PhilSci-L01 - Introduction and Logical Empiricism

Lecture 1: Introduction and Logical Empiricism

Overview

This lecture situates philosophy of science at the intersection of philosophy, modern logic, and modern physics. Its central historical case is logical empiricism, the movement associated with the Vienna Circle. The movement tried to make philosophy rigorous by tying meaningful factual claims to empirical testability and by treating logic and mathematics as analytically true. Its account of scientific theories begins with a syntactic picture: a theory is an axiomatized language connected to observation. The later semantic picture instead takes a theory to be a class of mathematical models representing possible physical systems.

The enduring questions are: What distinguishes science from non-science? What makes a claim meaningful? How do unobservable theoretical terms such as electron acquire empirical content? And should a theory be understood primarily as sentences or as models?

Lecture argument

Logical empiricism made science, observation, and formal logic the starting points for philosophy. Carnap’s early verificationist programme rejected metaphysics as pseudo-statement, but later developments in semantics and theory interpretation weakened the sharpness of that rejection. The syntactic and semantic conceptions of theories illuminate different aspects of science, and contemporary work questions whether their difference is ultimately substantial.

1. Historical roots of philosophy of science

Classical sources: reason, experience, and logic

Two classical orientations frame a recurring dispute.

• Rationalism, associated here with Plato, holds that reason can yield substantive knowledge independently of sense experience. Mathematics, axioms, clear and distinct ideas, and deduction are central. Rationalists seek certainty and necessity in at least some domains.
• Empiricism, associated here with Aristotle and later Locke, Berkeley, Hume, and van Fraassen, holds that experience and sense perception are the fundamental source of substantive knowledge about the world. Induction is central to inquiry. Empirical knowledge is normally contingent and fallible rather than certain.

Plato is presented as a rationalist. Aristotle is presented as an empiricist and as the source of syllogistic logic. A syllogism is a rule-governed deductive argument in which a conclusion follows from premises, classically through terms such as “all humans are mortal” and “Socrates is human.”

These labels do not settle every issue. Empiricism is compatible with idealism, and nominalism is compatible with realism about physical objects. It is therefore essential to distinguish epistemological from metaphysical positions.

Domain	Position	Core claim
Metaphysics	Realism about objects	Objects exist independently of minds and our knowledge; knowledge is constrained by, and aims to represent, them.
Metaphysics	Idealism	Objects, or objects as known, depend in some significant way on mind, experience, or conceptual activity.
Metaphysics	Realism about ideas	Universals, ideas, or general concepts correspond to mind-independent entities. Plato is the standard example.
Metaphysics	Nominalism	Only individual things exist; general terms do not correspond to mind-independent universals. Generality is explained by names, conventions, or mental concepts. Ockham is the standard example.
Epistemology	Rationalism	Reason provides at least some substantive knowledge independently of sensory experience.
Epistemology	Empiricism	Experience is the fundamental source of substantive knowledge of the world.

Kant’s attempted synthesis

Kant tried to combine empiricism about the content of knowledge with rationalism about its a priori conditions. His key claim is that some judgments are synthetic a priori:

• A judgment is synthetic if it extends knowledge, rather than merely unpacking what is already contained in a concept.
• A judgment is a priori if its justification does not derive from particular experiences.

For Kant, space, time, and the categories are conditions under which objects can be experienced by us. This is not the simple empiricist claim that all knowledge comes from experience, nor the simple rationalist claim that experience is dispensable. It says that experience has an a priori structure supplied by the conditions of possible experience.

Modern logic and language

Late nineteenth- and early twentieth-century logic gave philosophy of science a new formal toolkit.

• Gottlob Frege developed modern first-order logic, a formal language with variables, quantifiers such as $\forall$ and $\exists$, predicates, and rules of inference.
• Alfred North Whitehead and Bertrand Russell pursued the logicist project of deriving mathematics from logic.
• Ludwig Wittgenstein shaped the philosophical study of the relation between language and reality. His later slogan, “meaning as use,” stresses that a word’s meaning is connected to its role in practices rather than merely to an inner mental item.

The logical empiricists inherited the idea that logic is tautological or analytic: logic imposes no new factual content about the world. A logical truth is true in virtue of logical form and rules, unlike an empirical claim, whose truth depends on how the world is.

Modern physics and the emergence of the field

The revolutionary sciences of the period made foundational questions unavoidable: special relativity (1905), general relativity (1915), and quantum mechanics (developed roughly 1900 to 1927, associated with Bohr, Heisenberg, and Schrödinger). Philosophy of science grew from the need to understand concepts, evidence, laws, explanation, and theory change in sciences of this kind.

The twentieth-century division between analytic and continental philosophy is connected to a disagreement over what grounds objective knowledge: the priority of logic and analysis, or fundamental philosophical insight. Carnap, working in Jena, became a central analytic philosopher; Martin Heidegger, working in Freiburg, became a central continental philosopher. Their radically different approaches to metaphysics will matter below.

2. The Vienna Circle and the development of philosophy of science

The Vienna Circle was an intellectual group active from 1924 to 1938. Key figures included Rudolf Carnap, Otto Neurath, Moritz Schlick, and Kurt Gödel. Its early programme is commonly called logical positivism; the later and broader label is logical empiricism.

The Circle’s programme combined empiricism with modern symbolic logic. Its regulative ideal was that legitimate knowledge should be either empirically grounded or logically/mathematically necessary, and that philosophy should clarify scientific language and reasoning rather than compete with science by making speculative factual claims.

The Circle was destroyed as a local institution by political violence and fascism. Schlick was murdered by his former student Johann Nelböck in 1936. Following the Nazi annexation of Austria in 1938, its members dispersed: Carnap travelled through Prague to Chicago in 1936; Neurath travelled through The Hague to London in 1940; Popper went to Cambridge in 1937; and Gödel went to Princeton in 1938.

From justification to scientific practice

The first half of twentieth-century philosophy of science focused mainly on a normative or regulative ideal: how can scientific knowledge be justified, evaluated, and made rigorous? In the second half, the focus expanded toward science as actually practiced: What do scientists do? How do historical, social, technological, and institutional conditions shape inquiry?

Important subsequent currents include Popper, Quine and the Duhem-Quine underdetermination problem, Kuhn on scientific progress, Lakatos, work on explanation, science and technology studies, feminist philosophy of science, postcolonialism, and later work by Nancy Cartwright and Margaret Morrison. Recent philosophy also reconnects logical-empiricist themes with scientific practice, especially through models, explanation, and scientific understanding.

Six recurring course debates are:

1. What is science?
2. What is scientific progress, especially in Kuhn’s account?
3. Is theory choice underdetermined by evidence, as the Duhem-Quine thesis suggests?
4. What are explanation and understanding?
5. Is scientific realism defensible?
6. What are laws of nature?

Core vocabulary

• Metaphysics: the study of the nature of things or being. Typical questions concern space and time, causation, freedom and determinism, and the relation between mental and physical.
• Ontology: the branch of metaphysics concerning being as such, especially what entities there are.
• Epistemology: the study of knowledge and understanding.
• Semantics: the study of meaning, including meaning in formal and mathematical languages.
• Syntax: the formal rules governing the construction and transformation of expressions in a language, including rules of deduction. Syntax abstracts from what expressions mean.

3. Carnap: meaning, verification, and the rejection of metaphysics

The verificationist criterion in early logical empiricism

Carnap’s early view starts from the distinction between two kinds of meaningful statement:

1. Analytic or logically necessary statements, including logic and mathematics. They are true in virtue of rules, definitions, or logical form, and do not add factual information about the empirical world.
2. Empirically verifiable statements, whose truth can in principle be checked through observation.

The associated verificationist principle of meaning says, roughly, that a factual statement is meaningful only if it is empirically verifiable. The Vienna Circle’s thought is not merely that evidence is useful. It is that an alleged factual statement with no possible observational bearing fails to state a genuine fact at all.

For early Carnap, philosophy’s task is therefore logical or linguistic analysis. It should identify the forms of valid reasoning and distinguish scientifically meaningful discourse from merely grammatical but meaningless pseudo-discourse.

Carnap initially worked syntactically. He studied forms of expression, deduction, and rule-following, while treating meanings as matters for psychology or empirical science. This historical point matters: fully formal semantics developed only in the late 1930s and 1940s, particularly through Alfred Tarski. Carnap’s early efforts to define meaning consequently move through several notions that later philosophy carefully separates.

Carnap characterizes the meaning of a word through an elementary sentence $S$ containing it, by appeal to:

1. what is deducible from, or what entails, $S$;
2. the conditions under which $S$ is true;
3. the conditions under which $S$ can be verified; and
4. the meaning of $S$ itself.

This invites a critical question: are deduction, truth, verification, and meaning really the same thing? They are not obviously identical. Deduction concerns logical consequence, truth concerns satisfaction by the world or a model, verification concerns evidential procedures, and meaning concerns what an expression says. The early programme needs strong assumptions to connect them.

Reduction to observation: the arthropod example

Carnap’s account of a scientific theory is a set of sentences deduced from observation sentences. Observation verifies the truth of those observation sentences. A term is meaningful when sentences containing it can be reduced to observation sentences.

Consider the elementary sentence:

$x\text{ is an arthropod}.$

Its empirical content can be articulated by observable characterizations:

$x\text{ is an animal},x\text{ has a segmented body},x\text{ has jointed legs/appendages},x\text{ has an exoskeleton}.$

An arthropod is an animal with an exoskeleton, a segmented body, and jointed appendages. An exoskeleton is a hard outer shell made of chitin; a segmented body is divided into segments; jointed appendages are limbs with joints, such as legs or antennae. The example illustrates the reductionist aspiration: explain theoretical or classificatory vocabulary by linking it to observationally checkable conditions.

Carnap against metaphysics

Carnap argued that ordinary language can mislead us through type confusions. Expressions may be grammatically well formed but fail to translate into a logically correct form. He claimed that pseudo-statements occur especially often in certain metaphysical writing, including Hegel and Heidegger, where predicates appropriate to one kind of entity are applied to predicates, being, or existence itself.

His target example comes from Heidegger’s 1929 inaugural lecture What Is Metaphysics?: “Das Nichts selbst nichtet,” conventionally rendered “The Nothing itself noths” or “nihilates.” Carnap’s logical diagnosis is that “nothing” should not be treated as a name of a thing, nor as a verb. In modern logic, “nothing is $P$” is expressed by:

$\neg \exists xP(x),$

meaning: it is not the case that there exists an $x$ that is $P$. The word “nothing” does not here refer to a peculiar object called Nothing that could perform an action.

Similarly, existence is not an ordinary activity or property of an individual. The logically appropriate form for “something is $P$” is:

$\exists xP(x).$

Carnap applies this to Descartes’s “I think, therefore I am.” “I think” can be rendered as “there is something that thinks,” again $\exists xP(x)$ for an appropriate thinking predicate $P$. On this analysis, a substantial metaphysical ego is not needed in addition to the existence claim. The analysis eliminates a subjective metaphysical component that the grammar tempts us to posit.

The argument is not that every sentence containing “nothing,” “being,” or “I” is false. It is that some apparent metaphysical claims misuse grammar so that they do not express truth-evaluable propositions. Carnap’s early conclusion is therefore semantic and methodological: philosophy should expose such pseudo-statements rather than treat them as deep discoveries.

Carnap’s later moderation

In 1966, Carnap described his early work as a reaction to the climate of German idealism and acknowledged that it contained many prohibitions suited to that historical situation. He later relaxed strict verificationism, accepted interpretation of theoretical terms through a fuller semantics, and allowed partial interpretation rather than demanding complete reduction to observations.

This change makes logical empiricism more hospitable to theoretical science and closer to scientific realism. It also weakens the old anti-metaphysical diagnosis: once theoretical terms are semantically interpreted and ontology cannot simply be avoided, the line between meaningful theory and metaphysics becomes less brutally sharp. Quine’s revival of analytic metaphysics from 1948 onward is part of this later landscape. Carnap nevertheless retained the methodological demand that philosophy be logical, rigorous, and close to science.

4. What is a scientific theory?

The syntactic conception, or received view

The syntactic conception identifies a scientific theory with a collection of sentences formulated in an appropriate formal logical language. It is called syntactic because it emphasizes a vocabulary, formation rules, and inferential rules, including deduction.

Its structure is:

1. Axioms: theoretical starting points, interpreted as basic laws.
2. Derived sentences: further claims derived from the axioms by the formal rules.
3. Correspondence rules: rules relating theoretical, non-observational terms to observational terms.

The distinction between theoretical terms and observational terms is essential. Terms such as atom and electron are not directly observable, but are meaningful because correspondence rules link them to the empirical world. An observation sentence contains only observational vocabulary. This picture is often called the Received View of Theories, a label associated with Putnam.

The following schematic captures the diagrammatic structure of the view:

axioms (basic theoretical laws)
             |
             | deduction in a formal language
             v
derived theoretical sentences
             |
             | correspondence rules
             v
observation sentences and observable test conditions

This was a normative ideal: it proposed a clear and rigorous standard by which scientific knowledge and scientific progress could be evaluated. It need not describe every laboratory’s actual practice.

Example: kinetic theory and the ideal gas law

Kinetic theory illustrates the syntactic conception. Its axioms include:

• gases are systems of molecules in motion;
• energy and momentum are conserved;
• molecules are elastic and behave according to Newtonian mechanics.

Correspondence rules connect microphysical terms to macroscopic measurable quantities:

• pressure $P$ is the mean force with which molecules strike the container walls;
• absolute temperature $T$ is proportional to mean molecular kinetic energy:

$kT=\frac{2}{3}\left(\frac{1}{2}m{v}^2\right),$

where $k$ is Boltzmann’s constant, $m$ molecular mass, and $v$ molecular speed. Equivalently, the mean kinetic energy per molecule is $\frac{3}{2}kT$.

From the molecular axioms, kinetic theory yields:

$PV=\frac{2}{3}N\left(\frac{1}{2}m{v}^2\right),$

where $V$ is volume and $N$ is the number of molecules. Substitution of the temperature correspondence rule yields the empirical Boyle-Charles ideal-gas law:

$PV=NkT.$

The explanatory pattern is: postulate a microscopic molecular system, derive its consequences formally, and link the derived quantities to observable pressure, volume, and temperature.

The semantic conception

The semantic conception begins from criticisms of identifying theories with axiomatized sentences:

1. Many scientific theories are not axiomatized, and may never be put into a single axiomatized form.
2. One theory can have several equivalent syntactic formulations. It seems wrong to identify the theory with only one verbal or symbolic presentation.
3. Scientists routinely use models, and these models often represent the scientific content more directly than a fixed list of sentences.

Its slogan is: a scientific theory is a collection of models. A model is a structure, typically defined in mathematical or set-theoretic terms, that represents a possible system and satisfies the theory’s constraints. A vector space is a simple example of a mathematical structure: a set equipped with vector addition and scalar multiplication satisfying specified axioms.

Two senses of model must be kept separate:

1. Iconic or representational model: a concrete or visual representation that captures relevant features of what it represents. It can be structurally similar, or isomorphic, to its target. Isomorphism means there is a structure-preserving one-to-one mapping between the relevant parts and relations of model and target.
2. Mathematical or logical model: a formal structure in which a theory’s sentences are true. This is the principal technical sense in the semantic conception.

Kinetic theory under the semantic conception

On the semantic view, kinetic theory is not primarily a set of sentences. It is a class of mathematical models representing possible gases. For a gas of $N$ molecules occupying a region $C\subseteq {\mathbb{R}}^3$ of volume $V$, a microstate can be represented in phase space:

${\Gamma }_N=C^N\times {\mathbb{R}}^{3N}\subseteq {\mathbb{R}}^{6N}.$

Here $C^N$ records the three-dimensional positions of all $N$ molecules, and ${\mathbb{R}}^{3N}$ records their three-dimensional momenta or velocities. Together they give $6N$ coordinates: three position and three momentum coordinates per molecule. Laws of motion and collision specify how a point in phase space evolves.

Statistical mechanics then links these micro-models to thermodynamic quantities:

• $V$ is fixed by the container;
• $P$ is determined by average momentum transfer to the walls;
• $T$ is determined by the equilibrium distribution of molecular energies.

Under the idealizing assumptions of kinetic theory, the class of models yields $PV=NkT$. The difference in emphasis is therefore clear:

Syntactic conception	Semantic conception
Theory is a formal set of sentences.	Theory is a class of mathematical structures/models.
Emphasizes axioms, derivation, and correspondence rules.	Emphasizes representation of physical systems by mathematical structures.
Links theory to observation through a vocabulary distinction and rules.	Links models to target systems and conditions of application.

Does the distinction survive?

The debate remains lively. Halvorson argued that the semantic view, if plausible, is itself syntactic; replies came from Glymour and van Fraassen. Lutz argues that the syntax-semantics debate may not capture significant differences. Frigg’s balanced diagnosis is that the syntactic view is too strict if it requires only formal sentence systems, since science also needs models and natural language, while the semantic view also requires a language to specify structures and their application. The two conceptions may therefore collapse into one another, or at least be complementary descriptions of the same scientific practice.

5. Exam-ready takeaways

1. Logical empiricism combines empiricism with modern logic. It treats empirically testable statements and analytic/logically necessary statements as the legitimate forms of meaningful discourse.
2. Early Carnap’s verificationist programme identifies philosophy with rigorous logical and linguistic analysis and attacks metaphysical pseudo-statements generated by grammatical or type confusions.
3. Carnap’s analysis of “nothing” uses quantification: $\neg \exists xP(x)$ does not posit an object called Nothing. Existence is expressed by the existential quantifier, not by treating existence as an activity.
4. The syntactic conception presents a theory as axioms plus deductive consequences plus correspondence rules that connect theoretical and observational vocabulary.
5. The semantic conception presents a theory as a class of mathematical models representing possible systems. Its motivation is the non-axiomatized and multiply formulated character of actual science.
6. Carnap later softened strict verificationism and accepted fuller semantics and partial interpretation of theoretical terms. This increased realism’s room to breathe and weakened the blanket anti-metaphysical stance.

Further reading from the lecture

• Stanford Encyclopedia of Philosophy: Logical Empiricism and Vienna Circle.
• R. N. Giere and A. W. Richardson, eds. (1996), Origins of Logical Empiricism, Minnesota Studies in the Philosophy of Science, vol. XVI.
• A. W. Richardson, “Introduction: Origins of Logical Empiricism.”
• M. Friedman, “Overcoming Metaphysics: Carnap and Heidegger.”
• P. Galison, “Constructing Modernism: Cultural Location of Aufbau.”
• F. Suppe (1974), The Structure of Scientific Theories.
• R. Frigg (2022), “Models and Theories.”