I am going to try writing down pieces of advice that I give to students all the time, in the hopes that they might be useful for people who can’t make it to my office hours.
“Spelling out your theoretical and methodological assumptions — the contours of your conceptual equipment, so to speak — is a vital part of doing good social science, because if I don’t know what your assumptions are then I really can’t fairly evaluate your results. In fact, if I don’t know what your assumptions are, I probably have little choice but to apply my own standards, which may or may not be appropriate to your project. So being as clear as you can about your assumptions (with the caveat that it’s impossible to actually spell out *every* assumption that you’re making, both because that kind of self-awareness is a theoretical ideal rather than a live possibility, and because of the Wittgensteinian logical paradox involved in trying to endogenize every rule of a game) is critical.
However, spelling out your assumptions is not the same thing as establishing their validity or their value. Yes, your take on discourse is more pragmatic/Foucault than CDA/Wodak, but that’s not a conclusion of your research — it’s an assumption. Just like ‘individuals make rational choices under conditions of imperfect information’ or ‘human beings are meaning-making animals.’ The fact that you assume this tells me a lot about you, but basically zippo about whether you are right or, more to the point, about whether your assumption is a useful one for the research problem at hand. You can’t use a set of assumptions about discursive practices to conclude that discourse matters or that discourse works the way you think it does, because you already assumed that at the outset! Ditto assumptions about material factors, ideas, etc. “mattering.” You can and should be as detailed as you can be about your assumptions, but if you want anyone to appreciate them as anything other than an expression of your idiosyncratic aesthetic sensibilities, you need to show us what insight they generate in practice — and you have to refrain from overreaching and tautologically concluding that results generated by applying assumption X are an argument for the validity of assumption X. Those results might indeed contribute to an argument that it is useful to make assumption X when trying to explain what you’re trying to explain, but that’s as far as it goes.
Making ‘assent to assumption X’ a condition of membership in some fraternity helps you found or adhere to a school of thought, but whether it helps you explain anything is an entirely different issue. The fact that members of a school, like adherents of any other type of sect, will parade their results as if they constituted ‘evidence’ for their assumptions should be regarded in about the same spirit as any other testimonial, which is to say, compelling to believers but largely inscrutable to outsiders. Displaying your allegiance doesn’t contribute to knowledge, although it can get you into interesting conversations.”
Excellent post.
Can I just ‘like” here?
“Those results might indeed contribute to an argument that it is useful to make assumption X when trying to explain what you’re trying to explain, but that’s as far as it goes.”
As good a statement of instrumentalism as you’ll find anywhere, which is itself an assumption, and as you know not a particularly good one at that imo. I assume that gravity exists when I walk of the cliff which is why I don’t do it. And there’s plenty of evidence (which is not the same as an assumption) to say gravity is there, not that it’s just useful. Without this ‘added extra’ you are devoid of the possibility of criticising any one’s assumptions. Also, I’d differentiate assumptions from arguments. Not everything underpinning an approach is an assumption, and in fact uncritically taken assumptions are problematic. Sorry for messing up you page with some reality. Otherwise really good Patrick, just look at the assumptions that make it work, and give me arguments about why they are valid.
I think you’re conflating two different uses of the word “assumption.” What I am talking about here involves explanations; your example of not walking off of a cliff involves practical activity, Do I need a theory of gravity, or a theory involving gravity, in order not to walk off of that cliff? No, I do not. Indeed, I don’t need any kind of explanation; I just need the experience of stuff falling off of the cliff and not hovering in mid-air. There’s a difference between a presupposition made in action, and an assumption made for explanatory purposes.
Evidence that gravity “is there” is, I believe, evidence that “gravity” as a concept makes sense out of a bunch of disparate phenomena — that it has explanatory value. (The reality of an assumption used in an explanation is contingent on that explanation being a good one, after all.) Whether one wants to call that “really being there” or “being useful” is pretty much a matter of semantics, as far as I am concerned ;-)
Of course not everything underpinning an approach is actually an assumption, so we’re in agreement there. Arguments underpinning an approach would consist of assumptions and conclusions or derivations from those assumptions, and my point was only about the assumptions themselves. But I am not sure that I quite grok “criticizing … assumptions,” unless that means “showing how particular assumptions don’t actually afford the explanatory work that you are claiming that they do.” Otherwise it’s just my assumptions vs. your assumptions, my gods vs. your gods, and we never get anywhere.
Actually, Patrick I think you were conflating various uses of the word assumption in you original post; which was the point of my post. Do YOU need an explanation of gravity not to walk of a cliff? No! Do you need the assumption? Probably. Please don’t try it without that assumption Patrick, much as we disagree, I quite like you. What science does (and I’m assuming in academic social inquiry we want to go beyond what people just do in practice or else what are we doing? Redescribing in complex terms what is actually quite simple?) Science goes beyond what people just do to provide an explanation, so on that we agree. It takes what people think happens and gives it a deeper explanation. Also, if it were just semantics (which is an assumption not an explanation – or an argument) then why is it you can’t just say ‘it’s more useful if gravity weren’t there’. Again, you could try this by walking off the cliff and half way down saying, ‘ok, semantically, I think I want to change my mind now’; again I wouldn’t recommend this. But hey, you want to give it a go! Incidentally, instrumentalism necessarily leads to conservatism; there’s always a link between metaphysics and politics; more people should read the Marx and the German Ideology.
And no it’s not just my Gods vs yours; I don’t believe in them, but I can’t challenge your belief in one (or many). Assumptions, on the other hand, in a scientific context, are always open to challenge, and need defending; that’s what distinguishes one from the other. On this I’m with Dawkins.
As far as I’m concerned, challenging assumptions in a scientific context means raising questions about their explanatory utility. If we get into the rather silly territory of “realistic assumptions,” Nancy Cartwright shows up and reminds us all that the laws of physics lie and no assumption is realistic … so we aren’t disagreeing. What kind of a scientific defense of an assumption could there possibly be except for a demonstration that the assumption in question enables or affords a better explanation of something? Of course, treating the assumptions of one explanation as the object of another explanation has often led to some incisive insights (particularly when one re-rederives the assumption as a special case of another set of precepts, as when Newton’s laws of motion show up as a special case of relativistic mechanics when everything is moving at a low percentage of the speed of light), but as I said, in my view that just displaces the basic issue down a level: the new explanation makes its own assumptions, and the process continues.
Don’t you think it might sometimes be more useful if gravity weren’t there? ;-) But seriously, that’s not what I am saying or implying. My point is only — only! — about explanations, not about experiences or phenomena. If I walked off of a cliff and fell, the problem wouldn’t have anything to do with any potential explanation for my falling; it would have to do with my willful disregard of practical experience. The “semantic” issue only involves the explanation. Instrumentalism — or what Bas van Fraassen calls “constructive empiricism” — isn’t a species of linguistic determinism, after all. “Gravity” as part of an explanation for the observed trajectory of terrestrial objects can only be evaluated the same way as any other explanatory component, which is to say, in terms of what it contributes to the explanation. “Gravity” as a kind of popular shorthand for the experience of stuff falling when dropped is not involved.
My favorite definition of science, as you know, is Weberian: “a thoughtful ordering of empirical actuality,” or, paraphrased, the analytical ordering of experience. I’d rather not follow Dawkins into the messy realm of conflating scientific explanations with (anti-)theology; I’m willing to risk a certain kind of conservatism in a perhaps foredoomed-to-failure effort to carve out a space for non-partisan empirical investigations.