Robert W. Schuler
November 27, 1999
This paper reviews and critiques Theodore Roszak’s book The Cult of Information A neo-Luddite treatise on high-tech, artificial intelligence, and the true art of thinking 2nd Edition, 1994 University of California Press, Berkley and Los Angeles, California (ISBN 0-520-08584-1).
The Cult of Information is a refreshing book in an era of "How-to" books on computing. Rather than describing a new aspect of technology or software, this book takes a high level social look at computers, computing, and the industry that is based on information. Roszak starts the book by quoting "the emperor has no clothes" and maintains a healthy skepticism of computers throughout.
The book is divided into 11 numbered chapters and an unnumbered introduction to the Second Edition that is 50 pages long and may be considered Chapter 0. Each of these twelve chapters is divided into subsections which are essays relating to the chapter title. There are over 60 of these essays in the book and they range in length from a few pages to a whole chapter.
The author’s purpose is established in the preface and can be summarized by three statements: "The burden of my [Roszak’s] argument is to insist that there is a vital distinction between what machines do when they process information and what minds do when they think."; "But the Information Age has now entered the educational curriculum in an aggressive and particularly insidious way which could distort the meaning of thought itself."; and "Information technology has the obvious capacity to concentrate political power, to create new forms of social obfuscation and domination." (Roszak, xiv-xvi). In addition to these three statements, Roszak also defines a cult of information consisting of computer enthusiasts, data merchants, futurologists, hackers and hucksters. He claims that the cult of information has an agenda of shifting power to the technically elite by blurring the distinction between information and thought, by pushing technology in the school system, and by leveraging information technology in politics and government.
While I disagree with the Roszak about the existence of an organized cult with an agenda, I have found several of his points to be valid and several of the issues that he has raised to be worthy of further study. In the remainder of this paper, I will discuss:
The term Luddite is used to describe one who distrusts or fears changes brought about by new technology. The original Luddite revolt occurred in 1811 during the industrial revolution as an action against the English Textile factories that displaced craftsmen in favor of machines. Today's Luddites (or neo-Luddites) continue to raise moral and ethical arguments against the excesses of modern technology to the extent that it threatens our essential humanity. (Ryder, Paragraph 1). Roszak explains that the Luddites where not rebelling against the machines per se, but rather where rebelling against businesses who where using machines to replace workers who needed the income. He claims that the Luddite plea was simply for justice and humane treatment. (Roszak, xviii)
According to Albert F. Davenport, "While many today blindly fear technology, Neo-Luddites embrace it. We see the personal computer as a great emancipator. Within each computer is the power to fulfill a dream. We no longer need to work to enrich others, because we have limitless opportunities to work for ourselves. Like those Luddites who spun garments in their cottages, we present unique creations to a community of intellect from our own homes. We have come full-circle. It is Capitalism much the way Adam Smith envisioned it, with information as the medium of exchange." (Davenport, Neo-Luddite, Paragraph 2).
Davenport’s definition seems to disagree with Roszak’s. Roszak seems to encourage a healthy skepticism of the new technology rather than simply embracing it. Like so many other terms in the computer community’s vocabulary, neo-Luddite seems to be overloaded to the point of being ambiguous. The clear part of the definition is the reference to the early industrial age rebels who resisted new technology in favor of their ‘current’ way of life.
In a very real sense the use of the term neo-Luddite draws a parallel between the Industrial Revolution of the early 1800’s and the Information Revolution of the late 1900s. There are many reasons for examining this parallel since the Information Revolution promises to have the same sweeping impact on our society today that the Industrial Revolution had on society in the early 1800s. Personal computers have already dried up most of the typing pools, started putting draftsmen out of work, and thanks to touch pads with pictures of cheeseburgers, have reduced the math skills requirements for many cashiers.
From this perspective, Roszak is correct in adding ‘neo-Luddite’ to the subtitle of his book. His essays are largely fearful of the use of the information technology that is sweeping our new world-economy. It is not the technology itself that concerns Roszak just as it was not the power looms that upset the Luddites. Rather it is the application of these devices to change society that is raising concern.
In chapter five, Roszak explains that "information, even when it moves at the speed of light, is no more than its ever been: discrete little bundles of fact, sometimes useful, sometimes trivial, and never the substance of thought" and that "the mind thinks with ideas not information." He goes on to say, "Information may helpfully illustrate or decorate an idea; it may, where it works under the guidance of a contrasting idea, help to call other ideas into question. But information does not create ideas; by itself, it does not validate or invalidate them. An idea can only be generated, revised, or unseated by another idea. A culture survives by the power, plasticity, and fertility of its ideas. Ideas come first, because ideas define, contain, and eventually produce information." (Roszak, 87-88). On page 90, Roszak actually defines "Ideas are integrating patterns which satisfy the mind when it asks the question, What does this mean? What is this all about?"
This distinction between ideas and thought is important and Roszak does an excellent job in highlighting it. Chapter five is one of the jewels of the book. The study of thoughts is more within the domain of psychology and metaphysics and has been largely kept outside of the domain of computer science. Patterns of symbols, like the text you are reading now, are used to communicate ideas, but the creative process that generates these patterns is still outside of the grasp of computerized algorithms.
In the second essay of chapter five Roszak defines what he calls "master ideas—the great moral, religious, and metaphysical teachings which are the foundations of culture. . . Master ideas are based on no information whatever." For example, "all men are created equal." (Roszak, 91). Thomas Jefferson used the phrase "self evident" to refer to these types of ideas. This concept of a master idea is significant because it implies a human sensibility that reaches beyond information and therefore cannot be expressed as the result of computation.
Computers are incapable of addressing the human sensibility required to recognize and synthesize master ideas. Computers simply manipulate symbols. Since written text can be used to communicate master ideas between humans and since computers are very adept at manipulating the symbols of written language, there is a tendency to blindly assume that computers can manipulate master ideas. This has been the primary assumption that has driven artificial intelligence (AI) research for the past one-half century.
Roszak rightly accuses the cult of information of distorting reality to match the computer’s capability. "As so often happens when the computer’s narrow mechanical powers are applied to a complex problem, the success that results comes mainly by way of downgrading the task into something the computer can do—and then cheering loudly enough to drown out the sensible reservations." (Roszak, 216).
The ability of very smart people to perform this reality distortion contributes to the folklore of computers. "AI has been peculiarly characterized by extravagant, often propagandistic claims in its own behalf, with the result that authorities in the field have contributed as much to the folklore of computers as the advertisers, promising machines that would translate languages, understand speech, process visual images, make legal, political, and financial decisions, and, in general, outstrip human intelligence in every application."(Roszak, 122).
On page 124, Roszak quotes Herbert Grosch, formerly of IBM: "The emperor—whether talking about the fifth generation or AI—is stark naked from the ankles up. From the ankles down, the emperor is wearing a well worn and heavily guilded pair of shoes called expert systems. They are useful, but we’ve had them for over thirty years. All that the fifth generation boys have done is relabel them."
What has fallen out of AI research has been a healthy respect for the complexity of human intelligence along with a useful new class of software called expert systems. By restricting the domain of information required to a specific context, computers can assist humans in making complex decisions like diagnosing certain ailments or trouble shooting a piece of complex equipment.
Roszak’s points, while valid, are as propagandistic as those he is criticizing. It is true that AI has fallen short of its promises, but this is not due to a cult overstating the potential uses of technology. It is due to a misunderstanding of the complexity of thought. Technologists do not propagate the idea that computers will one day think like humans, at least not to the extent that Roszak would have us believe.
Roszak really pulls out the stops when preaching on the ills of computers in the classrooms in chapters 3 and 4. He emphasizes a fear that has shown up in several news media throughout the 1990s; specifically that due to the cost of computer equipment, only students in rich school districts will have access to computers in schools. "Writing in The New York Times (August 7, 1992) Charles Piller and Liza Weiman charge that computers may be reinforcing ‘a two-tier system of education for the rich and poor.’"(Roszak, 48)
The fact that the use of computers in our primary and secondary education systems needs to be examined for its social and economic impacts does need to be looked at. I believe that there is a discrepancy between the rich and the poor when it comes to computer access. Susan Rose supports this concept in her thesis, The Role of Community Access Centers in Bridging the Technology Gap: "As computers become the conduit through which society accesses information, and computer literacy becomes a baseline job skill, it is important to ensure that everyone has access to computers. At this time, everyone does not have access to computers. People who live in low-income communities are disproportionately cut off from computer training."(Rose, Title Clause)
Unfortunately, Roszak does not stop at the identification of a social gap. He goes on to blame the data merchants for creating an environment in which schools and parents where conned into believing that children need computers: ". . . the computer has made a wastefully ragged and disruptive entry into the school systems of the nation. For the most part, the schools (or mainly trend-conscious administrators and anxious parents, less so teachers) have responded with the promptness and the gullibility of well-trained consumers to the commercial pressure of the computer industry."(Roszak, 48-49)
"Initially, there was a simple justification for favoring computers over violins in budgetary priorities of the schools. It was embodied in the catchphrase computer literacy—a seemingly undeniable necessity in the Information Age. Lacking that skill, children would grow up to be unemployable."(Roszak, 49). Roszak goes on to explain how this catchphrase has been used and abused over the past two decades, but he fails to acknowledge that computers are on every desktop and in many homes in the U.S. I think this is because the original release of this book was 1986; before the Internet took off and personal computers became as powerful as they are today.
Whether the "trend-conscious administrators and anxious parents" created a self-fulfilling prophecy is irrelevant now. Computer skills are vital in today’s workforce, so the first part of Roszak’s argument is correct, "the computer has made a wastefully ragged and disruptive entry into the school systems of the nation", and this has resulted in a blurry definition of computer literacy. I disagree with Roszak’s approach of laying the blame for this solely at the feet of the data merchants, however.
In the third essay of Chapter 3, Roszak makes the claim that data merchants are eager to sell their wares to universities. "The schools are the gateway to a student market populated by prospectively high earning professionals and white collar workers, the ideal computer buyers of the future. . . schools . . .have seriously considered imposing [a requirement that students own computers]. Still more welcome than a literacy requirement would be a property requirement: namely that owning a computer should be a condition of admission. . . Has there ever been another instance of universities making ownership of a piece of equipment mandatory for the pursuit of learning?" (Roszak, 57-58). Roszak goes on to worry that information technology may convert professors from "the sage on the stage" to the "guide at the side of the magic box." (Roszak, 59).
Here again, I disagree with Roszak. The course for which this term paper is produced is an online course, and admission to the program requires me to own a computer capable of interacting on the Internet. The power of long distance asynchronous communication justifies this trend of requiring students to own computers. Again, I think that Roszak was not crazy when he wrote these words in 1986, I think he just turned out to be wrong.
Roszak makes a Luddite argument in his discussion of computers entering the classroom when he discusses EUIT, EDUCOM’s Educational Uses of Information Technology unit: "What EDUCOM’s bright prospectus leaves out of account are the ambitions that bring many students to school in the first place. If the computer is used to slash salaries and eliminate teaching positions, what is a college degree going to be worth to those whose goal it is to become educators? There comes a point in the computer enthusiast’s image of our cultural destiny that leaves creativity and imagination only one outlet: programming software, designing hardware."(Roszak, 59). Here we see the fear of a teacher about computers replacing teachers. Computer based training and multimedia courses (consisting of VHS tapes and/or audio tapes accompanying books) have been around for two decades now, and the educational system has not crumbled as a result.
A final point on which I do agree with Roszak that computer literacy is an ambiguous term. Many people know that computer literacy is important, but few people can agree on what it means to be computer literate. This confusion does not lessen the importance of developing curricula to educate students in the use of computers.
Susan Rose also supports this argument in her thesis: "One constant is that computer training is seen as a fundamental literacy skill along the lines of reading and writing. Yet, there is no established method of teaching this literacy skill. When school teachers teach students to read, they have baseline goals for each level. All teachers follow the same guidelines, even though teaching styles differ. Computer training could benefit from the same kind of structure. It seems that some standardization could aid CACs in reaching their goals. It is hard to say whether a student has mastered basic computer skills if there are no baseline guides as to what basic skills are." (Rose, Clause 6.1.1)
In chapter 9, Roszak dedicates several essays to the effect of information technology on public access to information. He explains that since the Age of Reason (Ben Franklin’s day) to the present, the public library has made information available to all citizens regardless of their income or social status. "Yet it is striking that, in the superheated public discussion of computer technology, the library is so rarely mentioned. To some degree the library may be a missing link between the computer and our society at large because, in the public mind—and perhaps in the minds of many librarians as well—it is too closely associated with print-on-paper to seem compatible with high tech." (Roszak, 173-174).
I think that this argument is dated. At the time the first edition of the book was published in 1986 I/T was still making its way into the libraries of our country. I have not been into a local library recently that did not offer Internet access from a publicly available terminal. Libraries are filling the missing link that Roszak discusses. He goes on to suggest that large databases should be made available free of charge in a similar manner to existing encyclopedias.
"A true citizen’s reference service of this kind clearly does not jibe with the thrust toward making information a proprietary benefit available only by license from those who own it, for whatever fee they care to charge for it. In times past, a book which contained information might be a commodity with a price upon it; but a book is cheap. If it finds its way into a public library, the library owns it and it is available free of charge. And whatever it contains—say, the definitions in a dictionary, the thought in an encyclopedia, the statistics in an almanac—is in the public domain. Nobody collects a fee each time you look up a word in Webster’s or consult the Britannica." (Roszak, 174).
This is an interesting argument, but it fails to reflect two simple economic factors: 1) it costs more to compile electronic information in a useable form than to publish a book, and 2) even when printed encyclopedias are freely available, there is a cost to the end user to a) find information available and b) search through indexes (or page by page) looking for key words. These factors have led companies to offer subscriptions rather than one-time purchases of electronic databases and the time savings in searching have justified end users spending the subscription fees. There is a public interest in guarding against "privatizing the public’s right to know" (Roszak, 178), but it must be weighed against the benefits that private databases offer to public library patrons.
Roszak goes on to identify a neo-Luddite perspective for the librarian: "In lieu of librarians we will have programmers and database experts." (Roszak,182). He then identifies two mindsets within the librarian profession: "conservative librarians" who wish to stick to the traditional distribution of information through printed material, and the "progressive librarians" who believe that "the commodification and privatization of data amounts to an insidious form of thought control by profit-based businesses that are unaccountable to any public authority."(Roszak, 182).
The role of a librarian to help people find information will never be replaced by computers. Roszak is right in suggesting that we should guard against this, but even in a world where all books have been transformed to electronic media, there will still be a need for learned people to help other navigate through the available resources.
From libraries, Roszak transitions to discussing the Internet, which was even less evolved in 1986. On page 186, Roszak makes a brilliant observation that: "government may need do nothing more than guarantee by law the context within which such a network runs itself: namely, that information shall be a public utility kept free because the people want it to be free. As it has always been in the free public library." This idea, in part, was signed into public law under the Internet Tax Freedom Act (P.L. 105-227) which is "based on the simple principle: Information should not be taxed" (Cox, paragraph 1). This is a very important idea, but must be balanced against the cost of providing information in a publicly accessible format. Web pages are not free. There is a cost associated with operating and maintaining the computers and the communications links to the computers that serve web pages. Unlike a printed book whose cost is fixed at the time a library purchases it, on-line electronic information has a continuous cost associated with it. The trick is to ensure that the information itself remain publicly available and that the people paying the costs do not acquire the power to control the information flow.
A final noteworthy subject that Roszak brings up while discussing libraries is the idea of "electronic Alzheimer’s", which addresses a very real problem in I/T. Information recorded electronically requires electronic equipment for viewing. The problem is that technology is changing very quickly and data stored on older electronic media are no longer accessible. I am unable to read several of my term papers from my undergraduate work because they are stored on 360k 5.25 inch floppy disks and all of my current computers have 3.5 inch disk drives. Roszak points out an even more expensive example on page 194: "New hardware cannot always read old software. NASA, for example, owns twenty thousand seven-track data tapes from its history of space exploration that can only be read on a seven-track tape drive. In 1992, when its last seven-track drive broke, NASA was able to find only one more such drive in existence. And when that breaks . . ."
Beyond shifting the power of information away from publicly available channels, Roszak fears that the use of polling is changing the mechanics of politics and the power of marketing too. He makes the claim that "What pollsters claim for their methods results from a drastic cheapening of the democratic process. In effect, they deflect attention from issues of substance by turning debate and electoral judgement into a vacuous statistical game."(Roszak, 216)
"John Kennedy was the first national candidate to make important use of polling. That was in 1960; his hired pollster was Louis Harris, who then became an independent expert. By the late 1970s, every serious candidate for office in the United States who could afford the price was following Kennedy’s lead; expensive polling along with lots of media exposure had become the prevailing campaign style."(Roszak, 213)
"What caused this change of tone and rhetoric [in Ronald Reagan’s 1980 presidential campaign]? It was done in response to a key campaign advisor, Richard Wirthlin. The advice might have been based, as such advice usually has been, on pure political instinct, which may or may not have been persuasive with the candidate and his many other counselors. Politicians always work at the center of rumors, guesswork, hunches, tested savvy, gut feelings. But in this case, Wirthlin’s advice was based on something else: numbers, lots of numbers. It arose from a barrage of public opinion polling all across the nation. Wirthlin commanded statistics. That gave his advice the appearance of something more than guesswork. It looked like science."(Roszak, 212).
From these examples and a discussion of the pollsters, Roszak draws the conclusion that "The pollsters . . . obviously bias elections in favor of those who can spend the most for the best services."(Roszak,214) and that "The result is a dismal style of politics that grows more and more obsessively focused on imagery, sloganeering, rhetorical legerdemain—in fact, on the huckstering skills of the marketplace. To be sure, politics has always been dogged by these vices in America, but the pollsters only ingrain the corruption deeper by claiming they can maneuver their candidates with pinpoint accuracy. . . For of course, none of the information is any better than the assumptions that program it."(Roszak, 215)
The target of the neo-Luddite is not the computer technology but the use of the technology. What is most interesting in this argument is that Roszak seems to be including political statisticians in the cult of information. This represents a shifting away from the stereotype of electronics fanatics and database weenies that are interested in developing and proving algorithms without concern for broader the social impact. Here, Roszak’s cult members are specifically interested in the broader social impact, but this is still bad.
I agree with Roszak’s assessment of polling, and offer my own example to support this argument. Consider the "Poll Taker" question from the Washington Post Web Site shown in the following figure:
What is this question really asking? What aspect of foreign policy is important? The result of this poll will simply be to direct candidates to say, "I am strong on foreign policy," or "I am more concerned with _______", where the blank will be filled in by what ever else the survey indicates. These statements are very shallow, and the energy spent worrying about them could be better spent discussing actual foreign policy issues.
Roszak ends his discussion by extending a hypothetical situation that illustrates the extreme. He offers a fictitious news article from the future: "It will begin with a poll asking people if they approve of the way the polls have been handling the president’s approval rating. Then there will be a poll measuring the public’s opinion about the results of that poll. Then there will be a poll about the poll about the poll. At last there will be an election in which the public will vote for the poll that it thinks most accurately reflects the publics opinion." (Roszak, 218) As crazy as this sounds, it does reflect a trend in poll taking and results reporting.
Theodore Roszak’s book The Cult of Information A neo-Luddite treatise on high-tech, artificial intelligence, and the true art of thinking is truly a neo-Luddite treatise. Although it contains some rather extreme negative views of technology, this book cannot be dismissed with a wave of the hand. It contains many useful insights into technology and its application in our society.
Roszak succeeded in showing the distinction between what machines do when they process information and what minds do when they think and he does an excellent job of explaining self-evident truths (or master ideas). So far, the human mind is the only known source of master ideas and the only known authority for recognizing master ideas.
Roszak succeeded in arguing that the use of computers in classrooms needs to be reexamined, but he fails to convince me that computers entered the classroom in an aggressive and particularly insidious way which could distort the meaning of thought itself. Rather, the second part of the argument is one of the extremes that must be weighed in the mind of each reader.
Roszak also succeeded in demonstrating the capacity of Information technology to manipulate political power, to create new forms of social obfuscation and domination. The control of public information should be of paramount concern to all citizens in a democracy and I agree with Roszak that the public library is an important institution in ensuring that public information remains public.
While computers are only one ingredient in the current use of political polls, they are an important ingredient. The use of statistics to confuse people and to direct attention away from important issues should also be of great concern in a democracy. Roszak’s argument about the weakening of our democracy is a little extreme, but again deserves the attention of each reader.
When one considers the context of the original work (1986) and the effort required to develop the second edition in 1994 this work shows great insight. Roszak wrote about the Internet before it exploded and for the most part, he was right. Even in 1994, but definitely in 1986, the work being done in laying the groundwork for the Internet must have looked like a for-profit-only enterprise being conducted by a cult of technologists.
Perhaps it was, but Information technology keeps doubling in power every 18 months following Moore’s law (Raymond, 1999) which has had an equalizing effect that is missing from Roszak’s work. Specifically, Information technology is no longer in the hands of highly trained technologists, but has started to show up in the less trained hands of ‘average’ citizens. As this trend continues, even the inequalities between the rich and the poor (as far as access to technology is concerned) will evaporate over the next 5 to 10 years.
One final note: Technologists and neo-Luddites can both learn something valuable by reading Roszak’s book. It is very healthy to look at the social implications of the technology that is driving the Information Age and the parallels to the Industrial Revolution are worth studying during the Information Revolution.
Cox, C. (1998, October 20). "Plain English" Summary of The Internet Tax Freedom Act (P.L. 105-277), [online].Available: http://www.house.gov/cox/nettax/lawsums.html [1999, November 27].
Davenport, A. (1997). THE NEO-LUDDITE MANIFESTO, [online].Available: http://www.personal.psu.edu/users/a/f/afd108/neo-ludd.htm [1999, November 27].
Poll Taker (1999, November 27). On Politics. Washington Post.com, [online]. Available: http://www.washingtonpost.com/wp-dyn/politics/ [1999, November 27].
Raymond, E. (1999, June 17). The Jargon Dictionary, [online].Available: http://www.netmeg.net/jargon/terms/m/moore_s_law.html [1999, November 27].
Rose, S. (1997). The Role of Community Access Centers in Bridging the Technology Gap, [online].Available: http://www.ctcnet.org/rose/06con.htm [1999, November 27].
Roszak, T. (1994). The cult of information: A neo-Luddite treatise on high tech, artificial intelligence, and the true art of thinking. Univ. of CA Press. ISBN 0520085841
Ryder, M. (1999, October 23). The Neo-Luddite Reaction, [online].Available: http://www.cudenver.edu/~mryder/itc_data/luddite.html [1999, November 27].