Verbs, semantics and (L)syntax: A proposal for Iberian Romance meaningful copulas

MANGIALAVORI RASIA, MARÍA EUGENIA

Nijmegen

Congreso; 29 Going Romance; 2015

Radboud University

Verbs, semantics and (L)syntax: A proposal for Iberian Romance meaningful copulas.Iberian Romance copulas expose a semantic contrast and selectional restrictions by now familiar, if not apparent. However, they also comprise a range of occurrences and effects that has proven extremely difficult to pin down. Here we tackle the task of trying to deal with ser/estar in two different but complementary ways. On the one hand, we aim at capturing the widest range of occurrences possible, also keeping a crisp differentiation between lexical and constructional aspect. On the other hand, we propose an account of their structural characteristics (semantically specified SC articulation) and further syntactically-accountable effects (e.g., categorial restrictions on the predicate, aspectual prevalence of the copula) via P-incorporation. The original motivation for P-incorporation theory lays in the attempt to provide a locus of semantic encoding (a particular non-eventive (stative) relation whereby ser/estar diverge); but in a way that makes it possible to assimilate the structure of semantically contentful copulas and (the encoding of) their distinct semantic properties to the syntax of small clauses in general, and at the same time preserving their structural syntactic symmetry and allowing for a (much desired) syn-sem homomorphism. As for the ontology of the incorporated element, the proposal of two different Central Coincidence [CC] Ps (PHAVE and PAT) for the encoding of ser and estar?s distinct semantic content ?as an alternative to a central/terminal P contrast proposed before? responds to the need to capture a significant portion of data remaining elusive to previous accounts; and, more importantly, keeping the contrast between two aspectually different predications within the stative domain.Starting point So far, it is clear that ser/estar depart from the standard definition of copula (i.e., a semantically dummy verb) in the sense that they bear semantic content relevant enough to draw mutually exclusive selectional patterns on APs with specific aspectual properties (1) and nontrivial differences between where they overlap (2), even in minimal pairs (i.e., where the only variable is the copula) and within a same stative type (3) ?thus bolstering the SLP/ILP classification long held. Further, this semantic specification overrides the one contributed by the head of the SC predicate (which should be the one prevailing over a verb classically defined as semantically irrelevant) (4) and allows both their partaking and mutual exclusion in non-stative environments (5)-(6). Finally, the distribution is category-sensitive (7). (1)a.{*Ser/Estar} contento, descalzo, harto, lleno, vivo, muerto, disponible, pendiente BESER BEESTAR happy, barefoot, sick, full, dead, available, pendingb.{Ser/*Estar} capaz, mortal, idóneo, válido, posible, probable, cuadrado, entrañable BESER BEESTAR able, mortal, apt, valid, possible, probable, square, endearing(2){Ser / Estar} apacible, feliz, cariñoso, fresco, gris. BESER/BEESTAR quiet, happy, affectionate, fresh, grey (3)a. Esta cartera {es/está} muy bonita SLP This wallet is very nice(estar=is nice [in my view])b. El editor ha {sido / estado} muy grosero con nosotros.ILP The editor has been been [to me] very rude to us(estar=was rude [in my view])(4)a. ser {despierto/espabilado} = to be smartser+PERF/SLPAP= ILPb. estar {sensible/deseante} = to be touchy, eager estar+IMPERF/ILPAP=SLP(5)a. La habitación terminará de {estar/*ser} limpia en 20 minutos (telicity) the room will end up clean in 20 minutes b. El editor estaba {*estando/siendo} grosero cuando llegamos. (progressive/durativity) The editor was being rude when we arrived(6)Los pasajes {*?son/están} baratos {lunes y jueves/a cada rato/una y otra vez/nuevamente} The tickets are cheap {on Mondays and on Thursdays/every once in a while/over and over/again}(7)Ser {PP/AP/NP/*AdvP}estar {PP/AP/AdvP/*NP} Against this background, we will defend the hypothesis that semantically contentful copulas like Iberian Romance ser/estar result from the combination of two elements: the categorial signature (in Hale&Keyser?s words) or (semantically dummy) lexical head (V0) and a birrelational non-eventive component (P*) instantiating two (diverse) CC relations (PAT/PHAVE, respectively). Hence, it is P*, governed by and incorporated into V0, that (i) heads the SC, establishing a distinct (aspectual) relation between the subject and the predicate and the consequent distribution; (ii) warrantees the correct semantic interpretation of the state rendered by the predicate; (iii) measures-out the event (cf. Harley 2005), thus explaining the relative insensitivity of V to the aspectual specifications of the AP and (iv) handles selectional restrictions not captured by aspect (categorial type), thus capturing (1)-(7) at once.The approach. P-incorporation, in conjunction with the CC/TC distinction has been attempted before on ser/estar (Gallego & Uriagereka 2009, Brucart 2012) and, in fact, the presence of a P-like (spatial) component (understood either semantically or syntactically) is particularly suitable for handling the aspectual/semantic properties of estar, from the inchoative/resultative flavor delivered by certain occurrences (terminal coincidence [TC] p à la Hale & Keyser 1993:72, cf. Gallego & Uriagereka 2009; INCHpfeature in Zagona 2011) as well as other instances falling out from this program. Yet, adjustments are still needed. For instance, a feature-checking approach is rapidly bolstered by the fact that not all (nor most) potential estar complements present either the expected (perfective) morphology or semantic features (e.g. inchoativity) (e.g. disponible and capaz in (1)); besides, complements not featuring these morphological/semantic properties still reflect the semantic partition at stake. Moreover, a [-CC] feature (as proposed by Brucart, 2012) is technically untenable (cf. Hale & Keyser 2005). As for aspectual semantics, a TC/CC contrast does not reflect the distribution nor correctly captures either the stative nature of copulas nor the atelicity of most estar occurrences. If P* is understood structurally, there is the problem raised by the PP/AP complement: as it stands, the sister-to-V position should be (pre)emptied in order to sit the copular predicate.Syntactic alternative: P-incorporation and P-cognation. This has two main advantages. First, it allows the differentiation of two local relations that permit the correct interpretation of the copular complement (especially in minimal pairs) and handle selectional restrictions (syntactic and semantically). Second, it places ser/estar a part of a more general phenomena affecting Romance languages (spatial PP cognation; cf. Mateu & Rigau 2009:135). Another advantage over a purely semantic or feature-based conception of P is that semantics is not acting freely; rather, it determined by syntax. Since syntactic positions are directly relevant in the measuring-out phenomena attested, it becomes possible to explain the triviality of the aspectual features of the AP (4) without appealing to extragrammatical devices (coersion). Semantics: the Ontology of P*. Empirical evidence stresses the need to abandon the previously held opposition between CC/TC (Gallego & Uriagereka 2009, Brucart 2012) or ascription of a terminal feature to estar (e.g. Zagona 2011) in favor of a distinction within the CC realm (and a structural conception of P rather than a featural one). This not only makes it possible to account for both stative and non-stative occurrences ?the latter featuring the CC expressed by the copula embedded within a complex (TC) relation? but also to properly predict the stative nature of both ser/estar, in the face of the long-held correlation between telicity and TC. As an alternative, we appeal to a pair of CC Ps (PAT/PHAVE) which not only has already been argued to lie in the heart of other systematic verbal alternations; but also match the semantic entailments of ser/estar predication pretty conveniently. In particular, a spatially bounded P* (PAT) matches the specific aspectual tests indicating that it is boundedness which (eventually) allows embedding under TC (thus capturing telicity/perfectivity as a result of constructional aspect) and consistently describes the range of estar?s occurrences, its contrastive flavor in non-telic and non-bounded predicates ((3)b)-(5)b) and further semantic implications (e.g. resultative flavor, repetitive and restitutive readings (6), etc.).Selected references: Brucart, J M. 2012. Copular alternation in Spanish and Catalan attributive sentences. Revista de Estudos Linguísticos da Univerdade do Porto. Vol. 7; 9-43 ? Gallego, Á.& Uriagereka, J. 2009. Estar=Ser+P. Talk held at XIX CGG. Vitoria-Gasteiz ? Hale, K & Keyser, S J. 2002. Prolegomenon to a Theory of Argument Structure, Cambridge: MIT Press ? Harley, H. 2005. How do verbs get their names? In Nomi Erteschik-Shir and Tova Rapoport (ed). The syntax of aspect. p42?65. Oxford:OUP.