|AUTHORS: Nekes, Hermann; Worms, Ernest A.
EDITOR: McGregor, William B.
TITLE: Australian Languages
SERIES TITLE: Trends in Linguistics. Documentation 24
PUBLISHER: Mouton de Gruyter
Claire Bowern, Department of Linguistics, Rice University
_Australian Languages_ is a compilation of primarily grammatical and lexical
materials from a number of Indigenous Australian languages. It is based on
materials collected in the 1930s by Hermann Nekes and Ernest Worms; their
manuscript was published in microfilm format in 1953, complete with hand-written
corrections to the typescript. While the 1953 manuscript was very useful to
those working on North-Western Australian languages, it was not very accessible.
The microfilm was hard to obtain and it was difficult to work with. The age and
orientation of the work also meant that one already needed a good deal of
background in the relevant languages before being able to use it, even though it
contained a great deal of valuable material.
The title of the work is a misnomer; the great majority of data come from
Nyulnyulan languages (spoken in far North-West Australia) and their immediate
neighbors, although sporadic other languages are also quoted, and there is an
appendix on Dyirbal. The book has the format of a comparative grammar, and
covers such topics as noun and verb morphology, ''phonetics'' (Nekes and Worms
have no concept of the phoneme), adjectives, pronouns, adverbs, exclamations
(interjections, curses and good wishes) and a text collection. The massive
comparative dictionary is not printed in the hard copy of the book but is
available on the CD-ROM which accompanies the book (and is available in XML,
formatted html, and in Toolbox backslash codes). The CD also has the texts, a
facsimile of the pages keyed to the original manuscript, and some sound files
William McGregor has edited and annotated the original manuscript. McGregor has
left the original largely intact, preferring to annotate rather than correct.
There is a general introduction to the book by McGregor, including the history
of publication, some background to data collection, information about the
authors and their knowledge of linguistics, and an explanation of the
conventions of the book. This is followed by the authors' (that is, Nekes and
Worms') introduction. Each chapter also has an editor's introduction. Therefore,
this book is an edited volume where the original authors' insights and analysis
come through, rather than a book which represents our current state of knowledge
of Nyulnyulan languages. This is simultaneously an advantage for students of the
history of linguistics and a disadvantage for those using the book as a
comparative grammar of Nyulnyulan.
The phonetics chapter covers the sounds found in the languages as a whole, with
notes on differences between the languages. There is some note of the major
correspondences (e.g. Bardi o: corresponding to Nyulnyul obo), although Nekes
and Worms do not draw any historical conclusions or reconstruct any forms. The
chapter on nouns includes not only inflectional morphology but also some
comments on lexical semantics and cultural items (for example, terms for
seasons). Again, comparisons are drawn but these are not specific claims of
cognacy. The chapter also covers derivational morphology, such as compounding
strategies and semantics. There is considerable over-etymologizing in this
chapter; for example, there is no reason to suppose that Bardi garrabal 'bird'
is bimorphemic, containing a morpheme -bal of unknown meaning.
In the verb chapter, there is extensive information about both regular and
irregular verbs, and here the differences among the languages are clearer. The
''adverbs'' chapter also includes conjunctions such as translation equivalents of
'or'. The texts section includes both traditional narratives (some of which
contain gender restricted material) and oral history, in several languages. In
the case of the Jabirr-Jabirr sources, these texts appear to be our only source
of extended language and so are especially valuable. There are also many example
sentences in addition to the texts at the end of the work, so there are plenty
of primary materials to work from.
I have split the critical evaluation into two sections. First I discuss Nekes
and Worms' component of the book. I then focus on McGregor's editorial work
before giving a general evaluation.
To be honest, Nekes and Worms' scholarship is pretty naive in some areas, even
taking into account when they were working. This is not a book by Boas, Sapir or
Bloomfield. I suspect that at least some of the problems can be attributed to
problems in transcription. For example, in not notating the difference between
homorganic nasal-stop palatal clusters and heterorganic apical nasal-palatal
stop clusters (nyj and nj in Bardi orthography) they miss an important
difference between intransitive and transitive verb conjugation for
palatal-initial roots. Compare, for example, inyjoogoolij 'it broke' with
injoogoolij 'he broke it'. (On the other hand, it's hard to imagine how a
statement like ''Australian aborigines attach a greater importance to consonants
than to vowels.'' (p58) can be attributed to transcription errors alone.)
I strongly suspect that Nekes and Worms regularized their comparative data; for
example they quote forms in Nyikina with long vowels, where only the cognates in
Bardi have long vowels (Nyikina has lost all vowel length contrasts). Some of
the Bardi forms are suspicious and may reflect standardization towards Nyulnyul
(the most clear are the subsection terms; although they are attributed to Bardi,
Bardi people have never used such terms, and have a different kinship
structure). Comparison with Nimanburru data recorded by Anthony Peile in the
1960s indicates that the Nimanburru forms in this book are in general closer to
Nyulnyul than Peile's data; however, because we know very little about the
circumstances of collection of either source, it isn't known whether these are
real (e.g. dialectal) differences or Neks and Worms' editorial intervention.
They are inconsistent in segmentation and they miss other morpheme boundaries
(e.g. the continuative suffix in 'present' stems, and the spurious 'ma-agolen'
vs 'm-agolen' 'break' (Bardi)).
Another problem is language attribution. For example, the Bardi dative is given
as -dj [-j in the current community orthography] but we are then given a whole
lot of examples from Nyulnyul. The dative in Bardi is extremely rare in both the
modern language and in texts recorded in the 1920s; therefore, although the
suffix exists in Bardi it does not have the same function and distribution as
the other Western Nyulnyulan languages. We therefore get a picture of
overgeneralized homogeneity and we should be cautious about extrapolating too
far about the behavior of one language in the family from another. Set against
this, however, are the number of parallel translations, which greatly facilitate
While I have been critical of the linguistic analysis in _Australian Languages_,
it is still a valuable work. This book is an important resource for a number of
languages which are no longer spoken and for which this book is almost the sole
source. It is an early example of comparative work and is of interest for the
history of linguistics (especially in Australia). Many of the problems I mention
here are also noted by McGregor.
I now turn to the editorial comments. McGregor has done a great deal to make the
1953 edition more usable, and the CD adds a lot to the book, for both audiences
using the book for language data and those who are focused more on the history
of linguistics and missionary linguistic work in Australia. The structure of the
original is maintained but the text is annotated with error corrections,
explanations for non-specialists, and references to more recent work. This is a
complex book to work from, though. For example, McGregor's notes appear in
several different places; at the bottom of tables, in the text itself, between
, in the endnotes, and some items are unmarked (but the reference to Metcalfe
1975 must be the editor's).
There are a few mechanical issues that I want to mention, too. These are minor,
but given the cost of Mouton's books, at more than 50c per page I think we have
a right to font kerning. The engma seems to have been inserted from a different
font. A syllabic mark is used instead of an underdot to mark retroflection, and
apparently [p xxix] it wasn't possible to use a capital engma in the text. There
is misalignment in the tables and numerous diacritics are poorly spaced with
respect to their host character. The index appears to be missing a lot of
languages; for example Goa (Guwa) is given in the language list but not in the
index. (This is where being able to search the pdf is useful.)
Finally, a book like this brings up questions of context. In 1953, the year the
manuscript first appeared, Aboriginal people in Australia could not vote, could
not enter pubs, and in many states could not marry without the consent of an
overseer. Opportunities were very limited and 'light-skinned' children were
still being taken from families without their consent and placed in orphanages
and foster homes (the Stolen Generations). The fact that anyone found the time
and energy beyond survival for language work is testament to a strength and care
for culture which should be acknowledged, and is reflected in McGregor's
dedication of the book to the Indigenous people with whom Nekes and Worms were
known to have worked.
ABOUT THE REVIEWER
Claire Bowern is an Assistant Professor of Linguistics at Rice University. She
is a historical linguist who works on Australian Indigenous languages. She is
the other Nyulnyulan specialist.