SpringerLink
Log in

Implicit arguments in English double object constructions

  • Published:
Natural Language & Linguistic Theory Aims and scope Submit manuscript

Abstract

Pesetsky (1995) argued that both objects in the double object construction must be selected arguments of the lexical verb, based on patterns of optionality. A closer examination shows that this is not correct. The second object of the double object construction and both the NP and PP of the PP frame behave like selected arguments of the lexical verb: the lexical verb determines both whether they can be implicit or not, and how they are interpreted when they are (indefinite versus definite). In contrast, particular lexical verbs determine whether the first object of the double object construction can be dropped, but not how it is interpreted. All implicit first objects are interpreted as pragmatically recoverable definites. Implicit first objects also do not license sluicing, unlike all other implicit objects. I propose a purely syntactic account of these patterns, using the ApplP analysis of double object constructions (Marantz 1993; Bruening 2001). In the analysis, arguments of functional heads like Voice and Appl can only be implicit in the presence of a Pass(ive) head, but arguments of lexical verbs are left implicit through adjunction of an operator to the lexical V. I show that implicit arguments are not projected as NPs in the syntax, contra works like Landau (2010). The failure of sluicing with implicit first objects is then analogous to the failure of sluicing with active-passive pairs. I also suggest that the identity condition on ellipsis makes reference to maximal projections, not to heads as in Rudin (2019).

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

Access this article

Log in via an institution

Subscribe and save

Springer+ Basic
EUR 32.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or Ebook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Price includes VAT (Germany)

Instant access to the full article PDF.

Institutional subscriptions

Similar content being viewed by others

Notes

  1. Some speakers may allow a preposition to pop up with some of these verbs in order to license sluicing, even when they do not otherwise allow a preposition (on this phenomenon, see Bruening 2010b and fn. 22). These speakers still do not allow sluicing without the P, though. Note also that sluicing is completely acceptable with an overt indefinite, so there is no problem with extracting the first object in the sluiced clause (not surprising, since sluicing can generally violate constraints on extraction).

  2. Chung et al. (1995:248) observe this effect with serve and Chung (2013:3) notes it with send, but neither of those works notes its full generality. Note also that the order P NP versus NP P (known as “swi**,” Rosen 1976; Culicover 1999; Merchant 2002) is just a matter of personal preference and is irrelevant here. I used the order that sounds most natural to me, but the other order is generally acceptable, too.

  3. The terms “indefinite” and “definite” are simplifications. There are cases where an indefinite implicit argument can apparently refer to something given, for instance in John picked up the glass of beer and drank (Groefsema 1995:145, (23a)). As Groefsema (1995:145) points out, the implicit argument is still indefinite in that it refers to an unspecified amount of beer, not to the whole glass of beer (unlike drank it, which would require drinking the entire glass of beer). I will ignore this complication here. See also Koenig and Mauner (1999).

  4. I have not included the obligatory double object verb bill in these lists. This verb does allow sluicing, but only with how much:

    1. (i)
      1. a.

        They can’t bill you without telling you how much.

      2. b.

        They will bill you, but I don’t know what.

    Bill only seems to take an amount as its second object (They billed me $200/*a pint of blood). The amount does not have to be present in the discourse, but the discourse does have to include that some amount is expected. For this reason I tentatively classify bill as taking a definite implicit second object, even though sluicing with how much is felicitous.

  5. Another way to tell the difference is with negation. With an indefinite implicit object and negation, for instance They weren’t eating, the interpretation is that they were eating nothing. With a definite, for instance They weren’t watching, the interpretation is instead that there is a particular thing that they weren’t watching, not that they weren’t watching anything. The reader can use this test to verify the facts reported here (as an example, They haven’t charged the entrance fee yet means there is a particular person they haven’t charged yet, not that they have charged no one).

  6. Some may find it relatively easy to accommodate with explain and introduce. Nevertheless, I believe these PPs are definite, as negation indicates: Barry didn’t explain the contract does not mean that he explained it to no one, it means there is some particular individual he did not explain it to (possibly the speaker). The other verbs behave similarly.

  7. An alternative that I will not pursue is that She wrote me is actually the PP frame but has a null preposition. For many speakers this construction is quite marked. It also does not passivize without the P: I was written *(to). One could propose that the P can be null just when the PP and its complement are pronounced together adjacent to the verb. If this sort of analysis is correct, then write falls into the same class as pass, and it might be significant that no verb allows an implicit direct object in either frame.

  8. A reviewer notes that when prepositions permit implicit objects, they are generally definite: On Mondays, I’m not in (the office); He has a hat on (his head); He took his clothes off (his body). This is unlike what we have seen with implicit second objects, which can be either definite or indefinite, so it is unlikely that licensing an implicit second object could be attributed to a null P within the small clause.

  9. In some versions of small clause analyses, a verb like give is analyzed as ‘cause to have,’ with the verb have heading the small clause. This analysis might try to relate the licensing of implicit arguments to the verb have. As a reviewer points out, however, have does not license implicit arguments: Do you have *(anything/something)? Where’s the screwdriver—Do you have *(it)? The analysis of double object constructions as ‘cause to have’ therefore does appear to be a viable one.

  10. There is another ApplP analysis of double object constructions, namely the “low applicative” analysis of Pylkkänen (2008). This is a variety of small clause analysis, and I do not adopt it for the reasons enumerated immediately above. See also Larson (2010) on issues with the semantics of the low applicative analysis.

  11. Oehrle (1976) claims that entailments of successful transfer of possession can depend on the choice of external argument (see also Martin and Schäfer 2017). I find these effects weak to non-existent with ditransitive verbs in English, and do not discuss them here (Martin and Schäfer 2017:note 1 cite an unpublished 2013 manuscript by Hans Kamp that also argues that any apparent such effects are cancellable).

  12. A reviewer suggests that Case Theory might account for some of these facts. I believe the idea is that a proposed null NP in (90) might not need, or even cannot receive, Case, and so another NP can move across it. In contrast, in (91), the overt direct object would not receive Case from anything, ruling the sentences out. The proposed null NP is now really a non-entity: it cannot bind, it cannot be modified by secondary predicates, it cannot receive Case,…. I cannot see any advantage to positing such a null NP, since it seems to be equivalent to nothing.

  13. Martí (2011) proposes that implicit indefinite objects are null objects that incorporate into the verb. Since in at least some languages, verbs with incorporated objects are treated as intransitive, this proposal is potentially compatible with the data just discussed, which indicates that verbs with implicit objects behave as intransitive. However, there are many problems for an incorporation account. One is that PPs can be implicit objects, but PPs do not typically incorporate in the world’s languages. Another is that some verbs permit multiple implicit objects (see Appendix B), but again multiple incorporation is not generally attested. For these reasons I reject an incorporation analysis.

  14. Implicit objects do seem to be able to control PRO, at least in some contexts. See Landau (2010) and the references there. The implicit objects under discussion here do not seem to be capable of this; see Rizzi (1986). I do not discuss control of PRO here because it is not clear that control really does require a syntactic relation, contra Landau (2010). There are clearly many instances of control where the controller is not syntactically represented at all. See Williams (1987) and McCourt et al. (2015), among many others. In addition, Pearson (2016) shows that partial control, the basis for Landau’s argument that control must be syntactic, is actually amenable to a semantic account.

  15. A reviewer asks why one of the functional heads ∃ and ι does not block a resultative secondary predicate, while a projected NP seems to (drink (*whiskey) the pub dry). What I suggested in the text was that the pub and whiskey compete for the same position. Apparently, adjoining ∃ or ι does not stop an NP from occupying the object position (sister of V); this just has to be the argument of some other predicate, like a resultative secondary predicate. For lack of space, I will have to leave the syntax of resultative secondary predicates unexplored.

  16. Notice also that verbs that have a recipient thematic role for the subject still have an indefinite in the passive: The package was received; Even the obsolete computer was bought. If something about the verb or its thematic role was what determined the interpretation of an implicit argument, then we might expect that some implicit external arguments of passives would be interpreted as definites. None are.

  17. While multiplying lexical entries appears undesirable, it is probably unavoidable. The verb believe, for instance, can combine with a CP or a raising-to-object IP/TP; it can also combine with just an NP. When it combines with just an NP, it seems to mean something slightly different from when it combines with a clause. Many other lexical items (if not most) are similar. It is then not surprising that ι and ∃ would have multiple (but related) lexical entries.

  18. The covalued option is for cases where an NP in the antecedent seems to have as its correlate an NP with some other form in the elided clause (see Rudin 2019, Sect. 1.3). For instance, it is possible to have apparent violations of Condition C in sluicing:

    1. (i)

      They rejected Sandy1, but she1 doesn’t know why.

    If the elided clause were identical to the antecedent, she would c-command an R-expression with which it is covalued. If the R-expression can instead be a pronoun in the elided clause (they rejected her1), then a Condition C violation can be avoided. Note that requiring identity of maximal projections might make specifying this option unnecessary, if pronouns involve ellipsis of the lexical material of their antecedent (e.g., Elbourne 2001). Then her in the elided clause would be something like [the Sandy], and it would be headed by a token of the same lexical item as its antecedent.

  19. Note that inflectional elements outside of the VoiceP or PassP can often differ; see (Rudin 2019:Sect. 2). I ignore this complication here.

  20. The idea that an Appl head can be passivized is also proposed in Legate (2014:133–140).

  21. I am very skeptical of the claim that the first object tends to be a topic. It is very easy to have it be a focus or new information, instead. The claim seems to be based almost entirely on the tendency for pronouns to occur as the first object, but this seems to be related to the relative weight of the two objects, not to topicality. Note also that Bruening (2010b, 2018b) argues that some apparent instances of the PP frame are actually heavy-shifted variants of the double object construction (see fn. 22); if correct, then this changes the empirical picture regarding which NPs tend to appear as first objects versus PPs. (I am also skeptical that the notion of a topic plays any role whatsoever in English syntax, but that is a discussion for another place.)

  22. See Bruening (2013) on the by-phrase. Also, this proposal may offer some insight into the operation of “R-Dative Shift” proposed by Bruening (2010b, 2018b). Those works propose that certain apparent instances of the PP frame (especially those that occur with otherwise obligatory double object verbs) are actually the double object construction, but with the specifier of Appl projected on the right and the preposition to appearing. If the current analysis is correct that there is a type of passive head that selects Appl, we could view this instance of a to phrase as being analogous to a by phrase in the passive. In the current analysis extended to “R-Dative Shift,” the to phrase would adjoin to ApplP prior to ApplPass combining with it.

  23. One might wonder why Pass seems to differ from ApplPass in this respect, such that ApplPass only combines with certain verbs, while Pass is not selective. Two observations are relevant: First, ApplPass occurs lower in the tree, and the V moves to (and through) it. Pass, as the highest head, and the phase head, might have to be less restricted. Second, Pass may be more selective and idiosyncratic than people think: there are many verbs that cannot passivize (e.g., Postal 1990), and there actually are some verbs that occur either exclusively in the passive, or with different meanings or properties in their active and passive variants (e.g., be born, say). See Bruening (2014b) for discussion and references.

  24. Alternatively, the second object is not entailed at all, and strike can be used as a simple transitive whose sole argument can be implicit; it is hard to tell.

References

  • Aoun, Joseph, and Yen-Hui Audrey Li. 1989. Scope and constituency. Linguistic Inquiry 20: 141–172.

    Google Scholar 

  • Barss, Andrew, and Howard Lasnik. 1986. A note on anaphora and double objects. Linguistic Inquiry 17: 347–354.

    Google Scholar 

  • Beavers, John. 2011. An aspectual analysis of ditransitive verbs of caused possession in English. Journal of Semantics 28: 1–54.

    Google Scholar 

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

    Google Scholar 

  • Bhatt, Rajesh, and Roumyana Pancheva. 2006. Implicit arguments. In The Blackwell companion to syntax, eds. Martin Everaert and Henk van Riemsdijk, Vol. 2, 558–588. Oxford: Blackwell.

    Google Scholar 

  • Bošković, Željko. 2014. Now I’m a phase, now I’m not a phase: On the variability of phases with extraction and ellipsis. Linguistic Inquiry 45: 27–89.

    Google Scholar 

  • Bresnan, Joan. 1978. A realistic transformational grammar. In Linguistic theory and psychological reality, eds. Morris Halle, Joan Bresnan, and George Miller, 1–59. Cambridge: MIT Press.

    Google Scholar 

  • Bresnan, Joan. 1982. Polyadicity. In The mental representation of grammatical relations, ed. Joan Bresnan, 149–172. Cambridge: MIT Press.

    Google Scholar 

  • Bruening, Benjamin. 2001. QR obeys superiority: Frozen scope and ACD. Linguistic Inquiry 32: 233–273.

    Google Scholar 

  • Bruening, Benjamin. 2010a. Ditransitive asymmetries and a theory of idiom formation. Linguistic Inquiry 41: 519–562.

    Google Scholar 

  • Bruening, Benjamin. 2010b. Double object constructions disguised as prepositional datives. Linguistic Inquiry 41: 287–305.

    Google Scholar 

  • Bruening, Benjamin. 2013. By-phrases in passives and nominals. Syntax 16: 1–41.

    Google Scholar 

  • Bruening, Benjamin. 2014a. Precede-and-command revisited. Language 90: 342–388.

    Google Scholar 

  • Bruening, Benjamin. 2014b. Word formation is syntactic: Adjectival passives in English. Natural Language and Linguistic Theory 32: 363–422.

    Google Scholar 

  • Bruening, Benjamin. 2018a. Depictive secondary predicates and small clause approaches to argument structure. Linguistic Inquiry 49: 537–559.

    Google Scholar 

  • Bruening, Benjamin. 2018b. Double object constructions and prepositional dative constructions are distinct: A reply to Ormazabal and Romero 2012. Linguistic Inquiry 49: 123–150.

    Google Scholar 

  • Bruening, Benjamin. 2020. Idioms, collocations, and structure: Syntactic constraints on conventionalized expressions. Natural Language and Linguistic Theory 38: 365–424. https://doi.org/10.1007/s11049-019-09451-0.

    Article  Google Scholar 

  • Carrier, Jill, and Janet H. Randall. 1992. The argument structure and syntactic structure of resultatives. Linguistic Inquiry 23: 173–234.

    Google Scholar 

  • Carston, Robyn. 2002. Thoughts and utterances: The pragmatics of explicit communication. Oxford: Blackwell.

    Google Scholar 

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

    Google Scholar 

  • Chung, Sandra. 2005. What fronts? on the VP-raising account of verb-initial order. In Verb first: On the syntax of verb-initial languages, eds. Andrew Carnie, Heidi Harley, and Sheila Anne Dooley, 9–29. Amsterdam: John Benjamins.

    Google Scholar 

  • Chung, Sandra. 2013. Syntactic identity in sluicing: How much and why. Linguistic Inquiry 44: 1–44.

    Google Scholar 

  • Chung, Sandra, William Ladusaw, and James McCloskey. 1995. Sluicing and logical form. Natural Language Semantics 3: 239–282.

    Google Scholar 

  • Chung, Sandra, William Ladusaw, and James McCloskey. 2011. Sluicing: Between structure and inference. In Representing language: Essays in honor of Judith Aissen, eds. Rodrigo Gútierrez-Bravo, Line Mikkelsen, and Eric Potsdam, 31–50. Santa Cruz: University of California, Santa Cruz, Linguistic Research Center.

    Google Scholar 

  • Culicover, Peter W. 1999. Syntactic nuts: Hard cases, syntactic theory, and language acquisition. Oxford: Oxford University Press.

    Google Scholar 

  • den Dikken, Marcel. 1995. Particles: On the syntax of verb-particle, triadic, and causative constructions. Oxford: Oxford University Press.

    Google Scholar 

  • Dowty, David. 1979. Word meaning and Montague Grammar. Dordrecht: Kluwer.

    Google Scholar 

  • Dowty, David. 1981. Quantification and the lexicon. In The scope of lexical rules, eds. Michael Moortgat, Harry van der Hulst, and Teun Hoekstra, 79–106. Dordrecht: Foris.

    Google Scholar 

  • Elbourne, Paul. 2001. E-type anaphora as NP-deletion. Natural Language Semantics 9: 241–288.

    Google Scholar 

  • Fillmore, Charles J. 1986. Pragmatically controlled zero anaphora. In Annual meeting of the Berkeley Linguistics Society (BLS) 12, 95–107. Berkeley: Berkeley Linguistic Society.

    Google Scholar 

  • Fodor, Jerry, and Janet Dean Fodor. 1981. Functional structure, quantifiers, and meaning postulates. Linguistic Inquiry 11: 759–769.

    Google Scholar 

  • Goldberg, Adele E. 2014. The information structure of ditransitives: Informing scope properties and long-distance dependency constraints. In Perspectives on linguistic structure and context: Studies in honor of Knud Lambrecht, eds. Stacey Katz Bourns and Lindsy L. Myers, 3–17. Amsterdam: John Benjamins.

    Google Scholar 

  • Goldsmith, John. 1980. Meaning and mechanism in grammar. In Harvard studies in syntax and semantics, ed. Susumu Kuno, 423–449. Cambridge: Harvard University.

    Google Scholar 

  • Green, Georgia M. 1974. Semantics and syntactic regularity. Bloomington: Indiana University Press.

    Google Scholar 

  • Groefsema, Marjolein. 1995. Understood arguments: A semantic/pragmatic approach. Lingua 96: 139–161.

    Google Scholar 

  • Gropen, Jess, Steven Pinker, Michelle Hollander, Richard Goldberg, and Ronald Wilson. 1989. The learnability and acquisition of the dative alternation in English. Language 65: 203–257.

    Google Scholar 

  • Harley, Heidi. 1997. If you have, you can give. In West Coast Conference on Formal Linguistics (WCCFL) 15, eds. Brian Agbayani and Sze-Wing Tang, 193–207. Stanford: CSLI Publications.

    Google Scholar 

  • Harley, Heidi. 2002. Possession and the double object construction. Yearbook of Linguistic Variation 2: 29–68.

    Google Scholar 

  • Harley, Heidi. 2008. The bipartite structure of verbs cross-linguistically, or, why Mary can’t exhibit John her paintings. In Conferências do v congresso internacional da associação brasileira de lingüística, eds. Thaïs Cristófaro Silva and Heliana Mello, 45–84. Belo Horizonte: ABRALIN and FALE/UFMG.

    Google Scholar 

  • Harley, Heidi, and Hyun Kyoung Jung. 2015. In support of the PHAVE analysis of the double object construction. Linguistic Inquiry 46: 703–730.

    Google Scholar 

  • Hornstein, Norbert. 1995. Logical form: From GB to Minimalism. Cambridge: Blackwell.

    Google Scholar 

  • Jackendoff, Ray S. 1990. Semantic structures. Cambridge: MIT Press.

    Google Scholar 

  • Janke, Vikki, and Ad Neeleman. 2012. Ascending and descending VPs in English. Linguistic Inquiry 43: 151–190.

    Google Scholar 

  • Johnson, Kyle. 1991. Object positions. Natural Language and Linguistic Theory 9: 577–636.

    Google Scholar 

  • Kayne, Richard. 1984. Unambiguous paths. In Connectedness and binary branching, 129–163. Dordrecht: Foris.

    Google Scholar 

  • Koenig, Jean-Pierre, and Gail Mauner. 1999. A-definites and the discourse status of implicit arguments. Journal of Semantics 16: 207–236.

    Google Scholar 

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

    Google Scholar 

  • Kratzer, Angelika. 2002. The event argument and the semantics of verbs. Draft accessed 5/20/2009 at http://semanticsarchive.net/Archive/GU1NWM4Z/.

  • Kratzer, Angelika. 2005. Building resultatives. In Event arguments: Foundations and applications, eds. Claudia Maienborn and Angelika Wöllstein, 177–212. Berlin: De Gruyter.

    Google Scholar 

  • Krifka, Manfred. 2004. Semantic and pragmatic conditions for the dative alternation. Korean Journal of English Language and Linguistics 4: 1–32.

    Google Scholar 

  • Landau, Idan. 2010. The explicit syntax of implicit arguments. Linguistic Inquiry 41: 357–388.

    Google Scholar 

  • Larson, Richard K. 1988. On the double object construction. Linguistic Inquiry 19: 335–391.

    Google Scholar 

  • Larson, Richard K. 2010. On Pylkkänen’s semantics for low applicatives. Linguistic Inquiry 41: 701–704.

    Google Scholar 

  • Larson, Richard K. 2014. On shell structure. London: Routledge.

    Google Scholar 

  • Larson, Richard K. 2017. On “dative idioms” in English. Linguistic Inquiry 48: 389–426.

    Google Scholar 

  • Legate, Julie Anne. 2014. Voice and v: Lessons from Acehnese. Cambridge: MIT Press.

    Google Scholar 

  • Levin, Lori S. 1982. Sluicing: A lexical interpretation procedure. In The mental representation of grammatical relations, ed. Joan Bresnan, 590–654. Cambridge: MIT Press.

    Google Scholar 

  • Maling, Joan. 2001. Dative: The heterogeneity of the map** among morphological case, grammatical functions, and thematic roles. Lingua 111: 419–464.

    Google Scholar 

  • Marantz, Alec. 1993. Implications of asymmetries in double object constructions. In Theoretical aspects of Bantu grammar, ed. Sam A. Mchombo, 113–150. Stanford: CSLI Publications.

    Google Scholar 

  • Martí, Luisa. 2006. Unarticulated constituents revisited. Linguistics and Philosophy 29: 135–166.

    Google Scholar 

  • Martí, Luisa. 2011. Implicit indefinite objects: Grammar, not pragmatics. Ms., Queen Mary Univeristy London. Available at http://ling.auf.net/lingbuzz/001317.

  • Martin, Fabienne. 2019. Non-culminating accomplishments. Language and Linguistics Compass 13: 12346. https://doi.org/10.1111/lnc3.12346.

    Article  Google Scholar 

  • Martin, Fabienne, and Florian Schäfer. 2017. Sublexical modality in defeasible causative verbs. In Modality across syntactic categories, eds. Ana Arregui, María Luisa Rivero, and Andrés Salanova, 87–108. Oxford: Oxford University Press.

    Google Scholar 

  • Massam, Diane, and Yves Roberge. 1989. Recipe context null objects in English. Linguistic Inquiry 20: 134–139.

    Google Scholar 

  • McCourt, Michael, Jeffrey J. Green, Ellen Lau, and Alexander Williams. 2015. Processing implicit control: Evidence from reading times. Frontiers in Psychology 6: 1629.

    Google Scholar 

  • Merchant, Jason. 2001. The syntax of silence: Sluicing, islands, and the theory of ellipsis. Oxford: Oxford University Press.

    Google Scholar 

  • Merchant, Jason. 2002. Swi** in Germanic. In Studies in comparative germanic syntax, eds. Jan-Wouter Zwart and Werner Abraham, 295–321. Amsterdam: John Benjamins.

    Google Scholar 

  • Merchant, Jason. 2013. Voice and ellipsis. Linguistic Inquiry 44: 77–108.

    Google Scholar 

  • Mittwoch, Anita. 1982. On the difference between eating and eating something: Activities versus accomplishments. Linguistic Inquiry 13: 113–122.

    Google Scholar 

  • Oba, Y. 2002. The double object construction and the extraction of the indirect object. Linguistic Analysis 32: 40–71.

    Google Scholar 

  • Oehrle, Richard. 1976. The grammatical status of the English dative alternation. PhD diss., Massachusetts Institute of Technology. Distributed by MIT Working Papers in Linguistics, Cambridge.

  • Pearson, Hazel. 2016. The semantics of partial control. Natural Language and Linguistic Theory 34: 691–738.

    Google Scholar 

  • Pesetsky, David. 1995. Zero syntax: Experiencers and cascades. Cambridge: MIT Press.

    Google Scholar 

  • Pinker, Steven. 1989. Learnability and cognition. Cambridge: MIT Press.

    Google Scholar 

  • Portner, Paul. 1998. The progressive in modal semantics. Language 74: 760–787.

    Google Scholar 

  • Postal, Paul M. 1977. Antipassive in French. Linguisticae Investigationes 1: 333–374.

    Google Scholar 

  • Postal, Paul M. 1990. Some unexpected English restrictions. In Grammatical relations: A cross-theoretical perspective, eds. Katarzyna Dziwirek, Patrick Farrell, and Errapel Mejias-Bikandi, 365–385. Stanford: CSLI Publications.

    Google Scholar 

  • Pylkkänen, Liina. 2008. Introducing arguments. Cambridge: MIT Press.

    Google Scholar 

  • Ramchand, Gillian Catriona. 2008. Verb meaning and the lexicon: A first-phase syntax. Cambridge: Cambridge University Press.

    Google Scholar 

  • Récanati, François. 2002. Unarticulated constituents. Linguistics and Philosophy 25: 299–345.

    Google Scholar 

  • Rizzi, Luigi. 1986. Null objects in Italian and the theory of pro. Linguistic Inquiry 17: 501–558.

    Google Scholar 

  • Rosen, Carol. 1976. Guess what about? In Proceedings of the annual meeting of the north east linguistic society NELS 6, 205–211.

    Google Scholar 

  • Rudin, Deniz. 2019. Head-based syntactic identity in sluicing. Linguistic Inquiry 50: 253–283.

    Google Scholar 

  • Shopen, Tim. 1973. Ellipsis as grammatical indeterminacy. Foundations of Language 10: 65–77.

    Google Scholar 

  • Thomas, Andrew L. 1979. Ellipsis: The interplay of sentence structure and context. Lingua 47: 43–68.

    Google Scholar 

  • Williams, Alexander. 2015. Arguments in syntax and semantics. Cambridge: Cambridge University Press.

    Google Scholar 

  • Williams, Edwin. 1980. Predication. Linguistic Inquiry 11: 203–238.

    Google Scholar 

  • Williams, Edwin. 1987. Implicit arguments, the binding theory, and control. Natural Language and Linguistic Theory 5: 151–180.

    Google Scholar 

  • Wurmbrand, Susi. 2017. Strip** and topless complements. Linguistic Inquiry 48: 341–366.

    Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Department of Linguistics, University of Delaware, Newark, DE, 19716, USA

    Benjamin Bruening

Authors
  1. Benjamin Bruening
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Benjamin Bruening.

Additional information

Publisher’s Note

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

Appendices

Appendix A: More on the semantics of implicit first objects

In the interest of completion, this appendix discusses the interpretation of implicit first objects in more detail. In the paper, I grouped all implicit first objects together as pragmatically recoverable definites. It appears that they actually fall into two groups. First, there is a group that allow an interpretation of the implicit first object as the addressee. With this interpretation, binomial each is licensed in the direct object, ranging over the implicit addressees:

  1. (141)
    1. a.

      This place will charge $20 each.

    2. b.

      The warden will permit one phone call each.

    3. c.

      The police will allow one phone call each.

    4. d.

      The pope will forgive one sin each.

The verbs that permit this are charge, permit, allow, forgive. In contrast, tip, deny, strike, and envy do not allow this:

  1. (142)
    1. a.

      We had two waiters. *I tipped $5 each.

    2. b.

      The boss will deny one request each (before permitting anything).

  1. (143)
    1. a.

      The black knight will strike one blow each.

    2. b.

      He envies one possession each.

In Sect. 2 we saw that strike and envy require a pragmatically identifiable (familiar) definite:

  1. (144)
    1. a.

      A: Why is this sword dented? B: #The black knight managed to strike a heavy blow.

    2. b.

      A: Why is the white knight bleeding so badly? B: The black knight managed to strike a heavy blow.

  1. (145)
    1. a.

      A: What’s up with Bill? B: #He envies the big house and expensive car.

    2. b.

      A: Why does Bill dislike Monty? B: He envies the big house and expensive car.

In contrast, tip and deny permit a uniquely identifiable but not necessarily familiar definite:

  1. (146)
    1. a.

      You’re supposed to tip at least 20% now.

      ≈ You’re supposed to tip the waiter/your waiter at least 20% now.

    2. b.

      Courts in other countries often deny due process.

      ≈ Courts in other countries often deny their charges due process.

The implicit first object of tip is interpreted as a unique definite, something like ‘the unique individual (singular or plural) who serves the subject in the given context.’ The same is true for deny, where it is plausibly something like ‘the unique individual who the subject has power over in the given context.’ Neither of these needs to be a familiar individual established in the discourse, instead they can be identified from the context and knowledge of the world. (Note that this is not the same as an indefinite; a paraphrase with someone is not appropriate, as was shown above.) In contrast, the missing first object with strike and envy does need to be familiar in the discourse.

What these four verbs have in common is that the missing argument is definite; definiteness encompasses both familiarity and uniqueness (in English). The verbs that permit binomial each, in contrast, permit a more specific interpretation, where the missing argument is specifically the addressee.

Some other verbs that are not obligatorily double object verbs also have the property of permitting an implicit you. Binomial each can range over this implicit you with these verbs as well:

  1. (147)
    1. a.

      They will provide one meal each.

    2. b.

      The teacher will assign one task each.

This means that the functional head that permits this addressee interpretation (ApplPass in the analysis) is not limited to occurring with obligatorily double object verbs. Some verbs will specify that they can have this interpretation when they occur with ApplPass and others will not. Some of the ones that do cannot appear in the PP frame and some can.

I will continue to treat this you interpretation as an implicit, non-syntactic argument of Appl, licensed by ApplPass. One reason to do this is that, while this implicit you licenses binomial each, it does not license other things that require a syntactically present NP. For instance, the null you in an imperative easily licenses anaphors, but the implicit first object you does not:

  1. (148)
    1. a.

      Hug each other!

    2. b.

      She will permit you each other.

    3. c.

      *She will permit each other.

  1. (149)
    1. a.

      Assign yourself to the team!

    2. b.

      I will assign you yourself.

    3. c.

      *I will assign yourself.

This means that it is not the case that implicit first object you is a null pronoun; it is truly an implicit argument (that is, not syntactically present).

Note also that this special addressee interpretation still does not license sluicing, even if we specify the wh-phrase to range over addressees:

  1. (150)
    1. a.

      This place will charge $20 (each), *but it’s not clear which of you.

    2. b.

      The warden will permit one phone call (each), *but it’s not clear which of you.

    3. c.

      The pope will forgive one sin (each), *but it’s not clear which of you.

    4. d.

      They will provide one meal (each), *but it’s not clear which of you.

    5. e.

      The teacher will assign one task (each), *but it’s not clear which of you.

    6. f.

      They will allow one phone call (each), *but it’s not clear which of you.

Again, sluicing seems to be ungrammatical, not pragmatically infelicitous. In this case, full pronunciation is much better, and even quite felicitous:

  1. (151)

    The warden will permit one last conjugal visit, but he didn’t say which of you he will permit one last conjugal visit.

The problem again is the failure of ellipsis licensing. The ApplPassP that permits the addressee interpretation has no correspondent in the elided clause, where a wh-phrase is projected in the specifier of Appl.

It should be noted that the addressee interpretation is not obligatory for the verbs that allow it. These verbs also permit the implicit first object to be an individual other than the addressee. We saw that for instance with bound readings with quantifiers:

  1. (152)
    1. a.

      No traveler should expect tourist destinations to charge (them) very little.

    2. b.

      No man on death row expects the warden to permit (him) one last conjugal visit.

    3. c.

      No teenage Catholic actually believes that God won’t forgive (him/her/them) premarital sex.

  1. (153)
    1. a.

      No passenger on a cruise ship should be shocked when they provide (him/her/them) less than three meals a day.

    2. b.

      No student should be shocked when the teacher assigns (him/her/them) a book to read.

I will continue to propose the same analysis. There is an Appl that does not project its specifier. The head ApplPass takes this ApplP as its complement and checks its unchecked [S:N] feature. The verb moves to Appl and combines with it semantically. Verbs specify in their lexical entries that they can appear in a structure with ApplPass. Some of these verbs specify that the unprojected argument can be a pragmatically identifiable (unique) definite (tip, deny), while others specify that they permit an interpretation where the unprojected argument is the addressee. The pragmatically familiar definite interpretation seems to be the default for ApplPass, so that all verbs allow it. Verbs that do not specify an addressee or unique definite interpretation only permit the familiar definite interpretation (strike, envy).

I noted in Sect. 5.6 that it is possible to passivize the second object across an implicit first object, and took this to indicate that an implicit first object is not syntactically present in any form. Otherwise it would block A-movement across it. One thing to note now is that not all double object verbs permit passivization of the second object across an implicit first object. For instance, envy and tip do not:

  1. (154)
    1. a.

      *The expensive car is envied.

    2. b.

      Twenty dollars was tipped.

In contrast, all of the verbs that permit an addressee interpretation permit passivization with promotion of the direct object when the first object is implicit:

  1. (155)

    implicit addressee

    1. a.

      A fee will be charged.

    2. b.

      A single phone call will be permitted.

    3. c.

      A single phone call will be allowed.

    4. d.

      Sins will be forgiven in the evening.

    5. e.

      Meals will be provided.

    6. f.

      Slee** quarters will be assigned.

(This is further evidence that the implicit addressee object is not syntactically present.)

However, it seems unlikely that the different definite interpretations of the implicit first object are due to a different syntax. First, as was shown above, all of the verbs that allow the implicit argument to be the addressee also allow a definite non-addressee interpretation. In (155), the implicit first object can be the addressee or some other recoverable definite.

Second, envy and tip differ in their semantics. As we saw, envy takes a familiar definite but tip permits a unique definite. Two other verbs are the same but do permit passivization of the first object, in contrast with envy and deny:

  1. (156)

    familiar definite

    1. a.

      A heavy blow was struck.

    2. b.

      *The expensive car is envied.

  1. (157)

    unique definite

    1. a.

      The request was denied.

    2. b.

      *Twenty dollars was tipped.

It therefore appears that the ability to passivize across an implicit first object does not correlate with the semantic interpretation. In principle, the second object can passivize across any implicit first object (addressee, familiar definite, unique definite). For some reason envy and tip do not permit this. All other verbs do, so the only condition on promotion in the passive seems to be that no other NP is crossed by that promotion. Since Appl’s specifier is not projected in the current analysis, there is no other NP in the way in the passive when the first object is implicit, and the lowest object can move to Spec-PassP and on to Spec-TP.

Appendix B: Multiple implicit arguments

We saw in the introduction that there are some verbs that allow both arguments to be implicit. It turns out that the data become quite complicated, and some verbs behave differently when both arguments are implicit, compared to when just one is.

First, strike and pass seem to allow both internal arguments to be implicit at the same time:

  1. (158)
    1. a.

      The knight struck. (definite first object, indefinite second)

    2. b.

      *The knight struck a heavy blow to his enemy.

  1. (159)
    1. a.

      She almost never passes when she gets the ball (direct object definite, PP indefinite)

    2. b.

      She wouldn’t pass *(to) me.

Since strike only occurs in the double object construction and not in the PP frame (158b), it must be a double object construction with two implicit arguments.Footnote 24 Verbs that behave like strike include ask (You can always ask), email (Will he email, do you think?), and tip (I don’t tip). With pass, the direct object can only be implicit in the PP frame and not the double object frame (159b), so it is likely that in this usage we have the PP frame.

Fillmore (1986) notes that the non-alternating verb contribute allows both its NP object and its PP object to be implicit at the same time. The NP object is interpreted as an indefinite, and the PP object is interpreted as a pragmatically recoverable definite:

  1. (160)

    We’re raising money for cancer, but Leo won’t contribute (anything) (to the cause).

The possibility of multiple implicit arguments in both frames follows naturally from the analysis. First, in the double object construction, ∃ or ι can attach to the complex [V Appl], while ApplPass combines with Appl prior to it projecting its argument:

  1. (161)

All selectional features are checked off, so this derivation converges syntactically. In the semantics, ∃ binds the first argument of Appl/V, making it indefinite, while ApplPass binds the second, making it definite.

In the PP frame, the ∃ or ι with the selectional feature [S:V(S:P)] has to attach first, since it binds the second argument of the V. This means it has to merge while the V still has two arguments, before the other operator merges with it. But then the ∃ or ι with the selectional feature [S:V(S:N)] can also attach, saturating the remaining argument (the NP):

  1. (162)

So far all works as one might expect. However, some verbs that allow multiple implicit arguments have unusual properties that will need to be accounted for. Consider first the verb teach. When both arguments are implicit, they both seem to be indefinite (163a):

  1. (163)
    1. a.

      He’s teaching right now.

    2. b.

      He teaches (*to) first-graders.

    3. c.

      He’s teaching a student one-on-one, but I’m not sure what.

    4. d.

      A: He’s teaching French right now. B: Oh? To who?

This should not be the double object frame, then, because we concluded above that an implicit first object is always definite. It would have to be the PP frame. However, if only the direct object is implicit, then the preposition is not allowed (163b); this means that the second object can only be implicit in the double object frame. This seems to be a contradiction. We also cannot say that teach is a simple transitive when only one NP is present, because sluicing is licensed (163c–d). The antecedent clauses in both cases must have implicit arguments. However, note that teach does allow an implicit PP, in (163d). This means that it is plausible that when both arguments are implicit, as in (163a), what we have is the PP frame. The trick is that an implicit direct object is only licensed in the double object frame in (163b) when it is the only implicit argument; but it can be licensed in the PP frame if the PP is also implicit. This will somehow have to be specified in the relevant lexical entries.

There are also two verbs that seem to behave differently when both arguments are implicit. These are sell and pay. Consider pay first. It allows two implicit arguments (164a). The direct object can only be implicit in the double object construction and not the PP frame (164b), so it is likely that (164a) is a double object construction. What is unusual is that, when both objects are implicit, they both seem to be pragmatically recoverable definites. Yet we concluded above that the implicit second object of pay is indefinite, based on sluicing and out-of-the-blue contexts (164c–d):

  1. (164)
    1. a.

      The government is charging her a fine, but she won’t pay. (both objects definite)

    2. b.

      She paid (*to) the bill collector.

    3. c.

      Ron has to pay the loan shark, but I don’t know how much.

    4. d.

      If you want a ride, you have to pay the wrangler.

It is possible that the listener accommodates in (164d), and we already concluded that sluicing does not necessarily diagnose an indefinite interpretation (especially with how much). Pay may then always have a pragmatically recoverable implicit direct object. It is also possible, however, that when it is the only implicit argument, it is indefinite, but when both arguments are implicit, it is definite. If this is true, then the analysis will again have to be complicated.

Sell behaves similarly. It is clear that its implicit goal is indefinite in (165a). If it has an implicit direct object (allowed only in the PP frame), that object is definite (165b). When both arguments are implicit, (165c–d), both of them seem to have to be definite. The goal in particular has to be given in the discourse; if it is not, it is infelicitous (165d):

  1. (165)
    1. a.

      They just sold the house! (no buyer necessary in discourse)

    2. b.

      We need this house, but the owner will never sell to a corporation like ours.

    3. c.

      We need to buy this house, but the owner won’t sell.

    4. d.

      I can’t believe someone built a house on this toxic waste dump! #But the owner won’t sell.

This means that we have a clear case of the interpretation of an implicit direct object changing depending on whether the PP is also implicit or not.

There is also one other oddity about sell, which is that both arguments can be implicit only in the presence of a modal. All of the examples in (165) have a modal. If there is no modal, the two objects cannot both be implicit:

  1. (166)
    1. a.

      A: They need to buy this house soon or the project timeline will be thrown off. B: The owner sold *(it).

    2. b.

      A: They need to buy this house soon or the project timeline will be thrown off. Will the owner sell?

    3. c.

      A: They needed to buy this house by Friday in order for the project to succeed. Did the owner sell *(it)?

(Note that modals do not license implicit arguments with other verbs: I asked him for his car *but he wouldn’t lend.) I will have to leave exploration of this to future work.

Additionally, the introduction noted that there is a large class of verbs that permits either argument to be implicit, but does not allow them both to be implicit at once. These include charge, cost, envy, fine, forgive, guarantee, give, serve, throw:

  1. (167)
    1. a.

      They charge visitors an entrance fee.

    2. b.

      They charge an entrance fee.

    3. c.

      They charge visitors.

    4. d.

      *They charge.

It is unclear how to capture this in the current analysis, other than by making yet more lexical stipulations. Since the facts do seem to be lexically idiosyncratic, this seems unavoidable.

Finally, the question has arisen at various points whether a ditransitive verb when it seems to have an implicit argument could actually be being used as a simple transitive. Bruening (2014b:note 23) notes one piece of evidence indicating that this is a possibility. This is that resultative secondary predicates cannot be predicated of the first object of a double object construction, but with feed, one can be predicated of the first object when the second is implicit:

  1. (168)

    (Bruening 2014b:402, (i))

    1. a.

      *Perhaps feeding the serpents corpses fat again would…

    2. b.

      Perhaps stop** a slaver or two and feeding the serpents fat again would restore their tractability. (Ship of Destiny, Robin Hobb)

This seems to indicate that there is no second object; the sole object is the complement of the lexical verb.

At the same time, an implicit second object can license sluicing:

  1. (169)

    We need to feed these serpents but we don’t know what.

Given that clauses with different argument structures do not license sluicing, the antecedent clause here must have the NP these serpents in Spec-ApplP, and an implicit second object, since the elided clause seems to require two NP positions.

Interestingly, sluicing seems unacceptable (at least to this author) if we combine it with a resultative secondary predicate:

  1. (170)

    We’ll try to feed the serpents fat again *but we don’t know what.

This suggests that both options are available to the verb feed: it can take a goal as the single argument in a monotransitive use, in which case a resultative secondary predicate is acceptable; or it can have the goal as the first object of the double object construction, with an implicit direct object, in which case it licenses sluicing. Once again, I will have to leave full exploration of this issue to future research.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Bruening, B. Implicit arguments in English double object constructions. Nat Lang Linguist Theory 39, 1023–1085 (2021). https://doi.org/10.1007/s11049-020-09498-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11049-020-09498-4

Keywords

Search

Navigation