Friday, 8 February 2013

'Anthropology Is Not A Science'

Gene Expression by Razib Khan is a wonderful blog about the human sciences by a sensible, scientifically-minded chap who also manages to write clearly about complex topics - even about race, a topic that is nothing if not a minefield.  I've followed Gene Expression on and off for years, and I was subscribed to it while it was part of the ScienceBlogs fold.  It left a long time ago, as did many other excellent blogs, causing an unfortunate fragmentation of the science blogosphere and making it harder to keep up with the blogs I used to follow; many of them have joined other blogging stables, including Freethought Blogs (which has largely abandoned the topics covered by ScienceBlogs) and Discover magazine (which has taken the lion's share of the good ScienceBlogs content).

Razib covers some very interesting material, and, fortunately, he's not afraid of a touch of controversy.    I was happy to see that he has his own take on the controversy surrounding Jared Diamond's latest book.  In looking at the book, he also takes on the furore among academic anthropologists (and Stephen Corry, an indigenous rights advocate) surrounding it.  He thereby addresses an issue close to my own heart: the fact that social anthropology is not taught in a scientific way.  His post has aroused plenty of controversy among cultural anthropologists, who don't see their work as scientific, but who are (inexplicably) quite proud of the fact.

The fact is, all the good things in anthropology are scientific, and all the bad things aren't.  The 'anthropology isn't a science' position is designed to make the bad stuff acceptable, as far as I can tell (although I've had heated discussions about this with academic anthropologists sympathetic to the scientific position, who nevertheless claim that anthropology isn't 'a science').

One thing I find annoying about this debate - not the Jared Diamond debate, but the anthropology-isn't-a-science debate - is the idea that science breaks up neatly into discrete sciences.  It doesn't.

Geology isn't a science; it is a part of science, the attempt to find out about the universe rationally and sensibly.  Science isn't divided up into things that are each 'a science'.  It seems pretty obvious that not only do scientific disciplines overlap considerably, there isn't actually any reason beyond the division of labour to see them as separate at all.  If we could become experts in everything, we wouldn't need these divisions: we would just 'do science', from particle accelerators to taxonomy to the analysis of sound changes and potsherds.  Science is a single project, an amazing project that makes life really worth living, and it is only the short duration of our lives and the inability of humans to master everything they encounter that prevents us from seeing this.  Anthropology and archaeology are not 'sciences'.  They are part of science, the great big project, in that their subject matter (human life, history, and diversity) is clearly part of the universe and can clearly be studied using essentially the same tools as any other subject studied scientifically: reason and evidence.

Anthropology isn't a science.  It's part of science.  And anthropologists shouldn't try to get a free pass from scientific standards by arbitrarily differentiating their discipline from 'science' as a whole.  I'm tired of people saying that an entire discipline - an ostensibly empirical one - isn't part of science.  It's especially stupid when it comes from a discipline that produces unemployable graduates (the least employable of all graduates, in fact) and refuses to provide any practical skills to its students.  Statistics isn't part of the anthropology curriculum, which would be fine if there were another formal method of equal proven validity from within the discipline (unlike linguistics, anthropology has no such method).  The earlier formal study of pre-state social structure ('kinship studies') used to provide a solid backbone to the discipline in encouraging logic and analysis, but this is now largely absent from anthropology curricula (as I found to my chagrin at Oxford - I find so-called kinship analysis exceptionally interesting and useful, and I was surprised to find it absent from exam papers).  Even critical thinking and logic are disparaged as attempts to police the thoughts of others and impose a mental hegemony.  This couldn't be less useful or interesting.

There isn't really a purpose to anthropology.  I've seen a good few blogs claiming that the purpose of anthropology is something along the lines of 'making the world safe for human difference', but this isn't the point at all.  That is a prescriptive point, not a descriptive one, and making that the mission statement of anthropology would take plenty of interesting anthropological topics away from the discipline of anthropology.  I don't want to make the world safe for the return of headhunting or human sacrifice, and yet studies of those topics are clearly key to understanding pre-modern human cultural and social diversity.  They are clearly anthropological topics.  Making the world safe for human diversity is a nice aim that can be pursued in other venues, but it isn't what anthropology is or should be.

Anthropology courses actually consist of nothing in particular.  The discipline isn't actually based on a subject, in the sense that geology is based on the study of the earth and biology is based on the study of life.  Anthropology as it is taught in most universities is based not on the study of human beings, or human beings in non-state or non-literate societies (as it once was), but rather on the study of anthropology.  What I mean is that the content of an anthropology course - and I wish I was making this up - is primarily the history of anthropology.  You hear about the 'structural-functionalists' and the 'structuralists' and the 'practice theorists' and so on; you read the books of prominent theorists from each school, going over their ideas without any attempt to refine or change them or place them in the light of scientific inquiry.

The purpose isn't to advance present day understandings of the subject matter of social anthropology, and in fact many of these 'theoretical approaches' aren't theoretical approaches to any subject at all, but rather wholly different ideas as to what the word 'anthropology' means (what a nihilistic process this is!).  Modules on the history of anthropology don't consist of a few brief lectures intended to ensure that students avoid making the mistakes their predecessors made in approaching a clearly-defined subject matter - they take up the bulk of the course itself, and even the bulk of the examination content.  The purpose seems to be to provide meat for a course in a discipline that has no real subject matter that can be agreed on by all of its practitioners.

This couldn't be less useful.  It couldn't less trite and stupid.  It amounts to this: 'Once upon a time, there was a discipline called anthropology that was in cahoots with colonialism and is therefore suspect.  Today, anthropology is a self-reflexive discipline that challenges earlier stereotypes of what anthropology is.  Here are some books to read that represent each level of change; we don't really have any representative works of the current consensus, because there is no current consensus, so try to make the best of these opaque, unscientific works and try to avoid despair when confronted with the absence of valuable subject matter.'

Some anthropologists claim that anthropology isn't in trouble.  But it is.  Employers believe (justifiably) that social anthropology graduates haven't really learnt anything and don't tend to have any practical skills beyond those they have taught themselves.  Other academics believe that anthropologists tend to produce politically-inspired works lacking scientific rigour.  They also know that anthropologists seldom look to other parts of science for ideas, and there is a clear and obvious (one-sided) enmity between social anthropologists and cognitive scientists, despite the equally clear and equally obvious applicability of insights from cognitive science to the subject matter of social anthropology.  This is all harmful to science.  And unfortunately, it seems as if employers and other scientists are justified in their view that anthropology is mostly rubbish, and I completely understand - although I don't condone - the fact that most of these academics and businessmen won't look through the shitpile to find the roses.

The whole enterprise is in trouble, and, from a student's perspective, there is absolutely no reason to study social anthropology as it is found in universities today.  It currently deserves its unfortunate and shameful reputation as a doss subject, and this is terrible for people like me with qualifications in social anthropology, terrible for the study of certain neglected topics (like non-industrial human societies), and terrible for science as a whole.


  1. Really enjoyed this article. "Anthropology courses actually consist of nothing in particular." I agree with this as I have dropped several upper level anth classes because nothing practical was being taught (stories about Nepal are very nice but...) but I have taken some great classes where practical and relatable information was taught.

  2. Well, I don't mean that anthropology courses should be 'practical', exactly, but just that they should aim to be correct, accurate, and interesting, and you should come away with some identifiable knowledge or skill that you couldn't get from some other activity. But an anthropology course is unlikely to give you this unless you have a fantastic tutor or are willing to sacrifice time and possibly grades to study certain topics on your own. And that's not how the academy is supposed to work.

    To me it seems a bit broken. I understand why other people would disagree, especially as there are plenty of people with a vested interest in keeping anthropology on the course it is currently on. But the fact that anthropology graduates are so unemployable is a sad indictment of the state of the discipline.

  3. Interesting article. What would you like to see incorporated into anthropology departments to improve both the study and employability of graduates? What sort of skill sets?

    1. If I started to change anthropology departments around, they wouldn't be recognisable (to anthropologists, at least). So I ought to hold myself back.

      But what I would include, if given free rein, is some understanding of phonology, semantics, linguistic typology, and language reconstruction - not a full linguistics degree, but components of one relevant to understanding culture history. Also units on excavation, seriation, chemical dating, etc, and on the problems of conducting and writing about ethnographic fieldwork (the latter being standard fodder). I would make exams and modules more knowledge-focused - the depth of area studies but with enough comparative content. A modern language - Spanish, Arabic, Indonesian, Chinese, Hindi, Russian - would be a plus for a number of reasons, but that's getting more into the realm of a joint degree.

      The skills I would ensure anthropology graduates had would not necessarily be directly relevant to jobs or careers outside of the academy. They would be useful for solving research questions, and would confer the kinds of general skills employers want. If you speak to HR departments at big firms, they don't often want marketing or management graduates - they want graduates with an interesting background and interesting skills who might be able to look at problems in different ways. I don't think anthropology departments currently prepare graduates for this, much as it is assumed that they do, so I'd actually try to make anthropology less relevant to the modern day, but also more problem-oriented and therefore more employable. Archaeology, linguistics, classics, and art history graduates are all more employable than anthropology graduates, and that's not because the skills classicists have are directly relevant to sales or HR.


You can post anonymously if you really want to, but I'd appreciate it if you could provide some means of identifying who you are, if only for the purpose of knowing who has written what.