Some of the Dutch is double

Presupposing no prior knowledge of linguistics... (this) book was designed to be as non-technical as possible", we are assured in the preface. Nothing could be further from the truth. Notational conventions pop up unannounced, prior knowledge assumed. Thus, in the chapter on writing systems, we meet for the first time with forward slashes, used for phonemes, and angle brackets, for graphemes, without any explanation of their meaning. An explanation of sorts does crop up 22 pages later, couched in the most awkward terms: "Each item in a class of - eme -designated things is an abstraction, its identity defined by its contrasts with all the other items in that class, and comprising a group of instantiations of the thing." A promise of a clarification is found 85 pages later, where John Laver, on linguistic phonetics, refers the reader to the next chapter "for definitions of basic phonological concepts used here, such as phoneme ". Eventually, phonemes do receive a cursory definition, half a paragraph long, buried in chapter eight, 142 pages after their first appearance. This is definitely not a handbook for hoi polloi , contrary to the assurances of its preface.

Is it then for the specialist, or even for those with a knowledge of languages?

Abigail Cohn's chapter on phonology left me speechless. She claims to have drawn her examples from American English, but her phonetic representations are at best broad Scots, for example (bet) for "bait", (bot) for "boat". And yet they cannot be, because Cohn remarks that "some of the vowels of English (for example, [i, e, o, u]) are perceived to be long in Japanese", implying that the vowel sounds in "bait" and "boat" are pure short vowels when, in fact, they are glides in most varieties of English and pure long vowels in broad Scots. Since elsewhere in her chapter Cohn clearly distinguishes between square brackets for phonetic representations and forward slashes for phonemic, one is left with no explanation other than that the printers have systematically replaced forward slashes with square brackets. But then, the remark about how /i, e, o, u/ are perceived by the Japanese makes no sense, since these would now represent either glides or long vowels. This whole chapter is absurd, and one emerges from it mentally exhausted, doubting one's sanity, one's knowledge of English and of phonology.

Bernard Comrie's chapter on the languages of the world is a 33-page-long list of language families and their members, from which you come out knowing nothing about those languages beyond what family and phylum they belong to, just as if, after having read a book on zoology, you had learnt that penguins and ostriches were birds, but did not know that they had feathers and laid eggs. More worrying are Comrie's uncritical references to Merritt Ruhlen's works, and especially to his On the Origins of Language: Studies in Linguistics Taxonomy , which has been denounced by comparative linguists as following no recognisable methodology and relying on tainted data.

The chapter on natural sign languages is not an introduction to sign languages, such as those that are used by the deaf, but an argument that they belong structurally to the family of spoken languages because they are recursive and they have syntax. The fact has long escaped some linguists that everything can be described as recursive, and that syntax can be read into any tree-shaped model, as William Calvin suddenly realised when playing Italian bowls and reported in his Lingua ex Machina , co-authored with Derek Bickerton: multi-jointed ballistics involving a tree, whence arose syntax. A leopard chasing an impala goes through multi-jointed movements, and those movements are recursive. In the process of their argument, Wendy Sandler and Diane Lillo-Martin overlook the remarkable properties of sign language - the signing space, and how those languages are fundamentally different from the corresponding spoken languages. French sign language, for instance, is as different from spoken French as Japanese is. The authors' treatment of sign languages here amounts to nothing but shoving them into the Procrustean bed of spoken languages, as all languages were once forced into the model of Greek or Latin.

Regrettably, the execrable chapters in this book cast doubts upon the rest. What should readers, ignorant as I am of clinical linguistics, think of David Crystal's treatment of the subject? Cohn on phonology is not to be trusted; is Crystal on clinical linguistics? This is unfair to Crystal.

Poor proof-reading and lack of attention to detail are evident, sometimes to the point of destroying what would otherwise be good, informative contributions. Peter Daniels's chapter on writing shows a large number of alphabets clearly laid out in comparative tables, so that you can see the resemblances between, say, Sanskrit, Tibetan and Javanese. Yet the Ogham alphabet has been subjected to such artistic licence as to be almost unrecognisable; no mention is made that Ogham was often written from bottom to top; and the claim that, like the Germanic runes, it "reveal[s]the influence of the Latin alphabet" can only be an editing mistake: Ogham looks nothing like Latin, Greek, Semitic or any other writing - it is unique. Likewise the list of the names of the Old Slavic letters has the first letter repeated at the end, four letters mentioned twice each, and uses the Russian soft and hard signs to represent consonant palatalisation or lack of it, so that this list is at once misleading and incomprehensible to anyone who does not know Russian. Add to that how the Korean word for "hand" has been mistranscribed as "so" instead of "son" - obviously a misprint - and you cannot help wonder how many more such inaccuracies have remained uncorrected throughout the book.

The 32 chapters of this volume, ranging in quality from excellent to execrable, written without regard for what precedes nor concern to pave the way for what follows, without discernible editorial guidance, make it a " Festschrift for the Unknown Linguist" rather than a handbook of linguistics, let alone The Handbook of Linguistics . Turn instead to David Crystal's more modestly titled Cambridge Encyclopaedia of Language .

Jacques B. M. Guy holds a PhD in linguistics from the Australian National University.