"Inchoativity, eventiveness and stativity: a challenge posed by participial constructions in Spanish

MANGIALAVORI RASIA, MARÍA EUGENIA; AUSENSI, JOSEP

Barcelona

Workshop; 15th Workshop on Syntax, Semantics and Phonology (WoSSP 2018); 2018

UAB-ICREA

Recent work on lexical aspect shows that what has been long considered a homogeneous event class (states) actually groups sufficiently distinct types. Namely, D(avidsonian) vs. K(imian)-states, which contrast in the presence (or absence) of eventive properties (Maienborn 2007 i.a.), as shown by their distinct behavior. Within D-States, a novel class identified by finer-grained analyses are inchoative statives [INS] (e.g. Reflexive Psychological Verbs like preocuparse ?worry? in Spanish (Marín 2015, Marín & McNally 2011), typically characterized by including a left boundary (Piñón 1997) marking the temporal onset of the state (see also Bar-el 2005 (Salish), Kiyota 2008 (Japanese); Choi 2010, Chung 2005, Lee 2006 (Korean)). Here, we draw on Old Spanish [OS] (12th-17th corpus data collected from http://parles.upf.edu/llocs/cqp/oldes/ and CORDE) to argue that states with a left boundary and no eventiveness (left-bounded K-states) are also possible. In particular, we observe that combinations of estar ?be.LOC?+past participle [PPT] of change of state/location verbs in OS?unproductive/ unavailable in Modern Spanish [MS]?yield pure (eventless) states. If correct, this would indicate that INS are not a D-state subclass, and that two distinct types of INS need to be thus differentiated, following the same relevant criteria (presence (or not) of eventive properties). Crucially, comparable constructions are missing in other Romance (nonproductive locative copula) and IberoRomance (estar-productive) languages (cf. *Està arribat ?isLOC arrived?(Cat), also in Old Cat., Port. etc.)Data: Degree/time/manner modifiers (Maienborn 2007, Rothmayr 2009) show that OS constructions also include reference to an initial interval at which a proposition φ (state/location) is true. Like MS INS (Marín & McNally 2011, Marín 2015) progressive forms (1),(6) only include in its denotation the state?s initial boundary (vs. ongoing event), also brought out by modifiers like desque (8). Importantly, however, OS INS do not involve the change-of-state entailment (transition from ¬φ to φ) seen in both MS estar (attributive and locative) constructions and INS verbs (e.g. se está enfadando ?[he] is getting angry?). Similarly, in both OS and MS INS, modification related to potential duration (e.g. for-x-time adverbials) is only possible on a stative reading (temporal frame within which the state holds); yet, if combined with in-x-time, MS INS and estar constructions yield the onset of a state coinciding with the endpoint of an entailed ?preparatory phrase? centrally involved in the establishment of a state characterized by a resultative-like flavor (e.g. Estuvo lista/en Francia en una hora ?She was ready in an hour? Brucart 2012). In the distinct type of INS productive in OS however, onsets are represented as initial points in the denotation of an eventless stative predication without including such (result/ change-of-state) entailments. Consequently, punctual adverbs (2) only specify a moment in which the state/situation φ holds (no endpoint reading). If correct, this suggests that a BECOME component marking the transition from ¬φ to φ, arguably found in estar+PPT constructions, would only apply to MS, where the process (transition) from ¬φ to φ is specified by the verb providing the PPT combined with the locative copula (as with INS verbs, e.g. estar enfadado ?be upset?). Conversely, in OS the contribution of the PPT is almost trivial; apparently, its only role is to refine the locative meaning of the predicate of spatial situation, with degree modification behaving accordingly, as in (3)-(4) (noneventive sense expected in an eventless state). We thus assume that in OS the left boundary is not necessarily related to a shift from ¬φ to φ, but merely to a situation that holds starting from a given moment in time (note the gloss ?once? in (1)). Quantifiers (especially PPTs from dynamic telic verbs) hence give K-state patterns, with time-span reading allowed only if understood as a progressless time frame for φ (i.e., a stage-level [SL] predicate, which is crucially considered a sort of K-state, Rothmayr 2009). In fact, the meaning coincides with the definition of K-states as abstract objects for the exemplification of a location z during a time t (Maienborn 2005: 47). Note that adverbs incompatible with punctual verbs like llegar ?arrive? such as while and much are readily accommodated by OS constructions under the non-eventive sense just described: hence, in (4) the quantifier expresses that the spatial situation extended for a considerable time span (not an event evolving over time), while (5) reinforces the temporal finiteness expected from a temporally-bounded stative SL situation.As expected from eventless eventualities, progressives/time span adverbials lack habitual readings (1); yet, unlike MS (D-State) INS (Marín&McNally 2011), OS INSs do not license repetitive readings either (e.g. aunque estuviese caída muchos días ?Even if it was there for several days? (16th c., CORDE) (Lit. fallen several days)); and allow spatial and comitative modification incompatible with both MS estar+PPT/INS and with the base verb itself even in OS (cf. OS/MS llegar a/*en/*?dentro vs. OS estar llegado a/en/dentro (Lit. be arrived at/in/within)), but natural in locative (nondynamic) constructions (estar en/dentro ?be in/inside?). Unlike ?standard? MS INS (Marín 2015) also, OS INSs can be located in space and reported as perceived (e.g. Que vieron los demás estar caida ?[the one] that the others saw [fallen] there? (16thc. CORDE)), two key characteristics of SL states. Just as in MS INS (SL states), temporal quantification with siempre is possible in OS (La anima en que siempre esta llegada ?The soul at which it always is? (15thc. CORDE) (Lit. at which it is always arrived)); yet lacking repetitive reading. Modifiers like mientras ?while? (if allowed) in MS INSs would be interpreted as the interval during which the transition leading to the designated state extends (standard MS INS: mientras se enfada ?while he gets angry?). With OS INSs, the combination only denotes that the object can be found at situation S during t time (i.e. LOCATEDObject and REFERENCEObject(location) coinciding at S for interval t, as in (5)). The otherwise unexpected distribution of spatial PPs ((6) but also nonspatial verbs like nacer ?born?: dentro está nacido este dicho árbol ?Inside there is this tree´ (Lit. inside is born this said tree (15th c., CORDE)) is consistent with the notion that the PPT does not introduce an event(uality) (reaching a place/coming into existence), but rather contributes to a spatial situation (notably, with llegar used a measure verb to refer to the amount of time that φ (location) held of X). Unselected predicative complements add a contrast with MS related to SL/IL-hood (Marín 2015) (7) in line with the nonresultative, motionless predicate (cf. MS caer enfermo ?get sick?; caer herido ?become hurt?).(1)Le detenían hasta que le creciesse el cabello, y estando crecido le embiavan.DAT detained until that DAT grow.SBJ the hairand be.ESTAR.GER grow. .PPT.3S DAT send.IMP.3P?[he] would be arrested until the hair grew. Once the hair was long, [he] was sent? (16th c., CORDE)(2)Et pero que estauan llegados estos poderesel dia que el çessar y llego.and but that beESTAR.IMP.3Parrive.PPT.3Pthese powers the day that the Caesar there arrived?But these powers were already there the day that the Caesar arrived (there)? (15th c., G5R)(3)Estoconocen mejor los que más llegados están, porqueparticipan mas de la lumbre.this know.3P better the that more arrive.PPT.3PbeESTAR.3Pbecauseparticipate.3Pmore of the light?The ones better located are the ones who see it clearer, because they take part in the light-up area? (16th c., Seis tratados muy devotos y útiles para cualquier fiel cristiano)(4)Tres fijosdalgo a otra noble table mucho están llegados (lit. much are arrived)?Three hidalgos located at another noble table for much time?. (14th c., LBG)(5)Fizo lo soterrar ; mientre llegados estauan.(lit. while arrived were). ?He/She had him buried while they were there? (13th c.,LBA)(6)Este dia estando llegados (en la sala del Ayuntamiento/dentro de la iglesia) con los señores This day being arrive.PPT in the room of-the city hall inside of the church with the lords ?That day, (while) being there (in the City Hall /inside the church) together with the councilors??(15th c., CORDE)(7)a. Está caído borracho.b. Otros estavan caídos muy mal llagados (16th c., CORDE)?He is there drunk?(lit. fallen drunk) ?Others were [fallen] there with severe ulcers? (lit. fallen bad ulcered) In sum: we argue that the locative copula estar ?be.LOC? is productively combined in OS with PPTs of change of state/location verbs to produce INSs that describe an eventless stative situation including in its denotation the starting point of a homogeneous non-dynamic stage-level locative predication. OS estar+PPT crucially differ from comparable constructions in MS (if allowed) by lacking a ?preparatory phase? and change/result entailment, as shown by the contrasting behavior of relevant modifiers. The puzzling distribution of spatiotemporal modification (e.g. locative PPs); and most notably, left-bound markers (9) (cf. MS INS, Marín 2015:30) are key in this respect, along with the sort of degree modification seen (cf. (8) vs. MS (*desde que) está (#algo) crecido). Importantly, even if similar patterns may be possible in MS estar constructions, these cannot be systematically formed from change-of-state/location unaccusatives generally lacking transitive variant (MS estar *llegado/*ido /?*crecido/*descendido/*partido/*venido/*vuelto/*restado/*quedado/*habido/*tenido/*llevado).(8)Y desque esten algo crecidos traspongan los antes queechenflor. (lit. since they are somewhat grown) ?Move them from the moment they are tall/mature, before they produce flowers? (16th c., GAH) (9)quando estouieren algo crescidas ylas escauen y les horaden las rayzes con vnas puas.?When they are a little big, dig them and pierce them with spikes? (lit. when they would be grown) (16th c., CORDE)