Books by Clifford Stoll

Silicon Snake Oil by Clifford Stoll

examines the computer networks that we have today, compares the actuality with the myths believed by many net-denizens and the hype presented to those who aren't connected yet, and considers the drawbacks and unintended consequences of the use of computers and networks. It is good to see a balanced treatment of such topics from someone who has been active upon Usenet and bulletin boards for years and cannot be dismissed as an ignorant Luddite. It is impassioned, carefully reasoned, and fun to read.

Stoll examines such topics as the culture of discussion nets (including both Usenet and local bulletin boards), how email compares with letters sent by post, the shallowness and narrowness of the information available online, and how schools, libraries and businesses are affected by the introduction of computers and networks. One underlying theme runs through every chapter: that contacts by network are, at best, no substitute for personal, face to face contacts. This should be obvious, but we often forget it in practice. He sets up this theme very memorably in his opening chapter, in which he contrasts the original Adventure text game with a speleological expedition which followed from a discussion about whether the cave descriptions in the game were geologically accurate.

He says little about the culture of Usenet that folks who have lurked there for a few months don't know already; he speaks of the rapid escalation of disagreement into flaming, of the periodic repetitiousness of nearly all newsgroups, and of the absurdly illogical newsgroup name hierarchy: "it's obvious that no librarian would ever develop such a freaky, byzantine system." He is fair, however; he also speaks of the more interesting, less flammable newsgroups, especially of mailing lists, and has high praise for local bulletin boards.

In comparing email to traditional postal letters, he points out a number of drawbacks to email and advantages of handwritten letters. For one thing, the effective bandwidth of letters is greater, because a handwritten letter is more expressive than plain ASCII. For another, the postal service is more robust than email forwarding software, less likely to lose a letter entirely because of a bad address, able to deliver a letter in spite of a slightly imperfect or incomplete address, and generally more helpful in explaining why an address did not work than the cryptic messages that most mail servers return in response to an email message with a bad address. Again, though, he is fair and points out the advantages of email - its usefulness for sending text that the recipient will need to incorporate into a document of his own, without the wasted time of retyping or risk of typos. And he qualifies his remarks on the greater expressiveness of handwritten letters by including an email message from a Bosnian refugee living in Croatia: "For a sufficiently powerful message, the medium is secondary."

He also comments upon the way email affects the way we communicate. People seem inclined to answer email and send off their response immediately upon reading a message. It discourages reflection upon what we have read and upon what we write, partly perhaps because we feel unconsciously that we have to write and read fast to match the speed of email transfer, partly because many people read and respond to email while logged on to their mail service. Handwritten (or word processed) letters, on the other hand, almost force reflection; we often re-read a letter just before we insert it in the envelope, and this gives us a chance to repent of a hasty reply or think of a better way to say something.

He speaks at length of how pressure to put computers into schools and libraries is not only diverting funds from more essential purposes, namely hiring teachers and librarians and buying books, but is distorting education and research. Aside from teaching the use of computers themselves, computers are not necessary for the learning of any subject, and are helpful in only a few. Most educational software is the expensive equivalent of flash cards or a multiple-choice quiz. It's a bit helpful in memorizing or reviewing facts, perhaps, but not for furthering real understanding.

He sees two main dangers that computer networks pose to libraries and research. The first is a diversion of attention to vaporware online libraries. Such online libraries can never be very extensive for a long while yet, partly because of copyright law and partly because digitizing and proofreading existing materials is time-consuming and expensive. But to the extent that they do exist, they tend to divert funding and attention from older libraries. And because they are quicker to use than existing libraries, they tend to distort research, narrowing it to a tunnel vision focused on the last fifteen years' periodicals and a few old public domain works. People often will use the information source that is easiest to use rather than the one that offers the most complete information. "Put something online -- anything -- and researchers will love it, whether or not it's right... There's a feeling that any answer is better than none."

The second is the damage done by computerized library catalogs, especially where they are intended to replace rather than supplement the paper card catalog. First, such databases cost money that could be better used for books. Would you rather have a computerized catalog with several terminals, or a few hundred more books in the library? Secondly, they are harder to learn to use than card catalogs. A card catalog may be slower than a computerized catalog, but it doesn't have downtime and the user interface is the same from library to library and from decade to decade. (I have used six different computerized catalog systems, three in county libraries and three in university libraries. Every one had a different user interface. Two of them were moderately easy to learn to use, but were correspondingly very limited in their ability to do complex boolean searches -- and if you can't do boolean searches, why computerize the catalog in the first place? The other four were somewhat more difficult -- not especially so for me, but undoubtedly offputting to many researchers who are less comfortable with computers or less experienced at learning user interfaces.) Because of the difficulty of using a computerized catalog, especially if the card catalog has been discarded or hasn't been updated since the computers were installed, many library users can't find what they want. Even at their best, and assuming that the patron is quick to learn the user interface, computerized catalogs are generally worse at subject listings than are card catalogs. The cross-referencing is rarely as complete as in an old, well-maintained card catalog, and many such systems don't allow a book to have more than a few subject headings. Such systems can be especially difficult for children. He tells of a study in which elementary and junior high school children were asked to look up several specific books. They were able to find more than two-thirds of what they were looking for in the card catalog, less than a fifth with the computers.

Silicon Snake Oil is very much worth reading. I borrowed it by interlibrary loan; I'm going to return it early, and suggest that the librarians read at least some of it before sending it back to the lending library. And I'm going to ask what they're planning for the card catalog when they finish entering all of the older books into the database. (10/95)

Follow-up: I talked with several of our librarians here at Clayton State about Silicon Snake Oil and Stoll's ideas. We now have a copy of it in the library; I strongly suggest that everyone involved with our "Information Technology Project" read at least part of it. The librarians haven't updated the card catalog in several years, and were still intending to do away with it the last time I asked them about it. Do we want this? How many times have you stood in line to use one of the computer terminals to look up a book? Think further back: how often did you have to wait for someone else to finish with the card drawer that contained the reference to the book you were looking for? Think about it.
(about summer 1997)

Follow-up to the follow-up: The librarians at Clayton State did throw out the card catalog after putting all the books in the electronic catalog. I've decorated my walls with cards retrieved from the recycle bins on the day they were thrown out (in September 1998).

%T Silicon Snake Oil: Second Thoughts on the Information Highway
%A Stoll, Clifford.  
%I Doubleday
%D 1995.
%G 0-385-41993-7
%P 234 pp.
%K schools education libraries research networks Usenet email

High Tech Heretic

In High Tech Heretic, Cliff Stoll takes the more important themes of Silicon Snake Oil and develops them more fully. About half the book details how limited education budgets are being wasted on unnecessary computers, network connections, and educational software. Attention is diverted from real life skills like writing clearly to the mechanics of a particular word processor for a particular system, from learning arithmetic tables that people need for making change and toting up one's checkbook to mechanics of using calculators or spreadsheets, and, perhaps worst of all, from real science in reasonably well-equipped physics, chemistry and biology labs to multimedia lab simulations.

The second half of the book consists of miscellaneous essays on the general theme of the hidden costs of computer use -- especially of people being pushed to use computers and software for things they aren't really that useful for. Some of them relate to libraries and how they become less useful as budgets are diverted from skilled librarian salaries, books, and periodicals to techie salaries, computers, and network connections. "New Uses for Your Old Computer" talks about how donating old computers to schools and other non-profits isn't that helpful - it may cost more in techie salaries to keep old computers running than to buy new ones. Stoll himself has turned his old Mac and PC cases into an aquarium and a litter box respectively.

If you heard Cliff Stoll talk during his 1995 Silicon Snake Oil tour, you might want to pick up High Tech Heretic just for the Bell Tower story - and other neat stories in the same vein. Both High Tech Heretic and Silicon Snake Oil are well worth reading, but if you aren't professionally involved with education, aren't a parent, and aren't enamored of Stoll's writing style (I am), maybe you can skip High Tech Heretic.

%T High Tech Heretic: Why Computers Don't Belong in the Classroom 
   and Other Reflections of a Computer Contrarian
%A Stoll, Clifford.  
%I Doubleday
%D 1999.
%G 0-385-48975-7
%P xv + 214 + 7 pp.
%K schools math science education libraries indexing all truth

Search on

To the Book Reviews page.
To my home page.
Email me: Jim Henry.

Get a GoStats hit counter