Determined Reference as a CI Component in the Meaning of Definites Evidence from languages with multiple definite articles Assif Am-David and Manfred Sailer,

Determined Reference as a CI Component in the Meaning of

Definites

Evidence from languages with multiple definite articles

Assif Am-David and Manfred Sailer,Goethe University Frankfurt

Determined Reference as a CI Component in the Meaning of Definites

• Introduction• Hierarchy of determined reference• Why CIs?• Typology of languages with multiple definite articles• Hierarchy of determined reference in multiple definite

articles• Evidence for CI interpretation• Fieldwork on Northern Frisian• Referential covariation under quantification• Summary

Introduction

Sioux (Lakota) (Siouan) (Rood & Taylor, 1996):• Articles :k?uN: anaphoric definitedifferent from demonstratives: le’, he’, ka’kiN: non-anaphoric definitewaN-: indefinite (1) SˆuN’ka kiN he’l yuNke’. SˆuN’ka k?uN thalo’ kiN thebye’. Dog the there lie. Dog the.past meat the eat.up‘The dog is lying there. The dog ate up the meat.’

Introduction

• Schwarz (2013) distinguishes two definite articles: weak and strong.

• Weak definite article marks uniqueness.• Strong definite article marks anaphoricity.

Introduction

Objectives: • Formally account for the distinction between

definite articles – Using different types of inferences (presuppositions and CIs [=conventional implicatures])

• Propose a fine-grained model of determined reference to replace the dual distinction

• Account for crosslinguistic variation in the domain of definite articles (including languages with three definite articles)

Hierarchy of determined reference

• Familiarity-identifiability based approaches (Maetzner, 1885; Christophersen, 1939; Jespersen, 1943; Strawson, 1950; Hawkins, 1978; Heim, 1982; Ariel, 1990; Gundel et al., 1993)

• Uniqueness-maximality based approaches (Russell, 1905; Sharvy, 1980; Link, 1983; Landman, 1991)


• Combining both approaches:Donnellan (1966): Ambiguous (referential vs. attributive)Roberts (2003): Both together (uniquely identifiable)


• Schwarz (2013): Weak definites mark uniqueness, strong definites mark familiarity

No apparent relation between the two.• Current attempt: combination of two

different meaning components – Maximality presuppositionA series of CIs – Degree of familiarity in the form of hierarchy of determined reference


• Maximality presupposition: !x P(x)• Affirmative CIs: A. discourse-oldness (=anaphoricity)B. hearer-oldness


A. Discourse-oldness (+DO): The speaker believes the addressee knows there is a maximal element satisfying the definite description belonging to some subset of the universe of discourse.

BELIEVE(I, (KNOW (you, (x )))))∈


• Good for: anaphorical definites(2) A student arrived. The student was helpless.• This allows extensionally non-referential

anaphoras in opaque contexts with empty descriptions.

(3) Everyone warned me of the witch in the cave. Yet, I was not afraid because I know the witch does not exist.


• For non-anaphorical definites the negative counterpart of the CI will be implied (-DO):

BELIEVE(I, (KNOW (you, (x )))))∈


• B. Hearer-oldness (+HO): The speaker believes the addressee knows there is a maximal element satisfying the definite description.

BELIEVE (I, (KNOW (you, )))


• No empty set denotation.• Does not require acquaintance.• More closely related to referentiality in

Strawson (1950) than Donnellan (1966) – No misdescriptions admitted.

• Good for: uniques(4) The sun is rising.


• For non-hearer-old definites the negative counterpart of the CI will be implied (-HO):

BELIEVE(I, (KNOW (you,


• The lack of both discourse-oldness and hearer-oldness is good for:

• situational definites(5) Give me the hammer on the table.


• (non-anaphorical) attributive definites (Russellian quantificational account)

(6) A: I do not know who the current king of Denmark is. B: In fact it is a queen – Margarethe II.No commitment to the existence of a maximal element satisfying the definite description.


Based on a distinction drawn by Schwarz(2013), bridging can evoke either +DO or -DO˄+HO depending on whether it is anaphoric or not.• +DO (strong definite): relational anaphoras Relation between antecedent and anaphora is semantically explicit. (7) I saw a book. Something about the cover kept intriguing me.


• -DO˄+HO (weak definite): non-relational anaphoras

Relation between antecedent and anaphora is semantically implicit.(8) I bought a very nice painting. My wife said she particularly liked the green spots.


• Hierarchy: Determined reference

Anaphoric Non-anaphoric

Hearer-old Non-hearer-old


• The hierarchy does not cover all logical options, but depicts the possible marking patterns in language.

• Introduction• Hierarchy of determined reference

• Why CIs?• Typology of languages with multiple definite articles• Hierarchy of determined reference in multiple definite articles• Evidence for CI interpretation• Fieldwork on Northern Frisian• Referential covariation under quantification• Summary

Why CIs?

Indicative tests for discerning presuppositions and CIs (Levinson, 1983; Potts, 2005):(I) Presupposition failure with a logically unembedded presupposition trigger leads to the loss of truth value in the assertion (or falsity).

Why CIs?

(9) # I stopped playing the piano, but I have never played the piano.I played the piano (the presupposition lexically triggered by stopped) is either false (Russell, 1905) or undefined (Strawson, 1950), never true.

Why CIs?

Conversely, infelicitous flouting of a CI does not affect the truth value of the assertion, which can be true independently of the truth value of the CI.

Why CIs?

(10) "Alex met Probal, the president of the World Esperanto Association."-> No, that's not true, Alex did not meet Probal.-> #No, that's not true, Probal is not the president of the World Esperanto Association.Alex met Probal could be true or false regardless of the truth value of the identity between Probal and the description, which is a CI triggered by the apposition.

Why CIs?

(II) Presuppositions can project, but do not have to - If the presupposition trigger is logically embedded, presuppositions can be cancelled (local accommodation). Example – polemic negation:(11) I did not stop playing the piano because I have never played the piano.

Why CIs?

Conversely, CIs must project and they can never be flouted (conjunctive addition) regardless of syntactic structure.(12) # It is not true he is smart but rich. Being smart has nothing to do with being rich.Here the CI is the contrary relation between smart and rich.

Why CIs?

Application to definites:(I) The cancellation of logically unembedded maximality leads to the loss of truth value in the assertion (Strawson, 1950) or its falsity (Russell, 1905).(13) # I saw the Harry Potter movie. In fact there are 8, of which I only saw the third.Here the assertion I saw the HP movie cannot be true.

Why CIs?

This does not happen with the infelicitous flouting of the CIs of determined reference.

(14) # A studenti came by. The studentj asked me a

question. In fact, it was not the student who came by.

Truth value of assertion the studentj asked me a question can be true regardless of the infelicitous reference! (This can be stipulated if pointing is applied, for example.)

Why CIs?

• Second indicative test will be applied to definites later.

• Introduction• Hierarchy of determined reference• Why CIs?

• Typology of languages with multiple definite articles

• Hierarchy of determined reference in multiple definite articles• Evidence for CI interpretation• Fieldwork on Northern Frisian• Referential covariation under quantification• Summary

Typology of languages with multiple definite articles

• Typology to provide morpho-syntactic categories.

• Semantic distinction between strong, weak and super-weak definite articles to be explained below.


(1) Strong and weak definite articles in a two definite articles system, phonologically unrelated:• Northern Frisian (Germanic, Indo-European) –

strong: di/det, weak: a/at (Ebert, 1970), but see p.c. Faltings.

• Sioux (Lakota) (Mississippi Valley, Siouan) – strong: k’uN, weak: kiN (Boas & Deloria, 1941; Rood & Taylor, 1996; van Vallin, 2012), both interpreted as topic marker in Ingham (2003).


(2) Strong and weak definite articles in a three definite articles system:• Forest Enets (Samoyedic, Uralic) – strong: r, weak:

da (Siegl, 2013; Gerland & Wratil, 2015)• Nganasan (Samoyedic, Uralic) – strong: rǝ, weak: zu

(Schröder, 2006)• Mari (Uralic) – strong: yd, weak: ys (Schlachter,

1960; Klumpp, 2009) • Northern Frisian (Germanic, Indo-European) strong:

di/det, weak: a, super-weak: a/at (p.c. Faltings)


(3) Strong and weak definite articles, weak definite is a contracted strong definite:• German dialects (Germanic, Indo-European) –

Alemannic German: strong: dε/di/das, weak: de/d/s (Studler , 2011), also Himmelmann (1996) for Franconian German.

• Hausa (Chadic, Afro-Asiatic) – strong: ɗin, weak: n/r (Jaggar, 2001), strong one identical with demonstrative (Buba, 1997)


(4) Strong and weak definite article, strong one identical with the demonstrative:• Malagasy (Malayo-Polynesian, Austronesian) –

strong: ilay, weak: ny (Ferrand, 1903; Fugier, 1999)

• Hausa (see (3))


(5) Only strong definite article:• Nenets (Samoyedic, Uralic) – strong: da/ta

(Hajdu, 1963)• Hidatsa (Missouri River, Siouan) – strong: s

(Matthews, 1965)• Fon (Fongbe) (Kwa, Niger-Congo) – strong: ɔ (Lefebvre & Brousseau, 2002)


• Twi (Akan) (Kwa, Niger-Kongo) – strong: no (Addai Boadi, 2010)

• Mupun (Mwaghavul) (Chadic, Afro-Asiatic) – strong: ɗi (Frajzynger, 1993)

• Mangarayi (Mangaray-Maran) – strong: gi (Merlan, 1982)

• Mauritian Creole (Bourbonnais Creoles) – strong: la, but extended (Wespel, 2008; Syea, 2013), intepreted as a specificity marker by Guillemin (2011)


(6) Two definite articles, unrelated to the hierarchy of determined reference:• Scandinavian (Swedish, Danish, Norwegian)

(Germanic, Indo-European) – • Distinction purely syntactic (cf. Dahl, 2004),

depending on the presence of a modifier for example.

• Introduction• Hierarchy of determined reference• Why CIs?• Typology of languages with multiple definite articles

• Hierarchy of determined reference in multiple definite articles

• Evidence for CI interpretation• Fieldwork on Northern Frisian• Referential covariation under quantification• Summary

Hierarchy of determined reference in multiple definite articles

Schwarz (2013) : • strong definite article – anaphoric definites • weak definite articles – non-anaphoric ones• Definite article can be phonologically empty.


In current model:• Languages with two definite articles:Strong definite article: +DOWeak definite article: -DO• Languages with three definite articles:Strong definite article: +DOWeak definite article: -DO,-HOSuper-weak definite article: -DO,-HO


Semantic maps – contiguous covering (Haspelmath, 2000).Lexical example: tree wood (stuff) firewood small forest large forest German Baum Holz Holz Wald WaldDanish trae trae trae skov skovFrench arbre bois bois forêt forêt Spanish árbol madera leña bosque selva


Grammatical example:

emphatic full grooming/ anticausative potential passive reflexive reflexive body motion passive

|——————————| English himself |—————————————| Classical Latin se

|—————————————————————| Late Latin se |————————————————————————|French se

Italian si |—————————————————————————————| Surselvan (Romansch) se-|——————————————————————|Swedish –s |——————————————————|


Application of semantic maps to the typology of definite articles.

1)

Languages: Northern Firian, Forest Enets, Mari

Uniqueness +HO˄-DO +DO


2)

Languages: Sioux, Alemannic German, Twi, Mangarayi

3)

Languages: English, Arabic, Q‘ekchi‘




• Languages with the mapping (2) or (3) are vague (underspecified) as to the grammatical distinction between the semantic categories.

• Attributive definites (-HO) appear not constitute a unit in the semantic map. For example, both strong and weak definites are allowed in Standard German in non-referential uses (own work).


• If a langugae has a phonologically realised indefinite article and no overt definite article a null definite article is assumed.

• If the realm of the definite article is divided into two morpho-syntactic categories and one of them is covert, it will be the weak one.

• If the realm of the definite article is divided into three morpo-syntactic categories and one of them is covert, it will be the super-weak one.

Evidence for CI interpretation• Introduction• Hierarchy of determined reference• Why CIs?• Typology of languages with multiple definite articles• Hierarchy of determined reference in multiple definite articles

• Evidence for CI interpretation• Fieldwork on Northern Frisian• Referential covariation under quantification• Summary

Evidence for CI interpretation

• Standard German: von dem/vom (from the)bei dem/beim (by the)zu dem/zum and zu der/zur (to the)in das/ins (to the)


• Both forms are very frequent.• Used to distinguish strong and weak definites

(Schwarz, 2013)(15) Hans ging zum/zu dem Haus. Hans went to the (weak/strong) house.


• Hypothesis: Non-anaphoricity (-DO) is a CI, therefore, its flouting should be infelicitous.

• Therefore: If the hypothesis is correct cataphoric definites with weak definite articles should be ungrammatical, even under the scope of negation.

• Contrary: If non-anaphoricity is a presupposition it should be cancellable if under the scope of negation.


(16) Weil ich eh nicht zu dem/zum Fest gegangen wäre, hat mein Ex-Mann mich gar nicht erst zu seiner Hochzeit eingeladen.

Because I wouldn‘t go to the(strong/weak) party anyway my ex-husband didn‘t invite me to his wedding in the first place.


(17) Weil ich nicht an dem/am Lauf teilgenommen habe, hat dieses Jahr jemand anderes den Berlinmarathon gewonnen.

Because I didn‘t participate in the(strong/weak) run someone else won the Berlin Marathon this year.


(18) Weil ich nicht sonderlich an dem/am Möbelstück hänge, werde ich den Schrank beim nächsten Sperrmüll rausschmeißen.

Because I‘m not particularly attached to the(strong/weak) piece of furniture, I will through away my cupboard with the next bulk trash.


• Plan to examine grammaticality judgments of native speakers of both contracted (weak) and full (strong) forms.

• Future fieldwork on:Northern Frisian (Volkert Faltings, Ferring Stiftung) Alemannic German (Elvira Glaser, Zurich University)

• Introduction• Hierarchy of determined reference• Why CIs?• Typology of languages with multiple definite articles• Hierarchy of determined reference in multiple definite articles• Evidence for CI interpretation

• Fieldwork on Northern Frisian• Referential covariation under quantification• Summary

Fieldwork on Northern Frisian

Northern Frisian: • West-Germanic language, most closely related to

the other Frisian languages (Eastern Frisian and Western Frisian) next to English

• Formerly considered a dialect of Frisian, now consensually classified as a language (certainly not mutually intelligible, substantial grammatical variation)

• Recognised minority language in Germany and EU


• 10000-8000 speakers (Schleswig-Holstein, 2015)• All speakers fluent in standard German as well

as Low German (highly diglossic environment)• Great influence from Low German and Danish• Potentially endangered, education available in

Northern Frisian (school as well as university)• Major break between insular and mainland

dialectal groups



• According to p.c. Faltings: threefold system of definite articles

• Strong article di/det diachronically derived from the demonstrative dihear/didiar.

• Weak article is a reduced super-weak article a/at.

• Ebert (1970) only mentions strong and super-weak forms.

• Indefinite article an.


• All data from Ferring-Amrum dialect• No data on other dialects (Sölring, Mooring)• Distinction missing in Eastern and Western Frisian• Similar distinction present in Southern Juttish

dialect of Danish (Ebert, 1970; Faltings, p.c.) – regional diffusion

• Threefold distinction disappearing in Ferring-Amrum, strong form disappearing in younger speakers (p.c. Faltings)


• Ferring Stiftung in Alkersum promotes research on Northern Frisian as well as school material and radio programmes

• Native speakers of Northern Frisian from all ages available – focus on speakers older than 40


• First Phase: Study of the use of the three definite articles through

• both judgments and elicitation• Second Phase: Judgments of cataphoric

sentences with negation in order to determine status of inference


Further plans pending financing:• Other dialects of Northern Frisian• Southern Juttish dialect of Danish• Language change in Northern Frisian

(synchronic and diachronic)

• Introduction• Hierarchy of determined reference• Why CIs?• Typology of languages with multiple definite articles• Hierarchy of determined reference in multiple definite articles• Evidence for CI interpretation• Fieldwork on Northern Frisian

• Referential covariation under quantification• Summary

Referential covariation under quantification

• Problem: Referential covariation under quantification

(19) Every student has a mentor and the mentor needs to meet her twice a year.• If referentiality is a CI how can the quantifier

take scope over it hence preventing it from projecting?


• Comparable case:(20) Jede Katze denkt, dass der Nachbarshund es zum Glück nicht merkt, wenn sie über den Zaun springt.Every cat thinks that the neighbour’s dog fortunately does not notice when it jumps over the fence.


• Zum Glück (fortunately) is an evaluative adverb which brings about a CI: It is fortunate for x that P.

• Conjunctive interpretation without covariation:

CI: It is fortunate for x (x=the speaker) that P.


• According to 20 readings judgments most natural reading:

Every cat thinks: It is fortunate that P.• Hence, CI: For every x that is a cat, it is

fortunate for x that P.


• The universal quantifier takes scope over the CI. The variable in the CI covaries with the variable bound by the quantifier.

• Conclusion: Covariation under quantification required independently of interpretation of referentiality in definites.

• Probably, quantifiers can (should?) take scope over CIs.

Conclusions

• Uniqueness and anaphoricity follow distinct inference patterns

• Two degrees of determined reference can account for more occurrences of definite article.

• Account for languages with three definite articles.

References

• Addai Boadi, Laurence (2010). The Akan noun phrase. Accra: Black Mask.• Ariel, Mira (1990). Accessing noun phrases antecedents. Cambridge: Cambridge

University Press.• Boas, Franz and Deloria, Ella Cara (1941/2011). Dakota grammar. Cincinnati:

Whitefish.• Buba, Malami (1997). “The deictic particle DI-N in Hausa”. African languages and

cultures. 10/1: 29-45.• Christophersen, Paul (1939). The articles. Copenhagen: Munksgaard.• Dahl, Östen (2004). “Definite articles in Scandinavian”. In: Kortmann, Bernd. (ed.).

Dialectology meets typology. Berlin: Gruyter.• Donnellan, Keith (1966/1998). “Reference and definite descriptions”. In: Ostertag,

Gary (ed.). Definite descriptions. pp. 173-193. Cambridge, Massachusetts: MIT Press.• Ebert, Karen H. (1970). Referenz, Sprechsituation und die bestimmten Artikel in

einem nordfriesischen Dialekt (Fering). Kiel: Nordfriisk Instituut.

References• Ferrand, Gabriel(1903/2010). Grammaire malgache. Munich: Lincom.• Frajzynger, Zygmunt (1993) A grammar of Mupun. Berlin: Reimer.• Fugier, Huguette (1999). Sytaxe malgache. Louvain-la-neuve: Peeters.• Gerland, Doris; Wratil, Melani (2015). Information structuring and

definiteness marking in Forest Enets. Ms.• Guillemin, Diana (2011). The syntax and semantics of a determiner

system. Amsterdam: Benjamins.• Gundel, Jeanette K.; Hedberg, Nancy; Zacharski, Ron (1993). “Cognitive

status and the form of referring expressions in discourse”. Language 69(2): 274–307.

References• Hajdu, Peter (1963). The Samoyed peoples and languages. Bloomington:

Indiana University Press.• Haspelmath, Martin (2000). “The geometry of grammatical meaning:

Semantic maps and cross-linguistic comparison”. In: Tomasello, Michael (ed.). The new psychology of language, vol. 2: Mahwah. New York: Erlbaum.

• Hawkins, John A. (1978). Definiteness and indefiniteness. London: Croom Helm.

• Heim, Irene R. (1982). The semantics of definite and indefinite noun phrases. Doctoral dissertation. University of Massachusetts.

• Himmelmann, Nikolaus P. (1996). Demonstratives in Narrative Discourse. In: Fox, Barbara (ed.). Studies in Anaphora. pp. 203–252. Amsterdam: Benjamins.

• Ingham, Bruce (2003). Lakota. Munich: Lincom.

References• Jaggar, Philipp J. (2001). Hausa. Amsterdam: Benjamins.• Jespersen, Otto (1943). A modern English grammar on historical

principles. Copenhagen: Munksgaard.• Klumpp, Gerson (2009). “Identifiability, givenness and zero-marked

referential objects in Komi“ . In: Dufter, Andreas; Fleischer, Jürg; Seiler, Guido (eds.). Describing and Modeling Variation in Grammar. pp. 325–359. Berlin: Gruyter.

• Landman, Fred (1991). Structures for semantics. Dordrecht: Kluwer.• Lefebvre, Claire and Brousseau, Anne-Marie (2002). A grammar of

Fongbe. Berlin: Gruyter

References• Levinson, Stephen C. (1983). Pragmatics. Cambridge: Cambridge University

Press.• Link, Godehard (1983). “The logical analysis of plurals and mass terms: a lattice-

theoretical approach”. In: Bäuerle, Rainer; Schwarze, Christoph; von Stechow, Arnim (eds.). Meaning, use and interpretation of language, Berlin: Gruyter.

• Maetzner, Eduard A. F. (1885). Englische Grammatik. Berlin: Weidmannsche Buchhandlung.

• Matthews, George H. (1965). Hidatsa syntax. London: Mouton.• Merlan, Francesca (1982/1989). Mangarayi. London: Routeledge• Potts, Christopher (2005). The logic of conventional implicatures. Oxford:

Oxford University Press.• Landesportal Schleswig-Holstein, http://

www.schleswig-holstein.de/DE/Fachinhalte/M/minderheiten/minderheiten_friesen.html, retrieved: 27.10.2015

http://www.schleswig-holstein.de/DE/Fachinhalte/M/minderheiten/minderheiten_friesen.html



References• Roberts, Craige (2003). “Uniqueness in definite noun phrases”. Linguistics

and Philosophy 26:287–350.• Rood, David S.; Allan R. Taylor (1996). “Sketch of Lakhota”. In: Goddard,

Ives (ed.). Hand- book of north American Indians. 17: 440–482. Washington: Smithsonian Institution. <lakxo- taiyapi.freecyberzone.com> retrieved: 23.6.2015.

• Russell, Bertrand (1905/1998). “On denoting”. In: Ostertag, Gary (ed.). Definite descriptions. pp. 35-49. Cambridge, Massachusetts: MIT Press.

• Schlachter, Wolfgang (1960). Studien zum Possessivsuffix des Syrjänischen. Akademischer Verlag: Berlin.

• Schroeder, Christoph (2006). “Articles and article systems in some areas of Europe”. In: Bernini, Giuliano (ed.). Pragmatic organization of discourse in the languages of Europe. Berlin: Gruyter.

• Schwarz, Florian (2013). “Two kinds of definites cross-linguistically”. Language and linguistics compass 7/10: 534-559.

References• Sharvy, Richard (1980). “A More General Theory of Definite Descriptions”.

The Philosophical Review 89: 607-23.• Siegl, Florian (2013). Materials on Forest Enets, an indiginous language

of northern Siberia. Tartu: Tartu University Press.• Strawson, Peter F. (1950/1998). “On referring”. In: Ostertag, Gary (ed.).

Definite descriptions. pp. 135-160. Cambridge, Massachusetts: MIT Press.• Studler, Rebekka (2011): Artikelparadigmen. Doctoral dissertation.

University of Zurich.• Syea, Anand (2013). The syntax of Mauritian Creole. London: Bloomsbury.• van Vallin Jr., Robert D. (2012). Definiteness in Lakhota. Ms.• Wespel, Johannes (2008). Descriptions and their domains. Doctoral

dissertation. University of Stuttgart.

Determined Reference as a CI Component in the Meaning of Definites Evidence from languages with multiple definite articles Assif Am-David and Manfred Sailer,

Documents