Open Collections

WCCFL 2020 UBC  March 6th, 2020  1 Comparatives in San Sebastián del Monte Mixtec: a Mixed Construction Iara Mantenuto imantenu@ucla.edu University of California, Los Angeles  1. Introduction This talk presents novel data on conjoined comparatives from San Sebastián del Monte Mixtec (SSM),1 an Otomanguean language spoken in Oaxaca, Mexico.  (1) Yu’ù jikò ka ì, òònjìví Chuchi.2 [TP1 yu’ù  jikò=ga=ì]  , [TP2 òònjìví  Chuchi]. 1SG.IND tall.CONT=GA=1SG   NEG.N  Chuchi ‘Yo soy más alto que Chuchi.’ ‘I am taller than Chuchi.’  The goals of this talk: • Typological contribution:  Puzzle: Previous work of conjoined comparatives do not have gradable predicates           [-DSP]3 (Motu - Beck et al. 2009, Fijian - Pearson 2009, Washo - Bochnak 2015, Walpiri - Bowler 2016, old Samoan - Hohaus 2018). (Davis and Mellesmoen (2019)) Observation: Davis and Mellesmoen (2019) observe that one could suggest, based on the data above mentioned, that the availability of a conjoined comparative construction might be enough for identifying a [-DSP]. Contribution: I will support the idea that conjoined comparatives are not a sufficient diagnostic for [-DSP] status. • Theoretical contribution: I offer an analysis for conjoined comparatives in SSM. I propose that the first conjunct is an incomplete comparative (also known as context comparative), the second conjunct in SSM conjoined comparatives negates an alternative of the first clauses.                                                         1 I would like to thank my language teachers Félix Córtes and Adrian Espinosa Davila, and the whole community in San Sebastián del Monte for welcoming me and teaching me Mixtec. A big thank you to my Research Assistants Angela Sicong Xu and Tatevik Shahinyan for organizing the data and discussing it with me. I am grateful to Prof. Pam Munro, Prof. Carson Schütze, Prof. Dylan Bumford, and Megan Gotowski for their guidance and for their suggestions. Finally, I would like to thank the members of the American Indian Seminar and of the Syntax-Semantics Seminar at UCLA and the three anonymous WCCFL reviewers for their observations. This work was funded by the Institute of American Cultures and the American Indian Studies Center at UCLA. Any mistakes are my own.  Abbreviations which are used in this paper: POT = potential, CONT = continuative, COMP = completive, M = masculine, F = feminine, COP = non-verbal copula, SG = singular, PL = plural, HON = honorific, CFR = classifier, HUM = human, BASE = pronominal base, NEG.N = nominal negation, IND = independent pronouns. 2 Each example from my own fieldwork is reported in five tiers: 1. Community orthography which I am developing with the speakers in the community. 2. Linguists’ orthography more faithful to the way the language is pronounced with segmentation. 3. Glosses 4. Spanish translation 5. English translation.  3 DSP is what distinguishes a degreeless language ([-DSP]) from a degreeful one ([+DSP]).  WCCFL 2020 UBC  March 6th, 2020  2 The structure of this talk:  §2: introduces a brief outline of SSM with the information necessary to follow    this talk.    §3: introduces argumentation in favor of two kinds of conjoined comparatives found  crosslinguistically: a [+DSP] one and a [-DSP] one.  §4: offers an analysis to account for the comparative construction in SSM.   §5: concludes and discusses future research questions.  2. The data San Sebastián del Monte is a town in the Santo Domingo Tonalá municipality of Oaxaca state, Mexico, approximately 45km SW of Huajuapan de León, with a population of 2000 people. San Sebastián del Monte Mixtec (ISO: mks) is part of the Mixtecan language family, Otomanguean stock.   It is a tonal language (three tones) and it has a VSO word order, though other word orders are available depending on information structure (focus and topic).   (2)  Sísi tìnà xìtà.  sísi  tìnà xìtà eat.CONT dog tortilla  ‘El perro come la tortilla.’ ‘The dog eats the tortilla.’       WCCFL 2020 UBC  March 6th, 2020  3 (3)  Sísi tí xìtà.  sísi=tí   xìtà eat.CONT=3animal tortilla    ‘Él (el perro) come la tortilla.’  ‘It (the dog) eats the tortilla.’   The element which is in focus or in topic position needs to be to the left of the verb.  (4)  Lupi tà´vi vásò.  Lupi tà´vi  vásò  Lupi break.COMP glass   ‘LUPI rompió el vaso.’  ‘LUPI broke the glass.’       SFOCUSVO  (5)  Lupi tà´vi ñá vásò.  Lupi tà´vi=ñá  vásò  Lupi break.COMP=3SG.F glass   ‘Lupi, la que rompió el vaso.’  ‘As for Lupi, she broke the glass.’4      STOPICVSO  Adjectives can function as predicates (6), in which case they occur before the subject, as with verbal predicates. In these cases no copula occurs with them.  (6) Jikó Chuchi. jikó  Chuchi tall  Chuchi ‘Chuchi es alto.’ ‘Chuchi is tall.’  2.1. San Sebastián del Monte Mixtec Comparatives SSM can express comparatives with three distinct constructions, all of which use the  comparative marker ga.  (7)  Yu’ù jiko ka ì, òònjiví mee nì.  yu’ù      jiko=ga=ì  , òònjiví mee=nì   1SG.IND        tall.CONT=GA=1SG    NEG.N  BASE=2SG.HON ‘Yo soy más alto que usted.’5 ‘I am taller than you.’      Conjoined comparative                                                           4 Although from the translation it seems that we are dealing with a contrastive topic, this is not the case. I use this kind of translation (“as for…,”) to indicate that the preverbal element is a topic.  5 To simplify things I am going to translate each comparative as an English particle comparative, even when the subject of the comparative is in focus or topic. WCCFL 2020 UBC  March 6th, 2020  4 (8)  Yu’ù jiko ka ì nòò mee nì.  yu’ù      jiko=ga=ì  nòò mee=nì   1SG.IND        tall.CONT=GA=1SG   on BASE=2SG.HON ‘Yo soy más alto que usted.’ ‘I am taller than you.’      Locative comparative  (9)  Yu’ù jiko ka ì ja mee nì.  yu’ù      jiko=ga=ì  ja mee=nì   1SG.IND        tall.CONT=GA=1SG   than BASE=2SG.HON ‘Yo soy más alto que usted.’ ‘I am taller than you.’      Particle comparative  In all three examples, the comparative morpheme is cliticized to the predicate (‘tall’) and precedes the clitic pronoun.   Conjoined comparative: In (7) a clause and a phrase are conjoined, although no conjunction appears. The clause, as I will show later, ends with the first person singular pronoun, while the phrase in the second conjunct starts with the negation òònjiví. Locative and particle comparatives: (8) and (9) are monoclausal comparatives, where the standard of comparison (‘you’) is introduced by two different standard markers (the locative nòò and the particle ja).   These three constructions express the same idea, but they are distinct in the way they are formed. At the end of this talk I will show the distribution of each construction depending on the age group which uses them; however, today’s talk will focus only on conjoined comparatives (for more information about the other two comparatives refer to my dissertation).  2.2. Conjoined comparatives: a background Conjoined comparatives are comparatives which use two conjoined clauses to associate the target of comparison and the standard of comparison (Stassen 1985).  - The two clauses can be prosodically separated and thus are reported orthographically with a comma in between, or they can be coordinated by a conjunction (e.g. ‘and’) as in (12).  - In the second conjunct they can either use an antonym of the predicate used in the first conjunct  as in Samoan and in ʔayʔaǰuθəm (10)-(13), or they can use negation as in Motu and in Menonimi (11)-(12).6  (10) Ua loa lenei va’a , ua puupuu lena. Is long this boat  is short  that ‘This boat is longer than that boat.’   Samoan (Stassen, 1985: 187)                                                            6 Although specifics about prosodic boundaries are not offered by Stassen (1985), other than by using a comma (no reference to the actual prosody), the clause boundary seems to match the position of the comma within the sentence. WCCFL 2020 UBC  March 6th, 2020  5 (11) Ina na namo herea , una na dia namo this is good more  that is not good  ‘This is better than that.’    Motu (Stassen 1985: 186)  (12) Apeqsek tata’hkesew ,  nenah teh kan  more  he-is-strong  I and not  ‘He is stronger than me.’    Menomini (Stassen 1985: 186)  (13) x̆ax̆aɬ  Tony,  titul  Laura.   tall Tony small Laura  ‘Tony is taller than Laura’ (Literally: ‘Tony is tall, Laura is small.’        ʔayʔaǰuθəm (Davis & Mellesmoen 2019: 47)  Like other conjoined comparative constructions, SSM uses a prosodic break between the two clauses (represented with a comma in writing, but produced with a pause orally); the same break is not available in the other two comparative constructions available in SSM (which are monoclausal). Furthermore, the second clause has a contrastive nominal negation (òònjìví), which negates the subject, the direct object or the indirect object of what is understood as the clause.   (14) Sìsi kà ì tako, òòn jìví7 mee ní.  sìsi=gà=ì  tako , òònjìví  mee=ní eat.CONT=GA=1SG taco  NEG.N  BASE=2SG.HON  ‘Comí más tacos que usted.’ ‘I ate more tacos than you.’  (15) Kási kuá’a ka tìna ndìka, òòn jivi tìkuaá.  kási-kuá’a=ga  tìna ndìka , òònjivi tìkuaá eat.CONT-much=GA dog banana  NEG.N orange ‘El perro está comiendo más bananas que naranjas.’ ‘The dog is eating more bananas than oranges.’                                                           7 Where else do we find òònjìví? It is a negation used specifically to negate DPs.  (i) Context: someone asks me if I ate the last apples in the kitchen.  Sàsì tikua’á, òònjiví manzana.  sàs=ì  tikua’á  , òònjiví manzana  eat.COMP=1SG orange   NEG.N apple  ‘Comí naranjas, no manzanas.’  ‘I ate oranges, not apples.’ (ii) Context: someone asks me if I ate the last apples in the kitchen.  Sàsì, òònjiví manzana.  sàs=ì  , òònjiví manzana  eat.COMP=1SG  NEG.N apple  Lit.: ‘Comí, no manzanas.’  Lit.: ‘I ate, [but] not apples.’  WCCFL 2020 UBC  March 6th, 2020  6 (16) Tantà’a kà ì letra Liya, òònjìví Chuchi.  tantà’a=gà=ì  letra Liya , òònjìví Chuchi send.CONT=GA =1SG letter Liya , NEG.N Chuchi ‘Estoy enviando mas letras a Liya que a Chuchi.’ ‘I am sending more letters to Liya than to Chuchi.’8   So far, then, the SSM construction fits the description of a conjoined comparative, by virtue of the prosodic break and the negation in the second conjunct.   In the following sections I will bring empirical evidence for the argument that indeed SSM conjoined comparative has gradable predicate and a comparative morpheme.   3. The proposal: not all conjoined comparatives comparatives are created equal  I will prove that SSM, similarly to ʔayʔaǰuθəm, has gradable predicates ([+DSP]). Thus confirming that conjoined comparatives are not a diagnostic for the [+/-DSP] parameter (Davis and Mellesmoen, 2019).  In the following section I will attest what typologically expected characteristics of comparatives the conjoined comparatives in SSM have. SSM conjoined comparatives: - ga functions as a comparative morpheme in the conjoined comparative construction, - SSM is a [+DSP] language   • Without the ga morpheme it is not possible for (17) to be interpreted in a comparative manner, as in (18).  (17) Chuchi jikò kà, òònjìví See.  Chuchi  jikò=gà  , òònjìví  See      Chuchi  tall.CONT=GA  NEG.N  See    ‘Chuchi es más alto que See.’  ‘Chuchi is taller than See.’  (18) Chuchi jikò, òònjìví See.  Chuchi  jikò  , òònjìví  See      Chuchi  tall.CONT  NEG.N  See    ‘Chuchi es alto, no See.’  ‘Chuchi is tall, [but] not See.’   Conclusion 1: The morpheme ga used in the first conjunct behaves as a comparative marker, and not simply as a contrastive or emphatic marker.9                                                           8 This sentence is ambiguous between the meaning offered above and ‘I am sending more letters to Liya than Chuchi is sending.’ One of my consultant suggested that we could disambiguate the two meanings by using the locative comparative. 9 That ga is a contrastive or emphatic marker has been previously raised for Chalcatongo Mixtec by Bobaljik (2012).  WCCFL 2020 UBC  March 6th, 2020  7 3.1. A language with gradable predicates [+DSP]  As demonstrated by Deal and Hohaus (2019), the presence of the comparative morpheme alone does not mean that we are dealing with a degreeful language.  In this section I will check whether predicates in SSM are indeed gradable or if they are vague.  The Degree Semantics Parameter (Beck et al., 2009) captures how languages vary in the semantics of their gradable predicates.  Degree Semantics Parameter [±DSP] (Beck et al., 2009: 19):  A language {does/does not} have gradable predicates (type <d, <e,t>> and related), i.e.  lexical items that introduce degree arguments.  Going back to two of the most reported examples of [+DSP] and [-DSP]10 languages in the comparative literature, English and Washo respectively.   (19) John is taller than Mary.  (20) ’é:liwhu delkáykayiʔ  k’éʔi  daʔmóʔmoʔ delkáykayiʔé:s  t’e:liwhu de-ʔil-kaykay-iʔ k’-eʔ-i  daʔmoʔmoʔ de-ʔil-kaykay-iʔ-e:s  man  nmlz-attr-tall-attr 3-cop-ipfv woman nmlz-attr-tall-attr-neg  k’áʔaš  k’-eʔ-aʔ-š  3-cop-aor-sr  ‘The man is taller than the woman.’  (Literally: ‘The man is tall, the woman is not tall.’)   Bochnak (2015:4)  Their predicates’ lexical entries are given in (21) and (22).   (21) [[tallWasho[-DSP]]]C=𝜆𝑥. 𝑥	𝑐𝑜𝑢𝑛𝑡𝑠	𝑎𝑠	𝑡𝑎𝑙𝑙	𝑤𝑖𝑡ℎ	𝑟𝑒𝑠𝑝𝑒𝑐𝑡	𝑡𝑜	𝐶    (22) [[tallEnglish[+DSP]]]C=𝜆𝑑. 𝜆𝑥.HEIGHT(𝑥) ≥ 𝑑  To distinguish between these two systems, where English involves degree arguments while Washo does not, I use differential comparatives as a test for [DSP]. It is known that although other tests (see appendix) have been successfully modeled in a [-DSP] system, the existence of differential comparatives is agreed upon as a sufficient condition for a language to be [+DSP].   Differential measure phrases: SSM conjoined comparative is capable of hosting a differential measure phrase as well.  (23) Jikó Liya iin metro kà, òònjìví Lupe.  jikó   Liya  iin  metro=kà  , òònjìví  Lupe tall.CONT Liya one meter=GA   NEG.N  Lupe  ‘Liya es un metro más alta que Lupe.’  ‘Liya is one meter taller than Lupe.’                                                         10 Predicates in [-DSP] languages are also known as vague predicates. WCCFL 2020 UBC  March 6th, 2020  8 - The differential measure phrase test is reported in (23) and the vague quantificational differential comparatives are also grammatical as reported in (24)-(25).  (24) Jìì cha ka inì, òònjìví Liya.  jìì  cha=ga  inì , òònjìví  Liya happy.CONT little=GA mind  NEG.N  Liya ‘Yo estoy un poco más féliz que Liya.’ ‘I am a little happier than Liya.’ (25) Yu’ù jiko kua’a ka ì, òònjìví mee ní.  yu’ù   jiko  kua’a=ga=ì   , òònjìví  mee=ní 1SG.IND tall.CONT much=GA=1SG   NEG.N  BASE=2SG.HON ‘Yo soy mucho más alta que usted.’ ‘I am much taller than you.’  Conclusion 2: SSM is a language with gradable predicates ([+DSP]).  4. Analysis of conjoined comparatives in SSM Theoretically a question arises: if SSM has gradable predicates ([+DSP]) and it uses a comparative morpheme, why do we have a conjoined comparative structure and what does that mean for the analysis of conjoined comparatives in SSM?  • In (26) there is an individual, or a set of individuals, which is salient in the discourse and that I am taller than.  (26) Yu’ù jikò ka ì.  yu’ù  jikò=ga=ì 1SG.IND tall.CONT=GA=1SG ‘Yo soy más alta.’ ‘I am taller.’  • When we only have the clause with the comparative marker, and we do not have a second coordinate clause, or a standard phrase, the meaning in Mixtec is that of an incomplete comparative (this term was created by Sheldon 1945 (as cited by Schwarzschild (2008)), but today it is also known as context comparative (Hohaus 2015)).  In (26) there is an implied completion to the comparative meaning, as previously indicated by Sheldon (1945) in English (27).  (27) {Come out onto the porch.} It’s cooler here.   (Schwarzschild 2008:89) In (27) the implied completion is “than inside” and it is made clearer by the preceding sentence. In SSM the implied completion was inserted in the contexts I have given before uttering (28) in Mixtec.  WCCFL 2020 UBC  March 6th, 2020  9 (28) Context 1: There is a group of tall people, all with different levels of tallness (I offered a drawing with a number equivalent to the height of each person). Can I say (26)?   (29) Context 2: Chuchi and I are both basketball players, I am 198 cm tall and Chuchi is 190 cm tall. Can I say (26)?  The answer to both contexts in (28) and (29) was affirmative, thus we can see that as Sheldon (1945) predicted for English, the completion of the comparative sentence is clear to the speaker and influenced by the context.  Moreover, using (26) in the context in (28) when no one else is as tall or taller than myself, shows that (26) also has a superlative meaning, as long as I allow the comparison set to have more than one member.  Conclusion 3: The context must be referred to by ga. Thus, a quantificational determiner (in our case ga – a degree quantifier) has an argument index whose value is determined by the pragmatics derived from the discourse (and I will add/clarify by the shared knowledge of the speakers) (vonFintel 1994, Schwarzschild 2010)  Moreover, as I proposed earlier, the predicates in SSM are gradable (30) and ga functions as a comparative morpheme in comparatives (31).  (30) [[jikò]]C=𝜆𝑑. 𝜆𝑥.HEIGHT(𝑥) ≥ 𝑑     - Jikò is a gradable predicate, d is a variable over degrees, x is a variable over individuals, and HEIGHT is a measure function relating x to d, a degree on the scale of height (Cresswell 1976, von Stechow 1984, Heim 1985, 2001, Kennedy & McNally 2005 among others).  I am going to assume the following semantics for the comparative morpheme in SSM, which is what usually we would use for the incomplete comparative’s operators (Hohaus 2015).  (31) [[gaC]]=𝜆𝑅<d,<e,t>>. 𝜆𝑥.𝑀𝐴𝑋(𝜆𝑑. 𝑅(𝑥)(𝑑) = 1) > 𝑀𝐴𝑋(𝜆𝑑. ∃𝑥𝜖𝐶[𝑅(𝑥)(𝑑)])  A comparative like (31) would mean ‘The maximal degree d such that Liya is d-tall exceeds some contextually provided height degree.’  (32) [[ [Liya [tall<d,<e,t>> comparison-ga] ] ]]C= 1iff  [𝜆𝑅<d,<e,t>>. [𝜆𝑥.𝑀𝐴𝑋(𝜆𝑑. 𝑅(𝑑)(𝑥) = 1) = 𝑐]]	(𝜆𝑑. 𝜆𝑥.𝐻𝐸𝐼𝐺𝐻𝑇(𝑥) ≥ 𝑑)(𝐿) = 1  Iff MAX(𝜆𝑑. 𝐻𝐸𝐼𝐺𝐻𝑇(𝐿) ≥ 𝑑) > 𝑐  Moreover, the semantics of the comparative is the same as the superlatives; we are going to assume that there is a set of possibilities, and in the comparative there is a set of two in there.  In the case of differentials we would add an extra variable.  WCCFL 2020 UBC  March 6th, 2020  10 (33) [[gaC]]C=𝜆𝑑. 𝜆𝑅. 𝜆𝑥.𝑀𝐴𝑋(𝜆𝑑. 𝑅(𝑑)(𝑥) = 1) 	≥ 𝑀𝐴𝑋(𝜆𝑑. ∃𝑥𝜖𝐶[𝑅(𝑥)(𝑑)]) +d  Let’s now add the second conjunct: (34) Jikò ka Liya, òònjìví Chuchi.  Jikò=ga Liya , òònjìví Chuchi. tall.CONT=GA Liya  NEG.N Chuchi ‘Liya es más alta que Chuchi.’ ‘Liya is taller than Chuchi.’  While the English than-clause specifies a standard of comparison directly, the second conjunct in SSM conjoined comparatives negates an alternative of the first clauses (where the DP in the second clause takes the place of one of the DP in the first clause).  Similarly to English:  (35) I ate more tacos, you did not eat more tacos.  When the second conjunct is missing, similarly to English, the context is relevant.   (36) {John is sick} I ate more tacos.  (37) {The burritos were awful} I ate more taco.   Observation: The second conjunct is formed by a nominal negation which takes a DP as its argument and it moves it to the left periphery of the clause. The rest of the clause can be elided under stripping.   Assumption: For the sake of concreteness one can assume, in line with the main analysis of fragments in stripping, that the negated DP has been fronted as it is the focused element, and everything else is elided. And as a matter of fact, independently in SSM we can find focused elements fronted.    (38)                TP       TP            (and) yu’ù               òònjìví   Chuchi      gaC       𝜆x             𝜆d               𝜆𝑥        𝜆d        x    x                      d       jikò  d jikò WCCFL 2020 UBC  March 6th, 2020  11 5. Conclusion SSM: is a [+DSP] language, has a coordination and a comparative marker. The first conjunct is an incomplete comparative, while the second conjunct negates one of the alternatives of the first clause.  The coordinate nature of the comparative does not inform us as to whether the languages is necessarily [-DSP] or on whether the second conjunct functions as an actual standard phrase; every point needs to be tested before concluding anything.  Whether SMM has undergone a shift from being [-DSP] to being [+DSP], similarly to Samoan (Hohaus 2018), is at this point pure speculation. Some support for this idea comes from the fact that the particle comparative started only with the new generations, while it did not previously exist.   40+ 20-40 11-20 Conjoined Yes No No Locative Yes Yes No Particle  No Yes Yes  Moreover, the particle comparative does not exist in closely related languages like Yucuquimi de Ocampo Mixtec (p.c. León Vázquez). It will be worthwhile to explore this point in the future, with both careful synchronic and diachronic comparison, in order to test this hypothesis.   As work done on ʔayʔaǰuθəm (Davis and Mellesmoen 2019), Samoan (Hohaus 2014, 2018), Washo (Bochnak 2014, 2018) and Nez Perce (Deal & Hohaus 2019) among others, has shown, to better inform our understanding of comparatives, and specifically of conjoined comparatives, we need to do more theoretically driven fieldwork. I look forward to provide more information to this comparative enterprise.   6. Reference Beck, S., Svetlana K., Daniel F., Remus G., Stefan H., Christiane S., John V. & Elisabeth V.  2009. “Crosslinguistic Variation in  Comparison Constructions.” Linguistic Variation  Yearbook  9: 1–66. Bierwisch, M. 1989. The semantics of gradation. In M. Bierwisch and E. Lang (Eds.),  Dimensional adjectives: grammatical structure and conceptual interpretation, pp. 71–262.  Berlin: Springer. Bobaljik, J. D. 2012. Universals in comparative morphology. Cambridge, MA: MIT Press. Bochnak, R. 2013. “Crosslinguistic Variation in the Semantics of Comparatives.” PhD thesis.  Chicago: University of Chicago. Bowler, M. L. 2016. The status of degrees in Warlpiri. In Proceedings of TripleA 2, pp. 1–17.  esswell, Max J. 1976. The semantics of degree. In Barbara Partee (ed.), Montague  Grammar, 261–292. New York: Academic Press. Davis, H., & Mellesmoen, G. 2019. Degree Constructions in Two Salish Languages.  In Papers for the International Conference on Salish and Neighbouring Languages 54. WCCFL 2020 UBC  March 6th, 2020  12 Deal, A. R., & Hohaus, V. 2019. Vague predicates, crisp judgments. In Proceedings of  Sinn  und Bedeutung (Vol. 23, No. 1, pp. 347-364). Fara, D. G. 2000. Shifting sands: An interest-relative theory of vagueness. Philosophical  Topics 28. 45–81. doi:10.5840/philtopics20002816. Originally published under the name  “Delia Graff”. Von Fintel, K. 1994. “Restrictions on quantifier domains.” Ph.D. thesis. Amherst:University of  Massachusetts at Amherst. Heim, I. 1985. Notes on comparatives and related matters. Ms. University of Texas at  Austin. Heim, I. 2001. Degree operators and scope. In Caroline F´ery & Wolfgang Sternefeld (eds.),  Audiatur vox sapientiae, 214–239. Berlin: Akademie-Verlag. Hohaus, V. 2015. “Context and composition: how presuppositions manipulate the interpretation  of free variables”. Ph. D. thesis. Tubingen: Universitat Tubingen. Hohaus, V. 2018. How do degrees enter the grammar? Language change in Samoan from [- DSP] to [+ DSP]. Kennedy, C. 2007. Modes of comparison. In Malcolm Elliott, James Kirby, Osamu  Sawada, Eleni Staraki & Suwon Yoon (eds.), Chicago Linguistic Society (CLS) 43,  Chicago: Chicago Linguistic Society. Kennedy, C. 2011. Vagueness and comparison. In Paul Egr´e & Nathan Klinedinst  (eds.),Vagueness and Language Use, 73–97. New York: Palgrave MacMillan. Kennedy, C & L. McNally. 2005. Scale structure, degree modification and the  semantics of gradable predicates. Language 81(2). 345–381.  Klein, E. 1980. A semantics for positive and comparative adjectives. Linguistics and  Philosophy 4(1). 1–46.  Mantenuto, I. to appear. “Other and More in San Sebastián del Monte Mixtec and Beyond.”  Ph.D. thesis. Los Angeles: University of California, Los Angeles. Pearson, H. 2010. “How to do Comparison in a Language without Degrees: A Semantics for  the Comparative in Fijian.” In: Proceedings of Sinn und Bedeutung (SuB) 14 . Ed. by  Martin Prinzhorn, Viola Schmitt & Sarah Zobel, 339–355. Pinkal, M. 1995. Logic and Lexicon. Dordrecht: Kluwer. Sapir, E. 1944. Grading, a study in semantics. Philosophy of Science 11(2). 93–116. Schwarzschild, R. 2008. The semantics of comparatives and other degree constructions.  Language and linguistics compass 2, 308–331.  Schwarzschild, R. 2010. Comparative markers and standard markers. In Proceedings of the  MIT workshop on comparatives. Sheldon, E. K. 1945. The rise of the incomplete comparative. American Speech 20(3), 161–  167. Stassen, L. 1985. Comparison and Universal Grammar. Oxford: Blackwell. von Stechow, A. 1984. Comparing semantic theories of comparison. Journal of Semantics 3.  1–77.  7. Appendix 7.1. Remaining puzzles  In SSM, when the meaning change of the second clause, compared to the first clause, involves something more or other than one DP we need to produce the full second clause and not just the WCCFL 2020 UBC  March 6th, 2020  13 phrase òòn jìví+DP. That is true when the aspect of the verb changes as in (39), when the verb changes as in (40)-(41) and when the predicate adjective changes as in (42). The result is that the second clause acquires the shape of a light headed relative, headed by the classifier that better fits the meaning.  (39) Sìsi ka ì tako, òònjiví ñá ko’o ìn kusì. sìsi=ga=ì  tako , òònjiví ñá-ko’o=ìn   kus=ì eat.COMP=GA=1SG taco , NEG.N CFR:3THING-go.POT=1SG eat.POT=1SG ‘Yo comí más tacos de que voy a comer.’ ‘I ate more tacos than the ones I am going to eat.’  (40) Jakua’a ka ì, òònjiví ñà kixì.  jakua’a=ga=ì  , òònjiví ñà-kix=ì study.POT=GA=1SG  NEG.N CFR:3THING-sleep.POT=1SG ‘Yo voy a estudiar más de que voy a dormir.’ ‘I will study more than I will sleep.’  (41) Kixi ka ì, òònjiví ñà jakua’a ì.  kixi=ga=ì  , òònjiví ñà-jakua’a=ì sleep.POT=GA=1SG  NEG.N CFR:3THING-study.POT=1SG ‘Yo voy a dormir más de que voy a estudiar.’ ‘I will sleep more than I will study.’  (42) Káá ka’ni kà mesa, òònjiví ña jiko tó.  káá  ka’ni=gà mesa , òònjiví ña-jiko=tó be.CONT large=GA table  NEG.N CFR:3THING-tall=3WOOD ‘La mesa está más larga que alta.’ ‘The table is larger than it is tall.’  Example (40) is a very interesting case because there is no overt object to “study,” yet the pronoun referring to THING is used. This is true even in example (41) with “sleep,” where “sleep” is an intransitive verb and it should not take any direct object. Thus, we can conclude that the third person pronoun referring to non-round inanimates can also refer to ideas or amounts. If the verb changes but the direct object stays the same the best fitting pronoun would be used, as the case with a round object, the orange, shows in (43)-(44). As we would not be able to normally use the pronoun referring to things (ñá) for oranges it follows that it cannot be used even in comparative constructions.  (43) Vìtì sàsì ò’òn kà tikuaa, òònjìví ti sàsì kònì. vìtì sàs=ì  ò’òn=gà tikuaa , òònjìví ti-sàs=ì today eat.COM=1SG five=GA orange  NEG.N CFR:3ROUND-eat.COMP=1SG kònì yesterday ‘Hoy comí 5 naranjas más de que comí ayer.’ ‘Today I ate five oranges more than I ate yesterday.’  WCCFL 2020 UBC  March 6th, 2020  14 (44) *Vìtì sàsì ò’òn kà tikuaa, òònjìví ñá.  *vìtì sàs=ì  ò’òn=gà tikuaa , òònjìví ñá-sàs=ì    today eat.COM=1SG five=GA orange  NEG.N CFR:3THING-eat.COMP=1SG kònì yesterday Since it is obligatory for the DP preceded by òòn jiví to move leftward in a focus position when in a comparative construction we can assume that òòn jiví+DP  will always occur to the left of the verb, whether the verb and the rest of the subordinate clause follows it overtly or it has been elided.   An additional puzzle is the impossibility of using the conjoined comparative when building a comparison with a degree as in (45). The ungrammaticality of (45) is repaired by an exceed comparative construction (thus adding a fourth kind of comparative to Mixtec).  (45) *Kà’vì kua’a kà ì, òònjìví ònì (libro).   *kà’vì-kua’a=kà=ì   , òònjìví  ònì  (libro)  read.COMP-many=GA=1SG  NEG.F  three book  Intended: ‘Yo leí más libros que tres.’  Intended: ‘I read more than three books.’  (46) Nìyà’a o ònì libro kà’vi.   nìyà’a   o  ònì  libro  kà’v-i  exceed.COMP COP three book read.COMP-1SG  Intended: ‘Yo leí más libros que tres.’  Intended: ‘I read more than three books.’  7.2. Kennedy’s (2007) tests (i) Crisp judgment test11 Implicit comparisons, where the predicate is in its positive form (“tall” rather than “taller”), is expected to be infelicitous in crisp judgement contexts, as the difference in height is minimal.12                                                           11 The Crisp Judgment Test relies on the Similarity Constraint: When x and y differ only to a very small degree in the property that a vague predicate G is used to express, speakers are unable or unwilling to judge the proposition that x is G true and y is G false (Klein 1980, Fara 2000, Kennedy 2011). 12 As the Washo case shows, when there is a minimum difference between the two gradients compared, and the sentence is infelicitous, then we are dealing with an implicit comparison. (i) Context: comparing two ladders, where one is only slightly taller than the other.  # wíːdiʔ  ʔitmáŋa  delkáykayiʔ   k'éʔi   wíːdiʔ  delkáykayiʔéːs  wiiːdiʔ  ʔitmáŋa  de-ʔil-kaykay-iʔ   k'-eʔ-i   wiːdiʔ  de-ʔil-kaykay-iʔ-eːs  this ladder NMLZ-ATTR-tall-ATTR 3-COP-IPFV this NMLZ-ATTR-tall-ATTR-NEG  k'éʔaš  k'-eʔ-aʔ-š  3-COP-AOR-SR  Intended: ‘This ladder is taller than that one.’  (lit.: ‘This ladder is tall, that one is not tall.’)    Washo (Bochnak 2014: 171) WCCFL 2020 UBC  March 6th, 2020  15 In SSM even when we are dealing with a minimum difference between two gradients, the comparative is felicitous.   (47) Context: I am 180cm tall and you are 179cm tall.  Yu’ù jiko ka ì, òònjiví mee nì.   yu’ù      jiko=ka=ì  , òònjiví  mee=nì.   1SG.IND        tall.CONT=GA=1SG    NEG.N  BASE=2SG.HON  ‘Yo soy más alto que usted.’  ‘I am taller than you.’   (ii) Absolute standard predicate test Implicit comparisons are expected to be infelicitous with absolute-standard predicates since their standard do not depend on the context; if we say that one bar is bent, it would be false to assert that the other one is not, when the context provides that both bars have some degree of bentness. On the other hand, explicit comparisons are supposed to be felicitous with absolute-standard predicates because their comparative morpheme needs two distinct degrees of “bentness” in the example offered where we are comparing bars that have different degrees of bendedness.13  In SSM the word ‘curved’ is used instead of ‘bent’ but it shares the same absolute nature. Also for this test, the conjoined comparative in SSM is felicitous, which means that there is a degree difference between the two objects compared.   (48) Nìkàvà kà barra yó’ò, òònjìví ñá seen.  nìkàvà=kà  barra yó’ò , òònjìví  ñá-seen curved.COMP=CM bar this  NEG.N  CFR:3THING-there ‘Esta barra esta más curva que esa (barra).’ ‘This bar is more curved than that one (bar).’  (iii) Differential measure phrases Measure phrases are capable to override the semantics of the positive form of the predicate. By composing a gradable adjective with a measure phrase we obtain a predicate which is not context dependent anymore (Pinkal 1995). Thus, Implicit comparison should not allow a non-comparative adjective to combine with a measure phrase, as there would not be any standard of comparison to manipulate, while in the Explicit comparison differential measure phrases are allowed.14  SSM conjoined comparative is capable to host a differential measure phrase as well.                                                         13 The example reported from Washo shows a case of an implicit comparison. (i) #wíːdiʔ  ʔilk'únk'uniʔaš    wíːdiʔ  ʔilšíːšibiʔi  wiːdiʔ  ʔilk'unk'uniʔaš    wiːdiʔ  ʔilšiːšibiʔi  this ATTR-bent-ATTR-AOR-SR this ATTR-straight-ATTR-IPFV  Intended: ‘This one is more bent than that one.’  (lit.: ‘This one is bent, that one is straight.’) Washo (Bochnack 2014: 172) 14 Washo, which only has implicit comparisons, does not allow the occurrence of measure phrases in its comparative constructions. WCCFL 2020 UBC  March 6th, 2020  16 (49) Jikó Liya iin metro kà, òònjìví Lupe.  jikó   Liya  iin  metro=gà  , òònjìví  Lupe tall.CONT Liya one meter=GA   NEG.N  Lupe  ‘Liya es un metro más alta que Lupe.’  ‘Liya is one meter taller than Lupe.’  7.3. Distribution of the ga morpheme in comparatives Depending on the element that we are comparing, the ga morpheme can combine with: - a verbal predicate (50),  - an adverb (51),  - an adjective, predicate (52) and attributive (53),  - or a numeral (54)  (50) Ñakaa sàsi ka ña kopaya, òònjìví me ní. Ña-kaa                        sàsi=ga=ña                 kopaya , òònjìví me=ní CFR.3HUM.SG.F-that   eat.COMP=GA=3HUM.SG.F   papaya  NEG.N BASE=2SG.HON ‘Ella comió más papayas que Usted.’   ‘She ate more papayas15 than you.’  (51) Chuchi kana kono kama kà rà, òònjìví Liya.  Chuchi kana  kono-kama=gà=rà  , òònjìví Liya Chuchi go.CONT run.CONT-fast=GA=3HUM.M  NEG.N Liya  ´Chuchi corre más rápido que Liya.’ ‘Chuchi runs faster than Liya’  (52) Yu’ù jiko ka ì, òònjìví mee nì.   Yu’ù      jiko=ga=ì  , òònjìví mee=nì.  1SG.IND        tall.CONT=GA=1SG    NEG.N BASE=2SG.HON ‘Yo soy más alto que Usted.’ ‘I am taller than you.’   (53) Sàsì ndika tyina’no ka, òònjìví mee ní.  Sàsì  ndika tyina’no=ga , òònjìví mee=ní eat.COMP banana big=GA  NEG.N BASE=2SG.HON ‘Yo como bananas más grandes que Usted.’ ‘I ate bigger bananas than you.’                                                             15 SSM does not distinguish between mass nouns and count nouns. (i) Yu’ù kua’a kopayá sàsì. Yu’ù   kua’a kopayá  sàs=ì.    1SG.IND  much papaya eat.COMP=1SG  ‘Yo comí muchas papayas.’/ ‘Yo comí mucha papaya.’    ‘I ate many papayas./I ate much papaya.’ WCCFL 2020 UBC  March 6th, 2020  17 (54) Tikaa sìsi rà ò’òn ka tako, òònjìví ñakaa.  Ti-kaa                      sìsi=rà       ò’òn=ga tako    , òònjìví  CFR.3HUM.SG.M-that  eat.COMP=3HUM.M    five=GA   taco    NEG.N  ña-kaa CFR.3HUM.SG.F-that    ‘Él comió cinco bananas más que ella.’  ‘He ate five tacos more than she did.’  7.4. Adjectives in SSM Adjectives can be attributive (55) and they can function as predicates (56), in which case they occur before the subject, as with verbal predicates. In these cases no copula occurs with them.  (55) Ñalo’ò jikó nìntivi nòò yá’vì.  ña-lo’ò   jikó  nìntivi   nòò  yá’vì  CFR:3HUM.SG.F-little tall enter.COMP land market  ‘La chica alta entró al mercado.’  ‘The tall girl entered the market.’  (56) Jikó ñálo’ò.  jikó  ñá-lo’ò  tall CFR:3HUM.SG.F-little  ‘La chica es alta.’  ‘The girl is tall.’  7.5. Various additional data In regular particle comparatives, where the second clause is a subordinate, the test generally used to check whether we are dealing with phrasal versus clausal comparative is an ambiguity test like the one reported in (57). The ambiguity can be seen by looking at the ambiguity that a sentence like (58) has in English, between external and internal reading.  (57) Mary bought a bigger car than John.     a. External reading: Mary’s new car is bigger than John’s car.  b. Internal reading: Mary’s new car is bigger than John. (Deal & Hohaus 2019:353) Although the same seems to be true also for SSM, at a more in depth analysis the ambiguity per se should be expected if we are dealing with coordination and with ellipsis.   (58) Jákákó Liya iin leè ka’no kà, òònjìví Lupi.  jákákó   Liya  iin  leè  ka’no=gà  , òòn-jìví  Lupi  give.birth.CONT Liya one baby small=GA  NEG.N-AFF Lupi  ‘Maria dió a luz un bebe más grande que Lupe.’  ‘Mary gave birth to bigger children than Lupe.’  a. External reading: Mary gave birth to bigger babies than Lupe did.  b. Internal reading: Mary gave birth to babies bigger than Lupe.  WCCFL 2020 UBC  March 6th, 2020  18 (59) Kòó jìkó kà ì, òònjìví mèní.  kòó  jìkó-gà=ì  , òònjìví mè=ní  NEG tall-GA=1SG  NEG.N BASE=2SG.HON  ‘Yo no soy más alto que usted.’  ‘I am not taller than you.’ (60) a. Chuchi sìsi kà táko nòò Lupe, òònjìví Liya.   Chuchi sìsi=gà  táko  nòò  Lupe ,  òònjìví Liya   Chuchi eat.COMP=GA taco on Lupe  NEG.N Liya   ‘Chuchi comió más manzanas que Lupe, not Liya.’   ‘Chuchi ate more apples than Lupe, not Liya.’  b. *Chuchi sìsi kà táko, òònjìví Lupe, òònjìví Liya.   *Chuchi  sìsi=gà  táko , òònjìví Lupe , òònjìví Liya   Chuchi  eat.COMP=GA taco  NEG.N Lupe  NEG.N Liya  (61) Ká’an Liya ñá jikò ní, òònjìví Chuchi.  Ká’an   Liya ñá  jikò=ní  , òònjìví  Chuchi.  speak.CONT Liya that tall=2SG.HON  NEG.N  Chuchi  ‘Liya dice que usted es alto, no Chuchi.’  ‘Liya says that you are taller than Chuchi.’ (62) a. Nájoò  ka’an Liya jiko kà nòò Chuchi?   nájoò  ka’an   Liya  jiko=gà  nòò  Chuchi   who speak.CONT Liya tall=GA on Chuchi   ‘Quién dice Liya que es más alto que Chuchi?’   ‘Who does Liya say is taller than Chuchi?’  b. *Nájoò ka’an Liya jiko kà, òònjìví Chuchi?   *Nájoò ka’an   Liya jiko=gà , òònjìví Chuchi   who speak.CONT Liya tall=GA  NEG.N Chuchi              WCCFL 2020 UBC  March 6th, 2020  19