The Traditional Conception of Science
Historical Foundations of the Traditional Conception of Science
The traditional conception of science refers to the view that dominated from the seventeenth century through the mid-twentieth century, persisting in some contemporary circles.
The seventeenth century marks the origin of modern science, characterized by a distinct shift in how nature was investigated.
This methodological shift was driven by a commitment to ensuring that scientific theorizing was objectively and rationally governed by nature through empirical observation.
The achievements of Galileo Galilei, Johannes Kepler, and Sir Isaac Newton served as a powerful vindication of this empirical-rational approach, establishing the traditional view as standard wisdom.
The Baconian Conception of Science
Early modern scientific methodology was formulated by Francis Bacon () in Novum Organum (Book 2, Section 10).
The Baconian methodology consists of a three-step procedure:
Objective Data Collection: Scientists collect observational data in a purely objective manner, free from all prejudices, prior preferences for specific theories, or surreptitious philosophical and religious presuppositions.
Transparent Data Organization: Collected data are organized in a naturally perspicuous manner without smuggled presuppositions or constraints.
Inductive Generalization: Correct generalizations and explanatory principles emerge out of the organized data through induction. If multiple explanations emerge, additional data collection resolves the issue in favor of a single explanatory principle.
Baconian science preserves three foundational characteristics:
Objectivity: Guaranteed by the total absence of presuppositions or a priori restraints.
Empiricality: Ensured by basing the entire process exclusively on empirical data.
Rationality: Maintained through strict adherence to the logical process of induction.
Inadequacies and Failures of Baconian Inductivism
Failure of Unguided Data Collection:
Collecting data without selection principles yields an unstructured collection of information that is largely unrelated and irrelevant to the subject of study.
Sorting relevant data from irrelevant data requires prior theoretical assumptions about causal mechanisms and related processes.
Example: Investigating tree growth rates across altitudes requires a prior decision that altitude is relevant and that the aesthetic beauty of the trees is irrelevant.
Scientific inquiry cannot proceed without presuppositions, including belief in the uniformity of nature and the consistent operation of causal mechanisms.
Failure of Unstructured Data Organization:
Data do not organize themselves; proper categorization depends on pre-existing or emerging theoretical principles.
Example: Categorizing heat generation remained ambiguous until theoretical principles of heat were discovered. It was unclear whether heat produced in warm puppies was akin to combustion, and solar heat was mistakenly classified as combustion heat for centuries.
Scientists organize data using prior theoretical suppositions regarding relationships and categories.
Failure of the Inductive Step:
Theories and explanatory principles are products of human invention and insight rather than automatic logical consequences of data.
There is no formal algorithm or "logic of discovery" to generate novel concepts or theoretical connections.
Going from observational data to a theoretical account requires an imaginative leap.
Data underdetermine theories: any finite body of observational data can support an infinite number of distinct theoretical explanations.
Graph Analogy: Given a finite set of points on a graph, an infinite number of distinct lines can be drawn through them. Some lines are smooth and simple, while others are complex or jagged like spaghetti. Similarly, theories explaining a set of observations can be simple and elegant or messy, arbitrary, and ad hoc.
Preference for Elegance: The preference for simple and elegant theories represents a philosophical preconception inherited from ancient Greek thought—the belief that nature is fundamentally simple.
Conceptual Limits: Inability to generate multiple explanatory theories reflects human cognitive limits rather than data forcing a single conclusion.
The Mid-Twentieth-Century Traditional View: Rationality
Modern symbolic logic in the early twentieth century influenced philosophy of science by defining rationality within formal logical structures.
Because science was considered the paradigm of rationality, scientific method was required to match formal logic arguments.
Deductive Prediction:
Prediction is modeled as a formal logical deduction from initial conditions and general laws.
Example 1: Predicting a cannonball landing position requires collecting data on initial conditions (cannon elevation angle, charge energy, cannonball mass, terrain configuration) and applying Newtonian laws to deduce the impact point mathematically.
Example 2: Deducing the composition of surface moon rocks from a hypothesis about solar system formation.
Covering-Law Model of Explanation:
Formulated by Carl Hempel, Ernest Nagel, and others, this model states that explaining an event consists of demonstrating that it was necessary given relevant laws and initial conditions.
Example: Explaining a cannonball landing site involves showing that given initial conditions (barrel elevation, charge size) and ballistics laws, the ball had to land at that exact location.
Prediction-Explanation Symmetry: Predictions and explanations are logically identical derivations from initial conditions and general laws or theories. The only distinction is temporal: predictions occur prior to the event, whereas explanations occur after the event.
Inductive Generalization and the Uniformity of Nature:
General laws arise from projecting observed regularities (e.g., water boiling at under standard conditions) to unobserved spatial and temporal domains.
Inductive generalizations rely on the Principle of the Uniformity of Nature—the unprovable assumption that unobserved regions of space and time resemble observed regions.
Because the uniformity principle cannot be logically proven, scientific conclusions retain an ineradicable tentativeness.
Testing and Coincidence:
Human observations are spatially, temporally, and instrumentally limited.
Regularities observed within a restricted sample domain may be coincidences rather than universal natural laws.
Hypothetico-Deductive Testing:
Scientists expand observational scope to evaluate whether perceived regularities are genuine laws or coincidences.
Procedure: Predictions are deduced from a hypothesis and evaluated under artificial experimental conditions or naturally occurring events (e.g., waiting for a typhoon).
Logical Outcome: If a deduced prediction proves false, the hypothesis is logically falsified (since a premise yielding a false consequence is false). If predictions hold, theoretical confidence increases, though absolute proof remains impossible due to finite sample sizes.
The Mid-Twentieth-Century Traditional View: The Empirical Element
Empirical centrality: Predictions, explanations, tests, theoretical confirmations, and hypothesis rejections rely entirely on empirical data.
Demarcation Criterion: Claims must be empirical or subject in principle to empirical testing to qualify as scientific. Nonempirical claims are excluded from scientific consideration.
The empirical base functions as the sole and ultimate arbiter of theoretical acceptability.
The Mid-Twentieth-Century Traditional View: Objectivity
Neutrality of Observation: Observational experience is considered inherently objective, neutral, and public. Observers perceive the same raw observational facts regardless of theoretical commitments or presuppositions.
Self-Correction: Public, neutral observational facts resolve theoretical disputes objectively, guiding scientists away from initial errors.
Reproducibility and Communal Objectivity: Individual subjective bias is controlled through communal standards requiring scientific results to be independently reproducible. Non-reproducible findings are rejected by the scientific community.
Theoretical Assumptions: Grounded in the assumption of an enduring, permanent reality (derived from ancient Greek philosophy) and the uniformity of nature.
Structural Elimination of Subjectivity: Relying on formal logic and empirical data prevents non-objective factors, such as philosophical or religious predispositions, from influencing theory selection.
Core Implications of the Traditional Conception
Structural Tentativeness: Because data underdetermine theories, empirical testing can never conclusively prove the truth of any scientific theory.
Cumulative Progress: Neutral, objective observational facts remain stable over time. Scientific progress occurs sequentially through the steady accumulation of facts, where successive generations correct past errors and append new knowledge to an enduring structure.
Logical Positivism
Logical positivism (or positivism) was the most influential school within the traditional view, developed in the 1920s and 1930s by the Vienna Circle (including Moritz Schlick, Rudolf Carnap, Otto Neurath, Herbert Feigl, Hans Reichenbach, and Carl Hempel).
Standard manifesto: A. J. Ayer's Language, Truth and Logic ().
Historical Evolution: John Locke's empiricism (inspired by Newton's exclusion of nonempirical concepts) David Hume's assertion that nonempirical ideas are empty $ ightarrow$ French Enlightenment's use of empiricism against ecclesiastical authority.
Verifiability Criterion of Meaning:
A statement is cognitively meaningful if and only if it is either:
Empirically verifiable or testable in principle.
Analytic (true solely in virtue of the definitions of its terms, e.g., "All bachelors are unmarried" or ).
Statements that are neither testable nor analytic express no cognitive content and are literally meaningless.
Theoretical and scientific terms must be strictly definable in terms of direct sensory observations.
Epistemological Reductionism: All legitimate human knowledge is reduced to empirical scientific knowledge.
Positivist Rationality and Formal Logic:
Scientific inferences and natural law statements were framed within formal symbolic logic.
Formal Confirmation Theory: Positivists used probability theory (notably Rudolf Carnap) to quantify how positive empirical evidence increases the formal probability of a theory.
Positivist Objectivity and Antirealism:
Extreme rejection of subjective intrusion, declaring metaphysics, philosophy, and religion to be literal nonsense.
Antirealism: Deeply suspicious of general theories and unobservable entities (such as atoms, electrons, and genes). Positivists denied that science yields knowledge of unobservable realms.
Phenomenalism: A. J. Ayer defined material objects as logical constructions of sense-contents.
The Collapse and Internal Incoherence of Logical Positivism
Conflict with Scientific Realism:
Practicing scientists are predominantly realists who seek to discover genuine, unobservable substructures and causal mechanisms (e.g., explaining balloon shrinkage via gas molecule leakage, chemical compound stability via valence electrons, or physical inheritance via genes).
Positivist antirealism fails to reflect actual scientific practice.
Loss of Explanatory Power:
By barring appeal to unobservable entities, positivism reduces science to describing empirical regularities without explaining them ("that's just how it is"), stripping science of theoretical depth.
Failure of Formal Logic Projects:
Attempts to construct an inductive logic of confirmation or reduce natural laws to formal symbolic logic failed.
Threefold Self-Destruction of the Verifiability Criterion:
Scientific Failure: Foundational scientific presuppositions (such as the uniformity of nature) are neither empirically testable nor analytic, classifying core scientific premises as meaningless under positivist standards.
Moral Failure: Moral statements (such as "Murder is wrong") are neither testable nor analytic, making the claim that ethics is cognitively empty philosophically and morally unacceptable.
Logical Incoherence: The Verifiability Criterion of Meaning is itself neither empirically testable nor analytic. By its own standard, the criterion is either meaningless or false.
The Decline of the Broader Traditional View
Asymmetry of Prediction and Explanation:
Predictions can be made from empirical regularities without possessing an explanation.
Example: Ancient astronomers predicted solar and lunar eclipses accurately using observed empirical regularities without knowing why eclipses occurred or what physical causes produced them.
Explanation cannot be reduced to simple deduction from empirical laws.
Failure of Confirmation Theory: Philosophers failed to produce a formal theory explaining how empirical evidence confirms general hypotheses.
Karl Popper and Falsificationism
Karl Popper () presented his philosophy in The Logic of Scientific Discovery () and Conjectures and Refutations ().
Rejection of Positivism: Popper acknowledged that nonempirical statements (philosophical, metaphysical) are meaningful, shifting focus to the Demarcation Problem—distinguishing scientific claims from non-scientific claims.
Critique of Induction and Confirmation:
Fallacy of Affirming the Consequent: Inferring the truth of a theory from verified predictions is logically invalid (, ).
Infinite Implications: Theories yield infinite observational consequences; checking a finite subset can never establish truth.
Nonexistence of Induction: Popper argued that inferring theories from empirical data is logically impossible and that "there is no such thing as induction."
Zero Probability Calculation: Because universal hypotheses cover infinite potential cases, checking a finite set of predictions yields a ratio of checked cases to unattempted cases of zero:
Theories can never be verified, confirmed, or rendered probable by empirical evidence.
The Falsificationist Paradigm:
Deductive Asymmetry (Modus Tollens): While positive empirical instances cannot verify a hypothesis, a single false prediction logically refutes it (, ).
Method of Conjectures and Refutations: Scientists propose bold, empirically falsifiable hypotheses and attempt to refute them through stringent testing. Scientific progress consists of eliminating falsified theories.
Demarcation Criterion: A statement is scientific if and only if it is empirically falsifiable (capable of being contradicted by empirical data).
Radical Popperian Insights:
Fallibility of Basic Statements: Empirical observational statements ("basic statements") are themselves falsifiable and subject to rejection.
Role of Convention: Accepting basic statements depends on human decisions and conventions to avoid an infinite regress of testing. The demarcation criterion is a conventional proposal based on value judgments and predilections.
The Swamp Metaphor: Science does not sit on solid empirical bedrock. It is like a building constructed on piles driven into a swamp. Piles are driven deep enough to support the structure for practical purposes without reaching solid ground.
Criticisms and Difficulties of Falsificationism
Disconnect from Scientific Practice: Practicing scientists view many scientific theories as well-established and strongly confirmed by evidence.
Entanglement with Nonempirical Factors: Scientific practice is interlinked with worldviews and philosophical presuppositions, complicating strict empirical falsifiability separation.
The Counterexample of Existential Statements:
Consider the claim: "Somewhere, sometime, the universe has contained or will contain a perfectly cubical piece of gold weighing exactly ."
This statement is empirical and compatible with natural laws, but cannot be proven false by any amount of empirical observation. Popular falsificationism fails to categorize such valid claims properly. (Popper classified existential statements as non-empirical metaphysical statements).
Impossibility of Conclusive Refutation: No logically consistent scientific theory can ever be conclusively falsified in isolation because empirical tests rely on background assumptions and auxiliary hypotheses.