Scientific Explanation I 11/PoS Week 11.pdf · Explanation of Laws: According to Hempel and Oppenheim, we can have DN explanations not only of particular facts but also of laws. Example:

1

PH201/400 – Week 11

Scientific Explanation I

The DN Model

Why-Questions and Scientific Explanation

Science not only tells us what in fact happens; we also expect it to tell us why things happen in the way they do.

So we expect science to answer why-questions.

2

That is, we expect science to answer questions like:

•  Why does Uranium 235 decay?

•  Why did the lunar eclipse occur?

•  Why did the prices go up by 0.1% last year?

•  Why did cholera break out?

•  Why do planes fly?

3

But not all explanations are scientific, and not all why questions demand scientific explanations:

•  Explain the plumber what is wrong with the shower à explanation as a specification of what is the case

•  Explain a poem by Borges à explanation as artistic interpretation

•  Explain the rules of chess à explanation as instruction

4

•  Explain why you are late à explanation as justification

•  Explain where the party takes place à explanation as providing information

•  Explain why you study philosophy – explanation as a specification of motives or motivations

Question: What is a scientific explanation? What do scientist do when they explain something?

5

Terminology:

Explanandum (Em): the thing that has to be explained

Explanans (Es): the thing does the explaining

6

Overview

Em particular facts regularities

Es

universal laws DN-model DN-model Lec. 8

statistical laws IS-model DS-model Lec. 9

causal relations Lec.10

unification

…

7

The Covering Law Model of Scientific Explanation

Hempel and Oppenheim 1948.

Leading idea:

•  We explain something by subsuming it under a general law

•  We show that it is an instance of a general pattern.

8

General Structure

L1, L2, …, Lk Laws

A1, A2, …, Am Auxiliary laws

B1, B2, …, Bn Boundary conditions

-----------------

E

Explanans

Explanandum

9

(1) The explanandum is a logical consequence of the explanans; i.e. the explanation is a valid deductive argument.

(2) The explanans must contain at least one law; and this law must be used in the derivation of the explanandum.

(3) The explanans must have empirical content; that is, it must at least in principle be empirically testable.

(4) The sentences contained in the explanans must be true.

Conditions of adequacy:

An argument of this sort is an explanation if the following four conditions are satisfied:

10

Simple example: The law of supply and demand: if there is increasing demand for a certain good and the supply is constant, then the price of that good will go up.

There was a rising demand for houses in London since the crises and the supply was constant.

---------------------------------------------------------------------

House prices in London were rising since the crises

11

Notice:

•  This is a ‘deductive-nomological explanation’ (DN-explanation): ‘deductive’ because it has the formal structure of a deduction; ‘nomological’ because it contains laws.

•  It’s also referred to as the “DN-model”.

•  On the DN-model, explanations are arguments.

•  Salmon: ‘the third dogma of empiricism’.

12

Explanation of Laws:

According to Hempel and Oppenheim, we can have DN explanations not only of particular facts but also of laws.

Example:

L: Newton’s axioms A: The law of gravitation B: The modelling assumption that heavenly bodies can

be treated as point masses; no forces other than gravitation is present; etc.

----------------------------------------------------------- E: Kepler’s laws

13

Problems for the DN Model Two kinds of difficulties:

Arguments that satisfy all requirements but nevertheless fail to be explanations. Such arguments show that these requirements are not sufficient.

Explanations are considered to be real explanation but do not satisfy the above requirements. Such explanations show that the requirements are not necessary.

14

(A) Against Sufficiency Problem 1: The Asymmetry of Explanation

Intuitively, explanation is asymmetric. That is, if A explains B then B does not explain A.

But this stands in contradiction to the DN model.

à Example with the Flagpole and the shadow.

15

sun

flagpole (of length Lf)

α

Ls

Explain the length of the shadow:

Laws of linear optics Lf α ------------- Ls

16

So the laws of optics, the height of the flagpole and the angle of incidence explain the length of the shadow.

But as matter of logic we also have:

Laws of linear optics Ls α ------------- Lf

So the length of the shadow explains the height of the flagpole!

17

Even worse:

Laws of linear optics Lf Ls ------------- α

So the length of the shadow and the flagpole explain the position of the sun!

Problem: we cannot rule spurious explanations of this kind with the means of the DN model.

18

Problem 2: Common Causes

Example:

Given that the barometer drops we know that the weather will get worse soon. The derivation of that fact has the form of a DN-explanation.

Nevertheless the reading of the barometer does not explain the bad weather. It is the drop in atmospheric pressure, which is registered by the barometer, that does.

19

Common Cause: an event that causes two other events is a common cause (of these two events). These other two events are then perfectly correlated.

Problem: Often correlated effects have a common cause. In these cases we cannot explain one in terms of the other, but a DN explanation can be formulated.

20

Problem 3: Retrodictive Explanation L: Laws of celestial mechanics

B: Positions of heavenly bodies in March

------------------------------------------

Eclipse in August

21

Problem 3: Retrodictive Explanation L: Laws of celestial mechanics Laws of celestial mechanics

B: Positions of heavenly bodies Positions heavenly bodies in March in October

------------------------------------------ -----------------------------------

Eclipse in August Eclipse in August

22

Problem 3: Retrodictive Explanation L: Laws of celestial mechanics Laws of celestial mechanics

B: Positions of heavenly bodies Positions heavenly bodies in March in October

------------------------------------------ -----------------------------------

Eclipse in August Eclipse in August

Problem: it seems that we legitimately use the constellation of the heavenly bodies in March to explain the eclipse in August; but one would not normally say that the constellation in October explains the eclipse in August.

23

Problem 4: Explanation vs. Prediction

The previous argument also highlights another issue: every prediction doubles as an explanation. Is this plausible?

à The ‘structural identity thesis’ (Hempel): explanation and prediction are indeed the same.

à Readings: Hempel 1965.

24

Problem 5: The Problem of Irrelevance

Example: Taking birth control pills prevents pregnancy

John took birth control pills for the last year

-----------------------------------------------------------

John did not get pregnant

Problem: One can construct valid DN-explanations which contain facts that are irrelevant.

25

Study question 1: How could a proponent of the DN-account reply to these criticisms? Is there a way to fix the account?

Study question 2: The aforementioned criticisms were directed against DN explanation of particular facts. Do they carry over to DN explanations of laws? If yes, how? Are there problems specific to DN explanations of laws?

26

(B) Against Necessity Problem 1: Regularities Do Not Explain

Subsumption under a general law does not explain anything – Cartwright (1983, 70-71): ‘[…] super laws […] may not be explanatory […] “Why does the quail in the garden bob its head up and down in that funny way whenever it walks?” … “Because they all do.” In the case of spin-orbit coupling it does not explain the fact the five energy levels that appear in a particular experiment to say “All carbon atoms have five energy levels”.’

Her conclusion: What we need are causes! 27

Problem 2: Singular Events

Example: Why did the markets crash in 2008? Because of sub-prime mortgages.

Problem: This seems to be a good explanation, but it does not contain a law.

Possible reply: It as it stands this explanation is incomplete; laws are tacitly assumed. This is controversial.

28

Problem 3: Quibbles with Laws

The conditions of adequacy for a bona fide DN explanation require the explanans to be true. But as we have seen earlier in the course (recall the discussion of the regularity view!), it is controversial whether laws are true.

Problem 4: Probabilistic Explanation

à next lecture

29