focus won t go away but f features might
play

Focus Wont Go Away, But F-Features Might Daniel B uring, Vienna - PDF document

Focus Wont Go Away, But F-Features Might Daniel B uring, Vienna Division of Labor, Jan 22/23, T ubingen 2015 1 Introduction 1.1 The World As We Know It Focus Projection Rules: Mediate between Accents (or other prosodic features)


  1. Focus Won’t Go Away, But F-Features Might Daniel B¨ uring, Vienna Division of Labor, Jan 22/23, T¨ ubingen 2015 1 Introduction 1.1 The World As We Know It • Focus Projection Rules: Mediate between Accents (or other prosodic features) and F-markers • Focus Interpretation Rules, e.g. [ [] ] F : Produce sets of focus alternatives (or structured meanings) • Focus Conditions, e.g. „ : Regulate when focus alternatives have to meet which (pragmatic) criteria 1.2 The Program Goal: Elimination of F-Markers as bona fide citizens of grammar. Precondition: (Relevant aspects of) Prosodic Structure must be part of the input to interpretations (e.g. interpret PF+LF pairs, or assume monostratal syntax) Challenge: Get the right Focus Alternatives, emulate the effects of Focus Projection Rules Prospect: See if we get additional milage out of the F-less set-up 1

  2. [ [] ] F F-pattern 1 Ñ set 1 of F-alternatives Ý [ [] ] F F-pattern 2 Ñ set 2 of F-alternatives F- Ý accent pattern projection [ [] ] rules F F-pattern 3 Ñ set 3 of F-alternatives Ý [ [] ] F . . . . . . Ý Ñ $ , set 1 of F-alternatives Set of ’ / [ [] ] A ’ / set 2 of F-alternatives & . Potential accent pattern Ý Ñ F-Values set 3 of F-alternatives ’ / (Spof) ’ / . . . % - Figure 1: Strategies for Deriving Focus Values 1.3 Strategy 2 Step 1: Collecting POFOS 2.1 F Style Object Focus ` a la Rooth (1992) S Focus Retrieval H „ C mary saw sue ✔ S mary saw x Alternative Propagation VP saw x Alternative Propagation DP V DP F Alternative Introduction { mary } { saw } x Mary saw J ohn * 2

  3. VP Focus (still Rooth) S Focus Retrieval H „ C mary saw sue ✔ S mary Q Alternative Propagation VP F Alternative Introduction Q DP V DP F Alternative Introduction { mary } { saw } x Mary saw J ohn * 2.2 F-Less Style: Collecting (Potential) Focus Values S Focus Retrieval mary saw x „ C mary took off her glasses ✔ S Alternative Introduction p, mary Q, mary saw x Alternative Propagation VP Alternative Introduction Q, saw x Alternative Propagation DP V DP Alternative Introduction { mary } { saw } x Mary saw J ohn * 2.3 Focus Projection Rules (If you must. . . ) (1) Alternative Introduction is limited to A(lternative)-Eligible node. (2) constituent N is A-eligible iff a. it is a lexical item and accented, or b. it is a syntactic node and has an A-eligible head-daughter, or c. it is a syntactic head and has an A-eligible internal argument 3

  4. 3 Turning Things Around 3.1 Introducing Exclusion Constraints Instead of collecting possible foci, we collect Restrictions on Possible Foci . The idea is this: At every branching node, we add a constraint that the focus cannot contain alternatives based on the literal meaning of the strong daughter and an alternative to the weak daughter. S Alternative Exclusion x R john VP Alternative Exclusion R john DP V DP Mary saw J ohn * Condition S Focus Retrieval on „ C : At H least one element in C „ C what happened ✔ meets all { it started to rain, mary saw bob, Exclusion fred kissed sue,. . . ,mary kissed john, sue saw john,. . . } Constraints S Alternative Exclusion x R john If So: You may remove VP Alternative Exclusion any R john Exclusion Constraint DP V DP met by at least one Mary saw J ohn element in * C . (3) strong or strong : introduces the Exclusion Constraints weak weak D D D D D s x and D s [ [D w ] ] A (or the one which subsumes the other) (4) Condition: At the root node, the set of restrictions is empty. 3.2 Overfocussing I (5) Mary kissed John. — No, #Mary saw JOHN. 4

  5. S mary Q ✔ ! „ C mary kissed john/ what did mary do to john Condition S (Rooth): mary Q C P [[ S ]] F or VP F C Ď Q [[ S ]] F DP V F DP F { mary } R x Mary saw J ohn * Condition on „ C : At S least one element in C „ C mary kissed john ✘ meets all S Exclusion Constraints x R john VP If So: You may remove R john any Exclusion DP V DP Constraint met by at Mary saw J ohn least one * element in C . 4 Prosodic Demotion (‘Deaccenting’) (6) If in or , D w would be the D s D w D s D w strong daughter by default (D w has been Prosodically Demoted ), D s must be a narrow focus.(cf. ? ) In Exclusion Constraint Talk: add x z D s D w Condition on Prosodic Demotion: D w must be given. 5

  6. weak strong functional lexical head complement left projection right projection Figure 2: Some structural defaults, in descending order of importance. S x R z saw john VP john R z saw john DP V DP t mary u t saw u { john } Mary s aw John * • R z saw is def the set of all relations other than ‘saw’ • R z saw john is ‘do something involving John, other than seeing him’ • R z saw john is the complement of that set, i.e. doing something not involving John, or seeing John • hence R z saw john (the Exclusion Constraint) allows the set R z saw john as focus alternatives. • For perspicuity, I will henceforth write R z saw john instead of R z saw john . • Luckily it turns out that x R john . is equivalent to x R z saw john , so I will write that instead. S x R z saw john VP john R z saw john DP V DP t mary u t saw u { john } Mary s aw John * (7) a. Did John’s sister run away when she saw John? — No, she KISSED John. b. Why wasn’t John there? — Because the police ARRESTED John. 6

  7. c. Did everybody leave when John arrived? — No, the police AR- RESTED John. (8) Did the police kill John? — No, the police ARRESTED John. 5 Question Answer: Forcing Deaccenting 5.1 Non-Final Focus Forces Deaccenting 5.2 No Random Shifts (Even in the Background) No Random Shifts (Even in the Background) (9) Who saw John? S l Given mary R z saw john „ C { bill saw john, jean saw ✘ john, mary saw john, fred saw john,. . . } S saw john x z mary saw john � Ø mary R z saw john VP John R z saw john DP V DP t mary u t saw u { john } M ary saw John * ¨ S l Given mary R z saw john „ C { bill saw john, jean saw john, ✘ mary saw john, fred saw john,. . . } S saw john � x z mary saw john Ø mary R z saw john VP John R z saw john DP V DP t mary u t saw u { john } M ary saw John ¨ * • the two Exclusion Constraints are incompatible: no single proposition can meet them both (nor can any regular question meaning) • Possibility 1: Completely ban inconsistent Exclusion Lists. 7

  8. • Possibility 2: Allow them, but insist that in this case we need two Focus Antecedents (on top of Givenness Antecedents) • Note that this only happens when you have Prosodic Demotion under Prosodic Demotion. In all other cases, the Exclusion Constraints are not only compatible, but in fact one always subsumes the other. 5.3 Hitting the Target (10) What did Mary do to John? S x z mary saw john, mary R z saw john „ C mary R john ✘ S x z mary saw john, mary saw john R z saw john VP john R z saw john DP V DP M ary saw John * ¨ 5.4 Non-Forced Deaccenting (11) What did John’s girlfriend do? 8

  9. S x R z kissed john „ C john’s girlfriend Q ✘ S x kissed john, x R z kissed john VP john R z kissed john DP V DP Mary kissed John * 6 Second Occurrence Focus {} (12) The kids only SKIMMED the S books. Even JOHN’s kid only w s even( C 1 ) {} skimmed the book. S z z john only „ C 1 s.b. else’s kid Prediction: If Prosodic Demotion only skimmed skimmed the book takes place within a P-Demoted ele- the books S s w ment, the focus introduced low needs m skimmed the book DP to be retrieved within the P-demoted ” skimmed the book VP element (which seems pretty accurate, JOHN w s cf. ? ) only( C ) {} VP R z skimmed the book „ C read the book VP s w skimmed DP the book 9

Recommend


More recommend