Chapter 12 – Sketch grammar
305
(4)
Be-an rake tea
nao,
ean paa nao papahiana.
if- 2
SG
want
COMPL
go 2
SG
TAM
go alone
‘If you want to go, you go by yourself.’
(Aro 5.142R)
where paa evidently does not mark past tense, and came to realize that
from the very beginning the exceptions should have been taken seriously.
Our current analysis suggests that paa does not locate an event in time, but
indicates a change of situation be it in the past or the future. Since this find-
ing also questions the glossing of other
TAM
markers, we now gloss all
TAM
markers simply as ‘
TAM
’ and compensate for this underspecified glossing
by a short description of each
TAM
marker in the sketch grammar. There we
precisely state our hypotheses and mention open questions such as the
compatibility of
TAM
markers with negations or temporal adverbs.
An alternative solution would be to give the
TAM
markers specific, but
semantically neutral glosses such as
T
1,
T
2, etc. which can be easily searched
for and be replaced by more meaningful glosses once their functions are
better understood. But also in this case, the LDSG should explicitly explain
the meanings of the glosses and describe the relevant grammatical phenom-
ena in a way that accounts for the preliminary status of the analysis and
thus minimizes the danger of misinterpretations.
2.3. User-friendliness
To be user-friendly, a LDSG must meet the user’s needs, i.e. to quickly get
an overview of the essential features of the language and all the information
necessary for using the annotated recordings in further linguistic and re-
lated research. Therefore the LDSG should be short and the grammatical
facts be clearly presented in a format that follows common practice, for
instance one that presents the description in an ascending manner with the
chapter on phonology first and that on complex sentences last.
All terms whose meaning is vague or variable in the linguistic literature,
e.g. adverb, particle, or is presumably only known to specialists, e.g. appli-
cative, should be defined, and all abbreviations be explained. The defini-
tions should, however, be short and only briefly state which language spe-
cific properties have been relevant for the selection of the terms in question.
306
Ulrike Mosel
3. Content
The content of the LDSG of a previously unresearched language depends on
–
the structure of the language;
–
the state of analysis;
–
how much and what kind of information is provided in the annotations
and the lexical database.
However varied these things are, a number of characteristics of the language
should be described in the LDSG in addition to the list of abbreviations
used in grammatical glossing. These are
–
charts of the consonant and the vowel system; a note on syllable struc-
ture, and the most important phonological processes; and a statement on
how the orthography and/or transcription used in the documentation re-
lates to these phonological characteristics (see Chapters 9 and 11);
–
an overview of the word classes and the grammatical categories (tense /
aspect, number, person, gender, case etc.) in order to facilitate a better
understanding of the glosses;
–
inflectional paradigms as these are very difficult to extract from text
corpora and are hardly ever fully represented there;
–
word and constituent order rules that would help the user to quickly
understand utterances.
In addition, it might be useful for linguists, especially typologists, to include
a list of important typological features and for ethnographers to add a sec-
tion with notes on lexical and phraseological characteristics (see Chapter 8).
Strictly speaking, the latter do not belong into a grammar but unless they
are dealt with in a different part of the language documentation, they may
be accommodated here.
3.1. The LDSG and the lexical data base
Since the entries in a lexical database contain information on word classes
and subclasses, the sketch grammar only needs to explain the principles of
word classification and briefly characterize each class to facilitate the un-
derstanding of the abbreviations used in the lexical database and the anno-
tations. For instance, in the Teop sketch grammar it is sufficient to say that
Chapter 12 – Sketch grammar
307
causatives are derived by the prefix vaa- from intransitive and transitive
verbs, but it is not necessary to investigate and describe which semantic
types of lexemes have causative counterparts and what kind of argument or
role structure they have. Any typologist interested in causativization can
find out the rules by checking the lexical database and the text corpus for
causatives.
Another example is the semantic motivation of noun class assignment to
one of the three noun classes in Teop, the a-class, the e-class and the o-class.
Since each noun in the dictionary is classified as n.a., n.e., or n.o., the sketch
grammar only needs to describe which role the noun classes play in syntac-
tic constructions and give a rough idea of their semantics. A thorough analy-
sis can wait until there is time for a specialist investigation.
3.2. The LDSG and the corpus of annotated recordings
To quickly find and understand interesting grammatical constructions, a
large corpus of recordings with interlinear glossings and a free translation is
not sufficient or at least not always practical. Imagine you are interested in
a very frequent conjunction like that in English. Your search gives you over
1000 examples, but 98% are uninteresting, you need the 20 exceptional ones
to find evidence for your hypothesis or falsify it (for a similar example see
Chapter 9).
Searching can be made easier when the annotation is accompanied by
notes on conspicuous grammatical phenomena and they are given easily
searchable names like
COMPLEMENT CLAUSE
. In this manner you can even
make notes on constructions that are not characterized by a particular lin-
guistic unit, e.g. a particle or bound morpheme, and for example note down
“juxtaposed/asyndetic
COMPLEMENT CLAUSE
.” Apart from creating a useful
tool for research on typological phenomena or the compilation of a refer-
ence grammar, making such notes is intellectually stimulating and helps to
cope with the sometimes boring task of transcribing and translating. Fur-
thermore, these notes give linguists the chance to document their insights
into the grammatical structure that they or others can make use of later. As
futher dicussed in Chapters 1 and 9, a language documentation requires a
thorough analysis of the language, but it does not necessarily leave room
for writing up a comprehensive reference grammar. A well-planned combi-
nation of a sketch grammar and grammatical notes in the annotated corpus
of recordings can to some extent compensate for this.
Dostları ilə paylaş: |