[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [xmca] Re: microcosm/unit of analysis and xmca discourse

I just got back from our two-day graduate "orientation training". There was a good presentation on "What kind of question is an open question?" which looked at what kinds of questions get many classroom answers (either from one student or from many students or even from the teacher). 
One of the good things about this question is that "What kind of question is an open question" is really an open question. It allows many kinds of answers: grammatical, discoursal, functional, formal, and conceptual. 
Let's take the least interesting of these answers as an example. It's possible to arrange questions grammatically in terms of degrees of freedom (df) where df = n-1, and n-1 represents the number of choices in the answer. Like this:
a) Are you well? (n-1 = 1)
b) How are you? (n-1 = x, where x is a restricted number of adjectives and adverbs for a preferred answer [e.g. "fine'], but an infinite number of adjectives, adverbs, and clausal constructions for a dispreferred one ["terrible", "so sad", "been down so long that it look like up to me"]. 
c) Tell me how you feel. (n-1 = infinity)
On the basis of this grammatical explanatory principle, we can then analyze exchanges into initiate-response units and look to see if there is a significant difference between the number of answers we get with a) and the number of answers we get with b). (Because we are using elementary school English data, there are no examples of c] in our data.)
The result is statistically significant (p <.007, t = 2.705), but not very substantial (an average of .9 answers for the wh-questions and .64 for the y/n questions). This suggests that we have a rather poor explanatory principle, and we need to keep looking; no teacher in her right mind is interested in the difference between .9 answers and .64 answers. 
But it also suggests that we know where to look, namely at the dispreferred results. We look, for example, at 
a) wh-questions which have FEW answers and 
b) y/n questions which have MANY answers and we try to figure out WHY. 
Now, it turns that in category a) the wh-questions that have FEW answers look like this:
T: Kyeong-rok! Kyeong-rok! Who is Kyeongrok?
S (silently raises a hand).
Category b) is somewhat more interesting because almost all of these are STUDENT questions to which the teacher provides multiple answers from which the kids choose. But let's stick with the boring bit.
On the way back from our grad retreat I got stuck in a four hour traffic jam with one of our foremost grammarians. He INSISTED that "Who is Kyeonrok?" is NOT, in fact, a wh-question at all, because it only admits a single answer. 
You can see IMMEDIATELY what this man has in mind. He wants to expand the category of grammar (the distinction between y/n questions and wh-questions) to INCLUDE the number of answers. In other words, he wants his own field, grammar, to take over my field, discourse.
So according to my colleague we can ONLY call wh-questions by the name wh-questions  when they have many answers.This means that our explanandum is now coterminous with the explanans. The analytical unit, the question-answer exchange, is isomorphic with the explanatory principle, and we find that some question-answer exchanges get many answers because...well, because they are wh-questions, i.e. many-answered exchanges. 
OK--now let me turn this posting over to my friend the Tortoise. I really only dip into philosophy the way I dip into the toilet when it's plugged. It's not that I disapprove of it; on the contrary, it seems to answer a very real and essential need. It's just that I'm slow, and I never seem to be able to catch up with the swift-footed ones on the other thread. 
ACHILLES: "But what we can say about non-tautological relations between object of psychology, explanatory principle and unit of analysis?"
TORTOISE: We can say that the object of psychology is not an object, but a process, and that a process changes in real time and in fact changes with the scale of resolution (delicacy) that we use. For example, the process we study when we look at the phylogenetic evolution of consciousness is adaptation. But the process we study when we look at the sociocultural development of consciousness is more like exaptation. And the process we study when we look at the ontogenetic growth of consciousness is development, while the process we study when we look at the microgenetic changes of consciousness is learning. 
Since the process changes, we can certainly expect that the analysis into units will have to change. For example, we may analyze phylogenetic evolution into species, but we cannot do this with sociocultural history. We can analyze sociocultural history into modes of production, but applied to ontogenesis this is Stalinist rather than Marxist. We can analyze ontogenesis into stages, but no teacher thinks of some of her children as sensorimotor and others as formal operational.
We can also expect that what was a explanandum on one semio-historical time scale will become a PARTIAL explanans in the next. For example, the (rather inadequate) physical endowment of newborn humans does PARTIALLY explain cultural-historical progress (but not without remainder). The cultural endowment of the child will PARTIALLY explain his or her development (but, pace Karpov, not without remainder). The ontogenetic endowment of a child will PARTIALLY explain microgenesis (but, pace Gesell, not without remainder).
ACHILLES: "I wonder that if the object cannot explain itself and demand a "extract of reality from which it is function" (explanatory principle), perhaps a criteria to think "unit of
analysis" could be its role in permits to the searcher establishes some
kind of indirect relationship between the object of study and its explanatory
principle... if we assume that these relations are not isomorphic nor
TORTOISE: Ah...my good friend Vygotsky writes that his good friend Marx writes that if objects could explain themselves science would be unnecessary. He's a careful man; he never gives us a functionalist explanation without adding a structural, a logical, and even a historical one.
Take evolution, something we tortoises are particularly fond of (because it has largely left us alone, and because like us it tends to take things fairly slowly). Functionalist explanations lead to Panglossian paradigms ("It evolved that way because everything always evolves for the best"). Panglossian paradigms cannot explain the many EXAPTATIONS we find in nature, when something evolves for one purpose (e.g. the tongue for eating and the lungs for breathing) and is co-opted for something quite different (e.g. dinnertable conversation).
ACHILLES: "Must the unity of analysis be ontologically pertinent to both
"planes" (or spinozian "modes") of reality: "object of study" and its "explanatory principle" (for instance, consciousness and the extract of reality from which its is function - perhaps "social relationships"(?)) to permit a concrete analysis?
TORTOISE: Answers have to be pertinent to questions. That's what coherent discourse means, and study has to be a coherent discourse. The plane we are on is always the same; we are studying.
ACHILLES: Or this ontological pertinence is not necessary and we can speak about more formal kinds of units?
TORTOISE: I never heard of form without content. Form is the how of a sculpture, and content is the what. Neither one is the material: a sculpture of Achilles is not a sculpture of a block of marble. I can't seem to imagine a sculpture of Achilles without Achilles. Wouldn't that just be a block of marble and not a sculpture at all?
When we are studying psychology, we have to have a coherent discourse in order to study it. Of course, we don't need this in order for there to be psychology; that will exist whether we study it or not. But in order to study it, we need questions that at least potentially have answers, and for that we need answers that will match the questions.
David (the Tortoise) Kellogg
Seoul National University of Education

xmca mailing list