Science Technology and Human Rights Notes | LL.M. Semester 3 | Mumbai University | munotes

What the paper actually covers

MU prints six topic labels and nothing else, so it is worth stating plainly what those six ask for, because the shape of the paper is not obvious from its name.

Notice what that structure is doing. Modules I and III are about principle, Modules II and IV are about the law of two particular fields, and Module IV closes by asking whether the international response has been adequate. An examiner can therefore ask a purely conceptual question, a purely statutory question, or a question that runs across both, and this book is built so that any of the three can be answered from it.

What this paper is not

It is not a science paper. No question will require you to explain how a polymerase chain reaction works. You must understand a technology well enough to see which right it touches, and no further.

It is not a general human rights paper. Your group has five other papers for that. Here, a right is discussed because a scientific or technological development bears on it.

It is not a medical jurisprudence paper. Medical jurisprudence, in the older sense, is about medical evidence in criminal trials. Module II is about the legal regulation of medicine and the rights of patients, which is a different subject with some overlap.

A worked example, to show the method

Take a single technology and run it through the questions this paper asks. A hospital installs a system that reads scans and reports whether a tumour is present.

Which rights does it engage? The right to health, because it may find cancers earlier. Privacy, because it processes the patient's medical images. Equality, because the system was trained on data from one population and may be less accurate on another. The right to a remedy, because if it is wrong it is not obvious who is answerable.

Who bears the duty? The hospital, as the body that chose to deploy it. The manufacturer, whose product it is. And the State, whose duty to protect means it must regulate both.

Is the benefit shared? If the system exists only in private hospitals in three cities, the technology has widened a gap rather than closed one, and Module I's right to enjoy the benefits of scientific progress is engaged directly.

What does the international framework say? The 1975 Declaration requires States to extend the benefits of science to all strata of the population and to protect them from the misuse of scientific developments. The Bioethics Declaration of 2005 requires risk assessment and ethics committees.

Every one of those four questions is answered somewhere in this book. That is the method the paper is testing.

What beginners get wrong here

Treating the subject as a list of technologies. It is not. A student who memorises facts about cloning, artificial intelligence and surveillance, and learns no framework, cannot answer a question about a technology the syllabus does not mention. The framework is the point; the technologies are examples.

Assuming science is the villain. The instruments in Module IV say the opposite. The right to enjoy the benefits of scientific progress is a human right, and a State that fails to develop and diffuse science is in breach of it. A one-sided answer loses marks.

Confusing ethics with law. Module III exists because much of what is objectionable in this field is not unlawful. Saying that something "violates human rights" when what you mean is that it is unethical, and not distinguishing the two, is the commonest weakness in answers on this paper.

Quick revision

• Science is knowledge and method; technology is application. Freedom questions attach to the first, harm and access questions to the second.
• The subject exists because science both delivers unequally and injures in ways ordinary law could not answer.
• Six printed topics: relationship in principle, relationship in practice, medicine, ethics, information and cyber terrorism, and the international response.
• The method: which rights are engaged, who bears the duty, is the benefit shared, what does the international framework say.
• Not a science paper, not a general human rights paper, not medical jurisprudence.

Test yourself

1. Distinguish science from technology, and say why the distinction matters in law. Science is organised knowledge and the method of producing it; technology is the application of that knowledge to do something. The distinction matters because science raises questions of freedom, chiefly the researcher's freedom to inquire and the State's duty to support inquiry, while technology raises questions of harm and of access, chiefly who is injured by a use and who is excluded from a benefit.

2. Why did human rights law become interested in scientific progress only in the twentieth century? Because progress became fast enough and powerful enough to matter in two ways at once. Its benefits became large enough that unequal access became a question of justice rather than of luck, and its harms became large enough, and were often inflicted lawfully by qualified people, that the ordinary law of wrongs could not answer them.

3. A State refuses to fund any medical research at all. Which limb of this subject is engaged? The right to enjoy the benefits of scientific progress, and specifically the State's duty to conserve, develop and diffuse science. This is the limb that shows the subject is not only about restraining science but also about requiring it.

4. Is this paper asking you to decide whether a technology is good or bad? No. It is asking you to identify which rights a technology engages, who owes duties in respect of it, whether the benefit and the risk are fairly distributed, and what the legal and ethical frameworks require. The evaluation follows from that analysis; it does not replace it.

Interdependent. Rights depend on each other. The right to health is worth little without access to information; privacy is worth little without an effective remedy.

The three generations

The classification is a teaching device, not a legal hierarchy, and it is commonly attributed to Karel Vasak, who borrowed the French revolutionary triad.

For this paper the classification is worth knowing for one reason above all. The right to science straddles the generations. Its freedom limb, the freedom indispensable for scientific research, behaves like a first generation right and asks the State to keep out. Its benefit limb asks the State to spend money and behaves like a second generation right. Its co-operation limb, and the idea that scientific knowledge is a common heritage, behave like third generation rights. A student who can say that has understood something that the printed topic label only implies.

Who owes the duty

A right without an identified duty bearer is a slogan. The modern formulation gives the State three duties, and the three recur throughout this book.

1. Respect. The State must not itself violate the right. It must not conduct experiments on prisoners.
2. Protect. The State must prevent third parties from violating it. It must regulate the private laboratory and the private hospital.
3. Fulfil. The State must take positive steps to realise it. It must fund research, train scientists and make the results available.

The second duty is what makes this a legal subject rather than a moral one. Most technology today is developed and owned privately. If the framework reached only the State's own conduct, it would reach almost nothing that matters. It is the duty to protect that makes a State answerable for what a company does with a genetic database.

A worked example

A private fertility clinic in Mumbai advertises that it can select embryos by sex. A couple pays for the service.

Whose right is engaged? The unborn child has no legal personality, so the analysis does not start there. It starts with the equality rights of women as a class, which sex selection attacks at scale, and with the rights of the woman undergoing the procedure.

Which duty is engaged? Not respect, because the State is not doing it. Protect is the operative duty: the State must prevent the private actor from doing it. That is precisely what the PCPNDT Act of 1994 does, and Module II examines how.

Would consent answer the complaint? No, and this is where inalienability does its work. The couple's consent cannot cure a harm whose victim is not the couple.

What it does NOT mean

It does not mean rights are absolute. Almost every right may be limited. What the framework requires is that a limitation be prescribed by law, pursue a legitimate aim, and be proportionate. Module IV turns that requirement into a four-part test.

It does not mean a right is only what a court will enforce. Much of the material in this book is in declarations that no court applies directly. That makes them weak, not meaningless, and Module IV's closing chapters ask honestly how weak.

It does not mean the State is the only actor. It is the primary duty bearer, but the instruments in Module III address researchers, institutions and corporations directly.

Quick revision

• A human right is inherent in the person, recognised rather than conferred, and imposes duties chiefly on the State.
• Four characteristics: universal, inalienable, indivisible, interdependent.
• Three generations: civil and political; economic, social and cultural; solidarity. The classification is a teaching device.
• The right to science straddles all three, which is why it is hard to classify and easy to neglect.
• Three duties: respect, protect, fulfil. The duty to protect is what reaches private technology.

Test yourself

1. Why does it matter that the Universal Declaration speaks of "recognition" rather than of grant? Because a right that the State grants is a right the State can withdraw, and the framework was built after a period in which a State used its own law against its own people. Recognition asserts that the entitlement exists independently of the State, so that denying it is a breach rather than a lawful policy choice.

2. Which of the three duties makes a State answerable for a private laboratory's conduct? The duty to protect. Respect governs the State's own conduct and fulfil requires positive provision; it is the duty to protect that requires the State to prevent third parties from violating rights, and it is the duty that makes regulation of private technology a human rights obligation rather than mere policy.

3. Explain why the classification into generations is awkward for the right to science. Because the right has limbs belonging to each generation at once. The freedom of research is a classic first generation liberty requiring abstention; the entitlement to the benefits of scientific progress is a second generation claim requiring expenditure; and international co-operation and the notion of a common heritage are third generation solidarity ideas. No single generation contains it.

4. A research participant signs a form agreeing to accept any risk whatever. Is that effective? Not entirely. Inalienability means some protections cannot be waived, and the instruments in Module III do not treat consent as making everything permissible. Consent is necessary, but the research must independently be ethically acceptable, which is what an ethics committee exists to decide.

"And its benefits." Two things are claimed, not one. A share in the advancement, which is the knowledge itself and the ability to take part in producing it, and a share in the benefits, which are its practical applications. A State that publishes research freely but prices the resulting medicine beyond reach has satisfied the first half and failed the second.

"The moral and material interests." Paragraph 2 is the intellectual property limb, and it is drawn deliberately narrowly. It protects the author, meaning the human creator. It speaks of moral as well as material interests, which is broader than a purely economic patent right and narrower in its beneficiary: a corporation holding a patent is not an author.

The tension inside the article

Paragraph 1 says everyone may share in the benefits. Paragraph 2 says the creator's interests are protected. Those two pull against each other, and the whole modern argument about patents and access to medicines lives in the gap.

The article does not resolve the tension, and it is important to say so rather than to pretend otherwise. What it does is refuse to let either side be forgotten. A student asked about access to medicines can therefore begin by observing that both the claim of the patient and the claim of the inventor are stated in the same article of the same instrument, which is why the question is genuinely hard.

A worked example

A laboratory in one country develops a vaccine using publicly funded research. Manufacturing capacity exists in a second country. A third country can afford neither the vaccine nor the licence.

Paragraph 1 gives the population of the third country a claim to share in the benefit. It is not charity that is being asked for; it is an entitlement stated in 1948.

Paragraph 2 protects the moral and material interests of the individual scientists who did the work. It does not, on its face, protect the balance sheet of the company that owns the patent, because that company is not an author. That is a genuine reading of the text and it is worth making, though it is not how the intellectual property system in fact operates.

The gap between the two is where later instruments do their work. The Covenant of 1966 turns the entitlement into a binding obligation with steps attached, General Comment 25 of 2020 spells out what those steps require, and the 1975 Declaration adds the duty of international co-operation.

Limits and what came after

The Declaration is a resolution of the General Assembly. It was not, when adopted, a treaty, and it created no directly enforceable obligation. Its authority has grown by other routes: through the argument that much of it now reflects customary international law, through its incorporation into national constitutions, and through the two Covenants of 1966 which turned its provisions into binding form.

For this paper the sequence to remember is short. Article 27(1) of the Declaration in 1948 becomes article 15(1)(b) of the ICESCR in 1966, which becomes the subject of General Comment 25 in 2020. The wording changes at each step, and the changes are examinable.

Distinctions

What beginners get wrong

Reading paragraph 2 as a patent clause. It protects authors, and it speaks of moral interests, which patent law does not recognise at all. Treating it as a straightforward endorsement of the patent system overstates it.

Forgetting that science sits in a cultural article. Students who quote article 27 for the science limb alone lose the argument that the drafters treated scientific knowledge as part of the common cultural inheritance.

Saying the Declaration is binding. It was not a treaty. Say instead that it has acquired authority through custom, constitutional reception and the Covenants.

Quick revision

• Article 27(1): everyone has the right freely to participate in cultural life, enjoy the arts, and share in scientific advancement and its benefits.
• Article 27(2): protection of the moral and material interests of the author of a scientific, literary or artistic production.
• Two claims in paragraph 1: a share in the advancement and a share in the benefits.
• Science sits in a cultural article, alongside art. That placement is doing work.
• The two paragraphs pull against each other, and the access-to-medicines debate lives in that gap.
• Declaration, 10 December 1948, General Assembly resolution 217 A (III). Not a treaty when adopted.

Test yourself

1. Set out article 27 and identify the two distinct rights it contains. Article 27(1) confers a right freely to participate in the cultural life of the community, to enjoy the arts and to share in scientific advancement and its benefits. Article 27(2) confers a right to the protection of the moral and material interests resulting from any scientific, literary or artistic production of which the person is the author. The first is a right of access to what science produces; the second is a right of protection for the person who produced it.

2. Why does it matter that article 27(1) claims a share in both the advancement and its benefits? Because the two can be satisfied separately. A State that publishes research openly gives a share in the advancement, and a State that makes the resulting product available gives a share in the benefits. A State can do the first and fail the second entirely, which is what happens when knowledge is public but the medicine derived from it is unaffordable.

3. Does article 27(2) protect a pharmaceutical company's patent? Not on its own terms. The paragraph protects the moral and material interests of the author, meaning the human creator, and it protects moral interests that patent law does not recognise. A corporate patent holder is not an author, so the paragraph is a weaker foundation for corporate intellectual property than it is often assumed to be.

4. What is the significance of science being placed in an article about cultural life? It treats scientific knowledge as part of what a community produces and holds in common, rather than as a commodity or a matter of industrial policy. That framing supports the claim that the benefits of science are owed to everyone as members of a community, and it is the textual root of the later idea that scientific knowledge is a common heritage.

Article 25 broken down

"An adequate standard of living." Health is not framed as a right to be treated when ill. It is framed as a right to the conditions in which a person can be well: food, clothing, housing and medical care in one list. That framing matters for this paper, because it means a State cannot answer a right-to-health claim purely by building hospitals.

"Health and well-being." Two words, and the second is wider. Well-being takes in mental health, which is why the Mental Healthcare Act's rights chapter in Module II belongs to this tradition and not to a separate one.

"Medical care and necessary social services." Medical care is named expressly. It is one item in a list, not the whole of the right.

"Security in the event of ... sickness, disability ... or other lack of livelihood." The article contemplates that illness is an economic event as well as a medical one. Catastrophic medical expenditure is therefore a rights question and not merely a policy failure.

"Motherhood and childhood are entitled to special care." The clause that carries into Module II's chapters on pregnancy, termination, surrogacy and assisted reproduction, and into Module IV's chapter on children online.

How the two articles frame this paper

The structure of the Declaration explains the structure of the syllabus more neatly than the syllabus does.

A student who can draw that table has a way into almost any question on this paper, because it shows that the three fields the syllabus examines are not an arbitrary selection. They are the three places the 1948 text already touched.

A worked example

A State introduces a national health database. Every citizen's medical record is uploaded, and researchers may apply for access to anonymised extracts.

Article 25 supports it. Better data means better treatment and better epidemiology, and the State's duty to secure an adequate standard of health is served by knowing what is making people ill.

Article 12 constrains it. Medical information is the most sensitive category there is. The interference must not be arbitrary: it needs a law, a legitimate purpose and proportionality, and the citizen needs a remedy if the safeguards fail.

Article 27 complicates it. The research done on the database will produce benefits, and the population whose records made the research possible has a claim to share in them. If the outputs are licensed exclusively to a private firm, the article is engaged.

Notice that the three articles do not give one answer. They give three questions that a good answer must address. That is what the Declaration supplies to this subject: not a rule, but the agenda.

What beginners get wrong

Treating article 12 as absolute. It forbids arbitrary interference. A student who writes that surveillance always violates the Declaration has misread the operative word and will be unable to handle any realistic problem.

Reading article 25 as a right to free treatment. It is a right to an adequate standard of living for health, of which medical care is one named component. It says nothing directly about who pays.

Forgetting the second sentence of article 12. The right to the protection of the law is what makes the article bite against private actors, and private actors hold most of the data.

Quick revision

• Article 12: no arbitrary interference with privacy, family, home or correspondence; no attacks on honour and reputation; and a right to the protection of the law against both.
• Arbitrary is the operative word: interference must be lawful, legitimate in purpose and proportionate.
• Article 25(1): a standard of living adequate for health and well-being, including food, clothing, housing, medical care and necessary social services, and security in sickness, disability and old age.
• Article 25(2): special care for motherhood and childhood.
• Articles 12, 25 and 27 map onto Modules IV, II and I respectively.

Test yourself

1. Why is the word "arbitrary" in article 12 more important than the list of four protected zones? Because the list reflects where a person could be intruded upon in 1948, and most sensitive information is no longer held at home or sent by post. The adjective, by contrast, states a test that survives any change of technology: interference must be lawful, must pursue a legitimate purpose, and must be proportionate. It is the part of the article that later law was able to build on.

2. What does the second sentence of article 12 add? A right to the protection of the law against interference and attacks. Without it the article would restrain only the State. With it, the State is obliged to provide legal machinery that works against private intruders too, which is what makes the article relevant to data held by companies.

3. How does article 25 frame the right to health, and why does the framing matter? It frames it as a right to a standard of living adequate for health and well-being, with food, clothing, housing, medical care and social services in a single list. The framing matters because it prevents a State from answering a right-to-health complaint solely by pointing to hospitals: the conditions of health are part of the right, not a separate policy area.

4. A national health database is challenged. Which Declaration articles are engaged and how do they pull? Article 25 supports the database, because better health information serves the State's duty to secure an adequate standard of health. Article 12 constrains it, because medical data is highly sensitive and the interference must not be arbitrary. Article 27 is engaged if research done on the data produces benefits, because the population whose records enabled it has a claim to share in those benefits.

"Scientific progress and its applications." Both the knowledge and what is done with it. A State that produces excellent research which never reaches a clinic has not discharged the obligation.

"(a), (b) and (c) together." The three sub-paragraphs are one right with three faces: cultural participation, scientific benefit, and protection of the creator. Sub-paragraph (c) carries the same tension with (b) that paragraph 2 of article 27 carried with paragraph 1.

The wording changed, and the change is examinable

A worked example

A State ratifies the Covenant. It has a well-funded space programme and almost no rural primary healthcare.

Is it in breach of article 15(1)(b)? The question cannot be answered by pointing to the space programme, because the obligation is not to produce science. It is to secure everyone's enjoyment of the benefits of scientific progress and its applications.

What would a complaint look like? That the applications of well-established science, in the form of vaccines, antibiotics and clean water, are not reaching a large part of the population, while resources are directed to research whose benefits reach almost none of them. That is an argument about distribution, and article 15(1)(b) is where it lives.

What is the State's answer? Progressive realisation. The obligation under the Covenant is to take steps to the maximum of available resources with a view to achieving progressively the full realisation of the rights. A State is not obliged to achieve everything at once, but it is obliged to move, to move in the right direction, and not to go backwards.

Limits

Progressive realisation is a real limit and a real obligation. It permits a State to plead resources. It does not permit inaction, and the Committee has consistently held that a deliberately retrogressive measure requires strong justification.

There is no direct individual remedy for most States. An Optional Protocol allowing individual communications was adopted in 2008 and is in force, but only for the States that have accepted it. Most complaints therefore surface, if at all, through the reporting process, and this is one of the honest weaknesses Module IV returns to.

The right was underdeveloped for over fifty years. Article 15(1)(b) attracted almost no jurisprudence until the Committee issued General Comment 25 in 2020. A student should say so, because it explains why the answers to many questions on this topic are recent.

What beginners get wrong

Saying the Covenant "created" the right. It recognised it, and the Declaration had stated it eighteen years earlier. The Covenant made it binding and attached duties.

Treating "progressive realisation" as an excuse. It is a standard against which a State is measured, not a defence to any complaint whatever.

Quoting article 15(1)(b) alone. The obligation is unintelligible without paragraphs (2), (3) and (4), which state what the State has to do. Those are the next three chapters, and an exam answer that omits them is half an answer.

Quick revision

• Article 15(1)(b) ICESCR, 1966: the right of everyone to enjoy the benefits of scientific progress and its applications.
• Binding on States Parties, unlike the Declaration.
• Wording changed from 1948: "advancement" became "progress", "share in" became "enjoy", and "applications" was added.
• Read with (a) cultural life and (c) protection of the author. One right, three faces.
• Subject to progressive realisation to the maximum of available resources; retrogression needs strong justification.
• Almost no jurisprudence until General Comment 25 in 2020.

Test yourself

1. State article 15(1)(b) and explain how it differs in legal character from article 27 of the Universal Declaration. Article 15(1)(b) provides that States Parties recognise the right of everyone to enjoy the benefits of scientific progress and its applications. It differs in that the Covenant is a treaty binding on the States that ratify it, whereas the Universal Declaration was a resolution of the General Assembly creating no directly binding obligation. The Covenant also attaches machinery, in the form of reporting to the Committee on Economic, Social and Cultural Rights, and states the steps required in paragraphs (2), (3) and (4).

2. What was lost and what was gained when the 1948 wording became the 1966 wording? The express reference to sharing in scientific advancement was lost, which weakened the participatory limb until interpretation restored it. What was gained was the express reference to applications, which makes clear that the right extends beyond knowledge to what is done with it, and the binding character of a treaty obligation with steps attached.

3. Does progressive realisation mean a poor State owes nothing under article 15? No. It means the State must take steps to the maximum of its available resources with a view to achieving the full realisation of the right progressively. Inaction is not permitted, the direction of travel is measured, and a deliberately retrogressive measure requires strong justification. Poverty affects the pace, not the existence, of the obligation.

4. A State has a large research budget and poor basic healthcare. Is that a breach? Possibly, and the argument runs through distribution rather than through the size of the research budget. Article 15(1)(b) protects everyone's enjoyment of the benefits and applications of scientific progress, so a State in which well-established applications such as vaccines and clean water do not reach a large part of the population may be failing the obligation regardless of how much science it produces.

"Shall include." A floor, not a ceiling. These three are compulsory, and a State that has done them is not thereby finished; other steps may still be necessary.

"Those necessary." The measure is what the object requires, not what is convenient or affordable in the abstract. Resources come in through article 2(1), not through a discount hidden in the word necessary.

"Conservation, development, diffusion." Three separate duties, and a State can be doing well at one and failing another. They are taken one at a time below.

"Of science and culture." Science is not separated out for weaker treatment. The same sentence governs both, which is the same structural point article 27 of the Declaration made by putting science in a cultural article.

What the three words mean

The Committee on Economic, Social and Cultural Rights, the treaty body that supervises the Covenant, put the distinction between abstaining and acting as plainly as it can be put:

States parties should not only abstain from interfering in the freedom of individuals and institutions to develop science and diffuse its results. States must take positive steps for the advancement of science (development) and for the protection and dissemination of scientific knowledge and its applications (conservation and diffusion).

Conservation is protection and preservation. It covers scientific knowledge already produced and the things that hold it: data sets, specimen and seed collections, archives, laboratories, instruments, and the trained people without whom none of it can be read. Knowledge that exists in a form nobody can any longer reach has not been conserved, and a State that lets that happen has failed this limb even though it never obstructed anybody.

Development is the advancement of science: research funding, research infrastructure, the training of scientists, and the institutional conditions in which inquiry can be carried on. The Committee treats this as a duty to promote actively, through education and investment in science and technology, through policies and regulations that foster research, and through the allocation of resources in budgets.

Diffusion is spreading science and its applications so that they reach people. The Committee names the instruments of diffusion in terms: libraries, museums and network access, a research infrastructure with adequate resources, and adequate financing of scientific education. It goes further and says that States should promote open science and open source publication of research, and that research findings and research data funded by States should be accessible to the public.

Which of the three duties this is

Every human right generates three duties on the State: to respect, to protect and to fulfil. Article 15(2) is the Covenant's own statement of the third. The Committee says so directly, that the duty to fulfil is reinforced and specified by article 15(2), which provides that States parties must take steps for the conservation, the development and the diffusion of science.

That is worth holding on to, because it tells a student where to put the paragraph in an answer. A question about censorship of a researcher is answered under article 15(3) and the duty to respect. A question about a private laboratory discriminating is answered under the duty to protect. A question about a State that produces no science, teaches no science and shares no science is answered here.

What a State actually has to do

The Committee's general comment turns the three words into obligations concrete enough to be argued about. Six of them are worth knowing by name.

1. Fund it. Allocate resources in budgets, and give priority in the allocation of public resources to research in the areas of greatest need in health, food and other basic needs, especially for vulnerable and marginalised groups.
2. Legislate for it. Develop a participatory national framework law on the right that includes legal remedies for violations, and adopt a national strategy or action plan that includes a strategy for the conservation, the development and the diffusion of science.
3. Teach it. Ensure that people have access to the basic education and skills necessary to understand and apply scientific knowledge, and that scientific education in public and private schools respects the best available scientific knowledge.
4. Open it. Make every effort to ensure equitable and open access to scientific literature, data and content, including by removing barriers to publishing, sharing and archiving scientific outputs.
5. Carry it to people. Make knowledge of scientific progress and its applications broadly available through schools, universities, technical colleges, libraries, museums, print and electronic media and other channels, with specific programmes to overcome problems of access related to age, language or cultural diversity.
6. Defend it. Promote accurate scientific information, refrain from disinformation and disparagement intended to erode public understanding of and respect for science, and adopt mechanisms protecting people from false, misleading and pseudoscience based practices.

Two further points from the same source correct a common assumption about who owes this. Open science, the Committee says, cannot be achieved by the State alone: scientists, universities, publishers, scientific associations, funding agencies, libraries and the media all have a role, especially for research financed with public funds. And it warns that all States should contribute to the common task of developing science to the maximum of their available resources, because recommending that poor States focus exclusively on applied science actually increases the gap and the unfair distribution of knowledge and power between States.

Going backwards

Because the obligation is to move forward, moving backward is itself a breach unless it is justified. The Committee holds that there is a strong presumption that retrogressive measures are not permissible, and it gives examples that map onto this paragraph precisely: removing programmes or policies necessary for the conservation, the development and the diffusion of science; imposing barriers to education and information on science; imposing barriers to citizen participation in scientific activities, including misinformation intended to erode public respect for science; and adopting legal and policy changes that reduce international collaboration on science.

Where such a measure is unavoidable, the State must show that it is necessary and proportionate, keep it in place only so long as it is necessary, ensure that disadvantaged groups are not disproportionately affected, and still guarantee the core obligations.

A worked example

The State of Ashwara ratified the Covenant. Three things happen in the same year.

The archive. Ashwara's National Crop Research Institute holds a seed collection assembled over sixty years, with the field records that make it usable. The Finance Ministry withdraws the maintenance grant, the cold storage fails, and part of the collection is lost. Nobody was obstructed and nobody was excluded. This is still a breach, of the conservation limb, because the State allowed knowledge and the material that carries it to be destroyed. It is also retrogressive: a functioning programme necessary to the conservation of science was removed.

The research budget. Ashwara funds a prestige robotics laboratory and nothing at all on a parasitic disease that affects two million of its poorest citizens and almost nobody elsewhere. That is a development question, and the answer is not that the robotics laboratory is forbidden. It is that public resources must give priority to research in the areas of greatest need in health and other basic needs, and a State that spends everything where the need is least has to justify that allocation.

The findings. Dr Meera Iyer, employed at a State university on public money, establishes a low cost diagnostic method for that same disease. Her results appear in a journal that charges a fee no district hospital in Ashwara can pay, and her data is not deposited anywhere. The State's science curriculum, meanwhile, has not been revised in twenty years. That is the diffusion limb twice over, once because research funded by the State is not accessible to the public and once because the population is not equipped to use it.

Ashwara's answer, and its limits. The State pleads progressive realisation and limited resources. That plea is available and is sometimes good: article 2(1) measures the obligation against the maximum of available resources. It does not answer the first item at all, because closing a working archive is a retrogressive act needing strong justification rather than a failure to expand. It answers the second only if Ashwara can show the allocation was reasoned, and it does the third no good whatever, because depositing publicly funded data and results costs almost nothing.

Distinctions

What beginners get wrong

Reading it as a paragraph about culture. The words are "science and culture" and the Committee's whole general comment is built on the science half. Nothing about the paragraph is confined to the arts.

Treating "shall include" as a complete list. It is a minimum. A State that has done these three things may still owe other steps under article 2(1).

Thinking a State discharges it by having universities. The duty is to conserve, develop and diffuse. A country with excellent universities whose findings never reach its own district hospitals has done the second and failed the third.

Assuming it is programmatic and unenforceable. The Committee states that this right, like every other right in the Covenant, is enforceable and therefore justiciable, and that States should establish mechanisms and remedies for violations by omission as well as by act. Whether courts in fact do this is a separate and much weaker story, which the closing chapters of Module IV take up.

Confusing diffusion with publication. Publishing a paper diffuses it to people who can already read papers. The Committee's examples run from libraries and museums to curricula and translation, and diffusion is measured by whether the knowledge arrives, not by whether it was released.

Quick revision

• Article 15(2) ICESCR: the steps to be taken shall include those necessary for the conservation, the development and the diffusion of science and culture.
• Conservation protects what exists; development advances it; diffusion spreads it and its applications.
• It is the Covenant's own statement of the duty to fulfil, reinforced and specified.
• Concrete duties: fund research where the need is greatest, legislate a framework and a plan, teach science, open access to publicly funded findings and data, carry knowledge through libraries, museums, media and curricula, and defend the public against pseudoscience and disinformation.
• Retrogression is presumed impermissible: closing programmes, raising barriers to science education, or cutting international collaboration all count.
• Resources are measured under article 2(1), to the maximum of what is available. Poverty affects the pace, not the existence, of the duty.

Test yourself

1. Set out article 15(2) and explain what it adds to article 15(1)(b). Article 15(2) provides that the steps to be taken by States Parties to achieve the full realisation of the right shall include those necessary for the conservation, the development and the diffusion of science and culture. Article 15(1)(b) states an entitlement; standing alone it could be satisfied by a State that obstructs nobody and provides nothing. Article 15(2) adds the positive side, requiring the State to preserve existing scientific knowledge, to advance it and to spread it and its applications, and it is the Covenant's own specification of the duty to fulfil.

2. A State says it cannot afford a national research programme. Is that an answer? Partly, and only to the extent it is true. The obligation under article 2(1) is measured to the maximum of available resources, so a poor State is not required to match a rich one. But the plea does not excuse inaction, it does not excuse allowing existing scientific collections and data to be destroyed, which is retrogressive rather than merely slow, and it does not excuse the steps that cost almost nothing, such as making publicly funded research findings and data accessible.

3. Distinguish conservation from diffusion, with an example of a State that satisfies one and breaches the other. Conservation is the protection and preservation of scientific knowledge and the means of producing it; diffusion is spreading that knowledge and its applications so that people can reach and use them. A State that maintains a well kept national data archive which is closed to the public, taught in no school and reproduced in no library has conserved and has not diffused. The reverse also occurs: a State may publicise findings widely while allowing the collections that produced them to decay.

4. Give three examples of a retrogressive measure under this paragraph. Removing a programme or policy necessary for the conservation, the development or the diffusion of science, such as closing a functioning research archive; imposing barriers to education and information on science, including official misinformation intended to erode public respect for scientific research; and adopting legal or policy changes that reduce international collaboration on science. Each is presumed impermissible and requires the State to prove necessity and proportionality.

5. Why is it wrong to say the diffusion duty is discharged by publishing research? Because diffusion is measured by arrival rather than by release. The Committee names libraries, museums, network access, scientific education and the media as instruments of diffusion, requires research findings and data funded by the State to be accessible to the public, and calls for specific programmes to overcome barriers of age, language and cultural diversity. Publication in a journal that a population cannot reach, in a language it does not read, and with no education equipping it to use the result, leaves the duty unperformed.

"The freedom." Singular, and undefined in the treaty. That is why the Committee on Economic, Social and Cultural Rights, the expert body to which States report under the Covenant, had to say what it contains. Its list is set out below.

"Indispensable." The adjective limits the protection and is often misread as weakening it. The Covenant does not protect every convenience a researcher would like. It protects what research cannot be carried on without. A denied promotion is not covered; a direction to reach a particular conclusion is.

"For scientific research and creative activity." The paragraph covers the artist as well as the scientist, which is again the Covenant treating science and culture as one subject. For this paper the scientific half is what matters.

What the freedom actually contains

The Committee's reading of article 15(3) lists the dimensions the freedom includes "at the least". They are worth learning as a list, because an examination answer that recites the bare words of the paragraph and stops has said almost nothing.

1. Independent judgment. Researchers are protected from undue influence on their independent judgment.
2. Autonomous institutions and self-chosen questions. Researchers may set up autonomous research institutions and define the aims and objectives of the research and the methods to be adopted.
3. Conscience. Researchers may freely and openly question the ethical value of particular projects, and may withdraw from a project if their conscience so dictates.
4. Collaboration. Researchers are free to co-operate with other researchers, nationally and internationally.
5. Sharing. Scientific data and analysis may be shared with policymakers, and with the public wherever possible.

To those the Committee adds the right to publish. As a consequence of the freedom of research and of the State's duty to disseminate science, scientists have in principle the right to publish the results of their research, and any restriction must comply with article 4. The Committee goes further and reaches private contracts: a State should ensure that any contractual restriction on publication is consistent with the public interest, is reasonable and proportionate, and provides for the appropriate crediting of the researchers who did the work.

The same passage that defines the State's positive duty describes what scientific freedom includes in practice: protection and promotion of academic and scientific freedom, including freedom of expression and freedom to seek, receive and impart scientific information, freedom of association and freedom of movement.

Whose freedom it is

A student writing on this paragraph will usually treat it as the scientist's professional privilege. That is the commonest error on the topic and it inverts the structure of the right.

The Committee treats the right to participate in and to enjoy the benefits of scientific progress as containing both freedoms and entitlements, and it puts the freedom to enjoy the freedom indispensable for scientific research in the first group. The protection of freedom of scientific research is an element of everyone's right, not a benefit conferred on a profession.

The reasoning is not sentimental. If the State silences a researcher who has found that a river is poisoned, the researcher loses a job and the population loses the knowledge. The people injured by the interference are the people who would have used the finding, which is why the freedom sits inside a right held by everyone.

The ground it stands on: thought and expression

Article 15(3) did not invent scientific freedom. Research is thinking, and then saying what the thinking found, and both of those are protected in their own right by instruments this paper has already met.

Freedom of thought. Article 18 of the Universal Declaration provides that everyone has the right to freedom of thought, conscience and religion. Article 18(1) of the International Covenant on Civil and Political Rights repeats it in binding form, and article 18(2) adds that no one shall be subject to coercion which would impair his freedom to have or to adopt a belief of his choice. Inquiry is the exercise of that freedom in an organised way.

Freedom of expression. Article 19 of the Universal Declaration provides that everyone has the right to freedom of opinion and expression, and that the right includes freedom to hold opinions without interference and to seek, receive and impart information and ideas through any media and regardless of frontiers. Article 19 of the Civil and Political Covenant makes it binding: paragraph 1 protects the holding of opinions without interference, and paragraph 2 protects the freedom to seek, receive and impart information and ideas of all kinds, regardless of frontiers, in any medium.

Three consequences follow, and they are examinable.

The three phrases do different work. To seek protects the gathering of data. To receive protects reading the literature, which is why access to journals and to the internet is a free expression question and not only a budget question. To impart protects publication.

"Regardless of frontiers" is the textual root of scientific collaboration. The right to receive and impart information does not stop at a border, which is why an obstacle to working with foreign colleagues is an interference with expression before it is anything else. That phrase is the bridge into the next chapter.

Scientific freedom is more than expression, not less. Article 19 protects what a researcher says. Article 15(3) protects the conditions in which the saying becomes possible: institutional autonomy, the choice of question and method, and freedom from undue influence on judgment. A student who answers a question on article 15(3) purely with free speech material has covered about half of it.

The limits, and how to test one

Freedom of scientific research is not absolute. The Committee says so in terms, and article 4 supplies the test. A limitation must satisfy every one of these:

1. Determined by law. An executive instruction, an informal warning from a ministry, or a funder's pressure is not law.
2. Compatible with the nature of the right. The Committee reads this as meaning that a limitation must respect the minimum core of the right.
3. Solely for the purpose of promoting the general welfare in a democratic society. Protecting a policy, an industry or a reputation from an unwelcome finding is not that purpose.
4. Proportionate. Where several means could achieve the legitimate aim, the least restrictive must be chosen, and the burden imposed on the right must not outweigh the benefit of the limitation.

The Committee then divides the field in a way worth memorising. Limits on the applications of science are relatively easy to justify: they can be used to guarantee the safety and quality of products, and human rights impact assessments may be needed to protect people against risky applications. Limits on the research process can also be necessary, particularly where research affects human beings, in order to protect their dignity, their integrity and their consent, and where research is conducted in countries or among populations other than the researchers' own, the State of origin must guarantee the rights and obligations of all parties involved. But any limitation on the content of scientific research implies a strict burden of justification by the State, in order to avoid infringing freedom of research.

The same division governs precaution. The Committee notes that limitations on the freedom of scientific research are compatible with the Covenant only in the circumstances set out in article 4, and that the precautionary principle applies more broadly to the use and application of scientific outcomes than to inquiry itself. The principle is taken on its own later in this module.

A worked example

Dr Farida Qureshi is a soil scientist at a public university. She begins a study of groundwater near a fertiliser plant. Five things happen. Each is tested the same way.

One. The funding contract. Part of her department's funding comes from the plant's owner, and the grant agreement says nothing may be published without the funder's written consent. This fails. A contractual restriction on publication must be consistent with the public interest and be reasonable and proportionate, and a private veto held by the party whose conduct is being measured is neither. The State's own duty here is the duty to protect: it must ensure that private investment in scientific institutions is not used to unduly influence the orientation of research or to restrict the scientific freedom of researchers.

Two. The stop order. The Ministry writes to the university, and the governing council directs her to close the study. This is a limitation on the content of research, so the strict burden of justification applies, and the limitation fails at the first step of article 4 before anything else is considered: a letter is not a law. Had there been a statute, it would still have to promote the general welfare in a democratic society, and suppressing a finding about contaminated water does the opposite.

Three. The conference and the co-author. Her application to travel to an international meeting is refused, and a rule forbids her to share raw data with a colleague abroad. Both bite on the collaboration dimension of the freedom, and the Committee treats the removal of obstacles to international collaboration among scientists as an example of the duty to respect, unless the restriction can be justified under article 4. It also treats the reduction of international scientific collaboration as a retrogressive measure.

Four. The blood samples. She proposes to take blood from villagers to measure exposure. Here the answer changes. This is a limit on the research process where research affects human beings, and it is permissible: she needs the free and informed consent of every participant and clearance from an ethics committee, and those requirements protect dignity, integrity and consent rather than suppressing a finding. The instruments that govern this are the subject of Module III, and freedom of research has never been a licence over other people's bodies.

Five. The test kit. A company wants to sell a testing kit built on her method, and the regulator requires certification before it goes on sale. That is a limit on an application and not on the research, and the Committee treats regulation and certification as part of ensuring the quality and the responsible development of science.

The pattern. Items one, two and three are interferences with the freedom the Covenant protects. Items four and five are not interferences at all in the relevant sense; they are the ordinary conditions under which research on people and products on the market are allowed to proceed. Sorting a fact pattern into those two piles is the whole skill this topic tests.

Distinctions

What it does NOT mean

It does not mean the researcher is above the law. The freedom is bounded by article 4 and by the rights of the people the research touches. Research on human beings without consent is not an exercise of scientific freedom; it is the wrong the Nuremberg Code was written to answer, and Module III sets that out.

It does not mean a right to be funded for anything. The claim to funding runs through article 15(2) and the duty to fulfil, and a State that declines to fund a particular project has not thereby interfered with anybody's freedom. What it may not do is fund on condition that the conclusion is agreed in advance.

It does not mean "respect" is a soft word. Respect is the strongest of the three duties in one specific way: it is immediate. A State may plead limited resources against a duty to fulfil. It cannot plead limited resources for a failure to leave a researcher alone.

It does not belong to scientists alone. The freedom is an element of everyone's right to participate in and enjoy the benefits of scientific progress. The public is the party that loses when it is denied.

Quick revision

• Article 15(3) ICESCR: States Parties undertake to respect the freedom indispensable for scientific research and creative activity. Duty to respect, immediate, not progressive.
• What it contains: independent judgment; autonomous institutions and self-chosen aims and methods; freedom of conscience to question and withdraw; national and international collaboration; sharing of data and analysis; and the right to publish.
• Its ground: freedom of thought (UDHR article 18, ICCPR article 18) and freedom of opinion and expression (UDHR article 19, ICCPR article 19), which protect seeking, receiving and imparting information regardless of frontiers.
• It is more than expression: it also protects institutional autonomy and the conditions of inquiry.
• Limits run through article 4: determined by law, compatible with the nature of the right, solely for the general welfare in a democratic society, and proportionate.
• Applications and the research process may be regulated. Limits on the content of research carry a strict burden of justification.
• The freedom is an element of everyone's right, not a professional privilege.

Test yourself

1. State article 15(3) and explain why its verb matters. Article 15(3) provides that the States Parties undertake to respect the freedom indispensable for scientific research and creative activity. The verb is "respect", which is the duty of non-interference, and it is the only limb of article 15 expressed that way: paragraph 1 recognises a right, paragraph 2 requires steps, paragraph 4 recognises benefits. Because abstention costs nothing, the duty is immediate rather than progressively realisable, so a State cannot plead lack of resources for interfering with research.

2. List the dimensions the freedom includes. Protection of researchers from undue influence on their independent judgment; the possibility of setting up autonomous research institutions and of defining the aims, objectives and methods of research; the freedom to question openly the ethical value of a project and to withdraw from it on grounds of conscience; the freedom to co-operate with other researchers nationally and internationally; and the sharing of scientific data and analysis with policymakers and, wherever possible, with the public. To these is added the right in principle to publish research results.

3. How does freedom of scientific research relate to freedom of expression? Freedom of expression is its ground. Article 19 of the Universal Declaration and article 19 of the Civil and Political Covenant protect the holding of opinions without interference and the freedom to seek, receive and impart information and ideas regardless of frontiers, which covers gathering data, reading the literature and publishing results, and the phrase "regardless of frontiers" is the textual basis for cross border collaboration. Article 15(3) adds what expression alone does not reach: institutional autonomy, freedom in the choice of question and method, and freedom from undue influence on judgment.

4. A State passes a law forbidding all research into a particular gene sequence, saying the subject is dangerous. How is that tested? Under article 4. The measure is determined by law, so it passes the first requirement. It must then be compatible with the nature of the right, which the Committee reads as respecting the minimum core; it must be solely for the purpose of promoting the general welfare in a democratic society; and it must be proportionate, meaning the least restrictive means capable of achieving the aim, with burdens that do not outweigh the benefit. Because it is a limitation on the content of research, the State carries a strict burden of justification, and a ban on inquiry will rarely be the least restrictive means when regulating the applications would meet the danger.

5. Is a requirement of ethics committee approval an interference with scientific freedom? No. It is a limitation on the research process of the kind the Committee treats as permissible, because the research affects human beings and the requirement exists to protect their dignity, their integrity and their consent. The distinction to draw is between limits that protect the people research is done on, which are legitimate, and limits on the content of the inquiry or on the publication of its results, which carry a strict burden of justification.

6. Why is it wrong to call article 15(3) a privilege of scientists? Because the Committee treats the freedom indispensable for scientific research as an element of the right of everyone to participate in and to enjoy the benefits of scientific progress. The reason is practical as well as doctrinal: when a researcher is silenced, the knowledge that would have been produced is lost to the whole population, so the injury falls on the people who would have used it and not only on the person prevented from working.

Broken down: what the Committee makes of it

General Comment 25 (2020) devotes its Part VI to international co-operation, and the argument runs in four steps. Each is a separate reason, and an answer that gives one of them has given a quarter of the material.

1. Some science can only be done together. Certain fields require a universal endeavour, so co-operation among scientists has to be encouraged if science is to progress at all. States should enable researchers to take part in the international scientific community, which means practical things: facilitating travel in and out of the territory, and policies that let researchers share data and educational resources across borders.
2. The disparities between countries are deep. Where financial or technological constraints require it, developing States should resort to international assistance and co-operation in order to comply with their Covenant obligations. Developed States should contribute to scientific development elsewhere by allocating aid and funding to scientific education, research and training, and by promoting collaboration between the scientific communities of developed and developing countries. Access to research results should be regulated so that developing countries and their citizens can reach them affordably, essential medicines being the example the Committee gives.
3. The benefits should be shared. The applications resulting from scientific progress should be shared with the international community, particularly with developing countries, communities living in poverty, and groups with special needs and vulnerabilities, especially where the benefits are closely related to the enjoyment of economic, social and cultural rights.
4. The worst risks are transnational. Climate change, rapid loss of biodiversity, dangerous technologies such as autonomous weapons based on artificial intelligence, and the threat of weapons of mass destruction cannot be addressed by any State alone. States should promote multilateral agreements to prevent those risks or mitigate them, and should act together against biopiracy and against illicit trafficking in organs, tissues, samples and genetic materials.

The Committee also names brain drain. While respecting a scientist's right to decide their own career, developed States should implement reasonable policies for identifying and countering, rather than fostering, its effects. That is a rare instance of a human rights body naming migration of skilled people as a distributive problem in science.

The 1975 Declaration says it in stronger words

The General Assembly's Declaration on the Use of Scientific and Technological Progress in the Interests of Peace and for the Benefit of Mankind, resolution 3384 (XXX) of 10 November 1975, is not a treaty and binds nobody. It is worth citing because its language is direct where the Covenant's is cautious, and because its preamble records the reasoning openly.

Its paragraph 1 requires all States to promote international co-operation to ensure that the results of scientific and technological developments are used to strengthen international peace and security, freedom and independence, and for the economic and social development of peoples and the realisation of human rights.

Its paragraph 5 requires all States to co-operate in the establishment, strengthening and development of the scientific and technological capacity of developing countries, with a view to accelerating the realisation of the social and economic rights of the peoples of those countries.

Its preamble records why: that scientific and technological progress is of great importance in accelerating the social and economic development of developing countries, and that the transfer of science and technology is one of the principal ways of accelerating that development.

Note the difference in force. The Declaration says States "shall co-operate". The Covenant says States "recognize the benefits". The Declaration is the stronger sentence and the weaker instrument, and a good answer says so rather than quoting whichever is more convenient.

The right to development

The phrase that gathers this material in the wider literature is the right to development: the claim that every human person and all peoples are entitled to participate in, contribute to and enjoy economic, social, cultural and political development in which all human rights can be fully realised. It was proclaimed by the General Assembly in 1986 and has been contested ever since, principally over whether it is a right at all and, if it is, who owes the corresponding duty.

Handle this claim carefully. The instrument that proclaims the right to development is not in this book's authority base, and nothing here quotes it or gives it an article number. What can be said on the strength of the texts we do hold is narrower and is enough for this paper: article 15(4) recognises international scientific co-operation; article 2(1) makes international assistance and co-operation part of the means by which Covenant rights are realised; the 1975 Declaration requires co-operation to build scientific capacity in developing countries; and General Comment 25 spells out what that requires of developed States. A candidate who argues from those four is arguing from authority. A candidate who recites article numbers of the 1986 Declaration from memory is doing the thing this paper's own sources forbid.

Worked example

A vaccine against a mosquito-borne disease is developed in a European university laboratory using public research funding. The disease is endemic in eight countries in South Asia and sub-Saharan Africa and effectively absent in Europe. Trials were conducted in two of the affected countries because that is where the disease is. The manufacturer prices the vaccine at a level the health budgets of those countries cannot meet, and declines to license production locally.

Step one, the right engaged. Article 15(1)(b): the people of the trial countries have a right to enjoy the benefits of scientific progress and its applications. A vaccine that exists and cannot be obtained is the paradigm case of a benefit withheld.

Step two, the co-operation limb. Article 15(4) with article 2(1) puts the developed State in the frame. General Comment 25 states that access to research results and their applications should be regulated so that developing countries and their citizens have adequate access in an affordable manner, and gives essential medicines as the example.

Step three, benefit sharing. The population that carried the risk of the trials is the population now unable to obtain the product. That is the point at which the general duty in paragraph 80 of the Comment, to share benefits particularly with developing countries and communities living in poverty, becomes concrete rather than aspirational.

Step four, what can actually be asked for. Not the abolition of the patent. The realistic asks are the ones the Comment names: public funding conditioned on affordable access, voluntary licensing for local manufacture, tiered pricing, and multilateral arrangements. The chapter on [Access to Medicines and Patents] takes the intellectual property side of this further.

What it does NOT mean

It is not a duty to give money. Article 15(4) creates no entitlement in one State to a payment from another. The Committee's language is about enabling environments, funding, collaboration and access regulation, not about a transfer that any court could quantify.

It is not only about money either. Half of what the Comment asks for costs very little: letting researchers travel, letting them share data, not obstructing collaboration. A State that funds nothing but obstructs nothing is in a better position under this limb than a State that writes a cheque and refuses visas.

Co-operation is not charity, and the Covenant does not frame it as such. The obligation is owed because the right exists, not because the giving State is generous. That is the whole point of putting it in a human rights treaty rather than in a development policy.

The developing State is not passive. Paragraph 79 puts an obligation on it too: where its own constraints require, it should resort to international assistance and co-operation. Failing to ask is itself a failure to take steps.

Limits and criticism

The limb is the weakest drafted in article 15 and has no enforcement machinery of its own. A State that does nothing under it faces the reporting procedure and nothing more.

It is also the limb where the gap between what the Committee says and what States do is widest. The Comment was adopted in April 2020, in the first weeks of a pandemic, and its paragraph 82 sets out at length what scientific co-operation against a pandemic should look like: sharing information about potential pathogens, early warning based on timely and transparent information, and sharing the best scientific knowledge and its applications once a pandemic develops. What followed is the strongest available illustration of the distance between the norm and the practice, and a good answer may say so, in measured terms, without turning into a political essay.

Quick revision

• ICESCR article 15(4): States Parties recognise the benefits to be derived from the encouragement and development of international contacts and co-operation in the scientific and cultural fields. The fourth and last limb of article 15.
• It is read with article 2(1) (steps individually and through international assistance and co-operation) and articles 55 and 56 of the UN Charter.
• General Comment 25, Part VI, gives four justifications: some science needs a universal endeavour; disparities between countries are deep; benefits should be shared, particularly with developing countries; the worst risks are transnational. It also names brain drain.
• 1975 Declaration, paragraph 5: all States shall co-operate in building the scientific and technological capacity of developing countries. Paragraph 1 ties co-operation to peace, development and the realisation of human rights.
• The Declaration is the stronger language in the weaker instrument.
• The right to development is the wider frame; argue it from article 15(4), article 2(1), the 1975 Declaration and General Comment 25, which are in front of you, and do not quote the 1986 instrument's articles.

Test yourself

1. Set out article 15(4) and explain why it cannot be read on its own. It provides that States Parties recognise the benefits to be derived from the encouragement and development of international contacts and co-operation in the scientific and cultural fields. Standing alone its language is weaker than the other three limbs, since it requires only recognition of benefits rather than steps or respect. It is read with article 2(1) of the Covenant, which obliges States to take steps individually and through international assistance and co-operation, and with articles 55 and 56 of the UN Charter. So understood it is the specific application in the scientific field of a general duty that already binds.

2. Give the Committee's four reasons for a reinforced duty of co-operation in science. First, certain fields of science require a universal endeavour, so co-operation among scientists must be encouraged for science to progress; States should facilitate researchers' travel and the international sharing of data. Second, there are deep disparities between countries: developing States should resort to international assistance where their constraints require it, and developed States should fund scientific education, research and training elsewhere and regulate access to research results so they are affordable. Third, benefits should be shared with the international community, particularly developing countries, communities living in poverty and groups with special needs. Fourth, the most acute risks, climate change, biodiversity loss, autonomous weapons and weapons of mass destruction, are transnational and need multilateral agreements.

3. A developed State funds no overseas research but freely admits foreign scientists, shares its published data and places no obstacle in the way of collaboration. A second State makes a large annual grant but refuses visas to researchers from the recipient countries. Which is closer to compliance with article 15(4)? The first. The Committee's account of the duty is not primarily financial: it asks States to enable participation in the international scientific community, specifically through facilitating travel and enabling the free sharing of data and educational resources. The second State has satisfied the funding dimension while defeating the participation dimension, and visas are named in the Comment where a grant is not. The better answer notes that neither is fully compliant, since the first does nothing about the disparity that paragraph 79 addresses.

4. Why should a candidate be cautious in writing about the right to development on this paper? Because the claim is easy to assert and hard to source. The right to development is contested both as to its existence as a right and as to who bears the duty, and the instrument proclaiming it is not among the texts this book works from, so any article number given for it would be recalled rather than read. The material that carries the same argument is available and binding or near-binding: article 15(4), article 2(1), paragraphs 1 and 5 of the 1975 Declaration, and Part VI of General Comment 25.

5. What does paragraph 82 of General Comment 25 say about pandemics, and why is its date significant? It states that pandemics are a crucial example of the need for scientific international co-operation, since pathogens do not respect borders and a local epidemic can rapidly become a pandemic. It calls for support for the World Health Organization, stronger commitment to scientific co-operation because national solutions are insufficient, sharing of scientific information about potential pathogens, improved early warning based on timely and transparent State information, and the sharing of scientific knowledge and its applications once a pandemic develops. The Comment was adopted on 30 April 2020, which makes the paragraph a statement of what the duty required at the moment it was being tested, and therefore a fair measure of the distance between the norm and State practice.

The Covenant was adopted by General Assembly resolution 2200A (XXI) of 16 December 1966 and entered into force on 3 January 1976.

Broken down

"The enjoyment of." Not the provision of. A facility that exists but that a person cannot reach, afford or use has not produced enjoyment of anything, and the obligation is stated in terms of the outcome for the individual.

"The highest attainable standard." Two limits sit inside one phrase, pulling opposite ways. Highest stops a State settling for a standard that is merely adequate or merely customary. Attainable lets in both what medicine has actually achieved and what the particular State can actually deliver, which is why the same article means different things in different countries in the same year.

"Physical and mental." Mental health is in the treaty text itself. It was not read in later by an interpreter. That is why the rights chapter of the Mental Healthcare Act 2017, taken in [The Rights of a Person with Mental Illness], is part of this tradition and not a separate compartment.

"The steps ... shall include those necessary for." The four sub-paragraphs are a floor and not a ceiling. "Shall include" means a State that has done these four things has not thereby exhausted article 12; it means a State that has done none of them is certainly in breach.

(a) Stillbirth-rate, infant mortality, healthy development of the child. A stillbirth is the birth of a child that has died before delivery, and infant mortality is the rate at which children die in the first year of life. Both are named because both were, and remain, the plainest available measure of whether a health system works at all.

(b) Environmental and industrial hygiene. Hygiene here means the conditions that keep disease from arising: clean water, drainage, air, and safe conditions in workplaces. Nothing in this sub-paragraph is about doctors.

(c) Epidemic, endemic, occupational and other diseases. An epidemic disease is one that spreads rapidly through a population over a limited period. An endemic disease is one that is constantly present in a given population or area. An occupational disease is one caused by the work a person does. The verbs are three: prevention, treatment and control.

(d) Conditions assuring to all medical service and medical attention in the event of sickness. The only sub-paragraph about clinical care, and the operative words are to all. It requires the creation of conditions that assure the service; it does not say who pays for it.

Read the four together and the Covenant's picture of health is mostly not a picture of hospitals. Three of the four sub-paragraphs are about children, sanitation, workplaces and disease control, which are questions of engineering, epidemiology and public administration. That is a scientific picture of health, and it is the reason this article can be taught in this paper at all.

What article 12 borrows from the rest of the Covenant

Article 2(1). Each State Party undertakes to take steps to the maximum of its available resources, with a view to achieving progressively the full realisation of the rights. Progressive realisation is explained at [The Right to Science in the Covenant], and it governs article 12 in exactly the way it governs article 15.

Article 2(2). The States Parties undertake to guarantee that the rights will be exercised without discrimination of any kind as to race, colour, sex, language, religion, political or other opinion, national or social origin, property, birth or other status. This obligation is not progressive. It bites at once, and it is the limb on which most realistic article 12 arguments are actually built.

Article 4. Limitations are permitted only where they are determined by law, only so far as compatible with the nature of the rights, and solely for the purpose of promoting the general welfare in a democratic society. A quarantine order is the standard example, and [Public Health Powers and Epidemics] takes the Indian version of it.

Why an article on health sits in a paper on science

The Committee that supervises the Covenant set out the links expressly in 2020, in the general comment on science taken in the next chapter. It gives four, at paragraphs 67 to 71.

First, science is the instrument. Scientific progress creates the applications that prevent disease, such as vaccination, or that let it be treated more effectively. The Committee therefore says States should promote research through funding or other incentives and make the resulting applications accessible and affordable to everyone, and that they should prioritise progress in the prevention, control and treatment of epidemic, endemic, occupational and other diseases. That last phrase is article 12(2)(c) quoted back into the law of science.

Second, science mediates between health and property. Some applications are protected by patents. The Committee calls the right to science a significant mediator between a human right, health, and a property right, and says States should use the flexibilities of the TRIPS Agreement, the World Trade Organization's agreement on trade related aspects of intellectual property rights, to secure essential medicines. It names compulsory licences, meaning a licence granted by the State allowing somebody other than the patent holder to make a patented product without that holder's agreement. It adds that States should refrain from granting disproportionately lengthy terms of patent protection for new medicines. [Access to Medicines and Patents] takes this in full.

Third, there is a duty to deliver what already exists. States must make available and accessible to all without discrimination, and especially to the most vulnerable, the best available applications of scientific progress needed for the highest attainable standard of health, and should prioritise safe and effective generic medicines over brand-name ones in national health plans.

Fourth, research itself carries health risks. The Committee requires States to prevent or mitigate them by careful application of the precautionary principle and by protecting research participants, and to ensure that medicines and treatments are evidence based and that risks are evaluated and communicated clearly enough for a patient to give properly informed consent. Those are the subjects of [The Precautionary Principle] and of [Clinical Trials and Research on Patients].

A worked example

A State has ratified the Covenant. In its capital, a government hospital tests every tuberculosis patient for drug resistance and prescribes the newer regimen where the test is positive. Four hundred kilometres away, Savitri, a farm labourer, is diagnosed at a primary health centre that has no such test and stocks only the older drugs.

Is article 12 engaged? Yes, and the reason is the word attainable. The State cannot argue that the newer standard is beyond it, because the State itself is already delivering that standard somewhere within its own territory.

Which limbs? Article 12(2)(c), on the prevention, treatment and control of epidemic and endemic disease, and article 12(2)(d), on conditions assuring to all medical service and medical attention in the event of sickness. The two words "to all" are what Savitri actually relies on.

What is the State's defence? Resources, under article 2(1). Rolling out testing to every primary health centre costs money, and the obligation is to move progressively towards full realisation rather than to arrive at it at once.

Why does that not close the argument? For two reasons. Article 2(2) makes non-discrimination immediate, so if the gap between the capital and the district tracks a prohibited ground, scarcity does not explain it. And the general comment on science puts a duty on the State to make the best available applications accessible to all without discrimination and especially to the most vulnerable, which is a duty about distribution rather than about totals.

What does the right to science add? The test Savitri was not given is an application of scientific progress. The duty in article 15(2) to diffuse science, worked in [The Duty to Conserve, Develop and Diffuse Science], is engaged by the same facts. Her complaint is a right-to-science complaint and a right-to-health complaint at once, and an answer that notices only one of them is half an answer.

Distinctions

What it does NOT mean

It does not mean a right to be healthy. No State can guarantee that outcome, and the article does not purport to. It guarantees a standard of provision and protection, measured against what is attainable.

It does not mean a right to free treatment. Sub-paragraph (d) requires the creation of conditions assuring medical service and attention to all. Who pays is left to the State, and a State that finances care through insurance rather than taxation is not thereby in breach.

It does not mean the four sub-paragraphs are the whole right. "Shall include" is inclusive language. Mental health, for instance, is named in paragraph 1 and appears in none of the four sub-paragraphs, and it is plainly inside the right.

It is not the same as the Indian right to health. India's enforceable right to health is read into article 21 of the Constitution, which is a different provision with a different history and a different remedy. That is [Science Under Article 21] and, in Module II, [The Right to Health Under Article 21].

Limits, and how the article reaches India

The weaknesses are the ones that attach to the whole Covenant. Realisation is progressive, article 4 permits limitations determined by law for the general welfare, and there is no general individual complaint procedure except for the States that have accepted the Optional Protocol, as [The Right to Science in the Covenant] records. Most supervision therefore happens through State reports rather than through anything resembling litigation.

India's Parliament has nonetheless written the Covenant into an Indian statute by name. Section 2(1)(f) of the Protection of Human Rights Act 1993 defines "International Covenants" to mean the International Covenant on Civil and Political Rights and the International Covenant on Economic, Social and Cultural Rights adopted by the General Assembly of the United Nations on the 16th December, 1966, together with any other Covenant or Convention the Central Government may notify. Section 2(1)(d) then defines "Human Rights" as the rights relating to life, liberty, equality and dignity of the individual guaranteed by the Constitution or embodied in the International Covenants and enforceable by courts in India.

Read the closing words of that definition carefully, because they carry the whole weight. The Act does not make everything in the Covenant enforceable. It brings in Covenant rights that are enforceable by courts in India, and in practice the route by which a health claim becomes enforceable here runs through article 21 of the Constitution. That is the bridge this chapter was written to build, and Module II crosses it.

Quick revision

• Article 12(1) ICESCR: the right of everyone to the enjoyment of the highest attainable standard of physical and mental health. Not a right to be healthy.
• Article 12(2): steps "shall include" those necessary for (a) stillbirth rate, infant mortality and the healthy development of the child; (b) environmental and industrial hygiene; (c) prevention, treatment and control of epidemic, endemic, occupational and other diseases; (d) conditions assuring to all medical service and medical attention in sickness.
• Three of the four sub-paragraphs are not about clinical care at all. The Covenant's picture of health is preventive and environmental.
• Article 2(1) progressive realisation, article 2(2) immediate non-discrimination, article 4 limitations by law for the general welfare.
• The science link, at paragraphs 67 to 71 of the 2020 general comment: science is the instrument; it mediates between health and patent; there is a duty to deliver what already exists; and research itself carries risk.
• Protection of Human Rights Act 1993, section 2(1)(d) and (f): the Covenants are named in Indian law, but only so far as the rights in them are enforceable by courts in India.

Test yourself

1. Why does article 12 speak of the "highest attainable standard" rather than of a right to health? Because health as an outcome cannot be guaranteed by any State. Whether a person is healthy depends on inheritance, accident and conduct. The article therefore fixes the obligation to a standard rather than to a result, and builds two limits into the phrase: highest, so that a State may not settle for the merely adequate, and attainable, which admits both the current state of scientific knowledge and the resources of the particular State.

2. What follows from the words "shall include" in article 12(2)? That the four sub-paragraphs are a floor and not an exhaustive statement of the right. A State that has taken all four categories of step may still be in breach in some other respect, and mental health, which appears in paragraph 1 and in none of the sub-paragraphs, is the clearest illustration.

3. A State says it cannot afford to extend a new diagnostic test beyond its capital. How would you answer? First, by article 2(2), which makes freedom from discrimination immediate rather than progressive, so that if the geographical gap tracks a prohibited ground the plea of scarcity does not explain it. Second, by the standard being demonstrably attainable, since the State is already delivering it internally. Third, by the duty stated in paragraphs 67 to 71 of the 2020 general comment on science to make the best available applications accessible to all without discrimination and especially to the most vulnerable, which is an obligation about distribution rather than about totals.

4. How does article 12 connect the right to health with the right to science? Through the word attainable, which makes the content of the health obligation depend on the state of scientific knowledge and on how far its applications have been diffused. The Committee's general comment on science states the connection in four ways: science supplies the applications that prevent and treat disease; it mediates between the right to health and patent rights; it generates a duty to deliver applications that already exist to everyone without discrimination; and research itself creates health risks that must be managed by precaution and informed consent.

5. Does the Protection of Human Rights Act 1993 make article 12 enforceable in India? Not by itself. Section 2(1)(f) names the Covenant, and section 2(1)(d) brings in rights embodied in the Covenants, but only those that are enforceable by courts in India. The Act supplies recognition and an institutional machinery rather than a new cause of action, and a health claim in India is in practice brought under article 21 of the Constitution.

Why it was written, and why so late

Two reasons appear in the opening paragraphs.

The first is that the subject matter had run away from the text. The Committee opens by observing that the intense and rapid development of science and technology has brought many benefits for economic, social and cultural rights, and that the risks, together with the unequal distribution of both the benefits and the risks, had produced a large and growing discussion.

The second is an admission. At paragraph 2 the Committee records that science is one of the areas of the Covenant to which States Parties give the least attention in their reports and in their dialogues with the Committee. The right had been in force since 1976 and had produced almost nothing, and the general comment exists to end that.

The Committee did not start from nothing. It names, among the material behind it, the Venice Statement on the Right to Enjoy the Benefits of Scientific Progress and its Applications of 2009, the UNESCO Universal Declaration on Bioethics and Human Rights of 2005, the UNESCO Recommendation on Science and Scientific Researchers of 2017, the report of the Special Rapporteur in the field of cultural rights on this right, and its own general comment No. 17 of 2005 on the protection of the moral and material interests of authors.

What the words mean

This is the part headed Normative content, paragraphs 4 to 14, and it is the part most likely to be examined directly.

Science. The Committee adopts the UNESCO definition, under which science is the organised attempt, by the objective study of observed phenomena and validation through the sharing of findings and data and through peer review, to discover and master the chain of causalities, relations or interactions. It then adds a test of its own at paragraph 5. Knowledge counts as science only if it is based on critical inquiry and is open to falsifiability and testability. Falsifiability means that the claim is of a kind that could in principle be shown to be false by evidence. Knowledge based solely on tradition, revelation or authority, which cannot be contrasted with reason and experience or checked by anybody else, is not science for the purposes of this right.

That paragraph does real work. It is what allows a State to refuse to treat an unfalsifiable claim as science without breaching anybody's right to science, and it is the foundation of the Committee's later insistence that States protect people from pseudoscience.

Both kinds, and both senses. Science here covers the natural and the social sciences. It also means two things at once: a process carried on by a method, which the Committee calls doing science, and the results of that process, which are knowledge and applications.

Progress. The Universal Declaration says scientific advancement and the Covenant says scientific progress. At paragraph 6 the Committee reads both words as emphasising the capacity of science to contribute to the well-being of persons and of humankind, and draws a striking conclusion from it: the development of science in the service of peace and human rights should be prioritised by States over other uses.

Applications. The particular implementation of science to the specific concerns and needs of the population, including the technology derived from scientific knowledge, whether medical, industrial, agricultural or in information and communications.

Benefits, in three senses. First the material results, such as vaccinations, fertilisers and technological instruments. Second the scientific knowledge and information itself, because science delivers by the spread of knowledge and not only by the sale of products. Third, and least expected, the role of science in forming critical and responsible citizens who are able to participate fully in a democratic society. A student who can give all three has understood why this right is not simply a right to receive goods.

The right got a longer name, and that is the biggest thing in the document

At paragraphs 9 to 11 the Committee rejects any reading of article 15(1)(b) that draws a rigid line between the scientist who produces science and a general population entitled only to enjoy the products.

It reaches that conclusion by four routes. It applies article 31 of the Vienna Convention on the Law of Treaties, which requires a treaty to be read in its context and in the light of its object and purpose. It reads sub-paragraph (b) with sub-paragraph (a), the right to take part in cultural life, and holds that scientific activity is part of cultural life. It uses the preparatory work on the drafting of article 15, the negotiating record known as the travaux préparatoires, which shows the article was intended to develop article 27 of the Universal Declaration. And it points out in a footnote that where the English text of article 27 says "share", the equally official French, Spanish and Russian texts say "participate".

The result, stated at paragraph 11, is that the right is properly named the right to participate in and to enjoy the benefits of scientific progress and its applications. Participation was not added by the Committee; it was recovered. The consequence, that States must not merely refrain from blocking ordinary people from scientific activity but must actively facilitate it, including what the Committee calls citizen science, is worked out in [Participation in Scientific Progress].

The other two limbs are read in the same section. Paragraph 13 lists what the freedom indispensable for scientific research contains, and paragraph 14 states that article 15(2) requires positive steps and not merely abstention. Those belong to [Freedom Indispensable for Scientific Research] and [The Duty to Conserve, Develop and Diffuse Science].

The structure it imposes on the right

Paragraph 15 is the hinge of that structure. The right contains freedoms and entitlements: the freedoms are the right to participate in scientific progress and the freedom indispensable for research, and the entitlement is to enjoy the benefits without discrimination. Both generate negative and positive obligations, and the right then has five interrelated and essential elements. Four of them are availability, accessibility, quality and acceptability, and they are the subject of the next chapter; the fifth, added at paragraph 20, is the protection of freedom of scientific research itself.

When the right may be limited

The Committee accepts at paragraph 21 that limitations may be necessary, because science and its applications can in some contexts damage economic, social and cultural rights. It then holds them to article 4 of the Covenant and adds a gloss.

The three requirements of article 4 are that the limitation is determined by law, that it promotes the general welfare in a democratic society, and that it is compatible with the nature of the right restricted. The Committee's gloss is that this implies two further things: the limitation must respect the minimum core obligations of the right, and it must be proportionate to the aim, which means that where several means could reasonably achieve the aim, the least restrictive must be chosen, and that the burdens imposed must not outweigh the benefits of the limitation.

Paragraph 22 then divides limitations by what they restrict. Limits on applications are relatively easy to justify and are used for safety and product quality, with human rights impact assessments where an application is risky. Limits on the research process are permissible to protect the dignity, integrity and consent of human participants. But any limitation on the content of scientific research carries what the Committee calls a strict burden of justification on the State, precisely so that freedom of research is not eroded.

The floor: core obligations

Paragraphs 51 and 52 identify core obligations, meaning the part of the right a State must implement as a matter of priority. A State that fails to satisfy them must demonstrate that it made every reasonable effort, taking account of all the rights in the Covenant and of the maximum of its available resources.

Twelve are listed. Among them: eliminate laws, policies and practices that unjustifiably limit access to scientific facilities, services, goods and information; identify and remove anything that undermines the participation of women and girls in science; remove limitations on freedom of research that article 4 cannot justify; develop a participatory national framework law on the right, including legal remedies; ensure people have the basic education and skills needed to understand and apply scientific knowledge; ensure access to those applications of scientific progress that are critical to the right to health; give priority in public resources to research in areas of greatest need in health and food; ensure health professionals are properly trained in modern technologies and medicines; refrain from disinformation intended to erode public respect for science; and protect people from false, misleading and pseudoscience based practices.

Notice the sixth of those. The Committee makes access to health-critical applications a core obligation of the right to science, which is the connection [The Right to Health in the Covenant] approached from the other side.

What a State should do at home

Part VII, paragraphs 85 to 89, is short and is the most quotable passage in the document for an Indian answer, because it is a list of things a legislature and an executive can be asked to produce.

A State should put in place a normative framework protecting non-discriminatory access to the benefits of science, protecting freedom of research within article 4, requiring ethics and human rights in research including ethics committees, and harmonising intellectual property with this right. It should adopt a national plan of action to promote scientific progress and disseminate its results, with public funding, scientific education, mechanisms against pseudoscience and measures on the conditions of researchers. It should identify indicators and benchmarks, with disaggregated statistics and time frames, so that implementation can actually be measured. And it should establish effective mechanisms and remedies, because, in the Committee's words at paragraph 89, this right like every other is enforceable and therefore justiciable, meaning capable of being decided by a court. The Committee adds that the right can be violated by omission as well as by action, so the remedies must reach both.

A worked example

Farida is an advocate. She acts for a group of patients challenging an executive order that bars the import of a class of diagnostic device, and she wants to run a right to science argument.

What the Covenant gives her. One sentence: the right of everyone to enjoy the benefits of scientific progress and its applications. On its own it does not tell a court what the State was obliged to do.

What the general comment gives her. A name for the right that includes participation; a definition of applications that plainly covers a diagnostic device; a limitation test, since the bar is a limitation and must therefore be determined by law, aimed at the general welfare, compatible with the nature of the right, proportionate, and the least restrictive means available; a core obligation about access to applications critical to health; and paragraph 89, which says the right is justiciable and that omissions count.

What she may not do with it. She may not plead that the State has violated General Comment 25. There is no such obligation. The breach is of the Covenant, and the general comment is the evidence of what the Covenant means. Nor is an Indian court bound to accept it, because it is not a treaty and India's courts apply Indian law; it is persuasive material about the content of an international obligation.

What it changes in practice. Before 2020 the argument would have been almost entirely her own construction. After 2020 every step of it can be attributed to the body that supervises the treaty, which is the difference between an argument and an authority.

What beginners get wrong

Treating it as binding law. Writing that "General Comment 25 provides that States must..." invites the reply that no State agreed to any such thing. Write instead that the Covenant obliges, and that the Committee has read the obligation as requiring the thing in question.

Thinking it created or renamed the right. Article 15(1)(b) reads today exactly as it read in 1966. The longer name is the Committee's description of what the article always contained, supported by the other language versions of the Universal Declaration and by the drafting record.

Thinking it covers the whole of article 15. It covers sub-paragraph (1)(b) and paragraphs (2), (3) and (4). The right to take part in cultural life in (1)(a) is the subject of the Committee's general comment No. 21 of 2009, and the protection of authors in (1)(c) is the subject of its general comment No. 17 of 2005, to which paragraph 12 expressly refers the reader.

Assuming science means natural science. The Committee says in terms that it includes the social sciences. The test is method, not subject matter.

Quick revision

• E/C.12/GC/25, 30 April 2020, Committee on Economic, Social and Cultural Rights, sixty-seventh session. Interprets article 15(1)(b), (2), (3) and (4).
• A general comment is the supervisory body's authoritative reading. Not a treaty, binds nobody by itself, but it is the reference text: the obligation is in the Covenant, the meaning is in the general comment.
• Written because science was the area of the Covenant States reported on least.
• Science requires critical inquiry, falsifiability and testability. Tradition, revelation and authority alone are not science. Natural and social sciences both.
• Benefits in three senses: material products, the knowledge itself, and the forming of critical citizens.
• The right renamed: to participate in and to enjoy the benefits of scientific progress and its applications, recovered from article 27 of the Universal Declaration, the other language versions and the drafting record.
• Five elements, four measurable ones plus freedom of research. Limitations under article 4 plus minimum core, proportionality and least restrictive means.
• Core obligations are the floor, twelve of them, one being access to applications critical to health.
• Part VII: normative framework, national plan of action, indicators and benchmarks, and remedies. The right is justiciable, and omissions violate it.

Test yourself

1. What is a general comment, and what authority does it have? It is a statement by a treaty's supervisory committee, addressed to all States Parties at once, of what a provision of the treaty means. It is not itself a treaty, no State ratifies it, and it creates no new obligation. Its authority is that it is the settled interpretation of the body to which every State Party must report and answer, so it is the reference text against which a State's conduct is measured in the only forum where that measurement happens.

2. What test does General Comment 25 apply to decide whether a body of knowledge is science? Whether it rests on critical inquiry and is open to falsifiability and testability, that is, whether it makes claims that could in principle be shown false by evidence and checked by others. Knowledge based solely on tradition, revelation or authority, which cannot be contrasted with reason and experience, is not science for the purposes of the right. Both natural and social sciences qualify, since the test is method rather than subject matter.

3. Explain the change of name of the right and the reasoning behind it. The Committee calls it the right to participate in and to enjoy the benefits of scientific progress and its applications. It reached that by reading article 15(1)(b) with 15(1)(a) on cultural life, applying article 31 of the Vienna Convention on the Law of Treaties so that the provision is read in context and in the light of its object and purpose, consulting the drafting record which shows article 15 was meant to develop article 27 of the Universal Declaration, and noting that the equally official French, Spanish and Russian texts of article 27 use the verb to participate where the English uses to share. Participation was therefore recovered rather than invented.

4. On what conditions may a State limit this right? Under article 4 of the Covenant the limitation must be determined by law, must promote the general welfare in a democratic society, and must be compatible with the nature of the right. The Committee adds that it must respect the minimum core obligations and be proportionate, so that the least restrictive of the available means is chosen and the burdens do not outweigh the benefits. A limitation on the content of research attracts a strict burden of justification, whereas limitations on applications for safety and quality are more readily justified.

5. A student writes that "General Comment 25 makes it compulsory for India to fund scientific research." What is wrong with the sentence? It attributes binding force to the wrong instrument. The obligation, if any, arises under the Covenant, chiefly article 15(2) read with article 2(1), and the general comment is the Committee's reading of what that obligation involves. The accurate form is that the Covenant obliges the State to take steps for the development of science to the maximum of its available resources, and that the Committee has read that as requiring positive promotion, appropriate budget allocation and an enabling environment, not merely non-interference.

1. Equal access to the applications of science, particularly where those applications are instrumental for the enjoyment of other economic, social and cultural rights.
2. Access to information about the risks and benefits of science and technology, without discrimination.
3. The open opportunity to participate in scientific progress, without discrimination. States should remove discriminatory barriers that stop people participating, for instance by facilitating access of marginalised populations to scientific education.

The third dimension is the one students most often miss. Accessibility is not only about receiving the products of science. It is about being allowed into the activity.

Quality

Quality refers to the most advanced, up to date and generally accepted and verifiable science available at the time, according to the standards generally accepted by the scientific community. It applies both to the process of scientific creation and to access to the applications and benefits.

Quality also includes regulation and certification, where necessary, to ensure the responsible and ethical development and application of science. States should rely on widely accepted scientific knowledge, in dialogue with the scientific community, to regulate and certify the circulation of new scientific applications that reach the public.

Two consequences follow that are worth stating. A State that gives its population access to something ineffective has not satisfied the right; access to a worthless remedy is not access to the benefits of scientific progress. And a State that certifies applications by a standard other than the accepted scientific one has failed this element even if the product is freely available.

Acceptability

Acceptability implies that efforts should be made to ensure that science is explained, and its applications disseminated in a way that facilitates their acceptance in different cultural and social contexts, provided this does not affect their integrity and quality. Scientific education and the products of science should be tailored to populations with special needs, and the Comment names persons with disabilities.

Acceptability also carries an ethical component. Scientific research has to incorporate ethical standards to ensure its integrity and the respect of human dignity, and the Comment points to the standards proposed in the Universal Declaration on Bioethics and Human Rights (2005). It lists some of them: benefits to research participants and other affected individuals should be maximised and possible harm minimised with reasonable protection and safeguards; the autonomy and free and informed consent of participants should be guaranteed; privacy and confidentiality respected; vulnerable groups or persons especially protected against discrimination; and cultural diversity and pluralism given due regard.

And a fifth: freedom of research

Paragraph 20 states that the protection of freedom of scientific research, dealt with at paragraph 13 and resting on article 15(3), is also an element of the right. It is treated in its own chapter, [Freedom Indispensable for Scientific Research], and is noted here so that the list is not given as four when the Comment gives five.

The proviso inside acceptability

The words "provided that this does not affect their integrity and quality" carry more weight than their length suggests. They are the answer to an obvious objection: if science must be made acceptable in different cultural contexts, may a State dilute what it teaches to suit local opinion?

The answer is no. Acceptability governs how science is explained and delivered, not what is true. Adapting the presentation is required; adapting the content until it is no longer the accepted science defeats the quality element. A candidate who can state that relationship between the third and fourth elements has understood the framework rather than memorised it.

Worked example

A State runs a childhood immunisation programme. Vaccines are procured centrally and are free. Uptake in three districts is under thirty per cent. Testing the programme element by element:

Availability. The vaccine exists, is procured and is distributed, and there is a cold chain. This element is largely satisfied. If the State had no procurement at all, the analysis would stop here.

Accessibility. The vaccine is free, so there is no cost barrier, but the nearest centre in the three districts is a day's travel and opens on two mornings a week. Access is formally equal and practically unequal. The second dimension is also engaged: is information about the risks and benefits reaching these districts in a language people read?

Quality. Is the vaccine the current one, stored correctly and administered by trained staff? A vaccine that has lost potency in an unreliable cold chain fails this element while appearing to satisfy the first two.

Acceptability. Has the programme been explained in a way that makes sense in the local context? Where refusal is driven by a rumour, the failure is at this element, and no amount of additional supply will fix it. The remedy is explanation, engagement with those the community trusts, and materials suited to the population, including those with disabilities.

The lesson of the example is the utility of the framework. "Low uptake" is not a finding a lawyer can work with. "Adequate on availability, deficient on accessibility in three districts, and failing on acceptability" is a finding that names the obligation breached and the remedy required.

What it does NOT mean

These are not four separate rights. They are elements of one right, and a State does not comply by satisfying three of them. The immunisation example is designed to show that a programme can be strong on the first two and fail entirely because of the fourth.

Availability is not the same as accessibility. A national library that exists in the capital is available. It is not accessible to a student six hundred kilometres away. The Committee treats the two separately because States routinely satisfy the first and report it as though it discharged the second.

Acceptability is not a veto. It does not entitle a community to demand that science be changed to match its beliefs. The proviso protects integrity and quality expressly.

Quality is not decided by the State alone. The standard is that generally accepted by the scientific community, and the Comment requires regulation to be undertaken in dialogue with it. A ministry that certifies a product against its own criteria has not met the element.

Distinctions

Quick revision

• General Comment 25 (2020), paragraphs 16 to 19, gives the right to science four essential elements: availability, accessibility, quality and acceptability. Paragraph 20 adds freedom of research as a further element.
• The framework is borrowed from General Comment 14 (2000) on the right to health.
• Availability: progress actually happening, knowledge protected and widely disseminated; libraries, museums, internet networks, research infrastructure, financing of scientific education; open science and public access to publicly funded findings and data.
• Accessibility: three dimensions, equal access to applications, access to information about risks and benefits, and the open opportunity to participate; remove discriminatory barriers, facilitate access of marginalised populations to scientific education.
• Quality: the most advanced, up to date, generally accepted and verifiable science available at the time; includes regulation and certification in dialogue with the scientific community.
• Acceptability: science explained and disseminated so as to be accepted in different cultural and social contexts, provided integrity and quality are not affected; tailored to persons with special needs; incorporates ethical standards, the Comment pointing to the Universal Declaration on Bioethics and Human Rights 2005.
• The Comment's own order is availability, accessibility, quality, acceptability.

Test yourself

1. Name the four elements and say where the framework comes from. Availability, accessibility, quality and acceptability, set out in paragraphs 16 to 19 of General Comment No. 25 (2020), with the protection of freedom of scientific research added at paragraph 20 as a further element of the same right. The framework is taken from General Comment No. 14 (2000) on the right to health, where the Committee first broke an economic and social right into these four elements so that compliance could be tested rather than merely asserted.

2. Give the three dimensions of accessibility. First, everyone should have equal access to the applications of science, particularly where those applications are instrumental for the enjoyment of other economic, social and cultural rights. Second, information about the risks and benefits of science and technology should be accessible without discrimination. Third, everyone should have the open opportunity to participate in scientific progress without discrimination, which requires States to remove discriminatory barriers, for example by facilitating access of marginalised populations to scientific education.

3. A government makes a traditional remedy freely available nationwide and promotes it energetically. Trials show it has no effect on the disease it is given for. Which element fails, and why does universal availability not save it? Quality fails. The element requires the most advanced, up to date, generally accepted and verifiable science available at the time, judged by standards generally accepted by the scientific community, and it applies to access to the applications and benefits of science as well as to the process of creation. An ineffective remedy is not a benefit of scientific progress, so distributing it universally satisfies availability and accessibility in form while defeating the right in substance. The element also covers regulation and certification, so a certification given against criteria the scientific community does not accept is itself part of the breach.

4. What work is done by the words "provided that this does not affect their integrity and quality" in the acceptability element? They prevent acceptability from becoming a licence to alter the content of science to suit local opinion. Acceptability governs how science is explained and how its applications are delivered, requiring adaptation of presentation to different cultural and social contexts and to populations with special needs. The proviso keeps that adaptation on the presentational side of the line, so that the quality element, which requires generally accepted and verifiable science, is not defeated by the element that follows it.

5. Which element does open access publishing belong to, and what exactly does the Comment ask for? Availability. The Comment states that States should promote open science and open source publication of research, and that research findings and research data funded by States should be accessible to the public. It sits under availability because that element concerns not only whether scientific progress is taking place but whether scientific knowledge is protected and widely disseminated, which is a question about diffusion rather than about individual access.

The word "benefits" is wider than products. The benefits of scientific progress "are not restricted to the material benefits or products of scientific advancement, but include the development of the critical mind and faculties associated with doing science". A person who has never been allowed near the method has been denied a benefit even if the medicine reaches them.

The drafting record points the same way. The Committee relies on the preparatory work of article 15, which shows the article was written to develop article 27 of the Universal Declaration of Human Rights, and article 27 recognises a right to participate in scientific advancement and not only to benefit from its applications. That earlier article is worked in its own chapter.

The other official texts say participate. The Committee notes that the English text of the Universal Declaration uses the word "share", while the French, Spanish and Russian texts, which are equally official versions of the same instrument, use the verb to participate. The participatory limb was in the 1948 text and was obscured by one language's choice of word.

Broken down: the four things participation means

One, doing science. Paragraph 10 says that doing science does not concern only scientific professionals but also includes "citizen science", which the Committee glosses in the same sentence as ordinary people doing science, and the dissemination of scientific knowledge. The obligation has two halves: States "should not only refrain from preventing citizen participation in scientific activities, but should actively facilitate it".

Two, deciding what is researched. This is the limb with teeth. Paragraph 55 provides that, with due respect to scientific freedom, some decisions concerning the orientation of scientific research or the adoption of certain technical advancements should be subjected to public scrutiny and citizen participation, and that scientific and technological policies should as far as possible be established through participatory and transparent processes and implemented with accompanying transparency and accountability mechanisms.

Three, knowing enough to take part. Participation without information is theatre. Paragraph 53 requires States to create an environment in which actual or perceived conflicts of interest are adequately disclosed and regulated, particularly where researchers advise policymakers. Paragraph 54 requires States to promote a culture of active citizen engagement with science through informed democratic debate and dialogue between the scientific community and society. Paragraph 49 requires equitable and open access to scientific literature, data and content, including by removing barriers to publishing, sharing and archiving scientific outputs.

Four, taking part in the control of risk. Paragraph 56 opens by saying that participation "also includes the right to information and participation in controlling the risks involved in particular scientific processes and its applications". Paragraph 57 completes the thought: in controversial cases the risks and potential of a technical advance should be made public so that society can decide, through informed and participatory public deliberation, whether the risks are acceptable. The principle that governs decisions under uncertainty is taken up in the chapter on the precautionary principle.

Participation is empty if whole groups cannot take part

Paragraph 28 identifies the groups that have experienced systemic discrimination in the enjoyment of this right, and names women, persons with disabilities, lesbian, gay, bisexual, transgender and intersex persons, indigenous peoples and persons living in poverty. It adds that temporary special measures might be necessary to achieve substantive equality.

Two of those are worked in a way this chapter needs. On women, paragraph 30 identifies a double discrimination: first, women have the right to participate in scientific research on an equal footing with men, so unequal access to scientific education or careers is discrimination in itself; second, because women are underrepresented in research, the research and the technologies that come out of it are commonly gender biased and insensitive to the needs of women. The exclusion damages the science as well as the excluded.

On indigenous peoples, paragraph 40 requires States to provide the educational and technological means to take part in the global scientific dialogue, and states that genuine consultation in order to obtain free, prior and informed consent is necessary whenever a State or a non-State actor conducts research, takes decisions or creates policies relating to science that have an impact on indigenous peoples, or uses their knowledge. Free, prior and informed consent here means agreement given without coercion, obtained before the activity begins, and with the facts disclosed.

The core obligations in paragraph 52 turn all of this into something a State must do first rather than eventually. Two of the listed bullets are participatory in terms: to identify and eliminate any law, policy, practice, prejudice or stereotype that undermines women's and girls' participation in scientific and technological areas, and to develop a participatory national framework law on this right which includes legal remedies for violations, together with a participatory national strategy or action plan.

A worked example

A State announces a national programme to sequence the genomes of ten thousand of its citizens, to build a database for medical research. Dr Iyer directs it. Kavita lives in a village selected for sampling, and her community has been asked to take part.

Was the decision to fund this taken participatively? Paragraph 55 is the provision. The choice to spend a research budget on genomic sequencing rather than on, say, drinking water quality is a decision concerning the orientation of scientific research, and it should have been subjected to public scrutiny and citizen participation, through a transparent process, before the money moved.

Are the conflicts of interest on the table? If Dr Iyer also advises the ministry that approved the funding, or if the sequencing contract goes to a company in which the programme's advisers have an interest, paragraph 53 requires that the interest be disclosed and regulated. Note that this obligation attaches whether or not anybody has behaved badly: the requirement is disclosure of actual or perceived conflicts.

Does Kavita's community get a say, or only a form to sign? If the community is an indigenous one, paragraph 40 requires genuine consultation in order to obtain free, prior and informed consent before the research is conducted or its knowledge used. That is a decision about whether the research happens there at all, and it is not answered by Kavita individually signing a consent form at the sampling tent.

Can Kavita take part in the science? Paragraph 10 says the State must actively facilitate citizen participation, not merely permit it. A programme that treats the sampled population purely as a source of material, and never as people who might learn the method, read the results or question the design, has satisfied the delivery half of the right and failed this half.

Are the results published? Paragraph 49 requires equitable and open access to scientific literature and data, particularly for research financed with public funds. A publicly funded database whose findings are available only to the institutions that hold it fails that obligation.

Distinctions

What it does not mean

It is not a vote on what is true. Paragraph 55 opens with the words "with due respect to scientific freedom". Participation governs decisions about the orientation of research and the adoption of technical advances. It does not put the content of a scientific finding to a show of hands, and paragraph 57 says in terms that the precautionary principle should not hinder and prevent scientific progress.

It is not the scientist's own freedom. The freedom indispensable for scientific research is a separate limb of article 15, stated in paragraph 3 of that article and worked in its own chapter. Participation is a right of the public, including the public who are not scientists; scientific freedom is a protection for the researcher against interference. They can pull against each other, which is precisely why paragraph 55 begins by acknowledging the tension.

It is not consent. Consent is the individual's control over an intervention on their own person or data, and it is exercised at the moment of the intervention. Participation is exercised earlier, collectively, over the design and direction of the enterprise. A study in which every participant signed a valid consent form may still have been decided on without any participation at all.

It is not satisfied by an announcement. Paragraph 55 requires participatory and transparent processes that are implemented with accountability mechanisms attached. A consultation held after the decision is taken, or one whose outcome nothing depends on, does not meet a requirement expressed in those terms.

Limits

The interpretation is authoritative, not binding. General Comment No. 25 is a document of the Committee on Economic, Social and Cultural Rights, adopted at its sixty-seventh session and issued as E/C.12/GC/25 on 30 April 2020. It is the treaty body's own reading of article 15, and it is the best evidence of what the provision requires, but it is not itself a treaty and it creates no new obligation. What binds a State is article 15, read as the Committee reads it.

Participation can be limited. Limitations on this right must satisfy article 4 of the Covenant, which permits only limitations determined by law, compatible with the nature of the right, and solely for the purpose of promoting the general welfare in a democratic society. The Committee adds in paragraph 21 that a limitation must respect the minimum core of the right, must be proportionate, and where several means could achieve the aim, the least restrictive must be chosen.

There is no ready enforcement. The Covenant's ordinary machinery is State reporting to the Committee, and paragraph 2 of the General Comment records that science is one of the areas of the Covenant to which States parties give the least attention in their reports. A right that depends on being reported on, in the field States report on least, is weakly enforced, and an honest answer says so.

Quick revision

• The right in article 15(1)(b) is the right to participate in and to enjoy the benefits of scientific progress and its applications: General Comment 25, paragraph 11.
• The passive reading is rejected in paragraph 9: no rigid distinction between the scientist who produces and a population entitled only to receive.
• Four sources for the participatory limb, all in paragraph 10: article 15(1)(a) cultural life; benefits are wider than products; the drafting record and article 27 of the Universal Declaration; and the French, Spanish and Russian texts which say participate where the English says share.
• Four contents: doing science (citizen science, actively facilitated); deciding the direction of research (paragraph 55); information and disclosure of conflicts of interest (paragraphs 53, 54, 49); and participation in controlling risk (paragraphs 56 and 57).
• Excluded groups: paragraph 28. Women suffer a double discrimination, paragraph 30. Indigenous peoples require free, prior and informed consent, paragraph 40.
• Core obligations, paragraph 52: eliminate what undermines women's and girls' participation, and adopt a participatory national framework law with remedies.
• Limits: article 4 of the Covenant, proportionality and the least restrictive means; the General Comment interprets rather than binds.

Test yourself

1. Article 15(1)(b) says "to enjoy the benefits". Where does a right to participate come from? From four arguments in paragraph 10 of General Comment 25. Article 15(1)(a) recognises the right to take part in cultural life, scientific activity falls within cultural life, and taking part in cultural life therefore includes taking part in scientific progress and in decisions concerning its direction. The benefits of science are not only its material products but include the development of the critical mind associated with doing science. The drafting record shows article 15 was written to develop article 27 of the Universal Declaration, which includes participation. And the French, Spanish and Russian texts of that article, equally official, use the verb to participate where the English uses share.

2. Distinguish participation in scientific progress from informed consent. Consent is individual, is exercised at the moment of an intervention on a person or their data, and protects control over what is done to that person. Participation is collective as well as individual, is exercised before and during the enterprise, and protects a say in whether the research is done, in what direction and on what terms. A trial in which every participant gave valid consent may still have been designed and funded without any public participation, and the right in article 15(1)(b) would still be unsatisfied.

3. A State consults the public after it has signed the contract for a new technology. Is that participation? On the wording of paragraph 55, no. The paragraph requires that decisions concerning the orientation of research and the adoption of technical advances be subjected to public scrutiny and citizen participation, that policies be established through participatory and transparent processes, and that they be implemented with accompanying transparency and accountability mechanisms. A consultation after the decision has been taken is not a process by which the policy was established, and nothing turns on its outcome.

4. Why does the Committee treat the exclusion of women from science as two wrongs rather than one? Paragraph 30. First, women have the right to participate in scientific research on an equal footing with men, so unequal access to scientific education and careers is itself discrimination. Second, because women are underrepresented in research, the research that gets done and the technologies that result are commonly gender biased and insensitive to women's particularities and needs. The first wrong is done to the excluded women; the second is done to everyone who then relies on science built without them.

5. Does participation mean the public can overrule a scientific finding? No. Paragraph 55 begins "with due respect to scientific freedom", and paragraph 57 states that the precautionary principle should not hinder and prevent scientific progress. What is opened to participation is the orientation of research, the adoption of particular technical advancements and the acceptability of the risks they carry, not the content of what the evidence shows. The public decides what shall be pursued and what risks are worth running; it does not decide what is true.

Two things are worth noticing in that sentence. The benefit and the protection are stated together, in one obligation, which is the structure of this whole paper. And the reach is "all strata of the population", which makes the distribution internal to a State as well as between States. Paragraph 3 of the same Declaration requires States to ensure that scientific and technological achievements satisfy the material and spiritual needs of all sectors of the population, and paragraph 5 requires co-operation in building the scientific capacity of developing countries. The nine paragraphs are taken one by one in Module IV.

The UNESCO declarations put it as a principle of bioethics. Article 15 of the 2005 Bioethics Declaration and Article 19 of the 2003 Genetic Data Declaration are both headed Sharing of benefits, and each sets out a list of the forms a benefit may take. Article 12(a) of the 1997 Genome Declaration says that benefits from advances in biology, genetics and medicine concerning the human genome shall be made available to all, with due regard for the dignity and human rights of each individual. Those three articles are worked in their own chapters in Module III; what this chapter takes from them is the principle they share.

General Comment 25 puts it inside the right to science. Paragraph 80 states the principle as one of the four justifications for the reinforced duty of international co-operation:

Third, the benefits and applications resulting from scientific progress should be shared, with due incentives and regulations, with the international community, particularly with developing countries, communities living in poverty and groups with special needs and vulnerabilities, especially when the benefits are closely related to the enjoyment of economic, social and cultural rights.

That placement matters for an examination answer. It means benefit sharing is not merely a rule of research ethics that happens to sit in some UNESCO texts. It is part of what article 15 of the Covenant requires, and article 15 is a treaty obligation.

Broken down: five questions the principle answers

One, shared with whom? With three groups, and a student who names only one has answered a third of the question. With society as a whole, because the knowledge is a common product. With the international community, particularly developing countries, because scientific capacity is unequally distributed between States. And with the persons and groups who took part in the research, because they carried the risk personally.

Two, what counts as a benefit? Not money, principally. Both UNESCO articles list forms, and read together they run to much the same thing: continuing assistance to, and acknowledgement of, the people and groups who took part; access to health care; the provision of the new diagnostic methods, treatments or products that came out of the research; support for health services; access to scientific and technological knowledge; and facilities that build the capacity of the host country to do research of its own. The 2003 Declaration adds, in terms, the strengthening of developing countries' own capacity to collect and process genetic data.

Three, may the benefit be cash? The 2005 Declaration answers this with a warning rather than a prohibition: benefits must not become an improper inducement to take part in research. A payment large enough to make a poor person disregard a risk they would otherwise refuse has not shared a benefit, it has bought a consent. That is why the enumerated forms are weighted towards health care, knowledge and capacity rather than towards a cheque.

Four, when is the benefit owed? The principle is engaged before the research begins and not only after it succeeds. The 2005 Declaration requires that when a research agreement is negotiated, the terms of the collaboration and the agreement on the benefits be settled with equal participation by the parties to the negotiation. Benefit sharing agreed after the result is known is agreed by the party that now holds all the cards.

Five, who owes the duty? States, in the first place, under the 1975 Declaration and under article 15 of the Covenant. But the 2005 Declaration is addressed to States and also gives guidance to individuals, groups, communities, institutions and corporations, public and private, so the funder and the institution carry it too. General Comment 25 adds an extraterritorial obligation at paragraph 84: a State must regulate and monitor the conduct of multinational companies over which it can exercise control, so that they exercise due diligence to respect this right when acting abroad, and it must provide remedies, including judicial remedies, for their victims.

Research done somewhere else

The hardest form of the problem is research carried out among a population that is not the researchers' own, which is the ordinary case when a wealthy institution studies a disease that is common somewhere poor.

General Comment 25 addresses it at paragraph 22: when research is done in countries or among populations different from those of the researchers, the State of origin must guarantee the rights and obligations of all parties involved. The obligation follows the researchers home; it does not stop at the border of the country where the work was done.

The 2005 Declaration adds two requirements of the same kind. Research funded from one State and carried out in another should be reviewed for ethics in both. And transnational health research should be responsive to the needs of the host countries, with the importance of research contributing to the alleviation of urgent global health problems recognised. A trial that recruits in a district for a condition common in that district, and produces a product that will never be sold there, fails that requirement squarely.

Who owns what science produces

Benefit sharing forces a prior question, and the instruments answer it in three places.

The human genome is not property. Article 4 of the 1997 Genome Declaration provides that the human genome in its natural state shall not give rise to financial gains, and Article 1 calls it, in a symbolic sense, the heritage of humanity. Something that is nobody's private property cannot generate a private entitlement to the profits of reading it. That is the strongest ownership statement in the whole authority base, and its own chapter in Module III works it out.

Traditional knowledge belongs to the communities that hold it. Paragraph 39 of General Comment 25 requires States to protect local, traditional and indigenous knowledge, including through special intellectual property regimes, and to secure the ownership and control of that knowledge by the communities and indigenous peoples concerned. The Indian examples of what happens when this fails are worked in the chapter on biopiracy and traditional knowledge.

Contribution must be credited. Paragraph 83 requires States, when negotiating international agreements or adopting a domestic intellectual property regime, to ensure that traditional knowledge is protected and that contributions to scientific knowledge are appropriately credited. Credit is itself a form of benefit, and it is the one most often withheld from the people at the collecting end of a study.

The intellectual property problem, at the level of principle

Benefit sharing and intellectual property are not natural enemies, and an answer that treats them as such is too simple. Paragraph 60 of General Comment 25 puts both sides: intellectual property enhances the development of science by giving inventors an economic incentive, and it can also work against the advancement of science and against access to its benefits in three specific ways.

The three, from paragraph 61, are worth learning as a set. It distorts what gets funded, because private money follows what is profitable and neglects what is not, which is what has happened with the diseases the Committee calls neglected. It restricts the sharing of research information for a period, through devices such as data exclusivity, and prices scientific publications beyond the reach of low-income researchers. And it can put the product itself out of reach, because a patent holder has a temporary exclusive right and can set the price, which is what has happened with new medicines essential to the life of persons with certain diseases.

Paragraph 62 states the answer as a balance rather than an abolition. States should fund the research that the market neglects, if necessary through international co-operation; they should guarantee the social dimensions of intellectual property in their national law and in international agreements; and the Committee reiterates that intellectual property is a social product with a social function, so States have a duty to prevent unreasonably high costs for essential medicines, seeds and learning materials from undermining the rights of large parts of the population. How that plays out in Indian patent law is the subject of a later chapter of this module.

A worked example

A company incorporated abroad funds a trial of a new treatment for a parasitic disease. The trial runs at a district hospital in India. Sunanda, a farm labourer, enrols. Two hundred others in her district enrol with her. The treatment works, is patented, and is launched at a price no one in the district can pay.

Has anything unlawful happened? Possibly nothing, on the ordinary law. Sunanda consented, was treated according to the protocol, and suffered no injury. That is exactly why benefit sharing is stated separately: the wrong it addresses survives a lawful trial and a valid consent.

Which claim does Sunanda have? The claim of a person who took part in the research. The forms of benefit contemplated are continuing assistance and acknowledgement, access to health care, and provision of the treatment that came out of the research. A trial that ends the day the last data point is collected, leaving the participants with nothing, is the situation the enumerated forms are aimed at.

Which claim does the district have? The claim of the host population. The 2005 Declaration requires transnational health research to be responsive to the needs of host countries. A trial run in the district for a disease of the district, whose product is priced for another market, is not responsive to those needs.

Which claim does India have? The claim of a developing country under paragraph 80 of General Comment 25, that benefits and applications resulting from scientific progress be shared with the international community and particularly with developing countries, especially where the benefits are closely connected to economic, social and cultural rights. Treatment for a disease that disables labourers is such a benefit.

When should this have been settled? At the negotiation, not now. The 2005 Declaration requires terms for collaboration and agreement on benefits to be established with equal participation by the parties to the negotiation, and the ethical review should have taken place in the funder's State as well as in India.

Who can be made to answer? India, as the State whose obligation under article 15 of the Covenant is engaged. And the funder's State, under paragraph 84, whose extraterritorial obligation is to regulate and monitor a company over which it can exercise control and to provide remedies for its victims.

Distinctions

What beginners get wrong

Treating benefit sharing as a money transfer. The enumerated forms in both UNESCO articles are mostly not money: access to care, provision of the resulting treatment, support for health services, access to knowledge, and capacity building. Cash is the form the instruments are most cautious about, because of the inducement problem.

Confusing it with consent. Consent decides whether this person may be included. Benefit sharing decides what is owed once the research has produced something, and it is owed to people who never signed anything, including the population of a country that hosted the work. Both can be satisfied and the other still breached.

Treating it as purely international. Paragraph 6 of the 1975 Declaration requires the benefits to be extended to all strata of the population, which is a duty a State owes internally. A country that acquires a technology and delivers it only to its richest cities has failed the principle without any foreign party being involved.

Saying intellectual property violates human rights. The Committee does not say that. It says intellectual property has positive and negative effects, identifies three specific negative ones, and calls for a balance in which access and the sharing of knowledge are guaranteed alongside the incentive. Stating the balance is the stronger answer.

Limits and criticism

The instruments say "should". Article 15 of the 2005 Declaration and Article 19 of the 2003 Declaration are drafted in the language of what ought to happen, and both instruments are declarations of the UNESCO General Conference rather than treaties. Article 19(b) of the 2003 Declaration expressly allows limitations to be provided by domestic law and international agreements, which leaves a wide door open.

The forms are open ended. Each list ends with a residual item covering any other form of benefit consistent with the declaration's principles. That flexibility is useful, because what a community needs varies, and it also means that a token gesture can be presented as compliance. Nothing in either instrument measures adequacy.

There is no machinery. No body receives complaints about a breach of benefit sharing. The nearest thing to enforcement is that the principle now forms part of the Committee's reading of article 15 of the Covenant, so a State's failure can be raised in its periodic report, and paragraph 2 of General Comment 25 records that science is the area States report on least.

The duty falls on States and the benefit is held by private parties. International human rights obligations bind States. The patents, the data and the manufacturing capacity are mostly held by companies. General Comment 25 answers this with the duty to protect and the extraterritorial duty in paragraph 84, but a duty to regulate is a weaker instrument than a duty to deliver, and the gap is the standing criticism of this whole area.

Quick revision

• Benefit sharing: the benefits of scientific research and its applications are to be shared with society as a whole, with the international community and particularly developing countries, and with the persons and groups who took part.
• Stated as Sharing of benefits in Article 15 of the UNESCO Bioethics Declaration 2005 and Article 19 of the UNESCO Genetic Data Declaration 2003; as benefits "made available to all" in Article 12(a) of the Genome Declaration 1997; as a State duty in paragraph 6 of the 1975 Declaration; and inside the right to science at paragraph 80 of General Comment 25.
• Forms of benefit: assistance and acknowledgement for participants, access to health care, provision of the resulting products, support for health services, access to knowledge, capacity building.
• Benefits must not become an improper inducement to take part in research.
• Terms must be settled at the negotiation, with equal participation, not after the result is known.
• Ownership: the human genome in its natural state shall not give rise to financial gains (Genome Declaration, Article 4); traditional knowledge is owned and controlled by the communities that hold it (General Comment 25, paragraph 39); contributions must be credited (paragraph 83).
• Intellectual property has three identified negative effects at paragraph 61: distorted funding, restricted sharing, and price. Paragraph 62 answers with balance, and calls intellectual property a social product with a social function.
• Extraterritorial duty at paragraph 84: regulate and monitor multinational companies, and provide remedies.

Test yourself

1. State the principle of benefit sharing and name the instruments in which it appears. Benefits resulting from scientific research and its applications should be shared with society as a whole and with the international community, particularly developing countries and communities living in poverty, and with the persons and groups who took part in the research. It appears as Article 15 of the UNESCO Universal Declaration on Bioethics and Human Rights 2005 and Article 19 of the UNESCO International Declaration on Human Genetic Data 2003, both headed Sharing of benefits; as Article 12(a) of the 1997 Genome Declaration; as paragraph 6 of the 1975 Declaration on the Use of Scientific and Technological Progress; and as paragraph 80 of General Comment No. 25 on article 15 of the Covenant.

2. A company pays each trial participant a large sum and provides nothing else. Has it shared the benefits? No, and it may have made the position worse. The 2005 Declaration warns that benefits must not constitute improper inducements to participate in research, and a payment large enough to lead a poor participant to accept a risk they would otherwise refuse is such an inducement. Benefit sharing is owed in addition to whatever a participant is paid for their time, is owed to people who were never enrolled, including the host community and the host country, and takes forms such as continuing access to the resulting treatment, support for health services and capacity building.

3. Why is it said that benefit sharing must be agreed before the research and not after it? Because the bargaining positions are not equal once the result is known. The 2005 Declaration requires that when a research agreement is negotiated, the terms of collaboration and the agreement on benefits be established with equal participation by the parties. After a successful trial the sponsor holds the data, the patent and the manufacturing capacity, and the host has nothing left to trade, so an agreement made then is not an agreement between equals.

4. How does General Comment 25 reconcile intellectual property with the right to science? Not by rejecting it. Paragraph 60 accepts that intellectual property gives incentives that drive research, and identifies three ways it can also work against the right: it distorts funding towards profitable subjects and away from neglected diseases; it restricts the sharing of research information, through devices such as data exclusivity and through the price of publications; and it can place the product beyond the reach of poor patients and poor countries. Paragraph 62 requires States to fund neglected research, to guarantee the social dimensions of intellectual property in domestic and international law, and to prevent unreasonably high costs for essential medicines, seeds and learning materials from undermining rights, because intellectual property is a social product with a social function.

5. A State imports a new diagnostic technology and installs it only in private hospitals in its three largest cities. Has benefit sharing been breached, given that no foreign party is complaining? Yes, on the wording of paragraph 6 of the 1975 Declaration, which requires States to extend the benefits of science and technology to all strata of the population. The obligation is internal as well as international. Paragraph 47 of General Comment 25 points the same way, requiring States to use the maximum of their available resources to overcome the hurdles that any person faces in benefiting from new technologies, with particular relevance for disadvantaged and marginalised groups.

Read paragraphs 6 to 8 together and the structure of the whole subject is visible: extend the benefits, prevent the harms, and do both by law.

The Covenant carries the same idea through its limitation clause rather than in terms. Article 15(1)(b) gives the right to enjoy the benefits. Article 4 permits limitations, and General Comment 25 applies it to science.

The limitation test

General Comment 25 is explicit that limits on this right may be necessary, because science and its applications can in certain contexts affect economic, social and cultural rights. Any limitation must satisfy article 4 of the Covenant:

1. It must be determined by law.
2. It must promote the general welfare in a democratic society.
3. It must be compatible with the nature of the right restricted.

The Committee reads the third requirement as importing two further constraints: limitations must respect the minimum core obligations of the right, and they must be proportionate to the aim pursued. Where several means are reasonably capable of achieving the legitimate aim, the least restrictive must be selected, and the burdens imposed on enjoyment of the right should not outweigh the benefits of the limitation.

Paragraph 22 then distinguishes three kinds of limit, and the distinction is the most useful thing in the chapter:

That last row is the point at which the framework refuses to let dual use become an excuse. A State may regulate what is done with knowledge far more readily than it may forbid the acquisition of knowledge.

Worked example

A university group develops a method for reconstructing a person's approximate facial appearance from a DNA sample. Two applications follow at once. A police force wants it to generate a likeness of an unknown suspect from material at a scene. A humanitarian organisation wants it to help identify the dead after a disaster so families can be informed.

One technology, two uses. There is no separate humanitarian version. The same method, the same reference database, the same laboratory.

Which rights are engaged, on each side? For the humanitarian use: the families' interest in knowing the fate of a relative, and the dignity of the dead. For the police use: the presumption of innocence, privacy, and equality, since the method's error characteristics will not be uniform across populations and a likeness that is systematically wrong for one group produces investigations that fall unequally.

Apply the limitation test to a proposal to permit the police use. Is it determined by law, or is it an operational decision taken inside a police force? Does it pursue a legitimate aim? Is it proportionate: is there a less restrictive means reasonably capable of achieving the same aim, and does the burden on the persons wrongly implicated outweigh the investigative benefit?

Note which limb of paragraph 22 is engaged. A rule restricting the police use limits an application, which is ordinary regulation. A rule forbidding the university from publishing the method limits the content of research, which carries a strict burden of justification. A candidate who treats those two as the same restriction has missed the distinction the Comment draws.

And the other direction. If the technique is dependable for identifying the dead and the State declines to make it available after a disaster, the complaint is not about misuse at all. It is a failure to extend the benefit, contrary to paragraph 6 of the 1975 Declaration and article 15(1)(b).

What it does NOT mean

It does not mean that every technology is neutral. Neutrality is a claim about moral responsibility and it is usually made by people who would rather not accept any. Dual use is a claim about capability: the same capability can serve either end. Where a technology has a single realistic use and that use is the infliction of harm, calling it dual use is an evasion, and the treatment of weapons in [Weapons, Warfare and the Right to Life] proceeds on that footing.

It does not mean the two directions are equally likely. Which use predominates is a question of fact about incentives, cost and who owns the technology, and the answer is often not evenly balanced. Saying "it can be used well or badly" and stopping there is the weakest available answer.

It is not a reason to refuse to regulate. The argument that a rule is pointless because the knowledge is already public confuses knowledge with application, and paragraph 22 keeps them apart precisely so that the first is not used as an excuse about the second.

It is not a reason to suppress research either. The strict burden of justification on limiting the content of research is there to stop dual use becoming a general licence for prohibition.

Limits and criticism

The framework tells you what questions to ask and not what answers to give. Whether a particular surveillance capability is proportionate is not settled by article 4; it is argued under it. That is a fair criticism of the whole limitation structure and it applies here.

The harder criticism is one of timing. Dual use is most easily managed early, when a technology is expensive and few people have it, and that is exactly when its harmful uses are hardest to foresee. By the time the harm is obvious the capability is cheap and widespread and the regulatory moment has passed. The 1975 Declaration's insistence in paragraphs 7 and 8 on legislative measures reads, in that light, as a demand that States act before they are certain, which is the same instinct as the precautionary principle discussed in [The Precautionary Principle].

Quick revision

• Dual use: one body of knowledge or one technique both advances rights and defeats them, so the analysis must be of a use by an actor in a setting, never of the technology in the abstract.
• A State can breach the right in either direction: by allowing harmful use, and by withholding a benefit.
• 1975 Declaration: paragraph 1 the beneficial direction; paragraph 2 prevent use, particularly by State organs, to limit or interfere with rights; paragraph 6 both directions in one sentence, naming privacy and physical and intellectual integrity; paragraphs 7 and 8 require measures including legislative measures.
• ICESCR article 4 governs limitations: determined by law; promoting the general welfare in a democratic society; compatible with the nature of the right. The Committee adds respect for the minimum core and proportionality, with the least restrictive means and a weighing of burdens against benefits.
• General Comment 25, paragraph 22 distinguishes limits on applications (safety and quality, human rights impact assessments), on the research process (dignity, integrity, consent; the State of origin's duty where research is done among other populations), and on the content of research, which carries a strict burden of justification.

Test yourself

1. Define dual use and explain why it prevents a rights analysis being done on a technology in the abstract. Dual use is the property by which a single body of scientific knowledge, or a technique or device derived from it, is capable both of advancing the enjoyment of human rights and of being used to defeat them. It prevents analysis in the abstract because the same capability yields opposite conclusions depending on who exercises it, against whom, for what purpose and with what safeguards. A question framed as whether a technology is compatible with human rights therefore has no answer; the answerable question is whether a particular use by a particular actor, with the safeguards proposed, satisfies the limitation test.

2. Show how the 1975 Declaration contains both directions, citing paragraphs. Paragraph 1 requires States to promote international co-operation so that the results of scientific and technological developments are used for peace and security, for economic and social development and for the realisation of human rights. Paragraph 2 requires appropriate measures to prevent the use of such developments, particularly by State organs, to limit or interfere with the enjoyment of human rights. Paragraph 6 states both together, requiring States to extend the benefits of science and technology to all strata of the population and to protect them from the harmful effects of misuse, with particular regard to privacy and the physical and intellectual integrity of the person. Paragraphs 7 and 8 then require measures, including legislative measures, to secure the beneficial use without discrimination and to prevent detrimental use.

3. Set out the three requirements of article 4 of the Covenant and the two constraints the Committee reads into the third. A limitation must be determined by law; it must promote the general welfare in a democratic society; and it must be compatible with the nature of the right restricted. The Committee reads the third as requiring, first, that limitations respect the minimum core obligations of the right, and second, that they be proportionate to the aim pursued, which means that where several means are reasonably capable of achieving the legitimate aim the least restrictive to economic, social and cultural rights must be chosen, and that the burdens imposed on enjoyment must not outweigh the benefits of the limitation.

4. A government proposes both to prohibit the sale of a genetic testing kit and to prohibit publication of the method by which it works. Are the two proposals assessed alike? No. The first limits an application of science, which paragraph 22 treats as ordinary regulation available to guarantee the safety and quality of products used by persons, and for which human rights impact assessments may be appropriate. The second limits the content of scientific research, which the same paragraph says implies a strict burden of justification by the State in order to avoid infringing freedom of research. Both must satisfy article 4, but the second faces a materially heavier justification, because the framework guards the acquisition of knowledge more jealously than its commercial application.

5. Why is the claim that a technology is "neutral" not the same as the claim that it is dual use, and when does the neutrality claim fail? Dual use is a statement about capability, that the same means can serve opposite ends. Neutrality is a statement about responsibility, that the person who supplies the means bears none for what is done with them, and it does not follow from dual use. The neutrality claim fails wherever a technology has one realistic use and that use is the infliction of harm, since there is then no second direction to point to; it also fails wherever the supplier knows the use to which the capability will in fact be put, because the duty to protect under the 1975 Declaration's paragraphs 2 and 8 is engaged by foreseeable misuse and not only by intended misuse.

"To develop." Not "to have" and not "to possess". The verb contemplates something acquired over time and by effort, which is why the duty reads naturally as a duty about education and habit rather than about a state of mind a person either has or lacks.

"The scientific temper." This is the phrase the examiner will ask about, so it has to be pinned down. It does not mean scientific knowledge, and it is not a duty to study science. A temper is a disposition, a settled way of approaching a question. The scientific temper is the habit of asking what the evidence for a claim is, of preferring an explanation that has been tested to one that has not, and of holding a belief open to correction when the evidence changes.

"Humanism." The placing of human welfare and human dignity at the centre of judgement. Its presence in the same clause is what keeps the scientific temper from being read as a licence for anything that can be measured, which is precisely the failure Module III exists to answer.

"The spirit of inquiry and reform." Two dispositions again, and they are the practical end of the clause. Inquiry is the willingness to ask a question about a settled practice. Reform is the willingness to change the practice when the answer shows it to be wrong. Inquiry without reform produces nothing, and reform without inquiry is merely change.

Four words, one duty. Clause (h) is not four separate obligations. Reading them together is what stops "scientific temper" from collapsing into "technical training", and it is the reading a good answer states expressly.

Where it sits, and when it arrived

The list therefore began with ten duties and has had eleven since 2010. The addition of clause (k) is the only change the Part has undergone, and it is a useful thing to know because it shows that Parliament treats the list as capable of growth.

Can it be enforced?

This is the question the article is really examined on, and the Constitution answers it by what it does not say.

Part III carries its own remedy. Article 32 guarantees the right to move the Supreme Court by appropriate proceedings for the enforcement of the rights conferred by that Part, and empowers the Court to issue directions, orders and writs.

Part IV disclaims enforcement in terms. Article 37 provides that the provisions of that Part "shall not be enforceable by any court, but the principles therein laid down are nevertheless fundamental in the governance of the country and it shall be the duty of the State to apply these principles in making laws".

Part IVA does neither. It contains no machinery of its own, no sentence like article 37, and no penalty. Article 51A states duties and stops. Three consequences follow, and a student should state all three rather than only the first. No court can be moved to compel a citizen to develop a scientific temper. Breach of the clause is not by itself an offence, because the article creates none. And because the duty lies on the citizen, it is not a foundation for a claim against the Government either.

What is left is not nothing. The clause is part of the text of the Constitution, and a litigant may argue that another provision should be read in a way that is consistent with it, or that a law made to give it content is supported by it. That is an argument about how the Constitution should be construed. It is not a rule of enforcement, and it should not be written up as one.

The State-facing half of the same idea

The Constitution does not leave science entirely to the citizen. Article 48 directs the State to organise agriculture and animal husbandry "on modern and scientific lines", which is a directive principle and so falls under article 37.

International law puts the corresponding duty squarely on the State. General Comment No. 25 of the Committee on Economic, Social and Cultural Rights, adopted in 2020, states at paragraph 42 that the obligation to respect the right to science requires a State to refrain from disinformation, disparagement or deliberate misinformation intended to erode citizen understanding of and respect for science and scientific research. Paragraph 43 requires the State to ensure that private persons and entities do not disseminate false or misleading scientific information. Paragraph 44 requires protection against messages from pseudoscience, meaning claims presented as scientific that the scientific community does not accept, and it gives as its example parents who refuse vaccination for their children on grounds the scientific community considers false.

Put those beside clause (h) and the picture is complete. The Covenant obliges the State not to corrupt public understanding and to protect people from those who would. The Constitution asks the citizen to bring a disposition that makes the protection work. The general treatment of General Comment 25 belongs to [General Comment 25 and the Content of the Right]; what matters here is that the two instruments meet on this point.

A worked example

Sunil Rane runs an unregistered clinic in a small town and advertises a mineral treatment that he says cures diabetes. He tells patients to stop taking insulin. Anjali Deshmukh's father is one of them, and after he is hospitalised she wants the practice stopped. She has read article 51A(h) and asks what it does for her.

Against Sunil Rane, directly, it does nothing. Clause (h) imposes a duty on him as a citizen. It confers no right on Anjali, states no sanction, and Part IVA supplies no machinery. A petition asking a court to enforce clause (h) against him has nowhere to go.

Against the State it does nothing either. The duty is addressed to citizens, so it is not a source of an obligation on the Government to act.

The enforceable law is elsewhere, and there is plenty of it. Article 21 and the obligation it casts on the State to preserve life are treated in [Science Under Article 21]. The regulation of drugs, of medical practitioners and of clinical establishments, and the liability of a person who holds himself out as competent to treat, are the subject matter of Module II.

What clause (h) contributes is the reason. It states, in the Constitution's own words, the interest that all of that machinery exists to protect, and it lets Anjali say that the Constitution itself treats the disposition Sunil Rane is exploiting as a matter of constitutional concern. General Comment 25 paragraph 43 points the same way, because it requires the State to ensure that private entities do not disseminate false or misleading scientific information.

The honest conclusion. Article 51A(h) explains why the problem is a constitutional problem. It is not the provision that solves it. An answer that says otherwise is overstating the article, and an answer that ignores the article altogether has missed why the Constitution mentions science at all.

Distinctions

What beginners get wrong

Reading it as a duty to study science. It is a duty about how a claim is examined, not about what a person knows. A person with no formal education can satisfy it and a person with a science degree can fail it.

Saying that unscientific belief is unconstitutional. The article creates a duty, not an offence, and the Constitution's own guarantees of conscience, belief and expression sit in the same document. What the clause states is an expectation, and reading it as a prohibition contradicts both its wording and the rest of Part III.

Using it to demand that the Government fund research. That claim exists, but it comes from article 15(2) of the Covenant and is dealt with in [The Duty to Conserve, Develop and Diffuse Science]. Article 51A(h) is addressed to citizens and cannot carry it.

Treating "not enforceable" as "of no effect". The absence of machinery is a statement about remedies. It is not a statement that the words are not part of the Constitution, and the two should not be run together in an answer.

Limits and criticism

No definition. The Constitution does not say what a scientific temper is. The content has to be built from the words the phrase is joined to, which is why the clause has to be read whole.

No sanction and no machinery. A duty with neither is an unusual legal instrument, and the standing criticism of Part IVA is that it states aspirations in the register of obligation.

Citizens only. The duty binds citizens. A subject whose harms fall on everybody within the territory is oddly served by a provision limited by nationality.

The company it keeps. Clause (h) sits in a list that also asks citizens to respect the National Flag and to defend the country when called upon. The variety makes it easy to read the whole Part as exhortation and to skip past the one clause in it that has real work to do in this paper.

It binds the wrong party for the biggest problems. The decisions this subject is most worried about, whether to deploy a surveillance system, whether to license a drug, whether to fund a line of research, are taken by governments and companies, not by individual citizens. Clause (h) reaches none of them.

Quick revision

• Article 51A(h): the duty of every citizen of India to develop the scientific temper, humanism and the spirit of inquiry and reform.
• Part IVA, inserted by the Forty-second Amendment, 1976, in force from 3 January 1977. Clause (k) added by the Eighty-sixth Amendment, 2002, in force from 1 April 2010.
• Addressed to the citizen, not the State. No machinery, no sanction, no equivalent of article 32 or article 37.
• Scientific temper is a disposition, not knowledge: evidence, testing, and a belief open to correction.
• Read all four words together: temper, humanism, inquiry, reform.
• The State-facing counterparts: article 48 in Part IV, and General Comment 25 paragraphs 42, 43 and 44 in international law.

Test yourself

1. State article 51A(h) and explain what "scientific temper" means in it. Article 51A(h) provides that it shall be the duty of every citizen of India to develop the scientific temper, humanism and the spirit of inquiry and reform. Scientific temper there means a disposition rather than a body of knowledge: the habit of asking what the evidence for a claim is, of preferring a tested explanation to an untested one, and of holding a belief open to correction. It is read together with humanism and with the spirit of inquiry and reform, which is what prevents it from being reduced to technical training.

2. Can a court enforce article 51A(h), and how do you know? No, and the answer comes from the structure of the Constitution rather than from any words in the article. Part III carries its own remedy in article 32. Part IV disclaims enforcement expressly in article 37. Part IVA does neither: it supplies no machinery, states no sanction and creates no offence. So no court can be moved to compel a citizen to develop a scientific temper, and because the duty lies on the citizen it founds no claim against the State either. It remains part of the constitutional text and may be used in argument about how another provision should be read.

3. A private hospital promotes an untested treatment. Is article 51A(h) breached, and what follows? The clause is addressed to citizens and speaks of developing a disposition, so calling this a breach of it adds nothing that can be acted on. The article supplies no remedy against the hospital and none against the State. What it supplies is the constitutional statement of the interest at stake, and the enforceable law has to be found in article 21 and in the statutes regulating medicine, drugs and clinical establishments. Internationally, General Comment 25 paragraph 43 requires the State to ensure that private entities do not disseminate false or misleading scientific information.

4. How does article 51A(h) fit the international framework on science? It is the citizen-facing half of a duty the international instruments place on the State. General Comment 25 requires a State to refrain from disinformation intended to erode public understanding of science, to stop private entities spreading false scientific information, and to protect people from pseudoscience. The Constitution asks the citizen to develop the disposition that makes those protections effective. Neither half works alone: a State that tells the truth to a population with no habit of testing claims achieves little, and a scientifically minded population cannot check a State that controls the information.

Once the guarantees of Part III are read together rather than in isolation, the content of "life" and "personal liberty" stops being fixed by a list. That is the whole mechanism, and a student who can state it in those terms can explain any of the extensions that follow without having to memorise them one by one.

The first limb: life, and therefore health

Pt. Parmanand Katara v. Union of India, AIR 1989 SC 2039, (1989) 4 SCC 286, decided on 28 August 1989.

Facts. The petitioner, describing himself as a human rights activist, filed a writ petition in the public interest on the basis of a newspaper report. A scooterist had been knocked down by a speeding car. He was taken to the nearest hospital, where the doctors refused to attend on him and said he should be taken to another hospital some twenty kilometres away that was authorised to handle medico-legal cases, meaning cases in which a criminal investigation is likely. He died before he could be taken there. The petitioner asked for directions that an injured person be given medical aid at once, with the procedural criminal law operating afterwards. The Union of India told the Court that no provision of the Indian Penal Code, the Code of Criminal Procedure or the Motor Vehicles Act prevented a doctor from attending a seriously injured person before the police arrived.

Held. Article 21 casts the obligation on the State to preserve life. Preservation of human life is of paramount importance, because once life is lost the position cannot be restored. Whether the patient is an innocent person or a criminal, those in charge of the health of the community are obliged to preserve life, since social laws do not treat death by negligence as a form of legal punishment. Every doctor, at a government hospital or otherwise, has the professional obligation to extend his services with due expertise for protecting life. No law or State action can be allowed to avoid or delay the discharge of that obligation, and laws of procedure that would interfere with it must give way.

Why it matters here. It is the step from a capability to a duty. Medical science had made it possible to save the scooterist; the hospital that could have saved him was standing in front of him; and article 21 is what turns "could" into "must". That step is the entire relationship between science and rights compressed into one case, and it is why a paper on science and human rights begins its Indian material here. The detail of the obligation to treat is taken up in [Emergency Treatment and the Duty to Treat], and the wider right to health under this article in [The Right to Health Under Article 21].

The second limb: privacy, and therefore data

Justice K.S. Puttaswamy (Retd.) v. Union of India, AIR 2017 SC 4161, (2017) 10 SCC 1, decided on 24 August 2017 by nine judges.

Facts. A Bench of three judges hearing the challenge to the Aadhaar scheme recorded, in its order of 11 August 2015, that the norms for and the compilation of demographic and biometric data by the Government were being questioned as a violation of the right to privacy. The Attorney General argued that the existence of a fundamental right of privacy was itself in doubt, relying on observations in the two early decisions described above. The petitioners answered that both rested on an approach later held not to be good law. Because the decisions that had since recognised privacy were given by smaller Benches, the question was referred upwards, and on 18 July 2017 a Constitution Bench directed that it be placed before nine judges.

Held. The Order of the Court disposed of the reference in four propositions. The decision holding that the right to privacy is not protected by the Constitution stands overruled. The second early decision stands overruled to the extent that it so holds. The right to privacy is protected as an intrinsic part of the right to life and personal liberty under article 21 and as a part of the freedoms guaranteed by Part III. The later decisions that had taken that position lay down the correct law. In the judgment delivered for four of the nine judges, Chandrachud J recorded that the Court had to be sensitive to the opportunities and dangers posed to liberty in a digital world; that information is nonrivalrous, invisible and recombinant, so that a person may never learn that their data has been taken; that a State intervening to protect a legitimate interest must nonetheless put in place a regime satisfying a threefold requirement of a law, a legitimate State aim and proportionality; and that a data protection regime must not permit collection that discriminates on grounds including genetic or health status.

Why it matters here. The case exists because of a technology. Nobody asked the Supreme Court an abstract question about privacy; a biometric identification programme was built, and the question could no longer be avoided. That is the pattern this whole paper describes, and here it is visible in the procedural history of a single case. The result matters twice over for what follows: it puts personal data inside article 21, and by naming genetic and health status it reaches the subject matter of Modules II and III as well as Module IV. The threefold requirement is worked out in [Informational Privacy and the Proportionality Test], and the right itself in [Privacy as a Fundamental Right].

The third limb: the body and the mind

Selvi v. State of Karnataka, AIR 2010 SC 1974, (2010) 7 SCC 263, decided on 5 May 2010.

Facts. A group of criminal appeals raised the legality of administering three scientific investigative techniques to a person who had not consented: narcoanalysis, in which a drug is used to reduce a subject's inhibition against speaking; polygraph examination, which measures physiological responses while a subject answers questions; and the Brain Electrical Activation Profile test, which measures electrical activity in the brain on exposure to stimuli associated with a crime. Accused persons, suspects and witnesses had been subjected to them without consent. The State defended the practice on the footing that the techniques help where evidence is hard to gather by ordinary means, cause no bodily harm, would be used only to guide the investigation rather than led in evidence, and are a gentler alternative to the third degree, meaning the physical ill treatment of suspects in custody.

Held. Compulsory administration of the techniques violates the right against self incrimination in article 20(3), whose protection extends to the investigative stage and covers accused persons, suspects and witnesses alike; the results are testimonial in character and are not material evidence. Forcing a person to undergo any of them also violates the substantive due process required before personal liberty is restrained, and does so whether the technique is administered in an investigation or for any other purpose, because the results can expose a person to adverse consequences that are not penal. Compulsory administration is an unjustified intrusion into the mental privacy of the individual, would amount to cruel, inhuman or degrading treatment in the language of evolving international human rights norms, and conflicts with the right to a fair trial. A compelling public interest cannot justify diluting those rights. No individual should be forcibly subjected to any of the techniques. Voluntary administration remains possible with safeguards, but even then the results by themselves are inadmissible, because the subject does not exercise conscious control over the responses.

Why it matters here. It is the clearest Indian case of a scientific technique meeting a constitutional right directly, and it settles two things this paper needs. Article 21 protects the mind as well as the body, so a technology that reads a person without touching them is still an intrusion. And the usefulness of a technique to the State is not an answer to the person's refusal, which is the proposition that every later argument about surveillance, genetic testing and compelled disclosure rests on. The forensic setting is taken further in [Forensic Science and the Rights of the Accused].

Why this is the hook for the rest of the paper

The table is worth learning, because a question on this paper about almost any technology can be opened by identifying which limb it touches. That is the practical value of putting article 21 in Module I rather than leaving it to the module where each application is discussed.

A worked example

A State government proposes to require every person receiving a food subsidy to submit to an iris scan at the ration shop, and to store the scans centrally so that duplicate claims can be detected.

Which limb is engaged? Privacy, on the authority of the nine judge decision above, because an iris pattern is personal data and its central storage is informational privacy. Bodily integrity is engaged more weakly, because the scan is taken from the person but nothing is done to them.

Is the scheme therefore unconstitutional? No, and a student who stops at "privacy is a fundamental right" has answered half the question. Article 21 permits deprivation according to procedure established by law, and the threefold requirement recorded in that case asks whether there is a law, whether the aim is legitimate, and whether the means are proportionate to it. Preventing the diversion of a subsidy is capable of being a legitimate aim.

Where the argument actually happens. On proportionality, and on the design of the scheme rather than on its purpose. Is central storage necessary, or would matching on a card achieve the same result? What happens to a person whose scan fails to read? Is there a remedy? Those are the questions [Informational Privacy and the Proportionality Test] works through.

What article 21 has done. It has converted a policy argument about administration into a constitutional argument with a burden on the State. That conversion is the whole function of the article in this paper.

Limits, and what article 21 does not do

It binds the State. Part III defines "the State" to include the Government and Parliament of India, the Government and Legislature of each State, and all local or other authorities within the territory of India or under the control of the Government of India. A private hospital, a private laboratory or a technology company is not, without more, bound by article 21. They are reached through statute, through the general law of wrongs, and through the State's own duty to make law that protects, which is why Module II is a module about Acts and not about the Constitution.

It permits deprivation. The article does not forbid interference with life or liberty. It requires a procedure established by law, and the case law above adds that the law must pursue a legitimate aim and be proportionate. An answer that treats article 21 as an absolute bar will be wrong on every realistic problem.

It is not a budget. A right that the State must not take away is easier to enforce than a claim that the State must provide. Article 21 works powerfully against action and less predictably against inaction, which is why the positive obligations in this subject are usually argued from the Covenant and from the directive principles alongside it. Those are taken up in [The Directive Principles on Health and Work].

What beginners get wrong

Saying article 21 was amended to include privacy or health. It was never amended. What changed was the reading of Part III as an integrated whole rather than as a set of watertight compartments.

Treating a Supreme Court decision as adding a new article. The decisions above did not create rights outside the text. They held what the words "life" and "personal liberty" already contain, which is why the reasoning matters more than the result.

Using article 21 against a private company. The Constitution restrains the State. The route to a private actor runs through legislation and through the general law, and confusing the two is the commonest structural error in answers on this paper.

Reading the nine judge decision as deciding the Aadhaar case. It decided the reference on whether privacy is a fundamental right. The validity of the scheme itself was a separate matter, decided later and by a different Bench.

Quick revision

• Article 21: no person shall be deprived of his life or personal liberty except according to procedure established by law. Never amended. Protects every person.
• The Constitution has no separate article on health, privacy or bodily integrity. All three sit inside article 21.
• The mechanism: Part III is read as an integrated whole, not as separate compartments.
• Life and health: article 21 casts the obligation on the State to preserve life, and every doctor has the professional obligation to protect it.
• Privacy: protected as an intrinsic part of the right to life and personal liberty under article 21, held by nine judges in 2017, with a threefold requirement of law, legitimate aim and proportionality.
• Body and mind: compulsory narcoanalysis, polygraph and brain profiling violate article 20(3) and intrude on mental privacy; a compelling public interest is no answer.
• Limits: it binds the State, it permits lawful and proportionate deprivation, and it is stronger against action than against inaction.

Test yourself

1. Article 21 does not mention health, privacy or the body. How does it protect them? By interpretation, not amendment. The article protects life and personal liberty for every person, and Part III came to be read as an integrated whole rather than as a set of self-contained provisions, so the content of those two words is not fixed by any list. On that footing the Supreme Court held that article 21 casts an obligation on the State to preserve life, that privacy is intrinsic to the right to life and personal liberty, and that the article protects mental as well as physical integrity.

2. State the facts and holding of the nine judge privacy decision, and say why a science paper cares about it. A three judge Bench hearing the challenge to a biometric identification scheme referred the question upward after the Attorney General argued that no fundamental right of privacy existed, and a Constitution Bench placed it before nine judges. They held that the two early decisions denying constitutional protection to privacy stand overruled, and that the right to privacy is protected as an intrinsic part of the right to life and personal liberty under article 21 and as part of the freedoms in Part III. A science paper cares because the question was forced by a technology, and because the judgment brings personal data, including genetic and health data, inside the right to life.

3. Police wish to administer a lie detector test to a witness who refuses. Advise. The test may not be administered. Compulsory administration of a polygraph examination violates the right against self incrimination under article 20(3), whose protection extends to the investigative stage and covers witnesses as well as suspects and accused persons, and it violates the substantive due process required before personal liberty is restrained, because it is an unjustified intrusion into the mental privacy of the individual. That the results might be useful to the investigation is not an answer. If the witness later consents, the test may be conducted with safeguards, but the results by themselves would still be inadmissible, because the subject does not exercise conscious control over the responses.

4. Why is article 21 a weaker instrument against a private technology company than against a government department? Because Part III restrains the State, which is defined to include the Government and Parliament of India, the Government and Legislature of each State, and all local or other authorities in India or under the control of the Government of India. A private company is not within that definition, so a claim against it does not lie directly under article 21. It has to be brought under a statute, under the general law of wrongs, or by attacking the State's failure to make law that protects, which is why the regulation of medicine, data and technology in this book is statutory rather than constitutional.

Article 21A is a fundamental right, not a directive principle. Before 2002 education appeared in Part IV, in article 45, as a direction to the State to endeavour to provide free and compulsory education for all children until the age of fourteen within ten years. The Eighty-sixth Amendment moved the entitlement for the six to fourteen group into Part III, where it is enforceable, and substituted a new article 45 directing the State to provide early childhood care and education for children below six.

The age band is narrow. Article 21A stops at fourteen. Scientific education in any serious sense begins around that age and continues well past it, so the fundamental right reaches the foundation of the diffusion duty and not its upper storeys, which remain in Part IV and in policy.

What the Committee asks for

General Comment 25 treats education as an element of availability. Availability requires instruments for the diffusion of science, and the Comment names libraries, museums and internet networks, a strong research infrastructure with adequate resources, and adequate financing of scientific education.

Under accessibility it requires States to remove discriminatory barriers that impede persons from participating in scientific progress, for instance by facilitating the access of marginalised populations to scientific education.

It is also specific about women. States must take steps to ensure women's substantive equality in access to scientific education and careers, for example by raising public awareness in order to eliminate stereotypes that exclude women from science, and by adopting policies allowing both men and women to balance domestic life with scientific careers. It states that temporary special measures, such as quotas for women in scientific education, might be necessary to speed up the attainment of substantive equality, and that the availability of kindergartens and other childcare institutions is key to the advancement of equality.

The Comment adds a proposition that is easy to pass over and useful in an answer: a gender-sensitive approach is not a luxury for scientific research but a crucial tool, because it ensures that scientific progress and new technologies adequately take account of the characteristics and needs of women and girls. It should be present from the choice of subject and the design of methodology onwards, and not relegated to the last stages.

The Indian frame around article 21A

Article 41 directs the State, within the limits of its economic capacity and development, to make effective provision for securing the right to work, to education and to public assistance in cases of unemployment, old age, sickness and disablement.

Article 45, as substituted in 2002, directs the State to endeavour to provide early childhood care and education for all children until they complete the age of six.

Article 51A(k), the fundamental duty inserted by the same amendment, requires a parent or guardian to provide opportunities for education to his child or ward between the ages of six and fourteen. It is the mirror of article 21A: the State must provide, and the parent must not withhold.

Article 51A(h), the duty to develop the scientific temper, humanism and the spirit of inquiry and reform, supplies the content that distinguishes education in general from scientific education in particular.

Read together these make the Indian position on diffusion legible: an enforceable right for one age band, directive principles above and below it, and two fundamental duties that say what the education is for.

Worked example

A State government proposes to close the science stream in rural higher secondary schools where enrolment falls below a threshold, transferring those students to the arts stream locally or offering science at a residential school in the district town. The measure is defended as an efficient use of scarce laboratory equipment and qualified teachers.

Is article 21A engaged? Not directly. The students are above fourteen, so the fundamental right does not reach them, and this is the limit noted above.

Is the Covenant duty engaged? Yes. Article 15(2) requires steps necessary for the diffusion of science, and General Comment 25 names adequate financing of scientific education under availability and the removal of barriers to participation under accessibility. Withdrawing the subject from a rural area is a reduction in diffusion, and the Committee's general approach treats a deliberate backward step as requiring justification.

Who is affected, and does that matter? It does. The measure operates only on rural schools, and the residential alternative will in practice be taken up less by girls than by boys. The Comment's requirement of substantive equality in access to scientific education, and its acceptance that temporary special measures may be needed, are directly in point. Equality analysis of this kind is developed in [Equality and the Digital Divide].

What follows. The State is not obliged to run an under-enrolled laboratory in every village. It is obliged to show that it considered the effect on diffusion and on equality, and that the alternative genuinely preserves access rather than relocating it beyond reach. The distinction between a resource decision and an abdication is exactly what the four elements are for.

What it does NOT mean

Article 21A does not guarantee scientific education. It guarantees free and compulsory education for a defined age band and says nothing about content. The scientific character of that education comes from article 51A(h), from policy, and from the Covenant, not from article 21A itself.

Diffusion is not publication. Putting research findings on a public website satisfies part of availability. It does not by itself diffuse anything to a population that cannot read the language, reach the network or understand the material.

The duty is not discharged by the school system alone. The Comment names libraries, museums and internet networks in the same breath as scientific education. Diffusion is a set of institutions, not a single one.

Free and compulsory does not mean free at the point of use only. The obligation runs to the conditions that make attendance possible, which is why the Comment's reference to childcare is not a digression.

Quick revision

• ICESCR article 15(2): the steps taken shall include those necessary for the conservation, the development and the diffusion of science and culture. Mandatory in form, unlike article 15(4).
• Article 21A: free and compulsory education for children of six to fourteen, inserted by the Eighty-sixth Amendment, 2002, which moved the entitlement from a directive principle into Part III.
• The same amendment substituted article 45 (early childhood care and education below six) and inserted article 51A(k) (the parent's duty for six to fourteen). Article 41 covers the right to education within economic capacity; article 51A(h) supplies the scientific temper.
• General Comment 25: adequate financing of scientific education under availability; removal of barriers and facilitating access of marginalised populations to scientific education under accessibility.
• On women: awareness-raising against stereotypes, policies to balance domestic life with scientific careers, temporary special measures such as quotas may be necessary, childcare is key, and a gender-sensitive approach is not a luxury and belongs from the first stage of research.
• Article 21A stops at fourteen, which is where serious scientific education begins.

Test yourself

1. State article 15(2) and explain how its language differs in force from article 15(4). Article 15(2) provides that the steps to be taken to achieve the full realisation of the right shall include those necessary for the conservation, the development and the diffusion of science and culture. It is mandatory in form: it prescribes what the State's steps must contain. Article 15(4) provides only that States Parties recognise the benefits to be derived from encouraging international contacts and co-operation, which on its own is close to a statement of policy and acquires its force from article 2(1) and the UN Charter.

2. What did the Eighty-sixth Amendment do, and why does it matter that education moved from Part IV to Part III? It inserted article 21A, guaranteeing free and compulsory education to children between six and fourteen; substituted article 45, directing the State to endeavour to provide early childhood care and education below six; and inserted article 51A(k), making it a fundamental duty of a parent or guardian to provide opportunities for education to a child between six and fourteen. The move matters because a directive principle under Part IV is not enforceable by a court, whereas a fundamental right under Part III is, so the entitlement changed from a direction to the State into a claim a child can assert.

3. Why is article 21A an incomplete answer to the diffusion duty in article 15(2)? Because of its age band and its silence on content. It ends at fourteen, and scientific education in any developed sense begins around that age and continues into higher secondary and university study, none of which article 21A reaches. It also guarantees education without specifying its character, so nothing in the article requires that the education be scientific. The scientific content comes from article 51A(h), from policy and from the Covenant, and the stages above fourteen remain in Part IV and in executive policy rather than in Part III.

4. Set out what General Comment 25 requires of States in respect of women and scientific education. It requires steps to ensure women's substantive equality in access to scientific education and careers, giving as examples raising public awareness to eliminate stereotypes that exclude women from science, and policies enabling both men and women to balance domestic life with scientific careers. It states that temporary special measures such as quotas for women in scientific education might be necessary to speed up substantive equality, and that the availability of kindergartens and other childcare institutions is key. Separately it states that a gender-sensitive approach is not a luxury but a crucial tool, so that scientific progress takes account of the characteristics and needs of women and girls, and that it must be present from the choice of subject and the design of methodology onwards.

5. A State publishes all publicly funded research on a free website and argues that it has discharged the diffusion limb. Assess the argument. It has taken a step the Comment expressly asks for, since availability requires the promotion of open science and open source publication and public access to publicly funded findings and data. It has not discharged the limb. Diffusion under article 15(2) is a broader obligation, and the same paragraph names libraries, museums, internet networks, research infrastructure and adequate financing of scientific education as its instruments. Publication does not reach a person who cannot read the language of the paper, has no network access, or lacks the education to evaluate it, and the accessibility element separately requires removal of the barriers that prevent participation. Publication is one instrument of diffusion and the State has confused it with the whole duty.

Part IV is built for that shape of problem. It is addressed to the State, it speaks of what the State must secure rather than what it must not do, and it does not need a named victim before it operates. That is why a science and technology paper reaches for it. The rights in Part III answer the question what may not be done. Part IV answers the question what must be arranged, and technological harm is nearly always a failure of arrangement.

The provisions themselves

Article 39 opens by directing the State to secure six things. The full clause list matters, because two of the six carry this chapter and a third carries the next one on automation.

The State shall, in particular, direct its policy towards securing:

(a) that the citizens, men and women equally, have the right to an adequate means of livelihood;

(b) that the ownership and control of the material resources of the community are so distributed as best to subserve the common good;

(c) that the operation of the economic system does not result in the concentration of wealth and means of production to the common detriment;

(d) that there is equal pay for equal work for both men and women;

(e) that the health and strength of workers, men and women, and the tender age of children are not abused and that citizens are not forced by economic necessity to enter avocations unsuited to their age or strength;

(f) that children are given opportunities and facilities to develop in a healthy manner and in conditions of freedom and dignity and that childhood and youth are protected against exploitation and against moral and material abandonment.

Article 42 is a single sentence.

The State shall make provision for securing just and humane conditions of work and for maternity relief.

Article 47 is two sentences joined into one, and the second half is the part everybody quotes.

The State shall regard the raising of the level of nutrition and the standard of living of its people and the improvement of public health as among its primary duties and, in particular, the State shall endeavour to bring about prohibition of the consumption except for medicinal purposes of intoxicating drinks and of drugs which are injurious to health.

Clause (f) of article 39 is in its present form because of the Constitution (Forty-second Amendment) Act 1976, which took effect on 3 January 1977. Articles 42 and 47 carry no amendment footnote in the official text at all, so both stand exactly as they were adopted.

Broken down

Article 39(e), first limb: health and strength. The State must secure that the health and strength of workers are not abused. Two words are doing work. "Abused" is wider than injured: it covers wearing a person out by the conditions of the work, not only hurting them by an accident. And the clause protects strength as well as health, which is capacity to go on working, not merely the absence of disease. A process that leaves a worker able to do the job at thirty and unable at forty has abused their strength even if no doctor ever diagnoses anything.

Article 39(e), second limb: coerced choice. The State must secure that citizens are not forced by economic necessity to enter avocations unsuited to their age or strength. An avocation is an occupation or calling. The wrong this limb names is not that dangerous work exists. It is that poverty leaves a person no other option, so their consent to the danger is worth nothing. That is the exact shape of the argument about a technology which removes the safe jobs from a labour market and leaves only the unsafe ones.

Article 39(a): livelihood. Citizens, men and women equally, are to have the right to an adequate means of livelihood. This is the clause on which the whole automation argument rests, and it is worked out in [Work, Automation and Livelihood]. Note that it speaks of an adequate means of livelihood and not of a particular job, which is both its strength and its weakness: it survives the disappearance of an industry, and it promises nothing to the individual whose trade has gone.

Article 42: just and humane conditions of work. Two provisions in one line. "Just" is about the fairness of the terms, "humane" is about what the work does to the body and the mind of the person doing it, and maternity relief is a separate and specific instruction. Article 42 also uses a stronger verb than most of Part IV. It says the State shall make provision, where article 41 promises the right to work only "within the limits of its economic capacity and development" and article 47 falls back on "shall endeavour" for its prohibition limb.

Article 47: public health as a primary duty. Three objects are named: raising the level of nutrition, raising the standard of living, and improving public health. The article then calls these "among its primary duties", which is the strongest description Part IV gives to anything. The particular limb that follows requires the State to endeavour to prohibit the consumption of intoxicating drinks and of drugs which are injurious to health, except for medicinal purposes. That exception is a science and medicine carve-out written into the Constitution itself, and the statutory regime that draws the line between a medicine and a poison is taken in [Drugs, Standards and Control].

The three routes from these articles into this paper

Occupational health. A new process brings a new exposure, and the people exposed are the people who need the wage. Articles 39(e) and 42 are the constitutional statement that arranging against this is the State's job. The international mirror is article 7(b) of the Covenant on Economic, Social and Cultural Rights, which recognises the right of everyone to "safe and healthy working conditions".

Public health. Technologies change what people eat, drink, breathe and are treated with, usually faster than the regulator moves. Article 47 is the general duty; article 12(2)(b) of the same Covenant requires States to take steps towards "the improvement of all aspects of environmental and industrial hygiene", and [The Right to Health in the Covenant] takes that article properly.

Livelihood. Article 39(a) and the second limb of 39(e) between them cover the case where a technology does not injure anybody but takes away the work. The Committee on Economic, Social and Cultural Rights has said what it thinks that requires, in paragraph 76 of General Comment No. 25 of 2020.

States parties should adopt policies to ensure that those vulnerable to temporary and long-term job loss as a result of scientific and technological advances are provided with and encouraged to pursue vocational training and other job placement opportunities.

That sentence is worth learning, because it is the only place in the material this paper works from where automation and the loss of a livelihood are named as a human rights obligation rather than as an economic misfortune.

A worked example

Anjali works in a small garment unit in Bhiwandi. The owner installs an automated cutting machine which does the work of six people and uses a bonding agent that gives off fumes in a room with two windows. Four of the six cutters are let go. Anjali keeps her job, and within a year she has a cough that does not clear.

Which articles of Part IV are engaged? For the fumes, article 39(e), because her health and strength are being abused by the conditions of the work, and article 42, because the conditions are neither just nor humane. For the four who lost their jobs, article 39(a), because their means of livelihood has gone. For the district as a whole, article 47, if the same agent is in use in thirty other sheds.

What does Part IV give Anjali? Not a case. She cannot file a petition asking the court to enforce article 42 against the State, and article 37 is the reason. The book takes the enforceable route separately: the right to life is in [Science Under Article 21], and the ways a person actually recovers something are in [Remedies When Technology Causes Harm].

Then what is Part IV for, here? Two things, and both are examinable. First, it is the answer to who was supposed to act: not the machine, not the market, but the State, and its failure to have regulated the agent is a constitutional failure even though it is not an actionable one. Second, suppose the State does act, and the owner challenges the new rule as an unreasonable restriction on his freedom to carry on his business. Articles 39(e), 42 and 47 are the State's answer for why the restriction is a reasonable one. That is the work directive principles do most often and most usefully.

Distinctions

What beginners get wrong

Treating unenforceable as unimportant. Article 37 makes Part IV unenforceable and in the same breath calls it fundamental in the governance of the country and a duty in making laws. An answer that quotes only the first half has quoted half a sentence.

Calling article 47 the prohibition article. Prohibition of intoxicating drinks and of injurious drugs is a particular application tacked on at the end. The general duty is nutrition, standard of living and public health, and article 47 is the only article in Part IV to describe what it demands as a primary duty.

Calling article 39(e) the child labour clause. It protects the health and strength of workers of both sexes first, and the tender age of children second. Its subject is abuse of health and strength, and its second limb is about economic coercion, not about the employment of children as such.

Reading article 42 as maternity leave. Maternity relief is the second of the two provisions in it. The first, just and humane conditions of work, is the wider one and is the one a technology answer needs.

Limits and criticism

There is no remedy, and seventy years of breach produce nothing. A directive principle that has never been implemented leaves the person it was meant to help exactly where they were. That is the standing objection to Part IV and it should be stated in an answer rather than avoided.

The text is of 1950 and names the hazards of 1950. Neither automation nor data appears anywhere in Part IV. Article 39(a) reaches the loss of a livelihood only at one remove, by promising an adequate means rather than a job, and nothing in the Part speaks to a person displaced by a machine that works better than they do.

The verbs are inconsistent and nothing turns on it. Article 42 says the State shall make provision, article 41 confines the right to work to the limits of the State's economic capacity, and article 47 drops to shall endeavour for its prohibition limb. In an enforceable Part those differences would be litigated; here they are not, because none of them is enforceable.

The international framework has moved further than the Indian text. General Comment 25 names job loss caused by scientific and technological advance and attaches an obligation to it. An answer on this topic is stronger for running the constitutional and the international material together, rather than treating them as two separate questions.

Quick revision

• Article 37: Part IV is not enforceable by any court, but is fundamental in the governance of the country and a duty of the State in making laws.
• Article 39(e): health and strength of workers, and the tender age of children, not to be abused; citizens not to be forced by economic necessity into avocations unsuited to their age or strength.
• Article 39(a): an adequate means of livelihood for citizens, men and women equally. The automation clause.
• Article 42: just and humane conditions of work, and maternity relief.
• Article 47: raising nutrition, the standard of living and public health, described as among the State's primary duties; prohibition of intoxicating drinks and injurious drugs except for medicinal purposes.
• International mirrors: article 7(b) of the Covenant (safe and healthy working conditions) and article 12(2)(b) (environmental and industrial hygiene). General Comment 25, paragraph 76, on job loss from technological advance.
• Part IV's practical use: it identifies who had to act, and it justifies a law that restricts a business.

Test yourself

1. Set out article 37 and explain why a directive principle is still worth citing. Article 37 provides that the provisions of Part IV shall not be enforceable by any court, but that the principles laid down in it are nevertheless fundamental in the governance of the country, and that it shall be the duty of the State to apply them in making laws. It is worth citing for three reasons: it identifies the State as the body that was required to act, it explains the purpose of a statute passed to give effect to it, and it supplies the justification when a law made in pursuance of it is attacked as an unreasonable restriction on somebody's freedom to trade.

2. A chemical used in a new manufacturing process is found to damage the lungs of the workers who handle it. Which parts of article 39 and article 42 are engaged, and what do they give the workers? Article 39(e) is engaged in both its limbs: the health and strength of the workers are being abused, and if the local labour market offers nothing else, they have been forced by economic necessity into an avocation unsuited to their strength. Article 42 is engaged because the conditions of work are neither just nor humane. What the articles give the workers is not a cause of action, because article 37 excludes one. They give the constitutional statement that regulating the chemical was the State's duty, and they answer the employer who says the resulting regulation is an unreasonable interference with his business.

3. Why is article 47 unusual within Part IV? Because of the phrase "among its primary duties". Part IV elsewhere tells the State to secure, to provide, or to endeavour, and article 41 expressly limits its promise to the State's economic capacity. Article 47 alone characterises what it demands, namely the raising of nutrition and of the standard of living and the improvement of public health, as a primary duty of the State. It also contains an express medical exception, permitting the consumption of otherwise injurious drugs for medicinal purposes.

4. A factory replaces two hundred workers with machines. No one is injured. Is any part of Part IV engaged? Yes. Article 39(a) directs the State to secure that citizens have the right to an adequate means of livelihood, and the second limb of article 39(e) protects citizens from being forced by economic necessity into work unsuited to their age or strength, which is what an unemployed skilled worker is commonly pushed towards. Nothing in Part IV entitles anybody to keep a particular job. The international material goes one step further: paragraph 76 of General Comment 25 requires States to ensure that those vulnerable to job loss caused by scientific and technological advance are provided with vocational training and job placement opportunities.

5. Distinguish a directive principle from a fundamental duty. A directive principle is in Part IV, is addressed to the State, and instructs it about the ends its laws should serve. A fundamental duty is in Part IVA, is addressed to every citizen, and instructs the individual. Neither is enforceable by a court. They frequently state the two sides of one aim: the State is told to protect and improve the environment and the citizen is told to do the same, which is the pairing examined in [The Environment in the Constitution].

Broken down

"Shall endeavour." This is the weakest verb Part IV uses. Article 42 says the State shall make provision; article 47 calls its objects primary duties; article 48A only requires an attempt. Nothing turns on the difference in a court, because article 37 makes none of Part IV enforceable, but it matters in an answer: article 48A is a statement of direction, not of standard, and it fixes no level of environmental quality that the State must reach.

"Protect and improve." Two duties, not one. Protection is conservative: keep what is there from getting worse. Improvement is restorative: make it better than it is now. The second is the more demanding, and it is the one usually forgotten. A State that merely stopped further damage would have satisfied half of article 48A.

"The environment." The article does not define the word, and no other article of the Constitution defines it either. That is a real gap, and it is also why the provision has stayed useful: an undefined term of 1976 can absorb harms that nobody had heard of in 1976, which a defined term listing smoke, effluent and noise could not have done.

"The forests and wild life of the country." Two things are named specially, and only two. Neither pollution, nor water, nor air, nor soil, nor hazardous waste, nor any technology appears anywhere in the text of article 48A. Everything this paper is about therefore reaches article 48A through the general words, not through the specific ones.

Article 51A(g) is wider in its list and narrower in its subject. It speaks of the natural environment, and it names lakes and rivers, which article 48A does not. It then adds a duty that has no counterpart in Part IV at all: to have compassion for living creatures. But its subject is the individual citizen, and a duty on a citizen carries no sanction. What a fundamental duty is, and why India has them, is taken in [Scientific Temper as a Fundamental Duty].

What article 48A does not do

It confers no right on anybody. It is addressed to the State, and a person who wants a court to act cannot ask it to enforce article 48A, for the reason given in [The Directive Principles on Health and Work]. If there is an enforceable claim in an Indian environmental case, its textual home is elsewhere, and this book takes that route in [Science Under Article 21].

It does not bind the polluter. A factory is not the State, and article 48A says nothing to it. The route from a constitutional direction to a private defendant runs through legislation, which is why the answer to "what does article 48A do about a chemical plant?" is "it obliges the State to have passed and enforced a law about chemical plants". [The State's Duty to Regulate Technology] takes that step, and [Environment and Technological Harm] takes the industrial disaster that made the question urgent in India.

It fixes no standard and provides no remedy. It tells nobody how clean the air must be, and it gives nobody anything when the air is dirty.

The international material that says the same thing

Three instruments in this paper's authority base carry the same idea, and each is stronger than article 48A in one respect.

The Covenant on Economic, Social and Cultural Rights, in article 12(2)(b), requires States to take steps towards "the improvement of all aspects of environmental and industrial hygiene" as part of the right to health. That is a treaty obligation rather than an unenforceable direction, and it is examined in [The Right to Health in the Covenant].

The Universal Declaration on Bioethics and Human Rights of 2005 devotes its article 17 to the protection of the environment, the biosphere and biodiversity, requiring due regard to "the interconnection between human beings and other forms of life" and to the role of human beings in protecting them. It is taken in [The Bioethics Declaration: Society and the Future].

General Comment No. 25 of 2020 supplies the standard article 48A lacks. In paragraph 56 the Committee describes the harm that must be guarded against in advance of full scientific certainty as harm to humans or to the environment that is "threatening to human life or health", "serious and effectively irreversible", "inequitable to present or future generations", or "imposed without adequate consideration of the human rights of those affected". Those four descriptions are the nearest thing this subject has to a definition of unacceptable environmental risk, and they belong to [The Precautionary Principle].

A worked example

An electroplating unit at Taloja discharges rinse water into a creek. The discharge is within the limits of the consent the unit holds. Over eleven years the creek silts up, the fish die, and the families in Sonavane village who fished it move to daily wage work. A resident, Suresh, wants to know what the Constitution gives him.

Can he sue on article 48A? No. It is in Part IV, it is addressed to the State, and article 37 forecloses enforcement. If his petition says nothing else, it fails.

Does article 51A(g) help him against the unit? No. It is a duty on citizens and it carries no sanction, and in any event the unit is a company rather than a citizen exercising a duty. Article 51A(g) is a reason, not a remedy.

Where does article 48A actually bite? In three places. It tells him whom to complain about, namely the State that set the consent limits. It tells a court what the pollution statute was for, so that a doubtful provision in that statute is read in the way that protects the creek rather than the way that does not. And if the State tightens the consent limits and the unit challenges the tightening as ruinous, article 48A is the State's constitutional justification for having done it.

What must Suresh add to make a case? The right to life, and the statutory scheme, and a claim for compensation. All three are elsewhere in this book, and the point of this chapter is that article 48A never travels alone.

Distinctions

What beginners get wrong

Saying article 48A gives a right to a clean environment. It gives a direction to the State. Any right in this area has to be found in Part III, and locating it correctly is half of a good answer.

Swapping the two provisions. Article 48A binds the State and article 51A(g) binds the citizen. The mistake is easy because both use the words "protect and improve", and it is expensive because the whole point of quoting them together is that they are addressed to different people.

Thinking "shall endeavour" makes the article decorative. It weakens the language and it does not remove the direction. Part IV is unenforceable whatever verb it uses, so the softness of the verb changes far less than students assume.

Treating 1976 as the moment India acquired environmental law. The amendment put the subject into the Constitution. It did not create the statutory or common law machinery, and it did not by itself make anything actionable.

Limits and criticism

A direction without a standard. The article says protect and improve and stops. It does not say from what, to what level, or by when, and a duty with no measurable content cannot be failed in any way a court could recognise even if Part IV were enforceable.

The named subjects are the wrong ones for this paper. Forests and wild life are named; industrial effluent, hazardous waste, emissions, radiation, chemicals and every technology that produces them are not. A student answering a technology question from article 48A is arguing from the general words of a provision whose specific words are about something else.

Nothing about future generations. The most distinctive feature of technological environmental harm is that it lands on people who cannot be consulted because they do not yet exist. Article 48A does not mention them. General Comment 25 does, in the phrase "inequitable to present or future generations", and the international instruments in Module III go further still.

Nothing about the causer. The article obliges the State and says nothing to the person who did the damage. That is a design feature of Part IV rather than a defect in article 48A, but it means the constitutional provision on the environment is silent about polluters, which surprises every student who reads it for the first time.

Quick revision

• Article 48A: the State shall endeavour to protect and improve the environment and to safeguard the forests and wild life of the country. Part IV, unenforceable by article 37.
• Article 51A(g): the duty of every citizen to protect and improve the natural environment including forests, lakes, rivers and wild life, and to have compassion for living creatures. Part IVA, also unenforceable.
• Both inserted by the Constitution (Forty-second Amendment) Act 1976, sections 10 and 11, with effect from 3 January 1977. The original Constitution of 1950 said nothing about the environment.
• Article 48A names only forests and wild life. Pollution, waste, water, air and technology appear nowhere in it.
• Two duties in the words "protect and improve": one conservative, one restorative.
• International mirrors: Covenant article 12(2)(b), environmental and industrial hygiene; Bioethics Declaration 2005 article 17; General Comment 25 paragraph 56 on serious, effectively irreversible harm inequitable to present or future generations.
• Article 48A never travels alone. It identifies the duty bearer, colours the reading of a statute, and justifies a restriction.

Test yourself

1. State article 48A, say where it sits in the Constitution, and say how it got there. Article 48A provides that the State shall endeavour to protect and improve the environment and to safeguard the forests and wild life of the country. It is in Part IV, the Directive Principles of State Policy, so by article 37 it is not enforceable by any court although it is fundamental in the governance of the country. It was inserted by section 10 of the Constitution (Forty-second Amendment) Act 1976 and took effect on 3 January 1977; section 11 of the same Act inserted Part IVA, which contains the citizen's parallel duty in article 51A(g).

2. A village downstream of a tannery petitions a court relying only on article 48A. What happens, and what should the petition have said? It fails, because article 48A is in Part IV and article 37 provides that Part IV is not enforceable by any court. The petition should have relied on a fundamental right in Part III, on the statutory pollution regime, and on a claim for compensation, using article 48A to show that the State was constitutionally directed to regulate, to argue for the reading of the statute that protects the watercourse, and to answer any complaint that the regulation is too strict.

3. Compare article 48A with article 51A(g). Why does it matter that they were inserted together? Article 48A directs the State; article 51A(g) imposes a duty on every citizen. Article 51A(g) is the wider list, adding lakes and rivers and a duty of compassion towards living creatures; article 48A is the one addressed to the body that can actually legislate. They were inserted by the same amendment, sections 10 and 11 of the Forty-second Amendment Act 1976, which shows that the framers of the amendment treated environmental protection as a shared burden rather than as a purely governmental one. Neither is enforceable.

4. What does article 48A fail to say that a technology question needs? It fails to define the environment, to name any source of harm, to fix any standard of quality, to say anything to the person or company that caused the damage, to provide a remedy, and to mention future generations. Every one of those gaps has to be filled from somewhere else: from a fundamental right, from a statute, or from the international material, where paragraph 56 of General Comment 25 supplies the account of harm that is serious, effectively irreversible or inequitable to present or future generations.

5. Why is it significant that article 48 speaks of organising agriculture on modern and scientific lines? Because it shows how the original Constitution understood science. In the text of 1950 science and technology appear in Part IV as instruments of development, something the State should apply more of. The idea that scientific and technological activity is itself a source of risk that the State must guard against enters the Constitution only in 1976, with article 48A and the fundamental duties. That change of posture, from science as a means to science as a hazard as well, is the whole of this paper in one amendment.

The third of these is the one that connects this chapter to the next section. The Comment protects the researcher who objects, which is a right that only makes sense on the assumption that a researcher has ethical duties in the first place.

Paragraph 20 adds that the protection of this freedom is itself an element of the right to participate in and enjoy the benefits of scientific progress, and not merely a separate guarantee sitting beside it. The full treatment is in [Freedom Indispensable for Scientific Research].

The moral and material interests of the author

Article 15(1)(c) recognises the right of everyone to benefit from the protection of the moral and material interests resulting from any scientific, literary or artistic production of which he is the author. It is the third limb of article 15(1) and the one most often confused with intellectual property.

The confusion is worth clearing up because it is a standard examination trap. The Covenant right is a human right of the author, a natural person, and it protects an interest connected to human dignity. A patent or a copyright is a legal entitlement, of limited duration, freely transferable, and commonly held by a company that is not the author at all. The two overlap and are not the same thing, and the point returns in [Access to Medicines and Patents].

The ordinary freedoms

Freedom of research is not self-executing. It operates through expression, association, movement and access to information, and it is usually attacked through those rather than directly. The Committee's insistence on facilitating researchers' travel in and out of the territory, discussed in [International Co-operation in Science], is an example of exactly this: the freedom is defeated at the visa counter rather than in the statute book.

The responsibilities of the scientist

The framework states these less systematically than it states the rights, and an honest answer says so. Three sources carry them.

The limitation clause. General Comment 25 at paragraph 22 accepts that limitations on the research process can be necessary, particularly where research affects human beings, in order to protect their dignity, integrity and consent. It adds a duty that is easy to miss and directly relevant to research conducted across borders: where the research is done in countries or among populations different from those of the researchers, the State of origin must guarantee the rights and obligations of all parties involved. That is aimed at the practice of conducting trials where regulation is weakest.

The research ethics instruments. The Nuremberg Code 1947 and the Declaration of Helsinki state the duties owed to a research participant. Both are taken in Module III, where the ethics of scientific development is the syllabus topic. Neither is a treaty; both are the professional statement of what a researcher owes, and both are treated by the Comment as feeding the acceptability element of the right.

The acceptability element. Acceptability requires that scientific research incorporate ethical standards in order to ensure its integrity and the respect of human dignity, and General Comment 25 points to the standards proposed in the Universal Declaration on Bioethics and Human Rights 2005: maximising benefits and minimising harm with reasonable safeguards, guaranteeing autonomy and free and informed consent, respecting privacy and confidentiality, protecting vulnerable groups against discrimination, and giving due regard to cultural diversity and pluralism.

Taken together the responsibilities are: to obtain consent, to protect the participant, to be honest about method and result, to consider the foreseeable uses of the work, and to speak when a project is wrong.

UNESCO's Recommendation, named and placed

The instrument that states the scientist's position most fully is the Recommendation on Science and Scientific Researchers, adopted by UNESCO in 2017, revising an earlier Recommendation of 1974. General Comment 25 names it in its opening paragraph among the important documents on the relationship between science and economic, social and cultural rights, and cites it repeatedly: for the dimensions of freedom of research, for the definition of technology as knowledge relating directly to the production or improvement of goods or services, and in the sections on obligations and on international co-operation.

Two things follow, and the second is a caution.

What it is. A recommendation of the UNESCO General Conference. It is not a treaty and creates no binding obligation, but it is the most detailed international statement of the status, freedoms and responsibilities of the scientific researcher, and the Committee treats it as an authoritative reference in construing article 15.

What this book can say about it. The Recommendation is not among the texts held in this subject's authority base. Everything stated about it here is what General Comment 25 records, and no paragraph of the Recommendation is quoted or numbered from any other source. A candidate should do the same: name it, place it, say what the Committee draws from it, and resist the temptation to attribute to it provisions recalled from elsewhere.

Worked example

A pharmaceutical company funds a university team to trial a new drug. The contract provides that the company owns the data, that publication requires the company's written consent, and that the lead researcher will not comment publicly on the trial for three years after its conclusion. Midway through, the researcher forms the view that an adverse event pattern is emerging and that the trial should be stopped.

Which of the researcher's rights are engaged? Independent judgment, first: a contract that makes publication conditional on the funder's consent is undue influence on the researcher's independent judgment in the sense of paragraph 13. The freedom to question the ethical value of the project and to withdraw is engaged directly by the researcher's own conclusion. The sharing of data and analysis with policymakers and, where possible, the public is engaged by the publication clause.

Which of the researcher's duties are engaged? The duty to the participants, who are being exposed to a risk the researcher now believes is materialising, and whose consent was given on a different footing. Consent that is not updated when the risk changes is not consent to the risk now being run.

Who else bears a duty? The State. The right in article 15(3) is a right against the State, and the Comment's account of obligations includes the duty to protect against the acts of private parties. A legal system in which such a clause is enforceable, and in which the researcher who speaks is liable in damages, is a system that has failed the duty to protect. The remedy is not to instruct the researcher to be brave; it is regulation of the terms on which publicly employed researchers may contract with funders.

Where does the analysis end up? The researcher's freedom and the participants' safety point the same way, which is the usual case. They are opposed less often than the drafting of this topic suggests, and an answer that assumes a standing conflict between the scientist's freedom and everyone else's rights has misread the framework.

Distinctions

What it does NOT mean

Freedom of research is not absolute. General Comment 25 says so in terms at the end of paragraph 13, and paragraph 22 identifies where the limits fall.

It is not an entitlement to funding. The freedom to define aims, objectives and methods is a freedom from direction, not a claim on the public purse for a particular project.

It is not a defence to harm. A researcher may not conduct an unethical experiment and plead article 15(3). The freedom is a freedom of inquiry, and the limits on the research process protecting dignity, integrity and consent are the very thing the Comment permits.

And the responsibilities are not merely professional. They appear in the acceptability element of a Covenant right, which means a State that permits research without them is in breach of the Covenant, not merely of good practice.

Quick revision

• The scientist is both rights holder and duty bearer, and the freedom is stated first because the duties are limits on it.
• Article 15(3): States undertake to respect the freedom indispensable for scientific research. General Comment 25, paragraph 13 gives five dimensions: independent judgment; autonomous institutions and the definition of aims and methods; freedom to question a project's ethics and to withdraw; freedom to co-operate nationally and internationally; sharing of data and analysis. Paragraph 20: the freedom is itself an element of the right.
• Article 15(1)(c): the moral and material interests of the author. Not the same as intellectual property: the Covenant right is a human right of a natural person, IP is transferable, time limited and often corporate.
• Responsibilities come from three places: paragraph 22 (limits on the research process to protect dignity, integrity and consent; the State of origin's duty where research is done among other populations), the ethics instruments (Nuremberg 1947, Helsinki), and the acceptability element, which points to the Universal Declaration on Bioethics and Human Rights 2005.
• UNESCO Recommendation on Science and Scientific Researchers 2017: not a treaty, the fullest statement of the researcher's status, named and relied on throughout General Comment 25. Not in this book's base; do not quote paragraph numbers for it.

Test yourself

1. Give the five dimensions of freedom of research in General Comment 25 and say which of them presupposes that the researcher has duties. Protection of researchers from undue influence on their independent judgment; the possibility of setting up autonomous research institutions and of defining the aims, objectives and methods of research; the freedom to question openly the ethical value of certain projects and the right to withdraw if conscience so dictates; the freedom to co-operate with other researchers nationally and internationally; and the sharing of scientific data and analysis with policymakers and, wherever possible, the public. The third presupposes duties, since a right to object on ethical grounds and to withdraw is only intelligible if the researcher is understood to owe ethical obligations in respect of the work.

2. Distinguish article 15(1)(c) from a patent. Article 15(1)(c) recognises the right of everyone to benefit from the protection of the moral and material interests resulting from any scientific, literary or artistic production of which he is the author. It is a human right, held by a natural person, connected to human dignity and to the author's relationship with the work, and it is not transferable in the way property is. A patent is a legal entitlement created by statute, of limited duration, freely assignable, and frequently owned by a corporation that is not the author. The two overlap where an author holds intellectual property in their own production, but the Covenant right neither requires a particular intellectual property regime nor is satisfied merely because one exists.

3. What does paragraph 22 of General Comment 25 require of the State of origin when research is conducted abroad, and what practice is it aimed at? It requires that, where research is done in countries or among populations different from those of the researchers, the State of origin must guarantee the rights and obligations of all parties involved. It is aimed at the practice of conducting research, particularly clinical trials, in jurisdictions where regulation, oversight and the practical ability of participants to obtain a remedy are weakest, so that the protections that would apply at home do not apply to the people actually exposed to the risk.

4. A university contract requires the funder's written consent before any publication. Assess it under article 15(3). It engages the first and fifth dimensions of the freedom in paragraph 13: protection from undue influence on independent judgment, and the sharing of scientific data and analysis with policymakers and with the public wherever possible. A power in the funder to withhold consent to publication places the researcher's conclusions under the control of a party with an interest in them. Because article 15(3) binds the State, the question is whether the State has discharged its duty to protect by regulating the terms on which researchers may contract; a legal system that enforces such a clause without qualification, and exposes a researcher who speaks to liability, has not.

5. What is the UNESCO Recommendation on Science and Scientific Researchers, and what is the correct way to use it in an answer on this paper? It is a recommendation adopted by the UNESCO General Conference in 2017, revising an instrument of 1974, and it is the fullest international statement of the status, freedoms and responsibilities of scientific researchers. It is not a treaty and binds no State, but General Comment 25 names it among the important documents in the field and relies on it repeatedly, including for the dimensions of freedom of research and for its definition of technology as knowledge relating directly to the production or improvement of goods or services. The correct use is to name it, place it as a non-binding but authoritative reference relied on by the Committee, and state what the Committee draws from it, without attributing to it particular numbered paragraphs unless its own text has been read.

Participation also includes the right to information and participation in controlling the risks involved in particular scientific processes and its applications. In this context, the precautionary principle plays an important role. This principle demands that, in the absence of full scientific certainty, when an action or policy may lead to unacceptable harm to the public or the environment, actions will be taken to avoid or diminish that harm.

That placement is the single most examinable thing about the paragraph. On this syllabus precaution is not an idea borrowed from environmental law and attached to science. It is a limb of the right to science itself: if you have a right to take part in scientific progress, you have a right to a say in the risks that progress creates. The Committee's own footnote attributes the formulation to the World Commission on the Ethics of Scientific Knowledge and Technology, "The precautionary principle" (Paris, UNESCO, 2005).

Broken down

The principle has four elements, and an answer that states all four is worth far more than one that states the slogan.

1. A threat of harm that can be identified in advance. Somebody must be able to say what the feared harm is and by what route it would occur. A vague unease is not a threat.
2. Absence of full scientific certainty. This is the trigger, and it cuts both ways. There must be enough science to raise the concern and not enough to settle it. Where the harm is proved, the principle is not needed; where nothing but imagination supports the fear, the principle does not apply.
3. The harm must be unacceptable. General Comment 25 gives four categories, and any one of them is enough: harm threatening to human life or health; harm that is serious and effectively irreversible; harm inequitable to present or future generations; and harm imposed without adequate consideration of the human rights of those affected. The fourth category is the one students miss, and it is the most useful, because it makes the manner of imposition a ground of objection by itself.
4. Action to avoid or diminish the harm. The Committee says that actions "will be taken", not that the activity will be forbidden. The response is graduated.

The same paragraph names the working tools: "Technological and human rights impact assessments are tools that help to identify potential risks early in the process and the use of scientific applications." An impact assessment is a structured study, done before a project proceeds, of what it is likely to do to a defined interest.

What the principle does to the burden of proof

The burden of proof is the answer to the question "if nobody can tell, who loses?" Ordinarily it lies on the person complaining: a claimant who cannot show that the factory caused the illness fails, however sympathetic the claim.

Precaution moves it. Once a credible threat of unacceptable harm is shown, the proponent of the activity is required to show that it can proceed safely, or on what conditions. The doubt no longer works in favour of the person creating the risk. That single shift is what the principle is for, and it is what its critics dislike about it.

Notice what it does not do. It does not reverse the burden in a criminal prosecution, where the accused is presumed innocent, and it does not make a regulator's opinion unchallengeable. It allocates the consequence of scientific uncertainty in a regulatory decision, and nothing more.

A worked example

Kaveri Agro Ltd applies to the licensing authority for permission to market a new seed coating in Maharashtra. Two published studies suggest that the compound persists in groundwater; a third finds no effect. Nobody has studied its effect on the children of the farming households that will handle it. Villagers in Warud object.

Element one, the threat. The objection is specific: persistence in groundwater, with a route of exposure through drinking water. That is a stated threat, not an unease.

Element two, the uncertainty. Two studies against one is exactly the state of knowledge the principle addresses. If the science were settled either way, the authority would simply apply it.

Element three, unacceptability. Harm through drinking water is threatening to health, which is the first category. Persistence makes it capable of being effectively irreversible, which is the second. And because nobody has studied the children who will handle the product, the harm would be imposed without adequate consideration of the human rights of those affected, which is the fourth.

Element four, the response, and it is graduated. The authority is not obliged to choose between a licence and a ban. It may require a human rights and technological impact assessment before deciding. It may licence on conditions: restricted quantities, mandatory protective equipment, monitoring of the village wells, and a reporting obligation. It may licence for a fixed period and review. It may suspend a decision until one study is repeated. Prohibition is the last of a series, not the first.

What the ordinary law would have done. A negligence claim would begin after the wells were contaminated, would require the villagers to prove that this compound caused their illness, and would probably fail on causation. The precautionary principle exists because that sequence arrives too late to be of use to anyone.

Where the principle already operates in Indian law

India has no single enactment that states the principle in general terms, but the shape of it is visible in statutes a student on this paper is already reading, and quoting one of them is worth more than an assertion that Indian law accepts precaution.

Section 26A of the Drugs and Cosmetics Act, 1940. The Central Government may act if it "is satisfied, that the use of any drug or cosmetic is likely to involve any risk to human beings or animals", and may then "regulate, restrict or prohibit" its manufacture, sale or distribution. The trigger is likelihood of risk, not proof of injury, and the response is expressly graduated in the same three words the principle uses. The chapter on drugs control in Module II works the section in full.

Section 2(1) of the Epidemic Diseases Act, 1897. The State Government may take special measures when it is satisfied that the State or any part of it "is visited by, or threatened with, an outbreak of any dangerous epidemic disease". A threat is put on the same footing as a visitation. Module II examines the public health powers that follow.

The Constitution. Article 51A(g) makes it the duty of every citizen "to protect and improve the natural environment including forests, lakes, rivers and wild life, and to have compassion for living creatures", and article 48A places the corresponding endeavour on the State. The chapter on the environment in the Constitution takes article 48A; what matters here is that the constitutional text supplies a footing for anticipatory protection rather than only for compensation after the event.

Distinctions

What it does NOT mean

It does not demand zero risk. Nothing human is riskless, and a principle requiring proof of complete safety would forbid everything, including the technologies that save lives. The test is unacceptable harm, and the word doing the work is unacceptable.

It does not licence a State to shut down research. General Comment 25 says in terms that the principle "should not hinder and prevent scientific progress, which is beneficial for humanity", and that limitations on the freedom of scientific research are compatible with the Covenant only within article 4 of the ICESCR, which permits limitations determined by law, solely for the purpose of promoting the general welfare in a democratic society, and compatible with the nature of the right. Paragraph 22 adds that any limitation on the content of scientific research "implies a strict burden of justification by States". Precaution bites hardest on the use and application of scientific results and least on the asking of questions.

It is not a substitute for evidence. It applies where there is a scientific basis for concern. Invoked against a fear with no scientific basis at all, it becomes a licence for whoever shouts loudest.

It is not a rule of criminal law. It governs what a regulator may do in advance. It does not lower the standard of proof for punishing anybody.

Limits and criticism

There is no single agreed formulation. Stated weakly, it says that uncertainty is not a reason to postpone cost effective measures against serious harm. Stated strongly, it says that no activity may proceed until its proponent proves it safe. The strong version is close to impossible to satisfy, because safety cannot be proved, only failed to be disproved, and a student who states only the strong version invites the obvious objection.

Inaction has victims too. Refusing or delaying a technology is itself a decision with a cost, measured in the harm the technology would have prevented. A delayed vaccine, a withheld diagnostic and an unapproved treatment all injure real people, and those people have the right to enjoy the benefits of scientific progress. The principle is a rule for deciding under uncertainty, not a presumption against novelty.

It can be captured. A rule that lets a State restrict a product without proving harm can be used for disguised protection of domestic industry, or by an established producer against a new competitor. The safeguard is procedural: reasons, evidence and review.

Somebody has to decide what is unacceptable, and that is a value judgment, not a scientific one. General Comment 25 accepts the difficulty and answers it with process. Because the application of the principle "is sometimes controversial", it says, participation and transparency become crucial: the risks and the potential of a technical advance should be made public so that society can decide, through informed, transparent and participatory public deliberation, whether the risks are acceptable. On this syllabus that is the answer to give. Precaution is legitimate when the people who will carry the risk have been told about it and heard.

Quick revision

• Precautionary principle: in the absence of full scientific certainty, where an action or policy may lead to unacceptable harm to the public or the environment, action is taken to avoid or diminish that harm.
• Stated for this paper in General Comment No. 25 (2020), E/C.12/GC/25, paragraph 56, under the heading "Participation and the precautionary principle". Precaution is a limb of the right to science, not an import.
• Four categories of unacceptable harm: threatening to life or health; serious and effectively irreversible; inequitable to present or future generations; imposed without adequate consideration of the human rights of those affected.
• Four elements: identified threat, scientific uncertainty, unacceptable harm, proportionate action. Tools: technological and human rights impact assessments.
• Its real effect is on the burden of proof: the proponent must justify the activity, not the exposed person prove the injury.
• Indian shape: section 26A of the Drugs and Cosmetics Act 1940 ("likely to involve any risk"), section 2(1) of the Epidemic Diseases Act 1897 ("visited by, or threatened with"), articles 48A and 51A(g) of the Constitution.
• Limits: it must not hinder scientific progress; restrictions on research need strict justification under ICESCR article 4; inaction has victims; acceptability is decided by participation and transparency.

Test yourself

1. State the precautionary principle and identify its four elements. In the absence of full scientific certainty, where an action or policy may lead to unacceptable harm to the public or the environment, measures must be taken to avoid or diminish that harm. Its elements are a threat of harm identified in advance; genuine scientific uncertainty, meaning enough evidence to raise the concern and not enough to settle it; harm that is unacceptable, which General Comment 25 defines as threatening to life or health, serious and effectively irreversible, inequitable across generations, or imposed without adequate consideration of the human rights of those affected; and a response proportionate to the threat.

2. Why does it matter that General Comment 25 places the principle under participation? Because it makes precaution part of the right to science rather than a borrowing from environmental law. If everyone has a right to participate in scientific progress, that participation includes a say in the risks the progress creates. It also supplies the answer to the principle's hardest question, which is who decides that a risk is unacceptable: the Committee's answer is informed, transparent and participatory public deliberation, not the unexplained opinion of a regulator.

3. A State bans all research on a new gene editing method because the public is anxious about it. Is that an application of the principle? Not as it stands. Precaution operates chiefly on the use and application of scientific results, and General Comment 25 says the principle should not hinder and prevent scientific progress. A limitation on the content of research must satisfy article 4 of the ICESCR, being determined by law, compatible with the nature of the right, and solely for the general welfare in a democratic society, and it carries a strict burden of justification. Public anxiety with no stated threat and no scientific basis does not discharge that burden.

4. How does the principle change the burden of proof, and why is that its most important feature? Ordinarily the person alleging harm must prove it, so that scientific doubt defeats the claim and benefits whoever created the risk. Under the principle, once a credible threat of unacceptable harm is shown, the proponent of the activity must justify proceeding and on what conditions. It matters because the evidence about a new technology usually matures only after a population has been exposed, so leaving the burden where it ordinarily lies means the harm must occur before the law can respond.

5. Give two Indian statutory provisions that are precautionary in structure, and say what makes them so. Section 26A of the Drugs and Cosmetics Act 1940 lets the Central Government regulate, restrict or prohibit a drug or cosmetic where it is satisfied that its use "is likely to involve any risk" to human beings or animals: the trigger is likelihood, not proof, and the three verbs are a graduated response. Section 2(1) of the Epidemic Diseases Act 1897 lets a State Government take special measures where the State is "visited by, or threatened with" an outbreak, placing a threat on the same footing as an actual outbreak.

A technology that changes none of the five is usually not a rights question. A technology that changes three or four of them is almost always one.

Step three: list the rights on both sides

Every serious technology advances some rights and threatens others, and an answer that gives only one side is incomplete on this paper. The chapter on dual use develops the point; here it is a discipline of the method.

Ask first which rights the technology serves: health, education, work, information, participation, the enjoyment of the benefits of scientific progress. Ask then which rights it threatens: life, privacy, equality, expression, fair trial, bodily integrity. Write both lists before choosing a side.

The examiner is not asking whether the technology is good. The examiner is asking whether the student can see both columns and then reason about the trade between them.

Step four: name the duty bearer and the duty

A right with no identified duty bearer is a complaint, not a legal claim. The chapter on what a human right is sets out the three duties of the State, to respect, to protect and to fulfil; the step here is to apply them.

Respect is engaged when the State itself deploys the technology: a police force running the facial recognition, a hospital run by the government holding the records.

Protect is engaged when a private actor deploys it, and this is the duty that does most of the work in this subject, because most technology is developed, owned and operated privately. The State's failure to regulate is itself the breach.

Fulfil is engaged when the complaint is that the technology is absent: no diagnostic equipment in the district hospital, no assistive technology in the school.

The 1975 Declaration writes the first two duties into an instrument. Paragraph 2 requires States to take appropriate measures to prevent the use of scientific and technological developments, "particularly by the State organs", to limit or interfere with the enjoyment of human rights. Paragraph 8 requires States to take effective measures, "including legislative measures", to prevent and preclude the use of scientific and technological achievements to the detriment of human rights and of the dignity of the human person. The words "including legislative measures" are the treaty language of the duty to protect.

Step five: test the interference

Almost no right in this field is absolute, so identifying an interference does not end the argument. An interference must be justified, and the four questions are always the same: is it prescribed by law, does it pursue a legitimate aim, is it necessary for that aim, and is it proportionate, meaning that no less restrictive measure would do and that the harm imposed does not outweigh the benefit gained.

Module IV develops that test in the form Indian law now applies. At this stage what matters is the discipline of asking it. A student who writes that a technology "violates the right to privacy", and stops, has not answered the question, because every lawful search and every lawful interception also interferes with privacy.

Step six: ask about distribution and remedy

Two questions close the analysis, and both are frequently forgotten.

Distribution. Who receives the benefit and who carries the risk? The two are often different people, and where the population that bears the risk is not the population that gets the benefit, the objection is one of equality and of benefit sharing rather than of safety. The chapters on benefit sharing and on the digital divide take this further.

Remedy. What can the affected person actually do, and against whom? A right without a remedy fails at exactly the point where it matters. Ask whether the person can find out that the technology was used on them, whether they can obtain the evidence, whether there is a body with power to order the use stopped, and whether compensation is available. The chapter on remedies when technology causes harm follows those routes through.

What the instruments add

Two treaty provisions are worth carrying into any answer on this topic, because they show that the method is not merely an academic construction.

CEDAW, article 11(3). The Convention on the Elimination of All Forms of Discrimination against Women, adopted by General Assembly resolution 34/180 of 18 December 1979 and in force from 3 September 1981, provides that protective legislation relating to the matters covered by article 11 "shall be reviewed periodically in the light of scientific and technological knowledge and shall be revised, repealed or extended as necessary". That is a binding obligation to keep the law in step with what science now knows. It is the clearest treaty statement that a change in scientific knowledge can, by itself, make an existing law unlawful.

CEDAW, article 14(2)(g). States must ensure to rural women the right "to have access to agricultural credit and loans, marketing facilities, appropriate technology and equal treatment in land and agrarian reform". Access to a technology is stated as a component of equality, which is the treaty form of step three's second column: the absence of a technology can breach a right just as its presence can.

The 1975 Declaration adds the general obligation in paragraph 7: States are to take the necessary measures, including legislative measures, to ensure that the utilization of scientific and technological achievements promotes the fullest realization of human rights "without any discrimination whatsoever on grounds of race, sex, language or religious beliefs".

A worked example

Godavari English School in Nashik installs a camera at its gate that records attendance by recognising each pupil's face. Mrs Kulkarni, the principal, buys the system from a private vendor, which stores the images on its own servers. Ayesha, aged fourteen, is marked absent three times because the system misreads her when she wears a headscarf.

Step one, the capability. The school, and in fact the vendor, can now identify and log the movement of every child at the gate, continuously, automatically, without any adult present, and at negligible cost per child.

Step two, what changed. Scale: every child, every day, instead of a register read by a teacher. Permanence: a biometric template of a minor is stored indefinitely. Invisibility: neither Ayesha nor her parents can see what is recorded or by whom. Delegation: the decision that Ayesha was absent is now taken by software. Four of the five changes are present.

Step three, both columns. Served: education administration, punctuality, the safety of children whose arrival can be confirmed. Threatened: privacy, the special protection owed to a child, equality, because the error rate is not evenly distributed, and the right to a remedy.

Step four, the duty. The school is private, so the primary duty engaged is the duty to protect: the State must regulate what a private school and a private vendor may do with a child's biometric data. That is the duty the 1975 Declaration puts in paragraph 8 as a duty to take effective measures, including legislative measures.

Step five, the justification. Is there a law permitting it? Is marking attendance a legitimate aim? Certainly. Is a facial biometric necessary for it, when a card, a register or a fingerprint-free roll call would do the same job? That is where the system probably fails, because a less restrictive means plainly exists. Proportionality asks whether a lifelong biometric record of a child is a fair price for saving a teacher five minutes a day.

Step six, distribution and remedy. The benefit goes to the school in saved time, and the risk is carried by the children. Ayesha's remedy is the weakest part: she cannot see the data, cannot test the error, and has been marked absent by a decision with no author. The absence of a remedy is not a side issue in this problem. It is the strongest part of her case.

When a technology is NOT a rights question

The method has to be able to say no, or it says nothing. Three filters, and all three must be satisfied.

It must touch an interest that a right protects. A better gearbox on a bicycle affects convenience and nothing else. Convenience is not a protected interest.

The effect must be attributable to somebody who owes a duty. Harm caused by nobody in particular is a misfortune. Harm caused by a State organ, or by a private actor the State was obliged to regulate, is a rights question.

The effect must be more than trivial. Every technology has some effect on somebody. A right is engaged when the effect on the protected interest is real, not notional.

Applying those three keeps the answer honest. Many complaints about technology are complaints about a policy choice, a commercial practice or a personal taste, and calling them human rights violations weakens the phrase for the cases that deserve it.

The three ways a technology enters rights law

Most examination questions are about the first route. The best answers notice that the second and the third exist, because that is what distinguishes a student who has understood topic 2 from one who has learnt a list of dangers.

What beginners get wrong

Describing the technology instead of analysing it. Half a page on how a system works earns nothing. One sentence naming the capability earns everything, because it is the sentence the rest of the answer is built on.

Assuming a new technology needs a new right. It almost never does. The existing rights are re-applied to new facts. Claiming a new right is an exceptional move that has to be argued for, and the chapter on access to the internet shows what that argument looks like.

Treating the State as the only actor. Most of the technology in this paper is private. The question is not only what the State did but what the State allowed, which is the duty to protect.

Stopping at "this violates a right". Interference is the beginning of the analysis. Justification is the rest of it, and an answer that skips step five has skipped the law.

Quick revision

• A new technology becomes a human rights question when it changes the practical conditions on which an existing right depends, by creating a capability that can be used against a protected interest or by supplying a means of realising a right.
• Six steps: name the capability; ask what it changed; list the rights on both sides; name the duty bearer and the duty; test the interference; ask about distribution and remedy.
• Five changes to test for: scale, cost, permanence, invisibility, delegation.
• The duty to protect does most of the work, because most technology is private. 1975 Declaration, paragraphs 2, 7 and 8.
• CEDAW article 11(3): protective legislation must be reviewed periodically in the light of scientific and technological knowledge and revised, repealed or extended as necessary. Article 14(2)(g): access to appropriate technology as a component of equality for rural women.
• Three filters before calling something a rights question: a protected interest, an attributable duty bearer, a more than trivial effect.
• Three routes into rights law: threat, means, re-reading.

Test yourself

1. Set out the six steps of the analysis and say why the order matters. Name the capability rather than the device; ask what the capability changed, testing scale, cost, permanence, invisibility and delegation; list the rights the technology serves and the rights it threatens; identify the duty bearer and which of the duties to respect, protect or fulfil is engaged; test whether the interference is prescribed by law, pursues a legitimate aim, is necessary and is proportionate; and ask who gets the benefit, who carries the risk and what remedy the affected person has. The order matters because each step produces the material the next one uses: without the capability there is no right to identify, and without the right there is no duty to allocate.

2. Why is the duty to protect the most important of the three duties in this subject? Because the great majority of technology is developed, owned and deployed by private actors. If the framework reached only the conduct of State organs it would leave the hospital chain, the platform, the laboratory and the vendor untouched. The duty to protect makes the State answerable for failing to regulate them, which is why the 1975 Declaration speaks in paragraph 8 of effective measures including legislative measures.

3. A district administration proposes to distribute tablets to schoolchildren. Is this a rights question, and if so which route? Yes, by the second route, technology as a means. The complaint in such cases is usually not that the tablet threatens a right but that its absence prevents the enjoyment of one, which engages the duty to fulfil. The equality analysis matters most: who is left out, and does the distribution widen an existing gap. CEDAW article 14(2)(g) is the treaty analogue, stating access to appropriate technology as part of equality rather than as a benefit.

4. Give an example of a technological change that is not a human rights question, and explain why. An improvement in the gearing of a bicycle. It fails the first filter, because the interest it affects is convenience, which no right protects. It also fails the second, since no duty bearer is responsible for anyone's ordinary preference between one gearbox and another. Keeping the filters strict matters: describing every commercial or policy complaint as a human rights violation devalues the term where it is genuinely needed.

5. What does article 11(3) of CEDAW add to this topic that no other provision in the base supplies? It states, in binding treaty form, that laws must be re-examined when scientific knowledge changes, and revised, repealed or extended accordingly. It is the textual foundation of the third route, in which a technology becomes a rights question not by creating a danger but by making an existing rule obsolete, and it shows that the obligation to keep law in step with science is a legal duty rather than good practice.

The word to notice is "arbitrary or unlawful". Two conditions, not one. An interference must be lawful, meaning provided for by law, and it must not be arbitrary, which imports reasonableness and proportionality. A State that passes a statute permitting mass interception has satisfied the first and not the second.

The 1975 Declaration, paragraph 6, is the provision that ties privacy specifically to this paper. States are to extend the benefits of science and technology to all strata of the population and to protect them from possible harmful effects of the misuse of scientific and technological developments, including their misuse to infringe upon the rights of the individual or of the group, particularly with regard to respect for privacy and the protection of the human personality and its physical and intellectual integrity. That is a 1975 instrument naming privacy as the leading casualty of technological misuse, before most of the technology existed.

The Indian position

Justice K.S. Puttaswamy (Retd.) v. Union of India, AIR 2017 SC 4161, (2017) 10 SCC 1, decided on 24 August 2017 by nine judges.

Facts. A Bench of three judges hearing the challenge to the Aadhaar scheme recorded that the norms for and the compilation of demographic and biometric data by the Government were being questioned as a violation of the right to privacy. The Attorney General argued that the existence of a fundamental right of privacy was itself in doubt, relying on M P Sharma and Kharak Singh. The petitioners answered that both rested on an approach later held not to be good law. Because the decisions that had since recognised privacy were given by smaller Benches, the question was referred upwards, and on 18 July 2017 a Constitution Bench directed that it be placed before nine judges.

Held. The Order of the Court disposed of the reference in four terms. M P Sharma, which holds that the right to privacy is not protected by the Constitution, stands over-ruled. Kharak Singh stands over-ruled to the extent that it so holds. The right to privacy is protected as an intrinsic part of the right to life and personal liberty under article 21 and as a part of the freedoms guaranteed by Part III. Decisions subsequent to Kharak Singh which have enunciated that position lay down the correct position in law.

Why it matters here. The case exists because of a technology. Nobody asked the Supreme Court an abstract question about privacy; a biometric identification programme was built and the question could no longer be avoided. That is the pattern this whole paper describes, visible in the procedural history of one case. It also settles the Indian starting point for everything in Module IV: personal data is inside article 21.

The three-part test

The instruments and the Indian cases converge on the same structure, and it is the structure an answer should use. An interference with privacy is permissible only if it satisfies three requirements.

1. Legality. There must be a law. Not a policy, not an executive instruction, not an internal standard operating procedure. The law must be accessible and must be precise enough that a person can foresee when the power may be used against them.
2. Legitimate aim. The purpose must be one the constitutional or treaty order recognises: national security, prevention of crime, public health, the protection of the rights of others.
3. Proportionality. The measure must be rationally connected to the aim; there must be no less restrictive means reasonably capable of achieving it; and the benefit must outweigh the burden imposed on the right.

The same three-limbed structure governs limitations on the right to science under article 4 of the ICESCR, set out in [Dual Use: the Same Technology Both Ways]. That is not a coincidence: it is the general shape of limitation analysis in human rights law, and a candidate who has learnt it once can apply it to any right on this paper.

Worked example

A municipal corporation installs cameras with automated face recognition at every entrance to a public market. The stated purpose is the detection of pickpockets. Images are matched against a database of persons previously arrested, and all footage is retained for two years. The scheme is created by a resolution of the corporation.

Legality. It fails at the first step. A resolution of a municipal body is not a law that authorises interference with a fundamental right, and there is no statute defining who may be enrolled in the database, on what grounds, for how long, or with what remedy for a person wrongly included. The analysis could stop here, and a good answer says so before going on.

Legitimate aim. Prevention and detection of crime is a legitimate aim. This limb is usually satisfied and is rarely where a case is won.

Proportionality. Rational connection is arguable. The failure is at necessity and at balance. Retaining footage of every visitor for two years to catch pickpockets is not the least restrictive means; the burden falls on the entire population of shoppers, none of whom is suspected of anything; and matching against a database of persons previously arrested, rather than convicted, imports the error rate of past policing into every future identification.

The equality dimension. Recognition systems do not perform uniformly across populations, so a scheme of this kind distributes its errors unevenly. That converts a privacy problem into a discrimination problem, which is taken up in [Equality and the Digital Divide].

What could survive. Cameras without automated matching, under a statute, with a defined retention period and an accessible complaints route, may well be proportionate. The point of the test is that it does not produce a yes or a no to "cameras"; it produces a yes or a no to a particular design.

What it does NOT mean

Privacy is not secrecy. A person who has told a doctor something has not made it public, and a person walking down a street has not consented to being tracked. The right protects control over access, not concealment.

"Nothing to hide" is not an argument the law recognises. It misstates the interest, which is not in hiding wrongdoing but in not being watched; and it misallocates the burden, since it is the State that must justify the interference, not the individual who must justify wanting to be left alone.

Consent does not cure everything. Consent obtained as the price of a service a person cannot do without is not a free choice, and the framework treats it accordingly.

Anonymised is not anonymous. Data stripped of names can frequently be re-identified by combining it with other data, which is precisely the property General Comment 25 notes when it observes that information is nonrivalrous and recombinant.

Surveillance is not only a State activity. Most of the capability is privately owned, which engages the duty to protect and makes regulation, rather than restraint, the primary obligation.

Limits and criticism

The three-part test is a method, not an answer, and its second limb is almost never contested. Cases are decided at legality and proportionality, and a candidate who spends an answer establishing that national security is a legitimate aim has spent it on the uncontested limb.

The deeper criticism is that the test is individual in shape and the harm is increasingly collective. It asks whether this person's rights were interfered with and whether that interference was justified. It is a poor instrument for a harm that consists in an entire population altering its behaviour because it knows it may be watched, since no individual can show that they in particular were observed. That is a real limitation of the framework and worth a sentence in an answer that has room for it.

Quick revision

• UDHR article 12 and ICCPR article 17: no arbitrary or unlawful interference with privacy, family, home or correspondence; and a right to the protection of the law against it. Two conditions: lawful, and not arbitrary.
• 1975 Declaration, paragraph 6: protect people from misuse of scientific and technological developments, particularly with regard to respect for privacy and the physical and intellectual integrity of the person.
• Puttaswamy (2017), nine judges: privacy is protected as an intrinsic part of article 21 and as part of the Part III freedoms; M P Sharma over-ruled, Kharak Singh over-ruled to the extent it held otherwise.
• The three-part test: legality (a law, accessible and foreseeable), legitimate aim, proportionality (rational connection, least restrictive means, benefit outweighing burden).
• The five changes technology made: scale, cost, permanence, invisibility, delegation. Surveillance engages all five.
• Not secrecy; "nothing to hide" is not an argument; consent as the price of a necessity is not free; anonymised is not anonymous; and most of the capability is private, so the duty to protect does the work.

Test yourself

1. Set out article 17 of the ICCPR and explain the significance of the words "arbitrary or unlawful". Article 17 provides that no one shall be subjected to arbitrary or unlawful interference with his privacy, family, home or correspondence, nor to unlawful attacks on his honour and reputation, and that everyone has the right to the protection of the law against such interference or attacks. The two adjectives impose two distinct conditions. Unlawful means the interference must be provided for by law, so an executive practice with no legal basis fails immediately. Arbitrary imports a further standard of reasonableness and proportionality, so that a State which enacts a statute authorising an interference has satisfied the first condition and must still show the interference is not arbitrary. A test that collapsed the two would allow any interference a legislature chose to authorise.

2. What did Puttaswamy decide, and why is it a case about technology rather than about privacy in the abstract? Its Order disposed of the reference in four terms: M P Sharma is over-ruled in so far as it holds that privacy is not protected by the Constitution; Kharak Singh is over-ruled to the same extent; the right to privacy is protected as an intrinsic part of the right to life and personal liberty under article 21 and as part of the freedoms guaranteed by Part III; and the decisions after Kharak Singh which took that view state the law correctly. It is a case about technology because the question reached the Court only because a biometric identification programme had been built and challenged; the constitutional question was forced by the capability rather than raised for its own sake.

3. Apply the three-part test to a proposal to require every mobile telephone connection to be linked to a biometric identity. Legality asks whether a statute authorises it, whether that statute is accessible, and whether it is precise enough that a person can foresee when and how the power will be used, including who may access the resulting database and for what. Legitimate aim is likely satisfied, prevention of crime and national security both being recognised. Proportionality does the work: whether biometric linkage is rationally connected to the aim; whether a less restrictive means, such as documentary verification without biometrics or retention of the biometric only at the point of verification, would reasonably achieve it; and whether the benefit outweighs the burden of creating a database that links every person's communications to their body. The retention period, the access controls and the remedy available to a person wrongly matched are all part of proportionality rather than separate questions.

4. Why is "nothing to hide" not an answer, and where does the burden of justification lie? Because it misdescribes the interest and misplaces the burden. The interest protected is control over access to oneself and one's information, not the concealment of wrongdoing, so a person with nothing to hide still loses something when they are watched. And under both article 17 and article 21 as construed in Puttaswamy, an interference must be justified by the State: it must point to a law, a legitimate aim and proportionality. The formulation reverses that, requiring the individual to explain why they should not be observed, which is the opposite of how a limitation clause operates.

5. What is the principal structural weakness of the three-part test as applied to population-scale surveillance? The test is individual in shape while the harm is collective. It asks whether a particular person's right was interfered with and whether that interference was justified, which suits a search of one house or the interception of one line. Where a capability is applied to a whole population, no individual can ordinarily show that they in particular were observed, and the characteristic injury, that people alter their behaviour because they know they may be watched, is not an injury to any identifiable claimant. The framework can address this through the requirement that a law be foreseeable and through proportionality at the level of the scheme rather than the instance, but it remains a poor fit for a diffuse harm.

Article 7 deals with those who cannot consent. Authorisation must be obtained in accordance with the best interest of the person and with domestic law, and the person must still be involved in the decision to the greatest extent possible. Research on such a person should only be carried out for their direct health benefit, and research without potential direct health benefit only by way of exception, with the utmost restraint, at minimal risk and burden, and subject to conditions prescribed by law. Refusal by such a person to take part is to be respected, which is a striking rule: a person who cannot give a valid consent can still give an effective refusal.

The right in the instruments

ICCPR article 7 provides that no one shall be subjected to torture or to cruel, inhuman or degrading treatment or punishment, and adds a sentence written for exactly this subject: in particular, no one shall be subjected without his free consent to medical or scientific experimentation.

That second sentence is unusual. Most treaty articles state a right in general terms and leave the applications to interpretation. Here the drafters put a specific scientific practice into the text of a prohibition on torture, which tells you how the practice was regarded in 1966 and how closely this subject sits to the core of the human rights framework.

The 1975 Declaration, paragraph 6, requires States to protect people from misuse of scientific and technological developments, particularly with regard to respect for privacy and the protection of the human personality and its physical and intellectual integrity. The phrase covers both halves of this chapter: the body, and the mind.

Constitution of India, article 21, protects life and personal liberty, and has been construed to include bodily integrity. Article 20(3) provides that no person accused of any offence shall be compelled to be a witness against himself. The two operate together in the leading Indian case.

The Indian case

Selvi v. State of Karnataka, AIR 2010 SC 1974, (2010) 7 SCC 263, decided on 5 May 2010.

Facts. A group of criminal appeals raised the legality of administering three scientific investigative techniques to a person who had not consented: narcoanalysis, in which a drug is used to reduce a subject's inhibition against speaking; the polygraph examination, which measures physiological responses while a subject answers questions; and the Brain Electrical Activation Profile test, which measures electrical activity in the brain on exposure to stimuli associated with a crime. Accused persons, suspects and witnesses had been subjected to them without consent. The State defended the practice on the footing that the techniques assist where evidence is hard to gather by ordinary means, cause no bodily harm, would be used to guide the investigation rather than be led in evidence, and are a gentler alternative to the third degree, meaning the physical ill treatment of suspects in custody.

Held, on article 20(3). Compulsory administration violates the right against self-incrimination, whose protective scope extends to the investigative stage and, read with section 161(2) of the Code of Criminal Procedure 1973, protects accused persons, suspects and witnesses alike. Article 20(3) protects the choice between speaking and remaining silent whether the resulting testimony is inculpatory or exculpatory, and the results of each technique bear a testimonial character and cannot be categorised as material evidence.

Held, on article 21 and beyond. Forcing a person to undergo any of them also violates the substantive due process required for restraining personal liberty, and does so whether the technique is administered during an investigation or for any other purpose, since the results could expose a person to adverse consequences of a non-penal nature. The techniques cannot be read into the Explanation to sections 53, 53-A and 54 of the Code by an expansive interpretation. Compulsory administration is an unjustified intrusion into the mental privacy of the individual, would amount to cruel, inhuman or degrading treatment in the language of evolving international human rights norms, and conflicts with the right to a fair trial. Invocations of a compelling public interest cannot justify diluting such rights.

The order. No individual should be forcibly subjected to any of the techniques, whether in a criminal investigation or otherwise. Room is left for voluntary administration with safeguards, but even where the subject has consented the results by themselves are not admissible, because the subject does not exercise conscious control over the responses.

Why it matters here. Every argument the State made is the argument that technological capability generates, and the Court rejected each of them. The technique is useful: not an answer. It causes no bodily harm: not an answer, because the integrity protected includes the mind. It is gentler than the alternative: not an answer, because the alternative is itself unlawful. And the utility of a technique to the State does not answer the person's refusal. That proposition is the foundation of every later argument in this paper about compelled testing and compelled disclosure.

The structure of the right

An answer is much stronger if it separates four questions that are commonly run together.

1. Is there an interference with the body or the mind at all? A blood sample is obvious. An image taken through clothing, a reading of brain activity and an analysis of a discarded sample are less obvious and are still interferences. Selvi settles that the mind counts.
2. Was there consent, and was it valid? Prior, free, informed and, for research, express, on adequate information in a comprehensible form: UNESCO 2005, article 6. Consent given by a person in custody to a technique proposed by their custodian raises an obvious question about whether it was free.
3. If there was no consent, is there a law that authorises the interference? Legality, in the sense used in [Privacy and Surveillance]. Selvi holds that the Code's provisions on medical examination cannot be stretched to cover these techniques, which is a holding about the limits of interpretation as much as about the techniques.
4. If there is a law, is the interference proportionate? And for a category of interference the framework regards as absolute, such as experimentation without free consent under article 7, the question does not arise at all, because no justification is available.

Point four contains the distinction that separates a good answer from an average one. Some interferences are subject to justification and some are prohibited outright. Compulsory medical treatment may in principle be justified in a public health emergency. Scientific experimentation without free consent may not be justified at all, because article 7 admits of no limitation.

Worked example

A State proposes a statute requiring every person arrested for a scheduled offence to provide a DNA sample, to be retained on a permanent database irrespective of the outcome of the case, and permitting familial searching, that is, searching for partial matches in order to identify relatives of the person sampled.

Interference. Yes, twice. Taking the sample interferes with bodily integrity. Retaining and searching the profile interferes with informational privacy, and familial searching interferes with the privacy of people who were never arrested and never consented.

Consent. Absent, and the setting makes it unobtainable in any meaningful sense.

Legality. A statute is proposed, so this limb can be satisfied if the statute is precise about who may be sampled, what may be retained, for how long, who may search and on what grounds, and what remedy a person has.

Proportionality. The failures are at necessity and balance. Retention irrespective of outcome means the profile of an acquitted person is kept, which is not necessary for the aim of solving crimes committed by convicted persons. Familial searching extends the burden to persons against whom nothing whatever is alleged, purely because of a biological relationship, and no consent from them is even contemplated.

Which prohibition applies. This is a justifiable interference, so it is assessed rather than forbidden. It is not experimentation, so article 7's absolute rule is not engaged. The correct answer is that the scheme is capable of being lawful in a narrower form: sampling on charge for serious offences, destruction on acquittal, judicial authorisation for familial searching, and an accessible route to correction. The forensic setting is developed in [Forensic Science and the Rights of the Accused].

What it does NOT mean

It does not mean nothing may ever be done without consent. Emergency treatment of an unconscious patient, compulsory vaccination in a properly framed public health scheme and the taking of a sample under a valid statute may all be lawful. The right requires authority and proportion; it does not require permission in every case.

It is not limited to the body. Selvi is the authority: mental privacy is within the protection, and a technique that touches nobody may still be an intrusion.

Consent is not a signature. A form signed by a person who did not understand it, or who could not realistically refuse, fails the Nuremberg conditions of information and freedom. This returns in Module II with Samira Kohli.

Absence of harm is not consent. The State in Selvi argued that the techniques caused no bodily harm, and the Court did not treat that as relevant to the question whether they could be imposed.

Distinctions

Quick revision

• The principle: that a technique works is not a reason for using it on a person. Authority plus consent, or authority plus proportionality.
• Nuremberg Code, point 1: the voluntary consent of the human subject is absolutely essential, and it stands first of the ten. Point 9: the subject is at liberty to end the experiment.
• UNESCO Bioethics 2005, article 6 defines the consent: prior, free and informed, on adequate information, express where appropriate, withdrawable at any time and for any reason without disadvantage or prejudice; for research, also express, with information in a comprehensible form including how to withdraw. Article 6(3): a community agreement or a leader's consent never substitutes for the individual's. Article 7: for a person without capacity, direct health benefit, minimal risk by exception, and refusal is respected.
• ICCPR article 7: no torture or cruel, inhuman or degrading treatment; and in particular no one shall be subjected without his free consent to medical or scientific experimentation. No limitation clause.
• 1975 Declaration, paragraph 6: protection of the human personality and its physical and intellectual integrity.
• Selvi (2010): compulsory narcoanalysis, polygraph and BEAP violate article 20(3), which reaches the investigative stage and covers accused, suspects and witnesses; violate substantive due process under article 21; are an unjustified intrusion into mental privacy; would amount to cruel, inhuman or degrading treatment; and a compelling public interest cannot justify dilution. Voluntary administration is possible with safeguards, but the results alone are still inadmissible.
• Separate the justifiable interference from the absolutely prohibited one. That distinction carries marks.

Test yourself

1. State the first point of the Nuremberg Code, and say where the content of the consent it requires is to be found. It states that the voluntary consent of the human subject is absolutely essential. It is stated as an absolute rather than as a factor to be weighed against the value of the research, and it stands first of the ten points, before those dealing with the design of the experiment, the avoidance of suffering and the qualifications of the investigator. The Code does not itself define the consent. Article 6 of the UNESCO Universal Declaration on Bioethics and Human Rights 2005 supplies it: consent must be prior, free and informed and based on adequate information, express where appropriate, and withdrawable at any time and for any reason without disadvantage or prejudice; for scientific research it must also be express, and the information must be adequate, in a comprehensible form, and must include how consent may be withdrawn.

2. Why is the second sentence of ICCPR article 7 unusual, and what follows from where it is placed? It is unusual because it names a specific scientific practice, medical or scientific experimentation without free consent, inside a general prohibition on torture and cruel, inhuman or degrading treatment, where treaty articles ordinarily state a principle and leave applications to interpretation. What follows from its placement is that the prohibition is absolute: article 7 contains no limitation clause and is non-derogable, so experimentation without free consent cannot be justified by any aim however compelling, whereas most interferences with bodily integrity fall to be assessed under a proportionality test.

3. Summarise the reasoning in Selvi and identify which of the State's arguments the Court rejected. The Court held that compulsory administration of narcoanalysis, polygraph testing and the BEAP test violates article 20(3), whose protection extends to the investigative stage and covers accused persons, suspects and witnesses alike, the results being testimonial rather than material evidence; that it violates the substantive due process required before personal liberty is restrained, whether administered in an investigation or otherwise, because the results may expose a person to non-penal adverse consequences; that it is an unjustified intrusion into mental privacy and would amount to cruel, inhuman or degrading treatment in the language of evolving international human rights norms; and that it conflicts with the right to a fair trial. It rejected the arguments that the techniques are useful where ordinary evidence is scarce, that they cause no bodily harm, that they would merely guide the investigation, and that they are a gentler alternative to the third degree, holding that invocations of a compelling public interest cannot justify diluting such rights.

4. A hospital wishes to enrol unconscious accident victims in a trial of a new resuscitation technique, on the ground that consent is impossible and the technique may save lives. Analyse. The proposal is experimentation, so the governing provision is the second sentence of article 7 and the first point of the Nuremberg Code, both of which treat free consent as essential rather than as a factor. The impossibility of obtaining consent is a reason why the research is difficult, not a reason why the requirement does not apply, and the potential benefit does not supply a justification because the prohibition admits of none. The analysis is different from that of emergency treatment, where a doctor may act without consent in the patient's interests, because treatment is directed to the benefit of that patient whereas research is directed to the production of generalisable knowledge. The route, if there is one, lies in the research ethics framework taken up in Module III, through mechanisms such as consent by a legally authorised representative and independent ethical review, and not in dispensing with the principle.

5. Distinguish an interference with bodily integrity that is subject to justification from one that is prohibited outright, and give an example of each. An interference subject to justification is assessed against the three-part test: it must be authorised by an accessible and foreseeable law, pursue a legitimate aim, and be proportionate, meaning rationally connected, no more restrictive than necessary, and productive of a benefit outweighing its burden. Taking a DNA sample from a person charged with a serious offence under a precise statute is an example. An interference prohibited outright is one for which no justification is available, because the governing provision contains no limitation clause: medical or scientific experimentation without free consent under the second sentence of ICCPR article 7 is the example, and torture is the other. The practical importance of the distinction is that in the first category a compelling public interest is weighed, and in the second it is irrelevant.

CEDAW supplies two provisions that are directly about technology and are worth citing by number because so few candidates do. Article 11(3) requires that protective legislation relating to matters covered by the employment article be reviewed periodically in the light of scientific and technological knowledge and revised, repealed or extended as necessary. Article 14(2)(g) requires States to ensure to rural women the right to have access to agricultural credit and loans, marketing facilities, appropriate technology and equal treatment in land and agrarian reform.

Constitution of India, articles 14 and 15, guarantee equality before the law and the equal protection of the laws, and prohibit discrimination on grounds of religion, race, caste, sex or place of birth. Article 15(3) and (4) permit special provision, which is the constitutional basis for the positive measures this chapter argues for.

What the Committee requires

General Comment 25 treats non-discrimination as running through the elements of the right rather than sitting beside them.

Under accessibility it requires that scientific progress and its applications be accessible to all without discrimination, in the three dimensions set out in [Availability, Accessibility, Acceptability and Quality]: access to applications, access to information about risks and benefits, and the open opportunity to participate. It requires States to remove discriminatory barriers, giving as its example the facilitation of access by marginalised populations to scientific education.

Under availability it requires that benefits be distributed especially to vulnerable and marginalised groups, which is a distributive instruction and not merely a prohibition on exclusion.

On women it goes further and permits temporary special measures, such as quotas for women in scientific education, to speed up substantive equality; and it states that a gender-sensitive approach is not a luxury but a crucial tool, because scientific progress and new technologies must take account of the characteristics and needs of women and girls from the first stage of research onwards.

Formal and substantive equality

This is where most of the marks are, and it is a distinction rather than a fact, so it can be learnt once and used everywhere.

Formal equality asks whether the rule treats people alike. A scheme that makes an online service available to every citizen on identical terms satisfies it completely.

Substantive equality asks whether people can actually enjoy the right on equal terms. The same scheme fails at once where a third of the population has no connection, a language barrier, no device, or no ability to read.

The instruments are drafted for substantive equality, not formal. Article 2(2) requires that the rights be exercised without discrimination, not merely that they be granted alike; the Committee's requirement to remove barriers and to distribute benefits especially to marginalised groups is a positive duty; CEDAW article 4 and the Constitution's article 15(3) and (4) authorise differential treatment for that purpose; and General Comment 25 accepts quotas in terms.

The practical consequence is that a State cannot answer a complaint about the digital divide by showing that its scheme applies to everybody. That answer establishes formal equality and is the beginning of the problem, not the end of it.

Indirect discrimination and the digital divide

The divide rarely takes the form of a rule that excludes anyone. It takes the form of a facially neutral measure with unequal effect, which is the classic shape of indirect discrimination.

Four dimensions are worth naming separately, because an answer that names only the first is thin.

1. Access. Whether a person has a connection, a device and reliable electricity. This is the dimension everyone writes about.
2. Skills. Whether a person can use them. A connection given to somebody who cannot read the interface transfers the hardware and not the benefit.
3. Quality of use. Whether what a person can do with the technology is comparable. A shared handset used for half an hour a day is not the same access as a private computer.
4. Representation in the technology itself. Whether the technology was built for them at all. A system trained on data that under-represents a population performs worse for that population, which converts the divide from a question of access into a question of design. This is the point General Comment 25 makes about gender-sensitive research from the first stage.

Worked example

A State moves applications for a food subsidy entirely online, with biometric authentication at the point of collection. The stated purposes are the elimination of fraud and administrative saving. The rule applies to every applicant identically.

Formal equality. Satisfied. There is no classification on any prohibited ground.

Substantive equality. Test the four dimensions. Access: applicants without a connection or a device must rely on an intermediary, at a cost. Skills: applicants who cannot read cannot complete the form unaided, and the intermediary is now in a position of power over them. Quality of use: an applicant with intermittent connectivity may fail at the last step and lose the entitlement. Representation: biometric authentication fails more often for manual labourers, whose fingerprints are worn, and for the elderly, and those are disproportionately the people the subsidy exists for.

Which right is engaged? Not only equality. The subsidy exists to secure an adequate standard of living, so a technological barrier to obtaining it is an interference with that right, effected through a means that bears unequally. Article 2(2) attaches the non-discrimination requirement to the exercise of the Covenant right itself.

What the State must show. That the measure pursues a legitimate aim, which it plainly does; and that the disparate effect is justified and mitigated. The mitigation is the whole of the practical answer: an offline route that is genuinely available rather than nominal, an exception procedure for authentication failure, and a rule that no entitlement is lost because a machine did not recognise a person.

The general lesson. The problem is not the technology. It is the removal of the alternative. A digital route added to a paper route enlarges access; a digital route that replaces a paper route redistributes it.

What it does NOT mean

It does not mean new technology should be withheld until everyone can have it. That would defeat the right in article 15(1)(b) rather than serve it. The duty is to extend the benefit and to attend to the distribution, which is what paragraph 6 of the 1975 Declaration says in one sentence.

It is not only about the internet. The divide in irrigation, in diagnostics, in cold chains and in laboratory access is older and in places more consequential. The digital case is the current instance of a permanent pattern.

Unequal outcome is not automatically discrimination. The framework asks whether the disadvantage is connected to a protected ground or to a group the Committee identifies as marginalised, and whether the State has taken the positive steps required of it. An answer that treats every inequality as a violation has stopped doing law.

And equality does not require identical treatment. CEDAW article 4, the Constitution's article 15(3) and (4), and the Committee's acceptance of temporary special measures all point the other way.

Quick revision

• The digital divide: the gap in access to information and communication technologies and in the education and infrastructure to use them; the leading instance of technology as a multiplier of existing inequality.
• ICESCR article 2(2): Covenant rights, including article 15(1)(b), to be exercised without discrimination on any of the listed grounds. The key provision.
• 1975 Declaration, paragraph 7: measures, including legislative measures, so that utilisation of scientific achievements promotes the fullest realisation of rights without any discrimination whatsoever on grounds of race, sex, language or religious beliefs.
• CERD article 5: equality before the law in the enjoyment of economic, social and cultural rights including public health, medical care, education and training.
• CEDAW article 11(3): protective legislation to be reviewed in the light of scientific and technological knowledge. Article 14(2)(g): rural women's access to appropriate technology.
• General Comment 25: remove discriminatory barriers; distribute benefits especially to vulnerable and marginalised groups; temporary special measures including quotas are permissible; a gender-sensitive approach from the first stage of research.
• Four dimensions: access, skills, quality of use, and representation in the technology itself.
• Formal equality asks if the rule treats alike; substantive equality asks if the right can be exercised alike. The instruments are drafted for the second.

Test yourself

1. Which single provision most directly makes unequal access to science a question of discrimination, and why? Article 2(2) of the ICESCR, which requires States Parties to guarantee that the rights in the Covenant will be exercised without discrimination of any kind as to race, colour, sex, language, religion, political or other opinion, national or social origin, property, birth or other status. It is the most direct because the right to enjoy the benefits of scientific progress in article 15(1)(b) is itself a Covenant right, so the non-discrimination guarantee attaches to it without any further step of reasoning. Its wording also matters: it requires that the rights be exercised without discrimination, not merely conferred alike, which is the textual basis for a substantive rather than a formal reading.

2. Distinguish formal from substantive equality and say which the instruments adopt. Formal equality asks whether a rule treats people alike, and is satisfied by a measure that applies to everyone on identical terms. Substantive equality asks whether people are able in fact to enjoy the right on equal terms, and is not satisfied by identical treatment where the starting positions differ. The instruments adopt the substantive conception: article 2(2) speaks of rights being exercised without discrimination; General Comment 25 imposes positive duties to remove discriminatory barriers and to distribute benefits especially to vulnerable and marginalised groups, and accepts temporary special measures including quotas; CEDAW article 4 and articles 15(3) and 15(4) of the Constitution authorise differential treatment for that purpose.

3. Name the two CEDAW provisions that bear directly on technology and state what each requires. Article 11(3) requires that protective legislation relating to the matters covered by the employment article be reviewed periodically in the light of scientific and technological knowledge and be revised, repealed or extended as necessary. It is significant because it makes keeping the law in step with science a legal duty rather than good practice. Article 14(2)(g) requires States to ensure to rural women the right to have access to agricultural credit and loans, marketing facilities, appropriate technology and equal treatment in land and agrarian reform, which states access to technology as a component of equality rather than as a benefit conferred.

4. A State replaces a paper application route with an online one that applies to all citizens identically. Identify the equality problem and the four dimensions along which you would test it. The problem is indirect discrimination: a facially neutral measure with unequal effect, which satisfies formal equality while redistributing practical access. The dimensions are access, whether the person has a connection, a device and reliable power; skills, whether they can use it unaided, since a person who cannot read the interface receives the hardware and not the benefit; quality of use, whether what they can do with it is comparable, a shared handset for a short period not being equivalent to private and reliable access; and representation in the technology itself, whether the system was designed and tested for them, since a system built on data under-representing a population performs worse for that population. The decisive fact is usually not the addition of the digital route but the removal of the alternative.

5. Why is it wrong to answer a digital divide complaint by saying that a new technology should not be introduced until everyone can have it? Because it treats the right as a right to equality alone, when article 15(1)(b) is a right to enjoy the benefits of scientific progress and article 15(2) imposes a duty to develop and diffuse science. Withholding a benefit from those who could have it does not confer it on those who cannot; it defeats the right for both groups while producing the appearance of equal treatment. Paragraph 6 of the 1975 Declaration states the correct obligation in a single sentence, requiring States both to extend the benefits of science and technology to all strata of the population and to protect them from harmful effects, so the duty is to extend and to attend to the distribution at the same time, not to choose between them.

Constitution of India, article 21, protects life and personal liberty and has been construed to include the right to health and to emergency medical treatment. Article 47, a directive principle, makes it a primary duty of the State to raise the level of nutrition and the standard of living and to improve public health. The Indian material is developed in [Science Under Article 21] and [The Directive Principles on Health and Work].

The three points of engagement

An answer is much stronger if it separates the three, because they have different tests, different duty bearers and different remedies.

As a benefit: distribution

The question is who gets it. The governing provisions are article 12, article 15(1)(b) and the non-discrimination guarantee in article 2(2). The four elements in [Availability, Accessibility, Acceptability and Quality] apply directly, and General Comment 25 makes the connection explicit when it says that everyone should have equal access to the applications of science particularly when they are instrumental for the enjoyment of other economic, social and cultural rights, health being the leading example.

The characteristic failures here are the ones in [Equality and the Digital Divide]: the technology exists and is concentrated where the money is. The characteristic remedy is distributive, not prohibitory.

As an intervention: consent

The question is whether it may be done to this person. The governing principles are in [Bodily Integrity and Intervention Without Consent], and the law is worked out in Module II through the doctrine of informed consent and the standard of care. Two Indian cases hold the ground and are taken there: Samira Kohli v. Dr. Prabha Manchanda on the consent a patient actually gave, and Jacob Mathew v. State of Punjab on when a failure becomes a crime rather than a tort.

The characteristic failure is the treatment given without a consent that was real. The characteristic remedy is compensatory, and sometimes criminal.

As a capability: the new questions

The question is what the law should be at all, because there was no rule before the capability existed. Life support made it necessary to decide when treatment may be withdrawn. Prenatal diagnosis made it necessary to decide what may be done with the result, which in India produced the Pre-conception and Pre-natal Diagnostic Techniques Act 1994. Assisted reproduction made it necessary to decide who a child's parents are, which produced the Surrogacy (Regulation) Act 2021 and the Assisted Reproductive Technology (Regulation) Act 2021. Transplantation made it necessary to define death, which the Transplantation of Human Organs and Tissues Act 1994 does.

The characteristic failure here is delay: the capability arrives, the law does not, and the gap is filled by contract, by professional guidance or by nothing at all. All four statutes named above are Module II.

Worked example

A district hospital acquires a diagnostic device that detects a treatable cancer several years earlier than existing methods. It is expensive, and the hospital can screen roughly one tenth of the eligible population each year.

As a benefit. Article 12 requires steps towards the highest attainable standard, and the arrival of the device raises what is attainable. That does not oblige the State to screen everyone immediately; economic and social rights are subject to progressive realisation within available resources. It does oblige the State to have a defensible basis for the selection, because a scarce benefit distributed without criteria is distributed by proximity, influence and wealth, which is where article 2(2) bites. Clinical risk is a defensible criterion; ability to reach the hospital on a working day is not.

As an intervention. Screening is done to a person, so consent is required, and consent to a screening test is not merely consent to the procedure. It includes understanding what a positive result will mean, what the false positive rate is and what will follow. A signature obtained without that is not informed consent in the sense of UNESCO 2005 article 6.

As a capability. The device generates knowledge that did not previously exist, and the law has to decide who else may have it. May the result be shared with an insurer? With an employer? The question is not medical at all: it is the informational privacy question from [Privacy and Surveillance], arriving through a clinic.

What ties the three together. One machine, three legal analyses, three different bodies of law, three different remedies. A candidate who runs them together produces an answer that is about health policy. A candidate who separates them produces an answer that is about law.

What it does NOT mean

The highest attainable standard is not a guarantee of health. It is a right to the conditions and services necessary for health, judged against what is attainable, and no State guarantees any individual good health.

Progressive realisation is not an excuse. It permits a State to move towards full realisation over time within its resources. It does not permit inaction, does not permit retrogression without justification, and does not displace the minimum core or the immediate obligation of non-discrimination in article 2(2).

Access to a doctor is not access to medical technology. The four elements are the reason: availability of a service does not establish accessibility, quality or acceptability, and a scheme that supplies a clinic without supplying what the clinic needs has satisfied one element.

And the arrival of a technology can reduce rights as well as expand them. A test that can be done can be demanded, by an employer, an insurer or a family, and the pressure falls hardest on those least able to resist it. That is the dual use proposition of [Dual Use: the Same Technology Both Ways] in its medical form.

Quick revision

• Medical technology engages the framework at three points: as a benefit (distribution, article 12 with article 15(1)(b) and article 2(2)); as an intervention (consent and bodily integrity); as a capability (new questions the old law never answered).
• ICESCR article 12(1): the highest attainable standard of physical and mental health. The content rises as science advances, which is the clearest link between the two halves of this paper. Article 12(2) lists steps: stillbirth and infant mortality, healthy development of the child, environmental and industrial hygiene, epidemic and occupational disease, and medical service and attention in sickness.
• UDHR article 25(1): an adequate standard of living including medical care.
• Constitution: article 21 (health and emergency treatment), article 47 (public health as a primary duty of the State).
• General Comment 25: equal access to applications of science particularly where instrumental for other economic, social and cultural rights.
• The three characteristic failures: maldistribution, treatment without real consent, and delay in legislating for a new capability.
• The four Indian statutes generated by capability, all Module II: PCPNDT 1994, Surrogacy 2021, ART 2021, Transplantation of Human Organs and Tissues 1994.

Test yourself

1. Why is the phrase "highest attainable standard" the most important wording in article 12 for this paper? Because it makes the content of the right depend on what is attainable, and what is attainable is a function of scientific progress as well as of available resources. A standard fixed absolutely would be unachievable and a standard fixed at current provision would be meaningless, whereas a standard of the highest attainable rises whenever science makes a new prevention, diagnosis or treatment possible. That is the precise mechanism by which an advance in science enlarges a legal obligation, which is the relationship this paper exists to study.

2. Set out the three points at which medical technology engages the human rights framework, with the governing test for each. As a benefit, where the question is distribution: the governing provisions are article 12, article 15(1)(b) and the non-discrimination guarantee in article 2(2), applied through availability, accessibility, quality and acceptability, and the remedy is distributive. As an intervention, where the question is whether it may be done to this person: the governing principles are consent and bodily integrity, worked out through informed consent and the standard of care, and the remedy is compensatory and sometimes criminal. As a capability, where the question is what the rule should be at all because none existed before: the characteristic failure is legislative delay, and the response is a statute.

3. A scarce diagnostic device can screen a tenth of the eligible population. Does article 12 require the State to screen everyone, and what does it require? No. The right is to the highest attainable standard and is subject to progressive realisation within available resources, so a State is not in breach merely because it cannot immediately extend a new technology to everyone. What it requires is that the State take deliberate steps towards fuller realisation, and that the selection among those who receive the scarce benefit rest on a defensible basis. That second requirement follows from article 2(2), which attaches non-discrimination to the exercise of Covenant rights immediately and not progressively: a scarce benefit distributed without criteria is in practice distributed by proximity, influence and wealth, and clinical need is a defensible criterion where ease of access to the hospital is not.

4. Explain the "delay" failure and give two Indian examples of statutes enacted to answer it. The delay failure occurs where a technological capability arrives and the law that should govern it does not, so that the gap is filled by contract, by professional guidance or by nothing, and the persons affected have no defined rights while it lasts. Prenatal diagnostic technology made it possible to determine the sex of a foetus and was answered by the Pre-conception and Pre-natal Diagnostic Techniques Act 1994. Assisted reproductive technology raised questions of parentage, of the position of the woman carrying the child and of commercial exploitation, and was answered by the Surrogacy (Regulation) Act 2021 and the Assisted Reproductive Technology (Regulation) Act 2021. Transplantation required a legal definition of death, supplied by the Transplantation of Human Organs and Tissues Act 1994.

5. In what sense can a medical technology reduce a person's rights rather than expand them? Because a test that can be performed can also be demanded. Once a condition can be detected, an employer, an insurer, a prospective spouse or a family may require the detection as a condition of something the person needs, and the resulting knowledge can be used to exclude them. The capability therefore creates a new pressure that did not exist before, and it falls hardest on those least able to refuse. The analysis is the dual use proposition in medical form: the same technology serves the right to health when used for the patient's benefit and defeats privacy, equality and sometimes livelihood when used at another party's demand, so the legal question is never whether the technology is good but who may require it and what may be done with the result.

Article 7 recognises the right of everyone to the enjoyment of just and favourable conditions of work, which ensure in particular fair wages and equal remuneration for work of equal value without distinction of any kind, women being guaranteed conditions not inferior to those of men with equal pay for equal work; a decent living for the worker and their family; safe and healthy working conditions; equal opportunity for promotion subject to no considerations other than seniority and competence; and rest, leisure, reasonable limitation of working hours and periodic holidays with pay.

Two limbs of article 7 do work in this chapter that is easy to miss. Safe and healthy working conditions is the provision under which the introduction of a hazardous process is challenged, which links this chapter to [Environment and Technological Harm]. And equal opportunity for promotion subject to no considerations other than seniority and competence is engaged where the assessment is made by an automated system, which links it to [Artificial Intelligence and Decisions by Machines].

UDHR article 23 states the right to work, to free choice of employment, to just and favourable conditions and to protection against unemployment.

CEDAW article 11(3) requires that protective legislation relating to the matters covered by the employment article be reviewed periodically in the light of scientific and technological knowledge and revised, repealed or extended as necessary. It is the only provision in the base that expressly requires labour law to be kept in step with technological change, and it should be cited whenever the question is what the State ought to have done.

The Indian position

Article 21 has been construed to include the right to livelihood, on the reasoning that the right to life would be reduced to animal existence if it did not include the means by which a person lives. The doctrine matters here because it converts the loss of a means of living into a question about a fundamental right, though it does not follow that any particular job is protected.

Article 39(a), a directive principle, requires the State to direct its policy towards securing that the citizens, men and women equally, have the right to an adequate means of livelihood.

Article 41 requires the State, within the limits of its economic capacity and development, to make effective provision for securing the right to work, to education and to public assistance in cases of unemployment, old age, sickness and disablement, and in other cases of undeserved want.

Article 43 requires the State to endeavour to secure to all workers a living wage, conditions of work ensuring a decent standard of life and full enjoyment of leisure and social and cultural opportunities.

Taken together the Indian frame is: a fundamental right to livelihood, with the machinery of protection sitting in Part IV. That asymmetry is the source of most of the difficulty, because the enforceable right is stated at a high level of generality and the specific obligations are in the unenforceable part.

What the right does and does not require

This is the heart of the chapter and the place to be precise.

It does not require that a technology be kept out. No provision in the base gives a worker a right that their employer continue using an obsolete process, and a right of that kind would defeat article 15(1)(b) and article 15(2), which require the development and diffusion of science. An answer that concludes that automation violates the right to work has reasoned from the wrong premise.

It does not guarantee any particular job. The right in article 6 is to the opportunity to gain a living by work freely chosen or accepted, which is a right to a functioning labour market and to the State's steps towards full and productive employment, not a right to remain in one post.

It does require the State to act on the transition. Article 6(2) puts technical and vocational guidance and training in the mandatory content. Article 41 requires provision for public assistance in unemployment. CEDAW article 11(3) requires the law itself to be reviewed as technology changes. The obligation is therefore concrete: retraining, income support during transition, and legislative review.

It does require attention to who bears the loss. Article 2(2) attaches non-discrimination to the exercise of Covenant rights. Where displacement falls disproportionately on women, on a caste or on a region, the equality analysis in [Equality and the Digital Divide] applies to the displacement itself and not only to access.

And it does require that the new work be work. Article 7's standards do not lapse because the work is mediated by a platform. Fair wages, safe conditions, rest and reasonable limitation of hours are stated as attributes of work, not of a contractual form, so an arrangement that reclassifies a worker without changing what they do engages article 7 directly.

Worked example

A State transport undertaking introduces automated fare collection and closes the conductor grade, which employs about four thousand people, most of whom joined between fifteen and twenty-five years ago and have no other qualification. It offers a redundancy payment calculated on years of service.

Is the introduction of the technology unlawful? No, and an answer should say so at once. Nothing in article 6, article 21 or the directive principles gives a worker the right to the continuation of a superseded process.

What obligations are engaged? Article 6(2) directly: technical and vocational guidance and training programmes are part of the mandatory content of the right to work, so the question is whether retraining was offered, whether it was real, and whether it led anywhere. Article 41: public assistance in cases of unemployment is what the transition period requires. Article 39(a): adequate means of livelihood, which is the standard against which the redundancy payment is measured, since a payment that does not bridge the period to new employment has not secured a means of livelihood, only softened its loss.

Is there an equality question? Test it rather than assume it. If the conductor grade was one of the few open to a particular group, its closure removes that group's route into the undertaking, and a facially neutral decision has an unequal effect.

What about those redeployed? Article 7 applies to the new posts: equal opportunity for promotion on seniority and competence, and conditions not inferior. A redeployment to a grade with no progression satisfies the form of continued employment and not the substance of article 7.

The realistic conclusion. The undertaking may automate. It must retrain, support the transition, examine the distribution of the loss, and preserve conditions for those who remain. That is a lawyer's answer, and it is more useful to the four thousand than an argument that the machines should be banned.

What it does NOT mean

Technology does not simply destroy work. It changes its composition, and the historical record is that total employment has not fallen. That is not a reason for complacency, because the framework's concern is the distribution and the transition, and both can be severe while the aggregate is unchanged.

"New jobs will appear" is not an answer to a rights claim. It may be true and still irrelevant, because the person displaced is not the person who takes the new job, and article 6(2) addresses precisely that gap.

The right to work is not a right to be employed by anybody in particular. The duty bearer is the State, and its duty is to take steps towards full and productive employment, not to direct a private employer to retain a post.

And a productivity gain is not automatically a public benefit. Whether it is depends on where it goes, which is a distributive question and therefore a question the equality provisions reach.

Quick revision

• ICESCR article 6(1): the right to work includes the opportunity to gain a living by work freely chosen or accepted. Article 6(2): steps shall include technical and vocational guidance and training programmes and policies for full and productive employment. Article 6(2) is the displaced worker's provision.
• Article 7: just and favourable conditions, including fair wages and equal pay for work of equal value, a decent living, safe and healthy working conditions, equal opportunity for promotion on seniority and competence only, and rest and reasonable limitation of hours.
• UDHR article 23; CEDAW article 11(3), labour law to be reviewed in the light of scientific and technological knowledge.
• India: article 21 includes the right to livelihood; article 39(a) adequate means of livelihood; article 41 public assistance in unemployment; article 43 living wage and decent conditions. The enforceable right is general, the specific duties sit in Part IV.
• The law does not forbid the technology and does not protect a particular job. It requires retraining, transitional support, legislative review, and attention to who bears the loss.
• The three constants of every wave: output rises; gains and losses fall on different people; the harm is in the transition.

Test yourself

1. Which provision of the Covenant is the displaced worker's provision, and why? Article 6(2), which provides that the steps to be taken to achieve the full realisation of the right to work shall include technical and vocational guidance and training programmes, and policies and techniques to achieve steady development and full and productive employment under conditions safeguarding fundamental political and economic freedoms. It is the displaced worker's provision because it places retraining inside the mandatory content of the right rather than leaving it to policy, so a State that permits a technological displacement without providing vocational guidance and training has failed a step the treaty says its steps shall include. It is also the provision that answers the objection that new work will appear, since it addresses the position of the person during the interval.

2. Does the right to work make the introduction of labour-replacing technology unlawful? Give reasons. No. Article 6(1) protects the opportunity to gain a living by work freely chosen or accepted, not the continuation of any particular post or process, and the duty bearer is the State rather than the individual employer. A right to the continuance of a superseded process would also conflict with article 15(1)(b) and article 15(2), which give everyone the right to enjoy the benefits of scientific progress and require States to take steps for the development and diffusion of science. What the framework requires is not prohibition but management of the transition: vocational training under article 6(2), public assistance in unemployment under article 41 in the Indian scheme, review of protective labour legislation in the light of technological knowledge under CEDAW article 11(3), and attention under article 2(2) to whether the loss falls disproportionately on a protected group.

3. Identify two limbs of article 7 that bear on technological change and explain how. Safe and healthy working conditions is engaged whenever a new process introduces a hazard, so it is the provision under which the introduction of a dangerous technology is challenged by those who must work with it, and it connects this topic to industrial disaster and environmental harm. Equal opportunity for everyone to be promoted to an appropriate higher level, subject to no considerations other than seniority and competence, is engaged where selection or assessment is performed by an automated system, because a system that scores on correlates of past outcomes introduces considerations other than seniority and competence, which is what the limb forbids.

4. Set out the Indian constitutional position and explain why its structure creates difficulty. Article 21 has been construed to include the right to livelihood, on the footing that the right to life would be reduced to mere animal existence if it excluded the means by which a person lives. Article 39(a) directs the State to secure to citizens, men and women equally, the right to an adequate means of livelihood; article 41 requires effective provision, within the State's economic capacity, for the right to work and for public assistance in unemployment; and article 43 requires endeavour towards a living wage and decent conditions. The difficulty is structural: the enforceable guarantee in Part III is stated at a high level of generality, while the specific obligations that would answer a displacement, retraining and unemployment assistance, sit in Part IV and are not enforceable by a court, so the right that can be asserted is vague and the duties that are precise cannot be compelled.

5. A platform reclassifies its delivery riders as independent contractors while their work is unchanged. Which provision is engaged and why does the contractual form not settle the question? Article 7 is engaged, which recognises the right of everyone to the enjoyment of just and favourable conditions of work, including fair wages and equal remuneration for work of equal value, a decent living, safe and healthy working conditions, and rest, leisure and reasonable limitation of working hours with periodic holidays with pay. The contractual form does not settle the question because article 7 states these as attributes of work and of the person who performs it, not as incidents of a particular legal relationship, and article 6(1) speaks of the opportunity to gain a living by work rather than of employment in a technical sense. A reclassification that changes the label while leaving the direction, the hours and the economic dependence intact therefore alters who bears the cost of the standards without altering the standards themselves, and CEDAW article 11(3) supplies the further point that protective legislation is to be reviewed and extended as technological knowledge changes, which is exactly what a new mode of engaging labour calls for.

Note the structure, because it is the same three-part test used in [Privacy and Surveillance]: a law, a listed aim, and necessity. Note also that the list of aims in article 19(3) is closed. A restriction serving an aim not on the list fails whatever its merits.

Constitution of India, article 19(1)(a), guarantees freedom of speech and expression, and article 19(2) permits reasonable restrictions on eight enumerated grounds. That list too is closed, which is the pivot of the Indian case.

The Indian case

Shreya Singhal v. Union of India, AIR 2015 SC 1523, (2015) 5 SCC 1, decided on 24 March 2015 by Chelameswar and Nariman JJ, judgment by Nariman J.

Facts. Writ petitions challenged section 66A of the Information Technology Act 2000, inserted by the 2008 amendment, which punished sending by a computer resource or communication device information that was grossly offensive or had menacing character, or that the sender knew to be false but sent for the purpose of causing annoyance, inconvenience, danger, obstruction, insult, injury, criminal intimidation, enmity, hatred or ill will. The petitions followed arrests for material posted online, and also challenged section 69A and the Blocking Rules 2009, section 79 and the Intermediary Guidelines Rules 2011, and section 118(d) of the Kerala Police Act.

Held, on section 66A. It is struck down in its entirety, being violative of article 19(1)(a) and not saved under article 19(2). The Court held it unconstitutional on the ground of overbreadth, because it takes within its sweep protected speech and speech that is innocent in nature, and is therefore liable to be used so as to have a chilling effect on free speech.

Held, on the rest of the scheme. Section 69A and the Blocking Rules 2009 are constitutionally valid. Section 79 is valid subject to section 79(3)(b) being read down to mean that an intermediary, on receiving actual knowledge from a court order or on being notified by the appropriate government or its agency that unlawful acts relatable to article 19(2) are going to be committed, then fails expeditiously to remove or disable access to the material; the Intermediary Guidelines Rules 2011 are valid subject to rule 3(4) being read down in the same way, the knowledge spoken of being only through the medium of a court order. Section 118(d) of the Kerala Police Act is struck down as violative of article 19(1)(a) and not saved by article 19(2).

Why it matters here. It is the clearest Indian instance of the sequence this paper describes: a technology creates a capability; Parliament legislates in wide terms because it cannot foresee the uses; a court cuts the legislation back on constitutional grounds. It also decides the practical question for every platform, by making a court order, rather than a private complaint, the trigger for an intermediary's duty to remove. A candidate should note both halves of the decision. Answers that say only that section 66A was struck down leave out the part that governs everyday practice.

A warning about the text. The consolidated copy of the Information Technology Act served by India Code still prints section 66A. It is void, and a candidate who paraphrases it as live law has made an error that the judgment itself forecloses.

Restriction in the new forms

The older law of speech contemplated a prosecution after publication, or a licence before it. Technology produced four mechanisms that fit neither description, and they should be distinguished in an answer because their legal treatment differs.

1. Blocking. The State orders access to specified material disabled. In India this is section 69A, upheld in Shreya Singhal, the Court noting the safeguards in the section and the Rules. It is prior restraint in substance and is treated as permissible where the statutory conditions are met.
2. Intermediary liability. The State makes the carrier answerable for what the user says, which converts a private platform into a censor with an incentive to remove first and ask later. The Shreya Singhal reading down of section 79(3)(b) is the answer to that incentive.
3. Shutdown. Access is suspended for a region or a population, restricting not one speaker but everybody's right to seek and receive. It is treated under proportionality, and the leading Indian authority is queued for Module IV.
4. Private moderation. The largest volume of restriction is imposed by platforms under their own terms, where no State action is involved and the human rights framework operates, if at all, through the State's duty to protect rather than its duty to respect. This is the point developed in [The State's Duty to Regulate Technology].

Worked example

A State issues an order requiring platforms to remove, within six hours of a complaint from any government officer, content that is "likely to disturb public tranquillity", failing which the platform loses its protection from liability for all user content.

Which right, and whose? Article 19(1)(a) and article 19(2) of the Covenant, of the person whose content is removed; and the right to seek and receive of everybody who would have read it. The second is usually forgotten and is the stronger point where the removal is at scale.

Legality. Is "likely to disturb public tranquillity" precise enough that a person can foresee what is prohibited? Shreya Singhal struck down section 66A partly because terms such as "grossly offensive" and "annoyance" left a person of ordinary intelligence unable to know what was forbidden.

Legitimate aim. Public order is on the list in both article 19(3) and article 19(2) of the Constitution, so the aim is available.

Necessity and proportionality. Six hours with liability for the whole platform on failure is an incentive to remove everything complained of without examining it, which is the structural objection to intermediary liability. The reading down in Shreya Singhal addresses exactly this by requiring a court order.

Overbreadth and chilling effect. The two doctrines the case turns on. The scheme reaches protected speech along with unprotected, and its effect is that speakers avoid lawful speech for fear of the consequence. Both are grounds of invalidity in themselves, independent of whether the power is in fact abused.

What it does NOT mean

It does not mean online speech has a special or lesser protection. The right applies through "any media"; the medium changes the analysis of proportionality, not the standard.

Striking down a provision is not deregulation. Shreya Singhal upheld the blocking power and preserved intermediary liability in a narrowed form. It replaced a bad rule with a workable one.

Private removal is not censorship in the constitutional sense, because the constitutional guarantee runs against the State. That is a reason to reach it through the duty to protect and through regulation, not a reason to treat it as unimportant, since it is quantitatively the largest restriction there is.

And the right to hold opinions is absolute. Article 19(1) has no limitation clause, which matters when a technology proposes to infer or influence what a person thinks rather than to restrict what they say.

Quick revision

• UDHR article 19 and ICCPR article 19(2): freedom to seek, receive and impart information and ideas through any media and regardless of frontiers. The technology expanded the first two, which the older law neglected.
• ICCPR article 19(1), the right to hold opinions without interference, is absolute. Article 19(3) permits restrictions only if provided by law and necessary for a closed list of aims: rights or reputations of others, national security, public order, public health or morals.
• Constitution: article 19(1)(a) with the eight grounds in article 19(2), also a closed list.
• Shreya Singhal (2015): section 66A struck down in its entirety, violative of article 19(1)(a), not saved by article 19(2), on the ground of overbreadth because it sweeps in protected and innocent speech and has a chilling effect. Section 69A and the Blocking Rules 2009 valid. Section 79 valid subject to section 79(3)(b) being read down to require actual knowledge by court order or government notification; rule 3(4) read down likewise. Section 118(d) of the Kerala Police Act struck down.
• India Code's consolidated Act still prints section 66A. It is void.
• Four new mechanisms of restriction: blocking, intermediary liability, shutdown, private moderation.

Test yourself

1. Which two of the three activities protected by article 19 were transformed by information technology, and why does that matter? The freedom to impart information and the freedom to seek and receive it. When the means of mass publication were expensive, the freedom to impart belonged in practice to those who owned presses and broadcasting licences, and the freedom to seek and receive was confined to what those owners chose to publish. Technology made imparting nearly costless and made seeking a real activity. It matters because it establishes that the relationship between this technology and this right is not only one of threat: the same development that supplied new means of restriction also enlarged the substance of the right, which is the dual use proposition applied to expression.

2. Set out what was decided in Shreya Singhal, on all four matters before the Court. Section 66A of the Information Technology Act 2000 was struck down in its entirety as violative of article 19(1)(a) and not saved by article 19(2), on the ground of overbreadth, because it takes within its sweep protected and innocent speech and is liable to be used so as to have a chilling effect. Section 69A and the Blocking Rules 2009 were held constitutionally valid. Section 79 was held valid subject to section 79(3)(b) being read down, so that an intermediary's duty arises only on actual knowledge received through a court order or on notification by the appropriate government or its agency that unlawful acts relatable to article 19(2) are going to be committed; the Intermediary Guidelines Rules 2011 were held valid subject to rule 3(4) being read down in the same manner. Section 118(d) of the Kerala Police Act was struck down as violative of article 19(1)(a) and not saved by article 19(2).

3. Explain overbreadth and chilling effect, and why each is a ground of invalidity independent of actual abuse. A provision is overbroad when its terms reach beyond the speech that may legitimately be restricted and take in speech that is protected or innocent. A chilling effect is the consequence that speakers, uncertain where the line falls and unwilling to risk prosecution, refrain from lawful expression. Each is independent of actual abuse because the injury is done by the existence of the provision rather than by its enforcement: a person who does not speak because of a law that might reach them has lost the right without any prosecution occurring, and the loss cannot be repaired by an assurance that the power will be used sparingly, since the assurance does not restore the certainty the speaker lacked.

4. Why is intermediary liability a structural threat to expression, and how did the Court answer it? Because it places the cost of a wrong decision entirely on one side. A platform that removes lawful content loses nothing, while a platform that leaves unlawful content in place risks liability for everything its users say, so any rule that makes liability turn on a private complaint gives the platform an overwhelming incentive to remove first and examine afterwards. The censor is then a private party with no duty to the speaker and no procedure the speaker can invoke. The Court answered it by reading down section 79(3)(b) so that the intermediary's obligation is triggered only by actual knowledge through a court order, or by notification from the appropriate government or its agency relating to acts covered by article 19(2), and by reading rule 3(4) of the Intermediary Guidelines Rules 2011 in the same way.

5. Why does it matter that the list of permissible aims in article 19(3) is closed? Because it removes from the State the argument that a restriction is justified by some purpose it considers valuable but that the instrument does not name. A restriction on expression must be provided by law and necessary for respect of the rights or reputations of others, or for the protection of national security, public order, public health or morals, and no other aim will do. A State that restricts speech to protect the dignity of an office, to prevent embarrassment, or to preserve an official account of events is therefore defeated at the second stage without any inquiry into proportionality. Article 19(2) of the Constitution works in the same way with its own eight enumerated grounds, which is why so much Indian argument is about whether a restriction can be fitted into one of them.

Constitution of India, article 21, protects life, and has been construed to include a right to a wholesome environment. Article 48A directs the State to protect and improve the environment and to safeguard the forests and wild life, and article 51A(g) makes it a fundamental duty of every citizen to protect and improve the natural environment and to have compassion for living creatures. These are worked in [The Environment in the Constitution].

The rule: absolute liability

M.C. Mehta v. Union of India, AIR 1987 SC 1086, (1987) 1 SCC 395, decided on 20 December 1986 by five judges, judgment by Bhagwati CJ.

Facts. The petitioners sought, by a writ petition under article 32, the closure of the units of Shriram Foods and Fertiliser Industries on the ground that they were hazardous to the community. While the petition was pending, oleum gas escaped from one of the units. The Delhi Legal Aid and Advice Board and the Delhi Bar Association applied for compensation for those harmed. A Bench of three judges, permitting Shriram to restart on conditions, referred the compensation applications to five judges because questions of great constitutional importance arose, including the scope of the Court's jurisdiction under article 32.

Held, the new rule. The Court refused to be confined to the rule in Rylands v. Fletcher as developed in England, saying that India no longer needed the crutches of a foreign legal order and could evolve a new principle to meet the situations thrown up by hazardous industry. It held that an enterprise engaged in a hazardous or inherently dangerous industry which poses a potential threat to the health and safety of those working in the factory and residing in the surrounding areas owes an absolute and non-delegable duty to the community to ensure that no harm results from the hazardous nature of the activity. The activity must be conducted with the highest standards of safety, and if harm results the enterprise is absolutely liable to compensate; it is no answer that it took all reasonable care and that the harm occurred without negligence.

Held, why the enterprise must bear it. Permission to carry on such an activity for private profit is to be treated as conditional on the enterprise absorbing the cost of accidents as an item of its overheads, and the activity can be tolerated only on condition that the enterprise indemnifies all who suffer, whether it is carried on carefully or not, since the enterprise alone has the resources to discover and guard against the hazard. The liability is not subject to the exceptions that qualify strict liability under Rylands v. Fletcher. The Court added that compensation must be correlated to the magnitude and capacity of the enterprise, so that it has a deterrent effect: the larger and more prosperous the enterprise, the greater the compensation payable.

Why it matters here. It is the point at which Indian law accepted that a technology's characteristic harm cannot be handled by a rule designed for careless individuals. Three moves in it are worth learning separately: fault is removed; the exceptions are removed; and the measure of compensation is tied to the size of the defendant rather than only to the loss of the plaintiff, which converts a compensatory rule into a deterrent one.

The failure: Bhopal

Union Carbide Corporation v. Union of India, AIR 1992 SC 248, (1991) 4 SCC 584, decided on 3 October 1991 by five judges.

Facts. Proceedings arising out of the Bhopal gas leak of December 1984. The Union of India sued Union Carbide Corporation in the United States District Court for the Southern District of New York; those proceedings ended when the company's plea of forum non conveniens, that another court was the more appropriate forum, was upheld. The Union of India then sued in the District Court at Bhopal, and the claim was settled by orders of the Supreme Court of 14 and 15 February 1989. The settlement was challenged by review and by writ petitions, on grounds including that no fairness hearing had been afforded to the victims, that there was no re-opener clause if the fund proved inadequate, that the medical documentation and categorisation of injuries were unreliable, and that the quashing of pending criminal proceedings as part of the settlement was beyond the Court's power and opposed to public policy.

Held. The Court examined the absence of a fairness hearing and of a re-opener clause, the adequacy of the fund and the position should it prove insufficient, and whether the mere possibility that medical documentation and categorisation were faulty could justify setting the settlement aside. On its own powers it held that the inherent jurisdiction under articles 136 and 142 to withdraw, transfer and finally dispose of the suits and the pending criminal proceedings was not taken away by article 139A, and it construed the words "cause or matter" in article 142. It restored the criminal proceedings that the settlement had quashed. It also stated that an award of compensation is to be proportionate to the economic superiority of the offender.

Why it matters here. It is the case in which the system met the largest industrial disaster of the technological age and found its own machinery unequal to it. Note what the substantive rule could not do. M.C. Mehta had been decided two years before the settlement and states a rule of absolute liability with compensation geared to the size of the enterprise. What defeated the victims was not the absence of a rule. It was forum, which moved the claim between two legal systems and cost years; representation, since the persons whose claims were settled were not heard before the settlement; proof, since the categorisation of injury on which every individual award depended was contested; and time. A candidate who can say that the failure was procedural rather than doctrinal has understood the pair of cases.

Worked example

A chemical plant on the edge of a town has an emission that damages crops on adjoining land over several years, and is later associated with a raised incidence of a respiratory illness among residents. The company shows that it operated within the limits of its consent to operate and that its equipment was maintained to the manufacturer's specification.

Does the compliance defence work? No, on M.C. Mehta. Absolute liability is not displaced by proof of reasonable care, and operating within a licence is a form of that proof. Compliance is relevant to prosecution and to the regulator; it does not answer a claim in absolute liability.

Who may sue and for what? The crop damage is a straightforward loss and is provable. The illness is where the difficulty lies: latency, multiple causes, and a population that is exposed rather than a claimant who is injured. This is the point at which the substantive rule is intact and the evidence is not, which is precisely the Bhopal problem in miniature.

What does the measure of compensation depend on? Under M.C. Mehta, on the magnitude and capacity of the enterprise as well as the harm, so that the award deters. That is a rule most students can state and few apply, and applying it is what makes an answer look like law rather than recitation.

Which rights are engaged beyond tort? Article 21 for the residents; article 7(b) for those working inside; article 12(2)(b) for the State's duty as to environmental and industrial hygiene; and paragraph 8 of the 1975 Declaration for the State's duty to take legislative measures rather than to wait for litigation. The precautionary principle in [The Precautionary Principle] governs what should have happened before the plant was permitted at all.

What it does NOT mean

Absolute liability is not strict liability. They are routinely confused and the difference is the whole point of M.C. Mehta: strict liability under Rylands v. Fletcher comes with exceptions, and the Court held that its rule is not subject to them.

It does not apply to every industry. The rule is framed for an enterprise engaged in a hazardous or inherently dangerous activity, and that characterisation is the first thing an answer must establish rather than assume.

A liability rule is not a safety system. It operates after the harm. Article 12(2)(b) and paragraph 8 of the 1975 Declaration are about prevention, and the practical protection of a population comes from the consent conditions, the inspection and the siting decision, not from the prospect of a suit.

And compensation is not restoration. Bhopal is the standing demonstration. A fund, however calculated, does not undo an exposure, and the framework's preference for prevention follows from that.

Distinctions

Quick revision

• ICESCR article 12(2)(b): steps for the improvement of all aspects of environmental and industrial hygiene. Article 7(b): safe and healthy working conditions. 1975 Declaration, paragraphs 6 and 8: protect from harmful effects of misuse; effective measures, including legislative measures.
• Constitution: article 21, article 48A (State to protect and improve the environment), article 51A(g) (fundamental duty of the citizen).
• M.C. Mehta (1986), five judges, Bhagwati CJ: a hazardous or inherently dangerous enterprise owes an absolute and non-delegable duty to the community; absolutely liable for resulting harm; no answer that all reasonable care was taken; not subject to the Rylands exceptions; compensation correlated to the magnitude and capacity of the enterprise to deter.
• Union Carbide (1991): the Bhopal settlement reviewed; fairness hearing and re-opener clause considered; articles 136 and 142 not taken away by article 139A; criminal proceedings restored; compensation proportionate to the economic superiority of the offender.
• The pair teaches the lesson: the substantive rule was strong and the outcome was still poor, because the failures were forum, representation, proof and time. Doctrine is worth what the procedure can deliver.

Test yourself

1. State the rule in M.C. Mehta and identify the three ways it departs from Rylands v. Fletcher. An enterprise engaged in a hazardous or inherently dangerous industry which poses a potential threat to the health and safety of those working in the factory and residing in the surrounding areas owes an absolute and non-delegable duty to the community to ensure that no harm results; it must conduct the activity with the highest standards of safety; and if harm results it is absolutely liable to compensate, it being no answer that all reasonable care was taken and that the harm occurred without negligence. It departs from Rylands v. Fletcher in that the liability is not subject to any of the exceptions that qualify strict liability; in that its basis is engagement in a hazardous activity rather than a non-natural user of land and the escape of a dangerous thing; and in that the measure of compensation is correlated to the magnitude and capacity of the enterprise so as to have a deterrent effect, rather than being confined to the loss suffered.

2. Why does a fault-based rule fail the victim of an industrial escape? Because both of its requirements are unattainable for that victim. It requires proof of want of care, and the knowledge of the process, the maintenance and the training lies entirely inside the enterprise, so the person harmed cannot identify what failed. And it excuses the defendant where care was taken, which leaves the loss with people who live nearby and had no part in choosing the hazard, even though a hazardous process remains hazardous when properly run. The Court's answer in M.C. Mehta was to change the rule rather than to try to improve the plaintiff's evidence, reasoning that the enterprise alone has the resources to discover and guard against the hazard and that the cost of accidents should be an item of its overheads.

3. What was decided in Union Carbide, and why is it said that the failure there was procedural rather than doctrinal? The Court reviewed the settlement of the Bhopal claims recorded by its orders of 14 and 15 February 1989, considering the absence of a fairness hearing and of a re-opener clause, the adequacy of the fund, and whether the possibility of faulty medical documentation and categorisation could justify setting the settlement aside; it held that its inherent powers under articles 136 and 142 were not taken away by article 139A, construed 'cause or matter' in article 142, restored the criminal proceedings that had been quashed, and stated that compensation should be proportionate to the economic superiority of the offender. The failure is called procedural because the substantive rule was already available: M.C. Mehta had been decided in December 1986, two years before the settlement, and states absolute liability with a deterrent measure of compensation. What defeated the victims was the forum contest that moved the claim between two legal systems, the settlement of their claims without their being heard, the contested categorisation of injury on which each individual award depended, and the passage of time.

4. A company shows it operated within the terms of its consent to operate and maintained its plant to specification. Is that a defence to a claim in absolute liability? No. Absolute liability is not displaced by proof that reasonable care was taken, and compliance with a licence and with a maintenance specification is a form of that proof. M.C. Mehta states in terms that it should be no answer to the enterprise to say that it had taken all reasonable care and that the harm occurred without any negligence on its part, and that the liability is not subject to the exceptions applicable under Rylands v. Fletcher, among which statutory authority is one. Compliance remains relevant to a prosecution and to the regulator's decisions, but it does not answer the civil claim.

5. Why does the framework prefer prevention to compensation in this area, and which provisions carry that preference? Because the characteristic harms of industrial technology cannot be undone: an exposure cannot be reversed, latency means the injury may appear years later, and the harm is spread across a population in increments too small to litigate individually, so that even a generous fund reaches some people late and others not at all. Bhopal is the standing demonstration. The preference is carried by article 12(2)(b) of the Covenant, which requires steps for the improvement of all aspects of environmental and industrial hygiene; by article 7(b), which requires safe and healthy working conditions; and by paragraph 8 of the 1975 Declaration, which requires effective measures including legislative measures to prevent and preclude the use of scientific and technological achievements to the detriment of human rights, prevention being the operative word in each.

That final phrase is the most useful sentence in the instrument for this topic. It does not merely prohibit the conduct; it characterises the use of science for those ends as a perversion of what science is for, which is a normative claim about the discipline itself and connects directly to the responsibilities discussed in [The Scientist: Rights and Responsibilities].

Paragraph 8 requires effective measures, including legislative measures, to prevent and preclude the utilisation of scientific and technological achievements to the detriment of human rights and the dignity of the human person.

The rights engaged

ICCPR article 6(1): every human being has the inherent right to life, this right shall be protected by law, and no one shall be arbitrarily deprived of his life. The word arbitrarily is what carries the analysis: the article does not prohibit every deprivation of life, and the question in any given case is whether the deprivation was arbitrary.

ICCPR article 7: no torture, and no cruel, inhuman or degrading treatment or punishment. Weapons that injure in particular ways, and weapons whose effects continue long after their use, are argued under this article as well as under article 6.

UDHR article 3: everyone has the right to life, liberty and security of person.

General Comment 25, paragraph 81, brings this within the right to science expressly. The Committee identifies among the most acute risks related to science and technology the development of dangerous technologies, such as autonomous weapons based on artificial intelligence, and the threat of weapons of mass destruction, especially nuclear weapons, notes that these risks are transnational and cannot be addressed without robust international co-operation, and requires States to promote multilateral agreements to prevent them or mitigate their effects.

That paragraph is worth citing precisely because it appears in a general comment about the right to enjoy the benefits of scientific progress. It establishes that the arms question is not outside this syllabus but inside the very right the syllabus is built on.

Where this paper stops

A boundary has to be stated, because the material invites an answer that belongs on a different paper.

International humanitarian law, the law of armed conflict, governs the conduct of hostilities and the lawfulness of particular weapons through the principles of distinction, proportionality in attack, precaution, and the prohibition of superfluous injury and unnecessary suffering. It is a distinct body of law with its own instruments and its own institutions, and in this LL.M. it is the subject of the separate paper on international humanitarian law and refugee law.

Human rights law, which is this paper, addresses the same facts from the position of the individual and continues to apply, on the accepted view, in situations of armed conflict alongside humanitarian law. What this paper asks is narrower and quite specific: what does the development, possession and use of a technology whose purpose is destruction do to the rights of the person, and what does it do to the obligations of a State that is developing it?

A candidate who writes an answer about distinction and proportionality in attack has written a good answer to the wrong question. The right frame here is the 1975 Declaration, the right to life, and General Comment 25's treatment of transnational risk.

The four categories, and what distinguishes them

An answer is stronger for treating these separately rather than as a single category of frightening things.

The third row is the one that belongs most naturally to this paper, because it is an instance of delegation, the fifth of the changes identified in [How a Technology Becomes a Rights Question], applied to the gravest decision there is. The argument against it is not that a machine will be less accurate than a person. It is that accountability requires an agent who chose, and there is none.

Worked example

A State's defence laboratory develops a system that identifies and engages targets without a human operator confirming each engagement. The system is more accurate in trials than a human crew. The State argues that fewer civilians will be harmed, so the technology serves the right to life rather than threatening it.

Take the argument seriously. It is not frivolous, and the strongest answers engage with it. If the claim about accuracy is true, then a refusal to deploy costs lives, and that is a rights consequence too.

Then locate the flaw. The right to life is protected by article 6 in two ways: substantively, by prohibiting arbitrary deprivation, and procedurally, by requiring that a deprivation be capable of being investigated and accounted for. Accuracy answers the first and not the second. Where a machine selects and engages, there is no person who formed the judgment, so there is no judgment to review, no reasons to test, and nobody whose responsibility can be established. A right that cannot be vindicated after its breach is diminished however low the breach rate.

Add the distributional question. Systems trained on data perform unevenly across the populations represented in that data, so a claim of superior average accuracy is compatible with markedly worse performance for some group. That is the argument of [Equality and the Digital Divide] transposed to the use of force.

And the obligation on this State. Paragraph 8 of the 1975 Declaration requires effective measures including legislative ones; General Comment 25 requires the promotion of multilateral agreements against exactly this class of risk. The obligation is not satisfied by an internal review conducted by the department that built the system.

What it does NOT mean

It does not mean scientists may not work on defence. Neither the Declaration nor the Covenant prohibits it. Paragraph 4 prohibits the use of scientific achievements for specified purposes, principally aggressive war, and the responsibility of the individual researcher operates through the freedom to question the ethical value of a project and to withdraw, which is a freedom rather than a prohibition.

It does not mean the right to life is absolute in the sense of prohibiting all killing. Article 6 prohibits arbitrary deprivation, and that word does the work.

It does not mean this paper answers the lawfulness of a particular weapon. That is a humanitarian law question, and the honest answer here says so and then addresses the question this paper does ask.

And it does not mean a technology is exempt because it has civil uses. Almost all of them do, which is the dual use problem; the response is control of the use and of transfer, not a classification of the underlying knowledge as forbidden.

Quick revision

• Weapons are the limiting case: destructive capacity is the object rather than a side effect, so the neutrality argument is unavailable. Harms are irreversible and cannot be confined, so the framework is preventive.
• 1975 Declaration, paragraph 4: States shall refrain from using scientific and technological achievements to violate sovereignty and territorial integrity, interfere in internal affairs, wage aggressive wars, suppress national liberation movements or pursue racial discrimination; such acts are a flagrant violation of the Charter and an inadmissible distortion of the purposes that should guide scientific development. Paragraphs 1, 2 and 8 supply peace and security, prevention of State misuse, and legislative measures.
• ICCPR article 6(1): inherent right to life, protected by law, no arbitrary deprivation. Article 7: cruel, inhuman or degrading treatment. UDHR article 3.
• General Comment 25, paragraph 81: names autonomous weapons based on artificial intelligence and weapons of mass destruction, especially nuclear weapons, as acute transnational risks, and requires States to promote multilateral agreements.
• Four categories: nuclear, chemical and biological, autonomous, and cyber or surveillance capabilities in conflict.
• The autonomous weapons objection is about accountability, not accuracy: no agent decided, so nothing can be reviewed and nobody can be held responsible.
• The boundary: distinction, proportionality in attack and superfluous injury belong to international humanitarian law, which is a different paper in this LL.M.

Test yourself

1. Set out paragraph 4 of the 1975 Declaration and explain why its closing words matter. It requires all States to refrain from any acts involving the use of scientific and technological achievements for the purposes of violating the sovereignty and territorial integrity of other States, interfering in their internal affairs, waging aggressive wars, suppressing national liberation movements, or pursuing a policy of racial discrimination, and states that such acts are not only a flagrant violation of the Charter of the United Nations and of the principles of international law but constitute an inadmissible distortion of the purposes that should guide scientific and technological developments for the benefit of mankind. The closing words matter because they go beyond prohibition to characterisation: they assert that science has purposes, that those purposes are the benefit of mankind, and that the military use in question is a perversion of the enterprise itself, which is the normative foundation for the responsibilities of the individual researcher.

2. Why are weapons described as the limiting case of this subject? Because the feature that makes every other topic difficult is absent. Elsewhere the harm is a by-product of an application intended to do something else, which is what makes the dual use analysis necessary and what supports the argument that a technology is neutral between uses. In a weapon the destructive capacity is the specification rather than a side effect, so there is no second direction to point to and the neutrality argument is unavailable. The other distinguishing features are irreversibility and scale: the harms of the most serious weapons cannot be undone and cannot be confined in space, in time or to the parties to a conflict, which is why the legal response is preventive rather than remedial.

3. What does General Comment 25 say about weapons, and why is it significant that it says it? Paragraph 81 identifies among the most acute risks to the world related to science and technology the development of dangerous technologies such as autonomous weapons based on artificial intelligence, and the threat of weapons of mass destruction, especially nuclear weapons; states that such risks are transnational and cannot be adequately addressed without robust international co-operation; and requires States to promote multilateral agreements to prevent them from materialising or to mitigate their effects. It is significant because the Comment is an authoritative interpretation of the right to participate in and enjoy the benefits of scientific progress under article 15 of the ICESCR. Its treatment of weapons therefore places the arms question inside the right this syllabus is built on, rather than leaving it to the law of armed conflict, and supplies a candidate with a Covenant-based route into the topic.

4. State the objection to autonomous weapons in its strongest form, and explain why a claim of superior accuracy does not answer it. The objection is that the decision to take a life is delegated to a system, so no person formed the judgment. Article 6 protects life both substantively, by prohibiting arbitrary deprivation, and procedurally, by requiring that a deprivation be capable of investigation and accounting. A claim of superior accuracy addresses only the substantive limb, by asserting a lower rate of wrongful killing. It leaves the procedural limb untouched, because where a machine selected and engaged there is no judgment to review, no reasons that can be tested, and no agent whose responsibility can be established, so a person wrongly killed has no remedy and the State has no answerable official. A further point is distributional: a system's accuracy is an average across the populations represented in its training data, so superior overall performance is consistent with markedly worse performance for a particular group.

5. Where does this paper stop and international humanitarian law begin, and how should a candidate handle the boundary? International humanitarian law governs the conduct of hostilities and the lawfulness of particular weapons, through the principles of distinction, proportionality in attack, precaution, and the prohibition of weapons causing superfluous injury or unnecessary suffering, and in this LL.M. it is the subject of a separate paper. This paper approaches the same facts from the position of the individual and asks what the development, possession and use of a destructive technology does to the rights of the person and to the obligations of the State that develops it, using the 1975 Declaration, the right to life in article 6, and General Comment 25's treatment of transnational risk. A candidate should state the boundary expressly, note that human rights law continues to apply in armed conflict alongside humanitarian law, and then answer the question this paper asks, because an answer written entirely in the vocabulary of distinction and proportionality in attack answers a question that was not set.

Article 2(a): everyone has a right to respect for their dignity and for their rights regardless of their genetic characteristics. Article 2(b): that dignity makes it imperative not to reduce individuals to their genetic characteristics and to respect their uniqueness and diversity.

Article 2(b) is the operative core of the whole instrument and the sentence to learn. The wrong it names is reduction: treating a person as their genotype. That is precisely what an insurer does in pricing a policy on a marker, what an employer does in screening, and what a State does in classifying a population.

Article 3: the genome evolves and is subject to mutations, and contains potentialities expressed differently according to each individual's natural and social environment, including health, living conditions, nutrition and education. This is a scientific proposition placed in a legal instrument in order to defeat determinism: the genome is not destiny, and a legal consequence attached to a marker rests on a claim the science does not support.

Article 4: the human genome in its natural state shall not give rise to financial gains. Note the qualification. It does not prohibit patents on inventions derived from genetic knowledge; it addresses the genome in its natural state, and the line between the two is the whole of the argument in [Access to Medicines and Patents].

The rights of the person

Article 5 governs intervention. Research, treatment or diagnosis affecting an individual's genome may be undertaken only after rigorous and prior assessment of the potential risks and benefits and in accordance with national law; in all cases the prior, free and informed consent of the person must be obtained, and where they cannot consent, consent or authorisation is obtained as prescribed by law, guided by their best interest; research protocols must be submitted for prior review; and where a person lacks capacity, research affecting their genome may be carried out only for their direct health benefit, research without expected direct benefit being permissible only by way of exception, with the utmost restraint, at minimal risk and burden, where it is intended to benefit others of the same age category or with the same genetic condition.

Article 5(c) states a right that has no real counterpart elsewhere in human rights law: the right of each individual to decide whether or not to be informed of the results of a genetic examination and of the consequences. It is a right not to know, and it exists because the information is predictive and unalterable, so that receiving it can injure a person who can do nothing with it.

Article 6: no one shall be subjected to discrimination based on genetic characteristics that is intended to infringe or has the effect of infringing human rights, fundamental freedoms and human dignity. Note "or has the effect of", which reaches indirect discrimination.

Article 7: genetic data associated with an identifiable person and stored or processed for research or any other purpose must be held confidential in the conditions set by law.

Article 8: every individual has the right to just reparation for damage sustained as a direct and determining result of an intervention affecting their genome.

Article 9: limitations on the principles of consent and confidentiality may be prescribed only by law, for compelling reasons, within the bounds of public international law and international human rights law. It is the limitation clause, and it is narrow.

Research and its limits

Article 10: no research or research application concerning the human genome, in particular in biology, genetics and medicine, should prevail over respect for the human rights, fundamental freedoms and human dignity of individuals or of groups.

Article 11: practices contrary to human dignity, such as reproductive cloning of human beings, shall not be permitted, and States and competent international organisations are invited to co-operate in identifying such practices and taking the necessary measures.

Article 11 is the clearest instance in the whole of this paper of dignity operating as an operative legal standard rather than as an aspiration. It does not say cloning is unsafe, or that it is unpopular. It says the class of prohibited practices is defined by contrariety to dignity, and gives reproductive cloning as an example of the class rather than as its whole content. That drafting is deliberate: it leaves the category open, so that practices not yet invented can fall within it.

Worked example

An insurer proposes to require applicants for life cover to disclose the results of any genetic test they have taken, and to price the policy accordingly. It argues that it does not require anyone to take a test, that it asks only for information the applicant already has, and that non-disclosure of known risk is a familiar ground for avoiding a contract.

Article 6 is engaged directly. Discrimination based on genetic characteristics that has the effect of infringing rights and dignity is prohibited, and the words "has the effect of" defeat the answer that no discrimination was intended.

Article 2(b) names the wrong precisely. Pricing on a marker reduces the individual to their genetic characteristics, which is what article 2(b) says dignity makes it imperative not to do.

Article 3 answers the actuarial argument. The genome contains potentialities expressed differently according to the individual's environment, health, living conditions, nutrition and education. A marker is a probability across a population, not a fact about this applicant, so the insurer's inference is weaker than it presents itself as being.

Article 5(c) shows the second harm. A rule of this kind destroys the right not to know, because a person who wishes to remain uninformed must now choose between the knowledge and the insurance. The pressure operates without anyone compelling a test, which is why the insurer's first argument does not answer the objection.

And the harm is not confined to the applicant. A test result discloses about siblings and children who are not parties to the contract. That is the shared property of genetic information doing legal work, and it is the reason this cannot be handled as an ordinary question of contractual disclosure.

What it does NOT mean

The genome is not the common property of humanity. Article 1 says heritage "in a symbolic sense", and the qualification is there precisely to prevent the property reading.

Article 4 is not a prohibition on biotechnology patents. It addresses the genome in its natural state. What may be patented, and on what conditions, is a question for patent law, and in India for section 3(d) of the Patents Act, taken up in [Access to Medicines and Patents].

Genetic non-discrimination is not a bar on all use of genetic information. Article 9 permits limitations on consent and confidentiality by law, for compelling reasons, within the bounds of international human rights law. Clinical use with consent is the ordinary case and is not touched.

And dignity is not a trump card. It is doing specific work here, identified in article 2(b) as the prohibition on reducing a person to their genetic characteristics. An answer that invokes dignity without saying what it forbids in the case at hand has used the word rather than the concept.

Quick revision

• Genetic information is shared, predictive, immutable and constitutive, which is why privacy and equality alone do not answer it and why the instruments use dignity.
• UNESCO 1997, adopted 11 November 1997, endorsed by GA resolution 53/152 of 9 December 1998. Not a treaty.
• Article 1: the genome underlies the unity of the human family; in a symbolic sense the heritage of humanity. Article 2(a): dignity and rights regardless of genetic characteristics. Article 2(b): imperative not to reduce individuals to their genetic characteristics. Article 3: potentialities expressed differently according to environment, so the genome is not destiny. Article 4: the genome in its natural state shall not give rise to financial gains.
• Article 5: prior risk assessment, prior free and informed consent, prior review of protocols, and for a person without capacity, direct health benefit save by exception. Article 5(c): the right not to know.
• Article 6: no discrimination on genetic characteristics intended to or having the effect of infringing rights and dignity. Article 7: confidentiality. Article 8: just reparation. Article 9: limitations on consent and confidentiality only by law, for compelling reasons.
• Article 10: no research application shall prevail over rights, freedoms and dignity. Article 11: practices contrary to human dignity, such as reproductive cloning, shall not be permitted, the class being left open.

Test yourself

1. What four properties of genetic information make it a distinct legal problem? It is shared, because a person's genome is substantially that of their parents, siblings and children, so a disclosure by one person discloses about others who have not consented. It is predictive, speaking to a future that has not occurred and which the person cannot disprove. It is immutable, so a harm caused by disclosure cannot be undone by changing anything. And it is constitutive, being in one sense the specification of the person rather than merely information about them. Together these defeat an analysis built only on privacy and consent, since consent by one person cannot cover the others implicated and no remedy can restore the position.

2. Set out article 2(b) of the 1997 Declaration and explain why it is the operative core of the instrument. It provides that the dignity of everyone, regardless of their genetic characteristics, makes it imperative not to reduce individuals to their genetic characteristics and to respect their uniqueness and diversity. It is the operative core because it names the specific wrong that genetic technology makes possible: reduction, that is, the treatment of a person as their genotype. That is what an insurer does in pricing on a marker, an employer in screening, and a State in classifying a population, so the article converts dignity from a preambular sentiment into a test that can be applied to a practice.

3. What is the right not to know, where is it found, and why does it exist? It is the right of each individual to decide whether or not to be informed of the results of a genetic examination and of the resulting consequences, stated in article 5(c) of the 1997 Declaration. It exists because genetic information is predictive and unalterable: a person may learn that they will probably develop a condition they can do nothing to prevent, and the knowledge injures without enabling any action. It has no close counterpart elsewhere in human rights law, where the usual concern is to secure access to information rather than to protect a person from receiving it.

4. Why is article 11 significant for the argument that dignity is an operative standard rather than rhetoric? Because it defines a prohibited class by reference to dignity alone. It provides that practices contrary to human dignity, such as reproductive cloning of human beings, shall not be permitted, and invites States and competent international organisations to co-operate in identifying such practices and in taking the measures necessary. It does not rest the prohibition on safety, on efficacy or on public opinion, and it gives reproductive cloning as an example of the class rather than as its whole content, which leaves the category open so that practices not yet invented may fall within it. Dignity is therefore performing the function of a legal standard, supplying the criterion by which a practice is judged.

5. An employer requires applicants to disclose existing genetic test results but requires nobody to be tested. Which provisions answer it? Article 6 is engaged directly, since it prohibits discrimination based on genetic characteristics that is intended to infringe or has the effect of infringing human rights, fundamental freedoms and human dignity, and the words "has the effect of" defeat the argument that no discrimination was intended. Article 2(b) identifies the wrong as reduction of the individual to their genetic characteristics. Article 3 answers the argument from actuarial or occupational rationality, because the genome contains potentialities expressed differently according to the individual's environment, health, living conditions, nutrition and education, so a marker is a probability across a population rather than a fact about the applicant. Article 5(c) shows the second harm, since a requirement of this kind destroys the right not to know by forcing a person who wishes to remain uninformed to choose between the knowledge and the employment. And article 7 is engaged because genetic data associated with an identifiable person must be held confidential in the conditions set by law, which a general disclosure requirement circumvents.

ICCPR article 26 guarantees equality before the law and the equal protection of the law without discrimination, and prohibits discrimination on any ground including "other status". ICESCR article 2(2) does the same for Covenant rights. Both reach measures that are neutral in form and unequal in effect.

ICCPR article 17 protects against arbitrary or unlawful interference with privacy, which is engaged by the collection and combination of the data such systems require, as set out in [Privacy and Surveillance].

ICESCR article 7(c) guarantees equal opportunity for promotion in employment subject to no considerations other than seniority and competence. It is an unexpectedly exact provision for automated assessment: a system that ranks on proxies correlated with past outcomes introduces considerations other than seniority and competence, which is what the limb forbids.

Constitution of India, article 14, guarantees equality before the law and the equal protection of the laws, and the settled doctrine that State action must not be arbitrary is directly engaged by a decision for which no reason can be given. Article 21 supplies the requirement of a fair procedure where life or personal liberty is affected.

General Comment 25, paragraph 22, supplies the regulatory route: limitations on the applications of science may be used to guarantee the safety and quality of products used by persons, and human rights impact assessments might be necessary to protect persons against risky applications. That sentence is the treaty-body basis for requiring an assessment before deployment, and it is the most directly useful citation available on this topic.

The four problems

Keeping these separate is what distinguishes a legal answer from a general essay.

1. Opacity. Where a system's output cannot be explained, the person affected cannot identify the error, and the reviewing body cannot test the reasoning. The right injured is procedural: not the right to a favourable decision, but the right to know why.
2. Bias. A system trained on records of past decisions learns the distribution in those records, including the effects of past discrimination. It then applies that distribution to new cases with the appearance of objectivity. The wrong is indirect discrimination, and the appearance of neutrality makes it harder to detect than the practice it automates.
3. Accountability. Where the decision is delegated, responsibility disperses among the body that deployed the system, the vendor that built it, and the people who assembled the data. Each can point to another, and a rights framework that cannot identify a duty bearer cannot deliver a remedy.
4. Scale. A human decision-maker who is wrong is wrong in a limited number of cases and inconsistently. A system that is wrong is wrong in every case of that kind, identically, until somebody notices. The consistency that is presented as the technology's virtue is what converts an individual error into a population-level harm.

Worked example

A State uses an automated system to decide which applications for a disability pension are examined by an officer and which are refused on the papers. The system was trained on the outcomes of previous applications. It is more consistent than the officers it replaced, and processing times have fallen.

State what is good about it. Faster decisions and consistency are real benefits to applicants, and article 15(1)(b) means the applicants have an interest in the State using what science has made available. An answer that treats the system as self-evidently objectionable has not applied the framework.

Opacity. Can a refused applicant be told why? If the answer is that the system scored the application below a threshold, no reason has been given, and the requirement of a fair procedure under article 21 and of non-arbitrariness under article 14 of the Constitution is not met.

Bias. The training data are past outcomes, so if a category of applicant was historically refused for reasons unconnected with entitlement, the system will reproduce that refusal rate and present it as a finding. Testing for this requires the outcomes to be examined by group, which requires that the data be collected and published; a State that does not measure cannot claim compliance with article 2(2).

Accountability. Who decided? If the department says the system did, and the vendor says it only supplied a tool, then no one is answerable, and the applicant's remedy is against a body that disclaims the decision.

Scale. An officer with a mistaken view of a condition affects the applications they see. A system with the same mistaken view affects every application of that kind in the State, immediately.

What the framework requires. Not prohibition. A human rights impact assessment before deployment, in the terms of paragraph 22; reasons capable of being given to the applicant; outcomes monitored by group; a real route to human reconsideration; and an identified body that owns the decision. Those five are the answer, and they are all available from provisions that already exist.

What it does NOT mean

It does not mean automated decisions are unlawful. No provision in the base prohibits them, and consistency and speed are genuine benefits. The framework regulates the conditions of use.

It does not mean accuracy is the test. A system that is more accurate on average may still fail, because the guarantees are procedural and distributive as well as substantive. This is the same point made about autonomous weapons in [Weapons, Warfare and the Right to Life].

A human in the loop is not automatically an answer. Where the person is presented with a score and has neither the time nor the information to go behind it, the human is a formality and the decision remains the machine's. The question is whether the human can and does exercise judgment.

And there is no treaty here. A candidate should not attribute to the ICCPR, the ICESCR or the UNESCO declarations any provision about artificial intelligence. What those instruments contain are general guarantees that apply, and General Comment 25's two references. Saying so accurately is worth more than a confident citation to something that does not exist.

Quick revision

• Later than the 2016-17 syllabus, and inside topic 2 because that topic is drafted openly; General Comment 25 (2020) refers to artificial intelligence, naming autonomous weapons based on artificial intelligence among acute transnational risks at paragraph 81.
• The principal effect is delegation, the fifth of the five changes; the others, scale, cost, permanence and invisibility, are all present too.
• No treaty on the subject. The provisions that apply: ICCPR article 14 (fair hearing), article 26 and ICESCR article 2(2) (equality, reaching indirect discrimination), ICCPR article 17 (privacy), ICESCR article 7(c) (promotion on no considerations other than seniority and competence), Constitution articles 14 and 21.
• General Comment 25, paragraph 22: limitations on applications to guarantee safety and quality, and human rights impact assessments might be necessary against risky applications. The most useful citation on this topic.
• Four problems: opacity, bias, accountability, scale.
• Five requirements: impact assessment before deployment, reasons that can be given, outcomes monitored by group, a real route to human reconsideration, and an identified body that owns the decision.

Test yourself

1. On what basis is a technology later than the syllabus properly examinable on this paper? Because topic 2 asks about the implications of the development of science and technology on human rights without naming any technology, and was printed in 2016-17 when the University could not have known which technologies a candidate would face. The topic is therefore drafted openly by design. The point is confirmed by General Comment 25, adopted in 2020, which is the authoritative interpretation of the right on which the paper rests and which itself refers to artificial intelligence, including autonomous weapons based on artificial intelligence among the most acute transnational risks at paragraph 81.

2. State the accountability problem and explain why it is distinct from the problem of bias. The accountability problem is that where a decision is delegated to a system, responsibility disperses among the body that deployed it, the vendor that built it and those who assembled the training data, so that each can point to another and no duty bearer can be identified. It is distinct from bias because it would remain even if the system were perfectly unbiased: a correct decision for which no one is answerable still deprives the affected person of reasons, of a route of challenge and of a respondent. Bias concerns the content of the decisions, accountability concerns the structure within which they are made, and a remedy for one does not supply a remedy for the other.

3. Why is ICESCR article 7(c) an unexpectedly precise provision for automated assessment in employment? Because it guarantees equal opportunity for everyone to be promoted in his employment to an appropriate higher level, subject to no considerations other than those of seniority and competence. An automated ranking system does not assess competence directly; it scores on features correlated with the outcomes recorded in its training data, which may include the length of a gap in employment, the institution attended or patterns associated with a group. Each such feature is a consideration other than seniority and competence, so the provision is breached by the method of assessment itself, independently of whether the outcome can be shown to be discriminatory on a prohibited ground.

4. A department says a refused applicant was scored below the threshold by its system. Has a reason been given? No. A score is a restatement of the outcome, not a reason for it: it tells the applicant what the system concluded and nothing about why, so it identifies no fact that might be wrong, no criterion that might be misapplied and nothing that can be contested. That defeats the guarantee of a fair hearing under article 14 of the ICCPR where the decision is one to which the article applies, and in India it engages both the requirement of a fair procedure under article 21 and the doctrine under article 14 that State action must not be arbitrary, a decision for which no reason can be given being the paradigm of arbitrariness. The practical requirement is that the department be able to state the factors that determined the outcome in this case in terms the applicant can address.

5. Is a human in the loop a sufficient safeguard? Give reasons. Not by itself. The safeguard is meant to preserve an agent who forms a judgment, and it does so only if that person has the information, the time and the authority to reach a different conclusion. Where an officer is presented with a score and a queue, lacks access to the factors behind the score, and is measured on throughput, the human presence is a formality and the effective decision remains the system's, so opacity and accountability are untouched while the appearance of human decision-making defeats scrutiny. The question is therefore not whether a person is nominally in the loop but whether they can and do exercise independent judgment, which requires that reasons be available to them, that departure from the recommendation be permitted and recorded, and that rates of departure be monitored.

Held, on limits. In the case of pregnant women there is a compelling State interest in protecting the life of the prospective child, so termination is permitted only where the conditions of the applicable statute are fulfilled, and the provisions of the MTP Act 1971 may be viewed as reasonable restrictions on the exercise of reproductive choices.

Why it matters here. It establishes the right in the harder direction. The choice the Court protected was the choice to continue a pregnancy, asserted by a woman whose capacity was in question, against a decision that others believed was in her interests, and supported by the Expert Body rather than opposed by it. That is the clearest Indian statement of the proposition running through this whole module: a technology or procedure may not be applied to a person because somebody else considers it good for them. It is also the case that answers the misreading, common in student work, that reproductive rights mean a right to terminate: the Court says expressly that the right runs both ways.

CEDAW

Article 16(1)(e) requires States Parties to ensure, on a basis of equality of men and women, the same rights to decide freely and responsibly on the number and spacing of their children and to have access to the information, education and means to enable them to exercise these rights.

Two features are worth noting. The right is to decide, which locates it in autonomy rather than in health provision. And it expressly includes access to the means, so a State that recognises the choice while making the means unavailable has not complied.

The four statutes, and what each answers

Each is a legislative response to a capability, and each is taken in full in Module II. What matters here is the pattern.

The PCPNDT Act repays attention because it is the clearest instance in Indian law of the dual use analysis from [Dual Use: the Same Technology Both Ways] being resolved by statute. The same ultrasound serves the health of the pregnancy and enables sex selection. Parliament did not ban the machine; it regulated the use, the communication of the result and the registration of the clinic. That is the model response to a dual use technology, and a candidate who can say so has understood both chapters.

Worked example

A married woman attends a fertility clinic. The consent form for the procedure requires the signature of her husband. She wishes to proceed; he refuses. Separately, the clinic offers to disclose the sex of the foetus once a pregnancy is established, describing it as information the family is entitled to.

The consent requirement. Suchita Srivastava locates reproductive choice in the woman's personal liberty under article 21, with privacy, dignity and bodily integrity as the crucial considerations, and CEDAW article 16(1)(e) states the right to decide on a basis of equality between men and women. A requirement that transfers the decision about her body to another person defeats both. Whether a particular statute or rule imposes such a requirement is a question for Module II; the constitutional analysis is the one above.

The offer to disclose. This is an offence, not a service. The PCPNDT Act prohibits the communication of the sex of the foetus, and the offer is the precise conduct the Act exists to prevent. Note the structure of the wrong: the diagnostic technology is lawful and valuable, and it is the use and the communication that are prohibited.

Whose autonomy is at stake in the second point? Not only the woman's. The harm the Act addresses is committed against the foetus and, at scale, against the sex ratio of a population, and the pressure to seek the information usually comes from the family. This is a case where a restriction on information protects rather than diminishes the woman's position, which is worth saying because it cuts against the assumption that autonomy always favours disclosure.

What it does NOT mean

Reproductive rights are not a right to terminate. Suchita Srivastava says in terms that the choices may be exercised to procreate as well as to abstain from procreating, and the case itself protected the choice to continue.

The right is not absolute. The Court recognised a compelling State interest in protecting the life of the prospective child and treated the MTP Act's conditions as reasonable restrictions.

Incapacity does not transfer the choice. The woman in the case was found to lack capacity in a general sense and her expressed willingness still governed, because the statute required her consent. Capacity is question-specific, and a finding of incapacity for some purposes does not licence a decision for her on this one.

And regulating a technology is not the same as prohibiting it. The PCPNDT Act is the model: the machine remains lawful, and the use, the disclosure and the premises are controlled.

Quick revision

• Reproductive technology separates conception from intercourse, gestation from parenthood, and diagnosis from birth; each separation creates a decision, and the legal question is who decides.
• Suchita Srivastava (2009): a woman's right to make reproductive choices is a dimension of personal liberty under article 21; exercisable to procreate as well as to abstain; privacy, dignity and bodily integrity are the crucial considerations; the right includes carrying to term, giving birth and raising children. Termination could not be ordered without her consent even though she was found to be mentally retarded, because the MTP Act itself requires it. There is a compelling State interest in the prospective child, and the MTP Act's conditions are reasonable restrictions.
• CEDAW article 16(1)(e): the same rights to decide freely and responsibly on the number and spacing of children and to have access to the information, education and means to exercise them.
• Four statutes, all Module II: MTP 1971, PCPNDT 1994, ART (Regulation) 2021, Surrogacy (Regulation) 2021.
• PCPNDT is the model dual-use response: the machine is lawful; the use, the communication of the result and the registration of the clinic are regulated.

Test yourself

1. What did Suchita Srivastava decide about reproductive choice, and why is the direction of the choice significant? It held that a woman's right to make reproductive choices is a dimension of personal liberty under article 21, that such choices may be exercised to procreate as well as to abstain from procreating, and that the crucial consideration is respect for her privacy, dignity and bodily integrity, reproductive rights including her entitlement to carry a pregnancy to full term, to give birth and to raise children. The direction matters because the choice protected was the choice to continue the pregnancy, asserted by a woman whose capacity was in question, against a High Court order for termination that others considered to be in her interests. It therefore establishes that the right is not a right to terminate but a right to decide, and it does so in the situation where the argument for overriding the person was at its strongest.

2. On what basis did the Court hold that her consent was required, given the finding about her capacity? On the basis of the statute. The Medical Termination of Pregnancy Act 1971 itself contemplates that a woman found to be mentally retarded should give her consent, so the requirement did not have to be constructed from the constitutional right alone. The Court accordingly respected her expressed willingness to bear the child notwithstanding her lack of understanding of the sexual act and the apprehensions about her capacity to carry the pregnancy to term and to assume maternal responsibilities, and it declined to use the parens patriae jurisdiction to reach a different result, the Expert Body having in any event reported in favour of continuation.

3. Explain why the PCPNDT Act is the model response to a dual use technology. Because it separates the technology from its use. Prenatal diagnostic technology is valuable for the health of the pregnancy and is the same technology that enables selection by sex. Parliament did not prohibit the equipment, which would have defeated the beneficial use and with it the right to enjoy the benefits of scientific progress. It regulated instead the conduct: the use of the technique for the prohibited purpose, the communication of the sex of the foetus, and the registration and conduct of the premises where the technique is performed. That is the structure a dual use analysis calls for, since the object of regulation is a use by an actor in a setting rather than a body of knowledge or a machine.

4. State CEDAW article 16(1)(e) and identify the two features that give it force. It requires States Parties to ensure, on a basis of equality of men and women, the same rights to decide freely and responsibly on the number and spacing of their children and to have access to the information, education and means to enable them to exercise these rights. Its first feature is that the right is a right to decide, which places it in autonomy rather than in health provision, so a State does not comply merely by providing services if the decision is made by someone else. Its second is that it expressly extends to the information, education and means, so a State that recognises the choice in principle while leaving the means unavailable, unaffordable or unreachable has not discharged the obligation.

5. A clinic requires a husband's signature before a woman may undergo a fertility procedure. Analyse the requirement. It transfers to another person a decision that Suchita Srivastava locates within the woman's own personal liberty under article 21, the crucial considerations being her privacy, dignity and bodily integrity, and it does so in respect of a procedure performed on her body. It also conflicts with CEDAW article 16(1)(e), which requires the same rights to decide on the number and spacing of children on a basis of equality of men and women, since a veto held by one spouse and not the other is the denial of that equality. The general principle from bodily integrity applies as well: an intervention on a person requires that person's consent, and a third party's consent is not a substitute for it but at most an additional formality. Whether a particular statute or clinic rule imposes such a requirement, and what follows if it does, is a question of the statutory scheme taken in Module II; the constitutional objection stands independently of it.

Held, on article 21 and beyond. Compulsion also violates the substantive due process required for restraining personal liberty, and does so whether the technique is administered during an investigation or for any other purpose, since the results could expose a person to adverse consequences of a non-penal nature. The techniques cannot be read into the Explanation to sections 53, 53-A and 54 of the Code by an expansive interpretation. Compulsory administration is an unjustified intrusion into mental privacy, would amount to cruel, inhuman or degrading treatment in the language of evolving international human rights norms, and conflicts with the right to a fair trial. Invocations of a compelling public interest cannot justify diluting such rights.

The order. No individual should be forcibly subjected to any of the techniques, whether in a criminal investigation or otherwise. Room is left for voluntary administration with safeguards, but even where the subject has consented the results by themselves are not admissible, because the subject does not exercise conscious control over the responses.

The distinction that decides these cases

The reason DNA and fingerprints are treated differently from narcoanalysis is not that one is more modern or more reliable. It is the testimonial line.

The Court's reasoning on the three techniques is that each produces a result that is testimonial in character even though nothing is spoken in the ordinary way. The polygraph and the BEAP test read a bodily response, but the response is a proxy for the content of the mind, so the result conveys personal knowledge. That step is the intellectual heart of the case and is what a good answer must reproduce.

The second holding, which is often missed

Even a consenting subject's test results are not admissible by themselves. The reason given is that the subject does not exercise conscious control over the responses. That is a statement about the reliability of the technique rather than about the rights of the accused, and it matters for two reasons.

It means consent does not convert the technique into ordinary evidence, so a practice of obtaining consent from persons in custody cannot be used to get the results before a court.

And it shows the Court doing something courts are often reluctant to do: assessing the scientific claim rather than accepting it. The techniques were not excluded only because using them would be unfair; they were also found not to be what they were said to be.

DNA, and where the line now sits

DNA identification is material evidence on the distinction above, so article 20(3) is not the obstacle. The questions it raises fall elsewhere and are the questions worked in [Bodily Integrity and Intervention Without Consent] and [Privacy and Surveillance]:

• Taking the sample is an interference with bodily integrity, so it needs statutory authority and proportionality.
• Retention of the profile after acquittal is an interference with informational privacy that is not justified by the aim of solving the offence charged.
• Familial searching, partial matching to identify relatives, extends the burden to persons against whom nothing is alleged and who have not consented.
• Error and interpretation: a match is a probability, and its weight depends on the size and composition of the reference database, which is a matter about which the accused must be able to cross-examine.

Worked example

An investigating officer proposes to put a suspect through a polygraph examination and, on the strength of the result, to search a location. The suspect signs a consent form at the police station in the presence of the officer.

Was the consent free? A consent given in custody to a technique proposed by the custodian raises the freedom question directly. This is the point where the consent requirements set out in [Bodily Integrity and Intervention Without Consent] do their work, and Selvi contemplates voluntary administration only with safeguards.

If the consent was valid, may the result be used? Not by itself. Selvi is express: even where the subject has consented, the results by themselves are inadmissible, because the subject does not exercise conscious control over the responses.

May the search proceed on the result? This is the practical question and it is distinct. The prohibition on compulsory administration is absolute in the sense the Court described, and the inadmissibility of the result is a rule about evidence. Whether information derived from a lawful voluntary test may direct an investigation is a separate question from whether the test itself may be led, and an answer should keep the two apart.

And what if there had been no consent? Then nothing in it is available. Compulsory administration violates article 20(3) and substantive due process under article 21, and a compelling public interest is not an answer.

What it does NOT mean

It is not a prohibition on forensic science. DNA, fingerprints and medical examination under the Code are untouched. The case is about techniques that read the mind.

Article 20(3) is not confined to the trial. Selvi holds that its protective scope extends to the investigative stage, and that read with section 161(2) it covers suspects and witnesses as well as the accused.

"It only guides the investigation" is not an answer. The Court rejected it, holding the violation complete whether the technique is administered in an investigation or for any other purpose, since the results can expose a person to non-penal adverse consequences.

And the absence of pain is not the test. The State argued that the techniques cause no bodily harm, and the Court held that the intrusion is into mental privacy, which the absence of physical harm does not answer.

Quick revision

• The line is material versus testimonial. Material evidence exists independently of the mind and may be compelled; testimonial evidence is the product of the accused's mental process and may not.
• Selvi (2010): compulsory narcoanalysis, polygraph and BEAP violate article 20(3), whose scope reaches the investigative stage and, with section 161(2), covers accused, suspects and witnesses; the results are testimonial, not material. They also violate substantive due process under article 21, are an unjustified intrusion into mental privacy, would amount to cruel, inhuman or degrading treatment, and conflict with the right to a fair trial. A compelling public interest cannot justify dilution.
• The techniques cannot be read into the Explanation to sections 53, 53-A and 54 of the Code.
• Voluntary administration is possible with safeguards, but the results by themselves are inadmissible, because the subject does not exercise conscious control over the responses.
• ICCPR article 14(3)(g): not to be compelled to testify against oneself or to confess guilt.
• DNA is material evidence, so the issues are taking, retention, familial searching and the weight of a match, not article 20(3).

Test yourself

1. State the distinction between material and testimonial evidence and explain how the Court applied it to the polygraph. Material evidence is a physical thing obtained from the body or the scene, such as blood, hair or a fingerprint; it exists independently of the accused's mind and would exist if they were unconscious, so compelling its production does not require the accused to furnish anything from their own knowledge and article 20(3) is not engaged. Testimonial evidence is the product of the accused's own mental process and conveys personal knowledge, and it is protected. The Court applied the distinction to the polygraph by looking past the fact that what is measured is a bodily response: the response is a proxy for the content of the mind, so the result conveys personal knowledge and bears a testimonial character, which is why it cannot be categorised as material evidence.

2. How far does the protection of article 20(3) extend, on the reasoning in Selvi? Its protective scope extends to the investigative stage of criminal cases, and not merely to the trial; read with section 161(2) of the Code of Criminal Procedure 1973 it protects accused persons, suspects and witnesses examined during an investigation alike; and it protects the choice between speaking and remaining silent irrespective of whether the testimony that would result proves inculpatory or exculpatory, since the aim is to prevent the forcible conveyance of personal knowledge relevant to the facts in issue.

3. Why are the results of a voluntarily administered test still inadmissible by themselves, and what is significant about that holding? Because the subject does not exercise conscious control over the responses, so the material produced is not the product of a deliberate act of communication even where the subject agreed to undergo the procedure. It is significant for two reasons. It prevents the prohibition being circumvented by obtaining consent from persons in custody, since consent secures the lawfulness of administering the test but not the admissibility of its product. And it shows the Court evaluating the scientific claim rather than accepting it: the exclusion rests not only on the unfairness of compelling the technique but on a finding about what the technique can actually establish.

4. The State argued that the techniques cause no bodily harm and are gentler than the third degree. Why did these arguments fail? The absence of bodily harm failed because the intrusion the Court identified was into mental privacy, and the integrity protected under article 21 extends to the mind as well as the body, so a technique that touches nobody may still be an intrusion. The comparison with the third degree failed because the alternative offered is itself unlawful, and the lawfulness of a practice is not established by showing that it is preferable to torture. The Court also rejected the related submission that the techniques would merely guide the investigation, holding that the violation is complete whether they are administered during an investigation or for any other purpose, because the results can expose a person to adverse consequences of a non-penal nature.

5. Since DNA identification is material evidence, what questions does it raise and where are they answered? Article 20(3) is not the obstacle, so the questions arise elsewhere. The taking of a sample is an interference with bodily integrity and requires statutory authority and proportionality. The retention of a profile after an acquittal is an interference with informational privacy that the aim of investigating the offence charged does not justify, since that aim has been exhausted. Familial searching extends the burden to relatives against whom nothing is alleged and who have given no consent, so the interference reaches persons entirely outside the process. And the weight of a match is a probability that depends on the size and composition of the reference database, which the accused must be able to test in cross-examination if the right to a fair trial is to be real. These are questions of bodily integrity, of privacy and of fair trial rather than of self-incrimination.

Its example is worth setting out because it is the clearest case in the whole Comment. Where parents decide not to have their children vaccinated on grounds the scientific community considers false, the parents' decision entails risks for the child and sometimes for society, from the possible resurgence of infectious diseases previously under control. In these cases the best interests of the child must be a primary consideration.

Paragraph 44 continues with two propositions that bear on children and on others: that people may be subject to great pressure from their social environment to undergo traditional treatment rather than benefit from the best available medical attention, and that States must guarantee everyone the right to choose or refuse treatment with full knowledge of the risks and benefits, subject to limitations meeting the criteria of article 4; and that States must establish protective measures in relation to messages from pseudoscience, which create ignorance and false expectations among the most vulnerable sectors of the population.

The Convention, named and placed

The Convention on the Rights of the Child of 1989 is the instrument that governs this area, and it is the most widely ratified human rights treaty there is. General Comment 25 relies on it, citing at footnote 12 particularly its articles 24 and 29, in support of the proposition in paragraph 38 about children in poverty, scientific education and the development of the child's abilities to their fullest potential.

Two things follow.

What can be said. The Convention is the governing instrument; the Committee treats it as directly relevant to the enjoyment of the right to science by children; and the standard of the best interests of the child, which paragraph 44 applies, is the Convention's central principle.

What this book will not do. The Convention is not in this subject's authority base, so no article of it is quoted here and none is paraphrased. A candidate should do the same: name it, state what the Committee draws from it, and avoid reciting article numbers that have not been read. That is a stronger answer than a confident but unsourced list, and it is the discipline this book applies to every instrument it does not hold.

The Indian provisions

Article 21A: free and compulsory education for children of six to fourteen, discussed in [Education and the Diffusion of Knowledge].

Article 24: no child below the age of fourteen years shall be employed to work in any factory or mine or engaged in any other hazardous employment. It is a fundamental right and it is directly a technology provision, since the hazard it addresses is industrial.

Article 39(e): the State shall direct its policy towards securing that the tender age of children is not abused and that citizens are not forced by economic necessity to enter avocations unsuited to their age or strength. Article 39(f): that children are given opportunities and facilities to develop in a healthy manner and in conditions of freedom and dignity, and that childhood and youth are protected against exploitation and against moral and material abandonment.

ICESCR article 10(3) requires special measures of protection and assistance on behalf of all children and young persons, and requires States to set age limits below which the paid employment of child labour is prohibited and punishable. Article 12(2)(a) requires steps for the reduction of the stillbirth rate and of infant mortality and for the healthy development of the child.

Worked example

A State proposes that every school pupil be issued a tablet preloaded with an application that records attendance, monitors reading progress and reports both to a central education database, retained until the pupil turns twenty-five. Parents consent on enrolment by signing a general form.

The benefit is real and should be stated. Article 15(1)(b) gives the child an interest in what science has made available, and paragraph 38 requires pedagogical tools and quality scientific education, particularly for children in poverty. An answer that treats the scheme as simply objectionable has ignored half the framework.

Consent. The child has not consented and cannot. The parent has consented, in general terms, at a moment when refusal would have meant not enrolling in school. Consent given as the price of a compulsory service is not free, and paragraph 44 is the provision that lets a State's obligation be asserted against the family's decision rather than through it.

Whose interests. The best interests of the child are a primary consideration. The interests served by retention until twenty-five are administrative, and the child's interest in not carrying a school record into adult life points the other way.

Duration. Retention for a decade after the child leaves school is the point at which the scheme fails proportionality most clearly, because the purpose that justifies collection expires long before the retention does.

Equality. Paragraph 38 requires attention to children living in poverty, particularly those with disabilities. A device-based scheme distributes by household, so a child without power at home is monitored as absent rather than assisted, and the tool that was meant to include them records their exclusion.

What survives. The tablets, the teaching application, and progress data held by the school for the school year. What does not survive is central retention to twenty-five on a consent the child never gave.

What it does NOT mean

It does not mean children should be kept away from technology. The Committee's demand is for access: pedagogical tools and quality scientific education, especially for children in poverty and children with disabilities. Exclusion is the harm paragraph 38 is directed against.

It does not mean parental consent is worthless. It is ordinarily the mechanism by which a child's rights are exercised. Paragraph 44 identifies the exception, where the decision made for the child is contrary to the child's interests on grounds the scientific community considers false.

Best interests is not the same as the child's wishes, and it is a primary consideration rather than the only one. The Committee's own example, vaccination, is a case where the standard operates against a decision made within the family and in favour of the child and of others.

And a child's inability to consent does not license intervention. The rule in [Bodily Integrity and Intervention Without Consent] applies: where a person cannot consent, the intervention needs authority, a best-interests basis and, in research, direct health benefit save by narrow exception.

Quick revision

• Children are treated separately because the conditions of exercising rights are absent: no capacity to consent, no ability to assess risk, dependence on adults, and a longer remaining exposure.
• General Comment 25, paragraph 38: measures so that children living in poverty, particularly those with disabilities, have full access; they are entitled to special care and assistance, through pedagogical tools and quality scientific education developing the child's personality, talents and mental and physical abilities to their fullest potential. Footnote 12 cites the Convention on the Rights of the Child, particularly articles 24 and 29.
• Paragraph 44: States may have to protect people within their familial, social or cultural context; those who cannot choose for themselves by reason of age or capacity need special protection; where parents refuse vaccination on grounds the scientific community considers false, the best interests of the child must be a primary consideration. Also: the right to choose or refuse treatment with full knowledge of risks and benefits, and protective measures against messages from pseudoscience.
• Constitution: article 21A (education six to fourteen), article 24 (no child under fourteen in a factory, mine or hazardous employment), article 39(e) and (f).
• ICESCR article 10(3) (special measures, age limits for paid employment), article 12(2)(a) (infant mortality and the healthy development of the child).
• The Convention on the Rights of the Child is not in this book's base: name it, say what the Committee draws from it, quote nothing.

Test yourself

1. Why does the framework treat children separately, given that they hold the same rights? Because the conditions on which rights are exercised are absent or impaired. A child lacks the legal capacity to consent, so the ordinary licence for intervention is unavailable; lacks the information and experience to assess a risk, particularly one that will mature years later; is dependent on adults whose interests may diverge from theirs; and has a longer remaining lifetime over which a harm can operate, so that data collected or a decision recorded now continues to affect them into adult life. The rights are the same; what differs is the person's ability to invoke them, which is why the framework supplements them with special protection and with the standard of the best interests of the child.

2. Set out paragraph 44 of General Comment 25 and explain why it is significant that it addresses the family. It states that States parties may sometimes have to protect people within their own familial, social or cultural context where their right to participate in and enjoy the benefits of scientific progress is affected, and that persons who cannot choose for themselves by reason of age or capacity must receive special protection. Its example is that where parents decide not to have their children vaccinated on grounds the scientific community considers false, the decision entails risks for the child and sometimes for society through the resurgence of infectious diseases previously under control, and that in such cases the best interests of the child must be a primary consideration. It is significant because human rights law ordinarily treats the family as the child's protection and the State as the potential threat. Here the Committee accepts that the decision made within the family can itself be the interference, which supplies the basis for a State obligation asserted against the parent's decision rather than exercised through it.

3. What does paragraph 38 require, and which treaty does the Committee rely on for it? It requires States to adopt measures ensuring that children living in poverty, particularly those with disabilities, have full access to the enjoyment of the right to participate in and enjoy the benefits of scientific progress and its applications, on the footing that they are entitled to special care and assistance, especially through pedagogical tools and quality scientific education that allow the development of the child's personality, talents and mental and physical abilities to their fullest potential. The Committee cites the Convention on the Rights of the Child, particularly its articles 24 and 29, at footnote 12.

4. How should a candidate handle the Convention on the Rights of the Child on this paper, if they have not read its text? By naming it and placing it rather than reciting it. The correct statement is that the Convention of 1989 is the governing instrument for children's rights and the most widely ratified human rights treaty, that General Comment 25 relies on it for the proposition in paragraph 38 and cites particularly its articles 24 and 29, and that its central principle, the best interests of the child, is the standard the Committee applies in paragraph 44. What should not be done is to attribute numbered provisions to it from memory, since an article number recalled rather than read is the commonest way an otherwise sound answer becomes inaccurate, and the argument can be made in full from the Comment, which is available.

5. A school scheme monitors pupils through issued devices and retains the data until they are twenty-five, with a general parental consent given on enrolment. Identify the three strongest objections. First, consent: the child has not consented and cannot, and the parent's consent was given in general terms at a moment when refusal meant forgoing schooling, so it is not free in the sense the framework requires, and paragraph 44 permits the State's obligation to be asserted against a decision made within the family. Second, duration and purpose: the justification for collection is educational and expires when the pupil leaves school, whereas the retention continues for years afterwards, so the measure fails proportionality at the point where purpose and retention diverge. Third, equality: paragraph 38 requires particular attention to children living in poverty and to those with disabilities, and a device-based scheme distributes by household, so a child without reliable power or connectivity is recorded as absent rather than assisted, and a tool intended to include them documents their exclusion instead.

Paragraph 19 adds the acceptability element: scientific education and the products of science should be tailored to the particularities of populations with special needs, such as persons with disabilities. So the requirement of adapted formats is not only in paragraph 35 but is built into one of the four elements of the right itself.

Two ideas that carry marks

Universal design means designing a product, environment or service so that it is usable by the widest range of people without adaptation. It operates before the person arrives. A building with a step and a later ramp has been accommodated; a building designed level was universally designed. In technology the difference is stark and cheap: software written so that a screen reader can interpret it costs almost nothing at the design stage and cannot be retrofitted without rewriting.

Reasonable accommodation is the modification made for a particular person in a particular case, where the general design does not serve them. It operates after the person arrives, and is limited by reasonableness.

The two are complements, not alternatives, and the Committee requires both, at paragraph 35(c) and (e) respectively. The distinction matters legally because they allocate cost and initiative differently: universal design is a standing obligation owed to everybody in advance, while reasonable accommodation is triggered by an individual and can be resisted on grounds of disproportionate burden. A State that relies on accommodation alone has left the excluded person to ask, every time.

The Convention, named and placed

The Convention on the Rights of Persons with Disabilities of 2006 is the governing instrument. General Comment 25 relies on it, citing at footnote 11 particularly its articles 1 to 9, in support of the requirement to raise awareness of capabilities and contributions and to combat stereotypes and harmful practices.

As with the Convention on the Rights of the Child in [Technology and the Rights of the Child], the Convention is not in this book's authority base. It is named, placed, and relied on through what the Committee records of it, and no article of it is quoted here. A candidate should do likewise, and should note that the Convention is the source of the two concepts above, which the Committee uses without defining them.

The Indian frame

Article 14 guarantees equality before the law and the equal protection of the laws, and equal protection has long been understood to permit and sometimes require differential treatment where circumstances differ. Article 15 prohibits discrimination on the enumerated grounds and permits special provision. Article 41 requires the State, within the limits of its economic capacity and development, to make effective provision for securing the right to work, to education and to public assistance in cases of unemployment, old age, sickness and disablement, and in other cases of undeserved want. Article 41 is the directive principle that names disablement expressly and is the natural citation for the distributive obligation.

Worked example

A State digitises its public services: applications, payments, certificates and grievances all move to a single portal. The portal is built to a modern visual design and is not compatible with screen readers. The State's answer is that any person unable to use it may attend a facilitation centre where an operator will complete the transaction for them.

Which right, and how is it engaged? Article 15(1)(b) read with article 2(2). The applications of science must be accessible without discrimination, particularly where they are instrumental for the enjoyment of other rights, and here they are instrumental for almost every dealing with the State.

Universal design. Paragraph 35(c) requires it, and this is the paradigm failure: compatibility with assistive software is a design property that costs little in advance and a great deal to retrofit. The exclusion was not intended and is complete.

Is the facilitation centre reasonable accommodation? It is an accommodation, and it is worse than it looks. It requires the person to travel; it makes them dependent on an operator; and it destroys the privacy of every transaction, since the operator sees the medical, financial and family information the portal handles. A blind applicant conducts their affairs in front of a stranger while a sighted applicant does so alone. That is not equal access; it is access on inferior terms, and the inequality is in dignity and privacy rather than in outcome.

Adapted formats. Paragraph 35(e) requires that the products of scientific development be disseminated in adapted formats, and paragraph 19 puts the same requirement inside acceptability. The obligation is to make the portal usable, not to supply a human substitute for it.

Participation. Paragraph 35(a) asks who was in the room. A portal that fails this way is almost always one designed without any user who needed a screen reader, which is the design-stage exclusion of paragraph 34 producing the access-stage exclusion.

Measurement. Paragraph 35(b) requires statistics disaggregated by disability. A State that does not measure who is failing to complete a transaction cannot know it has this problem, and cannot demonstrate compliance.

What it does NOT mean

It does not mean technology is uniformly good here. The same automation discussed in [Artificial Intelligence and Decisions by Machines] can exclude: a system trained on data that does not represent disabled users performs worse for them, and a verification step that assumes a particular physical capability locks them out entirely. The point is that the balance in this area is unusually favourable, not that the risks are absent.

Assistive technology is not charity. It is the means by which a right is exercised, which is why the Committee places it under an obligation rather than under welfare.

Reasonable accommodation does not discharge the design obligation. They are separate requirements in paragraph 35, and an accommodation that requires the person to ask, to travel, or to surrender privacy is unequal access even when it works.

And participation is not consultation after the fact. Paragraph 35(a) speaks of participation and contribution in decision-making procedures concerning science, which is a place at the table rather than a comment on a finished design.

Quick revision

• The area where technology is chiefly an enabler, so the legal problem is distribution and design, not restraint.
• Paragraph 34: deep discrimination, for two reasons, physical, communication and information obstacles, and products that do not take account of their specificities and particular needs; and persons with disabilities contribute to the promotion of the right.
• Paragraph 35, seven measures at the least: (a) participation in decision-making, including women with disabilities facing multiple discrimination; (b) statistics disaggregated by disability; (c) universal design; (d) promote technologies facilitating access to scientific education and employment; (e) reasonable accommodation, including dissemination in adapted formats; (f) raise awareness and combat stereotypes, citing the CRPD, particularly articles 1 to 9; (g) free, prior and informed consent where they are research subjects.
• Paragraph 19: acceptability requires science and its products to be tailored to populations with special needs, such as persons with disabilities.
• Universal design operates before the person arrives and is owed to everyone; reasonable accommodation operates after, for an individual, and is limited by reasonableness. Both are required.
• India: articles 14 and 15; article 41, which names disablement expressly.
• The CRPD is not in this book's base: name it, say what the Committee draws from it, quote nothing.

Test yourself

1. Give the Committee's two reasons why persons with disabilities have suffered deep discrimination in the enjoyment of this right, and say why the second is the more significant. Either because of severe physical, communication and information obstacles to accessing basic and higher scientific education and careers, or because the products of scientific progress do not take into account their specificities and particular needs. The second is more significant because it locates the exclusion at the design stage rather than at the point of access: nothing is placed in the person's way and nobody intends to exclude them, but a product built without them in mind is unusable by them, and the effect is as complete as a barrier. It follows that the remedy cannot be limited to removing obstacles after the fact and must reach the design itself, which is why the Committee requires universal design and participation in decision-making.

2. Distinguish universal design from reasonable accommodation and explain why both are required. Universal design is the design of a product, environment or service so that it is usable by the widest possible range of people without adaptation; it operates before any particular person arrives, is owed generally, and in technology is usually cheap in advance and expensive to retrofit. Reasonable accommodation is a modification for a particular person in a particular case where the general design does not serve them; it operates after the person arrives, is triggered by them, and is limited by reasonableness so that it can be resisted as a disproportionate burden. Both are required, at paragraph 35(c) and (e) respectively, because each fails where the other is absent: design cannot anticipate every individual need, and accommodation alone leaves the excluded person to ask every time, on terms that are often less private and less convenient than those enjoyed by everybody else.

3. List the measures paragraph 35 requires States to adopt at the least. To promote the participation and contributions of persons with disabilities, including women with disabilities who face multiple discrimination, in decision-making procedures concerning science; to develop statistics on access to science and its benefits disaggregated by disability; to implement universal design; to promote technologies that facilitate access to scientific education and employment; to ensure that reasonable accommodation is provided so that they can access scientific education and employment and benefit from the products of scientific development, including diffusion and dissemination in adapted formats; to adopt appropriate measures to raise awareness of their capabilities and contributions and to combat stereotypes and harmful practices; and to ensure that they have given their free, prior and informed consent when they are subjects of research.

4. A State offers a facilitation centre where an operator completes online transactions for anyone unable to use its portal. Is that equal access? No. It is an accommodation rather than access, and it is unequal in three respects that the framework recognises. It requires the person to travel, so the transaction costs them time and money a sighted user does not spend. It makes them dependent on another person's assistance, removing the independence the service gives everybody else. And it destroys privacy, since the operator sees whatever the transaction contains, including medical, financial and family information, so the person conducts their affairs before a stranger while others conduct theirs alone. The inequality is therefore in dignity and privacy rather than in outcome, and paragraph 35(c) and (e) require the portal itself to be usable and disseminated in adapted formats rather than a human substitute to be provided for it.

5. Why does the Committee say that persons with disabilities contribute to the promotion of this right, and which element of the right does that statement belong to? Because it treats them as participants in science and not merely as recipients of its products, stating in paragraph 34 that they bring unique perspectives and experiences into the scientific landscape and thereby specifically contribute to the promotion of the right. The statement belongs to the third dimension of accessibility, the open opportunity for everyone to participate in scientific progress without discrimination, under which States must remove discriminatory barriers that impede participation. Its practical importance is that it supports the obligation in paragraph 35(a) to include them in decision-making procedures concerning science, which is a claim to a place in the process rather than a claim to the output of it, and it supplies the answer to the design-stage exclusion identified in paragraph 34.

What the Committee says

Paragraph 17, under accessibility, requires that everyone have equal access to the applications of science, particularly when they are instrumental for the enjoyment of other economic, social and cultural rights. Medicines are the clearest case of such an application.

Paragraph 79 requires that access to research results and their applications be regulated in a form that allows developing countries and their citizens adequate access to these products in an affordable manner, such as access to essential medicines. The Committee names essential medicines expressly as its example.

Paragraph 80 requires that the benefits and applications resulting from scientific progress be shared, with due incentives and regulations, with the international community, particularly developing countries, communities living in poverty and groups with special needs, especially where the benefits are closely related to the enjoyment of economic, social and cultural rights.

The words "with due incentives and regulations" are the Committee accepting, in its own text, that the financing problem is real. That is the balanced position, and a candidate who quotes the phrase demonstrates it.

UNESCO 1997, article 4, adds a rule at the boundary: the human genome in its natural state shall not give rise to financial gains. It does not prohibit patents on inventions derived from genetic knowledge, and the qualification is where the argument lies.

The Indian answer: section 3(d)

India's legislative response is the most studied in the world, and this is the case.

Novartis AG v. Union of India, AIR 2013 SC 1311, (2013) 6 SCC 1, decided on 1 April 2013 by Aftab Alam and Ranjana Prakash Desai JJ.

Facts. Novartis applied for an Indian patent for the beta crystalline form of Imatinib Mesylate, marketed as Gleevec, used in the treatment of chronic myeloid leukaemia. The application had lain in the mailbox awaiting the amendments by which India introduced product patents for pharmaceuticals. It was refused, and the refusal upheld by the Intellectual Property Appellate Board, on the ground among others that the claimed product was a new form of a known substance which did not satisfy section 3(d) of the Patents Act 1970.

Held. Where a patent is claimed for a new form of a known substance with known efficacy, the product must satisfy, in addition to clauses (j) and (ja) of section 2(1), the test of enhanced efficacy in section 3(d) read with its Explanation. In the case of a medicine, efficacy means therapeutic efficacy. Increased bioavailability alone may not necessarily amount to an enhancement of therapeutic efficacy; whether it does in a given case must be specifically claimed and established by research data, and no material had been offered to show that the beta crystalline form would produce enhanced or superior therapeutic efficacy over the Imatinib free base. In whichever way section 3(d) is viewed, whether as setting up standards of patentability or as an extension of the definition of invention, the product failed its test.

Held, the caution. The Court added expressly that this does not mean section 3(d) bars patent protection for all incremental inventions of chemical and pharmaceutical substances, and that it would be a grave mistake to read the judgment as meaning that section 3(d) was amended so as to undo the change made by the deletion of section 5 of the parent Act.

Why it matters here. Section 3(d) is a domestic device that keeps a known medicine in the public domain unless a genuine therapeutic gain is proved, which is the practice sometimes called evergreening: extending a monopoly by patenting minor variations of an existing product. The case is important for this paper in two ways. It shows a legislature designing patent law with access in view, which is what the Committee asks for at paragraph 79. And the caution in the judgment shows a court refusing to convert a targeted rule into a general hostility to innovation, which is the balance paragraph 80 states in the words "with due incentives and regulations".

TRIPS and Doha, named and placed

The international framework is the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) of 1994, which requires members of the World Trade Organization to make patents available for inventions in all fields of technology including pharmaceuticals, and the Doha Declaration on the TRIPS Agreement and Public Health of 2001, adopted after the controversy over access to antiretroviral medicines, which affirmed that the Agreement should be interpreted and implemented in a manner supportive of members' right to protect public health and to promote access to medicines for all, and addressed the flexibilities available to that end, including compulsory licensing.

Neither instrument is in this book's authority base. They are named here because an answer on this topic that did not mention them would be incomplete, and nothing is quoted from either. A candidate should do the same: name them, state their effect in general terms, and avoid attributing numbered articles to a text they have not read. The compulsory licence, by which a State authorises manufacture without the patent holder's consent on stated conditions, is the mechanism most often discussed, and it is a creature of the domestic patent statute operating within that international framework.

Worked example

A company holds an Indian patent on an antiviral medicine. The annual course costs more than a district's entire drug budget. A domestic manufacturer can produce it at a fraction of the price. The company argues that a compulsory licence would destroy the incentive to develop the next antiviral.

Which rights, on each side? For the patients, article 12 and article 15(1)(b), reinforced by paragraphs 17 and 79 of the Comment, which name affordable access to essential medicines. For the company, no Covenant right of its own: article 15(1)(c) protects the author, and the corporate patent holder is not one. Its argument is a policy argument about incentives, and paragraph 80's reference to due incentives is where that argument properly belongs.

Is the patent valid at all? The first question, not the last. If the product is a new form of a known substance, section 3(d) applies and enhanced therapeutic efficacy must be established by research data, on Novartis. A patent that should not have been granted needs no licence to defeat it.

If the patent is valid. The question becomes the statutory conditions for a compulsory licence, which is domestic patent law, applied within the TRIPS framework as addressed at Doha. The human rights framework does not itself grant the licence; it supplies the reason the statutory power exists and the standard against which its exercise is judged.

And what else is available. The framework's own preferences are broader than litigation: public funding conditioned on affordable access, tiered pricing, voluntary licensing for local manufacture, and the international measures in [International Co-operation in Science]. An answer confined to compulsory licensing has taken the narrowest of the available routes.

What it does NOT mean

It does not mean patents violate human rights. The Covenant protects the author's interests in article 15(1)(c) and the Committee accepts the need for incentives in terms. The claim is about the design and the limits of the system.

Section 3(d) is not a bar on incremental innovation. Novartis says so expressly, and calls the contrary reading a grave mistake.

Efficacy is not any advantage. For a medicine it is therapeutic efficacy, and increased bioavailability does not by itself establish it; it must be claimed and proved by research data.

And a corporation is not an author. The most common error in answers on this topic is to give a company the benefit of article 15(1)(c).

Quick revision

• Three provisions: article 15(1)(b) (benefits of scientific progress), article 12 (highest attainable standard of health), article 15(1)(c) (the author's moral and material interests, a human right of a natural person, not a patent).
• General Comment 25: paragraph 17, equal access to applications particularly where instrumental for other economic, social and cultural rights; paragraph 79, access regulated so that developing countries and their citizens have affordable access, such as essential medicines; paragraph 80, benefits shared with due incentives and regulations.
• UNESCO 1997, article 4: the genome in its natural state shall not give rise to financial gains. Not a bar on biotechnology patents.
• Novartis (2013): a new form of a known substance must satisfy section 3(d) enhanced efficacy in addition to section 2(1)(j) and (ja); for a medicine, efficacy means therapeutic efficacy; increased bioavailability alone is not enough and must be specifically claimed and established by research data. Express caution: section 3(d) does not bar all incremental inventions, and reading it so would be a grave mistake.
• TRIPS 1994 and the Doha Declaration 2001, not in the base: name them, state their effect, quote nothing. Compulsory licensing is a creature of the domestic statute within that framework.

Test yourself

1. Why can a corporate patent holder not rely on article 15(1)(c) of the Covenant? Because article 15(1)(c) recognises the right of everyone to benefit from the protection of the moral and material interests resulting from any scientific, literary or artistic production of which he is the author. It is a human right, held by a natural person, connected to that person's relationship with what they created and to their dignity. A patent is a statutory entitlement of limited duration, freely assignable, and in the pharmaceutical field usually held by a company that did not author anything and could not be an author. A corporation invoking the article is therefore invoking a right that is not its own, and its true argument, that monopoly rents finance future research, is a policy argument about incentives rather than a rights claim, which is where paragraph 80's reference to due incentives places it.

2. State the holding in Novartis on section 3(d), including the caution. Where a patent is claimed for a new form of a known substance with known efficacy, the product must satisfy, in addition to clauses (j) and (ja) of section 2(1) of the Patents Act 1970, the test of enhanced efficacy in section 3(d) read with its Explanation; in the case of a medicine, efficacy means therapeutic efficacy; and increased bioavailability alone may not necessarily amount to an enhancement of therapeutic efficacy, whether it does in a given case having to be specifically claimed and established by research data. No such material having been offered for the beta crystalline form of Imatinib Mesylate, the product failed the test however section 3(d) is characterised. The Court cautioned expressly that this does not mean section 3(d) bars patent protection for all incremental inventions of chemical and pharmaceutical substances, and that it would be a grave mistake to read the judgment as undoing the change made by the deletion of section 5 of the parent Act.

3. What does paragraph 79 of General Comment 25 require, and why is its example significant? It requires that access to research results and their applications be regulated in a form that allows developing countries and their citizens adequate access to these products in an affordable manner, and gives access to essential medicines as its example. The example is significant because it moves the Committee's requirement from a general statement about the diffusion of science to the specific field where the conflict with intellectual property is sharpest, and because it identifies affordability rather than mere availability as the standard. A State or an international arrangement under which a medicine exists but cannot be bought therefore fails the paragraph even though nothing is physically scarce.

4. Set out the two true propositions that make this topic genuinely difficult. The first is that research is expensive and must be financed, so that a system which allowed no return would reduce the supply of new medicines, and the loss would fall on future patients who would never learn what they did not receive; a framework treating every patent as a violation would therefore defeat article 15(1)(b) over time. The second is that a monopoly over a medicine for a fatal condition is in practical terms a monopoly over survival, and those excluded by the price are excluded by a legal instrument rather than by any scarcity of the substance itself. The framework does not choose between them: it treats intellectual property as a policy instrument subject to the rights, requires the domestic law to be designed with access in view, and accepts in paragraph 80 that benefits are to be shared with due incentives and regulations.

5. How should a candidate handle TRIPS and the Doha Declaration if they have not read the texts? By naming and placing them without quoting them. The accurate statement is that TRIPS, concluded in 1994, requires members of the World Trade Organization to make patents available for inventions in all fields of technology including pharmaceuticals, and that the Doha Declaration on the TRIPS Agreement and Public Health of 2001, adopted after the controversy over access to antiretroviral medicines, affirmed that the Agreement should be interpreted and implemented in a manner supportive of members' right to protect public health and to promote access to medicines for all, and addressed the flexibilities available for that purpose including compulsory licensing. What should be avoided is attributing numbered articles to either text from memory. An answer that names them accurately and argues the substance from the Covenant, General Comment 25 and Novartis is stronger than one that cites article numbers it cannot vouch for.

The second is unusual and worth noticing. A duty on the State not to spread false scientific information is a duty about the State's own speech, and it is stated as an element of the duty to respect a Covenant right.

To protect

Paragraph 43. The obligation requires States to adopt measures to prevent any person or entity from interfering with the right, for example by preventing access to knowledge or by discriminating on grounds of gender, sexual orientation or gender identity or other circumstances. The Comment names the entities it has in mind: universities, schools, laboratories, cultural or scientific associations, patients in hospitals and volunteers participating in scientific experiments. Its examples of the duty are:

• ensuring that scientific associations, universities, laboratories and other non-State actors do not apply discriminatory criteria;
• protecting people from participating in research or tests that contravene the applicable ethical standards for responsible research, and guaranteeing their free, prior and informed consent;
• ensuring that private persons and entities do not disseminate false or misleading scientific information; and
• ensuring that private investment in scientific institutions is not used to unduly influence the orientation of research or to restrict the scientific freedom of researchers.

The last example is the one that answers the funding contract in [The Scientist: Rights and Responsibilities]. A State that permits such a contract to be enforced without qualification has not protected the researcher's freedom, and the Comment says so in terms.

Paragraph 44 extends the duty into the family and the community, and is worked in [Technology and the Rights of the Child].

To fulfil

Paragraph 45. The duty requires States to adopt legislative, administrative, budgetary and other measures and to establish effective remedies aimed at the full enjoyment of the right. The measures include education policies, grants, participation tools, dissemination, providing access to the Internet and other sources of knowledge, participation in international co-operation programmes, and ensuring appropriate financing.

Paragraph 46 states that the duty to fulfil is reinforced and specified by article 15(2), the duty to take steps for the conservation, the development and the diffusion of science, discussed in [The Duty to Conserve, Develop and Diffuse Science].

Where regulation actually bites

The obligations above are stated at the level of principle. In practice a State regulates technology through a small number of instruments, and an answer is stronger for naming them.

The last row is the one most often overlooked, and it is frequently the most effective. A State that funds research can set the terms of access at the outset, which avoids the far harder problem of extracting concessions from a monopolist afterwards.

Worked example

A private company operates a network of diagnostic laboratories. It offers a discounted genetic screening panel, sells anonymised aggregate results to insurers, and requires customers to accept those terms as a condition of the discount. There is no statute governing genetic data as such.

Is the company bound by the Covenant? No. That is the starting point, not the conclusion.

What was the State required to do? Under the duty to protect, to adopt measures preventing any person or entity from interfering with the right, including by ensuring that non-State actors do not apply discriminatory criteria and that consent is free, prior and informed. Consent obtained as the price of a discount, from customers who are not told what re-identification is possible, is not consent of that quality.

Is "there is no statute" an answer for the State? It is the opposite of an answer. Paragraph 45 requires legislative measures, and paragraph 8 of the 1975 Declaration requires effective measures including legislative measures. The absence of a statute is the breach, not the excuse for it.

And if a statute exists but nothing is done? Then paragraph 9 of the Declaration is engaged, which requires action to ensure compliance with legislation guaranteeing human rights in the conditions of scientific development. Enactment without enforcement is a distinct failure and should be identified as such.

What would compliance look like? A statutory rule on genetic data and consent; a prohibition on use for insurance underwriting or a rule of non-discrimination equivalent to article 6 of the UNESCO Declaration of 1997; a regulator with inspection powers; and a remedy for the individual, which is [Remedies When Technology Causes Harm].

What it does NOT mean

It does not mean private actors have no responsibilities. They have them in domestic law, in professional regulation and in the ethics instruments. What they do not have is direct obligations under the Covenant, which is why the analysis is routed through the State.

The duty to protect is not a duty to prohibit. The Comment's examples are about ensuring, requiring and guaranteeing. Prohibition is one instrument among several and usually the crudest.

It is not satisfied by enactment alone. Paragraph 9 of the 1975 Declaration and the reference to effective remedies in paragraph 45 both point to enforcement.

And the duty to respect is not merely passive. Paragraph 42 requires the State to refrain from interfering indirectly as well as directly, and lists eliminating barriers, eliminating censorship and eliminating obstacles, all of which require action rather than abstention.

Quick revision

• Human rights bind States; technology is mostly private; so the framework reaches it through respect, protect and fulfil, paragraph 41.
• Respect, paragraph 42: refrain from interfering directly or indirectly; eliminate barriers to science education and careers; refrain from disinformation, disparagement or deliberate misinformation; eliminate censorship or arbitrary limitations on internet access; remove obstacles to international collaboration, unless justified under article 4.
• Protect, paragraph 43: prevent any person or entity from interfering; non-State actors must not apply discriminatory criteria; protect people from research contravening ethical standards and guarantee free, prior and informed consent; prevent private dissemination of false or misleading scientific information; ensure private investment does not unduly influence the orientation of research or restrict scientific freedom.
• Fulfil, paragraph 45: legislative, administrative, budgetary and other measures and effective remedies; education policies, grants, participation tools, dissemination, access to the Internet, international co-operation, appropriate financing. Paragraph 46: reinforced by article 15(2).
• 1975 Declaration: paragraph 8, effective measures including legislative measures; paragraph 9, action to ensure compliance with legislation. Enactment without enforcement is a separate failure.
• Six instruments: prohibition of a use, licensing, standards and certification, mandatory assessment, liability rules, and procurement and funding conditions.

Test yourself

1. Why does the analysis of a private company's conduct proceed through the State, and why is that not merely a technicality? Because the Covenant and the other instruments impose obligations on States parties, and a company is not a party to them, so a claim that the company has violated the Covenant is misconceived. It is not a technicality because it changes the question asked and the party answerable. Instead of asking whether the company behaved badly, which no international body can determine and no individual can enforce against it under the treaty, the analysis asks what the State licensed, required, prohibited, inspected, funded or failed to legislate. The wrong is identified as a regulatory failure attributable to a party that is bound and that can be brought before a treaty body or a domestic court.

2. Set out the four examples the Committee gives of the duty to respect. Eliminating barriers to accessing quality science education and to the pursuit of scientific careers; refraining from disinformation, disparagement or deliberate misinformation intended to erode citizen understanding of and respect for science and scientific research; eliminating censorship or arbitrary limitations on access to the Internet, which undermines access to and dissemination of scientific knowledge; and refraining from imposing, or eliminating, obstacles to international collaboration among scientists, unless such restrictions can be justified in accordance with article 4 of the Covenant. Three of the four require the State to eliminate something, so the duty to respect is not purely one of abstention.

3. Which example under the duty to protect answers the problem of a funder controlling a researcher's conclusions? The requirement that States ensure that private investment in scientific institutions is not used unduly to influence the orientation of research or to restrict the scientific freedom of researchers. It matters because the researcher's freedom under article 15(3) is a right against the State, so a contractual restriction imposed by a private funder cannot itself breach the Covenant. The duty to protect supplies the missing link: a State whose law enforces such a clause without qualification, and which exposes a researcher who publishes contrary to it to liability, has failed to adopt the measures the Comment requires.

4. A State says it cannot act against a harmful commercial use of a technology because no statute covers it. Assess that answer. It is not a defence but a statement of the breach. The duty to fulfil in paragraph 45 requires States to adopt legislative, administrative, budgetary and other measures and to establish effective remedies aimed at the full enjoyment of the right, and paragraph 8 of the 1975 Declaration requires effective measures including legislative measures to prevent and preclude the use of scientific achievements to the detriment of human rights. The absence of a statute in an area where private conduct is interfering with the right is therefore the failure alleged. A separate failure arises where a statute exists and nothing is done, since paragraph 9 of the Declaration requires action to ensure compliance with legislation guaranteeing human rights in the conditions of scientific and technological development.

5. Name the six instruments through which a State regulates technology and identify the one most often overlooked. Prohibition of a particular use while the technology itself remains lawful; licensing and registration of those who may operate and on what conditions; standards and certification governing whether a product may reach the public; mandatory assessment before deployment, such as the human rights impact assessments contemplated by paragraph 22; liability rules allocating the cost when harm occurs, such as absolute liability; and procurement and funding conditions. The last is the one most often overlooked and frequently the most effective, because a State that funds research can fix the terms of access at the outset, which is far easier than extracting concessions afterwards from a party that has since acquired a monopoly, and paragraph 79's requirement that access be regulated so as to be affordable is naturally implemented in that way.

General Comment 25, paragraph 45, includes the establishment of effective remedies within the duty to fulfil.

The Indian routes, and where each fails

An answer that lists remedies is adequate. An answer that says where each one breaks is good.

Constitutional

Article 32 gives the right to move the Supreme Court for the enforcement of fundamental rights, and is itself a fundamental right; article 226 gives the High Courts a wider jurisdiction. Public interest litigation has made these the principal route for diffuse harms, and M.C. Mehta was itself a petition under article 32.

Where it fails. It reaches State action and public functions most comfortably; it delivers directions better than it delivers money to individuals; and it depends on a petitioner willing to litigate for years. In M.C. Mehta the Court declined to set up a special machinery for the compensation claims and instead directed the Delhi Legal Aid and Advice Board to file actions in the appropriate court, with the Delhi Administration funding them, which shows the constitutional court routing individual compensation back to ordinary litigation.

Civil

The action in tort, and in this field principally the rule of absolute liability established in M.C. Mehta, which removes fault and the Rylands v. Fletcher exceptions, and measures compensation with reference to the magnitude and capacity of the enterprise so that it deters.

Where it fails. Not on doctrine but on proof, delay and cost. This is the lesson of Union Carbide, where the substantive rule already existed and the victims were defeated by forum, by not being heard before their claims were settled, by the contested categorisation of injury on which every individual award depended, and by time.

Criminal

Prosecution of those responsible. In Union Carbide the Court restored the criminal proceedings that the settlement had quashed, holding that its powers under articles 136 and 142 were not taken away by article 139A.

Where it fails. The standard of proof is the highest; identifying an individual within a corporate structure is difficult; and a conviction compensates nobody. Jacob Mathew, taken in Module II, sets a deliberately high threshold before professional error becomes crime.

Statutory

Consumer protection, industrial and factory legislation, environmental statutes, and sector regulators. The Consumer Protection Act 2019 supplies a forum designed for exactly the small-value, high-volume claim that ordinary litigation cannot absorb.

Where it fails. Jurisdictional limits, the need to fit the claimant into a defined status such as consumer, and the absence of any statute at all in new fields, which is the gap identified in [The State's Duty to Regulate Technology].

The human rights commissions

The Protection of Human Rights Act 1993 constitutes the National Human Rights Commission and State Commissions, with functions of inquiry, intervention in proceedings and recommendation.

Where it fails. The Commission's principal outputs are recommendations, so this route supplies inquiry and publicity rather than an enforceable award, which is a real contribution where the difficulty is opacity and a limited one where the need is money.

Worked example

An industrial process contaminates groundwater in three villages over a decade. Illness appears in a proportion of residents at rates above the district average. The company denies causation, points to other sources of contamination, and its consent to operate was never revoked.

Constitutional. A petition under article 32 or 226 can obtain the things individuals cannot: an order for testing, disclosure of the company's records, an expert committee, and remediation. It is the right route for the facts and for the structural relief.

Civil. M.C. Mehta removes fault, so the claimants do not have to prove negligence, and compliance with the consent to operate is no defence. What remains is causation for the individual, which the absolute liability rule does not touch, and this is where the claims will actually be contested.

Criminal. Available in principle, and it will identify no individual within the company without the documents that the constitutional proceeding is best placed to extract. Note the ordering: the routes are complementary and are usually run together.

Statutory. Whether the residents can bring themselves within a statutory forum depends on their status, and people injured by contamination of a common resource frequently fit none of the definitions.

The Commission. An inquiry can establish and publish the facts, which is disproportionately valuable where the central problem is that nobody outside the company knows what happened.

The realistic assessment. The strongest position is the constitutional proceeding used to establish the facts and to obtain remediation, with compensation pursued behind it. Anyone who says the villagers have a good claim in tort and stops there has not understood what the last forty years of this litigation demonstrate.

What it does NOT mean

Absolute liability does not remove causation. It removes fault. The claimant must still show that this harm came from this activity, and in latent, multi-cause illness that is the whole battle.

A constitutional remedy is not a substitute for compensation. M.C. Mehta itself sent the compensation claims to the ordinary courts.

A settlement is not necessarily a remedy. Union Carbide is the demonstration: a settlement reached without the persons affected being heard, and without a re-opener clause if the fund proved inadequate, was the subject of the review.

And a remedy that is not enforced is not a remedy. Article 2(3)(c) makes enforcement a separate obligation, and it is the limb most often omitted from an answer.

Quick revision

• ICCPR article 2(3): three limbs, an effective remedy notwithstanding official capacity; determination by a competent authority and development of judicial remedies; and enforcement of remedies when granted.
• UDHR article 8: effective remedy by competent national tribunals. UNESCO 1997, article 8: just reparation for damage that is a direct and determining result of an intervention affecting the genome. General Comment 25, paragraph 45: effective remedies are part of the duty to fulfil.
• Five Indian routes and where each fails: constitutional (articles 32 and 226; strong on structural relief, weak on individual money, and M.C. Mehta sent compensation to the ordinary courts); civil (absolute liability; defeated by proof, delay and cost, as Union Carbide shows); criminal (highest standard, hard to attribute within a company, compensates nobody; Union Carbide restored the quashed proceedings); statutory (needs a defined status and an existing statute); Protection of Human Rights Act 1993 (inquiry and recommendation, not an enforceable award).
• Four features that defeat remedies: latency, diffusion, opacity, asymmetry.
• Absolute liability removes fault, not causation.

Test yourself

1. Set out the three limbs of article 2(3) of the ICCPR and say which is most often overlooked. First, to ensure that any person whose rights are violated shall have an effective remedy, notwithstanding that the violation has been committed by persons acting in an official capacity. Second, to ensure that any person claiming such a remedy shall have their right determined by competent judicial, administrative or legislative authorities, or by any other competent authority provided for by the legal system, and to develop the possibilities of judicial remedy. Third, to ensure that the competent authorities enforce such remedies when granted. The third is most often overlooked and is the most important in this subject, because a remedy that is pronounced and not enforced leaves the person exactly where they were, and the Covenant treats enforcement as a distinct obligation rather than an incident of the first.

2. Name the four features of technological harm that defeat remedies, and say which element of a claim each defeats. Latency, which defeats limitation and the availability of evidence, since the harm appears long after the exposure and records and defendants may be gone. Diffusion, which defeats the economics of litigation, since each individual injury is too small to justify a claim while the aggregate is very large. Opacity, which defeats pleading and proof, since the material facts about the process, the maintenance or the model lie inside the defendant. And asymmetry, which defeats endurance, since the claimant faces a party with far greater resources over a period measured in years.

3. Explain why absolute liability, though a strong rule, did not deliver in Bhopal. Because the obstacle was never the standard of liability. M.C. Mehta had been decided in December 1986, two years before the settlement, and it removes fault, removes the exceptions applicable under Rylands v. Fletcher, and measures compensation by reference to the magnitude and capacity of the enterprise so as to deter. What defeated the victims in Union Carbide was procedural: the forum contest that moved the claim between two legal systems and consumed years, the settlement of their claims without their being heard, the absence of a re-opener clause if the fund proved inadequate, the contested medical documentation and categorisation on which each individual award depended, and the passage of time. The case therefore demonstrates that a substantive rule is worth only as much as the procedure available to enforce it.

4. What is the significance of the causal standard in article 8 of the UNESCO Declaration of 1997? Article 8 provides that every individual has the right, according to international and national law, to just reparation for any damage sustained as a direct and determining result of an intervention affecting his or her genome. The standard is significant because it is demanding in precisely the field where causation is at its most difficult. Genetic harm is characteristically latent, probabilistic and multi-causal, so a requirement that the intervention be both the direct and the determining cause will exclude many claims that the rest of the Declaration's protections appear to contemplate. It illustrates the general point of the chapter, that the substantive protection and the remedial standard are drafted separately and can be pitched at different levels.

5. What did the Court do in M.C. Mehta about the compensation claims, and what does that show? Having declined to decide whether Shriram was an authority within the meaning of article 12, the Court did not set up a special machinery to investigate the claims of those affected by the oleum gas escape. It directed the Delhi Legal Aid and Advice Board to take up their cases and to file actions on their behalf in the appropriate court within two months, directed the Delhi Administration to provide the necessary funds for filing and prosecuting them, and asked the High Court to nominate judges so that they could be disposed of expeditiously. It shows that even a constitutional court laying down a far-reaching rule of liability routed the individual compensation claims back into ordinary civil litigation, which is why the constitutional route is strong for structural relief and for extracting facts, and weaker as a means of putting money into the hands of a particular injured person.