More notes on economics, language, religion, the media, childhood, democracy, and tools for hastening the decline of command-and-control computing. The materials for Matt Kirschenbaum's course on "Literary Narrative in an Information Age" are at http://www.engl.virginia.edu/courses/enlt248/s97/1 http://www.liszt.com is a large index of Internet mailing lists. http://www.hw.ac.uk/libWWW/irn/archive/a_z.html is one huge file of links to online publications of various organizations. It's unwieldy but good for exploring. I sent out a brief note about legal issues in digital signatures the other day; a much longer treatment by Michael Froomkin can be found on the Web at http://www.law.miami.edu/~froomkin/articles/trusted.htm The Digital Citizens Foundation (Netherlands) is at http://www.db.nl They have some useful links, e.g., on European information politics. New and recommended: Kwame Anthony Appiah and Henry Louis Gates, Jr., eds, The Dictionary of Global Culture, Knopf, 1997. This is an excellent bathroom book, a needed complement to the narrow guides to cultural literacy that have become established in recent years. Well written and cosmopolitan. I also recommend Vivian Paley's book "Boys and Girls: Superheroes in the Doll Corner" (University of Chicago Press, 1984). It is a fantastically observant account of kindergarten children by a celebrated teacher. I used this book in an introductory class for graduate students because it illustrates a form of academic writing that I wish would come back: although clearly motivated by several layers of theory about the psychology of gender differentiation, it is written in plain English and allows its theoretical claims to emerge in increments solely through the force of its narratives. One more recommendation, unfortunately not as accessible as the others: Susan Leigh Star and Karen Ruhleder, Steps toward an ecology of infrastructure: Design and access for large information spaces, Information Systems Research 7(1), 1996, pages 111-134. This paper describes an infrastructure that was built for a community of biologists. Its goal is to trouble the concept of infrastructure; it suggests that something is only "infrastructure" when it successfully fades into the background, and that this successful fading has many and varied conditions. A next step, I think, is to open up the dialogue between their interactional and phenomenological view of infrastructure and the economic and technical views. Analyses like Thomas Hughes' "Networks of Power" (an excellent book on the rise of the infrastructure of electrical distribution) provide a bridge. Liam Bannon reminded me of two more excellent and influential resources for those working to overcome command-and-control computing: Peter Naur, Computing: A Human Activity, Reading, MA: Addison-Wesley, 1992. Christiane Floyd et al, eds, Software Development and Reality Construction, Berlin: Springer-Verlag, 1992. Liam's own papers are also a good resource. Some of them can be found on the Web at http://www.ul.ie/~idc/library/papersreports/LiamBannon/LiamB.html Paul Krugman's rather nasty review of my hero Bill Greider's new book can be found at the Death Star, http://www.slate.com/dismal/97-01-23/dismal.asp In the way of many economists, Krugman raises a vital issue but then makes it seem much clearer than it really is. The argument goes: if technology throws 20,000,000 people out of work then we should rejoice because that means stuff is being produced more efficiently and so people can consume more stuff and that means that more jobs will be created to produce it all. Historically, as a grand general average, that's sort of what happens, and even though it's a lot more complicated than that, any author who issues warnings about the job-destructive effects of technology should certainly answer such arguments. The issue is particularly crucial for Jeremy Rifkin's book "The End of Work", which really does predict massive economic collapse from technology-driven unemployment. Rifkin's answer is that the number of jobs that are currently about to be axed is vastly greater than any historically precedented rate of job-creation could possibly compensate for. Maybe, maybe not. In any case, Greider's book is a lot more sophisticated than Rifkin's. Greider does not adequately answer Krugman's challenge, but Krugman (in my opinion) also grossly oversimplifies Greider's argument. On the general subject of economics, I neglected to mention that the text on path dependence that I sent out came from eh.net, the online resource for economic historians, http://www.eh.net/ An important subtext here is the difference between economic theorists and economic historians. Economic theorists, whose style of thinking is on display in Paul Krugman's review, place immense faith in the internally consistent logic of their theories. This logic is undeniably compelling at times, but then so is cocaine. Economic historians, on the other hand, operate under norms of humility that derive from the qualitative complexity of the phenomena they study. In my opinion, the very best economists are people like Paul David and Nathan Rosenberg who combine the two perspectives. Not surprisingly, these are also the people who are most impressed with the phenomenon of path dependence. Path dependence threatens the mainstream economic theorists because it tends to invalidate the most basic assumptions of their mathematics. This math was originally explicitly derived from physics. Its core concept is equilibrium; the clearing of the price system is treated as a sort of resultant of forces. Sure, you can say that a separate equilibrium exists within each possible path, but in doing so you've surrendered the whole game: the selection of paths, rather than the finding of equilibria, becomes the dominant explanatory principle. The bottom line, then, is this: if you think that equilibrium is the best concept for understanding information technology then you're in the camp with the mainstream economic theorists. If you think that some other concept might be needed then you're in the camp of the economic historians. The problem, of course, is that most of the grand claims that are made for markets depend on the concepts and results of conventional economic theory. Comical outcomes are obtained when these claims are applied mechanically to information technology, such as the assertion that Microsoft Windows must be the best personal computer operating system because it has the largest market share. The way it has evolved historically, computer system design has had a corrosive effect on language. Here is one common pattern: computer people draw on the vernacular lexicon for their technical terms, taking words that once referred to human activities and redefining them to refer to operations on computers, and then they turn around and pretend that the new technical terms actually refer to human activities. As an example, take the word "search". When a person -- a scholar, perhaps, or an entrepreneur -- searches for information, lots of complicated things can happen. The search can last moments or years; it can involve numerous media; it can involve other people in several different kinds of relationships; it can involve judgements of trust and risk; and so forth. For computer people, however, "search" names a command to a database program. This command might take three arguments, or else a Boolean expression whose predicates correspond to keys in the database. In such a context, the human activity of "searching" will now be narrowed to refer simply to the act of using that single computer command. As a result, most of the variety and complexity and social embedding of actual searching will vanish. It will almost literally become unthinkable. Nobody will admit that the word has changed its meaning, so the idea that something has vanished will itself become almost unthinkable, and someone who proposes to study the full social complexity of searching will be thought soft-headed, irrelevant, or weird. Fortunately, in the case of "search", some people -- librarians such as Marcia Bates and Chris Borgman -- have been willing to suffer those sorts of slings and arrows. As a result, something of the full human reality of searching has been brought back into the light. Their work is heroic in several ways, including the cognitive effort of pulling loose from computer-centered forms of language, so that people can once again become the center of our thinking. (For surveys, see Hewins in the 1990 issue of the Annual Review of Information Science and Technology and Star and Bishop in the 1996 issue.) Librarians are fiercely loyal to their patrons, and they pay a great price in professional prestige for it. This sucks. Yet the intellectual work they have begun is a crucial start toward recovering the aspects of human experience that technical language has obscured. How can we take this development further? I have a conjecture: recentering our thinking on people not computers requires more than studying people, more than making sure that user studies influence system requirements. It requires a separate intellectual center of mass: beyond just defining our studies in terms of "computer users" or "information seekers", it is important to define our studies in terms of some category of human life -- for example, museums as such, organizations as such, families as such, or even documents as such. Forget the computers for a while and study one of these phenomena at some other level. Let its full concrete detail emerge, and listen to what it says. Having built up some analytical weight on top of those specifically human categories, it becomes possible to negotiate between equals: our ideas about people and our ideas about computers will be equally powerful, and each side will be able to stand its ground in the negotiation through which technology is fitted to institutions and vice versa. So long as we automatically redefine the categories of human activities in terms of the mechanisms we happen to have available, it will remain impossible even to conceive of honest requirements specifications for the systems we build. This is the deep meaning of command-and-control computing, whose first step is to redefine the language in terms that fit with the mechanisms of command and control. In order to overcome command-and-control computing and its many pathologies, we should establish what Mike Robinson calls "double level languages" -- languages that draw a clear distinction between the vocabulary of the user community and the vocabulary of the formalism. That way it becomes possible to discuss and negotiate the frequently complex and shifting relationships between the two, as well as keeping track of what parts of experience and reality are lost -- as they will inevitably be -- in the transition from vernacular language to formalism. Wish list: Recently I had to secure permission to reprint figures from several books, some of them recent and others from the 1940's. It was an incredible pain in the butt. For one of the older publications, it proved impossible to identify the current copyright holder, since the original publisher was defunct and no successor firm could be identified. For one of the more recent publications, four branches of a global publishing conglomerate kept disagreeing about which one held the copyright, and they sent us in circles until a gruff but kindly gentleman in New York finally agreed to okay our request on his own authority because it was so trivial (which it really was). Another publisher warned of a 7-to-8 week delay unless we paid $50 for the two week "express service". In other words, they would sit on our request for 7-to-8 weeks unless we paid them the money. The whole process was insanely inefficient. I gave up wasting my own time after a while and hired a graduate student at $10 an hour (the customary rate for such things) to make all the phone calls; she arrived at my office at 5PM one morning to make several calls to London, then prepared and sent faxes to those people, after which it was time to make calls to New York and prepare and send faxes to those people, after which it was time to call San Francisco and send a fax to someone there as well. Heaven knows what the phone bills will look like, not to mention all the time spent by the people at the other end of the line. This business is screaming for automation. Some bits of automation do exist; a couple of publishers have complicated voice-mail systems that explain exactly where to address your fax, what to include in it, etc. But surely it is in everybody's interest to pitch in and create an Internet-based standard for processing reprint permissions, at least for the easy cases. Web pages for each publisher with their needs, Web forms to fill out, a catalog of old publishers and their successor firms, that sort of thing. Ideally such a system would connect to an actual catalog of the books with their ISBN's. Given such an interface and database, each publisher can implement its own policies. Most scientific and technical publishers routinely grant free permission to reprint single figures from their publications. A publisher could implement an expert system to determine the easy cases, and only the harder or more ambiguous cases would be passed along to a human being. If money is needed then it can be collected with digital cash or some similar scheme. None of this is very difficult technically. As with most Internet fantasies, the real trouble is institutional: how to get the ball rolling to agree on the necessary standard. Publishers need these standards for many reasons, not just for processing copyright permissions, and in my optimistic moments I like to think that they understand the urgency of such things. The 2/3/97 issue of The New Yorker contains (pp 41-42) the following sentence: If society's susceptibility to misinformation is like AIDS, then Web sites and Internet news groups and electronic bulletin boards -- a vast, thrilling, promiscuous commingling of facts with fabrications -- could be its bathhouses. I am not making this up. It appears in an article by one Kurt Andersen on the putative epidemic of bad facts in society, the supposed result of a "laissez- faire ultra-populism [that] finds its perfect medium in the Internet". His method is distilled to perfection in the following paragraph, which deserves to be read with full stops after every sentence. (The context is that he has reminded us about Pierre Salinger's TWA 800 scoop off the net.) Last summer's series in the San Jose Mercury News on the links between the CIA and Nicaraguan cocaine wholesalers is a very different case. The newspaper story was extensively reported and was probably true in many of its basic facts -- that is, some Contras were probably involved in the cocaine trade, and people on the CIA payroll probably looked the other way. As with Salinger, however, it posited an unholy federal conspiracy. And, as with Flight 800 fabulism, the story swelled with populist affirmation while circulating via the Internet and talk radio, especially among black Americans. In that parallel universe, where rumors of fantastic plots by government (a satanic Ronald Wilson Reagan -- six letters, six letters, six letters -- had crack invented as a genocidal weapon against blacks) are widely credited, a crack-dealing CIA can seem not just plausible but obvious. The paragraph condenses the well-tested methods for assimilating the SJMN's series to the alleged psychosis of black culture. The third sentence, with its unspecified "unholy federal conspiracy", is the pivot. You've heard the argument so I won't belabor it. Why is the New Yorker wasting space on such derivative stuff? The New Yorker is doing its job: making the media echoes into history by giving them the seal of established truth. (One weird note does intrude: "Disney had just bought the film rights to the CIA-Contra crack articles". I can't quite get my mind around this.) My point, once again, is not that reporters are evil. I have worked with scores of reporters, and the vast majority have been serious people. I'm in a good position, of course, in that my opinions on issues like privacy are popular among both reporters and their readers; I am also positioned as an expert rather than a combatant so I needn't worry as much about being slammed. Still, I am generally happy with the reporting in my area of expertise, given the institutional constraints that the reporters are up against. The point, rather, is that reporters participate in a larger system whose dynamics are largely beyond their control. Agendas arise in the media in complicated ways, and reporters are usually in the position of responding to those agendas, or being obliged to take the logical next step, rather than creating or shaping them on their own account. That's not true for every reporter, of course; the people who wrote the hit pieces for the New York Times and Washington Post on the Mercury-News series are fully responsible for their distortions. Those are clear examples of a conscious, coordinated strike against a reporter who got out of line. This doesn't happen very often, of course, but that's because it doesn't have to. My note endorsing Christian music sure confused some people. To answer the frequently asked questions: no, I am not a Christian, at least not yet; no, I am not proselytizing; no, I do not believe that by becoming a Christian one necessarily condones, abets, or benefits from the works of Pat Robertson; and no, I am not disparaging other religions -- this is, after all, the same mailing list that endorsed shamanism a while back. Some readers have figured out that I do not think in normal ways, and this might be an occasion to give some account of my methods. When something gets in my face, I try to get inside it. My forthcoming book, for example, argues that artificial intelligence research is dangerous nonsense and a powerful way of looking at the world. By immersing myself in AI and reconstructing its inner logic, I feel that I can steer intuitively through it, continually trying to explain to myself what feels right about it and what feels wrong. If a large social group finds something useful, I try to reconstruct its attraction from within. This means learning to make the arguments myself, use the language correctly, and find the part of myself to which the healthy side of it speaks. This is the approach that I take to economics (e.g., in my messages to this list), and to public relations (e.g., in the class I teach on the subject), and to conservative political uses of Christianity. I'm not just playing a part here -- I'm genuinely opening myself to being changed by each of these things. A couple of writers suggested that I'm talking about religion in a glib way here, not giving due weight to the destruction that has been loosed in the world by people acting in God's name. Let me assure you that's not the case. Religion has been a destructive force in my family for generations, and my own ability to engage with religion in a constructive way is hard-won. I'll tell one story about this. Last year I traveled to east central Norway to find my patrilineal ancestors' farm. Among the distant cousins I met in the village was a banker from Oslo who used one of the ancestral homes as a weekend retreat. He told me about my most illustrious relative, Ottar Akre. Ottar is largely forgotten now, but in his day he was the king of Norwegian accordion players. Back in the late 19th century, Norway was swept by hard economic times, and these produced a migration to the United States and a wave of fundamentalist Christianity. The religious shift had unfortunate aspects. It was ascetic in a bad sense, being strongly opposed in practice to most expressions of love and joy. The traditional instrument of Norwegian dance music, the fiddle, went out of fashion and was even burned as an instrument of the devil. Thus the accordion. I caught a ride back into Oslo with my cousin the banker, and along the way we happened to hear some of Ottar's music on the radio. It was pleasant and mournful and was certainly not going to excite any improper passions. Although Ottar himself presumably had no such intentions, the culture in which he participated was frequently destructive, particularly in its effects on relations between parents and children. Some sense of the patterns can be gotten from Norwegian literature (see for example Katherine Hanson, ed, An Everyday Story: Norwegian Women's Fiction, Seattle: Women in Translation, 1995), and they surely contributed to the thoroughgoing rejection of religion in contemporary Norwegian society. Some of the Norwegians who emigrated to the United States took these repressive patterns with them, and they have shaped my own life. Such patterns are not unusual in fundamentalist societies; at their worst they find expression in the delusion that a child's "will" literally reflects Satanic possession and must be physically beaten into submission. It will be objected that this belief is not true Christianity, and that's right, just as the Soviet Union was not true communism. If a significant social movement is going to argue that American society should throw off its traditions of tolerance and pluralism and be ruled by people who claim to speak in God's name, however, then it becomes fair to recall what has been said and done by others who have made such claims. A truly Christian society would probably be a good thing, just as a truly communist society would probably be a good thing. The well-known problem, however, is that societies take some extreme detours when they abandon democratic principles in the name of millenarian ends. Democracy comprises both political forms and cultural forms. Among its political forms are procedures for making collective choices about matters that affect the community; among its cultural forms is discouragement of the authoritarian repression of human feeling, as well as the libertarian celebration of addiction. Democracy thus requires a cultural consensus, and this consensus cannot be produced through coercion but must be won by the light of experience and reason. Achieving the necessary consensus has frequently seemed impossible, and it has very often seemed less likely than it does today. That's how it is with democracy, which is not a millenarian goal to be achieved once-and-for-all but a daily expression of maturity and solidarity. I personally believe that the skills and values of democracy require a spiritual life, but that democracy is doomed when spirituality is corrupted and devolves into a culturally organized excuse to control other people. That's the dividing line that I see between positive and negative expressions of religion in public life. And regardless of what the pundits say, I think it's also the overwhelming majority understanding of the matter among ordinary Americans as well. I am optimistic.