1.4 Use of this grammar.

Aside from the difficulty of arriving at satisfactory analyses of a phonologically and morphosyntactically complex language like this, I have the problem of a potentially diverse readership with quite different interests--teachers and literacy personnel at Numbulwar, academic linguists specialising in Australian languages (perhaps mainly interested in historical reconstructions), and other linguists with different theoretical specialties.

I think it would be advisable for most readers approaching the language for the first time not to try to read the grammar chapter by chapter, like a novel. I would suggest an initial skimming around, with subsequent more concentrated reading of those chapters or sections of most immediate interest. For example, someone interested in comparative Australian linguistics (e.g., the genetic position of Nunggubuyu) might find the verbal morphology, notably the pronominal prefixes (Chapter 9) and the inflectional suffixes (Chapter 11), the most useful for purposes of reconstruction and genetic classification, since these are relatively closed paradigmatic systems. Such a reader will also have to learn the phonological rules (Chapter 3) fairly well.

A teacher or literacy person would probably do best to concentrate on the morphology chapters, with occasional reference to relevant phonological rules; the outline of the verbal inflectional system in Chapter 8 should be learned well.

A theoretical linguist interested in morphosyntactic typology (or coming from a typologically concerned theory, such as relational grammar or ergative-syntax theory) should start from the list of major typological features in §1.2, above, and should begin by following up the references given there to various specific sections in later chapters. If such a reader has a functional-discourse orientation, or is at least willing to consider such an approach, I recommend Chapter 17 (mainly a discussion of a particular text and the various morphosyntactic features which interact in it to produce referential clarity and discourse coherence) as a basic point of departure, since it shows the various morphological forms discussed elsewhere "in action," so to speak.

A more traditional syntactician will probably be attracted mainly to Chapters 15 and 16. I hope, however, that those chapters make it clear that "syntax" is difficult to separate from "morphology" and "discourse" even to the extent this separation is possible in analysing English; some central conclusions of Chapters 15 and 16 suggest there is no purely syntactic solution to many of the problems raised there.

A theoretical phonologist will have to learn something about the morphology as well as examining Chapters 2 and 3, since many of the phonological rules are morphologically restricted, and since some "phonological" rules have a partly morphological character (both types of reduplication, ngu-Epenthesis, etc.). With regard to current debates about root-to-affix phonological cycles, such readers should note the general points in §3.54, and should go over the ordering relationships discussed in §3.52-53. It should be noted that the cases of cyclical application of rules affect the portion of the word going leftward from the root (for compound, the final root) in a somewhat different way from the rightward portion. In addition, it can be argued that a distinct "inner" cycle of rules applies within the component morphemes of the pronominal prefix of a verb, so there are two distinct clusters within with the earliest (morphophonemic) rules apply. The notions of 'contact', 'blind', and 'one-eyed' rules are developed in §3.54 as a way of looking at various boundary rules, depending on the extent to which a process affecting an outer segment of one morpheme requires knowledge of specific phonological segments on the other side of the boundary.

For linguists interested in metrical and "autosegmental" phonology, one of the more interesting rules is Initial Reduplication P-2 (§3.3), which in its productive application gives bisyllabic reduplicative segment when the stem begins with a stop, monosyllabic otherwise--thus the specification of CV(CV)-canonical shape for the reduplication is not independent of the specific segmental composition of the stem. (There is also a quite different Final Reduplication P-3). Such phonologists may also want to examine some of the insertion and epenthesis rules (P-1, -7, -20, and -21 in §3.2/8/21/22, respectively), as well as the rules affecting vowel length.

Aside from grist for particular existing phonological models, I would hope that some rather language-specific phonological patterns would be noticed. For example, a synchronically somewhat peculiar class of "paired continuants" including most but not all nonnasal sonorants acts as a set in some crucial rules (§2.5. §2.10). Related to this, a quite curious ngu-Epenthesis rule P-1 (§3.2), inserting a dummy morpheme /- ngu-/ in a partly phonologically defined environment, can be looked at as a kind of functional adjustment to the Hardening rule (P-18)--i.e., a device to counteract some threatened surface mergers. This kind of consideration is rarely noted in discussions of phonological theory.

While I welcome theoretical usage of the material in this grammar, I would suggest that modeling of the data be evaluated on the basis of whether it sheds light on some largish combination of Nunggubuyu data. Mere demonstrations that a given model, with or without patching up for the occasion, can "handle" some Nunggubuyu facts will be less impressive.

Because this is designed to be a permanent reference grammar of an interesting language, my discussion in chapters to follow is not overtly subordinated to current theoretical debates. However, many passages have been implicitly written with some such debates in mind, and usually it is not difficult to see how the data could be applied to them. The real test of this grammar, however, will be whether it continues to be useful some years hence for linguists interested in different issues and using different formalisms.