reading quiz 9/23 IBE

inference to the best explanation (IBE)

type of non-deductive inference. Nothing hangs on which choice of terminology we favour, so long as we stick to it consistently.

examples of IBE in science

Darwin’s theory of evolution. If we assume that current species have been separately created, it would be hard to explain the living world. But it makes perfect sense if current species have descended from common ancestors (theory).

IBE as a more fundamental pattern of inference

implicitly appealing to explanatory considerations (all pieces of metal examined so far conduct electricity, so all pieces of metal conduct electricity). Ordinary induction dependent on IBE.

IBE as being parasitic on ordinary induction

When we try to decide which of a group of competing hypotheses provides the best explanation for our data, we invariably appeal to knowledge that has been gained through ordinary induction. Thus, incorrect to regard IBE as a more fundamental mode of inference.

Criteria to determine which of the competing hypotheses provides the best explanation of the data?

The best explanation is the simplest and most concise one.

Scientific realism

the aim of science is to provide a true description of the world

Anti-realism

the aim of science is to provide a true description of the “observable” part of the world. It makes no odds whether what science says is true or not.

“observable part” of the world

things that can be directly perceived by human beings

science branches that are observable vs not observable

paleontology vs physics

realists and anti-realists disagree on ___ branches of science

unobservable, like physics

realists argument for branches like physics:

theories about electrons (ex.) are trying to provide a true description of the subatomic world, just like paleontologists

anti-realists for unobservable branches like physics:

unobservable entities are just convenient ways of predicting observations, not to truly describe the hidden facts. (e.g kinetic theory of gases)

first type of anti-realism:

talk of unobservable entities is not to be understood literally at all. not possible to make meaningful assertions about things that cannot be observed. unpopular now.

second type of anti-realism:

talk of unobservable entities should be taken at face value. scientists’ claims are either true or false, but we are incapable of finding out which. popular now.

theories are “empirically successful”

make excellent predictions about the behavior of objects in the observable world

‘no miracles’ argument + intention

would be a HUGE coincidence (like a miracle) if a theory that talks about electrons and atoms made accurate predictions about the observable world—unless unobservable entities actually exist. better not to believe in miracles —> be a realist!

intention: to give a plausible argument; an inference to the best explanation. NOT to prove that realism is right and anti-realism is wrong

anti-realists rebuttal for ‘no miracles’ argument

history of science shows that empirically successful theories have often turned out to be false. (e.g. wave theory of light: light consists of wave-like vibrations through ether, an invisible medium). this theory was empirically successful, but was not even approximately true.

main argument for scientific realism

not possible to draw the observable/unobservable distinction in a principled way.

problems associated w/ observable/unobservable distinction:

distinction between observation and detection (e.g. electrons)

Maxwell’s argument for scientific realism

If something can be seen with scientific instruments (binoculars, microscopes, etc.), does it count as observable or unobservable? How sophisticated can instrument be to distinguish from detecting than observing? Maxwell argued there’s no principle way of answering questions —> anti-realism’s attempt to classify entities as observable/unobservable is wrong.

van Fraasen’s argument for anti-realism

claims that Maxwell’s argument only shows ‘observable’ to be a vague concept. van Fraassen claims that vague concepts are perfectly usable and can mark genuine distinctions in the world. same applies to ‘observable’

vague concept

one that has borderline cases (cases that neither clearly do nor clearly do not fall under it). e.g. bald: hair loss comes in degrees.

anti-realists’ argument on underdetermination

observational data underdetermines the theories scientists put forward on unobservable entities. there will always be a number of competing theories that can account for that data equally well.

realists’ rebuttal on underdetermination

yes there will always be more than one theory that explains an observation, problem of underdetermination can be eliminated in practice by using criteria (simplicity, best explanation) to discriminate between competing explanations of observational data

ex: one theory might be simpler than the other, or might explain the data better.

additionally, if you apply underdetermination argument consistently onto observable world, forced to conclude that we can only acquire knowledge of things that have actually been observed