Clitics are often analyzed as D ele- ments that are head-adjoined to a functional head such as T0 (Uriagereka 1995; see also the references in Yuan 2018: 53). Crucially, subject-indexing morphemes in Semitic look nothing like determiners. First, there are, to my knowledge, no Semitic languages in which definite articles inflect for person or number.9 Determiners are largely invariant when compared to subject-indexing morphology on the verb.
Clitics. In this grammar, clitics are defined as grammatically separate morphemes that are phonologically dependent on another word (Xxxxx & Xxxxxxxxxx 2003). A clitic is attached to its host after phonological rules have been applied, so it usually does not receive primary stress even if it is the final syllable of a phonological word. Unlike affixes, which are more restricted in their choice of host, clitics can attach to words belonging to multiple lexical classes, or to entire clauses. There are eight categories of clitics in Sarikoli, as presented in Table 1.9: Table 1.9 Categories of clitics dicates aspect through form (perfective vs. imperfective form) and placement (attach- ing to the verb vs. preverbal element), in combination with the type of verb stem Introduced in §3.2
Clitics. An Introduction. Cambridge: Cambridge 41 University Press. 42 Sportiche, D. (1996). Clitic Constructions. In X. Xxxxxxx & X. Xxxxxx (Eds.), Phrase 44 Structure and the Lexicon (pp. 213-276). Kluwer, Dordrecht. 45 Suñer, M. (1988). The Role of Agreement in Clitic-doubled Constructions. Natural 46 Language and Linguistic Theory, 6 (3), 391-434. 47 Suñer, M. (1989). Dialectal Variation and Clitic-Doubled Direct Objects. In C. 48 Xxxxxxxx & X. Xxxxxxxxx (Eds.), Studies in Romance Linguistics (pp. 377-395). Selected 50 Papers from the 27th Linguistic Symposium on Romance Languages (LSRL XVII) 51 (Amsterdam Studies in the Theory and History of Linguistic Science), Amsterdam: Xxxx 52 Benjamins.
Clitics. Prosodic Studies. Holland Academic Graphics. In a recent paper Bo¨gel et al. (2009) outlined a new architecture for modeling the interaction between prosody and syntax. Rather than the co-description approach first suggested by Butt and King (1998), Bo¨gel et al. propose a pipeline arrangement of components in which prosodic information is developed in a module that operates independently of the syntax while still allowing for syntactic rules and pref- erences to be conditioned on prosodic boundaries and other features. This is made possible because the terminal string of the syntactic tree (the LFG c-structure) is a sequence of lexical formatives intermixed with elements inserted by the prosodic component. This architecture allows for misalignments between prosodic units and syntactic constituency because syntactic rules can simply ignore prosodic information that would otherwise disrupt syntactic patterns. But it also incorporates a Principle of Prosodic Prefer- ence that causes syntactic structures that do not coincide with prosodic boundaries to be dispreferred. By way of illustration, Bo¨gel et al. show how this new architecture provides for an insightful account of bracketing misalignments between the prosodic units of spoken English and syntactically motivated phrase structures. They also give an account of the bracketing misalignment exhibited by the Urdu clitic ezafe. The clitic ezafe attaches prosodically to a preceding host but it functions syntactically as an element of a following phrase (Bo¨gel et al. 2008). The present paper explores how the pipeline architecture can be applied to an additional class of syntax-prosody misalignments, the second position clitics that appear in many languages. Second posi- tion clitics have presented a challenge to many if not all theoretical frameworks, and there is a substantial literature on the subject (e.g. Xxxxxxx and Xxxxxx 1996, Xxxxxx and Xxxx 2000 and references therein). The crucial aspects of the problem, from an architectural point of view, are demonstrated by the following example from Serbian/Croation/Bosnian (SCB):
Clitics. Values and Uses of the Form se.
Clitics an introduction: Cambridge University Press.
Clitics can attach to material already containing clitics but affixes cannot. (Xxxxxx and Xxxxxx 1983:503f.)
3.1.1.1 Degree of host selection
Clitics show internal structure (they may be marked for gender and number) and they 24 are prosodically weak single words or affixes attached to the verbal host. As such, they cannot 25 be modified, conjoined, or topicalized and cannot appear in isolation (Kayne 1975). They are 26 also an important part of the argument morphology of Spanish (Xxxxxx 1995).3 28 Unlike subject agreement verbal morphemes that are suffixes, clitics can be prefixed or 29 suffixed onto a verb. In preverbal position they are labelled proclitics, (1) while in postverbal 30 position they are labelled enclitics (2). With simple verbs they appear as proclitics (1), but in 31 verbal periphrasis involving a conjugated modal verb and an infinitival form they may appear 33 as enclitics on the infinitival form (2) or as proclitics on the modal verb (3), and the same is 34 the case with auxiliaries and gerund forms (4), (5): 36 (1) La ve-o 38 37 CL3FSG see-1SG 39 ‘I see her.’ 40 41 (2) Xxxxx- o ver-la 43 42 want-1SG see-INF-CL3FSG 44 ‘I want to see her.’ 45 46 (3) Xx xxxxx-o v-er 47 CL3FG want-1SG see-INF 49 48 ‘I want to see her.’ 50 51 (4) La est-á v-iendo 52 CL3FSG be-3SG see- GER
55 3 From a lexical perspective, they encode the argument structure of the verb but can also be related to non- 56 argument projections such as ethical datives (Bonet i Alsina 1995; Franco & Huidobro 2008; Ormazabal & 57 Xxxxxx 2013), and applicatives (Cuervo 2003). 2 ‘(S/he) is seeing her.’ 5 (5) Est-á v-iéndo-la 6 be-3SG see- GER-CL3FSG 7 ‘(S/he) is seeing her.’ 9 This property, known in the literature as clitic climbing (Kayne 1989, 1991; Rooryck 2000; 11 Uriagereka 1995), is not found in Quechua, in which overt object agreement morphemes 12 appear only as suffixes onto the verb, as in the case of the first person object agreement 13 morpheme in: 15 (6) Maqa-wa-n4 17 hit- 1ACC-3SG 18 ‘S/he hit me.’ 20 The third person object marker is not overt when the subject is third person, as shown in: 23 (7) Maqa-Ø-n 24 hit-Ø-3SG 25 ‘S/He hits (him/her/it).’ 28 Unlike in Spanish, object agreement markers cannot be suffixed onto infinitival forms: 30 (8) *Maqa-wa-y muna-n 32 31 hit-1ACC-INF want-3SG 33 ‘(S/he) wants to hit me.’ 34 35 Nor can they be affixed to the main verb with the relevant meaning: 36 37 (9) *Maqa-y muna-wa-n 39 hit-INF want-1ACC-3SG 40 ‘(S/he) wants to hit me.’ 41 In addition to structures that involve only a clitic and in which the clitic receives an anaphoric
