/ instituut voor de Nederlandse taal / Taalportaal
bibliography glossary about
  • Dutch1
  • Frisian
  • Saterfrisian
  • Afrikaans
Show all
    4.2.2.2.Internal structure and distribution of the wat voor construction
    quickinfo

    The wat voor construction is a binominal construction that obligatorily contains the preposition voor'for'. The first noun in the phrase (N1) is always the interrogative pronoun wat'what'. The second noun (N2) can be a singular or plural count noun, a non-count noun, or the existentially quantified personal pronouns iets'something' or iemand'someone'.N2 is mostly optionally preceded by een, although this seems to be a less favored option if N2 is a quantifier. Some examples are given in (227).

    227
    a. [Wat voor (een) boek] lees jij?
    singular count noun
      what  for   a  book  read  you
      'What kind of book do you read?'
    b. [Wat voor (een) boeken] lees jij?
    plural count noun
      what  for   a  books  read  you
      'What kind of books do you read?'
    c. [Wat voor (een) koffie] drink jij?
    non-count noun
      what  for   a  coffee  drink  you
      'What kind of coffee are you drinking?'
    d. [Wat voor (?een) iets/iemand] is dat?
    quantified pronoun
      what  for     a  something/someone  is that
      'What kind of thing/person is that?'

    As pointed out in 4.2.2.1, the wat voor questions in (227) request a further specification of N2. The answer to (227a) could be, e.g., a childrenʼs book or a textbook in linguistics. The following subsections will discuss the syntactic properties of the construction.

    readmore
    [+]  I.  The string wat voor (een) N is a constituent

    The fact that the string wat voor een N occupies the initial position of the clause in the examples in (227) above suggests that we are dealing with a phrase. This conclusion is also supported by the fact illustrated in (228) that wat voor phrases can be coordinated (the constituency test).

    228
    Wat voor een vrouw en wat voor een man heb jij ontmoet?
      what for a woman and what for a man  have  you  met
    'What kind of woman and what kind of man did you meet?'

    The fact that the wat voor phrases in (227) can also be split (the so-called wat voor split) does not contradict this claim, since the split patterns can be and generally are analyzed as involving subextraction of wat, as indicated in (229).

    229
    a. Wati lees jij [ti voor een boek]?
    b. Wati lees jij [ti voor een boeken]?
    c. Wati drink jij [ti voor een koffie]?
    d. Wati is dat [ti voor iets/iemand]?

          Evidence in favor of this analysis comes from the fact that the split is possible only in certain syntactic configurations. For example, if the wat voor phrase is the complement of a preposition, as in (230), the split would be impossible because subextraction from an NP-complement of a preposition is generally excluded. Since much more can be said about the syntactic restrictions on the wat voor split, we will postpone further discussion of this to Section 4.2.2.3.

    230
    a. [PP Op [NP wat voor een bericht]] wacht je?
      for  what for a message  wait you
      'For what kind of message are you waiting?'
    b. * Wati wacht je [PP op [NPtivoor een bericht]]?
    [+]  II.  The semantic head of the construction

    The examples in (227) suggest that it is N2 that satisfies the semantic selection restrictions of the verb; this is also supported by the fact that the noun boek(en)'book(s)' in (227a&b) cannot be replaced by a noun like sigaar, which would violate these selection restrictions: *Wat voor een sigaar/sigaren lees je?'what kind of cigar(s) are you reading?'. It is therefore plausible to assume that N2 is the semantic head of the construction, not the interrogative pronoun wat. This assumption is also supported by the binding data in (231), in which coreference is indicated by italics.

    231
    a. Wie hebben elkaar gebeten?
      who  have  each.other  bitten
      'Who bit each other?'
    b. * Wat hebben/heeft elkaar gebeten?
      what  have/has  each.other  bitten
    c. Wat voor honden hebben elkaar gebeten?
      what for dogs  have  each.other  bitten
      'What kind of dogs bit each other?'
    d. * Wat voor hond heeft elkaar gebeten?
      what for dog  has  each.other  bitten

    The examples in (231a&b) show that the interrogative pronouns wie'who' and wat'what' differ in that the former can act as the antecedent of the reciprocal pronoun elkaar'each other', whereas the latter cannot (a difference which may be related to the fact that wat triggers singular agreement on the finite verb, whereas wie may trigger either singular or plural agreement; see the discussion under III). The acceptability of example (231c) therefore suggests that it is N2 that acts as the antecedent of elkaar; this is confirmed by the unacceptability of example (231d), in which the singular noun hond cannot be the antecedent of elkaar. These facts support the claim that it is N2 that functions as the semantic head of the wat voor phrase.

    [+]  III.  The syntactic head of the construction

    The examples in (232) show that the interrogative pronoun wat'what' differs from wie'who' in that it obligatorily triggers singular agreement on the finite verb.

    232
    a. Wat ligt/*liggen er op de grond?
      what lies/lie  there  on the floor
    b. Wie ligt/liggen er op de grond?
      who lies/lie  there  on the floor

    Consequently, if wat functions as the syntactic head of the construction, we would wrongly expect that a wat voor phrase would trigger singular agreement on the finite verb as well. The data in (233) therefore suggest that N2 is not only the semantic but also the syntactic head of the construction.

    233
    a. Wat voor een man loopt daar?
      what for a man  walks  there
      'What kind of man is walking there?'
    b. Wat voor een mannen lopen/*loopt daar?
      what for a men  walk/walks  there
      'What kind of men are walking there?'
    [+]  IV.  The status of the string wat voor een

    The conclusion that N2 is both the semantic and the syntactic head of the wat voor phrase has given rise to the assumption that the string wat voor een is a complex modifier. Apart from the fact that the interrogative pronoun wat cannot be replaced by any other pronoun, there are two arguments that support this assumption: the element een does not behave like a regular indefinite article, and the element voor lacks the case assigning property of prepositions. A problem for this assumption is, however, that wat can be extracted from the string wat voor een, which would be unexpected in view of the Lexical Integrity Constraint: if we are indeed dealing with a lexicalized form, extraction of wat should be blocked.

    [+]  A.  The article een

    Support for the assumption that the wat voor phrase is a complex modifier comes from the fact that een does not act like a regular indefinite article, which is clear from the fact, illustrated in (234a), that it may precede both singular and plural N2s, whereas indefinite articles preceding a plural noun normally have a null form. As a matter of fact, it may be the case that the null form may also appear in the wat voor construction (alternatively, of course, one may assume that no article is present at all), but the data in (234b) then show that this null form is not restricted to plural noun phrases, as would normally be the case.

    234
    a. Wat voor een hond/honden heb jij?
      what  for  a  dog/dogs  have  you
      'What kind of dog/dogs do you have?'
    b. Wat voor hond/honden heb jij?
      what  for  dog/dogs  have  you
      'What kind of dog/dogs do you have?'

    It is not entirely clear whether een can also precede N2 if the latter is an existential pronoun (which would be normally excluded: *een iets/iemand). Our intuitions are that this is impossible if N2 is the -human pronoun iets'something', but at least marginally possible if it is the +human pronoun iemand'someone'. This intuition seems to be confirmed by a Google search (June 2008): whereas the search on the string [wat voor een iets] resulted in only 3 wat voor constructions, the search on [wat voor een iemand] yielded 17 results. It can further be noted that in most of these cases the wat voor phrase was used as the predicate in copular constructions like wat voor een iets is dat?'what kind of thing is that?' and wat voor een iemand ben jij?'what kind of person are you?'

    235
    a. Wat voor (*?een) iets zou jij willen hebben?
      what  for     a  something  would  you  want  have
      'What kind of thing would you like to have?'
    b. Wat voor (?een) iemand zou jij willen uitnodigen?
      what  for     a  someone  would  you  want  invite
      'What kind of person would you like to invite?'

          Another argument in favor of the idea that een is a spurious article is that it cannot be replaced by any other determiner or any other element that may occur in the left periphery of the noun phrase; replacement of een by, e.g., a definite article or a numeral leads to an ungrammatical result.

    236
    Wat voor *de/*?drie honden heb jij?
      what  for    the/three  dogs  have  you

          It should be noted, however, that there is one apparent counterexample to the claim that N2 cannot be preceded by a numeral, viz., constructions involving an empty N2 licensed by quantitative er, as in (237). Een, which is normally pronounced with a schwa, must be pronounced in this construction like the numeral één'one', /e:n/. However, since één cannot be replace by a numeral like drie, it seems plausible that the occurrence of één in (237) is due to the fact that the empty noun must be preceded by some element carrying stress. Note that examples such as (237a) also occur without er: we found various instances of Wat voor een wil je (hebben)? on the internet.

    237
    a. [Wat voor één/*drie [e]] wil jij er hebben?
      what  for  a/three  want  you  er  have
      'What kind would you like to have?'
    b. Wat wil jij er [voor één/*drie e] hebben?

    Some speakers also allow examples such as (237a) without een being present, as shown in (238a). The split pattern in (238b), on the other hand, is consistently judged unacceptable, which might be related to the fact that the phonetic string in (238b) has a more prominent reading in which er ... voor functions as a pronominal PP: Wat wil jij ervoor hebben?'What do you want to have for it?'. Examples such as (238a) also occur without er: we found various instances of Wat voor wil je (hebben)? on the internet.

    238
    a. % [Wat voor [e]] wil jij er hebben?
      what  for  want  you  er  have
      'What kind would you like to have?'
    b. * Wat wil jij er [voor [e]] hebben?
    [+]  B.  The preposition voor

    The discussion in Subsection A suggests that een is a spurious indefinite article. Similarly, the preposition voor may not be a true preposition, which is suggested by the fact that it does not assign case. Unfortunately, this cannot be shown on the basis of Dutch since this language lacks morphological case, but we can show this on the basis of German. Whereas the German preposition für normally assigns accusative case, it does not assign accusative case to N2 in the was für construction. Instead, the case of N2 depends on the case of the complete was für phrase: if the was für phrase is a subject, N2 has nominative case; if it is a direct object, it has accusative case; and if it is the complement of a preposition like mit'with', it is assigned dative case. This is shown in (239).

    239
    a. Was für ein Mannnom hat das Buch gelesen?
    German
      what for a man  has  the book  read
      'What kind of man read the book?'
    b. Was für einen Mannacc hat sie geheiratet?
    German
      what for a man  has  she  married
      'What kind of man did she marry?'
    c. Mit was für einem Manndat hast du gesprochen?
    German
      with  what for a man  have  you  spoken
      'With what kind of man did you speak?'

    Another fact that might be taken to show that voor is not a true preposition is that the string voor + noun phrase cannot undergo R-pronominalization, which is normally possible with voor-PPs.

    240
    a. Wat voor een boek is dat?
      what for a book  is that
    b. * Wat ervoor is dat?
      what for-it  is that
    [+]  C.  The wat voor split

    The conclusions in A and B that eenis a spurious article and that voor is not a “true” preposition either could be seen as supporting the assumption that wat voor een is a complex modifier that is part of the lexicon as such: the availability of the string wat voor could then be accounted for by assuming that it is a reduced form of wat voor een. Analyses that adopt this assumption do, however, run into problems with the wat voor split. If wat voor (een) is a complex modifier, the examples in (241) would violate the Lexical Integrity Constraint, according to which parts of lexical items cannot undergo syntactic processes: in these examples, wat is extracted from the lexical modifier wat voor (een). Assuming that wat voor (een)is a complex modifier therefore forces us to introduce additional mechanisms to allow the violation of this constraint; see Corver (1990/1991) for a good overview of several proposals from the literature.

    241
    Wat heb jij voor (een) hond/honden?
      what  have  you  for   a  dog/dogs
    'What kind of dog/dogs do you have?'

    As an alternative, it has been proposed that wat must be considered a nominal predicate, that is, the wat voor construction should be analyzed like the N van een N construction in Section 4.2.1. Since arguing for this would lead us into a thicket of theory-internal issues of generative grammar, we cannot go into this matter here; for a discussion of this analysis, see Den Dikken (1995b) and Bennis et al. (1998), who provide more or less similar analyses for the two constructions in question.

    [+]  V.  Modification

    Being an interrogative pronoun, N1, of course, cannot be modified. Premodification of N2, on the other hand, does not seem to be restricted. Some examples of wat voor phrases with an N2 modified by an attributive adjective are given in (242a&b).

    242
    a. Wat loopt daar voor (een) rare man?
      what  walks  there  for   a  strange  man
    a'. Wat voor (een) rare man loopt daar ?
    b. Wat heb je daar voor (een) interessant pakje?
      what  have  you  there  for   an  interesting  parcel
    b'. Wat voor (een) interessant pakje heb je daar?

    Modification by means of a PP or a relative clause is possible as well, as is shown in (243a&b). However, in these cases, there seems to be a preference to split the wat voor phrase, which may be due to focus and to the general tendency to place longer phrases in the right periphery of the clause.

    243
    a. Wat loopt daar voor (een) rare man met een stok?
      what  walks  there  for   a  strange  man with a cane
    a'. ? Wat voor (een) rare man met een stok loopt daar?
    b. Wat heb je daar voor (een) interessant pakje in pakpapier?
      what  have  you  there  for   an  interesting  parcel  in wrapping paper
    b'. ? Wat voor (een) interessant pakje in pakpapier heb je daar?

    Postmodification by means of a relative clause is possible, provided that a split watvoor phrase is used; an example is given in (244a), although it should be noted that the most likely reading of this sentence is one in which the relative clause is interpreted as an apposition. Example (244b) shows that when the watvoor phrase is not split, use of a relative clause leads to an unacceptable result.

    244
    a. Wat is dat voor een man die daar met een stok loopt?
      what  is that  there  for a man  that with a cane walks
    b. *? Wat voor een man die daar met een stok loopt is dat?
    [+]  VI.  Syntactic distribution

    The wat voor construction can be used in all regular NP-positions, that is, both as an argument and as a nominal predicate. In (245), we give examples in which the construction functions as a subject, a direct object, an indirect object, the complement of a preposition, and the predicate in a copular construction.

    245
    a. Wat voor een kind heeft die lolly gestolen?
    subject
      what kind of a child  has  that lollipop  stolen
    b. Wat heb je voor een vaas gekocht?
    direct object
      what  have  you  for a vase  bought
    c. Wat voor een kind heeft hij die lolly gegeven?
    indirect object
      what kind of child  has  he  that lollipop  given
      'To what kind of child did he give a lollipop?'
    d. Op wat voor een bericht ben je aan het wachten?
    complement of P
      for what for a message  are  you aan het wait
      'For what kind of message are you waiting?'
    e. Wat voor een boek is dat?
    nominal predicate
      what for a book  is that
      'What kind of book is that?'
    References:
    • Bennis, Hans, Corver, Norbert & Den Dikken, Marcel1998Predication in nominal phrasesThe Journal of Comparative Germanic Linguistics185-117
    • Corver, Norbert1990The syntax of left branch extractionTilburgUniversity of TilburgThesis
    • Corver, Norbert1991The internal syntax and movement behavior of the Dutch 'wat voor'-constructionLinguistische Berichte133190-228
    • Dikken, Marcel den1995Copulas
    About this article
    Author(s): Hans Broekhuis
    Category: Dutch Syntax
    Keywords: ArticleAttributiveFocusIndefinite ArticlePersonal PronounPrepositionPronounR-pronominalizationReciprocal Personal PronounSpurious Indefinite Article
    Version: 1.0, December 2015
    Version history:
    version editor(s) date remarks
    1.0 Hans Broekhuis December 2015
    report errorprintcite
    Jump to:
    X