Title: The Manner/Result Complementarity Revisited: A Syntactic Approach. Affiliation: Centre de Lingüística Teòrica-Universitat Autònoma de Barcelona - PDF

Description
Title: The Manner/Result Complementarity Revisited: A Syntactic Approach Authors: Jaume Mateu, Víctor Acedo-Matellán Affiliation: Centre de Lingüística Teòrica-Universitat Autònoma de Barcelona 1 The Manner/Result

Please download to get full document.

View again

of 21
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information
Category:

Science & Technology

Publish on:

Views: 14 | Pages: 21

Extension: PDF | Download: 0

Share
Transcript
Title: The Manner/Result Complementarity Revisited: A Syntactic Approach Authors: Jaume Mateu, Víctor Acedo-Matellán Affiliation: Centre de Lingüística Teòrica-Universitat Autònoma de Barcelona 1 The Manner/Result Complementarity Revisited: A Syntactic Approach Structured abstract Purpose To show that the manner/result complementarity (Rappaport Hovav & Levin, 2010), that is, the fact that a verb cannot simultaneously lexicalize manner and result, need not be stipulated, but is derivable from general principles of syntactic computation and properties of the syntax-morphophonology interface. In particular, it derives from the formal (i.e., nonsemantic) fact that a single root cannot simultaneously undergo Conflation and Incorporation in the lexicalization of a verb (Haugen, 2009). Methodology/approach The approach to argument structure is syntactic: argument structure is built up through syntax (Hale & Keyser, 2002; Marantz, 1997; i.a.). A sharp distinction between syntactically nontransparent conceptual content and syntactically transparent semantic construal is assumed (Mateu, 2002; Ramchand, 2008; i.a.). Findings The manner/result complementarity derives from general syntactic principles and need not and, hence, must not be stipulated as a constraint on the structuring of events. These general principles are also responsible for other complementarities found in the lexicalization of verbs. If our proposal is correct, one should not pretend to explain the relevant constraint in mere event structure terms. 2 Originality/value of paper The syntactic approach to argument structure (Hale & Keyser, 1993f.; Marantz, 1997f.) assumed in this paper sheds light on recalcitrant cases like those ones involving manner/result complementarity. Keywords conflation, incorporation, argument structure, manner/result complementarity, syntax Categorisation Research paper 3 The Manner/Result Complementarity Rappaport Hovav & Levin (2010) make the interesting strong claim that no verb encodes or lexicalizes both manner and result (see also Levin & Rappaport Hovav, 1991, 2008). 1 The result is unspecified for MANNER verbs, i.e., those verbs that specify a manner of carrying out an action (see some examples in (1a)). Similarly, the manner in which something acquires a state is unspecified for RESULT verbs (see some examples in (1b)). (1) a. MANNER verbs (i.e., verbs that specify a manner of carrying out an action): e.g., wipe, scrub, walk, swim, etc. b. RESULT verbs (i.e., verbs that specify the result of an event): e.g., fill, clean, arrive, come, etc. Rappaport Hovav & Levin claim that the origins of the so-called manner/result complementarity in (2) can be found in the lexicalization constraint in (3): (2) Manner/result complementarity: Manner and result meaning components are in complementary distribution: a verb may lexicalize only ONE. Levin & Rappaport Hovav (2008, p. 1, ex. (6)) (3) The lexicalization constraint: A root can only be associated with one primitive predicate in an event schema, as either an argument or a modifier. Rappaport Hovav & Levin (2010, p. 25, ex. (12)) 1 According to Levin & Rappaport Hovav (2008, p. 1), lexicalized meaning refers to those meaning components entailed in all uses of a verb, regardless of context [ ]. In the unmarked case what is lexicalized in a verb is kept constant in all uses. (Our emphasis: XX.) 4 In their lexicalist model, these authors claim that the root can be associated as a modifier in the event structure pattern of manner verbs (see (4a)) or as an argument in the pattern of causative change of state predicates (see (4b)). Given the lexicalization constraint in (3), it is predicted that the root in a single verb cannot be associated to both modifier and argument positions (see (4c)): 2 (4) a. [x ACT ROOT ] b. [x CAUSE [ y BECOME ROOT ]] c. *[ [x ACT ROOT ] CAUSE [y BECOME ROOT ]] (* in a single verb) Our present syntactic proposal is that the constraint in (3) and its associated descriptive claim in (2) follow from how primitive elements of argument structure are composed in the syntax (Hale & Keyser, 2002; Harley, 2005; Marantz, 2005; Mateu, 2002; i.a.). We want to emphasize that the constraint in (3) must not be regarded as an inescapable stipulation (as in Rappaport Hovav & Levin s (2010) non-syntactic approach), but can be shown to be derived from the very nature of Conflation and Incorporation processes (Haugen, 2009). If our proposal is correct, one should not pretend to explain the relevant constraint in mere event structure terms (cf. (3)). A Syntactic Approach to the Manner/Result Complementarity To advance our main point, we will show how the descriptive observation in (2) can be accounted for in a syntactic model where notions like Manner and Result become grammatically relevant since they can be claimed to be relationally encoded in the relevant 2 Cf. Rappaport Hovav & Levin (2010, p. 26): [ ] assuming that manner roots modify the predicate ACT and result roots are arguments of BECOME, a root can modify ACT or be an argument of BECOME in a given event schema. A root cannot modify both these predicates at once without violating the lexicalization constraint. 5 syntactic argument structures: in particular, we claim that Manner can be read off the adjunction relation to v (cf. Den Dikken, 2008; Embick, 2004; Harley, 2005; Mateu, 2002f.; McIntyre, 2004; Zubizarreta & Oh, 2007), whereas Result can be read off a SCR(Small Clause Result)-like predicate (cf. Hoekstra, 1988; i.a.). For example, the same root BREAK is structurally interpreted as Manner in (5a) and as Result in (5b): (5) a. [ vp [ v BREAK v] [ SC [ DP he] [into the room]]] (He broke into the room) b. [ vp v [ SC [ DP the glass] [ BREAK]]] (The glass broke) We want to emphasize that our approach sharply contrasts with Levin and Rappaport Hovav s in that Manner and Result are not meaning components of the root, but interpretations derived from the position the root occupies in the structure. From now on, we use capital letters to refer to Manner and Result in this sense. It follows that, from this neoconstructionist perspective, expressions such as Manner root or Result root are oxymoronic; if any, we could refer to Manner constructions and Result constructions, that is, constructions where the root is adjoined to v and constructions where the root occupies the SCR-predicate position, respectively. By contrast, we use manner and result, in lowercase letters, to refer to the conceptual content of the root. In this sense, we stick to Grimshaw s (2005, p. 85) strong claim that there are no constraints on how complex the conceptual content of a root can be, unlike Rappaport Hovav & Levin (2010, p. 25), who claim that [m]anner/result complementarity, however, involves the root. As we will see, a root may certainly involve manner and result simultaneously; crucially, however, it may not be interpreted as Manner and Result simultaneously. 6 In our present theory, the constraint in (3) boils down to the syntactic fact that a single root cannot act both as a SCR-like predicate and as a v modifier at the same time (we ll exemplify it in more detail with the case study of the verb climb below). Importantly, the particular constraint in (3) should not be regarded as an inescapable stipulation (as in Rappaport Hovav & Levin, 2010), but can be shown to be derived from the formal fact that a root cannot be incorporated and conflated at the same time (that is, in a single verb). In particular, we follow Haugen s (2009) revisionist claim in (6): (6) Incorporation is conceived of as head-movement (as in Baker, 1988; Hale & Keyser, 1993), and is instantiated through the syntactic operation of Copy, whereas Conflation is instantiated directly through Merge (compounding). Haugen (2009, p. 260) According to Haugen (2009), there are two ways of forming denominal verbs: i.e., via Incorporation or via Conflation. Basically, in Incorporation cases, the denominal verb (e.g., see (7a)) is formed via copying the full matrix of the nominal complement into the null verb (see Hale & Keyser, 1993). In Conflation cases, the denominal verb (e.g., see (8a)) is formed via compounding a root with the null verb. (7) a. The boy danced. b. [ vp [ DP The boy] [ v [ v DANCE] [ DANCE]]] (8) a. The factory horns sirened midday (ex. from Clark & Clark, 1979, apud Borer 2005, p. 69) b. [ vp [ DP The factory horns] [ v [ v SIREN v] [ DP midday]]] 7 The reason why a root cannot be incorporated and conflated at the same time is morphophonological: a single null head, in this case v, may be specified with only one phonological matrix. Since both Incorporation and Conflation are aimed at filling up this null head v, they cannot apply simultaneously. It is important to emphasize the fact that a syntactic approach to the Manner/Result complementarity, like ours, is not equivalent to a purely semantic one. Thus, we disagree with Rappaport Hovav & Levin s claim in (9): (9) For the purposes of investigating manner/result complementarity, the specific type of predicate decomposition representation does not matter. The representations could be recast along neo-davidsonian lines [ ] or as minimalist syntactic structures. Rappaport Hovav & Levin (2010, p. 24, fn. 3) Rappaport Hovav & Levin s (2010) contention in (9) cannot be correct, since the predictions of the semantic and syntactic approaches can be shown to be quite different in an important way. For example, a brief comparison of Koontz-Garboden & Beavers s (2010) semantic approach with our syntactic one will be illustrative. As pointed out by these two semanticists, the manner/result complementarity in (2) cannot be said to hold as such when framed in truly semantic terms, contra Rappaport Hovav & Levin (2008, 2010). For example, Koontz- Garboden & Beavers point out that, conceived truth-conditionally, the prediction is that there should be verbs encoding both manner and result, and manner of death verbs can be claimed to fill in this gap, since they appear to lexicalize both the manner in which an action is carried out (e.g., for electrocute, electrocution) and the resultant state of an entity (e.g., for 8 electrocute, the state of being dead by electrocution), and neither meaning component can be dropped out. Thus, by using manner of death verbs like electrocute, drown or guillotine, Koontz-Garboden & Beavers (2010) claim that Rappaport Hovav & Levin s (2008, 2010) generalization with respect to the manner/result complementarity does not hold as such in semantic theory: the former point out that its scope is narrower than the latter assume. However, to our view, what Koontz Garboden & Beavers (2010) show is not that the complementarity in (2) is too strong; if any, what they show is that (2) cannot be formulated as such in purely semantic terms (contra Rappaport Hovav & Levin, 2010). Thus, Koontz- Garboden & Beavers (2010) conclude in (10): (10) We must admit the third and final logically possible class of eventive roots, namely manner+result roots, contra Rappaport Hovav & Levin s assumption that such roots should not exist. Koontz-Garboden & Beavers (2010, p. 34) We are happy with Koontz-Garboden & Beavers s conclusion in (10), since it is worded in terms of roots: as noted above, we have nothing to say with respect to how complex the conceptual semantics of a root element can be; in particular, a root can of course be claimed to encode manner and result simultaneously as part of its conceptual content, i.e., as part of the conceptual scene it invokes. 3 Rather our proposal here is that when Manner and Result are understood in syntactic terms, there is a validity for the descriptive generalization in (2) 3 Cf. Grimshaw s (2005, p. 75f.) important distinction between semantic structure and semantic content. Following Hale & Keyser (1993f.), we assume that only (part of) the former can be syntactized and then constrained by syntactic well-known principles. In contrast, the complexity of conceptual content (i.e., Grimshaw s [2005] semantic content) is not constrained by syntax. See also Borer, 2005 for extensive discussion on the need to sharply distinguish the meaning conveyed by grammatical structures from the grammatically inert, conceptual content encapsulated in roots (in her termes, listemes). 9 and the constraint in (3). Consider, for example, the manner of death verb guillotine in (11a). Our claim is that the syntactic argument structure corresponding to its use as a causative predicate of change of state is the one depicted in (11b), where the root is structurally interpreted as Result: in (11b) the root is the complement of an abstract P element that expresses Terminal Coincidence Relation (in Hale & Keyser s [2002] sense: a TCR involves a coincidence between one edge or terminus of the theme s path and the place, while a central coincidence relation (CCR) involves a coincidence between the center of the theme and the center of the place). The phonological matrix of the root in (11b) is copied into the null P en route to the null verb via Incorporation. (11) a. They guillotined Mary. b. [ vp [ DP They][ v GUILLOTINE [ PP=SC [ DP Mary] [ P P TCR GUILLOTINE]]]] That the predicate in (11a) is a change-of-state predicate and that it thus must receive an analysis along the lines of that in (11b) is born evidence to by the fact that it admits depictive secondary predication, as shown in (12). Following Rapoport s (1993, p. 179) proposal that only change of state verbs can have object-host depictives, Mateu (2002, pp ) claims that those telic verbs that coappear with these depictive predicates contain a TCR in their argument structure. In contrast, those atelic transitive verbs that involve a CCR cannot have object-host depictives (e.g,. cf. *John pushed the horse i tired i ). 4 (12) They guillotined the murderer i barefooted i. 4 See Hale & Keyser, 2002 and Mateu, 2002, for the claim that verbs like push involve a CCR. Cf. [ vp [ DP John][ v PUSH [ PP [ DP the horse] [ P P CCR PUSH]]]]] (cf. John provided the horse with a push; John gave it a push). But see Harley, 2005, for a different analysis of push-verbs. 10 We claim that the fact that the conceptual content of the root guillotine encodes manner/instrument is not structurally represented in (11b), although this fact could be said to have a linguistic effect: e.g., as is well-known (cf. Alexiadou 2010, Levin & Rappaport Hovav 1995, among many others), agentive change-of-state verbs do not enter into the causative alternation (see (13)). We claim that it is part of our world knowledge that one cannot become guillotined without the intervention of an agentive causer: (13) #Mary guillotined. (# on the reading: Mary became guillotined.) Similarly, in our neo-constructionist framework (see also Borer, 2005), the computational system allow us to generate the syntactic argument structures in (14a) and (14b), where the root is now structurally interpreted as Manner/Means since it is adjoined to v. As noted above, in these cases the root is argued to be compounded with the null verb via Conflation (see Haugen 2009; Mateu, 2005, 2008; McIntyre, 2004). Concerning (14b), our claim is that it is not syntactically but pragmatically ill-formed: its structural interpretation would be, roughly, They created Mary guillotining/with a guillotine (cf. example (8b) above and also (14c)): (14) a.. [ vp [ DP The guy][ v [ v GUILLOTINE v] [ PP=SC [ DP his way] [ P P TCR the list ]]]] 5 b. # [ vp [ DP They] [ v [ v GUILLOTINE v] [ DP Mary]]] (# on the reading: They created Mary guillotining/with a guillotine.) c. [ vp [ DP They] [ v [ v SMILE v] [ DP their thanks]]] 5 It s hard to believe that the same guy who drank, rocked, and guillotined his way onto the shit list of every right-thinking American parent 30 years ago [...]. (http://www.straight.com/article /alice-cooper). 11 The important point for us here is that manner of death verbs like the one exemplified in (11a) do obey the constraint in (3), since the root is only Incorporated (see (11b)) but not Conflated (see (14)). Notice then that the present theory predicts that a root cannot be interpreted as Manner and Result simultaneously, given that it cannot incorporate and conflate at the same time. In the same vein, we do not follow Harley & Haugen s (2007) analysis of instrumental verbs, which, according to them, involve Conflation rather than Incorporation (in Haugen s [2009] terms). Thus, Harley & Haugen (2007, p. 10) claim that English instrumental denominal verbs always involve roots conflating directly with v, indicating manner [...]. Haugen (2009, p. 254) also claims, for the same verbs, that the nominals are directly merged (or conflated) as adverbials directly into v. The representation in (15), taken from Harley, 2005, p. 61, could be said to be applied to the instrument verb in (11a), as shown in (16): 6 6 Harley s analysis in (15) raises a non-trivial theoretical problem: unlike Harley (2005) and Marantz (1997), Borer (2005) and Acedo-Matellán & Mateu (2010) argue that roots do not project (i.e., their non-relational nature prevents them from taking complements). Moreover, it is not clear to us which empirical evidence one could put forward in order to motivate a syntactic encoding of the root head hit besides the adjunct one encoded by HAMMER in (15). 12 (15) Sue hammered the metal. vp DP v hammering Sue v P (hit) DP the metal (16) They guillotined Mary. vp DP v guillotining They v P (hit?) DP Mary As pointed out by Acedo-Matellán & Mateu (2010), there is no empirical evidence for the Conflation analysis of instrumental denominal verbs depicted in (16). In other words, much as the conceptual content of the root GUILLOTINE includes the instrument with which an action is carried out, and that, thus, it codifies a manner, this does not necessarily mean that 13 this root need be merged as adjunct to v through Conflation. On the contrary, as claimed by Acedo-Matellán & Mateu (2010), a predicate like that in (16) is perfectly amenable to an analysis where the root occupies the complement position of a small-clause-like projection and is thus interpreted as Result (see 11b). As for Conflation (i.e., Compounding of a root with a null light verb), there are compelling reasons to claim that it is only to be found in those constructions that involve Talmy s (1991, 2000) Co-event pattern (see below). Be this as it may, for the purposes of the present paper it is not relevant whether Harley & Haugen s (2007) analysis of instrument verbs as those in (15) and (16) (based on Conflation) is correct or it is ours (based on Incorporation). What is relevant here is that a root cannot incorporate and conflate at the same time, which explains the Manner/Result complementarity. A final caveat is in order with respect to the Manner/Result complementarity. It is important to point out that such a complementarity only emerges in cases where a monomorphemic verb is involved: e.g., notice that (3) does not hold for resultative constructions like John wiped the table clean, where the verb only encodes Manner, the Result component being encoded by the adjective. Importantly, the word level is not relevant in establishing the complementarity (for a different view, see Rappaport Hovav & Levin, 2010, p. 26). For instance, in the outprefixation construction, exemplified in (17), the prefix encodes Result and the verb expresses Manner (Talmy, 2000): (17) John outswam/outdanced/outworked Mary. Another relevant case could be the one discussed by Marantz (2001, p. 21, 2005, p. 12): according to him, the verb destroy and related Latinate verbs (e.g., construct, instruct, restructure, etc.) involve the bimorphemic analysis in (18): STROY is a manner root that 14 incorporates a particle, spelled out de-, that takes an inner subject as the direct object of the syntactically derived verb destroy. According to Marantz, the presence of the root i
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks