Pounds of pretty paper.

PHAIDON ATLAS OF CONTEMPORARY WORLD ARCHITECTURE

London: Phaidon, 2004. [pounds sterling]100

Like so many departments of contemporary culture, architectural publishing has begun to embrace shock and awe. Rem Koolhaas's weighty but almost empty tomes have set a trend. Phaidon's Atlas of Contemporary World Architecture (Comprehensive Edition) is another (but much more massive) example of the same tendency. Compared to OMA's fission missiles, it is a fusion device, weighing in at a hefty 7kg, with pages measuring 460mm by 320mm. It is impossible to examine except at a large and sturdy table, or by lying on the floor.

As a publishing venture, it is most ambitious, trying to create a portrait of the present state of the art architecture with over 1000 entries of what the elusive editors have decided are the most important buildings to be completed in the last few years. The atlas is divided conventionally into six continents: Oceania, Asia, Europe and so on. Each continent is divided into chapters and subchapters on countries--though sometimes with some strange groupings, particularly so in Britain, which is divided into Scotland with England North, and United Kingdom South, which is bizarrely married to the Republic of Ireland, a country with a distinct and strong architectural culture that surely deserves a proper chapter to itself like Austria and South Korea.

Insistence on flat-footed categorization by geography is one of the great flaws of the mighty atlas. Geography precludes further analysis by climate, culture, politics or history. A superficial introductory section tabulates countries by, for instance, GDP per head, and an obscure measure of 'environmental sustainability'. But such information is not used at all in discussions of individual works.

Each building shown is given half a side, sometimes a whole page, and exceptionally two. Each is explained in good coloured pictures, a short description, and drawings: plans always, sections sometimes, with the odd elevation thrown in (why, if there are photographs?). I have not yet found a scale, nor a key to functions, nor a north-point (except when included by accident). So the drawings are largely decorative, and the usually small photographs are often not revealing enough to make much sense.

It is impossible to avoid the impression that the atlas has been composed by people who do not really understand architecture, but who know how to lay down pretty pages. The feeling is reinforced by the editors' choice of emphasis: an elegant hut is sometimes given a page; a complex housing scheme usually rates half (often with only one plan), perhaps because it is too difficult for the authors to interpret. Only fashionable icons get two pages.

What and whom is the atlas for? Can it be any use to practitioners? I doubt it, except as a random compendium of images. Is it a useful guide for students? Probably not, as it gives little or no notion of why and how its examples were created, nor indeed how they can be compared in sensible terms. Will future scholars and present hacks use it? Well perhaps, to check the field of runners in the early twenty-first century fashion stakes.

I suspect that the volume is fundamentally a website struggling to emerge from the dead weight of bound paper. Electronically, much more information on each building could be included, both written and visual. More examples could be included. More cross-referencing could be offered (though the indexes in the volume are not bad). With a comprehensive and more thorough electronic database, offering more examples than the atlas does, and describing them in more detail, a lightweight publication on paper could be made that would act as a guide and it could offer a variety of theoretical and descriptive interpretations of the contemporary scene.

The atlas shows how far paper compendia can go. For what it offers, it is not wildly expensive. It is a brave and, in some ways, even a heroic effort. Sad really: shock and awe have little place in creative cultural and professional discussion.