Transitivity alternation, nontrivial minimal pairs and a non-minimal variation

MANGIALAVORI RASIA, MARÍA EUGENIA

Bucharest

Congreso; 31 Going Romance; 2017

University of Bucharest

Transitivity alternation, nontrivial minimal pairs and a non-minimal variationThe causative-inchoative alternation?a key topic in the study of the lexicon-syntax interface?has been subject to extensive research. In Romance (Labelle 1992, Folli 2002 i.a.), a fact not commonly discussed is that the alternation in the syntactic frame can be argued to extend to a third variant, also related through the notion of cause. This little-discussed alternative features a unique argument, interpreted by default not as the undergoer, but as the entity with the relevant properties to instigate a change (cause/initiator) (1)c. The construction?called here Stative Causative (cf. Rothmayr 2009) Construction [SCC]?creates a problem for the structural characterization and current assumptions on change-of-state [COS] verbs in at least two respects. ❶ It raises the question as to whether the internal argument, a ?stable argument? (Hale&Keyser [HK] 1992:167) in the alternation if the analysis is limited to the variants normally considered ((1)a-b), is indeed a default constituent in the argument structure of these verbs, as generally assumed in both lexicalist and constructional accounts (HK 2002:112, Rappaport & Levin 2011:152). ❷ By suggesting that the causative component may be independently realized, SCCs challenge a basic rule of event composition shared by these two major lines of work, whereby: (i) the event structure of COS verbs combines two components (cause, process) in a hierarchical relation; (ii) the former, if present, causally implicates the latter (cf.(2)).(1) a. La comida chatarra engorda a los niños. ?Junk food fattens the kids? CAUSATIVE/TRANSITIVEb. Los niños engordan. ?The kids fatten [up]? INCHOATIVE/UNACCUSATIVEc. La comida chatarra engorda. [lit. Junk food fattens] ?Junk food is fattening? SCC/UNERGATIVE(2) a. ([[INITvP) [PROCVP[PROCvo, √]]] (Ramchand 2008)b. e1 → e2 (Hale & Keyser 1993:69) c.[xcause [become [ y ]]] (Rappaport & Levin 1998:108)I argue that SCCs can be explained by a null processless vo freely available for derivation. If correct, Romance languages like Spanish pose a nontrivial variation from languages like English by allowing this causative vo to combine directly with the root, thus producing an event and argument structure simpler than the one seen in (a), but at the same time, one which is semantically and syntactically different from the monoargumental structure in (b). Regardless of the specific implementation, using distinct v heads constraining argument selection and interpretation (VINIT (Ramchand 2008, McIntyre 2004); vCAUS (Folli & Harley 2007)), I work under the widely-accepted premise that syntactic projection of argument structure correlates non-trivially with event structure (Ramchand 2013 for summary). Arguably, SCCs provide new empirical validation for this notion, preserving a crucial event/argument structure relation. Three related facts are central to the analysis ((A)(B)(C)):(A)Monadicity correlates with unergativity. Null/implicit objects [NO] are a potential explanation for SCCs, especially under the view that COS verbs are bona fide unaccusatives/transitives. There are, nonetheless, contrasts. While NOs generally allow secondary (resultative/depictive) predication and AP modification (3), SCCs do not (4). Unlike NOs (Rizzi 1986), SCCs do not bind reflexives (5) nor allow NO quantification (6). Assuming that bare quantifiers behave as NOs (Cattaneo 2008), (7) shows that perfective inflection is crucially dependent on NOs; otherwise, the construction fails due to the incompatibility of NO-less construction (SCC) with perfect tense (*Questo ha infuriato.). Inasmuch as a focal post-V quantificational expression is missing (8), ne-cliticization, a common test in Italian for NOs and unaccusatives (Russi 2008:113, Borer 2003:37) does not apply. Note that unpassivizable verbs (e.g. enfurecer/infuriare) and verbs not allowing NOs (e.g. llenar) freely allow SCCs (e.g. Esta sopa llena ?This soup is satiating?[(lit.)This soup fills]). Further, resultatives like (9), generally allowed by unaccusative/transitive verbs (Rappaport & Levin 1995), are not allowed by SCCs.Crucially,(9)a is only possible on an undergoer (inchoative) reading of the DP, as in (9)b. Data also discourages an account based on generic (null) internal arguments (on Romance: Dobrovie-Sorin 1994, Bhatt&Pancheva 2006). Again, generic internal arguments [GA] shift the interpretation to an eventive (COS) type, yielding telicity independent of object quantification, as in (10), and licensing endpoint modifiers not admitted by SCCs. This agrees with (7): apparently, an internal argument cannot be invoked without yielding the consequent eventivity and (variable)telicity.(B)Unergativity correlates with stativity. Systematically, non-trivial minimal pairs are created by the internal argument (even NO/GA). Telicity and progression are only possible under this condition (11). Event modifiers, progressive and perfect tenses are only accepted on undergoer interpretation of the DP (12) (conceptually odd). If framing (for-x-time) occurs at all in SCCs, it is interpreted as a temporal bound to an individual-level property, not to an event evolving over time, just likegradability: note the scalar (stative) interpretation in (13). Individual-level predication fits well with the generic reading of the subject (cf. (11)) (not required by causative/inch.), predication over the lifetime of the subject, and incompatibility with perception reports (*He visto al chocolate engordar. ?I saw the chocolate fatten?) (Diesing 1992 i.a.). Further, insofar as potentially ambiguous structures get SCC readings (13), SCCs challenge not only (2)), but also the notion that it is default for these verbs to select internal arguments (e.g. Default Linking Rule, Levin & Rappaport 2010).(3) a. El chef cocina {abundante/rico/salado}. ?The chef cooks abundant/tasty/salty [meals]. RESULTb. El chef compra {orgánico/natural}. ?The chef buys organic/natural [items]. DEPICT(4) El horno calienta (*abundante/*rico/*natural). ?The oven heats abundant/tasty/natural?(5) La injusticia {entristece/enoja} (*consigo mismo). ?Injustice saddens/maddens with oneself?(6) El sol calienta *{todos/algunos}. cf. {Ví/Compré/Calenté} {todos/algunos}?The sun heats up all/some?. ?I saw/bought/heated all/some [of them]?(7) Questo ha {infuriato/impoverito/indignato}*(molti) ?This infuriates/impoverishes/outrages many?(8) La radiazione infrarossa ne riscalda/brucia *(la metà) ?Infrared radiation burns (half) [of them]?(9) a. El grafito calienta *(hasta quedar incandescente/fundido). SCC?Graphite heats [causes heat] (until [becoming] incandescent/fluid)?b. El grafito (se) calienta (hasta quedar incandescente/fundido). INCHOATIVE?Graphite heats (itself) up until incandescent/fluid?(10) El sol quemó (en un minuto/completamente) *(algo/alguno).?The sun burned *(some) in a minute/completely?(11) Estos payasos *(te) asustan (abruptamente/gradualmente/por un tiempo).`These clowns {*are scary [SCC]/scare you [INCH]} (suddenly/gradually/for some time)?(12) #El chocolate casi engorda/está engordando/engordó?Chocolate almost gets fat/is fattening up/fattened?(13) SPAEsta estufa calienta mucho/ITQuesta stufa scalda troppo.?This radiator causes too much heat?Namely, absence of int.-argument-introducing VP in SCCs would mean that interpretation of the DP as initiator/cause cannot be purely structural (defined by VP embedding in (2)b) (HK 1992:150).(C) Stativity: SCCs differ from stative predicates analyzed as the result of a causative vo in previous accounts (e.g. English Katherine fears nightmares) in two ways. ❶ In SCCs the subject is interpreted by default as a cause rather than as holder of a result state (Ramchand 2008:64). ❷Unergativity (which follows naturally under the assumption that causative vos introduce ext. arguments, cf. Harley 2014 for discussion) is not predicted by standard analyses and contrasts with the (dyadic) argument realization seen in English verbs (DP1{fears/annoys}DP2). In principle, SCCs show a configuration with the desired characteristics for CAUS/INITvo (unergative, DP initiator, stative) (e.g. vP[(INIT)vo √]]).Recap: SCC suggest a key event/argument structure relation (non-eventivity>unergativity) pointing to an internal-argument-licensing process component not included in the verbal configuration by default (cf. (2)). This is in contrast to an explanation building on implicit arguments and/or unrealized projections, not favored by empirical data, and is consistent with similar observations on atransitivity (cf. McIntyre 2004). By contrast, SCCs offer evidence suggesting that in Romance verbs commonly described as a basic transitive/unaccusatives may have unergative uses. Advantages: ❶ Stativity in SCCs is consonant with the stativity of the causative vo (Ramchand 2008 i.a.) and agrees with general principles relating unergativity and atelicity (Dowty 1979, Borer 2003:35). ❷The constructional status of the process subevent, together with the corresponding verbal projection, fits well with the relative morphological complexity of unaccusative frames in Romance (Haspelmath 1993). Namely, unaccusative morphology (se) could be related to the realization of a specific, non-defective vo (e.g. BECOMEvo) (cf.Harley 2006). ❸ Given the semantic properties of INITvo, the proposal correctly predicts that object-experiencer verbs (i.e., verbs with subject-initiator distribution) are productive in SCCs, as opposed to subject-experiencer verbs (Ana {asusta/*teme}?Ana {is fierce (lit. frightens)/*fears?}) ❹ Free composing with a causative component constrained by external semantic factors agrees with observations raised by proponents of different views on the topic (Harley & Noyer 2000, Hale & Keyser 2005; Rappaport & Levin 2011).❺ Finally, the analysis avoids a model of partial projection of lexical structure in order to capture relevant alternations. Thus. If correct, SCCs would contribute empirical evidence on the optionality of constituents largely seen as default in the literature. In this sense, and although further work is still needed, crosslanguage variation (e.g. the alternation allowed by Romance) seems relevant to a deeper discussion on the correct phrase structure of these verbs.