Encyclopedia of Vernacular Architecture of the World.

We are all aware of the disastrous way in which the world's plant and animal species are declining, and that by causing extinction on a massive scale we are depriving ourselves and our successors, not just of the wonders of nature, but of as yet unknown benefits to humankind. The vernacular architectures of the world are in a sense similar to natural species and they are just as hard pressed by crude industrialization.

Traditional buildings made without architects are precious because they are usually beautiful and fascinating in themselves. And, because their forms have been honed by a process analogous to natural selection, their examples may offer suggestions about how to make future architectures that are sustainably related to climate, topography and cultural tradition. Paul Oliver, editor of The Encyclopedia of Vernacular Architecture of the World believes that 'traditional wisdom and lore in building, using renewable resources and indigenous skills, may still offer wisely managed, economically effective and culturally appropriate solutions to the world's housing needs'.

Vernacular buildings are not precisely the same as natural species, in that while a plant is lost for ever when it becomes extinct, we can (up to a point) recreate destroyed artifacts - provided that they have been recorded. Oliver has laboured awesomely and produced in three volumes the most thorough record of the vast and heterogeneous field that will ever be collected in one book. Helped by some 800 scholarly contributors from over 80 countries, he has made a survey of all the different vernacular architectures of the globe. Perhaps as importantly, the Encyclopedia lists all key sources published in European languages, so that if you want to study a place in detail, find more measured drawings and so on, you know where to go.

Oliver has divided the vast mass of information and ideas into two main sections: the first on theories and principles and the second on cultures and habitats. The latter (which takes up volumes II and III) is an illustrated and painstaking survey of vernacular building, area by area over the globe. It suggests general lessons about each climatic and cultural zone and illustrates them with particular local examples. All very clear and straightforward, though of course you will sometimes disagree with the case studies chosen in the areas you know best - and certainly want to add to them. But that is the nature of such a book, and one of the reasons why it is so stimulating.

The first volume, on theories and principles, has a very different recipe. Oliver and his colleagues have analyzed the theories with which the vernacular architecture has been approached by scholars, thinkers and historians since the study of the subject emerged. The survey is fairly clearly organized: entries vary in perception with individual contributors.

The much longer 'principles' section is more problematic. The aim is to draw out laws about the nature of vernacular architecture, no matter what the location. Themes range from 'Environment' (basically climate and topography), to 'Production' (craftsmanship and production), to 'Typologies' (forms and structural types). The attempt to extract general lessons out of such a heterogeneous mass can sometimes lead to very bizarre conjunctions. For instance, under the general rubric of 'Schematic Decoration', the first sub-articles are (in order) 'Falun red' (the blood colour of the big Swedish barns); 'Colour notation' (a brief description of Natural Colour System theory); 'Colour: symbolic triad'; 'Facades: Chinese-American'; 'Facades: Ndebele'; 'Facades: Perambouco' (a place in Brazil); 'Fenestration: Tibetan'. There is no attempt to relate the different essays, and at first I found the minestrone of ideas and themes in the 'principles' section indigestible and irritating. Then it began to have some of the wonder of the extremely odd categorizations made by medieval naturalists which so impressed Borges. Now, I find the section much the most gripping, and dip into it regularly for pleasure, for one bizarre conjunction of ideas and forms leads to speculations about another, and pages begin to be turned furiously back and forth. Oliver emerges as a person of great quirkiness, charm and fun, as well as having huge erudition.

Two splendid indexes help you to plot paths through the jungles of words and pictures. There is a glossary, comparative lexicon (in seven languages, including Arabic), and of course, the magnificent bibliography. Clearly the whole enterprise should be put on CD-ROM as soon as possible (the entirely black and white illustrations are not generally good enough to worry about loss of quality on the screen). Cambridge University Press have made a good start in architectural electronic publishing by putting Ruskin's complete works on disk. A second or parallel CD edition of the great vernacular encyclopedia will be able to have an up-to-date bibliography and be much easier to use than three heavy volumes. But, sadly, I won't be able to take the 'principles' section to bed.