RE: [xmca] Kevin's paper for discussion: causality

From: O'Connor, Kevin (
Date: Mon Jul 10 2006 - 21:03:29 PDT

Hi all,

I'm enjoying the interesting discussion of my paper, and Iím sorry I havenít participated more up to now Ė I was out of town for the past week or so and had unexpected difficulty getting online. Iím now working my way through the discussion, which has given me a lot to think about Ė some of the directions the discussion has taken are not directions Iíve foregrounded, so itís been provocative.

I want to respond briefly to a request for clarification made by Steve early in the discussion, pertaining to how I used the notions of presupposing and entailing indexicality. The concepts are taken from semiotic accounts of culture within anthropology, originating in the work of Silverstein, and have an important place in that discipline that goes far beyond what I used them for. Iíll summarize only my more limited use of these ideas here.

Indexical signs take on meaning through a relationship of contiguity with their object, or what they represent. So indexical meaning is always context-bound. Clear examples of linguistic indexicals are deictics, such as pronouns, which can only be interpreted in the context of their occurrence. (These kinds of denotational indexicals have a particularly important role in semiotic theories of culture Ė and while Iíve written about these elsewhere, they were largely backgrounded in the paper under discussion.) All linguistic forms can serve as indexical signs, though Ė certain forms occur with greater than chance probability in certain places, among certain kinds of people, and in certain social practices, and become associated with those places, people, and practices. So the use of these forms indexes, points to, implicates the existence of, a certain kind of place, person, or practice; to speak in a certain way is, at least potentially, to make an occasion into a particular kind of occasion. It was these kinds of indexicals that were the main focus in the paper being discussed.

Indexical meaning isnít necessarily established straightforwardly, though - the use of a given form doesnít determine the context, since in any given stretch of discourse, there are always multiple potential projections of context, some of which are inevitably in tension with one another. For an interaction to proceed, there must be some presupposable context or frame within which utterances can be interpreted and that gives some degree of stability to the interaction. This presupposed context is always contingently achieved, though, and can always be transformed or dissolved if other potentially available indexical connections-to-other-contexts are made relevant - that is, if certain linguistic signs come to have entailments for the course of the interaction. So these presupposing and entailing aspects of indexicality are always in tension with one another in any interaction. Itís useful for me here to think of Bakhtinís tension between centripetal and centrifugal forces in an utterance, one tending toward closure and the other towards openness. My intention in using these dual aspects of indexicality was to try to demonstrate some limitations in the kinds of analyses which are typically done of ďsituated learning contextsĒ Ė that is, a primary focus on an official presupposed context, and a relative neglect of one type of indexical entailment, i.e., the intrusion of other, unofficial and most likely unwelcome contexts into this official construal of whatís happening.

I hope this clarifies somewhat. Iím enjoying the discussion, and will jump in again. Iím interested in how ideas of causality and units of analysis are being discussed here, and want to say something about those issues.


-----Original Message-----
From: on behalf of Steve Gabosch
Sent: Sun 7/9/2006 8:41 PM
To:; eXtended Mind, Culture, Activity
Subject: Re: [xmca] Kevin's paper for discussion: causality
Bill's 7/3 comments and questions on cause and effect (bottom of the
post) are a great starting place to look at Kevin's paper. In
addition to causes, Bill emphasizes the question of units of
analysis. What actual statements about cause and effect - and units
of analysis - are in this particular paper? As a kind of experiment,
I re-read the paper with this question in mind.

Below I take a detailed look at the 6 paragraph introduction, where
Kevin introduces many of his units of analysis and the causal
relations that guide his study. I have bolded various phrases and
passages as a way of taking notes. These can be ignored. My
comments are in *****italics. My purpose here is to drive out the
units of analysis and the causal statements to see if doing so is
useful. Along the way I make a few other comments while I am at
it. But the "units and causes" analysis is the main thing. See what
you think.

passages are copied from the pdf version at

Communicative Practice,
Cultural Production, and
Situated Learning

Constructing and Contesting
Identities of Expertise in a
Heterogeneous LearningContext

Kevin O'Connor

In this chapter, I consider some ways in which linguistic anthropology
can contribute to understanding "the cultural production of the educated
person" (Levinson & Holland, 1996), to use a phrase that nicely
captures an important focus in recent anthropological approaches to
education. Work on cultural production is part of a broader project in
the social sciences over the past three decades, a project that explores
how both persons and forms of social organization are constituted
through social practice. Among the major aims of this work has been to
challenge conceptions of culture as a stable and relatively unproblematic
body of knowledge that is transmitted from one genera tion
to the next. Instead, culture is seen as a dynamic process in which agents
create meaning by drawing on cultural forms as they act in social and
material contexts, and in so doing produce themselves as certain kinds
of culturally located persons while at the same time reproducing and
transforming the cultural formations in which they act.

Thus "cultural production" has a double meaning: it is concerned
with how persons are produced as cultural beings, and with how this
production of persons results in the (re)production of cultural formations.
Recent anthropological approaches to education have been concerned
with this process as it relates to learning and schooling. This
work has focused on the interplay between social structure and human
agency in sites in which "educated persons" are produced. In this view,
becoming "educated"-or "uneducated," or even "uneducable"-however
these might be locally understood, is an important way in which
persons become produced within cultural groups, and thereby contribute
to the production of the culture.

*****Kevin outlines his essential unit of analysis: cultural
production. Cultural production refers to two simultaneous realms:
how people create themselves, and how people reproduce
culture. Behind this unit of analysis is a causal theory. Persons
and forms of sociocultural organization are "constituted through"
("caused by") social practice. In particular, the "educated" and
"uneducated" person is "produced by" ("caused by") the interplay
between social structure and human agency.

This chapter focuses on how local processes of interaction are related
to broader, and often conflicting, conceptions of what it means to be
educated. Specifically, I examine the negotiation of "identities of expertise"
in one site designed to produce educated persons-a multi-institution
undergraduate engineering project. This site is of interest for
several reasons. First, the project was part of an attempt to challenge
overtly what it means to be an expert in the discipline of engineering.
It did this by attempting to elevate the status of traditionally devalued
"practical" aspects of engineering activity. That is, practical, and not
just theoretical, knowledge was taken as central to being educated.
Second, and closely related, the consortium was challenging traditional
views of how one becomes an expert by designing practical projects as
privileged sites for learning, as opposed to teaching engineering science
outside the context of "real world" activity.

*****Kevin expands on his unit of analysis, cultural production, and
introduces numerous elements of cultural production that are relevant
to this paper, such as "local processes of interaction", "conflicting
conceptions", "what it means to be" (educated), "negotiation",
identity, "identities of expertise", "expert", "status",
"traditionally devalued", "practical" vs "theoretical" "knowledge",
"traditional views" etc. In this shorthand, Kevin outlines how his
theoretical approach applies itself to real situations, in this case,
educating someone to be an "expert" engineer.

This reconceptualization of the nature of expertise and the process by
which it is attained drew heavily upon work by educational researchers
and designers who have adopted and developed a view of cognition
and learning that is itself grounded in theories of cultural production.
This allows for an examination of the cultural production of identities
of expertise, at a historical moment in which theories of cultural production
are themselves used as a partial basis for defining what it means
to be an "expert."

*****Kevin emphasizes that the conceptualization of "expertise" and
the "process by which it is attained" used in this paper is grounded
in theories of cultural production. What this means, I think, is
that Kevin is making it clear that the explanatory approach he is
using to guide his analysis of how expertise is created and
identified is based on the theory that social practice causes
(constitutes) cultural production.

The sponsoring consortium was also attempting to
promote "boundary crossing" between historically separate institutions.
In doing so, it had explicitly egalitarian objectives of providing
participants from lower-status and less technologically well-equipped
schools with access to the knowledge and the technological resources
of higher status schools.

*****Here, Kevin introduces a new dimension that is of special
interest to CHAT analysts: goals and motives. The objective of this
consortium project was to provide access to knowledge and resources
from higher-status, better-equipped schools to lower-status, less
well-equipped schools. The motive was (apparently and publicly)
"explicitly egalitarian" and is described as promoting "boundary crossing."

However, this had another, unintended effect.
It allowed for relationships to be negotiated among students who might
otherwise never come into contact with one another, thus creating new
possibilities for the construction of social identities.

*****Kevin explains that this attempt to provide resources from
better schools to other schools - cross boundaries - had an
"unintended effect." It allowed for students to "negotiate" the
"construction of their social identities". This process became the
central empirical aspect of Kevin's paper.

Thus, this project allows me to examine processes of "identification"
and "contextualization"
-that is, the mutual production of identities and contexts for
activity-under conditions of overt conflict and transformation.

*****Kevin introduces two key processes in the theory of cultural
production: "identification" and "contextualization." These refer to
"mutual production" of "identities" and "contexts for
activity". These processes took place "under conditions of overt
conflict and transformation." Later in the paper, Kevin will spend
some time explaining why investigating conditions of overt conflict
are needed to expand on the work of Lave and Wenger, who focused on
communities of practice in relatively "benign" conditions in their
studies. This sets up another series of cause and effect relations
that occupy this paper.

In analyzing this setting, I will draw on recent work in linguistic
anthropology. Scholars in this field have been centrally concerned, as
Duranti (1997) points out, with processes of cultural production, and
have been developing sophisticated theoretical and methodological
resources for understanding how language is involved in the construction
of meaning and the production of persons and cultures. I
will focus on how, in the detailed processes of moment-to-moment
interaction, language is used to produce a world in which certain
kinds of expertise are valued (or devalued) while at the same time
speakers position themselves and others within those ways of understanding

*****A comment: I find this a promising methodology (not a new idea,
of course!). I like the general idea of using tools of linguistic
analysis to understand how meaning is constructed, and how language
contributes to producing persons and cultures. This approach seems
especially useful for non-experts to learn how to use when expertise
is being claimed and questioned, and when speakers claiming expertise
are jockeying for position. What patterns in the use of words,
meanings and contexts can be used to help us make judgments about
"experts" and their claims and counterclaims?

*****But back to units of analysis and causes. Kevin draws an
explicit causal relationship. A world where kinds of expertise are
valued and devalued, and speakers position themselves in regard to
these understandings of expertise, is "produced" ("caused") by the
use of moment-to moment interaction (language).

*****I need to make an important point here. In causal relations,
described in everyday language, conditions and actions work together
to cause effects. In the kind of analysis Kevin is doing in this
paper, he is not trying to detail *all* the conditions and actions
that are necessary to create the effects he is describing. He is
usually focusing on one. In the immediate case, he is looking at the
use of language to value and devalue expertise. He is not saying
that language is the "only thing" needed to create a world where
certain expertise is valued, devalued, etc. Kevin knows, and assumes
the reader knows, that other conditions, like certain societies,
certain cultures, certain social formations, are also necessary
conditions. These are all implied in his opening paragraph, where he
frames this paper in terms of "anthropological approaches to
education." A potential problem emerges here, as with any
theory. Just what are these other necessary causes in addition to
language? In my opinion, they must inevitably be taken into account
for the theory to flourish. Not specifying these other necessary
anthropological/sociological causes (for the valuing and devaluing of
kinds of expertise, or any other effect) will eventually create
demands on the theory of cultural production to either create
explanations for these other necessary causes, or become subsumed
under a more viable theory that does. Many of Bill's questions in
his 7/3 post are exactly about searching for the relevant causes by
selecting the right units of analysis. This is the challenge every
theory must face.

The chapter is organized in the following way. First, I discuss recent
theories of situated learning, which attempt to conceive of learning in
terms of the kinds of processes of meaning-making that are central to
theories of cultural production. I outline two general ways in which
situated learning theories have been developed-which I call the "cognitive
apprenticeship" approach and the "cultural production" approach-
and argue that the second adopts a more adequate view of
contextualization and identification, with consequences for how learning
contexts should be examined.

*****Kevin discusses (pg 64 - 69) more than these two approaches - he
also discusses cognitivist approaches to learning and contrasts them
with situated learning approaches in general, of which he in turn
contrasts the two directions situated learning approaches have taken,
"cognitive apprenticeship" and "cultural production". A great deal
of Kevin's argument is based on his critique of the cognitivist
approach (abstract, inert, non-contextual learning) and the
advantages of the situated learning approach.

*****Causal statement here: Kevin emphasizes that the cultural
production approach "adopts" (roughly, "causes") a more adequate view
of contextualization and identification "with" (and "causes")
consequences for how to examine learning contexts.

I then turn to a discussion of linguistic
anthropological approaches to understanding contextualization, and
outline a view of this process as involving a tension between "presupposing
indexicality" and "entailing indexicality." Examination of
the dynamic interplay between these two kinds of linguistic signs can
usefully contribute to our understanding of the processes through
which both cultures and persons are produced and transformed
through activity, both within and across interactions.

***** I would have liked Kevin to explain "presupposing indexicality"
and "entailing indexicality" more. Unfortunately, I am not walking
away from this paper with a clear idea of what they are or how to use
them in linguistic analysis. Perhaps he can do this in our discussion.)

*****Returning to our "units and causes" analysis, Kevin draws a
clear causal relationship between examining these two linguistic
signs (key elements) and using them to contribute to our
understanding of cultural processes. Kevin maintains that examining
this dynamic interplay can "usefully contribute to" ("cause") a
better understanding of cultural processes.

I then illustrate these points by examining communicative practices
in two interactions
that took place during the student project that was the site of my

*****Kevin examines these interactions in detail (pg 76-88). His
most interesting and general comments on them for me appeared in the
second paragraph of the conclusion (pg 88-89). On a theme that Kevin
emphasizes in this conclusion, Kevin first discusses "working class
identity" on pg 83 as a key element in the student
interactions. This theme touches on a number of the cause and effect
relations that are at the heart of Kevin's study, and his theory of
cultural production, which takes into account - at least up to a
point - class differences and tensions.


******So, is it useful to apply these two scientific principles,
units of analysis and causality, to an examination of a paper like
this? A "units and causes" analysis? I found it to be
helpful. What do others think? I think it helped me see more
clearly what was being said, and what theories Kevin was using to
write his paper. It is certainly possible I didn't read something
quite right - please correct me - but that would be me, not this
"method" of looking for the units of analysis and statements about
causality. Or maybe it *is* the method. Perhaps someone is
uncomfortable with it? Sees problems with it? Thoughts?

- Steve


At 05:46 PM 7/3/2006 -0700, Mike wrote:
>Wonderful questions, bb
>I am almost certain yrjo has written on this topic, all it requires is all
>of google's search capacity to find where in that vast
>array of ideas!
>On 7/3/06, bb <> wrote:
>>I've been re-reading Granott's paper on units of analysis in mca from a
>>while back, and it would seem that one needs to think of what unit to use
>>before proceeding. Jay Lemke writes about downward [qualitative] causation
>>at what I think is a societal scale, as well as do Durkheim and
>>Weber. Arguably for Leont'ev, Activity is the unit of causation. But so to
>>reflect this back to Kevin's paper, how do we think of C&E at a community
>>[CoP] level, and is this a large enough unit to begin to determine
>>causation, or do we need more? Does the community have a history, in which
>>what has happened before is an essential ingredient in what is happening
>>now? How far back should one go in an historical analysis? Does the
>>community border on others in any way so that interaction with others is
>>significant, i.e. there are exchanges of practices? Perhaps Hutchin's
>>micronesian navigators were/still_are an isolated community, so that
>>causation is entirely internal to it?
>>What especially does causation have to do with symmetry?
>>xmca mailing list
>xmca mailing list
xmca mailing list

xmca mailing list

This archive was generated by hypermail 2b29 : Tue Sep 05 2006 - 08:13:16 PDT