Silicon Snake Oil
by Clifford Stoll
Reviewed by Mike Eisenberg (March, 1999)
Alongside the history of technological enthusiasm, there is an almost equally vigorous parallel history of technological critique. The tradition is as old as Plato, who wrote disparagingly of the value of the written word (note the irony there!) in the Phaedrus; and it has continued in the writings of (among many others) William Blake, Henry David Thoreau, Mary Shelley, and Lewis Mumford. The tradition has, if anything, flourished in the age of the computer. Indeed, to speak of "the computer" is too gross an abstraction: the instrument has presented a variety of different looks to the public over the past fifty years, and each new look-electronic brain, office leviathan, video game screen, window to the Internet-has spawned its own round of critique. Much of this writing-like Postman's Technopoly, or Slouka's War of the Worlds, is provocative and disturbing; some, like Weizenbaum's Computer Power and Human Reason or Roszak's The Cult of Information, is still provocative, though annoying. Clifford Stoll's book Silicon Snake Oil is, unfortunately, not even good enough to be annoying. It is essentially a 200-plus page top-of-the-head ramble on way too many subjects-the Internet, computers in the classroom, electronic card catalogs, email, cryptography-and the result is an intermittently informative hodgepodge with no coherence, no narrative path, and almost no sense of literature or history. When Stoll mentions, at the close of the book, that Silicon Snake Oil grew from scribbled notes on the back of an envelope, that admission explains a lot.
Stoll's basic mode of argument throughout the book is to compare two functionally-related technologies-one computational (e.g., email) and one less so (e.g., the U.S. postal service), and to point out several positive aspects of the latter that are missing in the former. For instance, in the case of email-vs.-the post office, Stoll argues that the post office allows a variety of unorthodox styles of addressing (e.g., an envelope with a picture of Alfred E. Neuman will be delivered to Mad Magazine); the post office is more reliable and faster than people generally give it credit for; physical letters have a marvelous sentimental value; and so on. There are similar comparisons of libraries with and without computerized card catalogs; darkrooms versus Photoshop; and typewriters versus word processors. Just to take the last of these as yet another example: Stoll writes of his Sears mechanical typewriter that
It came with a dustcover, a spare ribbon, and a cute bottle of oil. For a hundred and fifty dollars, it's entirely adequate for business applications. It delivers instant hard copy, never crashes, handles envelopes with ease, and anyone can adjust the margins. The sixteen-page instruction booklet includes diagrams of where to place your fingers, how to tab, and a drawing of correct posture. The carrying case even has handles.
There are several things that need to be said about this style of argument. First, it is hilariously easy-you can do a comparison of this sort to show, say, that the Ptolemaic picture of the solar system is preferable to the Copernican (the former provides a healthier sense of confidence in the importance of humankind, it trains young minds in thinking about the geometric intricacies of epicycles, and it's surprisingly useful in navigation). Or, if you like, you can show that the harpsichord is preferable to the piano (the former has a much statelier sound to it, and harpsichord music sounds all wrong when played on the piano). Or... well, you can try this yourself. The point is not that there aren't interesting things to say about any of these comparisons, but rather that Stoll's reflexive use of the "set-difference" technique-find whatever is in technology A that is not in technology B, and bemoan its loss-doesn't constitute much more than the beginning of an analysis.
The other thing that is strange about Stoll's use of this type of argument is his shifting view of what counts as "technology" versus "real life", or as "authentic" versus "inauthentic." This causes some oddly (and presumably unintentionally) ironic passages: Stoll mentions "the warmth of a voice across the telephone line" [p. 79] and states that "it's almost always faster to fax a single-page letter than to send Internet e-mail." [p. 16] As the book proceeds, and the examples proliferate, the reader may well experience a growing difficulty in remembering which technologies Stoll approves of, and which he doesn't, and which he is sort of ambivalent about. Certainly, many of his contrasts between computer-mediated experience and "real life" sound like they could equally well be made about reading and "real life". After all, the stereotypical bookworm is someone who has a dearth of physical experiences (he or she is more likely to read about the constellation Orion than to go outside at night and observe it); the too-avid reader is missing out on a variety of warm human interactions; much of the material printed in books is outdated, poorly written, mistaken, or even pernicious. Yet Stoll has (as far as I remember) only good things to say about the abstract symbols that appear in books and few good things to say about the abstract symbols that happen to appear on screens.
A corollary of the book's haphazard structure and style is that almost all of it is polemic, unburdened by any research beyond brief introspection. Thus, when Stoll writes sentences like:
"Computer networks isolate us from one another, rather than bring us together." [p. 58]
"Computers teach us to withdraw, to retreat into the warm comfort of their false reality." [p. 137]
"[C]omputers punish [those who are] imaginative and inventive by constraining them to prescribed channels of thought and action."[p. 46]
he offers no foundation in research, no explanation based in psychology or physiology or any relevant field, and no thoughtful comparison with what we might know of other technologies (do computers "teach us to withdraw" more than, say, books or movies?). This style of argument-by-fiat is, unfortunately, unlikely to do anything more than preach to the converted.
Silicon Snake Oil does have its better moments. Like a widely scattered fog that deposits a fine mist over everything in its path, the book occasionally does light upon something new and interesting. The discussion of the shifting formats of databases, for example, and the resulting difficulty of retrieving stored information, is the beginning of an important argument. And Stoll's description of the impact of computer databases on the practice of scientific research-he is an astronomer by profession-is especially useful; when he writes about the limitations of computer models, he's clearly drawing on a long professional history of reflection on the subject. Even here, though, it's a bit frustrating to see how brief and almost exclusively personalized the discussion of the subject is. (A book detailing the impact of computer models, for better or worse, on the practice of scientific research over the last two decades would be a much harder read than Stoll's book, but a much more rewarding one.)
Besides these occasional high spots, though, Silicon Snake Oil offers little sustenance. The style of writing is (to this eye) marred by the self-consciously folksy tone (the use of "'em", as in "give 'em hell" is overdone, and one "Yowsa" is enough for any book). The bibliography is strangely perfunctory, listing only a few book titles without publishers or dates. Indeed, the few books that are mentioned are all over the place-why are we pointed to Postman's Amusing Ourselves to Death and Mander's Four Arguments for the Elimination of Television, both of which are about television and not about computers? And why not (say) Weizenbaum, or Turkle, or Slouka, or Birkerts, or Cuban? And the less said about chapter 12 1/2, the better; let's just say that it would have been a better decision to leave it out altogether.
The various and protean forms of the computer need their technological critics. We need a Plato to look at the Internet (where it is and where it's headed); databases and their fragility; the impact of computational media on scientific thinking; the impact of email-writing on children's literacy and on their social styles; the use of computer simulations in classrooms; and many more. Stoll's book doesn't meet that need.