Depending on how you look at it, human-computer interaction has either no home or many homes. We're multidisciplinary without having become particularly interdisciplinary. The first HCI papers were Human Factors & Ergonomics, which is often located in Industrial Engineering departments. Business and Management schools then initiated relevant Information Systems research. CHI originally comprised mainly Cognitive Psychologists, with an influx from Computer Science some years later. HCI is now on the rise in Information and Design schools and departments.
Is this a positive instance of letting a thousand flowers bloom, or a dire Franklinian, "We must all hang together or we shall all hang separately?" If computers become invisible or disappear, the study and practice of human-computer interaction may be everywhere and nowhere. But HCI seems challenging enough to deserve to have a spotlight on it somewhere.
To try to understand why different strands of HCI research and practice have not converged, I have been examining the history and interviewing participants in it. Everyone has a unique perspective, choosing what to emphasize and what to skip over. I have collected some perspectives in this article (though they are not comprehensiveadd your own!). The measure of a perspective is less whether it is right or wrong than whether or not it is useful: In this case, does it help us understand the field's state of fragmentation and the possibilities for more effective integration?
Three Faces of HCI. I will focus on the three long-standing threads of research shown in Figure 1. Human Factors reaches back to the World Wars and earlier application of "scientific management" to the design of assembly lines and other work processes. Information Systems arose around the needs of business computing that spread in the late 1960s. These fields were active when a computer was so expensive that it had to be kept in continual use, with most hands-on users, in Brian Shackel's words, "almost slaves to feed it" . Non-discretionary use also marked the earlier human factors work on assembly lines and military technology (aviation psychologists formed the Human Factors Society after World War II). Efficiency and error reduction in skilled use were emphasized, with some focus on training to develop skill. To demonstrate small but valuable improvements in tasks such as data entry required controlled experiments and rigorous analyses of the sort identified by Don Norman in his column this issue .
Mandatory use was the norm through the mid-'70s, but visionary engineers, prototype developers, and writers anticipated a future in which people would use computers not because it was a job requirement but because they wanted to, because it "augmented their intellect"  or made jobs more enjoyable. When small mini-computers, home computers, personal computers, and workstations arrived, the visions began to be realized. Discretionary computer use has been a focus of CHI from the beginning. What difference does this make? For example, for understanding people's reactions to a new discretionary technology, controlled experiments and statistical analysis are less effective, and often less necessary.
Mandatory vs. Discretionary Hands-On Use. Figure 2 color codes these research threads to illustrate the distinction around discretion in use. Those interested in detailsfull names, acronyms and book titles; why items are positioned where they are; and what subtle complications arise (for example, discretion is not all or none, and a technology that is discretionary when it appears can become a job requirement later on)can consult two longer articles, one just published in IEEE Annals of the History of Computing [4, 5].
With the arrival of ecommerce, the Information Systems focus has expanded to include discretionary use. Previously concerned with internal use of systems in organizations, IS now examines corporate Web portals designed for finicky potential customers. The shared focus on discretionary use creates a new opening for collaboration between CHI and IS practitioners and researchers. Achieving this is a major goal of the Association for Information Systems' Special Interest Group on Human-Computer Interaction (AIS SIGHCI), formed in 2001.
When discretion is involved, aesthetic design matters. Slowly and not without reluctance, CHI loosened its ties to psychological theory and engineering and schooled itself in Design. With discretion also comes Marketing, as noted by Aaron Marcus in these pages last year . When CHI is ready for a crash course in marketing, IS may be in a position to help.
Attempted Bridges. The first official CHI conference was co-sponsored by ACM and the Human Factors Society. The program chair and many program members, tutorial instructors, and paper presenters were from the Human Factors community. But within a few years, Human Factors participation in CHI was negligible and computer scientists had arrived, many inspired by the early visionary writers and prototype builders and bringing decades of work on software engineering, artificial intelligence, and above all else, graphics.
In the late 1980s, Computer Supported Cooperative Work (CSCW) set about bridging Information Systems and CHI. Early conferences had strong IS representation, but again this only lasted for a few conferences.
In interviews, participants in these bridging efforts identified differences in methods that result from differences between non-discretionary operation and discretionary hands-on use as a significant factor in the lack of success. But it was not the only factor.
Journals vs. Conferences as the Principal Focus. As illustrated in Figure 3, Human Factors and Information Systems have retained the journal orientation familiar to the sciences. In the United States, archived ACM and IEEE conference proceedings have enabled Computer Science to regard selective conferences as the premier and final showcase for much research. Many people I spoke with identified this as a barrier to interdisciplinary communication. In fields in which conferences are for work in progress toward journal publication, the high rejection rates of ACM conferences are strange and off-putting: "If I had something polished enough to get into CHI or CSCW, I would be better off submitting it to a journal, which gets more recognition in my field." Conversely, CHI researchers attending conferences in other fields are dismayed by the unpolished work and uninterested in the goal of eventual journal publication. Some people see and understand these differences, but others have papers rejected or observe unpolished work and take offense.
As editor of Transactions on Computer-Human Interaction, I heard senior CHI people pronounce journals irrelevantslow and unnecessary. This was always from North Americans. In Europe and Asia, where professional societies have not routinely made conference proceedings available in perpetuity, work must still reach journal publication to get into the literature. This intercontinental difference creates substantial tension within the broad CHI community.
A Cultural Divide. The significance of Human Factors' origin in and ongoing ties to the military and other governmental agencies can hardly be exaggerated. Governments were the principal purchasers of early computer systems. Information Systems researchers extended their focus to other business settings, but came from the same generation, published in some of the same journals, and employed similar methods. For example, the waterfall development model that originated in government contracting was applied to in-house development, where it was less suitable.
John Markoff's book What the Dormouse Said is subtitled How the '60s Counterculture Shaped the Personal Computer Industry . This generational difference is illustrated in Figure 4. Many CHI researchers had been students in the Vietnam era and did not regard the military and government favorably. Human Factors and Information Systems continued to write about "man-machine interfaces," whereas CHI pushed for gender-neutral terminology from the start. These differences were not conducive to joint meetings and impeded forays into one another's literature.
Those who tried to bridge the fields encountered linguistic differences more extensive than disputes over male generics. Common termsuser, system, implementation, task analysis, evaluation, application, adoption, and so on had markedly different meanings or implications across research fields. This led to substantial misunderstanding and confusion. I observed an invited lecture, a conference panel, a journal submission, and a visiting professorship affected, in some cases seriously, by a misunderstanding over one of these terms. For example, interpreting task analysis in the CHI sense of a cognitive task analysis (e.g., is the command move better thought of as select, I showed up at a European panel on task analysis only to discover that they used the term to refer to an organizational task analysis). A major European researcher later expressed outrage over what he perceived as abuse of the term task analysis by CHI people. I have similar observations for each term listed above, and others.
Different Theoretical Foundations. The final schematic figure identifies fields of psychology that influenced these threads of human-computer interaction practice and research. Human Factors sprang from engineering psychology, aviation psychology, and industrial psychology, heavily influenced by the stimulus-response and radical behaviorist theories that dominated in the US in that era. CHI focused on cognitive psychology. The academic rivalry between these approaches impeded efforts to find common ground. Over time, antagonisms waned. Today two of the largest Human Factors and Ergonomics Society Technical Groups are Cognitive Engineering and Decision Making, and Human Performance Modeling, the latter organized only last year by people previously active in CHI.
Pre-CHI pioneers shared a focus on engineering psychology. In the late 1980s, as networking spread and CSCW became a significant subfield, CHI attracted more social scientists. The Information Systems field drew less directly from theoretical psychology. Some researchers applied elements of personality theory, and the applied specializations of industrial and organizational psychology had an influence. Social science came into play, as did a cognitive component that is evident in two of the most active AIS SIGs, SIGHCI, and SIGCORE (Cognitive Research).
HCI: Looking for a Home? In each of these fields, we are under pressure. The Computer Systems Technical Group of Human Factors has declined. Information Systems is under pressure in Management schools. CHI shifted quickly from Cognitive Psychology to Computer Science, and now shows signs of moving or diffusing further, to Information Science and Design.
This echoes the notion of the invisible computer/computation receding into the environment. Now, every branch of Human Factors, every specialization in Management schools, and diverse subdisciplines within Computer Science, Information Science, and Design must engage with human-computer interaction.
Information Science and Design. Information Science research began even earlier than the fields I've focused on, but it fully embraced digital technology only recently. Now it is pulling researchers from each of the threads I have highlighted, with IS and CHI refugees particularly numerous alongside library scientists. These schools and departments are interesting laboratories of disciplinary experimentation, mixing people from journal and conference cultures, from non-discretionary data entry and retrieval and discretionary information foraging cultures. In 2005, representatives gathered for a joint meeting to explore commonalities and differences .
Design has come into stronger focus within CHI because of its inescapable role in discretionary use. As a profession, though, Design introduces another cultural difference. Quality is assessed more by portfolio and reputation, less by conference or journal publication. Creating venues to attract both design and text cultures will be an ongoing challenge.
Conclusion: Opportunities for Inter-disciplinarity. These different views of human-computer interaction are presented as a way of understanding forces that have kept the field fragmented. In some cases, different theoretical and methodological approaches were sensible consequences of different priorities, and differences will remain. In other cases, the centripetal forces may be more happenstance in origin and might be overcome by better understanding. Nevertheless, obstacles such as the journal/conference orientation distinction are significant.
People ask for recommendations as to how to proceed. My response is that above all else, anyone wishing to bridge fields must think long and hard about the fields and the historical forces at work in them. Past efforts in which I participated and re-investigated more recently did not fare well. We must understand why. As you might expect, my response doesn't satisfy many who ask. They want to go out and DO something, and that's great.
3. Engelbart, D. A conceptual framework for the augmentation of man's intellect. In P. Howerton and D. Weeks (eds.), Vistas in Information Handling Volume I. Washington, D.C.:Spartan Books, 1963, pp. 1-29.
5. Grudin, Jonathan: "Human Factors, CHI and MIS," Human-Computer Interaction and Management Information Systems: Foundations, P. Zhang and D. Galletta, Eds., Advances in Management Information Systems, Vol. 4, M.E. Sharpe, in press.
About the Author:
Jonathan Grudin is a senior researcher in the Adaptive Systems and Interaction group at Microsoft Research. He has been active in CHI and CSCW since each was established, and participated (less frequently) in Human Factors and Information Systems activities. His Web page is http://research.microsoft.com/~jgrudin.
©2006 ACM 1072-5220/06/0100 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2006 ACM, Inc.