THEORIES OF COLLABORATIVE COGNITION: FOUNDATIONS FOR CSCL

AND CSCW TOGETHER

GERRY STAHL

Abstract Both computer-supported collaborative learning (CSCL) and

computer-supported cooperative work (CSCW) are centrally concerned with
teamwork, learning, problem solving, knowledge building, task accomplishment
and other cognitive achievements by small groups of people. There are many
theories useful for framing the cognitive work that groups undertake in CSCL and
CSCW settings, and they may in principle not be reducible to a single theory.
However, they seem foundational for both CSCL and CSCW, including their
synthesis in CSCL-at-Work. Collaboration research explores questions involving
numerous distinct—though interacting—phenomena at multiple levels of
description. It is important to conceptualize these group activities as involving
individual participants in group processes within larger community contexts. The
useful approach may be to clearly distinguish levels such as individual, small-group
and community units of analysis, and to differentiate terminology for discussing
these different levels. Theory in general has evolved dramatically over the ages,
with a trend to extend the unit of cognition beyond the single idea or even the
individual mind. Seminal theoretical works influential within collaboration research
suggest a post-cognitive approach to group cognition as a complement to—if not a
foundation for—analyzing cognition of individuals or of communities-of-practice.
While CSCL and CSCW can both build upon shared theories of cognition, they
may derive different implications from those theories as relevant for students
versus professionals and may converge in cases of CSCW at Work.

Theory for CSCL and CSCW

There is no one theory of collaboration in learning and working. Research in CSCL

and CSCW is guided by and contributes to a diverse collection of theories. Even
the word theory means different things to different researchers and plays various
distinct roles within collaborative-learning work. The reading of the history of
theory presented here is itself reflective of one theoretical stance among many held,
implicitly or explicitly, by collaboration researchers.
I originally tried to develop my theory of group cognition (Stahl, 2006) in
response to issues of CSCL and CSCW software design. In particular, one of my
research studies was an attempt to transform a basic CSCW system (BSCW) into a
basic CSCL system (BCSCL or Synergia) (Ch. 7), exploring both mutual
compatibilities and differences of emphasis between CSCL and CSCW. My other
case studies can be categorized as either CSCL (Ch. 1, 2, 6, 12, 21) or CSCW (Ch.
1
3, 4, 5) systems. These software development attempts—with their various
disappointments—led to my attempts to analyze the interaction and cognition
taking place (Ch. 8-13) and then—based on a recognition of the inadequacy of
available methods and conceptualizations—to investigations of relevant theory (Ch.
14-21), using data from these studies and later from a project specifically about
cognition in online teams (Stahl, 2009).
The nature and uses of theory have changed over history and continue to
evolve. The theories most relevant to computer-supported collaborative learning
and working—in the view developed in this chapter—concern the nature of
cognition, specifically cognition in collaborating groups.
Through history, the analysis of cognition has broadened, from a focus on
single concepts (e.g., Platonic ideas) or isolated responses to stimulae
(behaviorism), to a concern with mental models (cognitivism) and representational
artifacts (post-cognitivism). Theories that are more recent encompass cognition
distributed across people and tools, situated in contexts, spanning small groups,
involved in larger activities and across communities of practice. For collaborative-
learning and cooperative-work research, theory must take into account interaction
in online environments, knowledge building in small groups and cognition at
multiple units of analysis.

A brief history of theory

Consider the role of theory in a research field like CSCL or CSCW. These fields
are multi-disciplinary by their nature and as a result of their origins (see Stahl,
Koschmann & Suthers, 2006, for a history of CSCL from a perspective similar to
the one here). Consider the name of CSCL, Computer-supported Collaborative
Learning: it combines concerns with computer technology, collaborative social
interaction and learning or education—very different sorts of scientific domains.
CSCL and CSCW grew out of work in fields like informatics and artificial
intelligence, cognitive science and social psychology, the learning sciences and
organizational management—domains that are themselves each fundamentally
multidisciplinary. Theory in these fields may take the form of predictive
mathematical laws, like Shannon’s (1949) mathematical theory of information or
Turing’s (1937) theory of computation; of models of memory and cognition; or of
conceptions of group interaction and social practice. They may have very different
implications for research: favoring either laboratory experiments that establish
statistical regularities or engaged case studies that contribute to an understanding of
situated behaviors.
In the European tradition, theory (θεωρία) begins with the ancient Greeks—
especially Socrates, Plato and Aristotle—and continues through the 2,500-year-
long discourse of philosophy. In recent times, theory has veered into unexpected
directions as it has morphed into sciences based more on empirical research than on
intellectual reflection. For instance, the work of Freud, Darwin and Marx replaced
traditional philosophic assumptions about fixed natures of minds, organisms and
societies with much more dynamic views. Theory always transcended the opinions
of common sense—so-called folk theories based on the everyday experience of
individuals—to synthesize broader views. But folk theories have also changed over
time as they adopt popularized pieces of past philosophies; thus, a trained ear can
2
hear echoes of previous theories in the assumptions of common-sense perspectives,
including within current CSCL and CSCW research literature.
After the dogmatic centuries of the medieval period, philosophy took several
significant turns: the rationalism of Descartes, the empiricism of Hume, the
Copernican revolution of Kant, the dialectical development of Hegel, the social
situating of Marx, the existential grounding of Heidegger and the linguistic turn of
Wittgenstein. These all eventually led to important influences on theory in CSCL
and CSCW.
In particular, for instance, the field of educational research followed this
sequence of philosophic perspectives. Empiricism and positivism in philosophy of
science culminated in behaviorism in biology and the human sciences. The central
metaphor was that of stimulus provoking response, all objectively observable and
unambiguously measurable (as critiqued in Chomsky, 1959). The major theoretical
move of the generation before ours was to assert the necessity of taking into
account cognitive processes in studying human behavior, from Chomsky’s (1969)
theories of language based on deep grammar and brain mechanisms to the mental
models and internal representations modeled by artificial-intelligence programs.
Human-computer interaction, the part of computer science dealing with designing
for usage, has gone through a similar sequence of behaviorist and cognitivist
theories (see Carroll, 2003, for numerous examples). More recently, post-cognitive
theories have been influential in CSCL and CSCW, as will be discussed later.

The unit of analysis

The history of theory can be tracked in terms of the following issue: At what unit of
analysis should one study thought (cognition)? For Plato (340 BC/1941), in
addition to the physical objects in the world, there are concepts that characterize
those objects; philosophy is the analysis of such concepts, like goodness, truth,
beauty or justice. Descartes (1633/1999) argued that if there is thought, then there
must be a mind that thinks it, and that philosophy should analyze both the mental
objects of the mind and the material objects to which they refer, as well as the
epistemological relation between them. Following Descartes, rationalism focused
on the logical nature of mental reasoning, while empiricism focused on the analysis
of observable physical objects. Kant (1787/1999) re-centered this discussion by
arguing that the mechanisms of human understanding provided the source of the
apparent spatio-temporal nature of observed objects and that critical theory’s task
was to analyze the mind’s constructivist structuring-categorization efforts. Up to
this point in the history of theory, cognition was assumed to be an innate function
of the individual human mind.
Hegel (1807/1967) transcended that individualist assumption. He traced the
logical/historical development of mind from the most primary instinct of a living
organism through stages of intentional-consciousness, self-consciousness and
historical-consciousness to the most developed trans-national spirit of the times
(Zeitgeist). To analyze cognition henceforth, it is necessary to follow through its
biological unfolding and go beyond to the ultimate cultural understanding of a
society. Figure 1 identifies Hegel’s approach to theory as forming the dividing
line—or watershed—between philosophies or theories oriented on the individual
and those oriented to a larger unit of analysis.
3
Figure 1. Adapted from (Stahl, 2006, p. 289, Fig 14-1).

Philosophy after Hegel can be viewed as forming three mainstreams of

thought, following the seminal approaches of Marx (critical social theory),
Heidegger (existential phenomenology) and Wittgenstein (linguistic analysis). As
taken up within CSCL and CSCW, one can trace how these approaches established
expanded units of analysis.
Marx (1867) applauded Hegel’s recognition of the historical self-generation of
mankind and analyzed this historical process in terms of the dialectical co-
development of the social relations of production and the forces of production. His
analysis took the form of historical, political and economic studies of the world-
historical processes by which human labor produces and reproduces social
institutions. Here, the study of the human mind and its understanding of its objects
becomes focused at the epochal unit of analysis of social movements, class
conflicts and transformations of economic systems.
Heidegger (1927/1996) radicalized the Hegelian dialectic between man and
nature by starting the analysis of man from the unified experience of being-in-the-
world. The Cartesian problem of a distinction between an observing mind and an
objective world was thereby reversed. Heidegger, instead, had to show how the
appearance of isolated minds and an external world could arise through abstraction
from the primary experience of being-there, human existence inseparable from the
worldly objects that one cares for and that define one’s activity. The primordial unit
of analysis of cognition is the involvement of people in their world.
Wittgenstein (1953) focused increasingly on language as it is used to
accomplish things in the world through interpersonal communication. He rejected
his own early view (Wittgenstein, 1921/1974), which reduced a rationalist
conception of propositional, logical language to a self-contradictory position. Now,
linguistic meaning no longer dwelt in the heads of users or the definitions of the
words, but in communicational usage. Echoing the lived world of phenomenology,
4
Wittgenstein acknowledged the role of the human form-of-life. He also
conceptualized language as the playing of language games, socially established
forms of interaction. The unit of analysis shifted from mental meanings to
interpersonal communications in the context of getting something done together.
Marx, Heidegger and Wittgenstein initiated the main forms of post-Kantian,
post-Hegelian philosophy and scientific theory (Stahl, 2010c). Kant represents the
culmination of the philosophy of mind, in which the human mind is seen as the
active constructor of reality out of its confrontation with the objects of nature,
which are unknowable except through this imposition of human structuring
categories. With Kant—over two hundred years ago—the human mind is still a
fixed unit consisting of innate abilities of the individual person, despite how much
his philosophy differs from naïve realist folk theories, which accept the world as
fundamentally identical with its appearance to the human observer. Hegel
overthrows the Kantian view of a fixed nature of mind by showing how the mind
has itself been constructed through long sequences of processes. The Hegelian
construction of mind can be understood in multiple senses: as the biological
development of the brain’s abilities as it grows from newborn to mature adult; as
the logical development from simple contrast of being and non-being to the
proliferation of all the distinctions of the most sophisticated understanding; or as
the historical development from primitive homo sapiens to modern, civilized,
technological and cultured person. After Hegel, theory shifted from philosophy to
science, to explore the biological, logical and historical processes in more detail
and to verify them empirically. Followers of Marx, Heidegger and Wittgenstein
adopted approaches to this that can be characterized as social, situated and
linguistic. They are all constructivist, following Kant’s insight that the structure of
known objects is constructed by the knowing mind. However, they all focus on a
unit of analysis broader than the isolated individual mind of Descartes.

Seminal theories for CSCL and CSCW

The social, situated and linguistic theories of Marx, Heidegger and Wittgenstein
entered the discourse of CSCL and CSCW literature with researchers coming from
the various scientific traditions that went into forming these research domains,
including psychology, education, social science, design studies, computer science
and artificial intelligence (e.g., Dourish, 2001; Ehn, 1988; Floyd, 1992; Schön,
1983). Although these fields each introduced various theoretical perspectives, we
can see the major philosophic influences largely through several seminal texts:
Mind in Society (Vygotsky, 1930/1978), Situated Learning (Lave & Wenger,
1991), Lectures on Conversation (Sacks, 1962/1995) and Understanding
Computers and Cognition (Winograd & Flores, 1986).
Mind in Society is an edited compilation of Vygotsky’s writings from the early
1930s in post-revolutionary Russia, which has been influential in the West since it
appeared in English in 1978. Critiquing the prevailing psychology as practiced by
behaviorists, Gestalt psychologists and Piaget, Vygotsky did not try to fit
psychology superficially into the dogmatic principles of Soviet Marxism, but rather
radically rethought the nature of human psychological capabilities from the
developmental approach proposed by Hegel and Marx. He showed how human
perception, attention, memory, thought, play and learning (the so-called mental
5
faculties) were products of developmental processes—in terms of both maturation
of individuals and the social history of cultures. He proposed a dynamic vision of
the human mind in society, as opposed to a fixed and isolated function. The
Hegelian term, mediation, was central for Vygotsky, as it is for CSCL and CSCW.
Even in his early years still talking about stimulus and response, he asked how one
stimulus could mediate the memory of, attention toward or word retrieval about
another stimulus (p. iii). In Hegelian terms, this is a matter of mediating (with the
first stimulus) the relation (memory, attention, retrieval) of a subject to an object
(the second stimulus). This is fundamental to CSCL and CSCW because there the
learning of students or the work of professionals is mediated by technological
networking as well as by collaborative interaction. Another popular term from
Vygotsky is the zone of proximal development (pp. 84-91). This is the learning
distinction and developmental gap between what individuals can do by themselves
(e.g., on pre- and post-tests) and what they can do in collaboration (e.g., situated in
a small group). A group of children may be able to achieve cognitive results
together that they will not be able to achieve as individuals for a couple more years.
This is consistent with Vygotsky’s principle that people develop cognitive abilities
first in a social context—supported or mediated by peers, mentors or cognitive aids
like representational artifacts—and only later are able to exercise these cognitive
abilities as individuals. Vygotsky’s theory, if carried beyond where he had time to
develop it, implies that collaborative learning—including in workplaces—provides
the foundation upon which all learning is built. Methodologically, it argues against
judging the outcomes of collaborative learning by evaluating or assessing
individuals outside of their collaborative settings.
Situated Learning went beyond Vygotsky in expanding the unit of analysis for
learning at work. For Vygotsky and his followers, analysis must include the
mediating artifact (tool or word) and the mentor or group. For Lave and Wenger,
the unit of analysis is the even larger community-of-practice. Adopting the
theoretical and analytical centrality of social practices in Marx, they focused on
learning-at-work as the development of processes and relationships within the
communities in which individuals participated. Learning-at-work was viewed on
the model of apprenticeship, in which an individual gradually—and primarily
tacitly—adopts the practices that are established within the community in which the
individual is becoming a member. Within CSCL, this approach can be seen in the
idea that one learns mathematics by adopting the (predominantly discursive)
practices of mathematicians, such as using mathematical symbolisms, making
conjectures about mathematical objects and articulating deductive arguments
(Sfard, 2008). The CSILE project (Scardamalia & Bereiter, 1996), a pioneering
CSCL effort, tried to support the communicative practices seen in professional
research communities within the learning communities of school classrooms; the
unit of analysis for knowledge building mediated by the CSILE discussion software
was the discourse of the classroom as a whole. This illustrates a kind of CSCL-at-
Work in reverse, where learning incorporates work practices.
Lectures on Conversation laid the cornerstone of Conversation Analysis (CA),
which studies the linguistic practices of communities. It was based on the
ethnomethodological (Garfinkel, 1967) perspective, grounded in both
Wittgenstein’s linguistic analysis and Heidegger’s (1927/1996) and Husserl’s
(1936/1989) phenomenological approach. Like Wittgenstein, CA analyzed
language at a unit larger than the isolated word or speech act. CA often focuses on
adjacency pairs used in conversation—see (Schegloff, 2007) for a systematic
6
presentation based on 40 years of research by the CA community on adjacency-pair
structure. An adjacency pair is a sequence of two or three utterances that elicit or
respond to each other, such as a question and answer. The significance of the
adjacency pair as a unit of analysis is that it includes contributions by multiple
people involved in an interaction, and thereby avoids treating speech as an
expression of an individual mind. This is analogous to Marx’ (1867) focus on the
act of commodity exchange between people as a unit of interaction in contrast to
theories that dwell on rational decisions of an individual (Stahl, 2010c). What is
important in CA is the mode of interaction carried out by the adjacency pair
situated in its on-going, sequential discourse context. This should be contrasted
with approaches that code isolated utterances based on assumptions about mental
models inside the individual mind of the speaker. A CA analysis explicates how a
dyad or small group builds upon and solicits each other’s contributions, thus
providing insight into patterns of collaboration. In a sense, the CA unit of analysis
is not simply the adjacency pair, which includes multiple speakers, but the
linguistic community, which establishes the member methods underlying
adjacency-pair practices.
Understanding Computers and Cognition presents a Heideggerian critique of
the rationalist foundations of artificial intelligence by a leading AI researcher. The
book reviews three theories that endorse contextual analysis: Heidegger’s
(1927/1996) situated being-in-the-world, Gadamer’s (1960/1988) historically
grounded conception of interpretation and Maturana’s (1987) ecological version of
cognition. These theories emphasize the inseparability of the mind from its larger
context: human being engaged in the world, interpretation oriented within the
horizon of history and the organism bound in a structural coupling with its
environment. In contrast, AI software represents mental functions as isolatable
units of rational computation, which in principle cannot capture the richness and
complexity of situated human cognition and collaboration. The larger, primarily
tacit (Polanyi, 1966) unit of context cannot be adequately represented in a
computer system (Stahl, 2010d). Accordingly, the role of computer software should
be to support human interaction and collaboration, rather than to replace or fully
model human cognition.
The writings of Vygotsky, Lave & Wenger and Sacks further develop the
perspectives of Marx, Heidegger and Wittgenstein that view cognition as social,
situated and linguistic. Winograd—like other CSCW researchers, including Ehn
and Dourish—reviews the foundational post-cognitive theories and considers the
implications for computer-supported collaboration. But these theories can be—and
have been—taken in different directions by researchers when it comes time to
follow their implications for research conceptualizations and methods. These
directions can perhaps best be seen in terms of alternative theories of individual,
small-group and community cognition in collaboration research.

Theories of individual cognition

Many research questions within CSCL and CSCW involve individual cognition.
Collaboration research is often treated as a sub-discipline of educational or social-
psychological research, oriented to the mind of the individual student or worker,
within group contexts. Such research can follow traditional scientific research
7
paradigms based on pre-Kantian empiricism (Hume) and/or rationalism (Locke).
This research often adopts a constructivist approach, based on the Kantian principle
that the student or worker constructs his or her own understanding of reality. Such
constructivist theory is cognitivist, in that it involves assumptions about cognitive
processes in the mind of the individual underlying the individual’s observed
behaviors. For instance, a student’s responses in a test situation are assumed to be
reflective of the student’s mental models of some knowledge content, as construed
by the student.
Work within CSCL or CSCW certainly acknowledges the importance of the
larger social, historical and cultural context. However, it often treats this context as
a set of environmental variables that may influence the outcomes of individual
student or worker cognition, but are analytically separable from that cognition. In
this way, cognition is still treated as a function of an individual mind. This
approach may be called socio-cognitive. It acknowledges social influences, but tries
to isolate the individual mind as a cognitive unit of analysis by controlling for these
external influences.
Followers of Vygotsky, by contrast, are considered socio-cultural. They
recognize that cognition is mediated by cultural factors. Yet, they still generally
focus on the individual as the unit of analysis. They investigate how individual
cognition is affected by cultural mediations, such as representational artifacts or
even by collaborative interactions. Vygotsky himself—who was after all a
psychologist—generally discussed the individual subject. For instance, his concept
of the zone of proximal development measured an individual’s ability when
working in a group, not the group’s ability as such. Vygotsky was trying to
demonstrate that individual cognition was derivative of social or intersubjective
experiences of the individual, and so his focus was on the individual rather than
explicitly on the social or intersubjective processes in which the individual was
involved.
In this sense, much cognitive research investigates individual cognition in
settings of collaboration. In fact, if the research is based on testing of the individual
before and after a collaborative interaction and does not actually analyze the
intervening interaction itself, then it is purely an analysis at the individual unit of
analysis, where the collaboration is merely an external intervention measured by
presumably independent variables.
If one looks closely at most studies that claim to be about small-group
collaboration, one finds that they adopt this kind of methodical focus on the
individual within a group setting and treat the group interaction as an external
influence on the individual. This is particularly clear in the writings of cooperative
learning that preceded CSCL (e.g., Johnson & Johnson, 1989). As defined within
CSCL (Dillenbourg, 1999), in “cooperative” learning students divide up group
work and then put the individual contributions together, whereas in “collaborative”
learning students do the work together. Similarly on the methodological level, in
cooperative learning the analyst distinguishes the contributions to the work and
focuses on the learning by the individuals as a result of the cooperative experience,
whereas in collaborative learning the analyst may chose to focus on the group
processes. The same is true for small-group studies of sociology and social
psychology in CSCW: they usually treat the group as a context and analyze the
effects on the individual, rather than analyzing the group phenomena and treating
the individuals as contributors to the group processes.
8
 A final example of a theory of individual cognition is psycho-linguistic
contribution theory (Clark & Brennan, 1991). This particular paper is often cited in
CSCL and CSCW literature. Although the paper claims to be in the Conversation
Analysis tradition, it translates the adjacency-pair structure of grounding shared
understanding into the contributions of the individuals. It analyzes the individual
contributions as expressions of their mental representations or personal beliefs and
treats the resultant shared understanding as a matter of similar mental contents or
acceptance of pre-conceived beliefs rather than as a negotiated group product of
collaboratively co-constructed meaning making. In a later paper, Clark (1996) tries
to unite cognitivism with Conversation Analysis, but he analyzes the situated,
engaged interaction as an exchange of signals between rationally calculating minds,
who identify deliberate actions based on “knowledge, beliefs and suppositions they
believe they share” (p. 12). Interestingly, Clark (1996) concludes in favor of
recognizing two independent theories with different units of analysis (the
individual or the community, but ironically not the small group): “The study of
language use must be both a cognitive and a social science,” he says (p.25).

Theories of community cognition

In striking contrast to the insistent focus on the individual as the unit of analysis is
the social-science perspective on social processes. Marx provides a good example
of this. Where economists of his day analyzed economic phenomena in terms of
rational choices of individual producers and consumers, Marx critiqued the
ideology of individualism and in its place analyzed sweeping societal
transformations such as urbanization, the formation of the proletariat, the rise of the
factory system, and the drive of technological innovation. Lave and Wenger (1991)
brought this approach to educational theory, showing for instance how an
apprenticeship training system reproduces itself as novices are transformed into
experts, mentors and masters. Learning is seen as situated or embedded in this
process of the production-and-reproduction of structures of socially defined
knowledge and power. For Lave and Wenger, the community or community-of-
practice is the structure within social organizations (corporations, cultural
institutions, etc.) where interaction, task accomplishment, professional exchanges,
training and institutional learning take place; it is a prime location for CSCL-at-
Work.
The theoretical importance of the situation in which learning and work take
place is widely acknowledged in CSCL and CSCW. Suchman (1987) demonstrated
its centrality for human-computer interaction from an anthropological perspective
heavily influenced by both Heidegger (via Dreyfus, (1991) and Garfinkel (1967),
leading to conclusions similar to Winograd’s (Winograd & Flores, 1986). Suchman
(1987) and Nardi (1996) have helped to establish ethnographic methods—oriented
to community phenomena—as relevant to CSCL and CSCW research.
Unfortunately, even perspectives like situated cognition can take a reductive turn:
Recent commentaries on situated cognition (Robbins & Aydede, 2009) and
distributed cognition (Adams & Aizawa, 2008) frame the issues at the individual
level, to the extreme of reducing all cognitive phenomena to neural functions.
Building on Vygotsky and his Russian colleagues, Activity Theory
(Engeström, 1987; Engeström, 1999; Kaptelinin & Nardi, 2006) insists on taking
9
an entire activity system as the unit of analysis. In his triangular analysis rubric,
Engeström extends Vygotsky’s mediation triple of subject, mediator and object to
include mediating dimensions from Marx’s theory: the division of labor, the rules
of social relations and the community of productive forces. Like discourse analysis
(Gee, 1992), activity theory is repeatedly looking at small-group interactions but
only seeing the larger, societal issues. For instance, when activity theory addresses
the study of teams in the most detail in Chapter 6 of (Engeström, 2008), it is mostly
concerned with the group’s situation in the larger industrial and historic context;
rather than with analyzing how the group interactionally builds knowledge, it
paraphrases how the group deals politically with organizational management
issues.
There is something of this avoidance of the small group as the scientific focus
in other theories popular in CSCL and CSCW as well, for instance even in
distributed cognition. In seminal statements of post-cognitivist theory, Hutchins has
indeed explicitly pointed to group-cognitive phenomena:
• “Cognitive processes may be distributed across the members of a social group”
(Hollan, Hutchins & Kirsh, 2000, p. 176).
• “The cognitive properties of groups are produced by interaction between
structures internal to individuals and structures external to individuals”
(Hutchins, 1996, p. 262).
• “The group performing the cognitive task may have cognitive properties that
differ from the cognitive properties of any individual” (Hutchins, 1996, p. 176).
However, rather than focusing on these group phenomena in detail, he prefers to
analyze socio-technical systems and the cognitive role of highly developed artifacts
(e.g., airplane cockpits or ship navigation tools). Certainly, these artifacts have
encapsulated past cultural knowledge (community cognition), and Hutchins’
discussions of this are insightful. But in focusing on what is really the community
level—characteristically for a cultural anthropologist—he does not generally
analyze the cognitive meaning making of the group itself (but see his analysis of
group or organizational learning in Chapter 8 of Hutchins, 1996, for an exception
and an exemplary analysis of CSCL-at-Work).
Even ethnomethodology (Garfinkel, 1967; 2006) and conversation analysis
(Sacks, 1962/1995; Sacks, Schegloff & Jefferson, 1974; Schegloff, 2007) consider
themselves social sciences, versions of sociology or communication studies, but not
sciences of the small-group unit of analysis. They aim to analyze social practices,
defined across a whole society or linguistic community. This may be a quibble over
words, for they do in fact define many important processes at the group unit,
although they call them social.
Vygotsky, too, used the term social in an ambiguous way when he said that
learning takes place socially first and then later individually. Socially can refer to
two people talking as well as to transformations of whole societies. But for the sake
of distinguishing levels of description or units of analysis in CSCL and CSCW, it
seems important to make clear distinctions. Table 1 suggests sets of different terms
for referring to phenomena at the individual, small-group and societal levels. The
distinction of these three levels has previously been argued for by (Rogoff, 1995),
(Dillenbourg et al., 1996), (Stahl, 2006) and others. We start with these three levels,
which seem particularly central to much CSCL and CSCW work, although other
levels might also usefully be distinguished, such as “collective intelligence” at the
10
classroom/shop-floor level or “collective practices” at the school/company level
(Guribye, 2005; Jones, Dirckinck-Holmfeld & Lindström, 2006; Looi et al., 2011).
Perhaps consistent usage of such terminological distinctions would lend clarity to
the discussion of theories in CSCL and CSCW.
Table 1. Terminology for phenomena at the individual, small-group and community
levels of description. From (Stahl, 2010a, p. 27, Table 2.1 ).

Level of description Individual Small group Community

Role Person / student Group participant Community member
Adjective Personal Collaborative Social
Object of analysis Mind Discourse Culture
Unit of analysis Mental Utterance response pair Socio-technical activity
representation system, mediating
artifacts
Form of knowledge Subjective Intersubjective Cultural
Form of meaning Interpretation Shared understanding, Domain vocabulary,
joint meaning making, artifacts, institutions,
common ground norms, rules
Learning activity Learn Build knowledge Science
Ways to accomplish Skill, behavior Discourse, group methods, Member methods,
cognitive tasks long sequences social practices
Communication Thought Interaction Membership
Mode of construction Constructed Co-constructed Socially constructed
Context of cognitive Personal problem Joint problem space Problem domain
task
Context of activity Environment Shared space Society
Mode of Presence Embodiment Co-presence Contemporary
Referential system Associations Indexical field Cultural world
Form of existence Being-there Being-with (Mitsein), Participation in
(Heidegger) (Dasein) Being-there-together at communities of
the shared object practice (Volk)
Temporal structure Subjective Co-constructed shared Measurable objective
experiential temporality time
internal time
Theory of cognition Constructivist Post-cognitive Socio-cultural
Science Cognitive and Group cognition theory Sociology,
educational anthropology,
psychology linguistics
Tacit knowledge Background Common ground Culture
knowledge
Thought Cognition Group cognition Practices
Action Action Inter-Action Social praxis

Theories of small-group cognition

As suggested above, the CSCL and CSCW-related literature on small groups and
on post-cognitive phenomena provide some nice studies of the pivotal role of small
groups, but they rarely account for this level of description theoretically. They are
almost always in the final analysis based on either a psychological view of mental
11
processes at the individual level or a sociological view of rules at the community
level. They lack a foundational conception of small groups as a distinct level of
analysis and description. They often confuse analysis at the small-group level and
at the societal level, and they lack a developed account of the relationships among
the individual, small group and community of practice. Yet there are distinct
phenomena and processes at each of these levels, and analyses at different levels of
description reveal different insights.
It seems obvious that the small-group level should be considered particularly
central to CSCL and CSCW theory, because these fields are explicitly concerned
with supporting collaboration, knowledge co-construction or group cognition.
There are few other domains in which such activities by small groups are in
principle such a central concern. We have seen resistance to this focus on the small
group, for instance, in the case of activity theory—which could profitably be used
to investigate group processes—where Engeström (2008) argued against a focus on
small groups because workplace teams tend to come and go quickly, forming
changing knots of co-workers around ephemeral tasks.
Engeström’s argument echoes the attitude of Schmidt & Bannon (1992) in
their programmatic opening article of the inaugural issue of the CSCW journal. In
rejecting the use of the term “group” as a defining concept for CSCW, they reduced
the theoretical perspective to one focused on individuals “articulating” (i.e.,
coordinating) their “distributed individual activities” (p. 15). They made this move
despite claiming that their concept of “cooperative work” was congruent with
Marx’ (1867) definition of cooperative work as “multiple individuals working
together in a conscious way in the same production process.” In 1867, Marx was
analyzing in detail the historic shift of the unit of production from the individual to
the group, but in 1992 Schmidt & Bannon insist on still focusing on the individual.
They complain that the units of cooperative workers are not well-formed, clearly
defined, persisting groups. But that is beside the point.
The theoretical point is that interacting people accomplish work tasks and
associated cognitive tasks (including articulation tasks and power struggles)
through group interaction processes, and that these should be analyzed as such, as
local achievements of group interaction, not simply as sums of individual actions
and reactions or as effects of external societal forces. In particular, as cooperative
work shifts from the manual factory production of Marx’s time to knowledge
building and other forms of intellectual production in the information age, group-
cognition phenomena call more strongly for analysis at the small-group unit. A
small group is not defined ontologically as a certain number of human bodies
adjacent to each other for a certain period of clock time, but as a cognitive unit
capable of achieving specific tasks of cooperative work and collaborative
knowledge building through the interaction among individuals within a larger
community context.
There are distinct phenomena and processes at the individual, small-group and
community-of-practice levels, and analyses at these different levels of description
can reveal different insights. As Grudin (1994) put it in terms of the needs of
CSCW,
Computer support has focused on organizations and individuals. Groups
are different. Repeated, expensive groupware failures result from not
meeting the challenges in design and evaluation that arise from these
differences. (p. 93)
12
There are theoretical, methodological and practical reasons for both CSCL and
CSCW to focus on the small-group unit of analysis.
If group phenomena are treated seriously as first-class objects of theory, then
one can study how small groups engage in cognitive activities such as:
interpersonal trains of thought, shared understandings of diagrams, joint problem
conceptualizations, common references, coordination of problem-solving efforts,
planning, deducing, designing, describing, problem solving, explaining, defining,
generalizing, representing, remembering and reflecting as a group. In CSCL and
CSCW studies of text chat or discussion forums, for instance, analysis can show
group-cognitive accomplishments emerging from the network of meaningful
references built up by postings, demonstrating how the group’s self-formation and
its cognitive accomplishments are enacted in situated interaction. An analytic focus
on the group unit of analysis need not imply that groups exist as ontological entities
whenever people are observed in proximity or in communication with one another.
The small group is a theoretical construct, not a simple physical observable. Of
course, effective groups have to constitute themselves as such and they can change
dramatically over time. It is not the more-or-less persistent physical group that is
important, but the group processes, which may extend over seconds, days or years.
A single momentary exchange of greetings may be a group process of interest, as
shown by the early conversation analyses of telephone answering on a help phone
line (Hopper, 1992).
A theoretical approach that focuses on small-group interaction is that of
dialogicality (Linell, 2001; 2009; Mercer, 2000; Wegerif, 2007). Dialogical theory
goes back to Bakhtin (1986), a contemporary of Vygotsky. It stresses the linguistic
nature of interaction. It also reiterates the idea that a person’s identity as an
individual arises through the confrontation with one’s partners in dialogue—a view
that goes back beyond Mead (1934/1962) to Hegel’s (1807/1967) master-slave
dialectic (Stahl, 2006, p. 333f). The notion of dialogue partners coming from
different perspectives and negotiating from these is an important contribution of
dialogic inquiry (Wells, 1999). Another key concept is that of a shared dynamic
dialogic space, within which knowledge building can take place (Kershner et al.,
2010). This is similar to the joint problem space of (Teasley & Roschelle, 1993),
but now developed in an unambiguously post-cognitive manner.
The idea of an interactional space for interaction within a small group is
central to group-cognition theory (Stahl, 2006) as well. The term group cognition
was coined to stress the goal of developing a post-cognitive view of cognition as
the possible achievement of a small group collaborating so tightly that the process
of building knowledge in the group discourse cannot be attributed to any individual
or even reduced to a sequence of contributions from individual minds. For instance,
the knowledge might emerge through the interaction of linguistic elements, situated
within a sequentially unfolding set of constraints defined by the group task, the
membership of the group, and other local or cultural influences, as well as due to
the mediation of representational artifacts and media used by the group—within a
larger horizon of language and history (Gadamer, 1960; Husserl, 1936).
The theory of group cognition absorbs many ideas from the theories discussed
above, including that of a shared dynamic dialogical space. Despite some scattered
case studies by the authors already mentioned and their colleagues, there is yet not
much documentation and analysis of empirical instances of effective group
cognition. The analysis of group cognition needs not only specially focused
13
methods to track its occurrence, but even prior to that it needs appropriate
collaboration technologies, group methods, pedagogy and guidance to structure and
support groups to effectively build knowledge that can be shown to be a group
product not reducible to individual mental representations. The Virtual Math Teams
Project was launched to generate a data corpus that would allow for the analysis of
group cognition. This project and some analyses by a number of researchers are
documented in (Stahl, 2009). Group-cognition theory focuses on the sequential
team interaction within case studies of small-group collaboration. This takes place
within an interaction space or a world in the Heideggerian (1927) sense, which
opens up to allow the production of group-cognitive accomplishments. The
interaction that takes place within such a world—whether face-to-face or online—is
subject to a variety of constraints, as pictured in Figure 2.

Figure 2. A diagram of constraints on sequential small-group interaction. From

(Stahl, 2010b, p. 256, Figure 1).

Note that Figure 2 is not intended to be a model of objects and processes.

Rather it tries to present some of the complex constraints on the discourse through
which group cognition might be achieved. Neither the physical individuals nor their
group are represented here as such; the dialogical voices (Bakhtin, 1986) of the
individuals enter into the sequential team interaction and respond to it. Over time,
the sequential team interaction forms the central shared dynamic dialogic space
within which the group-cognitive constraints interact. Behind the individual voices
that enter into this interaction space are not so much minds containing mental
representations, as a fluid background of past experiences and developed resources
for action, which surface based on relevance to the interaction. The team discourse
14
is situated in the shared dialogical context generated by the on-going interaction
itself; the culture and history associated with the group’s community-of-practice;
and the socio-technical environment including the media of communication. The
interaction is goal-oriented toward the task—as given externally but as enacted by
the group—and mediated by a variety of kinds of artifacts, including codifications
of knowledge products previously generated by the group. These artifacts might
end up among the team outcomes, in relation to the guiding task. Of course, other
constraints and influences are possible as well, coming for instance from the
guidance of a teacher or the motivations of a reward system. The point is that one
can picture the whole system producing cognitive accomplishments without having
to postulate mental representations in individual minds, let alone to reduce the
whole system either to rational mental decisions or to regulation by rules of social
institutions.
The term constraint in Figure 2 is chosen to be a neutral term, not implicating
a notion of mechanistic causality. While it is clear that the traditional conception of
causality is inadequate—stemming back to Aristotle and metaphors of physical
mechanics from the everyday world—it is less obvious how to think about the
working of the constraints upon group cognition. Folk theory adopts a mechanistic
worldview, or even an anthropomorphic view of nature combined with a
mechanistic view of causality. Observable behavior of people is taken to be the
result of rational decision making in the heads of individuals causing the people to
behave as a result of the minds acting as the agency for causing words to be
produced and limbs to be moved. But the linguistic turn of Wittgenstein
(Wittgenstein, 1953) and even more so the recent practice turn (Schatzki, Knorr
Cetina & Savigny, 2001) have veered radically away from such a view.
Latour (1992) seems to be working toward a post-cognitive notion of causality,
perhaps relying heavily on Hegel’s notion of mediation. Interestingly, he not only
argues against the hegemony of individual minds as agents in the social world, but
he also argues against the adequacy of our notion of the social (Latour, 2007).
History is made neither by rational decisions of individual minds nor by the
workings of society. Rather, it is the result of a complex network of mediating
actors, including all kinds of artifacts as well as human actors. Thus, Latour seems
to be advocating an analytic approach that steers clear of both individual minds and
social institutions to focus on a middle ground. He selects the term ‘group’ for this
middle ground, precisely because it can be used without implying theoretical
preconceptions: “The word ‘group’ is so empty that it sets neither the size nor the
content…. This is exactly why I have chosen it” (Latour, 2007, p. 29). Figure 2
may illustrate the kind of network that he would endorse for picking apart and then
reassembling instances of group cognition.
Such new conceptualizations of cognition, agency and causality may be
particularly appropriate for collaborative learning and cooperative work, especially
as they are brought together in CSCL-at-Work. Here, the focus is on interpersonal
communication and work practices of groups. CSCL-at-Work should adopt these
new perspectives for facilitating computer-mediated discourse focused on
improving small-group practices.

15
A multiplicity of theories of cognition

In general, CSCL, CSCW and CSCL-at-Work raise many fundamental questions

for traditional theories, oriented as they are to small groups and to online
interaction. The accustomed characteristics of the physical world, in which
colleagues and interlocutors are embodied and visible to each other, are often
missing in these virtual settings, and that brings into question numerous
assumptions of folk theories and traditional approaches. The group itself has no
identity as a physical body and has no brain to possess its knowledge; it relies on
external memories, which differ essentially from personal memories (Donald,
1991). The online world—shared dialogical space—has no location or extension.
Group members can come from around the world and do not necessarily share local
connections and culture. CSCL and CSCW involve students and workers in
qualitatively different social relations of production, modes of being in the world or
forms of life; even Marx, Heidegger and Wittgenstein’s foundational philosophies
of post-cognitive theory need to be rethought for virtual groups. Concepts of
causality, world, knowledge, cognition, intersubjectivity, interaction and presence
need to be reconceptualized for theories of collaboration.
There are many avenues for developing theories of CSCL-at-Work, as
reviewed in this chapter. Although there are some similarities among these
alternatives—often in terms of their critiques of earlier theories—there are strong
differences of position and perspective. This is not necessarily a problem. There is
a huge assortment of processes taking place in successful collaborative events: at
multiple time scales and involving different aspects of interaction. It is possible to
raise innumerable research questions, each requiring possibly different methods of
investigation at various levels of analysis. It is likely that CSCL-at-Work requires
multiple theories, which are not reducible to one grand unifying theory and that
even seem incommensurate with each other. This goes essentially beyond the
common notion of mixed methods, in which two or more methods of analysis are
used to triangulate a single phenomenon from different angles. There are distinct
phenomena at different levels of description—and they interact with each other in
complex ways in group settings.
CSCL and CSCW study collaboration, from a design perspective. CSCL often
involves whole classrooms or schools and widespread educational practices;
CSCW often involves large departments or factories and widespread work
practices. At the opposite end of the spectrum, much of the actual work comes
down to tasks done by individuals. But much of the coordination, decision making,
articulation, brainstorming, discovery and knowledge building is accomplished by
small groups. Community accomplishments are thereby mediated by small groups,
which carry out the necessary activities and involve the individuals. Collaboration
involves a tight and complex integration of processes at the individual, small-group
and community levels. Computer support for collaboration must provide supports
at each level while also supporting the integration of the activities at all levels. To
provide insight for this, CSCL and CSCW research must recognize the levels as
distinct and conduct analyses at all levels.
Some time ago when I was a CSCL researcher working in a CSCW research
lab, I argued that CSCL and CSCW were closely related in terms of their
theoretical foundations, but that they differed in terms of the population on which
they focused (Stahl, 2002). CSCL is concerned with students who are learning
16
practices at which they are still novices. In the Vygotskian metaphor, they are
experiencing group-cognitive processes that they can subsequently internalize.
CSCW is concerned with professionals who are refining their skills as experts.
They may be engaged in group-cognitive processes which cannot be internalized by
an individual, such as navigating a large naval vessel (Hutchins, 1996). The theme
of this book, learning at work, demonstrates that even at the level of concrete
learning activities the two often separated research fields have much in common.
In CSCL-at-Work, there are many phenomena of interest, and they are largely
defined by the theories that conceptualize them. So different theories can be talking
about quite different phenomena (although they may unfortunately be calling them
by the same name). In order to avoid confusion and arguments about pseudo-
problems, we need to be clear about the theories behind research questions,
assumptions, methodologies, analysis tools, findings and claims in this research.
This chapter has sketched some of the theoretical landscape underlying CSCL
and CSCW research. Progress in further developing theories of CSCL-at-Work will
require careful analysis of case studies—such as those in this volume—and
experimental results guided by theoretical perspectives that are clearly enunciated.

References

Adams, F., & Aizawa, K. (2008). The bounds of cognition. Malden, MA:
Blackwell.
Bakhtin, M. (1986). Speech genres and other late essays (V. McGee, Trans.).
Austin, TX: University of Texas Press.
Carroll, J. (Ed.). (2003). HCI models, theories and frameworks: Toward a
multidisciplinary science. San Francisco, CA: Morgan Kaufmann
Publishers.
Chomsky, N. (1959). Review of verbal behavior, by B. F. Skinner. Language.
35(1), 26-57.
Chomsky, N. (1969). Aspects of a theory of syntax. Cambridge, MA: MIT Press.
Clark, H. (1996). Using language. Cambridge, UK: Cambridge University Press.
Clark, H., & Brennan, S. (1991). Grounding in communication. In L. Resnick, J.
Levine & S. Teasley (Eds.), Perspectives on socially-shared cognition. (pp.
127-149). Washington, DC: APA.
Descartes, R. (1633/1999). Discourse on method and meditations on first
philosophy. New York, NY: Hackett.
Dillenbourg, P. (1999). What do you mean by "collaborative learning"? In P.
Dillenbourg (Ed.), Collaborative learning: Cognitive and computational
approaches. (pp. 1-16). Amsterdam, NL: Pergamon, Elsevier Science.
Dillenbourg, P., Baker, M., Blaye, A., & O'Malley, C. (1996). The evolution of
research on collaborative learning. In P. Reimann & H. Spada (Eds.),
Learning in humans and machines: Towards an interdisciplinary learning
science. (pp. 189-211). Oxford, UK: Elsevier.
Donald, M. (1991). Origins of the modern mind: Three stages in the evolution of
culture and cognition. Cambridge, MA: Harvard University Press.
Dourish, P. (2001). Where the action is: The foundations of embodied interaction.
Cambridge, MA: MIT Press.

17
Dreyfus, H. (1991). Being-in-the-world: A commentary on Heidegger's Being and
time, division I. Cambridge, MA: MIT Press.
Ehn, P. (1988). Work-oriented design of computer artifacts. Stockholm, Sweden:
Arbetslivscentrum.
Engeström, Y. (1987). Learning by expanding: An activity-theoretical approach to
developmental research. Helsinki, Finland: Orienta-Kosultit Oy.
Engeström, Y. (1999). Activity theory and individual and social transformation. In
Y. Engeström, R. Miettinen & R.-L. Punamäki (Eds.), Perspectives on
activity theory. (pp. 19-38). Cambridge, UK: Cambridge University Press.
Engeström, Y. (2008). From teams to knots. Cambridge, UK: Cambridge
University Press.
Floyd, C. (1992). Software development and reality construction. In C. Floyd, H.
Zuellinghoven, R. Budde & R. Keil-Slawik (Eds.), Software development
and reality construction. (pp. 86-100). Berlin, Germany: Springer Verlag.
Gadamer, H.-G. (1960/1988). Truth and method. New York, NY: Crossroads.
Garfinkel, H. (1967). Studies in ethnomethodology. Englewood Cliffs, NJ:
Prentice-Hall.
Garfinkel, H. (2006). Seeing sociologically: The routine grounds of social action.
Boulder, CO: Paradigm Publishers.
Gee, J. P. (1992). The social mind: Language, ideology, and social practice. New
York, NY: Bergin & Garvey.
Guribye, F. (2005). Infrastructures for learning: Ethnographic inquiries into the
social and technical conditions of education and training. Unpublished
Dissertation, Ph.D., Department of Information Science and Media Studies,
University of Bergen. Bergen, Norway.
Hegel, G. W. F. (1807/1967). Phenomenology of spirit (J. B. Baillie, Trans.). New
York, NY: Harper & Row.
Heidegger, M. (1927/1996). Being and time: A translation of Sein und Zeit (J.
Stambaugh, Trans.). Albany, NY: SUNY Press.
Hollan, J., Hutchins, E., & Kirsh, D. (2000). Distributed cognition: Toward a new
foundation of human-computer interaction research. ACM Transactions on
Computer-Human Interaction. 7(2), 174-196.
Husserl, E. (1936/1989). The origin of geometry (D. Carr, Trans.). In J. Derrida
(Ed.), Edmund Husserl's origin of geometry: An introduction. (pp. 157-
180). Lincoln, NE: University of Nebraska Press.
Hutchins, E. (1996). Cognition in the wild. Cambridge, MA: MIT Press.
Johnson, D. W., & Johnson, R. T. (1989). Cooperation and competition: Theory
and research. Edina, MN: Interaction Book Company.
Jones, C., Dirckinck-Holmfeld, L., & Lindström, B. (2006). A relational, indirect,
meso-level approach to CSCL design in the next decade. International
Journal of Computer-Supported Collaborative Learning. 1(1), 35-56.
Kant, I. (1787/1999). Critique of pure reason. Cambridge, UK: Cambridge
University Press.
Kaptelinin, V., & Nardi, B. A. (2006). Acting with technology: Activity theory and
interaction design. Cambridge, Mass.: MIT Press. 333 p. pages.
Kershner, R., Mercer, N., Warwick, P., & Staarman, J. K. (2010). Can the
interactive whiteboard support young children’s collaborative
communication and thinking in classroom science activities? International
Journal of Computer-Supported Collaborative Learning. 5(4). 359-383.

18
Latour, B. (1992). Where are the missing masses? The sociology of a few mundane
artifacts. In W. E. Bijker & J. Law (Eds.), Shaping technology/building
society. (pp. 225-227). Cambridge, MA: MIT Press.
Latour, B. (2007). Reassembling the social: An introduction to actor-network-
theory. Cambridge, UK: Cambridge University Press.
Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral
participation. Cambridge, UK: Cambridge University Press.
Linell, P. (2001). Approaching dialogue: Talk, interaction and contexts in
dialogical perspectives. New York, NY: Benjamins.
Linell, P. (2009). Rethinking language, mind, and world dialogically: Interactional
and contextual theories of human sense-making. Charlotte, NC: Information
Age Publishing.
Looi, C.-K., So, H.-j., Toh, Y., & Chen, W. (2011). CSCL in classrooms: The
Singapore experience of synergizing policy, practice and research.
International Journal of Computer-Supported Collaborative Learning.
Marx , K. (1867). Das kapital: Kritik der politischen oekonomie (Vol. I). Hamburg,
Germany: Otto Meisner.
Maturana, H. R., & Varela, F. J. (1987). The tree of knowledge: The biological
roots of human understanding. Boston, MA: Shambhala.
Mead, G. H. (1934/1962). Mind, self and society. Chicago, IL: University of
Chicago Press.
Mercer, N. (2000). Words and minds. How we use language to think together:
Routledge.
Nardi, B. (Ed.). (1996). Context and consciousness: Activity theory and human-
computer interaction. Cambridge, MA: MIT Press.
Plato. (340 BC/1941). The republic (F. Cornford, Trans.). London, UK: Oxford
University Press.
Polanyi, M. (1966). The tacit dimension. Garden City, NY: Doubleday.
Robbins, P., & Aydede, M. (Eds.). (2009). The Cambridge handbook of situated
cognition. Cambridge, UK: Cambridge University Press.
Rogoff, B. (1995). Sociocultural activity on three planes. In B. Rogoff, J. Wertsch,
P. del Rio & A. Alvarez (Eds.), Sociocultural studies of mind. (pp. 139-
164). Cambridge, UK: Cambridge University Press
Sacks, H. (1962/1995). Lectures on conversation. Oxford, UK: Blackwell.
Sacks, H., Schegloff, E. A., & Jefferson, G. (1974). A simplest systematics for the
organization of turn-taking for conversation. Language. 50(4), 696-735.
Scardamalia, M., & Bereiter, C. (1996). Computer support for knowledge-building
communities. In T. Koschmann (Ed.), CSCL: Theory and practice of an
emerging paradigm. (pp. 249-268). Hillsdale, NJ: Lawrence Erlbaum
Associates.
Schatzki, T. R., Knorr Cetina, K., & Savigny, E. v. (Eds.). (2001). The practice
turn in contemporary theory. New York, NY: Routledge.
Schegloff, E. A. (2007). Sequence organization in interaction: A primer in
conversation analysis. Cambridge, UK: Cambridge University Press.
Schön, D. A. (1983). The reflective practitioner: How professionals think in action.
New York, NY: Basic Books.
Sfard, A. (2008). Thinking as communicating: Human development, the growth of
discourses and mathematizing. Cambridge, UK: Cambridge University
Press.

19
Shannon, C., & Weaver, W. (1949). The mathematical theory of communication.
Chicago, Il: University of Illinois Press.
Stahl, G. (2002). Groupware goes to school. In J. H. J. Pino (Ed.), Groupware:
Design, implementation and use: Proceedings of the 8th international
workshop on groupware (CRIWG '02). (pp. 7-24). La Serena, Chile:
Springer Lecture Notes in Computer Science 2440. Web:
http://GerryStahl.net/cscl/papers/ch11.pdf.
Stahl, G. (2006). Group cognition: Computer support for building collaborative
knowledge. Cambridge, MA: MIT Press.
Stahl, G. (2009). Studying virtual math teams. New York, NY: Springer.
Stahl, G. (2010a). Group cognition as a foundation for the new science of learning.
In M. S. Khine & I. M. Saleh (Eds.), New science of learning: Cognition,
computers and collaboration in education. (pp. 23-44). New York, NY:
Springer. Web: http://GerryStahl.net/pub/scienceoflearning.pdf.
Stahl, G. (2010b). Guiding group cognition in CSCL. International Journal of
Computer-Supported Collaborative Learning. 5(3), 255-258.
Stahl, G. (2010c). Marx and Heidegger. Philadelphia. PA: Gerry Stahl at Lulu. 217
pages. Web: http://GerryStahl.net/elibrary/marx.
Stahl, G. (2010d). Tacit and explicit understanding. Philadelphia, PA: Gerry Stahl
at Lulu. 374 + xii pages. Web: http://GerryStahl.net/elibrary/tacit.
Stahl, G., Koschmann, T., & Suthers, D. (2006). Computer-supported collaborative
learning: An historical perspective. In R. K. Sawyer (Ed.), Cambridge
handbook of the learning sciences. (pp. 409-426). Cambridge, UK:
Cambridge University Press. Web: http://GerryStahl.net/elibrary/global.
Suchman, L. (1987). Plans and situated actions: The problem of human-machine
communication. Cambridge, UK: Cambridge University Press.
Teasley, S. D., & Roschelle, J. (1993). Constructing a joint problem space: The
computer as a tool for sharing knowledge. In S. P. Lajoie & S. J. Derry
(Eds.), Computers as cognitive tools. (pp. 229-258). Mahwah, NJ:
Lawrence Erlbaum Associates, Inc.
Turing, A. M. (1937). On computable numbers, with an application to the
entscheidungsproblem. Proceedings of the London Mathematical Society.
2(1), 230.
Vygotsky, L. (1930/1978). Mind in society. Cambridge, MA: Harvard University
Press.
Wegerif, R. (2007). Dialogic, education and technology: Expanding the space of
learning. New York, NY: Kluwer-Springer.
Wells, G. (1999). Dialogic inquiry: Towards a socio-cultural practice and theory
of education. Cambridge, UK: Cambridge University Press.
Winograd, T., & Flores, F. (1986). Understanding computers and cognition: A new
foundation of design. Reading, MA: Addison-Wesley.
Wittgenstein, L. (1921/1974). Tractatus logico philosophicus. London, UK:
Routledge.
Wittgenstein, L. (1953). Philosophical investigations. New York, NY: Macmillan.

20