1
1
Lecture:Philosophy of Science (Schurz) Ss 2015 Wed 10.30-12 24.91/U1.61
Part I: General introduction and philosophical foundations
1) 04-15 Tasks and aims of Philosophy of Science. The method of rational reconstruction
2) 04-22 Common epistemological assumptions and methodological features of the sciences
3) 04-29 Classification of scientific disciplines and the demarcation problem: The requirement of value neutrality
4) 05-06 Scientific inference: Deduction, induction and abduction
Part II: Logical foundations
5) 05-13 Kinds of concepts
6) 05-20 Kinds of sentences
7) 05-27 Degrees of generality. Logical relations between sentences
Part III: Law hypotheses and their empirical testing
8) 06-03Testingfor Truth and Relevance 1: the deterministic case
9) 06-10 Testingfor Truth and Relevance 2: the statistical case
10) 06-17 Correlation and causality
Part IV: Scientific theories
11) 06-24Observation concepts, empirical disposition concepts and theoretical concepts
12) 07-01 Structure and methodological features of scientific theories
Example 1: Newtonian physics
13) 07-08 Example 2: Piaget's cognitive psychology. Theory evaluation and theory progress
14) 07-15 Exam
Book to the lecture: Gerhard Schurz: Philosophy of Science: A Unified Approach, Routledge,New York 2013.
Furtherliterature (in red: recommended):
Bird, A. (1998): Philosophy of Science, McGill-Queen's University Press, MontrealKingston.
Bunge, M.: Scientific Research, Springer, Berlin 1967.
Carnap, R.: Philosophical Foundations of Physics, Basic Books, New York 1966.
Curd, M., and Cover, J.A. (1998, ed.): Philosophy of Science, Norton, New York.
Feyerabend, P.: Against Method, New Left Books 1975.
Godfrey-Smith. P. (2003): Theory and Reality: An Introduction to the Philosophy of Science, University of Chicago Press, Chicago.
Hempel, C. G.:Philosophy of Natural Science, Englewood Cliffs, Prentice Hall 1966.
Kuhn, T.: The Structure of Scientific Revolutions, 2nd ed., Univ. of Chicago Press 1975 (orig. 1962).
Ladyman, J. (2002): Understanding Science, Routledge, London.
Lakatos, I., Musgrave, A.:Criticism and the Growth of Knowledge, CambridgeUniversity Press 1970.
Losee, J. (2001):A Historical Introduction to the Philosophy of Science, OxfordUniv. Press, Oxford (Orig. 1972).
Popper, K.: The Logic of Scientific Discovery, Basic Books, New York 1959.
Psillos, S. (1999): Scientific Realism. How Science Tracks Truth, Routledge, London and New York.
Reichenbach, H.: The Rise of Scientific Philosophy, Univ. of California Press, Berkeley and Los Angeles 1951.
Van Fraassen, B.: The Scientific Image, Clarendon Press, Oxford 1980.
Questions of the General Philosophy of Science
General versus special characteristics of the sciences including the natural sciences, social sciences and the humanities.
the objects
What arethe methods of the sciences ?
the goals and limitations
a scientific language
correct scientific argument
What is a scientific observation, fact or measurement?
a scientific law, a theory?
a prediction, explanation, causal relation ?
How are laws or theories empirically tested?
What are criteria for theory progress in science
What is objectivity and truth ? Philosophy of science epistemology
Values in science versus value-neutrality?Philosophy of science (meta-)ethics
•Applications internal to sciences:
Methodological advice and decision support on controversial questions in sciences
Interdiscplinary relations and transdisciplinary discoveries
Pioneer function for new sciences
•Applications external to sciences:
Demarcation problem (example: controversy surroundingcreationism)
Critical function: critique of political or economical abuse of the sciences
(example: systematic biases in pharmaceutical research)
The method of the philosophy of science
•The normative view (earlier): Karl Popper, Vienna circle (logical empiricism)
•The descriptive view: Thomas Kuhn ('historical turn'), W. Stegmüller, L. Laudan R. Giere ()
Argument of the normativists: Context of discovery (genesis)
versus
context of justification ( confirmation)
The method of rational reconstruction (Lakatos1971, Stegmueller1979):
NORMATIVE CORRECTIVE
Supreme epistemic goal (Minimal) epistemological model
(Revision?) (Justification) (Revision?)
RATIONAL RECONSTRUCTION
Philosophy of Science develops models of:
Observation, experiment, law, theory, confirmation, disconfirmation
falsification, explanation, causality, theory progress ....
(empirical support) (application)
Real (factual) sciences Real sciences vs. pseudo-sciences
positive and negative paradigm cases Controversial cases
DESKRIPTIVE CORRECTIVEUnsolved problems
Supreme Goal: (G) Finding true and content-rich statements
relating to the given domain of investigation
tension between probability of truth and richness of content
Minimal epistemological modell five assumptions:
(E1) Minimal realism (correspondence theory of truth)
(E2) Fallibilism (critical attitude)
(E3) Objectivity & Intersubjectivity (criterion)
(E4) Minimal empiricism (empirical testability)
(E5) Logic in the broad sence (concepts, statements, arguments)
Minimal methodology four methodological features:
(M1) Search of general and content-rich laws and theories
(M2)Search of actual observation statements
(M3) Attempt to explain the actual observation statements
and to predict potential observation statements,
with help of the laws or theories conjectured in (M1)
(M4) Attempt to test laws and theories by comparing the predicted (potential) observation statements with the
actual observation statements.
Agreement: Confirmation
Disagreement: Falsification or Disconfirmation
Three levels:
Scientific theories
Prediction explanation Confirmation disconfirmation
Empirical laws
Prediction explanationConfirmation disconfirmation
(Actual) observation statements
Classification of scientific disciplines, according to their domain of objects:
SCIENCES
1) of nature: physics, chemistry, biology, geology, medicine (astronomy, cosmology, geography, paleontology, history of biological evolution)
2)of technology:mechanical and electrical engineering …, also: computer science
3) of human beings: psychology (also: education, medicine, cognitive science)
4) of (human) society: sociology, economics, political science (also: anthropology, ethnology, geography)
5) of (human) history: history (also: anthropology, ethnology)
6) ofcultural (mental, social) artifacts:legal sciences, linguistics, literary science, sciences of fine arts and music,media studies (also: education, religious studies)
7) of formal structures (formal sciences): mathematics (logic, statistics, theoretical computer science, system theory…), formal methodology and philosophy of science
8) of the general foundations of human ideas: philosophy (epistemology, philosophy of science and theoretical philosophy; ethics, aesthetics and practical philosophy)
9) of God: theology (also: religious studies)
Natural sciences: only 1?, +2?, +7? +3?
Humanities: 5,6, 8 Why not 7? Or 3?
Human and social sciences: 3, 4 (5?, 6?) [Cultural sciences]
Factual sciences (as opposed to structural sciences): 1, 2, 3, 4, 5, 6 8?, 9??
Exceptional case of formal sciences 7: G; E1-E3, E5 (not E4); M1 (not M2-M4)
Demarcation:"Science" comprises all empirical sciences (satisfying G, E1-E5, M1-M4, plus their associated formal and methodological auxiliary sciences (7).
Limitations of science:
where evaluative statements enter the discipline ( 6, 8, 9 )
where assumptions of faith (religios creed) enter the discipline (9)
Example: Value judgement in jurisprudence (the science of legal judgment).
(Is a billboard advertising underwear still in accord with "common decency" or is it already a legal offense?)
Eike von Savigny et al. (1976)E. Hilgendorf and L. Kuhlen (eds., 2000)
Max Weber (1864-1920): Postulate of value-freedom
because: Values are not properties inherent to objects themselves, butbased on subjective interpretations by us humans.Ultimately, the decision for or against certain values is a question of personal freedom.
However, the scientist qua scientist
(1.) can study the factual presence of value and norm systems,
(2.) can discover logical relationships among value or norm sentences (test them for inconsistencies), and
(3.) ( most importantly for practical sciences): can infer derived norms from given fundamental norms and descriptive knowledge, by means of the so-called
Means-end inference:
Descriptive means-end hypothesis: M is in the given circumstances C a necessary
or alternatively an optimal means for the realization of end E.
Thus:Given (fundamental norm:) end E is to be realized,
then (derived norm:) means M should also be realized.
The requirement of value-neutrality (VN) in sciences:
EVexternal values (all values except those in IV)
IVinternal values (supreme goal (G) and all values following from (G) by means-end inferences)
CDcontext of discovery
CJcontext of justification
CAcontext of application
EV & IVonly IV EV & IV
CD CJ CA
(VN): A specific realm of scientific activity, namely their context of justification, should be free from fundamental science-external value assumptions.
( A normative recommendation. Not a generally followed practice)
The selection of the investigated objects and parameters in (CD) is not epistemologically neutral, but places limits on the results of the scientific investigation.
Therefore the selection in (CD) must be accessibleto subsequent correction by results in (CJ) (even when this goes against the external goals of the research project).
Selection of relevant variables
Example: causal theories of psychological depression
•Hippocrates: surplus of black bile
•Middle Ages: devil and demons; punishment for laziness
•Astrology: star constellation
•Freud: child development, lack of satisfaction in oral phase
•Beck: cognitive defects
•Seligman: uncontrolled fear
•Genetics: genetic dispositions
•Neurophysiology: low level of neurotransmitters
Further classifications of disciplines:
Classification of factual sciences:
Speculation
Empirical sciences
Experimental sciences
Dissecting sciences
Graduated division of scientific methodologies by increasing degree () of logical-mathematical and quantitative precision
logical-mathematical language natural language
"quantitative" "qualitative"
Logic Hermeneutics
Statistics and measurement theory Content analysis
Technology Field research
What is the characteristics of "(natural) science":
empirical? experimental? dissecting?quantitative?
+ yes- no n naïve ? between + and - ?+ tends towards + ?- tends towards -
PositionVNA1A2A3A4A5M1M2M3M4
Pre-modern History
Plato-n+-+?-+---
Aristotle?-n+-++?-+-+?-
Alexandria+n+-++n++++n+
MA ≤ 12th century-n+-+?-+-+-
Late scholasticism ?++?++?+n++++n+
Modern Times
Empiricism
Bacon, Locke?+n+-+-n++++n+
Hume+-++-+++++
Mill?++-++n+++++
Rationalism
Descartes, Leibniz?-+-++-+-+-
Kant-+-++?+++?
Contemporary Phil of Sci
Logical Empiricism+?+++n+++++
Post-Positivism++++++++++
Pragmati(ci)sm?-++++++++
Contemporary criticisms:
Relativism
+/- moderate/radical?+/-+---+?+-
Constructivism
+/- moderate/radical?+/-?+/-+/-+/?++/-++/-
Hermeneutics?-?+ ++-?+-+-?
Critical theory--+---?+?++?+
The inductive –The general (laws and theories)
deductive schema: inductive ascent deductive descent
(Aristotle)
The particular (observations)
Induction in the broad sense: Induction (i.n.s.)+ abduction
Three kinds of scientific inference and argumentation:
Deduction- certain: Logic in the narrow sense
All As are Bs, this is an A / therefore: this is a B
(Other kinds of ded. inference: fromgeneral togeneral, fromparticular toparticular)
Induction- uncertain:
Inductive generalization:
AllAs observed so far were Bs // therefore (probably): allAs are Bs
[Statistical version: r% of observed As were Bs // therefore: r(% of As are Bs]
Inductive prediction:
AllAs observed so far were Bs // therefore: the next A will be a B
Abduction - very uncertain: Inference to the best explanation, or inference to an unobserved cause (theoretical concept)
This is an A. Can be explained (in given background knowledge) by the
assumptionthat this is a B //(Conjecture:) this is a B
Controversial: is abduction a scientifically legitimate form of inference?
Popper: No: abduction discovery of hypotheses by trial and error
But oh! Popper and his students even doubt the scientific legitimacy of induction
Three Kinds of Induction:
1. Methodological induction
Induction as a method of "extracting"laws and theories from observations
Example:AllAs observed so far were Bs thereforeallAs are Bs
Major criticism of Popper:
Confusion of context of discovery and context of justification
Theories are not or better: not only discovered by induction
2. Logical Induction (Carnap, Reichenbach, Bayesianismus):
Induction as a methodof justification: determining the conditional probabilityof scientific hypotheses H given the observational data O:
Probability(H / O) = so-and-so (e.g., 0.9)
Major criticism of Popper: The space of all possible alternative theories is unlimited and cannot be probabilistically measured.
3. Epistemic Induction (or meta-induction):
Induction as a merely comparative evaluation of the probability of scientific
hypotheses (laws or theories):
Theory T1 has been empiricallyWe believe that T1will be empirically
more successful than theory T2 more succesful than T2in the future
current state of observational knowledge
the degree of confirmation of a
theory is alwaysdoubly relative: current state of alternative theories.
Abduction to theories - example:
Planets move around the sun in elliptic orbits.
This can be explained by Newton's force laws (2nd law & gravitational law & cp).
// Abductive conjecture: Newtons' force laws are approximately true
Interaction of epistemic induction and abduction
(1) Evidence: Tk is amongthe alternative theories T1,,Tnso far the empirically most successful.
epistemic inductive inference
(2) Instrumentalist conclusion: Tk is among T1,,Tn the most empirically adequate (therefore also in future the most empirically successful)
abductive inference to the best theory
(3) Realistic conclusion: Tk is among T1,,Tn the closest to the truth.
Empiricist instrumentalismversus realismin the philosophy of science
Kindsof concepts --- Classifications:
According to the logical type: (shorted presentation)
Singular concepts or terms(designate a single individual,
spatiotemporal location or situation)
Non-logical(individual constants: a, b, …)
concepts
General concepts
predicates one-place:express properties
or kinds (F, G, …)
n-place:express (n-ary)relations (R, …)
function symbols: expressfunctions (f, g, …)
LogicalTruthfunctional sentence operators not (), and (),
concepts inclusive or (), if-then (), (prop. logic)
Quantifiers "for all" (), "exists" () (pred.logic) Intensional sentence operators necessary (), possible () probable (modal logic)
Variables (for individuals x, y,…; forpredicates , ,…)
Mathematical concepts (set theory), +, (arithmetics), …
According to the content type:
(logical concepts)observation concepts empirical
concepts
descriptive concepts empirical disposition concepts
(non-logicaltheoretical concepts
concepts)
prescriptivenorm concepts
concepts
value concepts
criterion for observability: ostensive learnability
observable in the narrow sense versus empirically measurable in the wide sense
According to the gradation (scale) type:
classificatory concepts
nominal (categorial) scales
qualitative concepts
comparative concepts
ordinal (ranking) scales
interval (difference) scales
quantitative concepts
ratio scales
Kindsof Sentences --- Classifications:
According to the content type:
logically determined
analytic
determined by definition
observation sentences ()
empirical general emp. sentences (...)
descriptive
synthetic theoretical purely theoretical
mixed-theoretical
normative
purely prescriptive
evaluative
mixed-prescriptive
(Simplified) definitions:
Observation sentence:a singular sentence (*), which contains (apart from logical concepts) only observation concepts. Example: "this raven is black".
(*: or a localized-quantified sentence: e.g., "all apples in this basket are red")
Empirical sentence:a (possibly quantified) sentence, which contains (apart from logical concepts)only empirical concepts. Example: all raven are black.
Theoretical sentence: a sentence that contains theoretical concepts (besides logical concepts and possibly empirical concepts). Example: "atoms consist of protons, neutrons and electrons", or "in the center of our galaxy there is a black hole".
(T-theoretical concept, T-theoretical sentence)
A sentence is purely descriptive iff it
(it either contains no prescriptive concept, or if)
every prescriptive concept occurring in it lies in the scope of a subjective
attitude operator
A sentence is purely prescriptive iff all of its descriptive components (subsentence or subformulas) lie within the scope of a prescriptive operator.
(In other words, iff all of its elementary subsentences/subformulas are elementary prescriptive sentences.)
A sentence is mixed otherwise(it has descriptive and prescriptive elementary subsentences/subformulas).
Examples:
Peter believes that stealing is bad.
Stealing is bad.
Stealing is allowed for a person, if this person is suffering from hunger.
Peter believes that stealing is bad, although he is a thief himself.
If stealing is permitted, then there exists no right to private property.
Peter's car has good breaks.
Peter has a good character.
A sentence is logically true iff every sentenceof the same logical form is true.
in other words:iff its truth depends only on its syntactic structure and on the meaning of its logical concepts.
Logical form of a sentence:
Replace all nonlogical symbols by variables (dummy letters).
Example of a logically true sentence:
If all men are mortal, then there exists no man who is immortal.
Logical form: If all F are G, then there exists no F which is not a G
Formalization:x(FxGx) x(FxGx)
Example of a synthetically true sentence:
All men are mortal. (All F are G)
An argument (inference) is logically valid iff for every argument which has the same logical form the following holds: if all premises are true, theconclusion is true.
Example:Premise 1: All humans are mortal.
Premise 2: You are a human.
Conclusion: Therefore your are mortal.
Logical Form:All F are GFormalization:x(Fx Gx)
This a is an F Fa
Therefore this a is a G Ga
Example of an invalid argument:
All humans are mortal.All F are G
This living being is mortalThis a is a G
Therefore this living being is human.Therefore this a is an F
A sentence is definitorially(or: extralogically-analytically) true iff its truth is determined by certain conventions of meaning for its non-logical concepts (that are entrenchend in the underlying language or linguistic community)
Example: All bachelors are unmarried(Logical form: All F are G)
The length of the standard measure in Paris is one meter
Example of synthetical sentences:
All polar bears are white (Same logical form)
The length of this rod is one meter."
(Definiendum) (Definiens)
Explicit definitions:x: x is a bachelor defx is a hitherto unmarried man
Meaning postulates: If something is red, then it has a color.
Derived definitorially true sentences: If somebody is a bachelor, then he is male.
Requirements on definitions:
They must not be circular.
They must neither have empirical content, nor create new empirical content in combination with the given system S of accepted background beliefs beliefs.
Hence: No concept may be defined in two different ways.
Example: 1 Meter = the length of the platin-iridium-bar in Paris. (1)
1 Meter = the length of a pendulum at sea level with (2)
an oscillation frequency of one second.
(1) + (2) imply the following synthetical (empirical) consequence:
The length of the platin-iridium-bar in Paris= the length of a pendulum at sea level withan oscillation frequency of one second.
Classification of sentences
according to theirgenerality (logicalstrength):
Strict (or deterministic) generalizations
Purely universal sentences
e.g.: For all x: if x is A, then x is C x(AxCx))
'A' for 'antecedent', 'C' for 'consequent'
(spatiotemporally) unrestricted (spatiotemporally) restricted
Mixed-quantified generalizations; e.g. universal-existential (etc.).
General sentences
Non-strict generalizations
Statistical generalizations
e.g.: q % of all As are Cs p(C|A) = r (with: 0r1; q = 100r)
(spatiotemporally) unrestricted (spatiotemporally) restricted
Normic and ceteris paribus generalizations
e.g.: As are normally Cs and: C.P. As are Cs
Singular sentences e.g. This is an A, and it is (or is not) a C. Aa(Ca
Existential sentences e.g. There exists an A that is (or is not) a C. x(Ax(Cx)
(Mixed sentences)
Interlude: Two kinds of probability:
•1. Statistical (objective) probability:Always refers to classes, never to individual cases. "80% of all Italians are brown-eyed"
•2. Epistemic (subjective)probability:Rational degree of belief. Refers to individual cases. "With high probability Vicenco (from Italy) is brown-eyed"
Principle of narrowest reference class (Hans Reichenbach)
Important logical relations
Notation: "X Y" stands for "Y follows logically from X",
and "X Y" for "X and Y are logically equivalent".
(1) Universal sentence singular sentence
All As are Cs if a is A, then a is C
(2) Universal sentence & singular sentence singular sentence