The scientific American.

Charles Jencks is trying to find out how architecture can be related to a new understanding of the cosmos. `The world view that has grown along with contemporary science reveals a fundamentally creative universe - open, dynamic, surprising; and active not passive'; architecture should reflect such cosmogenesis. All that clockwork universe stuff of Newton and Maxwell is hopelessly simplistic; Darwin's gradual evolutionary progression must at least be radically reappraised and Freud's `mechanical view of the psyche' is ridiculously crude. Modernism was based on such concepts, so it can now be said to be scientifically invalid, as well as unfashionable, reductive and absurdly old hat.

To make sure that we have got architecture in the right postmodern, cosmogenic perspective, an avuncular and energetic Dr Jencks takes us on a brisk tour of some of the recent popular literature of science, pointing out salient features in the new landscape of knowledge: quantum mechanics, chaos theory, Gaia, punctuated equilibrium (in evolution), complexity theory, cosmology and The Theory of Everything. Most of us are vaguely aware of many of these ideas, but Jencks' attempt to unify and relate them to architecture is heroic, if sometimes Quixotic.

I am in no position to judge whether his brisk summaries of scientific thought are accurate or not, but I hope they are more precise than his description of Ruskinian `savageness', which Jencks says was'a raw relation to truth outside the human condition'. It was quite the opposite: by savageness Ruskin meant the imperfection of many Gothic artefacts caused by the human frailty of their creators. Such mistakes were the sign that Gothic was made by freepeople - not folk bound to the inhumane pursuit of perfection by Classical rules. Savageness was a mark of the nobility of the human condition.

But let's suppose that Jencks has got his science broadly right. Because the universe is progressing from the simple to the complex, he writes: `Perhaps "Complexitists" should come out of the closet and simply pronounce complexity better than simplicity.' Jencks' new architecture of complexity has several key characteristics, principally self-similarity, non-linearity, and organisational depth.

Self-similarity is a state in which parts resemble (but are not identical to) other parts and sometimes to the whole. The notion comes from the exploration of fractals and can be seen in nature in phenomena as different as clouds, snowflakes and ferns. In architecture, the most celebrated instance is Nicholas Grimshaw's train shed at Waterloo (AR September 1993), where to achieve a three-dimensional curved carapace capable of movement, elements of structure and skin were similar but all slightly different.

Non-linearity is a phenomenon that Jencks claims is perhaps characteristic of 99 per cent of nature: 'all that is living, chemical, unfolding and thermodynamic' - the tendency of systems in the real world to behave with sudden changes of nature and scale, rather than in conformance to simple growth curves. To reflect this, he calls for `an architecture of waves and twists, an architecture that undulates and grows and diminishes continuously and abruptly'. (Hence the book's title.)

Organisational depth is an aspect of the architecture of inclusion for which Jencks has been arguing ever since he emerged on the stage as a Post-Modernist. As he points out, it goes back a good deal further than that: to, for instance, Christopher Alexander's famous championing of the rich connections of the semi-lattice rather than the necessarily restricted linkages of the hierarchical tree as the organisational structure of a building. Organisational depth implies buildings `rich in linkage, with a high degree of redundancy' - not just in simple functional terms, but emotionally and visually. Jencks cites some of the recent work of Frank Gehry, such as Chiat/Day and the Vitra Headquarters as key examples of organisation in depth. But although he for once shows plans, his heart doesn't really seem involved in exploring the human relationships that a plan reveals. Images of the exterior remain primary, even in his discussion of the Vitra building, where Gehry's tender understanding of the variety of places needed by humans at work is discussed almost entirely in terms of external appearance. (Places are humanly enriching interactions of forms and spaces, Dr Jencks.)

Organisational depth, self-similarity and non-linearity are all useful concepts, and will doubtless be used in debating building morphology in future. But obsession with outsides has always been Jencks'strength and weakness. The early work on Post-Modernism was a welcome and necessary counter to the facelessness of bureaucratic modernity of the `60s. PoMo, which started with Jencks' full-blooded support, was almost entirely concerned with image. But to be fair, Jencks has become aware of some of PoMo's faults: ersatz, phoniness, camp, Kitsch, the lies perpetrated by an electronic society.' Yet in his latest phase, when he is trying to get to the roots of the universe, nature (and everything), the good Doctor is just as much a prisoner of appearance as he ever was.

His own artistic efforts which try to explain some of the understanding he has gained from his tour d'horizon of contemporary science are charming, but rather slight: a set of wrought-iron gates with twisted strips symbolises soliton waves; a pleasing landscape device of interwoven ramps and water is supposed to involve us with the notion of the folded wave; a sideboard incised with wavy lines illustrates non-linearity. There are astrolabe-like objects to represent cosmogenesis and Lovelock's Gaia. As Jencks says, `translating this story into architecture presents some difficulties. How do we interpret and depict it?'

Er, well, perhaps it would be useful to moderate the proposition that architecture can directly depict literary and scientific notions. Up to a point, of course, it does, but it has three-dimensional physical substance: material presence in space and time which simultaneously touches us in so many of our sensual and intellectual perceptions that to reduce the essence of buildings to literary one-liners is absurd.

To achieve some semblance of depiction of his cosmological preoccupations, Jencks has had to turn to some very strange sources. Though he says that he is concerned with getting beyond `the provincial [in the universe] concerns of the moment, beyond anthropomorphism and fashion', he uses two of the most fashion-conscious architects in the world today (Koolhaas and Eisenman) as exemplars of the new thinking. Rem Koolhaas'method of `panic design' (last-minute responses to huge programmes) is praised as a method of representing discontinuity and non-linearity. The results are almost always flashy and image-obsessed, detracting from the richness of human use and experience.

Much more worrying is Jencks' Eisenman worship. Peter Eisenman is held up as an example because `in an era when opinion and anthropocentrism dominate culture. [his work] returns us to a non-human standard for architecture that used to be the preserve of religions.' Can Eisenman's Nunotani Headquarters in Tokyo, built to look like a tower collapsed after an earthquake (a representation of punctuation to disturb the dreariness of equilibrium as if that were a threat in the contemporary world), or his Columbus Convention Center, with its overlapping wavy lines reminiscent of the Jencksian sideboard, really be considered to be our answers to the Parthenon, Chartres or Ronchamp?

The dark side of Jencks' new preoccupations is the dislocation of humanity from the centre of the stage. Eisenman and his friends have been going on about this for ages. The holocaust and Hiroshima have shown, they say, how fallible is our belief in the centrality of human values. Now Charles Jencks brings up the physicist Hans Christian von Baeyer to announce that `in the twenty-first century, the atom will replace man as the measure of all things'.

The prospect is terrifying: once you remove human values, what's left? Dear Charles Jencks is a kindly, charming, often intentionally funny man. He believes, with Ruskinian intensity, `that there is a covergence of the good, the true and the beautiful': that nature has only `occasional nastiness'. Yet the one thing that emerges with the greatest clarity between the lines of his tour of popular science is that universe is totally indifferent to any individual or species. Stephen Hawking may say that developments in physics may enable us to see the mind of God. But, on his measure, either God is a monster devoid of compassion and feeling, or we will have explored only a very small part of His mind. In either case, we have to try to generate decent ways of living with each other and with the biosphere without much guidance from Above.

Within an indifferent nature, the making of buildings has a vital role in creating places and sanctuaries for us as individuals and societies. Love, human understanding and generosity are the essence of this activity - and attempts to understand and represent the cosmos help us not a wit in three-dimensional interpretation of them.