Covert Emotive Modality Is a Monster

1
Covert Emotive Modality Is aMonster
Logic and Engineering of Natural Language
Semantics 2006 Tower Hall Funabori, Tokyo, June
5-6, 2006

• Sumiyo Nishiguchi
• Stony Brook University
• Osaka University
• snishigu_at_ic.sunysb.edu

2
Abstract

• It has been argued that attitude reports shift
  reference of indexicals in the embedded clauses
  in some languages (Schlenker 1999, 2003 Anand
  and Nevins 2004).
• I argue that implicit speaker attitudes on
  factive propositions are a context shifting
  operator which changes context parameters.

3

• I base my argument on the following three
  mono-clausal constructions
• i) fake past
• ii) fake present
• iii) out-of-the-blue wide-scope taking
  also/too
• N.B. I adapt the term fake' from Iatridou
  (2000).

4

• 1. Fake past
• Surprise licenses non-past interpretations of
  the past tense (Teramura 1984) with negative
  presuppositions.
• (1) Oh, it was here (all along).
• (2) A, koko-ni at-ta/a-ru (Japanese)
• Oh here-LOC be-PAST/be-PRES
• Oh, it was/is here'

5

• 2. Fake present
• Tense in narratives alternates between past and
  present (Klein 1994 Teramura 1984, among
  others).
• Alternation between past and present directs
  readers to re-experience narratives (Soga 1983).
• (3) Picchaa nage-ta. Ut-ta. Ichiro
  hashi-ru.
• Pitcher throw-PAST hit-PAST Ichiro
  run-PRES
• Oshii. Auto.
• sorry out
• The pitcher threw a ball. (Ichiro) hit it.
  Ichiro runs. Oh, no. He is out. (Japanese)

6

• 3. Wide-scope taking discourse initial too/mo
• Speaker's sentiments license wide-scope focus
  particle mo also/evenand too without explicit
  antecedents.
• (4) It's nice here, too.
• (5) Yo-mo huke-ta. (Japanese)
• night-also pass-PAST
• Its become late'

7
What is a Monster?

• Monster Def an operator on character which is a
  function from context to content/intension
• Kaplan (1977) there is no monster
• The indexicals, e.g., I, you, it that, this
  here, now, tomorrow, do not change the references
• Schlenker (1999,2003) All attitude predicates are
  monsters

8
Monster supporters

• Schlenker (1999,2003) Attitude verbs quantify
  over contexts of thought or of speech. Attitude
  predicates are monsters that shift the references
  of indexicals.
• As evidence,
• -Amharic first person pronoun shifts its
  reference into third person under attitude verbs
  (Schlenker 1999, 2003).
• Anand and Nevins (2004)
• -In Zazaki, the verb vano (say) shifts
  indexicals
• I, you, here and yesterday in its scope.

9
Limited evidence for a monster

• All supporting arguments for monsters have been
  based on the indexical shift in embedded context
  under attitude predicates.

10
Mono-clausal factive sentences

• The three constructions discussed in this paper
  are factive simple sentences, not embedded under
  attitude predicates, but temporal and world
  parameters shift.
• I argue that speaker's emotive/bouletic (in view
  of what I want) and epistemic speculative
  modality (in view of what I know, Kratzer 1991)
  is a context shiftable operator.

11

• Surprise, empathy and sentimentality affect
  temporal interpretations and satisfy
  presuppositions.
• (6) MODAL(fltlttc, wc, acgt, ltti, wigtgt)
• (fltltti, wc, acgt, ltti, wigtgt
• (t time, w world, a speaker, c context, i
  index)

12
Generalized quantifier

• Covert modal functions as a determiner taking
  negative presupposition in the restrictor and
  overt predicate in the nuclear scope (Kratzer
  1991 Berman 1991 von Fintel 1994 Ippolito
  2003).

13
Organization of the paper

• Sections 3 and 4 examine mono-clausal fake past
  and fake present sentences and show that modality
  distorts temporal interpretation.

14

• Section 5 shows that emotive modality
  accommodates the presupposition of too,
  wide-scope taking mo/to/ye also/too used out of
  the blue in Japanese, Korean and Chinese.
• I argue that speakers sentiments shift contexts
  so that presuppositions are satisfied.

15

• 1. Fake Past

16

• The past tense marker can receive non-past
  interpretation when associated with discovery,
  fulfillment of expectation, recalling of a plan
  (Teramura 1984, among others) often as
  exclamatives.

17
Fake Past and Aktionsarten Fake Past of
Discovery

• English
• Stative predicates
• (7) Oh, it was here (all along).
• Eventive predicates
• (8) Oh, the bus came/is coming.
• (NB The terminology fake' is taken from
  Iatridou (2000).)

18

• Japanese/Korean
• Stative predicates
• (9) A, koko-ni at-ta/a-ru. (Japanese)
• Oh here-LOC be-PAST/be-PRES
• Oh, it was here
• (10)Chek-i yogi iss-ot-ne. (Korean)
• book-NOM here be-PAST-EXC
• Oh, the was here'

19

• Eventive predicates
• (11) Basu-ga ki-ta. (Japanese)
• bus-NOM come-PAST
• The bus is coming
• (12) Ya ush-la. (Russian)
• I go-PAST
• I am leaving

20
The past tense refers to present, not simple past

• (13) Oh, the book was here. But it is not here
  anymore.
• (14) A, shinbun-ga koko-ni at-ta.
• Oh newspaper-Nom here-Loc be-Past
• Demo ima-wa mo nai.
• but now-Top already Neg
• Oh, the newspaper was here. But it's not
  here anymore'

21
Fake past of remembrance

• (15) What was your name? (Teramura 1982)
• (16) Where did you live?
• (17) Onamae-wa nan-deshi-ta-ka.
• Name-TOP what-HON-PAST-Q
• What was your name ?
• (18) Osumai-wa dochira-deshi-ta-ka-ne.
• residence-TOP where-HON-PAST-Q-PAR
• Where did you live?

22
Mismatched temporal adverbials

• Japanese
• (19) Asu-wa Maria-no tanjobi-dat-ta.
• tomorrow-Top Maria-GEN birthday-be-PAST
• Tomorrow is Maria's birthday
• Mandarin
• (20) Mintian you-le wanyan. tomorrow
  have-PERF party
• I had a party tomorrow'
• English
• (21) There was a party tomorrow.
• Antecedent of counterfactuals can (Ippolito
  2003)
• (22) If it rained tomorrow, I would go shopping.

23
Then,

• Tense is a shiftable indexical.
• What shifts tense?

24
Implicit attitude is a monsterous function that
changes a context parameter

• (23) fake( past flttc, wc, ac, hcgt, ltti,wigt)
• past fltti, wc, sc, hcgt,ltti,wigt
• (ttime, wworld, aspeaker, hhearer,
  ccontext, iindex, ti lt tc, cltwc, tc, acgt,
  iltwi, tigt)

25

• In the framework of the double index system
  (Lewis 1980), the ordinary past tense morphology
  shifts the temporal index into the prior time
• present f lttc, wc, sc, hcgt, lttc,wcgt
• past flttc, wc, sc, hcgt, ltti,wigt
• past f1 iff there is time ti prior to the
  utterance time tc

26

• Fake tense operator
• (24) Where cltwc, tc, acgt, iltwi, tigt,
• ti is prior to tc, c DcDsDe, s DsDwDt
• Fake ((cs)?t) ?((cs)?t)
• Fake (fltc, igt)1 iff fltcti/tc, igt1

27
Necessary condition for fake past interpretation
surprise

• Speaker's surprise due to negative presupposition
  causes fake past interpretation in simple
  sentences.
• (25) (Nai-to omotte-i-ta-ra,) at-ta.
• NEG-COMP think-be-PAST-then be-PAST
• It was here (surprisingly).'

28

• Without surprise, the fake past interpretation
  cannot be obtained.
• (1)-(22) would only refer to the past state or
  events.
• (26) Hon-ga at-ta.
• book-NOM be-PAST
• The book was (used to be) here

29
Or, the speakers expectation is realized

• (27) (Kuru-to omotte-i-ta basu-ga yappari)
  ki-ta.
• come-COMP think be-PAST bus-NOM as I expected
  come-PAST
• The bus is coming (as expected)
• The speaker doubted or has not been sure if p.
• The common ground contains both p worlds and
  non-p worlds
• Fake past assertions disambiguates the actual
  world (cf. Stalnaker 2004)

30
The data set is incomplete before the utterance
(Veltman 1981)

• The speaker does not know enough data but expects
  that the book is not here the bus is
  coming tomorrow is not Marys birthday
• The data set becomes complete by seeing the facts
  or remembering the facts

31

• Exp Expectation function based on the available
  data
• (28) Expa(wi)(ti)fwi,ti?Expa(wc)(tc)fwc,tc
  
• ?Knowa (wc)(tc)fwc,tc
• (29) Expa(wi)(ti)fwi,ti?Expa(wc)(tc)fwc,t
  c ?Knowa(wc)(tc)fwc,tc
• (tilttc, tc utterance time, wc actual world)

32
Unaccusativity Verb Classes of Fake Past

• Fake past predicates are mostly limited to
  unaccusative verbs such as be, exist, and come
  (cf. Kusumoto 2001 Ogihara 2004 for relative
  clauses).

33
Unaccusativity test

• VP internal numerals associate with the surface
  subject (Miyagawa 2004)
• (30)Honi-ga VPtsukue-no ue-ni ti ni-satsu
  at-ta.
• book-NOM desk-GEN up-LOC 2-CL
  be-PAST
• There were two books on the desk
• (31)Basu-ga VP ekimae-ni ti ni-dai ki-ta.
• bus-NOM station-front-LOC 2-CL come-PAST
• Two buses came in front of the station

34

• (32) A, warat-ta.
• oh smile-Past
• Oh, (the baby) is smiling
• (32) A, gakuseii-ga butai-de ti san-nin
  warat-ta. oh student-NOM
  stage-LOC 3-CL smile-PAST
• Oh, the three students laughed on the stage

35

• (33) Shimat-t-a.
• close-PAST-be
• Oh, no
• (33) Gakuseii-ga mae-de ti san-nin shimat-ta.
  student-NOM front-LOC 3-CL close-PAST
• The three students made a mistake in front

36
Adjectives (Individual-Level)

• (34) Yo-kat-ta.
• good-be-PAST
• Thanks goodness' (when a lost wallet was
  returned with money)
• (35) (While I expected it to be blue)
  Kiiro-kat-ta.
• yellow-be-PAST
• It is yellow'

37
Conversational Backgrounds

• 1. Speculative epistemic necessity/possibility
• must/probably/might f
• 2. Stereotypical conversational background (in
  view of the normal course of events)
• For all w, w?W, for any A?P(W) wAwiff
  pp?A and w ?p ?pp?A and w ?p
  (Kratzer1991)
• 3. Bouletic modality (in view of what I want)
• fworlds are ranked higher than f worlds

38

• Modal base f in view of the what I know
• (cs)?((cs)? t )? t)
• Ordering source g in view of normal course of
  events
• (cs)?((cs)? t )? t)
• Ordering source h in view of what I want
• (cs)?((cs)?t )?t)
• Where DsDwDt, DcDsDe

39

• (36) f
• h ordering source-bouletic
• g ordering source - stereotypical
• MODAL f modal base speculative
  modal
• (cf. Kratzer 1991 von Fintel and Iatridou 2005)

40

• (37) fake (past) (wc)(f)(g)(h)(f)
• 1 if wc?maxg(wi)(f (wi) ) ?wc?maxh(wi)(f (wi) )
  f(wc)1,
• past(wc)(f)(g)(f), otherwise.
• Where for a given strict partial order ltp on
  worlds, define the selection function maxp that
  selects the set of ltp -best worlds from any set X
  of worlds
• ? X? W maxp(X)w?X ? w'?X w'ltp w

41
Restrictive quantification

• Modal scopes over due to its quantificational
  force (Lewis 1968, 1973 Kripke 1972). Modal
  takes the presupposition as its restrictor, and
  the assertion in its nuclear scope (Berman 1991
  von Fintel 1994 Heim 1982 Diesing 1992).
• (38) MODAL ?i. fi ?i.fi
• determiner restrictor nuclear
  scope
• surprisingly while expecting f f is
  true

42
Negative presupposition as cataphora

• (39)
• MODAL P 1
• MODAL not it2 MODAL P1
  VP2
• ltMODALgt ltnot
  it2gt bus come

43

• 2. Fake Present

44
Narrative Present

• Narratives would freely alternate tense between
  past and present when storytelling as in (40)
  (Klein 1994 Mikami 1953 Nara 2001, among
  others).

45

• (40) Kocho-wa usuhige-no aru iro-no kuroi me-no
  okina
• principal-TOP mustache-GEN be color-GEN black
  eye-GEN big
• tanuki-no-yona otoko-de a-ru. Yani
  mottaibet-te-i-ta.
• badger-GEN-like man-be-PRES Terribly
  pompous-be-PAST
• The principal was a dark complexioned man, with
  a whistery mustache and large eyes like a badger.
  He was pompous.'
• (Soseki Natsume, Bocchan, quoted from Nara
  (2001), glossed by the author)

46

• Fake present tense invites readers into a
  depicted world so that the
• reader experiences the story as if present (Soga
  1983).

47
NARRATOR SAYSfunction and empathy

• The proposition is embedded under NARRATOR SAYS
  function which shifts the context.
• It is the empathy of both writer and reader which
  shifts the context parameter
• (41) NARRATOR SAYS((flttc, wc, sc, hcgt, ltti,
  wi, si, higt)
• f ltti, wi, si, higt ,ltti, wi, si, higt

48

• 3. Discourse Initial too/mo

49
Discourse Initial Too with Surprise

• (42) He is nice. He is a linguist, too.
• (43) Its nice here, too.

50
Wide-scope Mo also/even in Japanese

• Mo also/too, a focus marker or a
  quantifier-like element (Kuroda 1969) in
  Japanese, can associate with the whole
  proposition (Numata 1986, 2000).
• This mo takes wide scope over unaccusative
  predicates out of the blue without explicit
  antecedent that satisfies the presupposition. For
  example, (44) is usable without particular
  antecedent.

51

• (44) Yo-mo huke-ta.
• night-also pass-Past
• It grew late
• Semantically, mo takes wide scope.
• LF mo yo-ltmogt huke-ta
• also night pass-Past
• It grew late'

52
About mo

• A particle attached to noun phrases in Japanese.
• NPmo1 also
• (45) Ken-mo ki-ta.
• Ken-also come-Past
• Ken came, too'

53

• Mo2 obtains the meaning of even' when the NP is
  focused (Watanabe 2004).
• (46) KenF-mo ki-ta
• Ken-also come-Past
• Ken came, too'
• The even' mo2 forms NPIs with indeterminates
  (wh-words) (Kuroda 1965 Watanabe 2004 cf.
  Lahiri 1998)
• (47) Dare-mo ko-nai.
• who-mo come-Neg
• Nobody comes'

54
Mo3

• (48)Haru-mo takenawa-ni nari-mashi-ta.
  spring-also peak-GOAL become-HON-PAST The
  spring has reached its peak'
• (Numata 2000 172)
• (49) Ko-no saifu-mo huruku-nat-ta.
• This-GEN wallet-also old-become-PAST
• This wallet has become old

55

• Mo evokes sentiments that a nominative case
  marker ga would not.
• In (48), the speaker feels pleasant to find that
  spring has reached its peak.
• In (49), the speaker feels touched to see her
  worn bag, remembering the past. Covert emotive,
  e.g., Im glad that, happily, be touched with or
  I regret, is a monsterous function which
  satisfies the presupposition.

56
Korean to also/even

• Korean to also/even' has similar usage.
• (50) Pom-to wat-ta.
• spring-also come-PAST
• Spring came' (That's why I'm so sad)
• To also/even' demonstrates speaker's attitudes.

57
Sentence focus ye also' in Mandarin

• (51) Qiutian ye lai-le.
• fall also come-PERF
• (In view of the foregoing events) Fall came'

58

• Numata (2000) claims that such mo (also/too)
  attenuates the strength of assertion by giving
  rise to fictitious presupposed events.

59
Speaker sentimental modality shifts the context

• Speakers sentimental modality
• licenses too/mo3/to/ye
• shifts the context into a world in which the
  presupposed events exist so that the
  presuppositions of too/mo3/to/ye are
  accommodated.

60

• (50) MODALemotive( mo-flttc, wc, sc, hcgt, ltti,
  wi, si, higt)
• mo-f ltti, wi, sc, hcgt ,ltti, wi, si, higt

61
Unaccusativity

• The predicates of this type of mo are either
  unaccusative verbs with ta PAST, or adjectives.
  E.g., huke-ta have grown late, owari-ni
  chikazu-i-ta have neared the end,
  takenawa-ni-naru have reached the peak,
  iro-ase-ta have faded the color.

62
Restrictive Quantification by Emotive Modality

• Modality functions as a determiner which takes
  negative presupposition as the restrictor and the
  overt unaccusative VP in the nuclear scope (cf.
  Heim 1982 Berman 1991 von Fintel 1994).
• The negative counterpart is a copy of the overt
  proposition, which is a sentential cataphora
  subordinated under negative modality (cf. Roberts
  1996).

63

• (51) TP
• MODAL P VP
• MODALbouletic presupposition

64

• Modal also resembles psych-verbs such as surprise
  or affect in the argument structure. Emotive
  modal takes speaker as an experiencer and the
  event as the theme.

65
Speaker is an experiencer argument of psych-verbs
(cf. Belletti and Rizzi 1988)

• (52) FocP
• mo TP
• DP T
• yo ltmogt vP T
• speaker v ta
• v
• VP e
• NP V
• huke
• Mo adjoins to TP via internal merge and
  reprojects into the head of FocP (cf. Hornstein
  and Uriagereka 2002).

66
Conclusion

• The three constructions,
• fake past
• narrative present
• discourse initial mo/too
• discussed in this paper show that covert emotive
  modality interacting with bouletic, epistemic and
  circumstantial modality shifts context parameters
  in simple sentence.
• Surprise, empathy and sentimentality affect
  temporal interpretations and satisfy
  presuppositions.

67
Conclusion

• Implicit speaker attitudes shift context
  parameters.
• Modality shifts temporal interpretations, and
  contexts in order to satisfy presuppositions of a
  focus particle.