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

Prosodically marked mirativity Rett, Jessica 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

Download

Media
73804-Rett_J_Prosodically_marked_WCCFL38_2020_handout.pdf [ 2.69MB ]
73804-Rett_Jessica_Prosodically_marked_mirativity_slides.pdf [ 3.43MB ]
Metadata
JSON: 73804-1.0389873.json
JSON-LD: 73804-1.0389873-ld.json
RDF/XML (Pretty): 73804-1.0389873-rdf.xml
RDF/JSON: 73804-1.0389873-rdf.json
Turtle: 73804-1.0389873-turtle.txt
N-Triples: 73804-1.0389873-rdf-ntriples.txt
Original Record: 73804-1.0389873-source.json
Full Text
73804-1.0389873-fulltext.txt
Citation
73804-1.0389873.ris

Full Text

Prosodically marked mirativity˚Jessica RettWest Coast Conference in Formal Linguistics 38March 6, 20200 overview• mirativity is the implicit encoding of speaker surprise or exceededexpectation (DeLancey, 1997)• it’s encoded using a variety of strategies (Rett, 2012):– morphologically, in sentence particles (e.g. Finnish);– polysemously, in mirative evidentials (e.g. Cheyenne);– syntactically, in focus fronting (e.g. Spanish);– in prosody (e.g. English)• I’ll first present a semantic account of mirativity across these strate-gies qua ‘illocutionary content’ (Rett, 2019)• I’ll then present an investigation of the properties of prosodic mirativemarking in English• The Goals of the Talk:– a straightforward semantic analysis of mirativity;– a study of prosodically marked mirativity;– and a discussion of why the former might need to be supple-mented to fit the latter˚This talk is based in part on joint work led by Beth Sturman. It has also benefittedfrom comments from audiences at UCLA; the 6th Cornell Workshop in Linguistics andPhilosophy; the Berkeley Meaning Sciences Club; the Workshop on Encoding EmotiveAttitudes in Non-truth-conditional Meaning at DGfS 41; the Functional Categories andExpressive Meaning Workshop at the Universitat Auto`noma de Barcelona; and the Sec-ond UC Irvine Workshop in Logical Semantics. It was funded by a 2019 UCLA Councilon Research Faculty Research Grant.1Prosodically marked mirativity March 6, 20201 mirative marking1.1 defining mirativity• I use the term ‘mirative’ as a label for any natural-language expressionof exceeded expectation(1) a. Jane won the raceb. (Wow) Jane won the race!– the ‘expression’ bit (Kaplan, 1997; Castroviejo-Miro´, 2006)∗ like any expressive speech act, can be uttered insincerely(Searle, 1969)(2) (Wow) Airline seats are so tiny these days!∗ undeniable, non-negatable(3) A: (Wow) Those cupcakes are vegan!B: No, they’re vegetarian.B1:#No, you knew exactly how good they’d be.– the ‘exceeded expectation’ bit∗ Merin and Nikolaeva (2008); Rett (2011): speaker’s expec-tations violated or exceeded (cf. speaker surprise)“No matter how high my expectations might havebeen, what I have just heard exceeded them” (De-Lancey, 2001, 38).∗ a violated expectation can be flattering or insulting, depend-ing on the relevant expectations(4) You did better than the faculty expected you to.∗ generally speaker-oriented(5) a. #Mary said Jane won the race!b. *How did Jane won the race!?∗ tied to the here-and-now: Rett and Murray (2013); ‘novelinformation’ or ‘unprepared mind’ (DeLancey, 1997, 2001;Peterson, 2010); a “spontaneous reaction to a new, salient,often surprising event” (Aikhenvald, 2004, 197)WCCFL 38 2 rett@ucla.eduProsodically marked mirativity March 6, 20201.2 mirative strategiesthere are a variety of different strategies of mirativity marking (Rett, 2012):1. independent miratives• morphologic – bound: in Finnish, the sentence particle -pa¨ is afocus-sensitive mirative marker (Rett, 2019)(6) a. Ta¨a¨lla¨hereonbe-3rd.sgpaljona.lotkukk-ia.flower-prt.indf.pl‘There are lots of flowers here.’b. Ta¨a¨lla¨-pa¨here-paonbe-3rd.sgpaljona.lotkukk-ia.flower-prt.indf.pl‘(Wow) There are lots of flowers here!’• morphologic – free: in Mandarin there are mirative (and anti-mirative) adverbials (‘evaluative modals’; Wu, 2008).(7) ZhangsanZhangsanguoran/jingranmir/anti-mirlaicomele.part‘Zhangsan came (as expected/not expected by the speaker).’• syntactic: in Spanish, mirativity is marked via focus fronting(Cruschina, 2012, 2019):(8) ¡Imag´ınate!imagine.imp.2sg¡Conwithelthedirectordirectorquer´ıawant.impf.3sghablar!talk.inf‘Guess what! The director he wanted to talk to!’• prosodic: exclamation intonation, as in (1) (more soon)2. mixed-expression miratives• mirative conjunctions: second conjunct is surprising indepen-dently of the first (e.g. ‘lo and behold’; Russian, Malchukov 2004)(9) Onhezabolelfell.illdaconjiptclumer.died‘He fell ill and died (I didn’t expect it).’WCCFL 38 3 rett@ucla.eduProsodically marked mirativity March 6, 2020• expressive intensifiers like the German sau (‘female pig’); total(‘totally’); and voll (‘fully’; Gutzmann and Turgay 2011)(10) DiethePartypartywarwassauSAUcool.cool‘The party was very cool (I can’t believe how cool!).’3. dependent miratives• mirative evidentials receive their mirative interpretation in con-texts in which the content was recently learned through directevidence (Rett and Murray, 2013)(11) Motomotorcyclejo-nu-e.be-evid-declTsafiki (Dickinson, 2000)speaker hears motor: ‘It is apparently a motorcycle.’speaker thought he heard a car, but sees a motorcyclecoming: ‘It’s a motorcycle!’• more language-specific licensing requirements for mirativity:– Gitksan: licensed by the first person (Peterson, 1999)– Hare: licensed by the imperfect (DeLancey, 1997, 2001)(12) a. MaryMarye-we´’its-hidegha´layey˜idawork.perflo˜.lo˜‘Mary worked on hides (given what I’ve inferred/heard)’b. MaryMarye-we´’its-hidegha´layedawork.impflo˜.lo˜‘Mary is working on hides (I saw, to my surprise)’– Cheyenne: by the present tense (Murray, 2010, 2011)– Georgian: by individual-level predicates (Korotkova, 2012)2 a unified analysis• a discourse particle is a morphological strategy; discourse particlescan do a bunch of different stuff (Malamud and Stephenson, 2014)• mirativity is pretty semantically uniform across languages, but theway it’s encoded variesWCCFL 38 4 rett@ucla.eduProsodically marked mirativity March 6, 2020• intuitively, we should have a unified account of this ‘sememe’:– does mirativity behave the same way across languages/strategies?– if so, does it behave like any other sememe?– if so, how should we analyze it semantically?– (and how should we treat any differences?)• mirativity is not-at-issue (NAI) content– it’s undeniable in discourse– it cannot be targeted by truth-conditional operators(13) A: (Wow) Jane won the race!B: That’s not true, she came in second.B1:#That’s not true, you knew she would.(14) (Wow) Jane did not win the race!• the same goes for all of the mirative strategies listed previously– morphologically encoded independent markers like alas/pa¨;– dependent & mixed-expression markers like mirative evidentials2.1 illocutionary diagnostics• but mirative marking behaves differently from other types of NAI con-tent like appositives or evidentials, requiring distinct semantic treat-ment (Rett 2019)semantic contentat-issue not-at-issuedescriptive NAI illocutionary NAIWCCFL 38 5 rett@ucla.eduProsodically marked mirativity March 6, 2020• alternatively:semantic contentdescriptiveat-issue descriptive NAIillocutionaryillocutionary NAI• descriptive NAI markers: appositives, evidentials, utterance modi-fiers, CIs (Potts 2005)(15) a. Allegedly, Jane won the race.b. Frankly, Jane won the race.• illocutionary NAI markers: alas, (un)fortunately, mirative markerslike Finnish pa¨ (Rett 2019)(16) a. Alas, Jane won the race.b. (Wow) Jane won the race!• descriptive not-at-issue markers contribute to the descriptive contentof an utterance – what is said• “illocutionary content” is not-at-issue meaning pertaining to how thespeaker is using the utterance in the context• diagnostics of illocutionary content (Rett, 2019):1. susceptibility to Moore’s Paradox– standard Moore’s Paradox:(17) #It’s raining, but I don’t believe it’s raining.– Murray (2010): denial of mirativity results in Moore’s Para-dox; denial of evidentiality results in contradiction– to the extent that English speakers recognize this difference,I’ve found that the English data pattern the same way.WCCFL 38 6 rett@ucla.eduProsodically marked mirativity March 6, 2020(18) a. #KAllegedly, Jane lost the race, but no one al-leged she did.b. # Alas, Jane lost the race, but I’m not disap-pointed she did.c. # (Wow) Jane lost the race! But I’m not sur-prised she did.– MP is suspended in suppositional contexts (Yalcin 2007)(19) Suppose it’s raining, but I don’t believe it is raining.– so is the content encoded by emotive markers (20-b), butnot e.g. evidential adverbials (20-a)(20) a. #Suppose that, allegedly, Jane lost the race, but thatno one alleged that she did.b. Suppose that, alas, Jane lost the race, but that I’mnot disappointed she did.2. scope-taking– emotive markers associate with a single salient proposition...– ...this makes them incompatible with utterances associatedwith multiple propositions (Cheyenne, Rett and Murray 2013)(21) a. Mo´=e´-x-ho´1 ta˙heva´-hoo1oy/n=3-rem.pst-win-nar.3sgAe´nohe?Hawk‘Given the stories you heard, did Hawk win?’b. %Mo´=e´-ho´1 ta˙heva´-hoo1oy/n=3-win-nar.3sgAe´nohe?Hawk‘Given your surprise, did Hawk win?’– ...it also means they scope differently with sentential opera-tors than descriptive NAI markers(22) a. #Alas, if Jane loses, at least we’ll flip the Senate.b. If, alas, Jane loses, at least we’ll flip the Senate.(23) a. #Apparently, if Jane loses, I will run for office.b. #If, apparently, Jane loses, I will run for office.WCCFL 38 7 rett@ucla.eduProsodically marked mirativity March 6, 20202.2 a semantics for illocutionary content• analysis, informally:(24) (Wow) Jane, who hates politics, won the race!a. at-issue: Jane won the raceb. descriptive NAI: Jane hates politicsc. illocutionary NAI: speaker hadn’t expected Jane towin the race– at-issue content:∗ acts as a proposal to submit the proposition to the CommonGround (Stalnaker, 1973)∗ modeled in Farkas and Bruce (2010) as an update to theProjected Set– descriptive not-at-issue content:∗ directly updates the Common Ground (Murray, 2010)– illocutionary content:∗ updates speaker’s Discourse Commitments (Gunlogson, 2001)• a note on sincerity conditions– extant dynamic accounts of declarative mood do not formallymodel the sincerity condition, i.e. ‘Speaker believes that p’– arguably, emotive markers like mirativity do the same sort ofthing, contribute something like ‘Speaker is surprised that p’∗ Searle (1969): Moore’s Paradox arises when a sincerity con-dition is denied∗ emotive markers also susceptible to Moore’s Paradox (Searleand Vanderveken 1985: alas qua speech-act modifier)– Discourse Commitments (Gunlogson, 2001):∗ originally proposed for speaker bias in rising declaratives∗ originally characterized in terms of the speaker’s beliefs∗ public: speaker “mutually recognized as committed to them”– I treat public commitment as a proxy for belief, for the purposeof modeling (sincere) conversationWCCFL 38 8 rett@ucla.eduProsodically marked mirativity March 6, 2020• Farkas and Bruce’s model of a discourse structure K includes:1. the common ground (CG), the set of propositions believed byall discourse participants (for the purpose of the conversation);2. sets of discourse commitments (DC): for each participant x,the set of propositions x publicly commits to in the conversation;3. the Table T , modeling discourse salience;4. the projection set (ps), the set of beliefs that are being con-sidered for addition into the CG.• they adopt their treatment of illocutionary mood from Krifka (2001),but they do not differentiate between at-issue and not-at-issue con-tent; I borrow from Murray (2010, 2014) to do this (in (iv)).(25) Declarative operator (D), for sentences Sp with at-issue con-tent p and not-at-issue content q:D(Sp, a,Ki) = Ko such that(i) DCa,o “ DCa,i Y txbelieves, pyu(ii) To “ pushpxSp; tpuy, Tiq(iii) pso “ psi Y tpu(iv) CGo = CGi Y tqu• mirative markers and other emotive markers operate on propositionalcontent by placing that content on the top of the stack and addingthe speaker’s emotive attitude towards that content to her DC set(26) Mir, for clauses C with content p: MirpC, a,Kiq “ pC, a,Koqsuch that(i) DCa,o “ DCa,i Y txis-surprised, py}(ii) To “ pushpxC; tpuy, Tiq(27) JJane won the race!K = D(MirpS, a,Kiqq “ Ko such that(i) DCa,o “ DCa,i Y txbelieves, Jane won the racey}(ii) To “ pushpxS; Jane won the racey, Tiq(iii) pso “ psi Y {Jane won the race}(iv) DCa,o “ DCa,i Y {xis-surprised, Jane won the racey}• emotive markers interact scopally with descriptive NAI content andat-issue content via sub-sentential dynamic update (Appendix B)WCCFL 38 9 rett@ucla.eduProsodically marked mirativity March 6, 2020• Take Home Message: Mirative markers, like all emotive markers, en-code NAI content. But it’s qualitatively different than canonical,‘descriptive’ NAI content: it operates at an illocutionary level3 prosodically marked mirativity• unified semantic treatment of mirativity (and emotive markers):– some languages do things with morphemes and syntax that otherlanguages do with prosody– a good reminder that prosody needs to be represented in ourcompositional semantics (Pierrehumbert and Hirschberg, 1990)• previous work on the semantics of prosody:– prosody can mark illocutionary mood (Pierrehumbert, 1980; Jeongand Potts, 2016)– prosody can mark orientation (speaker or hearer; Gunlogson2001; Rudin 2018– prosody can mark other content like uncertainty, incredulity(Hirschberg and Ward, 1992)• my claim: English exclamation intonation is multi-faceted, markingsomething like illocutionary content as well as secondary effects likesurprise, emphasis• English mirativity marking– sentence exclamation(28) a. Jane carves gorgeous sculptures.b. (Wow) Jane carves gorgeous sculptures!– exclamatives(29) a. (Wow) What gorgeous sculptures Jane carves!b. (Wow) Can Jane carve gorgeous sculptures!c. (Wow) The gorgeous sculptures Jane carves!WCCFL 38 10 rett@ucla.eduProsodically marked mirativity March 6, 2020– discourse particles optional– distinct from rhetorical questions, e.g. How cool is that?– exclamatives differ from sentence exclamations, in English atleast, in that the object of surprise needs to be a degree insteadof a proposition (Rett, 2011)3.1 prosody: an overview• in English, utterances associated with intonational contours, i.e. tunes– tunes are composed of one or more intermediate phrases– intermediate phrases each have a pitch accent– pitch is measured wrt speaker’s fundamental frequency (f0)– pitch accents composed of one or more pitch targets, with thestar anchored to the stressed syllable– intermediate phrases are distinguishable in having:1. their own pitch accents (with a stressed syllable);2. their own phrase accent (with final lengthening);3. pitch range reset (a new ceiling after a down step)• Annotated with MAE ToBI (Beckman and Ayers Elam, 1997)– relatively coarse: can’t mark some things, e.g. speaker’s relativeheight (high vs. extra-high)• intermediate phrasesDoes the governor of Iowa endorse a radio program?L* H-H%– QUD: Which governors endorse a radio program?WCCFL 38 11 rett@ucla.eduProsodically marked mirativity March 6, 2020Does the governor of Iowa endorse a radio program?L* H- L* H-H%– QUD: out of the blue (Hirschberg and Ward, 1992)• the semantics of pitch accents– different canonical tunes (pitch accents`boundary tones) areassociated with different illocutionary moods (Pierrehumbert,1980; Pierrehumbert and Hirschberg, 1990)∗ declarative sentences: H* L-L%(30) Miriam made the lemonade.∗ polar questions & rising declaratives: L* H-H%(31) Miriam made the lemonade?∗ confirmation questions/rising imperatives: H* H-H%(32) Make the lemonade?– ...as well as perlocutionary effects (Jeong and Potts, 2016)∗ falling intonation: authoritativeness/assertion∗ level intonation: annoyance∗ rising intonation: politeness/positivity (cf. hedging)3.2 elicitation• research questions1. are English exclamations marked with uniform prosody?2. if so, how are they marked?3. how can we represent it semantically?4. (is the prosodic marking of mirativity semantically arbitrary?)WCCFL 38 12 rett@ucla.eduProsodically marked mirativity March 6, 2020• elicitation methodology– two consultants: one female, one male– conditions: 4 (construction) x 2 (˘discourse particle)– stimuli: 32 items per condition (256 total), between subjects(33) You dont expect Julian to make beautiful paintings, butyou find out he did. You tell Sara:a. (Wow) Julian makes beautiful paintings!b. (Wow)What beautiful paintings Julian makes!c. (Wow) Does Julian make beautiful paintings!d. (Wow) The beautiful paintings Julian makes!(34) You don’t expect Ariel’s stories to cause confusion, butyou find out they have. You tell Sara:a. (Wow) Ariel’s stories caused confusion!b. (Wow) What confusion Ariel’s stories caused!c. (Wow) Did Ariel’s stories cause confusion!d. (Wow) The confusion Ariel’s stories caused!– fillers: 4 construction conditions: declaratives; wh-questions; po-lar questions; definite subjects/clefts– 32 in each condition (128 total), btwn subjects (64 ea.)(35) a. Anna is good at chess.b. What is Anna good at?c. Is Anna good at chess?d. The game Anna is good at is chess.(36) a. James is an expert at knitting.b. What is James an expert at?c. Is James an expert at knitting?d. The thing James is an expert at is knitting.– consultants were instructed to read the sentences as they’d benaturally uttered in the provided context– the recordings were transcribed in Praat using MAE ToBI by oneof the authors, and confirmed by another trained transcriberWCCFL 38 13 rett@ucla.eduProsodically marked mirativity March 6, 20203.3 elicitation results• both consultants consistently:1. pronounced exclamations with L`H* pitch accents;L`H* H* H`!H* L* L*`HSE .96 .04 – – –WH .85 .10 .05 – –INV .94 .04 – – .02NOM .93 .03 – .04 –average .92 .05 .01 .01 ă.01Table 1: proportion of items with L`H* pitch accents, femaleL`H* H* H`!H* L* L*`HSE 1 – – – –WH .96 .04 – – –INV 1 – – – –NOM .94 .06 – – –average .98 .03 – – –Table 2: proportion of items with L`H* pitch accents, male2. pronounced exclamations with extra-high targets:– the high target exceeds expected height for position (after adownstep, etc.);– target exceeds default pitch maximum by at least 5%– cf. fillers, with zero extra-high targets`wow ´wowSE . 78 .78WH .94 .88INV .69 .53NOM .78 .94Table 3: proportion of items with at least one extra high targetWCCFL 38 14 rett@ucla.eduProsodically marked mirativity March 6, 20203. and inserted additional intermediate phrase boundaries– additional mid-sentence pitch-range resets– additional nuclear pitch accents (NPAs) relative to fillers– so words in exclamations were more prominent than fillers• example utterances(37) sentence exclamation (no sentence particle):(38) wh-exclamative (no sentence particle):(39) inversion exclamative (sentence particle):(40) nominal exclamative (sentence particle):• nuances across construction type– these three prosodic properties are necessary but not sufficientfor describing the intonational contour of the four constructionswe’re looking atWCCFL 38 15 rett@ucla.eduProsodically marked mirativity March 6, 2020∗ each construction manifests all three properties differentlyto form a unique tune∗ they differ in their macrorhythm (peak frequency)...∗ ...and in what types of words get marked as prominent– intonation is functioning to maximally differentiate each excla-mation type from its non-exclamation counterpart∗ sentence exclamation vs. simple declarative∗ wh-exclamative vs. wh-question∗ inversion exclamative vs. yes/no question∗ nominal exclamative vs. topicalized definite– the intonational patterns are the complete opposite of one an-other in terms of:∗ acoustic salience (for polar constructions: most/least salient);∗ what is prominent (for wh-constructions);∗ slow vs. fast macrorhythm (for polar constructions)(41) wh-question intonation∗ overall tune same as simple declaratives (H* L-L%)∗ wh-word is not prosodically prominent (Pierrehumbert, 1980),surprising given its discourse significance(42) wh-exclamative intonation∗ wh-word is highly prominent, marked with (L`)H* 87.5%∗ wh-word target is extra-high in 62.5% of wh-exclamativesWCCFL 38 16 rett@ucla.eduProsodically marked mirativity March 6, 2020(43) polar question intonation∗ canonical tune: L* H-H%∗ pitch accents are relatively sparse, resulting in slow macrorhythmfrequency (few peaks/valleys)(44) inversion exclamative intonation∗ most content words are marked with L`H*∗ results in a fast macrorhythm (many peaks/valleys)3.4 discussion• different roles for different prosodic properties– exclamation seems to be encoded in a specific pitch accent, L`H*– but it also seems to require super-tonal properties, namely extra-high targets and additional intermediate phrase boundaries∗ what’s the right treatment for this bundle of properties?∗ what’s the right semantic treatment for what might proveto be a gradient effect?• prosodic iconicity it seems possible that at least some of these char-acteristics of mirative prosody are non-arbitrary1. L`H* pitch accents – likely semantically arbitrary– there are other ways of marking mirativity cross- linguisti-cally that don’t involve L`H*WCCFL 38 17 rett@ucla.eduProsodically marked mirativity March 6, 2020– there are other uses of L`H*, even in English: prosodicfocus marking (Pierrehumbert, 1980; Selkirk, 1995)2. extra-high targets – likely semantically non-arbitrary– unclear whether they occur with other mirativity strategies,e.g. Finnish pa¨– but no other uses of extra-high targets in English3. additional intermediate phrases – ??– unclear whether they occur with other mirativity strategies– there are other uses of extra boundary insertion in focusmarking in English (Pierrehumbert, 1980) and other lan-guages (Royer and Jun, 2019)4 conclusions• mirativity is encoded:– morphologically – independently– syntactically – with other meaning (‘mixed expression’)– prosodically – polysemously (‘dependent’)• we need a semantic account that:– can treat a given semantic phenomenon across strategy types(morphologic, syntactic, prosodic)– in particular, that can treat prosody as one of the ‘parts’ in the‘parts and whole’ notion of compositionality– in terms of the formalism:∗ intuitively, mirativity is not part of descriptive content∗ diagnostics confirm differences between mirativity (and otheremotive markers) on the one hand and descriptive not-at-issue markers (e.g. evidentials) on the other– I model these differences in a dynamic context-based framework∗ with descriptive content updating the Common Ground,∗ and with illocutionary content – like mirativity – updatinga speaker’s Discourse CommitmentsWCCFL 38 18 rett@ucla.eduProsodically marked mirativity March 6, 2020• all exclamations, in English, are pronounced with:1. L`H* pitch accents;2. extra-high targets;3. additional intermediate phrase boundaries• there is a clear role for the L`H* pitch accent in exclamations– there’s other evidence that pitch accents operate at the illocu-tionary level (Jeong and Potts, 2016)– I analyze the pitch accent as the mirativity marker – effectivelyan illocutionary mood modifier – in English exclamations (Rett2019)• but there’s a question of what to do with the other prosodic compo-nents of mirativity marking in English– it’s intuitive to think that there’s something non-arbitrary aboutthe use of extra-high and additional targets to signify surpriseor unexpectedness– and while they aren’t necessary components of mirativity mark-ing, they seem to be sufficient• we’re currently running a naturalness rating task to see how impor-tant the super-tonal properties are for things like perceived speakersurprise, sincerity, etc.appendix A: flavored Discourse Commitments• to use DCs to model the content of emotive markers, I’ll need togeneralize them to propositional attitudes other than belief (inspiredby Portner 2006; see also Rudin 2018)(45) Flavored Discourse CommitmentsLet DCa be a set of pairs representing the public commit-ments of a with respect to a discourse in which a and b arethe participants, where:a. xbelieves, py is a public commitment of a iff ‘a believes p’is a mutual belief of a and b;WCCFL 38 19 rett@ucla.eduProsodically marked mirativity March 6, 2020b. xis-disappointed, py is a public commitment of a iff ‘a isdisappointed that p’ is a mutual belief of a and b; andc. xis-surprised, py is a public commitment of a iff ‘a is sur-prised that p’ is a mutual belief of a and b.• as a result, I’ll need to reformalize the sincerity condition on assertion(46) Declarative operator (D), for sentences Sp with at-issue con-tent p and not-at-issue content q:D(Sp, a,Ki) = Ko such that(i) DCa,o “ DCa,i Y xbelieves, py(ii) To “ pushpxSp; tpuy, Tiq(iii) pso “ psi Y tpu(iv) CGo = CGi Y tquappendix B: scope-taking• for an explicit discussion of how the analysis of mirative markersextends to questions, see Rett (2019)• encoders of illocutionary content contribute their own restriction tothe speaker’s DC set, which (in the case of embedded clauses) requiresa sub-sentential dynamic update(47) Jit’s raining alasK = A(S, a,Kiq “ pS1, a,Ko1q such that(i) DCa,o1 “ DCa,i Y txis-disappointed, It’s rainingyu(ii) To1 “ pushpxS1; {It’s raining}y, Tiq(48) JIt’s possible that it’s raining, alasK = D(S2, a,Ko1qq “ Ko2such that(i) DCa,o2 “ tDCa,i Y txis-disappointed, It’s rainingyuuY txbelieves, It’s possible that it’s rainingyu(ii) To2 “ pushpxS2; {It’s possible that it’s raining}y, ppushpxS1;{It’s raining}y, Tiqqq(iii) pso “ psi Y {It’s possible that it’s raining}(iv) CGo = CGi• while the proposition that it’s raining is pushed to the top of the stackin the update for the embedded clause, it is no longer at the top afterthe utterance of the matrix sentenceWCCFL 38 20 rett@ucla.eduProsodically marked mirativity March 6, 2020referencesAikhenvald, A. (2004). Evidentiality. Oxford University Press.Beckman, M. and Ayers Elam, G. (1997). Guidelines for ToBI labelling, version3.0. Ms., Ohio State University.Castroviejo-Miro´, E. (2006). Wh-Exclamatives in Catalan. PhD Thesis, Univer-sitat de Barcelona, Barcelona.Cruschina, S. (2012). Discourse-related features and functional projections. Ox-ford University Press.Cruschina, S. (2019). Focus fronting in Spanish: mirative implication and infor-mation structure. Probus, 31:119–146.DeLancey, S. (1997). Mirativity: the grammatical marking of unexpected infor-mation. Linguistic Typology, 1:33–52.DeLancey, S. (2001). The mirative and evidentiality. Journal of Pragmatics,33:369–382.Dickinson, C. (2000). Mirativity in Tsafiki. Studies in Language, 24:379–422.Farkas, D. and Bruce, K. (2010). On reacting to assertions and polar questions.Journal of Semantics, 27:81–118.Gunlogson, C. (2001). True to Form: Rising and Falling Declaratives as Questionsin English. Routledge.Gutzmann, D. and Turgay, K. (2011). Some (in)definiteness effects of expressiveintensifiers. Talk, Moscow Syntax & Semantics II.Hirschberg, J. and Ward, G. (1992). The influence of pitch range, duration, ampli-tude and spectral features on the interpretation of the rise-fall-rise intonationcontour in English. Journal of Phonetics, 20:241–251.Jeong, S. and Potts, C. (2016). Intonational sentence-type conventions for per-locutionary effects: an experimental investigation. In Proceedings of SALT 26,pages 1–22.Kaplan, D. (1997). The meaning of ‘ouch’ and ‘oops’. Ms., UCLA.Korotkova, N. (2012). The Georgian evidential perfect and how to analyze it.Ms., UCLA.Krifka, M. (2001). Quantifying into question acts. Natural Language Semantics,9:1–40.Malamud, S. and Stephenson, T. (2014). Three ways to avoid commitments:declarative force modifiers in the conversational scoreboard. Journal of Se-mantics, pages 1–37.Malchukov, A. (2004). Towards a semantic typology of adversative and contrastmarking. Journal of Semantics, 21:177–198.Merin, A. and Nikolaeva, I. (2008). Exclamative as a universal speech act cate-gory: a case study in decision-theoretic semantics and typological implications.Ms., University of Konstanz and SOAS London University.Murray, S. (2010). Evidentiality and the structure of speech acts. PhD Thesis,WCCFL 38 21 rett@ucla.eduProsodically marked mirativity March 6, 2020Rutgers University.Murray, S. (2011). A Hamblin semantics for evidentials. In Ito, S., Cormany, E.,and Lutz, D., editors, Proceedings of SALT 19 (2009). CLC.Murray, S. (2014). Varieties of update. Semantics & Pragmatics, 7:1–53.Nikolaeva, I. (1999). The semantics of Northern Khandy evidentials. JournalSociete´ Finno Ougrinne, 88:131–159.Peterson, T. (1999). Epistemic Modality and Evidentiality in Gitksan at theSemantics-Pragmatics Interface. PhD Thesis, The University of BritishColumbia.Peterson, T. (2010). Examining the mirative and non-literal uses of evidentials. InPeterson, T. and Sauerland, U., editors, Evidence from evidentiality. Universityof British Columbia Working Papers in Linguistics.Pierrehumbert, J. (1980). The phonology and phonetics of English intonation.PhD Thesis, MIT.Pierrehumbert, J. and Hirschberg, J. (1990). The meaning of intonational con-tours in the interpretation of discourse. In Cohen, P., Morgan, J., and Pollack,M., editors, Intentions in communication, pages 271–311. MIT Press.Portner, P. (2006). Comments on faller’s paper. Workshop on Philosophy andLinguistics at the University of Michigan.Rett, J. (2011). Exclamatives, degrees and speech acts. Linguistics and Philoso-phy, 34:411–442.Rett, J. (2012). Miratives across constructions and languages. Handout, CUSP5, UCSD.Rett, J. (2019). The semantics of emotive markers and other illocutionary content.Ms., UCLA.Rett, J. and Murray, S. (2013). A semantic account of mirative evidentials. InSALT XXIII Proceedings, pages 453–472. CLC Publications.Royer, A. and Jun, S.-A. (2019). Prominence marking in kazan tatarstan declar-atives. In Proceedings of the ICPhS 2019.Rudin, D. (2018). Rising imperatives. In Proceedings of SALT 28, pages 100–119.CLC Publications.Searle, J. and Vanderveken, D. (1985). Foundations of illocutionary logic. Cam-bridge University Press.Searle, J. R. (1969). Speech acts. Cambridge University Press.Selkirk, L. (1995). Sentence prosody: intonation, stress, and phrasing. In Gold-smith, J., editor, The Handbook of Phonological Theory, pages 550–569. Black-well.Stalnaker, R. (1973). Presuppositions. Journal of Philosophical Logic, 2:447–457.Wu, J.-S. (2008). Antonyms? Presuppositions? On the semantics of two evaluativemodals jingran and guoran in Mandarin. Journal of Chinese Language andComputing, 18:161–173.WCCFL 38 22 rett@ucla.edu Prosodically marked mirativity˚Jessica RettWest Coast Conference in Formal Linguistics 38March 6, 20200 overview• mirativity is the implicit encoding of speaker surprise or exceededexpectation (DeLancey, 1997)• it’s encoded using a variety of strategies (Rett, 2012):– morphologically, in sentence particles (e.g. Finnish);– polysemously, in mirative evidentials (e.g. Cheyenne);– syntactically, in focus fronting (e.g. Spanish);– in prosody (e.g. English)• I’ll first present a semantic account of mirativity across these strate-gies qua ‘illocutionary content’ (Rett, 2019)• I’ll then present an investigation of the properties of prosodic mirativemarking in English• The Goals of the Talk:– a straightforward semantic analysis of mirativity;– a study of prosodically marked mirativity;– and a discussion of why the former might need to be supple-mented to fit the latter˚This talk is based in part on joint work led by Beth Sturman. It has also benefittedfrom comments from audiences at UCLA; the 6th Cornell Workshop in Linguistics andPhilosophy; the Berkeley Meaning Sciences Club; the Workshop on Encoding EmotiveAttitudes in Non-truth-conditional Meaning at DGfS 41; the Functional Categories andExpressive Meaning Workshop at the Universitat Auto`noma de Barcelona; and the Sec-ond UC Irvine Workshop in Logical Semantics. It was funded by a 2019 UCLA Councilon Research Faculty Research Grant.1Prosodically marked mirativity March 6, 20201 mirative marking1.1 defining mirativity• I use the term ‘mirative’ as a label for any natural-language expressionof exceeded expectation(1) a. Jane won the raceb. (Wow) Jane won the race!– the ‘expression’ bit (Kaplan, 1997; Castroviejo-Miro´, 2006)∗ like any expressive speech act, can be uttered insincerely(Searle, 1969)(2) (Wow) Airline seats are so tiny these days!∗ undeniable, non-negatable(3) A: (Wow) Those cupcakes are vegan!B: No, they’re vegetarian.B1:#No, you knew exactly how good they’d be.– the ‘exceeded expectation’ bit∗ Merin and Nikolaeva (2008); Rett (2011): speaker’s expec-tations violated or exceeded (cf. speaker surprise)“No matter how high my expectations might havebeen, what I have just heard exceeded them” (De-Lancey, 2001, 38).∗ a violated expectation can be flattering or insulting, depend-ing on the relevant expectations(4) You did better than the faculty expected you to.∗ generally speaker-oriented(5) a. #Mary said Jane won the race!b. *How did Jane won the race!?∗ tied to the here-and-now: Rett and Murray (2013); ‘novelinformation’ or ‘unprepared mind’ (DeLancey, 1997, 2001;Peterson, 2010); a “spontaneous reaction to a new, salient,often surprising event” (Aikhenvald, 2004, 197)WCCFL 38 2 rett@ucla.eduProsodically marked mirativity March 6, 20201.2 mirative strategiesthere are a variety of different strategies of mirativity marking (Rett, 2012):1. independent miratives• morphologic – bound: in Finnish, the sentence particle -pa¨ is afocus-sensitive mirative marker (Rett, 2019)(6) a. Ta¨a¨lla¨hereonbe-3rd.sgpaljona.lotkukk-ia.flower-prt.indf.pl‘There are lots of flowers here.’b. Ta¨a¨lla¨-pa¨here-paonbe-3rd.sgpaljona.lotkukk-ia.flower-prt.indf.pl‘(Wow) There are lots of flowers here!’• morphologic – free: in Mandarin there are mirative (and anti-mirative) adverbials (‘evaluative modals’; Wu, 2008).(7) ZhangsanZhangsanguoran/jingranmir/anti-mirlaicomele.part‘Zhangsan came (as expected/not expected by the speaker).’• syntactic: in Spanish, mirativity is marked via focus fronting(Cruschina, 2012, 2019):(8) ¡Imag´ınate!imagine.imp.2sg¡Conwithelthedirectordirectorquer´ıawant.impf.3sghablar!talk.inf‘Guess what! The director he wanted to talk to!’• prosodic: exclamation intonation, as in (1) (more soon)2. mixed-expression miratives• mirative conjunctions: second conjunct is surprising indepen-dently of the first (e.g. ‘lo and behold’; Russian, Malchukov 2004)(9) Onhezabolelfell.illdaconjiptclumer.died‘He fell ill and died (I didn’t expect it).’WCCFL 38 3 rett@ucla.eduProsodically marked mirativity March 6, 2020• expressive intensifiers like the German sau (‘female pig’); total(‘totally’); and voll (‘fully’; Gutzmann and Turgay 2011)(10) DiethePartypartywarwassauSAUcool.cool‘The party was very cool (I can’t believe how cool!).’3. dependent miratives• mirative evidentials receive their mirative interpretation in con-texts in which the content was recently learned through directevidence (Rett and Murray, 2013)(11) Motomotorcyclejo-nu-e.be-evid-declTsafiki (Dickinson, 2000)speaker hears motor: ‘It is apparently a motorcycle.’speaker thought he heard a car, but sees a motorcyclecoming: ‘It’s a motorcycle!’• more language-specific licensing requirements for mirativity:– Gitksan: licensed by the first person (Peterson, 1999)– Hare: licensed by the imperfect (DeLancey, 1997, 2001)(12) a. MaryMarye-we´’its-hidegha´layey˜idawork.perflo˜.lo˜‘Mary worked on hides (given what I’ve inferred/heard)’b. MaryMarye-we´’its-hidegha´layedawork.impflo˜.lo˜‘Mary is working on hides (I saw, to my surprise)’– Cheyenne: by the present tense (Murray, 2010, 2011)– Georgian: by individual-level predicates (Korotkova, 2012)2 a unified analysis• a discourse particle is a morphological strategy; discourse particlescan do a bunch of different stuff (Malamud and Stephenson, 2014)• mirativity is pretty semantically uniform across languages, but theway it’s encoded variesWCCFL 38 4 rett@ucla.eduProsodically marked mirativity March 6, 2020• intuitively, we should have a unified account of this ‘sememe’:– does mirativity behave the same way across languages/strategies?– if so, does it behave like any other sememe?– if so, how should we analyze it semantically?– (and how should we treat any differences?)• mirativity is not-at-issue (NAI) content– it’s undeniable in discourse– it cannot be targeted by truth-conditional operators(13) A: (Wow) Jane won the race!B: That’s not true, she came in second.B1:#That’s not true, you knew she would.(14) (Wow) Jane did not win the race!• the same goes for all of the mirative strategies listed previously– morphologically encoded independent markers like alas/pa¨;– dependent & mixed-expression markers like mirative evidentials2.1 illocutionary diagnostics• but mirative marking behaves differently from other types of NAI con-tent like appositives or evidentials, requiring distinct semantic treat-ment (Rett 2019)semantic contentat-issue not-at-issuedescriptive NAI illocutionary NAIWCCFL 38 5 rett@ucla.eduProsodically marked mirativity March 6, 2020• alternatively:semantic contentdescriptiveat-issue descriptive NAIillocutionaryillocutionary NAI• descriptive NAI markers: appositives, evidentials, utterance modi-fiers, CIs (Potts 2005)(15) a. Allegedly, Jane won the race.b. Frankly, Jane won the race.• illocutionary NAI markers: alas, (un)fortunately, mirative markerslike Finnish pa¨ (Rett 2019)(16) a. Alas, Jane won the race.b. (Wow) Jane won the race!• descriptive not-at-issue markers contribute to the descriptive contentof an utterance – what is said• “illocutionary content” is not-at-issue meaning pertaining to how thespeaker is using the utterance in the context• diagnostics of illocutionary content (Rett, 2019):1. susceptibility to Moore’s Paradox– standard Moore’s Paradox:(17) #It’s raining, but I don’t believe it’s raining.– Murray (2010): denial of mirativity results in Moore’s Para-dox; denial of evidentiality results in contradiction– to the extent that English speakers recognize this difference,I’ve found that the English data pattern the same way.WCCFL 38 6 rett@ucla.eduProsodically marked mirativity March 6, 2020(18) a. #KAllegedly, Jane lost the race, but no one al-leged she did.b. # Alas, Jane lost the race, but I’m not disap-pointed she did.c. # (Wow) Jane lost the race! But I’m not sur-prised she did.– MP is suspended in suppositional contexts (Yalcin 2007)(19) Suppose it’s raining, but I don’t believe it is raining.– so is the content encoded by emotive markers (20-b), butnot e.g. evidential adverbials (20-a)(20) a. #Suppose that, allegedly, Jane lost the race, but thatno one alleged that she did.b. Suppose that, alas, Jane lost the race, but that I’mnot disappointed she did.2. scope-taking– emotive markers associate with a single salient proposition...– ...this makes them incompatible with utterances associatedwith multiple propositions (Cheyenne, Rett and Murray 2013)(21) a. Mo´=e´-x-ho´1 ta˙heva´-hoo1oy/n=3-rem.pst-win-nar.3sgAe´nohe?Hawk‘Given the stories you heard, did Hawk win?’b. %Mo´=e´-ho´1 ta˙heva´-hoo1oy/n=3-win-nar.3sgAe´nohe?Hawk‘Given your surprise, did Hawk win?’– ...it also means they scope differently with sentential opera-tors than descriptive NAI markers(22) a. #Alas, if Jane loses, at least we’ll flip the Senate.b. If, alas, Jane loses, at least we’ll flip the Senate.(23) a. #Apparently, if Jane loses, I will run for office.b. #If, apparently, Jane loses, I will run for office.WCCFL 38 7 rett@ucla.eduProsodically marked mirativity March 6, 20202.2 a semantics for illocutionary content• analysis, informally:(24) (Wow) Jane, who hates politics, won the race!a. at-issue: Jane won the raceb. descriptive NAI: Jane hates politicsc. illocutionary NAI: speaker hadn’t expected Jane towin the race– at-issue content:∗ acts as a proposal to submit the proposition to the CommonGround (Stalnaker, 1973)∗ modeled in Farkas and Bruce (2010) as an update to theProjected Set– descriptive not-at-issue content:∗ directly updates the Common Ground (Murray, 2010)– illocutionary content:∗ updates speaker’s Discourse Commitments (Gunlogson, 2001)• a note on sincerity conditions– extant dynamic accounts of declarative mood do not formallymodel the sincerity condition, i.e. ‘Speaker believes that p’– arguably, emotive markers like mirativity do the same sort ofthing, contribute something like ‘Speaker is surprised that p’∗ Searle (1969): Moore’s Paradox arises when a sincerity con-dition is denied∗ emotive markers also susceptible to Moore’s Paradox (Searleand Vanderveken 1985: alas qua speech-act modifier)– Discourse Commitments (Gunlogson, 2001):∗ originally proposed for speaker bias in rising declaratives∗ originally characterized in terms of the speaker’s beliefs∗ public: speaker “mutually recognized as committed to them”– I treat public commitment as a proxy for belief, for the purposeof modeling (sincere) conversationWCCFL 38 8 rett@ucla.eduProsodically marked mirativity March 6, 2020• Farkas and Bruce’s model of a discourse structure K includes:1. the common ground (CG), the set of propositions believed byall discourse participants (for the purpose of the conversation);2. sets of discourse commitments (DC): for each participant x,the set of propositions x publicly commits to in the conversation;3. the Table T , modeling discourse salience;4. the projection set (ps), the set of beliefs that are being con-sidered for addition into the CG.• they adopt their treatment of illocutionary mood from Krifka (2001),but they do not differentiate between at-issue and not-at-issue con-tent; I borrow from Murray (2010, 2014) to do this (in (iv)).(25) Declarative operator (D), for sentences Sp with at-issue con-tent p and not-at-issue content q:D(Sp, a,Ki) = Ko such that(i) DCa,o “ DCa,i Y txbelieves, pyu(ii) To “ pushpxSp; tpuy, Tiq(iii) pso “ psi Y tpu(iv) CGo = CGi Y tqu• mirative markers and other emotive markers operate on propositionalcontent by placing that content on the top of the stack and addingthe speaker’s emotive attitude towards that content to her DC set(26) Mir, for clauses C with content p: MirpC, a,Kiq “ pC, a,Koqsuch that(i) DCa,o “ DCa,i Y txis-surprised, py}(ii) To “ pushpxC; tpuy, Tiq(27) JJane won the race!K = D(MirpS, a,Kiqq “ Ko such that(i) DCa,o “ DCa,i Y txbelieves, Jane won the racey}(ii) To “ pushpxS; Jane won the racey, Tiq(iii) pso “ psi Y {Jane won the race}(iv) DCa,o “ DCa,i Y {xis-surprised, Jane won the racey}• emotive markers interact scopally with descriptive NAI content andat-issue content via sub-sentential dynamic update (Appendix B)WCCFL 38 9 rett@ucla.eduProsodically marked mirativity March 6, 2020• Take Home Message: Mirative markers, like all emotive markers, en-code NAI content. But it’s qualitatively different than canonical,‘descriptive’ NAI content: it operates at an illocutionary level3 prosodically marked mirativity• unified semantic treatment of mirativity (and emotive markers):– some languages do things with morphemes and syntax that otherlanguages do with prosody– a good reminder that prosody needs to be represented in ourcompositional semantics (Pierrehumbert and Hirschberg, 1990)• previous work on the semantics of prosody:– prosody can mark illocutionary mood (Pierrehumbert, 1980; Jeongand Potts, 2016)– prosody can mark orientation (speaker or hearer; Gunlogson2001; Rudin 2018– prosody can mark other content like uncertainty, incredulity(Hirschberg and Ward, 1992)• my claim: English exclamation intonation is multi-faceted, markingsomething like illocutionary content as well as secondary effects likesurprise, emphasis• English mirativity marking– sentence exclamation(28) a. Jane carves gorgeous sculptures.b. (Wow) Jane carves gorgeous sculptures!– exclamatives(29) a. (Wow) What gorgeous sculptures Jane carves!b. (Wow) Can Jane carve gorgeous sculptures!c. (Wow) The gorgeous sculptures Jane carves!WCCFL 38 10 rett@ucla.eduProsodically marked mirativity March 6, 2020– discourse particles optional– distinct from rhetorical questions, e.g. How cool is that?– exclamatives differ from sentence exclamations, in English atleast, in that the object of surprise needs to be a degree insteadof a proposition (Rett, 2011)3.1 prosody: an overview• in English, utterances associated with intonational contours, i.e. tunes– tunes are composed of one or more intermediate phrases– intermediate phrases each have a pitch accent– pitch is measured wrt speaker’s fundamental frequency (f0)– pitch accents composed of one or more pitch targets, with thestar anchored to the stressed syllable– intermediate phrases are distinguishable in having:1. their own pitch accents (with a stressed syllable);2. their own phrase accent (with final lengthening);3. pitch range reset (a new ceiling after a down step)• Annotated with MAE ToBI (Beckman and Ayers Elam, 1997)– relatively coarse: can’t mark some things, e.g. speaker’s relativeheight (high vs. extra-high)• intermediate phrasesDoes the governor of Iowa endorse a radio program?L* H-H%– QUD: Which governors endorse a radio program?WCCFL 38 11 rett@ucla.eduProsodically marked mirativity March 6, 2020Does the governor of Iowa endorse a radio program?L* H- L* H-H%– QUD: out of the blue (Hirschberg and Ward, 1992)• the semantics of pitch accents– different canonical tunes (pitch accents`boundary tones) areassociated with different illocutionary moods (Pierrehumbert,1980; Pierrehumbert and Hirschberg, 1990)∗ declarative sentences: H* L-L%(30) Miriam made the lemonade.∗ polar questions & rising declaratives: L* H-H%(31) Miriam made the lemonade?∗ confirmation questions/rising imperatives: H* H-H%(32) Make the lemonade?– ...as well as perlocutionary effects (Jeong and Potts, 2016)∗ falling intonation: authoritativeness/assertion∗ level intonation: annoyance∗ rising intonation: politeness/positivity (cf. hedging)3.2 elicitation• research questions1. are English exclamations marked with uniform prosody?2. if so, how are they marked?3. how can we represent it semantically?4. (is the prosodic marking of mirativity semantically arbitrary?)WCCFL 38 12 rett@ucla.eduProsodically marked mirativity March 6, 2020• elicitation methodology– two consultants: one female, one male– conditions: 4 (construction) x 2 (˘discourse particle)– stimuli: 32 items per condition (256 total), between subjects(33) You dont expect Julian to make beautiful paintings, butyou find out he did. You tell Sara:a. (Wow) Julian makes beautiful paintings!b. (Wow)What beautiful paintings Julian makes!c. (Wow) Does Julian make beautiful paintings!d. (Wow) The beautiful paintings Julian makes!(34) You don’t expect Ariel’s stories to cause confusion, butyou find out they have. You tell Sara:a. (Wow) Ariel’s stories caused confusion!b. (Wow) What confusion Ariel’s stories caused!c. (Wow) Did Ariel’s stories cause confusion!d. (Wow) The confusion Ariel’s stories caused!– fillers: 4 construction conditions: declaratives; wh-questions; po-lar questions; definite subjects/clefts– 32 in each condition (128 total), btwn subjects (64 ea.)(35) a. Anna is good at chess.b. What is Anna good at?c. Is Anna good at chess?d. The game Anna is good at is chess.(36) a. James is an expert at knitting.b. What is James an expert at?c. Is James an expert at knitting?d. The thing James is an expert at is knitting.– consultants were instructed to read the sentences as they’d benaturally uttered in the provided context– the recordings were transcribed in Praat using MAE ToBI by oneof the authors, and confirmed by another trained transcriberWCCFL 38 13 rett@ucla.eduProsodically marked mirativity March 6, 20203.3 elicitation results• both consultants consistently:1. pronounced exclamations with L`H* pitch accents;L`H* H* H`!H* L* L*`HSE .96 .04 – – –WH .85 .10 .05 – –INV .94 .04 – – .02NOM .93 .03 – .04 –average .92 .05 .01 .01 ă.01Table 1: proportion of items with L`H* pitch accents, femaleL`H* H* H`!H* L* L*`HSE 1 – – – –WH .96 .04 – – –INV 1 – – – –NOM .94 .06 – – –average .98 .03 – – –Table 2: proportion of items with L`H* pitch accents, male2. pronounced exclamations with extra-high targets:– the high target exceeds expected height for position (after adownstep, etc.);– target exceeds default pitch maximum by at least 5%– cf. fillers, with zero extra-high targets`wow ´wowSE . 78 .78WH .94 .88INV .69 .53NOM .78 .94Table 3: proportion of items with at least one extra high targetWCCFL 38 14 rett@ucla.eduProsodically marked mirativity March 6, 20203. and inserted additional intermediate phrase boundaries– additional mid-sentence pitch-range resets– additional nuclear pitch accents (NPAs) relative to fillers– so words in exclamations were more prominent than fillers• example utterances(37) sentence exclamation (no sentence particle):(38) wh-exclamative (no sentence particle):(39) inversion exclamative (sentence particle):(40) nominal exclamative (sentence particle):• nuances across construction type– these three prosodic properties are necessary but not sufficientfor describing the intonational contour of the four constructionswe’re looking atWCCFL 38 15 rett@ucla.eduProsodically marked mirativity March 6, 2020∗ each construction manifests all three properties differentlyto form a unique tune∗ they differ in their macrorhythm (peak frequency)...∗ ...and in what types of words get marked as prominent– intonation is functioning to maximally differentiate each excla-mation type from its non-exclamation counterpart∗ sentence exclamation vs. simple declarative∗ wh-exclamative vs. wh-question∗ inversion exclamative vs. yes/no question∗ nominal exclamative vs. topicalized definite– the intonational patterns are the complete opposite of one an-other in terms of:∗ acoustic salience (for polar constructions: most/least salient);∗ what is prominent (for wh-constructions);∗ slow vs. fast macrorhythm (for polar constructions)(41) wh-question intonation∗ overall tune same as simple declaratives (H* L-L%)∗ wh-word is not prosodically prominent (Pierrehumbert, 1980),surprising given its discourse significance(42) wh-exclamative intonation∗ wh-word is highly prominent, marked with (L`)H* 87.5%∗ wh-word target is extra-high in 62.5% of wh-exclamativesWCCFL 38 16 rett@ucla.eduProsodically marked mirativity March 6, 2020(43) polar question intonation∗ canonical tune: L* H-H%∗ pitch accents are relatively sparse, resulting in slow macrorhythmfrequency (few peaks/valleys)(44) inversion exclamative intonation∗ most content words are marked with L`H*∗ results in a fast macrorhythm (many peaks/valleys)3.4 discussion• different roles for different prosodic properties– exclamation seems to be encoded in a specific pitch accent, L`H*– but it also seems to require super-tonal properties, namely extra-high targets and additional intermediate phrase boundaries∗ what’s the right treatment for this bundle of properties?∗ what’s the right semantic treatment for what might proveto be a gradient effect?• prosodic iconicity it seems possible that at least some of these char-acteristics of mirative prosody are non-arbitrary1. L`H* pitch accents – likely semantically arbitrary– there are other ways of marking mirativity cross- linguisti-cally that don’t involve L`H*WCCFL 38 17 rett@ucla.eduProsodically marked mirativity March 6, 2020– there are other uses of L`H*, even in English: prosodicfocus marking (Pierrehumbert, 1980; Selkirk, 1995)2. extra-high targets – likely semantically non-arbitrary– unclear whether they occur with other mirativity strategies,e.g. Finnish pa¨– but no other uses of extra-high targets in English3. additional intermediate phrases – ??– unclear whether they occur with other mirativity strategies– there are other uses of extra boundary insertion in focusmarking in English (Pierrehumbert, 1980) and other lan-guages (Royer and Jun, 2019)4 conclusions• mirativity is encoded:– morphologically – independently– syntactically – with other meaning (‘mixed expression’)– prosodically – polysemously (‘dependent’)• we need a semantic account that:– can treat a given semantic phenomenon across strategy types(morphologic, syntactic, prosodic)– in particular, that can treat prosody as one of the ‘parts’ in the‘parts and whole’ notion of compositionality– in terms of the formalism:∗ intuitively, mirativity is not part of descriptive content∗ diagnostics confirm differences between mirativity (and otheremotive markers) on the one hand and descriptive not-at-issue markers (e.g. evidentials) on the other– I model these differences in a dynamic context-based framework∗ with descriptive content updating the Common Ground,∗ and with illocutionary content – like mirativity – updatinga speaker’s Discourse CommitmentsWCCFL 38 18 rett@ucla.eduProsodically marked mirativity March 6, 2020• all exclamations, in English, are pronounced with:1. L`H* pitch accents;2. extra-high targets;3. additional intermediate phrase boundaries• there is a clear role for the L`H* pitch accent in exclamations– there’s other evidence that pitch accents operate at the illocu-tionary level (Jeong and Potts, 2016)– I analyze the pitch accent as the mirativity marker – effectivelyan illocutionary mood modifier – in English exclamations (Rett2019)• but there’s a question of what to do with the other prosodic compo-nents of mirativity marking in English– it’s intuitive to think that there’s something non-arbitrary aboutthe use of extra-high and additional targets to signify surpriseor unexpectedness– and while they aren’t necessary components of mirativity mark-ing, they seem to be sufficient• we’re currently running a naturalness rating task to see how impor-tant the super-tonal properties are for things like perceived speakersurprise, sincerity, etc.appendix A: flavored Discourse Commitments• to use DCs to model the content of emotive markers, I’ll need togeneralize them to propositional attitudes other than belief (inspiredby Portner 2006; see also Rudin 2018)(45) Flavored Discourse CommitmentsLet DCa be a set of pairs representing the public commit-ments of a with respect to a discourse in which a and b arethe participants, where:a. xbelieves, py is a public commitment of a iff ‘a believes p’is a mutual belief of a and b;WCCFL 38 19 rett@ucla.eduProsodically marked mirativity March 6, 2020b. xis-disappointed, py is a public commitment of a iff ‘a isdisappointed that p’ is a mutual belief of a and b; andc. xis-surprised, py is a public commitment of a iff ‘a is sur-prised that p’ is a mutual belief of a and b.• as a result, I’ll need to reformalize the sincerity condition on assertion(46) Declarative operator (D), for sentences Sp with at-issue con-tent p and not-at-issue content q:D(Sp, a,Ki) = Ko such that(i) DCa,o “ DCa,i Y xbelieves, py(ii) To “ pushpxSp; tpuy, Tiq(iii) pso “ psi Y tpu(iv) CGo = CGi Y tquappendix B: scope-taking• for an explicit discussion of how the analysis of mirative markersextends to questions, see Rett (2019)• encoders of illocutionary content contribute their own restriction tothe speaker’s DC set, which (in the case of embedded clauses) requiresa sub-sentential dynamic update(47) Jit’s raining alasK = A(S, a,Kiq “ pS1, a,Ko1q such that(i) DCa,o1 “ DCa,i Y txis-disappointed, It’s rainingyu(ii) To1 “ pushpxS1; {It’s raining}y, Tiq(48) JIt’s possible that it’s raining, alasK = D(S2, a,Ko1qq “ Ko2such that(i) DCa,o2 “ tDCa,i Y txis-disappointed, It’s rainingyuuY txbelieves, It’s possible that it’s rainingyu(ii) To2 “ pushpxS2; {It’s possible that it’s raining}y, ppushpxS1;{It’s raining}y, Tiqqq(iii) pso “ psi Y {It’s possible that it’s raining}(iv) CGo = CGi• while the proposition that it’s raining is pushed to the top of the stackin the update for the embedded clause, it is no longer at the top afterthe utterance of the matrix sentenceWCCFL 38 20 rett@ucla.eduProsodically marked mirativity March 6, 2020referencesAikhenvald, A. (2004). Evidentiality. Oxford University Press.Beckman, M. and Ayers Elam, G. (1997). Guidelines for ToBI labelling, version3.0. Ms., Ohio State University.Castroviejo-Miro´, E. (2006). Wh-Exclamatives in Catalan. PhD Thesis, Univer-sitat de Barcelona, Barcelona.Cruschina, S. (2012). Discourse-related features and functional projections. Ox-ford University Press.Cruschina, S. (2019). Focus fronting in Spanish: mirative implication and infor-mation structure. Probus, 31:119–146.DeLancey, S. (1997). Mirativity: the grammatical marking of unexpected infor-mation. Linguistic Typology, 1:33–52.DeLancey, S. (2001). The mirative and evidentiality. Journal of Pragmatics,33:369–382.Dickinson, C. (2000). Mirativity in Tsafiki. Studies in Language, 24:379–422.Farkas, D. and Bruce, K. (2010). On reacting to assertions and polar questions.Journal of Semantics, 27:81–118.Gunlogson, C. (2001). True to Form: Rising and Falling Declaratives as Questionsin English. Routledge.Gutzmann, D. and Turgay, K. (2011). Some (in)definiteness effects of expressiveintensifiers. Talk, Moscow Syntax & Semantics II.Hirschberg, J. and Ward, G. (1992). The influence of pitch range, duration, ampli-tude and spectral features on the interpretation of the rise-fall-rise intonationcontour in English. Journal of Phonetics, 20:241–251.Jeong, S. and Potts, C. (2016). Intonational sentence-type conventions for per-locutionary effects: an experimental investigation. In Proceedings of SALT 26,pages 1–22.Kaplan, D. (1997). The meaning of ‘ouch’ and ‘oops’. Ms., UCLA.Korotkova, N. (2012). The Georgian evidential perfect and how to analyze it.Ms., UCLA.Krifka, M. (2001). Quantifying into question acts. Natural Language Semantics,9:1–40.Malamud, S. and Stephenson, T. (2014). Three ways to avoid commitments:declarative force modifiers in the conversational scoreboard. Journal of Se-mantics, pages 1–37.Malchukov, A. (2004). Towards a semantic typology of adversative and contrastmarking. Journal of Semantics, 21:177–198.Merin, A. and Nikolaeva, I. (2008). Exclamative as a universal speech act cate-gory: a case study in decision-theoretic semantics and typological implications.Ms., University of Konstanz and SOAS London University.Murray, S. (2010). Evidentiality and the structure of speech acts. PhD Thesis,WCCFL 38 21 rett@ucla.eduProsodically marked mirativity March 6, 2020Rutgers University.Murray, S. (2011). A Hamblin semantics for evidentials. In Ito, S., Cormany, E.,and Lutz, D., editors, Proceedings of SALT 19 (2009). CLC.Murray, S. (2014). Varieties of update. Semantics & Pragmatics, 7:1–53.Nikolaeva, I. (1999). The semantics of Northern Khandy evidentials. JournalSociete´ Finno Ougrinne, 88:131–159.Peterson, T. (1999). Epistemic Modality and Evidentiality in Gitksan at theSemantics-Pragmatics Interface. PhD Thesis, The University of BritishColumbia.Peterson, T. (2010). Examining the mirative and non-literal uses of evidentials. InPeterson, T. and Sauerland, U., editors, Evidence from evidentiality. Universityof British Columbia Working Papers in Linguistics.Pierrehumbert, J. (1980). The phonology and phonetics of English intonation.PhD Thesis, MIT.Pierrehumbert, J. and Hirschberg, J. (1990). The meaning of intonational con-tours in the interpretation of discourse. In Cohen, P., Morgan, J., and Pollack,M., editors, Intentions in communication, pages 271–311. MIT Press.Portner, P. (2006). Comments on faller’s paper. Workshop on Philosophy andLinguistics at the University of Michigan.Rett, J. (2011). Exclamatives, degrees and speech acts. Linguistics and Philoso-phy, 34:411–442.Rett, J. (2012). Miratives across constructions and languages. Handout, CUSP5, UCSD.Rett, J. (2019). The semantics of emotive markers and other illocutionary content.Ms., UCLA.Rett, J. and Murray, S. (2013). A semantic account of mirative evidentials. InSALT XXIII Proceedings, pages 453–472. CLC Publications.Royer, A. and Jun, S.-A. (2019). Prominence marking in kazan tatarstan declar-atives. In Proceedings of the ICPhS 2019.Rudin, D. (2018). Rising imperatives. In Proceedings of SALT 28, pages 100–119.CLC Publications.Searle, J. and Vanderveken, D. (1985). Foundations of illocutionary logic. Cam-bridge University Press.Searle, J. R. (1969). Speech acts. Cambridge University Press.Selkirk, L. (1995). Sentence prosody: intonation, stress, and phrasing. In Gold-smith, J., editor, The Handbook of Phonological Theory, pages 550–569. Black-well.Stalnaker, R. (1973). Presuppositions. Journal of Philosophical Logic, 2:447–457.Wu, J.-S. (2008). Antonyms? Presuppositions? On the semantics of two evaluativemodals jingran and guoran in Mandarin. Journal of Chinese Language andComputing, 18:161–173.WCCFL 38 22 rett@ucla.edu Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsProsodically marked mirativityJessica Rettrett@ucla.eduWCCFL 38March 6, 2020Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsoverviewI mirativity is the encoding of implicit speaker surprise orexceeded expectation (DeLancey 1997)I it’s encoded using a variety of strategies (Rett 2012):I morphologically, in sentence particles (e.g. Finnish);I syntactically, in focus fronting (e.g. Spanish);I prosodically (e.g. English)I I’ll first present a semantic account of mirativity acrossthese strategies qua ‘illocutionary content’ (Rett 2019)I I’ll then present an investigation of the properties ofprosodic mirative marking in EnglishI The Goals of the Talk:I a straightforward semantic analysis of mirativity;I a study of prosodically marked mirativity;I and a discussion of why the former might need to besupplemented to fit the latterProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsoverviewI mirativity is the encoding of implicit speaker surprise orexceeded expectation (DeLancey 1997)I it’s encoded using a variety of strategies (Rett 2012):I morphologically, in sentence particles (e.g. Finnish);I syntactically, in focus fronting (e.g. Spanish);I prosodically (e.g. English)I I’ll first present a semantic account of mirativity acrossthese strategies qua ‘illocutionary content’ (Rett 2019)I I’ll then present an investigation of the properties ofprosodic mirative marking in EnglishI The Goals of the Talk:I a straightforward semantic analysis of mirativity;I a study of prosodically marked mirativity;I and a discussion of why the former might need to besupplemented to fit the latterProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsoverviewI mirativity is the encoding of implicit speaker surprise orexceeded expectation (DeLancey 1997)I it’s encoded using a variety of strategies (Rett 2012):I morphologically, in sentence particles (e.g. Finnish);I syntactically, in focus fronting (e.g. Spanish);I prosodically (e.g. English)I I’ll first present a semantic account of mirativity acrossthese strategies qua ‘illocutionary content’ (Rett 2019)I I’ll then present an investigation of the properties ofprosodic mirative marking in EnglishI The Goals of the Talk:I a straightforward semantic analysis of mirativity;I a study of prosodically marked mirativity;I and a discussion of why the former might need to besupplemented to fit the latterProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsoverviewI mirativity is the encoding of implicit speaker surprise orexceeded expectation (DeLancey 1997)I it’s encoded using a variety of strategies (Rett 2012):I morphologically, in sentence particles (e.g. Finnish);I syntactically, in focus fronting (e.g. Spanish);I prosodically (e.g. English)I I’ll first present a semantic account of mirativity acrossthese strategies qua ‘illocutionary content’ (Rett 2019)I I’ll then present an investigation of the properties ofprosodic mirative marking in EnglishI The Goals of the Talk:I a straightforward semantic analysis of mirativity;I a study of prosodically marked mirativity;I and a discussion of why the former might need to besupplemented to fit the latterProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsoverviewI mirativity is the encoding of implicit speaker surprise orexceeded expectation (DeLancey 1997)I it’s encoded using a variety of strategies (Rett 2012):I morphologically, in sentence particles (e.g. Finnish);I syntactically, in focus fronting (e.g. Spanish);I prosodically (e.g. English)I I’ll first present a semantic account of mirativity acrossthese strategies qua ‘illocutionary content’ (Rett 2019)I I’ll then present an investigation of the properties ofprosodic mirative marking in EnglishI The Goals of the Talk:I a straightforward semantic analysis of mirativity;I a study of prosodically marked mirativity;I and a discussion of why the former might need to besupplemented to fit the latterProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsacknowledgmentsThis talk is based in part on joint research led by Beth Sturman,and was funded by a 2019 UCLA COR Faculty Research Grant.It has also benefitted from comments from audiences at UCLA; the 6th Cornell Workshop in Linguisticsand Philosophy; the Berkeley Meaning Sciences Club; the Workshop on Encoding Emotive Attitudes inNon-truth-conditional Meaning at DGfS 41; the Functional Categories and Expressive Meaning Workshopat the Universitat Auto`noma de Barcelona; and the Second UC Irvine Workshop in Logical Semantics.Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionswhat is mirativity?I use the term ‘mirative’ as a label for any natural-languageexpression of exceeded expectation(1) a. Jane won the race.b. (Wow) Jane won the race!I the ‘expression’ bit (Kaplan 1997; Castroviejo 2006)I like any expressive speech act, can be uttered insincerely(Searle 1969)(2) (Wow) Airline seats are so tiny these days!I undeniable, non-negatable(3) A: (Wow) Those cupcakes are good!B: #No, you knew exactly how good they’d be.Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionswhat is mirativity?I use the term ‘mirative’ as a label for any natural-languageexpression of exceeded expectation(1) a. Jane won the race.b. (Wow) Jane won the race!I the ‘expression’ bit (Kaplan 1997; Castroviejo 2006)I like any expressive speech act, can be uttered insincerely(Searle 1969)(2) (Wow) Airline seats are so tiny these days!I undeniable, non-negatable(3) A: (Wow) Those cupcakes are good!B: #No, you knew exactly how good they’d be.Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionswhat is mirativity?I use the term ‘mirative’ as a label for any natural-languageexpression of exceeded expectation(1) a. Jane won the race.b. (Wow) Jane won the race!I the ‘expression’ bit (Kaplan 1997; Castroviejo 2006)I like any expressive speech act, can be uttered insincerely(Searle 1969)(2) (Wow) Airline seats are so tiny these days!I undeniable, non-negatable(3) A: (Wow) Those cupcakes are good!B: #No, you knew exactly how good they’d be.Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionswhat is mirativity?mirativity = expression of exceeded expectationI the ‘exceeded expectation’ bitI Merin & Nikolaeva (2008); Rett (2011): speaker’sexpectations violated or exceededI “No matter how high my expectations might have been,what I have just heard exceeded them” (DeLancey ’01)I a violated expectation can be flattering or insulting,depending on the relevant expectations(4) You did better than the faculty expected you to.I generally speaker-oriented(5) a. #Mary said Jane won the race!b. *How did Jane win the race!?I tied to the here and now (Rett & Murray 2013);‘novelinformation’ or ‘unprepared mind’ (DeLancey 1997,2001; Peterson 2010);‘spontaneous reaction to a new,salient, often surprising event’ (Aikhenvald 2004)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionswhat is mirativity?mirativity = expression of exceeded expectationI the ‘exceeded expectation’ bitI Merin & Nikolaeva (2008); Rett (2011): speaker’sexpectations violated or exceededI “No matter how high my expectations might have been,what I have just heard exceeded them” (DeLancey ’01)I a violated expectation can be flattering or insulting,depending on the relevant expectations(4) You did better than the faculty expected you to.I generally speaker-oriented(5) a. #Mary said Jane won the race!b. *How did Jane win the race!?I tied to the here and now (Rett & Murray 2013);‘novelinformation’ or ‘unprepared mind’ (DeLancey 1997,2001; Peterson 2010);‘spontaneous reaction to a new,salient, often surprising event’ (Aikhenvald 2004)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionswhat is mirativity?mirativity = expression of exceeded expectationI the ‘exceeded expectation’ bitI Merin & Nikolaeva (2008); Rett (2011): speaker’sexpectations violated or exceededI “No matter how high my expectations might have been,what I have just heard exceeded them” (DeLancey ’01)I a violated expectation can be flattering or insulting,depending on the relevant expectations(4) You did better than the faculty expected you to.I generally speaker-oriented(5) a. #Mary said Jane won the race!b. *How did Jane win the race!?I tied to the here and now (Rett & Murray 2013);‘novelinformation’ or ‘unprepared mind’ (DeLancey 1997,2001; Peterson 2010);‘spontaneous reaction to a new,salient, often surprising event’ (Aikhenvald 2004)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsa taxonomy of mirativityI there are a variety of different strategies of mirativitymarking (Rett 2012):1. independent mirativesI morphologicI syntacticI prosodic2. mixed-expression miratives3. dependent mirativesI instead of providing semantics for a given strategy (e.g.discourse particles), we should aim to provide semanticsfor a given unit of meaning (i.e. ‘sememes’)I a clear need for a compositional semantics: mirativityinteracts not just with the proposition it ranges over,but with other properties of its marker (mixed-expression) or the context (dependent miratives).Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsa taxonomy of mirativityI there are a variety of different strategies of mirativitymarking (Rett 2012):1. independent mirativesI morphologicI syntacticI prosodic2. mixed-expression miratives3. dependent mirativesI instead of providing semantics for a given strategy (e.g.discourse particles), we should aim to provide semanticsfor a given unit of meaning (i.e. ‘sememes’)I a clear need for a compositional semantics: mirativityinteracts not just with the proposition it ranges over,but with other properties of its marker (mixed-expression) or the context (dependent miratives).Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsa taxonomy of mirativityI there are a variety of different strategies of mirativitymarking (Rett 2012):1. independent mirativesI morphologicI syntacticI prosodic2. mixed-expression miratives3. dependent mirativesI instead of providing semantics for a given strategy (e.g.discourse particles), we should aim to provide semanticsfor a given unit of meaning (i.e. ‘sememes’)I a clear need for a compositional semantics: mirativityinteracts not just with the proposition it ranges over,but with other properties of its marker (mixed-expression) or the context (dependent miratives).Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsa taxonomy of mirativityI there are a variety of different strategies of mirativitymarking (Rett 2012):1. independent mirativesI morphologicI syntacticI prosodic2. mixed-expression miratives3. dependent mirativesI instead of providing semantics for a given strategy (e.g.discourse particles), we should aim to provide semanticsfor a given unit of meaning (i.e. ‘sememes’)I a clear need for a compositional semantics: mirativityinteracts not just with the proposition it ranges over,but with other properties of its marker (mixed-expression) or the context (dependent miratives).Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsindependent mirativesI in Finnish, the focus-sensitive sentence particle -pa¨(6) a. Ta¨a¨lla¨hereonbe-3rd.sgpaljona.lotkukk-ia.flower-prt.indf.pl‘There are lots of flowers here.’b. Ta¨a¨lla¨-pa¨here-paonbe-3rd.sgpaljona.lotkukk-ia.flower-prt.indf.pl‘(Wow) There are lots of flowers here!’I Mandarin: (anti-)mirative adverbials (Wu 2008)(7) ZhangsanZhangsanguoran/jingranmir/anti-mirlaicomele.part‘Zhangsan came (as expected/not expected).’I Spanish: via focus fronting (Cruschina 2012, 2019)(8) ¡Imag´ınate!imagine.imp.2sg¡Conwithelthedirectordirectorquer´ıawant.impf.3sghablar!talk.inf‘Guess what! The director he wanted to talk to!’I English exclamation intonation, as in (1) (more soon)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsindependent mirativesI in Finnish, the focus-sensitive sentence particle -pa¨(6) a. Ta¨a¨lla¨hereonbe-3rd.sgpaljona.lotkukk-ia.flower-prt.indf.pl‘There are lots of flowers here.’b. Ta¨a¨lla¨-pa¨here-paonbe-3rd.sgpaljona.lotkukk-ia.flower-prt.indf.pl‘(Wow) There are lots of flowers here!’I Mandarin: (anti-)mirative adverbials (Wu 2008)(7) ZhangsanZhangsanguoran/jingranmir/anti-mirlaicomele.part‘Zhangsan came (as expected/not expected).’I Spanish: via focus fronting (Cruschina 2012, 2019)(8) ¡Imag´ınate!imagine.imp.2sg¡Conwithelthedirectordirectorquer´ıawant.impf.3sghablar!talk.inf‘Guess what! The director he wanted to talk to!’I English exclamation intonation, as in (1) (more soon)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsindependent mirativesI in Finnish, the focus-sensitive sentence particle -pa¨(6) a. Ta¨a¨lla¨hereonbe-3rd.sgpaljona.lotkukk-ia.flower-prt.indf.pl‘There are lots of flowers here.’b. Ta¨a¨lla¨-pa¨here-paonbe-3rd.sgpaljona.lotkukk-ia.flower-prt.indf.pl‘(Wow) There are lots of flowers here!’I Mandarin: (anti-)mirative adverbials (Wu 2008)(7) ZhangsanZhangsanguoran/jingranmir/anti-mirlaicomele.part‘Zhangsan came (as expected/not expected).’I Spanish: via focus fronting (Cruschina 2012, 2019)(8) ¡Imag´ınate!imagine.imp.2sg¡Conwithelthedirectordirectorquer´ıawant.impf.3sghablar!talk.inf‘Guess what! The director he wanted to talk to!’I English exclamation intonation, as in (1) (more soon)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsindependent mirativesI in Finnish, the focus-sensitive sentence particle -pa¨(6) a. Ta¨a¨lla¨hereonbe-3rd.sgpaljona.lotkukk-ia.flower-prt.indf.pl‘There are lots of flowers here.’b. Ta¨a¨lla¨-pa¨here-paonbe-3rd.sgpaljona.lotkukk-ia.flower-prt.indf.pl‘(Wow) There are lots of flowers here!’I Mandarin: (anti-)mirative adverbials (Wu 2008)(7) ZhangsanZhangsanguoran/jingranmir/anti-mirlaicomele.part‘Zhangsan came (as expected/not expected).’I Spanish: via focus fronting (Cruschina 2012, 2019)(8) ¡Imag´ınate!imagine.imp.2sg¡Conwithelthedirectordirectorquer´ıawant.impf.3sghablar!talk.inf‘Guess what! The director he wanted to talk to!’I English exclamation intonation, as in (1) (more soon)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsmixed-expression mirativesI mirative conjunctions: second conjunct is surprisingindependent of the first (‘lo and behold’), e.g. Russian(9) Onhezabolelfell.illdaconjiptclumer.died‘He fell ill and died (I did not expect it).’I expressive intensifiers, e.g. German sau (‘female pig’);total (‘totally’); voll (‘fully’; Gutzmann 2015)(10) DiethePartypartywarwassauSAUcool.cool‘The party was very cool (I can’t believe howcool!).’Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsmixed-expression mirativesI mirative conjunctions: second conjunct is surprisingindependent of the first (‘lo and behold’), e.g. Russian(9) Onhezabolelfell.illdaconjiptclumer.died‘He fell ill and died (I did not expect it).’I expressive intensifiers, e.g. German sau (‘female pig’);total (‘totally’); voll (‘fully’; Gutzmann 2015)(10) DiethePartypartywarwassauSAUcool.cool‘The party was very cool (I can’t believe howcool!).’Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsdependent mirativesI mirative evidentials are mirative only with directevidence (e.g. Tsafiki, Dickinson 2000)(11) Motomotorcyclejo-nu-e.be-evid-declspeaker hears motor: ‘It is apparently a motorcycle.’speaker thought he heard a car, but sees amotorcycle coming: ‘It’s a motorcycle!’I additional language-specific licensing requirements:I Gitksan: licensed by first person (Peterson 1999)I Hare: licensed by the imperfect (DeLancey 1997, 2001)(12) a. MaryMarye-we´’its-hidegha´layey˜idawork.perflo˜.lo˜‘Mary worked on hides (I inferred).’b. MaryMarye-we´’its-hidegha´layedawork.impflo˜.lo˜‘Mary is working on hides (I saw, to my surprise).’I Cheyenne: licensed by present tense (Murray 2010)I Georgian: by individual-level preds (Korotkova 2012)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsdependent mirativesI mirative evidentials are mirative only with directevidence (e.g. Tsafiki, Dickinson 2000)(11) Motomotorcyclejo-nu-e.be-evid-declspeaker hears motor: ‘It is apparently a motorcycle.’speaker thought he heard a car, but sees amotorcycle coming: ‘It’s a motorcycle!’I additional language-specific licensing requirements:I Gitksan: licensed by first person (Peterson 1999)I Hare: licensed by the imperfect (DeLancey 1997, 2001)(12) a. MaryMarye-we´’its-hidegha´layey˜idawork.perflo˜.lo˜‘Mary worked on hides (I inferred).’b. MaryMarye-we´’its-hidegha´layedawork.impflo˜.lo˜‘Mary is working on hides (I saw, to my surprise).’I Cheyenne: licensed by present tense (Murray 2010)I Georgian: by individual-level preds (Korotkova 2012)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsa unified semantic analysisI a discourse particle is a morphological strategy;discourse particles can do a bunch of different stuff(Malamud & Stephenson 2015)I mirativity is pretty semantically uniform acrosslanguages, but the way it’s encoded variesI intuitively, we should have a unified account of this‘sememe’:I does mirativity behave the same way across languagesand strategies?I if so, does it behave like any other sememe?I if so, how should we analyze it semantically?I (and how should we treat any differences?)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsmirativity is not-at-issueI mirativity is not-at-issue (NAI) contentI it’s undeniable in discourseI it cannot be targeted by truth-conditional operators(13) A: (Wow) Jane won the race!B: That’s not true, she came in second.B′:#That’s not true, you knew she would.(14) (Wow) Jane did not win the race!I the same goes for all of the mirative strategies listedpreviouslyI morphologically encoded independent markers like alasand Finnish pa¨;I dependent and mixed-expression miratives like mirativeevidentialsProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsmirativity is not-at-issueI mirativity is not-at-issue (NAI) contentI it’s undeniable in discourseI it cannot be targeted by truth-conditional operators(13) A: (Wow) Jane won the race!B: That’s not true, she came in second.B′:#That’s not true, you knew she would.(14) (Wow) Jane did not win the race!I the same goes for all of the mirative strategies listedpreviouslyI morphologically encoded independent markers like alasand Finnish pa¨;I dependent and mixed-expression miratives like mirativeevidentialsProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsa partial typology of NAI contentI but mirative marking behaves differently from othertypes of NAI content like appositives or evidentials,requiring distinct semantic treatment (Rett 2019)semantic contentat-issue not-at-issuedescriptive NAI illocutionary NAII alternatively:semantic contentdescriptiveat-issue descriptive NAIillocutionaryillocutionary NAIProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsa partial typology of NAI contentI but mirative marking behaves differently from othertypes of NAI content like appositives or evidentials,requiring distinct semantic treatment (Rett 2019)semantic contentat-issue not-at-issuedescriptive NAI illocutionary NAII alternatively:semantic contentdescriptiveat-issue descriptive NAIillocutionaryillocutionary NAIProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsdescriptive vs. illocutionary NAI contentI descriptive NAI markers: appositives, evidentials,utterance modifiers, CIs (Potts 2005)(15) a. Allegedly, Jane won the race.b. Frankly, Jane won the race.I illocutionary NAI markers: alas, (un)fortunately,mirative markers like Finnish pa¨ (Rett 2019)(16) a. Alas, Jane won the race.b. (Wow) Jane won the race!I descriptive not-at-issue markers contribute to thedescriptive content of an utterance – what is saidI “illocutionary content” is not-at-issue meaningpertaining to how the speaker is using the utterance inthe contextProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsdescriptive vs. illocutionary NAI contentI descriptive NAI markers: appositives, evidentials,utterance modifiers, CIs (Potts 2005)(15) a. Allegedly, Jane won the race.b. Frankly, Jane won the race.I illocutionary NAI markers: alas, (un)fortunately,mirative markers like Finnish pa¨ (Rett 2019)(16) a. Alas, Jane won the race.b. (Wow) Jane won the race!I descriptive not-at-issue markers contribute to thedescriptive content of an utterance – what is saidI “illocutionary content” is not-at-issue meaningpertaining to how the speaker is using the utterance inthe contextProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsdescriptive vs. illocutionary NAI contentI descriptive NAI markers: appositives, evidentials,utterance modifiers, CIs (Potts 2005)(15) a. Allegedly, Jane won the race.b. Frankly, Jane won the race.I illocutionary NAI markers: alas, (un)fortunately,mirative markers like Finnish pa¨ (Rett 2019)(16) a. Alas, Jane won the race.b. (Wow) Jane won the race!I descriptive not-at-issue markers contribute to thedescriptive content of an utterance – what is saidI “illocutionary content” is not-at-issue meaningpertaining to how the speaker is using the utterance inthe contextProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsdiagnostics of illocutionary content1. susceptibility to Moore’s ParadoxI standard Moore’s Paradox:(17) #It’s raining, but I don’t believe it’s raining.I Murray (2010): denial of mirativity is paradoxical (#);denial of evidentiality is contradictory (⊥)I to the extent that English speakers recognize thisdifference, the English data pattern the same way.(18) a. #⊥Allegedly, Jane lost the race, but no onealleged she did.b. # Alas, Jane lost the race, but I’m notdisappointed she did.c. #Wow Jane lost the race! But I’m notsurprised she did.Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsdiagnostics of illocutionary content1. susceptibility to Moore’s ParadoxI Moore’s Paradox is suspended in suppositional orconditional contexts (Yalcin 2007)(19) Suppose it’s raining, but I don’t believe it is raining.I so is the content encoded by emotive markers (20-b),but not e.g. evidential adverbials (20-a)(20) a. #Suppose that, allegedly, Jane lost the race, butthat no one alleged that she did.b. Suppose that, alas, Jane lost the race, butthat I’m not disappointed she did.Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsdiagnostics of illocutionary content2. scope-taking: illocutionary markers associate with asingle salient propositionI they’re incompatible with utterances associated withmultiple propositions (Cheyenne, Rett & Murray 2013)(21) a. Mo´=e´-x-ho´′ ta˙heva´-hoo′oy/n=3-rem.pst-win-nar.3sgAe´nohe?Hawk‘Given the stories you heard, did Hawk win?’b. %Mo´=e´-ho´′ ta˙heva´-hoo′oy/n=3-win-nar.3sgAe´nohe?Hawk‘Given your surprise, did Hawk win?’I they scope differently with other sentential operatorsthan descriptive NAI markers do(22) a. #Alas, if Jane loses, at least we’ll flip the Sen.b. If, alas, Jane loses, at least we’ll flip the Sen.(23) a. #Apparently, if Jane loses, I will run for office.b. #If, apparently, Jane loses, I will run for office.Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsanalysis, informally(24) (Wow) Jane, who hates politics, won the race!a. at-issue: Jane won the raceb. descriptive NAI: Jane hates politicsc. illocutionary NAI: speaker hadn’t expectedJane to win the raceI at-issue content:I acts as a proposal to submit the proposition to theCommon Ground (Stalnaker 1973)I descriptive not-at-issue content:I directly updates the Common Ground (Murray 2010)I illocutionary content:I updates the speaker’s Discourse Commitments(Gunlogson 2001)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsanalysis, informally(24) (Wow) Jane, who hates politics, won the race!a. at-issue: Jane won the raceb. descriptive NAI: Jane hates politicsc. illocutionary NAI: speaker hadn’t expectedJane to win the raceI at-issue content:I acts as a proposal to submit the proposition to theCommon Ground (Stalnaker 1973)I descriptive not-at-issue content:I directly updates the Common Ground (Murray 2010)I illocutionary content:I updates the speaker’s Discourse Commitments(Gunlogson 2001)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsanalysis, informally(24) (Wow) Jane, who hates politics, won the race!a. at-issue: Jane won the raceb. descriptive NAI: Jane hates politicsc. illocutionary NAI: speaker hadn’t expectedJane to win the raceI at-issue content:I acts as a proposal to submit the proposition to theCommon Ground (Stalnaker 1973)I descriptive not-at-issue content:I directly updates the Common Ground (Murray 2010)I illocutionary content:I updates the speaker’s Discourse Commitments(Gunlogson 2001)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsa note on sincerity conditionsI extant dynamic accounts of declarative mood do notformally model the sincerity condition, i.e. ‘Speakerbelieves that p’I arguably, emotive markers like mirativity do the samesort of thing, contribute something like ‘Speaker issurprised that p’I Searle 1965: Moore’s Paradox arises when a sinceritycondition is deniedI mirativity and other emotive markers also susceptible toMoore’s Paradox (Searle & Vanderveken 1985: alas quaspeech-act modifier)I Discourse Commitments (Gunlogson 2001):I originally proposed to account for speaker bias in risingdeclarativesI originally characterized in terms of the speaker’s beliefs(and “public in the sense that the participant ismutually recognized as committed to them”).I I treat public commitment as a proxy for belief, for thepurpose of modeling (sincere) conversationProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsa note on sincerity conditionsI extant dynamic accounts of declarative mood do notformally model the sincerity condition, i.e. ‘Speakerbelieves that p’I arguably, emotive markers like mirativity do the samesort of thing, contribute something like ‘Speaker issurprised that p’I Searle 1965: Moore’s Paradox arises when a sinceritycondition is deniedI mirativity and other emotive markers also susceptible toMoore’s Paradox (Searle & Vanderveken 1985: alas quaspeech-act modifier)I Discourse Commitments (Gunlogson 2001):I originally proposed to account for speaker bias in risingdeclarativesI originally characterized in terms of the speaker’s beliefs(and “public in the sense that the participant ismutually recognized as committed to them”).I I treat public commitment as a proxy for belief, for thepurpose of modeling (sincere) conversationProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsa note on sincerity conditionsI extant dynamic accounts of declarative mood do notformally model the sincerity condition, i.e. ‘Speakerbelieves that p’I arguably, emotive markers like mirativity do the samesort of thing, contribute something like ‘Speaker issurprised that p’I Searle 1965: Moore’s Paradox arises when a sinceritycondition is deniedI mirativity and other emotive markers also susceptible toMoore’s Paradox (Searle & Vanderveken 1985: alas quaspeech-act modifier)I Discourse Commitments (Gunlogson 2001):I originally proposed to account for speaker bias in risingdeclarativesI originally characterized in terms of the speaker’s beliefs(and “public in the sense that the participant ismutually recognized as committed to them”).I I treat public commitment as a proxy for belief, for thepurpose of modeling (sincere) conversationProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsanalysis, formallyI Farkas & Bruce (2010), with Gunlogson’s (2001)Discourse Commitments (and Murray’s (2014)treatment of descriptive NAI content)I with a declarative mood operator(25) Declarative operator (D), for sentences Sp withat-issue content p and not-at-issue content q:D(Sp, a,Ki ) = Ko such that(i) DCa,o = DCa,i ∪ {〈believes, p〉}(ii) To = push(〈Sp; {p}〉,Ti )(iii) pso = psi ∪ {p}(iv) CGo = CGi ∪ {q}I plus “flavored Discourse Commitments”: ordered pairsof propositional attitudes and propositions, e.g.〈believes, p〉; 〈is-disappointed, p〉; 〈is-surprised, p〉Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsanalysis, formallyI Farkas & Bruce (2010), with Gunlogson’s (2001)Discourse Commitments (and Murray’s (2014)treatment of descriptive NAI content)I with a declarative mood operator(25) Declarative operator (D), for sentences Sp withat-issue content p and not-at-issue content q:D(Sp, a,Ki ) = Ko such that(i) DCa,o = DCa,i ∪ {〈believes, p〉}(ii) To = push(〈Sp; {p}〉,Ti )(iii) pso = psi ∪ {p}(iv) CGo = CGi ∪ {q}I plus “flavored Discourse Commitments”: ordered pairsof propositional attitudes and propositions, e.g.〈believes, p〉; 〈is-disappointed, p〉; 〈is-surprised, p〉Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsanalysis, formallyI Farkas & Bruce (2010), with Gunlogson’s (2001)Discourse Commitments (and Murray’s (2014)treatment of descriptive NAI content)I with a declarative mood operator(25) Declarative operator (D), for sentences Sp withat-issue content p and not-at-issue content q:D(Sp, a,Ki ) = Ko such that(i) DCa,o = DCa,i ∪ {〈believes, p〉}(ii) To = push(〈Sp; {p}〉,Ti )(iii) pso = psi ∪ {p}(iv) CGo = CGi ∪ {q}I plus “flavored Discourse Commitments”: ordered pairsof propositional attitudes and propositions, e.g.〈believes, p〉; 〈is-disappointed, p〉; 〈is-surprised, p〉Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsanalysis, formally(26) Mir, for clauses C with content p:Mir(C , a,Ki ) = (C , a,Ko) such that(i) DCa,o = DCa,i ∪ {〈is-surprised, p〉}(ii) To = push(〈C ; {p}〉,Ti )(27) JJane won the race!K = D(Mir(S , a,Ki )) = Kosuch that(i) DCa,o = DCa,i ∪ {〈believes, Jane won the race〉}(ii) To = push(〈S ; Jane won the race〉,Ti )(iii) pso = psi ∪ {Jane won the race}(iv) DCa,o = DCa,i ∪ {〈is-surprised, J won the race〉}I emotive markers can interact scopally with descriptiveNAI and AI content via sub-sentential dynamic updateI THM: Mirative markers, like all emotive markers,encode NAI content. But it’s qualitatively different thancanonical, ‘descriptive’ NAI content: it operates at anillocutionary levelProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsanalysis, formally(26) Mir, for clauses C with content p:Mir(C , a,Ki ) = (C , a,Ko) such that(i) DCa,o = DCa,i ∪ {〈is-surprised, p〉}(ii) To = push(〈C ; {p}〉,Ti )(27) JJane won the race!K = D(Mir(S , a,Ki )) = Kosuch that(i) DCa,o = DCa,i ∪ {〈believes, Jane won the race〉}(ii) To = push(〈S ; Jane won the race〉,Ti )(iii) pso = psi ∪ {Jane won the race}(iv) DCa,o = DCa,i ∪ {〈is-surprised, J won the race〉}I emotive markers can interact scopally with descriptiveNAI and AI content via sub-sentential dynamic updateI THM: Mirative markers, like all emotive markers,encode NAI content. But it’s qualitatively different thancanonical, ‘descriptive’ NAI content: it operates at anillocutionary levelProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsanalysis, formally(26) Mir, for clauses C with content p:Mir(C , a,Ki ) = (C , a,Ko) such that(i) DCa,o = DCa,i ∪ {〈is-surprised, p〉}(ii) To = push(〈C ; {p}〉,Ti )(27) JJane won the race!K = D(Mir(S , a,Ki )) = Kosuch that(i) DCa,o = DCa,i ∪ {〈believes, Jane won the race〉}(ii) To = push(〈S ; Jane won the race〉,Ti )(iii) pso = psi ∪ {Jane won the race}(iv) DCa,o = DCa,i ∪ {〈is-surprised, J won the race〉}I emotive markers can interact scopally with descriptiveNAI and AI content via sub-sentential dynamic updateI THM: Mirative markers, like all emotive markers,encode NAI content. But it’s qualitatively different thancanonical, ‘descriptive’ NAI content: it operates at anillocutionary levelProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsprosodically marked mirativityI unified semantic treatment of mirativity (and emotivemarkers writ large):I some languages do things with morphemes and syntaxthat other languages do with prosodyI a good reminder that prosody needs to be representedin our compositional semantics (Pierrehumbert &Hirschberg 1990)I previous work on the semantics of prosody:I prosody can mark illocutionary mood (Pierrehumbert1980, Jeong & Potts 2016)I prosody can mark orientation (speaker or hearer;Gunlogson 2001, Rudin 2018)I prosody can mark other content like uncertainty,incredulity (Hirschberg & Ward 1992)My claim: English exclamation intonation is multi-faceted,marking something like illocutionary content as well assecondary effects like surprise, emphasisProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsprosodically marked mirativityI unified semantic treatment of mirativity (and emotivemarkers writ large):I some languages do things with morphemes and syntaxthat other languages do with prosodyI a good reminder that prosody needs to be representedin our compositional semantics (Pierrehumbert &Hirschberg 1990)I previous work on the semantics of prosody:I prosody can mark illocutionary mood (Pierrehumbert1980, Jeong & Potts 2016)I prosody can mark orientation (speaker or hearer;Gunlogson 2001, Rudin 2018)I prosody can mark other content like uncertainty,incredulity (Hirschberg & Ward 1992)My claim: English exclamation intonation is multi-faceted,marking something like illocutionary content as well assecondary effects like surprise, emphasisProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsprosodically marked mirativityI unified semantic treatment of mirativity (and emotivemarkers writ large):I some languages do things with morphemes and syntaxthat other languages do with prosodyI a good reminder that prosody needs to be representedin our compositional semantics (Pierrehumbert &Hirschberg 1990)I previous work on the semantics of prosody:I prosody can mark illocutionary mood (Pierrehumbert1980, Jeong & Potts 2016)I prosody can mark orientation (speaker or hearer;Gunlogson 2001, Rudin 2018)I prosody can mark other content like uncertainty,incredulity (Hirschberg & Ward 1992)My claim: English exclamation intonation is multi-faceted,marking something like illocutionary content as well assecondary effects like surprise, emphasisProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsprosodically marked mirativityI unified semantic treatment of mirativity (and emotivemarkers writ large):I some languages do things with morphemes and syntaxthat other languages do with prosodyI a good reminder that prosody needs to be representedin our compositional semantics (Pierrehumbert &Hirschberg 1990)I previous work on the semantics of prosody:I prosody can mark illocutionary mood (Pierrehumbert1980, Jeong & Potts 2016)I prosody can mark orientation (speaker or hearer;Gunlogson 2001, Rudin 2018)I prosody can mark other content like uncertainty,incredulity (Hirschberg & Ward 1992)My claim: English exclamation intonation is multi-faceted,marking something like illocutionary content as well assecondary effects like surprise, emphasisProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsprosodically marked mirativityI unified semantic treatment of mirativity (and emotivemarkers writ large):I some languages do things with morphemes and syntaxthat other languages do with prosodyI a good reminder that prosody needs to be representedin our compositional semantics (Pierrehumbert &Hirschberg 1990)I previous work on the semantics of prosody:I prosody can mark illocutionary mood (Pierrehumbert1980, Jeong & Potts 2016)I prosody can mark orientation (speaker or hearer;Gunlogson 2001, Rudin 2018)I prosody can mark other content like uncertainty,incredulity (Hirschberg & Ward 1992)My claim: English exclamation intonation is multi-faceted,marking something like illocutionary content as well assecondary effects like surprise, emphasisProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsprosodically marked mirativityI unified semantic treatment of mirativity (and emotivemarkers writ large):I some languages do things with morphemes and syntaxthat other languages do with prosodyI a good reminder that prosody needs to be representedin our compositional semantics (Pierrehumbert &Hirschberg 1990)I previous work on the semantics of prosody:I prosody can mark illocutionary mood (Pierrehumbert1980, Jeong & Potts 2016)I prosody can mark orientation (speaker or hearer;Gunlogson 2001, Rudin 2018)I prosody can mark other content like uncertainty,incredulity (Hirschberg & Ward 1992)My claim: English exclamation intonation is multi-faceted,marking something like illocutionary content as well assecondary effects like surprise, emphasisProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsEnglish mirativity marking,a.k.a. exclamation intonationI sentence exclamation(28) a. Jane carves gorgeous sculptures.b. (Wow) Jane carves gorgeous sculptures!I exclamatives(29) a. What gorgeous sculptures Jane carves!b. Can Jane carve gorgeous sculptures!c. The gorgeous sculptures Jane carves!I discourse particles optionalI distinct from rhetorical questions, e.g. How cool is that?I exclamatives differ from sentence exclamations, inEnglish at least, in that the object of surprise needs tobe a degree instead of a proposition (Rett 2011)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsEnglish mirativity marking,a.k.a. exclamation intonationI sentence exclamation(28) a. Jane carves gorgeous sculptures.b. (Wow) Jane carves gorgeous sculptures!I exclamatives(29) a. What gorgeous sculptures Jane carves!b. Can Jane carve gorgeous sculptures!c. The gorgeous sculptures Jane carves!I discourse particles optionalI distinct from rhetorical questions, e.g. How cool is that?I exclamatives differ from sentence exclamations, inEnglish at least, in that the object of surprise needs tobe a degree instead of a proposition (Rett 2011)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsEnglish mirativity marking,a.k.a. exclamation intonationI sentence exclamation(28) a. Jane carves gorgeous sculptures.b. (Wow) Jane carves gorgeous sculptures!I exclamatives(29) a. What gorgeous sculptures Jane carves!b. Can Jane carve gorgeous sculptures!c. The gorgeous sculptures Jane carves!I discourse particles optionalI distinct from rhetorical questions, e.g. How cool is that?I exclamatives differ from sentence exclamations, inEnglish at least, in that the object of surprise needs tobe a degree instead of a proposition (Rett 2011)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsEnglish intonationIn English, utterances are associated with varyingintonational contours, i.e. tunesI tunes are composed of one or more intermediate phrasesI intermediate phrases each have a pitch accentI pitch is measured by tracking the speaker’s fundamentalfrequency (f0)I pitch accents are composed of one or more pitchtargets, with the star anchored to the stressed syllableI intermediate phrases are distinguishable in having:1. their own pitch accents (with a stressed syllable);2. their own phrase accent (with final lengthening);3. pitch range reset (a new ceiling after a down step)Annotated with MAE ToBI (Beckman & Ayers-Elam 1997)I relatively coarse: can’t mark some things, e.g. speaker’srelative height (high vs. extra-high)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsEnglish intonationIn English, utterances are associated with varyingintonational contours, i.e. tunesI tunes are composed of one or more intermediate phrasesI intermediate phrases each have a pitch accentI pitch is measured by tracking the speaker’s fundamentalfrequency (f0)I pitch accents are composed of one or more pitchtargets, with the star anchored to the stressed syllableI intermediate phrases are distinguishable in having:1. their own pitch accents (with a stressed syllable);2. their own phrase accent (with final lengthening);3. pitch range reset (a new ceiling after a down step)Annotated with MAE ToBI (Beckman & Ayers-Elam 1997)I relatively coarse: can’t mark some things, e.g. speaker’srelative height (high vs. extra-high)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsEnglish intonationIn English, utterances are associated with varyingintonational contours, i.e. tunesI tunes are composed of one or more intermediate phrasesI intermediate phrases each have a pitch accentI pitch is measured by tracking the speaker’s fundamentalfrequency (f0)I pitch accents are composed of one or more pitchtargets, with the star anchored to the stressed syllableI intermediate phrases are distinguishable in having:1. their own pitch accents (with a stressed syllable);2. their own phrase accent (with final lengthening);3. pitch range reset (a new ceiling after a down step)Annotated with MAE ToBI (Beckman & Ayers-Elam 1997)I relatively coarse: can’t mark some things, e.g. speaker’srelative height (high vs. extra-high)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsEnglish intonationIn English, utterances are associated with varyingintonational contours, i.e. tunesI tunes are composed of one or more intermediate phrasesI intermediate phrases each have a pitch accentI pitch is measured by tracking the speaker’s fundamentalfrequency (f0)I pitch accents are composed of one or more pitchtargets, with the star anchored to the stressed syllableI intermediate phrases are distinguishable in having:1. their own pitch accents (with a stressed syllable);2. their own phrase accent (with final lengthening);3. pitch range reset (a new ceiling after a down step)Annotated with MAE ToBI (Beckman & Ayers-Elam 1997)I relatively coarse: can’t mark some things, e.g. speaker’srelative height (high vs. extra-high)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsintermediate phrasesDoes the governor of Iowa endorse a radio program?L* H-H%I QUD: Which governors endorse a radio program?Does the governor of Iowa endorse a radio program?L* H- L* H-H%I QUD: out of the blue (Hirschberg & Ward 1992)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsintermediate phrasesDoes the governor of Iowa endorse a radio program?L* H-H%I QUD: Which governors endorse a radio program?Does the governor of Iowa endorse a radio program?L* H- L* H-H%I QUD: out of the blue (Hirschberg & Ward 1992)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsintermediate phrasesDoes the governor of Iowa endorse a radio program?L* H-H%I QUD: Which governors endorse a radio program?Does the governor of Iowa endorse a radio program?L* H- L* H-H%I QUD: out of the blue (Hirschberg & Ward 1992)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsthe semantics of pitch accentsI different canonical tunes (pitch accents+boundarytones) are associated with different illocutionary moods(Pierrehumbert 1980, Pierrehumbert & Hirschberg 1990)I declarative sentences: H* L-L%(30) Miriam made the lemonade.I polar questions & rising declaratives: L* H-H%(31) Miriam made the lemonade?I confirmation questions/rising imperatives: H* H-H%(32) Make the lemonade?I ...as well as perlocutionary effects (Jeong & Potts 2016)I falling intonation: authoritativeness/assertionI level intonation: annoyanceI rising intonation: politeness/positivity (cf. hedging)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsthe semantics of pitch accentsI different canonical tunes (pitch accents+boundarytones) are associated with different illocutionary moods(Pierrehumbert 1980, Pierrehumbert & Hirschberg 1990)I declarative sentences: H* L-L%(30) Miriam made the lemonade.I polar questions & rising declaratives: L* H-H%(31) Miriam made the lemonade?I confirmation questions/rising imperatives: H* H-H%(32) Make the lemonade?I ...as well as perlocutionary effects (Jeong & Potts 2016)I falling intonation: authoritativeness/assertionI level intonation: annoyanceI rising intonation: politeness/positivity (cf. hedging)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsresearch questions1. are English exclamations marked with uniform prosody?2. if so, how are they marked?3. how can we represent it semantically?4. (is the prosodic marking of mirativity in Englishsemantically arbitrary?)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionselicitation methodologyI two consultants: one female, one maleI conditions: 4 (construction) x 2 (±discourse particle)I stimuli: 32 items per condition (256 total), btwnsubjects (128 per subject)(33) context: You dont expect Julian to make beautifulpaintings, but you find out he did. You tell Sara:a. (Wow) Julian makes beautiful paintings!b. (Wow)What beautiful paintings Julian makes!c. (Wow) Does Julian make beautiful paintings!d. (Wow) The beautiful paintings Julian makes!(34) context: You don’t expect Ariel’s stories to causeconfusion, but you find out they have. You tell Sara:a. (Wow) Ariel’s stories caused confusion!b. (Wow) What confusion Ariel’s stories caused!c. (Wow) Did Ariel’s stories cause confusion!d. (Wow) The confusion Ariel’s stories caused!Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionselicitation methodologyI two consultants: one female, one maleI conditions: 4 (construction) x 2 (±discourse particle)I stimuli: 32 items per condition (256 total), btwnsubjects (128 per subject)(33) context: You dont expect Julian to make beautifulpaintings, but you find out he did. You tell Sara:a. (Wow) Julian makes beautiful paintings!b. (Wow)What beautiful paintings Julian makes!c. (Wow) Does Julian make beautiful paintings!d. (Wow) The beautiful paintings Julian makes!(34) context: You don’t expect Ariel’s stories to causeconfusion, but you find out they have. You tell Sara:a. (Wow) Ariel’s stories caused confusion!b. (Wow) What confusion Ariel’s stories caused!c. (Wow) Did Ariel’s stories cause confusion!d. (Wow) The confusion Ariel’s stories caused!Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionselicitation methodologyI fillers: 4 construction conditions: declaratives;wh-questions; polar questions; definite subjects/cleftsI 32 in each condition (128 total), btwn subjects (64 ea.)(35) a. Anna is good at chess.b. What is Anna good at?c. Is Anna good at chess?d. The game Anna is good at is chess.(36) a. James is an expert at knitting.b. What is James an expert at?c. Is James an expert at knitting?d. The thing James is an expert at is knitting.I consultants were instructed to read the sentences asthey’d be naturally uttered (in the provided context)I the recordings were transcribed in Praat usingMAE ToBI by one of the authors, and confirmed byanother trained transcriberProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionselicitation methodologyI fillers: 4 construction conditions: declaratives;wh-questions; polar questions; definite subjects/cleftsI 32 in each condition (128 total), btwn subjects (64 ea.)(35) a. Anna is good at chess.b. What is Anna good at?c. Is Anna good at chess?d. The game Anna is good at is chess.(36) a. James is an expert at knitting.b. What is James an expert at?c. Is James an expert at knitting?d. The thing James is an expert at is knitting.I consultants were instructed to read the sentences asthey’d be naturally uttered (in the provided context)I the recordings were transcribed in Praat usingMAE ToBI by one of the authors, and confirmed byanother trained transcriberProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionselicitation methodologyI fillers: 4 construction conditions: declaratives;wh-questions; polar questions; definite subjects/cleftsI 32 in each condition (128 total), btwn subjects (64 ea.)(35) a. Anna is good at chess.b. What is Anna good at?c. Is Anna good at chess?d. The game Anna is good at is chess.(36) a. James is an expert at knitting.b. What is James an expert at?c. Is James an expert at knitting?d. The thing James is an expert at is knitting.I consultants were instructed to read the sentences asthey’d be naturally uttered (in the provided context)I the recordings were transcribed in Praat usingMAE ToBI by one of the authors, and confirmed byanother trained transcriberProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionselicitation resultsboth consultants consistently:1. pronounced exclamations with L+H* pitch accents;L+H* H* H+!H* L* L*+HSE .96 .04 – – –WH .85 .10 .05 – –INV .94 .04 – – .02NOM .93 .03 – .04 –average .92 .05 .01 .01 <.01Table : proportion of items with L+H* pitch accents, femaleL+H* H* H+!H* L* L*+HSE 1 – – – –WH .96 .04 – – –INV 1 – – – –NOM .94 .06 – – –average .98 .03 – – –Table : proportion of items with L+H* pitch accents, maleProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionselicitation resultsboth consultants consistently:2. pronounced exclamations with extra-high targets:I the high target exceeds expected height for position(after a downstep, etc.);I target exceeds default pitch maximum by at least 5%I cf. fillers, with zero extra-high targets+wow −wowSE . 78 .78WH .94 .88INV .69 .53NOM .78 .94Table : proportion of items with at least one extra high target3. and inserted additional intermediate phrase boundariesI additional mid-sentence pitch-range resetsI additional nuclear pitch accents (NPAs) relative to fillersI consequently, words in the exclamations sounded moreprominent than filler counterpartsProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionselicitation resultsboth consultants consistently:2. pronounced exclamations with extra-high targets:I the high target exceeds expected height for position(after a downstep, etc.);I target exceeds default pitch maximum by at least 5%I cf. fillers, with zero extra-high targets+wow −wowSE . 78 .78WH .94 .88INV .69 .53NOM .78 .94Table : proportion of items with at least one extra high target3. and inserted additional intermediate phrase boundariesI additional mid-sentence pitch-range resetsI additional nuclear pitch accents (NPAs) relative to fillersI consequently, words in the exclamations sounded moreprominent than filler counterpartsProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsexample utterancesI sentence exclamation (no sentence particle):(37) Angelique bakes delicious desserts!I wh-exclamative (no sentence particle):(38) What delicious desserts Angelique bakes!Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsexample utterancesI inversion exclamative (sentence particle):(39) Wow, does Angelique bake delicious desserts!I nominal exclamative (sentence particle):(40) Wow, the delicious desserts Angelique bakes!Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsnuances across construction typeI these three prosodic properties are necessary but notsufficient for describing the intonational contour of thefour constructions we’re looking atI each construction manifests all three propertiesdifferently to form a unique tuneI they differ in their macrorhythm (peak frequency)...I ...and in what types of words get marked as prominentI intonation is functioning to maximally differentiate eachexclamation type from its non-exclamation counterpartI sentence exclamation vs. simple declarativeI wh-exclamative vs. wh-questionI inversion exclamative vs. yes/no questionI nominal exclamative vs. topicalized definiteI the intonational patterns are the complete opposite ofone another in terms of:I acoustic salience (for polar constructions: most/leastsalient);I what is prominent (for wh-constructions);I slow vs. fast macrorhythm (for polar constructions)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsnuances across construction typeI these three prosodic properties are necessary but notsufficient for describing the intonational contour of thefour constructions we’re looking atI each construction manifests all three propertiesdifferently to form a unique tuneI they differ in their macrorhythm (peak frequency)...I ...and in what types of words get marked as prominentI intonation is functioning to maximally differentiate eachexclamation type from its non-exclamation counterpartI sentence exclamation vs. simple declarativeI wh-exclamative vs. wh-questionI inversion exclamative vs. yes/no questionI nominal exclamative vs. topicalized definiteI the intonational patterns are the complete opposite ofone another in terms of:I acoustic salience (for polar constructions: most/leastsalient);I what is prominent (for wh-constructions);I slow vs. fast macrorhythm (for polar constructions)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsnuances across construction typeI these three prosodic properties are necessary but notsufficient for describing the intonational contour of thefour constructions we’re looking atI each construction manifests all three propertiesdifferently to form a unique tuneI they differ in their macrorhythm (peak frequency)...I ...and in what types of words get marked as prominentI intonation is functioning to maximally differentiate eachexclamation type from its non-exclamation counterpartI sentence exclamation vs. simple declarativeI wh-exclamative vs. wh-questionI inversion exclamative vs. yes/no questionI nominal exclamative vs. topicalized definiteI the intonational patterns are the complete opposite ofone another in terms of:I acoustic salience (for polar constructions: most/leastsalient);I what is prominent (for wh-constructions);I slow vs. fast macrorhythm (for polar constructions)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsnuances across construction typeI wh-question intonationI overall tune same as simple declaratives (H* L-L%)I wh-word is not prosodically prominent (Pierrehumbert1980), surprising given its discourse significance(41) What does Greta know how to do?I wh-exclamative intonationI wh-word is highly prominent, marked with (L+)H*87.5%I wh-word target is extra-high in 62.5% ofwh-exclamatives(42) What delicious desserts Angelique bakes!Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsnuances across construction typeI polar question intonationI canonical tune: L* H-H%I pitch accents are relatively sparse, resulting in slowmacrorhythm frequency (few peaks/valleys)(43) Did Yolanda hear that the restaurant is closing?I inversion exclamative intonationI most content words are marked with L+H*I results in a fast macrorhythm (many peaks/valleys)(44) Wow, does Angelique bake delicious desserts!Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsdifferent roles for different prosodic propertiesin terms of the prosody:I exclamation seems to be encoded in a specific pitchaccent, L+H*I but it also seems to require super-tonal properties,namely extra-high targets and additional intermediatephrase boundariesI what’s the right semantic treatment for this bundle ofproperties?I what’s the right semantic treatment for what mightprove to be a gradient effect?Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsprosodic iconicityit seems possible that at least some of these characteristicsof mirative prosody are non-arbitrary1. L+H* pitch accents – likely semantically arbitraryI there are other ways of marking mirativity cross-linguistically that don’t involve L+H*I there are other uses of L+H*, even in English: prosodicfocus marking (Pierrehumbert 1980, Selkirk 1995)2. extra-high targets – likely semantically non-arbitraryI unclear whether they occur with other mirativitystrategies, e.g. Finnish pa¨I but no other uses of extra-high targets in English3. additional intermediate phrases – ??I unclear whether they occur with other mirativitystrategiesI there are other uses of extra boundary insertion in focusmarking in English (Pierrehumbert 1980) and otherlanguages (Royer & Jun 2019)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsprosodic iconicityit seems possible that at least some of these characteristicsof mirative prosody are non-arbitrary1. L+H* pitch accents – likely semantically arbitraryI there are other ways of marking mirativity cross-linguistically that don’t involve L+H*I there are other uses of L+H*, even in English: prosodicfocus marking (Pierrehumbert 1980, Selkirk 1995)2. extra-high targets – likely semantically non-arbitraryI unclear whether they occur with other mirativitystrategies, e.g. Finnish pa¨I but no other uses of extra-high targets in English3. additional intermediate phrases – ??I unclear whether they occur with other mirativitystrategiesI there are other uses of extra boundary insertion in focusmarking in English (Pierrehumbert 1980) and otherlanguages (Royer & Jun 2019)Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsconclusionsmirativity is encoded in different types of strategies.we need a semantic account that:I can treat a given semantic phenomenon across strategytypes (morphologic, syntactic, prosodic)I in particular, can treat prosody as one of the ‘parts’ inthe ‘parts and whole’ notion of compositionalityI in terms of the formalism:I intuitively, mirativity is not part of descriptive contentI diagnostics confirm differences between mirativity (andother emotive markers) on the one hand and descriptivenot-at-issue markers (e.g. evidentials) on the otherI I model these differences in a dynamic context-basedframework...I with descriptive content updating the Common Ground,I and with illocutionary content – like mirativity –updating a speaker’s Discourse CommitmentsProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsconclusionsmirativity is encoded in different types of strategies.we need a semantic account that:I can treat a given semantic phenomenon across strategytypes (morphologic, syntactic, prosodic)I in particular, can treat prosody as one of the ‘parts’ inthe ‘parts and whole’ notion of compositionalityI in terms of the formalism:I intuitively, mirativity is not part of descriptive contentI diagnostics confirm differences between mirativity (andother emotive markers) on the one hand and descriptivenot-at-issue markers (e.g. evidentials) on the otherI I model these differences in a dynamic context-basedframework...I with descriptive content updating the Common Ground,I and with illocutionary content – like mirativity –updating a speaker’s Discourse CommitmentsProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsconclusionsmirativity is encoded in different types of strategies.we need a semantic account that:I can treat a given semantic phenomenon across strategytypes (morphologic, syntactic, prosodic)I in particular, can treat prosody as one of the ‘parts’ inthe ‘parts and whole’ notion of compositionalityI in terms of the formalism:I intuitively, mirativity is not part of descriptive contentI diagnostics confirm differences between mirativity (andother emotive markers) on the one hand and descriptivenot-at-issue markers (e.g. evidentials) on the otherI I model these differences in a dynamic context-basedframework...I with descriptive content updating the Common Ground,I and with illocutionary content – like mirativity –updating a speaker’s Discourse CommitmentsProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsconclusionsmirativity is encoded in different types of strategies.we need a semantic account that:I can treat a given semantic phenomenon across strategytypes (morphologic, syntactic, prosodic)I in particular, can treat prosody as one of the ‘parts’ inthe ‘parts and whole’ notion of compositionalityI in terms of the formalism:I intuitively, mirativity is not part of descriptive contentI diagnostics confirm differences between mirativity (andother emotive markers) on the one hand and descriptivenot-at-issue markers (e.g. evidentials) on the otherI I model these differences in a dynamic context-basedframework...I with descriptive content updating the Common Ground,I and with illocutionary content – like mirativity –updating a speaker’s Discourse CommitmentsProsodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsconclusionsI all exclamations, in English, are pronounced with:1. L+H* pitch accents;2. extra-high targets;3. additional intermediate phrase boundariesI there is a clear role for the L+H* pitch accentI there’s other evidence that pitch accents operate at theillocutionary level (Jeong & Potts 2016)I I analyze the pitch accent as the mirativity marker –effectively an illocutionary mood modifier – in Englishexclamations (Rett 2019)I but there’s a question of what to do with the otherprosodic components of mirativity marking in EnglishI it’s intuitive to think that there’s somethingnon-arbitrary about the use of extra-high and additionaltargets to signify surprise or unexpectednessI and while they aren’t necessary components ofmirativity marking, they seem to be sufficientI we’re currently running a naturalness rating task to seehow important the super-tonal properties are for thingslike perceived speaker surprise, sincerity, etc.Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsconclusionsI all exclamations, in English, are pronounced with:1. L+H* pitch accents;2. extra-high targets;3. additional intermediate phrase boundariesI there is a clear role for the L+H* pitch accentI there’s other evidence that pitch accents operate at theillocutionary level (Jeong & Potts 2016)I I analyze the pitch accent as the mirativity marker –effectively an illocutionary mood modifier – in Englishexclamations (Rett 2019)I but there’s a question of what to do with the otherprosodic components of mirativity marking in EnglishI it’s intuitive to think that there’s somethingnon-arbitrary about the use of extra-high and additionaltargets to signify surprise or unexpectednessI and while they aren’t necessary components ofmirativity marking, they seem to be sufficientI we’re currently running a naturalness rating task to seehow important the super-tonal properties are for thingslike perceived speaker surprise, sincerity, etc.Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsconclusionsI all exclamations, in English, are pronounced with:1. L+H* pitch accents;2. extra-high targets;3. additional intermediate phrase boundariesI there is a clear role for the L+H* pitch accentI there’s other evidence that pitch accents operate at theillocutionary level (Jeong & Potts 2016)I I analyze the pitch accent as the mirativity marker –effectively an illocutionary mood modifier – in Englishexclamations (Rett 2019)I but there’s a question of what to do with the otherprosodic components of mirativity marking in EnglishI it’s intuitive to think that there’s somethingnon-arbitrary about the use of extra-high and additionaltargets to signify surprise or unexpectednessI and while they aren’t necessary components ofmirativity marking, they seem to be sufficientI we’re currently running a naturalness rating task to seehow important the super-tonal properties are for thingslike perceived speaker surprise, sincerity, etc.Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsthanks!contact me: rett@ucla.educontact Beth Sturman: bsturman@ucla.eduread about exclamations:Rett (2010), “Exclamatives, degrees, and speech acts”read about mirativity strategies:Rett (2012), “Mirativity across constructions and languages”read about mirative evidentials:Rett & Murray (2013), “A semantic account of mirativeevidentials”read about illocutionary content:Rett (2019), “The semantics of emotive markers and otherillocutionary content”Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsappendix A: flavored Discourse CommitmentsI I generalize Discourse Commitments to propositionalattitudes other than belief (inspired by Portner 2006):(45) Flavored Discourse CommitmentsLet DCa be a set of pairs representing the publiccommitments of a with respect to a discourse inwhich a and b are the participants, where:a. 〈believes, p〉 is a p.c. of a iff ‘a believes p’ is amutual belief of a and b;b. 〈is-surprised, p〉 is a public commitment of a iff‘a is surprised that p’ is a mutual belief of aand b.I I thus reformalize the sincerity condition on assertion(46) Declarative operator (D), for sentences Sp withat-issue content p and not-at-issue content q:D(Sp, a,Ki ) = Ko such that(i) DCa,o = DCa,i ∪ 〈believes, p〉 ...Prosodicallymarked mirativityJessica Rettoverviewmirative markingdefining mirativitymirative strategiesa unified semanticanalysisillocutionary contenta semantics forillocutionary contentprosodicallymarked mirativityprosody: an overviewelicitationresultsdiscussionconclusionsappendix B: scope-takingI emotive markers contribute their own restriction to thespeaker’s DC set, which (in the case of embedded clauses)requires a sub-sentential dynamic update(47) Jit’s raining alasK = A(S , a,Ki ) = (S1, a,Ko1 ) s.t.(i) DCa,o1 = DCa,i ∪ {〈is-disapp., It’s raining〉}(ii) To1 = push(〈S1; {It’s raining}〉,Ti )(48) JIt’s possible that it’s raining, alasK =D(S2, a,Ko1 )) = Ko2 such that(i) DCa,o2 = {DCa,i ∪ {〈is-disappointed, It’sraining〉}} ∪ {〈believes, It’s poss. it’s raining〉}(ii) To2 = push(〈S2; {It’s poss. it’sraining}〉, (push(〈S1; {It’s raining}〉,Ti )))(iii) pso = psi ∪ {It’s poss. it’s raining}(iv) CGo = CGiI while the proposition that it’s raining is pushed to the top ofthe stack in the update for the embedded clause, it is nolonger at the top after the utterance of the matrix sentence

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.73804.1-0389873/manifest

Comment

Related Items