Optional theme: Philosophy of science practice questions.

Popper's criterion of demarcation counts a theory as scientific only if it

• has been confirmed by a large and varied body of favourable instances
• rules out some observable outcome and so could be shown to be false
• can be conclusively verified by a finite set of observation reports
• is stated in a vocabulary whose terms all refer to measurable quantities

Answer: rules out some observable outcome and so could be shown to be false

Popper's criterion concerns the logical form of a theory rather than how well supported it is; confirmation by instances is the mark of the inductivist tradition, and conclusive verification is the strong criterion the early Vienna Circle aimed at, which Ayer himself gave up in favour of a weak verifiability requirement.

An astronomer's prediction fails. Rather than give up the theory, she concludes that the instrument was poorly calibrated. Which thesis about testing does this decision illustrate?

• observation statements are neutral records that any acceptable theory must accommodate
• a single conflicting observation is enough to refute a universal claim
• hypotheses meet experience as a body, so a failure does not say what to revise
• the terms of a theory are exhaustively defined by measurement operations

Answer: hypotheses meet experience as a body, so a failure does not say what to revise

This is the Duhem-Quine point that auxiliary assumptions are always under test alongside the hypothesis; it does not show that theories are immune to evidence, only that logic alone does not identify which premise to give up.

Two radiologists examine the same image and report different features, each guided by the categories of their training. Which claim about scientific method does the case support?

• a crucial experiment can settle a choice between two rival theories in one test
• generalisations are established once a sufficient number of instances is collected
• hypotheses must be deduced from higher laws before they may be tested
• what is registered as an observation depends on the theory brought to it

Answer: what is registered as an observation depends on the theory brought to it

Theory-ladenness undermines the idea of a neutral observational base against which theories could be checked one by one; it does not show that evidence is irrelevant, only that data and theory cannot be separated as cleanly as the empiricist picture assumed.

The pessimistic meta-induction argues that

• today's successful theories are probably false, as earlier ones proved to be
• the success of mature theories would be inexplicable unless they were roughly true
• the evidence available at any time is compatible with rival incompatible theories
• no finite body of observations can establish a statement of unrestricted generality

Answer: today's successful theories are probably false, as earlier ones proved to be

The argument runs from the track record of discarded successful theories to doubt about current ones, so it is the standard reply to the no-miracles argument for realism and is distinct from underdetermination and from Hume's problem.

A commentator claims that a question science cannot answer is not a genuine question about the world at all. This position is best described as

• naturalism, the view that there are no supernatural entities
• scientism, the view that science exhausts genuine knowledge
• instrumentalism, the view that theories are devices for prediction
• empiricism, the view that knowledge derives from sense experience

Answer: scientism, the view that science exhausts genuine knowledge

Scientism is a claim about the reach of scientific method, so it is stronger than naturalism, which is a claim about what exists; a naturalist can consistently deny that science answers every genuine question.

Nagel argues that an account of the world aiming at complete objectivity runs into a difficulty because

• no finite body of evidence can establish the universal laws it states
• every observation on which it rests is shaped by an accepted framework
• it must leave out the point of view from which experience is had
• its terms would have to be defined by measurement operations alone

Answer: it must leave out the point of view from which experience is had

Objectivity for Nagel proceeds by subtracting perspective, so the subjective character of experience is exactly what the method cannot capture; this is a claim about the limits of a method, not the sceptical claim that the external world is unknowable.