Philosophy of Science
What distinguishes science from pseudoscience? Can scientific theories ever be proven true? How do revolutions in scientific thought occur? This course explores the foundational questions about the nature, methods, and limits of scientific knowledge — from the logical positivists of the Vienna Circle to contemporary debates about realism, explanation, and the social dimensions of science.
Featured Lecture: Philosophy of Science
An introductory overview of the major themes in philosophy of science — the nature of scientific knowledge, the problem of induction, falsifiability, paradigm shifts, and the relationship between theory and observation.
Plato's Timaeus: The Creation of the Cosmos Explained
A detailed exploration of Plato's Timaeus — how the Demiurge shapes the cosmos from chaos using mathematical principles, the Platonic solids as the building blocks of matter, and the creation of the World Soul through harmonic ratios.
Featured Lecture: Foundations of Scientific Reasoning
A complementary lecture exploring the logical and philosophical foundations that underpin scientific inquiry — from deductive reasoning to the challenges of theory choice and scientific progress.
What Is Philosophy of Science?
Philosophy of science examines the foundations, methods, and implications of science. It sits at the intersection of epistemology (the theory of knowledge), metaphysics (the nature of reality), and logic. While scientists ask "How does the world work?", philosophers of science ask "What makes scientific knowledge scientific? How reliable are our methods? What do our theories actually tell us about reality?"
Epistemology of Science
How do we acquire scientific knowledge? What justifies our belief in scientific theories? Is there a logic of scientific discovery, or only a logic of justification? What role do observation, experiment, and theory play?
Metaphysics of Science
Do scientific theories describe reality as it truly is (realism), or are they merely useful instruments for prediction (instrumentalism)? What are laws of nature? Do theoretical entities like electrons "really exist"?
Methodology & Logic
Is there a universal scientific method? How should scientists test hypotheses? What makes a good explanation? How should we interpret statistical evidence? What role does simplicity (Ockham's Razor) play?
Plato, Mathematics, and the Timaeus
Long before the Vienna Circle or Karl Popper, Plato (c. 428–348 BCE) posed a question that remains at the heart of the philosophy of science: Why is mathematics so effective at describing the physical world? His dialogue the Timaeus — the most influential cosmological text in Western history until the Scientific Revolution — offers a breathtaking answer: the universe itself is a mathematical structure, crafted by a divine Craftsman (the Demiurge) according to eternal mathematical Forms.
The Demiurge and Mathematical Creation
In the Timaeus, Plato describes how the Demiurge gazes upon the eternal Forms (perfect, unchanging mathematical-rational archetypes) and imposes mathematical order on formless, chaotic matter (chōra, the "receptacle"). The cosmos is not a random assemblage — it is a living creature endowed with soul and intelligence (30b), structured through precise mathematical ratios.
The Demiurge first constructs the World Soul by blending Being, Sameness, and Difference into a mathematical mixture, then divides it according to a harmonic series: 1, 2, 3, 4, 9, 8, 27 — the powers of 2 and 3. The intervals are then filled with arithmetic and harmonic means, producing the musical ratios (octave 2:1, fifth 3:2, fourth 4:3) that govern both planetary motion and musical harmony. This is the origin of the music of the spheres — the idea that cosmic order is fundamentally harmonic.
The World Soul is then "cut lengthwise into two strips" and bent into circles — the celestial equator (the circle of the Same) and the ecliptic (the circle of the Different) — explaining the apparent motions of the sun, moon, and planets. Time itself is created as "a moving image of eternity," measured by these celestial revolutions.
The Platonic Solids: Geometry as Physics
The most celebrated passage of the Timaeus (53c–56c) presents what may be the first mathematical physics — a theory of matter based entirely on geometry. Plato assigns each of the four classical elements to one of the five regular polyhedra (the "Platonic solids"):
Crucially, Plato goes further — he decomposes the faces of three of the solids (tetrahedron, octahedron, icosahedron) into elementary right triangles (the 30-60-90 and 45-45-90 triangles). Because fire, air, and water share the same triangular building blocks, they can transmute into one another: water (20 faces) can decompose into 2 fire particles (4 faces each) + 1 air particle (8 faces), since 20 = 2(4) + 8 + 4 remaining triangles. Earth (built from a different triangle, the isosceles) cannot transmute into the others — a remarkable prediction of elemental asymmetry.
This is arguably the first reductionist physical theory: macroscopic properties of matter (heat, fluidity, solidity) are explained by the geometric structure of microscopic constituents. The parallel to modern particle physics — where the properties of matter arise from the symmetry group structure of fundamental particles — is striking.
Mathematical Harmony: From Plato to Kepler
The Timaeus established a research programme that lasted two millennia. The idea that nature is structured by mathematical harmony directly influenced:
- Pythagorean-Platonic astronomy: Eudoxus, Plato's student, developed the first mathematical model of planetary motion (homocentric spheres). The requirement to "save the phenomena" by combining uniform circular motions became the central programme of ancient astronomy, culminating in Ptolemy's Almagest.
- Kepler's Mysterium Cosmographicum (1596): Kepler attempted to explain the spacing of the six known planets by nesting the five Platonic solids between their orbital spheres. Though the specific model failed, Kepler's conviction that nature obeys mathematical law led him to discover his three laws of planetary motion — the foundation of Newtonian mechanics.
- Heisenberg and the Platonic solids: Werner Heisenberg, in Physics and Philosophy (1958), explicitly compared modern elementary particle physics to the Timaeus: "In modern quantum theory there can be no doubt that the elementary particles will finally also be mathematical forms, but of a much more complicated nature."
- Modern mathematical Platonism: Roger Penrose, in The Road to Reality (2004), defends a Platonic view in which mathematical truth exists independently of human minds, and the physical world is a "shadow" of mathematical reality — an explicit echo of Plato's Cave allegory and the Timaeus.
The Timaeus and the Philosophy of Science
The Timaeus raises several questions that remain central to the philosophy of science today:
The Applicability Problem
Why is mathematics "unreasonably effective" (Wigner, 1960) at describing the physical world? Plato's answer — that the world was deliberately constructed according to mathematical principles — is one solution. Modern alternatives include structural realism (the world is a mathematical structure), evolutionary epistemology (our mathematical intuitions evolved to track real patterns), and deflationary accounts (mathematics is effective because we develop it for physical applications).
Likely Stories (Eikōs Mythos)
Plato himself insists that the Timaeus offers only a "likely story" (eikōs mythos, 29d) — not certain truth. Physical cosmology, because it deals with the changing world of becoming rather than the eternal world of being, can yield only probable accounts, never demonstrative knowledge. This anticipates the modern distinction between mathematical proof (certain, deductive) and scientific theory (probable, inductive) — and resonates with van Fraassen's constructive empiricism, which holds that science aims for empirical adequacy rather than literal truth.
Teleology vs. Mechanism
The Timaeus presents two kinds of causation: the purposive intelligence of the Demiurge (teleological, "for the sake of the Good") and the brute necessity of material interactions ("the wandering cause"). This dual-aspect framework — intelligent design vs. mechanical necessity — prefigures the tension between teleological and mechanistic explanation that runs through the entire history of science, from Aristotle's final causes through Darwin's natural selection to modern debates about whether biology can be fully reduced to physics.
Mathematical Realism
If the fundamental constituents of matter are geometric forms (as Plato claims in the Timaeus, and as string theorists sometimes echo today), then the line between mathematics and physics dissolves. The Timaeus is the founding document of ontic structural realism — the view that the world is fundamentally constituted by mathematical structure rather than "stuff." This remains one of the most active areas of contemporary philosophy of science (Ladyman & Ross, Every Thing Must Go, 2007).
Key Passages from the Timaeus
"The god, wishing that all things should be good and nothing imperfect as far as possible, took over all that was visible — not at rest but in discordant and disordered motion — and brought it from disorder into order, since he judged that order was in every way better than disorder."— Timaeus, 30a
"For God desired that, so far as possible, all things should be good and nothing evil; and so, finding the whole visible sphere not at rest but moving in an irregular and disorderly fashion, out of disorder he brought order, considering that this was in every way better than the other."— Timaeus, 30a (Jowett translation)
"Now the nature of the ideal being was eternal, but to bestow this attribute fully upon a creature was impossible. Therefore he resolved to make a moving image of eternity, and when he set in order the heaven, he made this image — eternal but moving according to number — while eternity itself rests in unity. And this image we call time."— Timaeus, 37d
"And in the centre he put the soul, which he diffused throughout the body, making it also to be the exterior environment of it; and he made the universe a circle moving in a circle, one and solitary, yet by reason of its excellence able to converse with itself."— Timaeus, 34b
Gödel's Incompleteness Theorems: The Limits of Formal Systems
If Plato's Timaeus represents the ancient conviction that mathematics can fully describe reality, then Kurt Gödel's incompleteness theorems (1931) represent the modern discovery of its inherent limits. Gödel — a quiet, enigmatic member of the Vienna Circle — proved that any sufficiently powerful formal system is either incomplete or inconsistent, shattering the foundational programmes of both Hilbert and the logical positivists.
Lecture: Gödel Incompleteness
A detailed lecture on Gödel's incompleteness theorems — their construction, proof strategy, and devastating implications for the foundations of mathematics and science.
The Historical Context
By the late 19th century, mathematics faced a foundational crisis. The discovery of paradoxes in set theory (Russell's paradox, 1901), the challenge of non-Euclidean geometries, and the need to rigorize analysis all demanded a secure foundation. Three rival programmes emerged:
Logicism (Frege, Russell)
Reduce all mathematics to pure logic. Russell and Whitehead's Principia Mathematica (1910–13) took 362 pages to prove 1+1=2. Gödel showed this programme could never be completed.
Formalism (Hilbert)
Treat mathematics as a formal game of symbol manipulation. Hilbert's programme aimed to prove the consistency of mathematics using finitary methods. Gödel's second theorem showed this is impossible.
Intuitionism (Brouwer)
Mathematics is a mental construction; reject the law of excluded middle and non-constructive proofs. Gödel himself was sympathetic but his theorems apply to intuitionistic arithmetic too.
The Two Incompleteness Theorems
First Incompleteness Theorem (1931)
Any consistent formal system $F$ that is capable of expressing basic arithmetic contains statements that are true but unprovable within $F$.
More precisely: if $F$ is consistent and contains Robinson arithmetic $Q$, then there exists a sentence $G$ (the "Gödel sentence") such that neither $G$ nor $\neg G$ is provable in $F$. The sentence $G$ effectively says: "I am not provable in $F$."
Implication: Mathematical truth outstrips mathematical proof. No single formal system can capture all mathematical truths — there will always be "blind spots."
Second Incompleteness Theorem (1931)
No consistent formal system $F$ that contains basic arithmetic can prove its own consistency.
Formally: if $F$ is consistent, then $F \nvdash \text{Con}(F)$, where $\text{Con}(F)$ is the arithmetized statement "$F$ is consistent."
Implication: Hilbert's programme is impossible. We cannot use mathematics to prove that mathematics is consistent — we must always take something on trust.
The Proof Strategy: Gödel Numbering
Gödel's stroke of genius was to make mathematics talk about itself. He assigned a unique natural number (a "Gödel number") to every symbol, formula, and proof in the formal system, then showed that metamathematical statements ("this formula is provable") could be encoded as arithmetical statements about these numbers.
The key steps: (1) Assign Gödel numbers to all expressions. (2) Show that the relation "$x$ is a proof of $y$" is expressible as an arithmetical predicate $\text{Prf}(x, y)$. (3) Use the diagonal lemma to construct a sentence $G$ that says "there is no $x$ such that $\text{Prf}(x, \ulcorner G \urcorner)$" — i.e., "I am not provable." (4) Show that if $F$ is consistent, then neither $G$ nor $\neg G$ can be proved.
This technique of self-reference through arithmetization became one of the most powerful tools in mathematical logic, later exploited by Turing (the halting problem), Church (the undecidability of first-order logic), and Chaitin (algorithmic information theory).
Philosophical Implications for Science
The Limits of Formalization
If even arithmetic cannot be fully formalized, what hope is there for formalizing all of science? The logical positivists' dream of a unified, fully formalized science is provably unattainable. Every formal scientific theory will have "blind spots" — truths it cannot derive from its own axioms.
Mechanism and Mind
Gödel himself argued (following J.B. Rosser and later Roger Penrose in The Emperor's New Mind) that the theorems show the human mind is not a Turing machine — because we can "see" the truth of the Gödel sentence that the machine cannot prove. This remains deeply controversial: critics (Putnam, Benacerraf, Franzén) argue the argument assumes we are consistent, which we cannot know.
Gödel's Platonism
Gödel was a committed mathematical Platonist — he believed mathematical objects exist independently of human minds, and that the incompleteness theorems confirm this. If mathematical truth outstrips any formal system we can construct, then truth is not reducible to provability, and mathematical reality must be "out there" waiting to be discovered. This echoes the Timaeus and connects Gödel directly to the Platonic tradition.
Implications for Scientific Theories
Stephen Hawking invoked Gödel in arguing that a complete "Theory of Everything" may be impossible: any finite set of axioms capable of describing physics will leave some physical truths unprovable. Freeman Dyson: "Gödel proved that the world of pure mathematics is inexhaustible. I hope that an analogous situation exists in the physical world."
Key Quotes
"Either mathematics is too big for the human mind or the human mind is more than a machine."— Kurt Gödel (attributed, c. 1951)
"The human mind is incapable of formulating (or mechanizing) all its mathematical intuitions. I.e., if it has succeeded in formulating some of them, this very fact yields new intuitive knowledge, e.g., the consistency of this formalism."— Kurt Gödel, Gibbs Lecture (1951)
"Gödel's theorem shows that there can be no single formal system which is both consistent and complete for all of mathematics — but it doesn't show that humans are better than machines, only that both are limited."— Torkel Franzén, Gödel's Theorem: An Incomplete Guide to Its Use and Abuse (2005)
Historical Arc of the Discipline
Philosophy of science as a distinct discipline emerged in the early 20th century, though its roots stretch back to antiquity. The field has undergone its own "revolutions":
Ancient & Early Modern Foundations
Aristotle distinguished episteme (demonstrative knowledge) from doxa (opinion) and developed the first systematic account of scientific explanation through syllogistic reasoning and the four causes. Francis Bacon (1620) championed inductive method in the Novum Organum, arguing that knowledge must be built from careful observation rather than Aristotelian deduction. David Hume (1739) then devastated naive inductivism by showing that no amount of observed regularities can logically guarantee future ones — the problem of induction that remains central to philosophy of science today.
Logical Positivism (1920s–1950s)
The Vienna Circle (Schlick, Carnap, Neurath, Gödel) and the Berlin Circle (Reichenbach, Hempel) developed logical positivism — the view that meaningful statements are either analytically true (logic, mathematics) or empirically verifiable. The verification principle aimed to demarcate science from metaphysics. A.J. Ayer popularized these ideas in Language, Truth and Logic (1936). Logical positivism dominated Anglo-American philosophy of science for decades but eventually collapsed under internal contradictions and external criticism.
Falsificationism (1930s–1960s)
Karl Popper rejected both inductivism and verificationism. In The Logic of Scientific Discovery (1934/1959), he proposed that what makes a theory scientific is not that it can be verified but that it can be falsified. Genuine science makes bold, risky predictions that could be shown wrong. Theories that are unfalsifiable (like Freudian psychoanalysis or Marxist historical materialism, Popper argued) are not scientific. Popper's criterion of falsifiability became the most widely known philosophical account of science.
The Historical Turn (1960s–1970s)
Thomas Kuhn's The Structure of Scientific Revolutions (1962) fundamentally changed the field. Kuhn argued that science does not progress by the linear accumulation of facts but through "paradigm shifts" — revolutionary transitions between incommensurable frameworks. Normal science operates within a paradigm, solving puzzles according to shared rules and exemplars. Anomalies accumulate until a crisis triggers a revolution. Kuhn's work opened the door to historical, sociological, and psychological approaches to understanding science.
Contemporary Philosophy of Science (1980s–Present)
The field has diversified enormously. Major areas include: the realism/anti-realism debate (van Fraassen's constructive empiricism vs. scientific realism); Bayesian epistemology and formal models of confirmation; the new mechanistic philosophy of explanation; philosophy of specific sciences (physics, biology, cognitive science, economics); social epistemology and the role of values in science; feminist philosophy of science; the philosophy of scientific models and simulations; and the demarcation problem revisited (intelligent design, climate change denial, vaccine skepticism).
Course Structure
Part I: What Is Science?
The demarcation problem, the scientific method, observation and theory-ladenness. How do we distinguish science from non-science?
Part II: Logical Positivism & Its Critics
The Vienna Circle, the verification principle, the analytic/synthetic distinction, and the Quine-Duhem thesis.
Part III: Falsificationism
Popper's falsifiability criterion, the asymmetry of verification and falsification, corroboration, and verisimilitude.
Part IV: Scientific Revolutions
Kuhn's paradigm shifts, normal science vs revolutionary science, incommensurability, and Lakatos's research programmes.
Part V: Realism vs Anti-Realism
Scientific realism, instrumentalism, constructive empiricism, the no-miracles argument, and the pessimistic meta-induction.
Part VI: Explanation & Laws
The deductive-nomological model, causal and mechanistic explanation, laws of nature, ceteris paribus clauses.
Part VII: Probability & Confirmation
Bayesian confirmation theory, the problem of induction, the ravens paradox, the grue problem, statistical inference.
Part VIII: Philosophy of Physics
The philosophy of space and time, interpretations of quantum mechanics, symmetry principles, and the arrow of time.
Part IX: Philosophy of Biology
Natural selection and teleology, the species problem, reductionism vs emergence, the gene concept, evolutionary epistemology.
Part X: Science & Society
Values and objectivity in science, social epistemology, the role of trust and expertise, ethics of scientific research, science policy.
Central Questions of Philosophy of Science
The Demarcation Problem
What distinguishes science from pseudoscience, religion, and other forms of knowledge? Popper proposed falsifiability; Kuhn emphasized puzzle-solving within paradigms; Lakatos offered research programmes with progressive and degenerating problem shifts. No single criterion has proven universally satisfactory — the demarcation problem remains open.
"The criterion of the scientific status of a theory is its falsifiability, or refutability, or testability." — Karl Popper
The Problem of Induction
David Hume showed that inductive reasoning — inferring general laws from particular observations — cannot be logically justified. The sun has risen every morning, but this gives us no logical guarantee it will rise tomorrow. Yet science relies fundamentally on induction. How can we justify this?
Responses include: Popper's rejection of induction in favor of conjecture-and-refutation; Reichenbach's pragmatic vindication; Bayesian confirmation theory; Goodman's "new riddle of induction" (the grue problem); and evolutionary naturalized epistemology.
Underdetermination of Theory by Evidence
The Quine-Duhem thesis holds that any body of evidence is compatible with multiple, mutually incompatible theories. A failed prediction can always be blamed on auxiliary hypotheses rather than the core theory. Pierre Duhem showed this for physics; W.V.O. Quine generalized it to all of knowledge.
This raises deep questions about theory choice: if evidence alone cannot determine which theory is correct, what other factors (simplicity, elegance, unification, fruitfulness) legitimately guide scientific judgment?
Scientific Realism
Do mature scientific theories describe reality approximately as it is? The realist says yes: the success of science would be a miracle otherwise (Putnam's no-miracles argument). The anti-realist counters with the pessimistic meta-induction: many past "successful" theories (caloric, phlogiston, luminiferous ether) turned out to be fundamentally wrong. Why should we think our current theories are different?
Bas van Fraassen's constructive empiricism offers a middle path: science aims at empirical adequacy (saving the phenomena) rather than truth about unobservables. We can accept a theory as empirically adequate without believing in its theoretical entities.
Key Thinkers
Karl Popper (1902–1994)
Falsificationism, critical rationalism, the open society. Argued that science progresses by bold conjectures and severe attempts at refutation, not by inductive accumulation of confirming instances.
Thomas Kuhn (1922–1996)
Paradigm shifts, normal science, scientific revolutions, incommensurability. His Structure of Scientific Revolutions is one of the most cited academic books of all time.
Imre Lakatos (1922–1974)
Methodology of scientific research programmes. Synthesized Popper and Kuhn: theories have a hard core protected by auxiliary hypotheses. Progressive programmes make novel predictions; degenerating ones only accommodate known facts.
Paul Feyerabend (1924–1994)
Epistemological anarchism: "anything goes." Argued against universal methodological rules, claiming that every rule has been productively violated in the history of science. Against Method (1975).
Bas van Fraassen (1941–)
Constructive empiricism: science aims at empirical adequacy, not truth about unobservables. The Scientific Image (1980) revitalized anti-realism as a serious philosophical position.
Rudolf Carnap (1891–1970)
Logical positivism, the formal structure of scientific theories, inductive logic, probability as degree of confirmation. The most technically sophisticated of the Vienna Circle philosophers.
W.V.O. Quine (1908–2000)
Demolished the analytic/synthetic distinction ("Two Dogmas of Empiricism"), holistic underdetermination, naturalized epistemology. Made philosophy of science inseparable from the rest of philosophy.
Nancy Cartwright (1944–)
Argued that the laws of physics "lie" — they describe idealized, ceteris paribus situations that never actually obtain. Capacities and causal powers are more fundamental than covering laws.
Carl Hempel (1905–1997)
The deductive-nomological (D-N) model of scientific explanation, the ravens paradox, confirmation theory. A central figure in 20th-century philosophy of science who bridged logical positivism and its successors.
Formal Concepts in Philosophy of Science
Philosophy of science is not purely discursive — many of its central ideas have precise formal expressions:
Bayesian Confirmation
A hypothesis H is confirmed by evidence E if and only if the posterior probability of H given E exceeds the prior probability:
$$P(H|E) = \frac{P(E|H) \cdot P(H)}{P(E)} > P(H) \quad \Leftrightarrow \quad P(E|H) > P(E)$$
Evidence confirms H when it is more likely given H than it is overall. The degree of confirmation is measured by the likelihood ratio $P(E|H)/P(E|\neg H)$.
D-N Model of Explanation
Hempel's deductive-nomological model: an event E is explained by deducing it from laws L and initial conditions C:
$$L_1, L_2, \ldots, L_k \quad \text{(laws)}$$
$$C_1, C_2, \ldots, C_n \quad \text{(initial conditions)}$$
$$\therefore \quad E \quad \text{(explanandum)}$$
The explanans (laws + conditions) must be true, the derivation valid, and the laws essential (not eliminable).
Verisimilitude (Truthlikeness)
Popper proposed that even if no theory is perfectly true, some theories can be "closer to the truth" than others. If T is a theory and Cn(T) is the set of its consequences:
$$\text{Vs}(T) = |T_t| - |T_f| \quad \text{where } T_t = \text{Cn}(T) \cap \text{True}, \quad T_f = \text{Cn}(T) \cap \text{False}$$
Theory A is more verisimilar than B if A has more true consequences and fewer false ones. Tichý and Miller (1974) showed this naive definition is flawed — modern accounts use distance metrics in logical space.
Goodman's "Grue" Paradox
Define "grue" as: an object is grue if it is examined before time t and found to be green, or is not so examined and is blue. All emeralds observed before t are both green and grue. Both "all emeralds are green" and "all emeralds are grue" are equally confirmed by the same evidence. But they make incompatible predictions after t. This shows that not all inductive inferences are equally legitimate — something beyond mere evidence must ground induction.
Essential Reading
Primary Sources
- • Popper, The Logic of Scientific Discovery (1959)
- • Kuhn, The Structure of Scientific Revolutions (1962)
- • Lakatos, The Methodology of Scientific Research Programmes (1978)
- • Feyerabend, Against Method (1975)
- • van Fraassen, The Scientific Image (1980)
- • Hempel, Aspects of Scientific Explanation (1965)
- • Quine, "Two Dogmas of Empiricism" (1951)
- • Hume, An Enquiry Concerning Human Understanding (1748)
Textbooks & Surveys
- • Godfrey-Smith, Theory and Reality (2003) — best introductory text
- • Okasha, Philosophy of Science: A Very Short Introduction (2002)
- • Ladyman, Understanding Philosophy of Science (2002)
- • Rosenberg, Philosophy of Science: A Contemporary Introduction (2012)
- • Bird, Philosophy of Science (1998)
- • Curd & Cover (eds.), Philosophy of Science: The Central Issues (2012)
- • Stanford Encyclopedia of Philosophy — free, peer-reviewed articles on all topics