Special Session Summary Computers As Social Actors

John Deighton, Harvard University
[ to cite ]:
John Deighton (1999) ,"Special Session Summary Computers As Social Actors", in NA - Advances in Consumer Research Volume 26, eds. Eric J. Arnould and Linda M. Scott, Provo, UT : Association for Consumer Research, Pages: 392-393.

Advances in Consumer Research Volume 26, 1999      Pages 392-393

SPECIAL SESSION SUMMARY

COMPUTERS AS SOCIAL ACTORS

John Deighton, Harvard University

Are computers merely the medium in electronic commerce? Might they be better construed as actors? If they are so construed, to what extent do the regularities found in social psychological research apply? Do people perhaps treat consumers as people? Would you buy a used car from one computer but not another running the same software just because the first had previously been more polite to you? Two extensive programs of research, one at Stanford University and the other at MIT, find remarkable similarities between interpersonal conduct and person-consumer conduct. This session’s goal was to introduce these two provocative research programs to ACR members.

The Stanford lab, whose members have included Clifford Nass, Byron Reeves and Youngme Moon, has amassed a large body of research to support the contention that people treat media, particularly computers and television, as they do people, according them politeness, deference, the right to reciprocity in intimate disclosure, and so on. Some of the research is reported in a recent book, The Media Equation (Reeves & Nass, 1996).

Within the MIT Media Lab, a stream of research led by Rosalind Picard investigates how to endow computers with the ability to recognize emotion in people, and to express emotion in return. Again, the research is driven by the question of whether inter-human principles apply to human-computer interaction. For example, do people respond o a machine’s customer relations skills?

As the World Wide Web has blossomed in recent years, research on consumer response to web sites, to electronic commerce offerings, and to other phenomena of the online environment has grown commensurately and has become a significant focus of effort, particularly by doctoral students. However little of the research being conducted in marketing and consumer behavior departments has the history and programmatic character of the work being done in these two labs. The work from these two mature programs is instructive as much for its imaginative use of experimental methodology and their guiding theories as for the substantive findings. It is presented as an example of what can be done to study nascent communication technologies rigorously and empirically.

 

"FROM SOCIAL ATTITUDES TO SOCIAL BEHAVIORS: THE EVOLUTION OF THE #COMPUTERS AS SOCIAL ACTORS’ PARADIGM"

Clifford Nass, Stanford University

The book *The Media Equation* demonstrated that media equals real life, that is, that people apply the same social rules and responses toward computers and other advanced media that they apply to other people. The basic paradigm outlined in the book was: 1) pick a social science paper concerning individual’s attitudes toward other people; 2) in the theory section, change the words "other people" to "computers or other media"; 3) in the methods section, change the human object of attribution to a person, that is, run the same experiment described in the paper using a computer or other medium instead of a person; 4) demonstrate that individuals use the same rules for assessing the computer as they do for assessing other people; and 5) draw implications for design.

Using this paradigm, The Media Equation demonstrated that people are polite to computers, respond to flattery from a computer as they do to human flattery, react to text-based computer personalities as they do to other people, respond to computers as teammates, gender-stereotype computers, etc.

The presentation extended our research to human behaviors as well as attitudes. Thus, it discussed whether people are more helpful to a computer that provides a helpful web search as compared to a different computer, whether a Web TV labeled as a television leads to less willingness to order stock as compared to when the Web TV is labeled as a PC, whether consistency between character personality and personality of language lead to greater trust, and whether voice as compared to text input leads to different levels of socially desirable responses. For each study, both the theoretical and design implications were discussed.

 

"COMPUTERS WITH PERSONALITY"

Youngme Moon, Harvard University

On the surface, the fact that people respond socially to technologies is not altogether surprising. After all, almost all of us can think of situations in which we have muttered to our computers, cursed at our automobiles, or "talked back" to the television set. However, the conventional response to such social responses has been one of dismissal; indeed, the tendency to respond socially to technology is typically viewed as an aberration.

The dismissals generally fall into two categories. On the one hand are the scholars who argue that social responses to technologies are a function of deficiency, such as youth, lack of knowledge about technology, or psychological or social dysfunction (Barley, 1988; Turkle, 1984; Winograd & Flores, 1987; Zuboff, 1988). According to this view, "normal," well-adjusted individuals are unlikely to engage in social behavior toward machines. Others argue that this social behavior is, in fact, directed toward the human creator "behind" the machine. In other words, because the machine is logically perceived y the user to be a human artifact (Dennett, 1987; Heidegger, 1977), the "rational" response is to adopt an "intentional stance" (Dennett, 1987) toward the technology since it is simply a proxy for the creator or programmer (Searle, 1981).

What both of these explanations have in common is the assumption that individuals’ social responses to technologies are consistent with their beliefs about technology (see Nass, Moon, Morkes, Kim & Fogg, 1997). In other words, in the first case, behavior is presumably based upon a mistaken belief about the nature of machines; in the second case, behavior is presumably based upon a rational decision to orient toward the human creator behind the machine.

Neither explanation is sufficient to account for the extent to which people respond socially to computer technologies. Several studies were presented, all of which provide evidence that individuals engage in social behavior toward machines even though: (1) they are not deficient in any way; and (2) they do not perceive themselves as having a relationship with the machine’s creator. More specifically, the studies demonstrated that people respond to computer "personalities" in the same way that they respond to human personalities. Theoretical and practical implications of these studies were discussed.

 

"COMPUTERS THAT RECOGNIZE AND RESPOND TO EMOTION"

Rosalind W. Picard, Massachusetts Institute of Technology

New research in "affective computing" aims to give computers a variety of emotional skills, such as the ability to recognize, express, and in some cases "have" emotions. The emphasis is on developing the skills of emotional intelligence, which several scientists have argued are more important than traditional mathematical and verbal forms of intelligence for success in life. This presentation described a subset of our work in this area: giving computers the ability to recognize people’s emotions, and to respond appropriately to these emotions.

Emotion recognition involves processing many modalities: facial expressions, posture, gesture, vocal information, and more. In many cases a person is in physical contact with a computer or product, and this contact information also communicates emotion. The presentation showed examples of several means of affect recognition, including new wearable computers with customized pattern recognition and various sensors for analyzing signals carrying human affective information. These include eyeglasses that sense and communicate expressions such as confusion or interest, and a wearable "StartleCam," a first step toward a system that reduces information overload by recognizing and responding to affective responses of the wearer. It also described the results of experiments we ran to recognize frustration of people using computers.

Intelligent responses to emotional users pose a special challenge for computers. Our work in this area is just beginning, and is based on known principles that are effective in human-human interaction. The talk described the results of our first efforts in this area trying to develop software agents that attempt to implement some affective "customer relations" skills.

This technology has many applications for improving human-computer interaction as well as for assisting in gathering consumer information, both in a laboratory, and through distant, wireless, ongoing interactions with people going about their everyday activities.

REFERENCES

Barley, S. R. (1988). The social construction of a machine: Ritual, superstition, magical thinking, and other pragmatic responses to running a CT Scanner. In M. Lock and D. Gordon (Eds.), Knowledge and practice in medicine: Social, cultural, and historical approaches. Hingham, MA: Reidel.

Nass, C. I., Moon, Y., Morkes, J., Kim, E., and Fogg, B. J (1997), "Computers are Social Actors: A Review of Current Research," in Human Values and the Design of Computer Technology, Batya Friedman ed., CSLI/Cambridge University Press, 137-163.

Reeves, B. and Nass, C. I. (1996), The Media Equation. Stanford, CA: Cambridge University Press/CSLI.

Turkle, S. (1984). The second self: Computers and the human spirit. New York: Simon and Schuster.

Winograd, T., & Flores, C. (1987). Understanding computers and cognition: A new foundation for design. Reading, MA: Addison-Wesley.

Zuboff, S. (1988). In the age of the smart machine: The future of work and power. New York: Basic Books.

----------------------------------------