Skip to main content

Spans in South Caucasian agreement

Revisiting the pieces of inflection


I argue that a range of morphological phenomena sensitive to features of multiple arguments in Georgian (South Caucasian)—including Anti-Superiority ef- fects (Béjar 2003), and omnivorous number effects (Nevins 2011) – receive a unified account if spellout targets contiguous spans of maximally simple heads, in a fixed hierarchy. I introduce new data from a related language, Laz, and show that a close comparison of the two languages reveals that (i) number agreement is expressed om- nivorously only if the prefix is not sensitive to number, and that (ii) this number expression covaries with Tense only if the subject is third person. I argue that both Anti-Superiority and the facts about number expression should be interpreted as fu- sional morphology being limited to third person contexts, and that a principled expla- nation for such an asymmetry can be provided, if first and second person structurally contain third person, and the matching of exponents with syntactic structure is gov- erned by Overspecification (Starke 2009), such that a lexicalized span is a candidate of spellout for its sub-spans.

This is a preview of subscription content, access via your institution.


  1. 1.

    I use the notation X>Y for a transitive agreement context (such that X denotes the phi-features of the subject and Y denotes those of the object), as well as syntactic selection/sisterhood.

  2. 2.

    See also the collection of papers in Baunaz et al. (2018) for more recent work in this framework, Williams (2003) for a precursor that introduced a related notion of spanning, as well as Mirror Theory (Brody 2000; Adger et al. 2009) for related ideas.

  3. 3.

    That is, modulo notions such as zero affixes, zero heads/operators, etc.

  4. 4.

    As a notational convention, I will use the usual bracketing structure for spans throughout this paper, avoiding the n-tuple notation, <…>, despite the fact that a span is not usually/necessarily a constituent.

  5. 5.

    The original purpose of Pointers is related to certain types of idioms (Michal Starke, p.c.), but the idea to employ it for the present purposes goes back to Pantcheva and Caha (2012).

  6. 6.

    Since their proposal has not been published, I will give an extremely short recapitulation here: Blansitt’s (1988) generalization states that if a dative can be used for locative purposes in a language, then it can also be used for allative purposes. Caha and Pantcheva (2012) propose, based on a crosslinguistic study that these four cases correspond to the structures in (i).

    1. (i)

    Under the Pointer approach, the loc=dat=all syncretism arises when the case affix lexicalizes [dat[gen → [p]]]. Crucially, Blansitt’s generalization falls out from this, since matching both the dative structure and the locative structure is only possible if the vi also lexicalizes the allative structure without ‘unused’ heads. Unattested syncretisms like loc=gen=datall, or loc=datall, on the other hand, cannot be derived under this approach, i.e., in contrast to a subset-based approach to syncretisms that allows for free cross-classification of two independent systems, the Pointer approach is restrictive in a predictive fashion.

  7. 7.

    Note that the Pointer approach is limited in the kinds of cross-categorial syncretisms it allows for: If there is a cross-person syncretism in the plural that forms an L-shape with one of the singular forms, it has to be the largest one, since a hypothetical item /α/ ⇔ ‘[pl → [3]]’ could not apply in any person structure that is larger than the third person. The current system is thus more restrictive than a subset-based approach that allows independent cross-selection of the two systems.

  8. 8.

    On the Pazar dialect in particular, see also Öztürk and Pöchtrager (2011). Note that it contains a small number of errors with respect to the agreement accessibility of arguments, corrected in Demirok (2013).

  9. 9.

    The paradigms lack the cells that would correspond to first and second person reflexive forms. There are two ways to express reflexivity in Laz, either with reflexive pronouns, which uniformly trigger third person agreement, or with valency changing verbal morphology that makes the verb intransitive. First and second person can therefore never simultaneously trigger subject agreement and object agreement within a single verb form, i.e. the lacking cells are not a part of the language.

  10. 10.

    Note that the present stem is derived with a thematic suffix that is sensitive to argument structure, thematic roles, and lexical aspect (cf. Öztürk and Pöchtrager 2011; Öztürk and Taylan Erguvanlı 2017). The corresponding set of suffixes in Georgian have been dubbed present/future stem formants (cf. Aronson 1990: 40 and Harris 1982). Given that these suffixes are sensitive to argument structure, I assume that these spell out of Voice heads, contextually conditioned by linearly adjacent Tense in the spirit of Embick (2015); Kastner (2018), but due to their invariance with respect to agreement, I abstract away from them here.

  11. 11.

    Conversely, if an argument does not agree for person, it is also excluded from agreeing for number, as discussed by Demirok (2013: 79): In ditransitive constructions, it is the indirect object that triggers object agreement, and the direct object can mark neither number nor person:

    1. (i)
  12. 12.

    It should be noted that both languages show a second type of transitive agreement paradigm called Inversion. In these paradigms a dative subject triggers the kind of prefixal agreement usually found with objects. These are much more divergent between the two languages (cf. Öztürk and Pöchtrager 2011: 60ff and Aronson 1990: sections 10.1, 12.1), and I follow various other authors in excluding these here.

  13. 13.

    The system I propose below can be implemented in either variant. However, since an auth selecting implementation requires an additional zero affix, the add selecting implementation is slightly more elegant. The relevant notions for the analysis, however, are that there is a containment relation between participants, and that participant agreement properly contains the structure of third person agreement, i.e., independent of this choice. However, since both 1=3≠2 syncretisms and 2=3≠1 ones exist crosslinguistically (Cysouw 2003), this is one way for the present system to accommodate these facts.

  14. 14.

    The structure raises obvious questions as to its nature and the way it is constructed. One might interpret this structure either as the result of successive cyclic head movement of a set of person/number agreement heads (along the lines of Preminger 2011), or possibly Multiple Agree (Hiraiwa 2005) with the resulting structure reflecting relative heights. Under such a view, the target of spellout would be spans within a complex head, formed by the syntax, and a probe would be an instruction to the syntax to build such a structure under a matching requirement that pertains to a specific syntactic configuration, such as c-command. Alternatively, the agreement structure might in fact be part of the extended projection of the verb, possibly heads that provide phi values to initially unvalued pronominal elements, along the lines of Kratzer (2009), or Stegovec (2019), which might imply that agreement is somewhat reminiscent of Sportiche (2005, 2006) style determiners in the extended projection of the verb. The subpart of the analysis in the current section is agnostic about this question, as long as the relevant structural containment relations hold, i.e., as long as a specific theory of agreement is compatible with the syntax building the kind of structure given in (26), either view is compatible with the results derived here. In Section 5, I provide some arguments in favor of the latter hypothesis, based on morpheme order; the dependency, however, is asymmetric: While the account of linear order will crucially rely on phrasal movement, and therefore on these heads being part of phrasal syntax, the account of their paradigmatic distribution is independent of such an interpretation. I will largely leave the larger questions that this raises for the nature of Agree or agreement untouched, for now, and hope that future research may shed more light on these questions.

  15. 15.

    Phonologically null affixes are obviously motivated primarily on theory-internal grounds. Georgian has an overt counterpart to Ø- in the copula’s present Tense paradigm (Aronson 1990: 66), thus providing independent evidence for such an affix.

    1. (i)

    Note also that there is a rather curious prediction of this theory: This affix is able to spread into the 1sg>2sg cell, spelling out the subject features, although this could easily be avoided by postulating a second zero affix. I hope that these kinds of unexpected distributions will turn out to be useful in analyzing other complex agreement systems, but as it stands this is an unusual possibility predicted by the system.

  16. 16.

    Note that these affixes embed a structure [plo → [refo]], despite my claim that a configuration [plo [refo]] does not arise in transitive agreement. In the Laz inverse paradigms (discussed briefly in fn. 12), however, third person dative subjects do trigger plural object agreement. Whatever their precise syntax, their agreement morphology can therefore be captured in this system.

  17. 17.

    An interesting alternative to the zero affix would be to argue that the past tense properly contains the present tense, and to specify the relevant affixes for the local subjects for prs, but not pst, but the ones for third person for both. Under this perspective, -i would spell out only [pst], and we would not need another zero affix.

  18. 18.

    Note that in order for -en to lose out to -t due to the ep, we need to assume that -t is less specific than -en. I tacitly assume that this is due to the internal structure of the placeholder prs.

  19. 19.

    An interesting point about the occurrence of -t was brought to my attention by Thomas Wier (p.c.). Third person plural objects can exceptionally trigger plural agreement if they are focused and the subject is inanimate, as shown in (i). Note that a focused inanimate plural object appears to block a non-focused inanimate subject that would normally trigger plural agreement “optionally” from doing so (regardless of the object triggering plural agreement). In contrast, a human third person plural subject triggers -en, blocking -t (Léa Nash, p.c.).

    1. (i)

    As with second person plural objects, we see that -s disappears, and -t occurs, when there is exceptional plural agreement with third person objects. From the current perspective, it is expected that [refs[plo]] receive spellout by -t, but it raises the question of what spells out the object’s person feature. They need to be spelled out independently, in order to block -en from being anchored before -t can be. At the present moment I can only give an ad-hoc stipulation, suggesting that the system can be made to work by employing a zero affix that spells out a structure along the lines [foc[refo]], and that the element foc bleeds -en (cf. a similar ad hoc solution in Halle and Marantz (1993): fn. 6, which suggests that third person arguments can occasionally pattern with participants due to an additional feature).

  20. 20.

    In fact, some of the third person plural subject affixes appear to contain the TAM marker as well, e.g. the optative affixes -o, -os and -on. I will tentatively suggest that in these cases the -n spreads minimally into the TAM domain, but does not spell out all of it. If the vocabulary item -o contains a pointer to the same TAM structure -n spells out, below the remainder of the TAM domain, we derive the desired result. As this is not an investigation into the Georgian TAM structure, I will leave the precise formulation of this to future research.

  21. 21.

    Note that this variation is closely linked to another point of variation, namely the Pointer in Laz -an, and its respective absence in Georgian -en: It is only because -en cannot spell out the tense structure unless it also spells out pls that allows a tenseless -s to surface. Were the Georgian -en like the Laz -an, it would always overwrite a tenseless -s, even in the absence of any plurality. The tenseless -s, i.e., the -s that always co-occurs with the kind of tense marker that is limited to local subjects in Laz, is only possible with the pointerless -en, and the resulting conditional omnivorous number effect.

  22. 22.

    Presumably, -di itself is bimorphemic, with -d spelling out some higher structure. I once again abstract away from the details of TAM.

  23. 23.

    Note that the selection of the set of suffixes appears to depend on conjugation class, the phonological form of the stem, or irregular verbs. This does not seem to lend itself to an analysis in terms of spans, as advanced here. Whether an extension of an allomorphy approach (a notion that most Nanosyntax eschews), is a plausible way of accounting for these further details remains to be seen.

  24. 24.

    As an anonymous reviewer pointed out, the argument extends to material that is lower than the agreement structure, i.e., VoiceP-internal material: Given the claim in (69) that movement of VoiceP is the relevant operation, stranding of VoiceP-internal material via successive cyclic movement of a smaller phrase is ruled out. We thus predict two positions for VoiceP-internal material, in between the prefix and the stem (no movement/pied-piping), and after the stem but before the suffixes (snowball movement). As the reviewer points out, both positions are attested in, a.o., Laz causative morphology (their example):

    1. (i)
  25. 25.

    The preverbs serve a variety of functions beyond the future marking, such as spatial relations, or perfectivity. They are also present in the past Tense (aorist). Insofar as their function of future marking is concerned, they are “usually unpredictable” (Hewitt 1995: 145-169), i.e., specific to the verb, but selected from a small class. See ibid., 148-169 for details. Note also that certain verbs do always come with their respective preverb, and do not distinguish the present from the future, i.e., they exhibit a syncretism (Aronson 1990: 42ff, 61f).

  26. 26.

    A second caveat: Laz shows a set of thematic suffixes in certain configurations that denote properties of the event and argument structure. These are obligatory in the present tense Öztürk and Pöchtrager (2011, 69), but absent in the subjunctive forms, and again it is not a priori clear what the right analysis of this is under the present account. See Öztürk and Taylan Erguvanlı (2017) for a detailed analysis of the argument structure, event type and aspectual functions of the thematic suffixes.

  27. 27.

    See Svenonius (2016) for an alternative to evacuation that relies on specifier-complement asymmetries for the definition of a span, which might provide an alternative to evacuation analyses, and might offer a direction of research, under which the derivational timing as well as the construction of phonological words from syntactic structure might be understood in terms of PF-instructions relativized to heads.

  28. 28.

    For a general overview of DM see Harley and Noyer (1999), Embick and Noyer (2007).

  29. 29.

    Trommer’s approach is thus quite similar in spirit to the current one, insofar as it is an attempt at reduction to vocabulary insertion. For Trommer, Impoverishment reduces to insertion of a zero vi, and Fusion itself should also be eliminated from the system, and replaced with mutually conditioned contextual allomorphy. That is to say, rather than fusing two nodes, say the object’s [+pl] feature, and the Tns-Agr node, we would postulate two contextual allomorphy rules: [+pl] is spelled out as zero in the context of a third person subject, and a third person subject Tns-Agr is spelled out as -an in the context of a [+pl] node. Since we find ourselves in a system with omnivorous number, this might necessitate two rules for -an, one for [+pl] originating in the Tns-Agr node, and one for the contextual allomorphy, thus creating even more redundancy. It still fails to account for the person asymmetry.

  30. 30.

    As they note on page 45 this means divorcing the feature valuation from the resulting morphological expression, as a probe specified for only [] (German, Romance etc) can clearly result in morphological expression of the full feature structure.

  31. 31.

    Note that their approach thus breaks with the DM notion of spellout targeting terminals as well, while building a similar notion into the system as I did: v- is essentially the spellout of first person in the context of a third person probing.

  32. 32.

    The same point applies to Nevins (2011), which derives omnivorous number from Multiple Agree. For Nevins’s approach, the prefixes are clitic arguments that can be the targets of Multiple Agree; it is unclear why Multiple Agree would be blocked exactly in case the clitic itself has vocabulary items that make reference to the number distinction. Under said approach we would expect the Georgian first person objects to trigger omnivorous number the same way as second person subjects/objects and first person subjects do, and the proposal does not seem to offer a clear way of distinguishing these two cases. In addition, Nevins (2011: 962) also seems to suggest that Omnivorous Number and co-variance with Tense are in complementary distribution, but Laz clearly provides a counterexample to such a generalization.


  1. Adger, David, Daniel Harbour, and Laurel J. Watkins. 2009. Mirrors and microparameters: Phrase structure beyond free word order Cambridge studies in linguistics. Cambridge: Cambridge University Press.

    Book  Google Scholar 

  2. Aissen, Judith. 1999. Markedness and subject choice in optimality theory. Natural Language and Linguistic Theory 17: 673–711.

    Article  Google Scholar 

  3. Aissen, Judith. 2003. Differential object marking: Iconicity vs. economy. Natural Language and Linguistic Theory 21(3): 435–483.

    Article  Google Scholar 

  4. Anderson, Stephen R. 1992. A-morphous morphology. Vol. 62 of Cambridge studies in linguistics. Cambridge: Cambridge University Press.

    Book  Google Scholar 

  5. Aronson, Howard I. 1990. Georgian: A reading grammar. Corr. edn. Columbus: Slavica Publ.

    Google Scholar 

  6. Baunaz, Lena, Karen De Clercq, Liliane M. V. Haegeman, and Eric Lander, eds. 2018. Exploring nanosyntax. Oxford studies in comparative syntax, New York: Oxford University Press.

    Book  Google Scholar 

  7. Béjar, Susana. 2003. Phi-syntax: A theory of agreement. PhD diss., University of Toronto.

  8. Béjar, Susana, and Milan Rezac. 2009. Cyclic agree. Linguistic Inquiry 40(1): 35–73.

    Article  Google Scholar 

  9. Blansitt, Edward. 1988. Datives and allatives. In Studies in syntactic typology, eds. Michael Hammond, Edith A. Moravcsik, and Jessica Wirth. Vol. 17 of Typological studies in language, 173–191. Amsterdam: John Benjamins Publishing Company.

    Chapter  Google Scholar 

  10. Jonathan David Bobaljik. 2012. Universals in comparative morphology: Suppletion, superlatives, and the structure of words Current studies in linguistics. Cambridge: MIT Press.

    Book  Google Scholar 

  11. Brody, Michael. 2000. Mirror theory: Syntactic representation in perfect syntax. Linguistic Inquiry 31(1): 29–56.

    Article  Google Scholar 

  12. Caha, Pavel. 2009. The nanosyntax of case. PhD diss., University of Tromsø, Tromsø.

  13. Caha, Pavel, and Marina Pantcheva. 2012. Datives Cross-Linguistically.

  14. Chomsky, Noam. 1995. The minimalist program, Vol. 28. Cambridge: Cambridge University Press.

    Google Scholar 

  15. Chomsky, Noam. 2001. Derivation by phase. In Ken Hale, ed. Michael J. Kenstowicz, 1–52. Cambridge: MIT Press.

    Google Scholar 

  16. Cinque, Guglielmo. 2005. Deriving Greenberg’s universal 20 and its exceptions. Linguistic Inquiry 36(3): 315–332.

    Article  Google Scholar 

  17. Corbett, Greville G. 2000. Number. Cambridge: Cambridge University Press.

    Book  Google Scholar 

  18. Cysouw, Michael. 2003. The paradigmatic structure of person marking Oxford studies in typology and linguistic theory. Oxford: Oxford University Press.

    Google Scholar 

  19. Demirok, Ömer. 2013. Agree as a unidirectional operation: Evidence from Laz. MA thesis, Boğaziçi University, Istanbul.

  20. Embick, David. 2015. The morpheme: A theoretical introduction. Vol. 31 of Interface explorations. Boston: De Gruyter Mouton.

    Book  Google Scholar 

  21. Embick, David, and Rolf Noyer. 2007. Distributed morphology and the syntax-morphology interface. In The Oxford handbook of linguistic interfaces, eds. Gillian Ramchand and Charles Reiss, 289–324. Oxford handbooks in linguistics. Oxford: Oxford University Press.

    Google Scholar 

  22. Foley, Steven. 2017. Morphological conspiracies in Georgian and optimal vocabulary insertion. In 52nd Chicago Linguistics Society (CLS) meeting, eds. Jessica Kantarovich, Tran Truong, and Orest Xherija.

    Google Scholar 

  23. Halle, Morris, and Alec Marantz. 1993. Distributed morphology and the pieces of inflection. In The view from building 20, eds. Samuel Jay Keyser and Kenneth Hale, Current studies in linguistics, 111–176. Cambridge: MIT Press.

    Google Scholar 

  24. Harley, Heidi, and Rolf Noyer. 1999. State-of-the-article: Distributed morphology. GLOT 4(4): 3–9.

    Google Scholar 

  25. Harley, Heidi, and Elizabeth Ritter. 2002a. Person and number in pronouns: A feature-geometric analysis. Language 78(3): 482–526.

    Article  Google Scholar 

  26. Harley, Heidi, and Elizabeth Ritter. 2002b. Structuring the bundle: A universal morphosyntactic feature geometry. In Pronouns, eds. Horst J. Simon and Heike Wiese. Amsterdam: Benjamins.

    Google Scholar 

  27. Harris, Alice C. 1982. Georgian and the unaccusative hypothesis. Language 58(2): 290.

    Article  Google Scholar 

  28. Brian George Hewitt. 1995. Georgian: A structural reference grammar. Amsterdam: Benjamins.

    Book  Google Scholar 

  29. Hiraiwa, Ken. 2005. Dimensions of symmetry in syntax: Agreement and clausal architecture. PhD diss., Massachusetts Institute of Technology.

  30. Kastner, Itamar. 2018. Templatic morphology as an emergent property: Roots and functional heads in Hebrew. Natural Language and Linguistic Theory

    Article  Google Scholar 

  31. Kayne, Richard S. 1994. The antisymmetry of syntax. Vol. 25 of Linguistic inquiry monographs. Cambridge: MIT Press.

    Google Scholar 

  32. Kayne, Richard S. 2017. Antisymmetry and morphology: Prefixes vs. suffixes. Wiener Linguistische Gazette 82. (Festschrift für Martin Prinzhorn).

  33. Keine, Stefan. 2010. Case and agreement from fringe to core: A minimalist approach. Berlin: de Gruyter.

    Book  Google Scholar 

  34. Kiparsky, Paul. 1973. “Elsewhere” in Phonology. In A Festschrift for Morris Halle, eds. Morris Halle, Stephen R. Anderson, and Paul Kiparsky, 93–106. New York: Holt, Rinehart and Winston.

    Google Scholar 

  35. Kiparsky, Paul. 2008. Universals constrain change; change results in typological generalizations. In Linguistic universals and language change, ed. Jeff Good, 23–53. London: Oxford University Press.

    Chapter  Google Scholar 

  36. Koopman, Hilda. 2018. A note on Huave morpheme ordering: Local dislocation or generalized U20? In Perspectives on the architecture and acquisition of syntax, eds. Gautam Sengupta, Shruti Sircar, Madhavi Gayathri Raman, and Rahul Balusu. Berlin: Springer.

    Google Scholar 

  37. Kratzer, Angelika. 2009. Making a pronoun: Fake indexicals as windows into the properties of pronouns. Linguistic Inquiry 40(2): 187–237.

    Article  Google Scholar 

  38. Lacroix, René. 2009. Description du dialecte laze d’arhavi: Grammaire et textes. PhD diss., Université Lumière Lyon 2.

  39. Lacroix, René. 2011. Ditransitive constructions in Laz. Linguistic Discovery 9(2).

  40. Lebeaux, David. 1991. Relative clauses, licensing, and the nature of the derivation. In Syntax and semantics, Vol. 25, 209–239.

    Google Scholar 

  41. Lomashvili, Leila, and Heidi Harley. 2011. Phases and templates in Georgian agreement. Studia Linguistica 65(3): 233–267.

    Article  Google Scholar 

  42. McGinnis, Martha. 2008. Phi-feature competition in morphology and syntax. In Phi theory: Phi-features across modules and interfaces, eds. Daniel Harbour, David Adger, and Susana Béjar, 155–184. London: Oxford University Press.

    Google Scholar 

  43. McGinnis, Martha. 2013. Agree and fission in Georgian plurals. In Distributed morphology today, eds. Ora Matushansky and Alec Marantz, 39–57. Cambridge: The MIT Press.

    Chapter  Google Scholar 

  44. Nash-Haran, Léa. 1992. La catégorie agr et l’accord en géorgien. Recherches Linguistiques 21: 65–79.

    Google Scholar 

  45. Nevins, Andrew. 2011. Multiple agree with clitics: Person complementarity vs. omnivorous number. Natural Language and Linguistic Theory 29(4): 939–971.

    Article  Google Scholar 

  46. Öztürk, Balkız, and Markus Pöchtrager, eds. 2011. Pazar laz. Munich: LINCOM Europa.

    Google Scholar 

  47. Öztürk, Balkız, and Eser Taylan Erguvanlı. 2017. Omnipresent little v in Pazar Laz. In The verbal domain, Oxford studies in theoretical linguistics, eds. Roberta D’Alessandro, Irene Franco, and Ángel J. Gallego, 207–231. New York: Oxford University Press.

    Google Scholar 

  48. Pantcheva, Marina. 2011. The nanosyntax of directional expressions: Decomposing path. PhD diss., University of Tromsø.

  49. Pantcheva, Marina, and Pavel Caha. 2012. Contiguity beyond linearity: Modeling cross-dimensional syncretisms.

  50. Preminger, Omer. 2011. Asymmetries between person and number in syntax: A commentary on Baker’s SCOPA. Natural Language and Linguistic Theory 29(4): 917–937.

    Article  Google Scholar 

  51. Sportiche, Dominique. 2005. Division of labor between merge and move: Strict locality of selection and apparent reconstruction paradoxes. Ms., UCLA.

  52. Sportiche, Dominique. 2006. NP movement: How to merge and move in tough-constructions. Ms., UCLA.

  53. Starke, Michael. 2009. Nanosyntax: A short primer to a new approach to language. Nordlyd 36(1): 1–6.

    Google Scholar 

  54. Starke, Michal. 2018. Complex left branches, spellout, and prefixes. In Exploring nanosyntax, eds. Lena Baunaz, Karen De Clercq, Liliane M. V. Haegeman, and Eric Lander, Oxford studies in comparative syntax. New York: Oxford University Press.

    Google Scholar 

  55. Stegovec, Adrian. 2019. Taking case out of the person-case constraint. Natural Language and Linguistic Theory

    Article  Google Scholar 

  56. Stepanov, Arthur. 2001. Late adjunction and minimalist phrase structure. Syntax 4(2): 94–125.

    Article  Google Scholar 

  57. Stump, Gregory T. 2001. Inflectional morphology: A theory of paradigm structure. Cambridge: Cambridge University Press.

    Book  Google Scholar 

  58. Svenonius, Peter. 2016. Spans and words. In Morphological metatheory, eds. Daniel Siddiqi and Heidi Harley. Vol. 229 of Linguistik aktuell/linguistics today, 201–222. Amsterdam: John Benjamins Publishing Company.

    Chapter  Google Scholar 

  59. Taraldsen, Knut Tarald 2018. Spanning versus constituent lexicalization: The case of portmanteau prefixes. In Exploring nanosyntax, eds. Lena Baunaz, Karen De Clercq, Liliane M. V. Haegeman, and Eric Lander, Oxford studies in comparative syntax. New York: Oxford University Press.

    Google Scholar 

  60. Trommer, Jochen. 1999. Morphology consuming syntax’ resources: Generation and parsing in a minimalist version of distributed morphology. In ESSLI workshop on resource logic and minimalist grammars, eds. Christian Retoré and Edward Stabler. Utrecht.

    Google Scholar 

  61. Williams, Edwin. 2003. Representation theory. Current studies in linguistics. Cambridge: MIT Press.

    Google Scholar 

Download references


I would like to thank my consultant, İsmail Bucaklişi, as well as Balkız Öztürk, Aslı Göksel and Markus Pöchtrager, who introduced me to Laz. I am also indebted to my friend and colleague Ömer Demirok, without whom this paper could not have become what it is. I am grateful to Léa Nash and Mariam Matiashvili for providing me with judgements and comments on Georgian. In addition, many other people have provided helpful comments on and criticisms of this work at various points: Omar Agha, Jonathan Bobaljik, Pavel Caha, Aaron Doliana, Maria Gouskova, Daniel Harbour, Stephanie Harves, Nina Haslinger, Dalina Kallulli, Richard Kayne, Alec Marantz, Kate Mooney, Gereon Müller, Yining Nie, Jeffrey Parrott, Sarah Phillips, Martin Prinzhorn, Michal Starke, Guy Tabachnick, Gary Thoms, Adina Williams, Stanislao Zompì, and three anonymous reviewers have all been generous with their thoughts and suggestions.

Author information



Corresponding author

Correspondence to Hagen Blix.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Blix, H. Spans in South Caucasian agreement. Nat Lang Linguist Theory 39, 1–55 (2021).

Download citation


  • Georgian
  • Laz
  • Agreement
  • Syntax-morphology interface
  • Spans
  • Person hierarchy
  • Phi features
  • Nanosyntax
  • Distributed morphology