This is truly a magnum opus, and it is not at all surprising that it has been 20 years in preparation. The author is a professor of semiotics in Copenhagen Business School, and the founder and former Editor of the journal Cybernetics and Human Knowing, now edited by Ranulph Glanville. Some of the material in the book appeared first as articles in the journal.
The transition from first‐ to second‐order cybernetics brought the observer into the “universe of discourse” (a phrase often used by Gordon Pask), and the treatment presented here can be seen as progressing very much further in a similar direction. The discourse is extended to consideration of science, and indeed human culture generally, with consideration not only of formal theory but also of motivations and value judgments and the mental impressions denoted by qualia. Attention is focused on incompatibilities between the approaches of different disciplines. In particular, the bases of the social sciences are contrasted with those of the physical. The term hermeneutics, or study of interpretation, appears frequently. (The term is usefully discussed in Wikipedia and in the Stanford Encyclopedia of Philosophy at: http://plato.stanford.edu/entries/hermeneutics/).
It is claimed that the advent of computers and all that goes with them including artificial intelligence (AI), cybernetics, Shannon's information theory and the internet, demand new formulations of what is to be understood by such terms as information, cognition, communication, intelligence and meaning. Theory is developed on the basis of second‐order cybernetics, along with the topic of semiotics pioneered by Charles Saunders Peirce, with also reference to Niklas Luhmann and Humberto Maturana among others. A central consideration, obviously, is the nature of scientific investigation and theorizing. There is acceptance of Popper's assertion that a theory can never be proved but can be disproved, and that an accepted theory should be the hypothesis that best fits observations, but with the observation that much more is involved. Brier indicates this succinctly by observing that “enquiry is never disinterested.”
Developments in AI, which clearly has cross‐disciplinary connotations, are central to the arguments. In the list of references, the leading people in AI are conspicuous by their absence, except for three sources of an overview nature. One of these, acknowledged as influencing Brier at an early stage, is the book by Winograd and Flores (1986). Another is the work of Miller, Galanter and Pribram, and another the well‐known work of Penrose. Winograd and Flores are anxious to “debunk” what they see as the more extreme claims for possibilities of AI, and they follow Maturana closely in his treatment of perception.
The debunking of these more extreme claims supports the implication of the subtitle of the book: “Why information is not enough!” The reference to “information” is meant to indicate the possibilities opened up by computers and their handling of information, understood in the sense that makes it quantifiable in bits or bytes. There are undoubtedly gross mismatches between capabilities of brains and those of present‐day computer programs, but it is surely debatable whether this must always be so. To suggest that in principle, it must is to defend a dualist philosophical viewpoint. Such a viewpoint does not seem to me to be compatible with the current view of life as beginning in a primeval soup, followed by elaboration according to accepted evolutionary theory.
To my surprise, on following up the page numbers against “dualism” in the subject index, I found that at some points Brier explicitly defends Cartesian dualism against an alternative due to Laplace. Near the end of the book (p. 400), however, in discussing the evolution of language, he appears to modify the viewpoint, where he writes: “Thus, Cartesian dualism has in this conception been replaced by different kinds of semiotic webs interacting with one another inside, outside, and among nature, bodies and minds.” This seems to indicate an argument for dualism on the basis of complexity, which is more defensible than an “in principle” assertion.
From a practical viewpoint, the dualism‐monism distinction is not very important, since computer capabilities fall far short of those of biological intelligence in many respects and are likely to remain so for the reasonably foreseeable future. For that duration, distinct approaches to study of the two areas will be needed.
Although Brier writes in a clear and pleasant style this is certainly not an easy book to read, and I certainly need to spend more time with it. One reason for difficulty is illustrated by a diagram right at the beginning, in fact facing the title page, where “being” is opposed to “non‐being”, “manifesting” to “non‐manifesting”, and terms shown include “spirituality” and “utopia.” There is obviously room for variation in individual interpretations of these terms, and a similar observation can be made about other material in the book. Certainly, it would be wrong to insist that terms be precisely defined, since definitions can only combine other terms and must ultimately depend on understandings derived from usage (von Foerster and Poerksen, 2002). It would also be wrong to suggest that Brier should avoid the imprecise terms since that would restrict the “universe of discourse,” but this leaves the reader with the task of deriving some consensus from study of Brier's use of the terms.
I have the feeling that Brier is glib in accepting certain rather dubious assertions by other workers as though more firmly supported than they are. The extension of the principle of autopoiesis to social systems seems to be accepted without question although Maturana himself is distinctly lukewarm about it in his comments reported by Poerksen (2004). There is also reference to the Laws of Form and Varela's (1975) extension that is said to constitute a theory of self‐reference, but the debunking by Cull and Frank (1979) is ignored. There also seems to be rather uncritical acceptance of the still‐controversial assertion by Penrose that quantum mechanics play a part in perception and consciousness.
Although there is reference to evolution at a late stage, the treatment begins with consideration of different paradigms of science and culture. This seems wrong since these are late evolutionary developments that should not be viewed in isolation. The emphasis on semiotics as a theory of signs seems to show a similar bias, since a “sign” is likely to be interpreted as a discrete event (though it need not be) with an implication of advanced logic‐based processing. It is in fact easy to feel that the approach is misdirected in a way that corresponds to a mistake in early work in AI, where it was assumed that attention should be given to highly regarded manifestations of intelligence such as mathematical theorem‐proving, with an apparent assumption that when these had been realized as programs, means of achieving more everyday intelligent behaviour would be a simple “spin‐off.” Of course, this was quite wrong and although computers can be made to handle complicated formal mathematics they cannot for example control a robot so as to cross a street through traffic. Cybersemiotics may be subject to a similar mistake of aiming too high too soon.
These somewhat critical comments should not obscure the fact that the treatment in the book gives much food for thought and an introduction to important current thinking. It is possible that my misgivings will be allayed by further study.
