Abstract

Since Bowers 1993, it has been accepted that nonverbal small clauses are headed by a functional head, Pred0, whose function is to obligatorily mediate all nonverbal predication. I argue against this hypothesis by critically reanalyzing the original syntactic arguments for PredP, examining possible semantic support for mediated predication, and reviewing the putative crosslinguistic evidence for overt equivalence of Pred0. I first demonstrate that the facts originally taken as motivating a functional head in small clauses can now be accounted for by independently needed assumptions. I then show that standard Montagovian semantics treating NPs, APs, and PPs as unsaturated functions requires no mediating projection and that suggested alternative meanings for Pred0 either fail or cannot be used as motivation for its existence. Finally, I provide evidence that the syntax of copular particles and other “overt predicators” is different from that expected of Pred0 in such ways that they cannot be taken as prima facie evidence for it either. I sketch an alternative theory linking the use of predicative particles to nominal predication and provide evidence for it from crosslinguistic lexicalization patterns of copular particles. In sum, neither theoretical nor empirical considerations require a mediating functional head in small clauses, and therefore the PredP hypothesis should be abandoned.

1 The Structure of the Small Clause: An Introduction

Since Williams 1975 and Stowell 1981, 1983, it has been assumed that predication is possible in the absence of a verb in examples like (1a–c), but also those like (2a–b).

(1)

  • a.

    NP/DP predicate

    Alicei became [SC ti president/the head of the association].

  • b.

    AP/PP predicate

    This propositioni is/seems [SC ti preposterous/out of the question].

  • c.

    CP subject/ECM (exceptional case-marking) verb

    [CP That Jessie should fight]iwas considered [CP ti obvious].

(2)

  • a.

    With [this issue solved], we can turn to the real problem.

  • b.

    [Me mad]? Impossible!

The notion of a small clause has been proposed to describe such instances of propositional constituents not containing a verb or any (obviously visible) functional heads.1 The internal structure of small clauses was initially assumed to be simply that of the lexical predicate, with the subject appearing in the specifier of the maximal projection of the relevant lexical head (Stowell 1981, 1983).

(3)

graphic

Evidence for this view provided by Stowell (1981, 1983) comes from subcategorization: some verbs can combine with predicates only of a particular lexical category.

(4)

  • a.

    I expect [that man off my ship].

  • b.

    *I expect [that man very stupid].

(5)

  • a.

    We consider [it unlikely that John will win].

  • b.

    *I consider [John off my ship].

(6)

  • a.

    Alexandra proved [the theory false].

  • b.

    *I proved [the weapon in his possession].

(7)

  • a.

    We all feared [John killed by the enemy].

  • b.

    *We all feared [John unfriendly].

While for some of these cases semantics rather than lexical category may be the determining factor (see Bolinger 1972, Maling 1983, and Matushansky 2002 for the case of seem), for others no alternative explanations have been proposed. However, the hypothesis that small clauses are projections of their lexical heads ran into a number of theory-internal problems (discussed in detail below), which made a competing theory, due to Bowers (1993), all the more convincing. This proposal (further developed in Bowers 2001) argues that predication must be mediated by a functional head (originally Pr0, now Pred0), which has a semantic as well as a syntactic function.

(8)

graphic

In this article, I will argue that this theoretical move was a mistake and that no clear evidence can be provided for the presence of a functional head in small clauses. I will first (section 2) examine the original theory-internal arguments in favor of postulating PredP and show that either they are based on incomplete empirical generalizations or, in the current state of the theory, an independently available alternative explanation for the relevant facts can be found that does not require postulating PredP. I will then (section 3) examine the semantic motivation for Pred0 advanced by Bowers and demonstrate that it does not seem to be immediately clear what role the putative predication-mediating element should have. Next I will turn to the putative overt realizations of Pred0, focusing on copular particles (section 4) but also on other so-called overt predicators (section 5), and argue that their distribution is not that expected of Pred0 (as long as Pred0 is assumed to be present as a mediator in all small clauses). In sum, I will argue that there is no evidence for an obligatory functional head in small clauses and therefore a suitably Minimalist theory should not postulate one.

2 Theory-Internal Evidence for PredP

2.1 Against a PredP with Verbal Predicates

The core proposal advanced in Bowers 1993 and also adopted in Den Dikken 2006 is that all predication is mediated, and nonverbal predication is mediated by Pred0. If correct, this view provides a structurally unified analysis of the external argument as the specifier of PredP, for VP, NP, AP, and PP predicates. Before I proceed with the more standard perception of Pred0 as the head of nonverbal small clauses only, it therefore becomes necessary to provide a reason for not including verb phrases in the discussion of small clauses, contra Bowers 1993, which can now be regarded as a very clear and elegant argument for vP.

I contend that treating vP as a kind of PredP does not lead to any clear empirical generalizations. The hypothesis that the external argument of the verb must be introduced as the specifier of a dedicated functional head (vP or VoiceP) is based on a number of well-known distinctions between the semantics and the syntax of external vs. internal arguments (e.g., Harley 1995, Kratzer 1996, Marantz 1997), which either do not apply to nonverbal predication or have not been tested with it. While nonverbal small clauses with an AP, NP, or PP predicate all appear in many syntactic environments (complements of be (9), various raising and ECM verbs (10), the absolute construction (Van Riemsdijk 1978:62–86, Chung and McCloskey 1987) (11a), “Mad magazine” sentences (Akmajian 1984, Potts and Roeper 2006) (11b), resultatives (12a), and perhaps depictives (12b–c), bare VPs (in English) are restricted to “Mad magazine” sentences and complements of let, have, help, the active make, modals, and perception verbs. While some languages allow small-clause complements of modals ( possible in most Germanic languages; see, e.g., Barbiers 1995, Van Riemsdijk 2002, van Dooren 2014), others, such as Russian, disallow infinitival complements of perception verbs. Both facts strongly suggest that the difference should be attributed to c-selection rather than semantics and therefore that verbal and nonverbal complements do not belong to the same lexical category.

(9)

  • Copular clause

  • Sami is [SC tisad].

(10)

  • a.

    Raising/Stative

    Sami seems [SC timad].

  • b.

    Raising/Dynamic

    Sami became [SC timad].

  • c.

    Causative

    Sam made [SC Lee mad].

  • d.

    Denominative

    The people elected [SC Sam (??the) president].

  • e.

    Naming

    Carroll named [SC his heroine Alice].

  • f.

    Perception

    Kim saw [SC Sam mad].

(11)

  • a.

    Absolute construction

    [With John sick], we’ll never get the job done on time.

  • b.

    Mad magazine” sentence

    [Me mad]?! Ridiculous!

(12)

  • a.

    Resultative

    We painted [SC the room green].

  • b.

    Object depictive

    Sam ate the meatk [SC PROkraw].

  • c.

    Subject depictive

    Sami ate the meat [SC PROinude].

The fact that bare VPs are incompatible with copular and semicopular verbs/particles further suggests that nonverbal small clauses share properties that bare verb phrases do not have. The ability of v0/Voice0 to introduce an external argument is intimately linked to its ability to assign accusative case (Burzio’s Generalization), which is systematically not true for nonverbal predication. Finally, the external argument of a verb can be suppressed in passives or middles, but no effect of this kind is observed with nonverbal predicates.

This final striking difference between the external argument of verbs and the subject of nonverbal predicates might be taken as the first piece of evidence against treating the latter as the specifier of a functional head. However, it suffices for present purposes to accept, as was tacitly done in the literature on small clauses as PredPs, that it is only nonverbal predication that is mediated by Pred0; beyond Bowers’s and Den Dikken’s work, it has never been proposed that v0/Voice0 should be treated as a kind of Pred0, and the arguments given above reason against doing so independently of the question of whether nonverbal predication requires mediation. Further arguments against unifying v0/Voice0 and Pred0 can be found in Baker 2003a:37–39.

Having thus disposed of verbal predicates in small clauses, I now turn to dismantling the evidence provided for the presence of a functional head in nonverbal small clauses.

2.2 Coordination of Unlikes

The first argument that Bowers (1993) presents in favor of a functional head in small clauses comes from examples like (13a–b), where constituents of different lexical categories are coordinated despite the cited general prohibition on coordinating unlikes (examples (14a–b) via Sag et al. 1985; for other nonmainstream solutions, see Bayer 1996, Daniels 2002). The hypothesis that a functional head is present in the small clause resolves the issue, turning it into a coordination of two PredPs.

(13)

  • a.

    I consider Fred crazy and a fool.

  • b.

    I consider Mary both shrewd and in the know.

(14)

However, as noted already in Schachter 1977 and Maling 1983 (citing Dik 1968 and Peterson 1981), projections of different lexical heads may in fact be coordinated: in cases of adverbial modification (15), as well as in some more-complex cases, such as those in (16).

(15)

  • a.

    The surgeon operated slowly and with great care.

  • b.

    John ate quickly and with good appetite.

(16)

Schachter (1977) and later Whitman (2004) show that it is the semantics of the coordinated constituents that restricts coordination: the coordinated constituents should have the same semantic type. Once this constraint, independently required for proper composition, is taken into account, coordination of two predicates (or of two event modifiers) is naturally allowed because the conjuncts have the same type (⟨e, ⟨s, t⟩⟩ for predicates, ⟨v, ⟨s, t⟩⟩ for event modifiers), whereas ungrammatical cases fall out automatically: an entity-denoting NP or a proposition-denoting CP cannot be coordinated with an AP or a PP, as an argument, as a predicate, or as an adjunct.

The coordination of projections of distinct lexical categories may raise the question of assigning a (conventional) label to the constituent formed by the coordination of X′ and Y′. However, this question can easily be answered in the current state of the linguistic theory by appealing to the hypothesis that the conjunction itself is projected as a CoordP (for evidence that conjunctions project a binary structure, see, e.g., Schachter 1977, 1984, Munn 1987, 1993, Zhang 2009).

Summarizing, in the current state of the linguistic theory the projection problem can be resolved by assuming that the label is determined by the coordinating element (or, as suggested in the most recent work on labeling (e.g., Chomsky 2013, 2014), constructed from shared features), and the independently required semantic prohibition on coordinating items of different semantic types accounts for the fact formerly explained by the projection of PredP.

2.3 Movement

As Svenonius (1994) notes, the PredP hypothesis accounts for two more properties of the small clause that do not seem to be compatible with Stowell’s (1981, 1983) analysis on theoretical grounds: the position of its subject and the movement of its predicate. Starting with the latter, Williams (1983) observed that the ability of the small-clause predicate to move makes it a maximal projection. His theoretical reasoning relied on the fact that the small-clause predicate can be targeted by wh-movement, as in (17). If the subject of a small clause were located in its specifier (Spec,AP in (17)), it would have to be stranded to achieve the word order in (17), making a nonmaximal projection the target of the movement operation, which is prohibited by the constraint against movement of intermediate projections (X′ levels) in X-bar theory (a constraint that is now replaced by the prohibition on the movement of segments).2

(17)

  • a.

    What does John consider [AP Bill [A′what]]?

  • b.

    How do you want [AP your eggs [A′how]]?

  • c.

    How famous did the incident make [AP the criminal [A′how famous]]?

As Svenonius demonstrates, introducing PredP resolves this problem: the subject remains in Spec,PredP and what moves is then the AP in (17)—and, in general, a maximal projection.

(18)

  • a.

    What does John consider [PredP Bill [Pred′ Pred0 [APwhat]]]?

  • b.

    How do you want [PredP your eggs [Pred′ Pred0 [APhow]]]?

  • c.

    How famous did the incident make [PredP the criminal [Pred′ Pred0 [APhowfamous]]]?

In the current state of the syntactic theory, however, the problem of the stranded specifier is circumvented by independent developments—more specifically, by the general adoption of Postal’s (1974) raising-to-object analysis (see Runner 2006 for discussion). Raising-to-object analyses assume that the subject of a small-clause/infinitival complement of a verb moves into Spec,VP, with the verb raising to a higher head (e.g., Johnson 1991, Lasnik and Saito 1991, Koizumi 1993, Lasnik 1995). Under this analysis, (17a–c) can be analyzed as involving the movement of the entire small clause, still containing the trace of the subject.

(19)

graphic

In the Minimalist trend, we should also ask whether the prohibition against moving segments has any empirical support. Given the increase in the number of functional heads since the introduction of this constraint, a reevaluation seems to be in order.

2.4 Multiple Specifiers

As argued by Williams (1983), placing the subject of predication into the specifier of the lexical head gives rise to problems when nonbare predicates (20) are considered. Anglo-Saxon possessors (20a) are generally thought to occupy Spec,DP, which renders this position unavailable for the small-clause subject. Degree modification gives rise to the same issue: measure phrases, as in (20b–c), are generally thought to occupy Spec,DegP, which strongly suggests that the highest (functional) specifier in the extended projection of the predicate is filled, regardless of the structure adopted for comparatives. Thus, under one analysis (e.g., Bowers 1975, Jackendoff 1977, Heim 2000, Bhatt and Pancheva 2004), DegP is located in Spec,AP, leaving no free specifier for an AP-internal subject. Conversely, under the alternative analysis (e.g., Abney 1987, Bowers 1987, Corver 1990, 1991, 1997a,b, Matushansky 2013), Deg0 (-er) takes AP as its complement and Spec,DegP is filled with either the differential or the comparative clause, likewise leaving no empty specifier for the subject. Making small clauses projections of Pred0 solves the problem, since the subject can be merged in Spec,PredP.

(20)

  • a.

    I consider Josiah her father’s best friend.

  • b.

    Set the pole 15 inches to the right.

  • c.

    Ayelet was 2 inches taller than her friends.

However, the theory-internal prohibition against multiple specifiers has become obsolete for a number of theoretical and empirical reasons. From the theoretical point of view, such a prohibition seems like pure stipulation: if one maximal projection can merge with another maximal projection (which is required to enable the merger of the specifier), it is unclear why another such merger should be ruled out.3 To provide some precedents, multiple specifiers have been proposed to deal with multiple nominative constructions in Japanese and Korean (e.g., Fukui 1986, 1995, Heycock and Lee 1989, Heycock 1993, Koizumi 1995, Takahashi 1996, Ura 1996, Hiraiwa 2001) and multiple accusative constructions in Korean (Yoon 1989, 1990; see also Maling 1989); Chomsky (1995) assumes multiple specifiers to account for there-insertion; and multiple CP specifiers seem to be unavoidable when it comes to multiple wh-fronting (Rudin 1988). In the generally assumed phase-based approach to derivation, to enable movement out of the vP phase (Chomsky 2000) it must be postulated that vP has specifiers in addition to the thematic specifier hosting the subject.

It could be objected that we are dealing here with multiple nonthematic specifiers, whereas the subject position is necessarily thematic. However, empirical or conceptual reasons remain to be found for using this distinction to exclude multiple thematic specifiers, and the privileged status of a small-clause subject can be made to follow from the fact that it is the last such specifier.

2.5 Expletives in Small Clauses

As also observed by Svenonius (1996) and Grohmann, Drury, and Castillo (2000), nonthematic and expletive subjects can appear in small clauses (21)–(22), suggesting the presence of a functional head whose specifier must be filled as a result of the EPP requirement. However, Pred0 cannot be this head, given how the EPP is understood to interact with θ-theory in the current state of the linguistic theory. Specifically, to be able to be filled by an expletive, an EPP position should not be assigned a θ-role; however, this is not the case for Spec,PredP, as most theories of mediated predication also assume that Pred0 has a semantic role to play (see below). If Pred0 θ-marks the external argument (as in Bowers’s (1993) view), its specifier is a thematic position and so nonthematic subjects should not be introduced in Spec,PredP.

(21)

  • a.

    I want it cold (when I go skiing).

  • b.

    I saw some students shouted at.

  • c.

    I’ve never seen there be so many problems.

    (Svenonius 1996:499)

(22)

Grohmann, Drury, and Castillo propose to account for all of these cases by assuming that the subject of the small clause undergoes raising to object: it is Spec,VP of the main verb that contains these nonthematic subjects. However (Eddy Ruys, pers. comm.), this explanation cannot account for raising in the absolute construction (23a), as well as in Headlinese (23b), as in both of these cases no embedding verb is present and hence there is no Spec,VP to host a raised subject.

(23)

While these facts would seem to entail the presence of a functional projection in small clauses whose specifier would be targeted by the raised subject, this functional projection cannot be equated with PredP, since Spec,PredP is by hypothesis a thematic position.

2.6 Predicate Case

The PredP hypothesis has also been used as an explanation for predicate case marking. As the following examples show, NP and sometimes AP predicates can be case-marked, and the case on the predicate is subject to crosslinguistic variation:

(24) Russian

  • a.

    Ja sčitaju ee otličnoj lingvistkoj.

    I consider her.ACC great.INS linguist.INS

    ‘I consider her a great linguist.’

  • b.

    Ona vernulas’ aspirantkoj.

    she returned PhD.candidate.INS

    ‘She came back a PhD candidate.’

(25) Arabic

  • a.

    salma ʕayyanat walad-a-ha wazir-an.

    Salma nominate.CAUS.PFV child-ACC-her minister-ACC

    ‘Salma nominated her child to be a minister.’

  • b.

    walad-u-ha ʕuyina wazir-an.

    child-NOM-her nominate.PASS.PFV minister-ACC

    ‘Her child was nominated to be a minister.’

It seems altogether reasonable to assume (see, e.g., Bailyn and Rubin 1991 for Russian; but see also Abu-Joudeh 2013, where this analysis is rejected for Modern Standard Arabic) that the case on the predicate in (24)–(25) is assigned by Pred0. Predicate case can therefore be used as indirect evidence for the PredP hypothesis: a case-assigning functional head must be assumed anyway to account for the fact of case assignment, and placing this head between the subject and the predicate of a small clause, as is assumed for Pred0, straightforwardly explains why the small-clause subject does not intervene for predicate case assignment.

However, the facts are more complicated than the PredP hypothesis warrants. As far as I have been able to ascertain, in languages with a marked predicate case it is never only this case that surfaces on all predicates. In fact, there is systematically at least one environment where the predicate surfaces with the unmarked case: present tense copular clauses (26) that can, nonetheless, be shown to involve predication rather than equation (see Matushansky 2010 for a discussion of Russian).

(26)

  • a.

    Russian

    Vera assistent/*assistentom.

    Vera assistant.NOM/INS

    ‘Vera is an assistant.’

  • b.

    Arabic

    Zaydun waziirun/*waziiran.

    Zaydun.NOM minister.NOM/ACC

    ‘Zaydun is a minister.’

    (Maling and Sprouse 1995:182)

One strategy available for a PredP proponent would be to claim that examples like (26a–b) involve a non-case-assigning Pred0 (see, e.g., Bailyn 2001, 2002, Pereltsvaig 2007), yet such a move clearly undermines the role of predicate case as an argument for Pred0, as it requires postulating two lexical items with identical semantics and identical phonology (or rather, lack thereof) that can nonetheless be l-selected by the embedding verb. The situation is further complicated by the fact that in Finnish, Estonian, and Hungarian the nonverbal predicate can be marked by as many as five cases (in Hungarian), as a function of the environment (Fong 2003, Matushansky 2012a). Given that it is the least syntactically complex environment (depending on the language: present tense copular clauses, all copular clauses, the complement of be and become, raising verbs) that gives rise to the unmarked predicate, it seems reasonable to assume that the source of the predicate case lies outside the small clause itself.

To deal with this issue, I have proposed elsewhere (Matushansky 2012a) that morphological case realizes not individual features (e.g., [accusative]) but bundles of them (e.g., [transitive; predicative; change of state]), with different predicate cases realizing different combinations (with, e.g., the Finnish translative realizing the complex [predicative; change of state] and the Estonian translative realizing the complex [predicative; transitive]). Regardless of the precise implementation of this view, if we accept its key point—namely, that various predicate cases realize the feature [predicative] in combination with others or under various conditions—we can depart from the presupposition in Matushansky 2012a that this feature is assigned by Pred0. If we assume instead that the feature [predicative] is interpretable (i.e., is present on the predicate because it is a semantic predicate), then predicate case marking need not depend on case assignment by a functional head and therefore need not indicate the presence of such a head.

2.7 Summary

The hypothesis that small clauses are projections of a functional head whose specifier hosts their subject has been argued to successfully resolve a number of problems arising from the prior assumption that small clauses are (extended) projections of the lexical head. Since these problems may now be resolved in an independent fashion, motivation for Pred0 should come from its observable contribution: either on the semantic side (and then a proper lexical entry must be provided for it) or on the morphophonological side, where it would be visible as an independent lexical item. In the next section, I will discuss the potential semantic contribution of Pred0, postponing putative direct morphosyntactic evidence for it until section 4.

3 The Semantics of Pred0

It is a standard assumption in model-theoretic semantics starting from Montague that APs, NPs, and PPs denote unsaturated functions from entities to propositions (semantic type ⟨e, ⟨s, t⟩⟩). As a result, the interpretation of examples like (27) involves simple Function Application of the predicate to the subject, with the copula either semantically vacuous, as is often assumed (Frege 1892, Meillet 1906, Benveniste 1966; more recently, e.g., Pollard and Sag 1994, Heycock and Kroch 1999), or manipulating nonthematic argument slots, such as possible worlds, times, or events (e.g., Bierwisch 1988, Kamp and Reyle 1993, Rothstein 1999, Maienborn 2005a,b).

(27)

  • Kim is a student/tall/in Dublin.

  • [TPRES [be [Kim a student/tall/in Dublin]]]

    • a.

      [λx λw . x is a student/tall/in Dublin in the possible world w] (Kim)

    • b.

      {w : Kim ∈ {x : x is a student/tall/in Dublin in w}}

Maintaining this assumption obviously leaves Pred0 without a semantic role. While a number of syntacticians would take no issue with the hypothesis that phonologically null Pred0 makes no contribution on the LF side, it is generally asserted that Pred0 is required to mediate predication. Since the notion of predication is based in semantics, it seems reasonable to assume that the role of Pred0 is too. There are two possible reasons why Pred0 should be required to mediate the relation between an AP, NP, or PP predicate and its subject: either it enables the AP, NP, or PP to license the subject (thematic licensing) or it is required to permit the resulting proposition to combine with the embedding verb or tense. I will now show that neither of these two approaches can be motivated on semantic grounds.

3.1 A Change in Basic Type

To provide a semantic role for Pred0, Bowers (1993) proposes that APs, NPs, and PPs do not denote predicates (sets), but rather must be converted into predicates. The semantic function of Pred0 is therefore to create a predicate that can be combined with the subject. Appealing to the work of Chierchia (1985) and Chierchia and Turner (1988), Bowers proposes that APs, NPs, and PPs are introduced into the derivation as entities (semantic sort π) and that Pred0 converts these entity-correlates into propositional functions (type⟨e, p⟩).4 As such, they can themselves function as arguments (28).

(28)

While providing a semantic role for Pred0, the hypothesis that APs, NPs, and PPs are not underlyingly predicates gives rise to serious problems of composition. Starting with APs, the assumption that they have the semantic type π necessarily entails that they cannot be used as modifiers in cases like red books, and the same issue arises with PPs. For NPs, the situation is exacerbated because the semantic type of bare nouns becomes open to debate. While examples like (28b) suggest that they can or should also be treated as entity-denoting (semantic type π), additional mechanisms would then become necessary to enable them to combine with determiners, possessive phrases, and so on. In other words, under this view APs, NPs, and PPs would acquire the semantic type that precludes their linguistic use without further refinements, forcing the use of Pred0 in the predicate position and of some additional mechanism NP-internally. Not only does this approach look completely artificial, it also requires that the entire mechanism of NP-internal computational semantics be revised. Clearly, such a global overhaul of contemporary model-theoretic semantics can only be proposed if sufficiently motivated by other considerations, which I will show to be missing.

3.2 The Eventuality Argument

The alternative treatment of Pred0 would be to assume the standard ⟨e, ⟨s, t⟩⟩ type for APs, NPs, and PPs, which would enable NP-internal composition to proceed as usual, and to hypothesize instead that the role of Pred0 is to introduce the Davidsonian eventuality argument, which is otherwise missing from a nonverbal predicate. There are two reasons to reject this hypothesis. The first is theoretical: Bierwisch (1988), Kamp and Reyle (1993), Rothstein (1999), and Maienborn (2005a,b) suggest that it is the copula that has this role, which, given the close connection between eventualities and verbs, seems appropriate. The second reason is conceptual: linking Pred0 to Davidsonian eventualities does not predict that it should be obligatory, contrary to the standard view that PredP is an obligatory mediator.

To make the second reason explicit: the first and foremost property of a predicate is that it should be able to combine with an argument. It should therefore be possible to produce a true small-clause structure (P(x)) where a proposition is constructed without the introduction of an eventuality argument. This would be correct for be, which can indeed be omitted, but not for Pred0, which is claimed to be required for predication. In other words, even if there is a functional head introducing an eventuality argument, with this semantics its obligatoriness is still stipulative.5

3.3 More Than One Pred0

The hypothesis that Pred0 is necessary for predication on semantic grounds seems untenable, because the current state of model-theoretic semantics does not warrant postulating either a functional head that creates predicates out of entity-denoting APs, NPs, and PPs or one that introduces an event/state argument. Yet there is one more possibility to dismiss: that Pred0 exists but makes different semantic contributions when combining with AP, NP, and PP predicates. This is in fact the hypothesis that Baker (2003a) advances (without, however, going into detail). It could be postulated that, while Pred′ is uniformly a predicate (semantic type ⟨e, ⟨s, t⟩⟩), different terminal nodes realize Pred0 for different lexical categories. An additional advantage of this view is that it is natural to have several lexical entries for the same functional label. However, as reasonable as this proposal is, the semantic theory itself does not motivate appealing to it—in other words, it does not provide direct evidence for Pred0. Quite on the contrary, in fact: it requires a modification of standard assumptions regarding the semantic type of APs, NPs, and PPs, which does not seem to be motivated yet by anything. In other words, for the time being Pred0 cannot be argued to be motivated by theory-internal syntactic considerations (see section 2), nor is it required by semantics (i.e., either it must be semantically null, or the lexical entries for various instantiations of Pred0 and the underlying nonpredicative semantic types for APs, NPs, and PPs have yet to be determined).

Given that in most languages discussed so far Pred0 is also phonologically null, this lack of semantic contribution is highly suspicious: we do not expect a functional head to be vacuous at both interfaces.6 This is why it has been proposed that Pred0 can in fact have overt realizations. In the next section, I will argue that copular particles, frequently claimed to instantiate Pred0, do not in fact provide any evidence for it. First, I will argue that not all copular particles can in principle be used as such evidence: they need to appear in nonfinite small clauses (i.e., not be restricted to copular clauses). Then, I will demonstrate that even such copular particles are subject to additional constraints that raise doubts about whether their role is indeed that of an obligatory mediator for predication. Other potential Pred0 candidates—a heterogeneous class of functional items, including as and its crosslinguistic equivalents (Danish som, French comme, Russian kak, German and Dutch als, etc.), for and its crosslinguistic equivalents (Russian za, Dutch voor, etc.), and Russian v ‘in’—will be examined in section 5.

4 Copular Particles as Lexicalizations of Pred0

4.1 The Expected Distribution of Pred0

In many languages a functional element, which I will gloss uniformly as CPRT, appears between the subject and (some categories of) the predicate (29)–(30), leading Bowers (1993, 2001) to hypothesize that it realizes Pred0. If such an element has the distribution expected of Pred0, according to Bowers it provides independent evidence for it.7 We will see that such is not the case.

(29) Welsh

  • a.

    Mae Siôn *(yn) ddedwydd.

    is Siôn CPRT happy

    ‘Siôn is happy.’

  • b.

    Y mae Siôn yn feddyg.

    PRT is Siôn CPRT doctor

    ‘Siôn is a doctor.’

    (Rouveret 1996:128)

(30) Edo

  • a.

    Èmèrí *(yé) mòsèmòsè.

    Mary CPRT beautiful.A

    ‘Mary is beautiful.’

  • b.

    Úyì *(rè) òkhaèmwèn.

    Uyi CPRT chief.N

    ‘Uyi is a chief.’

    (Baker 2003a:40)

(31) Chichewa

  • a.

    M-kango *(ndì) w-a u-kali.

    3-lion CPRT 3-ASSOC 3-fierce

    ‘The lion is fierce.’

  • b.

    M-kango *(ndì) m-lenje.

    3-lion CPRT 1-hunter

    ‘The lion is a hunter.’

    (Baker 2003a:44)

Can these elements be taken as evidence for a functional head mediating predication? To provide such evidence, they would need to have the distribution expected of Pred0; that is, they would need to appear in all canonical small clauses ((9)–(12)). The Welsh copular particle yn shows just this kind of behavior, appearing in copular clauses (32), in small-clause complements of ECM and raising verbs (33), in resultatives (34a), in depictives (34b), in the absolute construction (35), with naming verbs (36), and in NP-internal reduced relatives (37), and could therefore be taken to instantiate Pred0.

(32)

  • a.

    Mae Siôn *(yn) ddedwydd.

    is Siôn CPRT happy

    ‘Siôn is happy.’

  • b.

    Y mae Siôn yn feddyg.

    PRT is Siôn CPRT doctor

    ‘Siôn is a doctor.’

    (Rouveret 1996:128)

(33)

  • Rydw i’n ystyried [Siôn yn niwsans].

  • am I+PROG consider Siôn CPRT nuisance

  • ‘I consider Siôn a nuisance.’

  • (Zaring 1996:139)

(34)

  • a.

    Peintia’r petryal bach yn goch.

    paint.IMP+the rectangle small CPRT red

    ‘Paint the small rectangle red.’

  • b.

    Dw i’n licio cwrw yn oer.

    be.1SG I+PROG like beer CPRT cold

    ‘I like beer cold.’

    (Bob Morris Jones, pers. comm.)

(35)

  • A mi yn ofnus, ni ddywedais ddim.

  • and I CPRT shy NEG said nothing

  • ‘Since I am shy, I said nothing.’

  • (Rouveret 1996:144)

(36)

  • Enwyd ef yn Siôn ar ôl ei dad.

  • name.PASS he CPRT Siôn after his father

  • ‘He is named Siôn after his father.’

  • (Matushansky 2008:582)

(37)

  • buddsoddi ym mhensaernïaeth fy ngwlad, yn hen ac yn newydd

  • invest.VN in architecture my country CPRT old and CPRT new

  • ‘to invest in the architecture of my country, old and new’

  • (Willis 2003:23)

Conversely, in a large number of languages a pronominal or demonstrative-like element appears to mark identity statements, generic attributions, or classificatory assertions (Berman and Grosu 1976, Doron 1983, 1986, Rapoport 1987, Rothstein 1995, Sichel 1997, Greenberg 1998, 2002, Heller 2002 for Hebrew, Soschen 2003, Rutkowski 2006, Citko 2008 for Polish, Markman 2008 for Russian, Wondem 2014 for Geez; see Li and Thompson 1977, Diessel 2009, Lohndal 2009, and van Gelderen 2011 for the general crosslinguistic phenomenon of the diachronic demonstrative-to-copula conversion).

In the next sections, I will differentiate between nonverbal copulas (which appear in the absence of a lexical verb or only with the copula be) and predicative particles (which appear in finite clauses in other small-clause environments). The former will be exemplified by two languages where they have been argued to instantiate Pred0: Chichewa and Edo (Baker 2003a). After showing that they cannot be taken as theory-independent evidence for Pred0, I will turn to languages with the latter: Eastern Riffian and Welsh. I will show that these languages also offer no evidence for Pred0 as an obligatory mediating functional head for predication. Finally, I will discuss crosslinguistic lexicalization patterns of both types of copular particles, showing that they are inextricably linked to NP predication, and I will hypothesize what role they might be playing that explains their distribution.

4.2 Bantu Copular Particles

Evidence from Chichewa features prominently in Baker’s (2003a) discussion of the PredP hypothesis. However, copular particles in Bantu languages cannot be taken as direct evidence for the existence of Pred0, because their distribution is restricted to only some instances of nonverbal predication:8 while in some Bantu languages (Zulu, Xhosa) copular particles appear only with NP predicates, in others (Swahili, Venda, Kinyarwanda) they are restricted to present tense copular clauses (and sometimes only to third person subjects). In Chichewa, these constraints are combined, allowing the copular particle only for NP predicates in present tense copular clauses. Kiso (2012) also discusses two languages closely related to Chichewa, Citumbuka and Cisena, where the copular particle appearing with NP primary predicates disappears in tenses other than the present; yet Kiso does not address its presence or absence with AP predicates.

(38) Zulu: Only with NP predicates (see also Posthumus 1978, 1988, 2006)

  • a.

    Ngi- mu- hle.

    AGRS1SG- AA1- beautiful

    ‘I am beautiful.’

  • b.

    Ngi- ngu- mfundisi.

    AGRS1SG- CPRT- 1.teacher

    ‘I am a teacher.’

    (Buell and de Dreu 2013:427)

  • c.

    Ng- a- be ngi- ngu- mfundisi.

    AGRS1SG- PAST- be AGRS1SG- CPRT- 1.teacher

    ‘I was a teacher.’

    (Buell and de Dreu 2013:427)

(39) Xhosa: Only with NP predicates

  • a.

    Ndi- za ku- ba ngu- mfundisi.

    AGRS1SG- FUTKU- COPCPRT1- 1.teacher

    ‘I will become a teacher.’

  • b.

    U- za ku- ba m- hle.

    AGRS1SG- FUTKU- COPAA1- beautiful

    ‘She/He will become beautiful.’

(40) Swahili: In present tense copular clauses, optional (Steere 1884/1930, Loogman 1965, Brauner and Herms 1986, Marshad and Suleiman 1991, McWhorter 1992)

  • a.

    Shati ni

    graphic
    - chafu.

    shirt5CPRTAA5- dirty

    ‘The shirt is dirty.’

  • b.

    Nguo zi safi.

    clothes10AGRS10 clean

    ‘The clothes are clean.’

  • c.

    Ali m- réfu.

    Ali AA1- tall

    ‘Ali is tall.’

    (Marshad and Suleiman 1991:30)

(41) Venda: With third person subjects in present tense copular clauses

  • a.

    Ni vhafunzi.

    AGRS2PL missionaries2

    ‘You are missionaries.’

  • b.

    Mutukana ndi mu- vhuya.

    boy1CPRTAA1- good-natured

    ‘The boy is good-natured.’

  • c.

    Mufunzi ndi tshihole.

    missionary1CPRT cripple7

    ‘The missionary is a cripple.’

  • d.

    Ndi do vha dokotela duvha linwe.

    AGRS1SGFUTCOP doctor day one

    ‘I will become a doctor one day.’

  • e.

    Nd- o- la nama mbisi.

    AGRS1SG- PAST- eat meat AA9.raw

    ‘I ate the meat raw.’

    (Pylkkänen 2002:34–35)

(42) Chichewa: Only with NP predicates in present tense copular clauses

  • a.

    M-kango *(ndì) w-a u-kali.

    3-lion CPRT 3-ASSOC 3-fierce

    ‘The lion is fierce.’

    (Baker 2003a:44)

  • b.

    M-kango *(ndì) m-lenje.

    3-lion CPRT 1-hunter

    ‘The lion is a hunter.’

    (Baker 2003a:44)

  • c.

    Nyumba-yi i-na-li yayikulu.

    house-DEMNCL.SBJ-RECENT.PAST-COP big

    ‘The house was big.’

    (Kiso 2012:84)

  • d.

    A-na-li mphunzitsi.

    3.SBJ-PAST-COP teacher

    ‘She/He was a teacher.’

    (Kiso 2012:25)

(43) Kinyarwanda: Only with third person subjects in present tense copular clauses

  • a.

    In-gofero ni mini.

    9-hat CPRT big

    ‘The hat is big.’

    (Jerro 2015:95)

  • b.

    Umu-curanzi njye nahuye nawe ni Michael Jackson.

    1-singer I met I CPRT Michael Jackson

    ‘The singer that I met is Michael Jackson.’

    (Jerro 2015:95)

  • c.

    Tw-a-li abana.

    1PL-PAST-COP children

    ‘We were children.’

    (Dekympe 2000:49)

  • d.

    Tu-ra-ba beza.

    1PL-FUT-COP beautiful

    ‘We will be beautiful.’

    (Dekympe 2000:49)

Summarizing, copular particles in Bantu languages are restricted to NP predicates (a predicative particle appearing in all small-clause environments) or to copular clauses (a nonverbal copula, for which further restrictions, such as present tense only, third person only, etc., are possible). Of course, the nonverbal copula could be argued to instantiate Pred0 and to disappear as a result of restructuring (Stowell 1991), which would take the form of Pred0-to-V0 head movement into the embedding verb (for which no morphological evidence is, however, available). But such reasoning already presupposes the existence of Pred0, which these facts must be reconciled with; they cannot be used as primary evidence. The same reasoning applies to the copular particle that appears only with NP predicates: as I will show in the next sections, while it is possible to hypothesize that Pred0 is overtly realized only with some lexical categories, this hypothesis is not supported by any independent evidence.

4.3 Edo Copular Particles

Another source of evidence for the PredP theory, according to Baker (2003a), is the Volta-Niger language Edo, where adjectival and nominal predicates are introduced by different copular particles. Arguing that the role of Pred0 is to introduce the external argument, Baker specifies that it does so differently for different lexical categories. As a result, different lexicalizations of Pred0 for AP and NP predicates are not unexpected.

(44)

  • a.

    Èmèrí *(yé) mòsèmòsè.

    Mary CPRTA beautiful.A

    ‘Mary is beautiful.’

  • b.

    Úyì *(rè)

    graphic
    kha
    graphic
    mw
    graphic
    n.

    Uyi CPRTN chief.N

    ‘Uyi is a chief.’

    (Baker 2003a:40)

It can be shown, however, that the copular particles of Edo cannot be taken as evidence for the PredP theory, either in its original form or in Baker’s version, for two reasons. First, the use of the copular particles is limited to copular clauses (potentially, only in the present tense), making them a nonverbal equivalent of be. Second, it can be demonstrated that the copular particle appearing with nonnominal predicates () forms part of verbal conjugation.

Starting with the former, both and are presented as verbs by Melzian (1937): “to be, indicating quality . . . [or] in a certain place” (p. 224)9 and “to be, with a noun predicate” (p. 180); is also used in identity clauses (Déchaine and Tremblay 2012), which clearly argues against relating it to predication.

(45)

  • a.

    Ùyì òré né!né

    graphic
    kha
    graphic
    mw
    graphic
    n.

    Uyi it.be the chief.N

    ‘Uyi is the chief.’

    (Baker 2003a:111)

  • b.

    Èvbáré òré Òzó lé.

    food.N it.be Ozo cook

    ‘It’s food that Ozo cooked.’

    (Baker 2003a:140)

Furthermore, as Baker himself notes, Edo copular particles disappear in resultatives. While under the causative verb ‘make’ is possible (although the inceptive particle dòó is required to render the sentence fully grammatical (46a)), in resultatives must be omitted, as in (46b) and (47) (see also (50a)).

(46)

  • a.

    Úyì yá èmát

    graphic
    n ?(dòó) yé p
    graphic
    rh
    graphic

    Uyi make metal INCEPCPRTA flat.A

    ‘Uyi made the metal be flat.’

    (Baker 2003a:42)

  • b.

    Òzó gbé èmát

    graphic
    n (*yé) p
    graphic
    rh
    graphic

    Ozo beat.PAST metal CPRTA flat.A

    ‘Ozo beat the metal, causing it to be flat.’

    (Baker 2003a:223)

(47)

  • Òzó gìá ìrhùnmwùn khéréé.

  • Ozo cut grass small.A

  • ‘Ozo cut the grass short.’

  • (Ogie 2009:256)

Though Baker claims that depictives are not possible in Edo, Ota Ogie (pers. comm.), while largely confirming this, provides (48a–b), also without the copular particle.

(48)

  • a.

    À bi

    graphic
    Èmérì
    graphic
    kha
    graphic
    mw
    graphic
    n.

    IMPRS give.birth.PAST Mary chief.N

    ‘Mary was born a chief.’

  • b.

    À bi

    graphic
    Èmérì mòsèè.

    IMPRS give.birth.PAST Mary beautiful.A

    ‘Mary was born beautiful.’

    (Ota Ogie, pers. comm.)

By the criteria adopted above, the lack of the copular particle in secondary predication10 means that Edo copular particles cannot be viewed as evidence for Pred0. While Baker attributes the distribution of the copular particle to incorporation of Pred0 into a higher head (cf. Stowell 1991), as discussed above, this hypothesis necessitates independent evidence for PredP, which I have shown to be lacking, or for incorporation.

Turning now to as a potential Pred0, it appears in the predicate position only with those adjectives that have a verbal equivalent, from which they differ by tonal pattern and reduplication. While Baker (2003a) glosses the verbal and adjectival variants of the same quality concept identically and Omoruyi (1986) claims that there is no noticeable difference between them (49)–(50), Ogie (2009) indicates that the adjectival variant is interpreted as a transitory state (51).

(49)

  • a.

    Òzó yé zùr

    graphic
    zùr
    graphic

    Ozo CPRTA foolish.A

    ‘Ozo is foolish.’

  • b.

    Òzó zùr

    graphic
    .

    Ozo foolish.V

    ‘Ozo is foolish.’

    (Baker 2003a:87)

(50)

  • a.

    Òzó kòkó Àdésuwa mòsèmòsè.

    Ozo raise.PAST Adesuwa beautiful.A

    ‘Ozo raised Adesuwa so that she was beautiful.’

  • b.

    Òzó kòkó Àdésuwa mòsé.

    Ozo raise.PAST Adesuwa beautiful.V

    ‘Ozo raised Adesuwa so that she was beautiful.’

    (Baker 2003a:227)

(51)

  • a.

    Òzó yè mòsèè.

    Ozo CPRTA beautiful.A

    ‘Ozo looks beautiful.’

  • b.

    Òzó mòsé.

    Ozo beautiful.V

    ‘Ozo is beautiful.’

    (Ogie 2009:162)

In addition to these low-tone predicatives that appear with the copular particle , Edo has a closed class of adjectives that are obligatorily attributive. While some of them cannot appear in isolation at all, others can, but only with an implied elided noun.

(52)

  • a.

    *

    graphic
    gb
    graphic
    n
    graphic
    ré Òsàr
    graphic
    d
    graphic
    rè.

    new it.is Osaro buy.PAST

  • b.

    graphic
    wi
    graphic
    graphic
    ré Òsàr
    graphic
    d
    graphic
    rè.

    old it.is Osaro buy.PAST

    ‘It is an old one that Osaro bought.’

    (Omoruyi 1986:299)

Omoruyi (1986) claims that parts of speech in Edo can be distinguished by the initial segment: all verbs are consonant-initial, whereas all nouns are vowel-initial. While some of these adjectives (e.g., (52)) are vowel-initial, like nouns, others are consonant-initial (like verbs, but without a verbal equivalent), as in (53). In neither case would their semantics need to be such as to preclude predicative use.

(53)

  • a.

    ìm

    graphic
    t
    graphic
    w
    graphic
    r
    graphic

    car long

    ‘long car’

  • b.

    èrhán túkpúrú

    tree short

    ‘short tree’

    (Omoruyi 1986:298)

If realizes Pred0, whose function is to turn an attributive adjective into a predicate, the correlation between the existence of a verbal form and the compatibility of its low-tone counterpart with is unexpected; moreover, it is unclear why Pred0 cannot freely make any potentially intersective adjective into a predicate. Conversely, if the only way of constructing a nonnominal predicate is by starting from a verbal stem, the particle can be viewed as a nonverbal copula c-selecting for APs that are already predicates, making the low-tone verbal forms participles or gerunds, which, as in English, can freely become attributive.

Summarizing, the distribution of Edo copular particles does not support the hypothesis that they lexicalize Pred0: not only do they appear only in copular clauses, but the ability of a given nonnominal stem to function as a predicate is determined by the existence of the corresponding verb, which shows that is not used to turn an AP into a predicate.

4.4 The Welsh Copular Particle

The first to treat the copular particle yn as Pred0 was Bowers himself (Bowers 1993, crediting Wayne Harbert). As examples (32)–(37) show, yn appears in all canonical small clauses. However, like other predicative particles crosslinguistically, it is obligatorily absent when the predicate is a PP (Jones and Thomas 1977:47, Jones 2009).

(54)

  • a.

    Mae Siôn (*yn) yn Lludain / o flaen y tŷ.

    is Siôn CPRT in London / of foremost the house

    ‘Siôn is in London/in front of the house.’

    (Zaring 1996:115)

  • b.

    A hwy yn yr eglwys, ysbeiliwyd eu tŷ.

    and them in the church was.looted their house

    ‘While they were in the church, their house was looted.’

    (Rouveret 1996:141)

Given yn’s putative origin as a preposition (but see Gensler 2002), this distributional fact can be motivated diachronically, but not synchronically. A defender of the PredP theory could argue, of course, that the preposition undergoes head movement and incorporates into Pred0, or that a null Pred0 c-selects for a PP predicate. However, for this argument to be valid, independent reasons have to be available to postulate Pred0, which I have shown to be missing. Furthermore, as examples (55)–(56) show, the distribution of yn in comparatives containing a differential is not that expected of Pred0: yn can (perhaps must) appear between the differential adverbial and the comparative that it combines with. The same effect can be observed with differential measure phrases, which require yn to appear after the measure phrase (57).11

(55)

  • a.

    ateb ychydig yn well

    answer little CPRT better

    ‘an answer slightly better’

  • b.

    ateb sydd ychydig yn well

    answer is.REL little CPRT better

    ‘an answer that is slightly better’

    (Mittendorf and Sadler 2008:379)

(56)

graphic

(57)

graphic

The compositional semantics of NP-internal modification requires the degree argument of an AP to be saturated before the AP can function as a predicate or a modifier. If yn were an overt realization of Pred0, we would expect to find it before the differential, which must form a unit with the comparative before the entire extended AP combines with Pred0. Given that this prediction is incorrect, movement of the differential from its base position to somewhere in the middle field could be hypothesized to account for this. Two facts make this proposal highly unlikely: on the one hand, Welsh does not have scrambling, and on the other hand, the informational status of adverbs and degree expressions makes them particularly resistant to scrambling in those languages that do have it (cf. Adger 1996).

4.5 The Eastern Riffian Copular Particle

As discussed by Oomen (2012), the Berber language Eastern Riffian has a predicative particle d that is used with both AP and NP predicates, yet not with PP predicates. This particle is obligatory in copular clauses (58), as well as in secondary predication (59) and, strikingly, in indefinite NPs (60).

(58)

  • a.

    netta d a-ryaz.

    he CPRTM-man

    ‘He is a man.’

  • b.

    netta d a-wessar.

    he CPRTM-old

    ‘He is old.’

    (Oomen 2012:8)

(59)

  • a.

    y-err-itd lmalik.

    3MSG-turn.into.PERF-3MSG.ACCCPRT king

    ‘He made him king.’

  • b.

    i-ssess lqehwa-nnes t ta-berkan-t.

    3MSG-drink.IMPF coffee-3MSG.POSSCPRTF-black-F

    ‘He drinks his coffee black.’

  • c.

    ta-mɣar-t-nnes t-err-itd a-wessar.

    F-woman-F-3SG.POSS 3FSG-turn.PFV-3MSG.ACCCPRTM-old

    ‘His wife made him old.’

  • d.

    twali-ɣ-td a-buhali.

    appear.IMPF-1SG-3SG.ACCCPRTM-mad

    ‘He seems mad to me.’

    (Oomen 2012:8)

(60)

  • ẓṛi-ɣ idž n we-xxam d a-zeggwaɣ.

  • see.PFV-1SG a of M-house CPRTM-red

  • ‘I saw a red house.’

  • (Oomen 2012:13)

An important property of Eastern Riffian is the nominal nature of its adjectives. As Oomen (2012) discusses, the very existence of the category adjective in Berber languages is controversial, and quality concepts are frequently expressed by nouns or verbs. For instance, Djemai (2008) argues that Kabyle Berber has no adjectives; deverbal nouns are used instead. In Tashelhiyt (Aspinion 1953, via Oomen 2012) and Tuareg (Prasse 1972, 2010, via Oomen 2012), adjectives are regarded as a subclass of (stative) verbs. In Eastern Riffian, as Oomen demonstrates, quality concepts are generally encoded as quality verbs, which can be distinguished from other verbs by their compatibility with the comparative phrase introduced by the preposition x ‘on’.

(61)

  • a.

    lbeṭṭix y-eɣla kteṛ zi lbanan / x lbanan.

    melon 3MSG-expensive.PFV more from banana / on banana

    ‘Melons are more expensive than bananas.’

  • b.

    y-uyur kteṛ zi uma-s / *x uma-s.

    3MSG-walk.PFV more from brother-3SG / on brother-3SG

    ‘He walked more than his brother.’

    (Stanly Oomen, pers. comm.)

The roots that are distinguished in this way can also be realized as morphological nouns, which exhibit adjectival behavior in that they, unlike other nouns, can function as modifiers in close apposition and also combine with degree operators.

(62)

  • a.

    a-ɣṛum a-berkan

    M-bread M-black

    ‘black bread’

    (Oomen 2012:5)

  • b.

    netta d a-meqqwṛan bezzaf.

    he CPRTM-big very

    ‘He is very big.’

    (Oomen 2012:7)

Strikingly, adjectives of Arabic origin, which are not morphologically integrated and are therefore nonnominal, appear without the predicative particle d (Oomen 2012).

(63)

  • y-etban eyyi mṭewweṛ.

  • 3MSG-appear.IMPF 1SG.DAT smart

  • ‘He seems smart to me.’

  • (Oomen 2012:9)

The distribution of the predicative particle does not follow from the hypothesis that it instantiates Pred0: not only does it fail to appear with PP predicates, it is also conditioned by the nominal nature of the predicate. While it could be argued that Pred0 is not lexicalized in the same way with different predicates (and is realized as the phonological zero where it fails to surface), independent grounds for such an assumption are necessary; moreover, this argument already presupposes the existence of PredP and therefore cannot be used as evidence for it.

4.6 Lexicalization Patterns

If predication must be mediated by a functional head whose semantics is the same for APs, NPs, and PPs, we expect either no categorial differences with respect to its overtness or random lexicalization (in some languages it would be overt only with APs, in others it would be overt with PPs and NPs, in still others three different lexical items would be used, etc.). The observed pattern, however, is regular in a way not expected from the PredP hypothesis because the presence of an overt predicative particle depends on the lexical category of the predicate in a systematic way (Hengeveld 1992, Stassen 1997, Pustet 2005). While PP predicates never appear with predicative particles,12 a predicative particle appears with AP predicates in a language only if it also appears with NP predicates. In other words, the lexicalization pattern of predicative particles does not follow from the hypothesis that they lexicalize a functional head that obligatorily mediates all predication. Some additional assumptions are necessary.

The crosslinguistic absence of an overt Pred0 with PP predicates could be attributed to its incorporation into the preposition itself, but once again this speculation places the burden of proof on the proponents of PredP analyses: some independent evidence for the existence of Pred0 with PP predicates must be provided before it can be argued that its lexicalization fails for an otherwise unmotivated reason. Likewise, the implicational hierarchy in the lexicalization of the copular particle with AP and NP predicates does not follow from the simple assumption that they both must be merged as complements to Pred0. While it is not unreasonable to hypothesize that the observed pattern will follow once the proper role of Pred0 (e.g., the factors enabling it to introduce AP and NP predicates) is established, this role and the existence itself of Pred0 should be established first.

One proposal to that effect comes from Hengeveld (1992), who notes the independently known fact that in different languages or within the same language adjectives can be “more verbal” or “more nominal.” Hengeveld proposes that it is the latter type that gives rise to an overt mediator in the predicative position, and shortly we will see support for this proposal.

Independent evidence for gradation in the concept of an adjective, from more nominal to more verbal, comes from languages where two types of adjectives are available: Japanese (e.g., Dixon 1977, Miyagawa 1987, Kubo 1992, Nishiyama 1999, Baker 2003b, Backhouse 2004), the Cariban language Macushi (Abbott 1991, via Dixon 2004), various Bantu languages including Zulu (e.g., Doke 1927, Stassen 1997:168, Posthumus 2000), the Tibeto-Birman language Manange (Genetti and Hildebrandt 2004), and so on. A clear illustration of the split comes from Japanese, where adjectives are divided into “verbal” and “nominal” on the basis of their morphology and the use of the copula.

(64) Canonical (“verbal”) adjectives

  • a.

    Yama-ga taka-i.

    mountain-NOM high-PRES

    ‘The mountain is high.’

  • b.

    Yama-ga taka-katta.

    mountain-NOM high-PAST

    ‘The mountain was high.’

(65) Nominal adjectives

  • a.

    Yoru-ga sizuka da.

    night-NOM quiet COP.PRES

    ‘The night is quiet.’

  • b.

    Yoru-ga sizuka datta.

    night-NOM quiet COP.PAST

    ‘The night was quiet.’

(66)

  • a.

    Taroo-ga [utukusi-i] tori-o mita.

    Taroo-NOM beautiful-PRES bird-ACC saw

    ‘Taroo saw a beautiful bird.’

  • b.

    Hanako-ga [kirei na] hana-o katta.

    Hanako-NOM pretty COP flower-ACC bought

    ‘Hanako bought a pretty flower.’

    (Yamakido 2000:588)

Both types of adjectives can combine with degree operators and modifiers or function as nonintersective modifiers (Yamakido 2000).

(67)

  • Max-ga kanzen-na baka da.

  • Max-NOM complete-PRES fool COP.PRES

  • ‘Max is a complete fool.’

  • (Yamakido 2000:593)

If it is a priori possible for adjectives to exhibit more or fewer nominal properties, it can be assumed that the predicative particle with adjectival predicates in a language reflects the more nominal nature of adjectives (as a category) in that language. This predicts that in languages where adjectival predicates are introduced by predicative particles, adjectives would generally exhibit a more nominal behavior. This hypothesis is confirmed for both Welsh and Eastern Riffian, where adjectives and nouns share a number of properties usually associated with nouns. In particular, as discussed in Matushansky 2012b, Welsh adjectives do not form adverbs (the particle yn is also used for VP-modification), they combine with prepositions in the absence of any noun, their mode of combination with nouns can reasonably be treated as compounding, and they trigger agreement on their own modifiers, as do nouns. Importantly, Welsh differs in this respect from other Celtic languages, which do not use predicative particles with adjectival predicates. The link between the use of the predicative particle and the nominal nature of an adjective is supported by Eastern Riffian integrated and nonintegrated adjectives, as discussed above, and by Telugu, where the marker -gaa, which is obligatory with AP predicates and induces transient interpretation of NP predicates (Balusu 2014), also correlates with the core semantics of adjectives as property concepts rather than predicates.

The semantic role of the predicative particle can therefore be hypothesized to be the enabling of predication, but only for NPs and most likely not in the sense envisaged by Bowers, since I will still assume that the core meaning of the noun is that of a predicate (as needed for NP internal composition). Instead, I hypothesize that the need for a predicative particle arises because true predication requires more transience than is natural for a nominal predicate. Therefore, the predicative particle is used to ensure the contrast between intrinsic and extrinsic properties (i.e., classification, including identity, vs. temporary property ascription). Hypothesizing that adjectives in Eastern Riffian and Welsh denote quality concepts rather than properties allows for a certain alignment of morphosyntactic and semantic factors here, although a proper analysis must be left for future work.

4.7 Copular Particles: Conclusion

Copular particles have been hypothesized to be lexicalizations of Pred0, but a careful crosslinguistic study shows that their distribution does not support this hypothesis. First of all, in no language do predicative particles proper appear with PP predicates,13 which entails that PP predicates should be treated differently (see Baker 2003a; potential evidence for this also comes from the crosslinguistically frequent requirement to use stance verbs rather than verbal or nonverbal copulas with PP predicates; see Stassen 1997). Second, the predicative particle appears to be strongly linked to nominal predication, and in two languages where AP predicates appear with a predicative particle, it can be shown that adjectives systematically manifest some nominal properties. Third, a case study of the Welsh predicative particle shows that its distribution does not in and of itself match what would be expected from Pred0: while some of these mismatches can be accounted for by assuming incorporation, others cannot. The lack of information on Eastern Riffian precludes a comparable in-depth analysis, but given the close link between the use of a predicative particle and the nominal nature of the AP predicates, the least that can be said is that the PredP analysis does not account for it. Conversely, it becomes possible to hypothesize that at least in Welsh and Eastern Riffian the predicative particle functions as either a marker of nominal predication or a semantic operator turning a time-stable property denoted by an NP into a more transient property that is required for predication (as opposed to identity). The proper formalization of this operator is impossible without a much more detailed discussion of the two languages than is possible here.

I emphasize again that for a number of cases, including those where the copular particle is used only in copular clauses, the distribution of the copular particle can be accounted for in the PredP approach—for instance, by hypothesizing its incorporation into a higher or lower head. However, given that the crosslinguistic distribution of copular particles does not correspond to what we would have expected for Pred0, it cannot be used as an empirical argument for Pred0. In other words, had we not started with the assumption that nonverbal predication has to be mediated, we would not have arrived at the PredP hypothesis as a natural explanation for copular particles. It would be missing an underlying generalization, I contend, to extend to all predication the pattern that can only be observed with nominal predicates.

5 Other Putative Overt Predicators

Summarizing what we have seen so far, in languages that have copular particles these do not have the distribution expected from Pred0. However, some elements in more familiar languages have also been hypothesized to lexicalize Pred0. Such elements include as and for (e.g., Emonds 1985, Aarts 1992, Bowers 1993, 2001, Starke 1995, Den Dikken 2006) and their crosslinguistic counterparts (Bailyn 2001, 2002 for Slavic, Eide and Åfarli 1999 for Norwegian), as well as the Russian v ‘in’ (Bailyn 2002).

(68)

  • a.

    Mary takes John for a fool.

  • b.

    Jessamine views her mother as her best friend.

(69) Russian

  • a.

    My sčitaem ego svoim.

    we consider him.ACC self.POSS-INS

    (Bailyn 2001:18)

  • b.

    My sčitaem ego kak svoego.

    we consider him.ACCAS self.POSS.ACC

    (Bailyn 2001:18)

  • c.

    My sčitaem ego za svoego.

    we consider him.ACCFOR self.POSS.ACC

    ‘We consider him as one of us.’

    (Bailyn 2002)

(70) Norwegian

  • a.

    Vi fant Marit (*som) naken / *(som) nervevrak.

    we found Mary SOM naked / SOM nervous.wreck

  • b.

    Vi så Jon (*som) rasende / *(som) spùkelse.

    we saw John SOM furious / SOM ghost

  • c.

    Vi returnerte pakken (*som) uåpnet / *(som) flypost.

    we returned parcel.DEFSOM unopened / SOM air.mail

  • d.

    Han ankom selskapet (*som) maskert / *(som) sjørøver.

    he arrived party.DEFSOM masked / SOM pirate

  • e.

    Hun levde og døde (*som) ensom / *(som) eneboer.

    she lived and died SOM lonely / SOM hermit

    (Eide and Åfarli 1999:160)

(71)

(72)

  • Russian

  • On rešil vybrat’sja v prezidenty.

  • he decided elect.INF.REFL in presidents.ACC=NOM

  • ‘He decided to get elected as president.’

Regarding the distribution of these elements, what is true for the English as and for—namely, that their use is heavily restricted—is also true for their counterparts in other languages. So far, I have argued that this restriction entails that the putative lexicalization of Pred0 cannot be used as evidence for the existence of Pred0, and the same argument can be used in this case: the distribution of such putative overt predicators as for, as, and in is not that expected of the head of a small clause. The argument can be made stronger in certain cases, where it can be demonstrated that the putative overt predicator does not look like Pred0 even if we assume that the head of a small clause is overt only in a subset of small clauses.

One case in point is the Russian v ‘in’ in examples like (72), where, Bailyn (2002) suggests, the case on the plural NP is nominative, agreeing with the subject of the small clause, rather than the accusative (which must surface as genitive for an animate NP, like presidents) that the preposition v ‘in’ should assign here. An alternative is proposed by Mel’čuk (1985:461–482), Franks and Pereltsvaig (2004), and Marelj and Matushansky (2010): if it is assumed that a normally animate NP can in this context be treated as inanimate (which its interpretation, the appropriate profession or calling, rather than a set of individuals, is fully compatible with), then the nominative case marking on it is the normal surface realization of accusative for inanimate NPs. Supporting this view is the fact, noted by Mel’čuk, that v ‘in’ can also combine with mass nouns denoting the entire profession—and the case is clearly accusative.

(73)

  • On pošel/xočet v aviaciju.

  • he went/wants in aviation.ACC

  • ‘He went/wants to go into aviation.’

The plural NP in (72) and the mass noun in (73) have a very similar interpretation, and both, as also noted by Mel’čuk, preserve it in two more environments. In addition to the ‘become’ use, contributed by the directional interpretation of the preposition, there exist parallel constructions of persistence in the state (74a) and discontinuation of the state (74b) with the corresponding locative prepositions.

(74)

  • a.

    On služil v soldatax / armii.

    he served in soldiers.LOC / army.LOC

    ‘He served as a soldier.’

  • b.

    Ego vygnali iz lëtčikov / aviacii.

    him.ACC chased.PL from pilots.GEN / aviation.GEN

    ‘He was kicked out of aviation.’

While the animacy of the plural NP complements of the locative prepositions cannot be determined on the basis of these examples, their interpretation as denoting the appropriate profession or calling is shared with that of the plural NP complement of v ‘in’ in examples like (72)—and with the corresponding mass nouns. Clearly, neither fact can be accounted for by the hypothesis that the preposition v ‘in’ in examples like (72) is Pred0: Pred0 is not expected to combine with mass nouns denoting professions, nor does it appear likely that the locative prepositions in (74) also constitute realizations of Pred0, which would need to be assumed: if the small-clause structure is hypothesized for ‘become’ examples like (72), it would need to be assumed in their ‘stay’ and ‘cease’ counterparts in (74).

Similarly, evidence can be provided (Marelj and Matushansky 2015) that in examples (75a–c) for and its counterpart za in Serbo-Croatian and Russian are just prepositions. It can be shown that for and za c-select NPs, to which they assign accusative, as za does in its directional usage in Russian and Serbo-Croatian. The NP hypothesized to be the subject of the for/za small clause does not behave as such with respect to anaphor binding, yet shows exactly the behavior expected from the direct object of the verb. The verbs that combine with for/za-PPs generally do not combine with clear small clauses, and vice versa: verbs that unquestioningly combine with small clauses generally do not combine with for/za-PPs.

(75)

  • a.

    English

    Imogen takes/mistakes him for a fool.

  • b.

    Serbo-Croatian

    Uzima me za budalu.

    take.3SG me.ACCFOR fool. ACC

    ‘She/He takes me for a fool.’

  • c.

    Russian

    My prinjali ego za duraka.

    we take.PAST.PL him. ACC FOR fool. ACC

    ‘We took him for a fool.’

Given these facts, it is simpler to assume that in examples like (75a–c) for/za is just a preposition, and Marelj and Matushansky (2015) also show that the semantics of attitude ascription can be derived from the component parts of a change-of-possession verb and the preposition for/za in its ‘replacement’ interpretation. In other words, for the specific case of for/za in these three languages, the PredP hypothesis does not explain the facts, which turn out to be easier to derive from the more straightforward assumption that for/za is a preposition.

It is unquestionable that evidence of this kind cannot be unhesitatingly extrapolated from one language to another. For instance, in French (see Starke 1995) pour ‘for’ has a less restricted distribution in that it can also combine with AP predicates.

(76)

  • a.

    Je le tiens pour fidèle à ses amis . . .

    I 3SG.ACC hold for faithful to his friends

    comme sa claymore l’est à sa main.

    as his claymore 3SG.ACC+is to his hand

    ‘I grant him true to friendly band, // As his claymore is to his hand.’

    (Sir Walter Scott, La Dame du Lac (The Lady of the Lake), canto II:XIV; transl.

    by Albert Montémont)

  • b.

    Son visage est sans attraits et je la tiens pour stupide.

    her face is without attractions and I her hold for stupid

    ‘Her face is unattractive and I consider her stupid.’

    (Étienne Léon Lamothe-Langon (baron de), Mémoires et souvenirs d’une femme de qualité, sur le Consulat et l’Empire, vol. 4, p. 274. Paris: Mame et Delaunay-Vallée, 1830.)

Does this mean that in French pour ‘for’ is more likely to be a lexicalization of Pred0? To answer this question, it is necessary to examine the distribution of pour in more detail. I contend, however, that at this point the issue should not even be raised. In the absence of evidence in favor of PredP in unquestionable small clauses (sections 23), the presence of a functional element in far less obviously predicational structures, even if proved, does not entail anything about canonical small clauses: even if it is demonstrated that the overt elements in examples like (68)–(72) are indeed required to mediate predication, their presence and role can be straightforwardly attributed to the fact that the structures they appear in are sufficiently different from canonical small clauses to require such mediating elements. I conclude that the so-called overt predicators discussed in this section cannot be taken as evidence for the existence of a functional head in small clauses.

6 Conclusion and Further Questions

Summarizing what has been established here on both the theoretical and the empirical sides, it can be concluded that the PredP hypothesis is not substantiated by facts. Given that no overt material intervenes between the subject and the predicate of a canonical small clause, the postulation of a null functional head there—which is, moreover, obligatory—must be motivated either by theoretical considerations or by empirical ones.

On the theoretical side, as shown in section 2, theory-internal arguments for Pred0 either have become obsolete or are amenable to independently motivated alternative analyses: for instance, raising to object can explain how what looks like a segment (the predicate to the exclusion of the subject of the small clause) can be fronted, while the constraints on the coordination of unlikes can be shown to be semantic in nature.

On the empirical side, from the point of view of compositional semantics there does not seem to be any role that Pred0 can fulfill. If we assume, with Bowers (1993, 2001), Baker (2003a), and Den Dikken (2006), that it enables the predicate to take a subject, and if we attempt to formalize this hypothesis in the standard Montagovian approach, a number of problems arise (section 3.1). An attempt to link Pred0 to event semantics (section 3.2) likewise fails. In other words, not only is there no developed theory of the semantic contribution of Pred0, but an attempt to create one leads to what seems to me to be insurmountable problems.

Turning now to the putative overt instantiations of Pred0 (i.e., copular particles; section 4), we have seen that their behavior does not follow from the PredP hypothesis. The most striking property of predicative particles is that they never appear with PP predicates. On the assumption that small clauses with PP predicates are PredPs, this gap in the distribution of the putative Pred0 is inexplicable. Setting PP predicates aside does not help: even if we limit our attention to AP and NP predicates, the distribution of copular particles is limited in ways that do not follow from the hypothesis that they lexicalize Pred0. More specifically, crosslinguistically copular particles generally occur with NP predicates only, and if they appear with AP predicates, then they also appear with NP predicates. This pattern would be unexpected if copular particles lexicalized Pred0; the alternative that I advance here is that copular particles are indeed limited to NP predicates and either mark ascriptive (as opposed to classificatory) predication or enable it (see also Adger and Ramchand 2003). Evidence for this view comes from the fact that in languages where predicative particles systematically occur in small clauses with AP predicates (a small set consisting of Welsh and Eastern Riffian), adjectives can be argued to be nominal on independent morphosyntactic grounds.

A brief incursion into other elements hypothesized to be “overt predicators,” such as the English as, for, and in and their crosslinguistic equivalents, has shown that these elements do not support the PredP hypothesis either: while an alternative analysis can be provided for some of them, for others it can simply be argued that the distribution of the constituents that they should be heading is not that of small clauses—for instance, they do not appear with verbs that normally take propositional arguments.

To conclude, the null functional head in small clauses is not a theoretical necessity in the current state of Minimalist syntax; it is semantically unmotivated to say the least; and, on the empirical side, equating Pred0 with any overt element does not seem to lead to correct predictions. The question now arises of establishing what the structure of a small clause is, if it is not headed by a functional head. The most straightforward answer would seem to be to reject a categorially unified analysis and to extend Stowell’s analysis of small clauses as lexical projections to more complex cases by treating predication as the last thematic merge to an extended projection of a lexical head.

(77)

graphic
graphic

The fact that the structures in (77) are nonuniform with respect to lexical category has positive consequences as well as potentially negative ones. Starting with the former, the fact that the lexical category of the predicate of a small clause is visible from the outside makes it possible for us to account for the c-selection restrictions discussed in section 1 (with all the caveats about the potential alternative explanation in terms of s-selection). The clearest case of such a restriction is the crosslinguistic categorial constraints on resultatives: while PP resultatives are generally allowed, AP resultatives are rare and NP resultatives are the rarest of all. A c-selectional restriction does not explain this hierarchy, but at least provides a handle on these cases (especially if we do not view lexical categories as atomic, regarding them instead as complexes of features, allowing, for instance, for nominal and verbal adjectives). Conversely, postulating the same head for all small clauses leads us to expect them all to have the same distribution, contrary to fact.

The hypothesis that small clauses have the lexical category of their predicate also appears to make some questionable predictions. One is that AP small clauses are expected to have the distribution of an AP; NP small clauses, that of an NP; and so on. However, semantic factors, can easily explain why this situation does not obtain: a DP or an NP denoting an entity (type e) is not expected to have the same distribution as a DP or an NP denoting a proposition (type ⟨s, t⟩) for reasons having nothing to do with their lexical category.

Less clear is the status of specifiers. To take an extreme case, consider (77a) and (77c), where DP1 and DP2 are specifiers of the same functional head (DP and DegP, respectively). At first blush, the same syntax is incorrectly predicted, yet closer consideration reveals that the two DPs differ in so many other properties (θ-role assignment, case marking, hierarchical position, semantic type, derivation by external or internal Merge, etc.) as to make this prediction unverifiable.

Finally, it could be objected that the nonuniform analysis in (77) fails to account for the fact that all of these structures are small clauses. I contend, however, that structural uniformity is not necessary. What small clauses have in common is their semantics: the predication relation, with its resulting propositional denotation. Under this view, the fact that the small-clause subject is the subject of a DP in (77a), of an NP in (77b), of a DegP in (77c), and of an AP in (77d) is immaterial because the distribution of a small clause is primarily determined by its propositional denotation (and other properties of small clauses follow from their distribution). Postulating syntactic uniformity in addition to this not only seems superfluous, but also, as we have seen, has no basis in fact—there is no evidence that small clauses are headed by one and the same functional head. I conclude that the PredP hypothesis, while esthetically pleasing, has no evidence in its favor.

Notes

1 Proposition-like constituents containing a verbal predicate (e.g., See [Spot run]) or constituents containing no functional projections or verbs (e.g., a [star visible]) have also been proposed as candidates for small clauses. As the latter have neither the same distribution nor the same semantic type as canonical small clauses, I disregard them here; verbal small clauses are discussed below.

2 In the current state of the theory, a degree word, such as more or how, indicates the presence of another functional projection, DegP, that it is either the head (Abney 1987, Bowers 1987, Corver 1990, 1991, 1997a,b) or the specifier (Bowers 1975, Jackendoff 1977, Heim 2000, Bhatt and Pancheva 2004) of. The problem is then exacerbated, as Spec,DegP is not available as a subject position under either view. I address this issue in section 2.4.

3 It is frequently (and incorrectly) claimed that the Linear Correspondence Axiom (Kayne 1994) is incompatible with multiple specifiers. For discussion, see Abels and Neeleman 2006.

4 It must be made clear at this point that Chierchia 1985 is concerned with the finite/nonfinite opposition in VPs, associating the standard property/set interpretation with finite verbs (e.g., in John runs) and attributing the entity-correlate interpretation to their nonfinite counterparts (e.g., in John tried to run). Which interpretation is derived from the other is not a question that is raised, since Chierchia assumes that a VP is ambiguous between the two and does not discuss the compositional semantics implicit in the finite/nonfinite opposition; if anything, the semantic relation between predicates and their entity-correlates derives the latter from the former by the mechanism of nominalization ().

Chierchia and Turner 1988, on the other hand, explicitly assumes that the entity-correlate interpretation of VPs is basic, but it also only discusses the finite/nonfinite opposition on the empirical side.

Finally, the assumption that lies at the core of Bowers’s (1993) proposal is also approximated by Bealer’s (1982) first-order theory of properties, which assumes that all properties are primitives combined with their subjects by a distinguished two-place logical predicate Δ that expresses the predication relation. However, given that this theory assumes (p. 82) that Δ is expressed by the copula and does not consider NP-internal composition or (obviously) small clauses, it need not concern us here.

None of this work, to my mind, can or should be taken as evidence for Pred0.

5 Some further doubts about this approach come from the observation (Rothstein 2005, Grashchenkov and Markman 2008, Uchihara 2010) that small clauses pass at least some diagnostics for eventualities, obviating the need for an eventuality argument introduced by be. On the other hand, Engelberg (2005) demonstrates that some of these eventuality diagnostics also apply to attributive adjectives. Therefore, the question remains open whether the eventuality argument is absent in the predicate to begin with.

6 It has been suggested to me (see also Baker 2003a) that the role of Pred0 is purely syntactic: it introduces a specifier to which the predicate assigns its external θ-role. For this proposal to work, it is necessary to first assume that nonverbal predicates cannot themselves project specifiers, and this assumption is not independently motivated. An alternative is that Pred0 assigns this θ-role, as has been assumed for v0; yet see section 2.1 for evidence against assimilating Pred0 to v0.

7 In addition to the Welsh yn, Bowers (2001) suggests that Pred0 can be realized overtly in Korean (kye). To the best of my knowledge, however, this particle (written key in Kim and Maling 1997, one of Bowers’s references, and in Wechsler and Boh 2001) appears only in resultative small clauses, in infinitive-like clauses, and to mark an adverb.

8 Generalizations and examples in this section not attributed otherwise come from Matushansky and de Dreu 2009.

9 While Baker consistently uses high tone on the copular particle () and low tone on the preposition (), Ogie (2009) indicates opposite tone markings. I will not attempt to regularize Edo phonology here.

10 While some potential evidence for the presence of in reduced relative clauses can be drawn from the fact that the underlyingly low-tone complementizer acquires the high tone in (i), Omoruyi (1986) does not posit an underlying copular particle there and does not discuss the tonal change; I do not have sufficient knowledge of Edo phonology to hypothesize what it is due to.

  • (i)

    òkhùò [nè mòsèmòsè] ⇒ némòsèmòsè

    woman that beautiful.A

    ‘a beautiful woman’

    (Omoruyi 1986:298)

11 For the Welsh data below, I am indebted to Peredur Davies-Webb and Gwenllian Awbery, where not indicated otherwise. While the latter rejects placing yn before a differential adverb, the former permits it, along with the possibility of having yn both before and after a differential measure phrase (57c). I have no explanation for this contrast.

12 Recall that by predicative particle I mean a functional item that occurs with nonverbal predication in small-clause environments not limited to copular clauses. The distribution of copular verbs and nonverbal copulas follows the same pattern, although occasional exceptions occur, such as the appearance of the Hebrew generic nonverbal copula ze with PP predicates (Greenberg 1994, 1998) and the need for a stance verb (e.g., stand, sit, lie) or a special copular verb for PP predicates.

13 The problem is most severe for the hypothesis (Den Dikken 2006) that all predication must be mediated: either PP predicates should be attributed some special status (which is difficult, given that even VP modifiers are now generally treated as predicates over events and therefore a linking element would be expected even if a PP is not regarded as the main predicate in a small clause), or some reason should be provided for why Pred0 is never overt with PPs. Obligatory incorporation is not independently motivated, nor is its result expected to always be identical to the preposition itself.

Acknowledgments

For the development of this article, I owe a debt of gratitude to a lot of people, who talked and listened to me about the head of small clauses and related topics. I thank, in no particular order, Merijn de Dreu for the joint analysis of Bantu copulas and all the data that he has so painstakingly collected; Marijana Marelj for sharing my pain and working with me on the so-called overt predicators in Slavic; Victor Manfredi for discussing Edo with me, sharing his intuitions, and verifying mine—may no blame be attached to him for my errors; Stanly Oomen for introducing me to Eastern Riffian and other Berber dialects; and, first and foremost, Eddy Ruys for unfailing support, listening, arguing, and proposing. Thanks are also due to the three anonymous reviewers for extremely detailed and helpful comments, and to audiences at the Research Institute for Linguistics, Budapest (20 May 2015), the University of Geneva (17 March 2015), IATL 30 (20–21 October 2014), Utrecht University (19 November 2014), and Paris VII (13–14 November 2014), where parts of this work were presented. Acknowledgments for judgments provided are given below.

References

References
Aarts,
Bas
.
1992
.
Small clauses in English: The nonverbal types
.
Berlin
:
Mouton de Gruyter
.
Abbott,
Miriam
.
1991
. Macushi. In
Handbook of Amazonian languages
, ed. by
Desmond C.
Derbyshire
and
Geoffrey K.
Pullum
,
3
:
23
160
.
Berlin
:
Mouton de Gruyter
.
Abels,
Klaus
, and
Ad
Neeleman
.
2006
.
Universal 20 without the LCA
.
Ms., University of Tromsø/University College London
.
Abney,
Steven
.
1987
.
The English noun phrase in its sentential aspect
.
Doctoral dissertation, MIT, Cambridge, MA
.
Abu-Joudeh,
Maisoun I
.
2013
.
Case checking in verbless sentences functioning as embedded clauses in Modern Standard Arabic: A Minimalist account
.
Theory and Practice in Language Studies
3
:
209
221
.
Adger,
David
.
1996
. Aspect, agreement and measure phrases in Scottish Gaelic. In
The syntax of the Celtic languages
, ed. by
Robert D.
Borsley
and
Ian
Roberts
,
200
222
.
Cambridge
:
Cambridge University Press
.
Adger,
David
, and
Gillian
Ramchand
.
2003
.
Predication and equation
.
Linguistic Inquiry
34
:
325
360
.
Akmajian,
Adrian
.
1984
.
Sentence types and the form-function fit
.
Natural Language and Linguistic Theory
2
:
1
23
.
Aspinion,
Robert
.
1953
.
Apprenons le berbère: Initiation aux dialectes chleuhs
.
Rabat
:
Editions F. Moncho
.
Backhouse,
Anthony E
.
2004
. Inflected and uninflected adjectives in Japanese. In
Adjective classes: A crosslinguistic typology
, ed. by
R. M. W.
Dixon
and
Alexandra Y.
Aikhenvald
,
50
73
.
Oxford
:
Oxford University Press
.
Bailyn,
John
.
2001
.
The syntax of Slavic predicate case
.
ZAS Papers in Linguistics
22
:
1
26
.
Bailyn,
John
.
2002
.
Overt predicators
.
Journal of Slavic Linguistics
10
:
23
52
.
Bailyn,
John
, and
Edward J.
Rubin
.
1991
. The unification of instrumental case assignment in Russian. In
Cornell working papers in linguistics, vol. 9
, ed. by
Almeida Jacqueline
Toribio
and
Wayne
Harbert
,
99
126
.
Ithaca, NY
:
Cornell University, Department of Modern Languages and Linguistics
.
Baker,
Mark C
.
2003a
.
Lexical categories: Verbs, nouns, and adjectives
.
Cambridge
:
Cambridge University Press
.
Baker,
Mark C
.
2003b
. Verbal adjectives as adjectives without phi-features. In
Proceedings of the Fourth Tokyo Conference on Psycholinguistics
, ed. by
Yukio
Otsu
,
1
22
.
Tokyo
:
Keio University
.
Balusu,
Rahul
.
2014
.
The overt predicator -gaa in Telugu
.
Paper presented at Syntax Brown Bag, New York University, 11 April 2014
.
Barbiers,
Sjef
.
1995
.
The syntax of interpretation
.
Doctoral dissertation, HIL. Leiden: Leiden University
.
Bayer,
Josef
.
1996
.
Directionality and Logical Form
.
Dordrecht
:
Kluwer
.
Bealer,
George
.
1982
.
Quality and concept
.
Oxford
:
Clarendon Press
.
Benveniste,
Emile
.
1966
. Structure de relations de personne dans les verbes. In
Problèmes de linguistique générale
, ed. by
Emile
Benveniste
,
225
236
.
Paris
:
Gallimard
.
Berman,
Ruth
, and
Alexander
Grosu
.
1976
. Aspects of the copula in Modern Hebrew. In
Studies in Modern Hebrew syntax and semantics
, ed. by
Peter
Cole
,
265
285
.
Amsterdam: North-Holland
.
Bhatt,
Rajesh
, and
Roumyana
Pancheva
.
2004
.
Late merger of degree clauses
.
Linguistic Inquiry
35
:
1
45
.
Bierwisch,
Manfred
.
1988
. On the grammar of local prepositions. In
Syntax, Semantik und Lexikon
, ed. by
Manfred
Bierwisch
,
Wolfgang
Motsch
, and
Ilse
Zimmermann
,
1
65
.
Berlin
:
Akademie Verlag
.
Bolinger,
Dwight
.
1972
.
Degree words
.
The Hague
:
Mouton
.
Bowers,
John
.
1975
.
Adjectives and adverbs in English
.
Foundations of Language
13
:
529
562
.
Bowers,
John
.
1987
. Extended X-bar theory, the ECP and the Left Branch Condition. In
WCCFL 6: The Proceedings of the Sixth West Coast Conference on Formal Linguistics
, ed. by
Megan
Crowhurst
,
47
62
.
Stanford, CA
:
CSLI Publications
.
Bowers,
John
.
1993
.
The syntax of predication
.
Linguistic Inquiry
24
:
591
656
.
Bowers,
John
.
2001
. Predication. In
The handbook of contemporary syntactic theory
, ed. by
Mark
Baltin
and
Chris
Collins
,
299
333
.
Oxford
:
Blackwell
.
Brauner,
Siegmund
, and
Irmtraud
Herms
.
1986
.
Lehrbuch des modernen Swahili
.
Leipzig
:
Verlag Enzyklopädie
.
Buell,
Leston Chandler
, and
Merijn
de Dreu
.
2013
.
Subject raising in Zulu and the nature of PredP
.
The Linguistic Review
30
:
423
466
.
Chierchia,
Gennaro
.
1985
.
Formal semantics and the grammar of predication
.
Linguistic Inquiry
16
:
417
443
.
Chierchia,
Gennaro
, and
Raymond
Turner
.
1988
.
Semantics and property theory
.
Linguistics and Philosophy
11
:
261
302
.
Chomsky,
Noam
.
1957
.
Syntactic structures
.
The Hague
:
Mouton
.
Chomsky,
Noam
.
1995
. Categories and transformations. In
The Minimalist Program
,
219
394
.
Cambridge, MA
:
MIT Press
.
Chomsky,
Noam
.
2000
. Minimalist inquiries: The framework. In
Step by step: Essays on Minimalist syntax in honor of Howard Lasnik
, ed. by
Roger
Martin
,
David
Michaels
, and
Juan
Uriagereka
,
89
155
.
Cambridge, MA
:
MIT Press
.
Chomsky,
Noam
.
2013
.
Problems of projection
.
Lingua
130
:
33
49
.
Chomsky,
Noam
.
2014
.
Problems of projection: Extensions
.
Ms., MIT, Cambridge, MA
.
Chung,
Sandra
, and
James
McCloskey
.
1987
.
Government, barriers, and small clauses in Modern Irish
.
Linguistic Inquiry
18
:
173
237
.
Citko,
Barbara
.
2008
.
Small clauses reconsidered: Not so small and not all alike
.
Lingua
118
:
261
295
.
Corver,
Norbert
.
1990
.
The syntax of left branch extractions
.
Doctoral dissertation, University of Tilburg
.
Corver,
Norbert
.
1991
. Evidence for DegP. In
NELS 21
, ed. by
Tim
Sherer
,
33
47
.
Amherst
:
University of Massachusetts, Graduate Linguistic Student Association
.
Corver,
Norbert
.
1997a
.
The internal syntax of the Dutch extended adjectival projection
.
Natural Language and Linguistic Theory
15
:
289
368
.
Corver,
Norbert
.
1997b
.
Much-support as a last resort
.
Linguistic Inquiry
28
:
119
164
.
Daniels,
Michael W
.
2002
. On a type-based analysis of feature neutrality and the coordination of unlikes. In
Proceedings of the 8th International Conference on Head-Driven Phrase Structure Grammar
, ed. by
Frank
Van Eynde
,
Lars
Hellan
, and
Dorothee
Beermann
,
137
147
.
Stanford, CA
:
CSLI Publications
.
Déchaine,
Rose-Marie
, and
Mireille
Tremblay
.
2012
. Deriving nominal reference. In
Proceedings of the fortieth Western Conference on Linguistics
, ed. by
Christina
Galeano
,
Emrah
Görgülü
, and
Irina
Presnyakova
,
419
430
.
Burnaby, BC
:
Simon Fraser University, Department of Linguistics
.
Dekympe,
Karel
.
2000
. Kinyarwanda–Wort für Wort.
Bielefeld
:
REISE KNOW-HOW Verlag Peter Rump GmbH
.
Diessel,
Holger
.
2009
.
The morphosyntax of demonstratives in synchrony and diachrony
.
Linguistic Typology
3
:
1
50
.
Dik,
Simon C
.
1968
.
Coordination: Its implications for the theory of general linguistics
.
Amsterdam
:
North-Holland
.
Dikken,
Marcel den
.
2006
.
Relators and linkers: The syntax of predication, predicate inversion, and copulas
.
Cambridge, MA
:
MIT Press
.
Dixon,
R. M. W.
1977
.
Where have all the adjectives gone?
Studies in Language
1
:
19
80
.
Dixon,
R. M. W
.
2004
. Adjective classes in typological perspective. In
Adjective classes: A cross-linguistic typology
, ed. by
R. M. W.
Dixon
and
Alexandra Y.
Aikhenvald
,
1
49
.
Oxford
:
Oxford University Press
.
Djemai,
Salem
.
2008
.
L’expression de la qualité en berbère: Étude de quelques aspects morphologiques, sémantiques et syntaxiques de l’adjectif en kabyle
.
Master’s thesis, INALCO
.
Doke,
Clement M
.
1927
.
Textbook of Zulu grammar
.
Johannesburg
:
Witwatersrand University Press
.
Dooren,
Annemarie van
.
2014
.
People cannot change–but cartoon animals can
.
On modality and dynamicity. Paper presented at TIN-dag, Utrecht, 1 February 2014
.
Doron,
Edit
.
1983
.
Verbless predicates in Hebrew
.
Doctoral dissertation, University of Texas, Austin
.
Doron,
Edit
.
1986
. The pronominal “copula” as agreement clitic. In
The syntax of pronominal clitics
, ed. by
Hagit
Borer
,
313
332
.
New York
:
Academic Press
.
Eide,
Kristin M.
, and
Tor
A.Åfarli
.
1999
.
The syntactic disguises of the predication operator
.
Studia Linguistica
53
:
155
181
.
Emonds,
Joseph E
.
1985
.
A unified theory of syntactic categories
.
Dordrecht
:
Foris
.
Engelberg,
Stefan
.
2005
.
Kimian states and the grammar of predicative adjectives
.
Theoretical Linguistics
31
:
331
347
.
Fong,
Vivienne
.
2003
. Resultatives and depictives in Finnish. In
Generative approaches to Finnic and Saami linguistics
, ed. by
Diane
Nelson
and
Satu
Manninen
,
201
234
.
Stanford, CA
:
CSLI Publications
.
Franks,
Steven
, and
Asya
Pereltsvaig
.
2004
. Functional categories in the nominal domain. In
Formal Approaches to Slavic Linguistics 12: The Ottawa Meeting
, ed. by
Olga
Arnaudova
,
Wayles
Browne
,
Maria-Luisa
Rivero
, and
Danijela
Stojanovic
,
109
128
.
Ann Arbor
:
Michigan Slavic Publications
.
Frege,
Gottlob
.
1892
.
Über Begriff und Gegenstand
.
Vierteljahrsschrift für wissenschaftliche Philosophie
16
:
192
205
.
Fukui,
Naoki
.
1986
.
A theory of category projection and its application
.
Doctoral dissertation, MIT, Cambridge, MA
.
Fukui,
Naoki
.
1995
.
Theory of projection in syntax
.
Stanford, CA/Tokyo
:
CSLI Publications/Kurosio Publishers
.
Gelderen,
Elly van
.
2011
.
The linguistic cycle
.
Oxford
:
Oxford University Press
.
Genetti,
Carol
, and
Kristine
Hildebrandt
.
2004
. The two adjective classes in Manange. In
Adjective classes: A cross-linguistic typology
, ed. by
R. M. W.
Dixon
and
Alexandra Y.
Aikhenvald
,
74
96
.
Oxford
:
Oxford University Press
.
Gensler,
Orin D
.
2002
.
Why should a demonstrative turn into a preposition? The evolution of Welsh predicative yn
.
Language
78
:
710
764
.
Grashchenkov,
Pavel
, and
Vita G.
Markman
.
2008
. Non-core arguments in verbal and nominal predication: High and low applicatives and possessor raising. In
WCCFL 27: Proceedings of the 27th West Coast Conference on Formal Linguistics
, ed. by
Natasha
Abner
and
Jason
Bishop
,
185
193
.
Somerville, MA
:
Cascadilla Proceedings Project
.
Greenberg,
Yael
.
1994
.
Hebrew nominal sentences and the nature of the stage-individual level distinction
.
Master’s thesis, Bar-Ilan University
.
Greenberg,
Yael
.
1998
. An overt syntactic marker for genericity in Hebrew. In
Events and grammar
, ed. by
Susan
Rothstein
,
125
143
.
Dordrecht
:
Kluwer
.
Greenberg,
Yael
.
2002
. The manifestation of genericity in the tense aspect system of Hebrew nominal sentences. In
Themes in Arabic and Hebrew syntax
, ed. by
Jamal
Ouhalla
and
Ur
Shlonsky
,
267
298
.
Dordrecht
:
Kluwer
.
Grohmann,
Kleanthes K.
,
John
Drury
, and
Juan Carlos
Castillo
.
2000
. No more EPP. In
WCCFL 19: Proceedings of the 19th West Coast Conference on Formal Linguistics
, ed. by
Roger
Billerey
and
Brook Danielle
Lillehaugen
,
153
166
.
Somerville, MA
:
Cascadilla Press
.
Harley,
Heidi
.
1995
.
Subjects, events and licensing
.
Doctoral dissertation, MIT, Cambridge, MA
.
Heim,
Irene
.
2000
. Degree operators and scope. In Proceedings of SALT 10, ed. by
Brendan
Jackson
and
Tanya
Matthews
,
40
64
. https://journals.linguisticsociety.org/proceedings/index.php/SALT/issue/view/24.
Heller,
Daphne
.
2002
.
On the relation of connectivity and specificational pseudoclefts
.
Natural Language Semantics
10
:
243
284
.
Hengeveld,
Kees
.
1992
.
Non-verbal predication: Theory, typology, diachrony
.
Berlin
:
Mouton de Gruyter
.
Heycock,
Caroline
.
1993
.
Syntactic predication in Japanese
.
Journal of East Asian Linguistics
2
:
167
211
.
Heycock,
Caroline
, and
Anthony
Kroch
.
1999
.
Pseudocleft connectedness: Implications for the LF interface level
.
Linguistic Inquiry
30
:
365
397
.
Heycock,
Caroline
, and
Young-Suk
Lee
.
1989
.
Subject and predication in Korean and Japanese
.
Ohak Yon’gu/Language Research
25
:
775
791
.
Hiraiwa,
Ken
.
2001
. Multiple Agree and the Defective Intervention Constraint in Japanese. In
Proceedings of the 1st HUMIT Student Conference in Language Research (HUMIT 2000)
, ed. by
Ora
Matushansky
,
Albert
Costa
,
Javier
Martin-Gonzalez
,
Lance
Nathan
, and
Adam
Szczegielniak
,
67
80
.
MIT Working Papers in Linguistics 40
.
Cambridge, MA
:
MIT
,
MIT Working Papers in Linguistics
.
Jackendoff,
Ray
.
1977
.
syntax: A study of phrase structure
.
Cambridge, MA
:
MIT Press
.
Jerro,
Kyle
.
2015
. Copulas and the semantics of location. In
Proceedings of the 15th Meeting of the Texas Linguistic Society
, ed. by
Christopher
Brown
,
Qianping
Gu
,
Cornelia
Loos
,
Jason
Mielens
, and
Grace
Neveu
,
91
105
. http://tls.ling.utexas.edu/2014tls/TLS15-Proceedings.pdf.
Johnson,
Kyle
.
1991
.
Object positions
.
Natural Language and Linguistic Theory
9
:
577
636
.
Jones,
Bob Morris
.
2009
.
The modification of adjectives in Welsh
.
Journal of Celtic Linguistics
13
:
45
116
.
Jones,
Bob Morris
, and
Alan R.
Thomas
.
1977
.
The Welsh language: Studies in its syntax and semantics
.
Cardiff
:
The University of Wales Press
.
Kamp,
Hans
, and
Uwe
Reyle
.
1993
.
From discourse to logic
.
Dordrecht
:
Kluwer
.
Kayne,
Richard S
.
1994
.
The antisymmetry of syntax
.
Cambridge, MA
:
MIT Press
.
Kim,
Soowon
, and
Joan
Maling
.
1997
. A crosslinguistic perspective of resultative formation. In
Proceedings of the 1997 TLS Conference: The Syntax and Semantics of Predication
, ed. by
Ralph C.
Blight
and
Michelle Y.
Moosally
,
189
204
.
Austin
:
University of Texas at Austin
.
Kiso,
Andrea
.
2012
.
Tense and aspect in Chichewa, Citumbuka and Cisena: A description and comparison of the tense-aspect systems in three Southeastern Bantu languages
.
Doctoral dissertation, Stockholm University
.
Koizumi,
Masatoshi
.
1993
. Object agreement and the split-VP hypothesis. In
Papers on Case and agreement I
, ed. by
Jonathan David
Bobaljik
and
Colin
Phillips
,
99
148
.
MIT Working Papers in Linguistics 18
.
Cambridge, MA
:
MIT
,
MIT Working Papers in Linguistics
.
Koizumi,
Masatoshi
.
1995
.
Phrase structure in Minimalist syntax
.
Doctoral dissertation, MIT, Cambridge, MA
.
Kratzer,
Angelika
.
1996
. Severing the external argument from its verb. In
Phrase structure and the lexicon
, ed. by
Johan
Rooryck
and
Laurie
Zaring
,
109
137
.
Dordrecht
:
Kluwer
.
Kubo,
Miori
.
1992
.
Japanese syntactic structures and their constructional meaning
.
Doctoral dissertation, MIT, Cambridge, MA
.
Lasnik,
Howard
.
1995
. Last resort. In
Minimalism and linguistic theory
, ed. by
Shosuke
Haraguchi
and
Micio
Funaki
,
1
32
.
Tokyo
:
Hituzi Syobo
.
Lasnik,
Howard
, and
Mamoru
Saito
.
1991
. On the subject of infinitives. In
Papers from the 27th Regional Meeting of the Chicago Linguistic Society
, ed. by
Lise
Dobrin
,
Lynn
Nichols
, and
Rosa
Rodríguez
,
324
343
.
Chicago
:
University of Chicago, Chicago Linguistic Society
.
Li,
Charles N.
, and
Sandra A.
Thompson
.
1977
. A mechanism for the development of copula morphemes. In
Mechanisms of syntactic change
, ed. by
Charles N.
Li
,
419
444
.
Austin
:
University of Texas Press
.
Lohndal,
Terje
.
2009
. The copula cycle. In
Cyclical change
, ed. by
Elly van
Gelderen
,
209
242
.
Amsterdam
:
John Benjamins
.
Loogman,
Alfons
.
1965
.
Swahili grammar and syntax
.
Pittsburgh, PA
:
Duquesne University Press
.
Maienborn,
Claudia
.
2005a
. On Davidsonian and Kimian states. In
Existence: Semantics and syntax
, ed. by
Ileana
Comorovski
and
Klaus
von Heusinger
,
107
130
.
Dordrecht
:
Springer
.
Maienborn,
Claudia
.
2005b
.
On the limits of the Davidsonian approach: The case of copula sentences
.
Theoretical Linguistics
31
:
275
316
.
Maling,
Joan
.
1983
. Transitive adjectives: A case of categorial reanalysis. In
Linguistic categories: Auxiliaries and related puzzles
, ed. by
Frank N.
Heny
and
Barry
Richards
,
253
289
.
Dordrecht
:
Reidel
.
Maling,
Joan
.
1989
. Adverbials and structural case in Korean. In
Harvard studies in Korean linguistics III
, ed. by
Susumu
Kuno
,
John
Whitman
,
Young-Se
Kang
,
Ik-Hwan
Lee
,
Sung-Yun
Bak
, and
Youngjoo
Kim
,
297
308
.
Cambridge, MA
:
Harvard University, Department of Linguistics
.
Maling,
Joan
, and
Rex A.
Sprouse
.
1995
. Structural case, specifier-head relations, and the case of predicate NPs. In
Studies in comparative Germanic syntax
, ed. by
Hubert
Haider
,
Susan
Olsen
, and
Sten
Vikner
,
167
186
.
Dordrecht
:
Kluwer
.
Marantz,
Alec
.
1997
. No escape from syntax: Don’t try morphological analysis in the privacy of your own lexicon. In
Proceedings of the 21st Annual Penn Linguistics Colloquium
, ed. by
Alexis
Dimitriadis
,
Laura
Siegel
,
Clarissa
Surek-Clark
, and
Alexander
Williams
,
201
225
.
University of Pennsylvania Working Papers in Linguistics 4.2
.
Philadelphia
:
University of Pennsylvania, Penn Linguistics Club
.
Marelj,
Marijana
, and
Ora
Matushansky
.
2010
.
Against overt predicators in Slavic
.
Paper presented at GLOW 33, Wrocław, 13 April 2010
.
Marelj,
Marijana
, and
Ora
Matushansky
.
2015
.
Mistaking for: Testing the theory of mediated predication
.
Linguistic Inquiry
46
:
43
76
.
Markman,
Vita
.
2008
.
Pronominal copula constructions are what? Reduced specificational pseudo-clefts!
In
WCCFL 26: Proceedings of the 26th West Coast Conference on Formal Linguistics
, ed. by
Charles B.
Chang
and
Hannah J.
Haynie
,
366
374
.
Somerville, MA
:
Cascadilla Proceedings Project
.
Marshad,
Hassan A.
, and
Saleh M.
Suleiman
.
1991
.
A comparative study of Swahili ni and Arabic kana as copulative elements
.
Language Sciences
13
:
21
37
.
Matushansky,
Ora
.
2002
.
Tipping the scales: The syntax of scalarity in the complement of seem
.
Syntax
5
:
219
276
.
Matushansky,
Ora
.
2008
.
On the linguistic complexity of proper names
.
Linguistics and Philosophy
31
:
573
627
.
Matushansky,
Ora
.
2010
. Russian predicate case, encore. In
Proceedings of FDSL 7.5
, ed. by
Gerhild
Zybatow
,
Philip
Dudchuk
,
Serge
Minor
, and
Ekaterina
Pshehotskaya
,
117
135
.
Frankfurt
:
Peter Lang
.
Matushansky,
Ora
.
2012a
.
On the internal structure of case in Finno-Ugric small clauses
.
Finno-Ugric Languages and Linguistics
1
:
3
43
.
Matushansky,
Ora
.
2012b
.
On the role of the copular particle: Evidence from Welsh
.
Paper presented at Celtic Linguistics Conference 7, Rennes, 22–23 June 2012
.
Matushansky,
Ora
.
2013
. More or better: On the derivation of synthetic comparatives and superlatives in English. In
Distributed Morphology today: Morphemes for Morris Halle
, ed. by
Ora
Matushansky
and
Alec
Marantz
,
59
78
.
Cambridge, MA
:
MIT Press
.
Matushansky,
Ora
, and
Merijn
de Dreu
.
2009
.
Some notes on non-verbal predication in Bantu
.
Paper presented at MUST, Utrecht University, 11 March 2009
.
McWhorter,
John
.
1992
.
NI and the copula system in Swahili: A diachronic approach
.
Diachronica
9
:
15
46
.
Meillet,
Antoine
.
1906
.
La phrase nominale en indo-européen
.
Mémoires de la société de linguistique de Paris
14
:
1
26
.
Mel’čuk,
Igor
.
1985
.
Poverxnostnyj sintaksis russkix čislitel’nyx vyraženij
.
Vienna
:
Institut für Slawistik der Universität Wien
.
Melzian,
Hans
.
1937
.
A concise dictionary of the Bini language of Southern Nigeria
.
London
:
Kegan Paul
.
Mittendorf,
Ingo
, and
Louisa
Sadler
.
2008
. NP would like to meet GF: A Welsh adjectival construction. In
Proceedings of the LFG08 Conference
, ed. by
Miriam
Butt
and
Tracy Holloway
King
,
373
393
.
Stanford, CA
:
CSLI Publications
.
Miyagawa,
Shigeru
.
1987
.
Lexical categories in Japanese
.
Lingua
73
:
29
51
.
Munn,
Alan
.
1987
.
Coordinate structure and X-bar theory
.
McGill Working Papers in Linguistics
4
:
121
140
.
Munn,
Alan
.
1993
.
Topics in the syntax and semantics of coordinate structures
.
Doctoral dissertation, University of Maryland, College Park
.
Nishiyama,
Kunio
.
1999
.
Adjectives and the copulas in Japanese
.
Journal of East Asian Linguistics
8
:
183
222
.
Ogie,
Ota
.
2009
.
Multi-verb constructions in Edo
.
Doctoral dissertation, Norwegian University of Science and Technology, Trondheim
.
Omoruyi,
Thomas O
.
1986
.
Adjectives and adjectivalization processes in Edo
.
Studies in African Linguistics
17
:
283
302
.
Oomen,
Stanly
.
2012
.
Adjectives as a distinct class in Eastern Riffian (Berber, Afroasiatic)
.
Paper presented at Syntax Interface Lectures, Utrecht, 16 April 2012
.
Pereltsvaig,
Asya
.
2007
.
Copular sentences in Russian: A theory of intra-clausal relations
.
Berlin
:
Springer Verlag
.
Peterson,
Peter
.
1981
.
Problems with constraints on coordination
.
Linguistic Analysis
8
:
449
460
.
Pollard,
Carl
, and
Ivan A.
Sag
.
1994
.
Head-Driven Phrase Structure Grammar
.
Chicago
:
University of Chicago Press
.
Postal,
Paul
.
1974
.
On raising
.
Cambridge, MA
:
MIT Press
.
Posthumus,
Lionel
.
1978
.
Die kopulatiewe in Zulu met verwysing na ander Bantoetale
.
Bloemfontein
:
UOVS
.
Posthumus,
Lionel
.
1988
.
Identifying copulatives in Zulu and Southern Sotho
.
South African Journal of African Languages
8
:
61
64
.
Posthumus,
Lionel
.
2000
.
The so-called adjective in Zulu
.
South African Journal of African Languages
20
:
148
158
.
Posthumus,
Lionel
.
2006
.
Identificative copulatives in Southern Ndebele: Evidence for diachronic postulations in Zulu
.
Literator
27
:
109
134
.
Potts,
Christopher
, and
Thomas
Roeper
.
2006
. The narrowing acquisition path: From expressive small clauses to declaratives. In
The syntax of nonsententials: Multi-disciplinary perspectives
, ed. by
Ljiljana
Progovac
,
Kate
Paesani
,
Eugenia
Casielles
, and
Ellen
Barton
,
183
201
.
Amsterdam
:
John Benjamins
.
Prasse,
Karl-G
.
1972
.
Manuel de grammaire touarègue (tăhăggart)
.
Copenhagen
:
Akademisk Forlag
.
Prasse,
Karl-G
.
2010
.
Tuareg elementary course (Tahăggart)
.
Cologne
:
Köppe
.
Pustet,
Regina
.
2005
.
Copulas: Universals in the categorization of the lexicon
.
Oxford
:
Oxford University Press
.
Pylkkänen,
Liina
.
2002
.
Introducing arguments
.
Doctoral dissertation, MIT, Cambridge, MA
.
Rapoport,
Tova R
.
1987
.
Copular, nominal and small clauses: A study of Israeli Hebrew
.
Doctoral dissertation, MIT, Cambridge, MA
.
Riemsdijk,
Henk van
.
1978
.
A case study in syntactic markedness: The binding nature of prepositions
.
Dordrecht
:
Foris
.
Riemsdijk,
Henk van
.
2002
.
The unbearable lightness of GOing
.
Journal of Comparative Germanic Linguistics
5
:
143
196
.
Rothstein,
Susan
.
1995
. Small clauses and copular constructions. In
Small clauses
, ed. by
Anna
Cardinaletti
and
Maria Teresa
Guasti
,
27
48
.
New York
:
Academic Press
.
Rothstein,
Susan
.
1999
.
Fine-grained structure in the eventuality domain: The semantics of predicative adjective phrases and be
.
Natural Language Semantics
7
:
347
420
.
Rothstein,
Susan
.
2005
.
States and modification: A reply to Maienborn
.
Theoretical Linguistics
31
:
375
381
.
Rouveret,
Alain
.
1996
. Bod in the present tense and in other tenses. In
The syntax of the Celtic languages
, ed. by
Robert D.
Borsley
and
Ian
Roberts
,
125
170
.
Cambridge
:
Cambridge University Press
.
Rudin,
Catherine
.
1988
.
On multiple questions and multiple wh-fronting
.
Natural Language and Linguistic Theory
6
:
445
501
.
Runner,
Jeffrey T
.
2006
.
Lingering challenges to the raising to object and object control constructions
.
Syntax
9
:
193
213
.
Rutkowski,
Paweł
.
2006
.
From demonstratives to copulas: A cross-linguistic perspective and the case of Polish
.
Journal of Universal Language
7
:
147
175
.
Sag,
Ivan A.
,
Gerald
Gazdar
,
Thomas
Wasow
, and
Steven
Weisler
.
1985
.
Coordination and how to distinguish categories
.
Natural Language and Linguistic Theory
3
:
117
171
.
Schachter,
Paul
.
1977
.
Constraints on coordination
.
Language
53
:
86
103
.
Schachter,
Paul
.
1984
.
A note on syntactic categories and coordination in GPSG
.
Natural Language and Linguistic Theory
2
:
269
281
.
Sichel,
Ivy
.
1997
. Two pronominal copulas and the syntax of Hebrew nonverbal sentences. In
Proceedings of the 1997 TLS Conference: The Syntax and Semantics of Predication
, ed. by
Ralph C.
Blight
and
Michelle Y.
Moosally
,
295
306
.
Austin
:
University of Texas at Austin
.
Soschen,
Alona
.
2003
.
On subjects and predicates in Russian
.
Doctoral dissertation, University of Ottawa
.
Starke,
Michal
.
1995
. On the format of small clauses. In
Small clauses
, ed. by
Anna
Cardinaletti
and
Maria Teresa
Guasti
,
237
269
.
New York
:
Academic Press
.
Stassen,
Leon
.
1997
.
Intransitive predication
.
Oxford
:
Clarendon Press
.
Steere,
Edward
.
1884/1930
.
Handbook of the Swahili language as spoken at Zanzibar
. 4th ed.,
revised and enlarged by A. C. Madan
.
London
:
Society for Promoting Christian Knowledge
.
Stowell,
Tim
.
1981
.
Origins of phrase structure
.
Doctoral dissertation, MIT, Cambridge, MA
.
Stowell,
Tim
.
1983
.
Subjects across categories
.
The Linguistic Review
2
:
285
312
.
Stowell,
Tim
.
1991
. Small clause restructuring. In
Principles and parameters in comparative grammar
, ed. by
Robert
Freidin
,
182
208
.
Cambridge, MA
:
MIT Press
.
Svenonius,
Peter
.
1994
.
Dependent nexus: Subordinate predication structures in English and the Scandinavian languages
.
Doctoral dissertation, University of California at Santa Cruz
.
Svenonius,
Peter
.
1996
. Predication and functional heads. In
Proceedings of the Fourteenth West Coast Conference on Formal Linguistics
, ed. by
José
Camacho
,
Lina
Choueiri
, and
Maki
Watanabe
,
493
507
.
Stanford, CA
:
CSLI Publications
.
Takahashi,
Chioko
.
1996
.
Multiplicity, optionality and constraints on the distribution of nominative case in Japanese
.
Doctoral dissertation, Cornell University, Ithaca, NY
.
Uchihara,
Hiroto
.
2010
.
Pronominals in Edo – A descriptive study
.
Ms., University at Buffalo, The State University of New York
. Available at https://www.academia.edu/1341848/Final_Paper_Pronominals_inEdo-A_descriptive_study.
Ura,
Hiroyuki
.
1996
.
Multiple feature-checking: Atheory of grammatical function splitting
.
Doctoral dissertation, MIT, Cambridge, MA
.
Wechsler,
Stephen
, and
Bokyung
Boh
.
2001
.
On resultative predicates and clauses: Parallels between Korean and English
.
Language Sciences
23
:
391
423
.
Whitman,
John
.
2004
.
Semantics and pragmatics of English verbal dependent coordination
.
Language
80
:
403
434
.
Williams,
Edwin S
.
1975
. Small clauses in English. In
Syntax and semantics
,
vol. 4
, ed. by
John P.
Kimball
,
249
273
.
New York
:
Academic Press
.
Williams,
Edwin S
.
1983
.
Against small clauses
.
Linguistic Inquiry
14
:
287
308
.
Willis,
David
.
2003
.
Antisymmetry and Welsh noun phrases
. .
Wondem,
Mulusew Asratie
.
2014
.
The syntax of non-verbal predication in Amharic and Geez
.
Doctoral dissertation, Utrecht University
.
Yamakido,
Hiroko
.
2000
. Japanese attributive adjectives are not (all) relative clauses. In
WCCFL 19: Proceedings of the 19th West Coast Conference on Formal Linguistics
, ed. by
Roger
Billerey
and
Brook Danielle
Lillehaugen
,
588
602
.
Somerville, MA
:
Cascadilla Press
.
Yoon,
James H.-S
.
1989
. The grammar of inalienable possession of constructions in Korean, Mandarin and French. In
Harvard studies in Korean linguistics III
, ed. by
Susumu
Kuno
,
John
Whitman
,
Young-Se
Kang
,
Ik-Hwan
Lee
,
Sung-Yun
Bak
, and
Young-joo
Kim
,
357
368
.
Cambridge, MA
:
Harvard University
.
Yoon,
James H.-S
.
1990
. Theta-theory and the grammar of the inalienable possession constructions. In
NELS 20
, ed. by
Juli
Carter
,
Rose-Marie
Déchaine
,
Bill
Philip
, and
Tim
Sherer
,
502
516
.
Amherst
:
University of Massachusetts, Graduate Linguistic Student Association
.
Zaring,
Laurie
.
1996
.
“Two be or not two be”: Identity, predication and the Welsh copula
.
Linguistics and Philosophy
19
:
103
142
.
Zhang,
Niina Ning
.
2009
.
Coordination in syntax
.
Cambridge
:
Cambridge University Press
.