Gert-Jan C. Lokhorst


G.J.C. Lokhorst. Review of Terrell Ward Bynum and James H. Moor, eds., The Digital Phoenix: How Computers are Changing Philosophy (Oxford (UK) and Malden MA (USA), Blackwell Publishers, 1998; ISBN 0-631-20352-4; 412 pp.). Ethics and Information Technology, 1 (1): 71-75, 1999. ISSN 1388-1957 (Paper) 1572-8439 (Online).


No technology has ever had a greater influence on philosophy than modern information and communication technology.

First, this technology has profoundly altered the ways in which philosophers carry out their daily work. Philosophers use computers to write their papers and books, they use email to keep in contact with their colleagues, they use the intranets to which they are connected to borrow books from the university library and to consult invaluable resources such as the Philosophers Index, and last but certainly not least, they have not lagged behind the rest of the educated world in discovering the World Wide Web as a rich source of information.

Secondly, the computer has taken a firm grip on the philosophical imagination. Are people computers? Are their minds comparable to computer software? Is the whole universe a kind of computer? Should we not try to formulate our philosophical theories (for example in epistemology) as computer programs, so as to make them precise and testable? (Formerly, formal logic was the only tool we had to regiment our philosophical thinking.)

Thirdly, the wide-spread use of information and communication technology raises its own ethical problems. One need only think of issues such as privacy protection, the gap between the information rich and the information poor and the continual well-intended attempts of governments to limit the freedom of speech on the Internet to get a feeling for what is at stake here. These issues have given rise to a new field of ethics, computer ethics, which has a similar status as other fields of applied ethics such as medical ethics and business ethics.

Bynum's and Moor's book The Digital Phoenix is a collection of 26 essays (plus an introduction) by 27 authors which gives a fine survey of the many ways in which information and communication technology are currently influencing philosophy.

The theme of The Digital Phoenix is not completely new. More than twenty years ago, Aaron Sloman wrote The Computer Revolution in Philosophy (Harvester Press, 1978), in which he stated: "I am prepared to go so far as to state that within a few years, if there remain any philosophers who are not familiar with some of the main developments in artificial intelligence, it will be fair to accuse them of professional incompetence, and that to teach courses in philosophy of mind, epistemology, aesthetics, philosophy of science, philosophy of language, ethics, metaphysics, and other main areas of philosophy, without discussing the relevant aspects of artificial intelligence will be as irresponsible as giving a degree course in physics which includes no quantum theory" (p. 5).

Even more closely related to The Digital Phoenix is Leslie Burkholder, ed., Philosophy and the Computer (Westview Press, 1992). This collection of 16 essays by 28 authors is similarly devoted to the "computational turn" in philosophy, as Burkholder put it.

Although there is some overlap between The Digital Phoenix and Sloman's and Burkholder's earlier books, the new volume nevertheless contains much new material. How could it be otherwise in such a rapidly developing new field? Books like The Digital Phoenix should appear every five years or so. (And indeed, Moor told me he is already working on a successor volume.)

There is one thing about The Digital Phoenix which is definitely not new: its title. In 1995, Bjørn Lynne brought out a CD, Dreamstate (Centaur Discs, CENCD009), whose sixth track has the same title. Bynum and Moor seem to have been unaware of this fact.

The Digital Phoenix is divided into two parts. Part I, the largest part of the book, is concerned with the impact of the computer on philosophical issues. It shows that the computer is having a great influence on the content of philosophy. The much smaller Part II discusses the more mundane issue of the computer's influence on the ways in which professional philosophers carry out their daily activities. I will discuss these two parts of the book in the same order.

Before beginning, one remark about Bynum's and Moor's introduction: on p. 2, they claim that "the widely accepted Church-Turing thesis states that whatever is computable is computable by a Turing Machine." The Church-Turing thesis states no such thing. It only states that whatever is "effectively computable" (i.e., computable by a human being who mechanically follows some set of instructions and uses only pencil and paper) is computable by a Turing Machine. This is well explained in Jack Copeland's article on the Church-Turing thesis (and its perversions) in the Stanford Encyclopedia of Philosophy at

The Impact of Computing on Philosophical Issues

Part I begins with two articles on epistemology, a discipline often regarded as the heart of philosophy.

In his "Procedural Epistemology", John L. Pollock describes his OSCAR project which is aimed at the construction of a general theory of rationality and its computer-implementation in an artificial rational agent. Human reasoning is defeasible in the sense that new information may cause us to retract previously held beliefs. Pollock's objective is to develop precise rules about how this can and should be done. He tries to formulate such theories in terms of computer programs for the following three reasons. First, this allows us to test whether the theory actually works. "As mundane as this constraint may seem, I am convinced that most epistemological theories fail to satisfy it." Secondly, in order to make a computer model, we have to make the theory precise and work out the details. "That can have a very therapeutic effect on a profession that is overly fond of handwaving." Thirdly, it is often difficult to foresee the consequences of one's theories when applied to complicated situations. If they are formulated in terms of computer programs, one may simply run the program and see what happens.

Pollock's work on defeasible reasoning is remarkable because it belongs to both philosophy and AI. There is certainly more work of this type. Much current work on conditional reasoning, epistemic logic, deontic logic and the logic of action similarly belongs both to philosophy and AI. It is regrettable that Bynum and Moor do not even mention this work.

In "Epistemology and Computing", Henry Kyburg agrees with Pollock that "fast digital computers are a wonderful boon to doing certain kinds of philosophy, for example epistemology, in the sense that they provide a kind of philosophical laboratory." However, he criticizes Pollock's approach on philosophical grounds. He sketches a different model, FLORENCE, that embodies some different principles. FLORENCE does not yet exist, so Kyburg is certainly doing some "handwaving" of the type described by Pollock. However, suppose that it existed and worked as well as OSCAR. Which model should one prefer in such a case? Interestingly, Kyburg suggests that "the answer to this question may well call for philosophical experimentation of a kind that can only be done by computers."

From epistemology, we move to a related field, the philosophy of science.

In "Computation and the Philosophy of Science", Paul Thagard briefly describes the computational approach to the philosophy of science of which he was one of the pioneers. In this field, one uses AI techniques to model, amongst other things, the context of discovery. (A subject that was more or less taboo in traditional philosophy of science.) In her "Anomaly-Driven Theory Redesign: Computational Philosophy of Science Experiments", Lindley Darden presents an example of a program of this type. Her TRANSGENE program addresses the problem of how a scientific theory (Mendelian genetic theory in this case) is properly modified given an anomaly. Her project is different from traditional philosophy of science in that she cannot afford to be vague about the details of the growth of scientific theories.

Next, we encounter two papers on reason and argument.

In his "Representation of Philosophical Argumentation", Theodore Scaltsas describes the Archelogos project, in which many scholars cooperate to generate a hypertext database which will contain analyses of the arguments used in ancient Greek philosophical texts, including commentaries on these arguments made by ancient commentators. This is a useful project which, however, uses the computer in a totally routine way.

In their "Computers, Visualization, and the Nature of Reasoning", Jon Barwise and John Etchemendy describe their well-known programs Turing's World, Tarski's World and Hyperproof, which are used in logic courses all over the world. The authors describe how the graphical output of their programs inspired them to explore a wholly new area of logic, namely the study of inference processes involving non-sentential representations (such as diagrams).

Even metaphysics is affected by the computer. The first paper in this category is Eric Steinhart's "Digital Metaphysics". This title sounds interesting, but it is not clear to me what the author wants to say. First, he says that all physically possible worlds consist of "universal computers" (p. 118). Right after this, he says that the basic components of these worlds "are not classical computers (i.e., not Turing or von Neumann machines), but are more powerful in ways not yet clear" (p. 119). A few pages later, however, we read that "actual infinities entail paradoxes" (p. 121), that "there are no infinitely complex things in nature" (p. 123) and that nature therefore consists of "finite state-machines" (p. 125, p. 129). It is hard to make sense of this inconsistent set of remarks. And even if it could be made sense of, it seems clear that these are empirical rather than metaphysical issues. Only physics can decide whether space-time is discrete, whether nature can be adequately modeled by finite automata, and so on.

Mark A. Bedau's "Philosophical Content and Method of Artificial Life", on the other hand, delivers exactly what its title promises. Artificial life studies computational structures and processes which exhibit lifelike behavior. One of its most fascinating discoveries is that complex global behavior may sometimes emerge from very simple rules. Philosophers should take notice of this. As Bedau puts it, "It is hard to avoid the fallacy of putting too much stock on our a priori intuitions when contemplating complex systems" (p. 147). He vividly illustrates this point by means of Daniel Dennett's Darwin's Dangerous Idea (Simon and Schuster, 1995). "Dennett assumes that evolution by natural selection can explain human concerns like mind, language, and morals. But Dennett's assumption is only an article of faith. He never attempts to construct an evolutionary explanation for mind, language, and morality; he never "puts his model where his mouth is" and checks whether natural selection really could explain these phenomena, even in principle... He's only guessing... Maybe natural selection can explain [these phenomena], maybe it can't; we just don't know yet" (p. 147). In other words, Darwin's Dangerous Idea is a good example of the typical behavior of the armchair philosopher which Pollock referred to--handwaving.

The section philosophy of mind contains a paper by Paul M. Churchland, "The Neural Representation of the Social World", and a paper by William G. Lycan, "Qualitative Experience in Machines". Churchland's paper, interesting though it is, is just an excerpt from chapters 6 and 10 of his The Engine of Reason, the Seat of the Soul (MIT Press, 1995). It describes how neural networks can be taught to recognize emotional facial expressions and how such networks can learn to make moral judgments without knowing any ethical rule. This is clearly relevant for moral philosophy. Lycan's very well-written article shows that there is no reason at all to maintain that machines cannot have qualitative experiences. Every undergraduate student should read this paper!

In the section philosophy of artificial intelligence, we first encounter the life-long critic of AI, Hubert L. Dreyfus. His "Response to My Critics" is difficult to understand because it presupposes knowledge of the critics' original critiques. Dreyfus is inspired by Merleau-Ponty and Heidegger, although he seems to have some difficulties reconciling these two. I think we may safely ignore these two figures from the past. Merleau-Ponty is based on antiquated neuroscience and Heidegger's work has led to so many conflicting interpretations that it makes one think of a Rorschach blot.

James H. Moor's "Assessing Artificial Intelligence: Chess and the Turing Test" is more interesting and informative. He discusses both Deep Blue's encounters with Kasparov and the annual Loebner Prize Contest (also known as the Turing tournament). Both series of events make it clear that there are huge differences between current artificial intelligence and human intelligence.

Under philosophy of computation, we first find a paper by Selmer Bringsjord entitled "Philosophy and "Super" Computation". In this paper, Bringsjord describes some abstract devices which are more powerful than the universal Turing machine. He argues that people are such devices. Unfortunately, the crucial passage of his argument is missing! It should have been at the bottom of p. 246, but it has disappeared as a result of erroneous cutting and pasting. Bringsjord's conclusion that people have "super-minds" seems to be based on his thesis that the set of interesting stories is decidable but not enumerable. I think this basis is too weak to support such a strong conclusion.

The second paper in the Philosophy of Computation section is James H. Fetzer's "Philosophy and Computer Science: Reflections on the Program Verification Debate". Here he recounts the early history of the program verification debate, which started with his observation that formal methods cannot guarantee that real life computers behave as they should.

The last section of Part I is devoted to ethics and creativity.

Terrell Ward Bynum's "Global Information Ethics" is a very fine introduction to computer ethics. It first discusses some historical milestones (such as Norbert Wiener's pioneering work), then discusses several alternative definitions of the field, next presents sample topics in computer ethics (computers in the workplace, computer security, software ownership, professional responsibility), and ends with a brief look at the future. According to Bynum, computer ethics is rapidly evolving into a broader field, namely Global Information Ethics. This transformation is due to the advent of global networks like the World Wide Web. Sample topics of study include global laws, global cyber-business, global education and the gap between the information rich and the information poor. Bynum's essay is highly recommended to anyone who wants a brief survey of computer ethics and a preview of developments in the near future.

In "How Computers Extend Artificial Morality", Peter Danielson briefly describes his work on the emergence of moralized interaction in populations of artificial agents. According to Danielson, "important parts of morality are artificial cognitive and social devices... which allow cooperation unattainable otherwise" and which thus indirectly benefit individual agents. He uses the computer to test his ideas because "Ethics is so charged with prejudice--intuition--that we need powerful tools to keep our theories honest and open to surprising--i.e. counter-intuitive--ideas. Ethics is an area where we should expect informal tools--such as the thought experiment--to be unreliable, because the equipment we run them on--our morally shaped minds, constrained by principle and norm--isn't up to the task of following out unwanted consequences" (p. 292). Danielson's chapter is perhaps too short to do him justice. It is better to read his book, Artificial Morality (Routledge, 1992).

Finally, in her "Computing and Creativity", Margaret A. Boden argues that we have gained much insight into creativity by trying to make creative computer programs.

The Impact of Computing on Professional Philosophy

Part II of The Digital Phoenix (which occupies only one-fifth of the book) has a completely different character than Part I. It contains the following chapters: "Teaching Philosophy in Cyberspace" (Ron Barnette); "Philosophy Teaching on the World Wide Web" (Jon Dorbolo); "Multimedia and Research in Philosophy" (Robert Cavalier); "Teaching of Philosophy with Multimedia" (John L. Fodor); "Resources in Ethics on the World Wide Web" (Lawrence M. Hinman); "The APA Internet Bulletin Board and Web Site" (Saul Traiger); "Using Computer Technology for Philosophical Research: An APA Report" (Robert Cavalier); "Using Computer Technology for Teaching Philosophy: An APA Report" (Ron Barnette); "Using Computer Technology for Professional Cooperation: An APA Report" (Lawrence M. Hinman). Developments are occurring so fast in these areas that much of this material is already outdated. As far as ethics is concerned the following points may be worth mentioning. First, multimedia CD-ROMs are ideal for the presentation of ethical cases in all their details. Secondly, the World Wide Web is a continually surprising source of information. For example, there is much more legislation and jurisdiction available on the Web than one might expect.


Bynum's and Moor's book does not present a complete survey of the interface between philosophy and computer technology. I missed the following topics. (1) The enormous influence which the computer metaphor has had in the philosophy of mind. (2) The interaction between logic and artificial intelligence. Modal logic, epistemic logic, deontic logic, non-monotonic logic, temporal logic, conditional logic and the logic of action are topics which originated within philosophy but which are nowadays being studied by both philosophers and artificial intelligence researchers and computer scientists. There is a lively exchange of ideas. There is much more in this area than just Pollock's work. (3) The philosophy of virtual reality and hypertext as described, for example, in Michael Heim's The Metaphysics of Virtual Reality (Oxford U.P., 1993). (4) The work by Patrick Grim and his collaborators on philosophical computer modeling.

Nevertheless, Bynum and Moor have done a fine job. Their book certainly delivers what its subtitle promises.

I have only one worry. At the present time, the whole world seems to be under the spell of information and communication technology. Is it wise for philosophers to follow this trend?

Take, for example, the case of Arthur Prior. The tense logic which he invented in the 1950s was the result of a love of ancient and medieval logic and a concern to make conceptual room for freedom of the human will. It is nowadays being used in formal reasoning about the behavior of concurrent programs. (See Jack Copeland's article on Prior in the Stanford Encyclopedia of Philosophy at It seems safe to say that Prior's thinking about ancient and medieval philosophy and the freedom of the will was in the long run more useful to computer science than any thinking of his about the computer would have been!

Now contrast this with the Mission Statement of the Philosophy Department at Carnegie Mellon, which states that they are interested in "automated theorem proving, machine learning, language technology, game and decision theory" and that "the teaching and learning of this material is supported by appropriate technology (computer tutors, interactive multimedia software)" (cited in The Digital Phoenix, p. 390). One wonders whether anything as truly original as Prior's tense logic can ever come out of such an environment.

As a second case in point, consider the answer of Donald Knuth, the acknowledged father of computer science, to the question why he does not use email. "I have been a happy man ever since January 1, 1990, when I no longer had an email address. I'd used email since about 1975, and it seems to me that 15 years of email is plenty for one lifetime. Email is a wonderful thing for people whose role in life is to be on top of things. But not for me; my role is to be on the bottom of things. What I do takes long hours of studying and uninterruptible concentration. I try to learn certain areas of computer science exhaustively; then I try to digest that knowledge into a form that is accessible to people who don't have time for such study" (Donald E. Knuth, "Email (Let's Drop the Hyphen)", at http://Sunburn.Stanford.EDU/~knuth/email.html).

The same seems to apply with even more force to philosophers. Aren't they precisely the people who are expected by society to be "on the bottom of things"? Shouldn't they turn their backs on these new technologies and work in relative isolation, just like Prior did and Knuth does?

Related to this is the worry, expressed by many, about the possibly stifling influence of the Internet on human creativity. As Tsichritzis put it, "The global networks help propagate innovation, but they breed conformity. How can researchers get something new and significant if they are in constant communication?" (Dennis Tsichritzis, "The Dynamics of Innovation", in Peter J. Denning and Robert M. Metcalfe, eds., Beyond Calculation: The Next Fifty Years of Computing, Springer-Verlag, 1997, quotation from p. 261.)

This brings me to my last point. Philosophers are supposed to be critical. It would have been nice if The Digital Phoenix had included at least one paper by a dissenter who is not at all happy with the current computer-related developments in philosophy.

Previous | Up | Next || July 17, 2015 || HTML 4.01 Strict