m.l.j. abercrombie on the difference between descriptions and inferences

While I’ve been gathering up final papers and portfolios for my graduate and undergraduate courses, I’ve been looking at a real gem of a book loaned to me by a colleague: M.L.J. Abercrombie’s Anatomy of Judgement (1960).  Abercrombie was an expert in perceptual disorders who spent years training medical students how to observe–and conceptualize their observations–in a more scientific way.  Both collaborative learning and problem-based learning pedagogies, practices that she pioneered, owe a great deal to this powerful little book.

Anatomy has some of the best descriptions of the give-and-take of classroom discussion that I’ve ever read, and is especially good on the technique of using students’ exchanges to alert them to their own unspoken assumptions, even in seemingly straightforward operations like reading medical instruments or interpreting charts.

Her fundamental insight is that “seeing” is not just a matter of passively registering what is out there in the external world, but represents a form of judgment that involves both selection and interpretation  of sensory data.  This (usually unconscious) process of selection and interpretation, however, is heavily influenced by what she terms “schemata,” which are experiences or conceptual schemes that cannot themselves be brought to light until verbal discussion, disagreement and elaboration bring them into the consciousness of those making the observations.

So, for example, she describes test subjects who were asked to read a story about a tragic family gathering, then asked to describe a cheerful Breughel painting.  The subjects tended to interpret the painting’s visual details in light of the earlier story, even when these interpretations were belied by the happy surface events of the story.  The subjects were not able to banish the effects of the earlier schemata, even when it contradicted what they saw in front of their eyes.

I became interested in this book’s treatment of the novice observer because it offered a potential response to Laura’s earlier post about student difficulties with metacognition: how does one make students aware of the gap between their reflexive responses to literature and the demands of genuine literary criticism?

Abercrombie’s answer  focuses upon her notion of a trained, disciplinarily informed “judgement” that really gets formed collaboratively, in the context of group discussions (modeled on that era’s group psychotherapy) facilitated by an expert who helps students to begin to recognize their own characteristic errors and patterns of judgement.  The role of both the other students and the facilitator is to draw everyone’s attention to alternative schemata and interpretations, which must then be tested by further discussion to determine their validity.  According to Abercrombie, the more schemata the student has potentially available to her, the better her chances of making an accurate judgement.

Where I see the relevance to Laura’s description of her students’ characteristic errors is in their struggle to perceive the difference between descriptions and  inferences.  When students reading the Country Wife assume wrongly that Horner is simply impotent, or generalize that “back then, women had no rights and were simply taken advantage of,” they are simply making an unwarranted leap to the inferential stage without ever noticing how far they have strayed from a faithful description of the play’s characters or events.  The more experienced–or trained–a student is, however, the more likely she is to base her inferences on accepted evidence like passages in the primary text, credible biographical and historical information, key critics, etc., and to build up her own inferences in ways that her teacher can recognize, even when she disagrees with those inferences.

But students are not likely to know where to find this information, or  how to use it, until they understand the difference between description and inference more generally in literary studies, and how to build up more successful and persuasive inferences from the appropriate and agreed-upon information and understandings (in other words, schemata) from their previous experiences of reading, writing, discussion, and research.  If Abercrombie is correct, and I believe she is, the best way for students to learn how to build up their inferences is in free group discussion, where they hear others test the various interpretations collectively.  In this way, what begins as an idiosyncratic observation is transformed into yet another alternative for the group to explore while testing its validity.

Abercrombie provides a suggestive description of a trained radiologist’s reading of radiographs:

A radiologist differs from laymen in having seen many radiographs, so that he has built up visual schemata of them.  But he differs also in that he knows (as laymen do not) how to get other information about the things shown in the radiograph: he correlates certain shadows with the the results of other investigations of the patient–his state of health, or reaction to clinical tests for instance.  He can therefore check whether a shadow which fits his schemata of say, tubercular lesions, indicates that the patient has tuberculosis (48).

I would compare the radiologist looking at a single x-ray to the literary scholar examining a particular literary work: both of them, because of the large number of alternative schemata they have accumulated over the years, are able to extract vastly more information from the single graph or paragraph than the neophyte  trying to figure out what’s important under a particular set of circumstances.  What would literary studies look like if we considered the literary work as not merely containing “meaning,” which after all can be produced with no knowledge at all of the work or its contexts, but the kind of “information” that can only be extracted by the trained eye?

As Abercrombie memorably says, “How to tell students what to look for without telling them what to see is the dilemma of teaching.”  And, I would add, this same dilemma faces anyone who would teach students about literature, when student themselves are unsure of what they are reading or how they should read it.  Rather than talk about literature’s “deep meanings,” which I think is misleadingly subjectivist, perhaps we should regard the literary work as full of potential information to the would-be scholar.  The would-be scholar, the position that we all inhabit, requires considerable training and practice and discussion to recognize all the varied information a literary work potentially contains, and the various schema that might accommodate it.



5 responses to “m.l.j. abercrombie on the difference between descriptions and inferences

  1. I remain fascinated by the similarities in analytical comportment between radiologists and art historians (in the latter group, especially connoisseurs, formalists, and others who attend very closely to visual detail) on the one hand and literary critics and their texts on the other. The coming of the literati to art history in the 1980’s (Norman Bryson et al.) has now been integrated into the discipline, but it seems like such fertile ground in terms of pedagogy.

  2. Laura Rosenthal

    Thanks for that interesting post. I will put the book on my list. It reminds me a little bit of one I read a while back called “Why Don’t Students Like School: A Cognitive Scientist Answers Questions About How the Mind Works and What It Means for the Classroom,” by Daniel T. Willingham. The part that really stuck with me about that book was the author’s argument that, essentially, contrary to some recent arguments and assumptions behind practices, having a foundation of knowledge makes a huge difference in student capacity to learn new things and truly comprehend them. Thus we shouldn’t pit abstract skills (critical thinking, for example) against a certain amount of contextual accumulation because the abstract skills are deeply dependent on the knowledge foundation.

  3. @Rex: some of those barriers between the verbal and non-verbal came down when literary critics became interested in semiotics, along with film and performance, while cultural studies scrambles things in its own way. I agree it’s probably underused in terms of pedagogy.

    @Laura: don’t know Willingham, but I think what you’re saying makes sense. I’m assuming that skills like critical thinking come from manipulating the knowledge students need to understand the complexity of the phenomena they are studying. I do think that the problem of “inert knowledge” means that anything learned as foundational has to be put to work almost immediately, if it is to be retained.

    • @ Dave: I should say that the verbal/visual nexus may be underutilized over here (art history), but see Rod Nevitt’s recent work on the creation of the 17C Dutch teenager (or young lover)…I don’t know how that seeps into his pedagogy, and we certainly haven’t formalized any sort of critical practice with the students.

  4. If you’ve got a cite handy for Rod’s work, post it in the comments. The topic sounds perfect for a seminar at an undergrad or grad level. I’m always tempted to do more with visual stuff in my own research and teaching, but I don’t know it well enough to use in a classroom context. Still, one of the great advantages of a “period” concept is the way in which words and images from the same chronological slice are felt to have some unstructured relation to one another, though one that needs to be specified somehow.