The art of science

By Razib Khan | August 14, 2012 8:26 pm

Over at Scientific American Blogs Maria Konnikova posts Humanities aren’t a science. Stop treating them like one. The whole write-up leaves me scratching my head, because I don’t really get what the whole point of all the prose is. This is a thesis that is as old as 19th century romantics, and not all too complicated. The author herself has an academic webpage which indicates she works within an analytic framework that’s anything but “soft.” There are huge confusions with terminology, and Jerry Coyne has a response which addresses many of my questions (e.g., what exactly is the alternative to doing statistical tests in psychology? Rely on the impressions and intuition of the researchers and just trust them?). But let me highlight one section:

… Societal conventions change. And is today’s real-world social network really comparable on any number of levels to one, say, a thousand, or even five or one hundred years ago?

Yes, today’s real-world social network probably is comparable to those of the past. There is some science on this issue. Not even rocket science with abstruse statistics. Science which is highly relevant today. Question science, and it may surprise you with what it has discovered!

On a real scientific note: the great thing about science as a system and method, rather than individual erudition, is that it advances. Consider the case in phylogenetics. Before the cladists you had to rely on the expertise of domain specialists, and perhaps some phenetic similarity methods. Now everyone is shifting toward Basyesian phylogenetics (or maximum likelihood if you want speed). Yes, the statistics can be daunting and the computation herculean, but this sort of progress actually makes knowledge more democratic and precise. You don’t have to have devoted your whole career to one small branch of the tree of life to say something about the nature of evolutionary processes.

I can grant that there are problems, and sometimes negative returns, on attempts to quantitize or formalize every domain of scholarship. And I do agree that I am skeptical of the drive for science to swallow up all the territory of scholarship. But Maria Konnikova’s broad and expansive argument rankles me because it seems reminiscent of reactionary nostalgia for the age when there was true wisdom in gurus and initiates, those who had developed powerful intuition about specific subjects through means esoteric or ascetic. I don’t assert here that the author has any such intent, rather, I suggest that her broadside against scientification substantively supports the repeal of much of the democratization of knowledge through the dissemination of models and methods.


Comments (7)

  1. I can grant that there are problems, and sometimes negative returns, on attempts to quantitize or formalize every domain of scholarship.

    I didn’t see her as saying anything more than this, albeit vehemently.”It seems reminiscent of reactionary nostalgia for the age when there was true wisdom in gurus and initiates, those who had developed powerful intuition about specific subjects through means esoteric or ascetic”is an overstatement. Take out the “gurus and initiates part” and just the part about powerful intuitions, reading and finding the value in literature or an idea, making a case without reference to statistics or number-crunching.

    The author is unhappy at the fact that the humanities has moved to the “science” side of things and that it’s moved there because of the perception that this form of “hard” analysis is more reliable. You (I think) agree that it is, in fact, more reliable. To argue about that is to argue about the heart of the difference between sciences and humanities. I think humanities lately has a deservedly bad rep, but I agree that the attempt to “quantitize” the humanities is more than a bit depressing, although I’m not a scholar in any subject.

    All that said, the “killing Hitler” cartoon cracked me up.

  2. #1, i think her net is WAY too wide. e.g., keep poli sci and psychology imprecise and verbal? i don’t think so.

  3. I read it, and had a very similar reaction. I’m preparing a course in theory of science for psychologists right now, and they treat the questions of hard sciences vs hermeneutics in chapter 2, which is basically dedicated to get the students up to speed on what has been chewed over for the past however many centuries of philosophizing about the nature of things, so we can get on with the business of theory of science and eventually current thinking in philosophy of psychology. Intro masters level. We have taught this for eons to our budding therapists (who we try to gently nudge out of that mind-set).

    Also seemed a bit like a warmed over bit from the science wars. Like, where have you been the past 15 years? (then again, I figure I’m twice as old as she is).

    I do have to encounter similar arguments – mainly from the therapy students, but it is also in some of the older faculty (on the therapy side).

    I also don’t understand why quantifying would ever be depressing. The work she described – using network theory to see if one could distinguish between old stories that could be based on real events, vs completely fictional is fucking AWESOME! For no reason other than it is AWESOME.

    And, having tried to quantify seriously (I mean, not just taking the old trusty anovas and applying them to experiments designed so ANOVAS will be the appropriate quantifying tool) in mathematical psychology (and god knows, I’m not particularly good at that one), and having encountered people who really did it (Like recently deceased Duncan Luce, Bill Estes, etc) what quantifying does is not lull you into some superior sense of certainty, but reveals very starkly the big gaps in your understanding of phenomena.

    The narrative and imprecise has a part in research, but the big problem with it is that it is so easy to fool ourselves into thinking we understand something, to miss gaping holes in logic, and justify our beliefs. You do that in the quantitative too, of course, but you have to work a bit harder at it.

    Well – maybe I should write a blog post about this after all…

  4. #1

    I think it’s depressing that we want to grant immunity to certain ideas & thoughts or disciplines from mathematical deconstruction just because it’s too difficult or messy. In reality, I hardly think that the concerns are truly motivated by a fear that we will get things wrongs or overly simplify things. I think the real motivations are the same as those of Keats when he accused Newton of reducing the rainbow to a prism etc.

    The entire tone of the article was so defeatist and dismissive of the scientific enterprise and the examples offered to ‘prove’ her thesis were even more puzzling. Linguists have a problem agreeing about statistical methods (referring to a crappy New Yorker article that gets basic facts wrong, that Pinker panned) Cliodynamics won’t ever tell us something until it has happened? What kind of a bullshit criticism is that? We didn’t have counterfactual definiteness in any domain of science in the past. A couple of thousand years later, we do have it in a lot of domains that go beyond STEM. All the people who are working in complexity sciences will be laughing at this article hundred years from now. According to this author, we should just give up right now, because it’s too hard.

    Was she just trolling?

  5. Patrick Wyman

    It’s not like qualitative and quantitative methods are mutually exclusive. This is what’s puzzling to me about the piece: is the author really so threatened by mathematical models and scientific methodologies that that she has to condemn them altogether? Quantitative analysis can’t answer every question, this is true, but neither can analyzing anecdotes. Depending on the questions you’re trying to answer and the body of evidence at your disposal, one or the other or *gasp* both might be the most appropriate tool for the job.

    I work in the humanities – history, specifically – and Konnikova raises some valid points about the limitations of quantitative analysis, some of which are common enough among historians (especially of the 70s and 80s) that they bear repeating. With that said, every methodology has limitations, and the point is figuring out how to limit the drawbacks while maximizing its potential to tell you meaningful things. Discomfort isn’t a valid reason to dismiss a whole realm of analysis.

  6. prasad

    I like how her go-to example of improper humanities-quantification is an applied mathematician publishing a paper in EPL applying statistical methods to networks. Which humanities department is even affected here? Also, trying to find patterns in relationships that separate factual from fictional narratives? That’s cool like sharks with lasers are cool.

  7. Anthony

    There is an argument that formalizing the “human” sciences (I don’t want to call them social sciences for several reasons) does not mean mathematizing them – the folks in the Austrian School of economics argue that due to the nature of the subject of study, mathematical prediciton is impossible in economics, sociology, and psychology. (See, for example, or )

    There is also the argument that literature and the other arts teach us things on an emotional level which would be hard to reach by using the analytical tools of math and logic. That argument certainly doesn’t save the social sciences from the needs for some formalism, whether or not it’s mathematical, but it is a reasonable argument for teaching of the humanities. Does it matter if the Iliad is “true”, or “truthy”? Yes and no. It is interesting, but the benefit one might gain from reading it does not depend on whether the story is literally true, “based on a true story”, or invented from whole cloth. It resonates the way it does, and the way it has over two and a half milennia because its characters are plausible – they tell us real things about the human condition even if there never were specific people who did those specific things. (This is, incidentally, why Richard III works *so* well when done in 1920s/30s costume, as I’ve seen done.)

    However, the blog post seems to be something more than that – an emotional reaction against math invading “her territory”, or something. Perhaps someone could do a quantitative analysis of math-phobic articles like this and the infamous Hacker op-ed against algebra, and find the psychological roots of this phenomenon.
    However, this


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


RSS Razib’s Pinboard

Edifying books

Collapse bottom bar