Dear David Kellogg,
Thanks for pressing me on the habituation/enculturation distinction--particularly with respect to my assertion that some domains of activity are "precisely defined and hence subject to being learned as a skill, outside of cultural context," whereas others are not. As you point out, my notion of "precisely defined" is far from precisely defined.
This is a tough distinction to draw clearly; my delay in responding has largely been taken up with working it through.
To begin, the idea of a distinction between a skill and a "disposition" (my word for a cultural practice requiring cultural context to learn) is something that comes to me from educational practice. The genres approach is my attempt to untangle the coherent threads of pedagogical intention that all too often are tangled up within our integrative discourse about good teaching--a discourse that takes "learning" to be a complex and multifaceted construct rather than recognizing the discrete metaphors for learning that actually drive our pedagogical interests.
Let me give a quick example to illustrate the distinction between skills and dispositions before attempting to characterize it analytically. For the past 3 decades, mathematics educators have been interested in promoting an agenda of non-routine problem solving competence in K-12 math students. Back in the 1930s and 1940s a distinguished Hungarian mathematician, George Pólya, had undertaken to explicate problem solving expertise and arrived at a set of heuristic questions that mathematicians tend to ask themselves to enrich their exploration and solution of difficult problems. What's important about these heuristics is that they are rules-of-thumb rather than algorithmic rules. Pólya recognized the context sensitive nature of these rules, and his own pedagogy always sought to engage students in the context of working on actual non-routine problems. Observing students, he would ask himself aloud the kinds of heuristic questions native to actual mathematical practice that would usefully arise at the point in problem solving that the student had reached. In this way, he was very successful in inculcating these powerful practices of self questioning.
Unfortunately, the enculturationist character of this pedagogical practice escaped most mathematics educators taking up Polya's heuristics in the 1980s. Textbooks were written that listed out Polya's heuristics and provided exercises in which each heuristic could be practiced over and over, one at a time. In this way, Polya's teaching of heuristics as cultural practices was replaced by a decontextualized pedagogy of skill development, one that failed to promote anything resembling the sought after expertise in non-routine problem solving.
Despite having many such examples that differentiate skills from dispositions, making an analytic distinction remains a daunting challenge. This is because none of the traditions of inquiry in psychology that address learning in these distinct senses cares to participate in the drawing of strong analytic boundaries. Rather, I argue from a Kuhnian perspective, the trajectory for research in the various subfields of psychology is always outward from locally coherent intuitions and instances toward embracing varied phenomena of learning being explored by theorists in other camps. It is only along this trajectory that psychology ever can hope to become paradigmatic, that is unified around a single perspective. In this respect, the work I do in attempting to establish firm analytic boundaries between notions of learning as "habituation of skills," "construction of concepts," and "enculturation of dispositions" is a kind of guerilla research--I call it crossdisciplinary research--that draws on psychological theory, but at the same time resists its integrative tendencies.
I draw on two theoretical traditions to help understand how skills are learned, and each can provide some guidance toward demarcating skills from dispositions. Behaviorism is one such tradition, and it is distinguished by its claims of scientific objectivity: a restriction of its interests to stimuli and responses that are objectively characterized and measurable. However, criticisms of the putative objectivity of behavioral theory go back to the early days for associationist psychology. For example Meiklejohn (1908) complained that the objectivity of stimuli is misleading: “Think of learning to drive a nail with a yellow hammer, and realize your helplessness if, in time of need, you should borrow your neighbor’s hammer and find it is painted red” (p. 126). The point, here, is that even as "objective" a stimulus as a hammer can be ambiguous with respect to how it is mentally represented.
Despite such criticisms, over the next 50 years behaviorist rhetoric increasingly extolled the importance of scientific "objectivity" even as behaviorists ventured further and further from the "natural" domain of habituationist learning to address behavior that is not simply constrained by environmental factors. The culmination was Skinner’s (1958) efforts to extend the analysis of behaviorism to verbal behavior famously beaten back in Chomsky’s (1959) book review. Skinner claimed, for example, that the response "beautiful" elicited by the stimulus of seeing a painting must be under "stimulus control." However, Chomsky convincingly argued that Skinner's analysis was entirely post hoc, without any explanatory or predictive value.
This sets the stage for localizing the domain of habituation learning to instances in which specific response repertoires can be seen as elicited by specific stimuli. Still, how one characterizes "specific" remains much in need of clarification. I turn to the Implicit Learning paradigm of cognitive psychology for further guidance. Study of implicit learning was initiated by Arthur Reber (1967) who demonstrated that when one organizes a set of stimuli according to a complex grammar, subjects will acquire competencies related to the grammar despite being unaware, consciously, of any organizational structure to the data they are observing. A particularly nice example comes from Lewicki, Czyzewski, & Hoffman (1987) who asked subjects to find the single instance of a particular alphanumeric character from among a large array of other characters. In repeated trials in which the target character is randomly repositioned in the array, subjects quickly achieve their minimum average rate of retrieval, r. This has to do with the average time it takes to scan the whole array, given the random placement of the target. Next, unbeknownst to subjects, Lewicki and his team manipulated the location of the target character, placing it in a specific quadrant of the display according to a bizarrely complex rule: For the first trial, the placement was random. For the second trial, the placement was again random, but excluding the quadrant of the first trial. Based on these first two randomly located choices, the next three were precisely determined. For example, if the first two were in quadrants I and IV, respectively, this would determine the placement of the next 3 as say in quadrants II, II, and IV in that order. However, if the initial two random placements were in quadrants III and II, respectively, then a different placement would occur for the next 3, say in quadrants IV, II, and I in that order. After the first five trials, the pattern would repeat, starting with a new first and second random placement, determining the following 3 placements according to the rule. Over hundreds of trials in which the location of the target was determined by this rule, Lewicki et al. found that r, the best rate that can be achieved if the location of the target character is truly random began to decrease for items 3, 4, and 5 in the sequences of 5 items. In other words, subjects were coming to anticipate the location of these items. In debriefing after the study, subjects indicated no awareness that the placement of the target was anything but random, or that their performance regularly was better on some items than on others.
The implicit learning studies indicate something for us about the nature of habituated learning, and the preconditions needed to elicit it. Habituated learning is the forging of unconscious connections between input stimuli and output responses through repetition. In terms of cognitive mechanisms, we don't really need to regard the learning of dispositions as inherently different from the learning of skills. Rather, the distinction rests on an educational question: What kinds of competencies can be facilitated by setting up regimes of practice; what kind require actual cultural context to acquire the needed experiences. As an example, let's take the quintessential American skill of hitting a baseball with a bat. The first precondition for being able to classify this as a skill rather than a disposition is the unproblematic character of the stimulus set. Having a baseball coming toward one while holding a bat over one's shoulder is unproblematically a distinct and identifiable experience. What is more, we can establish regimes of practice for this experience in a batting cage with a pitching machine. One doesn't need to acquire the experiences in the context of an authentic baseball game. Note that the key point here is not that the stimulus set is "objectively definable." Rather it is subjectively experienced as a distinct phenomenological situation.
What is it that actually constitutes the skill of hitting a baseball with a bat? This is murky territory. Presumably pitches coming in at different speeds, from different angles, and in different parts of the strike zone require a somewhat different configuration of motor responses. To set up regimes of practice, we don't have to figure all of this out explicitly. However, we do have to meet a second precondition in order for skills to develop through repetitive experience: a reliable feedback system in the form of a metric of the success of one's response. For example, in the case of hitting a baseball, whether one has or has not hit the ball is immediately apparent. As well, when one does hit the ball, one experiences the success of the hit in the organization of various percepts including the sound produced by the hit, the angle the ball takes after being hit, and the feel of the bat when the ball is well hit versus poorly hit. Putting this all together, we can classify hitting a baseball as a skill that can be practiced rather than a disposition that needs to be acquired in cultural context.
Let's run some of our other examples through this system of evaluation. The Lewicki experiment for the development of competence in scanning for a target letter on a screen laden with distracters is a skill. The stimulus set is cordoned off from other experience because it only occurs in the setting of Lewicki's lab when one is sitting down at a computer and presented with an array. What is more, the response latency--how long it takes to respond--is an automatically monitored variable that we humans--for whatever reason--are spontaneously interested in minimizing. Thus the two conditions are met. Note that if we wanted to train a pigeon to acquire this skill, we would have a problem with the second condition, the pigeon might not monitor response latency, and/or might not be independently motivated to optimize it. In this case, we would likely set up a reward system, say food pellets, that accrues to (what we see as) desired performance. Much of the behaviorist shaping of behavior is creation of artificial feedback systems designed to keep the subject attending to the right feedback variables.
Next let's turn to the example of Pólya's heuristics as related to competence in solving non-routine problems. First, the condition of being an unproblematic category of experience is met. We have no difficulty distinguishing when we are working on a non-routine math problem. However, problems do arise in the feedback system. Different situations of problem solving call for different heuristic questions. In this respect, problem solving is no different than hitting a baseball, different pitches call for different swing responses. The difference is that whereas the feedback in hitting the ball is instantaneous and reliable, the efficacy of selecting one heuristic question over another is delayed and unreliable. Not all problems are solvable, so missing doesn't give good feedback. And the efficacy of solution approaches includes much more than whether or not one has made a good choice of heuristic questions. In short, the whole dynamic of problem solving is too open-ended to yield to habituationist methods.
We're now in a position to analyze the effectiveness of Pólya's enculturationist approach against the failure of habituationist approaches presented by many textbook authors who provided repetitive practice in the various heuristics. As noted above, Pólya's method involved asking himself aloud heuristic questions useful at the point in the solution reached thus far by the student. Deploying these questions, herself, the student comes to experience success in the solving process sufficiently often for the feedback loop to have effect and needed connections established between problem-situation and heuristic question. The reason habituationist repetitive practice was ineffective is because there's no way to codify the subtle connections among problem types and heuristic questions as would be needed for successful habituationist instruction. Returning to the opening challenge, David, it is in this sense that competencies need to be "precisely defined" in order to yield to habituationist methods. I think a very similar analysis could be given for the example of politeness we were discussing. In general, politeness is too context sensitive to yield to habituationist learning methods, yet one can certainly pull out specific practices, such as correct wedding protocols, and teach them in a habituationist fashion.
This probably is enough said for now on the topic of skills versus dispositions. But I do want to reiterate the renegade nature of this kind of psychological theorizing. Examining the Implicit Learning literature, one finds the simple explanation I've offered here of the non-conscious character of the learning process to be largely discredited. After Reber introduced his first startling and compelling studies beginning in 1967, it took nearly 20 years before other psychologists paid any attention to it (Berry, 1997). When psychologists finally decide to pay attention, it was with a remarkable degree of resistance to the idea that implicit learning could be completely segregated from explicit knowledge. Berry’s (1997) edited volume, How Implicit Is Implicit Learning, and Frensch & Cleereman’s (2002) Implicit Learning and Consciousness show the range of opinion on the matter. Critics of the fully implicit character of implicit learning have gone to extraordinary lengths to document trace aspects of conscious awareness connected with implicit learning studies. In fact, this complexity of explanation is now embraced by Reber himself:
"As for the disputes about how implicit implicit knowledge is, you make the same mistake that almost everyone else makes. We do not and should not insist that all the apparent knowledge in these experiments is unconscious. Only that there is more that is 'known' implicitly than can be articulated --- and that this knowledge results in behaviors that cannot be understood purely by examination of consciously held representations." (Reber, September 17 2008, personal communication)
I do wonder if this resistance reflects preparadigmatic angst at the possibility of separate constructions of learning.
David Kirshner
-----Original Message-----
From: xmca-bounces@weber.ucsd.edu [mailto:xmca-bounces@weber.ucsd.edu] On Behalf Of David Kellogg
Sent: Monday, January 31, 2011 6:14 PM
To: lchcmike@gmail.com; Culture ActivityeXtended Mind
Subject: Re: [xmca] Ways With Words
Well, yes and no. That is, David Kirshner's post does clarify a lot of my misunderstandings of his work. But it also produces some new puzzles, at least for me.
For example, David says that A, perhaps even THE, main difference between a domain which is well framed for a skills teaching genre and one which is framed for enculturation/acculturation has to do with the way in which the domain is defined in words.
An "undefined" domain is conducive to enculturation. He gives the example of politeness. But of course there are far more (and even far better) definitions of politeness (e.g. wedding procedures) then there are of pecan-shelling processes, at least here in China.
It's not clear to me what it means to define a domain. You would think that marching together in unison is an almost perfect example of a skill. But when we actually try to "define" it, we don't end up giving recipes (e.g. "First everybody puts there right foot forward at exactly the same time"). That kind of thing might help you get the ball rolling but it in no way helps with the difficult part, which is what you do when you begin to march out of step.
In fact, the "definitions" we give are almost always tautological: they are not procedures so much as what Harris calls "constitutive rules"; they are descriptions of what it means to walk in unison. "When you begin to march out of step you have to look at the kid next to you and adjust your gait to his or hers by either slowing down or speeding up." That looks much more like conceptual knowledge, rather than procedural knowledge, so it belongs to the knowledge-as-concepts genre rather than the knowledge-as-skills genre.
But of course nobody actually teaches this way. For one thing, it presupposes a kind of descriptive knowledge that usually follows rather than precedes performance. If you gotta ask, you ain't never gonna know. For another, it assumes that the kid next to you is marching in step. Suppose he or she isn't?
So iInstead, we "just do it". You get a bunch of kids together and you get them to march more or less in unison, and they create a kind of culture-of-marching-in-step, using something like good old "Watch this" and "Follow me". That, of course, sounds like an acculturation genre followed by an enculturation genre. And of course it begs Salman Rushdie's question: How does newness come into the world?
David Kellogg
Seoul National University of Education
PS: Did anyone else feel that Fred Newman and Lisa Fulani's article was PIAGETIAN--that is, it separates development and learning, and places the former absolutely and logically prior to the latter? I did, and perhaps for that reason found it really VERY depressing.
dk
--- On Mon, 1/31/11, mike cole <lchcmike@gmail.com> wrote:
From: mike cole <lchcmike@gmail.com>
Subject: Re: [xmca] Ways With Words
To: "eXtended Mind, Culture, Activity" <xmca@weber.ucsd.edu>
Date: Monday, January 31, 2011, 7:42 AM
Both of you messages D&D, are helpful.
David Ki, do you have a manuscript on the en-ac-ulturation distinction and
the genre approach?
mike
On Sun, Jan 30, 2011 at 6:56 PM, David H Kirshner <dkirsh@lsu.edu> wrote:
David,
Thanks for the lovely wedding anecdote, and also the example of your
mother-in-law inducting you into practices of pecan cracking through her
verbal instructions. The model of enculturation that most of us have to go
by comes from the situated cognition literature grounded in examples of
craft apprenticeship. In this model, identity development (from peripheral
to central participation) as well as the production of skills and concepts
are incorporated into a complex integrative model of learning. This seems to
be the frame that you are bringing to bear in understanding my enculturation
genre.
It is precisely this integrative assumption that the genres approach
resists. The sculpted genres of teaching enable a parsing of the discrete
elements interacting even within complex settings. For instance, the
shelling of pecans, though certainly a cultural practice, is precisely
defined and hence subject to being learned as a skill, outside of cultural
context. (Contrast this with open-ended cultural practices like politeness
or approaches to solving of non-routine problems which cannot be precisely
specified and hence must be learned in cultural context.) Indeed, though
your pecan-shelling lesson did transpire in an authentic cultural locale, I
would want to argue that the structure of the learning support for your
pecan-shelling prowess is from habituation instruction, not from
enculturation.
The key to habituated learning is unconscious (subcognitive) association of
perceptual stimuli and motor responses. Your mother-in-law's directions for
how you should hold and operate the apparatus served to make perceptually
salient certain aspects of the stimulus and response domains, and your
practice served to establish the requisite subcognitive linkages between
them. I think we can probably rule out concept teaching, as presumably your
mother-in-law was telling you what to do, rather than explaining principles
to you (not discounting the possibility that you, independently, chose to
"make sense" of what you were being asked to do). From a genres perspective,
habituation would be a sufficient explanation to account for your newfound
skill in pecan shelling. In fact, the requirements for enculturational
learning of this "practice" probably were not present.
Let me take a moment to unpack the two enculturation-related pedagogies in
order to be able to continue the genres analysis of your pecan-shelling
learning episode. One of the difficulties, given the prior model of situated
cognition theory and craft apprenticeship, is to imagine how enculturational
learning could be separated from identity development. However, in the
genres analysis, identity becomes a salient concern in the case of
alternative identity possibilities. For instance, in entering a craft
apprenticeship, one makes a decision to "become" a craftsperson (of a
certain sort). Thus one is actively seeking to acculturate oneself to the
practices of the culture.
This dynamic helps structure the "acculturation pedagogy" genre that I will
soon distinguish from the "enculturation pedagogy." In acculturation
pedagogy, a bona-fide member of the culture models mature cultural practices
in order that novices seeking to acculturate themselves to the culture can
emulate those practices. In your case, David, it doesn't seem that you
considered this to be a Chinese cultural practice, or even that you expected
your mother-in-law to be proficient in it. If anything, what you most
admired about her was her ability to transfer from her prior experience with
cracking peanuts and pumpkin seeds to new nuts and new devices. However, the
ability to transfer was NOT what you were learning. You were learning to
shell pecans.
Enculturation is an even worse fit than acculturation to your
pecan-shelling episode. Enculturation is the process of cultural absorption
that comes about when one is immersed in a unitary cultural milieu, for
instance a child within the national culture adopting the characteristic
practices of the culture. This kind of learning is accomplished without
conscious intention or awareness. The associated pedagogical genre has the
teacher work surreptitiously to develop the classroom microculture so that
it gradually comes to resemble the reference culture with respect to valued
practices. Students learn not because of an intention to assume a new
identity, but because they're immersed in a classroom culture that they
gradually become enculturated to, even as it continues to evolve. For
instance, a math teacher might seek to shape the culture of argumentation in
the classroom so that it comes to more closely resemble the kinds of logical
chains of reasoning that characterize mathematical proof. This is a gradual
process over a long period of time--not a good fit for your pecan-shelling
experience.
Thanks for engaging with the genres approach. I hope this helps clarify
some of the genres, and the way the genres framework is used to analyze
situations of learning and teaching.
David
-----Original Message-----
From: xmca-bounces@weber.ucsd.edu [mailto:xmca-bounces@weber.ucsd.edu] On
Behalf Of David Kellogg
Sent: Sunday, January 30, 2011 6:20 PM
To: Culture ActivityeXtended Mind
Subject: [xmca] Ways With Words
David (Kirshner) is of course quite correct to point out that narrative is
not a necessary or sufficient element in his “enculturation” educational
genre. But it does seem to me that the “enculturation” educational genre is
distinguished by greater discursivity; it’s a much talkier model, and I
think it is for this reason we often find it in traditional, more oral
cultures. Unfortunately (I think) we also find it in foreign language
classrooms, where I think it is fundamentally inappropriate.
First of all, it seems to me that enculturation almost necessarily involves
some kind of legitimate peripheral participation, non-essential but
nevertheless participatory roles in an activity that can in theory be taken
by children and outsiders. This clearly suggests a very hierarchical set of
roles, which, since they are not set by skills or by knowledge must be set
by some other criterion (e.g. being a so-called “native speaker”)
Secondly, it seems to me that enculturation models place a premium on doing
fairly simple, general, everyday things with great adroitness, creativity,
and confidence. An obvious example of this would be cooking, something which
everybody has to do but which can be done either with routinism, or with
verve and inspiration, or with the consummate mastery that is born of
endless routines illuminated by flashes of inspiration.
Thirdly (and as David says, this is where narrative “kicks in”)
enculturation means learning what Shirley Brice Heath calls “ways with
words”. If it were simply a matter of “Watch this” “Now you try it”, then
there would be no difference between the discursive model and the skills
model. Even if we add “Now, what was the difference?” we only get a skills
model plus explicit knowledge, and that is not what the enculturation model
is really about.
That’s really all I have to say here. The rest of this post is just two
anecdotes to illustrate. and if I had any sense I would just shut up at this
point. I am sure that many readers will stop reading at this point, if not
long before. But of course in the enculturation model, ways with words are
very important, and sometimes anecdotes and illustrations are more important
than the actual skills and concepts imparted.
The other day I was sitting here at this very table cracking newly imported
American pecans for my mother-in-law, who has had a stroke and can barely
speak. She was watching me intently, having never seen either pecans or the
jar-opening device I was using to crack them, and began to make speaking
sounds. I leaned over to listen and suddenly realized she was giving very
precise instructions about how to use the device so that the meat would not
be shattered.
Her body no longer obeys her brain, and she has reacquired the skills that
an infant must have in getting others to obey it instead. But in normal
times this simply involves laughing or crying, not “ways with words”. The
unusual thing about this was the objectivity, the precision, and efficiency
of her instructions: as soon as I held the jar-opening device the way she
told me to, my speed doubled, my efficiency tripled, and not a single
nut-meat was broken.
I realized that cracking peanuts and pumpkin seeds with immense precision
is something she has spent a large stretch of her non-working life doing
(she retired from the textile mill where she worked at forty years of age)
and she obviously had very developed views, transferable to entirely new
products and even completely new tools, about how it should be done. In
normal times (when we were both twenty years younger) she would have simply
shoved me out of the way and done it herself. But in this situation,
absolutely no other way of transferring her knowledge than a slurred mixture
of Shaanxi and Henan dialects, to which I am normally fairly impervious.
This circumstance is probably not unique; over thousands of years of human
history there were probably many situations where knowledge had to be
transferred in this highly imperfect way from disabled elders to not yet
able juniors. And so ways with words turn out to be as important as skilled
performances.
But unskilled performances also have to be included, first of all, to
provide the contrast that we have in skills models (“Watch this” “Now you
try it”) and the explicit knowledge we have in conceptual models (“See the
difference”), secondly to allow the elders to show the mastery on which
their authority must ultimately be based (we cannot always live off of the
capital of social position), and thirdly to allow some means by which
outsiders can teach insiders, as well as insiders teach outsiders, making
the enculturation model not entirely a closed system and allowing the whole
to develop new forms of knowing.
Yesterday my brother-in-law and I went to a wedding in a nearby village
where he is doing some business with the local village head, whose friend
had a son getting married. Village weddings in China are what I would call
loosely scripted: certain things must be done, but they are not done to
schedule; they happen when all the principals are accounted for and there is
enough of an audience to make it worthwhile. In order to make sure that the
audience shows up and stays, a huge tub of “saozi mian” (noodles) is kept on
the boil all day, and anyone can eat as much as they like, whether they are
related to the bride and groom or not.
There are lots of roles that call for little skill, but there are also
roles which can be fulfilled very skillfully. For example, when we first
arrived at the wedding, they were carrying the bride’s gifts to her new
inlaws into a room where the inlaws sat before portraits of their ancestors
to receive them.
My brother-in-law and I, along with some neighborhood children, took some
of them in (I took a large, purple plastic thermos bottle) and in return the
male adults were given cigarettes and the children were given milk sweets.
While my brother-in-law was smoking his cigarette (I stuck mind behind my