Bonnie Nardi, Ravi Vatrapu, Torkil Clemmensen
Comparative informatics (CI) is the application of the comparative method to the study of information and communication technologies (ICTs) across diverse contexts . CI systematically examines similarities and differences in the ICT life cycledesign, development, deployment, adoption, use, impact, and evaluationin contexts including cultures, regions, nations, generations, socioeconomic classes, gender, organizations, and technologies. The objective is to generate nuanced, critical understandings of technology in human life in the world we inhabit together. The analytical aim is not laws or generalization, but heightened awareness of the uneven surface of ICT practices locally and globally. This surface is often taken unproblematically to be generically Euro-American in character, or even simply American.
CI is motivated in part by a practical challenge: Any company designing for multiple countries and cultures must confront difference. Localization is the typical response, but often localization efforts touch only the most superficial contours of migrating technologies, leaving in their wake underserved users struggling to adapt to powerful but misaligned technologies. Understanding when difference matters for design (“the difference that makes a difference,” to use anthropologist Gregory Bateson’s useful phrase) is at the heart of CI.
We believe CI should center on diligent, critical, empirical analysiswhat we call grounded comparison. To illustrate grounded comparison, we present a study of regional styles in word processing and an ethnographic investigation of online work groups. Both examples indicate the breadth of methodological approaches appropriate to comparative analysis, which also encompass traditional laboratory studies .
Grounded Comparison #1: A Comparative Case Study of Word Processing
We begin with recent work by Clemmensen , turning our attention to the “white rat” of human-computer interaction: word processing. Invented in California in the late 1970s, computer-based word processing rapidly attained status as a ubiquitous technology. Much of the founding research of HCI is concerned with this office-work task . Given its basis as an anchor of conceptual development in HCI, word processing might be expected to be a technology with few or insignificant differences in styles of usage across the world. However, when we compare how word processing is developed and used across regions, important and sometimes subtle differences in HCI appear.
National differences do not by themselves necessitate deep differences in the design of technology; such styles are themselves changed by new technologies. New forms of human-computer interaction may emerge, or are required . The question is: Will normal work processes and organizational changes, with sufficient time and effort, successfully overcome disruptions in the task-artifact cycle, or will forms of HCI emerge as negative and controversial?
Word processing in Japanese and Hindi. In HCI, the case of the introduction of the Japanese word processor is quite famous. Early on, Ito and Nakakoji pointed out that although the design of most American word-processing software was based on the typewriter metaphor, the use of this metaphor required a major shift in practice to print characters for Japanese people . Most used hand printing until word processors became available; typewriters were not common in Japanese society. Japanese handwriting was accomplished within a 20×20 grid, written from top to bottom, one column at a time, and right to left. Tab stops were unfamiliar, and document length was measured in characters, not words.
We can only guess how word processing might have looked had it been developed in the context of the Japanese writing grid. An age-old tradition was destroyed by the introduction of new technology. At this point, the Japanese do not desire a different kind of word processor; they are used to the typewriter-look-alike word processor . However, hybrid forms of word processing have emerged. MS Word (Japanese version) and Ichitaro (a Japanese word processor) allow Japanese writers to type vertically as well as horizontally, according to professor Masaaki Kurosu at The Open University of Japan . Today most academic papers are written horizontally, because such texts use alphanumeric words, except for some social science and literature papers. Newspapers and comics use the vertical format, simply because of tradition. The grid is used for hand writing, which is sometimes used by writers; Ichitaro displays that grid for special purposes. The grid is not widely used today, because horizontal writing is the default for many word processors.
The changes that word-processing technology brings about in non-Latinate language cultures may be difficult to assess and sometimes surprising. Katre compared the number of keystrokes necessary to write the same word in different languages and on different mobile phones , showing in one case that it requires 38 (!) key presses to enter a specific sign in Hindi on a mobile phone. We can hardly imagine how this shapes Indian users’ approaches to short text messaging in Hindi. Similar examples can be found in Indian PC word processing .
Word processing in Danish. Less obvious, but just as important, are discontinuities within a Euro-American cultural ambit. Countries that are supposed to share cultural values and practices with the U.S. include European countries like Denmark. However, in the design of word-processing technology, this doesn’t hold true. For example, a Danish blogger shares his joy in using his new smartphone; life has become much easier, as he can now process email more quickly. Among the many input methods, he prefers the PC-like keyboard. This keyboard, however, is English, without the Danish vowels æ, ø, and å. The blogger would have to switch to a Danish-language keyboard to key the Danish letters. Not wanting to do that, he has developed his own special solution:
[To switch to the Danish language keyboard] I do not want to waste efforts to do that, when I merely want to send a short message. So I practice writing without using words that include æ, ø or å. In fact it is really fun, when you have a positive attitude towards the task. It sharpens awareness of language when you have to throw away sentences because they include a word that has æ, ø or å in it 
The blogger, remarkably, adapted to his smartphone by “throwing away” hundreds of common words in his native tongue to avoid “wasted effort.”
Similarly, in a blog posting for a travelers club, Danish traveler Anders Andersen begins with an example in Danish: “Kaere Aase En lille hilsen fra Oestrig. Haaber du nyder Aegypten…”(Translation: “Dear Aase, A little greeting from Austria. Hope you enjoy Egypt…”) and uses it as an illustration of the fact that many Danish people recognize the old-fashioned characters used in the sentence, though they never use them in everyday writing . He then proceeds to explain it is possible to write æ, ø and å on computers with a keyboard lacking these lettersby using the numeric code for each letter, holding down the alt key while keying the code for the letters (see Figure 3). For example, press and hold “alt” while keying 0230 on the numeric keypad, and the letter æ will appear on the screen.
He further writes: “...it may sound a little troublesome, but in fact it is quite easy and after a couple of letters, you will remember these numbers in your head. You will write nearly as fast as when you used ae, oe, and aa, and it does indeed look much better.”
This is not the whole story; Andersen explains how to find the alt key and how to make sure the numeric keyboard is set to be numeric, and not navigation, and why entering zero as the first letter of the code is necessary to make the identification of the letter unambiguous. One could ask, which is more old-fashioned, using letter-writing forms that were outdated by the Danish language revision in 1948 (made publically available in Retskrivningsordbog udgivet af Dansk Sprognœvn 1955), or punching in codes like punching holes in a punch card? In both cases, California-bred word-processing technology transported to Denmark sets Danish word processing back 50 years.
Microsoft Word is widely used in Denmark. Both the user interface and the help system are localized to the Danish language. But Danish-language users of Word are left with a reduced version of the software compared with English language users. Up until recently, what Danish users could not doand American users could doincluded summarizing reports by using the software’s built-in auto-summarizing functions, translating words from Danish to other languages such as German or English, and dictating words into the word processor in their own language. While some Danish-language users will never discover what they do not get, Danish-language users who also write in English may become aware of these limitations in the Danish versions of Word.
These examples of declinations of Danish writing practice suggest the limitations and additional resources required to use local versions of global technologies. We believe such local differences can be appreciated through comparative studies to develop our critical sensibilities as designers and users.
Grounded Comparison #2: An Ethnographic Study of Online Workgroups
A second example is Kow and Nardi’s discussion of the rhetoric and values of an influential socio-technical paradigm: Web 2.0. With its celebration of sharing, interoperability, user-generated content, and collaboration , Web 2.0 has a normative structuring that presumes and codifies values so apparently sensible and humane it seems they scarcely need be examined. But empirical inquiry suggests Web 2.0 constructs are contingent and variable across nations and cultures in ways that have not been appreciated.
Comparative informatics positions itself relative to certain current efforts as open, broad, and inclusive in its analytic elements. We believe that openness and breadth are imperative at this stage of development.
In American and European contexts, online “sharing,” “user-generated content,” and “collaboration” evoke a scenario such as the development of Wikipedia, wherein members who actively contribute and adhere to established, publicly stated community norms are promoted to positions of increasing responsibility . Wikipedia is a meritocracy of independent, self-motivated contributors, democratic in spirit.
By contrast, online workgroups in China may emerge and operate quite differently . Chinese “modders” (people who write software extensions or “modifications” to commercial video games) share many of the challenges as Wikipedians: producing content, maintaining communication, creating a workable organization. But Chinese modding workgroups are organized around a charismatic leader who recruits friends to join him in developing mods. He often knows recruits in real life and socializes with them offline. While others join the group in online venues such as chat rooms, the core team, as it is known, is central to the functioning and success of the modding workgroup. The core team is small, exclusive, and tightly connected. It produces most of the mods, controls the agenda, and provides friendship and companionship. Core team members may go out to dinner and socialize informally. If they are from different locales, they gather for a festive meal when a member happens to be in town. They maintain private channels of online communication not open to the rank and file. Members of Chinese modding workgroups do not ascend to more important roles simply through hard work, as Wikipedians do; they must be on the core team leader’s radar.
Core teams are not peculiar to modding communities. The term “core team,” , is used pervasively in China to denote a group of key members in an online community or at a company. The term is frequently heard on Chinese television and seen in advertisements (e.g., restaurants advertise special rooms for core-team business dinners). Modders’ core teams are an adaptation of a strategic organizing principle found throughout Chinese society.
Positioning Comparative Informatics
CI positions itself relative to certain current efforts as open, broad, and inclusive in its analytic elements. While this generalized approach is at a disadvantage in generating tightly focused conversations stabilized around highly stylized categories of discourse, we believe that openness and breadthgeographically, historically, and intellectuallyare imperative at this stage of development.
CI aims to create and promote a global program of inquiry in which the analytic gaze may come to rest on any IT practice anywhere. This aim is particularly germane to research framed in a critical perspective in which relations of dominance and authority arise within and across First World localesas well as in the developing worldand in research in which powerful institutional actors impel IT practice. For this reason, postcolonial perspectives, while useful, are not exhaustive; other cultural and historical circumstances and events may be of more importance, such as current global political activity.
CI departs from approaches such as Hofstede’s, in which a small set of putatively stable cultural differences organize analysis . The translation and transportation of culture theory in other academic disciplines to ICT settings create a perpetual loop of self-reference and self-inference rather than advancing understanding . The purview of CI is larger than a discrete set of differences, including but going beyond culture to other differences, even differences that arise from a single individual, as with the Danish bloggers. CI is, furthermore, oriented toward change, regarding it as inevitable in IT practice.
CI is sympathetic to popular notions of appropriation and domestication of technology that permeates the literature, but at the same time, CI’s commitment to interrogating and conducting design, and accounting for powerful corporate and governmental institutions enmeshed in design activity, is paramount. In particular, providing academic resistance/dialogues to corporate and governmental interests is a key facet of CI’s critical attitude. End users and consumers constitute elements of a complex system to be apprehended through a logic in which institutional forces remain visible and potent.
Grounded comparisons have clear design implications and should shape discourse that shapes design, even if comparisons do not indicate immediate business strategies. A goal of research is to open novel vistas of thought and to populate them with rich descriptive and analytic materials that carry the capacity to inform, provoke, and edify discussions in arenas of practical activity. CI can admirably serve this role by underwriting reflection and analysis, heading off scenarios in which we are liable to uncritically export what is taken for granted in the West (or the U.S. or California) to the rest of the world, lacking sensitivity to geographies and other dimensions of difference. By the same token, just as universal design in architecture and interior design emerged from the specificities of disability, grounded comparison potentiates the discovery of new “universals” as it reports and reveals precise, distinctive human experience in diverse arenas, experiences to tactically appropriate in multiple contexts.
While the idea of comparison is hardly new, by putting a name and a face to it and deliberately placing it within the warrant of informatics, we hope to promote and systematize comparison as an essential element of our practice. We desire to build a community of scholars and practitioners laboring together to forge denser, more global networks of collaboration in order to animate and transform what we regard as necessary comparative inquiry in informatics.
The First International Workshop on Comparative Informatics was convened at the Copenhagen Business School August 2223, 2010. The Workshop was supported by a grant to Ravi Vatrapu under the First International Network Programme of the Danish Ministry of Science, Technology and Innovation. The Workshop brought together an international group of ICT researchers to develop a CI agenda. Participants were: Ravi Vatrapu, Torkil Clemmensen, and Janni Nielsen of the Copenhagen Business School; Ellen Christiansen of Aalborg University; Ruy Cervantes and Bonnie Nardi of the University of California, Irvine; Mara Miller and Scott Robertson of the University of Hawaii at Manoa; Zhengjie Liu of the Dalian Maritime University, China; and Yair Amichai-Hamburger of the Interdisciplinary Center, Herzliya, Israel.
1. Vatrapu, R. Explaining culture: An outline of a theory of socio-technical interactions. Proc. of the Third international Conference on Intercultural Collaboration (Copenhagen, Denmark, Aug. 1920). ACM, New York, 2010, 111120.
3. Clemmensen, T. Regional styles of human-computer interaction. Proc. of the Third International Conference on Intercultural Collaboration (Copenhagen, Denmark, Aug. 1920). ACM, New York, 2010, 219222.
9. Gupta, R. Indian language support in Microsoft products, 2004; http://tdil.mit.gov.in/Oct_2004/ILSMP-11.pdf/
11. Andersen, A. Æ, Ø og Å på udanske computere De berejstes klub, 2005; http://www.berejst.dk/209/Tastatur.htm/
13. Bryant, S., Forte, A. and Bruckman, A. Becoming Wikipedian: Transformation of participation in a collaborative online encyclopedia. Proc.of the 2005 international ACM SIGGROUP Conference on Supporting Group Work, GROUP’05. (Sanibel Island, FL, Nov. 69). ACM, New York, 2005, 110.
Bonnie Nardi is an anthropologist in the School of Information and Computer Science at the University of California, Irvine. Her research concerns online computer games, ethnographic methods, activity theory, and comparative informatics.
Ravi Vatrapu is an associate professor of Applied ICT and director of the Computational Social Science Laboratory at the Copenhagen Business School. His current research focuses on theory-based empirical studies of socio-technical affordances.
Torkil Clemmensen is an associate professor in the Department of Informatics, Copenhagen Business School, Denmark. His research includes interests in virtual communities and cultural-cognitive perspectives on user representations.
©2011 ACM 1072-5220/11/0300 $10.00
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 © 2011 ACM, Inc.