A Ulithian Grammar

Sohn, Ho-min (1973). A Ulithian Grammar. Canberra: Australian National University, Research School of Pacific Studies, Dept. of Linguistics. ISBN 978-0-85883-098-1.

• Has attachment: File:Z6ML3UC4.pdf

Abstract: The present work is a slighlty revised version of H.M. Sohn's University of Hawaii dissertation (1969) which was written under the direct supervision of and in close cooperation with B.W. Bender. In the writing of the original version, the phonology portion was constructed in large part along the lines of ideas provided by Bender, whereas the syntax portion generally followed those of Sohn. Since essentially nothing was known of the Ulithian language before Sohn's five months of field work on the island in 1968, a major aim of the work was to provide the reader with as much of an overall picture of the language as possible while using a generative-transformational model that was current in the mid-sixties. A wealth of example sentences are provided to assist the ambitious reader who wishes to reinterpret portions of the syntax along more recent models, as the authors themselves undoubtedly would if they were to start from the beginning again. As a matter of fact, an enormous number of works have recently appeared which provide far more refined options within linguistic theory than those available at the time when this work was done. Fresh arguments are being made in phonology regarding such topics as abstract ness in base form phonemicisation, rule ordering, and direction of rule operation. A number of proposals are being made in favour of generative semantics over generative syntax. If these recent theoretical developments were taken into account, this work would no doubt be refined and reorganised. For example, negation has been treated no higher than the level of a lexical category in view of its morphological characteristics. In the framework of generative semantics, negation would certainly be dealt with at a much higher level, which would show parallelism with the other languages of the world in terms of linguistic universality. The authors' recent work on Woleaian, the language closest to Ulithian, has revealed that various hypotheses made in this book are basically correct, in that similar hypotheses may be applicable to Woleaian also. For example, reconstruction of lost final vowels in Ulithian base forms is supported by devoiced reflexes of such vowels in corresponding Woleaian forms. On the other hand, certain hypotheses made with Ulithian evidence alone could be improved or modified if parallel Woleaian evidence were taken into account. For example, the final vowels in such Ulithian base forms as lage 'sky ' , tale 'rope ' , cale 'water', fase 'stone', and mwale 'man' are not pronounced but are reconstructed in view of their respective construct forms: l a g e l ' s ky of', talel 'rope of', calel 'water of', fasel 'stone of', and mwalel 'man of', while such particles as cox 'just' and gaag 'I' are not provided with a final vowel because they do not have a construct form on which the final vowel may be set up. If corresponding Woleaian voiceless final vowels were taken into consideration, the above Ulithian base forms would be modified as lagi, tali, calu, fasu, mwale, caxu, and gaagu, together with a rule that lowers i and u to e. In fact, the authors have already postulated such a rule (PRI9) to handle a minor related case. With the modified base forms as given above, PR19 would acquire greater generality, while many rules which were motivated by the existing set of base forms could be collapsed into a few general rules. For example, PRs 24, 25, 28, 29, and 31 could be combined into one rule of assimilatory vowel raising (a-raising between high vowels) if we revise the Ulithian base forms taking Woleaian evidence into account. Finally, the authors have noticed several places where Ulithian data themselves are not satisfactorily dealt with, due probably to incomplete elicitation or lack of deeper analytical insights on the part of the authors. For example, inclusion of capital letters in base forms to block compensatory lengthening will have to be re-examined, since a brief recent check has revealed that blocking of the lengthening might have been caused by inconsistent elicitation, hence a case of possible overdifferentiation. In spite of the abovementioned unsatisfactory areas which call for further study, the authors would consider it sufficient reward if this work were of any small help to the reader.

Extra details:

DOI: 10.15144/PL-C27
MAG: 659361241
CorpusID: 60260271