In the What is knowledge thread, this has come up now, and I think it should be headlined:
KF, 201: >> Can we regard scientific theories as factual knowledge?
This is a deep challenge, especially on the so-called pessimistic induction that historically theories in effect have hidden sell-by dates. That is, theories show more of a track record of replacement (sometimes presented as refinement) than we are comfortable with.
A first answer is that a theory, from the abductive angle, is a “best current explanatory framework,” often involving dynamics which may be deterministic or stochastic (or tempered by stochastic factors), and may be empirically reliable in a known or unknown range of circumstances. The turn of C20 surprises faced by Newtonian dynamics have been a major lesson.
The import is, that often theories are more like models that are “useful fictions”(with perhaps a few grains of deep truth in them) than descriptions of factors at work in reality that are all credibly true. This becomes especially so where theories address remote reaches of space or time where we cannot directly observe the actual circumstances. In these cases, we are limited to observations of traces of the circumstances, and we make models of the place and time, we have not got direct checks.
Scientific simulations or scenarios and visualisations tied to such, then become even more remote from the right to claim credible truth.
Of course, actual credible observations are much better as candidates for credible and reliable truth claims.
Such suggests that we need to be far more circumspect in our evaluation of scientific theories than we are sometimes wont to be, e.g. the tendency to say of climate dynamics models and projected developments of climate under human impact, that the science is “settled,” or that those who hold appropriate background — or even laymen expressing concerns — and raise questions on key issues are “deniers.”
The future is beyond current observations, so while we may be well advised to act with prudence, we should not exaggerate our knowledge claims on the future.
Similarly, we should be cautious about exoplanet studies and especially artistic renderings of suggested planets. These are — with a few exceptions — not direct observations, they are inferred from gravitational effects. We may be confident that planetary objects are there and may infer they are terrestrial or gas giant etc, but we should be cautious.
Reconstructions of the past of the cosmos, our solar system and planet, as well as the history of life are also beyond direct observation and should be presented with due cautions. Evidence such as the detection of clear cases of dinosaur soft tissues from a claimed 65+ MYA, should give us pause. And if there are cases where the smell of death/decay is still there, that should give us pause. I know there is a recent headline on a Triceratops horn being dated to 30+ kYA, but that should be taken with a grain of salt for the moment too.
When it comes to wider senses of science such as Economics, we should be even more cautious. Even something like GDP or an unemployment rate is a calculation not an observation. Often useful, but use with due caution.
I begin to suggest that we view theories more like models of high reliability that we hope capture something significant regarding the true dynamics of our world, but we are less than certain of that. The theories may be part of the body of knowledge of a field of study, but that is a matter of observing the field of study as itself a phenomenon subject to observation and evaluation. The credible truthfulness of the contents of a given theory and its key objects or processes and laws etc are something that we should likely take a very eclectic case by case view on. No-one has actually directly observed an electron, but we are highly confident that these entities exist, never mind weird quantum properties of such a “wavicle.” We can make a much better case for more or less observing an atom, given scanning techniques.
The remote future, or remote reaches of space or the remote past of origins, we do not directly observe. We would be well advised to be cautious, and to bear in mind the limitations of inductive methods of investigation.
Ironically, on the design inference debates, the reality of something like FSCO/I [= Functionally Specific Complex Organisation and/or associated Information] and its empirically observed origin are far better observed than the suggested deep-time powers of chance variation and differential reproductive success. But institutional power makes a big difference on how things are perceived. Which, is yet another caution: scientific “consensus” or the ex cathedra statements of august panels and their publicists should be taken with a grain of salt.
Science at its best is openly provisional and open-ended.>>
Again, food for thought. END