Science
Anti-Realism can generally be codified as a kind of instrumentalism: science gives us useful answers, that don’t have to be real and probably aren’t. Aiming mainly for empirically adequate descriptions.
Tbere is a difference between the aim of science and that of scientists.
Both realism and anti-realism:
We can have knowledge of observables
We should aim for true theories of observables
Though they disagree on:
Realists hold that we can have have knowledge of unobservables
Realists think we should aim for true theories of unobservables
So the issue hangs at the intersection of observables-unobservables.
No miracles argument by Putnam:
’Realism is the only philosophy that doesn’t make the success of science a miracle.’
If we look at the details of the theories which make accurate predictions, it wouldn’t be possible unless the theories were actually tangenting truth.
It relies on the idea that science is highly successful.
Success for science is to make super accurate predictions.
Anti-realist response:
Theories undergo a natural-selection process: no surprise that theories are accurate. Each theory is pruned so as to get closer to good estimations at each forward movement in time. Ie. It is not random, but small changes over a long time reach results which might at first hand seem improbable. Denying this would be to deny the adaptability of species because of evolution.
The base rate fallacy:
Imagine getting tested for covid:
Rate of false negatives = 0
Rate of false positives = 5%
Your result is positive!
What is the probability that you have covid?
Need more information:
You actually need to know how common it is in the population, if it is 90% of the population, false positives probably don’t matter, and if it is only 1%, it suddenly starts to matter a lot more. This is the base rate.
Given a base rate of 1/1000, we get a 2% chance that we have covid given that it is positive.
If 1000 people are tested, about 50 get a positive result, but only 1 person actually has it.
The positive result on the test does not automatically imply that you probably have a virus, as such, an accurate theory does not imply that it is approximately true. You need the base rate of the theory. But how can we know that in regards to theories?
Statistical argument
Theoretically independent techniques to view unobservables show the same thing: eg. light microscope and electron microscopes. → we should think that what these observe is ’true’
However, the techniques are often constructed for the sake of corroborating prior theories.
Phlogiston theory:
As a flame burns, it releases phlogiston, and if you trap it in a small point, the phlogiston will build up until the flame cannot release anymore and so is squeezed into burning out.
Thought that adding phlogiston to a rusted material will restore it to its normal weight, ie. Phlogiston has a negative mass.
The theory was quite successful, and not really that far from the current accepted truth.
Given that we have had empirically adequate theories in the past, but which turned out to be false, we should probably assume that the same remains true now.
-
Almost all theories in the history of science have turned out to be false.
-
There is no reason to think that current scientific theories are more likely to be true.
→ most current scientific theories are false.
Realist would say that theories are becoming more and more true throughout history.
Should we treat the unobservable like the unobserved?