132
John B. Haviland
which languages to include in a multilingual dictionary, raise obvious ques-
tions. For what sort of use is a lexical database produced? What knowledge
on the part of the user is presupposed in its design? Why did its compiler
produce it in the first place? Let me review several different kinds of field
dictionaries, related to my own research in Mexico and Australia. Especially
useful to me have been the introductions to two Tzotzil dictionaries by
Robert M. Laughlin (1975, 1988), one modern and the other based on a
sixteenth-century work.
In what I call the Colonial tradition, collecting vocabularies was always
a vocation of imperialists, often an accidental byproduct of exploration and
conquest. Explorers collected flora and fauna, and often they also collected
words. Somewhat less innocent were the wordlists created explicitly to aid
in conversion, conquest, and control. The friars’ dictionaries of Indian lan-
guages in the New World, or vernacular vocabularies destined for colonial
bureaucrats in Africa and India, represented unabashedly instrumental
“documentation,” often of languages whose eventual endangerment was a
byproduct of colonial expansion in the first place. Such wordlists were
plainly not made “for” the speakers of the languages so documented.
The missionary tradition continues to produce many field dictionaries,
and reading them gives some flavor of the purposes and
populations served
by this particular lexicographic practice. In Chiapas, Mexico, the Summer
Institute of Linguistics – a Protestant Bible-translating organization – has
published many dictionaries of Indian languages from the region (Delgaty
and Ruiz [1978] for Tzotzil, Aulie and Aulie [1978] for Chol, to mention
just two), and they are widely used even by speakers who do not share the
religious beliefs of the translators. Such dictionaries are subtly infused with
cultural metacomment and religious ideology.
Here, for example, is a translation of the entry in Aulie and Aulie (1978)
for the Chol word
ajaw, reflex of a root which means “lord, master, God”
in other Mayan languages. According to the Aulies, the Chol word means
“espíritu malo de la tierra,” and they go on to comment:
They call it lak tat ‘our father.’ It is believed that a person can make a pact
with it. Such a person can make requests of the spirit for or against another.
The person who establishes such relations with the ajaw is called a “sac-
ristán.” If a man or woman offends the sacristán, the latter appeals to the
spirit to curse the other, and in a short time the other person will die.
Here both the lexicographers’ voice and its underlying ideological accent
are plainly on display. Thus, for the Aulies there is no apparent dissonance
Chapter 6 – Documenting lexical knowledge
133
between their proposed gloss, “evil spirit of the earth” and the alternate
locution “our father” (with a first-person plural inclusive prefix). Further-
more, the ‘they’ of the comment is clearly someone other than the diction-
ary writers (though perhaps not different from the dictionary users). Note
finally an interesting voicing contrast. Although the possibility of “making
a pact” with ajaw is cited as something “believed” (presumably by ‘them’),
the consequences of the appeal on the part of the hypothetical sacristán (the
term itself a Spanish loan introduced into Chol during the Catholic conver-
sion of Chol speakers following the Conquest) are given a different episte-
mological status: “in a short time the other person will die.” The dictionary
thus incorporates different, perhaps mutually contradictory stances towards
Chol beliefs and practices into the lexical entries themselves.
Slightly different is the “ethnolinguistic” lexicographic tradition, whose
immediate origins are in ethnographic research. Sticking again to highland
Chiapas, Laughlin’s exhaustive dictionary of contemporary Zinacantec
Tzotzil (1975) has the form of a traditional bilingual dictionary. The first
section gives extensive glosses (in English) of Tzotzil words, both derived
and simple, and arranged under their putative underlying roots. There fol-
lows an English index to the Tzotzil section. Laughlin’s dictionary has over
35,000 Tzotzil to English entries, making it one of the largest dictionaries
of an indigenous language of the Americas. However, it is a bilingual dic-
tionary in Tzotzil and English, limiting its direct use to the handful of peo-
ple who speak those two languages.
3
It is also a defiantly dialect-bound
(and even gender-bound) dictionary, documenting the way middle-aged
men spoke during the 1960s and 1970s in just the single municipality of
Zinacantán, arguably a minority variant of what has since become a domi-
nant Indian language in highland Chiapas with a much larger number of
speakers from other dialects. Thus, the choice of language variety in the dic-
tionary reflects accidents of the background research rather than principled
lexicographic or sociolinguistic design. Moreover, grouping entries by a
theoretical underlying root (a form which does not occur in speech, having
only psychological rather than surface “reality”), and stripping words of all
affixes – i.e. lemmatizing them – makes locating a word in this dictionary
something of an analytical challenge, again, a reflection of the intellectual
priorities of its producers, but with possibly inconvenient consequences for
many potential Tzotzil-speaking users.
A different variant of the ethnolinguistic wordlist, from Australia, illus-
trates another aspect of the field lexicographer’s dilemma. Many linguists
have documented Australian Aboriginal languages with very few remaining