A Review of: Wertsch, J., Tulviste, P., & Hagstrom, F.
(1993). A sociocultural approach to agency.
In E. Forman, N. Minick, & C. Stone,
Contexts for learning (pp.336-356). New York: Oxford.
Stanton Wortham, Bates College
This is an ambitious chapter. It aims to use insights from the
sociocultural tradition to reframe three venerable questions in
the human sciences: (1) can we causally explain human action with
reference to agents, and if so what is the nature of these
agents? (2) can we give an integrated account of the micro-social
and the macro-social? (3) can intelligence be recast as
heterogeneous and not unified?
The authors do not claim to answer these questions definitively,
but they do sketch sociocultural approaches to each. Their
intuitions are appealing. Wertsch, Tulviste, and Hagstrom
suggest that we should not ground explanations of action in
isolated individuals, but instead in the indissoluble unit of
persons and their sociocultural situations. They suggest that we
must expand our accounts of socioculturally mediated action to
include the role of macrosocial factors in both intermental and
intramental functioning. And they suggest that a sociocultural,
macrosocially-sensitive account requires a heterogeneous, domain-
specific account of intelligence.
The authors begin with a question: "who is it who carries out
mental processes" (p.339)? Western psychologists traditionally
posit an individual agent, separated from social and historical
context. Against this view, the authors make two arguments.
First, "socially distributed cognition" cannot be explained
solely with reference to individual agents. Second, thought is
so bound up with mediators from outside the skin that we cannot
isolate a pre-social agent that drives intramental processes.
As an alternative construct, Wertsch, Tulviste, and Hagstrom
suggest "mediated agency" to explain the origin and structure of
mental processes. This construct acknowledges the essential
interrelation of sociocultural and intrapsychological factors in
cognition. Accounts of mediated agency do not take refuge in
decontextualized mental states, but instead explore the multi-
layered reality of thought. After sketching this alternative,
the authors claim that it should be expanded in two directions.
First, we must recognize that mediational means (cognitive tools)
carry influences from the larger society, and thus we need
macrosocial factors in our psychological explanations. Second,
our sociocultural accounts of intelligence should acknowledge the
domain specificity of human mental functioning. This domain
specificity fits neatly with a sociocultural account, as the
intermental and macrosocial precursors of individual thought are
themselves heterogeneous.
The authors should be commended for tackling such crucial issues
as agency, the macro-micro link, and the heterogeneity of
intelligence. I also sympathize with their larger project: to
recast such issues, using a more sociocentric account of thought
and action. In my view, however, two things must be done for
this project to bear fruit. First, each of these large issues
requires much more extended treatment, to work through the
implications of a sociocultural reframing. Second, the authors
need to push the more radical implications of their perspective
harder than they do. I will illustrate these two points with
reference to the sociocultural reframing of agency.
Wertsch, Tulviste, and Hagstrom give a plausible account of how
"the nature of individuals' mental functioning can be understood
only by beginning with a consideration of the social system in
which it exists" (p.340). Note, however, that this is much
weaker than the claim that all agency is essentially social. A
defender of individual agency could grant any amount of social
mediational means, but still claim that the causal origin of an
action lies in the individual's intention to perform the act.
The same could be said for distributed cognition. Perhaps the
cognitive processes are stretched over many people, but the
causal origin of the act might well still lie in an individual's
intention.
The construct "agency" is meant to explain what caused an act.
If we trace the causal chain behind an action, what entity
initiates that chain? What explains my smashing the piggy-bank
with the hammer? It's true that the hammer is a social tool, but
wasn't my intention the ultimate cause? To give a sociocentric
answer to the question of agency, we will have to move beyond
mechanisms, to an essentially collective account of causes.
There are alternative non-individualist accounts of agency. For
behaviorists, the ultimate causes of human action lie in the
environment, not in the individual. Sociological determinism
casts macrosocial forces as the causes of individuals' actions.
I am relatively sure that Wertsch, Tulviste, and Hagstrom do not
want to follow either the behaviorists or social determinists.
They want an account that genuinely bridges the inter- and
intramental. I worry, however, that their use of
"internalization" to connect the individual and the social locks
them into dualism, and thus blocks an adequate reconstrual of the
agency problem. My own predilections tend toward a more radical
attempt to overcome the dualism inherent in both individualist
and determinist accounts. Whether it involves internalization or
not, however, we should follow Wertsch, Tulviste, and Hagstrom in
working on a sociocultural alternative to individual intentions.