Peter Behnstedt is well known for his numerous contributions to Arabic dialectology, and especially for a series of dialect atlases, including the Sprachatlas von Syrien (Wiesbaden, 1997), and most recently, with Manfred Woidich, the Wortatlas der arabischen Dialekte in four volumes (Leiden, 2010–forthcoming). Fans of the latter (among whom I count myself) will come to this latest atlas with high expectations. They will find a work with a wealth of valuable information, as beautifully presented as the Wortatlas, but suffering from several significant flaws, as Behnstedt and his co-author Aharon Geva-Kleinberger (henceforth BG-K) themselves candidly admit.

The volume contains 207 plates, with 197 devoted to mapping specific linguistic features of the dialects of 104 localities in northern Israel, from Ǧisr az-Zarqāʔ and Bartaʕa in the south-west to ar-Rīḥāniyya and Ṭūba-Zanɣariyya in the north-east. About one-fifth of the work is devoted to lexicon – specifically meanings whose realization varies in interesting ways across the dialects investigated – while the rest is divided roughly equally between phonology and morphology. As with the Wortatlas, the lexicon section focuses mainly on body parts, flora, and fauna. For example, map 179 presents a rather clear split (but see below) on the realization of “cauliflower”, between the east with zahra, and the north and west with various reflexes of qarnabīṭ. The phonology section is especially detailed in its coverage of dialectologically relevant features, including highly localized features such as the diphthongization of long vowels in pausal position (e.g. the second person feminine singular pronoun inti# > intay#), which occurs with some speakers in north central Galilee (map 48). The morphology section chiefly deals with details of the phonological realization of various free and bound morphemes, with 31 maps devoted to the morphophonemics of verbs. There is a wealth of useful contextualization in the commentaries that accompany every map, often including notes on related items and processes which do not receive their own map (e.g. p. 121 on shortening of unstressed long vowels).

The question of the provenance of the data which the maps present is somewhat fraught. The preface informs us that Rafael Talmon of the University of Haifa was a key figure in the projects that generated it. Unfortunately, Talmon's untimely death in 2004 resulted in the disappearance of much of this data, with some later recoverable from his archives, but much of it apparently lost forever. The atlas clearly suffers as a result of this loss of material, and BG-K note that they are fully aware of the resulting deficiencies (p. xi). But even the original data collection process seems to have been rather haphazard, with BG-K admitting that, as a result of both lost data and difficulties with the original data collection process “the data for the area is completely heterogeneous, and in a way chaotic”. Readers need to take this context into consideration, and to see the work not simply as a dialect atlas, but additionally as a sort of salvage operation, presenting the data that is available, incomplete and patchy as it is.

A second significant flaw of this work, also forthrightly acknowledged by BG-K, is the failure to adequately disentangle variation that is best explained (and depicted) as a function of geography, from variation that is in fact sociolinguistic in nature. Thus they write (p. xii) that “in retrospect, the whole project, right from the beginning, should have been organised as a sociolinguistically orientated project and not as a project of traditional ‘dialect geography’”. They note (p. xii) that for many localities there is nevertheless a reasonable sociolinguistic spread of informants (generally each locality is represented by between two and five informants, often with a mix of genders and ages). But since this was never done systematically with sociolinguistic aims in mind, the outcome could be seen as the worst of both worlds: no serious investigation of sociolinguistic variables is possible but, at the same time, the appearance of dialect-geographical differences in the realization of a given feature risks in fact being a reflection of sociolinguistic differences between the informants that happened to be consulted.

It is not always straightforward to interpret the symbols and special formatting used on the maps. For example, the meaning of italic font on the maps (namely that the dialect whose number is italicized is a Bedouin variety) can only be found by consulting the list of abbreviations, which is not the most intuitively obvious place for this information. On a few maps (e.g. map 66 on interdentals), braces are used, but the meaning of their inclusion at one location versus their absence at another is not explained in the keys of the maps in question or anywhere that I could see in the introductory material. The precise meaning of the tilde used on some maps (e.g. map 79) is hard (for me) to discern. I was unable to understand the explanation (p. xiv) that “the tilde means that these forms are used in other places as parallel forms”.

Clearly, then, this work is not without its flaws. But the wealth of detail and information it contains means there can be no doubt that the field of Arabic dialectology is far richer with it than without it. It should be used judiciously, however, and readers should beware of treating the maps it contains as a straightforwardly reliable depiction of the contemporary Arabic dialects of the Galilee.