Philosophy of Science (Lecture)

I followed this course by, Dr. Sebastian de Haro Ollé during my 5th Semester at the ILLC Amsterdam for my M.Sc. Logic (UvA).

Assignments

Lecture Notes

Here I summarise the content of the lecture and state brief remarks on the papers read during the course.

02.09, I. Logical Empiricism

Frege, Russel, Wittgenstein: Logic has a tautological nature
- Russel: “math is the only science in which you don’t know what you are talking about”
  - content does not matter
Vienna Circle: every principle is either empirically verifiable or logically necessary
Carnap:
- Early Carnap: philosophy as linguistic analysis
  - syntactic, analysis of the form of reasoning
  - semantic, is part of psychology
  - of a word tfae: (i) dedicubility, (ii) conditions of truth, (iii) verifiability, (iv) meaning
    - why? given the following theory
  - scientific theory := set of sentences
  - sentences that are not observational (i.e. theoretical) shall be divisible into observational
  - strong position against metaphysics
Syntactic view
1. Axioms make a theory
2. Theorems (sentences) can be derived
3. Correspondences Rules bond the theorems to the observations
- Change in Theory
  - Positivists Periodisation: you do observations given a theory in the background
  - Kuhn: a change in the theory, changes the observation
  - Galisons: experiments and instruments shall change, accordingly to theories
    - there can be dialogue between sciences,
Semantic view
- theories are sets of models
  - not all theories can be axiomatised
  - no language dependence
- see Halvorson, H. (2012), What Scientific Theories Could not Be, Halvorson, H. (2013), The semantic view, if plausible, is syntactic, in general Structuralism.

5.09, II. Popper

Falsifiability: empirically provably false
- necessary statements are not fansifiable
  - what is necessary, might be a metaphysical matter
  - mathematical statements
observations rely on the theory
Lakatos: falsifiability should be reviewed
- no guidelines on what to do in response to an anomaly
- New Criteria
  - Naive falsificationist: falsifiable $\to$ acceptable (early Popper)
  - Sofisticated falsificationist: falsifiable, (i) progressive, (ii) heuristic value
    - broader view: one should look at research programs, not theories
- Parts of a theory
  - Hard Core: real essence of the theory
  - Protectvie Belt: what can change insignificantly

09.09, III. Kuhn

Paradigm: a framework for practicing science shared by the scientists in a particular discipline
- science is a social enterprise
- theoretical principles & assumptions, methods, tools,
science is conservative and seemingly dogmatic:
- Popper is on the opposite side
  - Kuhn is not normative, he describes how sciences are
observation is theory-laden, it cannot be completely detached from paradigms
- semantic incommensurability: words get different menaing
- methodological incommensurability: standards of measure and method change
Critiques to Kuhn:
- Kuhn’s paradigms compete and do not comunicate, that’s not true
- “paradigm” is not a precisely described enough word
- Lakatos: there are far less emotions involved than Kuhn claims, it’s not politics
a good sci.th. is: (i) accuracy, (ii) consistency, (iii) broad scope, (iv) simplicity, (v) fruitfulness
- it is not an algorithmic choice, different previous experience gives different valuations
Max Weber: values influence scientists’ actions, but (social) science is objective
- choice of topic is value-laden, research is value-free, application of results is value-laden again
Longino: objective knowledge can result at community level
- Criteria for objectivity:
  1. recognised avenues for criticism (peer review)
  2. shared standards
  3. community is responsive
  4. equality of intellectual authority

16.09, IV. Quine

Duhem: there is no crucial experiment, it cannot be the experiment that falsifies the theory
- “hypotheses can only be tested in combination with auxiliary assumptions”
- even if you believe to have falsified the theory, there could always be an unconceived alternative which saves the theory and of which nobody thought of.
Quinde: Quine, V.W. (1951), Two Dogmas
1. Reductionism: scientific statements are logically equivalent to statements in empiricist language
  - subj.lang. of immediate experience / obj.lang. of physical things
2. Analytic / Synthetic Distinction: difference between conceptual schemes and facts is of degree only
  - at a certain point, while defining an analytic statement, one will use some experience.
- in the end, the two dogmas are equiv.
Critique:
- Prof: all started with skepticism on meaning, if we have a theory of meaning, problem solved
  - intention of a sentence is in possible worlds semantics
    - max.set s.t. $S$ holds
Under-determination:
1. Empirically equivalent: same observational sentences
2. Theoretically inequivalent: different theoretical sentences
- hence theory is under-determined by the empirical data.

23.09, V. Scientific Explanation & Understanding

before ‘48: explanation is not a philosophical topic for logical empiricists.
C. Hempel, 1948: covering law model, epistemic explanation
- deductive argument: conditions & general laws explain (deductively prove) the phenomenon.
  - only general laws prove only general things
- Conditions of adequacy
  1. logical consequence {can you actually formalise the conditions? perhaps the general laws}
  2. general law, to guarantee necessity {trivially take spatio-temporally bounded general laws?}
  3. explanans must have empirical content
  4. true metaphysically {empirically accessible?}
- No need of causation, this is more general
- Problems:
  - It is not clear what explains what, one can change arguments and have symmetries
  - One get to correlations rather than explenations
- Alternatives
  - causation is needed to explain
    - Hume: deletes cause
      - Hume, counterfactual: if no $A$ , then no $B$ , hence $A$ causes $B$
      - Hume, regularity theory: usually the sun rises.
    - De Regt: causation is one way, not the only one
  - unification of how phenomena fit in a broader pattern
    - {this could be deductive, general law give the frame, conditions the instance
    - -: not really, not as formal and you can already assume $E$ to have happened}
  - pragmatic, basically contextualism
  - scientific explanation requires understanding
Understanding:
- Hempel’s Eliminativism: understanding is matter of the subject, it is psychological.
- Khalifa’s Reductvism: having an explanation could be enough
  - Lipton: need to know the cause
  - De Regt: have an oracle who, you know, tells you the truth, you have still no understanding
- Carnap: in order to have understanding one should
  1. capability of use for the description of facts
  2. metaphysical understanding (refused)
  3. knowing semantic truth conditions
- van Fraassen:
  - explanation & understanding are part of the pragmatic, subjective.
  - explanation is irreducible and special to each subject.
  - explanation is not a relation between theory and fact, it is a three term, with also context.
- De Regt: Pragmatic use of Understanding
  - macro-level: science as a whole
  - meso-level: scientific community
  - micro-level: individuals

17.10

Putnam & Oppenheim, Unity of Science as a Working Hypothesis
- One only discipline: Science
  - Same law, same laws, and coherent laws
- Reducibility of a theory: $T_{1}$ reduces $T_{2}$ iff
  - vocabulary in $T_{2}$ contains terms not in the vocabulary of $T_{1}$

🪴 Quartz 4.0

Explorer