Friday, April 15, 2005

Reading :: Situated Learning

Originally posted: Fri, 15 Apr 2005 19:14:24

Situated Learning : Legitimate Peripheral Participation

by Jean Lave, Etienne Wenger

This short (138pp, including index) monograph packs in a lot about how people learn. I first read it -- oh, probably in grad school, when I noticed it was being cited quite a bit by activity theorists, distributed cognitionists, and people interested in the social aspects of computing. Rereading it has helped me to see the huge impact it had on learning theory -- but also the problems of applying it uncritically.

The idea of the book is to examine learning in all its contexts as a sort of cognitive apprenticeship in which people have to participate in an activity, first peripherally (on the edges, with minor responsibilities), then in ways that increase complexity and responsibility. This participation must be legitimate, that is, sanctioned and validated so that the activity is properly continued. When all three of these are put together, you get "legitimate peripheral participation." That's not as easy as it might sound. Schooling allows learning that is legitimate and peripheral, but not participation. Apprentice meat cutters, in Marshall's 1972 study, engage in learning that is legitimate and participation, but not peripheral -- they start with simple tasks such as wrapping, but are physically separated from the journeymen who actually cut the meat in the back room.

LPP takes its inspiration from traditional apprenticeship, of course, and Lave and Wegner discuss historical apprenticeship at some length (Ch.3). They criticize functionalist and Marxist understandings of apprenticeship, which both treat it "as a historically significant object" that "connotes both outmoded production and obsolete education (p.62). In particular, they take issue with Engestrom's association of apprenticeship with craft production, "emphasizing the individual or small-group nature of production, the use of simple tools and tacit knowledge, a division of labor based on individual adaptation, and the prevalence of traditional protective codes" (p.63). In contrast, Lave and Wenger "emphasize the diversity of historical forms, cultural traditions, and modes of production in which apprenticeship is found" (p.63).

Apprenticeship, they argue, allows people to enter a community of practice. And when conditions "place newcomers in deeply adversarial relations with masters, bosses, or managers; in exhausting overinvolvement with work; or in involuntary servitude rather than participation," these conditions "distort, partially or completely, the prospects for learning in practice." In these cases, "communities of practice may well develop interstitially and informally" (p.64).

After reviewing five cases of apprenticeship -- including, interestingly, participation in Alcoholics Anonymous -- Lave and Wenger spell out other characteristics of LPP. Apprentices often "move backward through the production process" -- beginning their apprenticeship by finishing details on a product, then progressively mastering the previous steps in reverse order (p.96). They develop a "circumferential" perspective by taking part in "partial, peripheral, apparently trivial activities" such as "running errands, delivering messages, or accompanying others," and in doing so, they absorb "a first approximation to an armature of the structure of the community of practice" (p.96). (Formal learning curricula, in contrast, often evolve from a perscriptive understanding of participation (p.97).) In participating in an activity, apprentices also participate in a heritage:

A community of practice is a set of relations among persons, activity, and world, over time and in relation with other tangential and overlapping communities of practice. A community of practice is an intrinsic condition for the existence of knowledge, not least because it provides the interpretive support necessary for making sense of its heritage. Thus, participation in the cultural practice in which any knowledge exists is an epistemological principle of learning. The social structure of this practice, its power relations, and its conditions for legitimacy define possibilities for learning (i.e., for legitimate peripheral participation). (p.98)

Heritage is a big theme here, and the idea of apprentices slowly absorbing and becoming part of a community of practice is interesting because it reflects what I've been calling the weaving perspective, in which activities develop slowly and more or less linearly over long periods of time. It doesn't do a good job of examining spliced work in the so-called new economy, and it also doesn't do a good job of exploring the resistances, tensions, and contradictions involved in absorbing new members -- something for which Lave and Wegner have been criticized from a variety of perspectives, including activity theory.

In terms of technology, "understanding the technology of practice is more than learning to use tools; it is a way to connect with the history of the practice and to participate more directly in its cultural life" (p.101).

Lave and Wegner close the book with a discussion of a major contradiction in LPP: between the continuity of the practice over generations vs. the displacement as old-timers are replaced with new participants (p.114). They propose this contradiction in addition to the fundamental contradiction of commodity (use-value vs. production-value) that activity theorists such as Engestrom often cite (p.114). "The continuity-displacement contradiction is present during apprenticeship .... The different ways in which old-timers and newcomers establish and maintain identities conflict and generate competing viewpoints on the practice and its development" (p.115). So there is resistance, there is conflict and contradiction, even though critics of LPP have tended to downplay these and portray it as an unproblematic absorption of new members.

In the conclusion, Lave and Wegner explain that LPP "obtains its meaning ... in its multiple, theoretically generative interconnections with persons, activities, knowing, and world" (p.121). When they say "interconnections," evidently they mean dialectic, since they use the term elsewhere in this monograph. "Legitimate peripheral participation has led us to emphasize the sustained character of developmental cycles of communities of practice, the gradual process of fashioning relations of identity as a full practitioner, and the enduring strains inherent in the continuity-displacement contradiction. This longer and broader conception of what it means to learn, implied by the concept of legitimate peripheral participation, comes closer to embracing the rich significance of learning in human experience" (p.121).

All in all, this is a really thought-provoking book. But I find myself thinking, as I implied earlier, about the sort of work that seems not to lend itself to this model. In particular, I don't see much place for historically new activities; for innovations; for multiplicity (this is perhaps because of Lave and Wegner's reliance on dialectic); in a word, for splicing. I would have difficulty applying LPP, for instance, to a new or substantially changed profession brought about by new legislation, as is often true in the telecommunications industry, or to hybrid organizations brought about by mergers. Looking at Lave and Wegner's examples, I note that meat cutting and tailoring have both undergone radical changes in work organization and automation over the last few decades, changes that were in full swing by the time the book was written, and Lave and Wegner do little to address those changes. Is LPP applicable to distributed capitalism? I'm sure it is, but not as a general explanatory principle.

Blogged with Flock

Wednesday, April 13, 2005

Reading :: Aramis or the Love of Technology (supplemental notes)

Originally posted: Wed, 13 Apr 2005 00:37:20

Aramis or the Love of Technology

by Bruno Latour

I reviewed Aramis earlier, but since I have just finished rereading it along with my grad students, the time seems opportune for some follow-up observations.

First, I think Aramis can be read as a treatise on research methodology: much of it is about how qualitative researchers conduct their research, how they represent their participants, and what it means to "do research." Latour's comparison of traditional and relational sociology in particular cuts to some really interesting divisions with broad methodological repercussions.

Second -- and this is in conjunction with some other readings I've been doing recently -- Latour's text has to be carefully parsed when we read his declarations about power relationships and alliances. One of my grad students declared halfway through the book that Latour was really talking about manipulation: the many actors attempt to enact their programs by convincing others to do their will. But I tend to think that manipulation entails negotiations in which one party makes all the compromises -- and I don't think that happens in Aramis. All parties find themselves compromising, yet in the end, none of them compromise enough to save the project. And this leads us to the often-leveled accusation that Latour and other actor-network theorists are Machiavellian. There is some basis to these accusations, so it's surprising to find Latour's characters accusing each other of being Machiavellian (as if it were a bad thing) and Latour's protagonist dismissing such charges. Parse the book carefully, though, and you'll see that Latour is denouncing the caricature of Machiavelli, the amoral opportunist and manipulator; his methodology is still grounded in the historical Machiavelli, the pragmatic observer who constantly examines, tests, and forges alliances through open and covert negotiations. Latour gives us what I take to be sly, uncited references to that Machiavelli. For instance, Latour's protagonist says:

"That's exactly why military types have learned to draw up strategies and hierarchies, why they've invented uniforms and epaulets, why they've created the maps, why Gribeauval perfected the general staff, why military orchestras were signed up quite purposefully -- precisely so that a strategy could become possible no matter what." (p.177)

The topic is one that Machiavelli takes up in The Art of War.

Other points also surface on my third reading of this book. For instance, I have noted that Latour ends up blaming everyone -- even himself -- for Aramis' "death." So how do we interpret this passage from halfway through the book? "An accusation that's so watered down it blames the whole wide world is even more futile than one that picks a scapegoat" (p.198). I suppose Latour is contrasting the blame of everyone involved in the project (the network) with the blame of culture or society.

On the next page, Latour contrasts classical and relativist (relationist) sociology. Classical sociology presumes a social structure that functions as an explanatory device; it "can comment on what the patients say because it possesses metalanguage, while they have only language" (p.199). But his relationist sociology doesn't use this explanatory device: it "has no fixed reference frames, and consequently no metalanguage. It expects the actors to understand what they are and what it is" (p.200). Again, this relativist sociology should sound familiar. Here's a quote from Machiavelli's The Prince to drive things home:

But since my intention is to say something that will prove of practical use to the inquirer, I have thought it proper to represent things as they are in real truth, rather than as they are imagined. Many have dreamed up republics and principalities which have never in truth been known to exist; the gulf between how one should live and how one does live is so wide that a man who neglects what is actually done for what should be done moves toward self-destruction rather than self-preservation. (p.50)

Is it a coincidence that the social structure described in Latour's quote sounds a lot like an activity system? Or that activity systems have been notoriously hard to nail down, as some have recently discussed?

In the end, I was glad to take another crack at this book. It doesn't read as well as a novel, but it does represent an audacious piece of work.

Blogged with Flock

Reading :: Dune

Originally posted: Wed, 13 Apr 2005 09:12:25


by Frank Herbert

In class today, we were finishing up our discussion of Aramis and my grad students were complaining that although it was an interesting study, it was not an especially good novel. Yes, I agreed, it had the form of a novel but was still ensconced in the genre of ethnography, making it an uncomfortable and at times incoherent hybrid -- sort of like certain theme albums. Shortly after this exchange, one grad student pointed out that the glossary at the end, which defined a seemingly endless series of acronyms and participants, seemed a lot like the similarly endless back matter in Frank Herbert's Dune.

Oh yes, I agreed. How interesting that you should mention that. I'm reading that book right now.

He lifted his eyebrows and said something like: "Dune and Rush? We're getting some really scary insights into your personality." Ouch.

Why am I reading Dune? The truth is that my wife and I were cleaning out the garage the other day, opened a box, and discovered all of these books from our childhood. Correction: most were hers; I had cleaned house a while back and gotten rid of most of my collections. Now I have only a handful of books. But I just couldn't get rid of Dune.

That's not because it's a good book, although it is. It's because this book, this very copy, represented an entire youth spent reading countless paperbacks countless times. My older brother gave me Dune for Christmas one year, and I must have been quite young. Maybe fifth grade. I know that I read Tony Rothman's The World is Round in fifth or sixth grade -- I remember paging through it in Mr. Long's class -- and I'm quite sure I had read Dune before that. A few hundred pages' worth of assassinations, breeding programs, slavery, and apocalyptic imagery is probably not the ideal gift for a fifth-grader, but it did set the bar fairly high for science fiction; I remember being distinctly underwhelmed by Foundation in comparison.

How did it set the bar high? Like other disturbingly intricate works of fiction, Dune managed to describe an entire reality in detail. Herbert clearly had far more stories in his head than made it onto the page. So each section of each chapter starts with a quote from a book or religious text from this universe; characters made constant references to other happenings and to historical events outside the present story; and a string of appendices gave the most intrepid reader still more insight. Like Lord of the Rings or The Matrix, it created a universe with enough hinted detail that it could become the nexus of what Jason Craft calls a fiction network: an ever expanding set of stories told by other authors and fans. You can imagine.

Dune also has an interesting take on the future. It avoids becoming outdated by keeping the technology to a minimum (the plot device here is that humans had destroyed and banned all thinking machines centuries earlier, after being enslaved by them) and by using feudalism to underpin the social structure. The one thing that really dates the book is that the fanatical Fremen (who at one point are identified as descendants of Sunni) stun onlookers by being willing to sacrifice their own lives. In one scene, a battle-hardened observer sees a Fremen pilot a stolen airship right into a troop carrier, killing hundreds of enemy soldiers along with himself. "What kind of men are these?" he wonders. After 9/11, the act seems unremarkable.

Dune is Machiavellian in the crude sense: amoral rulers and elaborate treacheries abound, but we detect none of the enthusiasm for republican government that the historical Machiavelli showed. Indeed, the message seems to be that the strong were meant to rule over the weak, and the strongest of all is also one who has the divine right of rule. It is also sociological in the crude sense, postulating a complex set of social structures and a broader fate of the species that ultimately controls them. In this sense, it suffers in comparison with Aramis which (written by a sociologist) is much more sophisticated and much more Machiavellian: Aramis has the same republican urge and the same determination to follow the actors that old Machiavelli had. Dune, in contrast, is essentially the sociological landscape seen through the eyes of a fourteen-year-old boy. When we first meet Paul Atriedes, he's fourteen, and already amazingly intelligent, a proficient fighter, a man who has total control over his emotions and actions -- the kind of person a young male reader of science fiction would want to be -- and his world is tailor made for him, full of people who are alternately trying to kill him and amazed by his abilities and wisdom.

It's not surprising, then, that Paul Atriedes doesn't seem to change much as he matures into an eighteen-year-old emperor of the galaxy. Character development in Dune is about the same as in Stephen R. Donaldson's earlier works, which is to say, imperceptible. The characters all start out grim and with "terrible purpose," and Herbert tries to steadily ramp up this grim terrible-purposedness. When Paul Atriedes' mother is frightened near the end by his very grim grimness and his terribly terrible purpose, we have a hard time seeing why: Herbert stacks the adjectives like cordwood, but we don't see a whole lot of concrete change beyond that.

Unfortunately, Herbert's later work in this universe suffers from the same malaise as Donaldson's. The dramatic thread tapered off in the books Dune Messiah and Children of Dune, then falls off a cliff in works such as God Emperor of Dune, which I charitably like to believe was ghostwritten by Herbert's son Brian. Brian Herbert has subsequently written an enormous number of Dune-themed books, none of which I have read.

Dune is a guilty pleasure, and reading through it this last time, I wonder if its style and detail helped to prepare me for my current reading diet, which mostly consists of ethnographies constructed from similarly detailed fragments. It also reminds me of the many, many hours I spent reading similar fiction. I thought of Barlowe's Guide to Extraterrestrials, which led me to Ringworld and Childhood's End; Inherit the Stars, with its sublimely creepy cover art and wooden prose; The Starchild Trilogy, whose exploding collars were appropriated by The Running Man; and so many others. Childhood's end indeed.

Blogged with Flock

Monday, April 11, 2005

(Reading Roundup: Engestrom and Engestrom, Engestrom, & Karkkainen on the third generation of activity theory)

Originally posted: Mon, 11 Apr 2005 09:42:44

My reading of Between School and Work yielded several cites that seemed directly relevant to my recent interests: (1) the relationship between dialogism and activity theory and (2) work fragmentation. So I looked up a couple of the cites, both by Yrjo Engestrom and coauthors. They provided some insight into how Engestrom and other activity theorists have been handling these questions, but raised other questions.

Engestrom, Yrjo. "Developmental work research as educational research."

This is apparently an introductory essay for a special issue of Nordisk Pedagogik on developmental work research, which is the project that Engestrom and his collaborators have been developing for some time. This project has its roots in Engestrom's Learning by Expanding.

As the name implies, developmental work research uses a developmental perspective of work, even in increasingly fragmented work. The way to follow developments, Engestrom says, is by examining the contradictions or double binds that build within and between activity systems.

The concept of activity took the paradigm a huge step forward in that it turned the focus on complex interrelations between the individual subject and his or her community. In the Soviet Union, the societal activity systems studied concretely by activity theorists were largely limited to play and learning among children, and contradictions in activity remained an extremely touchy issue. (p.132).

But in the 1970s, researchers in the West picked up the concept and began applying it to different domains, such as work (p.132). Engestrom credits Ilyenkov with conceptualizing "the idea of internal contradictions as the driving force of change and development in activity systems" (p.133), but modestly avoids discussing his own considerable role in elaborating and popularizing the idea. This work led to the "third generation" of activity theory, which has begun to examine connections and contradictions between activity systems as well as within them. "The third generation of activity theory needs to develop conceptual tools in order to understand dialogue, multiple perspectives, and networks of interacting activity systems" (p.133). Engestrom discusses some of the work on developing each of these, including James Wertsch's and Ritva Engestrom's separate work on dialogue; Holland & Reeves' work on perspective; and the question of activity networks, which has been developed by Saarelma as well as Engestrom and his collaborators. A "discussion" between "Bruno Latour's" actor-network theory and activity theory "has been initiated" and Engestrom points to boundary-crossing as another "tool." In this third generation of activity theory, "the basic model is expanded to include minimally two interacting activities" (p.133).

Now Engestrom gets to the question of changes in work -- "radical changes," over "the last ten years," he says (p.133), changes that Engestrom examines in terms of flexibility and collectivity (p.134). He sees activity theory and developmental work research as playing a special role in the range of research approaches and theories brought to bear on work (p.134).

What special role?

Developmental work research is a multidisciplinary approach. However, it is also a distinctively educational approach. The educational quality of the approach is threefold. For the first, developmental work research studies processes of learning and development as its central objects. Secondly, the methodology of developmental work research is based on educational interventions. And thirdly, developmental work research also studies education as work and educational institutions as workplaces. (p.134)

Yes, developmental work research is educational in nature, hence the enduring focus on development. The notion of contradictions, as Engestrom hinted earlier, takes a major role here as the source of change and development; an activity system must continue to change so that it can loosen the double binds and reconcile the contradictions that dog it. A "third generation" analysis recognizes the contradictions set up across multiple activity systems, some of which become visible as the result of "boundary crossing" -- such as the boundary crossing individuals undergo as they move from school to work, or from job to job.

On the other hand, some organizations such as state monopolies "have multiple mechanisms to muffle and buffer the contradictions so that they do not take the form of mounting experienced disturbances in everyday work," "postponing a double bind inside the workplace while the situation may be approaching a crisis when observed from the outside" (p.136).

Engestrom goes on to describe a "change laboratory," which sounds very much like a participatory design activity to me (p.138).

So what do we get out of this overview? First, we get the general program Engestrom and his collaborators have set for a "third generation" of activity theory. Second, we get an idea of how changes in work, including what I'm loosely calling work fragmentation, have precipitated this third generation and developmental work research in particular. Third, we get the centrality of education and the importance of contradictions within that framework.

Remember that Latour looks askance at the notion of contradictions. I'll have to review these sections. But I doubt that it's a coincidence that Latour also avoids a developmental, educational perspective in his work. Latour's language of change is translations, which emphasize contingencies; Engestrom's language of change is contradictions, which emphasize orderly tensions and progressive development.

Engestrom, Yrjo, Engestrom, Ritva, and Karkkainen, Merja. "Polycontextuality and boundary crossing in expert cognition: Learning and problem solving in complex work activities."

This article is a bit more directly related to the themes of Between School and Work. It tackles the question of expertise. What is it? Engestrom et al. charge that expertise has been traditionally defined vertically, in terms of the stages a person passes as she becomes more expert in a specific domain: an expert chessmaster, an expert ditch digger, etc. But the authors argue that there is also a horizontal dimension to expertise:

In their work, experts operate in and move between multiple parallel activity contexts. These multiple contexts demand and afford different, complementary but also conflicting cognitive tools, rules, and patterns of social interaction. The criteria of expert knowledge and skill are different in the various contexts. Experts face the challenge of negotiating and combining ingredients from different contexts to achieve hybrid solutions. The vertical master-novice relationship, and with it, in some cases, the professional monopoly on expertise, is problematized as demands for dialogical problem solving increase. (p.319)

Yes, yes. I agree here: activity theory has had to evolve to take into account the increasing interpenetration of work activities. As Engestrom said in the previous article, this increasing interpenetration yields increasing potential for contradictions among the activities, and at the same time these activities are too massive and too interconnected with other activities to simply synthesize. Those separate "contexts" and separate dialogues aren't going to go away. So Engestrom et al. are positioning activity theory/developmental work research to address work multidimensionally.

To do so, they address "two central features of this newly emerging landscape": polycontextuality and boundary crossing (pp. 319-320). Polycontextuality involves working on tasks from different activities or frames of work simultaneously. And on the level of activity -- here's the authors' contribution -- polycontextuality "means that experts are engaged not only in multiple simultaneous tasks and task-specific participation frameworks within one and the same activity. They are also increasingly involved in multiple communities of practice" (p.320).

Polycontextuality leads to boundary crossing. When two different activities are linked together -- such as two organizations that must collaborate to provide a service, or two workplaces that have emerged separately but have been acquired by the same company and must now work together -- "the two contexts" in linked activities must "be iteratively connected." The tools, relationships, social languages, and so forth may be very different; the linked activities need "boundary crossers" who can mediate between them (p.321). The authors refer to Suchman's 1994 article in Computer Supported Cooperative Work, which I should read soon, along with a string of other cites which suggest that boundary crossing is becoming the hallmark of knowledge work.

The authors then turn to Susan Leigh Star's valuable work with boundary objects, which they characterize as "a useful attempt at identifying mediating artifacts that may help overcome 'groupthink' and fragmentation" (p.322). I don't think this was Star's original intent, but this is the potential Engestrom et al. see for it. (Latour talks about boundary objects approvingly as well, though I don't think he would agree with this characterization.)

The authors then go into three case studies that illustrate various aspects of boundary crossing. These are valuable, but not as valuable as the position-staking that goes on earlier in the chapter, which is really quite interesting. I come back to the point that the authors are trying to adapt activity theory in the face of increasing work interpenetration and fragmentation. The question I keep coming back to is the question of whether a developmental perspective buys us much in the face of increasingly interpenetrated work. As has been pointed out elsewhere, the level of activity is a very abstract thing, extrapolated and inferred rather than observed, and I am not sure how many abstract activity systems can be inferred with fidelity once we reach a certain threshold of interpenetration.

Blogged with Flock