The notion of progress in the social sciences

[Stephen Tashi]

What constitutes evidence for progress in the sciences? - including the social sciences!

I see two basic criteria.

1. The expert overview: An expert in a science can survey historical development and understand that current theories are better explanations that past theories. If the concept of progress applies then current theories are compatible with the theories they replaced, in the sense that although current theories may disagree with past theories, current theories also explain the experimental and empirical evidence that led to past theories and include past theories as approximations or special cases.

2. Technology: Progress in technologies that use a scientific discipline can be regarded as indirect evidence that there has been progress in the science.

Applying those criteria, do we find progress in disciplines like psychology, sociology, economics etc?

To do the expert overview would require an expert, so I can't apply that criteria.

Using the criteria of technology, one might argue that internet marketing, tracking of user's interests etc. is progress in technology. But is this sort of progress due to advances in psychology or economics? - or is it primarily the work of people untrained in those disciplines? I have the impression that it's mostly the work of people untrained in social sciences, but I don't have personnel data for any internet companies.

 

[TeethWhitener]

Applying those criteria, do we find progress in disciplines like psychology, sociology, economics etc?

Applying those criteria, do we find progress in biology, chemistry, and physics? I’m not sure spontaneous generation; the four humors; phlogiston; the elements of air, fire, water and earth; or Aristotle’s notion that $F=mv$ are approximations or special cases of subsequent theories.

 

[russ_watters]

Applying those criteria, do we find progress in disciplines like psychology, sociology, economics etc?

It may be more vague and takes longer to measure or make changes (because the applications change more slowly and in more difficult ways to measure than in science/technology), but I would say the way you do it is the same: by examining history and seeing if things are better now than in the past and in what ways.

[edit]
Hmm...Newton published Principia in 1686, and the next big breakthrough in gravity, as far as I know, was GR in 1915. Lots of improved measurements were made in between, but the theory itself didn't change. But an awful lot changed in both the theory and applications of the social sciences in that time.

 

[BillTre]

An additional consideration might be the expansion of what is included in the field.
It may be that there are a lot of new sets of data that are being considered that were not before and "progress" might just coming to a way to deal with the new information sets.
I don't really think a lot of social sciences are mature enough at this time to have an overall all encompassing theory that covers everything.
If so, then just bringing together a new kind of data might be considered progress.

 

[Stephen Tashi]

Applying those criteria, do we find progress in biology, chemistry, and physics?

I'd say yes, beginning at a certain point in history - not beginning at arbitrarily distant times in the past.

 

[Stephen Tashi]

[edit]
Hmm...Newton published Principia in 1686, and the next big breakthrough in gravity, as far as I know, was GR in 1915.

There need not be breakthroughs to have progress. The Newtonian approach to mechanics was greatly elaborated through the years. As one example: Hertz's work on contact mechanics, 1882 according to https://en.wikipedia.org/wiki/Contact_mechanics

 

[etotheipi]

I’m not sure ... Aristotle’s notion that $F=mv$ are approximations or special cases of subsequent theories.

This might not be so relevant here, but a while ago I came across this paper by Carlo Rovelli that discusses how some of Aristotle's results (e.g. that, in modern terms, $v \sim c(\frac{W}{\rho})^n$ for a body falling in a fluid) arise from approximations of Newtonian physics in a fluid (with a spherically symmetric gravitational field).

Although now we know that $F = mv$ is wrong, it would have been much harder to know this without an understanding of dissipative forces or indeed a framework for handling "forces" at all! But the result still carried with it some intuition of the physical world, some of which can be rationalised with full Newtonian mechanics treatments!

 

[russ_watters]

There need not be breakthroughs to have progress. The Newtonian approach to mechanics was greatly elaborated through the years. As one example: Hertz's work on contact mechanics, 1882 according to https://en.wikipedia.org/wiki/Contact_mechanics

I didn't mean to imply that there was no progress at all, just no major/groundbreaking progress. By contrast, all of the world's democracies are <250 years old. Switching from a monarchy to a democracy is a groundbreaking change in social/government structure.

 

[TeethWhitener]

I'd say yes, beginning at a certain point in history - not beginning at arbitrarily distant times in the past.

Isn’t “a certain point in history” a synonym for “arbitrarily distant times in the past?” As a possibly-relevant aside: vitalism (the theory that only living organisms can produce organic compounds) persisted well into the 19th century.

My point is that I’m not convinced that a new theory must incorporate past theories as approximations or limiting cases to supersede them. It only needs to explain observations with the same or better accuracy (and ideally predict new phenomena that older theories failed to anticipate).

 

[Stephen Tashi]

Isn’t “a certain point in history” a synonym for “arbitrarily distant times in the past?”

No. "There exists a point" doesn't mean "For any given point".

My point is that I’m not convinced that a new theory must incorporate past theories as approximations or limiting cases to supersede them.

If we don't require the notion of making progress with a new theory to mean "building upon past work", I agree. However, as I indicated in giving my criteria, I do include that notion. If we don't include the concept of building upon past work, then what an average person would call "fads" in psychology, sociology etc. would be progress simply by virtue of being new.


If past theories are based on empirical facts and somewhat successful at explaining them, a successful new theory should not make predictions that contradict those empirical facts. So, somehow, the new theory must explain why the old theory was successful in those cases where it worked.

 

[TeethWhitener]

So, somehow, the new theory must explain why the old theory was successful in those cases where it worked.

Ok but this could be as trivial as “the old theory was just wrong and somehow got lucky.” Hence my example of spontaneous generation.