When I think casually about people working with technology, it is not uncommon for me to instinctively choose a rather narrow view of that activity. I find myself imagining the simple case of "using" the technology. I imagine that:
- A single person is working
- He is working successfully using his technology
- The technology is already set up
- All needed supplies are on hand
- The technology is appropriately configured for the work task it is part of, and
- The user's work task is within the accepted norms of the work situation and is understood by all concerned.
Nevertheless, from years of experience with design of various kinds, I know that this view of technology in use is narrow in many ways. And when I stop to think about it, I am aware that I am not thinking things through as carefully as I might. When I think about it, I know that I might instead imagine:
- A task involving a number of people, wherein
- The users have difficulty with the technology while doing their work
- The technology is not ready to go and requires being set up
- Supplies are not on hand and the user has to be found
- The technology must be reconfigured before it can be put to use, and
- The work goes beyond what is expected or permitted, and people must negotiate with and about those norms in order to get the job done.
Over the years, through scientific studies and design practice, I have found myself organizing into six dimensions this vast range of human activity that design must address. I do not believe that this structure is complete or even the best; it is just the way I think about what I am designing. I use it as a checklist to help ensure that I am not inadvertently forgetting any of the lessons that, sometimes painfully, experience has taught me.
This column describes the six dimensions. They provide a high-level structure for the human activities associated with using technology as I have come to see them (science: the seeing of where you are). However, knowing what is there, despite being an excellent foundation for seeking solutions (design: the exploration of where you would like to be), does not dictate what those solutions should be. Therefore, in this column I do not suggest solutions; I only indicate that solutions are needed. And, further, implementing any such solutions (engineering: the getting from where you are to where you'd like to be) is another set of concerns that are not addressed here.
I will not go into detail on any of these dimensions. (A thorough treatment will appear in a forthcoming book on task analysis, edited by Gerrit van der Veer.) However, even without much detail, this essay is somewhat longer than is appropriate for the design column of interactions. Therefore, the essay is presented in two parts, each explaining three of the dimensions. The second part will appear in the next issue.
Narrowness of view can affect our ability to see all that is relevant to our analysis, even when we are thinking carefullyas, for example, when we are explicitly and purposefully observing and analyzing the tasks in which users are involved. As we analyze the tasks that people are doing as they use technology, our pretheoretical na*iuml;ve viewpoint can restrict our ability to see clearly all that should be seen for a thorough analysis.
Conversely, any activity is infinitely full-bodied; the structure and details of carrying it out are not only of arbitrary richness in themselves, they are also based in limitless physical and social arrangements that extend without bound. Most of this material will not be of concern most of the time. Further, when issues do become of concern, the user will usually deal with them only to the extent necessary to continue the job. However, at any time, any part of this extraordinarily complex reality can become of concern, and when it does, the user will have to address it. Thus, the activity in which the user is involved is unbounded in extent. A complete description is impossible, indeed unthinkable.
Still, attempting to lay some structure over this infinite mass of subject matter is valuable. With such a framework in hand, studies of activityand the resources brought to them by participantscan be focused, design can be undertaken, and systems can be implemented. Lacking such an orienting framework, our science, design, and engineering will remain hopelessly awash in important and relevant "details."
Each perspective is an extension of the analytical viewpoint, permitting the analyst to see activity more fully. Each perspective identifies a particular sort of work that people must do as part of making use of the machine. As a consequence, for the analyst, additional user activity means a broader subject matter for analysis, and for the designer, more user activity means more to be thought about in the development of the technology, including how the technology could be designed to support that activity.
A word of acknowledgment: These perspectives arose from my years of research on humancomputer interaction (HCI), mostly at Xerox Corporation. Although the crafting of the experience into perspectives is my own, the grounding experiences occurred during collaborations with Xerox colleagues. In particular, I want to thank Lucy Suchman, Annette Adler, Kathy Carter, and Jeanette Blomberg, and to acknowledge their profound contributions to my understanding of the world. Without them, these viewpoints would no doubt have remained literally unimaginable to a techie like me. Responsibility for what I make of it is, of course, my own.
The six dimensions of extending our framework for understanding people's activity in using technology are organized to address three aspects of use:
- Operating: the activities carried out by users to make the technology do what they want: roughly, "driving" the technology;
- Enabling: the activities of arranging things to enable the operating activities: roughly, "preparing" for operating the technology; and
- Empowering: the activities of establishing the social circumstances within which the activity has meaning and value: roughly, "justifying" the enabling and operating activities.
The "operating" aspect of the activity of using a machine deals with the mechanics of users making the machine do what they want. It includes the actions and thought involved in "driving" the machine.
Dimension 1: Trouble
Case: Operating the 8200 copier
In the spring of 1981 Lucy Suchman and I observed pairs of people operating the Xerox 8200 copier, the first machine with an automatic document handler for handling entire documents of loose sheets . One pair of users changed their minds about how many copies they wanted to make while they were part way through making them. But the copier was not prepared to accept the change; one user pushed the Clear button, and the machine responded by not changing anything. However, the users interpreted this inactivity (what the designers presumably expected them to see as a refusal) as a possible failure of the machine to "hear" them; they tried again. The machine interpreted the second attempt as a request for a complete restart of the job. The control panel reflected the change, but the operators did not notice the shift. The copy count on which they were focused was reset and subsequent entry of the new count was successful; they believed they had made their desired change. Thus, the machine was starting on a new job, whereas the users were continuing on the modified original job. Much difficulty resulted from this misalignment. Some time later, these folks came to the same juncture in another attempt to do their job, and again they attempted the same "illegal" change in copy count. The interactions were the same, but this time the users saw what had happened. Of course, now they wanted to undo the second push of the Clear button. But again, the machine was not prepared to do this.
Activity: People getting into trouble
Users rarely carry out an activity flawlessly. Rather, they act out of partial understanding (both learners and experienced users) and work it out as they go along. Further, the world in which they are operating is constantly changing and, as a result, presenting new situations. Designers embed into the technology particular views of the world. People will take actions that depart from those views; they will get into "trouble" of varying degrees of complexity and will have to deal with that trouble in one way or another. In this case, the users had trouble both with understanding the interaction with the machine and with the impasse they faced by having made a mistake and not being able to undo it.
Design requirement: Help people get out of trouble
As designers we should design technology to help users get out of the trouble that they inevitably get into. As designers we should consider how things can go wrong, as well as how they can go right. That is, once a design is completed, with scenarios of how users will complete tasks, a second phase must be entered. Consider all the actions that users might take that are not the ones we have addressed in our scenarios of "right" use. The reason that users might take each action are completely open ended, so designing around any particular reason is not wise. Instead, having thought about possible incorrect actions, we must ask how the user will experience them, detect that they are incorrect, and correct them. Machine interactions to support these activities must be added, be examined for how they too can go wrong, and be corrected.
Dimension 2: Users
Case: Connecting offices by video
In 1991 Annette Adler and I connected our offices at Xerox PARC with video [1, 5]. Our two offices became very much like a single two-person office with a "barrier across the middle" that restricted the motion of physical things. This round-the-clock video window was new to people in our laboratory, and many surprises occurred. Not least of these surprises was that we were not the only users. We began to notice that neighbors from down the hall came to my door and peered in as a way to reach Annette. They even would ask Annette to help them reach Annette's neighbors, or in some cases (usually when neither of us was there) they would call through, hoping that Annette's neighbors would hear. This spilled over into the hallways. When the fitness class near Annette's office started, the music would spill into the hallway outside my office. This connection created some embarrassing situations for those who did not know, or forgot, that the connection was there. On occasion, peopleincluding mewould rearrange clothing, forgetting that their actions could be observed beyond the physical room they were in; occasionally there was someone in the other office observing, or discretely not observing. We learned that we had to inform neighbors about the link. And we had to inform the cleaning staff too. At the largest and most diffuse level, everybody was affected by this technology and its use. People became aware of the link's existence and capability. The very possibility of being able to install holes between offices changed the shape of space in the building from being something fixed to something that was malleable. This slowly worked its way into people's awareness. During a space allocation meeting, it was even suggested that rather than move someone to be nearer to a colleague, the two should be provided with a video connection.
Activity: Technology has many users
The presence of technology affects many more people than those with their hands on the controls. It is so easy to think only of those identified as "the users" when justifying the purchase of the equipment. However, many others may use it, either directly or indirectly through others. Furthermorethe neighbors, the competitionmay be affected by it. Still moreleaders and managers, the community at largewill become aware of it.
Design requirement: Design for the needs of all the users
Many people use any given technology, with a correspondingly wide range of activities that they do with or through it. As analysts and designers we must watch not only those who have their hands on the technology, but all the others who are affected by its presence. The needs of everyone must be considered, and the support of their "operation" of the technology must be addressed and designed for, just as seriously as for those ostensibly involved more directly.
We now move beyond the activities of operating technology to examine activities carried out by users to enable the operating activities. Enabling activities include both the physical and the social and are often mechanistic. They are the small, often unconsidered, but usually critical arrangements that must be in place to enable users to operate the machine. The first dimension of enabling addresses resources; the second (in the next issue) addresses practices.
Dimension 3. Support
Case: Making a copy
Once in 1994, when doing a routine piece of office work, I intentionally and self-consciously tried to pay attention to everything I did that was not directly associated with operating the technology but that was still needed to get the job done. I wanted to print a letter on Xerox letterhead. First, I had to decide whether to change paper in the shared hallway printer, thereby risking someone else's job getting there first to our mutual annoyance, or to print on plain paper and copy onto letterhead. I chose the latter because I liked control. I sent the document to print and walked down the hall to the printer. Finding no output, I discovered that the machine was out of paper. I added paper, and the print appeared. I turned to the copier, and found not only no letterhead, but no table on which the letterhead usually sat. After some unsuccessful searching, I appealed to the administrator who "owned" the copier; she informed me that the repair people required more space around the machine, so the table had been moved around the corner. Returning, I found the table, but still no letterhead. More searching turned up the letterhead stored in shelving beside the table. I replaced the letterhead on the table, and put a sign above the machine pointing around the corner. Turning once again to the copier, I found my friend Steve doing a big job on it. We negotiated an interrupt because my job was small and because Steve knew how to do it without losing his place in his own job. Steve showed me how to use the by-pass tray for the originals and an auxiliary paper tray for letterhead. After getting the orientation wrong the first time, I successfully got my letter on letterhead. The elapsed time for all of this: about 5 minutes, with most of the time spent writing notices and talking to the administrator and Steve. I was amazed at how much supporting work there was, work that was normally unremarkable and hence invisible.
Activity: Understanding that the support of doing requires activities associated with knowing, changing, and managing
Supporting activities are usually more extensive and varied than operational activities. Support activities do not necessarily have to be carried out before the operation begins, and in fact often emerge as needing attention as the operation unfolds (e.g., the copier running out of paper leads me to find and load more; wanting to avoid unloading a big job leads me to learning how to use the by-pass and auxiliary paper trays). A rough breakdown of these support activities is as follows:
- Knowledge activities: things that need to be done to ensure that the people using the technology know how to do so. They include learning (which we tend to design into our technology), teaching (we help those who help others [my interaction with Steve])
- Organizing the information: (we add newly generated understanding that we want to share [my posting notes]), and taking responsibility for seeing that these knowledge activities happen and happen well (who is the local expert, either assigned or self appointed [knowing that the administrator was the person to talk to about letterhead]).
- Technology modification activities: things that need to be done to ensure that the technology is capable of doing the job. Although copiers are not usually adjustable, other technology can increasingly be tailored to individual needs and preferences (e.g., car seats, radio presets, preferences in software applications). These activities include: customizing the technology to its context (tailoring), arranging ancillary technology (moving the table around, repositioning the letterhead for others to find), maintaining the machine, and cleaning up the surrounding work area.
- Resource management activities: things that need to be done to ensure that the resources associated with operating the technology are managed. They include management of the resources required for the technology to work (ensure that the power is on, that paperincluding special kinds like letterheadis available, that toners of the right colors are available) and management of the resource that the working technology provides (the ability to make copies [my negotiation with Steve]).
Design requirement: Support knowing, technology modification, and resource management
People often spend considerable effort on such support activities, in fact more effort than in operating the technology when things have gotten properly set up. As designers, we must understand these support activities and address them in our designs as much as we address the activities of operation.
The second part of this column will appear in the next issue. It will address the other three dimensions: practices (a continuation of enabling) and values and designers (the empowering aspect).
The column will conclude with some comments about how all these dimensions interact.
2. Blomberg, J. Social interaction and office communication: effects on user's evaluation of new technologies. In Technology and the Transformation of White Collar Work, Lawrence Erlbaum Associates, Hillsdale, NJ, 1987, pp.195210.
6. Fikes, R. E., and Henderson, A. On Supporting the Use of Procedures in Office Work. In Proceedings of the First Annual National Conference on Artificial Intelligence, American Association of Artificial Intelligence, Menlo Park, CA, 1980.
7. Harris, J., and Henderson, A. A Better Mythology for System Design. In Proceedings of CHI'99: Human Factors in Computing Systems, Pittsburgh, PA; Association for Computing Machinery, New York, 1999.
8. Henderson, A and Card, S. K. Rooms: the use of multiple virtual workspaces to reduce space contention in a window-based graphical user interface., ACM Transactions on Graphics 5, 3 (July 1986), pp. 211243.
9. Henderson, A., and M. Kyng. There's No Place Like Home: Continuing Design in Use. In J. Greenbaum and M. Kyng, Design at Work: Cooperative Design of Computer Systems, Lawrence Erlbaum Associates, Hillsdale, NJ, 1991.
10. Henderson, A., Robertson, G. G., et al. Buttons as First Class Objects on an X Desktop. In Proceedings of UIST'91: Symposium on User Interface Software and Technology, Hilton Head, SC; Association for Computing Machinery, New York, 1991.
11. MacLean, A., Carter, K., et al. User-tailorable systems: Pressing the issues with buttons. In Proceedings of CHI'90: Human Factors in Computing Systems, Seattle, WA; Association for Computing Machinery, New York, 1990.
Design Column Editors
89 South St, 2nd Floor
Boston MA 02111
Rivendel Consulting & Design, Inc.
P.O. Box 334
8115 La Honda Rd. (for courier services)
La Honda, CA 94020 USA
©2000 ACM 1072-5220/00/0900 $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 © 2000 ACM, Inc.