PhilSci-L02: Kuhn on Scientific Practice

In short

How does science actually change, and if the real story is messy, can science still be rational and objective?

Kuhn: look at the history. Scientists normally work inside a shared framework, a paradigm, solving the puzzles it sets without questioning it. When stubborn failures pile up, the framework can be replaced as a whole, in a revolution. Old and new paradigms see the world differently, so no neutral test settles which is better. Kuhn still insists the choice is rational: shared values such as accuracy and simplicity guide it without dictating it.

Values in science: can science be value-free? Helen Longino says values enter through background assumptions, and objectivity comes from a community criticising each other’s work. Heather Douglas says values may set how much evidence is enough when being wrong is costly.

On the mock exam: Q4 on revolutions as Gestalt switches, Q5 on Longino’s “social character” of knowledge.

Kuhn’s question: what do scientists actually do?

The logical empiricists of L01 asked a normative question (about how things ought to be): how can scientific knowledge be justified? Popper in L01b asked another normative one: what makes a theory scientific, and how should it be tested? Thomas Kuhn, a physicist turned historian of science, asked a descriptive one: how does science actually work, and what do scientists do? His answer, in The Structure of Scientific Revolutions (1962, “SSR” below), made him one of the main opponents of logical empiricism.

He attacked the picture of science as steady accumulation, knowledge growing brick by brick. His alternative rests on four claims, each explained below.

• Science is dynamic: long stable periods, broken by revolutions.
• Progress is not simple accumulation: revolutions throw things away too.
• Observation is theory-laden: what scientists see depends partly on their theory. The logical empiricists needed a theory-neutral observation language to test theories against, and Kuhn says there is none.
• Science is social: frameworks, standards and verdicts belong to communities, not lone individuals.

Paradigms: what a scientific community shares

Kuhn’s central word is paradigm: everything a scientific community takes for granted in order to do its work. That includes a theory, but also methods, tools, standards for a good answer, model problems, and assumptions about what the world is like. Sharing it holds the community together.

The Newtonian paradigm

• Theoretical principles: Newton’s laws of motion and gravitation.
• Methods: Newton’s regulae philosophandi (his rules of reasoning in natural philosophy).
• Tools: differential equations.
• Exemplars (standard solved problems students learn to imitate): the Earth-Sun system; the harmonic oscillator (a mass bouncing on a spring).
• Metaphysical assumptions: absolute space and time.

Exemplars are how a paradigm is passed on: students absorb Newton by solving the mass-on-a-spring problem and its relatives until they see a new problem as “one like that”.

Paradigm

A framework for practising science, shared by the scientific community in a particular discipline. The paradigm holds the community together: science is a social enterprise. Its elements: theoretical principles, methods of investigation, standards, tools, exemplars, metaphysical assumptions.

Normal science: puzzle-solving inside the paradigm

Most of the time scientists do not test their paradigm. They use it. Kuhn calls this normal science. Its problems are puzzles: like a crossword, each is assumed to have a solution under the accepted framework, and what is being tested is the solver’s skill.

Normal science

Research that solves puzzles within the paradigm. The paradigm is accepted unconditionally.

Accepting the paradigm unconditionally makes normal science conservative, and this has two consequences.

Conservatism allows progress. Because the foundations are not reopened every week, researchers can work in great detail. The price: critics of the paradigm are marginalised or pushed out of the community.

A failure reflects on the scientist, not on the paradigm. If a puzzle will not come out, the community assumes the researcher went wrong: “only a poor workman blames his tools.” So normal science looks dogmatic, the opposite of Popper’s picture, in which scientists keep trying to refute their theories.

Popper and Kuhn on astrology. Both deny astrology is a science, for different reasons. Popper’s (L01b): astrologers escaped falsification with vague prophecies. Kuhn’s, from a 1970 reply to Popper outside the course readings: astrology did make predictions that failed, so by Popper’s test it was falsifiable. What it lacked was a puzzle-solving tradition: its failures never became puzzles a community could work on, so it never had normal science. That tradition is Kuhn’s own mark of a science, and his answer to Lakatos’s charge that he has no demarcation criterion.

From anomaly to revolution

Some puzzles refuse to come out. Normally the community blames error or poor technique. But a puzzle that resists the best people for a long time, and clashes with what the paradigm leads them to expect, looks like a fault in the paradigm itself: an anomaly.

When important anomalies pile up, confidence in the paradigm weakens and the paradigm itself becomes a topic of discussion: a crisis. Research becomes extraordinary science, which can end in a revolution, when a new paradigm wins and sets new puzzles. (Kuhn also allows, in a chapter of SSR outside the course reading, that a crisis can end without one: normal science sometimes solves the anomaly after all, or the problem is set aside for later.) The cycle: normal science, anomaly, crisis, extraordinary science, revolution, new normal science.

Kuhn's stages

• Anomaly: a puzzle that resists solution and is inconsistent with the paradigm’s expectations.
• Crisis: anomalies weaken confidence in the paradigm (with social and psychological dimensions).
• Extraordinary science: competing approaches explored, standards of assessment contested.
• Revolution: victory of a new paradigm. The change is non-cumulative: concepts, problems, methods and standards change with the theory.

Chemistry: phlogiston to oxygen

Kuhn reads the history of chemistry in three stages. Alchemy was preparadigmatic: no single framework united the field. In the phlogiston paradigm, associated with Joseph Priestley, burning was the release of a substance, phlogiston, from the burning body. In Antoine Lavoisier’s oxygen paradigm, burning is combination with oxygen.

Lavoisier did more than add a fact. A metal gets heavier when it burns (background from outside the course material): awkward if burning releases something, expected if it absorbs oxygen. So the key substance, the vocabulary of explanation and the reading of the weight gain all changed.

The other standard case is astronomy: Ptolemy’s geocentric world-picture gave way to Copernicus’s heliocentric one. Even “planet” changed meaning: before, the Sun and Moon were planets and the Earth was not; after, the reverse.

Why revolutions are necessary

Kuhn asks two questions about revolutions: why they are needed at all, and why they deserve the name. The first: why can’t new theories simply be added to the old ones? His answer: “cumulative acquisition of novelty is not only rare in fact but improbable in principle.” The argument:

1. Normal science is cumulative, and it succeeds because scientists choose problems they can solve with concepts and instruments close to the ones they already have. It does not aim at novelty.
2. So a genuine discovery can only arrive where the scientist’s expectations about nature and instruments turn out wrong: as an anomaly.
3. New theories could in principle come from three kinds of phenomena:
   • phenomena the paradigm already explains well: these seldom give a motive or a starting point for a new theory;
   • phenomena the paradigm points to but has not worked out in detail: that is normal science’s job;
   • recognised anomalies that refuse to fit: only these give rise to new theories.
4. A theory built to handle the anomalies must predict something different from the old one somewhere; if it agreed everywhere, it would not be needed.

So a new theory cannot simply be added to the old one. It comes from anomalies, and it contradicts the old paradigm somewhere, so one has to replace the other. That replacement is a revolution. Normal science cannot do this job itself: it can only notice, through anomalies, that its foundations have stopped working.

The positivist reply: Newton as a special case of Einstein. The logical positivists (the logical empiricists of L01) answered that Newton’s dynamics can be derived from relativity for slow bodies, so Newton was only restricted, never overthrown. Kuhn’s reply: the derived equations look like Newton’s laws but are not, because their terms mean different things. Newtonian mass is conserved; Einsteinian mass is convertible with energy. The two are measured the same way only at low speeds, and even then they are different concepts (semantic incommensurability, below). Besides, the defence works only by restricting Newton’s theory to cases already confirmed, and a theory that may not go beyond its evidence predicts nothing new: saved this way, no theory could guide research.

Do you agree? Kuhn himself allows that a new theory could in principle sit at a higher level and link existing theories without displacing any; his example is the law of conservation of energy. Critics add that history shows more continuity across revolutions than the argument predicts, as Kuhn’s own earlier history of the Copernican revolution does.

Why arguments between paradigms are circular

The second: why “revolution”? A political revolution changes institutions in ways those institutions themselves forbid, so normal political procedures cannot settle it, and the parties resort to mass persuasion. Likewise, the normal ways of evaluating a theory depend on a paradigm, and the paradigm is what is in dispute.

Kuhn

“Like the choice between competing political institutions, that between competing paradigms proves to be a choice between incompatible modes of community life.”

That is why paradigm debates are circular. When paradigms enter a debate about paradigm choice, “their role is necessarily circular”: each side uses its own paradigm to argue in that paradigm’s defence, because its paradigm supplies the standards it judges arguments by. For instance (an illustration, not Kuhn’s own), a phlogiston chemist and Lavoisier read the same weight gain differently: for one it is a puzzle to explain away, for the other it is the evidence.

Circularity does not make the arguments wrong or useless. It turns them into persuasion: powerful, but not logically or even probabilistically compelling for someone who refuses to step into the circle. “There is no standard higher than the assent of the relevant community.” Lakatos read passages like this as irrationalism (L01b); Kuhn’s reply comes below.

Revolutions as Gestalt switches

Gestalt psychology (German Gestalt, form or whole) studies how people perceive patterns as wholes. Its lesson for Kuhn: observation and interpretation are intertwined. You do not first see lines and then infer a face; you see a face.

The standard example is the duck-rabbit, a drawing that can be seen as a duck facing one way or a rabbit facing the other: the duck’s bill is the rabbit’s ears. Nothing on the page changes when you switch, but what you see does. Kuhn’s phrase: after a revolution, “what were ducks in the scientist’s world before the revolution are rabbits afterwards.”

He makes the comparison for two reasons.

1. A revolution is a radical, sudden, holistic shift. The duck turns into a rabbit all at once, not feature by feature. Likewise a new paradigm makes scientists see the world in a different light as a whole, without a piecemeal accumulation of new facts. Kuhn compares the change to a conversion experience that logic alone cannot force. The suddenness is in each scientist’s switch, which happens all at once or not at all (though not necessarily in an instant). Across the community the change is gradual, as scientists convert a few at a time (Kuhn says this in a later chapter of SSR, outside the course reading).
2. It expresses his view of observation, taken from the Gestalt psychologists: observation is theory-laden. What scientists see, and what they take their observations to show, depends on the paradigm they bring. After a revolution the same data are organised into a different whole.

Two experiments support the second point, and a third complicates it.

• The priming experiment. Half an audience first sees a clear drawing of a young woman, the other half one of an old woman. Then everyone sees the ambiguous drawing “My Wife and My Mother-in-Law”. People tend to see the version they were primed with, some unable to see the other until it is pointed out. The prior image plays the role of a paradigm.
• The playing-cards experiment (Bruner and Postman, used in SSR). Subjects saw cards in brief flashes, some anomalous, such as a red six of spades. At short exposures they identified the odd cards as normal ones without any sense of trouble; with longer exposures they got confused, and most eventually saw them correctly. Kuhn’s moral for discovery: an anomaly is at first not perceived at all, because expectation shapes perception. It is recognised only against that expectation, and with effort.
• The Müller-Lyer illusion. Two equal lines, one ending in outward arrowheads and one in inward ones, look unequal. For Kuhn’s side, perception is no passive copy: the mind organises what it sees. But the illusion survives measuring the lines, so knowing better does not change what you see. Jerry Fodor (outside the course material) used this to argue that perception is partly insulated from belief, against strong theory-ladenness.

Where the analogy stops, by Kuhn’s own account. A viewer of the duck-rabbit can switch back and forth and knows the lines stayed the same. A scientist has no standpoint outside both paradigms to check that the world stayed the same. Kuhn puts it this way: “Though the world does not change with a change of paradigm, the scientist afterward works in a different world.” So he does not mean perception is arbitrary, or that scientists inhabit different physical worlds. He means that no theory-neutral description can settle a revolutionary dispute from outside both sides.

Incommensurability: no common measure

The word comes from Greek mathematics. Two lengths are commensurable if some common unit fits a whole number of times into both. The side and diagonal of a unit square are not: the diagonal is $\sqrt{2}$, which is irrational, so no unit measures both exactly. They are still related; they just lack a common unit.

Kuhn borrows the word for paradigms. The tradition that comes out of a revolution is not only incompatible with the old one but often incommensurable with it.

Incommensurability

Two paradigms cannot be fully measured against each other by a common language or a fixed set of standards.

• Semantic incommensurability: changes of meaning. “Mass” in the Newtonian and the Einsteinian paradigm; “planet” before and after Copernicus.
• Methodological incommensurability: changes of standards, for instance standards of explanation before and after Newton.

Theory-laden observation feeds this. If what scientists observe depends on their paradigm, there is no theory-neutral observation language in which two paradigms could be compared on neutral ground.

Kuhn’s example of the methodological kind: Newton’s gravity was an attraction at a distance with no mechanism behind it, and critics who demanded mechanical explanations rejected it as an occult quality (a hidden power that names an effect without explaining it). Later the standard itself shifted and such a force became an acceptable explanation. The two sides disagreed on what an answer had to look like.

This creates the problem of progress. If meanings and standards change, in what sense is a new paradigm objectively better than the old one?

Kuhn later softened the claim. Incommensurability became local untranslatability: some terms cannot be translated without loss, and there may be no translation that keeps all their inferential relations (what follows from what). “Local” means the failure is confined to a cluster of interdefined terms, such as “mass”, “force” and “energy” in mechanics, while most of the two languages stay intertranslatable. So, like the side and the diagonal, incommensurable paradigms lack a common measure but can still be related and compared.

Is incommensurability overstated? A standard objection: scientists do communicate across paradigms in practice, and both sides recognise some of the same empirical successes.

Is Kuhn an irrationalist? The critics and his reply

The criticisms:

• No clear demarcation criterion (Lakatos’s charge; Kuhn’s answer is the puzzle-solving tradition above).
• Unclear, metaphorical language, above all “paradigm”: Margaret Masterman counted 21 different uses of the word in SSR.
• Relativism: no objective criteria to judge progress.
• Irrationalism: paradigm victory would be a social victory with no rational basis. Lakatos accused Kuhn of vindicating “mob psychology”, because of the comparison with political revolutions.
• The history is less tidy: a “winning” paradigm never eradicates all its competitors, and Kuhn’s own earlier work stresses more continuity than SSR.

Keep the views apart. The charge that theory change is irrational is Lakatos’s reading of Kuhn. Kuhn denied both relativism and irrationalism, and insisted science is rational. His reply, in “Objectivity, Value Judgment, and Theory Choice” (1977), names five shared values that scientists use to assess theories. He calls them objective criteria, objective in the limited sense that the whole scientific community shares them.

Kuhn's five values of theory choice

1. Accuracy within its domain of application.
2. Consistency, internally and with other accepted theories.
3. Broad scope: consequences beyond the observations it was designed for.
4. Simplicity: it brings order to phenomena.
5. Fruitfulness for new research findings.

Together they provide “the shared basis for theory choice”. They are norms or values, not rules.

The argument:

• The five are individually imprecise, so scientists apply them differently. (How simple is simple?)
• They may conflict: one theory can be more accurate but less simple, another broader but less consistent with accepted theories. So scientists rank them in different orders.
• Therefore scientists committed to the same criteria can choose differently, partly for contextual reasons such as previous experience. Theory choice is not algorithmic: it mixes shared and individual criteria.
• So rational disagreement is possible and legitimate, and progress needs it. If everyone always chose alike, at the same moment, competing possibilities would not be developed or tested.

So the conversion language and the claim of rationality fit together. No algorithm or neutral experiment forces the switch, so it is a matter of persuasion; but the persuasion appeals to shared values, so it is reasoned.

Can science be value-free?

Kuhn’s values raise a bigger question: what role may values play in science at all?

Max Weber, the German sociologist, defended the value-freedom of (social) science with two principles.

• Value relevance: values may guide which research problems are chosen.
• Value freedom: research must not present the researcher’s own commitments as empirical conclusions; empirical claims are assessed by evidential standards.

Behind this is the fact-value distinction: descriptive statements (how things are) versus normative ones (how they ought to be). The logical positivists agreed: value statements express subjective attitudes, not propositions, so they cannot be true or false.

This gives the triptych model of science, three panels side by side. Heather Douglas (a philosopher of science, below) calls it the externality model, because values are kept outside the middle panel.

Research agenda	Scientific research	Application and policy
May involve values	Value-free	May involve values

The controversial question is panel 2: can “pure” science be value-free? Weber and the logical positivists say yes. Many later philosophers say values can guide theory choice. Which are acceptable, and why? Two kinds:

• Epistemic (or cognitive) values serve the cognitive aims of science, such as truth and understanding: accuracy, consistency, explanatory power, broad scope, simplicity, fruitfulness, maybe beauty.
• Non-epistemic values: social desirability, consistency with political preferences, justice, and so on.

Letting epistemic values into research gives a revised triptych, the weak value-free ideal: non-epistemic values may enter the agenda and the applications, and only epistemic values may enter research itself. Kuhn’s five are usually read as epistemic, so they fit this ideal. They still show that theory choice inside panel 2 is not mechanical.

Longino and Douglas go further. They hold internality positions: non-epistemic values can sometimes play a legitimate role within scientific reasoning itself.

Longino: objectivity is social

Helen Longino’s contextual empiricism starts from a point about evidence. Data do not support a hypothesis on their own; they support it only against background assumptions. Those assumptions can include commitments to non-epistemic values. Longino calls such values contextual (from the personal, social and cultural context), as opposed to the constitutive values that come from the goals of science itself.

A drug trial (an illustration from outside the course material)

A trial tests a heart drug only on middle-aged men and measures only heart attacks. That assumes men’s results carry over to women and that heart attacks are the outcome that matters. These assumptions fix what the data are evidence for, and carry judgements about whose outcomes count. No political preference was added to the conclusion, yet values shaped the evidence.

Does that make science subjective? Longino says no, because objectivity is not a property of a single investigator’s mind. No individual is free of background assumptions, and others are better placed to spot ours than we are. So objectivity comes from transformative criticism by a community, which exposes and revises background assumptions.

Longino's conditions for objectivity

The normative criteria for objectivity are the transformative dimension of critical discourse. A community’s discourse is objective to the degree that it meets four necessary conditions:

1. Recognised avenues for criticism.
2. Shared standards that critics can invoke.
3. Responsiveness to criticism: the community changes its beliefs or practices in response.
4. Equality of intellectual authority among qualified members.

Each condition is necessary: a community that ignores criticism, or shuts qualified members out, falls short however careful its scientists are. And objectivity comes in degrees, depending on how deep and wide the critical scrutiny goes.

Together, this is what Longino means by the social character of scientific knowledge. Knowledge becomes scientific through the participation of two or more individuals, in a process of critical emendation (correction) and modification of the individual processes. Each step of that process is transformative. A claim becomes scientific knowledge by surviving this process: once it has been criticised from several points of view and modified in response, it is free of the subjective preferences of any one investigator. The social mechanisms that do this:

• Replication of experiments, with variations.
• Peer review, which brings another point of view on the phenomena, so the authors interpret the phenomena free of subjective preferences.
• Treatment after publication: others refine the new ideas and techniques.
• Absorption into the body of knowledge: later citation, use and modification by others.

Contextual empiricism is not an ethical theory

It is a theory of scientific objectivity and method. Whether research is ethically permissible is for ethical theories (harms and benefits, rights and consent, fair distribution of risk). Longino is at most indirectly relevant, by encouraging scrutiny of research assumptions and of whose perspectives are included in criticism.

Douglas: values set the bar for evidence

Heather Douglas distinguishes two roles for values.

• Direct role: values determine a decision or course of action, such as which problem to fund. This is compatible with the externality picture.
• Indirect role: when evidence is uncertain, the consequences of error may legitimately influence the evidential threshold for accepting or rejecting a claim. This is where inductive risk comes in: the risk of being wrong when a claim goes beyond the evidence.

An illustration from outside the course material: scientists must say, on uncertain data, whether a chemical causes cancer. Wrongly calling it safe risks illness; wrongly calling it dangerous costs jobs. How much evidence is “enough” depends on how bad each error would be, which needs non-epistemic as well as epistemic values to weigh.

Values do not thereby become evidence for the hypothesis: they do not make the chemical more or less carcinogenic, they set how high the bar is. Because this indirect role operates inside science, Douglas holds an internality position.

Where values enter

Values…	View	Role
set the agenda or applications	Value-free ideal	Outside research
guide theory choice	Kuhn, weak ideal	Epistemic values inside
set a threshold under uncertainty	Douglas	Indirect, inside
shape background assumptions	Longino	Contextual, inside the evidence

The difference between the last two: for Douglas, values decide how much evidence is enough. For Longino, they help decide what counts as evidence in the first place: which variables are measured, which models tested, which comparison groups chosen, what counts as explanation.

What to be able to write

• Kuhn’s cycle, with the Newtonian paradigm and phlogiston to oxygen.
• Why revolutions are necessary (a new theory must contradict and replace the old) and why paradigm debates are circular.
• Mock Q4: both reasons for the Gestalt comparison, an experiment as support, and the consequence: no theory-neutral observation language, hence incommensurability.
• Incommensurability, semantic and methodological; local untranslatability.
• Kuhn’s reply to the irrationalism charge: five values, imprecise and conflicting, so choice is rational but not algorithmic.
• The triptych and the weak value-free ideal.
• Mock Q5: two or more individuals, critical emendation, surviving it, the four mechanisms; for depth, the four conditions.
• Douglas versus Longino on where values enter.

Full model answers for Q4 and Q5 are the first two exam questions below.

Exam questions

Mock exam Q4 (10 points, half a page): Thomas Kuhn compares scientific revolutions, which consist in a change of paradigm, to visual "Gestalt switches", in which "what were ducks in the scientists' world before the revolution are rabbits afterwards". Explain why Kuhn makes this comparison.

Key points: revolution is a radical, sudden, holistic shift, like a conversion; observation is theory-laden (Gestalt-inspired); after a revolution the same data form a different whole; support with priming or playing-cards experiment; no theory-neutral observation language (incommensurability).

Follow the model answer (two reasons, both needed):

1. A revolution is a radical, sudden shift. Like a conversion experience, it makes one see the world in a different light all at once, without piecemeal accumulation of new facts. The duck becomes a rabbit as a whole.
2. It expresses his view of observation, inspired by the Gestalt psychologists: observation is theory-laden. What scientists see, and what they take their observations to show, depends on the paradigm they bring to it, so after a revolution the same data are organised into a different whole.

Support with the young woman / old woman priming experiment or the Bruner and Postman playing-cards experiment, and name the consequence: there is no fully theory-neutral observation language to adjudicate between paradigms (incommensurability).

Losing marks: giving only one of the two reasons.

Mock exam Q5 (10 points, half a page): What does Helen Longino mean by the statement that objective scientific knowledge has a "social character"? How does something become "scientific knowledge"? Give some examples of social mechanisms that secure objectivity.

Key points: two or more individuals; critical emendation and modification of the individual processes, each step transformative; knowledge is what survives this, free of subjective preferences; replication and peer review; treatment after publication and absorption into knowledge.

Follow the model answer:

1. Social character. Knowledge becomes scientific through the participation of two or more individuals, in a process of critical emendation and modification of the individual processes. Each such step is transformative.
2. How something becomes scientific knowledge. By surviving this process: a claim criticised from several points of view and modified in response is free of the subjective preferences of any single investigator. Objectivity is a property of the community’s practice.
3. Mechanisms: replication of experiments with variations; peer review, which brings another point of view on the phenomena so the interpretation is free of subjective preferences; treatment after publication (refining new ideas and techniques); absorption into knowledge (subsequent citation, use and modification by others).

For depth: tie it to contextual empiricism (data support a hypothesis only against background assumptions, which may be value-laden, so an individual cannot guarantee objectivity alone) and list the four conditions for transformative criticism: recognised avenues for criticism, shared standards, responsiveness to criticism, equality of intellectual authority.

Describe Kuhn's model of scientific development, illustrate it with one historical case, and explain why it challenges the idea of cumulative progress.

Key points: paradigm, normal science, anomaly, crisis, extraordinary science, revolution; one case (phlogiston to oxygen, or Ptolemy to Copernicus); revolutions are non-cumulative; theory-ladenness and incommensurability block neutral comparison; Kuhn’s five shared values as reply.

• Cycle: a shared paradigm organises normal science (puzzle-solving, paradigm accepted unconditionally); a puzzle that persistently resists solution becomes an anomaly; anomalies weaken confidence and cause a crisis; in extraordinary science rival approaches are explored and standards are contested; a crisis can end in a revolution, which installs a new paradigm that sets new puzzles, and normal science resumes.
• Case: chemistry went from a preparadigmatic stage (alchemy) to the phlogiston paradigm (Priestley: burning releases phlogiston) to the oxygen paradigm (Lavoisier: burning is combination with oxygen). The key substance, the explanatory vocabulary and the reading of the weight gain all changed. Alternatively Ptolemy (geocentric) to Copernicus (heliocentric), where the meaning of “planet” changed.
• Against cumulative progress: a revolution is non-cumulative. Only anomalies give rise to new theories, and a new theory must clash with the old one somewhere, so one displaces the other (“cumulative acquisition of novelty is not only rare in fact but improbable in principle”). Observation is theory-laden, and paradigms differ in meanings (semantic incommensurability) and standards (methodological incommensurability), so no neutral algorithm compares them and it becomes unclear in what sense the new paradigm is objectively better.
• Close with Kuhn’s reply: five shared values keep theory choice rational, though not algorithmic.

Why does Kuhn hold that scientific revolutions are necessary, and why are arguments in paradigm debates necessarily circular?

Key points: normal science does not aim at novelty, so novelty arrives as anomaly; only recognised anomalies give rise to new theories; new theory must be incompatible with the old; Newton is not a special case of Einstein (mass changes meaning); evaluation standards depend on the paradigm in dispute, so each side argues from its own, which makes argument persuasion.

• Necessity: normal science is cumulative because it picks problems solvable with existing tools; it does not aim at novelty. Discovery comes only where expectations fail, as an anomaly. Of three kinds of phenomena (already explained, which seldom prompt a new theory; indicated by the paradigm but needing articulation; recognised anomalies), only anomalies give rise to new theories. A new theory must predict differently from the old one somewhere, so it contradicts the old paradigm and must replace it. Hence “cumulative acquisition of novelty is not only rare in fact but improbable in principle”. Revolutions are how science corrects foundations that normal science cannot touch.
• Against the positivist escape: Newton derived from Einstein at low speeds is not Newton, since Newtonian mass is conserved and Einsteinian mass is convertible with energy; and a theory restricted to past observations predicts nothing new.
• Circularity: as with political revolutions, which change institutions in ways the institutions forbid, the normal procedures cannot decide, because they depend on the paradigm in question. “Like the choice between competing political institutions, that between competing paradigms proves to be a choice between incompatible modes of community life.” Each group uses its own paradigm to argue in its defence.
• Consequence: the arguments are not wrong or useless, but they are persuasion, not logically compelling to someone outside the circle; there is no standard higher than the assent of the relevant community. Lakatos read this as irrationalism (“mob psychology”); Kuhn replied with shared values.

Kuhn is often accused of relativism and irrationalism. State the criticisms and explain how Kuhn's account of theory choice answers them.

Key points: irrational social victory, relativism, unclear “paradigm”, no demarcation criterion; five shared values: accuracy, consistency, broad scope, simplicity, fruitfulness; values, not algorithmic rules: imprecise and can conflict; rational disagreement is legitimate and needed for progress.

• Criticisms: paradigm victory is an irrational social victory; no clear demarcation criterion; “paradigm” is unclear (Masterman counted 21 uses in SSR); no objective criteria for progress, hence relativism; Lakatos: Kuhn vindicates “mob psychology” by comparing science with political revolutions.
• Reply (“Objectivity, Value Judgment, and Theory Choice”, 1977): Kuhn denies both relativism and irrationalism and insists science is rational. Five shared criteria, accuracy, consistency, broad scope, simplicity, fruitfulness, are “the shared basis for theory choice”. They are objective in the sense that the scientific community shares them.
• They are values, not algorithmic rules: each is imprecise, and they can conflict (more accurate but less simple), so scientists weigh them differently. Every choice mixes shared and individual criteria, and previous experience counts.
• Hence rational disagreement is possible and legitimate, and it is needed for progress: if everyone switched at once, competing possibilities would not be developed or tested.
• On demarcation, Kuhn’s own criterion is a puzzle-solving tradition (astrology fails because its failures never became puzzles for a community to work on).

Can science be value-free? Compare the value-free ideal with the positions of Kuhn, Douglas and Longino.

Key points: value-free ideal: values only in agenda and application; weak ideal: epistemic values allowed in research; Kuhn: epistemic values, theory choice not mechanical; Douglas: indirect role via inductive risk; Longino: values in background assumptions, objectivity through transformative criticism.

• Value-free ideal (Weber, logical positivists): values may enter agenda setting and application (panels 1 and 3 of the triptych), never scientific research itself (panel 2).
• Weak value-free ideal: non-epistemic values allowed in panels 1 and 3, only epistemic values in panel 2.
• Kuhn: his five values are usually read as epistemic, and they show that internal theory choice is not mechanical.
• Douglas: values have an indirect internal role where there is inductive risk: under uncertainty, the consequences of error set how much evidence is required. Values do not become evidence.
• Longino: values can enter the background assumptions that connect data to hypotheses (what is measured, which models, which comparison groups, what counts as explanation). Objectivity is secured socially, by transformative criticism.
• Douglas and Longino are internality positions: non-epistemic values can legitimately play a role inside scientific reasoning.

Flashcards

What question does Kuhn put in place of logical empiricism's focus on the justification of scientific knowledge?

How does science actually work: what do scientists actually do, in real historical context?

What is a Kuhnian paradigm?

The shared framework through which a scientific community practises its science, and which holds that community together.

Besides theoretical principles, what elements does a Kuhnian paradigm include?

Methods, standards, tools, exemplars, metaphysical assumptions.

What does Kuhn call a concrete, standard problem-solution that students learn to imitate?

An exemplar.

What is Kuhn's normal science?

Research within an accepted paradigm that solves the paradigm’s puzzles instead of testing whether the paradigm is true.

What is a puzzle in Kuhn's sense?

A problem whose solution is expected in principle under the accepted framework.

In Kuhnian normal science, who is blamed when a puzzle is not solved?

The scientist, not the paradigm: “only a poor workman blames his tools.”

Why does Kuhn think the conservatism of normal science allows progress?

Because the foundations are not constantly reopened, researchers can work on problems in detail.

How does Kuhn's normal science contrast with Popper's ideal of science?

Kuhn: the paradigm is accepted unconditionally, so science is conservative. Popper: theories face permanent critical testing.

What is the difference between Popper's and Kuhn's reasons for denying astrology is a science?

Popper: astrologers escape falsification. Kuhn: astrology lacks a puzzle-solving tradition, so it never had normal science.

What is a Kuhnian anomaly?

A puzzle that persistently resists solution and appears inconsistent with the paradigm’s expectations.

What is a crisis in Kuhn's model?

Important anomalies undermine confidence in the paradigm, which itself becomes a topic of discussion.

What happens during Kuhn's extraordinary science?

Researchers explore competing approaches, and even the standards for judging theories become contested.

What is a scientific revolution for Kuhn?

A non-cumulative change in which a new paradigm replaces the old one.

Give Kuhn's cycle of scientific development in order.

Paradigm and normal science, anomaly, crisis, extraordinary science, revolution (then new normal science).

Why does Kuhn count the shift from phlogiston to oxygen as a revolution?

The framework changed: the key substance, the explanatory vocabulary and the reading of weight gain. It was more than a new fact.

What is the difference between the phlogiston and the oxygen accounts of combustion?

Phlogiston: release of phlogiston from burning bodies. Lavoisier: combination with oxygen.

Why does Kuhn hold that cumulative acquisition of novelty is improbable in principle?

Normal science does not aim at novelty, so novelty arrives only as an anomaly that the paradigm cannot absorb.

Why must a new theory displace the old one, according to Kuhn?

It must predict differently somewhere, or it would not be needed, so the two are incompatible.

Why is Newton's dynamics not a special case of Einstein's, according to Kuhn?

The terms change meaning: Newtonian mass is conserved, Einsteinian mass is convertible with energy.

Why does Kuhn call paradigm change a revolution?

Like a political revolution, it cannot be settled by normal procedures, because those depend on the paradigm in dispute; so the sides turn to persuasion.

Why are arguments in paradigm debates necessarily circular, for Kuhn?

Each side argues from its own paradigm, which supplies the standards for judging arguments.

Does circularity make arguments between paradigms worthless, for Kuhn?

No: they can be effective persuasion, but cannot be logically compelling to someone outside the circle.

Why does Kuhn compare a revolution to a Gestalt switch, as regards how the change happens?

Each scientist’s switch is sudden and holistic, like a conversion, not a piecemeal accumulation of new facts.

Which view of observation does Kuhn's Gestalt-switch comparison express?

Observation is theory-laden: what scientists see depends on the paradigm they bring to it.

Where does Kuhn's Gestalt-switch analogy break down?

The viewer knows the lines stayed the same, but a scientist has no external standpoint to check the world stayed the same.

What does the "My Wife and My Mother-in-Law" priming experiment show about paradigms?

The prior image acts like a paradigm: the same stimulus is seen as a different whole depending on what the viewer brings.

What does Kuhn take the playing-cards experiment to show about discovery?

Anomaly is at first not perceived: it is recognised only against the background of expectation, and only with effort.

Does Kuhn's theory-ladenness mean scientists in different paradigms live in different physical worlds?

No. The world does not change; what an observation is taken to show depends partly on the paradigm.

What is Kuhn's incommensurability?

Two paradigms cannot be fully measured against each other by a common language or fixed set of standards.

How does theory-laden observation feed incommensurability?

There is no theory-neutral observation language in which to compare two paradigms.

What is semantic incommensurability?

Meanings change across paradigms: “mass” in Newtonian and Einsteinian physics, “planet” across the Copernican shift.

What is methodological incommensurability?

Standards change: what counts as a satisfactory explanation before Newton may differ after Newton.

What problem about progress does incommensurability create?

If meanings and standards change, in what sense is a new paradigm objectively better?

What did Kuhn later soften incommensurability into?

Local untranslatability: some terms cannot be translated without loss of their inferential relations.

Why does the "local" in local untranslatability leave room for comparing paradigms?

The failure is confined to a cluster of interdefined terms (mass, force, energy), and most of the two languages stay intertranslatable.

What are the main criticisms of Kuhn?

Irrational social victory, no demarcation criterion, unclear “paradigm”, relativism (no objective criteria for progress).

What does Lakatos accuse Kuhn of vindicating?

“Mob psychology”, by comparing science with political revolutions.

Did Kuhn hold that theory change is irrational?

No. That is Lakatos’s reading. Kuhn denied it and insisted science is rational.

What are Kuhn's five values of theory choice?

Accuracy, consistency, broad scope, simplicity, fruitfulness.

In what sense are Kuhn's five values objective?

They are shared by the scientific community.

Why do Kuhn's values not work as an algorithm for theory choice?

Each is imprecise and they can conflict, so scientists weigh and rank them differently.

Why does Kuhn think rational disagreement is needed for progress?

If everyone switched at once, competing possibilities would not be developed or tested.

How can paradigm change be a matter of persuasion and still be rational, for Kuhn?

No algorithm forces it, but the persuasion appeals to shared values, so it is reasoned.

What is the difference between Weber's value relevance and value freedom?

Value relevance: values may select research problems. Value freedom: commitments must not be presented as empirical conclusions.

What did the logical positivists hold about value statements?

They express subjective attitudes, not propositions, so cannot be true or false.

In the triptych model of science, which panel must be value-free?

The middle one, scientific research. Research agenda and application may involve values.

What does the weak value-free ideal allow inside scientific research?

Epistemic values only. Non-epistemic values stay in agenda setting and application.

What is the difference between epistemic and non-epistemic values?

Epistemic serve cognitive aims (accuracy, simplicity). Non-epistemic are e.g. justice, political preferences.

What do internality positions (Longino, Douglas) claim against the value-free ideal?

Non-epistemic values can sometimes legitimately play a role inside scientific reasoning.

What is Longino's contextual empiricism?

Data support a hypothesis only relative to background assumptions.

How, according to Longino, do values enter scientific reasoning internally?

Through background assumptions, which shape what is measured, modelled, compared, interpreted or counted as explanation.

What does Longino call values from the personal, social and cultural context, as opposed to those from the goals of science?

Contextual values, as opposed to constitutive values.

For Longino, what secures objectivity if no individual is free of assumptions?

Transformative criticism within a community, which exposes and revises background assumptions.

What are Longino's four necessary conditions for objectively transformative discourse?

Recognised avenues for criticism, shared standards, responsiveness to criticism, equality of intellectual authority.

Is objectivity all-or-nothing for Longino?

No: it comes in degrees, with the depth and breadth of critical scrutiny.

For Longino, how does a claim become scientific knowledge?

By surviving critical emendation and modification by others, which frees it from any individual’s subjective preferences.

What social mechanisms secure objectivity for Longino?

Replication with variations, peer review, treatment after publication, absorption into the body of knowledge.

Does Longino's contextual empiricism say whether research is ethically permissible?

No. It is a theory of scientific objectivity and method, not an ethical theory.

What is a direct role for values in Douglas's account?

Values determine a decision or course of action, compatible with the externality picture.

What is inductive risk (Douglas)?

The risk of being wrong when a claim goes beyond the evidence.

What is the indirect role of values, for Douglas?

Under uncertainty, the consequences of error set how much evidence is enough to accept or reject a claim.

Do values become evidence for a hypothesis in Douglas's inductive-risk account?

No. They set how much evidence is enough, given how costly each kind of error is.

What is the difference between Douglas's and Longino's accounts of values in science?

Douglas: values set thresholds under uncertainty. Longino: values enter background assumptions that link data to hypotheses.

Links

• Course: Course overview · tutorial questions in PhilSci - Tutorials (Tutorial 2)
• Previous: PhilSci-L01 - Introduction and Logical Empiricism for the logical empiricism Kuhn opposes · PhilSci-L01b - Popper and Lakatos for falsification, Lakatos’s research programmes and his reading of Kuhn
• Next: PhilSci-L03 - Under-determination (Quine’s “On Empirically Equivalent Systems of the World” was assigned at the end of this lecture)
• Source: lecture slides, “Lecture 2: Kuhn’s Scientific Revolutions and Values in Science”, Sebastian De Haro, 8 September 2026
• Readings (Curd & Cover): Kuhn, “The Nature and Necessity of Scientific Revolutions”, pp. 79-93 (the circularity quote is on p. 88, the cumulative-novelty quote on p. 89). Longino, “Values and Objectivity”, pp. 144-164 (the mock model answer cites pp. 175-176 of an older edition).
• Further: Kuhn, The Structure of Scientific Revolutions (1962), Chapter X for the duck-rabbit passage. Kuhn, “Objectivity, Value Judgment, and Theory Choice” (given as a lecture, published 1977 in The Essential Tension). Kuhn, “Logic of Discovery or Psychology of Research?” (1970) for the astrology argument. Masterman, “The Nature of a Paradigm” (1970). Kuhn, The Copernican Revolution (1957). Kahneman, Thinking, Fast and Slow (2011). Stanford Encyclopedia of Philosophy, “Scientific Objectivity”, section on epistemic and contextual values: https://plato.stanford.edu/entries/scientific-objectivity/#EpisContValu