Scope, Time, and Predicate Restriction in Blackfoot using MC-STAG

We use STAG to model the interaction between demonstratives and tense found in Blackfoot (Algonquian). In clauses with no tense or aspect marking, past tense can be encoded through a distal demonstrative on either the internal argument (transitives, unaccusatives) or the external argument (unergatives). A fourth class of predicate, semantically transitive but syntactically intransitive with a pseudo-incorporated object, does not allow either argument to mark tense. Using the scope mechanics described in Frank and Storoshenko (2012), we first model the unique scope properties of pseudo-incorporation, following on the Bliss (2013) claim that predicate saturation in these cases derives from the Chung and Ladusaw (2004) operation of predicate restriction, not function application. Modelling the predicate restriction operation within a STAG derivation is shown to correctly predict the scope facts in a way such that the tense facts are also easily captured. 1 Tense and Demonstratives In this section, we present the basic tense data, and discuss the challenges it presents for a GB/Minimalism-style analysis before moving on to STAG. Blackfoot is an Algonquian language of Southern Alberta and Montana with a largely polysynthetic morphology. While the language has a range of tense and aspect markers, clauses may appear untensed. As described in Lewis (2014), untensed predicates may receive an obligatory temporal interpretation from a demonstrative on one of their arguments:1 (1) amo ninaa ispii DEM.PROX man dance ‘This man is dancing/danced.’ (either tense) (2) oma ninaa ispii DEM.DIST man dance ‘That man danced.’ (past only, Lewis 2014 ex19b) As shown, while the proximal demonstrative does not fix the tense, a distal demonstrative fixes the interpretation in the past. The task of writing a function which could accomplish this is simply a matter of lambda gymnastics: (3) JF K = λPλs′.P (s′) ∧ DIST (s′) → s′ < s∗ F , a recursive function of type 〈〈s,t〉,〈s,t〉〉 on the verbal projection inspects its complement for situation variables in the scope of a DIST function (assuming demonstratives to also carry situation variables, see Elbourne (2008)), dictating that such situations be interpreted as prior to the utterance time s∗. The proposal is not that demonstratives themselves carry any temporal interpretation, rather that is built into the verbal projection (and therefore inherently part of the predicate) which interacts with demonstratives in its scope. The same pattern of past tense following from distal demonstratives has been shown for unaccusative predicates: (4) amo ninaa o’too DEM.PROX man arrive ‘This man arrives/arrived.’(either tense) In the interests of space, we will only give the minimal morphemic breakdown necessary to illustrate the facts salient to our discussion. Readers are encouraged to consult the sources cited in this paper to appreciate the complexity of the morphology.