Hostname: page-component-cd9895bd7-8ctnn Total loading time: 0 Render date: 2024-12-25T18:29:18.514Z Has data issue: false hasContentIssue false
  • English
  • Français

The syntax of silent directional prepositions in Jordanian Arabic

Published online by Cambridge University Press:  22 May 2023

Mohammad Alhailawani Open the ORCID record for Mohammad Alhailawani [Opens in a new window]
Mohammad Alhailawani*
Affiliation:
University of Petra, Amman, Jordan
*
mohammad.alhailawani@uop.edu.jo
Rights & Permissions [Opens in a new window]

Abstract

In Jordanian Arabic (JA), the complement of some motion verbs optionally appears without a visible preposition in what is known as (P)reposition-drop (Ioannidou and den Dikken 2009). This article offers a detailed description of P-drop in JA, showing that the common properties of P-drop found in languages with P-drop hold in JA. I argue that Gehrke and Lekakou's (2013) pseudo noun incorporation approach to P-drop cannot account for the P-drop facts in JA. I show, through different diagnostics, that the prepositionless noun in JA does not exhibit the typical properties of pseudo-incorporated nouns. Instead, I argue that P-drop in JA involves a full PP–DP structure with a silent P head (Ioannidou and den Dikken 2009, Myler 2013, Biggs 2014, Bailey 2018). The findings of this article add insights to the ongoing debate regarding the underlying mechanisms involved in P-drop.

Résumé

Résumé

En arabe jordanien (AJ), le complément de certains verbes de mouvement peut éventuellement apparaître sans préposition visible dans ce qu'on appelle (P)reposition-drop ou ‘chute de P’ en français (Ioannidou et den Dikken, 2009). Cet article propose une description détaillée de la chute de P en AJ, montrant que les propriétés communes de la chute de P trouvées dans les langues avec chute de P tiennent aussi en AJ. Je soutiens que l'approche de pseudo-incorporation nominale de Gehrke et Lekakou (2013) pour la chute de P ne peut pas rendre compte des faits P-drop dans l'AJ. En utilisant divers diagnostics, je montre que les noms sans préposition en AJ ne comportent pas les propriétés typiques des noms pseudo-incorporés. Je propose plutôt que la chute de P en AJ implique une structure PP–DP complète avec une tête P silencieuse (Ioannidou et den Dikken, 2009; Myler 2013, Biggs 2014, Bailey 2018). Les conclusions de cet article ajoutent des éclaircissements au débat en cours concernant les mécanismes sous-jacents impliqués dans la chute de P.

Keywords

Jordanian ArabicP-dropdirectional PPssilent Ppseudo noun incorporationArabe jordanienchute de PPP directionnelsP silencieuxpseudo-incorporation nominale
Type
Article
Information
Canadian Journal of Linguistics/Revue canadienne de linguistique , Volume 68 , Issue 2 , June 2023 , pp. 229 - 249
Check for updates
Copyright
Copyright © Canadian Linguistic Association/Association canadienne de linguistique 2023

1. Introduction

In Greek and some varieties of British English, the complement of some motion verbs can optionally appear without an overt preposition or a determiner. The following examples illustrate this:Footnote 1, Footnote 2

This phenomenon is known as P(reposition)-drop (Ioannidou and den Dikken Reference Ioannidou, den Dikken, Halpert, Hartman and Hill2009). P-drop happens mostly in directional contexts in a number of languages like Greek (Terzi Reference Terzi2010, Gehrke and Lekakou Reference Gehrke and Lekakou2013), some varieties of British English (Haddican and Holmberg Reference Haddican and Holmberg2012, Myler Reference Myler2013, Biggs Reference Biggs2014, Bailey Reference Bailey2018, Hall Reference Hall2019);Footnote 3 and in some Italian dialects (Longobardi Reference Longobardi2001).

In Jordanian Arabic (JA), directional prepositions like ʕa and li ‘to’ can be dropped with certain motion verbs like yiruuħ ‘go’ and yewsal ‘arrive’, as seen in (3).Footnote 4

Research on P-drop in Greek and varieties of British English identified two main approaches to this phenomenon. A number of researchers argue that P-drop involves a full PP and DP structure with a silent P head (Ioannidou and den Dikken Reference Ioannidou, den Dikken, Halpert, Hartman and Hill2009, Terzi Reference Terzi2010, Myler Reference Myler2013, Biggs Reference Biggs2014, Bailey Reference Bailey2018). Others argue in favour of the radical absence of PP structure. Under this view, the prepositionless goal noun is a pseudo-incorporated nominal (Gehrke and Lekakou Reference Gehrke and Lekakou2013, Hall Reference Hall2019).

In this article, I offer a formal description of P-drop in JA. I show that the pseudo noun incorporation (PNI) analysis of P-drop cannot capture the JA P-drop facts. In particular, I show that the prepositionless goal noun in JA does not exhibit the typical properties of a pseudo-incorporated nominal. I apply a number of diagnostics to detect the presence of a silent PP structure, and argue that P-drop in JA involves a full PP and DP structure with a silent P head, in line with Myler’s (Reference Myler2013) analysis of P-drop in the Ormskirk variety of English spoken in the Northwest of England, and Biggs's (Reference Biggs2014) analysis of the same phenomenon in Liverpool English. The discussion in this article adds to the ongoing debate regarding the underlying mechanisms involved in P-drop.

This article is organized as follows. In section 2, I provide an overview of languages with P-drop. I also discuss the two main approaches to P-drop, with representative examples. In section 3, I set out the JA P-drop facts and note the similarities and differences between JA and other P-drop languages. In section 4, I discuss and refute an analysis of JA P-drop in terms of pseudo noun incorporation. Section 5 shows that a silent P analysis of P-drop fares better at capturing the JA P-drop data. In section 6, I argue for an analysis of P-drop in JA where PP and DP are present, showing that the analysis neatly captures the JA P-drop facts. Section 7 is a brief conclusion.

2. An overview of P-drop

There are a number of properties that broadly hold for languages with P-drop. These include: (i) the presence or absence of a determiner on the goal noun; (ii) the range of motion verbs available with P-drop; (iii) the possible interpretations of the goal noun; and (iv) well-establishedness/familiarity of the goal noun. Discussion of P-drop has consistently shown that there is language-specific variation with respect to these aforementioned properties. In London English (Hall Reference Hall2019), for instance, the goal noun in P-drop contexts must appear without a visible determiner (4), whereas in Ormskirk English (Myler Reference Myler2013), the determiner must be present on the goal noun (5).

  1. (4) I went (to the) chicken shop every day last week. (Hall Reference Hall2019: 1)

  2. (5) John came *(the) pub with me. (Adapted from Myler Reference Myler2013: 189)

As mentioned in section 1, there are two main approaches to P-drop:

  1. (6) Approaches to P-drop

    1. a. Absence of PP structure: the prepositionless goal noun undergoes pseudo noun incorporation (Gehrke and Lekakou Reference Gehrke and Lekakou2013, Hall Reference Hall2019)

    2. b. A silent P head: P-drop involves a full PP and DP structure with a silent P (Ioannidou and den Dikken Reference Ioannidou, den Dikken, Halpert, Hartman and Hill2009, Terzi Reference Terzi2010, Myler Reference Myler2013, Biggs Reference Biggs2014, Bailey Reference Bailey2018).

Each of these approaches will be discussed extensively below with representative examples.

2.1 Pseudo noun incorporation (PNI)

Gehrke and Lekakou (Reference Gehrke and Lekakou2013) show that P-drop in Greek is only possible with the preposition se ‘to/at’. Additionally, only certain motion verbs are allowed to P-drop (e.g., pao/pijeno ‘go’, ftano ‘arrive’, beno ‘enter’, epistrefo ‘return’). Gehrke and Lekakou (Reference Gehrke and Lekakou2013) argue in favour of a pseudo-incorporation analysis of P-drop in Greek, showing that P-drop in this language displays hallmark properties of pseudo-incorporation (Baker Reference Baker1988, Massam Reference Massam2001, Dayal Reference Dayal2011: among others).Footnote 5 In what follows, I report on the main properties of P-drop in Greek as discussed in Gehrke and Lekakou (Reference Gehrke and Lekakou2013).

Gehrke and Lekakou (Reference Gehrke and Lekakou2013) note that the goal noun in P-drop lacks definite marking, and that the noun must be a well-established (i.e., familiar) place (7a), similar to pseudo-incorporated nominals (Borik and Gehrke Reference Borik, Gehrke, Borik and Gehrke2015). Also, the goal noun lacks number marking, yet the noun can receive a non-singular interpretation, suggesting that the noun is number-neutral (7b).

According to Gehrke and Lekakou (Reference Gehrke and Lekakou2013), another shared property between pseudo-incorporated nominals and the prepositionless goal noun is the absence of nominal modifiers (apart from kinds modifiers). The examples in (8) show that adjectives and relative clauses cannot modify the goal noun.

Gehrke and Lekakou (Reference Gehrke and Lekakou2013) show that, like pseudo-incorporated nominals, the goal noun in P-drop obligatorily takes narrow scope with respect to quantificational elements in the clause. In (9), for instance, negation takes scope over paralia ‘beach’ such that the subject will not go to any beach, and not that there is a particular beach the subject will not go to.

Finally, the goal noun in Greek P-drop is discourse-opaque; that is, the noun cannot support pronominal anaphora (10).

For Gehrke and Lekakou (Reference Gehrke and Lekakou2013), the goal noun is a predicate, denoting a property, not an individual. Thus, the noun cannot serve as an antecedent of pronominal anaphora.

Hall (Reference Hall2019) offers a similar analysis of P-drop in London English. He shows that the prepositionless goal noun in this variety exhibits most of the properties associated with pseudo-incorporated nominals. First, the goal noun in P-drop contexts lacks both definite marking and number marking, as seen in (11).

  1. (11)

    1. a. We went (*the) pub last night.

    2. b. *We went pubs last night. (Hall Reference Hall2019: 6)

Second, despite the fact that the goal noun cannot be plural, Hall (Reference Hall2019) shows that a non-singular interpretation of the goal noun is available (12).

  1. (12)

    1. a. We all went pub yesterday.

    2. b. Possible continuation: Sam went to the Lord Tredegar, Katie went to the Morgan Arms, and I went to the Horn of Plenty. (Hall Reference Hall2019: 6)

Third, like pseudo-incorporated nominals, the goal noun cannot be modified by intersective modifiers (13a –c), and only a specific set of kind modifiers are allowed (13d).

  1. (13)

    1. a. *Can we go park with the big swings?

    2. b. *I'm going pub we met at last year.

    3. c. *We went new cinema last weekend.

    4. d. I went {corner, chicken} shop. (Hall Reference Hall2019: 7)

A fourth similarity between P-drop in London English and pseudo-incorporation is the well-establishedness of the goal noun. Hall (Reference Hall2019) shows that the set of possible goal nouns in P-drop is restricted to well-established (i.e., familiar) places (14).Footnote 6

  1. (14)

    1. a. We went {shop, school, church, cinema, post office, pub, park, chicken shop}.

    2. b. *We went car dealership.(Hall Reference Hall2019: 7)

Finally, Hall (Reference Hall2019) points out that the goal noun in P-drop obligatorily takes narrow scope with respect to negation, as seen in (15).Footnote 7

  1. (15) Sam didn't go pub yesterday. (￢>pub; *pub >￢)(Hall Reference Hall2019: 8)

Following Gehrke and Lekakou (Reference Gehrke and Lekakou2013), Hall (Reference Hall2019) argues that the goal noun in London English is a pseudo-incorporated nominal since it exhibits all properties of PNI objects. For Hall, pseudo-incorporation is a last-resort case-licensing mechanism that allows a structurally reduced nominal to be licensed. This way, a violation of the Case Filter is avoided (Levin Reference Levin2015).Footnote 8

2.2 A silent P

Several authors have argued that P-drop involves a silent PP structure (Ioannidou and den Dikken Reference Ioannidou, den Dikken, Halpert, Hartman and Hill2009, Terzi Reference Terzi2010, Myler Reference Myler2013, Biggs Reference Biggs2014, Bailey Reference Bailey2018).Footnote 9 Myler (Reference Myler2013), for instance, shows that in Ormskirk English, the definite article is obligatorily present in P-drop contexts (16).

  1. (16) John came the pub with me. (Myler Reference Myler2013:189)

In this variety, P-drop is possible with a wider range of motion verbs. This includes verbs such as go, run, drive, jog, pop, and nip (17).

  1. (17)

    1. a. He came/ran/jogged the pub.

    2. b. I haven't nipped the shops yet.(Myler Reference Myler2013: 196, 203)

Surprisingly, well-establishedness of the goal noun does not seem to hold in Ormskirk English, as seen in (18).

  1. (18) The ball went (to) the other end of the field (because I kicked it so hard). (Myler Reference Myler2013: 194)

Taking these facts into account, Myler argues in favour of a fully represented PP–DP structure with a silent P head (TO). The silent P incorporates into v in den Dikken's (Reference den Dikken2010) sense in order to be licensed. Myler (Reference Myler2013) shows that the goal noun has properties of direct objects in Ormskirk English. Thus, like direct objects, the goal noun moves to spec-vP where it is assigned accusative case in line with Johnson (Reference Johnson1991).Footnote 10

A similar analysis has been proposed to account for P-drop in Liverpool English. Biggs (Reference Biggs2014) shows that the definite article is obligatorily present in P-drop contexts, and manner of motion verbs are allowed (19a). Unlike Ormskirk English, P-drop is possible with stative at in Liverpool English (19b). Also, the goal noun can be embedded in the nominal domain (19c), and strict adjacency to the verb is not required (19d).

  1. (19)

    1. a. Swim *(the) end and back.

    2. b. She's staying (at) John's tonight.

    3. c. A trip (to) the pub is called for!

    4. d. Come with me (to) the pub.(Adapted from Biggs Reference Biggs2014: 53, 54, 63)

Biggs (Reference Biggs2014) argues that P-drop in Liverpool English does not involve a null lexical P, but rather that P-drop is licensed by a null case head dubbed k. According to Biggs (Reference Biggs2014), k corresponds to Svenonius’s (Reference Svenonius, Reuland, Bhattacharya and Spathas2007) higher p in a split p/P structure.Footnote 11 However, she relabels p as k to cover the distribution of both to and at in P-drop.

There are two main difference between TO in Ormskirk English and k in Liverpool English. First, TO is a null lexical P that theta-selects its complement goal DP, whereas k does not theta-select its complement. Also, k licenses and assigns case to its complement in situ, whereas TO must incorporate into the verb in order to be licensed, with concomitant movement of TO's complement into spec-vP for case assignment. These differences account for the differences between Ormskirk English and Liverpool English with respect to P-drop.

3. P-drop in Jordanian Arabic

In this section, I report on the main properties of P-drop in JA, which can be summarized as follows:

  1. (i) The definite article is obligatorily present on the goal noun.

  2. (ii) P-drop is only allowed with certain motion verbs.

  3. (iii) The noun must be interpreted as a directional Goal.

  4. (iv) The well-establishedness requirement is relatively less obvious in JA than it is in other P-drop languages.

In JA, the definite article is obligatorily present when the preposition is dropped (20).

Only motion verbs like yiruuħ ‘go’, yiwsal ‘arrive’, and yidxul ‘enter’ are allowed (21). Manner of motion verbs are disallowed in P-drop contexts (22).

Also, P-drop in JA is limited to the prepositions ʕa ‘to’, and -li ‘to’. The prepositionless noun cannot have source or locative readings.Footnote 12 Thus, the prepositions min ‘from’ and fi ‘in’ must be overt, as seen in (23a) and (23b) respectively.

Unlike most P-drop languages, the well-establishedness requirement on the goal seems to be less strict in JA. The prepositionless goal in JA can be an unfamiliar place or institution, as long as the directional interpretation is available (24).Footnote 13

Additionally, P-drop is possible with proper names, as seen in (25).

Terzi (Reference Terzi2010) notes that P-drop in Greek is not possible with parts of locations (e.g., ‘garage’, ‘balcony’, ‘garden’). Such examples are acceptable in JA:

Summing up, the broadly attested four main properties of P-drop seem to hold in JA with the exception of the well-establishedness requirement. The requirement seems to be less restrictive in JA than it is in Greek and in London English. Thus, JA seems to be more similar to Ormskirk English and Liverpool English in this regard.

4. Against pseudo noun incorporation

As mentioned in section 2, Gehrke and Lekakou (Reference Gehrke and Lekakou2013) and Hall (Reference Hall2019) argue in favour of a pseudo-incorporation analysis of P-drop in Greek and London English respectively. In this section, I investigate the possibility of analyzing the prepositionless goal in JA as being similar to PNI objects. The following properties are generally true of PNI objects:

  1. (27) Morphosyntactic and semantic properties of PNI (Borik and Gehrke Reference Borik, Gehrke, Borik and Gehrke2015)

    1. a. Absence of definite marking and number marking.

    2. b. Well-establishedness and restrictions on modification.

    3. c. Obligatory narrow scope.

    4. d. Discourse opacity (i.e., absence of pronominal anaphora).Footnote 14

In what follows, I show that the prepositionless goal noun in JA does not have the typical properties of PNI objects.

One of the defining properties of PNI is the absence of definite marking and number marking. As mentioned above, the definite article is obligatory on the goal noun. Also, the prepositionless goal noun in JA can be plural (28).

Another well known property of PNI is the absence of nominal modifiers (apart from kinds modifiers). In JA, the goal noun can be modified by an adjective (29), a PP (30), or a relative clause (31).

In addition to kind modifiers, the prepositionless goal noun in JA can be modified by intersective adjectives like old and new (32), which are are typically analyzed as predicates of individuals (Kamp and Partee Reference Kamp and Partee1995, Partee Reference Partee, B¨auerle, Reyle and Zimmerman2010).Footnote 15

As mentioned in section 3, well-establishedness (i.e., familiarity) of the goal noun is less restrictive in JA. Prepositionless goal nouns can be unfamiliar places (33a), parts of locations (33b), and proper names (33c).

Typically, proper names are assumed to be full referential DPs (Longobardi Reference Longobardi1994). The availability of proper names in P-drop suggests that the goal noun in JA is a full DP denoting an individual, as opposed to being a property-denoting NP, as in Greek (Gehrke and Lekakou Reference Gehrke and Lekakou2013).

Borik and Gehrke (Reference Borik, Gehrke, Borik and Gehrke2015) note that obligatory narrow scope is a stable cross-linguistic property of pseudo-incorporated nominals. In JA, a definite noun following overt ʕa ‘to’ would always have a wide scope reading with respect to logical operators, such as negation. This behaviour is not surprising given the neo-Russellian idea that definite descriptions presuppose both the existence and the uniqueness of the noun.Footnote 16 Nonetheless, it is possible to force a narrow scope reading for a definite goal noun with respect to negation. For instance, the sentence in (34) has two possible readings: (i) there is a specific mall that Faisal did not go to (i.e., wide scope); and (ii) Faisal did not go to any mall (i.e., narrow scope), with the latter reading being forced.

My JA informants had the exact same intuitions for the example in (35), where the preposition ʕa is absent.

Both (34) and (35) show that the presence or absence of an overt preposition does not affect the scope properties of the goal noun. Thus, the goal noun in P-drop seems to behave like a regular PP-object in this regard. This is unexpected if the prepositionless goal noun is a pseudo-incorporated nominal.

Finally, the prepositionless goal noun in JA can support pronominal anaphora (36). This lends further support to the status of the goal noun as a full referential DP.

Summing up, it is safe to say that the prepositionless goal noun in JA does not behave like a pseudo-incorporated noun. In fact, a number of observations suggest that the goal noun in P-drop is a full referential DP that is similar to regular PP objects. First, the noun can bear definite marking and number marking. Second, the goal noun in P-drop can be a proper name. Third, a variety of nominal modifiers (e.g., intersective adjectives, PPs, and relative clauses) can modify the goal noun. Fourth, the goal noun exhibits the same scope properties regardless of the presence or absence of an overt preposition. Finally, the goal noun can introduce a discourse referent, and as such, it can be the antecedent of pronominal anaphora. Given this, the presence of a full referential DP in JA P-drop is empirically motivated. In the next section, I provide evidence for a fully represented PP structure.

5. In favour of PP-hood

Myler (Reference Myler2013) shows that null prepositions have narrower semantics than overt prepositions. P-drop in Ormskirk English can only have a directional interpretation, and not a locative one (37).

  1. (37)

    1. a. Today I'm going the library.

    2. b. *Today I'm working the library (Intended: working at the library).

    3. c. *Today I'm staying the library. (Myler Reference Myler2013: 194)

The same is true in JA. The examples in (38) show that null P in JA can never have a locative interpretation.

The question that now arises is: What accounts for the narrower semantics of null TO? To answer this question, Myler (Reference Myler2013) adopts den Dikken's (Reference den Dikken2010) idea that directional PPs involve a Path head which can be overtly realized by to and from in English (see also Koopman Reference Koopman, Cinque and Rizzi2010). This head embeds a locative PP indicating Place instantiated by prepositions like in, on, and at.

  1. (39) [PPATH [PPLACE DP]]

    to/from in/on/at…(Adapted from Myler Reference Myler2013: 195)

Myler (Reference Myler2013) proposes that null TO is more semantically restricted than overt to because it is compatible with only one variant of the Place head, namely silent AT. On the other hand, overt to may co-occur with a much wider range of Place heads.Footnote 17

Hall (Reference Hall2019) notes that PPs and other adverbials cannot intervene between the verb and the prepositionless goal noun in London English (40), which suggests the absence of a PP structure.

  1. (40)

    1. a. *Come with me shop.

    2. b. *I went quickly shop.

    3. c. *Come with me quickly Stratford. (Adapted from Hall Reference Hall2019: 2)

In JA, strict adjacency between the verb and the goal noun is not always required. For instance, PP and adverbial intervenors may occur in multiple positions relative to the main verb. The following data show that the positioning possibilities of such intervenors are the same regardless of the presence or absence of an overt preposition.Footnote 18

Another argument for the presence of a PP comes from straight/right modification, a classic diagnostic of prepositions (Emonds Reference Emonds1985, Biggs Reference Biggs2014, Polinsky Reference Polinsky2016). In JA, overt PPs can be modified by duɣri ‘straight’, and in such a case, duɣri have three possible positions: (i) before the verb (43a); (ii) between the verb and the PP (43b); and (iii) following both the verb and the PP (43c).Footnote 19

Note that the use of duɣri as an adverbial requires the presence of a directional PP, as seen in (44).Footnote 20

As in overt PPs, duɣri ‘straight’ can appear in all three positions in P-drop contexts, as seen in (45).Footnote 21

The availability of modification by duɣri ‘straight’ in (45) is a robust piece of evidence for the presence of a silent PP structure.

6. Deriving P-drop in Jordanian Arabic

As shown in section 5, the P-drop patterns in JA suggest that there is a silent PP structure. Moreover, in section 4, I established that the prepositionless goal noun in JA P-drop is a full referential DP. In this section, I propose my analysis of P-drop in JA. In the remainder of this article, I will refer to the prepositionless goal noun in P-drop as “the goal DP”.

To begin with, I adopt the split p/P hypothesis of Svenonius (Reference Svenonius, Reuland, Bhattacharya and Spathas2007), for whom PPs and VPs share similarities in argument structure. In particular, adpositions have a split p/P structure that corresponds to split v/V in the verbal domain. Svenonius’s (Reference Svenonius, Reuland, Bhattacharya and Spathas2007) primary motivation for a split p/P is to capture the argument structure of adpositions consisting of Figures and Grounds (Talmy Reference Talmy1975). By analogy to v (or Voice0 in Kratzer Reference Kratzer, Rooryck and Zaring1996), p introduces the external argument (i.e., the Figure) as its specifier, whereas the lower lexical P introduces the internal argument (i.e., the Ground) as its complement. Both positions are depicted in (46).

  1. (46)

I argue that P drop in JA involves a silent P head that corresponds to an overt directional P (i.e., PDir in den Dikken's Reference den Dikken2010 sense). I label this head as ʕaNULL. I assume that ʕaNULL theta-selects the goal DP as its complement, similarly to overt PDir in JA.Footnote 22 Taking the structure in (46) into consideration, I propose that the P-drop example in (47) has the structure in (48).Footnote 23

At this point, two questions need to be resolved: (i) how ʕaNULL is licensed; and (ii) how case assignment to the goal DP takes place in P-drop. In addressing both questions, I will first examine two existing accounts of silent Ps in the literature, namely Myler (Reference Myler2013) and Biggs (Reference Biggs2014). Then I will show that the answers to both questions are found in both accounts. More precisely, I will show that ʕaNULL in JA is licensed similarly to TO in Ormskirk English (Myler Reference Myler2013), whereas case assignment to the goal DP in JA takes place as in Liverpool English (Biggs Reference Biggs2014).

Myler (Reference Myler2013) offers two options regarding the licensing of null TO in Ormskirk English. The first option is to assume that TO is a null lexical P that not only theta-selects the goal DP, but also assigns non-structural case to it. Under this option, null P is deleted under adjacency with the verb. The second option is to assume that null TO is licensed via P-incorporation into v (den Dikken Reference den Dikken2010).Footnote 24 Myler (Reference Myler2013) argues in favour of P-incorporation rather than deletion of P under adjacency with the verb by showing that linear adjacency is not always required. In particular, if linear adjacency between the verb and the goal DP is what licenses null P, then we predict that there should be no cases where the goal is separated from the verb. This prediction is not borne out, since P-drop is possible with ditransitives in Ormskirk English (49). As for case assignment, Myler (Reference Myler2013) argues that a goal DP in P-drop moves to direct object position Spec-v, where it is assigned accusative case. This means that the goal DP behaves like a regular direct object, as opposed to ordinary goal PPs. According to Myler (Reference Myler2013), the DP in P-drop displays an adjacency effect similar to direct objects, as seen in (50).

  1. (49) Me nan sent me the shops. (Myler Reference Myler2013: 189)

  2. (50)

    1. a. Come the pub with me.

    2. b. *Come with me the pub. (Myler Reference Myler2013: 198)

Now back to JA: I argue that ʕaNULL is licensed via P-incorporation into v (den Dikken Reference den Dikken2010), as opposed to linear adjacency with the verb. Evidence for P-incorporation into v comes from non-adjacency between the verb and the goal DP. In particular, P-drop in JA is possible in ditransitives, as seen in (51).

One advantage of P-incorporation is that it can explain the absence of P-drop with manner of motion verbs in JA. den Dikken (Reference den Dikken2010) assumes that manner of motion verbs have a manner head adjoined to v. Thus, incorporation of a silent P is blocked with manner of motion verbs, since it would violate the ban on multiple adjunction to a single host (Kayne Reference Kayne1994).

The second question to address is how the goal DP gets case. Based on the nonadjacency effects observed above (and also in section 5), it is safe to say that the case assigned to the goal DP in JA has to be structural. I adopt Biggs’s (Reference Biggs2014) Agree-based account of structural case assignment in Liverpool English. Biggs (Reference Biggs2014) argues that the higher p (K for her) is endowed with a set of unvalued [u-ϕ ] features, similarly to v in the verbal domain.Footnote 25 The unvalued features of p are valued via Agree with the goal DP in its c-command domain, and in return, p assigns case to the goal DP which bears an unvalued case feature [u-Case]. Agree between p and the goal DP is illustrated in (52).

  1. (52)

Before concluding this section, it is worthwhile highlighting the similarities/differences between ʕaNULL in JA, and TO and K in Ormskirk and Liverpool English respectively.

Under the analysis developed in this section, ʕaNULL is similar to Myler’s (Reference Myler2013) TO in that both heads are null Ps that theta-select the goal DP as their complement, but do not assign case to it. Moreover, both ʕaNULL and TO are licensed via P-incorporation into v. The difference between JA and Ormskirk with regard to Pdrop is in the way the goal DP is assigned case. In particular, Myler (Reference Myler2013) argues that the goal DP in Ormskirk is similar to regular direct objects, and as such, the DP raises to spec-vP for accusative case assignment. On the other hand, I argued above that the goal DP is assigned case via Agree with p.

As for the similarities/differences between P-drop in JA and Liverpool English, I argued above that ʕaNULL theta-selects its complement. For Biggs (Reference Biggs2014), however, the structure of the PP in P-drop lacks the lower lexical P found in a split p/P structure. Biggs’s K is a semantically bleached case head that corresponds to Svenonius’s (Reference Svenonius, Reuland, Bhattacharya and Spathas2007) higher p. This means that K does not theta-select its complement. The view that K is not a lexical head explains why P-drop in Liverpool English is possible with both directional to and stative at. Another difference between JA and Liverpool English is that ʕaNULL in JA is licensed via P-incorporation into v, whereas in Liverpool English, K is licensed in situ. This difference could potentially explain the availability of P-drop with manner of motion verbs in Liverpool English, and its absence in JA. Finally, both JA and Liverpool English are similar in the case licensing mechanism involved in P-drop. In particular, the goal DP in P-drop is assigned case in situ via Agree with the higher p in JA, which corresponds to K in Liverpool English.Footnote 26

7. Conclusion

In this article, I provided a detailed description of P-drop in JA. I showed that the main properties of P-drop found in other P-drop languages hold in JA. I refuted an analysis of P-drop in terms of pseudo noun incorporation (Gehrke and Lekakou Reference Gehrke and Lekakou2013, Hall Reference Hall2019), showing that the prepositionless goal noun in JA does not behave like a pseudo-incorporated nominal. I showed that the noun in JA P-drop exhibits the typical properties of a regular referential DP. I applied some diagnostics of PP-hood to show that there is a silent PP structure involved in JA P-drop. Thus, I argued for an analysis of P-drop in which a PP structure is present but unpronounced (Myler Reference Myler2013, Biggs Reference Biggs2014, Bailey Reference Bailey2018, among others). The silent P is licensed via incorporation of P into v (den Dikken Reference den Dikken2010), which in turn explains the absence of P-drop with manner of motion verbs. Also, the goal DP is assigned case in situ via Agree with a higher p head in a split p/P structure (Svenonius Reference Svenonius, Reuland, Bhattacharya and Spathas2007). P-drop is a relatively understudied phenomenon due to its restrictive nature. The arguments and analysis presented in this article contribute to the ongoing debate on the underlying mechanisms involved in P-drop.

Footnotes

An earlier version of this article was presented at the SynSem reading group at Queen Mary University of London. I would like to thank the audience at the SynSem reading group for their insightful comments. I also gratefully acknowledge the positive feedback of former CJL co-editor Daniel Siddiqi. Additionally, I thank three anonymous CJL reviewers for their helpful remarks and insightful suggestions.

1 Note that the preposition forms a clitic with the definite article in Greek.

2 Abbreviations used: AT: at; CS: Construct State; PNI: pseudo noun incorporation; TO: to.

3 In most dialects of English, P-drop is limited to nouns like home (Emonds Reference Emonds1985, Collins Reference Collins2007, Pearl and Caponigro Reference Pearl, Caponigro, Ashbury, Dotlačil, Gehrke and Nouwen2008).

4 The Arabic data used in this article are from JA, unless stated otherwise above any given example. The prepositions li- and ʕa- are used interchangeably to denote direction in JA, with ʕa being more widespread.

5 See Terzi (Reference Terzi2010) for an alternative analysis of Greek P-drop where she argues in favour of a silent PP structure.

6 Hall (Reference Hall2019) shows that proper names are allowed in P-drop in London English. According to him, this behaviour is very uncommon in pseudo-incorporated nominal contexts, since proper names are assumed to be full referential DPs. However, Hall follows Elbourne (Reference Elbourne2005) and Matushansky (Reference Matushansky2008) by assuming that proper names are definite descriptions that enter the syntax as simple predicates in a way similar to common nouns.

7 Hall (Reference Hall2019) notes that it is possible to force a wide scope reading for the goal noun when the preposition is overt. However, a wide scope reading for the goal noun is unavailable when the preposition is absent, as in (15).

8 As concerns the absence of pronominal anaphora as a diagnostic of PNI, Hall (Reference Hall2019) shows that Pdrop in London English can in fact support pronominal anaphora. In particular, he shows that it is possible to refer back to to the goal noun using the pronoun it. This is unexpected under a pseudo-incorporation analysis. Hall, however, questions the robustness of this diagnostic by referring to similar patterns observed with non-referential noun phrases (e.g., weak definites in English (Scholten and Guevara Reference Scholten and Guevara2010)), which can also support pronominal anaphora, similarly to the goal nouns found in P-drop.

9 Bailey (Reference Bailey2018) investigates P-drop in Southeast English. She shows that the determiner is obligatorily absent in P-drop contexts, and the verbs allowed in P-drop are always go or come. Also, the noun must be interpreted as a directional Goal, and well-establishedness of the goal noun in P-drop is also required in Southeast English. Bailey (Reference Bailey2018) argues in favour of a silent P, as in Myler (Reference Myler2013).

10 Myler (Reference Myler2013) motivates A-movement of the goal noun by showing that a goal noun in P-drop is allowed in gerunds, similarly to accusative objects. Also, the goal noun is disallowed in derived nominals. This behaviour is similar to ordinary direct objects, but is not similar to ordinary PP complements. See Myler (Reference Myler2013) for a detailed discussion of A-movement and accusative case assignment in P-drop.

11 Svenonius’s (Reference Svenonius, Reuland, Bhattacharya and Spathas2007) split p/P hypothesis is discussed in section 6.

12 The unavailability of P-drop with locative and source Ps in JA confirms Pearl and Caponigro’s (Reference Pearl, Caponigro, Ashbury, Dotlačil, Gehrke and Nouwen2008) observation that, cross-linguistically, only directional and unmarked stative prepositions can remain silent.

13 Note that the goal noun in (24a) and (24b) forms a Construct State (CS) with the following noun, thus explaining the absence of the definite article. The head noun in the CS cannot overtly realize the definite article, yet the noun is interpreted as definite (Ritter Reference Ritter and Rothstein1991, Borer Reference Borer, Lecarme, Lowenstamm and Shlonsky1996, Kremers Reference Kremers2003, Shlonsky Reference Shlonsky2004: among many others).

14 As mentioned above, absence of pronominal anaphora is not a robust diagnostic for PNI. Other nonreferential DPs can also support pronominal anaphora (e.g., weak definites in English; see Scholten and Guevara Reference Scholten and Guevara2010).

15 Modification of the goal noun by intersective modifiers and restrictive relative relatives clauses excludes the possibility of analyzing the noun as a “weak definite” (Carlson et al. Reference Carlson, Sussman, Klein and Tanenhaus2006, Aguilar Guevara and Zwarts Reference Aguilar Guevara and Zwarts2013).

16 See Rothschild (Reference Rothschild2007) for a detailed discussion of the scope properties of both definite and indefinite descriptions.

17 It is generally accepted that overt PPs are different from their silent counterparts. Polinsky (Reference Polinsky2016), for instance, shows that overt and silent PPs have different extraction and subextraction properties.

18 In JA, comitative PPs (e.g., maʕ ‘with’) cannot appear before the verb, even in the presence of an overt P (i).

This seems to suggest that comitative PPs occupy a position lower that T0, given the mainstream assumption that verbs in Arabic raise to T0 at least in past tense sentences (Aoun et al. Reference Aoun, Benmamoun and Choueiri2009).

19 In addition to its adverbial use, duɣri can be used as an adjective to describe an honest person, as seen in (i).

20 The adverb duɣri can sometimes be used without an overt PP with verbs like rawwaħ (went back), as in (i).

I do not have a concrete explanation for the acceptability of (i), but should point out that the absence of a PP with duɣri ‘straight’ is illicit with other verbs. I hypothesize that the semantics of the verb (or the verbalizing head v0) in (i) makes the sentence acceptable without an overt PP. In particular, the verb rawwaħ (went-back) is used only when referring to the action of going back home. The sentence in (ii), for instance, can only mean that I’m going back home, and not to some other place.

21 Some of my JA informants preferred to assign emphatic stress on duɣri ‘straight’ when it occurs in medial position, namely in (43b) and (45b). For those informants, stress assignment happens irrespective of the presence or absence of an overt preposition.

22 den Dikken (Reference den Dikken2010) argues that directional PPs have a fine-grained structure consisting of PDir and PLoc. den Dikken motivates the presence of PLoc in directional PPs by showing that locative Ps can sometimes be used in directional contexts. According to him, PLoc remains null in the presence of an intrinsically directional P. Motivating the presence of PLoc in directional PPs in JA is a task beyond the scope of this article. Nonetheless, one could entertain the possibility that PLoc is underlying in the structure of all intrinsically directional Ps in JA. For the purpose of the current discussion, I maintain that PDir immediately theta-selects the goal DP as its complement.

23 For ease of exposition, I put aside the issue of introducing the subject (i.e., the Figure) in spec-p, since it has no crucial effect on the licensing of ʕaNULL. I follow Jarrah (Reference Jarrah2017), who argues that the pre-verbal position of the thematic subject in JA is derived via movement of the subject from Spec-vP to Spec-SubjP (Rizzi Reference Rizzi, Brugé, Giusti, Munaro, Schweikert and Turano2005).

24 den Dikken (Reference den Dikken2010) argues that in directional PPs, an overt PDir can license a null PLoc in its command domain. A null PDir, on the other hand, is licensed via incorporation of PDir into v.

25 See Kayne (Reference Kayne2005) and Řezáč (Reference Řezáč, Harbour, Adger and Béjar2008) for evidence that adpositions can be Probes, carrying unvalued [ϕ] features.

26 I have nothing to say on P-drop in Greek and London English, but should point out that alternative Silent-P analyses of P-drop in Greek have already been proposed by Terzi (Reference Terzi2010) and Kouneli (Reference Kouneli2014). As for London English, it seems obvious that British dialects of English exhibit considerable variation with respect to P-(D)-drop. More research is needed to precisely identify the source of variation among British dialects of English in this regard.

References

Aguilar Guevara, Ana and Zwarts, Joost. 2013. Weak definites refer to kinds. Recherches linguistiques de Vincennes 42: 33–60.CrossRefGoogle Scholar
Aoun, Joseph E., Benmamoun, Elabbas, and Choueiri, Lina. 2009. The syntax of Arabic. Cambridge: Cambridge University Press.CrossRefGoogle Scholar
Bailey, Laura R. 2018. Some characteristics of Southeast English preposition dropping. Iberia: An International Journal of Theoretical Linguistics 10: 48–70.Google Scholar
Baker, Mark C. 1988. Incorporation: A theory of grammatical function changing. Chicago: University of Chicago Press.Google Scholar
Biggs, Alison. 2014. Dissociating case from theta-roles: A comparative investigation. Doctoral dissertation, University of Cambridge.Google Scholar
Borer, Hagit. 1996. The construct in review. In Studies in Afroasiatic Grammar, ed. Lecarme, Jacqueline, Lowenstamm, Jean and Shlonsky, Ur, 30–61. The Hague: Holland Academic Graphics.Google Scholar
Borik, Olga and Gehrke, Berit. 2015. An introduction to the syntax and semantics of pseudoincorporation. In The syntax and semantics of pseudo-incorporation, ed. Borik, Olga and Gehrke, Berit, 1–43. Leiden: Brill.CrossRefGoogle Scholar
Carlson, Greg, Sussman, Rachel, Klein, Natalie, and Tanenhaus, Michael. 2006. Weak definite noun phrases. In NELS 36: Proceedings of the thirty-sixth annual meeting of the North East Linguistic Society, vol. 1, 179. Citeseer.Google Scholar
Collins, Chris. 2007. Home sweet home. NYU Working Papers in Linguistics 1: 1–34.Google Scholar
Dayal, Veneeta. 2011. Hindi pseudo-incorporation. Natural Language & Linguistic Theory 29(1): 123–167.CrossRefGoogle Scholar
den Dikken, Marcel. 2010. Directions from the GET-GO: On the syntax of manner-of-motion verbs in directional constructions. Catalan Journal of Linguistics 9: 23–53.CrossRefGoogle Scholar
Elbourne, Paul D. 2005. Situations and individuals, vol. 90. Cambridge, MA: MIT Press.Google Scholar
Emonds, Joseph. 1985. A unified theory of syntactic categories. Dordrecht: Foris.CrossRefGoogle Scholar
Gehrke, Berit and Lekakou, Marika. 2013. How to miss your preposition. Studies in Greek Linguistics 33: 92–106.Google Scholar
Haddican, William and Holmberg, Anders. 2012. Object movement symmetries in British English dialects: Experimental evidence for a mixed case/locality approach. The Journal of Comparative Germanic Linguistics 15(3): 189–212.CrossRefGoogle Scholar
Hall, David. 2019. P-D-drop and Pseudo-incorporation in London English. https://ling.auf.net/lingbuzz/004771. [accessed 18-January-2021].Google Scholar
Ioannidou, Alexandra and den Dikken, Marcel. 2009. P-drop, D-drop, D-spread. In Proceedings of the 2007 workshop in Greek syntax and semantics at MIT, ed. Halpert, Claire Danielle, Hartman, Jeremy, and Hill, David, 393–408. Cambridge: Cambridge University Press.Google Scholar
Jarrah, Marwan Ali Saleem. 2017. Subject extraction in Jordanian Arabic. Doctoral dissertation, Newcastle University.Google Scholar
Johnson, Kyle. 1991. Object positions. Natural Language & Linguistic Theory 9(4): 577–636.CrossRefGoogle Scholar
Kamp, Hans and Partee, Barbara H.. 1995. Prototype theory and compositionality. Cognition 57(2): 129–191.CrossRefGoogle ScholarPubMed
Kayne, Richard S. 1994. The antisymmetry of syntax, vol. 25. Cambridge, MA: MIT Press.Google Scholar
Kayne, Richard S. 2005. Movement and silence. Oxford: Oxford University Press.CrossRefGoogle Scholar
Koopman, Hilda. 2010. Prepositions, postpositions, circumpositions, and particles: The structure of Dutch PPs. In Mapping spatial PPs: The cartography of syntactic structures, ed. Cinque, Guglielmo and Rizzi, Luigi, 26–73. Oxford: Oxford University Press.CrossRefGoogle Scholar
Kouneli, Maria. 2014. The syntax of silent locative prepositions in Greek. BA thesis, Yale University.Google Scholar
Kratzer, Angelika. 1996. Severing the external argument from its verb. In Phrase structure and the lexicon, ed. Rooryck, Johan and Zaring, Laurie, 109–137. New York: Springer.CrossRefGoogle Scholar
Kremers, Joost. 2003. The Arabic noun phrase: A minimalist approach. Doctoral dissertation, University of Nijmegen.Google Scholar
Levin, Theodore Frank. 2015. Licensing without case. Doctoral dissertation, Massachusetts Institute of Technology.Google Scholar
Longobardi, Giuseppe. 1994. Reference and proper names: A theory of N-movement in syntax and logical form. Linguistic Inquiry 25(4): 609–665.Google Scholar
Longobardi, Giuseppe. 2001. Formal syntax, diachronic minimalism, and etymology: The history of French chez. Linguistic Inquiry 32(2): 275–302.CrossRefGoogle Scholar
Massam, Diane. 2001. Pseudo noun incorporation in Niuean. Natural Language & Linguistic Theory 19(1): 153–197.CrossRefGoogle Scholar
Matushansky, Ora. 2008. On the linguistic complexity of proper names. Linguistics and philosophy 31(5): 573–627.CrossRefGoogle Scholar
Myler, Neil. 2013. On coming the pub in the North West of England: Accusative unaccusatives, dependent case, and preposition incorporation. The Journal of Comparative Germanic Linguistics 16(2–3): 189–207.CrossRefGoogle Scholar
Partee, Barbara H. 2010. 10: Privative adjectives: Subsective plus coercion. In Presuppositions and discourse: Essays offered to Hans Kamp, ed. B¨auerle, Rainer, Reyle, Uwe, and Zimmerman, Thomas Ede, 273–285. Leiden: Brill.Google Scholar
Pearl, Lisa and Caponigro, Ivano. 2008. Silent prepositions: Evidence from free relatives. In Syntax and semantics of spatial P, ed. Ashbury, Anna, Dotlačil, Jakub, Gehrke, Behrit, and Nouwen, Rick, Linguistics Today, vol. 120, 365–385. Amsterdam: John Benjamins Publishing.Google Scholar
Polinsky, Maria. 2016. Deconstructing ergativity: Two types of ergative languages and their features. Oxford: Oxford University Press.CrossRefGoogle Scholar
Řezáč, Milan. 2008. Phi-agree and theta-related case. In Phi-theory: Phi-features across modules and interfaces, ed. Harbour, Daniel, Adger, David, and Béjar, Susana, 83–129. Oxford: Oxford University Press.CrossRefGoogle Scholar
Ritter, Elizabeth. 1991. Two functional categories in noun phrases: Evidence from Modern Hebrew. In Syntax and semantics 25: Perspectives on phrase structure, ed. Rothstein, Susan, 37–62. New York: Academic Press.Google Scholar
Rizzi, Luigi. 2005. On some properties of subjects and topics. In Contributions to the XXX Incontro di Grammatica Generativa, ed. Brugé, Laura, Giusti, Giuliana, Munaro, Nicola, Schweikert, Walter, and Turano, Giuseppina, 203–224. Venice: Cafoscarina.Google Scholar
Rothschild, Daniel. 2007. The elusive scope of descriptions. Philosophy Compass 2(6): 910–927.CrossRefGoogle Scholar
Scholten, Jolien and Guevara, Ana Aguilar. 2010. Assessing the discourse referential properties of weak definite NPs. Linguistics in the Netherlands 27(1): 115–128.CrossRefGoogle Scholar
Shlonsky, Ur. 2004. The form of Semitic noun phrases. Lingua 114(12): 1465–1526. doi: 10.1016/j.lingua.2003.09.019.CrossRefGoogle Scholar
Svenonius, Peter. 2007. Arguments they introduce. In Argument structure, ed. Reuland, Eric J., Bhattacharya, Tanmoy, and Spathas, Giorgio, Linguistics Today, vol. 108, 63–103. Amsterdam: John Benjamins Publishing.CrossRefGoogle Scholar
Talmy, Leonard. 1975. Figure and ground in complex sentences. In Annual meeting of the Berkeley Linguistics Society, vol. 1, 419–430.Google Scholar
Terzi, Ahonto. 2010. On null spatial Ps and their arguments. Catalan Journal of Linguistics 9(1): 167–187.CrossRefGoogle Scholar