ࡱ > ~ @ , bjbjqq 32 Q
V V V j Q Q Q 8 Q R j h S N W :
X
X
X BY [ ^\ H n p p p 8 x H $ R n l i V "^ >Y BY "^ "^ l
X
X 4 m m m "^
<
X V
X
m "^ n m m | V
X vS Ef Q 4h 4 0 h M hk & M j j
M V 0 \ 0 \ " m \ ] \ \ \ l l j Z , > m p j j ; Lecture 11. Pronouns and Reflexives III: Typological Issues
TOC \o "1-2" \h \z \u HYPERLINK \l "_Toc196594200" 1. Background: typology and theory PAGEREF _Toc196594200 \h 1
HYPERLINK \l "_Toc196594201" 2. Kiparskys approach to anaphora typology PAGEREF _Toc196594201 \h 2
HYPERLINK \l "_Toc196594202" 2.1. On Obviation vs. Blocking PAGEREF _Toc196594202 \h 2
HYPERLINK \l "_Toc196594203" 2.2. The typology of anaphoric expressions PAGEREF _Toc196594203 \h 3
HYPERLINK \l "_Toc196594204" 3. Coargument Disjoint Reference. PAGEREF _Toc196594204 \h 4
HYPERLINK \l "_Toc196594205" 3.1. Four generalizations about disjoint reference PAGEREF _Toc196594205 \h 4
HYPERLINK \l "_Toc196594206" 3.2. Bound variable vs. coreferential (covaluation) readings PAGEREF _Toc196594206 \h 4
HYPERLINK \l "_Toc196594207" 3.3. Illustration: Swedish sig, an obviative reflexive. Cancellation of obviativity by sjlv. PAGEREF _Toc196594207 \h 8
HYPERLINK \l "_Toc196594208" 3.4. Gaps PAGEREF _Toc196594208 \h 9
HYPERLINK \l "_Toc196594209" 3.5. Morphologically marked Obviation PAGEREF _Toc196594209 \h 9
HYPERLINK \l "_Toc196594210" 4. The Typology of Pronouns PAGEREF _Toc196594210 \h 9
HYPERLINK \l "_Toc196594211" References PAGEREF _Toc196594211 \h 11
Readings: Full references and links are in References at the end. These are all on the CD except for Kiparsky 2002, which is on the web.
(1) ADDIN EN.CITE Kiparsky2002995599555Kiparsky, PaulKaufmann, IngridStiebels, BarbaraDisjoint reference and the typology of pronounsMore than Words179-226Studia Grammatica 53
2002BerlinAkademie Verlaghttp://www.stanford.edu/~kiparsky/Papers/anaph.hierarchies-t.pdf http://www.stanford.edu/~kiparsky/Papers/anaph.hierarchies-t.pdf (Kiparsky 2002) Disjoint reference and the typology of pronouns (not on your CD, but available on the web)
(2) ADDIN EN.CITE Testelets20059938993834Testelets, YakovBinding and Anaphora: Handouts for Five Lectures2005Handouts for a short course at NYInstitute in St Petersburg, July 2005(Testelets 2005), St.Petersburg Lectures on Binding Theory, Lecture 2: Domains, orientation, and a typology of pronouns and Lecture 3: Long-distance binding and Revisions of the Binding Theory
(3) Bring ADDIN EN.CITE 2004984298426Bring, DanielBinding TheoryCambridge Textbooks in Linguistics2004CambridgeCambridge University Press(2004), Chapter 11 (Exempt anaphora and reflexivity).
Optional readings:
(4) ADDIN EN.CITE Fischer20049921992117Fischer, SilkeOptimal bindingNatural Language & Linguistic TheoryNatural Language & Linguistic Theory481526222004(Fischer 2004) Optimal binding
(5) ADDIN EN.CITE Lee20039948994817Lee, FeliciaAnaphoric R-expressions as bound variablesSyntaxSyntax84-114612003(Lee 2003) Anaphoric R-expressions as bound variables
(6) ADDIN EN.CITE Cole2001997699765Cole, PeterHermon, GabriellaHuang, C.-T. JamesIntroductionLong-Distance Reflexives. Syntax and Semantics 332001New YorkAcademic Presshttp://www.ling.udel.edu/pcole/LD_Introduction.PDF http://www.ling.udel.edu/pcole/LD_Introduction.PDF (Cole et al. 2001) For next week: Introduction, which is online, plus possible articles relevant to your presentations that can be copied from the book (in LaTyp collection).
Background: typology and theory
Typology and theory are intimately linked, and this is especially clear in the study of anaphora! This point is emphasized from the very beginning in Testeletss Lectures on Binding Theory, and it is emphasized throughout Kiparskys article. (It can also be seen in many of the other works on anaphora, including Brings book, Reinhart and Reulands work, Pollard and Sags work, )
What has to be explained: Different languages have different kinds of pronouns and reflexives, but systems of anaphoric expressions are in fact remarkably similar. A good theory of anaphora should include a basis for predicting what kinds of pronouns and reflexives are possible in human language and what kinds are not. See Testelets, Lecture 1, for clear examples of imaginable systems which never occur, and which linguists quite confidently predict will never be found.
A theory which is too constrained will be falsified by anaphoric systems it predicted to be impossible. That happened with Chomskys early Binding Theory: since Principle A and Principle B operated in the same domain, it predicted that pronouns and reflexives should always be in complementary distribution, which we have seen is not true, even in English.
A theory which is not constrained enough may not be falsified directly, but it may predict too many possibilities, and will fail to explain why we find just the patterns that we do and not any of the patterns that are never attested.
A theory which is based on wrong notions, or is missing some of the important right ideas, may fail in both ways. Without a very wide array of typological data, its hard to distinguish right ideas from wrong ideas. Without a rich theoretical framework, its hard to understand the data and draw the right kinds of typological generalizations.
Kiparskys approach to anaphora typology
Note on terminology: Kiparsky uses the term pronoun as a general term for anaphoric expressions, including both plain pronouns and reflexives. We will mostly follow his terminology here; Ill write plain pronoun or non-reflexive pronoun when I want to refer to pronouns in the narrower sense.
2.1. On Obviation vs. Blocking
When it was discovered that pronouns and reflexives are not always in complementary distribution, linguists started looking for ways to make Principle A and Principle B distinct different local domains, different notions of binding, or something else.
Kiparsky makes the useful distinction between Obviation approaches and Blocking approaches, and himself argues that it shouldnt be viewed as an either-or matter, but that in fact both are real and a good theory should use both. Optimality theory makes that easier to do than some other theories.
The Obviation approach: Originating with Lasnik ADDIN EN.CITE 19769977997717Lasnik, HowardRemarks on coreferenceLinguistic AnalysisLinguistic AnalysisLinguistic Analysis1-2221976(1976), developed by Chomsky and others in the GB tradition: autonomous disjoint reference principles (like Principle B), which filter out illicit coindexations in certain structural domains. The argument that the relevant domain for Principle B effects should be defined in terms of a predicates arguments was argued by Hellan ADDIN EN.CITE 1983997899786Hellan, LarsAnaphora in Norwegian and Theory of BindingWorking papers in Scandinavian SyntaxWorking Papers in Scandinavian Syntax51983TrondheimUniversity of Trondheim1988997299726Hellan, LarsReflexives in Norwegian and the Theory of Grammar1988DordrechtForis(1983, 1988), Sells ADDIN EN.CITE 1986997999795Sells, PeterBerman, S.Choe, J.-W. McDonough, J.Coreference and bound anaphora: a restatement of the factsNELS 16434-446161986AmherstGLSA, University of Massachusetts(1986), and Farmer and Harnish ADDIN EN.CITE Farmer1987998099805Farmer, A.Harnish, M.Papi, M.Verschueren, J.Communicative reference with pronounsThe Pragmatic Perspective1987AmsterdamBenjamins(Farmer and Harnish 1987). They all proposed some version of the following principle, which is sometimes called Obviation, a term borrowed from Algonquian grammar.
( LISTNUM NumberDefault \l 1 ) Obviation: Coarguments have disjoint reference.
A second key idea, which we have discussed before, is that disjoint reference is a principle of semantic interpretation that requires certain types of arguments of multi-place predicates to be distinct ADDIN EN.CITE Montague1970391539155Montague, RichardVisentini et al., BrunoEnglish as a Formal LanguageLinguaggi nella Societ e nella Tecnica189-224Reprinted in Montague 1974, 188-2211970MilanEdizioni di ComunitKeenan1974866886685Edward KeenanThe functional principle: generalizing the notion of 'subject of'CLS 10: Papers from the Tenth Meeting of the Chicago Linguistic Society298-3091974ChicagoChicago Linguistic SocietyBach1980840284025Bach, EmmonPartee, Barbara H.Kreiman, JodyOjeda, Almerindo E.Anaphora and semantic structurePapers from the Parasession on Pronouns and Anaphora1-281980ChicagoChicago Linguistic Societyhttps://udrive.oit.umass.edu/partee/Bach_Partee80_Anaphora.pdfhttps://udrive.oit.umass.edu/partee/Bach_Partee80_Anaphora.pdfPollard19924381438117Pollard, CarlSag, Ivan A.Anaphors in English and the Scope of Binding TheoryLinguistic InquiryLinguistic InquiryLinguistic InquiryLinguistic Inquiry261-30323.2ReciprocalReflexive1992http://lingo.stanford.edu/sag/papers/LI-binding.pdfhttp://lingo.stanford.edu/sag/papers/LI-binding.pdfSells19914946494617Sells, PeterDisjoint Reference into NPLinguistics and PhilosophyLinguistics and PhilosophyLinguistics and PhilosophyLinguistics and Philosophy151-17014GerundNominals1991Reinhart19934548454817Reinhart, Tanya Reuland, EricReflexivityLinguistic InquiryLinguistic InquiryLinguistic InquiryLinguistic Inquiry657-72024BindingTheory1993(Montague 1970, Keenan 1974, Bach and Partee 1980, Sells 1991, Pollard and Sag 1992, Reinhart and Reuland 1993). Reinhart and Reuland explicitly argue, as we saw, that Principle A is defined on syntactic predicates and Principle B on semantic predicates.
Kiparskys approach adopts both of these ideas about obviation. Obviation will be argued to be a universal violable constraint interacting with other constraints in the ranked constraint systems that define the binding patterns of languages. (p. 3) (page numbers refer to the downloaded version from his website.)
Blocking: The second approach to disjoint reference replaces Principle B by a Blocking constraint that makes anaphors obligatory in their binding domain under the appropriate conditions, blocking the coreferential interpretation of pronominals in that domain. Blocking can be seen as either a grammatical condition applying to syntactic representations (we havent studied any of those) or as an extra-grammatical, pragmatic, condition involving conversational implicatures, hearer strategies, or the like ADDIN EN.CITE Dowty1980998199815Dowty, DavidKreiman, JodyOjeda, AlmerindoComments on the paper by Bach and ParteeCLS Parasession on Pronouns and Anaphora29-401980ChicagoCLSReinhart1983454745476Reinhart, TanyaAnaphora and semantic interpretation1983LondonCroom Helm19837614761417Reinhart, TanyaCoreference and bound anaphora: A restatement of the anaphora questionLinguistics and PhilosophyLinguistics and Philosophy47-8861983Farmer1987998099805Farmer, A.Harnish, M.Papi, M.Verschueren, J.Communicative reference with pronounsThe Pragmatic Perspective1987AmsterdamBenjaminsLevinson19873447344717Levinson, Stephen C.Pragmatics and the grammar of anaphora: a partial pragmatic reduction of Binding and Control phenomenaJournal of LinguisticsJournal of LinguisticsJournal of LinguisticsJournal of Linguistics379-43423PragmaticsGrice198719913448344817Levinson, Stephen C.Pragmatic reduction of binding conditions revisitedJournal of LinguisticsJournal of LinguisticsJournal of LinguisticsJournal of Linguistics107-16127AnaphoraPragmatics19912000998299826Levinson, Stephen C.Presumptive Meanings2000Cambridge, MABradford Books(Dowty 1980, Reinhart 1983a, 1983b, Farmer and Harnish 1987, Levinson 1987, 1991, 2000).
Kiparsky argues that both Obviation and Blocking constitute universal constraints that interact with each other and with other constraints in accord with rankings that are partly language-specific. He argues for their independence by showing that one applies to a syntactic representation (Blocking) and the other to semantic interpretation (Obviation), that they apply to different classes of elements, and that they apply in different domains.
Kiparsky sees Obviation and Blocking as conceptually different as well. Blocking is related to the very general Elsewhere principle that governs morphological expression in many domains; Kiparsky follows Burzio in treating it as an Economy principle. Obviation, on the other hand, is a constraint that is specific to binding theory.
2.2. The typology of anaphoric expressions
Kiparsky credits Faltz ADDIN EN.CITE Faltz19779963996332Faltz, Leonard M.Reflexivization: A Study in Universal SyntaxLinguistics DepartmentPh.D.1977University of California at BerkeleyPh.D. dissertation(Faltz 1977) with pioneering work showing that the binding properties of reflexives and reciprocals vary along at least two dimensions:
the size of the domain within which they must be bound
the nature of the antecedent in the clausal domain.
Faltz also observed that different anaphoric expressions within one language may differ in these respects, and that therefore binding domain and antecedent requirement are lexical properties of individual anaphors, not a syntactic parameter of the language as a whole. Later work on many languages confirmed these conclusions.
Kiparsky argues that the typology of pronouns comes from two cross-classifying properties:
whether the expression is obviative or not, i.e. whether it can license co-referential arguments;
what the expressions antecedent domain is.
With respect to antecedent domain, Kiparsky identifies a hierarchy of five successively more inclusive antecedent domains, each characterizing a class of pronouns. When these five types are further cross-classified as +/- obviative (also known as obviative vs. proximate), the resulting inventory predicts 10 types of anaphoric expressions (actually 9, since one is inherently impossible); Kiparsky argues that although some of the predicted types are rare, his typology matches the range of known anaphoric expressions fairly well. (Some of the data in ADDIN EN.CITE Huang2000998399836Huang, YanAnaphora: A Cross-Linguistic Study2000OxfordOxford University Press(Huang 2000) is acknowledged to pose difficulties.)
I will not try to summarize all of Kiparskys arguments, but will concentrate on some examples, with three goals: (i) to show how Optimality Theory works; (ii) to show how different rankings of constraints capture the behavior of different anaphoric expressions; and (iii) to highlight some of his typological distinctions. There are many interesting ideas in the paper that I will omit today.
3. Coargument Disjoint Reference.
3.1. Four generalizations about disjoint reference
In English, pronouns are obviative and reflexives (anaphors in Binding Theory) are proximate. But many languagse, including Swedish, distinguish between obviative and proximate reflexives (sig vs. sig sjlv), and some distinguish between obviative and proximate pronouns (Algonquian). So the obviation requirement and the choice of binding domain are in principle independent, though they may not be independent in each language.
Obviation accounts for Principle B effects. It also accounts for four disjoint reference phenomena that are not addressed in standard versions of Binding Theory.
( LISTNUM NumberDefault \l 1 ) Generalization 1: An anaphor whose antecedent is a coargument has a bound variable reading but not a coreferential reading. (This one is very important for us.)
( LISTNUM NumberDefault \l 1 ) Generalization 2: A plural or conjoined DP which overlaps in reference with a coargument has a collective reading but not a distributive reading. (We wont discuss this one.)
The other two generalizations are syntactic and are overtly instantiated in only a minority of languages; Generalization 3 concerns languages that distinguish +/- obviative reflexives, and Generalization 4 concerns languages that distinguish +/- obviative pronouns.
( LISTNUM NumberDefault \l 1 ) Generalization 3: An obviative and its coarguments have disjoint reference. (See Swedish example later below)
( LISTNUM NumberDefault \l 1 ) Generalization 4: At most one coargument may be proximate. (Algonquian; we wont discuss this one.)
3.2. Bound variable vs. coreferential (covaluation) readings
Generalization 1 is famously manifested in VP anaphora, as we have illustrated before. English reflexives whose antecedent is a coargument get only the bound variable reading (sloppy identity), as in ( REF JohnHatesHimselfAndsoDoesFred 6). What we have not emphasized before, though we mentioned it last week, is that English reflexives whose antecedent is not a coargumen t c a n b e a m b i g u o u s b e t w e e n b o u n d v a r i a b l e a n d c o r e f e r e n t i a l r e a d i n g s , a s i n ( R E F J o h n C o n s i d e r s H i m s e l f C o m p e t e n t A n d S o D o e s 7 ) .
( L I S T N U M N u m b e r D e f a u l t \ l 1 ) J o h n h a t e s h i m s e l f , a n d s o d o e s F r e d . ( u n a m b i g u o u s )
a . `" F r e d h a t e s J o h n t o o . ( n o s t r i c t i d e n t i t y , c o r e f e r e n t i a l r e a d i n g )
b . m e a n s : F r e d h a t e s h i m s e l f t o o . ( s l o p p y i d e n t i t y o n l y : b o u n d v a r i a b l e r e a d i n g )
( L I S T N U M N u m b e r D e f a u l t \ l 1 ) J o h n c o n s i d e r s [ h i m s e l f c o m p e t e n t ] , a n d s o d o e s F r e d . ( a m b i g u o u s )
a. Fred considers [John incompetent] too. (strict)
b. Fred considers [himself incompetent] too. (sloppy)
Other ambiguous examples cited by Kiparsky (some of the data may not be agreed on by everyone), all with reflexives whose antecedent is not a coargument, include the following.
( LISTNUM NumberDefault \l 1 ) a. John thought that Marys parents would approve of someone like himself, and so did Fred. (ambiguous)
b. John loves his wife, and so does Fred. (ambiguous; his is not marked for reflexivity)
c. John has a picture of himself, and so does Fred.
3.2.1. What is the local domain?
Kiparsky considers the binding in both ( REF JohnHatesHimselfAndsoDoesFred 6) and ( REF JohnConsidersHimselfCompetentAndSoDoes 7) to be local; he distinguishes local syntactic domains for binding (i.e. binding domains) from the semantic co-argument domain, which he does not consider to be a binding domain at all, but rather the domain of the universal Obviation principle. He argues that the real minimal binding domain is the accessible subject domain, the minimal domain in which there is a subject.
3.2.2. Disambiguation in Russian
In English, the distinction between the bound variable reading and the coreferential reading of pronouns normally has no direct formal reflex. But in Russian, the reflexives svoj (poss.) and sebja (acc.) can have not only third-person antecedents but also first and second person antecedents, and in that case they compete with regular first and second person possessives.
Kiparsky asserts, based on ADDIN EN.CITE Dahl19739984998417Dahl, stenOn so-called 'Sloppy Identity'SynthseSyntheseSynthese261973(Dahl 1973), that it is precisely in the coargument case ( REF JaLjubljuOdnogoMenjaOdnogoSebja 9) that the reflexive is obligatory, and there it has only the bound variable reading. (Class: does everyone agree?)
( LISTNUM NumberDefault \l 1 ) a. *Ja ljublju odnogo menja.
I like only me.
b. Ja ljublju odnogo sebja. (bound variable reading only right?)
I like only myself.
When the pronoun is not a co a r g u m e n t o f i t s a n t e c e d e n t , t h e c o n t r a s t b e t w e e n p r o n o u n a n d r e f l e x i v e ( b o t h f o r p l a i n p r o n o u n s a n d f o r p o s s e s s i v e s ) m a r k s t h e s e m a n t i c c o n t r a s t b e t w e e n c o r e f e r e n t i a l v s . b o u n d v a r i a b l e r e a d i n g :
( L I S T N U M N u m b e r D e f a u l t \ l 1 ) a . J a l j u b l j u m o j u ~e n u , i I v a n t o ~e . ( s t r i c t , c o r e f e r e n t i a l )
I l o v e m y w i f e , a n d I v a n a l s o ( l o v e s m y w i f e ) .
b . J a l j u b l j u s v o j u ~e n u , i I v a n t o ~e . ( s l o p p y , b o u n d v a r i a b l e )
I l o v e m y ( s e l f s ) w i f e , a n d I v a n a l s o ( l o v e s h i s w i f e ) .
Next step: to formulate constraints that will account for the Russian data, for generalizations 1-4 in Section 3.1 about English, and for Swedish and Algonquian data to follow.
3.2.3. Towards an analysis of the Russian data
First of all, how to conceptualize the three possible relations between a pronoun and a (potential) antecedent, namely bound anaphora, coreference, and disjoint reference?
Bound anaphora: coindexed; associated with the same variable. (cf. the discussion in Bring and in our previous lectures the antecedent will become sister to a binding operator, and just the anaphoric expression will be interpreted as a variable. But we wont try to make that explicit here.)
Coreference or covaluation: distinct variables, but given the same assignment (same individual, or perhaps same discourse referent)
Disjoint reference: distinct variables assigned to distinct individuals.
Kiparsky uses the following mnemonic notation:
( LISTNUM NumberDefault \l 1 ) a. Bound anaphora: Ax Bx (x ( a)
b. Co r e f e r e n c e : A x & B y ( x ( a , y ( a )
c . D i s j o i n t r e f e r e n c e : A x & B y ( x ( a , y ( b , a `" b )
H e n o t e s t h a t s i n c e w e n e e d t o d i s t i n g u i s h t h e s e t h r e e c a s e s , B i n d i n g T h e o r y c a n t b e b a s e d s i m p l y o n c o i n d e x i n g ; t h e r e a r e a l s o c o n s t r a i n t s o n t h e i n t e r p r e t a t i o n of non-coindexed arguments.
For his Optimality Theoretic approach, there will be inputs and outputs, and a given set of candidate input-output pairs will be evaluated to find the most harmonic output for a given input. The input candidates will be logical forms (semantically annotated syntax), with any coindexation and any assignment of values to variables. The outputs will be morphological forms.
Kiparsky proposes six constraints, in two groups. The first three are specific to binding theory, the second is a very general Parse constraint, and the last two are markedness (economy) constraints.
( LISTNUM NumberDefault \l 1 ) Binding Constraints:
a. Binding Domain: A pronoun has a compatible antecedent in a designated domain D.
b. Obviation: An obviative an its coarguments have disjoint reference.
c. Prox: A proximate is a bound anaphor (i.e. it is indexed to the same variable as its antecedent)
( LISTNUM NumberDefault \l 1 ) The Parse constraint: This bars the null candidate; it basically says that for every possible input, there must be an output. But it is outranked by some other constraints, so in fact some inputs have no outputs.
( LISTNUM NumberDefault \l 1 ) Economy Constraints:
a. Featural (Semantic) Economy: Avoid pronominals.
b. Morphological Economy: Avoid morphologically complex pronouns.
The idea behind the Featural Economy constraint is that pronominals have richer intrinsic semantic feature content than anaphors; anaphors are more fully dependent. Morphological Economy favors morphologically simple pronouns (like sig) over complex ones (sig sjlv).
Assumptions about Russian vs. English: Pronouns of Russian have the same properties as the corresponding pronouns of English, except that the reflexives svoj, sebja have no person features. As in English, the reflexives are proximate, and the pronominals (menja, moju) are obviative.
The tableau in ( REF RussianTableau 15) indicates the relevant properties of the input logical forms using the notation in ( REF NotationForLogicalFormsBoundCorefDisjRef \* MERGEFORMAT 11). Candidate sets 1 and 2 illustrate coargument reflexives, with bound anaphora and (the impossible) coreferential readings, respectively. Candidate sets 3 and 4 illustrate non-coargument possessive reflexives, with the same two readings (this time both possible, one reading with svoju and the other with moju). The single set of ranked constraints succeeds in picking out the right input-output pairs. (The Binding Domain constraint plays no role in these particular examples; it will later. The two Economy constraints are ranked together in the lowest position; when both are violated (see later Swedish example), there will be two *s.
How to read the OT tableau: First, think of it as four separate tableaus, each three lines long, that have been put together for efficiency. Each one concerns one input and a choice among two outputs plus a third choice no output: we actually have an example of no output in situation 2, meaning there is no way (at least not among those standard choices) to express that meaning in Russian.
The columns show the ranked constraints: the highest-ranked constraint is BD (Binding Domain) and the lowest ranked (pair of) constraint(s) is Economy.
The rows 1a, 1b, 1c show the three candidate input-output pairs for the input where antecedent and anaphoric expression are to express the bound variable reading, both assigned to a single variable x. (Imagine that input to be repeated on the left side of each line.) The three candidate outputs are Jax ljublju menjax, Jax ljublju sebjax, and ( (no output). In the cells on the right, each * marks a violation of the constraint that heads the given column.
The candidate that wins is the one that violates no constraint or violates a less-highly ranked constraint than any other candidate. (It doesnt matter how many constraints a candidate violates; only the highest-ranked constraint violated by each candidate matters.) For each of the losers, the highest-ranked violation is indicated by !*. The winner is marked on the left by the pointing hand (.
Gaps in the output (() violate the Parse constraint. Gaps occur when all of the available candidates violate some constraint thats higher ranked than Parse. The gap in the Russian data (which is paralleled by English) occurs when one tries to figure out how to express a coargument reflexive with a coreferential rather than a bound variable reading: at least with these choices, it cant be done.
( LISTNUM NumberDefault \l 1 )
Exercises for the reader: Ignoring constraint BD for the moment, are there any other rankings of the four relevant constraints that would give the same results? What would change if you interchange (just) the rankings of Obviation and Parse? What would change if you interchange (just) the rankings of Prox and Parse?
3.3. Illustration: Swedish sig, an obviative reflexive. Cancellation of obviativity by sjlv.
(41) (first half), p. 18
BDObviationProxParseEconomyCoarguments1a x ( a Johnx lskar ___x *sig *!1b ( sig sjlv*1c *honom*!*1d *honom sjlv**!1e ( (gap)? No *!2a x ( a, y ( a *sig*!2b *sig sjlv*!*2c *honom*!*2d *honom sjlv*!**2e ( ( (gap)? Yes*3a x ( a, y ( b *sig*!3b *sig sjlv*!**3c ( honom*3d *honom sjlv*!**3e ( (gap)? No*!
Consider 1a-1e first. Here we consider what forms can and cannot be used to express bound anaphora in a simple coargument sentence like Kiparskys (23a) *John fedrar sig John prefers self, (29a) Johni lskar honomi Johni loves himi, (31a) John lskar sig sjlv John loves refl+self, and (not cited in the article, but described) (29a) *Johni lskar honomi sjlv Johni loves himi+self.
The input is the logical form abbreviated by x ( a, i.e. bound variable anaphora. The candidate outputs are the four alternative forms for such a sentence with the four pronominal forms, plus the gap possibility (.
The ranking of the constraints is just the same as in the Russian example. What is different is the lexical specification of the candidate output forms: sig, like honom and unlike Russian sebja, is obviative. And sig sjlv and honom sjlv are both proximate. What makes sig and sig sjlv reflexives is Condition A, here represented as the Binding Domain constraint: that plays a role in Case 3, the Disjoint Reference situation: Both sig and sig sjlv violate that top-ranked constraint in 3, because with that indexing they have no compatible antecedent within the local domain.
See the article for the other half of this tableau, with examples involving non-coarguments.
3.4. Gaps
When we discussed earlier the fact that pronouns and reflexives are not in complementary distribution in many languages, we followed the literature in citing contexts where both are permitted. Kiparsky draws attention to another aspect of non-complementarity that emerges on his input-output perspective, namely situations where neither can occur. This shows that Blocking isnt enough, as Reinhart (and Bach and Partee) had hoped. The gap in the coreference alternative in both the Russian tableau, case 2, and the Swedish tableau, case 2, shows that Obviation can block a pronoun from occurring even in a case where a reflexive cannot occur either. Of course these gaps are not syntactic positions where neither can occur, but rather logical forms syntax plus interpretive choice that cant be expressed.
His paper gives further examples of such gaps in English and German.
As he notes, violation of either Obviation or Blocking (Economy) produces deviancy, but the sharpest judgments of ungrammaticality come from joint violation of both.
3.5. Morphologically marked Obviation
(Discussion of Algonquian languages such as Cree and Ojibwa, with further evidence that obviation and blocking are independent constraints in Ojibwa.)
4. The Typology of Pronouns
Table (59), p. 25
The table results from four binary divisions.
(60) a. A pronoun may be referentially independent or referentially dependent. Referentially independent pronouns can (but need not) introduce something new into the discourse for instance via deictic or demonstrative uses (in the narrow sense). Referentially dependent pronouns cannot introduce anything new they must have at least a discourse antecedent.
b. Referentially dependent pronouns may be reflexive or non-reflexive. (Referentially independent pronouns are necessarily non-reflexive.) Reflexive pronouns need a syntactic antecedent. Non-reflexives can (but need not) get their reference from context/discourse.
c. Reflexive pronouns may be finite-bound or non-finite-bound. (Non-reflexive are necessarily non-bound.) Finite-bound pronouns require an antecedent within the same finite clause. Non-finite-bound pronouns can (but need not) have an antecedent within the same finite clause.
d. Finite-bound pronouns may be subject to the requirement that they be locally bound, or not. Locally bound pronouns require an antecedent in the first accessible subject domain. Non-locally bound pronouns can (but need not) have a long-distance antecedent.
The category of a pronoun is defined by the maximum domain in which its antecedent may be found. Here are some illustrative diagnostic contexts, which define a hierarchy of five domains.
(61) (p.26) a. Referentially independent:
Its ____ !
We need to talk about ____, ____, and ____.
b. Referentially dependent:
1. Non-reflexive:
Johni is here. I saw ___i. (discourse antecedent)
2. Reflexive:
i. Non-finite-bound:
Johni thought that I would criticize ___i.
Johni was sad. Why didnt Mary love ___i?(logophoric in strict sense)
ii. Finite-bound:
A. Non-locally bound:
Johni asked me to criticize ___i.
B. Locally bound:
Johni criticized ___i. (reflexive)
Johni showed Billj ___i,j in the mirror.
The five pronoun types cross-classify with the obviation property [+/- Obviative], yielding ten logically possible types of pronominal elements. The full typology, with a representative of each type, is given in (62) (p. 27).
The gap on the top right edge represents the absence of a non-obviative referentially independent pronoun: if such a pronoun existed, it would be a universal pronoun, able to be used both as a demonstrative and a reflexive. It is said to violate an Expressiveness universal, namely that in every language, all pronominal systems are expressive in the sense that they provide a means to mark both coreference and non-coreference in any domain. (Formally, it is equivalent to the assumption that Parse universally outranks Economy.)
What about the potential gap on the bottom left edge? Are there obviative locally bound reflexives? Obviation should prohibit it from having a coargument antecedent while its local binding requirement can only be satisfied by an antecedent in the same clause. Kiparsky suggests that one can distinguish between German and English by calling the German locally bound subject-oriented reflexive Obviative to prohibit it taking a direct object antecedent, as English reflexives can. So in that narrowest domain, +/- Obviative can be used to indicate +/- Subject-oriented, which otherwise would need a separate feature. Subject-oriented is taken to be a relativized version of the coargument disjoint reference requirement (obviation).
(62), p. 27: The full pronoun typology: 5 local domains ( 2 values for [( Obviative]
There is much more in the paper, including examples of all the kinds of pronouns in the typology, but this much may give a hint of how Kiparskys ideas and their implementation in Optimality Theory can provide an interesting and potentially very fruitful approach to the typology of anaphora.
Note: in your presentations about your languages, I dont expect you to figure out exactly which of Kiparskys 10 classes your various pronouns and reflexives correspond to. But you may be able to narrow down the possibilities enough to make some interesting distinctions.
References
ADDIN EN.REFLIST Bach, Emmon, and Barbara H. Partee. 1980. Anaphora and semantic structure. In Papers from the Parasession on Pronouns and Anaphora, eds. Jody Kreiman and Almerindo E. Ojeda, 1-28. Chicago: Chicago Linguistic Society. [Reprinted in Partee, Barbara H. 2004. Compositionality in Formal Semantics: Selected Papers by Barbara H. Partee. Oxford: Blackwell Publishing, 122-152]. https://udrive.oit.umass.edu/partee/Bach_Partee80_Anaphora.pdf
Bring, Daniel. 2004. Binding Theory: Cambridge Textbooks in Linguistics. Cambridge: Cambridge University Press.
Cole, Peter, Gabriella Hermon, and C.-T. James Huang. 2001. Introduction. In Long-Distance Reflexives. Syntax and Semantics 33. New York: Academic Press. HYPERLINK "http://www.ling.udel.edu/pcole/LD_Introduction.PDF" http://www.ling.udel.edu/pcole/LD_Introduction.PDF
Dahl, sten. 1973. On so-called 'Sloppy Identity'. Synthse 26.
Dowty, David. 1980. Comments on the paper by Bach and Partee. In CLS Parasession on Pronouns and Anaphora, eds. Jody Kreiman and Almerindo Ojeda, 29-40. Chicago: CLS.
Faltz, Leonard M. 1977. Reflexivization: A Study in Universal Syntax, Linguistics Department, University of California at Berkeley: Ph.D. dissertation.
Farmer, A., and M. Harnish. 1987. Communicative reference with pronouns. In The Pragmatic Perspective, eds. M. Papi and J. Verschueren. Amsterdam: Benjamins.
Fischer, Silke. 2004. Optimal binding. Natural Language & Linguistic Theory 22:481526.
Hellan, Lars. 1983. Anaphora in Norwegian and Theory of Binding.vol. 5: Working papers in Scandinavian Syntax. Trondheim: University of Trondheim.
Hellan, Lars. 1988. Reflexives in Norwegian and the Theory of Grammar. Dordrecht: Foris.
Huang, Yan. 2000. Anaphora: A Cross-Linguistic Study. Oxford: Oxford University Press.
Keenan, Edward. 1974. The functional principle: generalizing the notion of 'subject of'. In CLS 10: Papers from the Tenth Meeting of the Chicago Linguistic Society, 298-309. Chicago: Chicago Linguistic Society.
Kiparsky, Paul. 2002. Disjoint reference and the typology of pronouns. In More than Words, eds. Ingrid Kaufmann and Barbara Stiebels, 179-226. Berlin: Akademie Verlag. HYPERLINK "http://www.stanford.edu/~kiparsky/Papers/anaph.hierarchies-t.pdf" http://www.stanford.edu/~kiparsky/Papers/anaph.hierarchies-t.pdf
Lasnik, Howard. 1976. Remarks on coreference. Linguistic Analysis 2:1-22.
Lee, Felicia. 2003. Anaphoric R-expressions as bound variables. Syntax 6:84-114.
Levinson, Stephen C. 1987. Pragmatics and the grammar of anaphora: a partial pragmatic reduction of Binding and Control phenomena. Journal of Linguistics 23:379-434.
Levinson, Stephen C. 1991. Pragmatic reduction of binding conditions revisited. Journal of Linguistics 27:107-161.
Levinson, Stephen C. 2000. Presumptive Meanings
# $ % ' ) ; < = > U V W X t u v w y ˸˥ޡtktQ=t 'j hs hhn 0J UmH nH u*2j hs hhn >*B*UmH nH ph u hhn mH nH uhs hhn 0J mH nH u$j hs hhn 0J UmH nH u hy j hy UhH %hpQ 5CJ \aJ fH q
%h2 5CJ \aJ fH q
%hi 5CJ \aJ fH q
hi hH 5CJ aJ hpQ 5CJ aJ hi h^ 5CJ aJ < = ) \
P <
P
R
^ n o gdM gd<
[$
[$
[$
gdH $a$gd> a, , , y z # $ % & ' ( ) * + G H ۻ̰̝zfṴ̝ jw h UmH nH u 'j hs hhn 0J UmH nH u*2j hs hhn >*B*UmH nH ph u hhn mH nH u$j hs hhn 0J UmH nH u h mH nH u j} h UmH nH u j hhn UmH nH uhhn mH nH u hs hhn 0J mH nH uhhn CJ mH nH u !H I J i j k Ĺ{p{gMĹ<