Context and Consciousness Activity Theory and Human-Computer Interaction

Review by Mary Brenner

SIGCHI Bulletin 29, Number 2 April 1997 49

This collection of articles, edited by Bonnie Nardi, provides a coherent and clear theoretical alternative to the dominant cognitive science research tradition in Human-Computer Interaction. Activity theory, as developed by the Soviet psychologists Vygotsky, Luria and Leontev and as extended more recently in Europe, is proposed as an alternative approach that overcomes the increasingly obvious limitations of a purely cognitive approach. The set of papers provides a nice balance between presentations of basic activity theory concepts, examples of applications within specific research projects and extensions of activity theory. The many examples of people, both computer users and application developers, engaged in computer-mediated activities demonstrate the promise of activity theory to elucidate new dimensions of technology use without hiding the current limitations of activity theory as both theory and method. Overall, the book will be useful to those who are learning about activity theory for the first time as well as those who are acquainted with the basics but want to know where activity theory can take us in the future.

The book is divided into three sections: Activity Theory Basics, Activity Theory in Practical Design, and Activity Theory: Theoretical Development. Nardi as editor provides both useful summary introductions to each section as well as beginning and ending essays that contextualize activity theory within the broader field of HCI research. Although Nardi's other contributions to the book demonstrate that she is an ardent proponent of activity theory, her introduction and conclusion provide a balanced view of the promise of activity theory. In Chapter 1 Nardi notes that activity theory's greatest contribution might be in its ability to provide disparate approaches to HCI with a common vocabulary for emergent issues in the natural study of technology usage. She also describes three issues directly addressed by activity theory-consciousness, the relationship between people and things, and the role of artifacts in everyday life  that have also been of concern in the field of HCI.

The other chapters in the first section clearly describe some of the basic concepts of activity theory and contrast them with other theoretical approaches to the same ideas. Kuutti in Chapter 2 provides a brief historical overview of the rising discontent with the cognitive science foundations of HCI research. He then introduces activity as the complex unit of analysis that enables activity theory to capture the dynamics of everyday practices. In what I consider the most useful part of this chapter, Kuutti describes the multilevel aspect of activity theory as embodied in the relations between operations, actions and activities. The chapter concludes with a discussion of how using a multilevel theory can enrich research on HCI.

Kaptelinin in Chapter 3 explores the issue of how computers are used as tools to mediate human relations with the world. This shifts the focus of research from the relationship of the user and the computer to a focus on how an integrated unit of computer/user participates in activities in the world. In particular Kaptelinin elaborates upon the concept of functional organ. A functional organ exists when resources internal and external to the human actor are combined to achieve a goal. Although Kaptelinin gives more complex examples, I think of my relationship with my email program as being that of a functional organ. Together we can communicate over the internet with a variety of people for a variety of purposes. I don't focus on the email program at all any more; it is more like an extension of my own capacity to communicate in writing. Kaptelinin explores the utility of this concept for HCI research and compares it to other approaches characterizing tools such as GOMS and Norman's work on cognitive artifacts.

Nardi's second contribution to this section of the book compares three different approaches to studying human actions in context: activity theory, situated action models and distributed cognition models. The first part of this chapter is a careful and clear delineation of the distinct characteristics of each of these approaches which are often carelessly lumped together when referring to research on everyday activities. The last part of this chapter is a somewhat partisan argument for activity theory that trivializes situated action models and virtually ignores distributed cognition models. The beginning part of the chapter is extremely illuminating, especially for those new to this kind of research, but the final part of the chapter casts an a shadow over the study of cognition in context.

Chapter 5 by Kaptelinin summarizes the value of activity theory for contextual studies of HCI, contrasting activity theory to cognitive theory. Although this chapter could stand alone as an argument for the benefits of activity theory, it also ties together the ideas that have been introduced in the first four chapters of the book.

The second part of this book examines practical applications of activity theory. In addition to giving research examples demonstrating the richness of activity theory for understanding different issues in the design and use of technology, it provides some clear suggestions for research methodology. As such, this section will interest both those who want to apply activity theory as well as those who more specifically want to study technology use in everyday contexts without necessarily buying into the framework provided by activity theory.

In Chapter 6 Bellamy discusses the development of educational software, taking into account how technology needs to act as a mediator of learning and change for an educational community that includes teachers and administrators as well as students. The chapter describes how Dinosaur Canyon was used by students in middle school to learn earth science and how Media Fusion, a program for video communication, was used to explore the topic of global warming. After describing the research on this software, Bellamy delineates three principles for designing educational environments conducive to educational change.

In Chapters 7 and 9, Bødker and Raeithel and Velichkovsky provide detailed suggestions for carrying out research. Bødker demonstrates the use of interaction analysis with videotapes using concepts from activity theory. Her examples are from a project in which she studies how employees of the Danish National labor Inspection Service use computer technology in their jobs. Raeithel and Velichkovsky suggest a number of research techniques that can be used to describe how groups of people jointly make meaning during interaction. Their detailed examples cover use of eyetracking methodology to explore noviceexpert collaboration in solving a puzzle and the use of the Repertory Grid Technique to interview people about their use of technology in communication. The chapter concludes with suggestions of how different research techniques can be used for the different process levels (activity, action, operation) of activity theory.

Christiansen uses more traditional ethnographic research methods in Chapter 8 to explore how a computer successfully becomes a mediating tool for a preexisting activity. She gives three case studies of a department of the Danish National Police. She advocates the use of activity theory as a prism for exploring multiple facets of everyday activity. Similarly, Nardi in Chapter 10 demonstrates how activity theory would have simplified the analysis from an interview study of how people use slide making software. Taken together, these two papers show how activity theory can facilitate generalizations about software use and design from a data base of very particular case studies.

The three chapters in the final section of the book are the widest ranging and thus the most difficult to briefly describe and evaluate. Each chapter introduces or develops new concepts that go beyond the basics of activity theory as introduced in the two earlier sections of the book. Holland and Reeves elaborate upon the concept of perspective as a way to link different activity systems to one another and to other aspects of social structure. Their ethnographic study traced the development of perspective among three different programming teams during a one semester course. Zinchenko's contribution deals with mediation between the internal and the external. He develops a model of spiritual development which draws from psychoanalysis, literature, mythology, history and culture. This essay is certainly the most complex and inscrutable in the book, but also most closely linked to the writings of the Soviet fathers of activity theory who also drew from a wide variety of sources in their work. As Nardi notes in her introduction to this section, Zinchenko's voice serves to push activity theory beyond reductionist schemes that are often proposed to explain the relation of humans consciousness to the external world.

In the final substantive chapter of the book, Engestrom and Escalante combine activity theory and network analysis to analyze the rise and fall of the Postal Buddy. The Postal Buddy was technological innovation that used computer kiosks to mechanize certain postal functions such as address changes and label production. As Engestrom and Escalante note, the Postal Buddy "was a substitute object of affection" (p. 366).

Just as the Postal Buddy was designed to be engaging, this chapter is particularly engaging as well. The activity systems come to life in this chapter and the utility of network analysis is well supported.

In summary, as many of the authors in this book note, the information processing models of cognitive science have been too restricted in their conceptualization of the unit of analysis for HCI research. The focus has traditionally been upon the computeruser dyed or the computerprogrammer dyed. In addition, the research methods have been based upon experimental models borrowed from the natural sciences. The result has been a research base with little utility for developers because it has excluded users engaged in real activities. According to the authors in this collection, the social sciences that have traditionally focused upon natural settings such as anthropology and sociology offer a conceptual framework that is too global to provide guidance to applied studies. In contrast, activity theory offers a unified framework for looking at natural behavior that can be explored using the methodology from a variety of disciplines. The empirical articles in this book substantiate this claim. Since the authors themselves are grounded in a wide variety of traditions including computer science, cognitive science, anthropology, psychology and communication, the image presented of activity theory is clearly a pragmatic one as opposed to the more doctrinaire image created when one reads the original theorists. As such I felt the book came across as a pleasant invitation to try activity theory on for size with ample room left for individual adaptation and growth.

About the Reviewer

Mary E. Brenner is a cognitive anthropologist who has used activity theory in both her research and in her classes. She is currently conducting research on after-school computer clubs as sites for learning math and science concepts in a structured play setting.

Review written by Mary E. Brenner. Assistant Professor, Graduate School of Education, Uniuersity of California, Santa Barbara.



back to ... Bonnie Nardi, Context and...