This article develops of an analysis of the clitic co-occurrence restrictions found on transitive unaccusative verbs in Choctaw, and how they are (or are not) repaired. It turns out that the repair strategy of Absolutive Promotion, by which a typically-absolutive argument becomes ergative, is sensitive to standard syntactic notions of intervention and locality, implying that it involves a syntactic Agree relation. Regarding the clitic co-occurrence restrictions, I show that they can be captured with the Condition on Clitic Hosts—a condition that syntactic heads can host at most one clitic, adapted from the condition of the same name developed by Arregi and Nevins (2012) for Basque. By detailed comparison with Basque, we see that the Condition on Clitic Hosts and Absolutive Promotion are found in both languages. However, they do not have entirely the same effect: Absolutive Promotion in Choctaw can repair a different set of structures from those it can repair in Basque, and clitics are hosted on a different set of heads in Choctaw from where they are hosted in Basque.
This is a preview of subscription content, log in to check access.
Buy single article
Instant access to the full article PDF.
Price includes VAT for USA
Subscribe to journal
Immediate online access to all issues from 2019. Subscription will auto renew annually.
This is the net price. Taxes to be calculated in checkout.
Unless explicitly stated, all the Choctaw data in this article comes from a number of one-on-one elicitation sessions conducted in 2016–2018 with several native speakers of Choctaw, living on the Mississippi Band of Choctaw Indians (MBCI) reservation near Philadelphia, Mississippi. Choctaw examples are written using the practical orthography from Broadwell (2006). Doubled vowels are long, doubled consonants are geminate, underlined vowels are nasal, the digraph <lh> represents [ɬ] and <’> represents a glottal stop. Pitch accent is marked with a ´ above the vowel. Pitch accent in Choctaw is a complex topic, and I follow Broadwell (2006) in marking it only where it is non-final on a verb or noun root. The addition of suffixes to verbs and nouns has complex effects on the placement of pitch-accent which are not well-understood, and, like Broadwell, I do not mark them here. Furthermore, if an example replicated from a published source lacks pitch accent marking, I do not add it on. I diverge from his notation in only marking glottal stops when they are morphemic (e.g. the jussive morpheme), since their realization is highly variable. Note also that the geminate vs. non-geminate status of vowels in certain lexical items may vary depending on morphophonological context. This is due to a process of iambic lengthening in which odd-numbered short vowels in sequences of short vowels become long, thus neutralizing the vowel length contrast in these positions (Nicklas 1974; Ulrich 1986). Following Broadwell (2006), I represent this lengthening orthographically, though I do not add it into examples from published sources where it is not represented.
The following non-transparent glosses are used for Choctaw. ds = different-subject switch-reference marker, loc = locative, exh = exhortative, irr = irrealis, juss = jussive, lg = l-grade aspectual form, ng = n-grade aspectual form, pc = paucal, poss = possessive, ptcp = particle, ss = same-subject switch-reference marker, tns = default tense. For non-Choctaw examples, I follow the glossing conventions of the author who supplies the example, with the exception of some Basque examples, where I have made some simplifying changes.
Rezac (2008b) uses the term Absolutive Displacement.
Broadwell and Martin (1993) report that erg forms may be omitted in the presence of an overt pronoun (except 1sg-li, which cannot be omitted). However, I was unable to replicate their consultants’ judgments, and for the speakers I consulted, allerg forms were obligatory regardless of the presence or absence of a corresponding overt pronoun. This is one of several instances reported in this article where the judgments of my consultants differ from those published in previous literature, likely indicating the presence of dialectal variation and/or generational change. Nonetheless, the judgments reported by Broadwell and Martin (1993) are fully consistent with the claim that the forms in (4) are clitics.
The 3rd-person dat clitic im- is often analyzed as a default or unmarked form, rather than an exponent of 3rd-person features (Ulrich 1986; Woolford 2008). This is likely the correct analysis, given the clear morphological decomposability of the dat clitics into ‘abs+(i)m-’, alongside the syntactic arguments presented by Ulrich (1986:241–243). However, for ease of exposition I simply gloss it as ‘3.dat’ for this article. This issue is taken up again in fn. 44.
Active alignment goes by many names, including split-S, active-stative, agent-patient and semantic, among others. See Mithun (1991) for discussion of the proliferation of terminology in this area.
Example (9b) contains an argument with a possessive marker, and the reader will notice that it is identical to the dat clitic. In fact, the entire paradigm of DP-internal possessive markers is identical to that of the dat clitics. This fact is discussed in the appendix, and used to support the clitic analysis of the dat forms over a potential alternative analysis in which they are agreement affixes.
In saying this, Broadwell contradicts Nicklas (1974), Dixon (1979) and Davies (1986), who state that Choctaw is a ‘fluid-S’ language, where the choice of clitic/agreement must affect the volitionality of the subject. Note that it is still possible that the choice of clitic could reflect an underlying difference in structure, as is the case with banna ‘want’ in Sect. 3.2. However, by allowing a raising-to-ergative derivation, we allow for the possibility that some erg/abs alternations have no relevance to thematic interpretation.
It is worth stating that in the absence of clear diagnostics for unergativity vs. unaccusativity that do not rely on the choice of clitic, the pairing of certain erg-assigning verbs with raising-to-ergative derivations will necessarily remain somewhat stipulative. The strongest kind of argument one can make in favor of a structural ergative case is the existence of a raising verb that assigns ergative to its subject, as in the case of the Basque auxiliary behar ‘must’ examined in Rezac et al. (2014). However, in Choctaw when subjects are clitic-doubled on auxiliaries, the clitic maintains the case it was assigned by the embedded predicate (Broadwell 2006:203), or it gets assigned (structural) dative, as in (24b).
Linker (1987) notes that clauses marked with -naare compatible with the modal future marker -aachi̲. It is possible, then, that while participial clauses lack the ModP projection, clauses marked with -cha or -na have it.
Since 3rd-person erg and abs arguments are not clitic-doubled, it’s necessary to use 1st/2nd-person focused pronouns to juxtapose the case and clitic systems within a single clause.
If evidence were to emerge that Choctaw does have ‘crossing’ derivations, in which a lower argument passes over a higher one, some parts of the analysis would have to be changed, though I set aside this possibility for now.
For these verbs, the abs experiencer argument is always obligatory, and the dat T/SM argument is always optional. From a typological perspective, Gerdts and Kiyosawa (2005) term this type of construction a ‘relational applicative,’ whose characteristic properties are (a) an applicative morpheme is added to an intransitive psych verb, (b) the experiencer argument functions as the subject, and (c) the applied argument—which I am calling the T/SM—functions as the syntactic object. They find that cross-linguistically it is rare but attested, being found in the Austronesian, Eskimo-Aleut, Muskogean (which includes Choctaw) and Salishan language families.
See Broadwell (2006) for evidence that Choctaw is a configurational language.
We cannot say that the examples in (26) are ungrammatical because the i̲-sa- clitic cluster is ruled out, since this is one of the clitic clusters that is permitted—see Sect. 3.1.
The reduction of banna to -nna before an experiencer-doubling abs clitic is a feature of Mississippi Choctaw but not Oklahoma Choctaw.
Davies (1986:70) states that different-subject marking is in fact possible when the experiencer argument of banna is coreferential with the subject of the adjoined or matrix clause. My consultants reported a different judgment, shown in (29), and this is also the one reported by Broadwell (1997). This is another likely point of dialectal variation.
For those Choctaw speakers who allow the forms in (32), there is variation as to whether the /s/ is inserted before am-. It is also notable that while Broadwell (2006) does not discuss any specific banned combinations of clitics, the only examples he provides of the dat>abs type are 1sg.dat>2sg.abs (linearized as 2sg.abs-1sg.datchi-am-).
The possessor-raising construction in (31a) patterns differently from other dat>abs transitive unaccusatives with respect to switch-reference, as discussed by Broadwell (1990, 2006). It turns out that the possessee (i.e. the lower argument), is in fact a better controller of switch-reference than the possessor. I do not have an account of this. However, the possessor-raising cases are less relevant for present purposes than the other dat>abs constructions. This is because in order to induce a clitic cluster, the abs argument would need to be 1st/2nd-person. Given that 1st/2nd-person arguments are not possessable, possessor-raising constructions are a priori prevented from inducing clitic clusters.
Camacho (2010) presents an analysis where switch-reference in Choctaw is controlled not by subjects but by arguments with nominative case. In double-nominative sentences like (31), then, either argument should be capable of controlling it.
A note is required on the fact that a dat clitic is retained on the embedded verb in (35a), while no abs clitic is found on the embedded verb in (25b). Ulrich (1986) provides the example in (i), in which a dat-subject verb in the complement of banna is marked with the 3rd-person or ‘default’ dat clitic i̲-, rather than the 2nd-person dat clitic chi̲-. My consultants found this option odd, and preferred to mark dat subjects on the embedded verb as well as the matrix verb, hence the example in (35a).
I thank an anonymous reviewer for bringing these facts to my attention.
A note on ditransitives, which could in theory also cause two internal-argument clitics to co-occur, is in order. Unfortunately I do not have clear and consistent data for clitic co-occurrence restrictions on ditransitives like I have for transitive unaccusatives. It’s clear that the restrictions in (40)–(44) hold for ditransitives too, so sentences like (i) are ruled out.
However, it is unclear whether the ‘1sg exception’, as in (45), holds in ditransitives. Speakers have varying judgments on the sentence in (ii) (note that pit ‘towards’ in (ii) is not an adposition but a directional particle, on which see Broadwell 1998, 2006). While some speakers will, on occasion, accept these structures as grammatical (though not readily), all speakers prefer to circumlocute them. Note also that this is another of the points where speakers’ judgments differ from those in Davies (1986), who offers the sentence in (iii). This topic requires further investigation.
Heath (1977:205) and Ulrich (1986:254) note that sentences like (48b) are possible. Heath characterizes it as something like a genuine ‘repair’ strategy (though he does not use that term), while Ulrich claims that it is a generally available alternative to sentences like (48a)—a characterization that does not apply to the Choctaw spoken by my consultants.
Davies (1986:5,131) provides examples like (i). Here, the usually-dat subject is realized by an erg clitic, and the usually-abs object is realized by a dat clitic. This is similar to how we might expect Absolutive Promotion operating on an dat>abs verb to look, with the higher (dat) argument being promoted, leaving behind the im- component of the dat clitic, which subsequently runs together with the 2sg abs clitic chi-.
However, the Mississippi Choctaw speakers I consulted did not accept this form or others like it as grammatical. I am therefore unable tell whether, for the speakers who allow it, (i) is found as the output of a repair operation, or as a generally-available alternative case frame for the dat>abs verbs.
There are some optionally transitive verbs that seem to freely alternate between taking erg and abs experiencers. For instance, nokoowa ‘be angry (at)’ works this way for all speakers, and some speakers treat noktalha ‘be jealous (of)’ as a member of this class too (e.g. (6c)). For these verbs, it would not be correct to say that when the experiencer is ergative, they have undergone Absolutive Promotion (which is the name for a repair operation). Instead, the option of raising to Spec-VoiceP is simply a generally available alternative structure, just as it is for habishko ‘sneeze’ in (13).
Dat>abs transitive unaccusatives are not subject to spurious repair, since they cannot be repaired by Absolutive Promotion at all—see (50).
One interpretation of this judgment is that the subject argument in (56) is base-generated in the external argument position, where it receives thematic ergative case. Alternatively, it could raise-to-ergative from a different internal argument position. Crucially, the form in (56) cannot be the result of Absolutive Promotion, since Absolutive Promotion does not alter the interpretation of the verb.
There are a number of possible factors that could underlie the rejection of (57). One is that the cluster is treated no differently from any other 2sg.dat>1sg.abs cluster, and is ruled out by the same principle. Another possibility is that the speakers I consulted do not use the verb noktakali ‘choke (on)’ as an abs>abs transitive, which is a necessary prerequisite for accepting (57). For my consultants, the only abs>abs verb in the language is banna ‘want’, and it is simply incompatible with object possessor raising:
This property essentially makes it impossible to test for the acceptability of clusters like that in (57), as they can never arise. Note that in the dialect reported by Davies (1986), however, speakers do permit object possessor raising with banna.
Ulrich also reports data from ‘Mississippi Choctaw of Oklahoma,’ a variety of Choctaw spoken on the Chickasaw nation in south central Oklahoma.
Absolutive Promotion is only found in some Western Basque dialects, including Berriatua (Aramaio 2001), Ondarru (Arregi 2004), Northern High Navarrese and Guispuscoan (Rezac 2008b), and Gernika, Mendata and Mundaka (Arregi and Nevins 2012). For simplicity, when I refer to ‘Basque’ in this article, I refer only to those dialects featuring Absolutive Promotion.
The caveat that the c-command relations hold within the VoiceP is important here, as it can be shown that in dat>abs verbs like (65a), the lower (absolutive) argument behaves as a subject with respect to control (San Martin 1999; Rezac 2008b). This could indicate that it crosses over the dative argument and occupies a subject position such as Spec-TP.
The glosses in Basque examples are simplified from their sources: details of tense and agreement have been stripped out, and all tense/agreement morphemes are glossed ‘(t.)agr’. The morpheme glossed ‘l’ is what Arregi and Nevins (2012) refer to as the ‘L-morpheme,’ a morpheme drafted in to occupy the leftmost position within the auxiliary under particular morphological conditions.
The optionality of ergative case-marking is not restricted to this structure in particular, and is a more general property of Basque (Arregi and Nevins 2012:72).
While Basque abs>dat verbs cannot be repaired by Absolutive Promotion, an alternative structure is available in which the goal receives allative case and does not cliticize to the auxiliary (Arregi and Nevins 2012:77).
In Tyler (2017), I proposed an alternative analysis of the contrast between Basque and Choctaw in the reparability of dat>abs verbs. The claim is that in Basque, dative is inherent, and so is invisible to Agree probes, following analyses such as McGinnis (1998a), McFadden (2004), Woolford (2006), Alexiadou et al. (2014). By contrast in Choctaw, dative is structural and defectively intervenes. However, no supporting evidence is offered for this difference—indeed, Rezac (2008a) shows that Basque dialects differ in the transparency vs. opacity of their dative arguments with respect to Agree probes and it is not clear that the Absolutive Promotion dialects are also the ones with opaque datives—so I set the proposal aside.
I follow Arregi and Nevins in leaving Spec-TP, the canonical subject position, unfilled, and take no position on whether the abs/erg argument undergoes subsequent movement to it.
In order for Arregi and Nevins’s explanation to hold, the Extension condition (Chomsky 1995) and the No Tampering condition (Chomsky 2008) must be abandoned or appropriately weakened to allow for a number of operations, including the formation of Agree relations and case-assignment, to take place countercyclically. It may make sense to restrict these apparent violations strictly to ‘repair’ contexts, but I do not pursue this idea here—see Rezac (2011) for discussion of the countercyclic character of repair.
The reader may be wondering how the claim that Choctaw clitics are D0s squares with the decomposability of the dat clitics into ‘abs+(i)m-’, discussed in fn. 4. I do not present an analysis here, but I suggest that dat clitics are underlyingly composed of an abs D0 incorporated into a prepositional head spelled out as (i)m-. In this way, dat clitics form part of the larger typology of D0+P0 complex clitics featuring a variety of different prepositions, including the comitative clitic (abs)-(i)baa-, the superessive clitic (abs)-o̲-, and the locative clitic (abs)-aa-. The behavior of these complex clitics with respect to clitic co-occurrence restrictions and Absolutive Promotion is discussed briefly in Tyler (2017), but requires further investigation.
Only two changes have been made from Arregi and Nevins’s original structure, neither significant. Firstly, their v has been relabeled Voice. Secondly, in (78) clitics are base-generated as adjuncts to DPs, while Arregi and Nevins have DPs sitting in the complement of K(ase)P or Part(icipant)P phrases, with clitics being base-generated as adjuncts to either KP or PartP. The proposal outlined here is fully compatible with a more articulated model of clitic generation such as theirs, I simply leave out the details here as they are not relevant to the analysis.
Arregi and Nevins (2012:63) also argue that clitic movement should proceed in two steps: an initial phrasal movement of an XP containing the clitic to a phrase immediately below the clitic host, followed by local head movement of the clitic. They argue that this two-step approach is necessary in order to avoid non-local head movement, which would violate the Head Movement Constraint (Travis 1984). However, for simplicity they do not discuss it further than this, and, here, neither do I.
Note that this description relies on Arregi and Nevins’s analysis of the decomposition of Basque auxiliaries being correct. This is not an uncontroversial point, as their analysis relies on Ø clitics (as in (79)), and an elaborate system of postsyntactic morphological and phonological rules and constraints. I nonetheless assume their account to be correct for present purposes: all arguments are obligatorily clitic-doubled in finite clauses, with the exception of 3rd person singular absolutive arguments.
The one exception here is the 1sgerg form -li, which does appear after the verb root. However, 1sg arguments have somewhat idiosyncratic properties across the whole language (Broadwell and Martin 1993), and I generally set aside issues of morpheme order in this article.
Choctaw’s low clitic-hosting heads can be considered equivalent to the vP-adjacent clitic phrase of Cardinaletti and Shlonsky (2004).
The reader may be concerned by the apparent movement of the erg clitic from inside Spec-VoiceP to Voice0. If clitic movement is like phrasal movement, then this is indeed an odd and potentially illegal movement, as the landing site does not c-command the launch site. One way of obviating the problem is to propose that ergative arguments are clitic-doubled not at Voice0 but some functional head closely above it. Alternatively, we could follow recent proposals arguing that clitic-movement involves an intermediate step in which the D0 clitic first becomes a specifier of VoiceP, above the external argument, before undergoing m-merger with Voice0 to form a complex head (Matushansky 2006). See Harizanov (2014) and Kramer (2014) for approaches to clitic doubling that exploit m-merger in this way.
Note that the Agree relation that H0 establishes with the dative argument, in order to clitic-double it, does not violate minimality. This is because the intervening absolutive argument moves to a position above H0, and A-movement traces are known not to intervene (Chomsky 2000; Anagnostopoulou 2003; Holmberg and Hróarsdóttir 2003).
There is an open question of how a probe could be relativized to Agree only with 1st-person singular arguments, while ignoring plural arguments. I leave this question open here.
I do not attempt to derive the correct morpheme order in (90) via e.g. head-movement, and assume that it must be determined templatically.
If the 1sg.abs>abs/dat clitic clusters really are only available when the 1sg clitic is doubling a subject, then we would predict that these clusters should be unavailable in ditransitives, where the subject position cannot be occupied by an internal argument. As for whether this prediction holds, the jury is still out, as I have been unable to collect consistent judgments (see fn. 26). However, it is potentially significant that, to the extent that clitic clusters in ditransitives are acceptable, speakers find them significantly degraded compared with the same clusters in transitive unaccusatives, which are fine and frequently volunteered.
This leaves the question of why 1sg ergative arguments couldn’t simply clitic-double at Voice0, just like any other ergative argument, in the absence of the Author0 head. I tentatively suggest that Broadwell and Martin’s (1993) analysis may provide a solution. They argue that the 1sgerg form -li is an agreement affix rather than a clitic. Under this analysis, then, there is no true 1sgerg clitic, meaning that 1sgerg arguments are base-generated without clitics, and clitic doubling to Voice0 is therefore not an option. See Tyler (to appear-a, to appear-b) for further support for the agreement-affix analysis of -li.
I leave open the question of why Absolutive Promotion is merely optional in these cases.
Sigurðsson (1996) finds that examples like (97b) improve if the verb agreement is 3sg, and in general, 1st/2nd-person nominative objects are acceptable so long as the verb is prevented from agreeing with them.
Ormazabal and Romero’s (2007) own Object Agreement Constraint is almost a conceptual mirror image of the CCH, in that it restricts licensing-by-object-agreement to one instance per verb, with clitic-doubled arguments being exempt.
Just as the difference in ordering in (101) was argued to derive from different underlying c-command orders within the VoiceP, so Stegovec argues that the different orders in (102) derive from the arguments having different c-command orders within VoiceP. However, while Stegovec argues that to derive the order in (102b), the direct object raises over the indirect object, I proposed that the different clitic orders in (101) derive from different base-generated c-command orders.
Broadwell and Martin show that abs clitics that reference objects most readily participate in this alternation. abs clitics that reference subjects are, for many speakers, restricted to appearing on the main verb. This is a mysterious gap for which I have no explanation. Similarly mysterious is the fact that dat clitics are limited to the lower verb for all speakers. I discuss these gaps in greater detail in Tyler (to appear-a, to appear-b).
Note also that the interaction of Choctaw clitics with oklah is very similar to that of French object clitics and the floating quantifier tous ‘all’, in that it can only associate with subjects or, as shown in (i), clitic objects. As shown in (ii), tous cannot associate with non-clitic objects.
Adger, David, and Daniel Harbour. 2007. Syntax and syncretisms of the Person Case Constraint. Syntax 10: 2–37.
Albizu, Pablo. 1997. Generalized Person-Case Constraint: A case for a syntax-driven inflectional morphology. In Theoretical issues on the morphology–syntax interface, eds. Myriam Uribe-Etxebarria and Amaya Mendikoetxea, 1–33. Donostia: Gipuzkoako Foru Aldundia/UPV/EHU.
Alexiadou, Artemis, Elena Anagnostopoulou, and Christina Sevdali. 2014. Opaque and transparent datives, and how they behave in passives. The Journal of Comparative Germanic Linguistics 17: 1–34.
Anagnostopoulou, Elena. 1999. On experiencers. In Studies in Greek syntax, eds. Artemis Alexiadou, Geoffrey Horrocks, and Melita Stavrou, 67–93. Dordrecht: Springer.
Anagnostopoulou, Elena. 2003. The syntax of ditransitives: Evidence from clitics. Berlin: Mouton de Gruyter.
Anagnostopoulou, Elena. 2005. Strong and weak person restrictions: A feature checking analysis. In Clitic and affix combinations: Theoretical perspectives, eds. Lorie Heggie and Francisco Ordóñez, 199–235. Amsterdam: Benjamins.
Aramaio, Itziar. 2001. Berriatuako aditz laguntzailea. BA thesis, University of Deusto.
Arregi, Karlos. 2004. The have/be alternation in Basque. Ms., University of Illinois, Urbana-Champaign.
Arregi, Karlos, and Andrew Nevins. 2012. Morphotactics: Basque auxiliaries and the structure of spellout. Dordrecht: Springer.
Artiagoitia, Xabier. 2001. Seemingly ergative and ergatively seeming. In Features and interfaces in Romance: Essays in honor of Heles Contreras, eds. Julia Herschensohn, Enrique Mallén, and Karen Zagona, 1–22. Amsterdam: Benjamins.
Artiagoitia, Xabier. 2003. Reciprocal and reflexive constructions. In A grammar of Basque, eds. José Ignacio Hualde and Jon Ortiz de Urbina, 607–632. Berlin: Mouton de Gruyter.
Baker, Mark. 2008. The syntax of agreement and concord. Cambridge: Cambridge University Press.
Baker, Mark. 2014. On dependent ergative case (in Shipibo) and its derivation by phase. Linguistic Inquiry 45: 341–379.
Baker, Mark. 2015. Case: Its principles and parameters. Cambridge: Cambridge University Press.
Bejar, Susana, and Diane Massam. 1999. Multiple case checking. Syntax 2 (2): 65–79.
Béjar, Susana, and Milan Rezac. 2003. Person licensing and the derivation of PCC effects. In Romance linguistics: Theory and Acquisition, eds. Ana-Teresa Pérez-Leroux and Yves Roberge, 49–62. Amsterdam: Benjamins.
Béjar, Susana, and Milan Rezac. 2009. Cyclic Agree. Linguistic Inquiry 40 (1): 35–73.
Belletti, Adriana, and Luigi Rizzi. 1988. Psych-verbs and θ-theory. Natural Language and Linguistic Theory 6 (3): 291–352.
Bonet, Eulalia. 1991. Morphology after syntax: Pronominal clitics in Romance. PhD diss., MIT.
Bonet, Eulalia. 1994. The Person-Case Constraint: A morphological approach. In The morphology–syntax connection, eds. Heidi Harley and Colin Phillips. Vol. 22 of MIT working papers in linguistics, 33–52. Cambridge: MITWPL.
Broadwell, George Aaron. 1988. Multiple theta-role assignment in Choctaw. In Thematic relations, ed. Wendy Wilkins, Vol. 21 of Syntax and semantics, 113–127. San Diego: Academic Press.
Broadwell, George Aaron. 1990. Extending the binding theory: A Muskogean case study. PhD diss., UCLA.
Broadwell, George Aaron. 1997. Binding theory and switch-reference. In Atomism and binding, eds. Hans Bennis, Pierre Pica, and Johann Rooryck, 31–50. Dordrecht: Foris.
Broadwell, George Aaron. 1998. Directionals as complex predicates in Choctaw. In 1998 Lexical-Functional Grammar (LFG) conference, eds. Miriam Butt and Tracy Holloway King.
Broadwell, George Aaron. 2006. A Choctaw reference grammar. Lincoln: University of Nebraska Press.
Broadwell, George Aaron, and Jack Martin. 1993. The clitic/agreement split: Asymmetries in Choctaw person marking. In Annual meeting of the Berkeley Linguistics Society (BLS) 19, 1–10.
Camacho, José. 2010. On case concord: The syntax of switch-reference clauses. Natural Language and Linguistic Theory 28 (2): 239–274.
Cardinaletti, Anna, and Ur Shlonsky. 2004. Clitic positions and restructuring in Italian. Linguistic Inquiry 35 (4): 519–557.
Cecchetto, Carlo. 2000. Doubling structures and reconstruction. Probus 12: 93–126.
Chomsky, Noam. 1995. The minimalist program. Cambridge: MIT Press.
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. Cambridge: MIT Press.
Chomsky, Noam. 2008. On phases. In Foundational issues in linguistic theory. Essays in honor of Jean-Roger Vergnaud, eds. Robert Freidin, Carlos Otero, and Maria Luisa Zubizarreta, 133–166. Cambridge: MIT Press.
Cuervo, María Cristina. 2003. Datives at large. PhD diss., MIT.
Dahlstrom, Amy. 1983. Agent–patient languages and split case marking systems. In Annual meeting of the Berkeley Linguistics Society (BLS) 9, 37–46.
Davies, William. 1986. Choctaw verb agreement and universal grammar. Dordrecht: Reidel.
De Cat, Cécile. 2000. Towards a unified analysis of French floating quantifiers. Journal of French Language Studies 10 (1): 1–25.
Deal, Amy Rose. to appear. Raising to ergative: Remarks on applicatives of unaccusatives. Linguistic Inquiry. https://doi.org/10.1162/ling_a_00310.
Dixon, Robert M. W. 1979. Ergativity. Language 55 (1): 59–138.
Doetjes, Jenny. 1997. Quantifiers and selection: On the distribution of quantifying expressions in French, Dutch and English. PhD diss., Leiden University.
Elordieta, Arantzazu. 2001. Verb movement and constituent permutation in Basque. PhD diss., Leiden University.
Fitzpatrick, Justin Michael. 2006. The syntactic and semantic roots of floating quantification. PhD diss., MIT.
Franks, Steven, and Catherine Rudin. 2005. Bulgarian clitics as K0 heads. In Formal Approaches to Slavic Linguistics (FASL): The South Carolina meeting, 2004, eds. Steven Franks, Frank Gladney, and Mila Tasseva-Kurktchieva, 104–116. Ann Arbor: Michigan Slavic Publications.
Gerdts, Donna, and Kaoru Kiyosawa. 2005. Halkomelem psych applicatives. Studies in Language 29 (2): 329–362.
Harizanov, Boris. 2014. Clitic doubling at the syntax–morphophonology interface. Natural Language and Linguistic Theory 32 (4): 1033–1088.
Heath, Jeffrey. 1977. Choctaw cases. In Annual meeting of the Berkeley Linguistics Society (BLS) 3, 204–213.
Holguín, Justin. 2007. The status of ergative case in Basque: A minimalist approach. B.A. thesis, Reed College.
Holmberg, Anders, and Thorbjörg Hróarsdóttir. 2003. Agreement and movement in Icelandic raising constructions. Lingua 113 (10): 997–1019.
Hualde, José Ignacio, Elordieta Gorka, and Elordieta Arantzazu. 1994. The Basque dialect of Lekeitio. Bilbao: UVP/EHU.
Jelinek, Eloise. 1989. The case split and pronominal arguments in Choctaw. In Configurationality, eds. László Marácz and Pieter Muysken, 117–141. Dordrecht: Foris.
Joppen, Sandra, and Dieter Wunderlich. 1995. Argument linking in Basque. Lingua 97 (2–3): 123–169.
Kramer, Ruth. 2014. Clitic doubling or object agreement: The view from Amharic. Natural Language and Linguistic Theory 32 (2): 593–634.
Laka, Itziar. 1993. The structure of inflection: A case study in X0 syntax. In Generative studies in Basque linguistics, eds. José Ignacio Hualde and Jon Ortiz de Urbina, Vol. 105, 21–70. Amsterdam: Benjamins.
Levin, Theodore. 2017. Successive-cyclic case assignment: Korean nominative–nominative case-stacking. Natural Language and Linguistic Theory 35 (2): 447–498.
Linker, Wendy. 1987. On the coordinating status of the switch-reference markers -chah and -nah in Choctaw. In Muskogean linguistics, ed. Pamela Munro, 96–110. Los Angeles: UCLA. UCLA Occasional Papers in Linguistics.
Marantz, Alec. 1997. No escape from syntax: Don’t try morphological analysis in the privacy of your own lexicon. In University of Pennsylvania working papers in linguistics 21, 201–225.
Matushansky, Ora. 2006. Head movement in linguistic theory. Linguistic Inquiry 37 (1): 69–109.
McFadden, Thomas. 2004. The position of morphological case in the derivation. PhD diss., University of Pennsylvania.
McGinnis, Martha. 1998a. Locality and inert case. In North East Linguistic Society (NELS) 28, 267–281. Amherst: GLSA Publications.
McGinnis, Martha. 1998b. Locality in A-movement. PhD diss., MIT.
McGinnis, Martha. 2000. Event heads and the distribution of psych-roots. In University of Pennsylvania working papers in linguistics 6, 107–144.
Mithun, Marianne. 1991. Active/agentive case marking and its motivations. Language 67: 510–546.
Munro, Pamela. 1999. Chickasaw subjecthood. In External possession, eds. Doris Payne and Immanuel Barshi, 251–289. Amsterdam: Benjamins.
Munro, Pamela. 2016. Chickasaw switch-reference revisited. In Switch reference 2.0, eds. Rik van Gijn and Jeremy Hammond, 377–424. Amsterdam: Benjamins.
Munro, Pamela, and Lynn Gordon. 1982. Syntactic relations in Western Muskogean: A typological perspective. Language 58: 81–115.
Myler, Neil. 2017. Cliticization feeds agreement: A view from Quechua. Natural Language and Linguistic Theory 35: 751–800.
Nevins, Andrew. 2007. The representation of third person and its consequences for person-case effects. Natural Language and Linguistic Theory 25 (2): 273–313.
Nevins, Andrew. 2011. Multiple agree with clitics: Person complementarity vs. omnivorous number. Natural Language and Linguistic Theory 29 (4): 939–971.
Nicklas, Thurston Dale. 1974. The elements of Choctaw. PhD diss., University of Michigan.
Ormazabal, Javier, and Juan Romero. 2007. The Object Agreement Constraint. Natural Language and Linguistic Theory 25 (2): 315–347.
Ormazabal, Javier, and Juan Romero. 2013. Object clitics, agreement and dialectal variation. Probus 25: 301–344.
de Urbina, Jon Ortiz. 2003. Causatives. In A grammar of Basque, eds. José Ignacio Hualde and Jon Ortiz de Urbina, 592–607. Berlin: de Gruyter.
Oyharçabal, Beñat. 1992. Structural case and inherent case marking: Ergaccusativity in Basque. In Syntactic theory and Basque syntax, eds. Joseba Andoni Lakarra and Jon Ortiz de Urbina, 309–342. Bilbao: UPV/EHU.
Oyharçabal, Beñat. 2003. Lexical causatives and causative alternation in Basque. In Inquiries into the syntax-lexicon relations in Basque, ed. Beñat Oyharçabal, 223–253. Bilbao: UPV/EHU.
Payne, Doris. 1982. Chickasaw agreement morphology: A functional explanation. In Studies in transitivity, eds. Paul Hopper and Sandra Thompson. Vol. 15 of Syntax and semantics, 351–378. New York: Academic Press.
Perlmutter, David M. 1971. Deep and surface structure constraints in syntax. New York: Holt, Rinehart & Winston.
Pesetsky, David. 1995. Zero syntax: Experiencers and cascades. Cambridge: MIT Press.
Pesetsky, David. 2013. Russian case morphology and the syntactic categories. Cambridge: MIT Press.
Poletto, Cecilia. 2000. The higher functional field: Evidence from Northern Italian dialects. Oxford: Oxford University Press.
Preminger, Omer. 2009. Breaking agreements: Distinguishing agreement and clitic doubling by their failures. Linguistic Inquiry 40 (4): 619–666.
Preminger, Omer. 2011. Agreement as a fallible operation. PhD diss., MIT.
Preminger, Omer. 2012. The absence of an implicit object in unergatives: New and old evidence from Basque. Lingua 122 (3): 278–288.
Preminger, Omer. 2014. Agreement and its failures. Cambridge: MIT Press.
Preminger, Omer. to appear. What the PCC tells us about ‘abstract’ agreement, head movement, and locality. Glossa.
Pylkkänen, Liina. 2008. Introducing arguments. Cambridge: MIT Press.
Rezac, Milan. 2008a. Phi-agree and theta-related case. In Phi theory: Phi-features across interfaces and modules, eds. Daniel Harbour, David Adger, and Susana Bejar, 83–129. Oxford: Oxford University Press.
Rezac, Milan. 2008b. The syntax of eccentric agreement: The Person Case Constraint and absolutive displacement in Basque. Natural Language and Linguistic Theory 26 (1): 61–106.
Rezac, Milan. 2009. Person restrictions in Basque intransitives. Lapurdum 13: 305–322.
Rezac, Milan. 2010a. On the unifiability of repairs of the Person Case Constraint: French, Basque, Georgian, and Chinook. In Festschrift for Beñat Oyharçabal, eds. Ricardo Etxepare, Ricardo Gómez, and Joseba Lakarra, 769–789. Bilbao: UPV/EHU.
Rezac, Milan. 2010b. ϕ-agree versus ϕ-feature movement: Evidence from floating quantifiers. Linguistic Inquiry 41 (3): 496–508.
Rezac, Milan. 2011. Phi-features and the modular architecture of language. Berlin: Springer.
Rezac, Milan, Pablo Albizu, and Ricardo Etxepare. 2014. The structural ergative of Basque and the theory of Case. Natural Language and Linguistic Theory 32 (4): 1273–1330.
Rivero, María Luisa. 2004. Spanish quirky subjects, person restrictions, and the person-case constraint. Linguistic Inquiry 35 (3): 494–502.
Rizzi, Luigi. 1990. Relativized minimality. Cambridge: MIT Press.
Roberts, Ian. 2010. Agreement and head movement: Clitics, incorporation, and defective goals. Cambridge: MIT Press.
San Martin, Itziar. 1999. Control in Basque. Ms., University of Maryland.
Schütze, Carson. 1995. Case, verb morphology and argument structure in Choctaw: A minimalist account. In North East Linguistic Society (NELS) 25, 441–456.
Sigurðsson, Halldór Ármann. 1996. Icelandic finite verb agreement. In Working papers in Scandinavian syntax, Vol. 57, 1–46.
Silverstein, Michael. 1985. Hierarchy of features and ergativity. In Features and projections, eds. Pieter Muysken and Henk van Riemsdijk, 163–232. Dordrecht: Foris.
Sportiche, Dominique. 1996. Clitic constructions. In Phrase structure and the lexicon, eds. Johan Rooryck and Laurie Zaring, 213–276. Dordrecht: Kluwer Academic.
Stegovec, Adrian. 2017. Personality disorders: The PCC and two kinds of clitic reordering in Slovenian. In Formal Approaches to Slavic Linguistics (FASL) 24: The New York meeting, eds. Yohei Oseki, Maria Esipova, and Stephanie Harves, 342–362. Ann Arbor: Michigan Slavic Publications.
Travis, Lisa. 1984. Parameters and effects of word order variation. PhD diss., MIT.
Tsakali, Vina. 2008. “Double” floating quantifiers in Modern Greek and Pontic. In Microvariation in syntactic doubling, eds. Sjef Barbiers, Olaf Koeneman, Marika Lekakou, and Margreet van der Ham, Vol. 36 of Syntax and semantics, 189–203. Bingley: Emerald.
Tyler, Matthew. to appear-a. In Choctaw, everyone’s a clitic. In Annual meeting of the Berkeley Linguistics Society (BLS) 43.
Tyler, Matthew. to appear-b. Choctaw as a window into the clitic/agreement split. Studia Linguistica.
Tyler, Matthew. 2017. PCC repair in Basque and Choctaw transitive unaccusatives. In Annual meeting of the Chicago Linguistics Society (CLS) 53.
Ulrich, Charles. 1986. Choctaw morphophonology. PhD diss., UCLA.
Uriagereka, Juan. 1995. Aspects of the syntax of clitic placement in Western Romance. Linguistic Inquiry 26 (1): 79–123.
Wood, Jim. 2015. Icelandic morphosyntax and argument structure. Dordrecht: Springer.
Woolford, Ellen. 2006. Lexical case, inherent case, and argument structure. Linguistic Inquiry 37 (1): 111–130.
Woolford, Ellen. 2008. Is agreement really independent of case in Choctaw? Ms., University of Massachusetts, Amherst.
Woolford, Ellen. 2010. Active-stative agreement in Choctaw and Lakota. Revista Virtual de Estudos da Linguagem 8: 6–46.
Yoon, H.-S. James. 2004. Non-nominative (major) subjects and case stacking in Korean. In Non-nominative subjects, eds. Peri Bhaskararao and Karumuri Venkata Subbarao, Vol. 2, 265–314. Amsterdam: Benjamins.
Yuan, Michelle. 2017. Diagnosing object agreement vs. clitic doubling: Evidence from Inuktitut. Ms., MIT.
Zwicky, Arnold M., and Geoffrey K. Pullum. 1983. Cliticization vs. inflection: English n’t. Language 59: 502–513.
My greatest thanks goes to Jim Wood for his insight, advice and encouragement at all stages of this project. I am extremely grateful to four anonymous reviewers and NLLT associate editor Daniel Harbour, who made this work immeasurably better. Thanks also to Stephen Anderson, Karlos Arregi, Claire Bowern, Aaron Broadwell, Bob Frank, Andrew Nevins, Michelle Yuan and Raffaella Zanuttini for comments, discussion and logistical help, as well as audiences at LSA 2017, CLS 53 and Yale University. This article would not have been possible without the Mississippi Choctaw speakers who took the time to share their language with me, always with patience, insight and good humor, and they have my deepest gratitude. They are: Elijah Ben, Patty Billie, Chris Chickaway, Shayla Chickaway, Zonie Isaac, Buck Willis and Darlene Willis. I am also grateful to the Language Program at the Mississippi Band of Choctaw Indians for generously hosting me during my visits to MBCI, and the Tribal Council and the Office of the Chief. This work was partly supported by a Pre-Dissertation Research Grant from the Yale MacMillan Center. All errors are my own.
Appendix A: Clitics vs. agreement affixes
Appendix A: Clitics vs. agreement affixes
In Sect. 2, it was asserted without further argument that the argument-referencing morphemes in (4) were clitics rather than agreement affixes, and at several points on the paper, this property forms a crucial part of the analysis. In particular, the account of the difference in the distribution of Absolutive Promotion in Choctaw vs. Basque, discussed in Sect. 4.3, relies on the intervention-voiding properties of clitic-doubling. Likewise, the explanation for the clitic co-occurrence restrictions presented in Sect. 5.3 crucially relies on this property. This appendix provides several independent arguments in support of the claim that at least the abs and dat forms are clitics. In Tyler (to appear-a, to appear-b), I provide a more extensive list of arguments for the clitichood of the Choctaw argument-referencing forms, including for the erg paradigm.
I begin with arguments for clitichood that do not rely on any particular syntactic implementation of the clitic/agreement split, before moving on to arguments that rely on the distinction between clitics as D0 heads that head A-chains, and agreement as D0-less bundles of ϕ-features.
A.1 General arguments for clitichood
The first argument provided here only applies to the dat clitics, and relates to their phonological status. Choctaw has a phonological process of iambic lengthening, by which even-numbered non-final syllables in a string of light syllables are lengthened (Nicklas 1974; Ulrich 1986; Broadwell 2006). Ulrich (1986), Broadwell and Martin (1993) and Broadwell (2006) show that dat clitics fall outside the phonological domain over which iambic lengthening takes place. This is evidence for their phonological clitichood, with which syntactic clitichood is often (though not necessarily) correlated. Note that while abs clitics do fall within the domain of iambic lengthening, there is other evidence suggesting their clitic rather agreement status.
The second argument for the clitichood of the abs and dat forms is that both possible morpheme orderings are attested, as shown in (101).
This reorderability would not be expected of agreement forms, which generally show rigid order. Clitics, on the other hand, are often available in different orders, as in (102), from Slovenian.Footnote 55
A third argument for the forms’ clitichood comes from the fact that they display no morphologically-conditioned allomorphy, and, for the dat forms at least, they do not trigger any allomorphy either. Zwicky and Pullum (1983) argue that this morphological isolation is characteristic of clitichood but not of agreement. Nevins (2011) refines this argument, proposing that clitics always remain invariant with respect to tense, but may take part in other allomorphy. In both Zwicky and Pullum’s and Nevins’s diagnostics, the Choctaw forms pattern as clitics rather than agreement forms.
A fourth argument is that both the abs and dat morphemes take place in alternations that resemble Romance clitic climbing. I take the characteristic property of clitic climbing to be that in a clause-union or restructuring clause, a clitic may appear in one of two positions—one attached to the higher verb, one to the lower—but not in both. (103) shows a representative example from Spanish.
While dat clitics do not participate in this alternation in clauses with auxiliaries, they do show a clitic-climbing-like alternation elsewhere. Specifically, when a jussive clause is embedded under the verb ahni, which here is best translated as ‘want’, an abs or dat clitic may surface on the embedded or matrix verb:Footnote 57
A.2 Theoretically-rooted arguments for clitichood
Section 5.1 outlined the theoretical assumptions of this article with respect to the clitic/agreement split. While clitics are the spellout of determiner heads (D0s) that head A-chains, agreement morphemes are the spellout of bundles of ϕ-features copied onto agreement probes.
Given this distinction, Kramer (2014) makes the argument that if clitics are D0 heads exponing the ϕ-features, and if possessive determiners are also D0 heads exponing ϕ-features, then we should expect them to share a paradigm. This is exactly what we find in Choctaw: the abs clitics lead doubles lives as DP-internal markers of inalienable possession, and dat clitics do the same for alienable possession:
A second argument comes from Preminger (2014). He points out that while agreement probes may be relativized to only probe for a subset of an argument’s ϕ-features—only number, for instance—clitics are necessarily featurally ‘coarse’, and must expone all the ϕ-features of their argument in a single bundle. The Choctaw abs and dat forms do just this.
A third theoretically-rooted argument comes from the interaction between the dat and abs forms and the plural-marking element oklah. Broadwell (2006) describes oklah as a ‘preverb’ that indicates that the subject of a clause is plural, as shown in (108).
However, in Tyler (to appear-a, to appear-b), I show that oklah may also appear in the presence of 1st and 2nd-person plural objects—i.e. those objects that are clitic-doubled by abs clitics. (109a) shows oklah being licensed by an object-doubling abs clitic, (109b) shows the same for an object-doubling dat clitic, and (109c) shows that 3rd-person (non-clitic-doubled) objects cannot license oklah.Footnote 58
Following the logic of Rezac (2010b), I propose that this pattern constitutes evidence for the clitichood of the abs and dat forms. I assume that oklah has the same licensing conditions as a floating quantifier—it needs to be c-commanded from an A-position by the argument it associates with (possibly, that argument must bind some variable in the structure of oklah, see Doetjes 1997; Fitzpatrick 2006). Typically, in-situ objects do not c-command oklah and so cannot license it. However, 1st/2nd-person objects are clitic-doubled at a higher clitic host (see Sect. 5.3). Given that argument-doubling clitics function as arguments in A-positions with respect to binding (on which see Sect. 4.3 and references cited there), the oklah-licensing properties of the clitics follow, provided that they are attached at a structural position from which they c-command oklah. This interaction, crucially, would be difficult to account for if the abs and dat morphemes were agreement affixes rather than clitics: Rezac (2010b) shows that agreement-bearing participles in French are unable to license floating quantifiers. Similarly, Tsakali (2008) and Harizanov (2014) show that clitics can license floating quantifiers in Greek and Bulgarian respectively.
About this article
Cite this article
Tyler, M. Absolutive Promotion and the Condition on Clitic Hosts in Choctaw. Nat Lang Linguist Theory 37, 1145–1203 (2019). https://doi.org/10.1007/s11049-018-9426-z