Wh-Fronting in a two-probe system


Prior work on wh-movement has distinguished among several types of wh-fronting languages that permit distinct patterns of overt and covert movement, instantiated for example by the Slavic languages, English, and German. This paper extends the cross-linguistic typology of multiple questions by arguing that Hebrew instantiates a new kind of wh-fronting language, unlike any that are discussed in the current literature. It will show that Hebrew distinguishes between two kinds of interrogative phrases: those that are headed by a wh-word (wh-headed phrases: what, who, [DP which X], where, how…) and those that contain a wh-word but are headed by some other element (wh-containing phrases: [NP N of wh], [PP P wh]). We observe the special status of wh-headed phrases when one occurs structurally lower in a question than a wh-containing phrase. In that case, the wh-headed phrase can be targeted by an Agree/Attract operation that ignores the presence of the c-commanding wh-containing phrase. The paper develops an account of the sensitivity of interrogative probing operations to the head of the interrogative phrase within Cable’s (2010) Q-particle theory. It proposes that the Hebrew Q has an EPP feature which can trigger head-movement of wh to Q and that a wh-probe exists alongside the more familiar Q-probe, and shows how these two modifications to the theory can account for the intricate dataset that emerges from the paper. The emerging picture is one in which interrogative probing does not occur wholesale but rather can be sensitive to particular interrogative features on potential goals.

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


  1. 1.

    I use the term interrogative phrase pre-theoretically to refer to the phrase that is observed to pied-pipe in overt wh-movement.

  2. 2.

    Here and throughout, potential interveners will be indicated in bold.

  3. 3.

    Here and throughout this paper I restrict my attention to multiple questions that contain exactly two interrogative phrases. I leave the investigation of more complex questions for future work.

  4. 4.

    Stylistic or triggered inversion has been analyzed as head movement of the verb to T (Borer 1995) or to C (Shlonsky and Doron 1992; Shlonsky 1997).

  5. 5.

    For reasons of brevity and readability, in what follows I only present examples with simplex wh-words. Questions with D-linked phrases show the same pattern of grammaticality as English D-linked phrases and as the questions with simplex wh-words presented here.

  6. 6.

    Note that here it is crucial to assume that the relevant operator for calculation of intervention effects is the semantic negation which occurs above the subject, and not the sentential negative marker: if the negative marker itself were the intervener then the intervention effect in (12c) would be unexpected, since lo is structurally lower than both of the interrogative phrases in the question.

  7. 7.

    Example (i) below shows that 2-order ditransitives pass the condition A binding test under both constituent orders. Example (ii) shows that 1-order ditransitives pass the test only under one constituent order.

    1. (i)
    1. (ii)
  8. 8.

    See Belletti and Shlonsky (1995) for a different view of ditransitives, under which ditransitives with NP and PP complements have the underlying order NP PP. The order PP NP is derived as the result of focalization. This theory predicts that 2-order ditransitives and 1-order ditransitives should pattern alike with regard to intervention effects. In particular, we expect questions with the order PP NP to be sensitive to intervention effects, contrary to fact. To maintain this theory we would have to explain why focalization bleeds intervention effects for verbs with NP PP complements, but not for verbs with PP PP complements. A theory that explains superiority in terms of scrambling (e.g. Wiltschko 1997; Fanselow 1998) faces a similar challenge. I leave this issue open at this time.

  9. 9.

    Recall that I use this phrase to refer to the phrase that would front, should overt wh-movement occur.

  10. 10.

    Unlike ma, the object marker cannot be omitted from mi ‘who’. Thus, the question in (ib) is degraded compared to (ia), whereas both (iia) and (iib) are grammatical questions in Hebrew. This effect, I believe, is due to the special status of mi and not that of ma: et must precede animate objects and proper names, whereas it can be omitted before inanimate objects under certain circumstances. As a result, et must always precede mi but it can be omitted before ma.

    1. (i)
    1. (ii)

    As an anonymous reviewer points out, this pattern follows the general typological pattern of differential argument encoding (Aissen 1999, 2003 and references therein). If et is viewed as a differential object marker, then, it follows that its presence or absence has visible effects on the syntax. Consequently, theories which suggest that differential object marking is purely a morphological phenomenon (e.g. Keine and Müller 2011) cannot be maintained.

  11. 11.

    Parallel examples with full-fledged prepositions replicate the judgments reported here. This data is omitted from the paper for space reasons—but, importantly, it again highlights the fact that the behavior reported in the paper is not a special property of the object marker et but rather of the structure of interrogative phrases in Hebrew.

  12. 12.

    Another reason why this proposal might be unlikely is that, in general, prepositions do not block c-command between a DP and an anaphor or bound-variable pronoun for the purpose of establishing binding relations, as shown in the data in footnote 7. Since Hebrew DPs that are contained inside PPs—including et-marked DPs—can serve as antecedents for anaphors and pronouns, we may infer that these prepositions also do not block c-command relations for the purpose of interrogative feature-checking. I thank Marcel den Dikken for pointing this out to me.

  13. 13.

    Furthermore, this approach predicts that we should not encounter an intervention effect in (ib) since it can be derived from the same superiority-obeying like structure as postulated in examples (24c) and (26b). However, (ib) is judged as ungrammatical.

    1. (i)
  14. 14.

    I thank Norvin Richards and an anonymous reviewer for a discussion of this alternative analysis.

  15. 15.

    Cable assumes existential closure over choice functions at the IP level. As a result, QPs that do not move to the CP periphery are interpreted as wh-indefinites and not as interrogatives.

  16. 16.

    Note that it is also possible to derive a superiority-obeying question from a structure that contains just one Q-particle which is merged with the interrogative phrase that is overtly moved. This option is already present in the analyses proposed by Pesetsky (2000) and Cable (2010). In fact, Cable (following Beck 2006) assumes that this is the only possible analysis of multiple questions in German. What is crucial is that the analysis presented in the text is available in Hebrew and English. Consequently, superiority-obeying questions in Hebrew and English can be derived from a structure in which all interrogative phrases are merged with Q and moved to C by LF. Hence, we can predict that superiority-obeying questions in Hebrew and English, but not German, are not sensitive to intervention effects.

  17. 17.

    This mechanism is equivalent to the use of Hamblin sets, and the semantics Cable assigns to the Force head that interprets these sets is equivalent to the Hamblin (1973) semantics for questions.

  18. 18.

    The need to treat all interveners, including negation, as focus-sensitive elements despite the fact that they do not show clear effects of association with focus is somewhat disturbing. In future work, I hope to develop a better characterization of these intervention effects, which I hope will not affect the analysis provided in this paper.

  19. 19.

    I note exceptions such as Hornstein and Weinberg (1990), who explain the pattern of English superiority effects—that is, the fact that simple questions do not allow superiority-violations but D-linked questions do—by exploiting the fact that in D-linked phrases, the interrogative phrase contains additional material beside the wh-word which. This allows Hornstein and Weinberg to develop a theory that explains superiority effects in D-linked questions by assuming that it is possible for which to move alone, stranding its sister NP in its base-generated position. Such an option is not available to simplex wh-words. While sensitivity to the presence or absence of additional material in the interrogative phrase is critical for this theory, the nature of that content is immaterial to the theory. Below I will argue that the nature of the interrogative phrase—that it, its headedness—is critical for the analysis of Hebrew. I thank an anonymous reviewer for brining this point to my attention.

  20. 20.

    A reviewer asks why intervening P and D heads should block head movement of wh to Q. If nothing blocks this movement, we predict a language that only contains whQPs. Such a language will behave the same as English; that is, it is not distinguishable from a language in which Q has no EPP feature at all. A different pattern would emerge from a language that has at its disposal a Q with an EPP feature and a second one that does not have an EPP feature, since this allows us to observe both QPs and whQPs within the same language. This is one way to interpret the optionality of wh-to-Q movement in Hebrew.

  21. 21.

    Alternatively one may assume that QP is a phase, and furthermore that it is only the complement of the head of the phase that becomes opaque when the phase is shipped to the interfaces. Consequently, wh will be visible for outside probing operations when it is the head of the phrase that Q merges with, following head movement of wh to Q. wh will be invisible if it is contained inside larger structure—that is, inside a wh-containing phrase—since it cannot head-move out of this structure and adjoin to Q. In that case, only the Q-feature is visible to outside probing operations but the wh-feature is not. In this way, we can derive the whQP-QP distinction without assuming co-headedness of interrogative phrases.

  22. 22.

    This proposal thus joins a large body of work that argues that different kinds of A-bar movement can be triggered in more than one way, for example by using different probes or by leaving behind gaps or resumptive pronouns (cf. McCloskey 1990; Manzini 1992; Postal 1998, among many others).

  23. 23.

    See Soare (2007) for a similar proposal in which both a Q-head and a wh-head exist in the left periphery of the clause, and for the possibility that one or both of these heads could have an EPP feature. This proposal is different than the one advanced here in that the two heads occur in different positions in the structure. As long as the two probes can be activated in either order, however, my proposal is consistent with this state of affairs. I note that my proposal predicts additional movement not predicted by Soare’s proposal because of the presence of parasitic Agree.

  24. 24.

    Such an option would predict the movement of single wh-words to C, stranding the interrogative phrases that contain them in their base-position, contrary to fact. The probe’s ability to trigger movement of phrase in addition to Agreement with it could be seen as a matter of cross-linguistic variation. Under this view, for example, Japanese would be treated as a language in which only the wh-Probe can trigger movement which the Q-probe may Agree with wh-phrases but not move them. This would result in movement of the -ka Q-morpheme alone to C, as argued e.g. in Watanabe (1992).

  25. 25.

    Note that this does not violate Freezing: while the wh-word itself cannot move any further following its Agree/move to Q, following Agreement the QP has its own active wh-feature which can be found by the wh-Probe.

  26. 26.

    An anonymous reviewer notes that parasitic Agree allows the MLC to be violated in certain configurations. I would like to argue that this outcome is in fact desirable. It does not require an abandonment of the MLC but rather suggests that optimal links can be obtained in more than one way in different derivations.

  27. 27.

    An anonymous reviewer suggests that the notion of Minimality (Starke 2001; Rizzi 2001, a.o.) can replace the notions of Parasitic Agree and Project both, if both Q and wh are viewed as interrogative features. Under this view, a phrase with more features is a “better” candidate for movement than one with fewer features. Consequently, a wh-headed phrases are better targets than a wh-containing phrases, even if they occur in a structurally lower position in the structure, predicting that they will be preferred by the interrogative probe. I note that this straightforward proposal makes the undesirable prediction that a derivation in which the wh-containing phrase is overtly moved must result in a superiority-violating structure, and should be sensitive to intervention effects. This is not the case, as we have seen above. As a result, the theory will have to be changed to accommodate the apparent optionality in movement operations in derivations with a base-generated whXP ≫ whP structure. I leave the matter of whether or not this optionality can be encoded in this framework open for future research.

  28. 28.

    In all following derivations I will ignore the possibility of extraneous probing operations that result in a failure to Agree and concentrate on operations that have consequences for the derivation.

  29. 29.

    Alternatively, one may imagine that while the Q-probe is necessarily present on C, the wh-probe may optionally occur on C, but if it does then it must be activated before the Q-probe. The resulting derivations are all equivalent to the ones presented in the text above, as far as I can tell. However, Preminger (2010, 2011) argues in detail for the need for a model of grammar in which a probe that has been activated but has failed to find a suitable target does not cause the derivation to crash. Consequently, I adopt a system that is consistent with this architecture of probing rather than assume optionality in the interrogative probes that can occur on C.

  30. 30.

    The derivations of (41b) questions are parallel to those of (41a) questions with the exception that it is also possible to begin probing with the wh-probe, not the Q-probe. The outcome is identical regardless of which probe is activated first: since both phrases in this question are whPs, the highest one that is merged with Q will be found and moved by whichever probe is used in a given derivation.

  31. 31.

    Note that this predicts that speakers for whom et is a genuine case marker and not a preposition will find (51a) grammatical and crucially different from questions with unambiguous full-fledged prepositions. So far I have only been able to find one speaker for whom this appears to hold. All the other speakers I have consulted with (who accept the Sect. 3 pattern) do not perceive a difference between et and other prepositions.

  32. 32.

    At this stage, the reader may wonder if this pattern is preserved in case the interrogative phrases are separated by a finite clause boundary. In general, superiority violations in such a configuration are ungrammatical. One example of this state of affairs is given in (i) below. This pattern is preserved in [XP wh ]≫[XP wh ], [whP]≫[whP] and [whP]≫[XP wh ] configurations. Crucially, this is caused by the fact that the lower phrase crosses a finite clause boundary as it moves over the higher phrase. If the two phrases originate from the lower clause, superiority violating questions become grammatical again.

    1. (i)

    There is just one exception to the ban on superiority-violating: in the configuration [XPwh] ≫ [whP], the superiority-violating question is grammatical, as shown in (ii). Surprisingly, it is sensitive to intervention effects. I leave the investigation of the factors that govern these phenomena to future work, but I note that they, too, support the special status of wh-headed phrases over superiority-violating phrases.

    1. (ii)
  33. 33.

    A third property—pronunciation of head vs. tail copies of movement chains—will predict the behavior of multiple wh-fronting languages like Bulgarian.

  34. 34.

    Note that in languages that only allow one Q-morpheme per question, the possible effects of an EPP feature on Q are impossible to detect. Only one QP (or perhaps whQP) may, and must, occur in the question in such a language, hence it will trigger movement after Agreement by either the Q-probe or the wh-probe.

  35. 35.

    Non-D-linked wh-phrases in English also have as many Q-morphemes but must be interpreted by a C head which requires multiple wh-phrases to appear in its specifiers. Hence, the superiority-violating structure is predicted to be impossible with these phrases. See Pesetsky (2000), Cable (2007, 2010) for a discussion of these structures.

  36. 36.

    Note that following Cable, we must also assume that Hebrew and English differ in the kinds of C heads in their lexicon: English uses a head which can only host one wh-phrase in its specifier, while Hebrew uses a head which can host multiple wh-phrases.

  37. 37.

    I note that the theory above makes one additional prediction: if we can find a two-place predicate that necessarily takes its complements in the order: [XPwh 1]≫[whP2], it should exhibit the same behavior we observed for questions in which the two phrases are separated by a non-finite clause boundary. As far as I can tell, however, all the Hebrew two-place predicates that take a whP as one of their complements are 2-order ditransitives and hence can have two underlying structures, as shown in Sect. 2 above. Consequently, it is impossible to test this prediction.

  38. 38.

    A reviewer calls the validity of these judgments into question. Similar concerns are also raised for the parallel Hebrew examples. I note that for Hebrew, I have found that the judgments in (57) represent a large proportion of the native speakers who I have been in contact with, including some who do not perceive the QP vs. whQP distinction. In fact, these speakers find the judgments about the availability of readings quite clear. However, still other speakers do not agree with these judgments and perceive no difference between superiority-obeying and superiority-violating questions. I have found similar behavior among English speakers who were asked about the English parallel of (57). At this time I cannot offer speculations as to what characterizes the different speakers.

  39. 39.

    A similar explanation for the data provided here may be given along the lines of the analysis in Fanselow (2004). This analysis argues for relativizing the MLC for interpretation: if two competing derivations yield the same LF but only one of them obeys the MLC, it is to be preferred. If the MLC-violating derivation yields an LF that cannot be obtained otherwise, it is to be allowed nonetheless.

  40. 40.

    As a reviewer notes, this is a transderivational constraint that must take into consideration not only the derivation at hand but also other potential derivations that could result in the same meaning. I refer the reader to Fox (2000) for extensive argumentation for this kind of economy constraint.

  41. 41.

    It has been reported that some speakers can avoid intervention effects in single-pair construals of multiple questions (Pesetsky 2000). This reading must therefore be ignored if the relevant judgments are to be collected.

  42. 42.

    That is, the necessary answer had to be of the form in (i) and not of the form in (ii):

    1. (i)
    1. (ii)


  1. Abney, Steven. 1987. The English noun phrase in its sentential aspect. Massachusetts Institute of Technology dissertation.

  2. Adger, David. 2007. A minimalist theory of feature structure. In Features, eds. Anna Kibort and Greville G. Corbett. London: Oxford University Press.

    Google Scholar 

  3. Aissen, Judith. 1999. Markedness and subject choice in Optimality Theory. Natural Language and Linguistic Theory 17: 673–711.

    Article  Google Scholar 

  4. Aissen, Judith. 2003. Differential object marking: iconicity vs. economy. Natural Language and Linguistic Theory 21: 435–483.

    Article  Google Scholar 

  5. Beck, Sigrid. 2006. Intervention effects follow from focus interpretation. Natural Language Semantics 14: 1–56.

    Article  Google Scholar 

  6. Belletti, Adriana, and Ur Shlonsky. 1995. The order of verbal complements: a comparative study. Natural Language and Linguistic Theory 13: 489–526.

    Article  Google Scholar 

  7. Bhatt, Rajesh. 2005. Long distance agreement in Hindi-Urdu. Natural Language and Linguistic Theory 23(4): 757–807.

    Article  Google Scholar 

  8. Borer, Hagit. 1995. The ups and downs of Hebrew verb movement. Natural Language and Linguistic Theory 13: 527–606.

    Article  Google Scholar 

  9. Cable, Seth. 2007. The grammar of Q: Q-particles and the nature of wh-fronting, as revealed by the wh questions of Tlingit. Massachusetts Institute of Technology dissertation.

  10. Cable, Seth. 2010. The grammar of Q: Q-particles, Wh-movement and pied-piping. Oxford: Oxford University Press.

    Google Scholar 

  11. Citko, Barbara. 2008. Missing labels. Lingua 118: 907–944.

    Article  Google Scholar 

  12. Chomsky, Noam. 1973. Conditions on transformations. In A festschrift for Morris Halle, eds. Stephen Anderson and Paul Kiparsky. New York: Holt, Rinehart and Winston.

    Google Scholar 

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

    Google Scholar 

  14. Chomsky, Noam. 1998. Minimalist inquiries: the framework. MIT Occasional Papers in Linguistics 15.

  15. Chomsky, Noam. 1999. Derivation by phase. MIT Occasional Papers in Linguistics 18.

  16. Chomsky, Noam. 2000. Minimalist inquiries: the framework. In Step by step: essays on minimalist syntax in honour of Howard Lasnik, eds. Roger Martin et al. Cambridge: MIT Press.

    Google Scholar 

  17. Danon, Gabi. 2001. Syntactic definiteness in the grammar of Modern Hebrew. Linguistics 39(6): 1071–1116.

    Article  Google Scholar 

  18. Dayal, Veneeta. 2002. Single-pair vs. multiple-pair answers: Wh in-situ and scope. Linguistic Inquiry 33(3): 512–520.

    Article  Google Scholar 

  19. Falk, Yehuda. 1991. Case: abstract and morphological. Linguistics 29: 197–230.

    Article  Google Scholar 

  20. Fanselow, Gisbert. 1998. Minimal Link effects in German (and other languages). Paper presented at the 1998 MLC conference, Potsdam.

  21. Fanselow, Gisbert. 2004. The MLC and derivational economy. In The Minimal Link Condition, eds. Arthur Stepanov, Gisbert Fanselow, and Ralf Vogel, 73–124. Berlin: de Gruyter.

    Google Scholar 

  22. Fanselow, Gisbert, Denisa Lenertovà, and Thomas Weskott. 2008. Studies on the acceptability of object movement to Spec, CP. In The discourse potential of underspecified structures, ed. Anita Steube, 413–437. Berlin: de Gruyter.

    Google Scholar 

  23. Featherston, Sam. 2005. Magnitude estimation and what it can do for your syntax: some wh-constraints in German. Lingua 115: 1525–1550.

    Article  Google Scholar 

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

    Google Scholar 

  25. Fox, Danny. 2011. Towards an alternative perspective on superiority. Ms., handout for 24.952: Advanced Syntax, Spring semester 2011, MIT.

  26. Golan, Yael. 1993. Node crossing economy, superiority and D-linking. Ms., Tel-Aviv University.

  27. Grimshaw, Jane. 2005. Extended projection. In Words and structure, ed. Jane Grimshaw, 1–73. Stanford: CSLI. Originally circulated as a manuscript in 1991.

    Google Scholar 

  28. Hagstrom, Paul. 1998. Decomposing questions. Massachusetts Institute of Technology dissertation.

  29. Hamblin, Charles Leonard. 1973. Questions in Montague English. Foundations of Language 10(1): 41–53.

    Google Scholar 

  30. Hornstein, Norbert, and Amy Weinberg. 1990. The necessity of LF. The Linguistic Review 7(2): 129–167.

    Article  Google Scholar 

  31. Karttunen, Lauri. 1977. Syntax and semantics of questions. Linguistics and Philosophy 1(1): 3–44.

    Article  Google Scholar 

  32. Keine, Stefan, and Gereon Müller. 2011. Non zero/non-zero alternations in differential object marking. In Proceedings of NELS 39, eds. Suzi Lima, Kevin Mullin, and Brian Smith, Vol. 2, 441–454.

    Google Scholar 

  33. Kishimoto, Hideki. 2005. Wh-in-situ and movement in Sinhala questions. Natural Language and Linguistic Theory 23: 1–51.

    Article  Google Scholar 

  34. É. Kiss, Katalin 1986. Against the LF-movement of WH-phrases. Ms, Hungarian Academy of Sciences, Budapest.

  35. Kuno, Susumu, and Jane J. Robinson. 1972. Multiple Wh questions. Linguistic Inquiry 3: 463–487.

    Google Scholar 

  36. Landau, Idan. 1994. Dative shift and extended VP-shells. Tel-Aviv University thesis.

  37. Manzini, Maria-Rita. 1992. Locality. Vol. 19 of Linguistic Inquiry monograph series. Cambridge: MIT Press.

    Google Scholar 

  38. McCloskey, James. 1990. Resumptive pronouns, A′-binding and levels of representation in Irish. In Syntax of the Modern Celtic languages, ed. Randall Hendrick. Vol. 23 of Syntax and semantics, 199–248. San Diego: Academic Press.

    Google Scholar 

  39. Miyagawa, Shigeru. 2001. EPP, scrambling, and wh-in-situ. In Ken Hale: a life in language, ed. Michael Kenstowicz. Cambridge: MIT Press.

    Google Scholar 

  40. Miyagawa, Shigeru. 2010. Why agree? Why move? Unifying agreement-based and discourse configurational languages. Vol. 54 of Linguistic inquiry monograph. Cambridge: MIT Press.

    Google Scholar 

  41. Pesetsky, David. 1987. Wh-in-situ: movement and unselective binding. In The representation of (in)definiteness, eds. Eric Reuland and Alice ter Meulen. Cambridge: MIT Press.

    Google Scholar 

  42. Pesetsky, David. 1998. Principles of sentence pronunciation. In Is the best good enough? eds. Pilar Barbosa, Daniel Fox, Paul Hagstrom, Martha McGinnis, and David Pesetsky. Cambridge: MITWPL/MIT Press.

    Google Scholar 

  43. Pesetsky, David. 2000. Phrasal movement and its kin. Cambridge: MIT Press.

    Google Scholar 

  44. Postal, Paul. 1998. Three investigations of extraction. Cambridge: MIT Press.

    Google Scholar 

  45. Preminger, Omer. 2005. Argument-mapping and extraction. Tel-Aviv University thesis.

  46. Preminger, Omer. 2010. Failure to agree is not a failure: phi-agreement with post-verbal subjects in Hebrew. In Linguistic variation yearbook, eds. Jeroen van Craenenbroeck and Johan Rooryck, Vol. 9, 241–278. Amsterdam: Benjamins.

    Google Scholar 

  47. Preminger, Omer. 2011. Agreement as a fallible operation. Massachusetts Institute of Technology dissertation.

  48. Radford, Andrew. 1993. Head-hunting: on the trail of the nominal Janus. In Heads in grammatical theory, eds. Greville G. Corbett, Norman M. Fraser, and Scott McGlashan, 73–113. Cambridge: Cambridge University Press.

    Google Scholar 

  49. Reinhart, Tanya. 1998. Wh-in-situ in the framework of the Minimalist Program. Natural Language Semantics 6: 29–56.

    Article  Google Scholar 

  50. Richards, Norvin. 1997. What moves where in which language? Massachusetts Institute of Technology dissertation.

  51. Richards, Norvin. 2010. Uttering trees. Vol. 56 of Linguistic inquiry monographs. Cambridge: MIT Press.

    Google Scholar 

  52. Rizzi, Luigi. 1990. Relativized minimality. Cambridge: MIT Press.

    Google Scholar 

  53. Rizzi, Luigi. 1997. The fine structure of the left periphery. In Elements of grammar, ed. Liliane Haegeman. Dordrecht: Kluwer Academic.

    Google Scholar 

  54. Rizzi, Luigi. 2001. Relativized minimality effects. In The handbook of contemporary syntactic theory, eds. Mark Baltin and Chris Collins, 89–110. Oxford: Blackwell.

    Google Scholar 

  55. Rooth, Mats. 1992. A theory of focus interpretation. Natural Language Semantics 1: 75–116.

    Article  Google Scholar 

  56. Shlonsky, Ur. 1997. Clause structure and word order in Hebrew and Arabic: an essay in comparative Semitic syntax. Oxford studies in comparative syntax. New York: Oxford University Press.

    Google Scholar 

  57. Shlonsky, Ur, and Edit Doron. 1992. Verb second in Hebrew. In West Coast Conference on Formal Linguistics (WCCFL), ed. Dawn Bates, Vol. 10, 431–435.

    Google Scholar 

  58. Soare, Gabriela. 2007. A cross-linguistic typology of question formation and the antisymmetry hypothesis. Vol. 5 of Generative grammar in Geneva.

    Google Scholar 

  59. Starke, Michael. 2001. Move dissolves into merge: a theory of locality. University of Geneva dissertation.

  60. Watanabe, Akira. 1992. Subjacency and S-structure movement of wh-in-situ. Journal of East Asian Linguistics 1: 255–291.

    Article  Google Scholar 

  61. Wiltschko, Martina. 1997. D-linking, scrambling and superiority in German. In Groninger Arbeiten zur germanistischen Linguistik, ed. Werner Abraham, Vol. 41. Germanistisch Institut, Rijksuniversiteit Groningen.

    Google Scholar 

  62. Winter, Yoad. 1999. DP structure and flexible semantics. In North East Linguistic Society (NELS), eds. Masako Hirotani, Andries Coetzee, Nancy Hall, and Ji-yung Kim, Vol. 30. Rutgers University.

    Google Scholar 

  63. Zeijlstra, Hedde. 2004. Sentential negation and negative concord. University of Amsterdam dissertation.

Download references


I am grateful to the audiences at WCCFL 30, GLOW 35, the Hebrew University/Tel-Aviv University syntax-semantics reading group, Syntax Square at MIT, and the Fall 2011 workshop at MIT. For many discussions and comments, I would like to thank David Pesetsky, Danny Fox, Norvin Richards, Martin Hackl, Shigeru Miyagawa, Irene Heim, Michael Yoshitaka Erlewine, Omer Preminger, Micha Breakstone, Hagit Borer, and Satoshi Tomioka. Finally, many thanks to three anonymous referees for Natural Language and Linguistic Theory and to Marcel den Dikken for their constructive and helpful comments on the pre-final version of this paper. All remaining errors are my own.

Author information



Corresponding author

Correspondence to Hadas Kotek.

Appendix: On collecting judgments

Appendix: On collecting judgments

In this section I expand on the conditions under which the judgments discussed in this paper were obtained. All of the sentences discussed in the paper contain multiple wh-questions, where two conditions are varied for each basic example: (a) whether it is superiority-obeying or superiority-violating, and (b) whether it contains an intervener or not. Several factors must be controlled in order to ensure that native speakers can access their intuitions about the sentences. Below I discuss each factor in turn, and discuss generalizations regarding the data.

  1. (61)

Acceptability of superiority-violating vs. superiority-obeying questions

As noted by Featherston (2005) and Fanselow et al. (2008), superiority-obeying questions are generally preferred over superiority-violating ones in languages that reportedly do not have superiority effects. That is, the superiority-obeying question (10a) (repeated here as (62a)) is considered by many speakers to be more natural than the superiority-violating questions in (62b–c) (that is, the superiority-violating variants of (62a) with and without stylistic inversion). Therefore, it is important to compare the acceptability of superiority-violating questions to the acceptability of other superiority-violating questions as well as to the corresponding superiority-obeying question, to bring out larger and smaller contrasts among pairs of similar examples.

  1. (62)

For many Hebrew speakers, the status of superiority-violating questions improved when they were introduced within supporting contexts or in embedded structures like “guess what read who” or inside structures that provide immediate context for producing the superiority-violating sentence, for example: “if I knew what read who I would know what paper topics to suggest to each of the class participants.” In some cases the added material, e.g. temporal or locational adjuncts appearing at the end of the question, improved the prosody of the question and helped sharpen the judgments.

Pair-list readings of multiple questions and their presuppositions

Since the examples discussed in the paper contrast the presence and absence of interveners in multiple questions, it was important to ensure that they all had pair-list readings.Footnote 41 As discussed in Sect. 5, Dayal (2002) shows that pair-list readings presuppose (a) Exhaustivity: that every member of the set quantified over by the overtly moved interrogative phrase is paired with a member of the set quantified over by the in-situ interrogative phrase; and (b) Functionhood: that every member of the set quantified over by the overtly moved interrogative phrase is paired with no more than one member of the set quantified over by the in-situ interrogative phrase. The element of the pair to which these presuppositions apply in the case of superiority-obeying questions and superiority-violating questions is marked with a box in (63)–(64) below.

  1. (63)

    Superiority-obeying question: which student cooked which dish?

  1. (64)

    Superiority-violating question: which dish did which student cook?

To accommodate these presuppositions and to allow the questions the best chance of being grammatical, each question discussed in the paper was presented in a supporting context that made the pair-list reading accessible and salient, and which satisfied the presuppositions of the question. Moreover, the questions all used singular wh-phrases; this was done in order to ensure that speakers accessed a pair-list answer to the question instead of a single-pair of pluralities.Footnote 42

Focus association of the intervener

In all of the examples, the association of negation and other interveners with focus was controlled. This ensured a natural reading of the questions, and that all speakers were judging questions with the same focus association and hence with the same truth conditions. In the case of negation, it was also important to ensure that negation was not construed as constituent negation, since constituent negation may occur too low to intervene (see discussion in Sect. 2.1).

Once all these factors were controlled, the following patterns emerged among the Hebrew speakers consulted for this paper:

  1. (65)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Kotek, H. Wh-Fronting in a two-probe system. Nat Lang Linguist Theory 32, 1105–1143 (2014). https://doi.org/10.1007/s11049-014-9238-8

Download citation


  • Wh-movement
  • Q-particles
  • Superiority
  • Probe-goal relations