158
John B. Haviland
seemed simply to name parts of the body. Whether or not, as anthropolo-
gists have sometimes suggested, these expressions represent an implicit
theory of the anatomical distribution of emotions and mental faculties (as
we might argue, for example, with English expressions like ‘hard-headed’
or ‘hard-hearted’), or instead are simply opaque culturally conventionalized
idioms (as we might argue for ‘green thumb’ or ‘lily-livered’
20
) it was clear
that Guugu Yimithirr had a semi-productive system for generating diverse
expressions based on “body-part” tropes. (11) gives an example based on
the Guugu Yimithirr word miil ‘eye’. The only way I could document the
system was to keep my ears open (as it were) for relevant expressions in
conversation, and to try systematically to force new combinations of body-
part words with adjectives and verbs, usually yielding only guffaws instead
of new lexemes.
(11) Guugu Yimithirr expressions based on miil ‘eye’
miilgu
= (lit., eye +
EMPHATIC
suffix) awake
miil warnggu
= (lit., ‘eye sleep’) sleepy
miil nhin-gal
= (lit., eye sit) watch out, keep an eye out
miil biyal
= (lit. eye sinew) staring all the time
miil ngamba
= (lit. eye careless) unobservant, shutting one’s eyes
to something
miil waarril
= (lit., eye fly) feel faint, go crazy, faint, get drunk
21
miil bagal
= (lit., eye poke) deceive, trick, become jealous
miil bathibay
= (lit., eye bone) sharp-eyed, always staring
miil biinii
= (lit., eye die) go blind
miil gulnggul
= (lit., eye heavy) sleepy
miilgu nhin-gal = (lit., eye-
EMPHATIC
sit) stay awake ….
Consider, too, the language of Tzotzil ritual (Gossen 1974, 1985; Haviland
1987, 1996, 2000). In contexts from prayer and song to formal denuncia-
tion, Tzotzil speakers abandon ordinary lexicon and grammar in favor of a
highly structured speech style that involves parallel lines which differ in
only a single word or phrase. These parallel lines are interpreted in terms of
a standard “stereoscopic” image (Fox 1977) invoked by the paired expres-
sions. Thus, to refer to the body one can use different doublets, depending
on the context. One is highly literal, using pat, xokon ‘back, side’ as a
metonym for the whole. Another is considerably more opaque, and sug-
gests an image of humility, as in the following extract from a curing prayer,
Chapter 6 – Documenting lexical knowledge
159
where the doublet lumal, ach’elal ‘earth, mud’ (both in possessed form)
refers to the patient’s body or self.
22
(12) From a Zinacantec curing prayer
ja’ me ta jmala lalumale
I am waiting for your earth.
ta
jmala
lavach’elale
I am waiting for your mud.
A further example is the doublet in Zinacantec ritual speech to refer to
liquor: xi`obil, sk’exobil, literally ‘cause for fear, cause for shame’. Such
expressions share properties with euphemism, always a problematic phe-
nomenon for lexicography that requires careful ethnographic fieldwork.
Systematic elicitation reveals little about the overall system of imagery in
ritual language, although it is an essential part of the language’s expressive
power. Laughlin’s (1975) dictionary of modern Zinacantec Tzotzil anno-
tates and illustrates words that participate in parallel constructions under
the rubric ‘ritual speech’. In my own work, I have relied on exhaustive re-
cording and transcription of prayer and other genres that employ parallel-
ism to expand on the list of doublets.
5. Conclusion
When does documentation of the lexicon end? While the lexicon is a re-
pository for the exceptional and the chaotic in language, it is also a site of
considerable regularity and productivity. Nonetheless, field lexicographers
like Laughlin express doubts about how well structured or widely-shared
lexical knowledge is across a speech community, basing his skepticism on
elicitation with both Zinacantec peasants and Washington D.C. university
students. Notoriously difficult even for well-studied languages is distin-
guishing between ‘literal’ and ‘figurative’ or tropic uses of words: older
Tzotzil speakers describe airplanes as xulem k’ok’, literally (as we say)
‘buzzard fire’ or telephones as ch’ojon tak’in ‘wire of metal’ – enduring the
giggles of younger speakers (who simply use a Spanish loan instead). Even
more difficult is distinguishing obscure polysemy from simple (but for-
mally unpalatable) homonymy. Laughlin’s Tzotzil dictionary posits two
homonymous roots, jav (2) – a positional root meaning ‘belly (or face) up’ –
160
John B. Haviland
and jav (1), a transitive verb root meaning ‘to chop in half’ because the two
meanings seem divergent enough to warrant separate entries. However,
Zinacantec folk etymology conjures a succinct image that connects the
senses: when you split, say, a log in two (using a verb based on jav(1)), the
two halves fall “belly up” (jav(2)). This is thus a case of covert polysemy,
23
or perhaps of underlying monosemy of a single root with different gram-
matical costumes. Such phenomena may remain intractable throughout a
lexical documentation project.
Similarly, how much ought the lexicographer to include of what might
be labeled “erroneous usage” – malapropisms, puns, or nonce creations?
Zgusta (1971: 56–57) distinguishes “systemic” from “occasional” uses of
words. An author may use ‘bondage’ occasionally to mean ‘marriage,’ with-
out thereby changing the systemic meaning of either term. Zinacantec men,
during several weeks of ribald gossip sessions in 1970, coined what was at
the time a highly creative Tzotzil sexual euphemism using a loan inyeksyon
from Spanish inyección, at a time when hypodermic injections were still a
relatively novel foreign introduction. Some of these men still jokingly use
the term almost 40 years later. The word is not in Laughlin’s Tzotzil dic-
tionary – but perhaps it should be.
Finally, questions already mentioned about aims and audience – for
whom is a lexical database produced? to what ends will it be put? – com-
plicate decisions about what words must be documented and how. The
problems are especially vexed when a lexical database may serve as the
basis for standardization or stabilization, especially in the form of a pub-
lished dictionary.
24
When people can use a dictionary to look up a word, to
see how it is spelled, and to read a definition, the speech community’s
authority over “proper” usage is irrevocably altered. How much belongs in
the lexical database of a language documentation project is thus never sim-
ply a matter of “completeness” or “coverage” but also involves ideological
decisions that may have far-reaching effects on the future of a language.
Building a lexical database is an expected part of any documentation
project, perhaps the final most demanding analytical task of all. It can be
aided by mechanical techniques applied to textual corpora and by familiarity
with the great lexicographic traditions, which have already grappled with
most of the problems a fieldworker is likely to encounter: lexical units, the
nature of meaning, the vagaries of usage, and, finally, ideologies of lan-
guage and social life. The end product is essential, but producing it relies
on both drudgery and ethnographic inspiration, on systematic elicitation
and serendipitous discovery. One inevitably (re)discovers that enough is
Dostları ilə paylaş: |