Measuring-out, scale boundedness and some alternatives

MANGIALAVORI RASIA, MARÍA EUGENIA

Caen

Conferencia; Chronos 12. 12th International Conference on Actionality, Tense, Aspect, Modality/Evidentiality; 2016

CRISCO research centre - Normandie Université, UCBN

Measuring-out, scale boundedness and some alternatives.This paper focuses on four different but (hopefully) related facts about deadjectival verbs [DV] in Romance along with some natural extensions for English:(1) a. The relevance of property-denoting lexical roots [Prop√] (i.e. scale boundedness) and the consequent root-event homomorphism noted in DVs in both Romance and English;b. The alternatives following from the availability of different non-eventive relational heads (p0) providing a locus of Path-encoding for DVs in Romance languages like Catalan and Italian;c. The consequences of the semantic and syntactic specifications of the Path-encoding constituent (p0) for the eventive and argumental structure of the DV, according to (b).d. The apparent failure of phrasal COS paraphrases (motion verb+adjective) to render non-resultative/atelic COS in English in contrast to analytic forms (DVs).Research Problem. DVs are not a homogenous class. On closer look, divergent aspectual entailments obtain. This difference has been widely argued to ensue from the scalar specification (boundedness) of the Property-denoting root (Prop√) (cf. Kearns 2007), or from the interpretation of the DV with reference to a closed/open property scale (Hay, Kennedy&Levin 1999). In this paper we will identify difficulties with this account, and argue instead that the relevance of scale boundedness for telicity is determined by syntax based on two facts. On the one hand, Catalan and Italian DVs comprise an alternative derivation for Prop√s which presents consistent aspectual/eventive (atelic) properties regardless of the type of scale contributed by the root/adjective. A second question has to do with the principles involved in the circumstance that the analytic paraphrase regularly employed in English (motion verb+adjective) appears to be (unexpectedly) constrained to resultative/telic COS regardless of open/closed-nature of the property scale?and especially when English DVs find no trouble in delivering the two aspectual types according to the Prop√ involved (thus accommodating to the aspectual partition observed).Data & Analysis. We will analyze alternative COS built on Prop√s in order to cast light on the syntactic and/or semantic factors involved and the way they affect the aspectual contour of the (COS) event delivered. Facts enumerated in (1) will guide the discussion.((1)a) Compatibility with proportional/maximality/intensive modifiers shows two DV types: open scale Prop√ essentially correlates with atelicity/unboundedness (2), closed-scaled prop√ results in telic COS (3). Crucially enough, further tests (negation scope, subinterval property, etc.) show that ±gradability of Prop√ correlates with ±homogeneity and ±resultativity. In our case, ascription of the divergence to the Prop√ is encouraged by four specific facts: (i) verbs in (2)-(3) share the same derivational constituents; (ii) the selection of alternating affixes (other DVs use ?a/ø) is not relevant; (iii) neither is categorial type (actually, the split is also reflected by denominal verbs, cf. Hale & Keyser 2002, Harley 2005); (iv) argument structure realization patterns do not correlate with the division either: both (2) and (3) comprise verbs entering the causative alternation, as well as (only) transitive and (only) ergative verbs.(2) embellecer [embellish], ensanchar [widen], endurecer [harden], endulzar [sweeten]{#enteramente[entirely]/#parcialmente[partially]/#absolutamente[absolutely]/#completamente [completely]/apenas[barely]/mucho[a lot]/demasiado[too much]/un poco [a little]}(3) emblanquecer [turn white], enverdecer [turn green], palidecer [turn pale], enloquecer [go crazy], enmudecer [become mute], ennegrecer [turn black], enrojecer [turn red], enronquecer [grow hoarse], ensordecer [go deaf], enturbiar [become murky]{enteramente[entirely]/parcialmente[partially]/absolutamente[absolutely]/completamente[completely]/totalmente[totally]/#apenas[barely]/#mucho[a lot]/#demasiado[too much]/#un poco [a little]}If this is correct, and Prop√ is the element triggering the delivery of a telic/atelic predication, then the situation has an interesting parallel in the domain of motion verbs and the way telicity is established. Mutatis mutandis?if we accommodate the affix as a path-encoding head (p0); thus accounting for the transitional character of DVs as opposed to Adjs built on the same Prop√?, the position of the Prop√ in DVs (prior to conflation) on the l-syntactic approach of Hale and Keyser (2002) is equivalent to the position of (unincorporated) measuring-out arguments such as Goals in motion verbs (e.g. The cart rolled *(to the park) in an hour)?assuming also that motion is comparable to change. Since property-denoting elements have been long-studied by differing in properties relevant to measuring-out, such as inherent boundedness (the simplest assumption being that scales may or may not have maximal/minimal elements; cf. Kearns 2005 i.a.); we can expect Prop√s with different scalar specification (±bounded) to deliver verbs with different aspectual/eventive properties, and that such properties will be reliably determined in pretty much the same way that such properties in Goal arguments determine the aspect of motion verbs (cf.Harley 2005 on denominals). If the telicity/resultativity of a motion/change event hinges on the availability of an endpoint (+max degree) for the Path, we should expect DVs derived from [+max] Prop√s to be telic and resultative, whereas DVs derived from [-max] roots shall render atelic and nonresultative COS, as further tests (negation scope, subinterval property) confirm. Hence, event-root homomorphism can be claimed to accommodate the distribution in (2)-(3).((1)b) However, Catalan and Italian DVs suggest that event?root homomorphism may not be a general or defective phenomenon, but a consequence of a (single) type of P0 being used in Spanish/English. Catalan and Italian offer two productive affixes for the derivation of DVs (4): an em? prefix [Type1]?comparable to the Spanish affix in (2)-(3)? and a suffix (?eggiare/?ejar ) [Type2]. Crucially, the resulting verbs show different aspectual properties even if built on the same Prop√. If event-root homomorphism held invariably, then DVs derived from color-denoting roots, namely, should always be resultative/telic, as generally agreed according to the bounded nature of the scale imputed to color-denoting roots/Adjs (cf. (3)). Yet, pairings like (5) show that a bounded [+max] root delivers a telic/resultative predication only when it sits at the right of the affix [Type1], but not when it precedes it [Type2]: the ?eggiare/?ejar alternatives are not telic nor resultative. Unlike Type1(classically defined as ?becoming x?, where x amounts to the property denoted by the root), Type2 is defined by denoting a state, essentially described as close or similar to the property denoted by the root (cf. Oltra & Castroviejo 2013). Specific diagnostics converge: proportional/maximality modifiers are, as expected, not possible with Type2, but natural with Type1; in turn, intensive modifiers are at least odd with Type2 and natural with Type1 (5). Consistent effects arise in the interaction with perfectivity (e.g. past perfect is only possible with Type 1 and imperfect tenses are preferred with Type2), telicity/culminativity environments (6)-(7). The fact that definiteness of the internal DP is not relevant either in Type1 ((7)b) seems to concur with the hypothesis that the Prop√ is the one measuring-out the event.(4) Type1. ITA. arrossire, ingiallire, sbiancare, inverdireCAT. enrogir, engroguir, emblanquir, enverdirType2. ITA. rosseggiare, gialleggiare, biancheggiare, verdeggiareCAT. vermellejar, groguejar, blanquejar, verdejar(5) a. {*abbastanza/*troppo/completamente/totalmente/parzialmente} inverditoquite too completely totally partially turn-green-PASTb. {abbastanza/troppo/*completamente/*totalmente/*parzialmente} verdeggiatoquite too completely totally partially be-greenish-PAST(6) Dopo aver finito di {arrossire/*rosseggiare}after have-INF ended of redden seem reddish ?After having finished reddening](7) a. {La campagna/alcune campagne} {verdeggia(no)/*inverdiscono} tutto l?anno.The countryside/Some countryside are green(ish)/turn green all the-year?The/some countryside look(s) green all year round]b. Il sole sbiancò *(due) pareti in 5 minuti. Cf. Il sole biancheggiò due pareti in 5 minutithe sun whitened two walls in 5 minutes the sun be-white(ish) two walls in 5 minutes?The sun turned [the] (two) walls white in 5 minutes]((1)c) Interestingly enough, this divergence is not confined to aspectual semantics and correlates with syntactic structure in relevant ways?in fact, the divergence would be wrongly confined to the semantics of the affix (e.g., claiming a neutralizing effect for ?eggiare). Delivery of COS vs stative predicates out of Prop√s can be easily handled by the difference between T(erminal) C(oincidence) and C(entral) C(oincidence) relations encoded (we propose) by the affix (P0) according to the layout sketched in (12). The eventive and argument structure of the resulting DVs present a relevant correlation: the complexity of the TC fits both the transitional character (COS) and the availability of a projection embedding the (result) state (the simpler CC necessarily comprised in TC relations, cf. Hale & Keyser 2002 i.a.); plus, it hosts the DP in the correct position (the internal subject of the ergative verb). In turn, a CC P0 (eggiare) conflates with the phonologically null V0 giving a simpler eventive and argument structure, corresponding to a unergative non-eventive (stative) predicate. Auxiliary distribution (8), ne-clitization (9), Causative/Inchoative alternation (10) and the fact that Type1, unlike Type2, allows passivization more or less successfully (11) are consistent. If correct, then Prop√ does not measure-out the event in type2 (i.e., the correlation with motion verbs and goal arguments does not hold) basically sinceit is not the complement of a directional non-eventive predicate (PT). Plus, the unergativity of Type2 contributes to the aspectual divergence noted, for the DP would not be in a measuring-out (internal) position in these DVs either.((1)d) Eventually, the general approach could be pushed further to accommodate the (apparently unexpected) fact that English analytic paraphrases are consistently telic and resultative, even if taking [-max]Prop√s (rendering non-resultative/atelic DVs) unless a comparative (-er) form is added (13) (cf. also ?ish). Two facts converge in this circumstance: English Path-conflated verbs like go feature a V-conflated TC P (PT) which accounts for the telic/directional reading in contrast to the well-known ambiguity between a directional and a locative reading in Romance equivalents. Second, the Prop√ is not acting freely, but contained in a further projection (AdjP); this is crucial not only configurationally, but because adjectival categorization has been proved to determine the aspectual the type of the predicate (thus, whereas Adjs trigger IL predication; DVs render, at least, SL predicates). Yet, further work on this point is still in order.(8) a. {ha/*è} rosseggiato, verdeggiato, biancheggiato, gialleggiato, bruneggiatob. {*ha/è} arrossito, inverdito, sbiancato, ingiallito, imbrunito(9) a. Trattamenti che ne {sbiancano/arrossiscono/*biancheggiano/*rosseggiano} la pelle?Treatments that make the skin red/white]b. Non ne {arrossisce/*rosseggia} nessuno di coraggiosi?None of them blushes [reddens] out of courageousness](10) La paura lo {sbiancò/*biancheggiò}. Si è {sbiancato/*biancheggiato} di paura.?Fear turned him white [pale]] ?He turned white with fear](11) Lo zucchero di canna grezzo non è stato {sbiancato/*biancheggiato}?Brown cane sugar has not been whitened].(12) a. [VP [ V0 [PT en-/a-/Ø [PC [DP [PC Prop√]]]]]]] b.[VP [V0 [PC [PCProp√, PC ?eggiare/ejar]]]](13) a. The sky darkened, but is it not dark [though].b. The sky {went/turned/grew/got}dark*(er), but it isn]t dark [though].Results. Spanish and English DVs suggest that the partition between two aspectual/eventive types is predicted by event-root homomorphism ((1)a); yet, Catalan and Italian DVs indicate that this effect actually hinges on the situation of the Prop√ within the VP ((1)b). The TC/CC opposition correctly handles the semantics and syntax?specifically, the argument structure realization patterns and eventive structure?of DVs involving different Ps in the derivation ((1)c): whereas a CC (Post)position (eggiare/ejar) leads to stative, unergative verbs where the Prop√ does not get to measure the event; a directional (TC) Preposition (en-/a-/Ø) gives an ergative COS verb where the open/closed nature of the property scale is directly relevant to the eventive determination of the predicate (event-root homomorphism). In turn, English analytic constructions ((1)d) originally accommodate to the proposed analogy with motion verbs, but they also show that independent realization of the Prop√ as AdjP is not trivial in this respect.Significance. The alternative DV derivations contributed by Italian and Catalan suggest that the correlation between scale boundedness and telicity is not as generalizable as thought. This provides independent evidence in favor of previous works claiming the relevance of scale boundedness to be determined by syntax; and that the position of Prop√s is comparable the one occupied by unincorporated measuring-out Goal arguments in motion verbs, triggering similar (measure-out) effects. Plus, they also establish relevant connections between eventive structure and argument structure. The fact that both eventive and argument structure realization patterns can be predicted from the structural nature of the P0 based on the CC/TC opposition does away with the need to postulate different V0s or additional covert (and hard to prove) components in accounting for the divergences noted between DVs. More importantly, both DVs and analytic COS could ultimately be seen as part of a much more general phenomenon related to the way that the element in sister-to-V position determines telicity in other (comparable) verbs. In any event, (aspectual) semantics is not acting freely, but is (transparently) determined by syntax.Selected References:Hale, K & SJ Keyser. 2002. Prolegomenon to a theory of argument structure. MIT PressHarley, H. 2005. How do verbs get their names? In The syntax of aspect. OUPKearns, K. 2007. Telic senses of deadjectival verbs. Lingua 117, 27-66Oltra, I. & E. Castroviejo. 2013. Approaching Results in Catalan and Spanish Deadjectival Verbs. CJL 12: 131-154.