/ instituut voor de Nederlandse taal / Taalportaal
bibliography glossary about
  • Dutch
  • Frisian
  • Saterfrisian
  • Afrikaans
Show all
    Clitics
    quickinfo

    Clitics are the weak forms of function words (such as pronouns, determiners, auxiliaries, particles, conjunctions, prepositions). Whereas the function and meaning of their weak form is identical to their full/ strong form, clitics are not autonomous words and phonologically resemble affixes. This entails that they are phonologically dependent on a host word to which they attach and with which they form a prosodic constituent. Moreover, they cannot carry main emphasis, accent or focus. What makes clitics special is the fact that they can attach to a wider range of word classes, which makes them (morphologically) different from regular affixes. Furthermore, the syntax of clitics can differ from the syntax of the corresponding strong/ full forms (see section about non-isomorphism below). Introductions to clitics and general discussions of the status of clitics can be found in Booij (1995), Wackernagel (1892), Zwicky (1977), Klavans (1982,1985), Spencer and Luís (2012) and Anderson (2005).

    The examples in tables (1)-(4) show which word sorts can form clitics.

    Table 1: Personal pronouns
    Weak form Strong form
    'k /k/or ik /ɪk=/, /ək/ ik /ɪk/ 1SG.SBJ
    me /mə/ mij /mɛi/ 1SG.OBJ
    je /jə/ jij /jɛi/ 2SG.SBJ
    je /jə/ jou /jɑu/ 2SG.OBJ
    ie /i/ hij /hɛi/ 3SG.SBJ.M
    'm /əm/ hem /hɛm/ 3SG.OBJ.M
    ze /zə/ zij /zɛi/ 3SG.SBJ.F
    'r /ər/, d'r /dər/ haar /har/ 3SG.OBJ.F
    't /ət/, /t/ het /hɛt/ 3SG.N
    we /ʋə/ wij /ʋɛi/ 1PL.SBJ
    ons /ɔns/ 1PL.OBJ
    jullie /jʏli/ 2PL.SBJ/OBJ
    ze /zə/ zij /zɛi/ 3PL.SBJ
    ze /zə/ hen /hɛn/and hun /hʏn/ 3PL.OBJ
    Table 2: Possessive pronouns
    Weak form Strong form
    m'n /mən/ mijn /mɛin/ my
    je /jə/ jouw /jɑu/ your
    z'n /zən/ zijn /zɛin/ his
    'r /ər/, d'r /dər/ haar /har/ her
    Table 3: Adverbs
    Weak form Strong form
    er /ər/ /ɛr/ there
    daar /dər/ /dar/ there
    eens /əs/ /ens/ once
    Table 4: Determiners
    Weak form Strong form
    't /ət/ het /hɛt/ DEF.ART.SG.N

    The full forms of the adverbial pronouns can be used in stressed and unstressed positions.

    er is both the allomorph of het which is used in prepositional phrases, always occuring before the preposition and the adverbial pronoun for there. It has the full form /ɛr/ and the weak form /ər/. In contrast to the personal pronouns, the pronouns het and er, cannot be used in stressed positions but instead have to be substituted by their corresponding demonstrative pronoun forms dat /dɑt/ and daar [dar] (with its weak form /dər/ also functioning as a stylistic variant of er).

    Since many of the weak forms listed above cannot be derived from their strong counterparts by simple phonological processes of reduction, they are presumed to be listed in the lexicon as such (indicated above by denoting both forms as phonologically stored). This is supported by the fact that in some idiomatic expressions only the weak form is possible:

    1
    a. Je vergist je.
    * Je vergist jou.
    You make a mistake.
    b. Schaam je!
    *Schaam jou!
    Shame upon you!
    readmore
    [+]Properties and status of clitics

    Most clitics, except ie /i/ and ik /ɪk/, either have schwa as their only vowel or are totally vowelless. They do not form prosodic words of their own since these would require at least one full vowel. Since schwa never bears stress (cf. topics about the mid-central vowel /ə/ and schwa epenthesis) and it is the prosodic word that is the domain of word-stress rules, clitics cannot carry stress either.

    It is debated whether clitics have to be treated as a special prosodic catergory. According to Nespor and Vogel (1986) and Hayes (1989) amongst others, the 'clitic group' occupies its own tier between the phonological phrase and the prosodic word (ω) in the prosodic hierarchy. Booij (1995;1996) argues against treating clitics as a separate group but rather for seeing it as a matter of prosodic integration into an adjacent prosodic word.

    [+]Prosodic integration of clitics

    Another phonological process that applies to a host+clitic-combination after schwa is Prevocalic Schwa Deletion, which states that schwa is deleted if it occurs before another vowel:

    2
    a. ik zette het
    /zɛtə ət/
    [zɛtət]
    put it
    b. pakte ik
    /pɑktə ɪk/
    [pɑktɪk]
    took I

    Another possibility to resolve hiatus is /n/-insertion, as shown in (3a) and (3b) (cf. also the section about the non-isomorphism between syntactic and prosodic structure below).

    3
    a. ik zette 't /ɪk zɛtə ət/ [ɪkzɛtənət] I put it
    b. wilde ie /ʋɪldə i/ [ʋɪldəni] wanted he

    Since the domain of this phonological process is the prosodic word, it does not take place between a clitic and a following word. Instead, like in unstressed prefixes, a glottal stop is inserted:

    4
    je eet (lekker) /jə et/ [jəʔet] [*jənet] you eat (nicely)

    Another process that also applies to clitics in Dutch is Homorganic Glide Insertion.

    5
    a. (Ik) zie 't.
    /si ət/
    [sijət]
    (I) see it
    b. (Ik) doe 't.
    /du ət/
    [duwət]
    (I) do it
    [+]Direction of cliticization: proclitics vs. enclitics

    Clitics attaching to the right edge of a word are called enclitics; clitics attaching to the left edge of a word are called proclitics.

    Figure 1: Procliticization/ adjunction (ω = prosodic word, σ = syllable)
    [click image to enlarge]
    Figure 2: Encliticization/ incorporation (ω = prosodic word, F = foot, σ = syllable)
    [click image to enlarge]
    extra

    For more information concerning this structure and other possibilities of representing encliticization see Booij (1996: 12f.)

    For more information about the representation in figure (1) and the discussion about F-level skipping and thus violating the Strict Layer Hypothesis see Booij (1996).

    Thus, Dutch proclitics are adjoined to the following prosodic word while Dutch enclitics are incorporated into the preceding prosodic word. Prosodically, proclitics are identical to prefixes while enclitics are identical to suffixes, being supported by evidence from morphologization that diachronically proclitics change to prefixes and enclitics to suffixes (Hopper and Traugott 2003). That there is a difference between pro- and encliticization is also shown by the fact that phonological processes apply to a different degree within the two groups. While, as shown above, enclitics undergo processes such as prevocalic schwa deletion, homorganic glide insertion, /n/-insertion etc., these are not obligatory in proclitics. Here prevocalic schwa deletion only occurs in very casual speech

    According to Booij (1995, 1996), in Dutch there is no absolute parameter for the direction of cliticization as proposed by Klavans (1985: 98) but 'a preferred direction of attachment' (ibid) which is encliticization in Dutch. Booij (1996) gives a classification of clitics in Dutch:

    1. words that encliticize,
    2. prosodically deficient words that cliticize in the direction in which a host is available, i.e. they preferably encliticize (especially clitics beginning with schwa), but procliticization is possible,
    3. non-prosodically deficient function words with an optional encliticization.

    Clitics can thus function both as proclitics and enclitics in Dutch, except group 1 comprising ie he, its allomorph die and pronominal der her which require a host on its left.

    6
    a. Komt-ie? /kɔmt i/ [kɔm.ti] Does he come?
    b. *Ie komt He comes

    Function words (group 3) do not have to cliticize at all since they can form prosodic words of their own.

    The preference for encliticization in Dutch is due to the fact that the language provides left-headed feet (aka trochees), which makes the prosodic incorporation of an (unstressed) clitic easier at the right edge of a word. Consequently, since Dutch generally disprefers prosodic words beginning with a schwa like 't [ət] it(Booij 1995), schwa-initial clitics will preferably be encliticized and resyllabified obligatorily ((Jan) ziet 't ((zi)σ (tət)σ)ω (Jan) sees it).

    The determiner de /də/ the shows both directions of cliticization, as the phonological proces of voice assimilation in clitics proves: In (7a) word-internal voicing is shown, while in (7ai) regressive voice assimilation is exhibited.

    7
    a. lees de (krant)
    read the (newspaper)
    lestə
    (lees de)(krant)
    a.' lees de (krant)
    read the (newspaper)
    lezdə
    (lees)(de krant)

    While pronominal clitics (see table (1) above) can exhibit both directions of cliticization, example (8a) shows that consonantal, i.e. vowelless clitics, such as 'k I or 't it, preferably are subject to encliticization, as claimed by amongst others Booij (1995) and Gussenhoven (1985). That consonantal clitics are indeed left-headed, is indicated by their use in combination with words ending in a consonant where the clitic variant with schwa must be used: Ik pak 't [ɪk pɑk ət]/ [*ɪk pɑkt] I fetch it.

    8
    a. (Dat) zal 'k eten
    (that) will I eat
    that I will eat
    zalk.e.ten
    a.' (Jan) zal keten
    John will fool
    zal.ke.ten.

    As the example above shows, these vowelless clitics can create consonant clusters which do not form well-formed onsets, like /kb-/'k ben I am or /tf-/'t valt it falls. All these co-occurence restrictions between a clitic and the preceding phonological material suggests that they indeed form a unit together.

    Sentence-initially, proclitics are used exceptionlessly, while the same holds true for enclitics in sentence-final position.

    9
    a. Er staat een paard in de gang.
    [ər]
    There stands a horse in the corridor.
    b. 't gaat (goed). [ət xat]  / [txat] It goes (well).
    [+]Non-isomorphism of syntactic and prosodic structure

    The prosodic and syntactic structure of clitics is not isomorphic. An example of this fact is found in (10) where the determiner 't the syntactically belongs to the following noun boek book. Prosodically however, since this clitic needs a host, it gets incorporated into the preceding verb kocht bought, together with which it subsequently forms a prosodic word.

    10
    Jan kocht 't boek.
    Syntactic structure:
    [(NP)Jan] [(V)kocht] [(NP)'t boek]
    Prosodic structure:
    [jɑn](ω) [[kɔx](σ)[tət](σ)](ω) [buk](ω)
    Jan bought the book.

    Evidence for incorporation is found in resyllabification where a blocking of other phonological processes such as optional /n/-deletion after schwa in (11) can be observed:

    11
    Zij kochten 't boek.
    [zei](ω) [[kɔx](σ)[tə](σ)[nət](σ)](ω) [buk](ω)
    They bought the book.
    References
    • Anderson, Stephen R2005Aspects of the theory of cliticsNew YorkOxford University Press
    • Booij, Geert1995The phonology of DutchOxfordClarendon Press
    • Booij, Geert1995The phonology of DutchOxfordClarendon Press
    • Booij, Geert1995The phonology of DutchOxfordClarendon Press
    • Booij, Geert1995The phonology of DutchOxfordClarendon Press
    • Booij, Geert1995The phonology of DutchOxfordClarendon Press
    • Booij, Geert1996Cliticization as prosodic integration: the case of DutchThe Linguistic Review13219-242
    • Booij, Geert1996Cliticization as prosodic integration: the case of DutchThe Linguistic Review13219-242
    • Booij, Geert1996Cliticization as prosodic integration: the case of DutchThe Linguistic Review13219-242
    • Booij, Geert1996Cliticization as prosodic integration: the case of DutchThe Linguistic Review13219-242
    • Booij, Geert1996Cliticization as prosodic integration: the case of DutchThe Linguistic Review13219-242
    • Booij, Geert1996Cliticization as prosodic integration: the case of DutchThe Linguistic Review13219-242
    • Gussenhoven, Carlos1985Over de fonologie van Nederlandse cliticaSpektator15180-200
    • Hayes, Bruce1989The prosodic hierarchy in meterKiparsky, Paul & Youmans, Gilbert (eds.)Phonetics and phonology1: Rhythm and meterSan DiegoAcademic Press201-260
    • Hopper, Paul J. & Traugott, Elisabeth C2003GrammaticalizationCambridge U.K.Cambridge University Press
    • Klavans, Judith L1982Some problems in a theory of cliticsUniversity of MichiganThesis
    • Klavans, Judith L1985The independence of syntax and phonology in cliticizationLanguage6195-120
    • Klavans, Judith L1985The independence of syntax and phonology in cliticizationLanguage6195-120
    • Nespor, Marina & Vogel, Irene1986Prosodic phonologyStudies in generative grammar 28DordrechtForis Publ.
    • Spencer, Andrew & Luís, Ana R2012Clitics: an introductionCambridge, UKCambridge University Press
    • Wackernagel, Jacob1892Über ein Gesetz der indogermanischen WortstellungIndogermanische Forschungen1333-436
    • Zwicky, Arnold M1977On CliticsBloomington
    About this article
    Author(s): Kathrin Linke, Anna Kirstein
    Category: Dutch Phonology
    Keywords: clitic
    Version: 1.0, December 2015
    Version history:
    version editor(s) date remarks
    1.0 Kathrin Linke, Anna Kirstein December 2015
    printreport errorcite
    Jump to:
    X