As I enter my second year as a student of Human Computer Interaction, I’m enjoying having the chance to read foundational and modern papers that represent the field. These are part of the HCI Research class I’m in, and each week we write up little reviews of the papers. I decided to take the opportunity to share my thoughts more broadly! So now all three people who stumble upon this blog will have access to my evolving thoughts and opinions about these papers, and HCI more generally.
Bush, Vannevar (1945). “As We May Think” Atlantic Monthly 176 (July 1945) pp. 101-108. (Original in Context Printing 1945)
Ivan E. Sutherland. 1964. Sketch pad a man-machine graphical communication system. In Proceedings of the SHARE design automation workshop (DAC ‘64). ACM, New York, NY, USA, 6.329-6.346. DOI=http://dx.doi.org/10.1145/800265.810742
Kiesler, S., Siegel, J., & McGuire, T. W. (1984). Social psychological aspects of computer-mediated communication. American Psychologist, 39(10), 1123-1134. http://dx.doi.org/10.1037/0003-066X.39.10.1123
Many HCI courses start out by reading Vannevar Bush’s classic 1945 piece “As We May Think”. It was written by the former lead engineer for the R&D efforts of World War 2. During WW2, Vannevar was in charge of coordinating the many scientists and researchers doing work for the war effort, and was instrumental in keeping the Manhattan Project at a high priority. Historical context is always important, but I think the particular details of this essay are especially interesting. Since he was part of the team that recommended deployment of the nuclear weapons to Harry S. Truman, Vannevar was among the first to know that the bombs would be used. This article came out after the decision was made, but before Hiroshima and Nagasaki were destroyed. I don’t know how obvious it was in July 1945 that the dropping of the bombs would be a decisive end to the war, but it seems that Vannevar’s mind was already looking beyond the war to how society and technology would integrate in the coming years and decades. Some even interpret this essay as a kind of apology for how much damage scientific efforts had inflicted on the world during the war years, and an attempt at leveraging the awesome war-time scientific coordination for more positive means. In trying to find some way to salvage the infrastructure and intellectual prowess that was used in such a destructive way, Vannevar latches onto this notion of intellectual contributions. His techno-optimism is both refreshing and depressing (since technology has not solved many of the problems he hoped it might).
Remarkably, for someone sometimes considered the “father of HCI”, many of his beliefs and actions run counter to what the field has become. For instance, as president of the Carnegie Institute of Washington in the 1930’s, he evidently saw little value in the humanities and the social sciences, and cut funding for a journal dedicated to describing the cultural influences of technology. These beliefs of his do come through in the descriptions of the technology in “As We May Think”: highly technical, designed for the work environment, with little to no thought as to how the technology might affect life outside of the lab.
The essay was a product of its time, and it’s important to we may recognize various limitations such as the fact that scientists are always referred to as “he” throughout the piece. Still, there is much to be lauded about the forsight and vision shared by Vannevar. I also find his enthusiasm – especially knowing his role in overseeing thousands of research projects – refreshing. Take this passage, where he marvels at the fact that complex machinery works so well:
Machines with interchangeable parts can now be constructed with great economy of effort. In spite of much complexity, they perform reliably. Witness the humble typewriter, or the movie camera, or the automobile. Electrical contacts have ceased to stick when thoroughly understood. Note the automatic telephone exchange, which has hundreds of thousands of such contacts, and yet is reliable. A spider web of metal, sealed in a thin glass container, a wire heated to brilliant glow, in short, the thermionic tube of radio sets, is made by the hundred million, tossed about in packages, plugged into sockets—and it works! Its gossamer parts, the precise location and alignment involved in its construction, would have occupied a master craftsman of the guild for months; now it is built for thirty cents. The world has arrived at an age of cheap complex devices of great reliability; and something is bound to come of it.
The first thing that strikes me about this paper is how closely it relates to the idea of Direct Manipulation - even though Schneiderman’s paper on that subject wasn’t published until nearly 20 years later! Specifically, choosing a graphical representation of data that is best shared visually is a perfect example of moving away from the “conversation metaphor” and towards the “model-world” metaphor described by Hutchins, et al. in their 1985 paper. I think Hutchins, Hollan and Norman would agree that Sutherland’s work successfully shrinks the gulfs of execution and of evaluation.
I’m also drawn to the extraordinarily tangible interface Sutherland describes: zooming in by turning knobs, flipping switches to change functions: the whole system sounds as complicated as a pilot’s cockpit (or the modern implementation of most CAD programs #burn #shade #iwentthere). Sutherland goes on to describe in great detail how one would construct a perfect hexagon or a perfect circle, and I’m very curious about how the average person in 1964 - who had never seen a computer, much less a graphical user interface like this one - would possibly interpret this. Would the system seem useful or just silly? Would it seem overly complicated? Graphical User Interfaces are obviously still a huge part of how humans interact with computers. I don’t know what most initial reactions to GUIs were by the average lay person, but I imagine it wasn’t a totally smooth introduction. It’s a reminder for myself to keep an open mind about how useful new interfaces may appear when first introduced: and especially that we may not be able to predict the effect they’ll have. (I always think of the inventor of the vacuum tube – which later enabled society-changing tech like the television – who died before it was used for almost anything at all!)
I’m also fascinated by the particular level of detail in the paper. It seems clear that the audience was expected to be those who had access to and familiarity with computers and compilers. Sutherland blithely references pointers with no explanation of what those are, apparently assuming that his audience needs no explanation.
I LOVE that the system basically teaches the user about one of its limitations by using that limitation as a control signal: the user indicates that they are finished drawing by flicking the pen too quickly for the system to track. This automatically provides the user with a mental model of how the system works by exposing a “flaw”. I don’t think it was intentionally done that way, but it’s still very satisfying!
I deeply appreciated this paper as yet another stepping stone along the path towards modern-day HCI. The combination of many fields - especially those well-equipped to study humans - is an enormous part of what I love about HCI.
One of my favourite sections in this paper is where they describe some of their findings in a study they did. The authors were interested in exploring computer-mediated communication and the effect of online systems on group decisions. They brought in groups of 3 people who were instructed to come to consensus on a well-studied choice-dilemma problem. There were three different scenarios: one where the subjects worked together in person, on where they worked together anonymously on the computers, and one where they worked together non-anonymously. Prior work showed a tendency for people to shift towards extreme opinions in group settings, if they are exposed to those extreme opinions. That often doesn’t happen in face-to-face conversations due to social pressures to maintain more moderate initial positions. However, the authors expected that people in the computer-mediated cases would be less inhibited since there were fewer social norms around computer use, and would therefore be more likely to share extreme versions of their views. This would lead to greater choice-shift. This hypothesis was supported. The authors speculate that the depersonalized nature of computer-mediated-communication contributed to a greater tendency to change their minds. They also suspected the computers lowered the cost of ignoring social pressure to reach consensus.
Reading this now, in 2017 is almost terrifying. The authors seem to implicitly assume that new social norms around using the computer will eventually develop. I guess they have, but reading this paper it seems that norms simply solidified into the worst case of what the authors call “uninhibited behavior”. This description of subject behavior in the anonymous-computer scenario eerily matches my own experience with Twitter:
subjects swearing, individuals shouting at their terminals, and groups refusing to make a group decision until a group member gave in
The description is comical, especially when constrasted with the formal, academic style of writing. But it also made me question whether or not we’ve actually progressed in terms of the way we interact through computational devices. In fact, the authors identified a number of problems that I experience every day on platforms like Twitter and Tumblr:
People did not know exactly when their arguments were understood or agreed to, and consequently everyone believed they had to exert more effort to be understood.
In other words the authors correctly identify online communication as being good for talking but terrible for listening. Truly, their descriptions of the challenges are almost eerie:
A final explanation for our results is that electronic communication involves a process of depersonalization or a redirection of attention away from one’s audience. Suppose computer-mediated communication prevented personal feedback and individuating information and at the same time lacked a shared etiquette and, further, was influenced by norms from the computer subculture. This could have made group members more responsive to immediate textual cues, more impulsive and assertive, and less bound by precedents set by societal norms of how groups should come to consensus. This explanation fits our data.
How have we still not solved this issue? How can we solve this issue? Is it possible for humans to build online systems that support better forms of communication? What would that look like, and how can we ensure systems that facilitate positive, calm, and felicitous online communication?
© 2023 / Molly Jane Nicholas / email