CONICET | Buscador de Institutos y Recursos Humanos

This paper is the second part of a larger, ongoing research about syntactically relevant aspectual properties of verbs contributed by different (linguistic and non-linguistic) domains. In particular, I will focus on the conformation of two aspectual classes of change-of-state deadjectival verbs [DVs] (1). (1)a. [- max] entibiar, endulzar, ensuciar, enflaquecer, engordar, embellecer, empobrecer, engrosar, enturbiar, achicar, agrandar, aclarar, acortar, adelgazar, afinar, alargar, alivianar, amansar, atontar, atenuar, aflojar b. [+max] ensordecer, enmudecer, enloquecer, enrojecer, emblanquecer, ennegrecer, emborrachar, empalidecer DVs are particularly salient in Romance languages, considering that the derivation of a synthetic verb represents a regular and rather productive device for the expression of change of state, in opposition to other languages (e.g. English) which, although featuring DVs as part of their lexical inventory, show marked preference for an analytic form combining a Path verb (go, turn) and an AP (2). (2)Enloquecer, secar, blanquear, ennegrecer, empeorar, aquietar, aflojar, tensar, enfriar, suavizar, agriar Go {crazy/dry/blank/impatient/bad/quiet/black/slack/numb/stiff/cold/soft/sour} Abstracting away from the fact that this parametric difference may also be predicted by well-known Path conflation patterns in Romance (cf. Mangialavori 2014), data shows that there are aspectual differences consistently determining the grammatical architecture and behavior of these verbs, and that these features may actually be rooted outside grammar. Specifically, the case posited by DVs suggests that the relevant aspectual properties of the scale involved (±maximal degree) may not be originally encoded by closed-class (grammatical) nor by open-class (lexical) forms (in the spirit of Talmy 2000), but shaped by our knowledge of the world and (cultural) conventions (cf. Kearns 2005). Otherwise, how can we account for the fact that solidification is conceived as a closed-scale kind of change ([+max]) but warming is not, and, therefore, the former admits endpoint modifiers not allowed by the latter (3)? Hence, aspectual constrains and patterns would then have to be due to the relation between human conception of change and of properties with/without maximal degree [+max] and the structure of the scale determined by it. Secondly, I will argue that this information is reflected by the syntactic and semantic structure of the verb, considering that telicity involves specific semantic/syntactic projections which are actually only present in the first type (solidificar [solidify]) as well. (3)a. {[+max]Solidificar/[-max]*?Entibiar} completamente b. Totalmente {[+max]sólido/[-max]*?tibio} On this account, the presence of a bounded scale [+max] ?and the corresponding resultative layer? would be explained by extragrammatical (conceptual) semantic material being mapped onto the process. The aspectual structure of the event would then be established via the scalar structure of that property, whose limits are shaped by real-world knowledge and conventions (the parameters according to which solid is associated with a bounded kind of change and warmth is not). Eventually, the process is defined by its progress through the scale contributed by the root ([±max]) and then mapped onto the Time Line of the event (±EndState), therefore making conceptual information crucial (and homomorphic) to the temporal structure of the process. What is more, the proposal of this difference being external to grammar matches the fact that this distinction crosscuts languages with different parameters regarding path encoding and verbal derivation, as English equivalents suggest. Additionally, in order to single out the grammatically relevant properties associated with verbs conforming each class, and the aspectual patterns behind this division, I will aim to show that (i) verbs delivering a gradable change do not necessarily involve the attainment of the property denoted by the A, but a mere displacement along a vector ([-max] scale); and that (ii) this correlation deals with the felicity facts concerning the combination with adjuncts associated with culminativity, gradability, etc. Accordingly, negation has two possible entailments in open-scale DVs ((5)b), since the property may be attained or not, whereas one of these two reading (the half-way, non-resultative interpretation) is not possible in DVs derived from [?max] roots, as ((5)a) shows. In fact, further evidence analyzed as part of the ongoing research may show that event type might shift by virtue of mere conceptual/conventional aspects of meaning (e.g. enfriar may accommodate both telic and atelic readings depending on cognitive standards). (4)a. La sopa fría ?→ La sopa {es/está} fría [cf. Demonte 1999] b. La sopa se enfrió ?/→ La sopa {es/está} fría [e.g., la sopa está tibia] (5)a. La sopa no se enfrió ?→ the soup didn?t get colder ?/→ the soup began to cool down, but it did not get cold b. La sopa no se congeló ?→ the soup didn?t get frozen ?→ the soup began to freeze but didn?t get frozen In short, facts indicate that the conceptual content contributed by the lexical root is not a property that the object is asserted to possess by virtue of participating in the event (although incrementality may be involved), but actually the parameter specifying the kind of transition [±max] undergone. In consequence, I will advocate in favor of the idea that event path associated with the (scale introduced by) the root denotes some variable trait; but this property is not necessarily inherent to the undergoer of the change, and it does not have to be monotonic with respect to its part?whole structure, as defined by our human perception of the notion of change (in the spirit of Ramchand 2007). At this point, a crucial question is whether the syn/sem structure (and truth conditions) require that the path covered grow homomorphically with the event, with the location identified by the root overlapped at the end of the event. Data in (5) indicates that, for the cases marked [-max], the answer is no. On this account, empirical data conforms to the claim that while the attainment of a result state can give rise to telicity, mere gradual change on the part of an argument is a distinct aspectual property and one which is logically separable from the attainment of the property denoted by the A derived from the same root as the DV. Hence, change is also compatible with a lack of temporal bound (cf. Hay, Kennedy & Levin 1999) and, therefore, the distinction we need to make is that of: (i) the type (±max) of scale mapped onto the process (value set outside language becoming syntactically and semantically decisive); and about (ii) the presence of a path covered homomorphically with the event, with a location identified by the root overlapped at the end of the event ([+max]) on the one hand, and unbounded change of state ([-max]) on the other. From our perspective, the key here is how these properties ?and the contribution made by extragrammatical domains? might be deployed to create both syntactic and semantic generalizations allowing us to escape an undesired proliferation of lexical specifications and/or syntactic projections without evident empirical instantiation.

enviar mensaje