CHAPTER 9
The English conative as a family of constructions
Towards a usage-based approach
Pilar Guerrero Medina. University of Cordoba, Spain
This chapter aims to provide a constructionist usage-based analysis of English conative expressions, arguing that a family of related constructions is required to account for the semantico-pragmatic properties of the at- frame in English. Drawing mainly on Broccias's (2001) and Perek and Lemmens's (2010) analyses, I challenge Goldberg's (1995) monosemic analysis of the conative construction, where the 'directed action' meaning remains invariable, highlighting the essential role played by the verb's inherent lexical semantics in determining the specific constructional senses that can be subsumed under the rubric of conative tises. Three distinct configurations are posited:.the allative af-construction, instantiated by non-resultative verbs (Tsunoda 1985), the ablative a£-construction, instantiated by resultative verbs, and the directional at-conscruction, compatible with intransitive verbs of Visual perception'.
Keywords: ablative at-construction, ailative at-construction, conative construction, Construction Grammar, directional at-construction, family-resemblance
l, Introduction
This chapter explores the issue of lexical-constructional integration in the English, conative construction, which in the linguistic literature is often presented as the intransitive variant of the so-called 'conative alternation* illustrated by Levin's (1993:6) examples in (1) and (2):
(1) a..   Margaret cut the bread. b.   Margaret cut at the bread.
(2) a.   Carta hit the door.
b.   Carla hit at the door.
DOi l0.L075/hcp.5B,10gue
© 2017 John Benjamins Publishing Company
»31
178  Pilar Guerrero Medina
The English conative as a family of constructions 279
In this type of object-oblique alternation the subject of the transitive variant, regarded as more basic than its counterpart, bears the same semantic relation to the verb as the subject of the derived intransitive variant. The conative alternation affects the syntactic expression and semantics of the second argument, which is rendered by a prepositional phrase (most frequently at, as in (lb) and (2b)) or on {with certain verbs of 'ingesting' like bite or nibble and the push/pull verbs):
(3) a. I pushed the table.
b. Ipusked at/on the table.
(4) a. The mouse nibbled the cheese.
b. The mouse nibbled at/on the cheese.
(Levin 1993:42) (Levin 1993:42)
The intransitive variant of this alternation is frequently associated with 'endeavor' or attempted action in contrast to the success' expressed in the transitive pattern (see Huddleston 2002:298, Levin 1993:42 and Perek 2015:90-91). According to Levin. (1993:42), the verb in the intransitive variant "describes an 'attempted' action without specifying whether the action was actually carried out". In (lb), which could be paraphrased by "Margaret tried to cut the bread," there is no entailment that the action denoted by the verb to cut was completed. Similarly, Dixon's account of the difference between He kicked the ball and He kicked at the ball is as follows:
A semantically canonical sentence with kick is something like He kicked the bait, one assumes that he aimed his foot at the ball, it made contact, and the ball flew off. On hearing He kicked at the ball one might infer that the aim was not achieved, i.e. he missed making contact (Dixon 2005:298)
However, the label 'conative' is not entirely appropriate (or indeed "misleading" as claimed by Broccias 2001:67), since the association of the at-construction with these meanings of 'non-achievement" or 'attempted action is clearly an oversimplification (see Perek and Lemmeos 2010: §19 and Perek 2014:63-64). As Pinker (1989:108) points oat, "if John cuts at the bread, it's not that the knife never arrives at the bread; rather, the bread was not properly cut".
In this chapter, I adopt a constructionist view of the conative pattern in English, arguing that a family-resemblance analysis, where groups of constructions are "united by related but not identical syntax and by related but not identical semantics" (Goldberg and fackendoff 2004:563), is required to account for the seman-tico-pragmatic properties of the so-called conative construction in English. The remainder of this chapter is organized as follows: Section 2 presents an overview of a number of non-derivational approaches to the conative construction, which can be regarded as compatible to some extent. Section 3 is concerned with the issue of meaning construction and lexical-constructional interaction in the family of English conative expressions. Starting from Broccias's (2001) and Perek and
Lemmens's (2010) accounts of the at-construction, three main configurations are posited: the aliative af-constructioti, the ablative af-construction and the directional at-construction. Section 4 explores the interaction between the semantico-prag-rnatic value of the conative construction and the semantics of three of the verb classes which enter into each of the three subconstriictions, i.e. verbs of 'hitting' (instantiating the aliative pattern), verbs of 'cutting' and 'ingesting' (instantiating the ablative pattern), and verbs of 'visual perception (instantiating the directional pattern). I have drawn mainly on corpus data from the Corpus of Contemporary American English (Davies 2008; henceforth COCA) and from the XML edition of the British National Corpus (henceforth BNC).1
2,  The conative construction in the linguistic literature
In this section I present a general overview of some of the main different non-derivational analyses of the conative or at-construction in the linguistic literature.2
2.1    The lexical rule approach
Proponents of the projectionist approach claim that the syntactic realization of arguments is projected from the verb's lexical entry. For example, on Pinker's (1989) approach to argument structure, alternations are analysed as involving the application of (language-specific) lexical rules to a verb's lexical semantic structure. As Levin and Rappaport (2005:10) note, one of the most explicit statements of the assumption that only certain components of verb meaning are relevant to argument realization is made by Pinker. In the author's (1989:108) words, "it's not what possibly or typically goes in an event that matters; it's what the verb's semantic representation is choosy about in that event that matters". For the conative alternation Pinker (1989:104) postulates a (broad-range) lexico-semantic rule that takes as input a thematic core of the form x acts ok y and produces the semantic structure x goes towards x acting-on y.3 Pinker also posits narrow-range lexical
1. BNC examples have been identified by means of a three-letter code and the sentence number within the text where the hit was found. The source information for COCA examples includes the date and the type of text fiom which the hit was extracted (SPOK: spoken; FIC; fiction).
1. Throughout this chapter I use the name conative as this is the term commonly used to refer to the construction in the linguistic literature (see Van der Leek 1996:376; m 3).
3. In the output of Pinker's lexical rule, 'goes' and 'toward' are "interpreted in a semantic field where locations are treated as intended states or events" (Pinker 19B9:104).
■I
280 Pilar Guerrero Medina
rules, which apply to "narrow conflation classes" of verbs (1989:103). The narrowly defined subclasses of verbs which are eligible to enter into the conative alternation "must signify a type of motion resulting in a type of contact" (1989:105). Verbs of 'cutting' (cut, slash, chop, hack, chip, etc.) and verbs of 'hitting' (hit beat, elbow, kick, punch, poke, etc.) are compatible with the construction. Verbs of 'touching' (touch kiss, hug, stroke, contact, etc.) and verbs of 'breaking" (breah shutter, crack, split, crumble, etc.) are not allowed in the construction:
(5) a.   'Nancy touched at the cat. b. *Jerry broke at the bread.
(Pinker 1989: i 04) {ibid.)
A similar analysis is put forward by Guerssel et aL (1985:58-59). The conative alternation is accounted for by a lexical rule which must be sensitive to the verb's lexical properties as indicated in its ICS (Lexical Conceptual Structure), The conative rule operates on the Predicate Argument Structure of the verb and modifies the LCS of verbs involving notions of contact and effect: "The conative construction only occurs with verbs whose LCS includes both an effect clause, 'x produce effect on y and a contact clause, 'by entity coming into contact with y'" (Guerssel et al. 1985:59). Both clauses are essential: verbs of contact", whose LCS lack an ebfect clause, and verbs of 'breaking', whose LCS lack a contact clause, do not undergo the conative rule.
As Goldberg argues, the lexical rule approach is largely compatible with her own approach, the main difference stemming from the increased focus of the constructional analysis "on the nature of the relation between the verb and the construction (...)" (1995:9).4 (See Section 2,4 for discussion of Goldberg's constructional analysis of the conative construction).
2.2   Tenny's aspectual approach
According to Tenny's Aspectual Interface Hypothesis, "the universal principles of mapping between thematic structure and syntactic argument structure are governed by aspectual properties" (1994:2). On Tenny's aspectual approach to event conceptualization, 'delimitedness' and measuring out" are the properties that lead us into the nature of the syntax/semantics interface". The term 'delimitedness' "refers
4. According to Rappaport and Levin (1998:129), the constructionist and projectionist approaches share many elements. Both recognize the same basic distinction between the 'structural and the 'idiosyncratic' aspects of meaning, and the major difference between the two concerns whether the structural aspect of meaning (ie. the lexical semantic template) resides in the lexical entry of individual words or is rather associated with the syntactic structure.
The English conative as a family of constructions 281
to the property of an event's having a distinct, definite an inherent endpoint in time" (1994:4).5 The direct internal argument plays the role of'measuring out' the event denoted by the verb: "'measuring out' entails that direct argument plays a particular role in delimitingthe event" (Tenny 1994:5).6 Tenny (1994:123) accounts for the aspectual and non-aspectual elements of meaning that "combine to isolate the class of verbs that undergo the conative alternation in English". The constraint that verbs of 'motion-towards' and 'contact' "may not unambiguously specify a measuring argument is an aspectual condition on the alternation" (1994:123; my emphasis).
According to Tenny (1994:47), "the conative alternation applies to verbs that are ambiguous with respect to measuring arguments (...) or to verbs that have non-measuring arguments". Verbs of change of state (such as break, crack, splinter, etc.) which obligatorily require a 'measuring argument', as they "enforce a delimiting change of state or impart an endstate entailment on the interpretation" (Tenny 1994:46), cannot enter into the conative alternation.
(6)  a.   Janet broke the bread, b. 'Janet broke at the bread.
(Tenny 1994:46) (ibid.)
It shouldbe observed, however, that a change-of-state verb like tear (also excluded from the conative alternation by Levin 1993 and Pinker 1989) is compatible with the at-construction, as shown in (7):7
(7)  a.   tie kept tearing at the wrapping paper until it was all removed.
(Dixon 2005:11)
b.   They simply tear at the flesh of the victim until the animal is weak from loss of Mood (...). (BNC BLX 1773)
5. Jackendoff (1996) presents a critique of Tenny's (1994} treatment of aspectual notions such as 'measuring out' and 'incremental theme' to account for syntactic argument selection. See also Bois (2003:62-65) for critical discussion of Tenny's account of argument alternations (e.g. her analysis of the resultanve construction) in terms of aspectual structure.
6. The notion of 'delimitedness' also plays important role in Vendler's verb typology. Vendler (L967) proposed four aspectual classes: activities, accomplishments, achievements and states, which continue to he the most widely identified classes (see Levin and Rapapport 2005:88). In Vendler's typology of verbal aspect, accomplishments and achievements are delimited events; states and activities are non-delimited.
7. The argument that measures out the event in (7a), the wrapping paper, acts here like an Incremental Theme, undergoing a definite change of state "in distinguishable separate stages" (Dowty 1991:568): the action was done several times, "bit by bit until a result was achieved" (Dixon 2005:119).
1 i
282 Pilar Guerrero Medina
"Die English coaatrve as a family of constructions 183
As Boas observes in this connection, Tenny's aspectual approach fails to account "for the different senses that are inherent to a verb's lexical semantics" (Boas 2003:65). I share Boas's view that "once mote attention is paid to the lexical semantic information associated with the constituents of the construction, especially the verb, a better understanding of the relevant constraints and exceptions -will follow" (2003:101; my emphasis) (see Section 3 for further discussion).
2.3   Van der Leek's compositional approach
Van det Leek (1996) presents a compositional analysis of the conative construction, suggesting two possible readings of the tit-pattern; an 'estimated point of contact' reading, exemplified in (8), and a 'point of contact' reading, which implies a 'bit-by-bit' interpretation, as in (9).
(8) a.   Sam threw a handful of mud at Sandy. (Van der Leek 1996:368) b.  Sam sprayed at the trees with some insecticide. (Van der Leek 1996:370)
(9) a.   She pulled at the thread until it came out of the piece of cloth. (ibid.)
b. [Sandy was] sipping at her drink just to be polite. (Van der Leek 1996:367)
c. The mouse rubbled at a piece of cheese. (Van der Leek 1996:374)
Van der Leek (1996:368) starts from the assumption that "meaning (...) is essentially compositional, with senses partly created in context", and suggests that the meaning of the at-construction is obtained by merging the 'skeletal meaning' of the verb with that of the ar-phrase. She analyses the behavior of at with non-cona-tive verbs, such as throw in (8a), where "an flf-phrase identifies a point of contact without signaling a path" (1996:368), and concludes that "at designates a point of contact and that the sense of estimated-point-of-contact is a 'sense in context' (Pustejovsky 1995:60) that gets created (...) in combination with throw" (Van der Leek 1996:369).
Broccias (2001) discusses the weaknesses of Van der Leek's analysis. In Broccias's view, Van der Leek's compositional analysis is not sufficient to motivate all the attested examples, "because it bvergenerates' (...), excludes plausible readings (...), and does not tackle the question of metaphorical usage" (Broccias 2001:72) (see Section 2.4 for an overview of Broccias's alternative approach).
2.4   The constructional approach
The variation inmeaning that accompanies many argument alternations has led many researchers to argue for a constructional approach, where "the verb is integrated into the construction rather than determining the construction, and the construction itself licenses some of the complement structure" (Levin and Rappaport 2005:190).
On Goldberg's (1995) constructional approach,8 the conative construction is not exclusively dependent on the semantics of the main verb. In fact, Goldberg (1995:1) defines constructions as "form-meaning correspondences that exist independently of particular verbs".9 Figure 1 represents how the semantics of the verb class integrates into the semantics of the conative construction (Goldberg 1995:64):'°
DIKECT-ACTCON-AT <   agt   theme > W
R*: instance, intended result [+motion +contact]
Syn
PRED
SUBJ OBL'ať
* It symbolizes the relation between the meaning of the vert" arid the morning of the construction.
Figure 1. Conative Construction (Goldberg 1995:64)
8. According to Levin and Rappaport (2005:191-192), Goldberg (1995) represents the 'traditional' version of constructional theories. Proponents of the 'traditional' constructional approach view syntactic structures as meaningful constructions (or form-meaning pairings). 'Neoconstructionists' on the other hand, assume that the meaning encoded in the construction is compositionaliy derived from the meaning of the verb together with the meaning of the syntactic structure (see Borer 2003).
9. Accordingto Goldberg (1995:229; fn 6) the concept of meaningis "construed broadly enough to include contexts of use, as well as traditional notions of semantics*.
10. The fact that the Agent is the profiled argument role (expressed by a direct grammatical relation) is indicated by bold type. As Perek and Lemmens {2010; fn 9) point out, Goldberg's use of the term 'theme', introduced by Gruber (1965) Co refer to the role of an entity "undergoing motion", is infelicitous here as this argument does not necessarily undergo a change of location. The authors replace it with the more fortunate label 'target'.
284 Pilar Guerrero Medina
The English conative as a family of constructions »85
In Goldberg's system, "the semantics of the verb classes and the semantics of the constructions are integrated to yield the semantics of particular expressions" (199S: 60). As shown in Figure 1, the meaning of the conative construction (represented as direct-action-at) remains constant independently of the type of relation existing between the meaning of the verb and the meaning of the construction. The verbal subevent can be related to the constructional event type in two ways: when a verb of 'attention (such as look or aim) fuses with the construction, the verb's semantics (represented by PRED) is an 'instance' of the semantics of the construction (see Section 3,1.3). On the other hand, [+rnotion, +contact] verbs like hit or shoot are related to the construction by a relation of intended result' as in (10):
(10) a.   In front and to each side, people were shooting at him, but they all missed.
(BNC GOL 3632) b.   He hit at her face with the gun, but she jerked her head back,
(BNCH85 837)
However, the question arises whether other subtypes of prototypically transitive verbs (e.g. stab, hack, cut), which dearly "entail physical change in the object" (Dowty 2001:184), can actually be seen as instantiating the 'intended-result' relation in the conative pattern:
(11) a.   She walked a few feet and stabbed at the earth with the fork.
(BNC HP 1853)
b. Catradine hacked at the other man's tick and a length of wood flew away.
(BNC 1391 GVL)
c. As the sea flooded around them, the boy grabbed an axe and hacked at the slithery tentacles. As he cut at them, the monster released a jet of blank ink, withdrew and disappeared. (BNC G3P 1424)
Finally, in the ar-pattern with verbs of'ingesting,' which is not discussed by Goldberg in her analysis, the label 'conative' is again clearly inappropriate. In the BNC examples with gnaw, nibble andst^ in (12) there is indeed consumption of food and wine:
(12) a.   He smiles a marrnite-stained smile and gnawed at another comer of his
sandwich. (BNC FT9 1289)
b. He nibbled at his food and studied her reflectively. (BNCFNT388)
c. Nick sipped at his glass of white wine. (BNC BPT 61)
The 'intended-result' reading, which in Goldberg's analysis results from the coercion of the verb's semantics to the semantics of the construction, clearly does not account for the conative uses of the construction in (12), where the oblique
objects denote affected entities and where the 'attempted action reading is thus insufficient.11
Broccias (2001) proposes a constructional (or schematic) analysis of so-called conative constructions, refining Van der Leek's and Goldberg's analyses:u
The problem of why certain verbs cannot appear in conjunction with an ar-phrase must be evaluated on the basis of a (non-deterministic) interplay between the scenario(s) coded by the af-construction(s) and the scenario(s) associated with a certain verb. (Broccias20Ql:72)
Broccias distinguishes three scenarios (or schemas): the allative schema, "where af-fectedness is possible but not necessary" (2001:74), the ablative schema, associated with continuous actions and with a component with refers to the attempted movement of an entity (2001:76), and the allative-ablative schema, which results from the interaction of the allative and ablative scenarios. Examples from Broccias, reproduced in (13), (14) and (15), illustrate the linguistic instantiations of these three schemas:13
(13) a.   Sally kicked at the wall. (Broccias 2001:73)
b. Sally shouted at Chris. (ibid.)
c. He clutched desperately at the branch as he fell. (ibid.)
(14) a.   The horse pulled at the cart. (Broccias 2001:75)
b. His tongue tickled at his hand. (Broccias 2001:77)
c. Sam chipped at the rock. (Broccias 2001:76)
(15) a.   Sam sprayed at the trees with insecticide. (Broccias 2001:79) b.   He was working athis painting. (ibid.)
The allative schema in (13) is associated with "translational movement" construed as the emission of a force (Broccias 2001:73) and with "lack of necessary affected-ness" (Broccias 2001:74). The ablative scenario in (14) is associated with the feature of continuity "either by repetition or prolonging one instance of the action" It comprises three cases: removal (14a), release (14b)14 and creation/destruction (14c).
11. In Construction Grammar coercion can be understood as "the resolution of conflict between constructional and lexical denotata" (MichaeSis 2004:8) in such a way that the conflicting element ends up conforming to the semantico-pragmatic profile of the construction.
12. Broccias (2001:68) replaces the label 'conative' with the "more neutral label of flf-construction"
13. Instances of shouting at1 are subsumed by Broccias under the allative partem. Verbs of perception are not mentioned in his analysis.
14. The release case includes verbs of 'touching' (e.g. stroke, tickle, finch), which authors like Levin (1995:42), Pinker (1989:105) and Goldberg (1995:63) regard as incompatible with the conative construction.
3
386 Pilar Guerrero Medina
The English conative as a family of constructiorLs 187
The aUative-ablative schema iu (15) shares features with both scenarios. It involves translationa] allative motion and contains an ablative component referring to the notions of continuous action and attempt.
3.   Lexical-constructional integration in the English conative pattern
Verb-centered lexical-semantic and aspectual approaches have proved insufficient to give a full account of the df-pattern, as they fail to observe that "the constructions in which verbs occur are meaningful in and by themselves" (Lemmens 1998:38). However, in my view the Goldbergian approach still provides an incomplete account of the semantico-pragmatic value of the conative construction. Goldberg's representation in Figure 1 still needs to be further refined in order to fully reflect the semantic and discourse-pragmatic properties of the English conative.
Goldberg (1995:67} conceives of grammar as a repertoire of constructions, which "form a network and are linked by inheritance relations which motivate many of the properties of particular constructions". In Goldberg's words, "[rjela-tionships between and among constructions are captured via a default [i.e. partial] inheritance network" (2013:21). Similarly, in Fried's (2007:727) view, "grammar is seen as consisting of networks of constructions, related through shared properties".
In this context, I evaluate Goldberg" s 1995 monosemic analysis of the conative construction, where the meaning of the construction remains invariable, arguing that the family of at-constructions involves at least three different configurations which trigger different aspectual and semantic interpretations:15 (i) the allative sf-pattem (or conative proper type), as illustrated in (16); (ii) the ablative af-pat-tem, as in (17), and (iii) the directional sf-pattern, instantiated by (intransitive) verbs of 'attention' (including verbs of 'visual perception' of the 'look' type), as exemplified in (18):
(16) a.   With an uncoordinated reflex, Ikickedatihe thing, knocking it a few meters away. (COCA 2011, FIC)
b.   Soldiers also shot at mourners burying the dead at two cemeteries in Bamako.
(BNC A03 86)
15. Adams (2001:28, 64) distinguishes four main flf-forms carrying different meanings: the contact frame', which negates the change-of-state implicature; the exerted-force' frame, which Implicates change of location of the object; the 'impact frame', where contact with the direct object is entailed, and the 'genericat-frame', which carries a more repetitive reading. Within each group the author distinguishes further subtypes with respect to the nature of the verbs that participate in them.
(17) a.   1 was nervous, so I sat by Reef, stroking him and talking to him while my dad
cut at the net bindinghim. (COCA 2003, FIC)
b.   He hacked at the man's legs, but the blows did no harm.
(COCA 2012, FIC)
(18) a.   They all looked disdainfully at the boy's father. (COCA 2010, FIC) b.   Plaster was falling from a small hole in the ceiling; (...) Rory aimed at the
small hole, pulled the trigger. (BNC GOA 1237)
According to Goldberg (1995:24), "the meanings of constructions and verbs interact in nontrivial ways", and therefore there is bound to be "some cross-reference between verbs and argument structures". However, as Boas (2008; 125) points out, "the interactions between verbs and constructions are difficult to constrain". Following the author's (2008:125) suggestion that "it is still necessary to include more specific semantic and syntactic information in a verb's lexical entry in order to be able to predict its distribution with a variety of different constructions", I adopt a usage-based 'splitting' approach to lexical description, exploring the interaction between the 'idiosyncratic' elements of verb meaning - what Rappaport and Levin (1998:106) term 'constants' - and the overall semantico-pragmatic import of the conative construction.
The 'directed action schema is in fact too general to characterize the semantics of all conative sentences. In Perek and Lemmens's (2010: §59) words, "an account taking such a general meaning as basis runs the risk of being too permissive and overgeneralizing by allotting too much power to coercion processes", Actually, as argued in Section 2.4, the 'intended-result' interpretation does not account for the conative uses of verbs of ingestion like sip or nibble, which clearly involve an affected object and a 'bit-by-bit' interpretation (see also Guerrero 2011:192 and Perek 2014:63-64).
In Goldberg's constructional approach to argument realization, coercion "is only licensed by particular constructions in the language. That is, coercion is only possible when a construction requires a particular interpretation that is not independently coded by particular lexical items" (1995:159). As Panther (2005:363) points out, one insufficiency of Goldberg's approach lies precisely in the umdirec-tionality of the coercion process, always working from constructional meaning to lexical meaning:
Constructions are assumed to have meaning and lexical items inserted in a construction do not necessarily have to fit the construction meaning 'perfectly" but may, under certain circumstances, be coerced into a meaning determined by the construction meaning (see Goldberg 1995). But it is not impossible to imagine that lexical meaning might also 'nibble at' constructional meaning and change it met-onymically. (Panther 2005:363: my emphasis)
.;
r
-■3
=35
I
■ ; 1
288 Pilar Guerrero Medina
The English conative as a family of constructions 289
In fact, Goldberg (1995:65) herself acknowledges this possibility when she says that we may "consider certain verb inherent semantics to bear a metonymic relation to the semantics of the construction"
Following Perek (2014:63), we can conclude that the semantic meaning contributed by the conative construction is indeed highly variable, "which makes a maximally general semantic characterization of the construction challenging, if possible at all"ls In line with Pereks proposal, in what follows I explore the issue of meaning construction in the conative (or 'conative-type') pattern analyzing the interaction between the semantics of the verb classes which appear subsumed in conative structures and the semantico-pragmatic value of the construction itself.
l,i    The role of the verb's semantics
Tsunoda (1985: 387) makes a distinction between resultative transitive verbs, i.e., prototypical transitive verbs "which describe an action that not only impinges on the patient but necessarily creates a change in it" and non-resultative transitive verbs, i.e., verbs which do not necessarily involve any physical change in the patient. Broccias's illative and ablative patterns subsume verbs from these two semantic types: the allative schema is associated with inherendy non-resultative verbs (e.g, kick, shoot), whereas the ablative schema is associated with inherently resultative verbs (e.g. cut,gnaw, nibble).
It should be observed that the distinction between resultative and non-resultative verbs correlates only partially with the Aktionsart distinction between 'telic or delimited eventualities' {accomplishments and achievements in Vendler's 1967 classification) and 'atelic or nondelimited eventualities' (Vendler's activities), as non-resultative verbs like hit and kick may be ambiguous between a delimited and non-delimited reading, e.g. hit/kick the wall once (delimited) vs hit/kick one wall (delimited, non-delimited) (see also Tenny 1987:119).
In the remainder of this section I illustrate the interaction between verb and constructions with corpus data including representative verbs from four semantic types: verbs of lutting' (kick, shoot), verbs of cutting' (cut, hack), verbs of 'ingesting' (sip, nibble), and verbs of 'attention' (look, aim).
16. Drawing oq earlier work by Croft (2003), Perek (2014) suggests a variant of collexeme analysis where the verbal distribution of the conative construction is split into semantic classes and 'verb-dass-specifk' constructions axe considered independently. This method is applied to three classes of verbs: verbs of 'cutting', verbs of 'pulling1 and verbs of 'striking'.
3.U The allative ^-construction: non-resultative verbs of 'hitting' Verbs in the lexical subdomain of fitting' (e.g. kick, shoot) are non-resultative verbs which refer to the Manip "being brought through the air" to impact on some thing or person (Dixon 2005:110).17 Dixon (2005:110) classifies "hit' verbs as a subtype of prototypical transitive 'affect1 verbs. However, as Tsunoda (1985:387) rightly points out, verbs Like hit and kick are not prototypical transitive verbs, as they "do not always imply a change in the patient".
The verb shoot is particularly interesting in this regard: it is inherently resultative with the reading of 'shooting dead', where the Object is the animate patient killed with a bullet, as illustrated in (19a), and inherently non-resultative when it means "fire a bullet, an arrow, etc." (citing from the Oxford Advanced Learner's Dictionary), as in (19b).18
(19)  a.   The IRA in Belfast yesterday shot the wrongman when they failed a laboratory analyst, writes David McKittrick. (BNC A2P 504)
b.   Gourmets lured them to piles of turnips, then shot at them with duck-guns.
(BNC F9H 1749)
With the reading of 'shooting dead' the conative construction is not possible, as shown in (20):
(20)  a.   shoot the cow dead b. *shoot at the cow dead
(Tenny 1994:46) (ibid.)
The allative at-pattern is, as Pinker (1989:68) puts it, "compatible with an absence of any effect at all", as shown in (21), where the action does not have an immediate effect on the object and is represented as unbounded as a result:19
(21)  a.   On 2 January 1964 a police constable shot at the President from close range.
He missed, killing a security guard, (BNC FAN 2326)
b. She shot at him, trying to kill him. (COCA, 2012 SPOK)
c. I always carried a half clothes-peg for the purpose - and no matter how he blasphemed and kicked at the door he couldn't get in.   (BNC 1069 ALH)
17. Dixon (2005:10) defines the M&oip role as the role of the entity which the Agent manipulates to come into contact with some thing or peTSon (the Target),
18. The manipulated entity {i.e. the bullet, arrow or projectile) may occupy the Object slot, as in I saw the movie once when Martians shot a gun at you and turn you into a statue (COCA 2000, HQ.
19. According to Capelle and Declerck (2005:893), an event is represented as bounded "if the clause describing it represents the event as reaching a terminal point, i.e. as coming to an end"
Pilar Guerrero Medina
The Eügüsh conative as a family of constructions J91
As argued by Broccias (2001:74-75), the lack of necessary affectedness of the object "correlates with notions such as randomness, attack and difficulty" Similarly, Ikegami (1985:280) observes that the combination of try to with the oblique v. + at expression "implies some obstacle or difficulty".
The allative at-frame does not necessarily remove the implicated contact of the transitive frame, as in (21 c), but emphasizes the greater involvement of the Agent. A sentence like He kicked at the door "could be used to focus on the factthat he was angry and just kicking out in fury, with what the kicks made contact with being of secondary importance" {Dixon 2005:298), Similarly the examples with kick in (22) show that the emphasis is on "the subject's engaging in the activity" {Dixon 2005:299) and on his/her reaction, rather than on the possible effect that the action may have on the object.
(22) a.   For a moment she almost hated him. She kicked at the hay in a burst of frustration. (BNC CSS 2351)
b. I kept cursing, and I kicked at the sidewalk with all the anger I had.
(COCA 2011.F1C)
c. He lacked at the chair-leg and the black, angry words burst out of her.
(BNC254AC4)
3.1.2 The ablative st-construction; rtsultative verbs of 'cutting' and 'ingesting' Dixon (2005:113) classifies verbs of cutting like cut and hack as prototypical transitive verbs of the 'stab' subtype, which "refers to a pointed or bladed Manip penetrating below the surface of the Target":
(23) a.   She cut at a knee, smelled the scorch of flesh as the blade severed through
armor. (COCA 2003, FIC)
b.   Dad hacked at an arm with his knife. (COCA 2008, PEC)
The meaning of these two verbs "involves notions of motion, contact and effect" (Levin 1993:157). They are accomplishment predicates that lexicalize both manner and result. Verbs of 'cutting' can thus be said to consist of two subevents: an activity and a resulting state (Mairat and Faber 2002:53).
The ablative af-frame with resultative verbs of 'cutting' cancels the implica-ture of an endstate or end-location which can be present in the transitive frame and emphasizes the manner of contact variable lexicalized by the verb (see Adams 2001:64). The examples in (24) do not necessarily signal lack of affectedness, but they still do not implicate an endstate or result:
(24) a.   A figure popped out of a doorway to Alexei's left, and he cut at it with his
sword and ran on. (BNC G17 1991)
b. There was frost on his hair, on the fur of his hood, but soon, as he hacked at the tree, his skin began to glow and a fine, warm moisture gleamed on his face. " (BNC HTM 2495)
There are two possible readings for these examples: either the action of 'cutting' or fiacking" is understood to be repeated longer, or there is movement of the Manip towards the object without contact necessarily being entailed. As in (22), the emphasis now seems to be on the undertaking of the action by the agent, rather than on the achievement of a particular result.
In the ablative pattern with verbs of 'ingesting' the conative seems to impose "a partial completion sense" (Van der Leek 1996:375), as illustrated by Huddleston's examples in (25):
(25)  a.   He nibbled the biscuit away.
b. *He nibbled at/on his biscuit away.
(Huddleston 2002:298) {ibid)
The sentence pair with nibble in (26) also exemplifies this 'hohstic/partitive' effect
(26) a.   [...] when he could not immediately think of an answer he nibbled the canapi and viewed the questioner with cold button eyes. (BNC CDN 1047)
b.   Moodie, more like a mouse than ever, nibbled at a bit of cheese.
(BNC HU0156)
In the af-pattem in (26b) the effect is that the entity is ingested partially, in a "bit-by-bif fashion (Van der Leek 1996:374). Verbs ]ike gobble and devour, which cannot occur without a holistic object, would be filtered out by the uncompleted entailment of the construction (see Dowty 2001:184). In the words of Van der Leek "We can only identify an event as a gobbling event if the food stuff mentioned as a whole disappears fast into the gobbler's mouth" (1996:375).
3.1.3    the directional ^-construction: verbs of attention The directional af-pattern is instantiated by verbs of 'attention' (such as look and aim), where the action does not impinge on the patient (Tsunoda. 1985:389). In Goldbergs (1995:64) analysis, look and aim are both instances of direct-action-at (see Figure 1), as shown in (27):
(27)
b.
Fred looked at Ethel Ethel aimed at Fred.
(Goldberg 1995:64) {ibid.)
According to Goldberg (1995:64), these verbs (which are not [+motion, +contact] verbs) "bear an obvious similarity" to other uses of the construction where the relation between the meaning of the verb and the meaning of the construction is one of 'intended result', as in (28):
2^3 Pilar Guerrero Medina
The English conative as a family of constructions 193
(28)  a.   Ethel struck at Fred. b.   Ethel shot at Fred.
(Goldberg 1995:63) {ibid.)
I
In Dixon's semantic classification, verbs of 'attention of the 'look' subtype refer "to the Perceiver directing their attention in order to connect with some Impression" (2005:133).20 Collocations of the verbs look snAgaze with agent-oriented manner adverbials such as thoughtfully and carefully reveal the agent-profiling semantico-pragmatic function of the construction. I present some examples in {29):11
(29)  a.   The Pastor paused for a moment and looked thoughtfully at his flock and, Erika was convinced, piercingly at her. (BNC A7A 426)
b. Miss Honey looked carefully at the tiny girl with dark hair and a round serious face sitting in the second row. (BNC CH4 795)
c, Tobby looked meaningfully at me and changed the subject as fit moved towards the table for lunch. (BNC 2144 CS4)
d. Alex gated thoughtfully at the empty doorway.        (COCA 2007, FIC)
e, She lifted flabby corrugated skin from her arm and gazed meaningfully at Lil. (COCA 2001, FIC)
Drawing on Goldberg*s construction grammar account, Perek and Lemmens (2010: §24) relate the conative frame {comprising Broccias's ablative and ailative schemas to other instances of the ^construction instantiating the constructional meaning 'directed action'.22 Verbs of voluntary visual perception (look, glance, stare, gaze, peer), verbs of 'shouting' (e.g. shout at a person, cry at a person, scream at a person), verbs of laughing' (e.g. laugh at a person) and 'smiling' (smile at a person), verbs of 'growling' (e.g. growl at a person, groan at a person) and 'scoffing" (e.g. sneer at
10. Ikegami (1985:283) mentions that the contrast between see and look at reflects the implicit semantic opposition between achievement and non-achievement, which is typically associated with the unmarked and marked variants of verbs in the conative alternation. As Gruber (1967:942) points out, "one can look at something without seeing it." On the other hand, see is obligatorily non-agentive, as opposed to look: e.g., What John did was to look at Biit vs. *What John did was to see Bill (Gruber 1967:943).
21. The meaning of these verbs contains the parameter of long duration and lexicalizes an "intentional act of visual perception" (Faber and Mairal 1999:152).
u. However, Perek and Lemmens (2010: §52) exclude visual perception from the central meaning of the construction, as it is only the orientational component of look (and not its visual component) which determines its compatibility with the at-construction. The authors present evidence against an extreme interpretation oFGoldberg's (2006) Lexical Origin Hypothesis, according to which "generalisations may arise developmentally as generalizations over lexical items in particular patterns" (2006:92), and conclude that the conative alternation could be viewed as a "higher-level contrastive form-meaning pair" unifying the ailative and ablative patterns (Perek and Lemmens 2010: %66).
a person, jeer at a person), which can be used intransitively with the preposition at but do not allow transitive uses ("Hook a person, * shout a person; *cry a person, *scream a person, *laugh a person,{smile a person, *growl a person; 'groan a person; ""sneer a person, "jeer a person) can also be regarded as instances of the conative construction as they also involve some notion of 'directed action' and even 'attack' (Ikegami 1985:282), as shown in (30):
(30) a.   They shouted at him and he rode off on a bicycle, latter to be stopped by the police (...). (BCNBM4107)
b. She came home and screamed at her husband. Then she screamed at the children. (BNC ACL 1274)
c. Whenever we move down a street, enter a pub, club or restaurant we risk being laughed at, sworn at, spat at. (BNC 266 C9S)
3.1   The fusion of verbal and constructional semantics: A family-resemblance analysis
Figure 2 represents the family of English conative constructions, connected though "principles of inheritance" (Fried 2007:237)M and family resemblance links.
The three configurations which form the family of af-constructions can be considered to be "the result of the modulation of the lexical semantics of the main verb •with the overall constructional meaning" (Gonzalvez 2008:119). Each sub-construction inherits the general semantic and pragmatic specifications of the construction (action with a goal-achieving intention / agent-profiling and patient-defocusinq) , and these general dimensions are further modulated by the lexical semantics of each verb class in each of the three subconstructions involved:
Along the lines of Fried's (2007) usage-based analysis, where evidence is provided by actual usage in specific communicative situations,241 assume that generalizations about the conative pattern revolve mainly around semantic and pragmatic notions. As I have argued in Guerrero (2011:189-199),35 Cooreman's definition
13. In Fried's (2007:737) words, the notion of inheritance "is used in constructional analysis as one way of expressing generalizations about grammatical patterns that show partial overlap of features, whether formal or functional"
24. In line with the basic methodology of usage-based approaches, where "item-specific knowledge exists alongside generalizations" (Goldberg 2006:44), in this chapter I have based my assumptions on authentic data, drawn from corpora.
25. In Guerrero (2011) I explore the functional dimensions of the English conative alternation, examining to what extent this object-oblique alternation fulfils similar functions to the antipas-sive in other languages.
294 Pilar Guerrero Medina
Syntax: NPVatPP
Semantics: Action with a Goal-Achieving Intention Pragmatics: Agent-Propiling and Patient-Defocusinc
ablative
Lack o? Completion, Partial Result
, resultative verb I t,      , ,
semantics 1 ,mgeSt,n«; j cutting
allative
LACK OP NECfiSSAJtf AFFECTEDWZSS
rnori-resultative 'contact1
directional. Directed Action
j intransitive ] attention'
verbs of 'shouting', 'growling^ 'smiling', laughing' •I*,: metaphorical extension links (Goldberg 1995)
Figure 2. A family- resemblance analysis of the English at-construction
of the general function of the antipassive captures the semantic and pragmatic dimensions of the English conative construction:
The antipassive which is used for semantic/pragmatic reasons is best described as indicating a certain degree of difficulty with which an effect stemming from an activity by a on an identifiable o can be recognized. (Cooreman 3 994:65)
The necessary level of difficulty for the conative construction to be used in English is to a great extent dependent on those semantic and aspectual components of verb meaning which determine the way in which the entailment of Incompletion of the construction is manifested. According to Dowty (2001:184), "the meaning of the construction entails that the action is incomplete in one of two ways":24
i ■
The English conative as 8 family of constructions 19s
With verbs that entail physical change in the patient, the derived construction means that some but not all of the patient is affected, and is consistent with the possibility that very iittie is affected (cf. eat at the cake}.
The remaining verbs in this construction entail motion and contact but not necessarily any physical change in the Patient at all, but they do involve a distinguishable manner or shape of movement by the agent even if contact fails to be achieved (...). With these verbs, the action is understood not to involve contact but only to involve this characteristic movement (cf. He swatted at the fly but missed (...)). (Dowty 2001:184-185)
The first group of verbs mentioned by Dowty corresponds to Tsunoda's (1985) resultative type, whereas the remaining verbs are non-resultative verbs.
As shown in Figure 2, the allative er-pattern instantiates the most prototypical realization of the meaning action with a goal-achieving intention (Ikegami 1985:280), With non-measuring verbs of contact (Tenny 1994), the at-construction may be used to indicate that contact is not achieved, as illustrated in (22) in Section 3.1.1.
Scholars Like Givon (2001) and Croft (1998) regard the demotion of the patient into an oblique case to cancel the affectedness entailment of the verb as the main function of the conative alternation. In Croft's (1998:45) words: "In general, an object-oblique alternation is associated with a difference in degree of affectedness of the participant alternatively encoded, so that the object encoding indicates a high(er) degree of afFectedness". However, the conative frame does not always render "a sense of low affectedness" (Givon 2001:171), but rather confers on the agent "the salience and accessibility that comes with starting-point status" (Langacker 1987:396). The deliberateness of the agent - the "volitional element" in Hopper and Thompson's (1980:264) terms - is inherently part of the semantico-pragmatic meaning of the conative pattern.
The broken arrow in Figure 2 indicates metaphorical extension (IM) links (Goldberg 1995:81) between the so-called directional af-pattern and the af-pattem with agentive uses of verbs of shouting' or 'growling' laughing' or 'smiling! As observed by Perek and Lemrnens (2010: §51), apossible motivation for the metaphorical use of these verbs in the directional at-construction could be to see "mental objects as entities in the mind with which the individual can interact".
16. In Dowry's (2001) meaning-driven analysis, the semantic change that arises through a particular alternation 'filters' the verbs that can enter into it. In the English conative construction, the 'incompleted entailment' of the at-variant dashes with the entailment of completion of action of a verh like break.
4.  Pinal remarks
In this chapter I have argued that the conative construction can be analyzed as a family of constructions comprising three main subconstructions with instantiate related meanings: the allative pattern, the ablative pattern, and the directional
-
296 Püar Guerrero Medina
The English conative as a family of constructions 297
pattern, with values such as lack of necessary a jfecteonbss, lack of completion or partial hesuit and directed action, as shown in Figure 2.v
Although I adhere to Goldberg's (2006:3) notion of grammatical constructions as "conventionalized pairings of form and function", I do not fully support the author's (2006:25) claim that "each argument structure pattern is best analyzed on its own terms, without relying on explicit or implicit reference to a possible alternate paraphrase (Goldberg 2006:25; my emphasis). I concur with Perek and Lemmens (2010: §65) that "not all alternations should be rejected on the grounds of the surface generalization hypothesis" and also agree with Davidse's claim that "alternations are indeed relevant both to verb meaning and to the semantics of constructions" (2Q11:12).2a
The conative construction can be understood as a high-level conceptual configuration capable of accommodating low-level structures of the kind provided by the lexical predicates which are coerced into the construction (see Ruiz de Mendoza and Gonzalvez 2011:192), as long as they do not clash with the incompleted entailment of the construction.
Acknowledgements
I am grateful to two anonymous reviewers for their valuable comments and suggestions oq earlier versions of this chapter. The research reported here was conducted under the auspices of the project EFI2012-36S23/FILO, funded by the Spanish Ministry of Economy and CompetitivenesS-
27. One reviewer points out that the inclusion of constructions expressing perception (e.g. X looked at Y) in the conative family of constructions may be problematic, as the directional pattern with verbs of perception is not marked in any way. Although I share the reviewer's Levjnian position that the meaning of the conative construction emerges from the contrast between the marked and unmarked variants of the conative alternation (see also Perek 2015:101), I believe that the allattve, ablative and directional st-patterns build up a family-resemblance group of structures. The inclusion of verbs of volitional visual perception within the conative family of constructions would be motivated by their agentive nature and by their orientational and directional components. Contra Gotdberg, however, I claim that the semantics of all conative uses cannot be accounted for by the schematic notion of 'directed action' (see Section 3).
28. See also Praost (this volume: 17-51), where the author emphasizes the role of verb classes and verb-dass-specific alternations to identify the different variants of the German search-construction.
References
Adams, CM. 2001. The conative alternation: An exploration of semelfactives and the elusive non-Theme Patient, Unpublished MA. thesis. University of Canterbury, Christchurch, New Zealand-Boas, H. 2003. A constructional approach to resultaiives. Stanford, CA: CSU Publications. Boas, H. 2008. Determining the structure of lexical entries and grammatical constructions in Construction Grammar. Annual Review of Cognitive Linguistics, 6.113-144. doi: to.io75/ard.6.o6hoa
Borer, H. 2003. The grammar machine. In A. Alexiadou, E. Anagnostopoulou, & M. Everaert (Eds.), The unaccusativity puzzle (288-331). Oxford: Oxford University Press,
Breccias, C, 2001. Allative and ablative ar-constructions. In M, Andionis, C. Ball, H. Elston, & S.Neuvel (Eds.), CLS 37: The main session. Volume 1 (67-82). Chicago: Chicago Linguistic Society.
Capelle, B.: & Dederdc, R. 2005. Spatial and temporal boundedness in English motion events.
Journal af Pragmatics, 37, 889-917, doi:io.ioi6/j.pragma.2004.io.oi2 Cooreraan, A. 1994. A functional typology of antipassives. In B. Fox, & P. I. Hopper (Eds.), Voice.
Form and function (49-88). Amsterdam & Philadelphia: John Benjamins.
doi:io.io75/tsi^7.o5coo
Croft, W. 1998, Event structure in argument linking. In M. Butt. & W. Geuder (Eds.), The projection of arguments, lexical and syntactic constraints (2 1 -63). Stanford, CA: CSLI Publications.
Croft, W. 2003. Lexical rules vs. constructions: afalse dichotomy. In H. Cuyckens.T. Berg, R. Dirven, & K-U Panther (Eds.), Motivation in language. Studies in honor of Günter Rodden (49-6S). Amsterdam & Philadelphia: John Benjamins, doi:io.io75/ciit243.07cra
Davies, M. 2008. The corpus of contemporary American English: 425 million words, 1990-present. Available online at http://corpusbyu.edu/cocay
Davtdse, K. 2011. Alternations as a heuristic to verb meaning and the semantics of the construction. In P. Guerrero (Ed.), Morpkosyntactic alternations in English. Functional and cognitive perspectives (11-37). Sheffield: Equinox Publishing.
Dowty, D.R. 1991. Thematic proto-roles and argument selection. Language, 57(3), 549-619. doi:io.i353/lan.i99i.O02i
Dowty, D. R. 2001. The semantic asymmetry of argument alternations (and why it matters). In G. van der Meer, & A. G. B. ter Meulen (Eds.) Making sense. Front lexeme to discourse 171-186. [Groninger Arbeiten zur gennanistischen Linguistik 44], Groningen: Centre for Language and Cognition.
Dixon, R. M. W 2005. A semantic approach to English grammar. Oxford Oxford University Press.
Faber, P., 8c MaiTal, R. 1999. CoHsfrucfiny a lexicon of English verbs. Berlin & New York: Moutan de Gruyter. doi: 10.1515/9783110800623
Fried, M. 2007. Constructing grammatical meaning. Isomorphism and polysemy in Czech re-fiexiviiation. Studies in Language, 31(4), 721-764, doi:io.i075/sl.3i4.02fri
Givon, T. 2001. Syntax- An introduction. Volume It. Amsterdam & Philadelphia: John Benjamins.
Goldberg, A. 1995. Constructions: A Construction Grammar approach to argument structure. Chicago: University of Chicago Press.
Goldberg, A. 2006. Constructions at work. The nature af generalization in language. Oxford: Oxford University Press.
ia8 Pilar Guerrero Medina
Goldberg, A, 2013. Constructionist approaches. In T. Hoffmann, & G. Trousdale (Eds.), The Oxford handbook of Construction Grammar (15-31). Oxford: Oxford University Press.
Goldberg, A., & Jackendoff, R. 2004. The English resultative as a family of constructions. Language, S0(3), 532-567. doi:io.i353/ian.2004-a^9
Gonzilvez, E 2008. Towards a constructionist, usage-based reappraisal of interpersonal manipulation: evidence from secondary predication in English and Spanish. Pevista Canaria de Esiudios Ingleses, 57,109-L36.
Gruber, J. 1965. Studies m lexical relations. Doctoral dissertation, MIT.
Gruber, J. 1967. Look and see. Language, 43,937-947. doi:io.2307/4ii974
Guerrero, P. 20 LI. An anttpassive Interpretation of the English "conative alternation". In P. sGuer-rero Medina (Ed.), Morphesyntactic alternations in English. Functionaland cognitive perspectives (182-203). Sheffield: Equinox Publishing.
Guerssel, M., Hale, K., Laughren, M., Levin, B., & White Eagle, J. 1985. A cross-linguistic study of transitivity alternations. In W.H. Eilfbrt, P.D. Knoeber, &K.L. Peterson (Eds.), CIS 21, Part 2: Papers on the parasession on causatives and agentivity (43-63). Chicago: Chicago Linguistic Society.
Hopper, P.J., 8c Thompson, S.A. 1980. Transitivity in grammar and discourse, Language, 56(2),
251-299. dai:io.i353/lan,i98o.ooi7 Huddleston, R. 2002. The clause: complements. In R. Huddieston, & G. K. Pullum (Eds.), The
Cambridge grammar of the English language (213-321). Cambridge: Cambridge University
Press,
Ikcgami, Y. 1985. 'ActMt/'-'acconiphsbjneQt'-'achievement'- A language that can't say 'I burnt it, but it didn't bum' and one that can. In A. Makkai, & A. K. Melby (Eds.), Linguistics and philosophy. Festschriftfor RuknS. Well (265-304). Amsterdam & Philadelphia: lohn Benjamins, cioi: io.io7s/cilt,42 .ziike
Jackendoff, R. 1996. The proper treatment of measuring out, telicity, and perhaps even quantification in English, Natural language and linguistic theory, 14,305-354. doi:io.ioo7/BFoovi3686
Langacker, R. 1987. Foundations of Cognitive Grammar, Volume I: Theoretical prerequisites. Stanford, CA: Stanford University Press.
Lemmens, M. 1998. Lexical perspectives on transitivity and ergativity. Causative constructions in English [Current issues in Linguistic Theory 166}. Amsterdam 8c Philadelphia: John Benjamins. doi:io.1075/cilt.i66
Levin, B. 1993. English verb classes and alternations. Chicago: The University of Chicago Press.
Levin, B., SrRappaport, M. 2005. Argument realisation. Cambridge: Cambridge University Press. dol:io.ioi7/CB097805ii6i0479
Mairal, R., & Faber, P. 2002. Functional Grammar and lexical templates. In R. Mairal Usdn, & M.J. Perez (Eels.), New perspectives on argument structure in Functional Grammar (39-94). Berlin & New York: Mouton de Gruyter.
Michaelis, L. 2004, Type shifting in Construction Grammar: An integrated approach to aspectual coercion. Cognitive Linguistics, 15(1), 1-67. doi.10.1515/cogU004.001
Panther, K-U 2005. The role of conceptual metonymy in meaning construction. In F. J. Ruiz de Mendoza, 8c M. S. Pena (Eds.), Cognitive linguistics. Internal dynamics and interdisciplinary interaction (355-386). Berlin & New York: Mouton de Gruyter.
Perek, F. 2014. Rethinking constructional polysemy: The case of the English conative construction. In D, Glynn, 8c J, Robinson (Eds.), Polysemy and synonymy: Corpus methods and applications in cognitive linguistics (61-85). Amsterdam Sj Philadelphia: John Benjamins.
"I
The English conative as a family of constructions 299
Perek, F. 2015. Argument stmcture in usage-based Construction Grammar. Experimental and
corpus-based perspectives [Constructional Approaches to Language 17J. Amsterdam Sc
Philadelphia: John Benjamins. Perek, E &Lemmens, M. 2010. Getting at the meaning of the English af-construction: the case of
a constructional split" CogniTextes [on-line] Volume 5/2010 Available at http://cognitextes.
revue5.org/33I. Last accessed on 1 October 2014. Pinker, S. 1989. Ltarnabiltty and cognition: The acquisition of argument structure. Cambridge,
Mass.: The MIT Press. Pustejovsky, J. 1995. The generative lexicon. Cambridge, Mass.: The MIT Press. Rappaport, M„ & Levin, B. 1998. Building verb meanings. In M. Butt, 8c W. Geuder (Eds.),
The projection of arguments: Lexical and compositional factors (97-134), Stanford: CSLI
Publications.
Ruiz de Mendoza, F. J., 8c Gonzalvez, F. 2011. Constructional Integration in the Lexical Constructional
Model. British and American Studies, 17,191-208. Tenny, C. L. 1987. Grammatkalizing aspect and affecUdness. Doctoral dissertation, MIT.
Cambridge, Mass.
Tenny, C. L. 1994. Aspectual roles and the syntax-semantics interface. Dordrecht- Muwer.
Tsunoda, T. 1985. Remarks on transitivity. Journal of Linguistics, 21(2). 385-396.
Van der Leek, F. 1996. The English conative construction: a compositional account In L. M.
Dobrin, It Singer, 8c L. McNair (Eds.). CLS 32: The main session (363-378). Chicago, IL:
Chicago Linguistic Society. Vendler, Z. 1967. Linguistics in philosophy. Ithaca, NY: Cornell University Press.