01 April 2012

Physics and the Social Sciences

I am writing an article whose first sentence reads: "The differences between the social sciences and the natural sciences have been greatly exaggerated."
After a couple of years reading physics and biology I got the sense that many of the presumed weaknesses of the social sciences are shared by those disciplines as well. You will have notice about it in due time.

Now, two colleagues from Rochester are making a related point in The New York Times:
"Overcoming 'Physics Envy'
"How scientific are the social sciences? Economists, political scientists and sociologists have long suffered from an academic inferiority complex: physics envy..."
See the rest here:

Previous posts on the subject in this Blog:
Cause and Effect: CLICK

How Relevant is Political Science?: CLICK

Politics and Physics: CLICK

More Scientific Politics: CLICK

The Best Academic Partners: CLICK

More Political than Science: CLICK


Helen Margetts said...

In your article, don't forget also to consider that the difference may be reducing from the other direction - now that the internet and related technologies are generating so much ('big') data - social scientists have the same sort of data as scientists - with all its advantages (huge scale 'whole of population' transactional real-time data instead of survey data) and disadvantages ('snapshot' data with no demographics or linkages to past events or other social activities, huge complexity, need for sophisticated visualization tools).

Which is why, much to my surprise, I am in the process of trying to appoint a physicist as the research fellow to work on three of my projects!

Helen Margetts
Director, Oxford Internet Institute

Rein Taagepera said...

Actually,  J of Theoretical Politics rejected a paper of mine because it DID test the model proposed -- so the authors do have  a safe outlet for untested models. 

Why reject any approach completely, on doctrinaire grounds? Here the authors go overboard.
Their description of the scientific method is simplistic to the point of being a cartoon -- and yes, all too many  poli sci publications pattern themselves on such a cartoon. 
Such sad misuse of the method is not grounds to reject it;  it can be used in a more fruitful way.

Re Downs: Presenting a model and testing it need not be done within the same article or even by the same person. Einstein never measured anything, it seems to me. 
But testing must eventually be done, one way or another. Observing several gas stations at the same corner already is an informal test. 
Re Arrow: Distinguish mathematical proofs from models about the socio-physical world. (Yes, some overlap may occur, but still..)

Likewise, the analysis of empirical data can be valuable even in the absence of a grand
theoretical model. Did the welfare reform championed by Bill Clinton in the 1990s reduce
Here I fully agree. Clinton had a model in mind, before acting. Such a study tests Clinton's model.

Rather than attempt to imitate the hard sciences, social scientists would be better off doing what they do best: thinking deeply about
what prompts human beings to behave the way they do.
This is precisely what good model building is about, involving qualitative and quantitative aspects.
 There also deep thinking in art, philosophy and religion. If this is one's approach to society, then call it social art, social philosophy or social religion, but not social science.
(Indeed, my personal religion could be called "social religion", but I do not confuse these unproved deep beliefs of mine with science.) 

U. California-Irvine, and Tartu-Estonia

Steve Coleman said...

A few days ago I heard an interview with several prominent economists on the BBC. The interviewer questioned how truly scientific economics was (or other social sciences). The economists defended the relative lack of successful predictions in economics compared to physical sciences by the fact that humans can change their actions in response to expected economic directions or policies. But is not invariance at the heart of every physical law? Can one have a scientific discipline without a principle of invariance? I would say No. Does that mean a science of social or economic behavior is impossible? No. But one must  build it on what is invariant. Fortunately, cognitive psychology and brain research have given us a lot of information to work with about invariant aspects of human cognition and behavior. This line of research may not tell us who will be the next president, or how long the recession will last, but as interesting as such questions are, likely they are not the right questions to ask of a science. 

Steve Coleman

This Blog said...

This is an interesting comment: I am still struggling with some findings in neuroscience and implications for determinism...

Steve Coleman said...

Yes, that's a great question about determinism--one that social scientists have yet to bite into. One might ask, in fact, is democracy reconcilable with what we are learning about how the brain works?

I was thinking about the question recently after the latest round of Russian elections. Many in Russia were protesting, believing that election fraud had undermined the democratic process. I interjected myself into their discussion by referring a number of Russians to the article I had published in Russia after the previous elections. A number of bloggers and election analysts then replicated my findings for the new elections. Basically I show that Russian voters are exceptionally dominated by social conformity, which substantially controls the outcome of their elections. Fraud had little to with the outcome. That sounds good at first--the democratic process is not undermined. But a bit more reflection might lead one to consider whether conformity is a more fatal blow than fraud. For conformity occurs without voters' conscious awareness, and in that sense is partly deterministic. (For links to some of the Russian blogs and my original article, see my website www.populardelusions.org).

Here are two books I have found especially good:

Timothy Wilson, Strangers to Ourselves: Discovering the Adaptive Unconscious
Chris Frith, Making up the Mind: How the Brain Creates our Mental World.

Wilson is psychology, Frith neurobiology. My book starts about where they leave off.




Anonymous Anthony McGann said...

The authors are right that the social science are different to the natural sciences, but for the wrong reasons.

The problem is not hypothetico-deductivism, but combining this with dogmatic empiricism. The problem is not that we posit hypotheses, see what they imply and then use this to test our hypotheses. The problem is that we seem to believe that the implications have to be empirical observations. As the authors point out, social scientists can sometimes prove something is impossible without needing empirical evidence (as do mathematicians and theoretical physicists).

The problem is not one of emulating physics, but of trying to emulate what many social scientists seem to think physics is.

1:10 PM  

Post a Comment

Links to this post:

Create a Link

<< Home