BALEE & MOORE: SIMILARITY AND VARIATION IN PLANT NAMES 259

whether literal plant words in Ka'apor refer to highly useful plants in a higher
proportion than Ka'apor metaphorical plant words. We limited this question
to the non-domesticated plant category, where the factor of use can be isolated
from that of plant management. A high use value for any species would be 1.0
or above; a low use value would be 0.5 or below. Of 92 names in the Ka'apor
sample of non-domesticated plant names, 25 refer to plants with a high use
value and 67 refer to plants with a low use value. Of the names denoting plants
with a high use value, 13 are metaphorical, while 12 are literal. Of names for
plants with a low use value, 37 are metaphorical while 30 are literal. The
relative proportions of literal to metaphorical plant names in the two
categories, high and low use value, do not significantly differ (X2 = .08, p >
.05, df = 1). This means that the usefulness of a plant is not a factor in why its
name is literal or metaphorical and, by inference, in why its name is retained or
not.
 Why are literal plant terms more stable? One hypothesis is that of
Alphonse de Candolle: they are shorter. While there may be some truth in
this, there is probably more to it, since the shorter metaphorical words in our
sample (one morpheme excluding any life-form or common plant part
morpheme) do not seem to have a higher similarity rate than the longer words
(two or three morphemes, excluding any life-form or common plant part
morpheme). Another possibility is that the literal terms endure because of
their arbitrariness--the metaphorical terms involve a cultural interpretation of
the plant which is susceptible to change. In spite of the strong correlations
observed in the section on results, there still remains some degree of
unpredictability and possibly still unidentified factors at work in determining
naming patterns. For example, some undomesticated species (in particular,
several palms) show stable literal names. It is not yet clear what causes such
exceptions.
 The general patterns explained above, however, appear to be also present
in the Mayan Tzeltal and Tzotzil as well as in the Tupi-Guarani languages
studied. Perhaps this is the general case in Neolithic societies. It would be
instructive to see whether similarity and variation in other semantic fields, such
as birds, fish, or mammals, can be analyzed along the same general principles.


 NOTES


 1. The standardized symbols for consonants are as follows: p, t, k,
 kw labializedd velar stop), glottall stop), b, d, g, c (dental
 affricate), c (alveopalatal affricate), s, 9, z, d (voiced interdental
 fricative), h, m, n, fi, g velarr nasal), gw labializedd velar nasal) w,
 r, 1, and y. The vowel symbols are: i, e (mid or low-mid front