There Is No Interface (Without a User). a Cybernetic Perspective on Interaction
Total Page:16
File Type:pdf, Size:1020Kb
INTERFACE CRITIQUE JOURNAL – VOL. I – 2018 THERE IS NO INTERFACE (WITHOUT A USER). A CYBERNETIC PERSPECTIVE ON INTERACTION By Lasse Scherffig “Interaction is seen as a one-way street, conveying a design model to a user, who is acting by that model either because they adapted to it, or because the model replicates their given structure. This is the cognitivist heritage of the HCI discourse responsible for the idea that interfaces can actually be designed.” Suggested citation: Scherffig, Lasse (2018). “There is no Interface (without a User). A cybernetic Perspective on Interaction.” In: Interface Critique Journal Vol.1. Eds. Florian Hadler, Alice Soiné, Daniel Irrgang DOI: 10.11588/ic.2018.0.44739 This article is released under a Creative Commons license (CC BY 4.0). 58 SCHERFFIG: THERE IS NO INTERFACE The interface in itself does not exist. rected development from the compu- This is not to say that any phenome- ting machinery of the past towards a non must be perceived in order to ex- future of interaction. ist, but rather that interfaces quite This trajectory constitutes the literally only come into being if they standard account of the history of in- are used. They are effects of interac- teraction. Often the field is seen as tion and thus they are ultimately cre- following a teleological development ated by their users. of progress, during which computers Of course, academic and pro- became more and more interactive, fessional disciplines like human- and interaction became more intui- computer interaction and interaction tive, rich, and natural. This develop- design assume the opposite: namely ment is often explicated as a that interfaces are designed (and ex- genealogy. Depending on the focus ist) before they are used, possibly and goals of their narrators, there are even creating their users. This article1 genealogies of interaction focusing traces the development of this view, on a succession of hardware genera- as well as offering an alternative to it tions, interaction paradigms, theo- that fundamentally understands any retic frameworks, or visionaries interface as “cybernetic interface.”2 pushing the field to the next level. An early and paradigmatic ac- count that focuses on hardware is John Walker’s genealogy of five “User GENEALOGIES OF Interaction Generations”4 published in the early 1990s. This account starts INTERACTION with the “plug boards” and “dedicated setups” of early computing.5 These When during the late 1990s, the new were followed by the (in)famous era millennium prompted countless ret- of batch processing – a time when rospectives and outlooks, Terry Wino- programming meant punching holes grad contributed a chapter to a book into cards, handing batches of these about the next fifty years of computer cards to a mainframe operator and science titled, “From Computing Ma- 3 waiting for hours to be handed back a chinery to Interaction Design.” Fol- printed result. lowing an idea of evolutionary It is only the third of these progress, this title described a goal di- generations that was interactive. As –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 1 A German language discussion of similar 3 Terry Winograd, “From Computing questions, that is much broader in scope, can be Machinery to Interaction Design,” in Beyond found in Lasse Scherffig, “Feedbackmaschinen. Calculation: The Next Fifty Years of Computing, ed. Kybernetik und Interaktion” (Dissertation, Peter Denning and Robert Metcalfe (Berlin and New Kunsthochschule für Medien Köln, 2017). York: Springer, 1997). 2 Søren Bro Pold, “Interface Perception: The 4 John Walker, “Through the looking glass,” in Cybernetic Mentality and Its Critics: The Art of Human-Computer Interface Design, ed. Ubermorgen.com,” in Interface Criticism: Aesthetics Brenda Laurel (Redding, MA: Addison-Wesley, 1990), Beyond Buttons, ed. Christian Ulrik Andersen and 439. Søren Bro Pold (Aarhus: Aarhus University Press, 5 Ibid. 439-440. 2011), 91. 59 INTERFACE CRITIQUE JOURNAL – VOL. I – 2018 to be expected from a proper genera- defined by “electrical interaction” (us- tion, it was a child of the previous one ing cables, plugs, and the soldering and generated by it: When, this ac- iron) with Walker’s dedicated set-ups. count goes, the algorithms allocating This was followed by the era of “sym- a mainframe’s computing time to bolic interaction” that was marked by several batch jobs got more and more the use of punch cards and batch pro- advanced, it became clear that it cessing – which were often pro- would be possible to divide the com- grammed using the symbols of puting time of a mainframe even fur- assembler languages instead of the ther. Divided into small enough raw zeros and ones of machine code. pieces, that follow each other in rapid This generation, in turn, led to the succession, it would seem to several “textual interaction” with the termi- people that they would have exclu- nals of time sharing systems. “Graph- sive control over the whole machine. ical interaction” here again marks the This idea, named “time sharing,” did final step in an evolution starting not divide a computer’s resources be- with machinery and ending with to- tween several batch jobs but between day’s interactive surfaces. several humans (re-defining these, as Apart from their implicit as- we will see below, as “users”) who sumption that interactivity progres- could now engage in “conversational sively increases, these, and most interactivity” with the machine.6 other histories of human-computer While the conversations of the time interaction (HCI) have one thing in sharing generation happened as ex- common: They all assume an origin changes of written text, the fourth of of interaction. While operating a com- Walker’s generation of interaction in- puter during Walker’s second genera- troduced graphical displays that con- tion meant batch processing, the centrated textual commands into third generation introduced time visual menus. The fifth and final gen- sharing and with it, interaction. In eration then spawned the graphical Dourish’s terminology, this transition user interfaces of personal compu- corresponds to the shift from “sym- ting that, in various iterations, keep bolic” (based on assembler language accompanying us on our desktops, and punchcards) to “textual interac- laptops, and phones until today.7 tion” (based on conversational inter- Paul Dourish’s “History of In- activity via command line). Only teraction,”8 which is much more con- when computers, after time-sharing temporary in style, follows a very was introduced, started to react similar path “from soldering to (seemingly) exclusively and directly mouse.” Focusing on the mode of in- to human input, did they become in- teraction instead of hardware gener- teractive: “Arguably, this is the origin ations, the first generation here was of »interactive« computing.”9 –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 6 Ibid. 441. 8 Paul Dourish, Where the action is. The 7 Ibid. 441-442. foundations of embodied interaction (Cambridge, MA: MIT Press, 2001), 1-17. 9 Ibid. 10. 60 SCHERFFIG: THERE IS NO INTERFACE Of course, history is not that ing with analog (i.e. continuous) val- simple. A closer look behind the nar- ues, it rather implied an entirely dif- ratives postulating a teleological de- ferent approach toward calculation: It velopment of interaction instead relied on building electrical and me- reveals contradictory and asynchro- chanic systems, that, as analogues or nous developments, as well as chron- analogies, could stand in for the sys- ological overlaps.10 Especially the tems they were built to simulate. Be- origin of interactive computing itself cause building such analog can be described differently, for in- computers entailed accurately fol- stance, by looking at the first interac- lowing and amplifying changing tive computer ever built – which physical signals, it largely depended happens to be one of the first comput- on another development: the rise of ers at all. It is, as this look reveals, the the use of negative feedback as the de very first generation of computing facto standard method for handling machinery that defined interactivity electro-mechanical systems. In fact, up to this day, including its problems. during the early twentieth century, negative feedback became so im- portant in both control and communi- cation engineering that both INTERACTIVE disciplines merged into one feedback COMPUTERS AS based control theory – in a paradigm shift that yielded the era of “classi- FEEDBACK cal”12 control. This development, in turn, constituted the nucleus of what MACHINES Norbert Wiener would later call cyber- netics – a science of “control and During the 1940s, the MIT Servomech- communication in the animal and the anisms Lab started to build a flight machine”13 that would become think- simulator. As the leading paradigm able mainly because the application for automatic computation at that of negative feedback and the associ- time was analog computing, the flight ated mathematical formalisms simulator was planned to be based on seemed to be powerful enough to that: “a cockpit or control cabin con- tackle any form of “behavior” – of liv- nected, somehow, to an analog com- ing and non-living systems.14 Be- puter.”11 “Analog computing” in this cause feedback implies using the context did not only imply calculat- output of a system as its own input, the systems of cybernetics exhibited –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 10 Hans Dieter Hellige, “Krisen- und Innovati- 12 Stuart Bennett, A History of Control onsphasen in der Mensch-Computer-Interaktion,” in Engineering 1930–1955 (Hitchin: Peter Peregrinus Mensch-Computer-Interface. Zur Geschichte und Zu- Ltd., 1993), 17. kunft der Computerbedienung, ed. Hans Dieter Hellige 13 Norbert Wiener, Cybernetics or: Control and (Bielefeld: Transcript, 2008), 15-20. Communication in the Animal and the Machine 11 Kent C. Redmond and Thomas M. Smith, (Cambridge, MA: MIT Press, 1961). Project Whirlwind: the history of a pioneer computer 14 Arturo Rosenblueth, Norbert Wiener, and (Bedford, MA: Digital Press, 1980), 32. Julian Bigelow, “Behavior, Purpose and Teleology,” Philosophy of Science 10 (1943): 18–24.