Wouldn't Karl Popper's work on emphasizing the importance of falsification be an example of how philosophy of science was useful to scientists?
My parents, both scientists educated in the 1950s, never heard of them. They followed a pretty simplistic notion of scientific methodology, that mainly revolved around intellectual honesty: Report everything that you observe, and be willing to change your beliefs when confronted with new information.
I never heard of Popper or Kuhn until after I finished my science degree, and then, only in the context of debates about evolution versus creationism. But I don't know if I'm typical or not.
I honestly don't know. I'd always assumed that all of that happened well before I was born, but never bothered looking for any exact dates. It's just something I've always been exposed to, and have always taken the importance of falsification as a given.
A quick search turns up this[1] from Wikipedia:
"In work beginning in the 1930s, Popper gave falsifiability a renewed emphasis as a criterion of empirical statements in science."
OTOH, the entry on Popper at Plato[2] says this:
"However, in the 1970's a series of papers published by researchers such as Miller, Tichý, and Grünbaum in particular revealed fundamental defects in Popper's formal definitions of verisimilitude."
which suggest that his work had achieved enough signficance by the 1970's that other people were basing work on it (and in this case, perhaps refuting bits of it).
The idea of falsification suffers from three issues: (1) it's binary true or false (2) it's asymmetric (falsifying a theory T vs falsifying the theory not T) and (3) it's kind of vague what constitutes falsification. Bayesian incorporation of evidence into beliefs arguably gives a superior solution to the problem, and Laplace (1749 - 1827) was already fully aware of this.
I don't know the exact history, but I suspect that people were doing quantitative science before we developed a good grasp of statistics.
Also, both Kuhn and Popper endeavored to describe science as it was already being done.
Philosophy may not be directly beneficial to the narrow practice of science, but it is critical to almost every activity surrounding it.
For example, philosophy can help scientists answer questions like:
- Does the pain caused during animal research justify the knowledge obtained from it?
- Should we fund scientists to study better methods of farming, or atomic bombs?
- Should we publish the details of the specific mutations required to produce a superbug from bird flu, or omit those details from our methods?
- What is scientific knowledge, and how does it differ from other types of knowledge?
And on and on and on.
Let's look at Feynmann's statement again, that "Philosophy of science is about as useful to scientists as ornithology is to birds."
We're supposed to read that and think 'Hah, yeah - birds don't know and can't know anything about ornithology, how can it be useful to them?!'
Of course it can be useful to them. We study birds to better understand them and their behavior. We use the results of ornithology to make their lives better by answering questions like:
- What can we do to minimize the impact of wind-power generators on the migratory patterns of birds?
- How can we save and preserve this near-extince species of bird?
- How do birds influence the spread of jungles and other forests?
And on and on and on. Ornithologists study birds, and the interaction between birds and humans - of course such knowledge will be useful to birds.
In the same way, an individual scientist can be completely ignorant of philosophy and still be a good scientist - but philosophy (and other non-scientific disciplines) support and inform every aspect of life around her job. If she wants to explore the ethical ramifications of her work, needs to justify support of her work, or discuss the process of science contrast to other disciplines, she must begin to use the terminology and processes of philosophy, not science.
[1] 'Soft sciences', such as psychology, sociology, economics, etc. attempt to study subjective phenomena but do so by directly observing behavior of the entities/systems under study.
> We're supposed to read that and think 'Hah, yeah - birds don't know and can't know anything about ornithology, how can it be useful to them?!'
I think you misunderstood the quote. Birds will not fly any differently whether ornithologists study them or not. That is what it means that ornithology is not useful to birds: the birds will not use ornithology to make decisions about flying. Feynman, as a practicing scientist, was well-placed to say whether philosophy of science informed his practice of science, so his comment absolutely does not reflect his ignorance.
> but philosophy (and other non-scientific disciplines) support and inform every aspect of life around her job.
The things that you listed won't inform how a scientist does science, so they don't really address the core of the question.
It just says "useful to scientists". I feel that my argument shows quite clearly how philosophy can be useful to scientists - maybe not in the actually practice of doing science (as that is practicing science - not practicing philosophy), but definitely useful for the scientist when he attempts to integrate his scientific findings or practice with the surrounding culture.
One might take this question, 'How is philosophy useful to the practice of science', flip it around, and come at it using your approach:
How is science useful to to the practice of philosophy?
Well, the answer there, since we've focussed on the actual practice instead of general utility, is that science isn't very useful to the practice of philosophy.
EDIT: I should qualify that last statement - scientific discoveries, when they overlap with assertions made by philosophers, can help to invalidate inaccurate philosophical assertions, and lend support to valid statements. So it can be helpful during the validation stage of the philosophical process.
If you believe that science can explain everything, your conceptual map widens the concept of 'science' to include practically everything that refers to the objective world, making the concept less useful for precise thought about epistemology.
And philosophy is more directly useful to scientists than cars are to philosophers. For one thing you're comparing a concrete artifact to epistemological processes. We're comparing processes, not artifacts. The concepts developed by philosophers have direct relevance to scientists, see my examples above.