Version classiqueVersion mobile

La Prédication existentielle dans les langues naturelles : valeurs et repérages, structures et modalités

 | 
Tatiana Bottineau

The existential construction and its type‑shifting instances in colloquial Modern Hebrew

Rivka Halevy

Résumé

Le chapitre s’attache à souligner les propriétés importantes des expressions existentielles couramment employées en Hébreu parlé. Leur structure morphosyntaxique spécifique ne se réduit pas à celle d’une proposition rendant l’existence, telle qu’elle cette notion est communément comprise. L’argument essentiel réside dans le fait que l’Hébreu est une langue au système verbal synthétique flexionnel et que le sujet est toujours indiqué par la forme du verbe, ce qui rend inutile l’encodage d’une explétive et se répercute sur l’analyse de plus longue portée de l’alignement atypique de la construction en question. Plus important encore est le fait que soit asserté l’objet qui indique le comportement et les propriétés de l’entité dont l’existence (ou l’occurrence) est dénotée par le NP. L’étude proposée réfute l’opinion selon laquelle ce seul NP thématique soit un sujet revu comme objet et offre une nouvelle analyse syntaxique et pragmatique de la construction prise dans son ensemble, tout en fournissant une explication de la motivation de son alignement atypique.

Texte intégral

1. Introduction

  • 1 E.g., McNally, 2011, a.o.
  • 2 Cf. Croft, 2001; Fillmore, 1988 ; Goldberg, 1995, a.o.

1The term “existential sentence” is often used by scholars1 to refer to a specialized construction that expresses a proposition about the existence or presence of someone or something. Pursuing Construction Grammar,2 the existential construction (henceforth ext) is viewed as a cognitive form‑meaning‑function unit that is not restricted to a bare assertion of existence, but rather refers to a specialized morphosyntactic construction built around a monovalent predicator, whereby a discrepancy between the grammatical form and the logical content is grammaticalized.

  • 3 Sasse, 1987, p. 568
  • 4 On the distinction “thetic” vs. “categorical” see Lambrecht, 1987; Sasse, 1987, a.o.

2The linguistic and philosophical literature subsumes ext sentences under the category traditionally known as thetic. Sentences in the thetic format are event‑centric, in the sense that the utterance expressed reports an event or state of affairs which is viewed from without, as an undifferentiated unit. To quote Sasse3, “the thetic type of statement is used whenever the speaker assumes that the hearer expects unitary information to be given about the whole situation in question.”4

  • 5 Williams, 1984; Francez, 2007, 2009, a.o.
  • 6 But see Williams (Williams, 1984, p. 135) who argues that, e.g., there are problems involves a nor (...)
  • 7 See Goldenberg, 2006, p. 335. From a different perspective, Francez (2007‑2009) argues that in sen (...)
  • 8 In some respects, comparable to phraseological expressions, in which the informational value of th (...)

3The hallmark of a sentence in the “thetic‑format” is the absence of the basic relation between subject and predicate. In our case, the entity or the individual about which there is an assertion of existence, presence or occurrence (hereafter e‑np)—defined by some scholars as the pivot5—is not in the default position of a topic. It means that the e‑np, which is the only obligatory theme in the construction (in the sense that the sentence would not be comprehensive without it), does not play the role of a “predication base” (hypokeimenon, subiectum). An additional modifying phrase such as a locative phrase, also termed as coda, is found in the ext, but its presence is not requested in all circumstances. The main contention of this study is that the ext does not involve an act of predication, in the sense of a semantic relation between a subject and a predicate.6 In other words, sentences in the ext do not fit into a normal thematic pattern consisting of a predicate and argument(s), hence the thematic and syntactic status of its e‑np appear to be problematic. The present study argues that the syntactic and semantic structure of the ext is not subject to the decompositional mechanisms employed in a regular predication. Instead, it is suggested to interpret it as one block predication.7 In what follows, I demonstrate that the idiosyncrasy of the ext derives from the fact that it integrates its parts into one “block predication,” both conceptually and structurally.8

4On a par with this assumption, this study seeks to advocate that the e‑np does not play in the construction at stake a role of an argument of any sort, rather it is to be understood as an integral part of the predication as a whole. In mh, as in many languages worldwide, as cognitive a form‑meaning‑function unit, the ext construction allows for a wide range of other readings besides bare assertion of existence.

  • 9 Borschev & Partee, 2002, 2007.

5Sentences in the ext invoke a holistic, collective interpretation of activities in a given space. The participation in the event of individual entities that are regularly in a position of a topic, or in terms of “perspectival structure”9—the figure (or “primary figure”), is backgrounded, and viewed as part of the ground. Furthermore, in most of the cases the e‑np is not mentioned again in the discourse. Thus, based on discourse functional considerations, it seems that there is little support to analysis this element as an s. But, then, is it to be analyzed as an o or, diachronically, as an s that became an o? The present investigation aims to cope with this controversial issue.

  • 10 On the problematic notion of “markedness” see Haspelmath, 2006.

6Apparently, in most cases the ext has an unmarked or rather “less marked”10 counterpart in the plain “categorical format” that renders a similar propositional meaning. Nevertheless, there are subtle differences between the ext and its alternates in the plain construction that derive not only from their morphosyntactic properties but also from their compositional meaning, information structure and the perspectival structure according to which the event is presented. However, it is important to keep in mind that the ext may in some cases have no plain alternant, or the alternant may function in a complete different meaning. Finally, it is interesting to note that, counter to the common intuition, in many cases in spontaneous oral production of mh speakers employ the ext as the unmarked or default option, while the “plain” construction is considered as the “more marked” option.

  • 11 Cf. Givón, 1990, I, p. 380; II, pp. 743‑744.
  • 12 Cf. Mikkelsen, 2002

7Across many languages there is a conflict between the syntactic “default rules” of the language and the meaning and function of the ext. In mh and in many European languages the e‑np seems to fluctuate between an S and an O. Furthermore, both synchronically and diachronically, the ext exhibits inconsistency and instability in terms of case marking, verb agreement, and in configurational languages (prototypically English) in terms of word‑order, and in a number of languages also in terms of pronominalization. For example, in English, like in mh, there is variability in agreement control—that in standard English agreement is displayed while in colloquial usage it tends to be ignored.11 For example, there are some difficulties versus there’s some difficulties. In colloquial English in presentational constructions, the e‑np may feature in the acc, nevertheless only in person pronominal, for example, “There’s him, and there’s you and me” (Bolinger, 1977, p. 116). Variability and inconsistency may be encountered even in cognate languages. Compare, for example, the difference in person pronominal case‑marking of the e‑np in Danish versus Swedish.12

(1a) Dan. Der var kun hende.acc hjemme
(1b) Swed. Det var bara (*henne.acc) hon.nom hemma
  There was only her at home.

8Presumably, a bare ext reading is more frequently available with weak referents (non‑salient and non‑affected), while with definite nps it lends itself to a presentational or locative reading, as well as various thetic‑eventive readings that are structured like the ext or attracted into the ext as a highly productive schema.

  • 13 Cf. Lambrecht, 1994, 2000.

9In exploring the mh ext, the informational structure13 and the speaker’s perspectival choice with respect to the setting of the event or state are equally taken into consideration.

  • 14 Most of the data adduced here is extracted from authentic conversations in MH, other examples are (...)

10The main questions addressed here are: the motivation for the discrepancy tolerated between the syntactic “default rules” of mh language and the meaning and function of the ext construction and its type‑shifting instances (equivalent to there‑constructions in European languages); the “coercive force” of the ext; the construction‑specific status of the e‑np and its atypical behavioural and coding properties.14

11After having introduced the ext in mh and spelling out the main questions of the present study, section 2 delineates the defining structural and discourse functional properties of the construction, with some comparisons with the semantically equivalents in major European languages. Taking the constructionist‑based approach, section 3 demonstrates the “coercive force” of the ext and reveals its inconsistency in spoken language. Section 4 constitutes the heart of the paper. It discusses the properties of the ext from the typological perspective, information structure and perspectival structure, while proposing a comprehensive explanation to the overall form‑meaning of the construction. Section 5 concludes and summarizes the main arguments of the paper.

2. The ext and extposs constructions in present‑day Hebrew

2.1. The typological perspective

  • 15 Gast & Haas, 2011.
  • 16 See also Russian, e.g., prišlo tri čeloveka “there came three men” (came3SG three manPL), quoted f (...)

12Before spelling out the properties of the mh ext, including the extposs as its subtype, it is important to distinguish between “thetic‑XV languages” and “thetic‑V1 languages.”15 The former (e.g., French, English and other Germanic languages) require an expletive/proform (X) in initial position (e.g., there is/are; es gibt; il y a), while the later, namely languages of the “thetic‑V1” type, such as Hebrew and Romance languages except French, do not require the occurrence of an expletive/proform to fulfill the s position.16

  • 17 Goldenberg, 1998, 2013. A similar view has been traditionally held in Latin and Ancient Greek gram (...)
  • 18 Abbreviations in interlinear glosses are according to the Leipzig Glossing Rules. For other abbrev (...)

13Being a synthetic inflectional language Hebrew does not require the encoding of an overt, independent, s constituent that syntactically, though not pragmatically, is appositional to the s morpheme incorporated into the finite verb form.17 By contrast, in analytic inflectional languages like French, English and Germanic languages, a non‑referential s placeholder must be encoded to satisfy the verbal analytic structure requirement. Consider the following examples from mh and Spanish, both representing “thetic‑V1” languages (though, unlike Hebrew, Spanish habere exts never exhibit the acc preposition a):18

(2a) haya ‘az gam 'et ha‑milxam‑ot še‑yeš hayom
  be.pst.3sg then also acc def‑warf.pl that‑ext today
(2b) hubo3sg entonces también las guerrasf.pl que también hay hoy (S. Nemirovsky)

14Conversely, consider the following from colloquial mh vs. its parallels in “thetic‑XV” languages like German, Swedish and French, in which a definite np is infelicitous in “there‑construction”:

(3a) yarad 'et ha‑me'a dollar me‑ha‑xešbon šeli !
  drop.pst3.m.sg acc def‑hundred dollar from‑def account mine
  (Those) hundred dollars were withdrawn from my account !
(3b) Es wurde (?die/diese) hundert Dollar von meinem Konto abgezogen !
(3c) Det avdrogs (?dessa) hundra dollar från mitt konto !
(3d) Il était dépouillé (?les/?ces) de cent dollars de mon compte!
  • 19 Milsark, 1974; Keenan, 2003; Leonetti, 2008, a.o.

15The abovementioned examples sustain the assumption that the typological distinction between “thetic‑V1” and “thetic‑XV” languages is fundamental, particularly when trying to account for the alleged restriction on the occurrence of definite nps in ext and other “there‑constructions,” which is known in the literature as the “Definiteness Effect.”19

2.2. Properties of the EXT construction in current Hebrew

  • 20 The negator 'en is not just the opposite of ext yeš. It has different pragmatic functions in mh, f (...)
  • 21 The particle 'et is a differential direct object marker (dom) indispensable in mh before definite (...)

16Let us now probe into the mh ext in order to better understand how the functional demands conflict with syntactic default rules of agreement control and case marking. As already seen in (2), the ext consists of a dedicated ext predicator yeš (neg. 'en) “there‑is” (is not),20 inflected in past and future tenses in the invariable, non‑referential 3m.sg of the verb haya “be,” followed by the e‑np (by default in postposition). When the e‑np occurs in the definite or pronominal form it is marked by the dom 'et21 alongside ignoring agreement control. For example:

(4) Q. yeš 'et ha‑pasta me‑'etmol?
    ext acc def‑pastaf.sg from‑yesterday?
    Is there (still) the pasta from yesterday?
  A. lo. 'en. haya 'ota 'ad ha‑boker.
    no. neg.ext be.pst.3m.sg acc‑pro3f.sg until def‑morning
    No, there’s not. It was (there) until this morning
  • 22 This construction is already attested in bh, for example in 2 Kgs 10: 15, and in Num 5: 10 in the (...)

17That is, the grammaticalized pattern of the mh ext is [v(sø) + {acc def‑} e‑np]22, where sø stands for the invariable, non‑referential s incorporated into the verb form.

18To mark possession, Hebrew employs a construction equivalent to the French être à, comprised of yeš (neg. 'en) and the possessor introduced by the dative clitic l‑ + the possessed marked by dom 'et (if definite), for example: 

(5) yeš l‑i 'et ha‑sefer šelxa ba‑bayit
  ext to‑me acc def‑book yoursm.sg in.def‑home
  I have your book at home

19In the marked word‑order, whereby the e‑np is thematized, acc 'et particle is nevertheless inserted before the np, as in:

(6) rak 'et ha‑ma'amar‑im be‑'anglit 'od 'en l‑i
  only acc def‑articlem.pl in‑English still neg.ext
  It is only the articles in English that I don’t have yet’.

20All the more so, acc 'et can optionally precede complementizer phrases, as in:

(7) kše‑higa'‑nu kvar lo haya 'et ma še‑ratsi‑nu
  when‑arrive.pst1pl already not pst3m.sg acc what that want pst1.pl
  When we arrived, there already wasn’t anything left of what we wanted

21Apparently, flagging the sole thematic np by the acc is anomalous, as the the ext is available only with monovalent predicators. In other words, inasmuch as the e‑np does not function in the discourse as the element whose referent the predication is about, or regarding which the hearer’s knowledge is increased, there is no motivation for its indexing on the verb.

  • 23 SAE comprises Romance and Germanic languages as well as English, but to the exclusion of Maltese, (...)
  • 24 But see on this controversial issue, Ziv, 1982a and 1982b.

22Regarding what is known as the “Definitness Effect,” as already observed in (3), compared to many of sae languages mh seems to be more permissive in this respect.23 Definite nps are not completely excluded from the construction, though ostensibly they cannot be portrayed with strongly quantified nouns, as is the case with salient s entities.24 Consider the following utterance adduced from a conversation in a medical clinic:

(8) yeš rak 'et Sarah (ve‑gamar‑nu le‑hayom)
  ext only acc Sarah (and‑finish pst1pl for‑today)
  There is only Sarah (left) (and we are done for today).
  • 25 Gundel (1988, p. 210): “An entity E is the topic of a sentence, S, if in using S the speaker inten (...)

23This utterance is not a statement about the specific human referent Sarah, rather it provides a piece of information about a situation that is believed to be unknown to the hearer.25 In other words, the e‑np does not have to be new, as it is not the speaker’s intent to present it as a topic, but rather as part of the new information conveyed by the entire predication.

24The same is true for pronominalization. While pronouns referring to the e‑np are allowed in English and other Germanic languages only on a limited scale, substantially when the pronoun refers to a human referent, alongside optionally agreeing with the verb to be, mh is much more permissive in this respect, as shown in the following:

(9) Q. 'efo yeš 'et ha‑sefer šelo?
    where ext acc def‑book his?
    Where can I find his book?
  A. yeš 'oto ba‑sifriya
    ext acc.pro3m.sg in.def‑library
    You can find it in the library

25A further hybrid alignment may recur in spontaneous usage of mh, whereby agreement between the verb and the e‑np is obtained, while at the same time acc 'et is inserted in front of the e‑np (if definite), as in:

(10) 'aval hay‑u gam 'et 'ele še‑lo hiskim‑u 'it‑a
  but be.pst3pl also acc those that‑not agree.pst3.pl with‑her
  But there were also those who did not agree with her
  • 26 Goldenberg, 2013, p. 180. This phenomenon is also known in other languages, and is not restricted (...)

26Incongruent agreement between the verb and the e‑np as in (10) above, can be accounted for as attraction of agreement by a non‑s constituent, in our case the e‑np.26 That is, in such cases, the NP attracts agreement even though it does not behave as a full‑fledged S. However, according to the view being advanced here, this does not seem to change the fact that the e-np is not understood as an S, as also from the speaker’s perspectival choice is to not foreground it is as the “primary figure.”

27It is here the place to note that in current spoken Hebrew, variability and inconsistency, and in some cases hesitation from the part of the speaker, are frequently part of the encoding of the ext, and especially its “type‑shifting instances” whose uttering is less conventionalized in comparison to that of the ext type‑construction. This fact seems to prove that speakers regard the structuring of the ext as pragmatically conditioned. More will be said on this later (cf. infra).

  • 27 This construction is already attested in late BH, e.g., EST. 3: 8.

28Another construction, ubiquitous in mh, built likewise around verboid yeš, however not inflected for the future and past tenses, occurs in a complex form with a bound pronoun in the third person that agrees with the post‑positional e‑np, and thus closer to an O than to an S,27 for example:

(11) yeš‑nam šinuy‑im ba‑toxnit
  ext‑pro3m.pl changem.pl in.def‑program
  There are changes in the program’.

29This construction is puzzling no less than the ext type‑construction under discussion. The reason for that is that the e‑np is construed in the indefinite form, while the bound pronoun, which is attached to yeš, is coreferential to the e‑np. All the more so, in accordance with the perspectival structure and the impersonal meaning of the overall construction, in the neg.ext 'en “(there) is not,” the preference in current spoken Hebrew is almost exclusively for third person. Thus, strikingly, in (12) below, a contradiction between first person singular 'ani, and third person singular in the neg.ext bound form 'enena (“is not‑pro3f.sg”) is created:

(12) tagid še‑'ani 'ene‑na
  sayfut2sg.m that‑I neg.ext‑pro3f.sg
  Say (tell him/her) that I am not here.

30A special construction emerged in mh in a tell-tale function. This context‑restricted construction in patterned in the form of hayo haya + np “once there was an np” (lit. be was a np), which is comprised of a “tautological infinitive” expression that intensifies the verb haya “be” (i.e., the infinitive absolute of the verb is reduplicated by the inflected verb form) and followed by an e‑np, as in:

(13) hayo haya dayag 'ani
  be be.pst3m.sg fisherman poor
  Once there was a poor fisherman.

31However, unfortunately, these variants of the ext, briefly introduced, are beyond the scope of the present study.

3. ext‑like constructions in colloquial mh

  • 28 Croft, 2001; Fillmore, 1988; Goldberg, 1995; a.o.
  • 29 The “coercive force” of a construction is understood here as a cognitive, interpretive and creativ (...)
  • 30 Cf. Borschev & Partee, 2007, p. 160. [quoting Jackendoff’s (1992, 1990) and Paducheva (1992, 1997) (...)
  • 31 Cf. Kuzar, 2012, pp. 103‑114, 117‑179.

32In line with Construction Grammar, a construction is assumed here to be a form/meaning‑function complex.28 It is further assumed that the formal meaning of a construction as a whole can modify, and even override, the composite denotation of its specific instantiation. Hence, as a highly productive schema, the ext exerts its coercive force upon sentences that do not necessarily assert existence or existence of possession in the literal sense.29 The coercive force of the type‑construction changes the argument structure of the type‑shifting instances, resulting in a co‑composition of the lexical meaning of the verb and the entire structural meaning of the construction. Through metaphorical/structural extension “being in a location” may be interpreted as “being in some state,” or “occurring in some spatiotemporal region;” by the same token, “being in someone’s possession” may extend to “being in the speaker’s (or an observer’s) perceptual field.”30 In such type‑shifting instances there is an expansion of the ext meaning of the construction to denote occurrence and disappearance, (in)sufficiency, commencement and termination of a situation, informative statements including impersonal passives, statements of desired state of affairs, and more.31

33In current spoken Hebrew, the expanded uses of the ext are construed from a similar perspective, centering on the event or state as a whole, and manifest a comparable putative mismatch between grammatical relations, structural positions, and case‑marking. As seen in the following:

(14) niš'ar raq 'et ha‑ 'uga
  remain.pst3m.sg only acc def‑cakef.sg
  There’s only the cake left.
(15) rašum kan 'et kol ha‑prat‑im?
  writtenm.sg here acc all def‑detailm.pl?
  All the details are written down here?
(16) lo mofia l‑i 'et h a‑ktovet šelo
  not appearm.sg l‑i to‑me acc def‑addressf.sg his
  I don’t see his address’ parents address (it does not appear to me).
(17) betax kore 'et ha‑ta'ut ha‑zot le‑harbe
  surely happenm.sg acc def‑errorf.sg def‑thisf.sg to‑many
  For sure this happens to many (people).
(18) yatsa 'et ze be‑miqre
  come.out.pst3m.sg acc it by‑accident
  It happened by accident.
(19) yeš ne mal‑im ba‑gina. male 'otan be‑kol pina
  ext antf.pl in.def garden full(of) acc‑pro 3f.pl in‑every corner
  There are ants in the garden. They are everywhere/in every corner (lit. it is full of them) 
(20) magia l‑o 'et ha‑matana ha‑zot
  arrive3m.sg to‑him acc def‑presentf.sg def‑thisf
  He deserves this present.

34And strikingly:

(21) ba‑sof nolad l‑a bat
  in.def‑end born3m.sg to‑her girl
  Eventually she had a girl.

35The atypical acc flagging of the sole thematic participant is all‑pervasive in oral elicitation of mh, especially in anaphoric and deictic forms, alongside neutralization of agreement, for example:

(22) haya 'et ze gam ba‑bayit šelanu
  be.pst3m.sg acc this/it also in.def‑house ours
  It was also in our house.

36Turning now to the analysis of the mh grammaticalized ext, I will elaborate on factors beyond syntax that interact in the structuring and meaning and function of the ext.

4. Discussion

37Our analysis takes as its starting point the hypothesis that the hybrid alignment of the ext, crucially the acc flagging of the sole thematic np, is due to discourse functional considerations and not to any formal or semantic requirement. In other words, it is affected by information structure (the np is non‑topic) and perspectival structure.

38Clearly, insofar as the ext denotes an “instatiation” or state of affairs, rather than an activity, there is no motivation for referring to any specific individual. Formally, this is manifested in mh by uttering the verb in the neutral, invariable 3m.sg form. Furthermore, by cancelling the canonical bipartite division of the sentence into subject/topic and predicate/comment, a conflict is created between the formal demands of agreement and case marking and the functional demands.

  • 32 The recommendation of “purists” of mh to retain agreement and avoid using the direct object marker (...)

39However, while the “hybrid alignment” of the ext type‑construction has more or less conventionalized in standard mh,32 inconsistency is most frequently detected in spontaneous production of ext‑like constructions. Intuitions about when to flag the sole thematic np by particle 'et alongside override agreement vary greatly among speakers and depend on the register and type of the predicate involved. It is important to stress here that genuine direct objects do not display in standard mh such inconsistency.

  • 33 Cf. Henkin, 1994, p. 45; who rightly alludes to the possibility that 'et is a “general non‑subject (...)
  • 34 Halevy, 2007, 2013: § 16.

40In many languages case markers are employed not only for formal reasons, but also for semantic and pragmatic reasons. Furthermore, it is not coincidental that languages vary as to what case they assign to grammatical relations in non‑canonical positions. I contend that the reason why the e‑np in the construction under examination is flagged by the acc lies in the fact that it is the default overtly coded case in Hebrew for marking non‑s constituents (as is probably true for other nom‑acc languages where the nom is zero coded).33 All the more so, in both classical and mh, particle 'et is employed in different functions, not only as a dom, notably for deploying an intensive, or all‑inclusive and holistic effect.34 For the sake of brevity, it would suffice to cite the following examples from colloquial mh:

(23) halax‑nu 'et kol Manhattan ba‑regel
  walk.pst1.pl acc all Manhattan by.def‑foot
  We walked [the length and width of] all Manhattan by foot.
(24) tsarax‑nu 'et ha‑xayim šel‑ anu (www.nana10.co.il/)
  scream.pst1.pl acc def‑life ours
  We yelled like mad’ (lit. we screamed our lives).
  • 35 Cf. Lazard, 1994; Creissels, 2009, 2014; Faarlund, 2001; Bentley, Ciconte & Cruschina, 2015; a.o. (...)
  • 36 Hebrew is a nom‑acc language. Proto‑Semitic morphological case‑marking has disappeared in Hebrew, (...)

41Given the observation above, I argue that the e‑np that signifies the only obligatory participant, albeit a non‑salient and non‑affected entity, is to be regarded as a clause‑internal constituent. Counter to the more common view, predominantly regarding the parallel construction in sae,35 which is biased by the position of the e‑np in the linear ordering of the sentence, I contend that the e‑np in the mh construction is not an s that became an o. In fact, I argue that it is neither s nor o. I suggest the following scenario: speakers of mh associate high degree of referentiality and individuation with typical properties of subjetcs, therefore they feel that they need to explicitly mark this sole obligatory np that, against our expectations, is designating a non‑s constituent. In sum: the motivation behind divesting the e‑np of the status of agreement controller, alongside flagging it by the acc,36 derive from the need to position its referent in the discourse in such a way that it does not coincide with the non‑referential s incorporated into the verb form. Undoubtedly, this marking strategy is invoked by the typology of Hebrew as a “thetic‑V1” language and by the fact that the s morpheme is as a rule couched in the verb form, contrary to “thetic‑XV” languages where s must be encoded separately, even when it qualifies as a “dummy‑s.

  • 37 The default word‑order in classical Hebrew, like in other Semitic languages, was VSO, however, pro (...)

42As to the structural position of the e‑np, I am reluctant to account for it in terms of “subject inversion,” as often subsumed regarding its parallel in European languages. Such an account is not feasible for mh for the following reasons: svo is not a strict word‑order in mh;37 Hebrew discerns grammatical functions via morphology (viz. agreement, case/clitics) rather than positionally, through word‑order, as is common in configurationally sae languages.

43In conclusion, while being morphosyntactically realized, the ext is pragmatically determined. Along these lines, I suggest to regard the anomalous acc marking of the e‑np as a device for marking a pragmatic inverse.

5. Conclusion

44The category of exts has been defined based on structural and functional, rather than semantic, notions.

45Being a highly productive construction the ext contributes its schematic meaning to a wide range of sentence types that instantiate it, and go beyond the meanings contributed by their components. The “coerced” sentences of the mh construction include statements of occurrence, subjective states, modal‑evaluative perspectives, as well as a subset of impersonal passive constructions involving verbs denoting transmission of information. However, whereas in standard mh the formal encoding of the ext is relatively conventionalized, this is not the case in oral spontaneous production, especially in uttering the “coerced” instances of the ext where variability and inconsistency, depending to a large extent on the speaker’s intuition, are frequently detected. The variability and inconsistency in the alignment of the ext construction suggest that it is anchored in discourse and speakers’ perspectival choices rather than in grammar.

46It was argued that in dealing with the atypical properties of the ext it is essential to analyse them in relation to the semantic and pragmatic value of the construction, as well as in accordance with the language‑specific grammaticalization path of the clause type as a whole.

47I challenged the view according to which the e‑np is an s that became an o. Instead, I proposed that, synchronically, it is to be treated as a non‑argumental constituent that functions as an internal part of the “block predication.”

48As to the puzzling flagging of the e‑np, it was ascribed to the function of the acc in mh as the default overtly coded case, and, subsequently, the default choice for identifying and distinguishing non‑s constituents, crucially when displayed in the definite form. From the discourse‑functional perspective, it was suggested to regard the atypical acc flagging as a device for divesting the non‑topical and backgrounded e‑np from the non‑referential s incorporated into the verb form. It was suggested that employing this strategy for marking a “pragmatic inverse” is instigated by the typology of Hebrew as a “thetic‑V1” language.

49Ultimately, it can be posited that the “hybrid alignment” of the ext corresponds to the meaning and function of the construction and does not contradict it.

Bibliographie

Bentley Delia, Ciconte Francesco Maria & Cruschina Silvio, 2015, Existentials and Locatives in Romance Dialects of Italy, Oxford University Press, Oxford, 331 p.

Bolinger Dwight L., 1977, Meaning and Form, Longman (coll. English Language Series), London & New York, 212 p.

Borschev Vladimir & Partee Barbara H., 2002, “The Russian Genitive of Negation : Theme-Rheme Structure or Perspective Structure?” in Journal of Slavic Linguistics, no 1/2, vol. 10, pp. 105‑144.

Borschev Vladimir & Partee Barbara H., 2008, “Existential Sentences, BE, and the Genitive of Negation in Russian” in Comorovski Ileana & Heusinger Klaus von (eds.), Existence: Semantics and Syntax, Springer, Dordrecht, pp. 147‑190.

Creissels Denis, 2009, “Uncommon Patterns of Core Term Marking and Case Terminology” in Lingua, no 3, vol. 119, pp. 445‑459, DOI : 10.1016/j.lingua.2008.09.007.

Croft William, 2001, Radical Construction Grammar: Syntactic Theory in Typological Perspective, Oxford University Press, Oxford & New York, 416 p.

Faarlund Jan Terje, 1998, « L’actance dans les langues germaniques » in Feuillet Jack (dir.), Actance et Valence dans les langues de l’Europe, Mouton de Gruyter, Berlin & New York, p. 789‑809.

Fillmore Charles, 1988, “The Mechanism of Construction” in Proceedings of the Fourteenth Annual Meeting of the Berkeley Linguistic Society, University of Berkeley, Berkeley, pp. 35‑55.

Francez Itamar, 2007, Existential Propositions, Stanford University, Stanford, 151 p.

Francez Itamar, 2009, “Existentials, Predication, and Modification” in Linguistics and Philosophy, no 1, vol. 32, pp. 1‑50, DOI : 10.1007/s10988-009-9055-4.

Gast Volker & Haas Florian, 2011, “On the Distribution of Subject Properties in Formulaic Presentationals of Germanic and Romance: A Diachronic-Typological Approach”, in Malchukov Andrej & Siewierska Anna (eds.), Impersonal Constructions: A cross linguistic perspective, John Benjamins Publishing Company, Amsterdam, pp. 127‑166, DOI : 10.1075/slcs.124.05gas.

Givón Talmy, 2001, Syntax: an Introduction, vol. I, John Benjamins, Amsterdam, 500 p.

Goldberg Adele E., 1995, Constructions: A Construction Grammar Approach to Argument Structure, University of Chicago Press, Chicago, 265 p.

Goldenberg Gideon, 1998, “On Verbal Structure and the Hebrew Verb” in Goldenberg Gideon (ed.), Studies in Semitic Linguistics: Selected Writings, Magness Press, Jerusalem, pp. 148‑196.

Goldenberg Gideon, 2006, “On Grammatical Agreement and Verb-Initial Sentences” in Borbone Pier Gorgio, Mengozzi Alessandro & Tosco Mauro (eds.), Loquentes linguis: Studi linguistic e oriental in onore di Fabrizio A. Pennacchetti, Harrassowitz, Wiesbaden, pp. 329-335.

Goldenberg Gideon, 2013, Semitic Languages: Features, Structures, Relations, Processes, Oxford University Press, Oxford, 363 p.

Gundel Jeanette K., 1988, “Universals of Topic-comment Structure” in Hammond Michael, Moravcsik Edith A. & Wirth Jessica (eds.), Studies in Syntactic Typology, John Benjamins, Amsterdam, pp. 209‑239, DOI : 10.1075/tsl.17.16gun.

Halevy Rivka, 2007, “Transitive Verbs With Non Accusative Alternation in Hebrew : Cross Language Comparison with English, German and Spanish” in Delbecque Nicole & Cornillie Bert (eds.), Trends in Linguistics—On Interpreting Construction Schemas, from Action and Motion to Transitivity and Causality, Mouton de Gruyter, Berlin & New York, pp. 61‑102.

Halevy Rivka, 2013, “Syntax: Modern Hebrew” in Khan Geoffrey (ed.), Encyclopedia of Hebrew Language and Linguistics, vol. 3, Brill, Leiden, pp. 707‑723.

Haspelmath Martin, 1998, “How Young is Standard Average European?” in Language Sciences, no 3, vol. 20, pp. 271‑287, DOI : 10.1016/S0388-0001(98)00004-7.

Haspelmath Martin, 2006, “Against Markedness (and What to Replace it With)” in Journal of Linguistics, no 1, vol. 42, pp. 25‑70.

Hazout Ilan, 2004, “The Syntax of Existential Constructions” in Linguistic Inquiry, no 35/3, pp. 493-430.

Henkin Roni, 1994, “Yeš gam ’et ze [EXT also ACC this]” in Balshanut Ivrit, no 38, pp. 41‑54.

Keenan Edward L., 2003, “The Definiteness Effect : Semantics or Pragmatics ?” in Natural Language Semantics, no 2, vol. 11, pp. 187‑216, DOI : 10.1023/A:1024400425028.

Kuzar Ron, 2012, Sentence Patterns in English and Hebrew, John Benjamins, Amsterdam & Philadelphia, 254 p.

Lambrecht Knud, 1987, “Sentence Focus, Information Structure, and the Thetic Categorical Distinction” in The Annual Proceedings of the Berkeley Linguistics Society, vol. 13, pp. 366‑382, DOI : 10.3765/bls.v13i0.1800.

Lambrecht Knud, 2000, “When Subjects Behave Like Objects: An Analysis of the Merging of S and O in Sentence-Focus Constructions Across Languages” in Studies in Language, no 3, vol. 24, pp. 611‑682, DOI : 10.1075/sl.24.3.06lam.

Langacker Ronald W., 1987, Foundations of Cognitive Grammar, vol. 1, Theoretical Prerequisites, Stanford University Press, Stanford, 516 p.

Langacker Ronald W., 1993, “Universals of Construal” in The Annual Proceedings of the Berkeley Linguistics Society, no 1, vol. 19, pp. 447‑463, DOI : 10.3765/bls.v19i1.1532.

Lauwers Peter & Willems Dominique, 2011, “Coercion: Definition and Challenges, Current Approaches, and New Trends” in Linguistics, no 6, vol. 49, pp. 1219‑1235, DOI : 10.1515/ling.2011.034.

Lazard Gilbert, 1994, « L’actant H  : sujet ou objet ? » in Bulletin de la société de linguistique de Paris, no 1, vol. 89, p. 1‑28, DOI : 10.2143/BSL.89.1.2013024.

Leonetti Manuel, 2008, “Definiteness Effects and the Role of the Coda in Existential Constructions” in Müller Henrik Høeg & Klinge Alex (eds.), Essays on Nominal Determination, John Benjamins, Amsterdam, pp. 131‑162.

Malchukov Andrej L. & Ogawa Akio, 2011, “Towards a Typology of Impersonal Constructions. A Semantic Map Approach” in Malchukov Andrej L. & Siewierska Anna (eds.), Impersonal Constructions: A Cross-Linguistic Perspective, John Benjamins Publishing Company, Amsterdam, pp. 19‑56, DOI : 10.1075/slcs.124.02mal.

McNally Louise, 2011, “Existential Sentences” in Maienborn Claudia, Heusinger Klaus von & Portner Paul (eds.), Semantics: an International Handbook of Natural Language Meaning, vol. 2, Mouton de Gruyter, Berlin & Boston, pp. 1829‑1848.

Mikkelsen Line, 2002, “Reanalyzing the Definiteness Effect: Evidence from Danish” in Working Papers in Scandinavian Syntax, no 65, pp. 11‑75.

Milsark Gary Lee, 1974, Existential Sentences in English, Ph.D Thesis, MIT, Cambridge, 260 p.

Sasse Hans-Jürgen, 1987, “The Thetic/Categorical Distinction Revisited” in Linguistics, no 3, vol. 25, pp. 511‑580, DOI : 10.1515/ling.1987.25.3.511.

Taube Moshe, 2015, “The Usual Suspects: Slavic, Yiddish, and the Accusative Existentials and Possessives in Modern Hebrew” in Journal of Jewish Languages, no 1‑2, vol. 3, pp. 27‑37, DOI : 10.1163/22134638-12340035.

Traugott Elizabeth Closs, 2007, “The Concepts of Constructional Mismatch and Type‑Shifting from the Perspective of Grammaticalization” in Cognitive Linguistics, no 4, vol. 18, pp. 523‑557, DOI : 10.1515/COG.2007.027.

Williams Edwin, 1984, “There-Insertion” in Linguistic Inquiry, no 1, vol. 15, pp. 131‑153.

Ziv Yael, 1982a, “Another Look at Definites in Existential” in Linguistics, no 18, pp. 73‑88.

Ziv Yael, 1982b, “On so Called Existentials: A Typological Problem” in Lingua, no 56, pp. 261‑281.

Annexes

do direct object
dom differential object marker
e‑np (the) existee’s noun phrase
ext existential construction
ext‑poss existential possession
o object
mh Modern Hebrew
s subject

Notes

1 E.g., McNally, 2011, a.o.

2 Cf. Croft, 2001; Fillmore, 1988 ; Goldberg, 1995, a.o.

3 Sasse, 1987, p. 568

4 On the distinction “thetic” vs. “categorical” see Lambrecht, 1987; Sasse, 1987, a.o.

5 Williams, 1984; Francez, 2007, 2009, a.o.

6 But see Williams (Williams, 1984, p. 135) who argues that, e.g., there are problems involves a normal subject predicate relation, i.e., of expletive there as subject, and problems—the “pivot” as predicate. Cf. also Hazout, 2004 for a similar approach.

7 See Goldenberg, 2006, p. 335. From a different perspective, Francez (2007‑2009) argues that in sentences of the form “There be NPpivot XPcoda” the pivot is a “second order predicate” (and not an argument of some predicate) and the coda is a modifier.

8 In some respects, comparable to phraseological expressions, in which the informational value of the construction and its form‑meaning are interpreted in terms of the whole. It may not be a coincidence that the ext shares with such expressions a similar linear vs ordering.

9 Borschev & Partee, 2002, 2007.

10 On the problematic notion of “markedness” see Haspelmath, 2006.

11 Cf. Givón, 1990, I, p. 380; II, pp. 743‑744.

12 Cf. Mikkelsen, 2002

13 Cf. Lambrecht, 1994, 2000.

14 Most of the data adduced here is extracted from authentic conversations in MH, other examples are taken from www.israblog.co.il and from mh corpora in www.sketchengine.co.uk.

15 Gast & Haas, 2011.

16 See also Russian, e.g., prišlo tri čeloveka “there came three men” (came3SG three manPL), quoted from Malchukov & Ogawa, 2011, p. 30.

17 Goldenberg, 1998, 2013. A similar view has been traditionally held in Latin and Ancient Greek grammar, where the representation of the S on the verb was referred to as agreement, even though this marking could occur without they’re being any coreferent NP.

18 Abbreviations in interlinear glosses are according to the Leipzig Glossing Rules. For other abbreviations see the Appendix.

19 Milsark, 1974; Keenan, 2003; Leonetti, 2008, a.o.

20 The negator 'en is not just the opposite of ext yeš. It has different pragmatic functions in mh, furthermore, it occurs also in copular and nominal predications. Across many languages negative exts constitute a separate grammatical and conceptual construction.

21 The particle 'et is a differential direct object marker (dom) indispensable in mh before definite nouns.

22 This construction is already attested in bh, for example in 2 Kgs 10: 15, and in Num 5: 10 in the ext‑poss pattern.

23 SAE comprises Romance and Germanic languages as well as English, but to the exclusion of Maltese, Basque, and Celtic and Finno‑Ugric languages. Haspelmath, 1998.

24 But see on this controversial issue, Ziv, 1982a and 1982b.

25 Gundel (1988, p. 210): “An entity E is the topic of a sentence, S, if in using S the speaker intends to increase the addressee’s knowledge about, request information about, or otherwise get an addressee to act with respect to E.”

26 Goldenberg, 2013, p. 180. This phenomenon is also known in other languages, and is not restricted to pure ECs. For example, French Il sont venus de nouveaux membres; and German Es kommen so viele Menschen.

27 This construction is already attested in late BH, e.g., EST. 3: 8.

28 Croft, 2001; Fillmore, 1988; Goldberg, 1995; a.o.

29 The “coercive force” of a construction is understood here as a cognitive, interpretive and creative way of using a linguistic construction inductively (Goldberg, 1995; Traugott, 2007; Lauwers & Willems, 2011; a.o.), rather than as a theory of grammar that accounts for sentences that are not predicted by the basic rules of grammar.

30 Cf. Borschev & Partee, 2007, p. 160. [quoting Jackendoff’s (1992, 1990) and Paducheva (1992, 1997)].

31 Cf. Kuzar, 2012, pp. 103‑114, 117‑179.

32 The recommendation of “purists” of mh to retain agreement and avoid using the direct object marker is still valid, though its implementation has long been faded. This view is based on the assumption that the proliferation of the acc in ext and Ext‑poss constructions is due to the influence of mh contact‑languages, most crucially Yiddish, in which the possessee in the construction with hobn (“have”) is flagged by acc (cf. Taube, 2015). However, as noted above, in fn. 22, the atypical acc marking in exts is attested already in bh (though infrequently).

33 Cf. Henkin, 1994, p. 45; who rightly alludes to the possibility that 'et is a “general non‑subject case‑marker” in Hebrew.

34 Halevy, 2007, 2013: § 16.

35 Cf. Lazard, 1994; Creissels, 2009, 2014; Faarlund, 2001; Bentley, Ciconte & Cruschina, 2015; a.o. Lambrecht, 2000, pp. 623‑625; argues that, diachronically, the postverbal np is a detopicalized S devoid of (some or all of) the S properties; consequently it has a status somewhere “between subjecthood and objecthood.” But see Lødrup (1999, pp. 206‑208) who suggests defining the np in the equivalent construction in Norwegian and other Germanic languages in negative terms, as a “non‑subject,” but, crucially, not as an object.

36 Hebrew is a nom‑acc language. Proto‑Semitic morphological case‑marking has disappeared in Hebrew, or has been replaced by adpositional markers. The nom is unmarked (i.e. appears in zero coding).

37 The default word‑order in classical Hebrew, like in other Semitic languages, was VSO, however, probably due to contact with European languages, it adapted the SVO as the default word‑order. Cf. Halevy, 2013, § 13.

Auteur

Hebrew University/Jerusalem

CC-BY-NC-SA-4.0

Le texte seul est utilisable sous licence CC BY-NC-SA 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search