West Coast Conference on Formal Linguistics (WCCFL) (38th : 2020)

Nominal ellipsis reveals concord in Moksha Mordvin Privizentseva, Mariia 2020-03-06

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


73804-Privizentseva_M_Nominal_ellipsis_WCCFL38_2020.pdf [ 342.33kB ]
JSON: 73804-1.0389868.json
JSON-LD: 73804-1.0389868-ld.json
RDF/XML (Pretty): 73804-1.0389868-rdf.xml
RDF/JSON: 73804-1.0389868-rdf.json
Turtle: 73804-1.0389868-turtle.txt
N-Triples: 73804-1.0389868-rdf-ntriples.txt
Original Record: 73804-1.0389868-source.json
Full Text

Full Text

WCCFL UBC ¨ March 6-8, 2020Nominal ellipsis reveals concord in Moksha Mordvin∗Mariia Privizentseva (Leipzig University, mprivizentseva@uni-leipzig.de)1 IntroductionInflection under ellipsis• In some languages modifiers generally do not show concord with the noun, butare inflected if the noun is elided:(1) a. [ adjective noun-infl ] / * [adjective-infl noun-infl ]b. [ adjective-infl noun-infl ] / * [adjective noun-infl ]Why a modifier is inflected only under ellipsis?Previous research:• The pattern is not new, see, e.g., Hungarian (Kester, 1996a; Saab & Lipta´k,2016), Persian (Ghaniabadi, 2010), Turkish (Bosˇkovic´ & S¸ener, 2014), Ossetic(Hettich, 2002).• There is a number of existing approaches (see Kester (1996a,b), Bosˇkovic´ & S¸ener(2014), De´ka´ny (2011), Saab & Lipta´k (2016), Ruda (2016), Murphy (2018), andSaab (2019)).Here I will:• Present the original data on nominal ellipsis in Moksha Mordvin;• Show that the existing approaches do not cover a full range of data;• Develop a new account.Sketch of the analysis1. Inflection in elliptical contexts is nominal concord• Evidence: Inflection has the same distribution as regular concord.2. Concord is a regular property of Moksha nominal syntax.• Features are present on a nominal modifier in non-elliptical contexts as well.3. Valued concord probes remain without morphological realization.(a) Spell-Out applies locally.(b) Shortly after valuation probe features are still identifiable as such and arenot yet subject to Vocabulary Insertion.∗I am very grateful to native speakers of Moksha I worked with for their excellent linguisticintuitions and all the time they have spent with me. I also would like to thank Mark Baker,Rajesh Bhatt, Seth Cable, Pavel Caha, Kyle Johnson, Maria Kouneli, Maria Kholodilova, EkaterinaLyutikova, Franc Marusˇicˇ, Beata Moskal, Gereon Mu¨ller, Andrew Murphy, David Pesetsky, MartinSalzmann, Sergei Tatevosov, and Svetlana Toldova for comments and suggestions.2 Data• Moksha belongs to the Mordvin group of Finno-Ugric languages. It is spoken inthe Republic of Mordovia, Russia. The data come from my own fieldwork.• Basics: SOV / SVO; genitive is the case of the direct object.2.1 Nominal ellipsis• Nouns in Moksha are inflected for case, definiteness and number– Inflection is fusional, there are restrictions on which features can be expressedtogether (e.g., definiteness can be only marked in structural cases).• All inflection appears on the noun:adj noun-infl / *adj-infl noun-infl / *adj-infl noun(2) ravzˇ@blackpinj@-nj@-njdjidog-def.pl-dat/ *ravzˇ@-nj@-njdjiblack-def.pl-datpinj@-nj@-njdjidog-def.pl-dat/*ravzˇ@-nj@-njdjiblack-def.pl-datpinj@dog‘to the black dogs’• If the noun is elided, its modifier is inflected for features of elided noun:(3) Pakayetzvonj-cj@-sjcall-freq-pst.3[sg]anjcj@konly[kaft-nj@-njdji].two-def.pl-dat‘{Context: My mom is calling to her friends.} By now she called only to the two[friends].’• If there is more than one remaining modifier, only the linearly last modifieris inflected:(4) MonIand-inj@feed-pst.3.o.1sg.s[maziniceaksˇ@-tj]white-def.sg.gen/ *[mazi-tjnice-def.sg.genaksˇ@]white/*[mazi-tjnice-def.sg.genaksˇ@-tj].white-def.sg.gen‘{Which cat did you feed?} I fed the beautiful white one.’• Inflection appears on the head of a branching modifier even if its head isnot the linearly closest element to the ellipsis site.– An argument of the participle can precede or follow it:(5) MonIrama-jnj@buy-pst.3.o.1sg.s1a. [keluv-@njbirch-genlopa-st@leaf-elti-fmake-ptcp.resnastojka-tj]liquor-def.sg.genb. [ti-fmake-ptcp.reskeluv-@njbirch-genlopa-st@leaf-elnastojka-tj].liquor-def.sg.gen‘I bought the liquor made from birch leafs.’• If the noun is elided, morphological exponents appear on the participle ratherthen on its argument in both cases:(6) MonIrama-jnj@buy-pst.3.o.1sg.sa. [keluv-@njbirch-genlopa-st@leaf-elti-f-tj]make-ptcp.res-def.sg.genb. [ti-f-tjmake-ptcp.res-def.sg.genkeluv-@njbirch-genlopa-st@]leaf-elc.*[ti-fmake-ptcp.reskeluv-@njbirch-genlopa-st@-tj].leaf-el-def.sg.gen‘{Context: Which liquor did you buy?} I bought the [liquor] made from birchleafs.’• If the elative form modifies the elided noun directly, inflection is possible:(7) MonIrama-jnj@buy-pst.3.o.1sg.s[keluv-@njbirch-genlopa-st@-tj].leaf-el-def.sg.gen‘{Context: Which liquor did you buy?} I bought the one from birch leafs.’Summary: Inflection is on the head of the modifier that is closest to the ellipsis site.2.2 Structure in the ellipsis site?• It is often argued for the unpronounced syntactic structure in the ellipsissite (see Merchant (2001), and also the recent overviews by van Craenenbroeck& Merchant (2013) and Merchant (2019)).• Also a common assumption in the literature on nominal ellipsis (see Corver &van Koppen (2009), Alexiadou & Gengel (2012), Merchant (2014), Saab & Lipta´k(2016), Saab (2019), see also Cinque (2012)).• Elided noun shows connectivity effects to the rest of the noun phrase:1. Elided noun can assign a Θ-role to its argument:(8) MonImuj-inj@find-pst.3.o.1sg.s[tjEthispisatj@lj-tjwriter-def.sg.genskucˇn@-st@].boring-el‘{Context: In which novel did you find a mistake?} I found in this author’s boring[novel].’2. A modifier of the elided noun can be extracted as in non-elliptical contexts:(9) MonIafnegsoda-sa,know-npst.3sg.o.1sg.skinjwho.genkolgaaboutKatiaKatiarama-zj@buy-pst.3sg.o.3sg.s[sj@thisocju-tj]big-def.sg.gen‘{Context: Katia bought books.} I don’t know, about whom Katia bought thisbig one.’3. Idiosyncratic markings of arguments are preserved under ellipsis:A direct object of an atelic verb can be marked by the postposition es@, markingis preserved with nominalization and under ellipsis.(10) a. Sonshesˇuv-sdig-pst.3[sg]tjEthislotk-tjhole-def.sg.genes@in.iniandlotka-sj.spot-pst.3[sg]‘She was digging this hole and then stopped’. (Kozlov, 2018, 423)b. [TjEthiszadacˇa-t’task-def.sg.genes@in.inkuvakalongaz-@n-ksˇnj@-ma-sj]say-freq-freq-nzr-def.sgizjneg.pst[3sg]pomaga.help.cn‘This long explanation of the task didn’t help.’ (Zakirova, 2018)c. [TjEthiszadacˇa-tjtask-def.sg.genes@in.inkuvaka-sj]long-def.sgizjneg.pst[3sg]pomaga.help.cn‘{Context: Did you read explanations?} The long [explanation] of this taskdid not help.’Conclusion: Diagnostics show that the elided noun is syntactically present.2.3 Restrictions on inflection• There are two types of nominal modifiers in Moksha. Modifiers of the first typeshow inflection under ellipsis. Modifiers of the second type are not inflected.• The first type can be exemplified by adjectives, numerals, participles, or modifiersmarked for the indefinite genitive1 (see the full list in the appendix):– Adjective(11) MonImaks-@njgive-pst.1sg[kodam@whichb@dj@indefaksˇ@-njdji]white-dat‘{Context: To which cat did you give food?} I gave to a white one.’– Indefinite genitive(12) Minjwerama-sjkbuy-pst.3.o.3pl.s[pona-njnj@-tj].wool-gen-def.sg.gen‘{Context: Which hat did you buy?} We bought the woolen hat.’• Definite genitive and lative are among modifiers that do not show inflection(again see appendix for the full list).• These modifiers can still license inflection: inflection ‰ licensing of ellipsis– Definite genitive:(13) TjEthisava-tjwoman-def.sg.gen( / *ava-tj-@twoman-def.sg.gen-pl/*ava-tj-@nz@)woman-def.sg.gen-3sg.poss.plasˇcˇ-˚j-tjbe-npst.3-plmorksˇ-tjtabledef.sg.genlank-s@on-in‘{Context: Whose books are on the shelf? I don’t know} This woman’s [books]are on the table’.1 The genitive marker in the presence of the noun is -(@)nj, but it is -(@)njnj@- before inflection ofthe elided noun. The geminated allomorph is used when a genitive exponent is not word-final2– Lative:(14) Sonshear˚t-@zj@paint-pst.3sg.o.3sg.s[sportzal-u]gym-lat( / *sportzal-u-tj)gym-lat-def.sg.genravzˇ@blackkraska-s@.paint-in‘{Context: Which door did she paint black?} She painted [the door] to the gymblack.’What derives the split between inflecting and non-inflecting modifiers?• In languages with regularly overt concord, modifiers that have their own φ-features cannot agree with the noun (see Baker (2008)).– The inherent φ-features intervene and block agreement with another noun.(15) Generalization:A modifier is inflected under ellipsis unless it has its own φ-features.• Adjectives or numerals do not have their own φ-features and they get inflected.• Definite genitive and lative are nouns with their own features and they cannotshow inflection.• Potential complication: indefinite genitive.• I suggest that this form lacks φ-features and functions as an attributivizer.2– Indefinite genitive can be attached to adverbs, such as ‘yesterday’ and turnthem into nominal modifiers.(16) a. Sonshesa-sjcome-pst.3[sg]isjak.yesterday‘She came yesterday.’b. Sonsherama-zj@buy-pst.3sg.o.3sg.s[isjak-@njyesterday-genksˇi-tj].bread-def.sg.gen‘She bought yesterday’s bread.’• Additionally: Inflection under ellipsis correlates with predicative agreement.– Agreement in the predicative position is another property that follows fromthe presence / absence of φ-features (see Baker (2008) and a large-scalesurvey by Stassen (1992, 2005)).• Adjectives and indefinite genitive show number agreement in the predicativeposition:(17) Sjinjtheyjomla-t.small-pl‘They are small.’(18) Kud-tjnj@house-def.plsˇuft@-njnj@-t.wood-gen-pl‘The houses are wooden.’• Definite genitive and lative do not agree in number:(19) Kolj@ndj@ma-tjnj@toy-def.pltjEthisstj@rj-njE-tjgirl-dim-def.sg.gen/ *sjt’@rj-njE-tj-(@)tjgirl-dim-def.sg.gen-pl‘The toys are this girl’s.’2 The peculiarity of this form is reflected in Moksha grammars: Indefinite genitive is not includedin the list of cases in some descriptions of Moksha grammar; see Kolyadyonkov & Zavodova (1962,189-192) and Cygankin (1980, 112).(20) TjEthiski-tjnj@road-def.plvirj-iforest-lat/ *virj-i-t.forest-lat-pl‘These roads are to the forest.’Summary: Inflection under ellipsis is restricted as nominal concord:• Inflection appears on the head of the modifier.• Modifiers with their own φ-features are not infected.3 Existing approachesLicensing of pro (see Kester (1996a,b), see also Lobeck (1995))• Ellipsis site is occupied by pro and that pro has to be identified and licensed.The modifier agrees with pro to license it.Substantivization (see Bosˇkovic´ & S¸ener (2014))• Modifiers are substantivized and therefore marked for nominal features.Main problem:• The data in section 2.2 have shown that the ellipsis site contains a full-fledgednominal structure; i.e., there is no pro or nominalization of the remnant.Cliticization (see De´ka´ny (2011, 51-53, 2015), Lipta´k & Saab (2014), Ruda(2016), Saab & Lipta´k (2016), Murphy (2018), and Saab (2019)).• The Lowering of the number features is blocked by ellipsis.3• ‘Stranded’ affix is repaired by Local Dislocation (see Embick & Noyer (2001);Embick (2007)).(21) EllipsisDPNumPNumPnPnNumrpls?+nAPD5(22) Linearizationadjective * plÑ Local Disclocationadjective-plSome problems:• Approach does not capture inflection with complex modifiers.– Inflection is predicted to appear on the argument of the participle, ratherthan on the participle.3 See Georgieva et al. (2019) for an evidence against the assumption that ellipsis block Lowering.3• Inflection is over-generated on all nominal modifiers.(23) Summary: Existing approaches to inflection under ellipsispro nmn cliticizationInflection only under ellipsis 3 3 3On the head of the branching modifier 3 3 7Connectivity effects 7 7 3Ellipsis without inflection 7 3(?) 7Correlation to the predicative agreement 3 7 74 Ellipsis reveals concord• Existing approaches share a idea that a nominal modifier receives nominal featuresbecause the noun is absent. Inflection is necessary to satisfy some constraint.1. Moksha has nominal concord.2. Concord features are not spelled out if the noun is present.4.1 Concord• Nominal concord is derived by Agree.– See Carstens (2001, 2018), Baker (2008), Kramer (2009), Danon (2011),Toosarvandani & van Urk (2014), Landau (2016), Ingason & SigurDsson(2017), Pusˇkar (2017, 2018), pace Pesetsky (2013), Norris (2014, 2018), Baier(2015), Bayırlı (2017), Hanink (2018) and Ackema & Neeleman (2019)• AP-over-NP structure (see, e.g., Abney (1987), Bosˇkovic´ (2005), Murphy (2018),and Salzmann (2018)).• Number, case and definiteness features (inMoksha) originate in the n head.• Nominal modifiers have unvalued probes forthe corresponding features.4• Probes on a nominal modifier always targetthe features on the same node (i.e., on thenoun), I assume that they all probe together.• Probe features are indicated as [˚F˚] (follow-ing the notation in Heck & Mu¨ller (2007)).(24) Nominal concordAPnP?rootn„#:plκ:genδ:defAr˚C: ˚sLow origin of definiteness and case is not problematic (just looks unusual).• Definiteness: It originates on n, but it is theninterpreted on D (cf. Hankamer & Mikkelsen (2005)and Heck et al. (2009)).• Case: Heads that are traditionally conceived ofas case assigners have in fact an unvalued (or,alternatively unchecked) case feature.(25) Case assignmentvPVPDPrκ:accsVvr˚κ: ˚s4 See Wintner (2000), Kramer (2010) for other examples of definiteness agreement.4.2 Morphological realization• In Moksha, concord is not realized on a modifier if the noun is present.Spell-Out• Syntactic structure is spelled out in steps (see Chomsky (2000, 2001), and alsoUriagereka (1999)).• What constitutes the spell-out domain: C and v* ; also complements of thecategory-defining heads (see Marantz (2007) and Embick (2010) among others);each phrase (see Mu¨ller (2011)); each Merge induces Spell-Out (see Wojdak(2008) and Starke (2009)); each syntactic operation (see Epstein & Seely (2002))• I pursue a local approach to Spell-Out:(26) Spell-Out:Spell-Out applies to a node that has no unsatisfied features, where a featurecounts as unsatisfied if it can induce operations.• Features that trigger Agree ([˚F: ˚]) and Merge ([‚F‚]) count as unsatisfied.These features are satisfied after the operations that they bring about apply.(27) No Spell-Out of XXPYPrF:αsXr˚F: ˚s(28) Spell-Out applies to XXPYPrF:αsXrF:αsPF• Spell-Out domains do not correspond to syntactically inaccessible domains (see,e.g., Dobler et al. (2011), Piggott & Travis (2017), Martinovic´ (2019), and alsoChomsky (2008, 143)).• There are different ways of deriving opacity in syntax without appealing toSpell-Out; see Rackowski & Richards (2005), Mu¨ller (2011), and Keine (2019)for some options.Probe Conversion• A life cycle of a probe includes two operations:5 Valuation and Conversion.6• Probes are valued by Agree.• After this they are subject to Probe Conversion:5 Another splitting up of Agree was proposed by Arregi & Nevins (2012). They suggest that Agreeis a two-step process that consists of Agree-Link and Agree-Copy.6 Analyzing the mechanism for the deletion of uninterpretable features suggested by Chomsky(2001), Epstein & Seely (2002) come to the conclusion that it requires probes to be differentfrom originally valued features shortly after valuation. Chomsky suggests that unvalued featurescorrespond to uninterpretable features and have to be deleted before transfer to LF. Deletionshould apply after valuation, but originally unvalued features should still be detectable, so that itcan be ensured that the right kinds of features undergo deletion.4(29) Probe Conversion:Probe Conversion applies to valued probes and makes them identical to originallyvalued features.(30) AgreeXPYPrF:αsXr˚F: ˚s(31) ValuationXPYPrF:αsXr˚F:α˚s(32) Probe ConversionXPYPrF:αsXrF: αs• Features that trigger internal and external Merge are not morphologically realized.• I suggest that this property is generalized over all operation inducing features:– Only converted probes can be morphologically realized.Ellipsis• [E]-features responsible for different types of ellipsis have a different featurespecifications (see Merchant (2001, 2005)).• Nominal ellipsis triggering [E] has an unchecked nominal feature [E[˚n˚]]that ensures the local presence of a noun.(33) Under ellipsisAPnPrF:αsArE[˚n˚]s(34) [E]-licensingAPnPrF:αsArE[˚n˚]s(35) Ellipsis of nPAPnPA4.3 AnalysisNo ellipsis• A nominal modifier has an unvalued [˚C: ˚]• After its agreement with the noun, there are not unsatisfied features on themodifier ÝÑ Spell-Out can apply• Concord probe is not converted at this point, which means that it is not subjectfor Vocabulary Insertion.• This generates an absence of concord exponents in non-elliptical con-texts in Moksha.(36) Step I: Unvalued CAPnP?rootn„#:plκ:genδ:defAr˚C: ˚s(37) Step II: AgreeAPnP?rootn„#:plκ:genδ:defAr˚C:pl/gen/def˚s(38) Step III: Spell-OutAPnP?rootn„#:plκ:genδ:defAr˚C:pl/gen/def˚sPF• C undergoes Probe Conversion, but this comes too late to feed realization.• After Conversion, a probe can serve as a goal for further agreement.(39) Step IV: Probe ConversionAPnP?rootn„#:plκ:genδ:defArC:pl/gen/defs(40) Step V: Further AgreeAPAPnP?rootn„#:plκ:genδ:defArC:pl/gen/defsAr˚C: ˚sUnder ellipsis• Concord exponents are present under ellipsis because a concord probe is notthe last unsatisfied feature on a nominal modifier.• If the noun is elided, its modifier, bears an [E]-feature with unchecked nominalsub-feature.• I assume that features on a nominal modifier are ordered: C probes first.5(41) Step I: Unvalued CUnchecked [E[˚n˚]]APnP?rootn„#:plκ:genδ:defA”˚C: ˚E[˚n˚]ı(42) Step II: AgreeAPnP?rootn„#:plκ:genδ:defA”˚C:pl/gen/def˚E[˚n˚]ıCprobes• When a concord feature is valued by Agree and thus satisfied, there is still an[E]-feature present.• Its presence prevents application of Spell-Out immediately after [C] is valued.• At the next step, [C] is converted, and then derivation proceeds to the nextunsatisfied feature on the nominal modifier.(43) Step III: Probe ConversionAPnP?rootn„#:plκ:genδ:defA”C:pl/gen/defE[˚n˚]ı(44) Step IV: [E]-licensingAPnP?rootn„#:plκ:genδ:defA”C:pl/gen/defE[˚n˚]ı(45) Step V: Spell-OutAPnP?rootn„#:plκ:genδ:defA”C:pl/gen/defE[˚n˚]ıPF• This produces overt nominal concord in elliptical contexts.Other properties1. Concord is overtly realized only on the linearly last remnant:(46) MonIand-inj@feed-pst.3.o.1sg.s[maziniceaksˇ@-tj]white-def.sg.gen/ *[mazi-tjnice-def.sg.genaksˇ@]white/*[mazi-tjnice-def.sg.genaksˇ@-tj].white-def.sg.gen‘{Which cat did you feed?} I fed the beautiful white one.’• This restriction follows from requirements on ellipsis licensing:– One [E]-feature is enough to trigger ellipsis of the noun.– It immediately precedes the ellipsis site (see Merchant (2001, 2005) andAelbrecht (2011)).(47) Higher adjectiveAPAPnP?rootn„φ:sgκ:genδ:defA”C:sg/gen/defE[˚n˚]ıAr˚C ˚s2. Inflection appears on the head of the branching modifiers (recall the examplewith a participle phase).(48) MonIrama-jnj@buy-pst.3.o.1sg.sa. [keluv-@njbirch-genlopa-st@leaf-elti-f-tj]make-ptcp.res-def.sg.genb. [ti-f-tjmake-ptcp.res-def.sg.genkeluv-@njbirch-genlopa-st@]leaf-elc.*[ti-fmake-ptcp.reskeluv-@njbirch-genlopa-st@-tj].leaf-el-def.sg.gen‘{Context: Which liquor did you buy?} I bought the [liquor] made from birchleafs.’• Complex modifiers were argued to be challenging for AP(/PartP/NumeralPetc.)-over-NP (see Kayne (1994), Alexiadou & Wilder (1998), Cinque (2010),and Roehrs (2018)).6• The modifier first combines with itsargument and then with the noun.• Probes from Part head project toPart1 that c-commands nP (see, e.g,Be´jar & Rezac (2009), Carstens(2016), and Keine & Dash (2019) onprobe projection)• The directionality of branching inPartP is not fixed. This allows theparticiple to be located before or afterits argument.(49) [ participle argument ]PartPnP?rootn„φ:sgκ:genδ:defPart1vPDP V+vPartr˚C: ˚s3. Modifiers that have their own φ-features cannot show concord inflection underellipsis.• Baker (2008) shows that the restriction on inflection is due to intervention:Concord probes will always encounter the features on the modifier first.• The same logic is applicable here. Independently of the exact position of theprobes within the modifying DP, they will first encounter the features fromwithin this DP.5 Conclusions5.1 Cross-linguistics variation• There are two types of languages with nominal concord:– Concord exponents are always presentEstonian, German, Russian etc.– Concord is morphologically realized only if the noun is elidedMoksha and potentially other languages with inflecting ellipsis, i.e. Hungar-ian, Turkish, Ossetic etc.Why concord is always present in languages of the first type?• I assume that the order of some operations is not universally determined, and canbe fixed language-specifically (see Georgi (2014, 2017), Assmann et al. (2015),and Murphy & Pusˇkar (2018))– In Moksha, Spell-Out can apply between two steps of Agree, i.e., afterValuation and before Probe Conversion.– In Russian, Valuation and Probe Conversion cannot be separated by Spell-Out.(50) Morphological realization of concord exponentsnoun present noun elidedI. Spell-Out follows AgreeRussian-type` `II. Spell-Out splits AgreeMoksha-type´ `5.2 Valuation does not imply realization• It is often assumed that features can be present in syntax but not morphologicallyrealized.• The analysis here is an attempt to develop to a more principled approach tonon-realization of features present in syntax.• It is based on two assumptions:1. Spell-Out is local. It applies to a node that has no unsatisfied features,where a feature counts as unsatisfied if it can induce syntactic operations(Agree or Merge).2. Agree is split into Valuation and Probe Conversion.– After Valuation, a probe has a value.– After Conversion, it is morphologically realized and can be targeted byfurther Agree.ReferencesAbney, S. 1987. The English Noun Phrase in its Sentential Aspect. Doctoral thesis, MassachusettsInstitute of Technology, Boston.Ackema, P. & A. Neeleman. 2019. Unifying Nominal and Verbal Syntax: Agreement and FeatureRealisation. Ms. University of Edinburgh, UCL. Available at https://ling.auf.net/lingbuzz/004921.Aelbrecht, L. 2011. The Syntactic Licensing of Ellipsis. Amsterdam: John Benjamins.Alexiadou, A. & K. Gengel. 2012. NP ellipsis without focus movement/projections: the role ofClassifiers, 177–205. Cambridge University Press.Alexiadou, A. & C. Wilder. 1998. Adjectival modification and multiple determiners. In Possessors,Predicates and Movement in the Determiner Phrase, eds. A. Alexiadou & C. Wilder, 303–332.Amsterdam: John Benjamins.Arregi, K. & A. Nevins. 2012. Morphotactics: Basque Auxiliaries and the Structure of Spellout.Dordrecht: Springer Netherlands.Assmann, A., D. Georgi, F. Heck, G. Mu¨ller & P. Weisser. 2015. Ergatives move too early: On aninstance of opacity in syntax. Syntax 18:343–387.Babby, L. H. 1975. A Transformational Grammar of Russian Adjectives. Mouton, The Hague.Berlin, Boston: De Gruyter Mouton.Babby, L. H. 2009. The Syntax of Argument Structure. Cambridge: Cambridge University Press.Baier, N. 2015. Adjective Agreement in Noon: Evidence for a Split Theory of Noun-ModifierConcord. In Proceedings of the 45th Meeting of the North East Linguistic Society, volume I,eds. T. Bui & D. O¨zyIldIz, 67–80. Amherst: GLSA Publications.Bailyn, J. F. 2012. The Syntax of Russian. Cambridge: Cambridge University Press.Baker, M. 2008. The Syntax of Agreement and Concord. New York: Cambrige University Press.Bayırlı, I˙. K. 2017. The universality of concord. Doctoral thesis, Massachusetts Institute ofTechnology, Boston.Be´jar, S. & M. Rezac. 2009. Cyclic agree. Linguistic Inquiry 40:35–73.Bond, O. & M. Chumakina. 2016. Agreement domains and targets. In In Archi: Complexitiesof agreement in cross-theoretical perspective, eds. O. Bond, G. G. Corbett, M. Chumakina &D. Brown, 43–76. Oxford: Oxford University Press.Borik, O. 2014. The argument structure of long and short form adjectives and participles in Russian.Lingua 149:139 – 165. SI: Aspect and Argument Structure of Adjectives and Participles.Bosˇkovic´, Zˇ. 2005. On the locality of left branch extraction and the structure of NP. StudiaLinguistica 59:1–45.Bosˇkovic´, Zˇ. 2014. Now I’m a Phase, Now I’m Not a Phase: On the Variability of Phases withExtraction and Ellipsis. Linguistic Inquiry 45:27–89.Bosˇkovic´, Zˇ. & S. S¸ener. 2014. The Turkish NP. In Crosslinguistic studies on nominal reference:With and Without Articles, eds. P. C. Hofherr & A. Zribi-Hertz, 102–140. Leiden: Brill.Carstens, V. 2001. Multiple agreement and case deletion: Against φ-incompleteness. Syntax4:147–163.Carstens, V. 2016. Delayed Valuation: A Reanalysis of Goal Features, “Upward” ComplementizerAgreement, and the Mechanics of Case. Syntax 19:1–42.Carstens, V. 2018. Concord and labeling. To appear in Agree to Agree: Agreement in the MinimalistProgramme, eds. P. W. Smith, K. Hartmann, and J. Mursell.7Chomsky, N. 1973. Conditions on transformations. In A Festschrift for Morris Halle, eds.S. Anderson & P. Kiparsky, 232–286. New York: Academic Press.Chomsky, N. 1995. The Minimalist Program. Cambridge: The MIT Press.Chomsky, N. 2000. Minimalist Inquiries: The Framework. In Step by Step, eds. R. Martin,D. Michaels & J. Uriagereka, 89–155. Cambridge: MIT Press.Chomsky, N. 2001. Derivation by phase. In Ken Hale. A Life in Language, ed. M. Kenstowicz,1–52. Cambridge: MIT Press.Chomsky, N. 2008. On Phases. In Foundational Issues in Linguistic Theory: Essays in Honor ofJean-Roger Vergnaud, eds. R. Freidin, C. P. Otero & M. L. Zubizarreta, 133–166. Cambridge:MIT Press.Chomsky, N. 2019. Lectures. Paper presented at University of California, Los Angeles, April 29 –May 2.Cinque, G. 2010. The Syntax of Adjectives: A Comparative Study. Cambridge: MIT Press.Cinque, G. 2012. A generalization concerning DP-internal ellipsis. Iberica: An InternationalJournal of Theoretical Linguistics 4:174–193.Citko, B. 2011. Multidominance. In The Oxford Handbook of Linguistic Minimalism, ed. C. Boeckx,119–142. Oxford: Oxford University Press.Corver, N. & M. van Koppen. 2009. Let’s focus on noun phrase ellipsis. Groninger Arbeiten zurGermanistischen Linguistik 48:2–26.van Craenenbroeck, J. & J. Merchant. 2013. Ellipsis phenomena, 701–745. Cambridge Handbooksin Language and Linguistics, Cambridge University Press.Cygankin, D. V. 1980. Grammatika mordovskih yazykov. Fonetika. Grafika. Orfografiya. Mor-fologiya. [Grammar of the Mordvin languages. Phonetics. Graphics. Orthogrphy. Morphology.].Saransk: Mordovskij gosudarstvennyj universitet.Danon, G. 2011. Agreement and DP-Internal Feature Distribution. Syntax 14:297–317.De´ka´ny, E´. 2011. A profile of the Hungarian DP: the interaction of lexicalization, agreement,and linearization with the functional sequence. Doctoral thesis, CASTL, University of Tromsø,Tromsø.De´ka´ny, E´. 2015. The syntax of anaphoric possessives in Hungarian. Natural Language & LinguisticTheory 33:1121–1168.Dobler, E., H. Newell, G. L. Piggott, T. Skinner, M. Sugimura & L. d. Travis. 2011. Narrow syntacticmovement after spell-out. Ms. McGill University, Montreal.Egorova, A. D. 2018. Sentencial’nye aktanty [sentential actants]. In Elementy mokshanskogo yazykav tipologicheskom osveshchenii [Elements of Moksha language in the typological perspective],eds. S. I. Toldova & M. A. Kholodilova, 666–706. Moskva: Buki Vedi.Embick, D. 2007. Linearization and local dislocation: Derivational mechanics and interactions.Linguistic analysis 33:303–336.Embick, D. 2010. Localism versus globalism in morphology and phonology. Cambridge: MIT Press.Embick, D. & R. Noyer. 2001. Movement Operations after Syntax. Linguistic Inquiry 32:555–595.Epstein, S. D. & T. D. Seely. 2002. Rule Applications as Cycles in a Level-Free Syntax, chapter 3,65–89. John Wiley & Sons, Ltd.Frampton, J. & S. Gutmann. 2000. Agreement is feature sharing. Ms. Northeastern University,Boston.Frampton, J. & S. Gutmann. 2006. How sentences grow in the mind: Agreement and selectionin an efficient minimalist syntax. In Agreement Systems, ed. C. Boeckx, 121–157. LinguistikAktuell/Linguistics Today 92, Amsterdam: John Benjamins.Geist, L. 2010. The argument structure of predicate adjectives in Russian. Russian Linguistics34:239–260.Georgi, D. 2014. Opaque interactions of Merge and Agree: On the nature and order of elementaryoperations. Doctoral thesis, University of Leipzig, Leipzig.Georgi, D. 2017. Patterns of movement reflexes as the result of the order of merge and agree.Linguistic Inquiry 48:585–626.Georgieva, E., M. Salzmann & P. Weisser. 2019. Ellipsis does not bleed Lowering - Evidence fromdo-support and fragment answers in Finno-Ugric. Paper presented at North East LinguisticsSociety (NELS 50), Massachusetts Institute of Technology, Boston, October 25–27.Ghaniabadi, S. 2010. The Empty Noun Construction in Persian. Doctoral thesis, University ofManitoba, Winnipeg.Hanink, E. A. 2018. Structural Sources of Anaphora and Sameness. Doctoral thesis, University ofChicago, Chicago.Hankamer, J. & L. Mikkelsen. 2005. When Movement Must Be Blocked: A reply to Embick andNoyer. Linguistic Inquiry 36:85–125.Heck, F. & G. Mu¨ller. 2007. Extremely local optimization. In WECOL 34: Proceedings of the 34thWestern Conference on Linguistics, eds. E. Brainbridge & B. Agbayani, 170–183. CaliforniaState University: Fresno.Heck, F., G. Mu¨ller & J. Trommer. 2009. A phase-based approach to Scandinavian definiteness mark-ing. STUF – Language Typology and Universals Sprachtypologie und Universalienforschung64:258–268.Hettich, B. G. 2002. Ossetian: Revisiting Inflectional Morphology. Master’s thesis, University ofNorth Dakota.Ingason, A. K. & E. F. SigurDsson. 2017. The interaction of adjectival structure, concord andaffixation. In Proceedings of the 47th Meeting of the North East Linguistic Society, eds.A. Lamont & K. Tetzloff, 89–98. Amherst: GLSA Publications.Johnson, K. 2017. Rethinking linearization. Ms. University of Massachusetts at Amherst.Kashkin, E. V. 2018. Opredelennost’ [definiteness]. In Elementy mokshanskogo yazyka v tipologich-eskom osveshchenii [Elements of Moksha language in the typological perspective], eds. S. I.Toldova & M. A. Kholodilova, 122–153. Moskva: Buki Vedi.Kayne, R. 1994. The antisymmetry of syntax. Cambridge: MIT Press.Keine, S. 2019. Selective Opacity. Linguistic Inquiry 50:13–62.Keine, S. & B. Dash. 2019. The ups and downs of agreement. Ms. University of Southern California,Los Angeles.Kester, E.-P. 1996a. Adjectival Inflection and the Licensing of Empty Categories in DP. Journal ofLinguistics, 32:57–78.Kester, E.-P. 1996b. The nature of adjectival inflection. Doctoral thesis, Utrecht University, Utrecht.Kholodilova, M. 2016. Moksha Non-Verbal Predication. In Mordvin languages in the field, eds.K. Shagal & H. Arjava, 229–259. Uralica Helsingiensia 10, Helsinki: University of Helsinki(Department of Finnish, Finno-Ugrian and Scandinavian Studies) & Finno-Ugrian Society.Kholodilova, M. A. 2018. Imennaja predikatsija [nominal predication]. In Elementy mokshan-skogo yazyka v tipologicheskom osveshchenii [Elements of Moksha language in the typologicalperspective], eds. S. I. Toldova & M. A. Kholodilova, 616–632. Moskva: Buki Vedi.Kolyadyonkov, M. N. & R. A. Zavodova. 1962. Grammatika mordovskih (mokshanskogo i erzyan-skogo) yazykov., volume I. Saransk: Mordovskoe knizhnoe izdatel’stvo.Kozlov, A. A. 2018. Akcional’naya semantika osnov i derivacionnyj vid [aktionsart of verbal stemsand derivational aspect]. In Elementy mokshanskogo yazyka v tipologicheskom osveshchenii[Elements of Moksha language in the typological perspective], eds. S. I. Toldova & M. A.Kholodilova, 417–437. Moskva: Buki Vedi.Kramer, R. 2009. Definite Markers, Phi Features and Agreement: A Morphosyntactic Investigationof the Amharic DP. Doctoral thesis, University of California, Santa Cruz, Santa Cruz.Kramer, R. 2010. The Amharic Definite Marker and the Syntax–Morphology Interface. Syntax13:196–240.Landau, I. 2016. DP-internal semantic agreement: A configurational analysis. Natural Language &Linguistic Theory 34:975–1020.Lipta´k, A. & A. Saab. 2014. No N-raising out of NPs in Spanish: ellipsis as a diagnostic of headmovement. Natural Language & Linguistic Theory 32:1247–1271.Lobeck, A. 1995. Ellipsis: Functional Heads, Licensing, and Identification. New York: Oxforduniversity Press.Marantz, A. 2007. Phases and words. In Phases in the theory of grammar, ed. S.-H. Choe, 199–222.Seoul: Dong In.Martinovic´, M. 2019. Interleaving syntax and postsyntax: Spellout before syntactic movement.Syntax 22:378–418.Matushansky, O. 2004. Going through a phase. In Perspectives on phases, eds. M. McGinnis &N. Richards, 157–181. MIT Working Papers in Linguistics 49, Cambridge: MIT, MIT WorkingPapers in Linguistics.Merchant, J. 2001. The syntax of silence: Sluicing, islands, and the theory of ellipsis. Oxford:Oxford University Press.Merchant, J. 2005. Fragments and ellipsis. Linguistics and Philosophy 27:661–738.Merchant, J. 2014. Gender mismatches under nominal ellipsis. Lingua 151:9–32. StructuralApproaches to Ellipsis.Merchant, J. 2019. Ellipsis: A survey of analytical approaches. In The Oxford Handbook of Ellipsis,eds. J. van Craenenbroeck & T. Temmerman, 19–45. Oxford: Oxford University Press.Mu¨ller, G. 2011. Constraints on Displacement. A Phase-Based Approach. Amsterdam: Benjamins.Murphy, A. 2018. Pronominal inflection and NP ellipsis in German. The Journal of ComparativeGermanic Linguistics 21:327–379.Murphy, A. & Z. Pusˇkar. 2018. Closest conjunct agreement is an illusion. Natural Language &Linguistic Theory 36:1207–1261.Norris, M. 2014. A Theory of Nominal Concord. Doctoral thesis, University of California, SantaCruz, Santa Cruz.Norris, M. 2018. Unmarked case in Estonian nominals. Natural Language & Linguistic Theory36:523–562.Pesetsky, D. 2013. Russian Case Morphology and the Syntactic Categories. Cambridge: MITPress.Pesetsky, D. & E. Torrego. 2007. The syntax of valuation and the interpretability of features. InPhrasal and Clausal Architecture: Syntactic derivation and interpretation. In honor of JosephE. Emonds, eds. S. Karimi, V. Samiian & W. K. Wilkins, 262–294. Linguistik Aktuell/LinguisticsToday 101, Amsterdam: John Benjamins.8Piggott, G. & L. Travis. 2017. Wordhood and word-internal domains. In The Structure of Words atthe Interfaces, eds. H. Newell, M. Noonan, G. Piggott & L. d. Travis, 41–74. New York: OxfordUniversity Press.Pusˇkar, Z. 2017. Hybrid agreement: modelling variation, hierarchy effects and phi-feature mismatches.Doctoral thesis, Universita¨t Leipzig, Leipzig.Pusˇkar, Z. 2018. Interactions of Gender and Number Agreement: Evidence fromBosnian/Croatian/Serbian. Syntax 21:275–318.Rackowski, A. & N. Richards. 2005. Phase edge and extraction: A tagalog case study. LinguisticInquiry 36:565–599.Roehrs, D. 2018. Adjectives are in phrasal positions. Ms. University of North Texas, Denton.Ruda, M. 2016. NP ellipsis (effects) in Polish and Hungarian: FFs on Fs, Agree, and ChainReduction. The Linguistic Review 33:649–677.Saab, A. 2019. Nominal ellipsis. In The Oxford Handbook of Ellipsis, eds. J. van Craenenbroeck &T. Temmerman, 526–561. Oxford: Oxford University Press.Saab, A. & A. Lipta´k. 2016. Movement and deletion after syntax: Licensing by inflection reconsidered.Studia Linguistica 70:66–108.Salzmann, M. 2018. Revisiting the NP vs. DP debate. Ms. Leipzig University, Leipzig.Starke, M. 2009. Nanosyntax: A short primer to a new approach to language. Nordlyd 36:1–6.Stassen, L. 1992. A hierarchy of predicate encoding. In Meaning and Grammar: Cross-linguisticperspectives, eds. M. Kefer & J. van der Auwera, 179–201. Berlin: Mouton de Gruyter.Stassen, L. 2005. Predicative Adjectives. In The World Atlas of Language Structures, eds.M. Haspelmath, M. S. Dryer, D. Gil & B. Comrie, 478–481. New York – Oxford: OxfordUniversity Press.Svenonius, P. 2004. On the edge. In Peripheries: Syntactic edges and their effects, eds. D. Adger,C. de Cat & G. Tsoulas, 261–287. Dordrecht: Kluwer.Toosarvandani, M. & C. van Urk. 2014. Agreement in Zazaki and the nature of nominal concord.Ms. Univercity of California, Santa Cruz and Massachusetts Institute of Technology, Boston.Uriagereka, J. 1999. Multiple spell-out. In Working minimalism, eds. S. D. Epstein & N. Hornstein,251–282. Cambridge: MIT Press.Wintner, S. 2000. Definiteness in the Hebrew noun phrase. Journal of Linguistics 36:319–363.Wojdak, R. 2008. The Linearization of Affixes: Evidence from Nuu-chah-nulth. Dordrecht:Springer Netherlands.Zakirova, A. N. 2018. Nominalizacii [nominalizations]. In Elementy mokshanskogo yazyka vtipologicheskom osveshchenii [Elements of Moksha language in the typological perspective], eds.S. I. Toldova & M. A. Kholodilova, 753–778. Moskva: Buki Vedi.6 Appendix A: More on Moksha• Direct objects are genitive or unmarked. Verbs agree with marked direct objects(cf. (51a)) and do not agree with unmarked direct objects (cf. (51b)).(51) a. MonInjEj-sasee-npst.3sg.o.1sg.sknjiga-tjbook-def.sg.gen/ *knjiga.bookb. MonInjEj-ansee-npst.1sgknjigabook/ *knjiga-tj.book-def.sg.gen‘I see a / the book.’Noun phrase(52) Part of the Moksha nominal paradigm illustrated by the noun velj@ ‘village’Indefinite declension Definite declensionsg pl sg plnominative velj@ velj@-t velj@-sj velj@-tjnj@genitive velj@-nj velj@-tj velj@-tjnj@-njdative velj@-njdji velj@-tji velj@-tjnj@-njdjiablative velj@-d@inessive velj@-s@elative velj@-st@• Nominal modifiers like adjectives, numerals and demonstratives are obligatorilyprepositional:(53) MonInjEj-inj@see-pst.3.o.1sg.s[ravzˇ@blackpinj@-tj]dog-def.sg.gen/ *[pinj@-tjdog-def.sg.genravzˇ@].black‘I saw the black dog.’• Possessors and arguments are usually prepositional, but postposition is possibleas well:(54) KoljEKoliakep@dj-@zj@grab-pst.3sg.o.3sg.s[tjEthisava-tjwoman-def.sg.gensumka-nc]bag-3sg.poss.sg.gen/ [sumka-ncbag-3sg.poss.sg.gentjEthisava-tj].woman-def.sg.gen‘Kolia grabbed this woman’s bag.’Ellipsis• Exponents that appear on the nominal modifier under ellipsis can differ fromexponents on the noun in the corresponding non-elliptical context.(55) a. MonIsoda-saknow-npst.3sg.o.1sg.s[tjEthisava-tj].woman-def.sg.gen‘I know this woman.’b. MonIsoda-saknow-npst.3sg.o.1sg.s[tjE-nj]this-gen/ *[t’E-tj].this-def.sg.gen‘{Which of these women do you know?} I know this one.’Non-verbal predication• In the third person present tense, adjectives cannot take verbal tense inflection,but agree with a third person subject in number. (see Kholodilova (2016, 2018)on a detailed description of non-verbal predication in Moksha).(56) Sonhejomlasmall/ *jomla-j.small-npst.3[sg]‘He is small.’(57) Sjinjtheyjomla-tsmall-pl/ *jomla-˚j-tj.small-npst.3-pl‘They are small.’• If the subject is a first or second person pronoun or if the predication has referenceto the past, the predicate is obligatorily marked for tense.• Agreement for number and person then does not depend on φ-features on thenon-verbal predicate. This due to the tense marking.• The T head that is higher than the subject is responsible for the predicativeagreement, so that the subject is the closest goal for agreement, and features onthe non-verbal predicate cannot intervene (see also Baker (2008, 56-63)).(58) Minj tjE ucˇitj@l˚j-nj@-tam@.we this teacher-def.pl-npst.1pl‘We are these teachers.’(59) Minj tonj ucˇitj@l˚j-nj@-lj-@m@.we you.gen teacher-def.pl-impf-pst.1pl‘We were your teachers.’9Appendix B: Restrictions on modifiersInflection under ellipsis – yes• Adjectives, numerals, demonstratives, participles, indefinite genitive, and elativemarked modifiers (see examples above).• Nouns without a case marker:(60) Panjcˇf-tflower-plrama-sjbuy-pst.3[sg][senj@mbluesjeljm@-sj].eye-def.sg‘{Context: Which girl bought flowers?} The [girl] with blue eyes bought flowers.’• Caritive:(61) Sonhemaksjgive.pst.3[sg][zonjtjik-ft@m@-tji].umbrella-car-def.sg.dat‘{Context: To whom did he give his coat?} He gave to the [person] without anumbrella’.Inflection – no• Examples with definite genitive and lative are given above.• Definite dative:(62) MonInjEj-sasee-npst.3sg.o.1sg.s[virj-tjiforest-def.sg.datki-tj]road-def.sg.gen/*[virj-tji-tj].forest-def.sg.dat-def.sg.gen‘{Context: Which road do you see?} I see [the road] to the forest.’• Non-definite dative:(63) MonIjuma-ft-inj@disappear-caus-pst.3.o.1.sg.s[kodam@whichb@dj@indefsjtj@rj-njE-njdjigirl-dim-datkazjnj@-tj]present-def.sg.gen/ *[sjtj@rj-njE-njdji-tj].girl-dim-dat-def.sg.gen‘{Which present did you loose?} I lost [a present] for some girl.’Non-verbal predication• Inflection under ellipsis correlates with agreement in the predicative position:– Inflection under ellipsis ÐÑ Agreement in the predicative position– No inflection under ellipsis ÐÑ No agreementAgreement – yes• Genitive of the indefinite declension:(64) Kud-tjnj@house-def.plsˇuft@-njnj@-t.wood-gen-pl‘The houses are wooden.’• Caritive:(65) TjEthiskaza-tjnj@goat-def.plsjura-ft@m@-t.antler-car-pl‘The goats are without antlers.’• Elative:(66) TjEthisnastojka-tjnj@liquor-def.plkeluv-@njbirch-genlopa-st@-t.leaf-el-pl‘These liquors are from birch leafs.’Agreement – no• Genitive of the definite declension:(67) Kolj@ndj@ma-tjnj@toy-def.pltjEthisstj@rj-njE-tjgirl-dim-def.sg.gen/ *sjt’@rj-njE-tj-(@)tjgirl-dim-def.sg.gen-pl‘The toys are this girl’s.’• Dative of the definite declension:(68) Kol@ndj@ma-tjnj@toy-def.pltjEthissjtj@rj-njE-tjigirl-dim-def.sg.dat/ *sjtj@rj-njE-tji-t.girl-dim-def.sg.dat-pl‘The toys are for this girl.’• Dative of the indefinite declension:(69) Kolj@ndj@ma-tjnj@toy-def.plkodam@whichb@dj@indefsjtj@rj-njE-njdjigirl-dim-dat/ *sjtj@rj-njE-njdji-t.girl-dim-dat-pl‘The toys are for some girl.’• Lative:(70) TjEthiski-tjnj@road-def.plvirj-iforest-lat/ *virj-i-t.forest-lat-pl‘These roads are to the forest.’Summary• Modifiers that show inflection under ellipsis, also show number agreement in thepredicative position.• Agreement is ungrammatical for forms that are not inflected under ellipsis.(71) Inflection on an element under ellipsis and in the predicative positionUnder ellipsis In predicative positionAdjective yes yesIndefinite genitive yes yesCaritive yes yesElative yes yesUnmarked noun yes yesDefinite genitive no noDefinite dative no noIndefinite dative no noLative no no10Against the silent noun analysis• Babby (1975; 2009, 93-110) and Bailyn (2012, 68-70) suggest that adjectives inthe predicative position modify a silent noun.7(72) [ modifier Ønoun ]• If so, restrictions on agreement in the predicative position can be reduced torestrictions on inflection under ellipsis.Empirical evidence against the presence of null noun:1. No usage that is restricted to adnominal modification• A form marked for elative can be used in the adnominal position to markclothes.• Such use of the elative form is ungrammatical otherwise.(73) a. Sjinjtheysenj@mbluepanar-st@dress-elsjtj@rj-njE-tjnj@.girl-dim-def.pl‘They are the girls in blue dresses.’b.*Sjtj@rj-njE-sjgirl-dim-def.sgsa-sjcome-pst.3[sg]sen’@mbluepanar-st@.dress-el‘The girl came in the blue dress.’c.*Sjinjtheysenj@mbluepanar-st@dress-el/ *panar-st@-t.dress-el-pl‘They are in blue dresses.’This restriction is unexpected if the elative form modifies a silent noun.2. Differences in inflection• Inflection may differ from the one that is expected in an elliptical context.(74) SjinjtheycjebErjgooddokt@r˚j-n’@.doctor-def.pl‘They are the good doctors.’(75) SjinjtheycjebEr˚j-tjgood-pl/ *cjebEr˚j-nj@.good-def.pl‘They are good.’Conclusion: Number inflection cannot result from ellipsis. It is subject agreement.Appendix C: Ways to rescue cliticization• The cliticization analysis as it stands over-generates inflection on all modifiers.• One might think that their analysis may be easily fixed by adding of somerestrictions on the positioning of affixes.• What how such restriction could look like? Given that the necessity for the newhost arises only quite late at PF1. Something like ‘One case rule’ (see Pesetsky (2013)) would still not derive thedata though, because there are not necessary 2 case affixes:7 The analysis is designed to account for differences between long and short form adjectives inRussian; see Geist (2010) and Borik (2014) for some empirical shortcomings of this analysis.(76) [Sportzal-u]gym-lat/ *[sportzal-uf-tgym-lat-plar˚t-f-t]paint-ptcp.ressEngErjEgreenkraska-s@.paint-in‘{Context: Those doors are red} and [the doors] to the gym are painted green.’2. A filter that prohibits two sets of φ-features from different noun phrases tobe realized within one phonological word will also block subject and objectagreement:(77) SonshenjEj-@zj-nj@see-pst.3sg.s.3pl.otjEthislomatj-tjnj@-nj.people-def.pl-gen‘She saw these people.’Appendix D: ConcordAgree derives concord• Above I pursue the Agree based approach to concord.– No additional redundancy: Can the new operations (i.e., post-syntacticfeatures copying or downwards feature percolation) derive other phenomenacaptured by Agree?– There are no good reasons to assume that concord is different from otherinstances of agreement.Norris’s 4 arguments that concord is different:1. In some languages, concord is realized on multiple elements within DP, whileclausal agreement appears only on the predicate.• Clausal agreement can also appear on multiple hosts: on the main verb andon the auxiliary, or on other elements such as adverbs and postpositions(see Bond & Chumakina (2016) on these phenomena in Archi).2. Only heads participate in predicative agreement, while elements showing nominalconcord can occupy a specifier and an adjunct position as well.• This depends on assumptions about the architecture of DP, cf. the analysisdeveloped below.3. Predicative agreement takes place between two distinct extended projections, buta probe and a goal are within one extended projection under nominal concord.• An interesting observation, but how this could be problematic for anyexisting implementation of Agree?4. Predicative agreement may be restricted by the case of a potential goal, but suchrestrictions are not attested for nominal concord.• Case sensitivity of predicative agreement is sometimes attributed to the factthat oblique nouns are embedded in PP/KP and this prevents probes fromreaching the features of DP. Given that all nominal modifiers are introducedbelow a PP/KP, no connection to case is expected.11Low origin of case• High origin of case is violates SCC and PIC.(78) Case assignmentvPVPDPAPnPr˚κ˚: sAr˚κ˚: sDr˚κ˚: sVvrκ:accs(79) Illegitimate concordvPVPDPAPnPr˚κ˚: sAr˚κ˚: sDrκ:accsVvrκ:accs????• DP constitutes a proper sub-part of the structure when the case is assigned, sothat case concord will violate SCC.• If DP (or any highest nominal projection) is a phase (see, e.g., Svenonius (2004),Matushansky (2004), and Bosˇkovic´ (2014)), case concord within the direct objectDP also violates even the weakest version of the PIC.(80) Strict Cycle Condition (SCC): (Chomsky, 1973, 1995, 2019)Within the current domain ∆1, no operation may exclusively affect positionswithin another domain ∆2 that is dominated by ∆1.(81) Phase Impenetrability Condition (PIC): (Chomsky, 2001)Given the structure [ZP Z ... [HP α [ H YP ] ] ] , where H and Z are phase heads,the domain of H is not accessible to operations at ZP, only H and its edge areaccessible to such operations.• There is one existing solution to this problem: Feature Sharing (see Frampton& Gutmann (2000, 2006), Pesetsky & Torrego (2007), and Kramer (2009) andDanon (2011) for such analysis of concord).– Problem for PIC stands: Some nodes dominating the shared probe shouldnot be accessible to operations at vP.– A challenge for the realization: A dominated by multiple nodes constituentis typically spelled out only in one of its positions (see Citko (2011), Johnson(2017)), but a shared feature is morphologically realized in all of them.Appendix E: Inflection in the verbal domain• Moksha has overt predicative agreement with respect to φ-features.(82) MonIluv-an.read-npst.1sg‘I read.’(83) Tjinjyou.plluv-i.read-npst.2pl‘You (pl) read.’• The analysis as it stands predicts absence of overt agreement morphology in theverbal domain as well.• Solution: Case and φ-features do not probe together in the clausal domain.8• Empirical evidence:– Case assignment does not require presence of φ-agreement; cf. (84).– φ-agreement can also proceed without case assignment; cf. (85).(84) Sjtj@rj-njE-sjgirl-dim-def.sgmasˇt-ican-npst.3[sg]pEnjakud-@njchimney-genusˇ-nj@-m@.fire up-freq-inf‘A girl can fire up a chimney.’ (Egorova, 2018)(85) Modamar˚-nj@-njpotato-def.pl-genmozˇn@-tcan-plvatka-m-s.peel-inf-ill‘One can peel potatoes.’ (A. Kozlov p.c.)(86) φ-agreementTPvPv 1VPvDP„φ:2plκ:genδ:defT”˚φ:2pl˚˚κ: ˚ ıφprobes(87) φ ConversionTPvPv 1VPvDP„φ:2plκ:genδ:defT”φ:2pl˚κ: ˚ıκprobes(88) κ-agreementTPvPv 1VPvDP„φ:2plκ:genδ:defT”φ:2pl˚κ:gen˚ıκprobes(89) Spell-OutTPvPT”φ:2pl˚κ:gen˚ıPF8 Another option: Order of operation is determined for the domain (e.g., a phase) rather than forthe language. In verbal domain, Probe Conversion must apply before Spell-Out.12


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items