Around the globe in glottal stops

Professional linguists are pulled in the two seemingly incompatible directions of working on specific languages while thinking about language in general. On the one hand, the data corpus on which their discipline is predicated relies on carefully gathered details of individual spoken forms, embedded in social contexts and historical processes. On the other, linguists must think big and engage in comparative, historical or typological analyses. It is rare that linguists have the chance to unite their understandings of specific languages with comparative insights into the geographical distribution of linguistic features and structures, particularly in association with other colleagues. The visually rich and intellectually weighty World Atlas of Language Structures provides an opportunity for such collaboration and is a marvellous resource for any reader with an interest in linguistic forms.

The volume, referred to as Wals by its editors, is a book and CD-Rom package that displays the structural properties of the world's languages. There are some 142 world maps and numerous regional maps - all in colour - that show the geographical distribution of features of pronunciation and grammar, such as number of vowels, tone systems, gender, plurals, tense, word order and body-part terminology. Each world map shows an average of 400 languages and is accompanied by a fully referenced description of the structural feature in question. To avoid confusion, this last point should be underscored: Wals is an atlas of "abstract features of the language system that can be compared across unrelated languages", not a dialect or population atlas showing where speakers of individual languages live and interact.

The book's 700 or so pages are organised according to conventional linguistic features: phonology, morphology, nominal categories and syntax, verbal categories, word order, simple clauses, complex sentences and lexicon. Data are also provided on sign languages, writing systems and clicks. The editors have selected known authorities to write sections on their specialities, such as Ian Maddieson on phonology, and have also written a number of chapters themselves, making for a well-rounded set of contributions. Readers interested in linguistic universals will be delighted to see sections on colour terminology by Paul Kay and Luisa Maffi and a brief aside on tea by Östen Dahl.

There is a natural variation in the number of languages represented in each map, and the degree to which the features illustrate areal patterns also ary. While some maps show up to 1,370 languages, others - on specific features such as irregular negatives in sign languages, for example - draw on fewer than 40. On account of the maps, this atlas can claim to make a significant contribution to the field of areal typology, which seeks to establish whether particular geographical distributions are the result of language contact among neighbouring languages or whether these concordances have other causes. As the editors explain, one way to determine whether a shared trait among languages of a certain area can be put down to coincidence is to see if there are other unrelated traits also shared among the same, or even a similar, set of languages. Multiple shared traits in a defined geographical area provide evidence for what is commonly referred to as a linguistic area. The editors have selected features to represent as many of the major domains and subdomains of language structure as possible, relying on reference grammars of individual languages written in "theory-neutral form facilitating cross-linguistic comparison".

The interactive reference tool that accompanies the volume on CD-Rom is powerful and adaptable and allows a user to view all the maps in a variety of forms and in combination with other features. The interactive database contains additional information on languages (genealogical classification, publications, references). The maps can be zoomed, panned and saved, while dot colours and shapes can be customised, map properties are switchable (rivers, country names) and languages can be searched for by language name, family and genus name, country and region within country. A helpful Guided Tour takes a novice user through the possibilities on offer, while more advanced researchers may want to dive into the Feature Viewer, the Language Viewer or the Composer, which allow one to interrogate the database with sophisticated queries. Monolingual English users may be surprised to find certain menu items and commands in German, such as " ausblenden " and " abbrechen ", but most will find nothing wrong with a little linguistic diversity in this digital product. As with all such initiatives, one fears that this otherwise impressive CD-Rom is doomed to be out of date rather soon on account of future software compatibility issues and because the corpus of data on lesser known languages is swelling by the day.

An important question, particularly in regard to such a massive undertaking, is that of the projected audience. While it is clear who will benefit from this atlas (students, scholars, interested lay people), it is less clear how they will use it. I venture to disagree with one reviewer who suggested that "many linguists will not be able to resist curling up with this massive volume on rainy days just for the fun facts", since curling up with 4kg of book measuring 36cm by 24cm is an awkward task even for the able-handed at the best of times. In any case, linguistics is not a subject that one normally associates with coffee tables. As the editors write, the "interactive electronic version (on the accompanying CD-Rom) provides much more information, including the possibility of zooming in on particular geographical areas that seem of particular interest". Given that the CD-Rom and a website ( www.wals.info ) offer the user a richer and more immersive experience, one wonders about the utility of publishing such a heavy and expensive book. A lighter, cheaper introductory text with a few sample maps augmenting an interactive CD-Rom might have been more marketable and user friendly. The editors disagree. "The data in the electronic version of the database should be thought of as an appendix to Wals and its individual chapters" and "it is not sufficient just to refer to 'the electronic version of Wals '". But it is difficult to see what the book can offer that the interactive tool cannot.

Regardless, university libraries and linguistics departments will surely want to acquire a copy of this important atlas, which sheds light on both the typological diversity of the world's languages and how this diversity is geographically patterned.

Mark Turin is co-director, Digital Himalaya Project, department of social anthropology, Cambridge University.