Skip to main content
SpringerLink
Log in

The meaning of the tough-construction

  • Published:
Natural Language Semantics Aims and scope Submit manuscript

Abstract

A formal semantic analysis of the tough-construction is provided building on the well-known observation that events play a central role. A close look at the semantic characteristics of the class of tough-predicates and the syntactic and semantic properties of nonfinite clauses reveals the link between these pieces, expanding on recent advances in the semantics of clauses (Moulton in Natural selection and the syntax of clausal complementation, PhD thesis, University of Massachusetts at Amherst, 2009). Building on Salzmann (Reconstruction and resumption in indirect A′-dependencies: On the syntax of prolepsis and relativization in (Swiss) German and beyond, de Gruyter, Berlin 2017b), a formal semantic and syntactic analysis of prolepsis is provided to explain the antecedent gap chain in the tough-construction. In total, this paper offers a description and explanation of (i) the class of tough-predicates; (ii) the properties of the nonfinite clause that appears in the tough-construction; (iii) why no other predicates or clauses are permitted in the construction; (iv) the many properties of the antecedent gap chain in the tough-construction, primarily the fact that the chain is “weakly” unbounded; and (v) the semantic contribution of prolepsis as it applies in the tough-construction.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others

Notes

  1. Members of the class of complement object deletions differ among each other in other significant ways as well. This class minimally includes the tough-construction (adjectives, nouns, and verbs), pretty-class adjectives, too/enough-clauses, and potentially other cases like nonfinite relative clauses, purpose clauses, and nonfinite predicate CPs (Landau 2011). For instance, while the tough-construction does not permit gaps in existential constructions (cf. (26, 27) below), too/enough-clauses do; cf. Mary is too dull for there to be a book about __ (pace Lasnik and Fiengo 1974:538). Moreover, too/enough-clauses often permit resumptive pronouns and subject gaps (control of PRO by the main clause subject) where the tough-construction systematically does not (Cinque 1990; Mulder and den Dikken 1992). Fleisher (2014:76) observes that pretty-class adjectives do not admit unbounded dependencies at all, while tough-constructions do. And while the tough-construction systematically fails to show reconstruction effects, nonfinite object relative clauses do (Bhatt 1999, 2006). This is all to say that the account of the tough-construction provided is not meant to cover all of these constructions, and indeed, despite surface similarities, it is not clear whether a unified account is even warranted. Still, see Sect. 8 for possible extensions of the analysis provided here to at least pretty-class predicates, which are the closest cousins of tough-predicates.

  2. We may trivially confirm that the proleptic objects in the tough-construction are arguments of the matrix clause and not, say, topicalized constituents of the for-CP by noting that they are not permitted with for-CPs in other contexts.

    1. (i)
      figure h

  3. Many of the parallels discussed here are originally found in Salzmann (2017b:325ff). See that work for additional parallels and discussion. It seems that Salzmann is unaware that the tough-construction permits a “true” proleptic object in addition to the tough-subject; the discussion in Salzmann (2017b) compares properties of the tough-subject to that of prolepsis in finite clauses. While there are indeed many shared properties, some, like being weakly bounded, are not shared across constructions.

  4. There are various reported cases of bound variable interpretations, but as Poole et al. (2017) point out (attributing the observation to a blog post by Benjamin Bruening), all such cases involve picture-NPs, and so are confounded by the well-known logophoric properties of such nouns.

    1. (i)
      figure l

    Controlling for this factor, the bound variable interpretations disappear. See Poole et al. (2017) for additional arguments against a low reading of the tough-subject.

  5. For instance, Postal (1971) lists such dependencies as ungrammatical, as do Bresnan (1971) and Nanni (1978) (though Kaplan and Bresnan 1982 later accept similar gaps). Dalrymple and King (2000) (correctly) observe that there is considerable speaker variation, and that such dependencies are marginal for many speakers.

    1. (i)
      figure q
    1. (ii)
      figure r

    Since the possibility of an unbounded dependency with finite clauses does not appear to be systematic, I will put it aside here in the hopes that further research may be able to more accurately define when this is permitted. See footnote 49 for a possible explanation, and discussion in Sect. 7 on variation in the acceptability of the antecedent gap chain.

  6. Lasnik and Fiengo (1974), among others, take such data to indicate that the antecedent gap chain in the tough-construction cannot cross a CP boundary. The reasoning is that the expletive-there cannot be parsed as the “experiencer” argument of the adjective, and so must be the subject of the nonfinite clause, which in turn forces C to be present (for case-marking reasons). When there is no overt subject, no C layer is present. Rezac (2006), following Heim (1987), suggests that such examples may be attributed to a definiteness effect, rather than an inability to cross a clausal boundary.

  7. I note in passing that there are some cases where prolepsis is possible but gaps are not which are ruled out for independent reasons. For instance, as a rule, English does not allow possessor gaps, which is enough to rule out (ii), even when (i) is fine.

    1. (i)
      figure ao
    1. (ii)
      figure ap

    Such examples may be taken as evidence against a movement/Agree analysis of prolepsis (as proposed in Salzmann 2017b).

  8. It is worth noting that the crosslinguistic facts illustrate that the tough-alternation can occur independent of category. For example, there are languages like Korean and Japanese with well-documented cases of the tough-construction, but which lack any sort of adjectival predicate in the alternation, instead utilizing verbal suffixes or verbs proper. Thus, the concept of “difficult” or “easy” is what matters, not the category of the predicates themselves. The same point of course can be made for English: the class of tough-predicates includes adjectives, nouns, and verbs (Gluckman 2019).

  9. The exceptions to this are the analyses that propose that the nonfinite constituent found in the tough-construction is smaller than a CP when there is an antecedent gap chain (Bresnan 1971; Lasnik and Fiengo 1974; Nanni 1980; Mulder and den Dikken 1992; Jacobson 1992; Longenbaugh 2016b; Wurmbrand 2001). However, this assumption is probably incorrect—at least in English. Though this is a contentious issue, there is considerable evidence that the nonfinite constituent is (or at least can be) a full clause; see footnote 21. I note that whether the constituent is analyzed as a full CP or smaller cross-cuts movement and predication approaches. There are movement approaches which analyze the nonfinite constituent as smaller than a clause (Bresnan 1971; Jacobson 1992; Longenbaugh 2016a,b), and there are likewise predication approaches which make this assumption (Nanni 1980; Mulder and den Dikken 1992).

  10. I will restrict discussion to adjectival tough-predicates, though everything said below applies to non-adjectival tough-predicates as well, as discussed in Lasnik and Fiengo (1974) and Gluckman (2019). Note that some of the predicates in (37) have non-adjectival variants, e.g., the psych-verbs. See Pesetsky (1987) for discussion of the relationship between psych-verbs and the tough-construction. Finally, there are many tough-predicates that exhibit significant interspeaker variation, e.g., the positive forms of impossible, illegal, inconvenient. Though these are traditionally believed not to participate in the tough-alternation, in fact speakers readily accept and produce them in such constructions, as a simple Google search illustrates.

    1. (i)
      figure as
    1. (ii)
      figure at

    Though this paper will not investigate all sub-classes of tough-predicates, I take it as given that certain predicates impose further restrictions, such as Fleisher’s rare-class predicates discussed below. I assume that (im)possible, (il)legal, (in)convenient constitute a definable sub-class as well, though I will not address how precisely this class is defined. Note though that there is systematicity in this class: negative forms are always tough-predicates. This suggests a solution that goes beyond mere selection (cf. Keine and Poole 2017). Finally, it is worth noting that in some languages, possible is perfectly acceptable as a tough-predicate (Tayalati and Danckaert 2020).

  11. I use the term “event” to mean “eventuality,” since tough-predicates may also describe states, as in (i) and (ii).

    1. (i)
      figure av
    1. (ii)
      figure aw

    I assume that states and events are of the same semantic class, i.e., eventualities (Bach 1986; Rothstein 2004). A reviewer asks how the characterization here differs from what is proposed in Keine and Poole (2017), transposed to an event-less (situation-based) semantics. The difference is discussed more in Sect. 4. Anticipating that discussion, I make the empirically motivated assumption that eventualities (events and states) are not propositional; they are ontologically distinguishable from the type attributed to worlds. Since Keine and Poole do not make this distinction, their account over-generates what is, and is not, a tough-predicate.

  12. A reviewer questions whether “intensionality” might be a better term than “relative truth,” given the eventual analysis offered below. As a descriptive label, I believe that “relative truth” is appropriate; I intend this to be a neutral term for any sort of non-objective assessment (see, e.g., MacFarlane 2014). Though later I adopt an intensional semantics, it is important to recognize that the proper treatment of judge dependence and the various other subjective/modal meanings found with tough-predicates is still an open issue. It is of course possible to translate the formal analysis below into other frameworks. Since in this section I aim for descriptive adequacy, I will use the more neutral term to describe this class of predicates.

  13. See Keine and Poole (2017) for a similar observation. In the syntax literature on the tough-construction, this argument is typically called the “experiencer.” I note that there is considerable debate concerning the status of this experiencer argument and its role in the tough-construction (cf. Levine and Hukari 2006; Hartman 2011; Keine and Poole 2017; among others). The debate concerns whether for should be understood as a complementizer, and hence part of the nonfinite clause, or as a preposition introducing the experiencer. For space reasons, I put this debate aside here; cf. footnote 8 and footnote 21.

  14. It is worth noting that vagueness (as in Kennedy 2007) is one type of relative truth that does not appear to matter for the tough-construction. Presumably, this is due to the fact that this kind of subjectivity is not associated with the event argument, but comes as a result of the functional morpheme pos (Bylinina 2017). Because vagueness is not construed as being a property of the event, purely vague predicates are precluded from functioning as tough-predicates, for the reasons presented in this article. The point highlights the importance of tying relative truth to the event argument itself.

  15. This of course does not mean that such predicates lack an event(uality) argument; rather, the distinction is one of selection: non-tough-predicates select for an individual-denoting subject (and assign it a thematic role).

  16. Still, there are different “kinds” of judge dependence. In being able to license a judge PP, tough-predicates proper are classified as experiential (Bylinina 2017). If I say The test was difficult, I must have some first-hand experience that allows me to determine whether the test was in fact difficult (either I took the test, or I constructed the test, etc.). Non-experiential judge-dependent event descriptions can be tough-predicates as well, like Stowell’s (1991) adjectives describing mental properties, e.g. kind, mean; cf. footnote 18.

  17. Tim Stowell (p.c.) points out that predicates that describe “mental properties”, like kind, mean, brave, etc., may also describe events and involve relative truth (they are judge dependent according to Bylinina 2017), but do not easily participate in the tough-alternation, as (i) and (ii) show. However, (iii) illustrates that grammatical instances of an antecedent gap chain with such adjectives are possible.

    1. (i)
      figure be
    1. (ii)
      figure bf
    1. (iii)
      figure bg

    Again, it is likely that there are further restrictions on what is allowed to be a subject for such predicates, perhaps tracking a kind/type distinction as well.

  18. A reviewer points out some cases of tough-predicates which are predicated of non-event-denoting individuals, with examples attributed to Williams (2003); judgments are from the source.

    1. (i)
      figure bh
    1. (ii)
      figure bi

    As Williams notes, it may often be difficult to determine whether there is an elided/implicit clause in such cases; such clauses would clearly be needed to interpret (ii). I take the grammaticality of (ii) to reflect what Pustejovsky (1996) refers to as metonymic reconstruction. Individual-denoting nouns can be coerced into an event reading. This is most readily found with aspectual predicates: Mary began the test/contest/chose/task/errand/puzzle. For pragmatic reasons, some nouns may stand for an act, e.g. “doing the __”. This is supported by the fact that predicates which strictly require eventive nouns, like take an hour, accept puzzle and a cup of coffee on their event readings, but not uncoerced non-eventive nouns, like the tree:

    1. (iii)
      figure bj

  19. Of course, some predicates are truly polysemous, e.g., hard (wood).

  20. A reviewer notes that canonical A-raising also does not allow movement out of finite clauses, and so wonders whether (47) is a useful observation. However, (47) is important precisely because the tough-construction is not canonical A-raising; there is a demonstrable A′-step in the lower clause. Since A′-movement is perfectly natural out of finite clauses, (47) is unexpected. The reviewer furthers notes that this may simply be attributed to selection. However, as an explanation this is not sufficient, because (47) is systematically ruled out. We have yet to find a language with an alternation like (47). While it may be that examples like (47b) are eventually discovered, as it stands, (47) seems to go beyond a mere idiosyncratic selectional property of certain adjectives. This is not to say that syntactic selection does not play a role in the tough-construction. As illustrated above, tough-predicates may impose selectional restrictions on their subjects (e.g., rare-class adjectives). It also matters that for-CPs allow merging of an operator but gerunds do not, again presumably a selectional fact. However, selection does not provide an adequate explanation for the class of tough-predicates and their restriction to nonfinite clauses.

  21. This is a controversial claim. Many authors have concluded that the nonfinite clause associated with the tough-construction must be smaller than a CP—at least when there is an antecedent gap chain (Bresnan 1971; Lasnik and Fiengo 1974; Nanni 1980; Longenbaugh 2016b). Evidence in favor of this view comes from cases where the noun after for cannot possibly be construed as an argument of the adjective, but the antecedent gap chain is not permitted. The reasoning is that the presence of for in its role as C somehow leads to ungrammaticality. I will not rehash the debate here; it suffices to show that there are clear examples where for-as-C is permitted.

    1. (i)
      figure bo

    See discussion in Levine and Hukari (2006), Keine and Poole (2017), and Gluckman (2018) for additional evidence against such “reduced clause” analyses.

  22. Note that even on more sophisticated analyses of PRO, the above treatment holds: once the PRO argument is bound or its reference is otherwise established, the resulting phrase is taken to be of type <s,t>.

  23. In Sect. 8 I provide a discussion on differentiating classes of clause-embedding adjectives. Briefly, the distribution is consistent with treating some predicates (certain, sure) as descriptions of “contentful” individuals, and some predicates as descriptions of “contentful” events (the tough-predicates). The distinction can be observed in which CPs they embed and also in which nominals they can be predicated of. Importantly, some tough-predicates are ambiguous between being properties of contentful events and contentful individuals; thus important may embed a finite or nonfinite clause and may also be predicated of a contentful nominal individual or contentful nominal event. This accounts for why the examples in (54) are felicitous with important, as pointed out by a reviewer. I have controlled for this ambiguity by using tough-predicates which do not exhibit this ambiguity.

  24. This fact is almost certainly related to the observation that for-CPs independently appear to involve modal quantification, as observed in their use with nonfinite relative clauses (Bhatt 1999, 2006; Hackl and Nissenbaum 2012). See also Jones (1991) for discussion of the syntax of such clauses.

  25. Kratzer (2006, 2013) in fact argues that that-CPs should be properties of situations, which then suggests that we should collapse the distinction between CPs as properties of individuals versus CPs as properties of events. I have no qualm with this, as long as we can still define an “individual/event” distinction at some level. The distributional differences between for-CPs and that-CPs, as well as contentful-event nouns (plan) and contentful-individual nouns (story), necessitate a dichotomy of this kind. See further discussion in Sect. 8.

  26. I assume that properties of type <v,st> can be lowered into definite descriptions of events when needed.

  27. The version of Predicate Modification that is assumed here includes a judge argument:

    1. (i)
      figure cd

  28. This is consistent with approaches to the tough-construction that involve complex predicate formation (Nanni 1980) and approaches that explicitly treat the CP as a modifier (Williams 1983; Wilder 1991; Hornstein 2001). While my analysis shares many traits of the proposal in Keine and Poole (2017), the most notable difference is that the embedded clause is not selected by the tough-predicate. Tough-predicates thus are always of the same basic type, rather than always ambiguous. See Wilder (1991), Hornstein (2001), and Gluckman (2018, 2021) for arguments that the for-CP does not act as an argument of the tough-predicate.

  29. I note that Keine and Poole in fact need a third version of tough in order to account for cases when the tough-predicate takes an event-denoting subject, as in (38).

  30. The overall goal of Keine and Poole’s paper is an explanation of the intervention effect observed in Hartman (2011) and Bruening (2014). Keine and Poole argue that the “intervention” is in actuality a type mismatch. I will not address intervention in this paper.

  31. Keine and Poole further argue that the for-CP is selected rather than an adjunct. This is probably wrong, as discussed in Wilder (1991), Hornstein (2001), and Gluckman (2021). The nonfinite clauses associated with the tough-predicate bears hallmarks of being adjuncts.

  32. This in fact distinguishes for-CPs from gerunds with tough-predicates. In (i) the climbing event and the difficult event are the same event.

    1. (i)
      figure ch

  33. See Cable’s (2011) de re reading under think and Hacquard (2006:60) for discussion of de re readings of events.

  34. This observation explains a noted pattern concerning which infinitives are acceptable in the tough-construction: it has been observed that tough-predicates tend to prefer “volitional” verbs in the lower clause (Dalrymple and King 2000; Nanni 1978)

    1. (i)
      figure ck

    How can John lack money in a tough way, or Mary want that expensive dress easily, or the teacher prefer the hardcover edition in a hard way? Tough-constructions are overall felicitous when the event of the for-CP is felicitous with the tough-predicate as a modifier, i.e., the event described by the infinitive bears the property ascribed to the event in the main clause.

  35. I am using the term “content” in a slightly different way than it is employed by Kratzer (2013). In her paper, the “content” function, fcont, refers strictly to epistemic modality, while factuality modality uses ffact. I am using the term “content” more broadly to refer to any event which has associated propositions, whether circumstantial or epistemic. The crucial factor here is that such modal bases are “anchored” to the actual world. In the proposal in Grano (2016), a Root function is employed which is only defined for non-epistemic worlds. It is identical to what is proposed below, except that it leaves out the counterpart relation assumed in Kratzer’s proposal.

  36. For readability, I will leave out the interaction of PED and the judge argument. Note that PED probably has other (good) consequences, in particular for the tense specifications of the matrix and nonfinite clause. It may help explain the fact that for-CPs in the tough-construction are “tensed” nonfinite clauses (Stowell 1982; Wurmbrand 2014). This requires a more thorough investigation than can be given here.

  37. It is also independent of whether we adopt trans-world events or not. In a theory which permits trans-world events, we would still need Hacquard’s Event Preservation. I will continue to use world-bound events simply because there is a need for multiple events, as the data in (66) demonstrate. We could also formulate this by expanding on the idea of an event having sub-events, which would be consistent with Hacquard’s proposal. I see no strong reason to prefer one option over the other, so I will stick with what I believe is the simpler theory.

  38. The idea of thematic underspecification has precedent in Beavers (2006, 2010, 2011), who defines an “unspecified” event relation as one that simply maps a participant to an event; the participant is not affected by (and does not affect) the occurrence/outcome of the event. See further discussion in Sect. 7 below.

  39. A reviewer notes that, as defined, Participant works differently from a more standard definition for a thematic relation, like Agent, which specifies a unique individual as Agent of the event, Agent(e)(w)=x. Participant only makes an existential claim: “x is an event participant.” However, as I discuss in Sect. 7, it is ultimately necessary to be more specific about how precisely the individual is an event participant. We need to specify a precise way in which the individual is an event participant, which in turn likely necessitates a more standard understanding of this thematic relation.

  40. I make the standard assumption that the distribution of each Voice head is subject to lexical semantic considerations. For example, murder allows VoiceAgent but not VoiceCause, while kill allows either. Tough-predicates impose no thematic restrictions on their subjects, or indeed any argument, and thus may freely merge VoiceParticipant, but because other predicates dictate their argument structure, VoiceParticipant will be constrained in when it can be merged. Note that tough-predicates may in fact impose other restrictions; see Fleisher’s (2014) rare-class predicates discussed earlier. I note that the argument-introducing head may be better categorized as an Applicative head, though the difference is not important here. Note that the specific issue of thematic role disappears in a theory that eschews the ontological status of thematic roles in general, e.g. Brody (1993), but the principles behind the analysis remain, namely, that the tough-subject/proleptic object is an event participant and that this relation must hold across modal worlds. As I discuss in Sect. 7, it is likely that the tough-construction is not sensitive to particular thematic roles at all, but instead is tied to a more general notion of “intrinsic event relation,” more commonly thought of as the argument/adjunct distinction.

  41. As a concrete example, note that instrumental phrases may be introduced as with-prepositional phrases or as subjects. The swarm-alternation also exhibits this pattern.

    1. (i)
      figure cx
    1. (ii)
      figure cy

    The alternation found in the tough-construction between with and Voice is not accidental, and is moreover sensitive to the lexical semantics of the predicate. Both with and Voice are heads that relate individuals to events in general, and can alternate where possible. It is, moreover, worth pointing out that the fact that with is what introduces a proleptic object in the tough-construction is also not accidental. With in general can be used as an event modifier, introducing instrument, manner, and comitative phrases. Taking the choice of preposition seriously suggests that the preposition used to introduce a proleptic object with finite clauses (of and about) suggests a different relationship between the proleptic object and the embedding predicate. Since of/about in general are associated with nominal syntax, prolepsis with finite clauses should be defined with respect to nominal syntax and semantics. This is of course consistent with the observation that finite clause embedding involves nominal semantics, as argued in Moulton (2009) and Elliott (2017).

  42. A reviewer asks why this abstraction cannot target a subject position. This is a well-known issue, and I will have nothing new to say about it. One option is that there is a conflict between syntactic and semantic requirements here: syntax requires that a subject displace to an A-position higher in the clause (i.e., the EPP), but an operator is only capable of A′-movement. Directly moving the operator to spec-CP would violate the EPP, while moving the operator first to spec-TP would require A-movement of an operator.

  43. One way that the argument/adjunct alternation in the tough-construction clearly patterns like other argument/adjunct alternations is that with some alternations, negative quantifiers are only allowed in the argument position:

    1. (i)
      figure dc
    1. (ii)
      figure dd

    I will not attempt to address this precise fact, but see the discussion in Sect. 7 and references therein.

  44. Note that this extends to cases of prolepsis which cannot otherwise be gapped, for independent reasons, like possessives. (ia) is infelicitous because mailing a picture of Mary is an event that does not (necessarily) involve Mary, while taking a picture of her does.

    1. (i)
      figure de

    Of course, (ia) does have a meaning as long as Mary can be understood to be involved in the event of mailing, for instance if she is doing something to hinder the speaker’s ability to mail her picture. I discuss this more in Sect. 7.

  45. A reviewer notes possible cases of prolepsis with certain, like (i), that have finite clauses and with.

    1. (i)
      figure df

    These are different from “true” cases of prolepsis with finite clauses. Rather, (i) is likely a hanging topic that has been post-posed: With this recent update, you can be certain that it should be available soon. If (i) is a hanging topic, then perhaps it may be explained along the lines speculatively suggested in footnote 50.

  46. Kayne (1989:251): “In both Italian and French, the ‘easy-to-please’ construction is possible with two levels of embedding only if the highest infinitive is of the class of verbs compatible with a CP complement having an empty head bound from without [= a restructuring predicate].”

  47. As noted earlier, there are also cases of unbounded dependencies that involve non-restructuring predicates; for many speakers, extraction out of at least some finite clauses is acceptable. Judgments in (i) are from the source.

    1. (i)
      figure dh

    Such examples may be accounted for by assuming that the tough-subject is related to the gap via prolepsis inside the nonfinite clause. That is, (ia) means something like “That problem is such that it is easy to think about it that you understand it.” (See e.g. Salzmann 2017b: Sect. 4.5.2.2.) Suggesting that this is on the right track, I note that explicitly giving the nonfinite embedding verb a different proleptic object significantly degrades the sentences.

    1. (ii)
      figure di

    We may view the speaker variation found with such sentences as a result of variation in how willing speakers are to accommodate such prolepsis in the lower clause. Still, given the level of speaker variation concerning the acceptability of long-distance extraction, I am wary of making too strong a claim about the analysis of such sentences. Moreover, if the analysis of (ia-c) suggested above is on the right track, then it rests on an analysis of prolepsis with finite clauses, which I cannot explore here.

  48. A reviewer points out the interesting cases noted in Grano and Lasnik (2018) where a bound pronoun may ameliorate the ungrammaticality found in certain long-distance dependencies, including the tough-construction.

    1. (i)
      figure dj

    I have nothing to say about these observations other than to point out, first, that the examples have parallel acceptance with their proleptic object variants, as seen in (iia) and (iib), and second, that the judgments do not generalize to other embedding verbs, as seen in (iic) and (iid). (This appears to distinguish the tough-construction from other cases, as experimentally shown by Grano and Lasnik.)

    1. (ii)
      figure dk

    Clearly, something about the bound pronoun in the lower clause suggests a “restructuring effect” between the two clauses for certain verbs, though I do not know how such an effect should be properly articulated or formalized.

  49. Similarly, more discussion should be given to whether the analysis may cover the fact that amounts and predicates are not permitted (in (28) and (29)). It seems that predicates are correctly ruled out: how can a predicate be an event participant? It is less clear whether amounts are excluded in the same way, though see discussion in the next section concerning the argument/adjunct distinction.

  50. A reviewer notes that the following example from Partee (1970) might be challenging to the analysis, since it involves coordination of a non-tough-predicate and a tough-predicate. This ought to lead to a conflict in thematic requirements.

    1. (i)
      figure dn

    I do not think (i) is particularly difficult to explain, since it seems plausible that explicit assigns an underspecified thematic role to its subject. However, the point is well taken that, in theory, sentences with coordinated predicates each of which assigns a (different) thematic role to the subject may raise issues. Ultimately, though, the treatment of such sentences hinges on the proper understanding of Across-the-Board movement, which is known to display unique and poorly understood “asymmetries” in the interpretation of the gap position (Munn 1994; Zhang 2010; Salzmann 2012). The same reviewer also wonders whether examples like (ii) are not even more problematic.

    1. (ii)
      figure do

    As far as I can tell, (ii) is not (uniquely) problematic for the proposal above either, though it raises questions about how coordinated propositions are handled on the assumption that C houses modal meaning. This is a good question, but it is again not one that I have the space to address here.

  51. By ‘argument’ I mean “a constituent that is syntactically selected,” and by ‘adjunct’ I mean “a constituent that is not syntactically selected.” The terms ‘argument’ and ‘adjunct’ are highly contentious, having changed meaning over time as theories and assumptions changes—as the discussion in this section emphasizes. Neither term is settled: there are “semantic arguments” and “syntactic arguments,” and there are optional adjuncts and adjuncts which are obligatory (Sailor and Schütze 2013). As I discuss below, I take (I believe) a more realistic view of the argument/adjunct distinction: it is gradient, not categorical. In the end, for me the terms ‘argument’ and ‘adjunct’ are descriptive rather than formal labels, equivalent to Beavers’ (cited below) use of ‘direct’ versus ‘oblique,’ respectively.

  52. For instance, Schütze (1995) observes the ungrammaticality of (i), though this particular example is accounted for on the above analysis given that the tough-subject is not an argument of the event of leaving.

    1. (i)
      figure dp

  53. Alternatively, we might suppose that all adjuncts are Late-merged (Stepanov 2001), and so are unable to enter into a dependency with the tough-subject. I am not aware of any work that has specifically applied this idea to the tough-construction.

  54. Cinque (1990:152) suggests that the tough-construction is unique in this regard even among other complement object deletions, most of which permit resumptive pronouns and gapless nonfinite clauses (save pretty-class predicates). My own judgments concerning the general class of complement object deletions is that, when there is a gap, it is sensitive to the same argument/adjunct asymmetry. The analysis proposed below may be one way to explain these facts, though these cases deserve their own separate study; see also Mulder and den Dikken (1992:306, footnote 5) for a similar point.

  55. The issue is not solved by adopting a movement account. See Fleisher (2013:327ff.) for discussion.

  56. Ultimately, the account offered here is similar in spirit (but not formalism) to Rizzi’s (1990) understanding of weak islands, which, in his view, can be reduced to an understanding of how thematic roles are syntactically assigned. In more recent work, the ideas below are consistent with Truswell’s (2011) view of weak islands, which relates weak islands to how events across clauses are related and perceived. However, I also stress that I do not intend this analysis here to be an account of all weak islands. I am concerned solely with the facts of the tough-construction.

  57. I feel it is worth mentioning that some of the verbs which Beavers lists as selecting for an unspecified object actually participate in an argument/adjunct alternation, suggesting that there are different “levels” of the unspecified event relation. This is what the account of the tough-construction below suggests as well.

    1. (i)
      figure ed
    1. (ii)
      figure ee

  58. Beavers’ formulation of the argument/adjunct distinction makes a clear prediction which I do not have the space to explore in full: in any argument/adjunct alternation, if the adjunct version can be gapped in the tough-construction, the argument version can as well. But the reverse should not hold. We should find instances in which the argument version can be gapped, but the adjunct cannot. And indeed, we find that only the Theme in the spray/load alternation is possible as a gap in the tough-construction, as seen in (i). Similarly, with verbs of consumption, only the argument can be gapped, as seen in (ii). There are also cases where both positions can be gapped, shown in (iii). But there are no cases among the verbs listed in Beavers’ work in which the adjunct is gappable but the argument is not.

    1. (i)
      figure ef
    1. (ii)
      figure eg
    1. (iii)
      figure eh

    This follows from the one-way entailment of Beavers’ analysis: the argument version is equivalent to or (monotonically) stronger than the adjunct version. Note that some argument/adjunct alternations are impossible to test because there are additional confounds that rule one of the sentences out. For instance, applicative benefactives are not possible in the tough-construction, as (iv) shows, but this is due to the fact, illustrated in (v), that indirect objects in general cannot be A′-extracted in American English.

    1. (iv)
      figure ei
    1. (v)
      figure ej

    The instrumental and passive alternations discussed in the text are also confounded, in that the adjunct alternates with a subject.

  59. I assume that pronouns in argument positions (i.e., *The knife was easy to cut with it) are ruled out due to standard economy principles (e.g., Avoid Pronoun; Chomsky 1981). When possible, gaps are preferred over pronouns. The opposite does not hold: a proleptic object can in principle target an argument position. This follows from the analysis in the text since a stronger event relation entails the weaker one. It was easy with this knife to cut with it is grammatical because the argument event relation associated with the pronoun entails the adjunct event relation associated with the proleptic object.

  60. This general discussion is also related to what Lasnik and Fiengo (1974:548, footnote 6) observe concerning the relationship between pretty-class predicates with and without a nonfinite clause. They point out that the expected entailment relationship between a pretty-class adjective with and without a nonfinite clause does not always hold, noting that The floor is slippery to dance on __ does not entail The floor is slippery, or else we would expect a contradiction in The floor is slippery to dance on __ but fine to walk on __. This then suggests that in the presence of a nonfinite clause, the thematic role attributed to the subject of a pretty-class adjective is different than when the subject is only predicated of the adjective.

  61. This oversimplifies the facts of control among tough-predicates. Space prevents a close examination of control in light of the proposal above. Broadly, PRO with tough-predicates exhibits (implicit) obligatory control: PRO in the lower clause must be controlled by some argument of the higher clause, whether overt or implicit. That is, PRO must be controlled by a particular event participant in the higher clause. Because of this restriction, the subject position is thus exempt from being controlled by any other element in the clause, e.g., the tough-subject. I must assume therefore that the “calculus” of control is distinct from that of operator–gap dependencies in general.

References

  • Anand, Pranav, and Natasha Korotkova. 2018. Acquaintance content and obviation. In Proceedings of Sinn und Bedeutung 22, eds. Uli Sauerland and Stephanie Solt, vol. 1, 55–72. Berlin: Leipniz Centre General Linguistics.

    Google Scholar 

  • Arregui, Ana. 2007. When aspect matters: The case of would-conditionals. Natural Language Semantics 15(3): 221–264.

    Article  Google Scholar 

  • Bach, Emmon. 1986. The algebra of events. Linguistics and Philosophy 9(1): 5–16.

    Article  Google Scholar 

  • Baker, Mark. 1988. Incorporation: A theory of grammatical function changing. Chicago: The University of Chicago Press.

    Google Scholar 

  • Bayer, Samuel. 1990. Tough movement as function composition. In Proceedings of the Ninth West Conference on Formal Linguistics, ed. Aaron Halpern, 29–42. Stanford: CSLI Publications.

    Google Scholar 

  • Beavers, John. 2006. Argument/oblique alternations and the structure of lexical meaning, PhD thesis, Stanford University.

  • Beavers, John. 2010. The structure of lexical meaning: Why semantics really matters. Language 86(4): 821–864. https://www.jstor.org/stable/40961719.

    Article  Google Scholar 

  • Beavers, John. 2011. On affectedness. Natural Language & Linguistic Theory 29: 335–370. https://doi.org/10.1007/s11049-011-9124-6.

    Article  Google Scholar 

  • Bhatt, Rajesh. 1999. Covert modality in non-finite contexts, PhD thesis, University of Pennsylvania.

  • Bhatt, Rajesh. 2006. Covert modality in non-finite contexts. Berlin: de Gruyter.

    Book  Google Scholar 

  • Bogal-Albritten, Elizabeth. 2016. Building meaning in Navajo, PhD thesis, University of Massachusetts at Amherst.

  • Bresnan, Joan. 1971. On sentence stress and syntactic transformations. Language 47: 257–281.

    Article  Google Scholar 

  • Bresnan, Joan. 1972. Theory of complementation in English syntax, PhD thesis, MIT.

  • Brody, Michael. 1993. Theta-theory and arguments. Linguistic Inquiry 24(1): 1–23.

    Google Scholar 

  • Browning, Marguerite. 1987. Null operator constructions, PhD thesis, MIT.

  • Bruening, Benjamin. 2014. Defects of defective intervention. Linguistic Inquiry 45: 707–719. https://doi.org/10.1162/LING_a_00171.

    Article  Google Scholar 

  • Bylinina, Lisa. 2017. Judge dependence in degree constructions. Journal of Semantics 34: 291–331. https://doi.org/10.1093/jos/ffw011.

    Article  Google Scholar 

  • Cable, Seth. 2011. A new argument for lexical decomposition: Transparent readings of verbs. Linguistic Inquiry 42(1): 131–138. https://www.jstor.org/stable/23011911.

    Article  Google Scholar 

  • Chomsky, Noam. 1973. Conditions on transformations. In A Festschrift for Morris Halle, eds. Stephen Anderson and Paul Kiparsky, 232–286. New York: Rinehart and Winston.

    Google Scholar 

  • Chomsky, Noam. 1977. On Wh-movement. In Formal syntax, eds. Peter Culicover, Tom Wasow, and Adrian Akmajian, 71–132. San Diego: Academic Press.

    Google Scholar 

  • Chomsky, Noam. 1981. Lectures on government and binding. Dordrecht: Foris.

    Google Scholar 

  • Cinque, Gennaro. 1990. Types of A′-dependencies. Cambridge, MA: MIT Press.

    Google Scholar 

  • Collins, John. 2013. The syntax of personal taste. Philosophical Perspectives 27: 55–103. https://www.jstor.org/stable/26382670.

    Article  Google Scholar 

  • Comrie, Bernard, and Stephen Matthews. 1990. Prolegomena to a typology of Tough Movement. In Studies in typology and diachrony: Papers presented to Joseph H. Greenberg on this 75th birthday, eds. William Croft, Suzanne Kemmer, and Keith Denning, 43–58. Amsterdam: Benjamins.

    Chapter  Google Scholar 

  • Creissels, Denis. 2014. Crosslinguistic variation in the treatment of beneficiaries and the argument vs. adjunct distinction. Linguistic Discovery 12(2): 41–55. https://doi.org/10.1349/PS1.1537-0852.A.445.

    Article  Google Scholar 

  • Dalrymple, Mary, and Tracy Holloway King. 2000. Missing-object constructions: Lexical and constructional variation. In Proceedings of the LFG00 conference, eds. Miriam Butt and Tracy Holloway King, 82–103. Stanford: CSLI Publications.

    Google Scholar 

  • Davies, William D. 2005. Madurese prolepsis and its implications for a typology of raising. Language 81: 645–665. https://www.jstor.org/stable/4489968.

    Article  Google Scholar 

  • Dowty, David. 1991. Thematic proto-roles and argument selection. Language 67(3): 547–619.

    Article  Google Scholar 

  • Dowty, David. 2001. The semantic asymmetry of ‘Argument Alternations’ (and why it matters). GAGL: Gröninger Arbeiten zur germanistischen Linguistik 44: 171–186.

    Google Scholar 

  • Duffley, Patrick. J. 2003. The gerund and the to-infinitive as subject. Journal of English Linguistics 31(4): 324–352. https://doi.org/10.1177/0075424203257833.

    Article  Google Scholar 

  • Elliott, Patrick. 2017. Elements of clausal embedding, PhD thesis, University College London.

  • Epstein, Samuel. 1989. Quantification in null operator constructions. Linguistic Inquiry 20(4): 647–658.

    Google Scholar 

  • Fleisher, Nicholas. 2013. On the absence of scope reconstruction in tough subject A-chains. Linguistic Inquiry 44(2): 321–332. https://doi.org/10.1162/LING_a_00130.

    Article  Google Scholar 

  • Fleisher, Nicholas. 2014. Rare-class adjectives in the tough-construction. Language 91: 73–108. https://www.jstor.org/stable/24672221.

    Article  Google Scholar 

  • Forker, Diane. 2014. A canonical approach to the argument/adjunct distinction. Linguistic Discovery 12(2): 27–40. https://doi.org/10.1349/PS1.1537-0852.A.444.

    Article  Google Scholar 

  • Gluckman, John. 2018. Perspectives on syntactic dependencies, PhD thesis, University of California, Los Angeles.

  • Gluckman, John. 2019. The natural class of tough-predicates, and non-finite clauses. In Proceedings of the 36th West Coast Conference on Formal Linguistics, eds. Richard Stockwell, Maura O’Leary, Zhongshi Xu, and Z. L. Zhou, 149–158. Somerville: Cascadilla Press.

    Google Scholar 

  • Gluckman, John. 2021. Taking time with the tough-construction. Journal of Linguistics, First View. https://doi.org/10.1017/S002222672100013X.

    Article  Google Scholar 

  • Goh, Gwang-Yoon. 2000. Pragmatics of the English tough-construction. In Proceedings of the North East Linguistics Society 30, eds. Masako Hirotani, Andries Coetzee, Nancy Hall, and Ji Yung Kim, 219–230. Amherst, MA: GLSA.

    Google Scholar 

  • Grano, Thomas. 2016. Semantic consequences of syntactic subject licensing: Aspectual predicates and concealed modality. In Proceedings of Sinn und Bedeutung 20, 306–322. https://doi.org/10.18148/sub/2016.v20i0.265.

    Chapter  Google Scholar 

  • Grano, Thomas, and Howard Lasnik. 2018. How to neutralize a finite clause boundary: Phase theory and the grammar of bound pronouns. Linguistic Inquiry 49(3): 465–499. https://doi.org/10.1162/ling_a_00279.

    Article  Google Scholar 

  • Grimshaw, Jane. 1990. Argument structure. Cambridge, MA: MIT Press.

    Google Scholar 

  • Grover, Claire. 1995. Rethinking some empty categories: Missing objects and parasitic gaps in HPSG, PhD thesis, University of Essex.

  • Hackl, Martin, and Jon Nissenbaum. 2012. A modal ambiguity in for-infinitival relative clauses. Natural Language Semantics 20: 59–81. https://doi.org/10.1007/s11050-011-9075-9.

    Article  Google Scholar 

  • Hacquard, Valentine. 2006. Aspects of modality, PhD thesis, MIT.

  • Hacquard, Valentine. 2009. On the interaction of aspect and modal auxiliaries. Linguistics and Philosophy 32(3): 279–315. https://doi.org/10.1007/s10988-009-9061-6.

    Article  Google Scholar 

  • Hacquard, Valentine. 2011. Modality. In Semantics: An international handbook of natural language meaning 2, eds. Klaus Von Heusinger, Claudia Maienborn, and Paul Portner, 1484–1515. Berlin: de Gruyter.

    Google Scholar 

  • Hartman, Jeremy. 2011. (Non-)Intervention in A-movement. Linguistic Variation 11(2): 121–138. https://doi.org/10.1075/lv.11.2.01har.

    Article  Google Scholar 

  • Hartman, Jeremy. 2012. Varieties of clausal complementation, PhD thesis, MIT.

  • Haspelmath, Martin. 2014. Arguments and adjuncts as language-particular syntactic categories and as comparative concepts. Linguistic Discovery 12(2): 3–11.

    Article  Google Scholar 

  • Heim, Irene. 1987. Where does the definiteness restriction apply? Evidence from the definitness of variables. In The representation of (in)definites, eds. Eric Reuland and Alice G. B. ter Meulen, 21–42. Cambridge, MA: MIT Press.

    Google Scholar 

  • Hedberg, Nancy, and Richard C. DeArmond. 2009. On complements and adjuncts. Snippets 19: 11–12.

    Google Scholar 

  • Heim, Irene, and Angelika Kratzer. 1998. Semantics in generative grammar. Malden: Blackwell.

    Google Scholar 

  • Heycock, Caroline. 1991. Layers of predication: The non-lexical status of clauses, PhD thesis, University of Pennsylvania.

  • Hicks, Glyn. 2009. Tough-constructions and their derivations. Linguistic Inquiry 40(4): 535–566. https://doi.org/10.1162/ling.2009.40.4.535.

    Article  Google Scholar 

  • Höhle, Tilman. 1978. Aktiv-Passiv-Relation und andere Infinitkonstruktionen im Deutschen. Berlin: de Gruyter.

    Google Scholar 

  • Holl, Daniel. 2012. Modale Infinitive und dispositionelle Modalität im Deutschen. Berlin: Akademie Verlag.

    Google Scholar 

  • Hornstein, Norbert. 2001. Move! A minimalist theory of construal. Oxford: Blackwell.

    Google Scholar 

  • Inoue, Kazuko. 1978. ‘Tough’ sentences in Japanese. In Problems in Japanese syntax and semantics, eds. John Hinds and Irwin Howard, 122–154. Tokyo: Kaitakusha.

    Google Scholar 

  • Inoue, Kazuko. 2004. Japanese ‘tough’ sentences revisited. Scientific Approaches to Language 3: 75–111.

    Google Scholar 

  • Jacobson, Pauline. 1992. The lexical entailment theory of control and the tough-construction. In Lexical matters, eds. Ivan Sag and Anna Szabolsci, 269–300. Stanford: CSLI Publications.

    Google Scholar 

  • Jones, Charles. 1991. Purpose clause: Syntax, thematics, and semantics of English purpose constructions. Dordrecht: Kluwer.

    Book  Google Scholar 

  • Kaplan, Ronald, and Joan Bresnan. 1982. Lexical-functional grammar: A formal system for grammatical representation. In The mental representation of grammatical relations, ed. Joan Bresnan, 173–281. Cambridge, MA: MIT Press.

    Google Scholar 

  • Kayne, Richard S. 1989. Null subjects and clitic climbing. In The null subject parameter, eds. Jaeggli Osvaldo and Ken Safir, 239–261. Dordrecht: Kluwer.

    Chapter  Google Scholar 

  • Keine, Stefan, and Ethan Poole. 2017. Intervention in tough-constructions revisited. The Linguistic Review 34(2): 295–329. https://doi.org/10.1515/tlr-2017-0003.

    Article  Google Scholar 

  • Kennedy, Chris. 2007. Vagueness and grammar: The semantics of relative and absolute gradable adjectives. Linguistics and Philosophy 30: 1–45. https://doi.org/10.1007/s10988-006-9008-0.

    Article  Google Scholar 

  • Klingvall, Eva. 2018. Agreement and reconstruction correlate in Swedish. Natural Language & Linguistic Theory 36: 1165–1205. https://doi.org/10.1007/s11049-017-9398-4.

    Article  Google Scholar 

  • Kratzer, Angelika. 1996. Severing the external argument from its verb. In Phrase Structure and the Lexicon, eds. Johan Rooryck and Laurie Zaring, 109–138. Dordrecht: Kluwer.

    Chapter  Google Scholar 

  • Kratzer, A. 2006. Decomposing attitude verbs. Manuscript, University of Massachusetts at Amherst. Accessed online at http://semanticsarchive.net/Archive/DcwY2JkM/attitude-verbs2006.pdf.

  • Kratzer, Angelika. 2013. Modality for the 21st century. Presented at The Language-Cognition Interface: 19th International Congress of Linguists, Geneva, July 2013.

    Google Scholar 

  • Landau, Idan. 2011. Predication vs. aboutness in copy raising. Natural Language & Linguistics Theory 29(3): 779–813. https://doi.org/10.1007/s11049-011-9134-4.

    Article  Google Scholar 

  • Lasersohn, Peter. 2005. Context dependence, disagreement, and predicates of personal taste. Linguistics and Philosophy 28(6): 643–686. http://www.jstor.org/stable/25001976.

    Article  Google Scholar 

  • Lasnik, Howard, and Robert Fiengo. 1974. Complement object deletion. Linguistic Inquiry 5(4): 535–571.

    Google Scholar 

  • Lee, Sun-Hee. 2003. Korean tough constructions and double nominative constructions. In The proceedings of the 9th international conference on HPSG, eds. Jong-Bok Kim and Stephen Wechsler, 187–208. Stanford: CSLI Publications.

    Google Scholar 

  • Levine, Robert, and Thomas E. Hukari. 2006. The unity of unbounded dependency constructions. Stanford: CSLI Publications.

    Google Scholar 

  • Lewis, David. 1968. Counterpart theory and quantified modal logic. The Journal of Philosophy 65(5): 113–126.

    Article  Google Scholar 

  • Lewis, David. 1971. Counterparts of persons and their bodies. The Journal of Philosophy 68(7): 203–211.

    Article  Google Scholar 

  • Lewis, David. 1983. Survival and identity. In The Identities of Persons, ed. Amelie Oksenberg Rorty, 17–40. Berkeley: University of California Press.

    Google Scholar 

  • Lewis, David. 1986. On the plurality of worlds. Oxford: Blackwell.

    Google Scholar 

  • Longenbaugh, Nicholas. 2016a. Rethinking the A/A′-distinction: Evidence from English tough-movement. Presented at GLOW 39, April 7, Göttingen, Germany.

  • Longenbaugh, Nicholas. 2016b. Non-intervention in tough-constructions. In Proceedings of the North East Linguistics Society (NELS) 46, eds. Christopher Hammerly and Brandon Prickett, 206–293. Amherst, MA: GLSA Publications.

    Google Scholar 

  • Manning, Christopher. 2003. Probabilistic syntax. In Probabilistic Linguistics, eds. Rens Bod, Jennifer Hay, and Stefanie Jannedy. Cambridge, MA: MIT Press.

    Google Scholar 

  • MacFarlane, John. 2014. Assessment sensitivity: Relative truth and its applications. Oxford: Oxford University Press.

    Book  Google Scholar 

  • Moulton, Keir. 2009. Natural selection and the syntax of clausal complementation, PhD thesis, University of Massachusetts at Amherst.

  • Moulton, Keir. 2015. CPs: Copies and compositionality. Linguistic Inquiry 46(2): 305–342. https://doi.org/10.1162/LING_a_00183.

    Article  Google Scholar 

  • Mulder, Rene, and Marcel den Dikken. 1992. Tough parasitic gaps. In The Proceedings of the North East Linguistics Society (NELS) 22, ed. Kimberley Broderick, 303–317. Amherst, MA: GLSA Publications.

    Google Scholar 

  • Munn, Alan. 1994. A minimalist account of reconstruction asymmetries. In Proceedings of the North East Linguistic Society (NELS) 24, ed. Mercé Gonzàlez, 397–410. Amherst, MA: GSLA Publications.

    Google Scholar 

  • Nanni, Deborah. 1978. The ‘easy’ class of adjectives in English, PhD thesis, University of Massachusetts at Amherst.

  • Nanni, Deborah. 1980. On the surface syntax of constructions with easy-type adjectives. Language 56(3): 568–581.

    Article  Google Scholar 

  • Nissenbaum, Jon. 2000. Investigations of covert phrase movement, PhD thesis, MIT.

  • Partee, Barbara. 1970. Negation, conjunction, and quantifiers: Syntax vs. semantics. Foundations of Language 6: 153–165.

    Google Scholar 

  • Pearson, Hazel. 2013. A judge-free semantics for predicates of personal taste. Journal of Semantics 30: 103–154. https://doi.org/10.1093/jos/ffs001.

    Article  Google Scholar 

  • Perlmutter, David, and Scott Soames. 1979. Syntactic argumentation and the structure of English. Berkeley: University of California Press.

    Book  Google Scholar 

  • Pesetsky, David. 1987. Binding problems with experiencer verbs. Linguistic Inquiry 18(1): 203–238.

    Google Scholar 

  • Pesetsky, David. 1991. Zero Syntax II: An essay on infinitives. Manuscript, MIT.

  • Pesetsky, David. 2013. Phrasal movement and its discontents: Diseases and diagnostics. In Diagnostics in syntax, eds. Lisa Cheng and Norbert Corver, 123–157. Oxford: Oxford University Press.

    Google Scholar 

  • Pietroski, Paul. M. 2000. On explaining that. The Journal of Philosophy 97(12): 655–662.

    Article  Google Scholar 

  • Poole, Ethan, Stefan Keine, and Jon Ander Mendia. 2017. More on (the lack of) reconstruction in tough-constructions. Manuscript, UCLA, USC, and University of Düsseldorf. Accessed online at https://ling.auf.net/lingbuzz/003702.

  • Portner, Paul. 1997. The semantics of mood, complementation, and conversational force. Natural Language Semantics 5: 167–212.

    Article  Google Scholar 

  • Portner, Paul. 2009. Modality. Oxford: Oxford University Press.

    Google Scholar 

  • Postal, Paul. 1971. Cross-over phenomena. New York: Holt, Rinehart and Winston.

    Google Scholar 

  • Postal, Paul 1974. On raising. Cambridge, MA: MIT Press.

    Google Scholar 

  • Pustejovsky, James. 1996. The generative lexicon. Cambridge, MA: MIT Press.

    Google Scholar 

  • Pylkkänen, Lina. 2008. Introducing arguments. Cambridge, MA: MIT Press.

    Book  Google Scholar 

  • Rezac, Milan. 2006. On tough movement. In Minimalist essays, ed. Cedric Boeckx, 288–325. Amsterdam: Benjamins.

    Chapter  Google Scholar 

  • Rizzi, Luigi. 1982. Issues in Italian syntax. Dordrecht: Foris.

    Book  Google Scholar 

  • Rizzi, Luigi. 1990. Relativized minimality. Cambridge, MA: MIT Press.

    Google Scholar 

  • Rothstein, Susan. 2004. Structuring events: A study in the semantics of lexical aspect. Oxford: Blackwell.

    Book  Google Scholar 

  • Runner, Jeffrey. 2006. Lingering challenges to the raising-to-object and object-control constructions. Syntax 6(2): 193–213.

    Article  Google Scholar 

  • Sæbø, Kjell Johan. 2009. Judgment ascriptions. Linguistics and Philosophy 32: 327–352. https://doi.org/10.1007/s10988-009-9063-4.

    Article  Google Scholar 

  • Sailor, Craig, and Carson Schütze. 2013. What is an adjunct? Manuscript, UCLA. Accessed online at https://ling.auf.net/lingbuzz/002180/current.pdf.

  • Salzmann, Martin. 2006. Resumptive prolepsis: A study in indirect A′-dependencies, PhD thesis, Universiteit Leiden.

  • Salzmann, Martin. 2012. A derivational ellipsis approach to ATB-movement. The Linguistic Review 29: 397–438.

    Article  Google Scholar 

  • Salzmann, Martin. 2017a. Prolepsis, 2nd edn. In The Wiley Blackwell companion to syntax, eds. Martin Everaert and Hans Riemsdijk. https://doi.org/10.1002/9781118358733.wbsyncom062.

    Chapter  Google Scholar 

  • Salzmann, Martin. 2017b. Reconstruction and resumption in indirect A′-dependencies: On the syntax of prolepsis and relativization in (Swiss) German and beyond. Berlin: de Gruyter.

    Chapter  Google Scholar 

  • Schachter, Paul. 1981. Lovely to look at. Linguistic Analysis 8: 431–448.

    Google Scholar 

  • Schütze, Carson. 1995. PP attachment and argumenthood. In MIT Working Papers in Linguistics 26, eds. C. Schütze, J. Ganger and K. Broihier, 95–151.

    Google Scholar 

  • Sigurðsson, Einar Freyr. 2016. Circumventing the ban on improper movement in Icelandic. In Proceedings of the 46th annual meeting of the North East Linguistic Society, eds. Christopher Hammerly and Brandon Prickett, 179–192. Amherst, MA: GLSA Publications.

    Google Scholar 

  • Stepanov, Artur. 2001. Late adjunctions and minimalist phrase structure. Syntax 4(2): 94–125. https://doi.org/10.1111/1467-9612.00038.

    Article  Google Scholar 

  • Stephenson, Tamina. 2007. Judge dependence, epistemic modals, and predicates of personal taste. Linguistics and Philosophy 30: 487–525. https://doi.org/10.1007/s10988-008-9023-4.

    Article  Google Scholar 

  • Stephenson, Tamina. 2010. Control in centred worlds. Journal of Semantics 27: 409–436. https://doi.org/10.1093/jos/ffq011.

    Article  Google Scholar 

  • Stowell, Tim. 1982. The tense of infinitives. Linguistic Inquiry 13(3): 561–570.

    Google Scholar 

  • Stowell, Tim. 1991. The alignment of arguments in adjective phrases. In Perspectives on phrase structure: Heads and licensing, ed. Susan Rothstein, 105–138. San Diego: Academic Press.

    Google Scholar 

  • Tayalati, Fayssal, and Lieven Danckaert. 2020. The syntax and semantics of Modern Standard Arabic resumptive tough constructions. Folia Linguistica 54(1): 197–238.

    Article  Google Scholar 

  • Truswell, Robert. 2011. Events, phrases, and questions. Oxford: Oxford University Press.

    Book  Google Scholar 

  • Van Valin, Robert, and Randy J. LaPolla. 1997. Syntax: Form, meaning, and function. Cambridge, MA: Cambridge University Press.

    Book  Google Scholar 

  • Wichmann, Søren. 2014. Arguments and adjuncts crosslinguistically: A brief introduction. Linguistic Discovery 12(2): 1–2.

    Article  Google Scholar 

  • Wilder, Christopher. 1991. Tough movement constructions. Linguistische Berichte 132: 115–133.

    Google Scholar 

  • Williams, Edwin. 1983. Semantic vs. syntactic categories. Linguistics and Philosophy 6(3): 336–423.

    Google Scholar 

  • Williams, Edwin. 2003. Representation theory. Cambridge, MA: MIT Press.

    Google Scholar 

  • Wurmbrand, Suzanne. 2001. Infinitives: Restructuring and clause structure. Berlin: de Gruyter.

    Google Scholar 

  • Wurmbrand, Suzanne. 2014. Tense and aspect in English infinitives. Linguistic Inquiry 45(3): 403–447.

    Article  Google Scholar 

  • Yeo, Seungju. 2005. The syntax of the tough-construction in Korean. Studies in Generative Grammar 15(1): 77–95.

    Google Scholar 

  • Zhang, Niina Ning. 2010. Coordination in syntax. Cambridge, MA: Cambridge University Press.

    Google Scholar 

Download references

Acknowledgements

Thanks to Dominique Sportiche, Yael Sharvit, Tim Stowell, Jesse Harris, Roumi Pancheva, Maayan Abenina-Adar, Margit Bowler, audiences at WCCFL 38 at UCLA, three anonymous reviewers for Syntax, and three anonymous reviewers for Natural Language Semantics.

Author information

Authors and Affiliations

  1. Department of Linguistics, University of Kansas, Blake Hall, Room 423, 1541 Lilac Lane, Lawrence, KS, 66045, USA

    John Gluckman

Authors
  1. John Gluckman
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to John Gluckman.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Gluckman, J. The meaning of the tough-construction. Nat Lang Semantics 29, 453–499 (2021). https://doi.org/10.1007/s11050-021-09181-3

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11050-021-09181-3

Keywords

Search

Navigation