Interpreting verb clusters


We argue for syntactic verb cluster formation in certain restructuring configurations, the result of which is interpreted via function composition. This cluster formation can be diagnosed by its semantic consequences. In particular, we observe that in these configurations all embedded elements must receive a matrix interpretation even if there is no evidence that these elements leave the embedded VP at any stage of the derivation. We show that verb cluster formation and function composition provide a solution to this puzzle. We propose that the process of cluster formation takes place whenever two lexical verbs are part of the same phasal Spell-Out domain, which we relate to Richards’ (2010) distinctness proposal. Our analysis entails that (i) some instances of head movement have semantic effects and hence cannot take place at PF; (ii) the set of rules of semantic composition must include function composition; and (iii) it provides additional support for the notion of distinctness and extends its application to head movement.

This is a preview of subscription content, access via your institution.


  1. 1.

    In the example sentences, we provide the grammatical case of a DP regardless of whether or where it manifests itself morphologically. In, e.g., (1a) den Traktor is glossed as ‘the tractor.acc’ despite the fact that morphologically the accusative case is realized on the determiner and not the head noun. In other examples (e.g., (9c)), case is not realized on a DP at all but can be deduced from verbal agreement. We do so to have a consistent glossing scheme and because the morphological expression of case is irrelevant for our concerns here.

  2. 2.

    A remark about the use of long passive constructions in this paper is in order: It is often acknowledged that for some speakers the long passive is rather marked. For other speakers it is entirely impossible. These facts are sometimes taken to show that the long passive either is not a proper part of German grammar (Höhle 1978, 177) or few theoretical conclusions can be drawn from it (Kiss 1995, 137). We decidedly disagree with this sentiment. First, the majority of the speakers we have consulted accept the construction. Second, for speakers whose grammar contains the long passives the semantic restrictions reported here and in the previous literature are remarkably clear and robust. Third, large scale data collections carried out by Wöllstein-Leisten (2001) and Schmid et al. (2005) confirm that the long passive is readily accepted by speakers. According to Wöllstein-Leisten (2001, 86), the set of verbs allowing long passives include ablehnen ‘reject’, anbieten ‘offer’, anfangen ‘start’, aufhören ‘stop’, beabsichtigen ‘intend’, beginnen ‘begin’, beschließen ‘decide’, geloben ‘pledge’, probieren ‘try’, vergessen ‘forget’, versuchen ‘try’, and vorschlagen ‘suggest’.

  3. 3.

    While we will primarily focus on the long passive here, verb cluster analyses are of course not limited to this construction (see, e.g., Evers 1975; Haegeman and van Riemsdijk 1986; Fanselow 1987; Bayer and Kornfilt 1990, 1994; von Stechow 1992; Grewendorf and Sabel 1994; Williams 2003).

    Complex head approaches have likewise been explored for the clause-final clusters of a lexical verb along with modals and auxiliary verbs. Various authors argue that these verbs are either base-generated as a complex head that subsequently combines with its thematic arguments (Steedman 1985; Jacobs 1992; Haider 1993, 2003; Hinrichs and Nakazawa 1994; Kiss 1995; Müller 1999, 2002, 2006; Meurers 2000; Williams 2003) or that this complex head is syntactically derived (Evers 1975; Haegeman and van Riemsdijk 1986; von Stechow and Sternefeld 1988; Bayer and Kornfilt 1990, 1994; Salzmann 2011, 2013). In this paper we will limit our attention to cases of prima facie VP/vP complementation, viz., the combination of various lexical verbs.

  4. 4.

    As is typically the case for dichotomies of this type, some accounts fall between the two poles. Sabel (1996), for instance, takes the infinitival complement to be a full-fledged CP, which is rendered transparent by incorporation of an embedded functional head into the matrix predicate. As such, this account relies on complex head formation but not between the two lexical verbs. Similarly, Wurmbrand (2013) takes the embedded complement to be a vP, the v head of which incorporates into the matrix verb.

  5. 5.

    The prosodic boundaries in (6b) is the unmarked option. For many speakers it is the only one. There are, interestingly, some speakers that marginally allow a long passive sentence to have a prosody similar to the local passives, viz., with a prosodic break between the two verbs. This issue is addressed in fn. 32.

  6. 6.

    Haider (2010, 307), who presents a verb cluster account, treats the topicalized verbal projection in (7a) as base-generated in its surface position, semantically linked to a silent verb in the base position which forms a cluster with the in-situ verb. As he himself notes, no current conception of movement is able to derive structures like (7) from a verb cluster structure. Moreover, this analysis requires the existence of pronominal verbs that can be bound by a verbal projection. It is not at all clear that there is evidence for such elements beyond the fact that a verb cluster account requires them.

  7. 7.

    Scenarios unambiguously diagnosing the scope relations in (10) are given in (i):

  8. 8.

    The reverse holds in local passives, where non-focus scrambling out of the verbal complement is impossible:

  9. 9.

    We provide here the sentences we used in our questionnaire study for ease of comparability. As a consequence, we diverge from the standard practice of giving embedded clauses. The material erst gestern wieder in (13) serves the dual purposes of (i) filling the prefield, and (ii) providing for a minimal pair with the topicalization structures to be discussed in Sect. 5. Analogous judgments hold for V-final clauses. In the interest of space, we will not present the active clauses corresponding to the passive examples in this section. For all the tests we are aware of, the active clause allows whatever reading is available in either the local or the long passive. We have also refrained from giving example analogous to (13) for the universal quantifier. Due to the confound identified above, these examples are less informative than ones employing nur ‘only’.

  10. 10.

    Examples (14b) and (15b) are grammatical under a non-NPI reading of auch nur ein einziger Traktor.

  11. 11.

    We should point out that there are some speakers for whom an embedded construal of the adverb is marginally possible in (18c) under certain prosodic conditions (cf. fn. 5). This will be briefly addressed in fn. 32 below.

  12. 12.

    Wurmbrand (2001) suggests instead that in long passives the verbal complement is too small to contain a layer of negation whereas local passives, on the other hand, are structurally large enough to host a negation.

  13. 13.

    All else equal, we would expect four different possible interpretations of wieder (restitutive vs. repetitive crossed with embedded vs. matrix). It is not clear to us what a restitutive reading with vergessen would mean and we will ignore this reading here as it is irrelevant for our purposes. Second, no judgment is provided for the embedded repetitive reading in (23a). This is because the repetitive reading is entailed by the restitutive reading. Every scenario verifying the repetitive reading necessarily also verifies the restitutive one. As a result, the existence of the latter prevents us from diagnosing the former. Nothing hinges on this complication as all that matters for the point at hand is that (23a) has a restitutive reading.

  14. 14.

    Questions arise with respect to the status of the infinitival marker zu that appears in both local and long passives. Wurmbrand (2001, 109–115) argues against the common assumption that either zu or its English counterpart reside in T. Instead, she proposes that zu is part of the projection of the lexical V. Its presence in long and local passives is hence consistent with the structures assumed here. That zu occupies a very low position is supported by the fact that zu might appear between a verbal stem and a prefix that are non-compositionally related. An example is abzurichten ‘to train’ in (11), where the verb abrichten ‘train’ morphologically consists of the stem richten ‘align’ and the prefix ab ‘away’, combined in an idiomatic way. Pullum (1982) and Pollard and Sag (1994) argue for a related position for English to. We will adopt these claims and will thus leave aside considerations about zu for the remainder of this paper.

  15. 15.

    Because the long passive structure in (25) involves recursive VP embedding, it predicts that long passive formation is an iterative process. While judgments decline, it does seem possible to go beyond one level of embedding as the following example from Haider (2010, 313) shows:

  16. 16.

    Function composition is a powerful device. If left unrestricted, it would allow for the formation of syntactic dependencies across island boundaries. A proper assessment of the distribution of function composition in the grammar is beyond the purview of this paper. For the purposes of our account it is sufficient if function composition is limited to elements within a single, though possibly complex, syntactic head.

  17. 17.

    This is not meant to be a general claim about control. Cases like (i.a, b) show that non-agents can be controlled.


    This is not the case in restructuring environments where something like agent identification seems to be at play. There are well developed semantic proposals for control (Dowty 1985) that could in principle be modified for current purposes. However, there is one potential complication that we would like to point out: our proposal is couched in a system where the external argument is not an argument of the ‘control’ predicate. This complicates the statement of the Meaning Postulates that Dowty’s approach uses. We are thankful to Annabel Cormack for making us aware of this issue.

  18. 18.

    Note that there is no independent evidence (word order, intonation) that the quantificational object needs to move into the matrix clause.

  19. 19.

    Notice that we assume here that the indirect object is directly introduced by the lexical verb instead of a designated functional Appl head. The reason for this assumption is entirely presentational. Under the more complex structure, the embedded verb first incorporates into the Appl head and the resulting complex head moves to the matrix verb (as forced by the Head Movement Constraint). Function composition and λ-reconstruction would thus apply iteratively. While entirely feasible, the denotations for such structures quickly become very cumbersome. We have therefore opted for the simpler structure for the sake of exposition.

  20. 20.

    The semantics we assume for only one is an oversimplification. As it stands, VP3 denotes a truth value. A more accurate treatment would have the quantification range over situations (Elbourne 2005) or make use of intensional function application (Heim and Kratzer 1998). We are also abstracting away from considerations of focus. We adopt the simpler semantics in (37) for the purposes of simplicity here.

  21. 21.

    Frey (1993) and Krifka (1998) argue that in German an element α can take scope over another element β only if α c-commands β or β’s trace. If the present account is correct, the scope restriction in the long passive constitutes a counter-example to this generalization because elements within the verbal complement take scope over the matrix predicate without ever c-commanding it. Thanks to Joachim Sabel for pointing this out to us.

  22. 22.

    As a reviewer points out, distinctness is conceptually similar to Hoekstra’s (1984) Unlike Category Condition and Riemsdijk’s (1988, 1998) Unlike Feature Condition. Yet they differ from distinctness in a crucial way: they do not straightforwardly differentiate between complements of different sizes because all of them would count as verbal. Because verb incorporation is size-sensitive in that it applies if the verbal complement is a VP but not if it is a vP, a more fine-grained differentiation within verbal projections is required. Richards’ (2010) notion of distinctness provides such a differentiation.

  23. 23.

    Richards attributes the importance of category labels to the temporal ordering of operations. Adopting the framework of Distributed Morphology (Halle and Marantz 1993, 1994), vocabulary insertion into functional heads is taken to be post-syntactic. Linearization, on the other hand, is taken to apply before vocabulary insertion. As such, linearization does not have access to information contributed by vocabulary insertion. Richards concludes from this that only coarse-grained information like the category label can enter into the computation of distinctness.

  24. 24.

    The reason for this prediction is as follows: Richards argues that linearization takes place before vocabulary insertion into functional heads. Following some work in Distributed Morphology, he also claims that vocabulary insertion into lexical heads invariably applies pre-syntactically, thus necessarily preceding linearization. As a result, linearization has access to a much richer array of information regarding these heads that allows it to distinguish them even if they bear the same category label.

  25. 25.

    Marantz (1996, 1997), for instance, notes that Distributed Morphology is consistent with both pre- and post-syntactic insertion of lexical heads. Thus, the distinction does not follow from the basic assumptions of Distributed Morphology but is an axiom itself.

  26. 26.

    To give just one example, the projecting element is consistently linearized to the left at the super-head level in English but to the right at the sub-head level. That linearization and distinctness seem to operate differently in these two domains clearly indicates that more work needs to be done towards a fully general theory of linearization.

  27. 27.

    An anonymous reviewer wonders about the possibility of having adjacent prepositions or adjectives (e.g., from under the bed, a tasty red apple). It has been argued for on independent grounds that prepositions constitute phases (Abels 2003) and that adjectives do as well (Preminger 2011). The reviewer’s observation can hence be construed as providing converging evidence for these views.

  28. 28.

    While the verbal complement in (44) is a bare VP, it might also contain more functional projections (e.g., Appl), as long as these projections are not phasal.

  29. 29.

    A reviewer wonders whether verb incorporation requires string adjacency or only structural proximity and suggests coordinate structures as a suitable test case. Consider, for example, the examples in (i):


    Due to the various well-known intricacies of coordinations, the predictions of our account depend on the structure of the conjuncts in (i). First, (i) could involve a right node raising structure. In this case, fünfmal should take wide scope within its respective conjunct. Second, (i) could involve bare VP coordination. If verb incorporation is thus enforced but at the same time prevented by the Coordinate Structure Constraint, the entire structure should simply be ineffable. A third option is that coordination itself constitutes a phase (Reich 2007), in which case fünfmal should be able to take narrow scope in both conjuncts. Overall, the one core prediction of the present analysis is that the two conjuncts should pattern symmetrically. An adjacency based characterization, on the other hand, would lead one to expect that wide scope of five times is possible in (i.b) but not (i.a). However, judgments are tricky and additionally clouded by a strong inclination to have fünfmal in (i.a) take scope over the entire coordination. That said, narrow scope seems to be a possibility in both cases. Based on these admittedly unclear judgments there is no evidence for a direct role of string adjacency.

  30. 30.

    A reviewer wonders whether verb-second movement could likewise bleed verb incorporation and thereby obviate its semantic repercussions. Unfortunately, this prediction is impossible to test because passives are periphrastic. The main verb appears as a participle, which never undergoes verb-second movement.

  31. 31.

    Movement of the verbal complement also makes it clear that not all instances of wide scope in the long passive can be reduced to verb incorporation. Consider as an example the sentence in (7a), repeated here as (i):


    (i) only has a reading in which nur blaue Autos ‘only blue cars’ takes scope over vergessen ‘forget’. Because (i) involves VP topicalization we can infer that no verb incorporation takes place. Instead, the matrix scope of nur blaue Autos is the direct result of its raising into the matrix clause and an independent constraint against reconstruction into a remnant (Barss 1986; Lechner 1998; Sauerland 1998). That not all types of scope reversals can be reduced to head movement is unsurprising, of course.

  32. 32.

    As noted in fn. 5, a few speakers marginally allow for a prosodic break to occur between the two verbs in the long passive (we are grateful to Klaus Abels, Gereon Müller and Susi Wurmbrand for pointing this out to us). Interestingly, it seems that in this case the wide scope and construal restrictions noted in Sect. 2 disappear: Only a narrow construal of embedded material is possible:


    While this observation is tentative at this point, the interactions between phrasal and head movement discussed in this section may extend to this further intricacy as well. Given the independent availability of string-vacuous scrambling in German, pointed out to by an anonymous reviewer, it is plausible to assume that the embedded VP in (i) has undergone short scrambling, which is masked in the surface string. This short scrambling step obviates the distinctness violation in the way just discussed and a low construal of the adverb is possible. In addition, because the embedded VP is thus moved away from its base position, it is no longer placed into a prosodic unit with the matrix verb. This captures the codependence between prosodification and construal.

  33. 33.

    Our work thus adds to a line of work by Wechsler (1991), Truckenbrodt (2006), and Lechner (2006) that associates certain instances of head movement with semantic effects. Our proposal differs though in how it does so. In Wechsler (1991) and Truckenbrodt (2006), verb movement itself does not have a semantic effect but the presence/absence of its trigger has a semantic effect. In Lechner’s (2006) treatment, the movement of a modal verb is semantically contentful but the movement does not leave a trace; the modal can simply be completely interpreted in its landing site. Note, however, that we are not claiming that all instances of head movement are semantically contentful. Some instances of head movement might also be motivated by purely morphosyntactic concerns and not distinctness. In other words, not all instances of head movement are amenable to the interpretation procedure developed here.

  34. 34.

    To give just one example, Bech (1955/1957, 66) gives a coherent example with glauben ‘believe’, a verb that cannot occur in the long passive (Reis and Sternefeld 2004, 473).


  1. Abels, Klaus. 2003. Successive cyclicity, anti-locality, and adposition stranding. Ph.D. dissertation, University of Connecticut, Storrs, Connecticut.

  2. Ades, Anthony, and Mark Steedman. 1982. On the order of words. Linguistics and Philosophy 4: 517–558.

    Article  Google Scholar 

  3. Alexiadou, Artemis, and Elena Anagnostopoulou. 2001. The subject in situ generalization, and the role of case in driving computations. Linguistic Inquiry 32: 193–231.

    Article  Google Scholar 

  4. Alexiadou, Artemis, and Elena Anagnostopoulou. 2007. The subject-in-situ generalization revisited. In Interfaces + recursion = language?, eds. Uli Sauerland and Hans-Martin Grätner, 31–60. Berlin: de Gruyter.

    Google Scholar 

  5. Barrie, Michael. 2011a. Dynamic antisymmetry and the syntax of noun incorporation. Dordrecht: Springer.

    Google Scholar 

  6. Barrie, Michael. 2011b. Review of uttering trees by Norvin Richards. Journal of Linguistics 47: 736–741.

    Article  Google Scholar 

  7. Barss, Andrew. 1986. Chains and anaphoric dependence. Ph.D. thesis, MIT Press, Cambridge.

  8. Bayer, Josef, and Jaklin Kornfilt. 1990. Restructuring effects in German. In Parametric variation in Germanic and Romance: Proceedings from a DYANA workshop, eds. Elisabeth Engdahl, Mike Reape, and Martin Mellor, 21–42. Edinburgh: University of Edinburgh, Centre for Cognitive Science.

    Google Scholar 

  9. Bayer, Josef, and Jaklin Kornfilt. 1994. Against scrambling as an instance of move-alpha. In Studies on scrambling: Movement and non-movement approaches to free word-order phenomena, eds. Norbert Corver and Henk van Riemsdijk, 17–60. Berlin: Mouton de Gruyter.

    Google Scholar 

  10. Bech, Gunnar. 1955/1957. Studien über das deutsche Verbum Infinitum. Copenhagen: Munksgaard, 2 vols.

  11. Beck, Sigrid, and Kyle Johnson. 2004. Double objects again. Linguistic Inquiry 35: 97–123.

    Article  Google Scholar 

  12. Bobaljik, Jonathan, and Susanne Wurmbrand. 2005. The domain of agreement. Natural Language and Linguistic Theory 23: 809–865.

    Article  Google Scholar 

  13. Chomsky, Noam. 1993. A minimalist program for syntactic theory. In The view from building 20: Essays in linguistics in honor of Sylvain Bromberger, eds. Ken Hale and Samuel Jay Keyser, 1–52. Cambridge: MIT Press.

    Google Scholar 

  14. Chomsky, Noam. 1995. The minimalist program. Cambridge: MIT Press.

    Google Scholar 

  15. Chomsky, Noam. 2000. Minimalist inquiries: The framework. In Step by step: Essays in syntax in honor of Howard Lasnik, eds. Roger Martin, David Michaels, and Juan Uriagereka, 89–155. Cambridge: MIT Press.

    Google Scholar 

  16. Chomsky, Noam. 2001. Derivation by phase. In Ken Hale: A life in language, ed. Michael Kenstowicz, 1–52. Cambridge: MIT Press.

    Google Scholar 

  17. Chomsky, Noam. 2013. Problems of projection. Lingua 130: 33–49.

    Article  Google Scholar 

  18. Di Sciullo, Anna-Maria, and Edwin Williams. 1987. On the definition of word. Cambridge: MIT Press.

    Google Scholar 

  19. Dowty, David. 1979. Word meaning and Montague grammar. Dordrecht: Kluwer.

    Google Scholar 

  20. Dowty, David. 1985. On recent analyses of the semantics of control. Linguistics and Philosophy 8: 291–331.

    Article  Google Scholar 

  21. Elbourne, Paul D. 2005. Situations and individuals. Cambridge: MIT Press.

    Google Scholar 

  22. Evers, Arnold. 1975. The transformational cycle in Dutch and German, Ph.D. dissertation, University of Utrecht, reproduced 1975 by, Indiana University Linguistics Club, Bloomington.

  23. Fabricius-Hansen, Cathrine. 1983. Wieder ein wieder? Zur Semantik von wieder. In Meaning, use and interpretation of language, eds. Rainer Bäuerle, Christoph Schwarze, and Arnim von Stechow, 26–41. Berlin: de Gruyter.

    Google Scholar 

  24. Fanselow, Gisbert. 1987. Konfigurationalität: Untersuchungen zur Universalgrammatik am Beispiel des Deutschen. Tübingen: Narr.

    Google Scholar 

  25. Fox, Danny. 2000. Economy and semantic interpretation. Cambridge: MIT Press.

    Google Scholar 

  26. Frey, Werner. 1993. Syntaktische Bedingungen für die semantische Interpretation: Über Binding, implizite Argumente und Skopus. Berlin: Akademie Verlag.

    Google Scholar 

  27. Gärtner, Hans-Martin. 2011. Function composition and the linear local modeling of extended neg-scope. In Local modelling of non-local dependencies in syntax, eds. Artemis Alexiadou, Tibor Kiss, and Gereon Müller, 337–352. Berlin: de Gruyter.

    Google Scholar 

  28. Grewendorf, Günther, and Joachim Sabel. 1994. Long scrambling and incorporation. Linguistic Inquiry 25: 263–308.

    Google Scholar 

  29. Haegeman, Liliane, and Henk van Riemsdijk. 1986. Verb projection raising, scope, and the typology of rules affecting verbs. Linguistic Inquiry 17: 417–466.

    Google Scholar 

  30. Haider, Hubert. 1993. Deutsche Syntax – generativ. Tübingen: Narr.

    Google Scholar 

  31. Haider, Hubert. 2003. V-clustering and clause union: Causes and effects. In Verb constructions in German and Dutch, eds. Pieter Seuren and Gerard Kempen, 91–126. Amsterdam: John Benjamins.

    Google Scholar 

  32. Haider, Hubert. 2010. The syntax of German. Cambridge: Cambridge University Press.

    Google Scholar 

  33. Halle, Morris, and Alec Marantz. 1993. Distributed morphology and the pieces of inflection. In The view from building 20: Essays in linguistics in honor of Sylvain Bromberger, eds. Ken Hale and Samuel Jay Keyser, 111–176. Cambridge: MIT Press.

    Google Scholar 

  34. Halle, Morris, and Alec Marantz. 1994. Some key features of distributed morphology. In Papers on phonology and morphology, eds. Andrew Carnie, Heidi Harley, and Tony Bures. Vol. 21 of MIT working papers in linguistics, 275–288. Cambridge: MITWPL.

    Google Scholar 

  35. Heim, Irene, and Angelika Kratzer. 1998. Semantics in generative grammar. Oxford: Blackwell.

    Google Scholar 

  36. Hinrichs, Erhard, and Tsuneko Nakazawa. 1994. Linearizing AUXs in German verbal complexes. In German in head-driven phrase structure grammar, eds. John Nerbonne, Klaus Netter, and Carl Pollard, 11–38. Stanford: CSLI.

    Google Scholar 

  37. Hoekstra, Teun, ed. 1984. Transitivity: Grammatical relations in government-binding theory. Dordrecht: Foris.

    Google Scholar 

  38. Höhle, Tilman. 1978. Lexikalistische Syntax: Die Aktiv-Passiv-Relation und andere Infinitivkonstruktionen im Deutschen. Tübingen: Niemeyer.

    Google Scholar 

  39. Jacobs, Joachim. 1992. Bewegung als Valenzvererbung. Linguistische Berichte 138: 82–122.

    Google Scholar 

  40. Jacobson, Pauline. 1990. Raising as function composition. Linguistics and Philosophy 13: 423–475.

    Article  Google Scholar 

  41. Jacobson, Pauline. 1992. Raising without movement. In Control and grammatical theory, ed. Richard Larson, 149–194. Dordrecht: Kluwer.

    Google Scholar 

  42. Kiss, Tibor. 1995. Infinite Komplementation. Tübingen: Niemeyer.

    Google Scholar 

  43. Kornfilt, Jaklin. 1996. NP-movement and “restructuring”. In Current issues in comparative grammar, ed. Robert Freidin, 121–147. Dordrecht: Springer.

    Google Scholar 

  44. Kratzer, Angelika. 1996. Severing the external argument from its verb. In Phrase structure and the lexicon, eds. Johan Rooryck and Laurie Zaring, 109–137. Dordrecht: Kluwer.

    Google Scholar 

  45. Kratzer, Angelika, and Lisa Selkirk. 2007. Phase theory and prosodic spellout: The case of verbs. The Linguistic Review 24: 93–135.

    Article  Google Scholar 

  46. Krifka, Manfred. 1995. The semantics and pragmatics of polarity items. Linguistic Analysis 25: 209–257.

    Google Scholar 

  47. Krifka, Manfred. 1998. Scope inversion under the rise-fall contour in German. Linguistic Inquiry 29: 75–112.

    Article  Google Scholar 

  48. Lebeaux, David. 2009. Where does binding theory apply? Cambridge: MIT Press.

    Google Scholar 

  49. Lechner, Winfried. 1998. Two kinds of reconstruction. Studia Linguistica 52: 276–310.

    Article  Google Scholar 

  50. Lechner, Winfried. 2006. An interpretive effect of head movement. In Phases of interpretation, ed. Mara Frascarelli, 45–70. Berlin: de Gruyter.

    Google Scholar 

  51. Lidz, Jeffrey, and Alexander Williams. 2002. Reflexivity and resultatives. In West Coast Conference on Formal Linguistics (WCCFL) 21, eds. Line Mikkelsen and Christopher Potts, 250–263. Somerville: Cascadilla Press.

    Google Scholar 

  52. Lidz, Jeffrey, and Alexander Williams. 2005. C-locality and the interaction of reflexives and ditransitives. In North East Linguistic Society (NELS) 34, eds. Keir Moulton and Matthew Wolf, 389–404. Amherst: GLSA.

    Google Scholar 

  53. Lødrup, Helge. 2014. Long passives in Norwegian: Evidence for complex predicates. Nordic Journal of Linguistics 37: 367–391.

    Article  Google Scholar 

  54. Marantz, Alec. 1996. ‘Cat’ as a phrasal idiom: Consequences of late insertion in distributed morphology. Ms., MIT Press, Cambridge.

  55. Marantz, Alec. 1997. No escape from syntax: Don’t try morphological analysis in the privacy of your own lexicon. In UPenn working papers in linguistics, vol. 4.2, ed. Alexis Dimitriadis, 201–225. Philadelphia: University of Pennsylvania. PLC.

    Google Scholar 

  56. Meurers, Walt Detmar. 2000. Raising spirits (and assigning them case). Groninger Arbeiten zur Germanistischen Linguistik 43: 173–226.

    Google Scholar 

  57. Moro, Andrea. 2000. Dynamic antisymmetry. Cambridge: MIT Press.

    Google Scholar 

  58. Müller, Stefan. 1999. Deutsche Syntax – deklarativ. Tübingen: Niemeyer.

    Google Scholar 

  59. Müller, Stefan. 2002. Complex predicates: Verbal complexes, resultative constructions and particle verbs in German. Stanford: CSLI.

    Google Scholar 

  60. Müller, Stefan. 2006. Complex predicates, 2nd edn. In Encyclopedia of language and linguistics, ed. Keith Brown, 697–704. Oxford: Elsevier.

    Google Scholar 

  61. Nissenbaum, Jon. 1998. Movement and derived predicates: Evidence from parasitic gaps. In The interpretive tract, eds. Uli Sauerland and Orin Percus. Vol. 25 of MIT working papers in linguistics, 247–295. Cambridge: MITWPL.

    Google Scholar 

  62. Ott, Dennis. 2012. Local instability: Split topicalization and quantifier float in German. Berlin: de Gruyter.

    Google Scholar 

  63. Ott, Dennis. 2015. Symmetric Merge and local instability: Evidence from split topics. Syntax 18: 157–200.

    Article  Google Scholar 

  64. Pollard, Carl J., and Ivan A. Sag. 1994. Head-driven phrase structure grammar. Chicago: University of Chicago Press.

    Google Scholar 

  65. Preminger, Omer. 2011. Asymmetries between person and number in syntax: A commentary on Baker’s SCOPA. Natural Language and Linguistic Theory 29: 917–937.

    Article  Google Scholar 

  66. Pullum, Geoffrey. 1982. Syncategorematicity and English infinitival to. Glossa 16: 181–215.

    Google Scholar 

  67. Rackowski, Andrea, and Norvin Richards. 2005. Phase edge and extraction. Linguistic Inquiry 36: 565–599.

    Article  Google Scholar 

  68. Reich, Ingo. 2007. From phases to “across the board movement”. In Chicago Linguistic Society (CLS) 43, 217–232. Chicago: CLS.

    Google Scholar 

  69. Reis, Marga, and Wolfgang Sternefeld. 2004. Review article of S. Wurmbrand “Infinitives: restructuring and clause structure”. Linguistics 42: 469–508.

    Google Scholar 

  70. Richards, Norvin. 2010. Uttering trees. Cambrige: MIT Press.

    Google Scholar 

  71. Riemsdijk, Henk van. 1988. The representation of syntactic categories. In Conference on the Basque language, Basque World Congress, Vol. 1, 104–116.

    Google Scholar 

  72. Riemsdijk, Henk van. 1998. Categorial feature magnetism: The endocentricity and distribution of projections. Journal of Comparative Germanic Linguistics 2: 1–48.

    Article  Google Scholar 

  73. Romero, Maribel. 1997. The correlation between scope reconstruction and connectivity effects. In West Coast Conference on Formal Linguistics (WCCFL) 16, eds. Emily Curtis, James Lyle, and Gabriel Webster, 351–366. Stanford: CSLI Publications.

    Google Scholar 

  74. Sabel, Joachim. 1996. Restrukturierung und Lokalität: Universelle Beschränkungen für Wortstellungsvarianten. Berlin: Akademie-Verlag.

    Google Scholar 

  75. Salzmann, Martin. 2011. Resolving the movement paradox in verb projection raising: in favor of base-generation and covert predicate raising. In Empirical issues in syntax and semantics 8, eds. Olivier Bonami and Patricia Cabredo Hofherr, 453–486. Paris: CSSP.

    Google Scholar 

  76. Salzmann, Martin. 2013. New evidence for verb cluster formation of PF and a right-branching VP. In North East Linguistic Society (NELS) 42, eds. Stefan Keine and Shayne Sloggett, 135–148. Amherst: GLSA.

    Google Scholar 

  77. Sauerland, Uli. 1998. On the making and meaning of chains. Ph.D. dissertation, MIT Press, Cambridge.

  78. Schmid, Tanja, Markus Bader, and Josef Bayer. 2005. Coherence: An experimental approach. In Linguistic evidence: empirical, theoretical, and computational perspectives, eds. Stephan Kepser and Marga Reis, 435–456. Berlin: de Gruyter.

    Google Scholar 

  79. Sportiche, Dominique. 2006. Reconstruction, binding, and scope. In The Blackwell companion to syntax, vol. IV, eds. Martin Everaert and Henk van Riemsdijk, 35–93. Oxford: Blackwell.

    Google Scholar 

  80. Stechow, Arnim von. 1992. Kompositionsprinzipien und grammatische Struktur. In Biologische und soziale Grundlagen der Sprache, ed. Peter Suchsland, 175–248. Tübingen: Niemeyer.

    Google Scholar 

  81. Stechow, Arnim von. 1995. Lexical decomposition in syntax. In Lexical knowledge in the organization of language, eds. Urs Egli, Peter Pause, Christoph Schwarze, Arnim von Stechow, and Götz Wienold, 81–118. Berlin: John Benjamins.

    Google Scholar 

  82. Stechow, Arnim von. 1996. The different readings of wieder: A structural account. Journal of Semantics 13: 87–138.

    Article  Google Scholar 

  83. Stechow, Arnim von, and Wolfgang Sternefeld. 1988. Bausteine syntaktischen Wissens: Ein Lehrbuch der generativen Grammatik. Opladen: Westdeutscher Verlag.

    Google Scholar 

  84. Steedman, Mark. 1985. Dependency and coordination in the grammar of Dutch and English. Language 61: 523–568.

    Article  Google Scholar 

  85. Takahashi, Masahiko. 2010. Case, phases, and the nominative/accusative conversion in Japanese. Journal of East Asian Linguistics 19: 319–355.

    Article  Google Scholar 

  86. Takahashi, Masahiko. 2011. Some consequences of case-marking in Japanese. Ph.D. dissertation, University of Connecticut, Storrs.

  87. Takahashi, Masahiko. 2012. On restructuring infinitives in Japanese: Adjunction, clausal architecture, and phases. Lingua 122: 1569–1595.

    Article  Google Scholar 

  88. Travis, Lisa. 1984. Parameters and effects of word order variation. Ph.D. dissertation, MIT, Cambridge.

  89. Truckenbrodt, Hubert. 2006. On the semantic motivation of syntactic verb movement to C in German. Theoretical Linguistics 32: 257–306.

    Google Scholar 

  90. Wechsler, Stephen. 1991. Verb second and illocutionary force. In Views on phrase structure, eds. Katherine Leffel and Denis Bouchard, 177–191. Dordrecht: Kluwer.

    Google Scholar 

  91. Williams, Edwin. 2003. Representation theory. Cambridge: MIT Press.

    Google Scholar 

  92. Wöllstein-Leisten, Angelika. 2001. Die Syntax der dritten Konstruktion: Eine repräsentationelle Analyse zur Monosententialität von ‘zu’-Infinitives im Deutschen. Tübingen: Stauffenberg.

    Google Scholar 

  93. Wurmbrand, Susanne. 2001. Infinitives: Restructuring and clause structure. Berlin: Mouton de Gruyter.

    Google Scholar 

  94. Wurmbrand, Susanne. 2007. How complex are complex predicates? Syntax 10: 243–288.

    Article  Google Scholar 

  95. Wurmbrand, Susanne. 2013. Complex predicate formation via voice incorporation. Ms. University of Connecticut, Storrs.

  96. Wurmbrand, Susi. 2014. Restructuring across the world. In Complex visibles out there: Olomouc Linguistics Colloquium 2014: Language use and linguistic structure, eds. Ludmila Veselovská and Markéta Janebová, 275–294. Olomouc: Palacký University.

    Google Scholar 

Download references


We are greatly indebted to three anonymous NLLT reviewers and our editor Marcel den Dikken for extensive comments and valuable suggestions on earlier versions of this paper. We also benefited from very helpful discussions with Klaus Abels, Alan Bale, Jonathan Bobaljik, Annabel Cormack, Vincent Homer, Norbert Hornstein, Kyle Johnson, Winnie Lechner, Gereon Müller, Ethan Poole, Joachim Sabel, Martin Salzmann, Bernhard Schwarz, Dominique Sportiche, Gary Thoms, Lisa Travis, Michael Wagner and Susi Wurmbrand, all of whom we would like to thank. We are also indebted to the people who have agreed to take our questionnaire. We are also grateful to audiences at UMass Amherst, UCL, McGill, Chicago, Tübingen, GLOW 36, WCCFL 32 and LISSIM 8. All errors are our own.

Author information



Corresponding author

Correspondence to Stefan Keine.

Appendix:  Questionnaire study

Appendix:  Questionnaire study

As we have observed throughout this paper, there exists some amount of variation between speakers with respect to the scope and construal restriction discussed here. For example, we have noted in fn. 5, 11 and 32 that it is possible for some but not all speakers to obviate the wide scope/construal restriction in long passives by placing a prosodic break between the two verbs. Second, speakers differ in the acceptance of long passives with a moved embedded clause, as noted in Sect. 5. To provide a first approximation to the variability in these domains, we report here grammaticality and truth value judgments from 12 linguistically sophisticated German speakers. No attempt was made to systematically control for dialectal affiliation. We thus stress that this questionnaire was of a preliminary nature, and that a more systematic investigation would be required to pinpoint dialectal differences with any precision.

The questionnaire comprised 35 sentences, all of which involved long passives, local passives or their active counterparts. Both grammaticality and truth-value judgments were carried out on a 5-point Likert scale, with ‘5’ indicating ‘grammatical’/‘definitely true’ and ‘1’ representing ‘ungrammatical’/‘definitely false’. For truth value judgments, participants had the additional option of indicating that a sentence is ungrammatical for them and not give a truth judgment. Only one lexicalization per condition was used, except for the establishment of the baseline acceptability of long passives. Here two items were used: one in which nominative case is reflected on the DP itself, and another where it is reflected in verb agreement only (see fn. 1). Because it is well-known that not all speakers of German allow for the long passive, the questionnaire contained two simple examples of the construction without scope-bearing elements, such as (1c) above, in order to allow us to identify speakers who fall into this group. Two of 14 speakers who filled out the questionnaire gave both sentences a rating of ‘2’ or ‘1’ and were excluded from analysis. In what follows, we report the mean score for each construction we tested, together with the sample standard deviation and the absolute range of judgments.

Regular active clauses such as (54) obtained an average rating of 5 points (sd: 0.2, range: 4–5). A corresponding long passive as in (1c) elicited a rating of 4.0 (sd: 1.1, range: 2–5) and the rating of the corresponding local passive as in (1b) was 4.1 (sd: 1.4, range: 1–5).


A.1 Embedded direct objects (Sect. 2)

Bobaljik and Wurmbrand (2005) observe that an embedded direct object has to take matrix scope in the long passive. This claim is corroborated by our study. The scope options of embedded direct objects was assessed using (55). Scenarios analogous to those in fn. 7 were used for disambiguation.


This demonstrates that wide scope of the embedded direct object is obligatory in long passives but impossible in local passives, in line with Bobaljik and Wurmbrand’s (2005) generalization.

A.2 Embedded indirect object (Sect. 2.1.1)

Section 2.1.1 claims that indirect objects are subject to the same scope constraint as direct objects in the long passive. We used the sentences in (56):


The constraints observed for direct objects thus extends to indirect objects. It is worth noting that the variability in the judgments of the long passives increases compared to (55b). This is consistent with the observations in fn. 5, 11 and 32 and the claim there that speakers differ with respect to the availability of string-vacuous scrambling.

A.3 NPIs (Sect. 2.1.2)

The key observation of Sect. 2.1.2 is that embedded NPIs can be licensed by vergessen ‘forget’ only in the local passive, and not in the long passive. Example (57) shows this to be correct for direct objects and (58) demonstrates the same point for indirect objects.


Note: The relatively high rating for the long passive is presumably a consequence of the fact that the non-NPI reading of (58b) (cf. fn. 10) is easily accessed due the possibility of satisfying the using Fritz to satisfy the presupposition of auch locally, in contrast to (57b). This may have artificially increased the rating of (58b). Yet in any case, there is a clear contrast between the local and the long passive.

A.4 Adverb construal (Sect. 2.3)

The central claim of Sect. 2.3 is that adverbs have to be construed with the matrix predicate in long passives. We employed the sentences (59) to assess this claim. These are particularly suitable because one of the two readings is nonsensical, a judgment that is easier to arrive at than the discrimination between two plausible readings.


Note: Again, the fact that the latter rating is relatively high is plausibly due to the fact that the sentence is grammatical, albeit on a non-sensical matrix construal of mit einem Spezialwerkzeug. In either case, the contrast between the two is clear: The sentence is degraded in the long passive relative to the local passive.

A.5 Moved verbal complements (Sect. 5)

As noted in the main text in Sect. 5, while long passive sentences with a moved verbal complement are often judged grammatical in the literature (see (7)), they are somewhat degraded compared to their non-moved counterparts. Thus, the in-situ long passive in (60a) received a higher rating than a structure in which the embedded VP is topicalized (60b). The questionnaire did not only test topicalization of the complement but also scrambling and extraposition, not reported here. While these constructions elicited lower ratings overall, the basic patterns in the judgments reported here for topicalization holds for these other constructions as well.


Against this background, we tested analogous sentences with scopal material inside verbal complement. All 12 participants accepted the in-situ variant in (61a) as grammatical on at least one reading, while only 8 did so for the moved variant (61b). Example (61) gives the average truth value ratings of the speakers that do accept the respective sentences.


While the in-situ variant (61a) is much more acceptable on a wide scope reading of only, the opposite is true for the topicalization structure (61b), where the narrow scope reading receives a substantially higher rating than the narrow scope interpretation. This supports our characterization in the main text that wide scope reading of only is possible only in the in-situ variant, whereas narrow scope is possible only in the moved counterpart.

Finally, we tested embedded adverbs. The in-situ variant (62a) was accepted by 12 participants and the topicalization structure (62b) by 9. Limiting our attention to the participants who do accept these structures, the average acceptability ratings are given in (62):


This contrast is again remarkable in light of the ratings for (60). There, topicalization incurred a lower rating. It is striking, therefore, that topicalization is associated with an increase in acceptability, albeit only a slight one, when the verbal complement contains an adverb that requires low scope. Inspecting the range of judgments is revealing here: While judgments range from 1 to 5 in the absence of movement, they uniformly fall in the upper half the scale if topicalization takes place. This pattern is expected on our characterization in the main text: An embedded construal is possible only if the verbal complement is moved.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Keine, S., Bhatt, R. Interpreting verb clusters. Nat Lang Linguist Theory 34, 1445–1492 (2016).

Download citation


  • Head movement
  • Function composition
  • Restructuring
  • Semantics of movement
  • Distinctness
  • Phases
  • Long passive