6 Adjectives in Papantla Totonac 155'You don't understand! It is not possible that whoever is a rabbit will be an
elephant!' (COy io)^6(15) shows CCs with overt copula, (153) as NPs, (i5b) as adjectives. Notice that if the
plural occurs explicitly, the different pluralization technique distinguishes between
the two sorts of CCs, but notice also that pluralization is optional.
(15) (a) kinata:kamd:n,
kin-na-ta:-kam-d:n
lPOSS-KIN:pl-COM-SOn-pl
aqsquyund'(ni:n) ixtawani:t
aq-squyu-nan-f-(ni:n) ix-ta-wan-ni:t
head-smoke-iNDEROBj-NR-plpST-3pl-cop-PF
'My brothers, they were traditional doctors.'
(b) kinata:kamd:n, (laq)sqalala ixtawani:t
kin-na-ta:-kam-d:n (laq)-sqalala ix-ta-wan-ni:t
iPOSS-KiN:pl-coM-son-pl DiST-intelligent PST-3pl-cop-PF
'My brothers, they were intelligent.'
Another diagnostic to distinguish copula clauses with nominal and adjectival com-
plements is the co-ocurence of adjectives,but not nouns, with certain adverbials (see
Levy 1992: 284, where I showed different patterning of nominal and adjectival CCs
wiihyaj 'not anymore'). The best diagnostic, though,because it is cross-linguistically
more prevalent (see Dixon, Chapter i, this volume), is furnished by Beck (2000:233-
4). Adjectives, but not nouns, co-occur with the intensifier 'very'. So (i6a) is gram-
matical, while (i6b) is not. But as Beck clearly points out, the intensifier co-occurs
also with verbs, crucially with intransitives (i6c). So in and of itself, it is a good diag-
nostic for separating nouns from adjectives, but not adjectives from verbs.
(16) (a) Snu:n caka:t namd: ix-kdm
very playful DEM 3POSS-son
"That son of his is very playful.'
(b) snu:n ma:qalhtawaqi:na Juan
very teacher Juan
'Juan is very teacher'
(^6) Beck (2000: 226) finds that for UNT, a bare noun cannot be a complement of a headless relative
clause, presumably because they are not interpretable as copular sentences. It would be interesting to
see whether a copular sentence with overt copula and an NP CC is possible in that frame, as the com-
plement of a headless relative clause, since for UNT this environment is the only one in which copula
sentences with nominal complement and copula sentences with adjectival complement do not pat-
tern alike. I suspect that the difference in the copula sentences with nominal CC in the two languages
may be due to the different extent of the use of xa- in the two varieties. In PT it is very frequent and
it has developed quite a range of syntactic, semantic, and pragmatic functions (Levy 2oo2a). So struc-
tures like (14) do occur in text. However, according to Beck (p.c.), in UNT xa- is far less frequent in
text than in PT, and I suspect that there might be also differences in the number and kinds of con-
structions that require it. If these differences in the copula and specifier are linked, this supports the
idea that the use of the specifier is tied to the grammaticization of the adjective category.