Skip to main content
Log in

Anti-locality and optimality in Kaqchikel Agent Focus

  • Published:
Natural Language & Linguistic Theory Aims and scope Submit manuscript


Many Mayan languages show a syntactically ergative extraction asymmetry whereby the A̅-extraction of subjects of transitive verbs requires special verbal morphology, known as Agent Focus. In this paper I investigate the syntax of Agent Focus in Kaqchikel, a Mayan language spoken in Guatemala. I argue that this extraction asymmetry in Kaqchikel is the result of a particular anti-locality constraint which bans movement that is too close. Support for this claim comes from new data on the distribution of Agent Focus in Kaqchikel that show this locality-sensitivity.

The distribution and realization of Agent Focus will then be modeled using a system of ranked, violable constraints operating over competing derivations. This theoretical choice will be supported by details in the pattern of agreement in Agent Focus. I will then show how rerankings of the proposed constraints can model the attested distribution of Agent Focus in a number of other Mayan languages. I also discuss extensions of this approach to other patterns of anti-agreement.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others


  1. Abbreviations used: A = Set A agreement, AF = Agent Focus, B = Set B agreement, COM = completive aspect, INC = incompletive aspect, FOC = focus marker, RC = relative clause marker, ∅ = empty string, for phonologically null morphemes. The aspectual terms “completive” and “incompletive” are used in the Mayan literature and are adopted here. The semantics of the aspectual system is not relevant to discussions here.

  2. With the notable exception of topicalization, which will be discussed in Sect. 4.4.

  3. The precise morphological realization of AF and the syntactic constructions which trigger it differ across the various Mayan languages. Stiebels (2006) presents a cross-Mayan survey of these aspects of AF but Kaqchikel is not included in that study. This section thus also acts to contribute a missing data point in this cross-Mayan look at AF.

  4. The Set A and Set B ϕ-agreement series are also observed in the nominal domain. Possessor agreement on DPs uses the Set A markers and free pronouns are based on the Set B markers. Here I will limit attention to the Set A and Set B markers in the verbal complex which cross-reference arguments of the verb.

  5. Full paradigms for the agreement markers in this variety of Kaqchikel are given in Preminger (2011).

  6. Relative clauses are introduced by ri. This could potentially be analyzed as another use of the definite determiner ri, as suggested by a reviewer, but I descriptively gloss them here as RC.

  7. But note that similar existential constructions which trigger AF are also attested in Popti’ (Craig 1979, fn. 8), K’iche’ (Campbell 2000, fn. 13), Poqomam (Dayley 1981, discussed in Stiebels 2006), Tzotzil (Aissen 1999a), Tz’utujil (Dayley 1985, fn. 8, discussed in Duncan 2003), and Yucatec Maya (Tonhauser 2003). However, some of these existential constructions in other Mayan languages may be a biclausal combination of an existential predicate taking a relative clause introduced by a relative pronoun or wh-word. We will see later this section that argument existential constructions in Kaqchikel do not underlyingly involve the formation of a relative clause. See also Hedberg (1988) who observes that transitive subject existentials trigger AF in another variety of Kaqchikel.

  8. Note that Kaqchikel also has an existential predicate k’o, which is a different lexical item than the existential operator k’o. We can distinguish these two items by their inflection and lack thereof: the predicate k’o exhibits Set B agreement with its argument, but the argument existential k’o never shows agreement. See also footnote 14 for an additional argument that argument existentials are not built using the existential predicate.

    Note also that majun looks like the negation ma and the numeral jun ‘one.’ However, I argue that it is not compositional in the synchronic grammar: when the numeral ‘one’ is actually compositionally negated, it shows the irrealis clitic ta which normally cooccurs with the negation ma. The use of majun, however, does not trigger the use of ta.

  9. The first-person subject pronoun yïn in the examples in (22) is topicalized. This is important for the derivation of (22a), which does not involve AF, as will be discussed in detail in Sect. 4.4. The subject pronoun is strongly preferred in (22a) but optional in (22b–e), which can be explained by a requirement that there be at least one constituent before the verb in declarative clauses (Clemens 2013).

  10. The relative clause in the baseline sentence (24a) is a subject relative headed by “man” with a preverbal object existential. The relative clause in (24a) notably lacks AF on its verb form. If AF is used, the relative clause is no longer grammatical with the intended interpretation. Relative clauses of this form are an important part of the argumentation in this paper and will be discussed in detail in Sect. 3.2. Here I focus on the contrast between (24a) and (24b) to establish the A̅-movement properties of the argument existential k’o. Note also that the structures in (24b) and (25b) continue to be ungrammatical even if the embedded verb is changed to AF in (24b) or the embedded verb is changed to be non-AF in (25b).

  11. Many but not all preverbal adverbs have this effect. These do not form natural classes—for example, aninäq ‘quickly’ obviates AF in this way, but the synonym jonamin does not. Formally, I adopt a suggestion by an anonymous reviewer that the adverbs which do not obviate AF are those which are adjoined to projections such as TP, vP, and VP that are always present, in contrast to those which obviate AF which require the projection of an additional functional projection (AdvP) (Cinque 1999). Future work is required to identify independent correlates of the distinction between these two classes of adverbs.

  12. With one exception: in matrix multiple wh-questions, only one wh-word fronts.

  13. I assume that multiple CP maximal projections will be projected in order to host the multiple A̅-operators in the periphery, with one specifier per CP projection (Watanabe 1992; Rizzi 1997). This makes the projections labeled “CP1” and “CP2” here distinct for the purposes of evaluating whether a movement step is too short or not. More detailed derivations will be presented below.

  14. A reviewer asks whether the argument existential k’o could actually be an existential predicate k’o (discussed in footnote 8) taking a headless relative clause. The grammaticality of examples such as (33) with multiple, crossing argument existentials shows that this cannot be the case, because argument existentials are sensitive to relative clause islands (24).

  15. Example (35) follows the orthography given in Broadwell (2000, 2007). Note that man jun in (35) is the non-existential operator in Patzicía Kaqchikel, corresponding to majun in the Patzún variety that I focus on here.

  16. Broadwell (2000, appendix) also makes a similar observation in passing, citing the pair repeated here as (35): “In sentences with multiple foci, it is the closest focus that determines whether the actor focus form is used.”

  17. The argument existential for the indirect object in (37b) involves k’o, the posessor wh-word achoj, and the preposition che. K’o achoj che together acts as “to someone.”

  18. The contraint XRef applies on an abstract level of feature agreement, rather than evaluating the output morphological form. In particular, we note that the Set B marker is ∅ in its third-singular form. In such cases, agreement of a third-singular goal by the B probe will count for satisfaction of XRef, even though the resulting morphology is ∅.

  19. The use of the B probe in intransitives could alternatively be described as the result of an Economy constraint on movement, as the A probe will trigger movement of its target but the B probe does not. See footnote 29.

  20. This is based on the analysis for Mayan VOS word order proposed in Aissen (1992), which models post-verbal subjects as a right specifier of VP.

  21. Subject-initial word orders will be discussed in Sect. 4.4.

  22. Tada (1993) observes a generalization across Mayan between the position of the absolutive (Set B) agreement marker and the existence of AF: with few exceptions, Mayan languages with Set B markers preceding the verbal root exhibit AF and Mayan languages with Set B markers following the verbal root do not. See Coon et al. (2014) for further discussion. Kaqchikel follows this generalization, having its Set B marker before the verbal root and exhibiting AF.

    My proposal here will not attempt to explain this correlation, describing the position of the Set B marker as simply part of the shape of the verbal complex in (48). I instead focus primarily on the Kaqchikel-internal distribution of AF, although the distribution of AF in the Mayan languages of Popti’ and Akatek will be discussed later in Sect. 6.1.

    It is not clear whether or not Tada’s generalization is a problem for my analysis. As noted by Aldridge (2012), Tada’s generalization does not hold outside of Mayan: in the Austronesian language family, there are languages with structurally higher and lower sources of absolutive which have a syntactically ergative extraction asymmetry similar to AF. (See Aldridge (2008) for discussion of both types of languages in Austronesian.) It is possible that the correlation observed by Tada reflects a historical codevelopment between AF and the preverbal Set B markers, rather than a deep fact about the source of syntactic ergativity. See also Henderson et al. (2013) for evidence that AF in Mayan may not be a unified phenomenon, as well as footnote 56 for a note on Ixil, a Mayan language that does not obey Tada’s generalization.

  23. In some cases, there will be multiple potential ϕ-agreement targets for the B probe. As mentioned above, in such cases the Set B agreement target will be determined, descriptively speaking, following the salience hierarchy in (13). The effect of this salience hierarchy on the choice of agreement target will be formally modeled in Sect. 5.

  24. The realization of the AF suffix depends on the verb stem type: if the verb stem is V-final or Vj-final, the AF suffix is -n (in the latter case, replacing the final -j consonant); otherwise it is .

  25. Note that Kaqchikel lacks ditransitives such as English tell which take both DP and CP internal arguments. Kaqchikel also lacks true ditransitives with two DP arguments, with verbs such as send instead having a DP direct object which is targeted by Set B agreement and a PP indirect object which is not ϕ-agreed with.

  26. Bošković (2003) cites Stjepanović and Takahashi (2001) and Lee (2003) as precursors for this view. Bošković (2003, 2007) provides a theoretical motivation for this difference.

  27. Similar “skipping” strategies are surveyed cross-linguistically in Rizzi and Shlonsky (2007). See also Schneider-Zioga (2007) for a similar skipping derivation forced by an anti-locality constraint in Kinande and Rouveret (2002) for brief discussion of a similar skipping derivation for anti-agreement in Celtic languages. Similar interactions between agreement and anti-locality yielding anti-agreement effects in languages other than Kaqchikel will be discussed in Sect. 6.

  28. The situation is the same with an unaccusative verb: T’s B probe will agree with the subject of the unaccusative verb, even though it is in a lower phase.

  29. An alternative view would be to describe both the A probe and B probe as optional, and derive the use of the B probe in intransitive clauses through Economy. The idea is that ϕ-agreement of the intransitive subject by the A probe or the B probe would satisfy the constraint XRef equally well in the derivation in (55)—in either case, the verb’s single argument will be cross-referenced by the verb. However, the A probe necessarily triggers movement of the subject, because the A probe has the EPP property. Since this alternative derivation using the A probe involves an additional movement step, it will be dispreferred by Economy considerations.

    In Optimality-Theoretic terms, this alternative derivation which uses the A probe will be harmonically bounded by the proposed derivation which uses the B probe. Because of this, in my formalization of this system with ranked, violable constraints in Sect. 5, the particular ranking of such an Economy constraint will not be important and therefore will not be illustrated. I thank an anonymous reviewer and Ellen Woolford (p.c.) for their thoughts on this point.

  30. See also Woolford (2015) for arguments against postulating covert ergative case in Mayan languages.

    The effects of the constraint XRef here are similar to a violable Case Filter (Grimshaw 1997; de Hoop and Malchukov 2008). I have chosen to state this constraint in terms of cross-referencing here as ϕ-agreement is clearly visible in Kaqchikel, whereas the assignment of Case is not.

    The logic of this system is similar to the dependent case system of Marantz (1991), which models ergative case in an ergative/absolutive system and accusative case in a nominative/accusative system as “dependent” cases, whose occurance depends on the existence of a competitor in the same local domain. The proposal here applies a similar logic to ϕ-agreement instead of morphological case marking. The effects of this system are also similar to proposals which prohibit two structural case targets (Alexiadou and Anagnostopoulou 2001, 2007) or two nodes which are categorically non-distinct (Richards 2010) from both being in vP, which may lead to one of them vacating the vP.

  31. I remain agnostic here as to whether the multiple CP-level projections are part of a split left periphery that contains multiple heads (Rizzi 1997) or an extension of a single C head (Watanabe 1992). This choice is not crucial to the analysis presented.

  32. Such subject initial word orders are common in this variety of Kaqchikel (Clemens 2013), and in modern Kaqchikel more generally (England 1991; García Matzar and Rodríguez Guaján 1997). Other researchers, however, report that verb-initial orders are still very productive in at least some varieties of the language (Broadwell 2000; Brown et al. 2006).

  33. An anonymous reviewer asks whether movement of an argument from inside CP to Spec,TopP without passing through Spec,CP will violate Phase Impenetrability. It need not, by adopting the view that the highest head in an extended projection will behave as the phase head (Bošković 2014).

  34. Based on my observation that only immediately preverbal subjects trigger AF, reported here in Sect. 3.2, Clemens (2013) suggests a similar explanation for SVO word order clauses without AF. Clemens (2013) proposes that non-AF-triggering preverbal subjects are in a structurally higher position than preverbal A̅-operators which do trigger AF. See Clemens (2013, Sect. 4.1) for details.

  35. Note that this characterization does not extend to all Mayan languages with AF. For example, AF is optional in all triggering contexts in Poqomam and Poqomchi’ (Stiebels 2006).

  36. Coon et al. (2014) argues for this view that, in Mayan languages with AF, the subject of transitive clauses cannot be straightfowardly extracted due to a problem with Case-licensing the object. However, they then state that languages may have different strategies for extracting the subject argument in such cases. They specifically note that their analysis of AF as a last-resort case assigner, put forth for Q’anjob’al, may not extend to other Mayan languages with AF, including Kaqchikel.

  37. See in particular Ordóñez (1995) which uses Agent Focus in Popti’ as an explicit argument for Chomsky’s (1991) notion of “last-resort”; Coon et al. (2014) follows Ordóñez (1995) in describing this strategy as a “last-resort.” Assmann et al. (2013:404) also acknowledges this “repair strategy” question.

  38. There are two previous approaches to aspects of the distribution and realization of AF in an Optimality-Theoretic syntax framework.

    Stiebels (2006) presents an analysis in terms of Lexical Decomposition Grammar, where each argument in the input’s Semantic Form is valued for certain features. Stated in terms of arguments having higher roles or lower roles ([±hr], [±lr]), its logic is similar to a dependent case theory à la Marantz (1991). This system, as stated in Stiebels (2006), is insensitive to the locality of movement involved, and is therefore unable to derive the pattern of AF documented here.

    Preminger (2011:98–100) offers a sketch for the agreement alignment on AF verbs in K’ichean, based primarily on Kaqchikel data, but does not attempt to derive the general distribution of AF at the same time. Preminger (2011, 2014) ultimately argues against the type of violable constraints model that I propose here; see footnote 42 below.

  39. I will refer to the candidates here as derivations, to reflect the derivational nature of the Spec-to-Spec Anti-Locality constraint as defined in (68a). See Müller (1997), particularly footnote 20, for discussion of the use of derivational constraints in Optimality-Theoretic syntax. Note, however, that if the candidate set were made up of representational outputs of syntax which retain information on movement chains, we could think of the Spec-to-Spec Anti-Locality constraint as a representational constraint.

    As noted in Müller (1997), Müller and Sternefeld (2001) and others, the consideration of transderivational constraints is certainly not incompatible with derivational syntax in the Minimalist tradition, and is not without precedent. See for example the notion of “reference set” in Chomsky (1995), which selects for converging derivations with the same input numeration. See also Fox (1995) and Reinhart (1998) for examples of competition between derivations with identical semantic interpretations.

    See further discussion in Sect. 5.3.

  40. Note that this formulation of XRef is functionally equivalent to Stiebels’s (2006) Max(ϕ) and Preminger’s (2011) HaveAgr.

  41. Preminger (2011, 2014) presents arguments against modeling this pattern of agreement in AF, which descriptively follows the salience hierarchy in (74), through a set of violable constraints such as {XRef-Participant, XRef}. (A constraint equivalent to (75) is called HaveAgrWith1/2 there.) In brief, he notes a striking similarity between the first- and second-person forms for Set B agreement markers and pronouns, but not for third-person forms. He proposes that Set B agreement is actually a combination of clitic doubling for first- and second-person arguments and a number probe for third-person arguments, and that both the preference for agreeing with first- and second-person arguments in AF as well as the morphological form of Set B markers can be explained together through this proposal. Instead, under my proposal, Set B agreement is a uniform process, whose target is governed by the constraint interaction modeled here, and the apparent connection between the agreement alignment in AF and the form of Set B markers would be a morphological accident. Preminger (2014) also notes a possible extension of his approach to patterns of argument licensing in Zulu.

    While this argument makes Preminger’s approach attractive, note that he limits his attention to discussing the pattern of agreement under AF. He offers no explanation for why AF generally appears obligatorily when transitive subjects are extracted and why this requirement for AF is obviated under certain person configurations, let alone its locality-sensitive distribution presented here. The proposal here instead offers a comprehensive approach to the distribution of Kaqchikel AF and its agreement facts.

  42. Here I will abstract away from the preference to agree with third-plural arguments over third-singular ones. Formally, it suffices to introduce a similar constraint, XRef-Plural, which assigns violations for plural ϕ-agreement targets which are not agreed with. I will not discuss the effect of this constraint, however, as the preference for agreeing with third-plural arguments over third-singular arguments will not yield the interesting AF-overriding behavior which I will present with first- and second-person arguments below.

  43. Recall that, unlike the A probe, the B probe is able to probe into the lower phase, so both targets are possible. See discussion at the end of Sect. 4.1.

  44. The same pattern has been observed in some dialects of the sister language K’iche’ by Mondloch (1981:223), as discussed in Stiebels (2006), and is also observed in Chuj (Robertson 1980:144). Focus extractions of transitive subjects requires AF morphology, with one exception—when both the subject and object are first- or informal second-person, in which case regular transitive verbal morphology is used instead.

  45. As noted in footnote 37 above, Coon et al. (2014) also put forth an analysis of AF as a last-resort case assigner for Q’anjob’al, but makes it clear that this approach is not intended for AF in other Mayan languages including Kaqchikel.

  46. In addition, the Case-based approaches do not predict the generalization that AF is sensitive to the locality of extraction, as shown in previous sections.

  47. The Spec-to-Spec Anti-Locality constraint proposed here may also have positive consequences beyond the derivation of anti-agreement effects. David Pesetsky (p.c.) notes that this constraint makes the prediction that if the heads X and Y in schema “[\(_{\mbox{\tiny YP}}\) α [ Y [\(_{\mbox{\tiny XP}}\) \(t_{\alpha}\) [ X” are both phase heads, the configuration will be an island for extraction. The logic is as follows: as movement out of phases must proceed through their edge, movement out of YP must necessarily stop at both Spec,XP and Spec,YP. However, Spec-to-Spec Anti-Locality explicitly bans movement from Spec,XP to Spec,YP. Thus this configuration is predicted to necessarily be an island. This configuration may indeed obtain under certain head-raising analyses of relative clauses (if the DP projection immediately dominates a CP) or with adjunct clauses (if a CP is adjoined at the vP level), which are both known islands for extraction. Note that using the combination of an anti-locality constraint and Phase Impenetrability to capture extraction restrictions is also the logic of Abels’s (2003) work on stranding. However, Abels considers a different formulation of anti-locality, explaining a different set of facts.

    The English that-trace effect may also be conducive to an anti-locality-based explanation. The idea would be that the extraction of a subject across the complementizer that is movement from Spec,TP to Spec,CP and therefore violates Spec-to-Spec Anti-Locality. Support for such a view comes from the fact that the addition of certain adverbs can obviate that-trace effects, as discussed by Bresnan (1977), Culicover (1993) and others. Movement of non-subjects to Spec,CP does not violate Spec-to-Spec Anti-Locality and therefore is not subject to restrictions on the form of complementizer chosen. See Erlewine (2014b) for such a proposal for the English that-trace effect based on the Spec-to-Spec Anti-Locality.

  48. As noted by an anonymous reviewer, the ranking of XRef-Participant ≫ XRef may be predicted to be universal following Aissen (1999b, 2003). Aissen derives similar rankings based on universal prominence scales and Prince and Smolensky’s (1993) alignment process (ch. 8) for the generation of corresponding constraint rankings. See Aissen (1999b, 2003) for details.

  49. Basic word order in both Popti’ and Akatek is VSO (Ordóñez 1995; Zavala 1997). Popti’ has also been called “Jakaltek” and “Jacaltec” in previous literature.

  50. This final point is important when comparing AF constructions across Mayan, as some languages utilize a general antipassive strategy, including demotion of the object to an oblique, instead of a dedicated AF construction in cases of transitive subject extraction. See Sect. 2.3.

  51. The motivation for this ranking is the same as for Kaqchikel in Sect. 5. Spec-to-Spec Anti-Locality ≫ XRef forces the subject to skip the Spec,TP position in transitive subject extractions.

  52. A reviewer asks how Set B agreement in Popti’ and Akatek can be ensured to strictly target the object. One option would be to locate the B probe in Popti’ and Akatek on a lower head than in Kaqchikel: for example, if the B probe is located on v, when v is merged into the structure, the B probe will probe down and find the object. This suggestion, however, is speculative at this point. I will leave for future research the precise mechanism ensuring this alignment.

  53. The orthography here for Popti’ follows Craig (1979). The symbol 7 is a glottal stop. Craig (1979) writes example (87) with a space in the middle of the verbal complex, but not (88). The order of morphemes in the verbal complex is, however, identical across Popti’, Akatek, and Kaqchikel.

    The same pattern of Popti’ data as in (87)–(88) with first-person used instead of second-person is documented in Robertson (1980:20).

  54. The orthography here for Akatek follows Zavala (1997). The directional suffix -toj is glossed by Zavala (1997) as “DIR:thither” and naj is a nominal classifier.

  55. According to Craig (1979), Ayres (1977) offers a description of Ixil which patterns with Akatek. That is, the AF form in Ixil “exhibits the same pattern of ergative deletion with the absolutive marker cross-referencing the patient” and furthermore uses the AF verb form for focus extractions of participants (Craig 1979:12). Robertson (1980:145) independently gives Ixil data bearing out this claim, reproduced in (i–ii) below. (Robertson’s (1980) English translation for example (ii) is “I hit you” with underlining on the subject to indicate focus.)

    figure bw

    It is worth noting that Ixil is an exception to Tada’s generalization (footnote 22), as its Set B marker follows the verbal root and yet it exhibits AF (ii).

  56. I preliminarily note that the types of arguments I present in Sect. 3 for Kaqchikel may not all be testable in other languages, for independent reasons. For example, multiple simultaneous A̅-extractions as in Sect. 3.2 are not possible in Q’anjob’al (Pedro Mateo Pedro, p.c.).

  57. Stiebels (2006) briefly discusses the similarity of Mayan AF to anti-agreement effects, but ultimately concludes that they cannot be analyzed together.

  58. We can imagine other strategies for avoiding the anti-locality violation besides this skipping technique. This will be discussed in the conclusion, Sect. 7.

  59. In these examples, cl is a preverbal clitic which also agrees with a preverbal subject. For our purposes, the behaviors of Trentino and Fiorentino are identical.

  60. An important question is why fronting of the subject is apparently optional, as exemplified by the “inversion” structures in (97), in violation of XRef. As noted in Brandi and Cordin (1989:fn. 6), inversion only occurs when the subject is focused. It is therefore a certain information-structural requirement, that focused arguments stay low, which outranks XRef.

  61. The formal similarity of Mayan AF and Karitiâna anti-agreement was first discussed in Hale and Storto (1996) and Richards (1997).

  62. The morpheme -mon appears in absolutive wh-questions but not other wh-questions. Storto (1999) identifies the suffix as a copula and therefore analyzes absolutive wh-questions as wh-clefts.

  63. Note that Storto (1999) abandons this view of Karitiâna from her previous work, with absolutive arguments moving to Spec,TP. I believe the facts discussed in Storto (1999) which motivated this theoretical change can also be accounted for under the logic of anti-agreement introduced here. I will however not present a reanalysis of the full Karitiâna facts here, as it is far beyond the scope of this paper.

  64. This is the “tentative” conclusion of Cinque (1999:133ff): “though attractive, I think that [the alternative where not all projections are always present] is more costly than the idea that functional notions are always all structurally represented.”


  • Abels, Klaus. 2003. Successive cyclicity, anti-locality, and adposition stranding, Doctoral dissertation, University of Connecticut.

  • Adger, David. 2003. Core syntax: a minimalist approach. London: Oxford University Press.

    Google Scholar 

  • Aissen, Judith. 1992. Topic and focus in Mayan. Language 68: 43–80.

    Article  Google Scholar 

  • Aissen, Judith. 1999a. Agent focus and inverse in Tzotzil. Language 75: 451–485.

    Article  Google Scholar 

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

    Article  Google Scholar 

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

    Article  Google Scholar 

  • Aldridge, Edith. 2008. Generative approaches to ergativity. Language and Linguistics Compass 2: 966–995.

    Article  Google Scholar 

  • Aldridge, Edith. 2012. Two types of ergativity and where they might come from. Presentation at Massachusetts Institute of Technology.

  • Aldridge, Edith Catherine. 2004. Ergativity and word order in Austronesian languages, Doctoral dissertation, Cornell University.

  • Alexiadou, Artemis, and Elena Anagnostopoulou. 2001. The subject in situ generalization and the role of Case in driving computations. Linguistic Inquiry 32: 193–231.

    Article  Google Scholar 

  • Alexiadou, Artemis, and Elena Anagnostopoulou. 2007. The subject-in-situ generalization revisited. In Interfaces + recursion = language? Chomsky’s minimalism and the view from syntax-semantics, eds. Uli Sauerland and Hans-Martin Gärtner. Vol. 89 of Studies in generative grammar, 31–60. Berlin: de Gruyter.

    Google Scholar 

  • Assmann, Anke, Doreen Georgi, Fabian Heck, Gereon Müller, and Philipp Weisser. 2013. Ergatives move too early: on an instance of opacity in syntax. In Rule interaction in grammar, eds. Fabian Heck and Anke Assmann. Vol. 90 of Linguistische Arbeits Berichte, 363–412. Leipzig: Institut für Linguistik, Universität Leipzig.

    Google Scholar 

  • Ayres, Glenn. 1977. ¿Es el antipasivo siempre una voz? Paper presented at the Second Mayan Workshop, San Cristobal de las Casas.

  • Bittner, Maria, and Kenneth Hale. 1996a. Ergativity: towards a theory of a heterogeneous class. Linguistic Inquiry 27: 531–604.

    Google Scholar 

  • Bittner, Maria, and Kenneth Hale. 1996b. The structural determination of case and agreement. Linguistic Inquiry 27: 1–68.

    Google Scholar 

  • Bošković, Željko. 1994. D-structure, θ-theory, and movement into θ-positions. Linguistic Analysis 24: 247–286.

    Google Scholar 

  • Bošković, Željko. 1997. The syntax of nonfinite complementation: an economy approach. Cambridge: MIT Press.

    Google Scholar 

  • Bošković, Željko. 2003. Agree, phases, and intervention effects. Linguistic Analysis 33: 54–96.

    Google Scholar 

  • Bošković, Željko. 2007. On the locality and motivation of Move and Agree: an even more minimal theory. Linguistic Inquiry 38: 589–644.

    Article  Google Scholar 

  • Bošković, Željko. 2014. Now I’m a phase, now I’m not a phase: on the variability of phases with extraction and ellipsis. Linguistic Inquiry 45: 27–89.

    Article  Google Scholar 

  • Brandi, Luciana, and Patrizia Cordin. 1989. Two Italian dialects and the null subject parameter. In The null subject parameter, eds. Osvaldo A. Jaeggli and Kenneth J. Safir. Vol. 15 of Studies in natural language and linguistic theory, 111–142. Berlin: Springer.

    Chapter  Google Scholar 

  • Bresnan, Joan. 1977. Variables in the theory of transformations. In Formal syntax, eds. Peter Culicover, Thomas Wasow, and Adrien Akmajian, 157–196. San Diego: Academic Press.

    Google Scholar 

  • Broadwell, George Aaron. 2000. Word order and markedness in Kaqchikel. In Lexical Functional Grammar (LFG00), eds. Miriam Butt and Tracy Holloway King, 44–62. Stanford: CSLI.

    Google Scholar 

  • Broadwell, George Aaron. 2007. Kaqchikel word order: preliminary observations. Manuscript, SUNY Albany.

  • Brown, R. McKenna, Judith M. Maxwell, and Walter E. Little. 2006. La ütz awäch?: introduction to Kaqchikel Maya language. Austin: University of Texas Press.

    Google Scholar 

  • Campbell, Lyle. 2000. Valency-changing derivations in K’iche’. In Changing valency: case studies in transitivity, eds. Robert M. W. Dixon, and Alexandra Y. Aikhenvald, 236–281. Cambridge: Cambridge University Press.

    Chapter  Google Scholar 

  • Campos, Héctor. 1997. On subject extraction and the antiagreement effect in Romance. Linguistic Inquiry 28: 92–119.

    Google Scholar 

  • Carstens, Vicki. 1985. Wh-movement in Yoruba. Studies in African Linguistics 9: 39–44.

    Google Scholar 

  • Carstens, Vicki. 1987. Extraction asymmetries in Yoruba. In Current approaches to African linguistics, ed. David Odden. Vol. 4, 61–72. Berlin: De Gruyter.

    Google Scholar 

  • Chomsky, Noam. 1991. Some notes on economy of derivation and representation. In Principles and parameters in comparative grammar, ed. Robert Freidin, 417–454. Cambridge: MIT Press.

    Google Scholar 

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

    Google Scholar 

  • Cinque, Guglielmo. 1999. Adverbs and functional heads. London: Oxford University Press.

    Google Scholar 

  • Clemens, Lauren Eby. 2013. Kaqchikel SVO: V2 in a V1 language. In Studies in Kaqchikel grammar, ed. Michael Kenstowicz. Vol. 8 of MIT working papers on endangered and less familiar languages, 1–24. Cambridge: MIT Working Papers in Linguistics.

    Google Scholar 

  • Cole, Peter, and Gabriella Hermon. 1981. Subjecthood and islandhood: evidence from Quechua. Linguistic Inquiry 12: 1–30.

    Google Scholar 

  • Coon, Jessica, Pedro Mateo Pedro, and Omer Preminger. 2014. The role of case in A-bar extraction asymmetries: evidence from Mayan. Linguistic Variation 14: 179–242.

    Article  Google Scholar 

  • Craig, Colette Grinevald. 1979. The antipassive and Jacaltec. In Papers in Mayan linguistics, ed. Laura Martin. Vol. 7, 139–164. Lucas Brothers.

    Google Scholar 

  • Culicover, Peter W. 1993. Evidence against ECP accounts of the that-t effect. Linguistic Inquiry 24: 557–561.

    Google Scholar 

  • Dayley, Jon P. 1981. Voice and ergativity in Mayan languages. Journal of Mayan Linguistics 2: 6–82.

    Google Scholar 

  • Dayley, Jon P. 1985. Tzutujil grammar. Berkeley: University of California Press.

    Google Scholar 

  • Duncan, Lachlan. 2003. The syntactic structure of Tz’utujil Maya. In Lexical Functional Grammar (LFG03), eds. Miriam Butt and Tracy Holloway King, 164–183. Stanford: CSLI.

    Google Scholar 

  • England, Nora. 1991. Changes in basic word order in Mayan languages. International Journal of American Linguistics 57: 446–486.

    Article  Google Scholar 

  • Erlewine, Michael Yoshitaka. 2014a. Anti-locality and Kaqchikel Agent Focus. In West Coast Conference on Formal Linguistics (WCCFL) 31, ed. Robert E. Santana-LaBarge, 150–159.

    Google Scholar 

  • Erlewine, Michael Yoshitaka. 2014b. Why the null complementizer is special in the English that-trace effect. Manuscript, MIT.

  • Fanselow, Gisbert, and Damir Ćavar. 2001. Remarks on the economy of pronunciation. In Competition in syntax, Vol. 49 of Studies in generative grammar, 107–150. Berlin: de Gruyter.

    Google Scholar 

  • Fox, Danny. 1995. Economy and scope. Natural Language Semantics 3: 283–341.

    Article  Google Scholar 

  • García Matzar, Pedro, and José Obispo Rodríguez Guaján. 1997. Rukemik ri Kaqchikel Chi’: gramática Kaqchikel. Guatemala City: Cholsamaj.

    Google Scholar 

  • Grimshaw, Jane. 1997. Projection, heads, and optimality. Linguistic Inquiry 28: 373–422.

    Google Scholar 

  • Grohmann, Kleanthes K. 2003. Prolific domains: on the anti-locality of movement dependencies. Amsterdam: Benjamins.

    Book  Google Scholar 

  • Hale, Kenneth, and Luciana Raccanello Storto. 1996. Agreement and spurious antipassives. Manuscript, MIT.

  • Heck, Fabian, and Gereon Müller. 2001. Successive cyclicity, long-distance superiority, and local optimization. In West Coast Conference on Formal Linguistics (WCCFL) 19, eds. Roger Billerey and Brook Danielle Lillehaugen, 101–114.

    Google Scholar 

  • Hedberg, Nancy. 1988. Discourse function, ergativity, and agreement in Cakchiquel Mayan. Manuscript, University of Minnesota.

  • Henderson, Robert, Jessica Coon, and Lisa Travis. 2013. Micro- and macro-parameters in Mayan syntactic ergativity. Presentation at Towards a theory of syntactic variation, Bilbao.

  • Hermon, Gabriella. 1984. Syntactic modularity. Dordrecht: Foris.

    Google Scholar 

  • de Hoop, Helen, and Andrej L. Malchukov. 2008. Case-marking strategies. Linguistic Inquiry 39: 565–587.

    Article  Google Scholar 

  • Kenstowicz, Michael J. 2013. Realize morpheme in Kaqchikel. In Studies in Kaqchikel grammar, ed. Michael Kenstowicz. Vol. 8 of MIT working papers on endangered and less familiar languages, 67–80. Cambridge: MIT Working Papers in Linguistics.

    Google Scholar 

  • Koopman, Hilda. 1982. Control from COMP and comparative syntax. The Linguistic Review 2: 365–391.

    Article  Google Scholar 

  • Koopman, Hilda. 1984. The syntax of verbs: from verb movement rules in the Kru languages to universal grammar. Vol. 15 of Studies in generative grammar. Dordrecht: Foris.

    Google Scholar 

  • Larsen, Tomas W., and William M. Norman. 1979. Correlates of ergativity in Mayan grammar. In Ergativity: towards a theory of grammatical relations, ed. Frans Plank, 347–370. San Diego: Academic Press.

    Google Scholar 

  • Lawal, Nike. 1987. Yoruba relativisation and the continuous segment principle. Studies in African Linguistics 18: 67–80.

    Google Scholar 

  • Lee, Jaecheol. 2003. Phase sensitivity in wh-dependencies. Korean Journal of Linguistics 28: 67–89.

    Google Scholar 

  • Marantz, Alec. 1991. Case and licensing. In Eastern States Conference on Linguistics (ESCOL) 8, eds. Germán F. Westphal, Benjamin Ao, and Hee-Rahk Chae, 234–253.

    Google Scholar 

  • Mondloch, James Lorin. 1981. Voice in Quiché-Maya. Doctoral dissertation, State University of New York at Albany.

  • Murasugi, Keiko, and Mamoru Saito. 1995. Adjunction and cyclicity. In West Coast Conference on Formal Linguistics (WCCFL) 13, eds. Raul Aranovich, William Byrne, Susanne Preuss, and Martha Senturia, 302–317.

    Google Scholar 

  • Müller, Gereon. 1997. Partial wh-movement and Optimality Theory. The Linguistic Review 14: 249–306.

    Article  Google Scholar 

  • Müller, Gereon, and Wolfgang Sternefeld. 2001. Competition in syntax. Vol. 49 of Studies in generative grammar. Berlin: de Gruyter.

    Book  Google Scholar 

  • Norcliffe, Elisabeth. 2009. Head-marking in usage and grammar: a study of variation and change in Yucatec Maya, Doctoral dissertation, Stanford University.

  • Ordóñez, Francisco. 1995. The antipassive in Jacaltec: a last resort strategy. In Catalan working papers in linguistics, Vol. 4, 329–343.

    Google Scholar 

  • Ouhalla, Jamal. 1993. Subject-extraction, negation, and the anti-agreement effect. Natural Language and Linguistic Theory 11: 477–518.

    Article  Google Scholar 

  • Pesetsky, David, and Esther Torrego. 2001. T-to-C movement: causes and consequences. In Ken Hale: a life in language, ed. Michael Kenstowicz, 355–426. Cambridge: MIT Press.

    Google Scholar 

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

  • Preminger, Omer. 2014. Agreement and its failures. Cambridge: MIT Press.

    Book  Google Scholar 

  • Prince, Alan S., and Paul Smolensky. 1993. Optimality theory: constraint interaction in generative grammar. Manuscript, Rutgers University and University of Colorado.

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

    Article  Google Scholar 

  • Richards, Norvin. 2001. Movement in language: interactions and architectures. London: Oxford University Press.

    Google Scholar 

  • Richards, Norvin. 2010. Uttering trees. Cambridge: MIT Press.

    Book  Google Scholar 

  • Richards, Norvin Waldemar, III. 1997. What moves where when in which language? Doctoral dissertation, Massachusetts Institute of Technology.

  • Rizzi, Luigi. 1982. Issues in Italian syntax. Dordrecht: Foris.

    Book  Google Scholar 

  • Rizzi, Luigi. 1997. The fine structure of the left periphery. In Elements of grammar, ed. Liliane Haegeman, 281–337. Dordrecht: Kluwer Academic.

    Chapter  Google Scholar 

  • Rizzi, Luigi, and Ur Shlonsky. 2007. Strategies of subject extraction. In Interfaces + recursion = language? Chomsky’s minimalism and the view from syntax-semantics, eds. Uli Sauerland and Hans-Martin de Gärtner. Vol. 89 of Studies in Generative Grammar, 115–160. Berlin: de Gruyter.

    Google Scholar 

  • Robertson, John S. 1980. The structure of pronoun incorporation in the Mayan verbal complex. New York: Garland.

    Google Scholar 

  • Rouveret, Alain. 2002. How are resumptive pronouns linked to the periphery? Linguistic Variation Yearbook 2: 123–184.

    Article  Google Scholar 

  • Saito, Mamoru, and Keiko Murasugi. 1999. Subject predication within IP and DP. In Beyond principles and parameters: essays in memory of Osvaldo Jaeggli, eds. Kyle Johnson and Ian Roberts. Vol. 45 of Studies in natural language and linguistic theory, 167–188. Berlin: Springer.

    Chapter  Google Scholar 

  • Schneider-Zioga, Patricia. 2007. Anti-agreement, anti-locality, and minimality. Natural Language and Linguistic Theory 25: 403–446.

    Article  Google Scholar 

  • Smith-Stark, Thomas. 1978. The Mayan antipassive: some facts and fictions. In Papers in Mayan linguistics, ed. Nora C. England, 169–187.

    Google Scholar 

  • Sonaiya, Remi. 1989. Wh-movement and proper government in Yoruba. In Current approaches to African linguistics, Vol. 5, 109–126. Dordrecht: Foris.

    Google Scholar 

  • Starke, Michal. 2001. Move dissolves into Merge: a theory of locality. Doctoral dissertation.

  • Stiebels, Barbara. 2006. Agent focus in Mayan languages. Natural Language and Linguistic Theory 24: 501–570.

    Article  Google Scholar 

  • Stjepanović, Sandra, and Shoichi Takahashi. 2001. Eliminating the phase impenetrability condition. Paper presented at the Motivating Movement conference, University of Ulster, Jordanstown (January 26–28).

  • Storto, Luciana Raccanello. 1997. Verb raising and word order variation in Karitiana. In Boletim da Associacao Brasileirade Linguistica (ABRALIN), Vol. 20.

    Google Scholar 

  • Storto, Luciana Raccanello. 1998. Karitiana: a verb-second language from Amazonia. In Conference of Students of Linguistics of Europe (CONSOLE) 6, ed. Tina Cambier-Langeveld.

    Google Scholar 

  • Storto, Luciana Raccanello. 1999. Aspects of a Karitiana grammar. Doctoral dissertation, Massachusetts Institute of Technology.

  • Storto, Luciana Raccanello. 2012. Information structure in Karitiana. In Conference on the Indigenous Languages of the Americas (CILLA) 5.

    Google Scholar 

  • Tada, Hiroaki. 1993. A/A-bar partition in derivation. Doctoral dissertation, Massachusetts Institute of Technology.

  • Tonhauser, Judith. 2003. F-constructions in Yucatec Maya. In Semantics of Under-Represented Languages in the Americas (SULA) 2, eds. Jan Anderssen, Paula Menéndez-Benito, and Adam Werle, 203–223. Amherst: GLSA.

    Google Scholar 

  • Watanabe, Akira. 1992. Larsonian CP recursion, factive complements, and selection. In North East Linguistic Society (NELS) 23, 523–537.

    Google Scholar 

  • Woolford, Ellen. 2003. Clitics and agreement in competition: ergative cross-referencing patterns. In Papers in Optimality Theory II, eds. Angela C. Carpenter, Andries W. Coetzee, and Paul de Lacy. Vol. 26 of University of Massachusetts occasional papers, 421–449. Amherst: GLSA.

    Google Scholar 

  • Woolford, Ellen. 2015. Ergativity and transitivity. Linguistic Inquiry 46: 489–531.

    Article  Google Scholar 

  • Zavala, Roberto. 1997. Functional analysis of Akatek voice constructions. International Journal of American Linguistics 63: 439–474.

    Article  Google Scholar 

Download references


This work would not have been possible without the patience and kindness of my consultant Ana López de Mateo. For comments and discussion, I thank Judith Aissen, Lauren Eby Clemens, Jessica Coon, Robert Henderson, Yusuke Imanishi, Hadas Kotek, Pedro Mateo Pedro, Junya Nomura, David Pesetsky, Omer Preminger, Norvin Richards, Daeyoung Sohn, Coppe van Urk, Ellen Woolford, some very helpful but unfortunately anonymous reviewers, editor Marcel den Dikken, and audiences at Tohoku University, National Tsing Hua University, the University of Toronto, CUNY Graduate Center, MIT, the 2012 American International Morphology Meeting, LSA 2013, and WCCFL 31. Portions of this work were supported by a Ken Hale fieldwork grant at MIT. A much shorter presentation of parts of Sects. 3 and 4 appeared in the Proceedings of WCCFL 31 as Erlewine (2014a). All errors are mine.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Michael Yoshitaka Erlewine.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Erlewine, M.Y. Anti-locality and optimality in Kaqchikel Agent Focus. Nat Lang Linguist Theory 34, 429–479 (2016).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: