1.2 Main features of the language.

Although attaching simple typological labels to a language as immensely difficult as this one would be arbitrary and unhelpful, I will try to mention a number of features which characterise it and which perhaps cohere with each other in some way, in order to provide an initial orientation to readers:

  1. word order flexible (§15.4);
  2. multi-word syntactic units like NP, VP, and clause are difficult to justify except as informal approximations or potentialities (for ex., coreferential elements in a "NP" may be scattered, and can be analysed as appositional), see §15.2, §15.3, §15.5;
  3. intonational/breath groups ('strings') only occasionally coincide exactly with (intuitive) "clauses," because of frequent pauses within a "clause" and because two or more "clauses" may be run together (§15.5);
  4. direct marking of cases is also rather casual, since zero (Nominative) is used for subject and direct object, and often as an option for other case categories (§4.18ff.);
  5. elaborate system of noun-class affixation, with different subsystems used for nouns, pronouns, demonstratives, and pronominal prefixes (subject-object) in verbs (§4.7ff.);
  6. proliferation of morphology, with each word-class (nouns of several types, personal pronouns, demonstrative pronouns, cardinal-direction adverbs, verbs) developing its own highly unique set of forms and categories;
  7. direct influence of discourse factors in much of the morphology--choice among two series of noun-class prefixes or absence of prefix (§4.8), use of a whole series of personal pronoun formations expressing different kinds of focus and emphasis (§6.8-10), use of demonstratives separately marking nuances of discourse status in roots (§7.1), prefixes (§7.4), and suffixes (§7.6-7), use of reduplication and compounding in some types of repetitions of verbs (§8.3-4, §14.8, §17.5-6), use of distinctive intonation patterns (§17.4), and use of a number of particles with framing effects (Chapter 12, §17.9), in addition to partial discourse determination of word order (§15.4);
  8. extensive morphological and syntactic interactions between tense-aspect, mood, and negativity (with effects on noun-class prefixes on nouns and pronominal concord prefixes on verbs), so that instead of separate formal expression these categories tend to blend into interactive configurations (Chapter 8, §15.6);
  9. "possession" not a unified phenomenon, rather an umbrella label for a number of quite distinct formal mechanisms, including propositus[Ego-of-reference] indexing in kin terms (Chapter 5), derivational ("inner") noun-class concord for whole-part harmony (§4.7, §4.9), and genitive usage of a broader "Relative" case marker (§4.30, §6.4), along with special forms for predicate genitive (§6.5), see overall discussion in §15.8;
  10. use of direct rather than indirect quotation (§16.3);
  11. "subordinated" clauses including Relatives having semi- autonomous status, formally with normal inflected verbs and nouns plus a subordinator (§16.2, §16.4ff.);
  12. use of a small number of interjection-like "root forms" as uninflectable forms with the meaning of an inflected verb, for certain meanings (§12.2);
  13. use of some lexical nouns as abstractives with verbal meaning (§4.33);
  14. use of some lexical nouns with agentive meaning (§14.21).

It is equally important to indicate some features of the more familiar languages which are not present:

  1. no infinitive or similar reduced complement clause type;
  2. no productive participial (deverbative) formation;
  3. no productive verbal noun form (derived from verb);
  4. no relative-clause construction requiring coreferential head noun (though such coreference is occasionally overtly marked in the Nunggubuyu "Relative" construction, §16.4);
  5. no switch-reference system or other cross-clause rules triggered by coreferentiality (for a minor and somewhat dubious exception see the 'to want' construction described in §16.8);
  6. no copula;
  7. no formal distinction between imperative and future (except for 'come:', §8.5);
  8. aside from dubious 'to want' construction mentioned in (s), no complex surface constructions with a "higher" verb like 'can', 'must', 'promise to', 'persuade to', etc. (see §16.9 for Nunggubuyu counterparts);
  9. no passive or antipassive rules creating demoted (chômeur) nouns (for derivations roughly similar to agentless passive and objectless antipassive see §10.5 and §14.9).

It is not appropriate to launch here into a long typological disquisition involving data from many other languages. However, we can say that some of these positive and negative features make sense, either by direct reference to nonlinguistic way of life (use of a few lexical, unsegmentable nouns with abstractive verbal or agentive participial meaning, instead of having productive deverbative formations, reflecting small number of traditionally institutionalised "activities" and absence of occupational specialisation), or by reference to others among the features listed. In particular, the prolific noun-class indexing system (e) and the use of discourse-controlled affixation in all major word-classes (g) seem to be centrally important, compensating for (and thus permitting) fluid word order (a), looseness of phrasal and clausal structure (b-c), casualness of case marking (d), looseness of subordination (j-k), and several of the negative features, cf. also the use of noun-class derivational marking for one type of "possession" (i).

In these respects, Nunggubuyu seems to have gone farther than even the neighbouring Aboriginal languages which I have also worked on (Ngandi, Warndarang, Mara, Ritharngu, Dhuwal, etc.). For this reason, and because these publications on Nunggubuyu are much more detailed and reflect much more effort than earlier ones on those languages, it would seem to me that Nunggubuyu is a language which deserves recognition by serious typologists, as representing maximal development of a certain discourse/grammatical strategy.