Several Timelines columns have argued that we fail to notice indirect consequences of successive waves of new hardware. This essay explores the dynamics of technology change, illustrated by an oscillation in the conditions affecting collaboration across continents in the computer-supported cooperative work community.
Major technology shifts affect technology producers and consumers, of course, but they also disrupt professional organizations and research fields. For example, the leading computer science research conference and trade show of the 1960s and 1970s, the National Computer Conference (NCC), collapsed suddenly in the late 1980s. Its losses brought down the American Federation of Information Processing Societies (AFIPS), which had represented the U.S. internationally. AFIPS was the parent organization of ACM and IEEE, both of which survived. NCC had tied its fortunes to the mainframe industry and resisted including vendors and researchers focused on personal computers and workstations. Evident as it is in hindsight, surviving meeting records of the time do not tie the collapse to technology change.
Why did it happen suddenly? With Moore's Law and related legislation operating steadily, why are there sudden disruptions rather than gradual evolution? An analogy with biological evolution suggests a possible explanation.
Punctuated equilibrium is the theory that animal and plant species exhibit periods of evolutionary stasis interrupted by short periods of rapid evolutionary change. The fossil record shows little change for a long time, then a dramatic shift. Is the record deceptive? Could the species have gradually evolved elsewhere and migrated to the location? If the theory is correct, what is the mechanism? Do mutations accumulate until they collectively provide a significant selective advantage? Does a sudden change in climate or environment enable a new configuration of characteristics to thrive? The fossil record is inconclusive.
Computer systems have also exhibited periods of stasis followed by disruptive change. Transistors replaced vacuum tubes. Lumbering mainframe computers built with integrated circuits spread across the land like dinosaurs, seemingly unstoppable, until minicomputers halted their progress. PCs and workstations, in close symbiosis with the pesky mammals called users, drove minis to extinction and penned in the mainframes. Today, smaller form factors are gaining the upper hand.
Our driving force is not random mutation. Moore's Law and other supralinear or exponential growth has been continuous. New chips appeared every couple years, but the major shifts appeared every decade (Figure 1). Many mutations occurred before a new entry appeared in our fossil recorda machine that delivered sufficient added value to reorganize industry and behavior around it.
Each new form factor took longer to achieve maturity in the market. Illustrated by the rise time for successive entries in Figure 1, this is presumably because 1) The installed base and market expanded. When integrated circuits arrived, there were few computers and computer professionals to convert. Once numbering thousands, users now surpass a billion. 2) A more extensive software infrastructure was built on each platform. 3) The ambition and potential of each new generation to change how we live and work was greater. Each successive platform encountered greater inertia. Handhelds splashed into view in the 1990s on schedule with the Apple Newton, PalmPilot, and other devices, but widespread adoption of mobile computing awaited better networking, more usable interfaces, greater functionality, and people trained to enter text with their thumbs.
Figure 1 omits the expansion of bandwidth and wireless access. Many Web applications are platform-agnostic, enabled by the bandwidth increases since 1990. For example, social networking sites can be accessed from various device types, although mobility is essential to exploit some features.
A previous Timelines column noted that the successive technology-driven waves created new software industries, use contexts, and the research fields shown in Figure 1 . In other disciplines, young researchers need permission from senior surfers to ride an established wave. Not so for young computer scientiststhey can move to an exciting new wave as it gathers strength, when it is still viewed with disdain by the old-timers. Designers and developers started new companies; researchers formed new conferences and journals.
Thoughtful human control of technology use is a worthy goal, but some technologies appeal to us in such fundamental ways that resistance is futile. Pizza, beer, and sugared water come to mind. People are drawn to efficient transportation, communications, and weaponry like bees are drawn to flowers, or moths to a flame. The digital fossil record supports a moderate degree of technological determinism.
A soft version of determinism holds that dropping a specific powerful technology into a particular culture will set dominoes falling that can't be stopped. Movable type in hierarchical and relatively static China did not lead to a printing revolution. But in Europe, efforts to suppress the printing revolution failed, and over time print has reached most cultures. Digital technologies are spreading faster than print, faster than television, or anything save infectious diseases. Shortly after reading The Lost City of Z, I winced as I read an account of the chief of an Amazon tribe using digital technology to preserve a record of a disappearing culture .
Cultureorganizational, ethnic, or nationalcan deflect or postpone what seems inevitable. Japan kept Western technologies and influence at bay for a century and a half. When change cannot be held off, there may be a survivor from the old school. Returning to biological evolution, terrestrial dinosaurs are extinct, but one class of dinosaur survived: birds. (Wikipedia: "From the point of view of cladistics, birds are dinosaurs... and dinosaurs are, therefore, not extinct.") IBM is a surviving mainframe vendor. Fujifilm (3 percent of it) is a surviving photographic film manufacturer. Mechanical watches haven't disappeared entirely.
Culture plays a role, at least in the short term. Some countries and companies try to prevent citizens or employees from using social networking sites. If an entire industry clings to archaic practices, natural selection can't operate within ittrade book publishers still force authors to adhere to some obsolete practices.
I will now focus on one research field, computer-supported cooperative work (CSCW). It is not especially important, but I have analyzed its participation and attended organizational meetings for 20 years. In doing so I have found that periodic turmoil that was generally attributed to personalities or conceptual differences may have been a consequence of technology changes. This analysis leads to a prediction about its future.
The history of CSCW was outlined in the July 2010 Timelines column; key markers are listed in Table 1.
The split into North American and European CSCW communities around 1990 seemed to reflect personalities and conceptual differences. Over time, differences diminished. A surprising reappearance of polarization 20 years later prompts reconsideration.
Did competitive young researchers in 1990 become mature collaborators by 2000, and 10 years later assume the curmudgeon roles that befit aging researchers? Is the prediction that in 10 years they will have been carried or pushed off the field by more amicable young researchers? No. The differences emerged during two waves of technological change, with the convergence occurring in the intervening period of equilibrium. A future convergence seems likely. The transitions are covered in the following sections.
CSCW origins. Our understanding of the basic issues surrounding digitally mediated communication and collaboration advanced steadily after Doug Engelbart's renowned 1968 demo, in which researchers miles apart collaborated online in real time. CSCW formed as the vision was slowly being realized, as research prototypes gave way to commercial systems. Central to this was the shift from mainframes and minicomputers to networked PCs and workstations in the late 1980s. But this technology shift was not experienced the same way or at the same time in North America and Europe .
1990: A divided field. The dominant commercial hardware and software PC and workstation companies were North American, including IBM, Sun, Apple, and Microsoft. Of course there were developers in Europe, Asia, and elsewhere, but the largest vendor companies were American. They sought "killer apps" to support the millions of small groups around the world. They hired experimental psychologists and software developers to refine interaction models and interfaces. The term CSCW was coined by employees of Digital Equipment Corporation and Lotus. North American CSCW participation was overwhelmingly corporate, and from telecoms and companies developing commercial software, not user organizations.
Europe was in a different space. When I worked there from 1989 to 1991, the focus was on in-house development and use in large companies and government agencies. The orientation was organizational, and projects were long-term and large system, with an emphasis on getting the system functionality right. This contrasted sharply with the American small-group focus on quick projects to refine an interface to gain a competitive advantage in the commercial market. My Danish colleagues regarded CHI as lively but preoccupied with shallow UI issues.
CSCW 88 reached out to European researchers, who were virtually all from universities and government laboratories. Their reaction was to initiate the ECSCW series in 1989. Although the CSCW and ECSCW conferences alternated years, their constituencies were not close. Early CSCW conferences had few European committee members. Apart from liaisons, ECSCW included few North Americansin fact, none were included in its initial conference and program committees. All of the founding editorial board members of the Kluwer journal CSCW were working in Europe as it was organized. Satisfaction was expressed that ECSCW was distinctly European.
Differences covered content, methods, and analytic approaches. Many North American researchers and developers sought quick prima facie or statistical support for ideas by building prototypes and conducting usability tests. Most were engineers or former experimentalists who had adopted less rigorous testing methods. Few showed any interest in theory. Europeans were concerned with theory and hypothesis testing, as the difficulty of generalizing broadly from one organization or system can leave theory construction as a way to claim a contribution to science. Skeptical of quantitative studies, the Europeans did not have statistical tests as an accoutrement of science. (North American organizational scientists, who published in information systems, not CSCW, insisted on both statistics and theory, the latter somewhat elastically defined.)
2000: Convergence. Ten years later, differences had all but vanished. CSCW and ECSCW program committees and content overlapped strongly. At the time, I attributed the convergence to mediation by industry-sponsored groups in the U.K., but in retrospect it seems more basic. We were in a state of equilibrium. The PC wave had engulfed the developed world. Large European organizations no longer built systems from scratch; they acquired commercial collaboration software of the type studied in North America. There, killer groupware apps had not materialized, and researchers realized that understanding organizational context, a focus of European research, was critical to further progress in small-group support. The communities converged. Ethnographers studying technology use were warmly embraced by both.
2010: Division reappears. Surprisingly, by 2010 the continental divide had returned. Many European researchers dropped the CSCW conference, participating instead in other North American conferences, such as ACM GROUP or Collaboration Technologies and Systems (CTS), the latter heavily attended by defense contractors and other industry researchers. North American participation in ECSCW declined and CSCW became an annual conference in 2011.
Perhaps it should not have been a surprise. A new wave of technology had arrived, Web based and increasingly mobile. Once again, the dominant commercial hardware and software developers were North American: Google, Facebook, YouTube, Twitter, Wikipedia, Blizzard (World of Warcraft), and so on.
As before, interests were pushed in different directions by technological change. Studies of consumer phenomena and online communities moved to the fore in CSCW. In Europe, companies and agencies were not interested in these. They focused on improving the collaborative use of complex internal systems and promoted sophisticated basic research in specific domains such as medicine, manufacturing, and telecommunications. Similar concerns naturally affect U.S. organizations, but that research appears in GROUP, Collaboration Technologies and Systems, and information systems conferences, not CSCW.
At a CSCW 2010 Town Hall meeting, European participants expressed unhappiness over the diminished presence of research into workplace activity Taking the "W" in CSCW seriously, they objected to studies of multiplayer gaming and other consumer or online community activity. An ECSCW 2011 manifesto by Bannon, Schmidt, and Wagner concluded "CSCW's focus is not just any kind of 'socially organized activity' but ordinary cooperative work: work in hospitals and factories, administrative agencies and research laboratories, software engineering bureaus and lawyers' offices, and so on" . This view has no vocal adherents in the North American CSCW community. CSCW 2012 had few sessions on these topics and many on studies of Wikipedia; social networking in war, crisis, and consumer settings; online communities; crowdsourcing; and multiplayer games. In the 1980s, networked PCs and workstations were found only in workplaces, but today, studies of multiplayer games, leisure, and consumer activity are of high interest to the American companies and academics that dominate CSCW.
CSCW 2010 Town Hall participants seemed to blame each other for choices over which we may have had little control. When this latest wave of technology has swept across the world, researchers may coalesce as they did 20 years earlier. Having harvested the low-hanging fruit, North Americans will realize that deeper understanding of Web 2.0 phenomena requires the domain-specific knowledge that Europeans are accumulating. Large European organizations will be using social networking software, productivity games, and other technologies that CSCW is exploring. Data miners, hopefully in concert with ethnographers, will assist in building a new bridge. And then, when amicability reigns, another disruptive wave may arrive, adding another layer to the digital fossil record.
CSCW is one example; other fields and regions might be similarly reassessed. Punctuated equilibrium is found in nature. Pressure on tectonic plates builds steadily beneath the Earth's crust, but instead of gradual change at the surface, earthquakes punctuate periods of equilibrium. It could help explain why so often organizations and entire industries are caught flat-footed by waves of technology change. Explorations of history can reveal the power of unseen forces to shape events. To influence where we go, we must understand the waves we surf.
2. Biuso, E. The 100 most creative people in business: Chief Almir / Surui Amazon Tribe; http://www.fastcompany.com/most-creative-people/2011/chief-almir-surui-amazon-tribe
Jonathan Grudin (http://research.microsoft.com/~jgrudin) is a principal researcher in the Natural Interaction Group at Microsoft Research. He attended the first CSCW conference in 1986, co-chaired CSCW '98, and was program co-chair of CSCW 2012. He has participated in ECSCW as author and program committee member.
In periods of equilibrium, mutations occur beneath the surface. In the case of technology, the mutations are not random: Their internal organs are getting smaller, cheaper, and faster. The PC's external appearance may have changed little, but in the early 1980s the "box" was crammed with boards, whereas two decades later it held mostly empty space. From 1984 to 1987, the original Macintosh, the 512K Mac, the Mac Plus, and the Mac II were released; all looked much the same on the outside. Inside, memory increased from 128K to 512K to more than 1MB, and processor speed from 8MHz to 16MHz. These changes were crucial. The 1984 Mac was a commercial failure, with too little memory and speed to both handle a GUI and do real work. (If you are skeptical, look up Apple's stock-price history or reports from that time.) By 1986, the Mac had succeeded. Its new hardware could support products that arrived in 1985: laser writers, Aldus Pagemaker, and Microsoft Word and Excel for Macintosh. A few more cranks of Moore's Law were needed to launch the GUI era.
Is the parallel between biological and computer systems just a metaphor, or is there an underlying principle? Why don't fossil records change steadily? For example, why didn't PC boxes shrink gradually? Were efficient manufacturing supply chains for boards, boxes, and other components too costly to abandon? Had offices been designed with niches for PCs that size? Was it marketing, the size inspiring a sense of heft and value? Correspondingly, why do animal species seem to be static, although they too may be undergoing internal changes not apparent in the fossil record? A significant change in size might catch the attention of new predators and make hiding more difficult. It might frighten familiar prey, require new feeding habits, and so on. Only if changes bring substantial advantage can they overcome the drawbacks to abandoning a successful niche (or, environmental change could remove the niche itself). Being able to slip a computer into your pocketthat's cool enough to motivate a break from the past, even if it requires learning to type with your thumbs!
©2012 ACM 1072-5220/12/0900 $15.00
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 © 2012 ACM, Inc.