This book presents a new theory of grammatical categories - the Universal Spine Hypothesis - and reinforces generative notions of Universal Grammar while accommodating insights from linguistic typology.
Date: Fri, 18 Apr 2003 13:18:32 +0200 From: W. Schulze <W.Schulze@lrz.uni-muenchen.de> Subject: Australian Languages. Their Nature and Development (2002)
Dixon, R.M.W. (2002) Australian Languages. Their Nature and Development. Cambridge University Press, hardback ISBN 0 521 47378 0, xIii, 734pp, 34 maps, Cambridge Language Surveys.
Wolfgang Schulze, University of Munich
The book under review has an interesting title ('Australian Languages' (henceforth 'AL')): It lacks what is found with most of the other volumes in the Cambridge Language Survey series, namely the definite article (compare 'The Celtic Languages' etc.). Assuming that this type of indefinite reference is chosen deliberately, one may infer from the title that Dixon's book is not just another reference book about the totality of Australian languages. In fact, Dixon's book considerable differs from the standard 'encyclopedic' presentation of a language 'group'. What Dixon aims at is to do two things at the same time: To acquaint the reader with the structural and categorial properties of the 'autochthonous' language in Australia and to both illustrate and corroborate his central thesis that the evolution of these languages cannot be accounted for in terms of simple stammbaum models. Instead, Dixon makes extensive reference to his equilibrium model that has found its first comprehensive treatment in Dixon 1997. Consequently, Dixon's 'Australian Languages' rather is a unique study in diachronic areal linguistics rather than a simple taxonomy of 'the' languages of Australia. This fact may perhaps disappoint those readers who are primarily interested in some kind of ^Ñcheck list^Ò of Australian languages, but it turns the book into an extremely interesting and fascinating 'reading book': I had difficulties to lay aside the book once I had started to read it. It tells us a comprehensive and in parts even thrilling story about the Australian languages, but it also calls for the reader's permanent attention not to lose the thread.
Fortunately enough, the book at issue represents only one side of the albeit not fully coined medal: The other side will be a 'companion volume' ('Australian languages: a complete catalogue') that 'will consist of a short account of each of the 240-50 languages, giving tribal and dialect names, traditional territory and current situation, plus a summary of the main phonological, morphological and syntactic features. Together, the two volumes will most likely satisfy all the demands of contemporary linguistics when referring to 'the' Australian languages.
The reader should also be aware of the fact that AL does not simply represent an actualized version of Dixon's 1980 volume ('The languages of Australia' (LoA)). In the 'Preface' to AL, the author makes clear that LoA can only be seen as a preliminary approach to a methodologically and conceptually validated description of the Australian languages. Accordingly, AL represents the 'mature' version of LoA. Nevertheless, the reader should not expect that Dixon simply dresses the LoA material with a new 'robe'. In fact, Dixon rarely refers to the LoA data and thus importantly extends the data base available to non-Australianists. Doing so, Dixon naturally benefits from the enormous progress the linguistics of Australian languages has made since the 80ies. Dixon thus covers most of the important findings that have come to the public since the appearance of LoA, claiming that 'in this volume I take account of all published and unpublished materials.' But he adds that he has only included 'bibliographical references that are strictly relevant to the overall thesis which is developed in the volume.'
Although AL introduces the reader to the specific perspective Dixon has taken, the book is nevertheless (more or less) theory-neutral with regards to the linguistic framework applied. This means that AL can be used by readers which camp so ever they have chosen. Nonetheless, it comes clear that Dixon's perspective neatly fits into the standard descriptive and analytic paradigm of Language Typology or ' to put it in Dixon's terms ' into Basic Linguistic Theory (see Dixon 1997).
The book is organized in fourteen chapters, preceded by lists of maps, of abbreviations, and of languages and language groups. At the end of the book, a comprehensive bibliography (roughly 600 entries) is followed by a language index and by a subject index (somewhat modest in seize). Most importantly, AL offers a number of highly informative and well-drawn maps (34 maps in sum) which allow the reader to graphically trace many of Dixon's observations and claims. Quite in accordance with the general goals of AL, the book does not give us a description of Australian languages 'language by language'. Instead, Dixon refers to a number of category-like features that serve as an anchor for the individual chapters. From a systematic point of view, both the choice of 'anchors' and their ordering are somewhat unconventional. For instance, there are word class related anchors (such as Chapter 6 'Verbs', chapter 7 'pronouns'), anchors related to morphology (Chapter 5 'Case and other nominal suffixes', Chapter 9 'Prefixing and fusion', and one explicit 'syntactic' anchor (Chapter 11 'Ergative/accusative morphological and syntactic profiles'). It is not always clear to me, why Dixon has chosen just these anchors (or: labels) instead of following a more 'traditional' arrangement. Nevertheless, the reader will soon get used to Dixon's way of presenting his findings, especially because it is coherent with the basic lines of his argumentation. Interestingly enough, it is the last major section of AL (Chapter 12), where we find a comprehensive description of the phonology of Australian languages. This ordering stands in the tradition of Dixon's presentation of Dyirbal (Dixon 1972) and other grammars of the 70ies. In parts, it reflects a 'top-down' argumentation (from larger to smaller units)' however in the case of AL, this line of arguments is not fully observed: The description of the syntactic profile of Australian languages is not put in the beginning, but 'in between'. In sum, AL first acquaints the reader with lexical features (Chapter 4 'Vocabulary'), then turns to Morphosyntax and Morphosemantics (Chapters 5 through 9), before coming back to semantically relevant features in Chapter 10 ('Generic nouns, classifiers, genders, and noun classes'). The chapter on the syntactic profile of Australian languages (Chapter 11) follows these lexical and morphological studies and again precedes the section of Phonology. The book ends with a section on 'Genetic subgroups and small linguistic areas' (Chapter 13) and with a brief summary given in Chapter 14.
The great number of Australian languages dealt with by Dixon has forced him to choose a special coding system for the language names. This system is introduced and explained on pages xxx-xIii (in addition, the author gives the relevant literature for each language). Although Dixon's system of classifying the totality of Australian languages is by itself extremely well-done, the convention he uses to refer to individual languages is somewhat difficult to assimilate by the reader. The abbreviations are hardly ever mnemotechnic (e.g. 'NG1' for Worrorra, North Kimberley Areal Group). As a result such phrasings like 'Languages with enclitic pronouns at Stage II include those in groups O, Q, T. W1, WGd, WI and NAB2. Prefixing languages at this stage include NB/f/g/h/i/k and ND-NK' (p.357) force the reader to again and again turn to the language list (as long as (s)he does not have memorized the abbreviations) - a fact that may impede the pleasure of reading AL for those not used to the impressive universe of Australian languages (note that with examples from individual languages, Dixon usually gives both, the 'code' and the language name).
The main objectives of AL are described by the author as follows: 'I attempt to characterize what the indigenous languages of Australia are like, how individual languages have developed their particular structural profiles, and the ways in which the languages are related. A portrait is provided of the Australian linguistic area, which is certainly the longest-established linguistic area in the world' (p.1). This quote illustrates the three major perspectives, Dixon has taken: The book elaborates the major typological features of Australian languages in their areal and historical settings. Hence, it is both a synchronic and a diachronic study. However, the reader should not expect that these two levels of description and explanation are dealt with separately. Rather, Dixon integrates the synchronic description into a general diachronic perspective, which concentrates on the alleged dichotomy 'genetic relationship' vs. 'areal diffusion'. Accordingly, AL frequently refers to the historical setting in which the Australian languages are thought to have evolved. The reader will thus enjoy not only the wealth of linguistic data and their historical background, but also a great number of non-linguistic references towards the emergence and diffusion of Australian cultural practices. Dixon starts with a brief portrait of the 'language situation in Australia' (pp.1-19). This short chapter prepares the reader for the presentation of Dixon's Equilibrium Model in Chapter 2. It informs on the diffusion of some basic social and other non-linguistic features of Australian societies and thus illustrates the assumption that language diffusion may coincide with the diffusion of cultural practices and cultural knowledge. In Chapter 2 ('Modelling the language situation'), the author deepens the historical perspective by introducing his concept of Equilibrium and Punctuation. The concept that has its prolegomenon in Dixon 1997 is based on the assumption that traditional stammbaum (family tree) models cannot reflect the long range history of languages, especially when they are related to a specific area. Dixon argues: 'The family tree idea is an important and useful model of one kind of linguistic relationship. It is appropriate for describing a period of population expansion and split, with concomitant split of languages. It is not, however, an appropriate model for dealing with every kind of language situation' (p.23). In fact, Dixon assumes that family tree models are especially helpful to describe periods of split that are related to his stage of punctuation. The short phases are ' according to Dixon ' conditioned by at least the following non-linguistic factors (p.33-34): Natural causes (droughts, floods etc.), material innovation, development of aggressive tendencies, and territorial expansion. The longer stages of equilibrium are characterized by a rather homogenous or contiguous cultural habitus, by the lack of dominant political structures, and by a relative high degree of interethnic mobility. After having elaborated some key arguments for this model, the author relates it to the 'Australian scenes' (pp35-40). He clearly argues in favor of an Equilibrium model to describe most of the stages of Australian history and arrives at a 'tentative scenario for the development of languages in the Australian linguistic area' (p.38-40). Accordingly, the first population of the Australian/New Guinea area would have started some 40-50.000 thousands years ago and would have been marked by a first punctuation situation. The basic topographical division of the Australia/New Guinea landmass (flat, open regions towards the Southwest, mountainous rain forests towards the Northeast, to put it into simple terms) would have caused two different types of development: The linguistic area in the flat, open regions would have been 'maintained for tens of millennia' (p.39), whereas the linguistic area in the mountainous regions would split up into more local groups. After New Guinea became separated from the Australian landmass (between 14.000 and 7.000 BC), the languages spoken in the Australian landmass (between 14.000 and 7.000 BC), the languages spoken in the forest areas of the Northeast of Australia became part of the linguistic area of Australia, 'with Australian languages infiltrating it from both north and south' (p.39). Next, Dixon gives illuminating examples for the processes of language split and language merger, both of which can be described for Australian languages. In an appendix to Chapter 2, the author carefully discusses possible aspects of punctuation that would argue in favor of the Pama-Nyungan idea and arrives at the conclusion that ''ÑPama-Nyungan' cannot be supported as a genetic group (p.53).
The objective of Chapter 3 ('Overview') is to 'provide an initial perspective on the nature of Australian languages' (p.55). Although this 'overview' is very helpful to readers not acquainted with Australian languages, Dixon nevertheless warns that 'the reader will be able to get the maximum out of the survey in the chapters which follow if they have studied one or more good grammars of Australian languages' (p.56). Perhaps, Dixon's warning is too strong: AL tells its complete story at least to those readers who are used to some kind of 'typological' argumentation. True, it would have been useful if the author had provided the reader with descriptive sections on a (limited) number of Australian languages in order to tell them 'how the systems work'. However, such sections would have expanded the volume to a dimension that would have been beyond the rational. It addition, one might have wondered which language to choose viewing the fact that hardly any Australian language can serve as an etalon for the whole linguistic area.
Chapter 3 first describes three salient semantic features of Australian languages, namely the opposition between 'actual' and 'potential', the 'volitional/non-volitional' parameter, and the general 'trend' to use 'generic terms' instead of or going with specific terms. 'Genericity' turns out to be relevant for nearly all Australian languages: 'In summary we posit an original scheme whereby great use is made of a smallish number of generic nouns and verbs, with wide meanings' (p.62). After having described some basic properties of the phonology of Australian languages, Dixon turns to a number of grammatical features: He discusses word classes (hinting at the relevance of ideophones), the relational role of nouns and adjectives, the architecture of pronouns and demonstratives etc., verbs and verbal inflection, derivational strategies, marking of possession, clause structure and constituent order, aspects of modal variation (commands, negation, questions), and the organization of complex clauses. The chapter ends with a brief consideration of special speech styles such as song style, initiation styles, and avoidance, or respect styles.
Chapter 4 (pp. 96-130) deals with the 'Vocabulary' of Australian languages. Here, Dixon not enumerates the different semantic 'classes' relevant in the languages at issue (such as kin terms, generic terms, names, adjectives etc.), but also illustrates metaphorical preferences and other semantic processes. The following domains are extensively illustrated: Flora and fauna, body parts, kin terms, artefacts, other nouns, adjectives, and verbs. Dixon usually refers to a number of 'key terms' to illustrate the areal distribution and possible genetic relations. This section includes very helpful taxonomic lists, among them a list of 68 verbal stems ordered in from a semantic perspective (motion, rest, giving, affect etc.). The chapter ends with a discussion of the lexical survey from a phonological point of view.
Chapters 5 through 11 concentrate on specific grammatical issues. In Chapter 5, Dixon describes basic patterns of case marking and other paradigms related to this topic. The main purpose of this section is not to simply list the case paradigms of individual languages, but to propose a comprehensive areal typology and functional explanation of the patterns in question. Accordingly, the first part of Chapter 5 is devoted to a portrait of case functions in Australian languages, distinguishing core clausal functions from peripheral functions, and phrasal functions from local functions. Here, Dixon adopts the descriptive and analytic framework of syntactic functions (or relational primitives) as developed in Dixon 1994. He makes a clear distinction between 'syntactic' and 'Semantic' functions and suggests that the accusative/ergative terminology should be confined to syntactic marking (p.133), whereas the semantic domain is to describe with the help of more 'semantic' terms. One might ask whether this distinction is appropriate from a cognitive point of view (see Schulze 2000): It would suggest that case markers in Australian languages are in parts homonymous (e.g. the ergative marker '((y)inggu in the Southern Kimberley subgroup which is said to have semantic rather than syntactic functions' (p. 133)). In addition, one might wonder whether it is useful to apply terms for case forms to the functional domain: For instance, Dixon uses the term 'Genitive' to indicate both the formal and the functional category. The same holds for terms like 'ergative' and 'absolutive' etc.. Nevertheless, the taxonomy of noun phrase functions as given in AL is a rather helpful tool that is applicable to languages outside the Australian area, too. This taxonomy includes the following functions: S, A, O (core), Purposive/Dative, Instrumental, Causal, Aversive (peripheral), Genitive, Comitative, Privative (phrasal), and locatives. Note that for locatives, Dixon describes the fact that '[e]very language has some grammatical marking for the three basic spatial functions (a) locative ; (b) allative ; and (c) ablative (p.142). Accordingly, Australian languages are marked for a basically 'triptotic' system as opposed to 'diptotic' systems that would show the merger of the locative with one of the directional functions. In the second part of Chapter 5, Dixon discusses the distribution of case forms in Australian languages. The main advantage of this section is that Dixon does not try to map the great variety of case forms onto a single set of 'proto-Australian' case markers. Instead he makes clear that there areal diffusion is crucial for the explanation of the current patterns. Nevertheless, he concludes: ' The evidence points towards there having been a small number of nominal suffixes at an earlier stage of the Australian linguistic area ' perhaps just our foundational cases (ergative, locative and purposive) plus comitative' (p.173).
The sixth chapter touches upon the relational domain ('Verbs'). Interestingly enough, Dixon also considers adverbs in this context, which makes sense from a functional point of view. In addition this section illustrates that many Australian languages are marked for Path Conflation rather than for Manner Conflation. The section starts with a brief discussion of transitivity in Australian languages. Accordingly, the (in)transitivity dichotomy is strictly observed in most languages, although there are several 'exceptional' valence types (such as ambitransitives, unusual case frames etc.). Simple verbs are marked by a 'simple verb root' to which derivational and inflectional affixes may be added (among them reflexes of the pan-Australian (!) derivational suffix *-dharri 'which may originally have had a basically semantic effect (indicating, say, that an action which is normally volitional is in this instance non-volitional)' (p.183). Complex verbs consist of one or more coverbs and one simple (often generic) verb. Dixon uses the simple/complex dichotomy to describe seven basic types (based on the question how many simple verbs, derived verbs and coverbs are present in a given language). Again this typology serves to describe the diffusion in terms of areal features. The section on 'Verb forms and inflections' (pp. 209-237) informs on the distribution of TAM-forms in Australian languages. The degree of variation ranges from just two categories (e.g. in Wik-Ngathan) to 'a dozen or more terms' (e.g. Panyjima, p.212). The author relates the TAM-morphology to a rather complex hypothesis about the emergence of 'conjugation classes' in Australian languages. Dixon suggests that at an earlier stage verbs could end both in a vowel or in a nasal or liquid. The non-vocalic element would later have been reanalyzed as a separate element that could merge with TAM-elements (such as imperative, purposive, or irrealis). The interaction of now 'conjugational markers' and TAM-morphemes would have led to a great number of TAM-allomorphs and new TAM-morphemes.
In Chapter 7, Dixon deals with pronouns. He starts with a discussion of categorial features, that is with the degree of semantic subclassification within the set of personal pronouns (dual, inclusive/exclusive etc.). Although in 'almost every Australian language there are different roots for sg and n[on]-sg or for min[imal] and non-minimal pronouns' (p.246), the typology of pronouns can be scrutinized with the help of the standard parameter 'lexically vs. morphologically based extension of number features'. Dixon arrives at the following conclusion: '[A]t an earlier stage, the pronoun system had fewer number distinctions, probably just sg and n[on]-sg, and that 'rrV was the n[on]-sg marker. A du[al]/pl[ural] distinction developed later and spread by areal diffusion ' (p.255-6). In order to account for the complex world of personal pronouns in Australian languages, Dixon gives a detailed discussion of both pronominal stem forms and of derivational element. This includes a highly sophisticated analysis of paradigms that lack morphological means to produce non-sg forms. Dixon concludes that the evolution of pronominal paradigms has started with a simple system (1/2(/3)). In addition, there may have been an inclusive 'as an extra-systemic term' (p.292). The interaction of the elements in this basic system has led to various kinds of reanalysis and extension resulting in the present-day paradigms. In addition to this discussion, Dixon surveys the evolution and diffusion of pronominal case forms starting with the hypothesis that at an earlier stage, '[s]g pronouns ha[d] distinct forms for S, A and O functions' (p.299), in other words that they were marked by a tripartite paradigm. Unfortunately, Dixon does not pay the same degree of attention to demonstratives as to personal pronouns. He states: 'The forms of demonstratives vary widely . There is need for a full survey across the Australian linguistic area . All I offer here are a few exploratory remarks' (p.335). Hopefully, the task of approaching demonstratives from the 'Dixonian' point of view will soon be accomplished.
Chapter 8 ('Bound pronouns') nicely extends the question of personhood in Australian languages to verbal morphology. Although bound pronouns are typical for prefixing languages, they can be nevertheless described for a number of suffixing languages, too. Usually, bound pronouns mirror bipersonal agreement (S in intransitive clauses, A+O in transitive clauses). Yet, a number of languages have extended their agreement system to peripheral function, while others (though limited in number) may perhaps show an accusative pattern (S, A). Finally, '[a] sprinkling of languages have a limited (and often irregular) set of bound pronouns' (p.345). The chapter extensively reports on the formal, functional, and categorial properties of bound pronominal paradigms in Australian languages. Most importantly, Dixon also compares these properties to the corresponding sets of free pronouns showing that there frequently is a mismatch between these two instantiations of personhood.
Chapter 9 is devoted to prefixing techniques that are relevant for verbal inflection. The number of prefixes may range from fourteen prefix slots in Tiwi to just two (fusing) prefixes in e.g. Alawa (Arnhem Land Group). In order to explain prefixing techniques, Dixon refers to basically two operations: 1) The development of bound personal pronominal clitics into prefixes, and 2) the compounding of coverb plus simple verb into a single unit (p.409). The author extensively portrays the structure of prefixing chains and also considers nominal incorporation as it shows up in about twenty of the prefixing languages (organized in three geographical blocks in Arnhem land). As to expected, noun incorporation mainly concerns nouns in S or O function. Nevertheless, nouns in peripheral function (such as instrumental or locative) may be incorporated, too, e.g. in Warray, Tiwi, Emmi and Patjtjamalh (p.427). Semantically speaking, 'a body part noun is most typically incorporable into the verb, a generic noun often is, and an adjective occasionally is' (p.427). AS can be expected from the discussion of the heuristic value of 'Pama-Nyungan', see above), Dixon also argues against the interpretation of prefixing techniques as an evidence for genetic relationship among prefixing Australian languages (in the sense of 'proto-prefixing').
Chapter 10 ('Generic nouns, classifiers, genders and noun classes') brings the reader back to semantic (and lexical) issues. Dixon starts with an analysis of generic nouns and classifiers as they frequently show up in Australian languages. He then turns to the question of the 'feminine' marker 'gan' found in some languages of eastern Australia and to gender in free pronouns (especially in the third person singular). Noun classes (marked on bound pronouns) are current in prefixing languages. Semantic 'gender' is present in at least five of the non-prefixing languages, including Wagaya, Diyari, Wangkumara, Bandjalang, and Dyirbal. It is interesting to see that Dixon illustrates the famous four class system of Dyirbal without alluding to Lakoff 1987 together with the assumption of radial categories as proponed by Lakoff. With prefixing languages, noun classes may vary from two to eight in number. Dixon claims that '[t]he variation in noun classes is consistent with the hypothesis presented here, that noun classes have developed recently, as an areal phenomenon, within the prefixing region. It is basically the category of noun classes that has diffused, with each language developing the actual marking for itself, out of its own internal resources' (p.471). On p.515 Dixon states: 'A pervasive theme of this book is the alternation between ergative and accusative schemes of morphological marking in Australian languages'. In fact, many of the parameters, categories and semantic or syntactic features referred to so far are structurally coupled with strategies of clausal organization (see Schulze 2000). Chapter 11 ('ergative/accusative morphological and syntactic profiles') is intended to bring the reader back to this central point of grammatical organization. Here, Dixon first recapitulates the means used to mark relational behavior in terms of morphology. The author assumes that originally, nouns were marked for an ergative behavior, whereas personal pronouns were marked for an accusative behavior. This common pattern would then have been rearranged and reanalyzed in different ways, leading to both 'pure' ergative and 'pure' accusative patterns in some languages of Australia. In addition, Dixon discusses the question of how the syntactic feature of pivothood interacts with clause internal strategies of marking relational behavior. He distinguishes languages with no syntactic pivot from languages with an S/O syntactic pivot and from languages with a 'mixed' pivot (e.g. S/O pivot for nouns and S/A pivot for pronouns in Yidinj). Switch reference as a specific type of pivotal behavior can be found especially in the central and western areas of Australia. The author then correlates pivothood to strategies of antipassivization (S/O pivot) and passivization (S/A pivot). Finally, he looks at a number of shifts in profile. He comes to the conclusion that '[t]here does appear to be something of an overall trend towards a more fully acc[usative] system, but there are also languages moving in the opposite direction. The great majority of languages retain both erg[ative] and acc[usative] elements in their grammatical profile' (p.545-6).
The survey of linguistic features in Australian languages ends with an in-depth study of phonology (Chapter 12). Dixon starts with two important phonetic observations that stem from the feather of A. Butcher (forthcoming): First, 'the lowering of the velum for nasal consonants tends to be delayed as long as possible' (p.547). As a result, nasalized vowels rarely occur. In addition, this tendency may result in 'prestopped nasals' that may become distinct phonemes in a number of languages. Second, 'in a stressed syllable, the pitch peak tends to occur relatively late in the syllable' (p.547). As a result, the syllable onset is relatively weak, whereas a coda consonant tends to be strengthened. These two tendencies can be related to a more general articulatory 'habitus' that accounts for the relative close similarities in the phonological systems of Australian languages. Dixon describes the 'canonical' sytem of Australian phonological organization and then relates individual phonetic features (laminals, apicals and rhotics) to areal distributional patterns. Other phonetic aspects referred in this chapter include initial dropping and medial strengthening, stop contrasts, fricatives and their historical development, the question of glottals, vowels, and phonotactic features. The chapter on 'Phonology' is especially important because here Dixon illustrates and discusses a vast number of phonetic processes that are typical for certain 'blocks of languages'. It comes clear that many such processes cannot be accounted for in terms of simple 'sound changes' as described for punctuated situations of language split.
This point brings us back to the overall model of how languages have evolved in Australia. In Chapter 13 ('Genetic subgroups and small linguistic areas'), Dixon resumes this question. He again stresses that the view of an Australian genetic macro-family 'cannot be sustained when the proper methodology of comparative and Areal linguistics is applied to the Australian situation' (p.659). Nevertheless, Dixon does not claim that the 'Australian situation' is characterized by complete entropy. He shows that a number of low-level genetic subgroups can still be described. Here, he discusses the following groups: the north Cape York subgroup, the Cairns subgroup, the Maric proper subgroup, the Central Inland New South Wales subgroup, the Wannji/Garrwa subgroup, the Yolngu subgroup, the Northern Desert Fringe (putative) subgroup, the Ngarna subgroup, the Tangkic subgroup, the Maringrida (putative) subgroup, the Mnid subgroup, the Kitja/Miriwung subgroup, the South Kimberley subgroup, and finally the North-west Arnhem Land (putative) subgroup. In sum, the author thinks of about forty low-level genetic subgroups ('mostly consisting of just twoo or three languages' (p.691). On the other hand, Dixon suggests a number of smaller linguistic areas with languages 'hav[ing] much greater similarities to other languages in the area than to anything outside the area' (p.668-9). The following smaller areas are described: Lower Murray, Arandic, North Kimberley, and Daly River. For a number of (genetic) subgroups, Dixon develops a scenario of expansion. For instance, he claims that North Cape York 'is basically of non-Australian type, but with some Australian substratum' (p.681). Maric seems to have expanded from the coasts of the Coral Sea to the inlands. The books ends with a brief summary (pp.690-699), which gives a fairly good though rather condensed overview of the claims and analyses put forward in AL. In his final paragraph, Dixon states: 'The Australian linguistic area poses problems of investigation and analysis unlike those found anywhere else in the world. The established methods of historical and comparative linguistics, which can be applied so successfully elsewhere, have limited appropriateness in Australia' (p.699). Perhaps, this claim too strongly emphasizes the uniqueness of the 'Australian situation'. A cursory look at for instance the 'East Caucasian situation' will reveal that other non-Indo-European areas, too, face the same kind of problems as they have been described by Dixon. In other words: It may well be that the success of the 'Indo-European' comparative method mirrors nothing but the peculiarities of the Indo-European type of language change and diversification. The worth of AL is ' among others ' the fact that Dixon proposes and applies a methodological alternative that departs from the given linguistic situation itself rather than from generalized hypotheses about language change that stem from a linguistic area quite different from the Australian situation in space and time. This does not necessarily mean that we have to adopt Dixon's methodological pathways as such to other linguistic areas. It may well be that this approach is fruitful for such other areas, too. But we should also take into account that different historical setting may result in different types of language change, diffusion and diversification. The main lesson non-Australianists learn from Dixon's book is that just as historical developments may follow both more general patterns and idiosyncratic lines, the language(s) of speech communities not necessarily develop alike.
In sum, AL is an extremely important contribution to both Australianists and non-Australianists. Australianists will probably have to work through the many details to judge whether all of Dixon's hypotheses and claims will finally 'pass the examination'. Non-Australianists will profit from AL in at least three respects: First, they are introduced to the 'Australian situation' in a way that is generally easy to follow. Sure, one has to get involved in Dixon's descriptive and analytic arguments. One has to accept that Dixon's way is stony and full of deviations, windings, and sometimes perhaps too suggestive short cuts. Many will object to some aspects of this way, but for the time being it seems that there is no other way to go. Second, the book can be used as a good instruction to the typology of Australian languages, disregarding whether or instruction to the typology of Australian languages, disregarding whether or not one accepts Dixon's 'Diachronic Areal Typology'. Third, the book also shows how to approach linguistic categories from a descriptive point of view. Not every category or function discussed by Dixon will withstand the critics of Theoreticians, which camp so ever they belong to. Yet, AL opens the way towards a descriptive mode that seems to be applicable to other linguistic areas, too.
Butcher, A. (forthcoming). The phonetics of Australian languages. Oxford: OUP. Dixon, R. M. W. 1972. The Dyirbal language of North Queensland. Cambridge: CUP. Dixon, R. M. W. 1980. The languages of Australia. Cambridge. CUP. Dixon, R. M. W. 1994. Ergativity. Cambridge: CUP. Dixon, R. M. W. 1997. The rise and fall of languages. Cambridge: CUP. Lakoff, G. 1987. Women, Fir, and Dangerous Things. What Categories reveal about the Mind. Chicago: CUP. Schulze, W. 2000. The Accusative Ergative Continuum. General Linguistics 37:71-155.
ABOUT THE REVIEWER:
About the reviewer: Wolfgang Schulze is the Head of the Institute for General Linguistics and Language Typology at the University of Munich (German). His main research topics include among others Language Typology, Cognitive Typology, Historical Linguistics, language contact, the languages of the Eastern Caucasus, and 'Oriental' languages. He currently works on a Functional Grammar of Udi and on a comprehensive presentation of the framework of a Grammar of scenes and Scenarios in terms of Cognitive Typology.