1 Continuity and Change in the Formation of Complex

2 downloads 0 Views 1MB Size Report
texts available from the Ben Yehuda Project,26 in order to see which verbs appear with the directional adverbial .... R. Ben-Shahar & G. Toury. Tel Aviv: The.
To appear in Linguistic Contact, Continuity and Change in the Genesis of Modern Hebrew edited by Edit Doron, Malka Rappaport Hovav, Yael Reshef and Moshe Taube, John Benjamins

Continuity and Change in the Formation of Complex Motion Event Descriptions in Hebrew Malka Rappaport Hovav The Hebrew University of Jerusalem

1.

Introduction

An assumption underlying much work in lexical semantics is that the basic conceptual components of event descriptions are drawn from the same inventory cross-linguistically and that the mapping to morpho-syntactic constituents is subject to constrained cross-linguistic variation. The challenge is to articulate what the basic components are and what the factors are which are responsible for this constrained cross-linguistic variation. The conceptual categories of MANNER and CHANGE have figured prominently in the literature addressing these questions. There exists a very large literature on the different ways languages encode event descriptions which include both the expression of some change and a specification of the manner in which the change is brought about. Beginning with work by Talmy (1975, 1985, 2000) on motion events, a typology of language types has been developed and a variety of theories have been put forth to account for the systematic ways in which languages differ. The basic distinction concerns whether in descriptions of motion events including a specification of manner the verb encodes the path or the manner. The former type of language has come to be called V-framed (since the path is encoded in the verb) and the later S-framed (since the path is encoded in a satellite, as will be described below). Research has uncovered particular syntactic and lexical properties which tend to cluster together in languages of both sorts and a variety of theories have been put forth to explain the constellation of properties which cluster together. A test case for theories meant to account for the documented patterns can be a language which in the course of its development underwent a change in its typological profile. AcedoMatellán and Mateu (2013), Kopecka (2013) and others have described the change in typological profile which took place with the development from Latin (an S-framed language) to the Romance languages (all V-framed). Acedo-Matellán and Mateu (2013) in particular stress that it is a cluster of properties which changed in all the descendants of Latin and that the particular theory they propose to characterize the difference between the two types of languages actually predicts this clustering properties. This paper brings some results of an attempt to look systematically at Biblical Hebrew (BH) and Modern Hebrew (MH) from this typological perspective. The interest in the (indirect) transition from BH to MH lies in the fact that during this transition, the language underwent a shift in typological profile, but in the opposite direction of that displayed in Romance. BH is shown to display V-framed behavior, whereas MH is developing S-framed behavior, though elements of V-framed behavior remain. It is thus of theoretical interest to see if the same cluster of properties which changed in one direction with the transition from Latin to Romance, changed in the opposite direction with the transition from BH to MH. Acedo-Matellán and Mateu (2013) suggest that the shift in profile from Latin to Romance was an internal development that had to do with morphological re-analysis. However, no evidence for such morphological reanalysis is apparent in Hebrew. In the current study a distinction is made between lexical and compositional factors underlying the typological split. The hypothesis put forth is that some time before the stabilization of MH, certain lexical changes having to do with the expression of direction and location in the satellite system of the language began and that the change in

1

compositional factors followed later on. It remains for future work to determine how and whether the lexical factors influenced the compositional factors. This paper is organized as follows. § 2 presents the typological profiles of what have come to be termed V-framed and S-framed languages in the literature. § 3 reviews the lexical and compositional factors which have been claimed by researchers to be at the heart of the split between the two typological profiles. § 4 scrutinizes these lexical and compositional factors in BH, and § 5 reviews these properties in MH. The conclusion is that indeed BH shows a Vframed profile while MH shows clear S-framed properties. The main lexical change appears to be the shift that certain locational satellites underwent to directional satellites. § 6 looks at more recent developments in MH involving constructions with unselected objects, constructions predicted to exist in S-framed languages under certain theories. § 7 contains some speculations about the historical development of the shift in typological profile and § 8 concludes the paper. 2.

Complex Motion Event Descriptions – A Typology

Talmy (1975, 1985) first offered an explication of the conceptual components which comprise descriptions of motion events (see also Talmy 2000). The following discussion is based on his analysis with slight adjustments in light of subsequent work by various researchers. A motion event description consists minimally of the following conceptual components1: (1)

figure, path

The path, in turn, is composed of two constituents: (2)

direction, reference object

The following examples include descriptions of motion events in English and Modern Hebrew with an indication of which syntactic constituents correspond to which conceptual components. (3) (4)

(5) (6)

John [FIGURE] went to [DIRECTION] the supermarket [REFERENCE OBJECT] dan[FIGURE] halax la[DIRECTION] -supermarket [REFERENCE OBJECT] Dan went.3MSG to.DEF-supermarket 'Dan went to the supermarket.' The jug[FIGURE] fell off [DIRECTION] the table [REFERENCE OBJECT] ha-kad[FIGURE] nafal me[DIRECTION] -ha-šulxan[REFERENCE OBJECT] DEF-jug fell. 3MSG fromDEF-table 'The jug fell from the table.'

There are verbs which express the manner in which a figure moves.2 (7) (8)

John [FIGURE] ran [MANNER] dan[FIGURE] rac[MANNER] Dan ran. 3MSG

1

There are verbs like jump, dance, and even run, which do not strictly entail a path (one can run, dance or jump in place). Sentences with these verbs, when used without a recoverable path (e.g. Marcy danced for hours), are not taken to express motion event descriptions in the strict sense, since a path is criterial for a motion event on this approach. 2 See fn. 1 on the status of sentences with these verbs. Verbs typically encode either the manner of motion or the direction of motion (see Rappaport Hovav and Levin 2010, and Levin and Rappaport Hovav 2012, 2013 for discussion).

2

(9)

(10)

'Dan ran' John[FIGURE] limped[MANNER] Dan[FIGURE] calaʕ[MANNER] Dan limped.3MSG 'Dan limped.' The jug[FIGURE] rolled[MANNER]. ha-qad[FIGURE] hitgalgel[MANNER]. DEF-jug rolled.3MSG 'The jug rolled.'

A motion event description may include all of these components - the figure, the direction, the reference object and a manner – in conjoined clauses. (11) (12)

John [FIGURE] ran [MANNER] and reached [DIRECTION] Dan[FIGURE] rac[MANNER] ve-higia[DIRECTION] Dan ran.3MSG and-reached.3MSG 'Dan ran and reached the supermarket.'

the supermarket[REFERENCE OBJECT] la-supermarket[REFERENCE OBJECT] to.DEF-supermarket

However, we can also include all these components in a single clause: (13) (14)

John [FIGURE] ran [MANNER] to [DIRECTION] the supermarket [REFERENCE OBJECT] Dan[FIGURE] rac[MANNER] la[DIRECTION] supermarket[REFERENCE OBJECT] 'Dan ran to the supermarket.'

I refer to sentences of this sort as containing COMPLEX MOTION EVENT DESCRIPTIONS, since they express in a single nuclear clause what can be expressed in two separate clauses. Languages differ in how the conceptual components are distributed across the syntactic constituents of a sentence encoding a complex motion event description. The contrast is particularly striking when we compare English and Romance languages, illustrated below by French. (15) (16)

(17) (18)

We jogged [MANNER] back [DIRECTION]3 Nous revînmes [DIRECTION] au petit trot [MANNER] we returned at.the jog 'We came back at a jog.' (French: Vinay and Darbelnet 1958:106) The bottle floated [MANNER] into [DIRECTION] the cave [REFERENCE OBJECT] La bouteille est entrée[DIRECTION] dans la caverne[REFERENCE OBJECT] en flottant[MANNER] the bottle is enter.PTC in the cave at floating 'The bottle entered the cave floating.' (French: Talmy 1985:69)

This difference between French and English has been recognized in the translation stylistics literature, where it is known by the French term chassé -croisé, or ‘crisscross’; see for example, Vinay and Darbelnet (1958) and Tesnière (1959). (19) blown

emporté

away

par le vent

(French: Cummins 1996)

3

The interpretation of a sentence with the expression of direction without a reference object necessarily involves a contextually determined reference object.

3

The contrast between English and French becomes of wider theoretical interest when two additional points are taken into consideration. First, the difference between English and French is found in a wider range of event-type descriptions. (20-22), for example, are caused(as opposed to non-caused) motion event descriptions. (20) (21)

(22)

John[CAUSE] shook[MANNER] the dice[FIGURE] onto[DIRECTION] the table[REFERENCE OBJECT]. *Jean a secoué les dès sur la table. John has shake.PTC DEF.PL dice on DEF.FSG table 'John has shaken the dice on the table.' (Jones 1996:394) Jean a jeté les dès sur la table. John has throw.PTC DEF.P dice on DEF.FSG table 'John threw the dice on the table.'

In (20) the manner component describes the action of an agent (a kind of cause), not the manner of motion of the dice. Compare this with (17) and (18) in which the verb describes the manner of motion of the figure. Though (20) expresses a caused-motion event, shake is not a manner of motion verb – since it doesn't describe the manner of motion of the figure. Nonetheless, the same difference between English and French is apparent here. The manner (of the agent causing a motion event) can be expressed in the verb in English but not in French (21). The natural rendition for (20) in French is (22), in which the manner is not expressed at all. The same constraint on the expression of manner is found in descriptions of changes of state in French. The English (23) expresses a change of state – rather than a change of location brought about in a specified manner. Here, too, the verb describes the action of the agent, rather than something about the entity undergoing the change. (23) (24)

Mary[CAUSE] sponged [MANNER] the table clean [CHANGE] Marie[CAUSE] a nettoyé [CHANGE] la table avec une éponge [MANNER] Marie have clean.PTC DEF.FSG table with a sponge 'Marie cleaned the table with a sponge.'

(French)

English and French continue to show a similar contrast: the verb expresses the change in French and the manner is expressed in a subordinate constituent, while the verb expresses the manner in English. This leads one to believe that there is some fundamental difference between English and French which does not make reference specifically to motion events, but to the more basic components of event descriptions of which motion event descriptions turn out to be a special case. In general, a sentence which expresses a change of certain kinds (change of location or state) in some manner, can in English encode the manner in the verb and the change in some other constituent. This is not possible in French: the change must be expressed in the verb and the manner, if at all expressed4, must be expressed in some kind of adjunct5. The second reason for the wider theoretical interest of these data is that English and French appear to be representatives of two language types which contrast in the way English and French contrast. (25), for example, illustrates the Catalan rendition of (18) and (26), the Spanish rendition of the same sentence. 4

As illustrated amply by Slobin (e.g., 2004) and colleagues, Romance languages include a specification of manner in motion event descriptions to a much lesser degree than languages like English. 5 Of course, as we will see, the manner verbs can be used alone in French, as in English, without the further expression of change of state or location, as in La bouteille a flotté ('The bottle floated') or Jean a couru ('John ran.') But as mentioned in footnote 1, we do not consider these to be motion events strictly speaking.

4

(25)

La botella entrà a la cova flotant the bottle entered.3FSG at DEF.FSG cave floating 'The bottled entered the cave floating.' (Catalan: Mateu and Rigau 2009: 227)

(26)

La botella entró en la cueva flotando. the bottle entered in DEF.FSG cave floating 'The bottle entered the cave floating.' (Spanish: Mateu and Rigau 2010: 255)

This leads us to seek the principles which govern this mapping and to determine the factors which underlie the typological variation in mapping. Talmy (1985, 2000) suggested classifying languages according to whether they encode direction or change (together referred to as result6) in the verb (as in French), or in what he calls a satellite (as in English), thus distinguishing between what he calls V(erb)-framed, and S(atellite)-framed languages. Despite the clumsy and somewhat opaque nature of the terminology, I continue to use it as it has become prevalent in the literature. Verb-framed languages: Romance languages, Japanese, Korean, Turkish, Hebrew7, Arabic… Satellite-framed languages: English, German, Russian, Mandarin, Latin, Finnish, Hungarian… As can be seen from these lists, genetically unrelated languages appear in both lists, and, interestingly, genetically related languages do as well. There is an asymmetry between the two types of languages. Most languages have result verbs (verbs of change of state and directed motion) and verbs of manner (though more on this later). Therefore, languages of both types can express simple activity events (with manner verbs; see fns. 1 and 6) and simple events of change (directed motion or change of state): (27)

(28)

a. John swam. b. Jean a nagé John have swim.PTC 'John swam' (French) a. John entered the room. b. Jean est entré dans la salle. John is enter. PTC in DET.FSG room. 'John entered the room.' (French)

As we have already seen (16, 18, 24 -26 above), V-framed languages can add an expression of manner to clauses encoding directed motion and change in adjunct phrases. However, Vframed languages appear to have a "deficiency" in not being able to fully express what Sframed languages can express productively: the encoding of the manner and the change in the

6

The term result used to refer to both directed motion and certain kinds of change of state reflects the idea that both these types of events have shared semantic base. The type of change we refer to here is what is sometimes called directed change (Levin and Rappaport Hovav 1995) or the more commonly accepted term scalar change (Rappaport Hovav 2008; McClure 1994; Beavers 2008). Activity verbs like laugh, write, run and wiggle also involve change (Dowty 1979), but they are not considered scalar changes. Verbs which lexically encode manner form activity event descriptions when used in isolation. 7 Talmy does not indicate whether he is classifying Biblical or Modern Hebrew.

5

nuclear components of the same clause (as in 17, 20 and 23 above) 8. Croft et al. (2010) survey the range of other strategies available to languages to express the manner component in a complex event description, including coordination. Some of these are systematically available in V-framed languages. For example, the most natural way to render "I rolled the barrel into the basement" in Japanese is: (29)

Watashi wa taru o korogashi -te chikashitsu ni ireta. I TOP barrel ACC roll -and basement to put.into.PST 'I rolled the barrel into the basement.' (Croft et. al. 2010: 217)

As stated in §1, the purpose of this paper is to consider the profile of Hebrew in two of its stages – Biblical Hebrew and Modern Hebrew – from the point of view of the Talmy-based typology. Talmy, Berman and Slobin (1994), Slobin (2004) classify Modern Hebrew as a Vframed language9. Doron and Dubnov (D&D; 2017) examine the status of BH and conclude that it shows the profile of a V-framed language. However, they suggest that Modern Hebrew shows certain characteristics of S-framedness. As D&D show, the contrast between BH and MH can be seen in certain passages from BH and their more natural MH renditions: (30)

(31)

(32)

a. way-yirkav yehu way-yelɛk yizrǝʕɛl-ɔ̄ and-rode Jehu and-went Jezreel-LOC 'Jehu rode and went to Jezreel.' (Kings II 9:16; D&D:70) b. yehu raxav le-yizraʔel Jehu rode.3MSG to-Jezreel 'Jehu rode to Jezreel.' (Modern Hebrew) a. yaḡbihu ʕup̄ ascend.3PL.IMP fly.INF 'They will fly upward.' (Job 5:7; D&D:72) b. yaʕufu lemaʕla fly.3PL.FUT upward (Modern Hebrew) 'They will fly upward.' a. way-yimšǝħu way-yaʕălu ʔet yosep̄ min hab-bor and-drew and-raised ACC Joseph from DEF-pit 'They drew and raised Joseph out of the pit.' (Genesis 37:12) b. mašxu et yosef me-ha-bor pull.3PLM ACC Josef from-DEF-pit 'They drew Joseph out of the pit.' (Modern Hebrew)

Henkin (1998) presents a more nuanced picture of MH. She takes the distinction between the two types of languages not to be a strict dichotomy (see also Beavers, Levin and Tham 2010), and suggests that languages fall on a continuum between strongly V-framed and strongly Sframed. This is because most languages have elements of both types and for her the major question is what kinds of structures languages use in unmarked contexts. She also distinguishes, as do Berman and Slobin (1994), between the grammatical profile and the rhetorical profile of languages. She suggests that in the case of complex motion event 8

Croft et. al. (2010) stress that it is more appropriate to talk about S- and V-framed constructions in particular languages rather than to classify languages as a whole; see also Beaver, Levin and Tham (2010). This may indeed be the case. The reason I continue to classify languages is that given the systematic 'deficiency' of V-framed languages, it is of interest to determine what underlying factors are responsible for this deficiency and to test hypotheses regarding the factors responsible for the deficiency in light of the historical development of Hebrew. 9 The astute reader may find this fact surprising in light of the example in (14) above, in which a manner verb combines directly with a directional phrase in Hebrew. This example is less indicative since, as shown by numerous authors, verbs like run often appear with directional phrases even in prototypical V-framed languages. See 3.2 below for some discussion.

6

descriptions, grammatically, MH is predominantly V-framed, though it has S-framed properties as well, and correspondingly, the rhetorical style of Hebrew is still predominantly that of V-framed languages, but it allows elements of the rhetorical style of S-framed languages as well when appropriate. The fact that most languages have elements of both V-framed behavior and elements of Sframed behavior has been well-known from the outset of the study of the Talmy-based typologies. Nonetheless, it remains a striking fact that languages which have been classified as V-framed consistently lack structures of a particular kind, as illustrated earlier in this section. It is then of interest to figure out what the underlying elements of grammar are which prevent V-framed languages from having certain kinds of structures. As Acedo-Matellán and Mateu (2013) have pointed out, theories which purport to account for this difference between language-types can be put to test by languages which have changed their typological profile, since the theories should be able to predict which cluster of properties will change together. Therefore, before we take a closer look at the development of Hebrew, we take a look at what scholars have taken to be the underlying factors which determine whether a language shows V- or S- framed behavior.

3.

Factors contributing to V- and S-framedness

Scholars have identified two types of factors which underlie the typological split between Sframed and V-framed languages: the first has to do with the nature of inventories of the functional and verbal lexicon. This is discussed in 3.1. The second has to do with capabilities of composition, discussed in 3.2. 3.1

Lexical factors

The first lexical factor which characterizes V-framed languages as opposed to S-framed languages is the absence or paucity of directional satellites in the former and the abundance of such satellites in the latter. French, for example, has no real directional satellites. Researchers have suggested that à as in (33) has basic locative semantics (Bouchard 1995; Cummins 1996, Song 1997, Song & Levin 1998, Fábregas 2007 and Son 2007). (33)

La

bouteille est retournée à la rive en flottant (Talmy 1985: 70) bottle is returned.PTC at DEF.FSG shore by floating 'The bottle floated back to the bank.' DEF.F

When it appears with a verb encoding direction, the location functions as the reference object for the path. The comparison between (33) and (34) is particularly instructive. The former includes a verb which encodes direction - to the direction of the reference object which appears as the complement of à - and the latter includes a verb which does not encode direction. In the former we get the sense that à has a directional meaning, but this is in fact to be attributed to the verb, not to the preposition. In (34), where the verb does not encode direction, à has a locational reading (Cummins 1996). (34)

Anne a marché à la plage (location only) Anne has walked. PTC at DEF.FSG beach 'Ann walked on the beach.'

More specifically, V-framed languages typically lack a real telic directional preposition, that is, prepositions which form phrases denoting bounded paths. They often have atelic directional phrases (which form phrases denoting unbounded paths), such as vers in French (35) or hacia in Spanish (36). 7

(35)

(36)

Ophélie a flotté vers/jusqu' à Hamlet Ophelia has float.PTC towards/up to Hamlet 'Ophelia floated towards/up to Hamlet.' (Bouchard 1995: 191) La botella flotó hasta/hacia la cueva. the bottle float.PTC until/towards DEF.FSG cave 'The bottle floated to/towards the cave.' (based on Beavers, Levin and Tham 2010: 347)

Many V-framed languages have phrases corresponding to jusqu'à in French (35) or hasta in Spanish (36). Beavers (2008) calls these 'general delimiters' which provide a static boundary point for some event participant that has physical or abstract extent. The precise form of delimitation is inferred from the nature of the event and the complement of the delimiter; when a motion predicate takes a delimiter with a ‘place’ as complement, the inference is that the complement names the endpoint of the path of motion, i.e. it is understood as the goal. But there are many other uses of these general delimiters, as in (37)-(40). (37)

cire coule jusqu’au bord de la table flow.PTC until.at.the edge of DEF.FSG table 'The wax flowed to the edge of the table.' (French – Cummins 1996) Maria a dormi jusqu’à dix heures Maria has sleep.PTC until-at ten hours 'Maria slept until ten o’clock.' (temporal delimitation) Combien de mètres y a-t-il du plancher jusqu’au how-many of meters are there from.the floor until-at.the plafond ceiling 'How many meters from the floor to the ceiling?' (static spatial numeral)(Beavers 2008:311) J’ai parlé avec mon ami jusqu’à ce que l’avion soit arrivé´ I-have speak.PTC with my friend until-at that DEF.MSG-plane is.SUJV arrive.PTC 'Until the plane arrived, I talked with my friend.' (propositional) (Beavers 2008: 311) La

DEF.FSG wax

(38)

(39)

(40)

In contrast, S-framed languages usually have articulated satellite systems distinguishing location and direction. For example, English has a telic goal marker to, as opposed to at and distinguishes between in and into, on and onto. Finnish, yet another S-framed language (Acedo-Matellán 2010), systematically distinguishes between directional and locational cases: the locational inessive and adessive and the directional elative, illative, ablative and allative. The second relevant lexical factor is the size of the manner verb vocabulary. S-framed languages typically have large manner vocabularies (Weinold 1995, Slobin 2006, Verkerk 2014). For example, English has well over 100 manner of motion verbs listed in Levin (1993). In contrast, V-framed languages have a much more restricted range of manner verbs and often encode manner by other means (e.g. adverbially or by ideophones (Weinold 1995)). I elaborate on this point further in section 4.1 below. V-framed languages typically have more direction verbs than S-framed languages, but this class tends to be relatively small to begin with because there are only so many types of directions to express in comparison to the number of manners a language may express (see Rappaport Hovav 2014 for discussion). How do these lexical factors impact on the typological profile of the language? Clearly, if a language lacks directional satellites, then when a verb encodes manner and not direction, it will not have the means to construct complex motion event descriptions with a manner verb. It is less clear why such languages should have small manner vocabularies, but see fn. 25.

8

3.2

Compositional Factors

English allows any manner of motion verb to compose with any directional PP – this is a strikingly fully productive process with no inherent limitation. English also allows a few other types of verbs to appear productively with directional phrases, most saliently sound emission verbs: (41)

a. The bullets whistled through the air. b. The truck rumbled into the driveway.

In addition to sound emission verbs, English allows a variety of other non-motion verbs to take directional complements10, as in (42)

She filed the serial number off.

This kind of structure is disallowed in French, which would have to render (42) as (43). (43)

Il a enlevé à la lime le numéro de série he has remove.PTC with DEF.FSG file DEF.MSG serial number 'He removed the serial number with a file.' (Green 1973:273)

French and Italian have some directional (typically atelic) Ps, but their combinatorial properties are much more limited. (44)

(45)

(46)

(47)

*L'eau remuait vers la rive DEF.MPL moved towards DEF.FSG bank 'The water moved towards the bank.' *L 'ivrogne trébuchait vers la bar DEF.MSG drunkard stumbled towards DEF.FSG bar 'The drunkard stumbled towards the bar.' (French: Cummins 1999:43) Gianni è corso via Gianni is run.PTC away ‘Gianni ran away’ *Gianni è ballato via Gianni is dance.PTC away ‘Gianni danced away.' (Italian: Mateu and Rigau 2010: 243)

As Alonge (1997), Folli & Ramchand (2005) and Mateu and Rigau (2010) point out for Italian, and Fábregas (2007) for Spanish, only verbs which strongly imply translational movement can appear with directional complements (see also Beavers, Levin and Tham 2010) . The verbs which typically appear in these structures are the translation equivalents of 'run' and 'fly'. This kind of constraint is not found in English and other English-type languages. How can one account for this contrast between English-type languages and French-type languages? A common suggestion is that English-type languages have a structure-building 10

Nonetheless, many non-motion verbs resist appearance with directional phrases. (i) *John laughed to the bank Almost all eventive verbs in English can appear with a directional phrase along with "(all) the way." It appears that in such cases the to-phrase is a complement of "the way," and not a complement of the verb: He cried all the way to the bank; All the way to the bank, he cried/*All the way, he cried to the bank. The way phrase seems to be an adjunct. For a suggested explanation of the distribution verbs in directed-motion constructions see Rappaport Hovav and Levin (2001).

9

operation which is lacking in French-type languages. We can understand the logic of this suggestion by looking at the basic combinatorial properties of locational and directional PPs. Locational PP's can appear freely with NPs and VPs as adjuncts; directional PPs cannot11. (48) (49)

a. the table on/under/behind the porch b. *the table to/toward the porch a. eat/sing/laugh on/under/behind the porch b. *eat/sing/laugh to/toward/along the porch

From this we can conclude that directional PPs can only appear with heads which select them – in the case of verbs, those which encode directed motion. Locational phrases, in contrast, are often adjuncts. Since manner verbs do not encode directed motion, they do not select directional PPs. This seems to be an assumption either explicitly made or implicitly held in almost all work on directed motion. Therefore, in a language which productively allows the addition of directional PPs to manner verbs, there must be a 'structure-building process' which adds a constituent which selects the directional complement. Processes of this sort are commonly expressed morphologically in languages of the world: when a manner verb appears with a telic directional complement, we often find an added constituent which selects the directional PP (or which directly encodes direction, selecting some place-denoting phrase into the reference object of a path-denoting phrase). Languages vary widely in the strategy they employ to license the directional PP. Russian (50; and other Slavic languages) uses directional perfective prefixes, as does Latin (51). Chinese uses a compound verb – a verb composed of a directed motion verb and a manner verb (52). Tswana (53) makes use of an applicative affix on the verb. In each case, the 'direction selecting' unit is italicized. (50)

Ja vy- bežal iz doma I out- ran from house.GEN 'I ran out of the house.'

(Russian; Talmy 1985:105)

(51)

Repente ex omnibus partibus ad pabulatores ad-volaverunt suddenly out all.ABL.FPL quarter(F)ABL.PL at forager.ACC.PL at-fly.PRF.3PL 'Suddenly, they flew upon the foragers from all quarters.' (Latin; Acedo Matellán and Mateu 2013: 11)

(52)

pingzi piao-jin le dongxue bottle float-enter ASP cave 'The bottle floated into the cave.'

(53)

(Chinese; Tai 2003:310)

mò-símàné ó-tábóg-èl-à kwá-gòdímò gá-thàbà CL.1-boy he-run-to-IMP DIST-top LOC-mountain 'The boy is running to the top of the mountain.' (Tswana; Schaefer 1985:66)

Despite the diversity in the morphological realization of the constituent chosen to augment the argument structure of the manner of motion verb, at an abstract level scholars have suggested that all these examples have a similar syntactic structure. Abstracting away from differences in implementation, Snyder (2012), Mateu (2010; 2012), Zubizaretta and Oh (2007) all suggest that S-framed languages have some kind of compounding (structure- building) process which allows a manner verb to be the syntactic sister of another direction/result-selecting constituent12. 11

There are some exceptions to this generalization, discussed in Fong (1997) and Beavers, Levin and Tham (2010). The important point is that directional phrases are licensed in NPs under different, and very restricted, conditions. 12 See Folli and Harley (2016) for a different approach; on their approach head movement of the root of a verb of change is mandatory in V-framed languages. The 'deficiency' of these languages turns out

10

English is in some sense an outlier from a cross-linguistic perspective since it does not mark the addition of direction complements on the verb. However, since the process is productive in English, scholars have suggested there is indeed such a complex morphological structure, where one component of the complex structure is not phonetically realized. For example, Mateu (2010) provides the following structure for a sentence such as "The bottle floated into the cave."

(54)

Mateu 2010:99

In this example, the verb float, which does not lexically encode any direction, forms a compound with an empty v node, which, in the context of a directional complement is interpreted as expressing movement. This directional complement would be uninterpretable without the addition of the v which forms a compound verb with the lexical verb. The directional phrases which appear in the examples studied so far are predicated of an argument of the verb in the main clause of the sentence. Languages which have such structure-building processes, have the additional ability to build structures in which a constituent expressing change (either change of location or change of state) is predicated of a non-selected argument. This is illustrated in (55) for Russian and in (56) and (57) for English. There is not complete consensus on the appropriate analysis of (55), but its similarity to the examples in (56) and (57) is worth pointing out. The verbs in (56, 57) are followed by a DP and a predicate which expresses the change (real or metaphorical) which the entity denoted by the DP undergoes. In the Russian example, the prefix on the verb serves the same function as the phrase which follows the post-verbal DP in the examples in (56) and (57). (55)

Ona is-pisala / *pisala svoju ručku she out-write.PST.3SG her.ACC pen.ACC 'She has made her pen to run out of ink through writing.' (Russian; Spencer and Zaretskaya 1998: 17)

(56)

a. b. c. d.

You can dance [your socks off]! You can dance [your fat off]! Let's dance [the night away] Michael Phelps swam [his way to five Olympics].

(57) The slave who wrote [his way out of slavery] http://www.theroot.com/articles/history/2012/11/100

then to be a syntactic parameter, similar to ones which control other types of head-movement, such as V-to-I or I-to-C movements.

11

Acedo-Matellán and Mateu (2013) point out that the assumption that what characterizes Sframed languages is a structure-building process makes a very strong prediction: we expect Sframed languages to have a particular cluster of constructions which are analyzed as involving such a structure-building process. They show, for example, that in the transition from Latin to the Romance languages, the Romance languages all lost a range of syntactic constructions which were available in Latin as a result of their shift to V-framed languages. In particular, they argue that one would expect unselected object constructions to go hand in hand with complex directed motion constructions. As mentioned earlier in this section, both BH and MH are classified as V-framed in much of the literature. D&D demonstrate in a detailed fashion the V-framed characteristics of BH. They point out, however, that in many ways MH appears to show properties of S-framed languages despite being classified as V-framed in much of the literature. In the §5 and 6, we will examine these claims from the perspective of the discussion in this paper13. In particular, we examine the lexical and compositional factors contributing to the typological profile reviewed in this section for both BH and MH. 4.

Biblical Hebrew

When we scrutinize BH from this typological perspective we see that in many ways it is similar to prototypical V-framed languages like French. Some of these points have been made in D&D but others are noted for the first time here. We look at BH in terms of both the lexical and the compositional factors which have figured in the discussion of other languages in this context. 4.1

Lexical factors

BH had no fully articulated distinction between location and direction. I claim that the situation in BH is similar to that in many V-framed languages in that almost all the items which can express direction have a basic stative use and receive a directional interpretation in composition with verbs which encode direction. As far as I know, this position has never been articulated in the literature on BH. The preposition lə- is typically translated as 'to'; there is good reason to assume, however, that it is basically a locational, not a directional, preposition. Jenni (2013) suggests that this preposition denotes the bare relation between two entities (‘with regard to’). It can, for example, express simple possession as well, as in mizmōr le-dawid 'a psalm of David'. Even ʔɛl, most often translated as 'to', and taken by scholars (Jenni 2013) to mean 'towards', has clear stative uses as in (58).14,15 13

D&D suggest that V-framed languages tend to have the same preposition used for the two variants of the locative alternation (what they call Locative Alternation Preposition Identity) and they tie this property to V-framedness. Though both BH and MH display this feature, the morphological exponence of a preposition is less likely to change with the change in the typological profile of a language. Acedo-Matellán and Mateu (2013) claim that a productive locative alternation is a property of S-framed languages, not V-framed languages. As D&D demonstrate, BH has a fair number of locative alternation verbs. D&D count 45, and there may be more since if a verb appears only in one frame in the Biblical Hebrew corpus, this does not mean that it did not necessarily appear in the other frame; it may just not have appeared in that frame in the corpus. On the other hand, it may be that the nature of the locative alternation in BH was different than it is in MH. In particular, BH lacks locative alternation verbs which denote surface contact and motion (D&D), while MH has a fair number of such verbs. In fact, some of the verbs which used to alternate in BH no longer alternate in MH (himṭir "rain down", śam "put", hesir " remove") and others are no longer used in MH (šillaћ "to send"). More work is needed on the nature of the alternation itself and its connection to the V- or S- framed profile of languages. 14 All Biblical translations are taken from the King James Version. 15 See also Bar-Asher Siegal (2017) who points out that the reciprocal expression ʔīš ʔɛl rēʕēhū, lit. 'man ʔɛl his friend' sometimes appear with stative verbs, as in pɔ̄ћădū ʔīš ʔɛl rēʕēhū " they were afraid

12

(58)

qɔ̄bar ʔabrɔ̄hɔ̄m ʔɛt śɔ̄rɔ̄h ʔišt-ō ʔɛl maʕăra-t śǝdē ha-makpɛlɔ̄ buried.PST.1MSG. Abraham ACC Sarah wife-POSS ʔɛl cave-CST field.CST DEFMachpela 'Abraham buried Sarah his wife in the cave of the field of Machpela.' (Gen. 23:19)

The verb natan, often translated as give, can be used in BH to mean put. In this use, it takes ʔɛl to mark the indirect object. Verbs of putting in general prefer locative, not directional phrases. we-nɔ̄tata ʔɛl ћošɛn ham-mišpaṭ ʔɛt hɔ̄-ʔurim we-ʔɛt hatand-put. PST.2MSG ʔɛl breastplate DEF-judgment ACC DEF-Urim and- ACC DEFtummim Thummim 'And thou shalt put in the breastplate of judgment the Urim and the Thummim.' (Exodus28:30)

(59)

BH also has what we have called above a general delimiter ʕad 'until' which is used both for spatial and non-spatial senses (as discussed in section 3.1 above), the latter illustrated in (60). (60)

ʕad ћăṣī ham-malkūt till half.CST DEF-kingdom 'It shall be even given thee to the half of the kingdom.' (Esther 5:3)

As argued by Beavers et al. (2010), and mentioned above, the translation equivalents of ʕad in other languages denote static delimitation, consonant with the position we hold here. BH has a number of substantives ending with -ɔ̄, which are often taken to be directional. However, they can also be found with stative uses. I suggest that they only assume directional senses with verbs which encode direction, as in the case with French à. Examples of these general direction markers are listed in (61) followed by clear stative uses of some of them16. The stative translation of these markers is set in bold. (61) (62)

(63)

(64)

(65)

(66)

ʔarṣ-ɔ̄ (earth-LOC) ; qadīm-ɔ̄ (east/front-LOC); lǝ-maʕăl-a (at-above-LOC); bayt-ɔ̄ (house-LOC); pǝnīm-ɔ̄ (inside-LOC); ћuṣ-ɔ̄ (DEF-outside-LOC) kī-migraʕot natan lab-bayit sabib ћuṣ-ɔ̄ for-rests give.PST3MSG to.DEF-house around outside-LOC 'for without in the wall of the house he made narrowed rests.' (Kings I 6:6) way-yǝṣap šǝlomoh ʔɛt hab-bayit mip-pnim-ɔ̄ zahab sagur and-overlay.IMP3MSG Solomon ACC DEF-house from-inside.loc gold pure 'Solomon overlaid the house within with pure gold.' (ibid:21) wǝ-hinnē sullam muṣṣab ʔarṣ-ɔ̄ and-behold ladder set.PSSPTC earth-LOC 'behold a ladder set up on the earth' (Genesis 28:12) bǝ-šɔ̄rtɔ̄m bǝ-šaʕărē hɛ-ћɔ̄ṣer hap-pǝnimi-t wɔ̄-bɔ̄yt-ɔ̄ in-serve.INF.3MPL in-gates.CST DEF-court DEF-inner.F and-house-LOC 'whiles they minister in the gates of the inner court, and within.' wǝ-hiqṭartɔ̄ ham-mizbeћ-ɔ̄

both one and other" (Jer.36:16), in which case he suggests that the ʔɛl marks an associative meaning, consonant with the position put forth here. 16

It should be pointed out, however, that the distribution of these elements is different from that of complements of à in French. We don't normally find them, for example, marking locative adjunct with verbs of position or verbs of living. More work is needed on the nature of these elements.

13

and make.smoke.PST.2MSG DEF-altar-LOC 'and you shall make (it) smoke on the altar.'

(Ex 29:13)

They appear with a directional meaning only with verbs which encode direction17 as we will see. We can conclude then that in BH the expression of direction was encoded in the verb, the pattern typical of V-framed languages. Furthermore, as with other V-framed languages, the language lacks an articulated array of directional satellites. Stative complements receive directional meaning from the verb; they mark a place which constitutes the reference object for the construction of the specific direction expressed in the event description. We turn now to the second lexical factor which contributes to the profile of a language as Vor S- framed. Below is a near-exhaustive set of BH manner of motion verbs. Most of them have only very few occurrences in the corpus18. (67) raṣ 'run'; rakab 'ride'; ʕap 'fly' ; nasaʕ 'journey'; ṣalaʕ 'limp'; qippēṣ' 'leap'; dillēg 'skip'; ṣaʕad 'step'; ṭapap- 'walk in circles'; šaṭ 'flow' ; zaram 'flow'; ṣap 'float'; zaħal 'creep'; daʔah 'hover'; ramaś 'creep'; riqqēd 'dance'; paśaʕ 'step'; dahar 'gallop'; gal 'roll'; riħēp 'hover'; mihēr 'hurry' pizzēz 'dance'; kirkēr 'leap'; hillēk 'walk' Some of these verbs appear to be virtual synonyms but it is difficult to determine their exact meanings, given the small number of appearances in the corpus. Slobin (1997) distinguishes between Tier 1 and Tier 2 manner of motion verbs. – Tier 1: neutral, everyday verbs: walk, fly, climb – Tier 2: more expressive verbs: amble, prance, scurry, tiptoe, trudge Cross-linguistic comparisons have revealed that typically S-framed languages have a large, elaborated inventory of Tier 2 manner of motion verbs, while V-framed languages have a small inventory of Tier 2 verbs (Slobin 1997: 459). For example, consider the following closely related English manner of motion verbs: scuttle, scurry, scamper, scramble, and clamber. These have no Hebrew counterparts. Hebrew has a single verb meaning 'to hurry' – mihēr (which, as mentioned in fn.18, is not really a verb of motion). Looking at the list of verbs in (67) above we see that BH has some verbs which can be considered Tier 2 verbs, such as ramaś 'creep', riqqēd 'dance', pizzēz 'dance' and kirkēr 'leap' but this inventory is strikingly smaller than its English counterpart. As we shall see in section 4.2 below, most of the verbs in this class never appear with directional phrases. Jackendoff (1990) analyzes these verbs as involving the basic primitive predicate MOVE, which does not select a path 17

I have not found any stative uses of the suffix when the reference object is the name of a place (miṣraym-ɔ̄ – to Egypt), or heaven (šɔ̄maym-ɔ̄ 'to heaven') or to Sheol (šǝʔol-ɔ̄ 'to Sheol'). However, there are examples such as yɔ̄šūḇū rǝšɔ̄ʕīm li-šǝʔol-ɔ̄ 'the wicked shall return to-Sheol-LOC' (Psalms 18:9) and hinnē ḵǝlē ḇēṯ JHVH mūšɔ̄ḇīm mib-bɔ̄ḇɛl-ɔ̄ 'behold the vessels of the house of the Lord are being returned from Babel-LOC' (Jer 27:16) which suggest that even place names can appear with -ɔ̄ in a stative sense. 18 The verb hɔ̄lɔ̄ḵ often translated as walk, is probably not a manner of motion verb but a general verb of motion as can be seen by the following verses: wat-tēlɛḵ hat-tēḇɔ̄ ʕal pǝnē ham-mayīm 'and the ark went (wat-tēlɛḵ) upon the face of the waters' (Gen 7:18); kol han-nǝћɔ̄līm holḵīm ʔɛl hay-yɔ̄m 'All the rivers flow (holḵīm) into the sea.' (Ecc. 1:7). The same root in the hiphil template does seem to translate as 'walk.' The verb mihēr 'to hurry' is included here (taken from the list in Doron and Dubnov 2017) but it probably does not belong, as it takes non-spatial complements as well as in way-yǝmmahēr parʔo liqro lǝ-moše 'Then Pharaoh called for Moses in haste.'

14

argument, as opposed to directed motion verbs, such as enter, go, come, arrive, advance and return, which are based on the predicate GO, which does select a path argument19. 4.2

Compositional factors:

Recall that a striking difference between English and French is that all manner of motion verbs in the former can appear in complex directed motion event descriptions, whereas in the latter only a small number of verbs can do so, and that this situation appears to be representative of S- and V-framed languages respectively. The situation in BH is in this regard similar to the situation in French. Most manner of motion verbs appear only with locational phrases headed by prepositions such as bə- 'in' and ʕal 'on'. These are probably best considered adjucts. The only manner of motion verbs which appear with ʔɛl in the sense of 'to, toward' are raṣ 'run', rakab 'ride', and ʕap 'fly,' probably the type which Romance languages allow to appear with directional particles since they so clearly imply translational movement (cf. 46 above). The substantives which I argued above (61) to have basic locative readings, receive directional meaning only with verbs which already encode direction. (68)

(69)

(70)

hašlik-e-hu ʔarṣ- ɔ̄ cast it ground-LOC 'Cast it to the ground.' (Exodus 4:3) ʔal yip-pol damm-i ʔarṣ-ɔ̄ NEG not fall.IMP3MSG blood-my the earth-LOC 'Let my blood not fall to the earth.' way-yɔ̄bō yōsēp hab-bayṯ-ɔ̄ and -come. IMP3MSG Joseph DEF-house-LOC 'And Joseph came home.'

(I Sam. 26:20) (Gen 43: 26)

Furthermore, the verbs which appear to be counterparts of Tier 2 verbs in Slobin's sense never appear with phrases interpreted directionally – though they do appear with locative adjuncts. The examples below illustrate their use with locative adjuncts and I have not found a single use of any of these verbs with directional complements. (71)

(72)

(73)

mǝqappēṣ ʕal hɛ-hɔ̄r-īm mǝḏallēg ʕal hag-gǝvɔ̄ʕ-ōṯ leap. PTCP.MSG on the-mountain-PL skip. PTCP.MSG on the-hill-PL 'leaping upon the mountains, skipping upon the hills.' (Songs 2:8) kōl rōmeś ʕal hɔ̄-ʔɔ̄rɛṣ all creep. PTCP.MSG on the-earth 'Whatsoever creepeth upon the earth' (Gen 8:19) dɔ̄wiḏ mǝḵarkēr ū-mǝpazzēz lip̄ nē JHVH david leap. PTCP.MSG and-dance. PTCP.MSG before 'David leaping and dancing before the Lord' (Samuel II 6:14)

Furthermore, as predicted by the analysis of the typological split presented in Acedo Matellán and Mateu (2013), BH has no constructions with non-subcategorized objects, nor does it have any AP resultatives of the type illustrated by English (23) above. Summarizing, this review substantiates the claim by previous researchers that BH has the profile of an S-framed language. 19

This is in fact what is predicted by Rappaport Hovav and Levin's (2010) thesis of manner/result complementarity. If a verb lexicalizes a manner, it will not encode a result; a directional complement is analyzed as a kind of lexicalized result.

15

5.

Modern Hebrew

We now look at how MH expresses directed motion along with manner from the perspective that we have been developing in this paper. 5.1

Lexical factors

MH has a more systematic distinction between location and direction than BH. Some of the expressions in (61) are still ambiguous between location and direction (such as lemaʕla 'up-loc' or 'up-dir' (74); see also Henkin 1998), but some have a clear location/direction contrast. This is the case, for example, with the contrast between bi-fnim 'inside.loc' and pnim-a 'inside.dir.'20 and ba-ħuc 'outside.loc' and ha-ħuc-a 'outside.dir.' as the examples in (75-78) illustrate. Note that MH –a corresponds to BH -ɔ̄. (74)

(75)

(76)

(77)

(78)

a. ani yošev lemaʕl-a I sit. PTCP.MSG up-LOC 'I am sitting up above/upstairs.' b. ani ʕole lemaʕl-a I ascend. PTCP.MSG up-DIR "I am going up(stairs).' šev bi-fnim/ * pnim-a sit.IMP.MSG in-inside inside-DIR 'Sit inside.' hikanes pnim-a/*bi-fnim enter. IMP.MSG inside-DIR 'Come in.' šev baħuc/ *ha- ħuc-a sit. IMP.MSG in.DEF outside DEF-outside-DIR 'Sit outside.' ce ha- ħuc -a/ *ba-ħuc exit DEF.outside-DIR in.DEF-outside 'Go outside.'

(62) above, repeated here as (79) is rendered as (80) in a recent sermon on the web: (79)

wǝ-hinnē sullam muṣṣab and-behold ladder set.PSS.PTC 'behold a ladder set up on the earth'

ʔarṣ-ɔ̄ earth-LOC (Genesis 28:12)

(80)

ha-sulam mucav ʕal ha-ʔarec DEF-ladder set.PSS.PTC on DEF-earth 'The ladder is set on the earth.' http://www.tzura.co.il/tshsd/yezira.asp?codyezira=42445&t=1

That is, -a, the MH counterpart to -ɔ̄, is now only directional and not locative. Therefore, in the modern rendition -ɔ̄ is replaced with a locative preposition. MH ha-bayt-a corresponding to BH bɔ̄yt-ɔ̄ which I analyzed as basically stative (see 61 above), has only a directional use in MH. The stative use is ba-bayit 'in the house.' 20

pnim-a still has something of a locative sense, where it means 'the inner/hidden part of something; within,' as in ba-lev pnim-a 'in the heart, within.' According to Even Shoshan (1986) dictionary, this use dates from the Hebrew of the revival period.

16

(81)

(82)

hišaʔer stay.IMP.MSG 'Stay inside.' ruc run.IMP.MSG 'Run home.'

ba-bayit/*ha-bayt-a in.DEF-house/*DEF-house-a ha-bayt-a /ba-bayit def-house-a in.DEF-house 'Run in the house.'

The result of this change is that MH does not have to rely on the verb to bring in the notion of direction; direction can be increasingly directly encoded in the satellites. In addition to the satellites ending with –a, we find that ʔɛl in MH has lost all its locative uses and is strictly directional. The modern parallel of (58) above requires a locative preposition such as be- (83). (83)

Avraham qavar et sara išt-o be-meʕara-t ha-maxpela Abraham bury.PST.3MSG ACC Sarah wife-POSS in-case.CS DEF.Machpela 'Abraham buried his wife Sarah in the cave of Machpela.'

No verb of putting takes the preposition ʔel in MH. For example in the following verse with word ʔɛl would be rendered with the preposition be- 'in' in MH. (84)

wat-tiqaħ miḵal ʔet hat-tǝrɔ̄pīm wat-tɔ̄śɛm ʔɛl ham-miṭɔ̄ and-took.3.FSG Michal ACC DEF. image and-put.3.FSG ʔɛl DEF.bed 'And Michal took an image, and laid it in the bed.' (1Sam 19:13)

The set of MH motion verbs has grown a bit. Some rarely used non-motion verbs from BH have been re-interpreted as manner of motion verbs. For example śaħa which seems to have meant 'to be immersed', today means 'to swim.' The verb zaħal 'creep' which is used solely for creeping animals now has a more general use, and can also describe the crawling of babies. tipes 'climb', from Rabbinical Hebrew, is a common manner of motion verb today. So, the size of the manner lexicon has grown, but not significantly. One common strategy for coining new manner verbs in English is denominal derivation. For example, English has a number of denominal manner of motion verbs such as skate, bicycle, paddle, pedal, waltz and tango. Denominal verbal derivation is much more restricted in Modern Hebrew, although there are occasional coinages of denominal verbal derivation such as ledaveš, 'to pedal', from davša 'pedal, n.' Another modern coinage is lezagzeg, 'to zigzag', though it is used more often in a metaphorical sense of acting inconsistently (though see (89) below). Summarizing this section, it appears that the significant change in MH has less to do with the size of the manner verb inventory and more with the growing systematic differentiation between location and direction satellites. 5.2 Compositional factors As mentioned above, most manner of motion verbs in BH did not appear with directional satellites directly. BH formed complex motion event descriptions via conjunction (30 and 32 above), though even this pattern is not particularly frequent. Mishnaic Hebrew21 developed a new strategy for forming complex motion event descriptions with the introduction of a productive process of derived nominal formation (Bar Asher 2015). These derived nominals are often used as adverbials when appearing as objects of the preposition be- "in". In the

21

Mishnaic Hebrew is the Hebrew spoken and written around the time of the destruction of the second temple in 70 CE and for a few centuries thereafter.

17

examples below, the derived nominal combines with a direction motion verb to build directed manner of motion event descriptions. (85)

yikanes lǝ- ḇet kneset bǝ-rica enter.IMPFV to synagogue in-running 'One should not enter the synagogue running.' (Yalqut Yosef Tefilla) baʔu me ha-gǝšam-im šam bǝ-zħila came water.CS DEF-rain-PL there in-creeping 'The rain waters came there creeping (slowly).' (Iggrot Moshe, Yore De'a) lo

NEG

(86)

This structure was not available in BH, since BH did not have derived nominals of this sort. It is worth noting, however, that the structure is still a V-framed structure, in that it is the verb which encodes the direction and the manner is relegated to an adverbial. This structure still exists in Modern Hebrew. However, MH has begun to show many signs of S-framedness. It is striking that today manner of motion verbs can appear with directional phrases. This is true even for the semantically heavier Tier 2 verbs: (87)

(88)

(89)

ke - še barux calaʕ ʔel tox ha-klinika when that Baruch limp.PST3MSG to inside DEF-clinic 'When Baruch limped into the clinic.' https://craniobalanceheb.wordpress.com/tag/%D7%9B%D7%90%D7%91%D7%92%D7%91/ hu zaħal pnim-a ʕim ha-panas šelo he crawl.PST3MSG inside-DIR with DEF-flashlight POSS.3SG 'He crawled inside with his flashlight.' http://www.mako.co.il/news-israel/local/Article-3504510f89ef141004.htm hu zigzez haħuc-a mi-simtaʔot meʔa šeʕarim he zizag. PST3MSG outside-DIR from-alleys.CSTR Mea Shearim 'He (a taxi driver) zigzagged out of the narrow streets of Mea Shearim.' www.maveze.co.il/mi-panuy-beme'a-shearim

The two kinds of structures – V-framed with the derived nominal (as in 83 and 84) and Sframed with directional satellites (as in 87- 89) – appear to be living side by side today in MH. However, the direction in which MH is developing can perhaps be seen from the following graph taken from Google N-gram viewer which compare zaħal haħuca – "crawled out" (S-framed), with yaca bizħila – 'exited crawling' (V-framed):

18

zaħal haħuca (S-framed)

yaca bizħila V-framed

Figure 1 From this we see that the S-framed structure does not appear before the 1970's 22, and though the S-framed and the V-framed structures live side by side, there is a growing preference for S-framed structures. It should be pointed out that while there is some disagreement about the telicity of sentences with PP's headed by ʔel 'to/toward' (Acedo-Matellán 2010: 246), the directional phrases such as ha-bayt-a 'home-DIR', ha-ħuc-a 'outside-DIR' form unambiguously telic sentences. These structures are then clear S-framed structures. The following is a report of a cat which crawled out of a washing machine after being washed in the machine. (90)

hu pašut zaħal ha-ħuc-a ba-rega še-ha-delet he simply crawl.PST.3MSG outside-DIR in-instant that-DEF-door nifteħa open.MID.PST.3FSG 'He simply crawled out the minute that the door opened.' https://news.walla.co.il/item/597705

Summarizing this section, we see that, in contrast to BH, MH has complex motion event descriptions in which the verb encodes the manner and the satellite in the same nuclear clause encodes the direction. This is the hallmark of S-framedness. While V-framed structures and S-framed structures appear in MH, there is some indication that the S-framed structures are gaining currency at the expense of the V-framed structures.

6.

Further S-framed properties in MH

In section 3.2, we saw that many researchers suggest that the difference between clear Vframed and clear S-framed languages is that that latter have a structure-building process lacking from the former, though the morphological realization of this structure-building process varies widely between languages. As Acedo-Matellán and Mateu (2013) point out, there is a range of construction types which one would expect to find in language which have at their disposal such a structure-building process. For example, one expects directional phrases with a wider range of non-motion verbs, illustrated for English in (42) above. 22

Note, however, that in section 7 below, I report the appearance of zaħal ha-ħuc-a "crawl out-DIR" already in the late 19th century. It is probably the case that such combinations of Tier 2 manner verbs with directional ha-ħuca appeared only sporadically and it is only more recently that it is gaining currency.

19

As expected, in MH we find directional phrases with other, non-motion, verbs which do not encode direction. All of the following verbs have been documented, for example, with haħuc-a 'outside-dir': (91)

baʕat 'kick'; niʕer 'shake'; šifšef 'rub'; sinen 'filter'; tite 'sweep'; pirsem 'advertise'; šider 'broadcast'; geref 'rake'; caʕak 'yell'; timren 'maneuver'

Some of these verbs may be thought to strongly imply translational movement, but others like caʕak ‘yell’; timren 'maneuver' and pirsem 'advertise', do not. The following is an example with the verb le-naʕer 'to shake,' with a directional phrase, followed by an example with letamren 'to maneuver' with a directional phrase. le-naʕer ha-ħuc-a tipot mayim me-ha-ʔozniyot to-shake.INF DEF-outside-LOC. drops water from-DEF-earphones 'Shake the water drops out of the earphones.' https://megasport.co.il/mwhqim-vpnai/mvcri-pnai/audio-products-70/headphones/ngn-whiihmp3-dgm-1015.html (92)

(93)

…ve-ex le-tamren qadima ve-axora im ha-hege and-how to-maneuver.INF forward and-back with def-steering-wheel '…and how to maneuver back and forth [on a boat- MRH] with the steering wheel.' https://www.haaretz.co.il/magazine/sayed/.premium-1.4321010

We saw further in section 3.2. that S-framed languages typically have the option of adding directional phrases predicated of non-subcategorized objects, as in the examples (55) – (57) above. This is a natural development in the case of a language which has a structure-building process which licenses the addition of non-selected constituents. And indeed, MH does seem to be developing directional phrases with non-subcategorized objects, as expected if it has a structure-building operation. Rappaport Hovav (2015) describes a MH construction corresponding to the English way-construction, which takes non-subcategorized objects, like its English counterpart. (94)

intel roxešet ʔet dark-a ʔel dor he-ʕatid šel yicur Intel acquiring ACC way-POSS to generation.CS DEF-future of production.CS ha-švavim DEF.chips 'Intel is acquiring its way to the new generation of chip production.' (Rappaport Hovav 2105: 334)

Furthermore, MH has developed a productive constructional idiom23 based on Rabbinic Hebrew ʔibbed ɛt ʕacmo la-daʔat 'he committed suicide', lit. 'he obliterated himself to-mind', or 'he killed himself knowingly.' The MH constructional idiom allows any verb to appear and the PP is re-interpreted as a resultative phrase meaning roughly 'to one's detriment.' What is interesting is that in this resultative constructional idiom the verb can be intransitive, in which case the reflexive is interpreted as a non-selected object (as in 95), corresponding to nonselected reflexives objects in English non-selected resultatives (96). (95)

bogi yaʕalon mefatpet

ʔet

23

ʕacmo

la-daʕat

A constructional idiom is a construction with fixed constituents and an open position, in which the fixed constituents contribute a meaning which is not fully compositional. See Goldberg (1995), Jackendoff (1997; 2002), and Espinal & Mateu (2010), among others for elaboration on the notion of constructional idioms.

20

(96)

bogi ya'alon blabber.PTC.3MSG ACC self.3MSG to-mind 'Bogi Ya'alon blabbers himself senseless.' (Rappaport Hovav 2015: 326) She drank herself senseless.

In this constructional idiom the object (subcategorized or not) is typically fixed as a reflexive pronoun. However, this structure has become more productive. The following sentence appears in an article by the Minister of Justice, Ayelet Shaqed: (97)

ha-kneset mevaqešet le-ħoqeq et ħaye- nu la-daʕat DEF-knesset seek.PTC.FSG to-legislate.INF ACC lives-POSS.1PL to-mind 'The Knesset (parliament) wants to legislate our lives to death.' (Hashiloach 1, 2016: 37).

In this example the post-verbal NP is non-subcategorized and is not the fixed reflexive from the traditional form of the constructional idiom. Moreover, there are sporadic examples of non-subcategorized objects in resultative constructions, something which Hebrew has never manifested. These examples have an innovative flavor to them but they are produced on the fly and do not draw much attention. The following is taken from the popular political satire TV program Gav Ha'uama (The Back of the Nation): (98)

la-ħtox et ha-taʕamula ha-ħuc-a to-cut.INF ACC DEF-propaganda DEF-outside-DIR 'to cut the propaganda out'

In another installment of this TV program, one of the participants feigned anger at a character from a joke she told. The MC drew her attention to the fact that she was expressing anger at a character that she herself created. When the participant said that she really wants to express her anger, the MC told her24: (99)

ba-paʕam ha-baʔa tixtevi et acm-ex pnim-a le-tox-a in-time DEF-coming write.2FS ACC self-2FS inside-DIR to-inside.POSS.3FS 'Next time write yourself into it (the joke)!' (Gav Hauma 22/2/2015)

In these examples the adverbial ha-ħuc-a and pnim-a function in a way which is similar to the English particles out and in. Acedo-Matellán and Mateu (2013) point to two other properties which they suggest cluster with those already mentioned. First, they claim that S-framed languages have productive locative alternations, whereas V-framed languages do not. I leave aside this alternation in this paper, as BH did indeed show a rather productive locative alternation, but it apparently had different properties from those in typical S-framed languages, as D&D have shown; see fn. 13. A more extensive comparison between the locative alternation in BH and MH is still needed. Second, they claim, following Talmy (1985), that S-framed languages allow aspectual satellites, or, more specifically, the directional satellites take on aspectual meanings. In Latin, these were aspectual prefixes, similar to aspectual prefixes in German and Russian, and in English these are particles. It is striking that MH directional phrases do not show, as far as I am aware, any tendency to express aspectual notions. It remains to be seen why it is that MH shows exactly the S-framed properties that it does, while not displaying others.

24

It is striking that this participant responded: 'I cannot write myself!', apparently giving expression to the fact that the reflexive is non-selected and that this in indeed an innovation.

21

7.

Historical development

The changes discussed in this paper in the (indirect) transition from BH to MH is of theoretical importance since it illustrates a change in the typological profile of a language which is the mirror image of the change which Latin underwent in the transition to its daughter Romance languages. Significantly, the preliminary results of the study here indicate that MH is developing many of the constructions which the Romance languages lost as they shifted away from S-framedness to V-framedness, as documented in Acedo-Matellán and Mateu (2013). These facts lend support to an account of the difference between V- and Sframed languages which does not make do with differences in lexical inventory, but rather points to a syntactic difference between the two kinds of languages25. It is of course of interest to determine what led to the shift in profile in the two cases. Acedo-Matellán and Mateu (2013) attribute the change in the case of the transition between Latin and the Romance languages to a semantic bleaching of the aspectual/directional prefixes of Latin. In the case of Hebrew, there does not seem to be any significant morphological re-analysis of the verbs in question. While Latin lost the ability to express direction outside the verb, Hebrew gained this ability, apparently with the transition of what used to be strictly locative satellites to directional satellites, as documented in § 5. It remains to be determined when this shift took place. A reasonable hypothesis is that the Hebrew of writers and subsequent speakers in the late 19th and early 20th centuries, many of whom spoke S-framed languages such as Slavic languages like Russian and Ukranian, and German and Yiddish, was influenced by their native languages. In particular, it may be the case that the use of directional prefixes and particles in these languages aided in the reinterpretation of locative particles as directional. Perhaps, in addtion, the current close contact between Hebrew and English, another S-framed language, is strenghthening this shift. While a more systematic study of the historical development is the topic for a separate paper, there are few comments worth making at this point. First, I conducted an informal search of texts available from the Ben Yehuda Project,26 in order to see which verbs appear with the directional adverbial ha-ħuc-a – 'outside-DIR' during the early period of MH. Already in the late 19th century, one finds a wide variety of verbs with ha-ħuc-a. These include the verbs listed in (100), in addition to the expected examples of li-vroaħ 'to flee', la-nus 'to flee', la-cet 'to go out' le-hoci – to take out.' (100) lizħol 'to crawl', lisħov 'to drag', lidħof 'to push', lehitparec 'to burst forth', lištof 'to wash', lehabit 'to look', lišmot 'to drop', lehitgalgel 'to roll', livlot 'to protrude', likpoc 'to jump', lehiyot munaħ 'to be placed.' Notice that two of these predicates are stative (livlot 'to protrude' and lehiyot munaħ 'to be placed), so that the directionality must be coming from the satellite. Other of these verbs – lizħol 'to crawl', lisħov 'to drag', lidħof 'to push', lišmot 'to drop' and lištof 'to wash' - are not the kind that one would necessarily expect to appear with a directional complement in a Vframed language. The combination of these verbs with ha-ħuc-a in the early stages of Modern Hebrew seems to point to a possible influence of contact languages such as Russian 25

A variety of scholars have suggested that there is a connection between aspects of the lexical inventory and the existence of the kind of compositional process discussed here. This is beyond the scope of the present paper, but see Kageyama (2003), Lepetit and Schosler (2009), Beavers, Levin and Tham (2010), Verkerk (2013) for suggestive ideas. 26 The Ben-Yehuda Project, launched in 1999, is an open virtual library of cultural heritage in Hebrew. It provides open access over the internet to a wide range of literary creations in the Hebrew language, including poetry, essays, reference works and translations. The online corpus of the BYP constantly grows as new works are transcribed. Due to its open access policy, the BYP includes only content that is old enough to be in the public domain (or for which specific distribution rights have been granted). I thank Aynat Rubinstein for help in the search.

22

on Early Modern Hebrew, though certainly more careful historical documentation needs to be carried out. On the other hand, the constructions with unselected objects are not at all evident until fairly recently. Rappaport Hovav (2015) reports that the resultative constructional idiom V NP ladaʕat discussed in § 6 does not appear in any sources she was able to uncover before 1980 and that the construction becomes common only after circa 2000. The Hebrew counterpart to the English X's way constuction is even newer. Borochovksy and Sovran (2003) claim that the English construction has no Hebrew counterpart, though, this construction is quite common today. Speculating then somewhat, it may be that the lexical change of the reinterpretation of locative satellites as directional satellites was the initial stage of the change. This then perhaps, brought about the option of a structure-building process, which later laid the foundation for allowing constructions with nonselected objects. This second stage in the development seems to have come about as a result of contact with English. Of course, a more careful documentation of the diachronic aspect of these developments needs to be undertaken in order to substantiate these hypotheses. 8.

Conclusion

In this paper I have scrutinized the lexical and compositional factors which have been claimed to be at the basis of the distinction between S- and V-framed languages in Biblical Hebrew and in Modern Hebrew. BH, as claimed by previous researchers, has the profile of a Vframed language. I claim that satellites which have been traditionally analyzed as directional are in fact locational, receiving their directional intepretation only in the context of verbs which encode direction. I show that in MH, some of these expressions now have a directional meaning. These directional phrases now combine with a wider variety of manner verbs, thus allowing MH to have complex motion event descriptions of the type found in S-framed languages. Analyses of complex motion event descritions of this sort involve a structurebuilding process available to S-framed languages. These analyses predict that S-framed langauges should also have a range of unselected object constructions. I have shown that in the last few decades two constructions involving unselected objects have indeed entered MH. *Acknowledgments Some of the research for this article was done during my stay at the Mandel Scholion Interdisciplinary Research Center in the Humanities and Jewish Studies, whose support I gratefully acknowledge. Thanks to the members of the Research Group on the Emergence of Modern Hebrew – especially to Edit Doron and Aynat Rubinstein - for helpful discussion. Thanks also to Elitzur Bar-Asher Siegal for insight and help. I am greatly indebted to Beth Levin for reading and providing much helpful feedback on an earlier version. Many thanks to the two reviewers of this paper for comments which led to significant improvements.

References: Acedo-Matellán 2010. Argument Structure and the Syntax-Lexicon Interface : A case study in Latin and other languages. PhD thesis, Universitat de Barcelona, 2010. Acedo-Matellán, Victor, and Jaume Mateu. 2013. Satellite-framed Latin vs. Verb-framed Romance: A Syntactic Approach. Probus 25, 227-265. Bar-Asher, Moshe. 2015. Morphology of Mishnaic Hebrew. Jerusalem: Bialik Institute & the Academy of the Hebrew Language. [in Hebrew] Bar-Asher Siegal, Elitzur A. 2017. Diachronic Syntactic Studies in Hebrew Pronominal Reciprocal Constructions”, in Diachrony in Biblical Hebrew ,edited by Cynthia Miller and Ziony Zevit, Winona Lake: Eisenbrauns. 209-244. Beavers, John. 2008. On the nature of goal marking and delimitation: Evidence from 23

Japanese. Journal of Linguistics 44, 283–316. Beavers, John, Beth Levin, and Shiao Wei Tham. 2010. The Typology of Motion Events Revisited. Journal of Linguistics 46, 331-377. Berman, Ruth A. and Dan I. Slobin. 1994. Relating Events in Narrative: A Crosslinguistic Developmental Study. Hillsdale, NJ: Lawrence Erlbaum Associates. Borochovsky, Esther and Tamar Sovran. 2003. Hebrew Construction Grammar. In Hebrew—A Living Language III, eds. R. Ben-Shahar & G. Toury. Tel Aviv: The Porter Institute, Tel Aviv University, and Hakibbutz Hame'uxad, 31–50 (in Hebrew). Bouchard, Denis. 1995. The Semantics of Syntax. Chicago: Chicago University Press. Cummis, Sarah. 1996. Movement and direction in French and English. Toronto Working Papers in Linguistics, 15, 31–54 Croft, William, J. Barðdal, W. Hollmann, V. Sotirova, and C. Taoka .2010. Revising Talmy’s Typological Classification of Complex Events”, in Hans Boas, ed., Contrastive Construction Grammar,. Amsterdam: John Benjamins, 201-235. Doron, Edit and Keren Dubnov. 2017. The locative alternation in Biblical (and Modern) Hebrew. In A. Moshavi and T. Notarius (eds.) Studies in Biblical Hebrew Linguistics.Winona Lake, IN: Eisenbrauns. 321-360. Dowty, David. 1979. Word Meaning and Montague Grammar. Dordrecht: Kluwer. Espinal, Maria Theresa and Jaume Mateu. 2010. On Classes of Idioms and Their Interpretation. Journal of Pragmatics 42: 1397–1411. Even Shoshan, Avraham.1986. Hamilon Hehadash. Jerusalem: Kiryat Sefer. Fábregas, Antonio. 2007. The Exhaustive Lexicalisation Principle. Nordlyd: Tromsø Working Papers in Linguistics 34:2, 165-199. Folli, Raffaella and Gilliam Ramchand. 2005. Prepositions and results in Italian and English: an analysis from event decomposition. In Henk Verkuyl, Henrie¨tte de Swart & Angeliek van Hout,eds., Perspectives on aspect. Dordrecht: Kluwer.81–105. Folli, Raffaella and Heidi Harley. 2016. Against deficiency-based typologies. Manner-alternation parameters in Italian and English'', in E. Carrilho, A. Fieis, M. Lobo, and S. Perei, eds., Romance Languages and Linguistic Theory 10: Selected papers from `Going Romance' 28, John Benjamins, Amsterdam, 103-120. Fong, Vivienne. 1997. The order of things: What directional locatives denote. Ph.D. dissertation, Stanford University Goldberg, Adele E. 1995. Constructions: A Construction Grammar Approach to Argument Structure. Chicago: University of Chicago Press. Green, Georgia. 1973. A syntactic syncretism in English and French. In Braj B. Kachru, Robert B. Lees, Yakov Malkiel, Angelina Pietrangeli and Sol Saporta, eds. Issues in linguistics. Urbana, IL: University of Illinois Press.257–278. Henkin, Roni. 1998. Motion in Hebrew: issues in the typology and rhetoric of motion expressions. Lešonenu 61.4. The Academy for the Hebrew Language. [in Hebrew]. Jackendoff, Ray. 1990. Semantic Structures. Cambridge, MA: MIT Press. Jackendoff, Ray. 1997. Twisting the Night Away. Language 73: 534–559. Jackendoff, Ray. 2002. Foundations of Language: Brain, Meaning, Grammar, Evolution. Oxford: Oxford University Press Jenni, Ernst. 2013. Preposition: Biblical Hebrew. Encyclopedia of Hebrew Language and Linguistics, Brill. Jones, M.A. 1996. Foundations of French Syntax. Cambridge: Cambridge University Press. Kayegama, T. 2003. Why English Motion Verbs Are Special. Korean Journal of English Language and Linguistics 3. The Korean Association for the Study of English Language and Linguistics, Seoul, 341-373. Kopecka, A. 2013. Describing motion events in Old and Modern French: Discourse effects of a typological change. In J. Goschler and A. Stefanowitsch, eds., Variation and Change in the Encoding of Motion Events, John Benjamins, Amsterdam, 163-184. 24

Lepetit, X. and L. Schosler. 2009. Verbs of movement in a diachronic and typological perspective. In P. Bernardini, V. Egerland, and J. Granfeldt, eds. Mélanges plurilingues offerts à Suzanne Schlyter à l'occasion de son 65 ème anniversaire. Lund: Lunds Universitet, Sweden, 269-280. Levin, Beth. 1993. English Verb Classes and Altrnations. Chicago: Chicago University Press. Levin, Beth and Malka Rappaport Hovav. 1995. Unaccusativity. Cambridge, MA: MIT Press. Levin, Beth, and Malka Rappaport Hovav. 2012. Lexicalized Meaning and Manner/Result Complementarity. In Berit Gehrke and Boban Arsenijevic Studies in Composition and Decomposition of Event Predicates. Amsterdam: John Benjamins. Levin, Beth. and Malka Rappaport Hovav. 2013. Manner and Result: A View from clean. Language Description Informed by Theory, in D. Guillemin, R. Pensalfini, and M. Turpin, eds., John Benjamins, Amsterdam. Levin, Beth. and Malka Rappaport Hovav. in press. Lexicalization Patterns, in Robert Truswell, ed., Oxford Handbook of Event Structure. Oxford: Oxford University Press. Masini, Francesca. 2005. Multi-word expressions between syntax and the lexicon: The case of Italian verb particle constructions. SKY Journal of Linguistics 18. 145–173. Mateu, Jaume. 2010. On the l-Syntax of Manner and Causation. In M. Duguine, S. Huidobro,N. Madariaga, eds., Argument Structure and Syntactic Relations: A crosslinguistic perspective. Amsterdam: John Benjamins. 89-112. Mateu, Jaume. 2012. Conflation and Incorporation Processes in Resultative Constructions. In Violete Demonte and Louise McNally, eds., Telicity, Change, and State: A CrossCategorial View of Event Structure. Oxford: Oxford University Press.252-278. Mateu, Jaume and Gemma Rigau. 2009. Romance Paths as Cognate Complements. Dins P. Masullo Erin, O'Rourke & Chia-Hui Huang, eds. Romance Linguistics: Structures, Interfaces, and Microparametric Variation. 227-242. Amsterdam: John Benjamins. Mateu, Jaume and Gemma Rigau. 2010. Verb-particle constructions in Romance: a lexicalsyntactic account. Probus 22, 241-269. McClure, William. 1994. Syntactic Projections of the Semantics of Aspect, Doctoral dissertation, Cornell University, Ithaca, NY. McIntyre, Andrew. 2004. Event Paths, Conflation, Argument Structure, and VP Shells. Linguistics 42, 523-571. Rappaport Hovav, Malka. 2008. Lexicalized Meaning and the Internal Temporal Structure of Events. In Susan Rothstein, ed. Theoretical and Crosslinguistic Approaches to Aspect. Amsterdam: John Benjamins. Rappaport Hovav, Malka. 2014. Building Scalar Changes. In The Roots of Syntax and the Syntax of Roots, Hagit Borer, Artemis Alexiadou and Florian Schaeffer, eds. Oxford: Oxford University Press. Rappaport Hovav, Malka. 2015. A Constructional Idiom in Modern Hebrew: The Influence of English on a Native Hebrew Collocation. Journal of Jewish Languages 3. 325–336. Rappaport Hovav, Malka and Beth Levin. 2001. An Event Structure Account of English Resultatives," Language 77:4. Rappaport Hovav, Malka and Beth Levin. 2010. Reflections on the complementarity of manner and result, in Malka Rappaport Hovav, Edit Doron and Ivy Sichel, eds. Syntax, Lexicon and Event Structure. Oxford: Oxford University Press Schaefer, Ronald. 1985. Motion in Tswana and its characteristic lexicalization. Studies in African Linguistics 16, 57–87. Slobin, Dan. 1997. Mind, code, and text. In Joan Bybee, John Haiman and Sandra A. Thompson eds., Essays on language function and language type, 437–467. Amsterdam: John Benjamins. Slobin, Dan. 2000. Verbalized events: a dynamic approach to linguistic relativity and determinism”, in S. Niemeier and R. Dirven, eds., Evidence for Linguistic Relativity, 25

John Benjamins, Amsterdam, 107-138. Slobin, Dan. 2004. The Many Ways to Search for a Frog: Linguistic Typology and the Expression of Motion Events. In S. Strömqvist and L. Verhoeven, eds., Relating Events in Narrative 2: Typological and Contextual Perspectives. Mahwah, NJ: Lawrence Erlbaum, 219-257. Sobin, Dan. 2006. What Makes Manner of Motion Salient? Explorations in Linguistic Typology, Discourse, and Cognition'', in M. Hickmann and S. Robert, eds., Space in Languages: Linguistic Systems and Cognitive Categories. Amsterdam: John Benjamins. 59-81. Snyder, William. 2012. Parameter Theory and Motion Predicates. in Violete Demonte and Louise McNally, eds., Telicity, Change, and State: A Cross-Categorial View of Event Structure, Oxford University Press, Oxford, 279-299. Son, Minjeong. 2007. Directionality and resultativity: The cross-linguistic correlation revisited. Nordlyd: Tromsø Working Papers in Linguistics 34.2, 126–164. Tromsø: University of Tromsø. Song, Grace. 1997. Cross-linguistic differences in the expression of motion events and their implications for second language acquisition. Ph.D. thesis, Northwestern University. Song, Grace and Beth Levin. 1998. A compositional approach to cross-linguistic differences in motion expressions. Presented at the 72nd Annual Meeting of the Linguistic Society of America, New York. Spencer, A. and M. Zaretskaya. 1998. Verb Prefixation in Russian as Lexical Subordination, Linguistics 36, 1- 39. Tai, J.H.-Y. 2003. Cognitive Relativism: Resultative Construction in Chinese. Language and Linguistics 4:2, 301-316. Talmy, Leonard. 1975. Semantics and Syntax of Motion. In J.P. Kimball, ed., Syntax and Semantics 4. New York: Academic Press. 181-238. Talmy, Leonard. 1985. Lexicalization patterns: Semantic structure in lexical forms. In Timothy Shopen (ed.), Language typology and syntactic description, vol. 3: Grammatical categories and the lexicon, 57–149. New York: Cambridge University Press. Talmy, Leonard. 2000. Toward a Cognitive Semantics, vol. 2: Typology and process in concept structuring. Cambridge, MA: MIT Press. Tesnière, L. 1966. Éléments de syntaxe structurale, Librairie C. Klincksieck, Paris. Verkerk, Annemarie. 2013. Scramble, Scurry and Dash: The Correlation between Motion Event Encoding and Manner Verb Lexicon Size in Indo-European. Language Dynamics and Change 3. 169-217. Verkerk, Annemarie 2014. Diachronic change in Indo-European motion event encoding. Journal of Historical Linguistics 4, 40-83. Vinay, Jean-Paul and Jean Darbelnet. 1958. Stylistique comparée du français et de l’anglais. Didier. Weinold, Götz. 1995. Lexical and conceptual structures in expressions for movement and space: with reference to Japanese, Korean, Thai, and Indonesian as compared to English and German. In Urs Egli, Peter E. Pause, Cristoph Schwarze, Arnim von Stechow & Götz (eds.), Lexical knowledge in the organization of language, 301– 340. Amsterdam: John Benjamins. Zubizarreta, Maria Louisa and Eunjeong Oh. 2007. On the Syntactic Composition of Manner and Motion. Cambridge, MA: MIT Press.

26

Suggest Documents