|
THE
REVIEW OF
COMMUNICATION Blaming Computers for Our Own Data DietLeonard Shyles
C. A. Bowers. Let Them Eat Data: How Computers
Affect Education, Cultural Diversity, and the Prospects of Ecological
Sustainability. Athens: University of Georgia Press, 2000. viii +
216 pages. Bibliography and Index. $40.00 (cloth); $18.95 (paper). Let Them Eat Data, by C. A. Bowers, is a vague polemic against the computer culture, written in a style Hegel and Marx would likely applaud. Unfortunately, the book fails to make its case, while sounding more definitive than it should. In short, there is “data” here, but not enough warrants to support the central claims. The title promises an assessment of the impact of computers on a broad array of human concerns including education, cultural diversity, ecology, and prospects for ecological sustainability. However, the book leaves many areas unresolved. For this reason, the wise reader will enjoy this book for the wealth of questions it raises rather than the answers it provides. The danger is that it may seem to the casual reader to be a valid critique of the issues it addresses; however, on closer analysis, its theses are asserted without support. Among the book’s strengths is its stunning review of the dangerous ecological practices of industrial corporations leading to inevitable ecological calamity, a trend not likely to abate any time soon. The book delineates clearly what should be viewed as a monumental scandal, namely, that so much raping of the world’s resources continues despite what we know about our interdependency with the environment. Is it really the position of environmental extremists to require that for every tree logging companies chop down, one be planted? Or that since the half-life of nuclear poisons and other chemical wastes in some cases exceeds nineteen thousand years, it might be in everyone’s long-term interest to curb their widespread production and use? Bowers presents a breath-taking list of serious environmental violations that threaten humanity’s long-term survival, including over-stressing our supplies of fresh water, fishing resources, and agricultural lands, a set of trends eventually destined to bring more widespread starvation than is currently in evidence. Global warming caused by the burning of nonrenewable fossil fuels threatens to make breathing conditions even more difficult as we move into the new millennium (and what good is unprecedented economic growth if no one can breathe?). Bowers makes clear that industrialized nations are now producing millions of tons of hazardous waste each year, and “disposing” of such toxins through export to developing countries. Cancers and other health risks are highly correlated with synthetic chemical use and waste disposal (including pesticides, herbicides, and nuclear waste). Yet the federal agencies charged with regulating chemical and pharmaceutical companies and other government agencies (the Department of Defense, for example) in order to protect the general welfare, continue to permit violators to profit from the production, distribution, and disposal of unarguably poisonous materials. Genetic engineering has begun creeping into our food chain while risks of that trend remain unknown. And the practice of feeding beef cattle bone meal combined with traditional grain feed is believed to have caused Mad Cow Disease, resulting in the slaughter of thousands of cattle in Europe, and the banning of beef exports, thus reducing that sphere of business across Europe. Perhaps government regulation banning the use of bone meal in such applications would actually be good for business. Such dangers to the public health are well documented, yet most were underway well before the personal computer won a ubiquitous place in homes and offices around the first world. But nowhere do any of these facts support the book’s central thesis, namely, that computers are responsible for exacerbating our current condition. Just what is the impact of computers on our Eco- and education systems, and on our consumer culture? While admittedly worthy of the asking in terms of appearing important and substantial, is this question well formed? Is it answerable? Bowers charges that computers do violence to local communities by disengaging users from their local context, while propelling them into a context-free zone. In Bowers’ words, in cyber-space, “there are no shared memories, moral responsibilities, or even known identities” (45). In agreement with Turkle and Rheingold, he asserts that cyber-space substitutes a decontextualized realm for culturally grounded narratives (45). That is, computer users become alienated and detached from their Gemeinschaft because they are using devices that ignore “cultural differences in the patterns of community life” (21). Bowers wonders what the impact is on “natural systems” of “a cultural lifestyle reinforced in cyberspace” (25). It appears audacious to assume such questions to be answerable—therefore, perhaps Bowers’ real objective is to be provocative, not definitive. It has long been popular in intellectual writing to consider the extent to which root metaphors subtly influence the scholarly enterprise, usually for the worse. Examples include the hegemony postulated by phenomenologists of the subject/object dichotomy suffusing critical thinking from Descartes to the present; or those analyzed so clearly by Stephen Pepper in his exquisite book World Hypotheses: formism, mechanism, contextualism, and organicism. Pepper’s definition of a dogmatist is relevant to how even the most rigorous critic’s habits of thought may be dulled by tacit presuppositions: dogmatists are those whose beliefs exceed their grounds for belief. Perhaps ideologues (Bowers included?) may be similarly afflicted. One position difficult to sustain in Let Them Eat Data is the false view that cyber-space is context-free. I suspect that computer programmers only wish that were so. One of the most insoluble problems software designers have had is programming software sensitive to a flexible event horizon. Instead what they have found is that there are ineluctable, irreducible cultural taken-for-granteds that software is incapable of avoiding; it is just this incapacity that limits its application—in short, software is brittle, not soft (for example, this software plays chess, not tennis; that program plays music, does not take food orders). And to the extent that it is limited in its application, users’ ability to relate to their own cultural context will not likely be supplanted by software. This is a key difference between human intelligence (HI) (flawed as it may be) and “artificial intelligence” (AI): frequently, human beings can better recognize the relevant aspects of a situation and discount the irrelevant ones in fashioning appropriate responses to sets of circumstances. In other words, compared to machines, people are better at understanding the event horizon. By contrast, computers are unsurpassed at performing singularly specific functions (like computing compound interest, for example), and they do their tasks more quickly and more accurately than human beings. Yet computers are colosally moronic at accommodating new contexts for which they have not been algoristically endowed by their creators. Terry Winograd and Fernando Flores recognized the problem of getting computers to behave in a context-free manner, attributing the difficulty to the impossibility of articulating completely all background assumptions in tasks requiring intelligent responses. For Winograd and Flores, the scope of comprehension remains severely limited for AI software. Hence, while computers may be very good in limited domains and closed contexts, AI quickly falters when its software is tested in open domains and broader (more fluid) contexts, where there is unpredictability in the shifting horizon or panorama of intention. The significance of this for software developers is the realization that the presence and pervasiveness of a shared background has major consequences for the design of computers and computer programs. As Winograd and Flores put it, if AI is conceived as an attempt to build a full account of human cognition into a formal system, thereby becoming context-free, the effort of articulation “can never be complete” (75). They add, “it is fruitless to search for a full formalization of the pre-understanding that underlies all thought and action” (99). I agree. As a humorous example to illustrate computers’ lack of sensitivity to background meanings in human situations (and therefore how context-bound software really is), I paraphrase Winograd and Flores, who offer the following scenario. Imagine a friend of yours returning from jogging on a hot day who, while sweating profusely, between gasps of air, asks, “Is there any water in the fridge?” You look up, notice his condition, and answer, ”Yeah. There’s some.” Your friend eagerly searches the refrigerator for water to quench his thirst only to find there is none available for such a purpose. He comes over to you and says: “I thought you said there was some water in the fridge. You must have been mistaken.” You look up again and reply, “I’m not mistaken. There are some molecules of water on the freezer coils” (105-107). This explanation strikes your friend as tactless, callous, and annoying. Yet that is precisely what a computer program might do if it were programmed to sense the presence of water, just as a thermostat might be used to turn on an air conditioner when the room reaches a pre-selected temperature, even when the occupant who has set it might contract a terrible case of the flu and need the room temperature to be much higher. It should be clear that computers are anything but context-free; in fact, the opposite is true. Computers are terrible at reacting appropriately to situations where the goal horizon among human agents is constantly shifting. This is because computers are tightly tied to the specific context for which they were designed. It is this very circumstance that convinced Wittgenstein to abandon his quest to find the atomic facts and objective starting points his theories in the Tractatus required, and argued instead that everyday situations are only meaningful in some context and for some purpose. In contrast to Bowers’ claims that computers threaten ecological sustainability, one must recognize that some of the effects of computers on the environment are directly positive. For example, it is clear that through telecommuting, workers avoid adding air pollution to the environment each time they do not drive their car to work. Further, the time expended doing actual work rather than just getting to work is arguably a boon to office efficiency. And, of course, the internal combustion engine has been spewing pollution into the atmosphere for decades before computers existed. From this perspective, there are obvious environmental benefits provided by computers as well as some savings of wear and tear on workers. It can easily be argued therefore that teleconferencing and email accrue to positive outcomes on the environment. However, some effects of the information explosion brought on by the Internet and computer technologies pose problems, especially now that such a blizzard of information is so readily available to the public. For example, it is now more difficult than ever to cut through the volumes of web pages, databases, and other information sources to find material relevant to a particular question a user might have. What strategies should one use to determine what is valid and what is junk? One ecologist I spoke to complained that she was less able to find the time to go out on field assignments because her computer databases were so crammed with information that she was now too busy reading reports to get out to do primary research. But she was not inclined to say that this was the fault of computers—rather it was her assessment that she needed to find ways of coping with the added burden, and that eventually she would adjust. In education, computers are emerging to serve the promising area of distance learning, a development I am convinced is utterly inevitable. For housebound learners, the handicapped, the elderly, and able-bodied students too far from a university to attend a physical campus, the digital telecommunications infrastructure offers entrée to levels of educational attainment never available before. But here it is the ethical responsibility of the users (both senders and receivers) to make the educational enterprise sound. The Internet and digital telecommunications infrastructures are connecting homes to audio-visual databases, for the first time enabling real-time interaction among teachers and students thousands of miles apart, and potentially offering immediate access to libraries and dialogue once available only to college students wealthy enough to afford to pay tuition and attend a physical campus. These developments have democratized education as never before. Some theorists (Bowers and I among them) have asserted that the interposing of such infrastructures into our educational system will inevitably redefine the role of teacher and student and even the meaning of the term “education.” However, of paramount concern is maintaining the quality of the education that is dispensed, even through shifts in the meaning of fundamental terms. Bowers is concerned that the values and lessons of the local community will be put at risk or lost by the new arrangements. However, it is important to realize that it is in the best interests of both the software designer and intended users that the materials fabricated for such functions be crafted in line with the cultural values and needs of the target audience if they are to succeed. If done properly, the promise of distance education can bring great rewards to those willing to use its gifts. Bowers is right to note the potential risks cyber-space brings to traditional and tenured professors who reside in the safety of the Ivy-walled academy. All teachers should recognize the risks of having their lessons cherry-picked by software companies. But how can they guard against becoming irrelevant in the brave new world of the cyber-school? Some suspect that the traditional university will not be replaced by the new order but will merely be supplemented by it. In my view, it will be a long time before interfacing with a screen will become a functional alternative for live face-to-face interaction. The intimacy is just not there. But some courses of study and some student-types do not require the traditional arrangement, and in these cases, replacement to a greater degree may occur. Bowers critiques cyber-schooling in an analysis of several specific software programs (Storybook Weaver, DynoPark Tycoon, and SimEarth). However, his critique does not do justice to the threat or promise of distance education as a whole. This is because distance education depends on the telecommunications network and not just packaged programs. Further, there are hundreds of educational software programs available, with more being developed each year. As a result, unfortunately this section on the false promises of computers in the educational setting presents too myopic a view of educational software to be taken seriously. In summary, Let
Them Eat Data provokes the reader’s thinking about the impact of
computers on the worlds of work, education, culture, and the
environment. But it does not support convincingly the positions it
stakes out. For readers looking for clarity in the arguments presented
and the conclusions reached in the areas broached, the book may be
unsatisfying. Leonard Shyles is associate professor of
communication at Villanova University. REFERENCES
Pepper, S. C. (1970). World Hypotheses. Berkeley: University of California Press. Winograd, T.; Flores, F. (1986). Understanding Computers and Cognition: A New Foundation for Design. Newark, NJ: Ablex.
|