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

A semantic constraint on the interpretation of pronominal elements Tatsumi, Yuta 2020-03-07

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

Download

Media
73804-Tatsumi_Y_Pronominal_WCCFL38_2020.pdf [ 58.66kB ]
Metadata
JSON: 73804-1.0389914.json
JSON-LD: 73804-1.0389914-ld.json
RDF/XML (Pretty): 73804-1.0389914-rdf.xml
RDF/JSON: 73804-1.0389914-rdf.json
Turtle: 73804-1.0389914-turtle.txt
N-Triples: 73804-1.0389914-rdf-ntriples.txt
Original Record: 73804-1.0389914-source.json
Full Text
73804-1.0389914-fulltext.txt
Citation
73804-1.0389914.ris

Full Text

WCCFL 38 March 7, 2020A semantic constraint on the interpretation of pronominal elementsYuta TatsumiUniversity of Connecticut1 Introduction• Two strategies of nominal ellipsis have been proposed in the literature: delation andpronominalization (Corver and Koppen (2011) and references therein).(1) Two strategies of nominal ellipsisa. [Deletion][ Mod [ N ] ]b. [Pronominalization][ Mod [ Npro ] ]• Hiraiwa (2016) argues that Japanese uses the pronominalization strategy.• Based on Hiraiwa (2016), this study investigates what kind of modifiers can combinewith the pronominal item.• Specifically, I argues that the condition in (2) holds for at least two types of pre-nominal modifiers in Japanese; numeral classifiers and temporal adjectives.(2) The pronominalization strategy is allowed when a pronominal item combineswith a local modifier of type 〈e,t〉.2 Pre-nominal numeral classifier phrases2.1 Unavailability of the quantificational reading• The first example of the condition in (2) comes from interpretations of pre-nominalnumeral classifiers in Japanese.• Japanese pre-nominal numeral classifier phrases can yield two interpretations ((3)).(3) Hanako-waHanako-TOP[go-satsu-nofive-CLS-GENhon]-obook-ACCkatta.bought‘Hanako bought five books.’ [Quantificational]‘Hanako bought {a book | books} composed of five volumes.’ [Property]The abbreviations are as follows: ACC = accusative; CLS = classifier; COP = copular; GEN = genitive linker;NOM = nominative; TOP = topic marker.• Kamio (1983) observes that the quantificational reading is unavailable in the nominalellipsis construction, as in (4b).• Japanese numeral classifier phrases have another type of anaphoric use as in (4c).When (4c) is uttered after (4a), only the quantificational reading is available.(4) a. Taro-waTaro-TOP[san-satsu-nothree-CLS-GENhon ]-obook-ACCkattaboughtkedobut...‘Taro bought three books, but ...’b. Hanako-waHanako-TOP[go-satsufive-CLSno]-oNO-ACCkatta.bought*‘Hanako bought five books.’ [Quantificational]‘Hanako bought {a book | books} composed of five volumes.’ [Property]c. Hanako-waHanako-TOP[go-satsu]-ofive-CLS-ACCkatta.bought‘Hanako bought five books.’ [Quantificational]*‘Hanako bought {a book | books} composed of five volumes.’ [Property]• The fact that the pronominal no is incompatible with the quantificational interpreta-tion is also confirmed by the quantifier subete ‘all’.• As shown in (5b), subete ‘all’ cannot be followed by the pronominal no.(5) a. Taro-waTaro-TOP[san-satsu-nothree-CLS-GENhon ]-obook-ACCkattaboughtkedobut...‘Taro bought three books, but ...’b. *Hanako-waHanako-TOP[subeteallno]-oNO-ACCkatta.bought‘Hanako bought all books.’c. Hanako-waHanako-TOP[subete]-oall-ACCkatta.bought‘Hanako bought all books.’• The quantifier subete ‘all’ has only the quantificational meaning, unlike pre-nominalnumeral classifier phrases. (5b) is unacceptable because the pronominal no is incom-patible with a modifier which has a quantificational meaning.• As shown in (5c), the ellipsis construction, which lacks both an overt noun and no,is acceptable under the quantificational reading.12.2 Analysis• Following Kamio (1983) and Hiraiwa (2016), I assume that there are two positionsfor Japanese pre-nominal numeral classifiers. The object noun phrases in (3) and (4)have one of the structures represented in (6).(6) Analysisa. [QP [ClsP three-CLS ] [Q’ [N {book | *no} ] Q ]] (Quantificational)b. [NP [ClsP three-CLS ] [N {book | no} ]] (Property)• I assume that the classifier phrase in (6a) is of type 〈〈e,t〉,〈〈e,t〉,t〉〉, like other quanti-fiers such as subete ‘all’.• In (6a), the condition in (2) is not satisfied, and the pronominal no is not licensed.• When the classifier phrase functions as an attributive modifier of type 〈e,t〉 (i.e.λx[five-volume′(x)]), it directly modifies the noun via Predicate Modification (Heimand Kratzer (1998)), as in (6b).• The ClsP in (6b) licenses the pronominal no, respecting the condition in (2).• The condition in (2) includes a locality constraint which requires a licensing modifierto be local to a pronominal item.• This requirement is confirmed by the example in (7).(7) a. Taro-waTaro-TOP[san-satsu-nothree-CLS-GENhon ]-obook-ACCkattaboughtkedobut...‘Taro bought three books, but ...’b. Hanako-waHanako-TOP[ takaiexpensivego-satsufive-CLSno]-oNO-ACCkatta.bought*‘Hanako bought five expensive books.’ [Quantificational]‘Hanako bought {an expensive book | expensive books} composed of fivevolumes.’ [Property]c. Hanako-waHanako-TOP[go-satsu-nofive-CLS-GENtakaiexpensiveno]-oNO-ACCkatta.bought‘Hanako bought five expensive books.’ [Quantificational]‘Hanako bought {an expensive book | expensive books} composed of fivevolumes.’ [Property]• The elliptical noun phrase in (7b) behaves like the one in (4b) regarding its interpre-tation.• In this case, the pre-nominal numeral classifier phrase, which is a property-denotingmodifier of type 〈e,t〉, licenses the pronominal no.• The adjective takai ‘expensive’ intervenes between the pre-nominal numeral classi-fier phrase and the pronominal no in (7c).• Crucially, (7c) exhibits the ambiguity of the interpretation of the pre-nominal nu-meral classifier.• The current analysis can capture the ambiguity of (7c).• The pre-nominal numeral classifier in (7c) is not local to no and hence is not subjectto the condition in (2). It can yield the two interpretations.• The absence of the quantificational reading in the nominal ellipsis constructioncan be explained by the condition in (2).• When a pre-nominal numeral classifier phrase is a licensing modifier of thepronominal item, it must be of type 〈e,t〉 due to the condition in (2).3 Temporal adjectives3.1 The ambiguity of Japanese temporal adjectives• The second example which exhibits the effect of the condition in (2) comes from(im)possible interpretations of temporal adjectives.• As shown in (8), the temporal adjective old in English exhibits an ambiguity.(8) a. (Larson and Cho (2003))This is John’s old car.b. ‘This is a car that John possesses and that is old.’ [N-mod.]c. ‘This is a car that John formerly owned.’ [POSS-mod.]• A similar ambiguity is observed in Japanese, as shown in (9).(9) kore-wathis-TOP[Taro-noTaro-GENhuruioldkurumacar] desu.COP‘This is a car that Taro possesses and that is old’ [N-mod.]‘This is a car that Taro formerly owned.’ [POSS-mod.]• I propose that Japanese temporal adjectives have one of the structures represented in(10).(10) a. [Indirect modification] / [N-mod.][NP [RC pro1 OLD ] [NP N1 ] ]b. [Direct modification] / [POSS-mod.][NP [AP OLD ] [NP N ] ]2• I refer to (10a) as the indirect modification structure, and (10b) as the direct modifi-cation structure (Cinque (2010)).• When the temporal adjective occurs in the indirect structure as in (10a), the resultingphrase receives the N-modifying reading.• When the temporal adjective occurs in the direct modification structure as in (10b),the resulting phrase receives the POSS-modifying reading.• There is evidence that the ambiguity of (9) arises from the structural ambiguity rep-resented in (10).• Firstly, when hurui ‘old’ is used as a predicate, only the N-modifying reading isavailable as in (11).(11) [Taro-noTaro-GENkuruma]-gacar-NOMhurui.old‘The car owned by Taro is old.’ [N-mod.]• The temporal adjective in (10a) is used as a predicate of the relative clause. There-fore, the temporal adjective in (10a) is expected to yield the N-modifying reading,like the one in (11).• Another piece of supporting evidence for the structures in (10) comes from adverbialexpressions in Japanese.• In Japanese, tokubetsu ‘special’ can be used as an adjectival expression or an adver-bial expression, depending on the suffix follows it.• The contrast in (12) shows that tokubetsu must be followed by the suffix -na, in orderto function as an adjectival expression modifying a noun.• The suffix -ni must be used for the adverbial use of tokubetsu, as shown in (13).(12) a. Yuta-waYuta-TOP[Hiro-noHiro-GENtokubetsu-naspecial-NAuta]-osong-ACCkiita.listened‘Yuta listened Hiro’s special song.b. *Yuta-waYuta-TOP[Hiro-noHiro-GENtokubetsu-nispecial-NIuta]-osong-ACCkiita.listened‘Yuta listened Hiro’s special song.(13) a. *Hiro-gaHiro-NOM[ tokubetsu-naspecial-NAutatta].sang‘Hiro specially sang.b. Hiro-gaHiro-NOM[ tokubetsu-nispecial-NIutatta].sang‘Hiro specially sang.• Crucially, tokubetsu-ni blocks the POSS-modifying reading of the temporal adjec-tive, as shown in (14a), while tokubetsu-na allows both the N-modifying reading andthe POSS-modifying reading as in (14b).(14) a. Taro-noTaro-GENtokubetsu-nispecial-NIhuruioldkurumacar‘a speciallly old car that Taro possesses’ [N-mod.]*‘a car that John owned a long time ago’ [POSS-mod.]b. Taro-noTaro-GENtokubetsu-naspecial-NAhuruioldkurumacar‘a special old car that Taro possesses’ [N-mod.]‘a special car that John formerly owned’ [POSS-mod.]• The current analysis can capture the contrast in (14).• The two structures in (15) are available for (14b).(15) a. [N-mod.][NP tokubetsu-na [NP [RC pro1 OLD ] NP1 ] ]b. [POSS-mod.][NP tokubetsu-na [NP [AP OLD ] NP ] ]• (14a) should have the structure in (16a), but not the one in (16b), because of thepresence of tokubetsu-ni.(16) a. [N-mod.][NP [RC pro1 [ tokubetsu-ni OLD ] ] NP1 ]b. [POSS-mod.]*[NP tokubetsu-ni [NP [AP OLD ] NP ] ]• Since tokubetsu-ni is an adverbial expression modifying a predicate, it cannot occurinside the extended nominal projection containing kuruma ‘car’, as in (16b).• In (16a), which yields the N-modifying reading, the temporal adjective is used as apredicate in the relative clause, and tokubetsu-ni modifies the predicate.• Under the current analysis, the POSS-modifying reading is unavailable in (14a) be-cause (16b) is impossible.3.2 Temporal adjectives and pronominalization• Crucially, the ambiguity of temporal adjectives disappears when the temporal adjec-tive occurs with the pronominal no.• When (17c) is uttered after (17a), only the N-modifying reading is available.3(17) a. are-wathat-TOP[Hanako-noHanako-GENkurumacar] desu.COP‘That is Hanako’s car.’b. kore-wathis-TOP[Taro-noTaro-GENhuruioldkurumacar] desu.COP‘This is a car that Taro possesses and that is old.’ [N-mod.]‘This is a car that John formerly owned’ [POSS-mod.]c. kore-wathis-TOP[Taro-noTaro-GENhuruioldnoNO] desu.COP‘This is a car that Taro possesses and that is old.’ [N-mod.]*‘This is a car that John formerly owned’ [POSS-mod.]• The noun phrase in (17c) should have the indirect modification structure given in(18a), due to the condition on the pronominalization strategy in (2).(18) a. [NP Taro-no [NP [RC pro1 OLD ] no1 ] ]b. *[NP Taro-no [NP [AP OLD ] no ] ]• Following Larson (1998), I assume that the temporal adjective in the direct modifi-cation structure ((10b) and (18b)) is not of type 〈e,t〉.• Larson (1998) proposes that old in English is a doublet, as in (19).(19) a. J old1 friend K = λx[old1′(x) & friend′(x)]b. J old2 friend K = old2’(ˆfriend′)• As shown in (19a), old1 is analyzed as a modifier of type 〈e,t〉, yielding the N-modifying interpretation.• On the other hand, old2 combines with ˆfriend′, and hence is not of type 〈e,t〉.• I implement Larson’s analysis syntactically.• Old1 occurs in the indirect modification structure ((10a) and (18a)), whereas old2occurs in the direct modification structure ((10b) and (18b)).• When old1 is used as a predicate in a relative clause, the relative clause functions asa modifier of type 〈e,t〉 and licenses the pronominal no, as in (18a).• Old2 in (18b) is not of type 〈e,t〉, and the pronominal strategy is unavailable due tothe condition in (2).• The unambiguity of (17c) can thus be explained by the condition in (2).4 Summary & implication• This paper argued that the condition in (2), repeated here as (20), holds for two typesof pre-nominal modifiers in Japanese; numeral classifiers and temporal adjectives.(20) The pronominalization strategy is allowed when a pronominal item com-bines with a local modifier of type 〈e,t〉.• The current analysis can be associated with the gist of Tomioka’s (2003) analysisof null pronouns. Boškovic´ (2017) also offers the following semantic condition onargument ellipsis based on Tomioka’s insights.(21) Only elements of type 〈e,t〉 can be copied in LF.• It is not unreasonable to pursue a unified analysis in which (20) is reduced to a moregeneral semantic condition like (21). I leave this for future research.ReferencesBoškovic´, Željko. 2017. On clitic doubling and argument ellipsis. In Proceedings ofthe 34th west coast conference on formal linguistics, ed. Miranda K. McCarvelAaron Kaplan, Abby Kaplan and Edward J. Rubin, 97–106. Somerville, MA: Cas-cadilla Proceedings Project.Cinque, Guglielmo. 2010. The syntax of adjectives: A comparative study. Cambridge:MIT Press.Corver, Norbert, and Marjo van Koppen. 2011. NP-ellipsis with adjectival remnants: amicro-comparative perspective. Natural Language and Linguistic Theory 29:371–421.Heim, Irene, and Angelika Kratzer. 1998. Semantics in generative grammar. Malden:Blackwell.Hiraiwa, Ken. 2016. NP-ellipsis: a comparative syntax of Japanese and Okinawan.Natural Language and Linguistic Theory 34:1345–1387.Kamio, Akio. 1983. Meisiku no koozoo [The structure of noun phrases]. In Nihongo nokihon koozoo [The basic structure of Japanese], ed. Kazuko Inoue, 77–126. Tokyo:Sanseido.Larson, Richard. 1998. Events and modification in nominals. In Proceedings fromsemantics and linguistic theory 8, ed. Devon Strolovitch and Aaron Lawson, 145–168. Ithaca, N.Y.: Cornell University: CLC Publications.Larson, Richard K., and Sungeun Cho. 2003. Temporal adjectives and possessive DPs.Natural Language Semantics 11:217–247.Tomioka, Satoshi. 2003. The semantics of Japanese null pronouns and its cross-linguistic implications. In The interfaces, ed. Kerstin Schwabe and Susanne Win-kler, 321–339. Amsterdam: John Benjamins.4

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

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"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.73804.1-0389914/manifest

Comment

Related Items