For whom does ‘theory’ work, and for how long?

Long-time readers of this blog will maybe have noticed that I’ve had a long journey with regards to what the discipline classifies as ‘theory’, which is probably to say, any use of terms outside our evidence to explain it. When I started writing for the Internet I was basically hostile to such approaches, which got me into a certain amount of argument. That argument however, because it was with people I respected and of whose work and findings I could often see the point, and finding some such work that did something useful for me, got me looking for more such work. This also entailed me becoming more ready to admit that, of course, we all carry round interpretative models as a result of our education and that in many ways an explicitly theorised approach may be more intellectually rigorous than one which doesn’t recognise a source for its ideas. (Mine seem largely to be Marx, tempered also with Foucault and Bourdieu, all of which I acquired second-hand but whose effect on my thinking I can’t deny even so.) By now, I am not only ready to admit that there is a lot of useful ‘theoretical’ work out there, even if amid a sea of stuff that serves no purpose other than to badge its producers as belonging to a group, but to complain that we need more theory and even, tentatively, to start trying to work it out.

It was in more or less this case that I had one of several avid conversations early last year on such matters with an archæologist, who would not thank me for naming them. I had at that point just read a piece by Chris Scull which struck me as being an excellent case of the adaptation of theories: it tries to understand the formation of kingdoms in southern Anglo-Saxon England in terms of systems theory, finds that inadequate and modifies it with some ideas on peer-polity interaction to produce a result that seems to me to make sense.1 It is, however, an explicitly processual appproach, which we are of course supposed to be over now, and so I perhaps shouldn’t have been surprised at said archæologist’s reaction, which was basically to tell me the approach was too old to be valid.

Now, perhaps I’m just a historian and thus automatically interested in the idea that knowledge of the past can inform the present but that didn’t seem to me to be a priori true. Surely an idea is an idea is an idea however old it is, and while the context in which it was generated will be important to it a good one might still be transportable. Further argument about this established that said archæologist believed that processual models were to be discarded because they had often been proved to be wrong by later work, which perhaps simply exposes my vulnerability here: since we will, really, never know how kingdoms actually formed in southern England in the sixth century, by finding Scull’s version plausible all I’m really doing is stating a preference in the secure knowledge that it doesn’t matter a damn one way or another. The archæologist’s position here was implicitly founded on the axiom that there is actually a right answer, about which the historical discipline can waver towards the agnostic to say the least, but in this case that right answer is probably beyond recovery, making all guesses much more equivalent in worth. But even if the archæologist here wasn’t doing it, there are others out there who think that novelty is sometimes too important a factor in interpretation.

Now, this is a thing to which medieval historians seem especially vulnerable. I have taught on methods courses where nothing published in the last twenty years was involved except from the literature side. Marx is still a big force on us (and rightly so, in my view; even if we don’t like his answers any more, he asked the right questions). Our anthropologists of resort are still Clifford Geertz and Pierre Bourdieu, where they’re not E. E. Evans-Pritchard or even James Frazer (a situation to which an anthropologist I used to be attached to reacted with the words, “Oh, it makes the Baby Jesus cry!”). And where frontiers come under discussion, the name of Frederick Jackson Turner still gets invoked, bewildering any Americanists we know, they having forgotten the man years ago. We are awfully outmoded. But it is a question at least partly of ‘mode’, in the French sense, I think, not utility, because the reason we keep these tools around is because they are useful to us. The reasons for that use may often be somewhat ugly and political, but one could be cynical and say that so is our work, again whether we realise it or not.

So I suppose this is one of those posts where rather than having opinions of my own, I’m interested in yours. The unnamed archæologist’s critique of Scull’s piece is as close as I’ve so far met to an argument against ‘old’ theory that wasn’t fundamentally about fashion-currency: when its results are found wrong old theory must be deprecated. Fair enough, except that as I say, just because Communism failed and the working-class revolution seems further away than ever doesn’t mean that it’s not important to ask who controls the means of production, does it? Do you feel guilty whenever you cite Bourdieu or Geertz for knowing they’re dead and that you’re ignoring two generations of subsequent work by people presumably as brilliant? Or will those tools still serve, and if so for what? Are our enquiries themselves two generations out of date, and is that a problem if so? What do you think about what you think?

1 C. Scull, “Social Archaeology and Anglo-Saxon Kingdom Origins” in Tania Dickinson & David Griffiths (edd.), The Making of Kingdoms. Papers from the 47th Sachsensymposium, York, September 1996, Anglo-Saxon Studies in Archaeology and History 10 (Oxford 1999), pp. 17-24.

23 responses to “For whom does ‘theory’ work, and for how long?

  1. I have the same divided view of theory. What I point out to students is that the social sciences are model building, trying to find one size fits all. When that fails or you encounter an exception, you either modify the theory or get a new one. But historians are all about the exceptions to the rule. So if the theory fits our particular, histoically unique situation, we are happy to use it just once. And we get annoyed when social scientists try to squash our unique past events into a model by stripping out the non-essential bits that don’t fit (the bits we care about).

    • That point about single-use theory is very fair, actually. The result, though, I suppose, is that we can wind up with a whole selection of bits of models and no coherent structure to link them except the infinite variability of the human being…

  2. We do need a bit of theory to help make sense of the individual bits of evidence. I’ve found my own youthful distaste for theory coming around to bite me later on, as I simply managed to come up with explanations of what I saw in evidence without consciously trying to conform to any particular theory. As you say, we all internalize some theory here and there, but when you just forge ahead and say “this is what I think” without reference to the cornerstones of said theory (which you didn’t actually use firsthand because it was internalized), you get in a bit of trouble.

    I guess I’m saying is that theory is useful (well, some theories), but we can’t go running around finding new places to which to apply the theory. Rather, we should select the theory that seems most useful for our set of questions, and remain more loyal to the questions than to the theory. I bet I wouldn’t mind this piece by Scull for doing just that.

    • Well, things one didn’t realise one had internalised is also a danger I’ve come to recognise! And this also raises an issue about how many generations one has to trace back. I borrow lots and lots of stuff from Matthew Innes, and I never really followed up any of his social theory references, but he was getting quite a lot of stuff from Chris Wickham, as now do I, and while Chris is an avowed Marxist there’s much more than Marx in there. I also use lots of Wendy Davies’s methods, and she’s much less explicit about having an avowed theoretical base. But even she has influences. Heck, even big theoretical names like Marx or Foucault were doing anything but writing in a void. Are any of these safe to use without having read into their roots? And if so, then the door is right open to only using the new stuff and not knowing that older stuff did the job already. This is of course a weakness we have generally, but it’s not usually thought helpful…

  3. One of the things that make History diferent form other sciences, imo, is that in History it’s almost impossible to avoid questioning the basis of our own understanding; call it conceptualization, categorization, world view, or wahtever you like. As individuals we all have our own set of ‘modern/current’ cultural settings, our own glasses to look at the universe, but the fact is that those glasses are just a set of changing conventions. The extra difficulty of History lies in to be able to approach older/stranger conventions, the ones that modeled the societies we are trying to understand. So it’s not only about our own theoric baggage and his options, it’s also about being able to change baggages…

    • Oh yes, I do like that! I may have to borrow that formulation should I get to teach such stuff again at some point…

    • Yes. IMO history is a cultural lens. The culture changes, so does our view of history. As it’s written down, though, we don’t lose previous hypotheses (I can’t bear to call them theories, sorry!) though we might forget them for a while.

      That’s why I love reading not only modern scholarship but long-out-of-print titles that the profession might find today find embarrassing. It all keeps me mind levered open, relatively nimble, and being a novelist I honestly don’t mind a professional (archaeologist or historian) bursting out laughing and telling me my ideas are ridiculous. Don’t mind, that is, as long they explain *why* they’re ridiculous. Because I learn from that…

      • Perhaps they won’t seem ridiculous in thirty years though! (I’m not sure I believe you anyway: with the amount of research you did for Hild you must be as well informed as many a practitioner…)

        There is a bit of a line in work these days exploring the preconceptions of earlier scholarship on our chosen subjects. The problems of such are especially obvious where issues of race or ethnicity come up but there’s plenty of hangovers from colonialism too. And these we can see, though whether such work is thus without value is a question with answers that vary from scholar to scholar (both the scholars being deconstructed and the scholars using them). The problem such work doesn’t usually manage to address, however, is whether we can deconstruct our own stuff… (Which is why once I tried!)

        • I did–am still doing!–a lot. But I have huge gaps in what I know because I didn’t get a formal education in history. I love it (once I get past hating it; I’m as human as anyone else) when someone does me the courtesy of pointing out that gap and makes a suggestion as to how to fill it. I love learning how it all fits together.

          And I’m of the opinion that idiocy in one aspect of a practitioner’s perspective doesn’t obviate their entire. Racist, sexist, imperialist (etc.) people–or should I say those who have expressed racist, sexist (etc) opinions–might have some good ideas about some things. People are…complicated.

      • I fully agree, completely. One of the conconsequences of this cultural dependence works at the anthropological level, that is, the part of History that intersecs Anthropology (ie: things like being able to understand at a personal level that polytheism implies a diferent range of vital experiences than monotheism -. lets say ‘tempestarii vs. presbyters’ -, or being able to understand things like astrology or incubatio as something more than superstitions ). Without some degree of mental aperture, we will not be able to evaluate ancien cultural patterns. And the same can be said for other intersections of History. So yes, every opinion is welcomed, ours, usually, are not better, only later.

  4. Long time fan, first time commenter!

    I’ve been struck by the question posed at the end of the post mainly because I can’t satisfactorily answer it. Having recently come through a theory/methodology/hisotriography course (where we were repeatedly told exactly what you’ve been saying – everyone internalises theoretical assumptions, so at least put some thought into it and make yours explicit) I’m now struck by how dated that course was – I read only one book published in my own lifetime (Buc’s ‘Dangers of Ritual’) and several that were published before my parents were born. Geertz was our required reading for the “History and Anthropology” seminar. To my shame, I’ve even cited Jackson Turner on frontiers.

    Does this matter? Well, I think like most of the comments here I have to make a qualified yes and no.

    Yes, because we shouldn’t ignore the work going on in cognate fields (like anthropology etc.) just because it’s easier to pull out the old reliables when a deadline’s looming than to read the latest works in an unfamiliar discipline (because it’s difficult enough to keep up with the latest in out own).But there are obviously a whole range of potentially useful approaches out there just waiting for an enterprising medievalist to get up of her backside and read about them.

    But no, because I do feel like the older theories have their uses and shouldn’t be discarded just because they have been overtaken. As long as theory informs questions rather than enforces the evidence into pre-existing paradigms, then I see no reason why historians shouldn’t still be reading Marx for many more years to come. It’s certainly possible to argue that Thucydides, for instance, still poses valuable theoretical questions for historians, though their answers may be very different than Thucydides’ own. As long as the models we construct (from the historical data, not the theory) are disposable, then the theories which inform them and the questions asked are viable. And this is probably a good thing, since I’ve yet to find a single theory which asks all the right questions!

    • Well, first off, welcome to the blog, glad to have you commenting! I suppose that the thing about disposable models is that it does imply that my archæologist interlocutrix was right in suggesting that some need throwing away. Do we deep-freeze them or deep-six them, though?

      I think that one thing I could have, meanly, said more about in the post that you now make me think is also the extent to which a ‘new’ theoretical resource needs to be accepted beyond its initial practitioners. This is probably something that is partly to be blamed on the kind of teaching you describe, where people are told certain things are standard tools and use them because their peers also accept them as such, whereas someone trying to bring in a new approach will have to face down initial scepticism that this brings us anything we didn’t already know, especially if it’s the kind of approach that coins its own terms of art. That may be part of the pressure to use old or shared models. All the same, the kind of theorist’s flag-waving that goes on among the cohort in literature who like Deleuze or similar uses flags much newer than the medieval historians seem to…

      • On your first point, and related to the discussion above about outdated work which remains valuable, is I think deep freezing (or even just refrigeration) is the way to go. Theory in history has to be ‘done’ in order to be fully understood; it’s much more effective to read Braudel’s ‘Mediterranean’ than it is to read ‘The “Annales” School: An Introduction’ if you want to understand the theories behind Annales historiography. And even if the models established by such theories are now defunct – say Bloch’s ‘Ties of Dependance’-style feudalism – it doesn’t make ‘Feaudal Society’ any less of a vital book for modern scholars because the theory and the questions which informed the defunct model are still valid. This obviously varies (not every book written in the ’30s will be as useful as Bloch’s) but as a general principle those theories that can help provide answers, and the models they generated, can guide or suggest the route forward (or mistakes to be avoided) to modern historians.

        On the second point, it is surely more difficult to establish a ‘new’ theories, but again if the stick against which we measure success is the utility of the theory to explain what we find in the data. If the new theory can produce good history, then I would remain optimistic that it would find acceptance, even if it takes a few years. Old models are hard to get rid of, but in general the academy seems to reject entrenchment in a model so deep that it obscures the history (see Chris ‘avowedly Marxist’ Wickham’s take down of Guy Bois’ Marxist interpretation of Cluny c. 1000). As long as whatever new theory isn’t so dense as to need to be read with a dictionary, then I don’t think older models are so inflexible that peers and doyens wouldn’t at least give a workable theory grudging respect, even if they don’t become ardent enthusiasts. But that could just be my youthful naivety talking…

        • It seems to me that there’s a link between those two points that is probably quite important. I’m sure that you’re right that ‘theory’ is better received in its demonstration than its statement. It’s not just cynics like me going, “what can this do for me?” but the simple fact that it’s easier to see the implications of a way of thinking if some have been worked out for you! and some of the theorists the medievalists seem to like, I suppose most especially Bourdieu who had a much large social project that was essentially (and sometimes explicitly) about ‘us’, not whichever ‘them’ he started with, still used their working sample to do that kind of demonstration in practice. It seems intuitive to me that an attempt to write ‘theory’ alone is therefore less likely to be picked up than one that was already solving questions for its originator. But that also gives me worries, connected with today’s post, about for example using someone else’s database design for your own project. The problems that tool was invented to solve won’t be the same, and using it may give answers that look like what your toolmaker was designing for, not those that you vaguely saw in the material. This would not be a failing in the tool, but with something as complex as a humanities project (in terms of unrestrictable variables, I mean, rather than in intellectual depth) is a tool ever properly transportable between projects?

  5. Like others of your commentators, I get a bit unhappy using the term “theory”, because most of the ideas we’re working with aren’t theories in the scientific/mathematical sense: falsifiable statements. I think it might be better to say that we’re using models or frameworks from other disciplines. And I think most historians end up being so eclectic because we need such models/frameworks for at least three different types of questions:

    1) how individuals behave (studied in psychology, but also subsections of religious studies and economics)
    2) how societies work (studied in sociology, anthropology, some subsections of economics)
    3) how traces of human activity are preserved in the historical/archaeological record (studied in archaeology, history, literary studies)

    So since there are almost no models/frameworks that cover all of this (I’d say Marxism tries to cover most, but still doesn’t do so), we have to borrow models from multiple sources. Or alternatively we use folk psychology, folk sociology etc: our intuitions about how people and societies behave, which are inevitably personal, and affected by our own experiences. Because I am a practising Christian, my views on whether people “really believed” in the Middle Ages is almost inevitable going to be slightly different from that of most atheist scholars on the period.

    As for the age of our models, I’ll start by making an obvious point: not all of us are using “theory” from several generations ago. If you work on gender and sexuality, for example, you’re very unlikely to be using really old models, because these issues weren’t being theorised fifty years ago, or only in completely discredited ways (e.g. sex role theories, psychoanalytical understandings of homosexuality). But even here, there’s still a surprising reliance on “classic” models: I’ve lost count of how many times people refer to Joan Wallach Scott, whose article on gender is now nearly thirty years old.

    I think there are several reasons, mainly justifiable, why historians mostly tend to use models that aren’t that up to date. One is the simple point that the latest models/frameworks may have serious problems, so it’s worth waiting a few years to see whether experts in the field point out flaws. (I’m less likely to spot subtle problems with a sociological argument than those who know more about it).

    A second issue is that models/frameworks tend to get progressively more complex as a field develops, and the more complex a model is, the harder it is to use. The kind of generalisations you need to produce a useful model, one potentially applicable to many situations, tend not to hold in some cases, but if you introduce more factors to allow for that, you get something that’s only applicable in one narrow domain. Foucault’s model of knowledge-as-power, for example, is probably not completely accurate anywhere, but it’s a reasonable fit for a lot of cases.

    A third aspect is that in some fields most of the models now being created are not easily applicable to premodern societies. Any economic theory that relies on statistical data or presupposes an efficiently-functioning market is irrelevant to us, and only marginally relevant to those in the later Middle Ages. Any gender model that presupposes the existence of capitalism is not likely to provide many useful insights for me. In contrast, Marx and Weber took premodern societies seriously: they may have been wrong in many ways, but they produced analytical tools that recognised the existence of the types of situation we deal with.

    Finally, I think historians also tend to be better (or at least less purist) at picking the useful bits out of scholars who come from deeply dodgy ideological backgrounds. We’ve had to be: we still of necessity use editions and data compiled and collected by racist, sexist bigots. (I’ve mentioned before the Nazi bookplates I’ve seen in some volumes of the Monumenta Germaniae Historica). A whole lot of modern German constitutional history of the early Middle Ages has been working out which ideas you can use from early twentieth century German historians and which are too “Zeitbedingt” (the standard euphemism for racist/Nazi-tainted). Whenever I pick up a book on sociology or religion from the 1960s or 1970s that refers as a matter of course to humanity as male, I’m alienated to some degree. How much of the discarding of older models/frameworks in other fields is due to a more self-conscious desire to avoid the taint of such biases?

    • Wow, there’s a lot there, thankyou Magistra. Much to think both with and about! I completely agree about the inapplicability of many models to pre-modern societies, though in cases like the rational economic actor I think it’s also inapplicable to modern ones! I also agree that complexity is inimical to applicability, though I don’t know how many new developments of a ‘theoretical’ kind are refinement rather than replacement, to which that argument wouldn’t apply. As for biases and taint, I suppose the place where I meet that most (since I am not as theorised as I should be about gender and tend just to cite Jinty) is in the reluctance of some archæologists to abandon culture-historical approaches, but we’ve seen here how I can get into deep water by suggesting that nationalism should be seen that way, given my area of focus

      One bit of what you say that I would contest is this, however:

      One is the simple point that the latest models/frameworks may have serious problems, so it’s worth waiting a few years to see whether experts in the field point out flaws.

      This is the core disagreement out of which this post sprung, isn’t it? An expert in the field pointed out flaws in processualism to me, but it seemed to me that it still did work I needed it to do for me to understand something better. Would you then be arguing that I ought, we ought, in fact to be using newer theory when the relevant experts say this one is worn out? Because that seems to me to push in the opposite direction of your other points about the retained utility of these ideas even once partly discredited.

  6. I think I wasn’t clear enough in my statement. What I meant is that experts could point out flaws in the original argument so that I am then better placed to adjudicate mentally between positions (e.g. this model is useful or is not). What I tend to find is that when looking at a situation/problem in which I’m not an expert it is harder to think of my own counterarguments than to assess the value of counterarguments produced by others. I think this is a common mental problem: that’s why we tend to give first-year undergraduates two opposing articles, rather than just give them one and say “what are the problems with this approach?”

    So with the anti-processual archaeology, if experts are against it because it doesn’t allow enough role for subjectivity, the question then for you would be whether the mix of models you’re using (and as I’ve said, I think historians have to use a mix of models) still allows room for that or not. i.e. you’re aware of the flaws in that particular model but can counterbalance them. One of the reasons I read some pieces of “theory” on gender wasn’t so I could use them, but so I could say explicitly why I didn’t use them: because their flaws/irrelevance for my project were sufficient that they were not of use. For example, we have almost no examples of “third-sex” people in the West in my period (whereas eunuchs are a factor in early medieval Byzantium).

    • Aha, the old adage by which it’s easier to acquire a balanced view when one has more than one source; fair enough! In that case I suppose the problem is just in deciding whether two theories are really sources for the same thing, though I suppose that with an identifiable referent like gender they should at least be possible to use as sources for it.

  7. Pingback: How to escape one’s theoretical baggage in four pages | A Corner of Tenth-Century Europe

  8. Pingback: Theory Trends in Archaeology - Rantin' and Rovin'

  9. Pingback: Seminar CCXVIII: Byzantine frontier badboys | A Corner of Tenth-Century Europe

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s