6 Adjectives in Papantla Totonac 169(43) kp'aqlhni:t mi-xa:lu
k-p'aqlh-ni:tan min-xa:lu
i-break(Vtrans)-PF iposs-cup
'I have broken your cup.'
tapa'qlhni:t mixadu
ta-pa'qlh-ni:tan min-xa:lu
iNGR-break-PF iposs-cup
'Your cup broke.'
(44) palha' 'hard' palha'-n 'S hardens'
sqalala 'intelligent, savvy' sqalala-n 'S becomes savvy'
sa:sti' 'new' sa:sti'-n 'S regenerates, gets
renovated'
maq-sa:sti '-n
(maq- 'body') 'S has her first menses'
sta'ranqa' 'white' sta'ranqa'-n 'S whitens'
The suffix -n forms intransitive verbs whose overt Subjects are underlying under-
goers; these verbs have an inchoative reading: the only participant, semantical-
ly an undergoer, enters into a process that culminates with it having the property
denoted by the base, e.g. palha'ni:t qi:la (harden-PF porridge), 'the porridge hard-
ened, became hard'. Crucially, no Agent is either present or implied. In contrast,
fa-, with stative bases, forms intransitive verbs which can be considered seman-
tically middles, because their subject is both semantically an Agent and a Patient,
e.g. tanu:-lh Juan (enter-cpL John), 'John entered', as in (42). And in (43), fa- back-
grounds Agent. It is the resource to derive an intransitive verb from a transitive
one, with the peculiarity that the only syntactial participant left, the Subject, is se-
mantically an undergoer. The verb, nevertheless, entails that the change of state did
not come about spontaneously, did not happen by some sort of inner causation,
but that there was an external cause that brought about the change of state. In cer-
tain terminologies (Haspelmath 1993, Dixon and Aikhenvald 2000, inter alia), this
is called the anticausative alternation: the transitive verb is unmarked, its intransi-
tive counterpart has only one syntactic participant that is syntactically a Subject,
but semantically an undergoer, and the change in the alternation is marked on the
intransitive counterpart of the pair.^10 The prefix fa- is multifunctional and a full
discussion of its behaviour is beyond the scope of this chapter, but in all its func-
tions, it signals in the semantic structure one of the implicatures of Proto-Agent,
in Dowty's (1991) terminology (see Levy forthcoming, for details).
(^10) Dixon and Aikhenvald (2000: 8) point out that while the passive signals that the '(derived S)
came into certain state because of the involvement of an agent [...] in contrast, the anticausative im-
plies that it came into the state spontaneously.' This is certainly the most frequent case typological-
ly. PT makes a different distinction. It codes spontaneous or internal causation with -n, and external
force with ta-, which is involved whenever the event is construed as having a potential Agent, even if
the actual derivation backgrounds it. PT has no demoting passive.