Academia.edu no longer supports Internet Explorer.

To browse Academia.edu and the wider internet faster and more securely, please take a few seconds to upgrade your browser.

Log In
Sign Up

Figure 7 – uploaded by Angelika Kiss

See full PDF downloadDownload figure

Figure 2.7: The mean pitch contour of an information-seeking question. — Figure 2 7: The mean pitch contour of an information-seeking question.

Related Figures (54)

Table 2.1: A speaker’s possible epistemic stances on a and scenarios with and without encountering contextual evidence. are listed in Table 2.1. then S may felicitously use an incredulous declarative question. For example, if S, who believes that it is not raining, is told by someone who is likely to be joking that is raining, they may choose to not accept the fact that it is raining. If S judges the contextual evidence for @ stronger than the reason to keep @ in Doxs, then S may felicitously use a surprise declarative question. An example of this would be the case when S, who falsely believes that it is not raining, steps out and sees the rain for herself. The above-described potential scenarios, together with the matching declarative question types. The above-described potential scenarios, together with the matching declarative question types.

Figure 2.1: The locus of the falling intonational contour (IC-1) in declarative clauses with basic word order conveying assertions (Munteanu and Kiss, in prep). The two contours are represented schematically in Figure 2.1 and Figure 2.2.

Figure 2.2: The locus of the rise-fall intonational contour (IC-3) in polar interrogative clauses with a basic word order conveying information-seeking questions (Munteanu and Kiss, in prep).

_ — “ ea a“ Based on Bryzgunova’s (1977) intonational system, there are two intonational parameters to con sider in Russian utterances: the shape of the contour and its alignment within the utterance. Munteanu and Kiss (in prep) argue that Russian declarative questions exhibit a mix of the properties of Bryz gunova’s (1977) IC-1, a low tune aligned to the rightmost stressed syllable in the utterance that marks assertions, and IC-3, a rise-fall contour aligned to the stressed syllable of the verb. Specifically, the declarative component is provided by the alignment of the peak, in that the peak appears in the posi tion where IC-1 would appear in a declarative sentence, which is the rightmost stressed syllable; and the interrogative component is contributed by the shape of the contour, exhibiting IC-3. the interrogative component is contributed by the shape of the contour, exhibiting IC-3. In other words, based on Bryzgunova’s descriptive analysis, Munteanu and Kiss (in prep) hypoth- esize that the declarative-and interrogative-like properties of Russian declarative questions consist in having the question-like intonational contour aligned to the position of the intonational contour of an unmarked declarative clause. esize that the declarative-and interrogative-like properties of Russian declarative questions consist in

Table 2.4: Example of a context, interlocutor utterance and target sentence of an item in the assertion condition. and the R code are available at https://osf.io/cdszb/. producing the target sentence in response to the interlocutor utterance. The stimuli, the measurements

Figure 2.4: Boxplot of the target sentence duration by sentence type (abbreviations: ISQ = information- seeking questions, CDQ = confirmative declarative questions, SDQ = surprise declarative questions, A = assertions) (Munteanu and Kiss, in prep).

Figure 2.5: Average FO curves for each of the four sentence types. FO, on the y-axis, was normalized by speaker. Time, on the x-axis, was normalized such that each of the three words was represented by 20 measurements (Munteanu and Kiss, in prep).

‘igure 2.6: Utterances plotted by PC1 on the x-axis and PC4 on the y-axis. Intended sentence type is ndicated by color. Each utterance is a single point. The labels correspond to by-sentence means (as in ‘able 2.4). Ellipses are drawn around each sentence type region with a .95 confidence interval using he stat_ellipse function in R (Munteanu and Kiss, in prep).

surprise questions, the speaker signals that this is a new and unexpected commitment. The end result in both cases is the same, the speaker makes it public that they now believe that @ is true. Accordingly, the authors conclude that a commitment to the highlighted alternative is indicated by a high pitch accent on the subject in Russian SVO sentences. This is summarized in Table 2.5. Table 2.5: Form-meaning relations in Russian declarative questions based on (Munteanu and Kiss, in prep). Abbreviations: ISQ = information-seeking question, DQ = declarative question accent on the subject in Russian SVO sentences. This is summarized in Table 2.5.

Figure 2.8: The mean pitch contour of a confirmative declartive question.

Figure 2.9: The mean pitch contour of a surprise declarative question.

Figure 2.10: The mean pitch contour of an assertion.

Figure 2.15: The five conditions of the target strings (szotag = ‘syllable’, konturok = ‘contours’) in Kiss (2022) The fundamental frequency of the fourth and fifth syllable is the same in each case, only the values preceding the vowel of the fourth syllable are different. The first and second contour are both somewhat rising, while the third one is a plateau contour, however, they all start at 300Hz which is where they arrive in the fourth syllable (conforming to H-), and therefore they all can be transcribed as HxH-L%. Contours 4 and 5, on the other hand, start out at 180Hz, which is the pitch height of the target of the

Table 2.6: Form-meaning relations in Hungarian declarative questions following Kiss (2022) Kiss (2022) found that in both experiments, participants were significantly less likely to choose the ‘Surprise’ option when the target stimulus involved a low rise (L*H-L%, contours 4-5), which confirms Varga’s claim that end-falling contours that do not start with a rise (which he calls ‘monotone-fall’ and ‘descent-fall’ contours) express surprise and/or exclamation. Speakers associate thus two different sub- types of declarative questions with different intonation. Contextual confirmative declarative questions were associated with L*H-L%, and surprise declarative questions, with HxH-L%. Based on these find- ings, certain key components of the character contours are tied to meaning components of declarative questions (and information-seeking questions and assertions) as shown in Table 2.6. questions (and information-seeking questions and assertions) as shown in Table 2.6. The fact that confirmative and surprise declarative questions were associated with different into- types of declarative questions with different intonation. Contextual confirmative declarative questions

Figure 2.17: The (trivial) special discourse effects of assertions expressed by the range of possible values for the credence levels in the truth of the highlighted alternative.

Figure 2.19: The special discourse effects of contextual confirmative declarative questions expressed by the range of possible values for the credence levels in the truth of the highlighted alternative. Since a contextual confirmative declarative question signals a lack of commitment to a, the sul ective probability of it being true is the same as the probability of it being false. I represent a neutré stance as the full range of credence levels. Contextual evidence points to the truth of a, its biasin orce is represented by excluding the neutral value (0.5) and including the extreme value of 1, whic ‘epresents a full commitment. However, the speaker does not find e compelling but simply biasing: - ndicates that given Doxs, a is considered to be more likely to be true than @, and so it does not includ he value of 0.5, but it does also not include the value of 1. As for the speaker’s resulting commitmen t is non-neutral, because it is biased towards a, but it also does not express full commitment, hence it i ‘epresented as (0.5, 1), a subinterval from which 0.5 and 1 are excluded. Note that the particular value of credence levels associated with the contextual evidence and the resulting commitment pertaining t 1 given speaker S in a given discourse d are always some subintervals of the intervals represented i ‘igure 2.19. Thus the speaker may express different levels of bias: a slight bias could be represented a 0.5, 0.6], while a strong bias could be [0.9, 1), etc.

Figure 2.20: The special discourse effects of epistemic confirmative declarative questions expressed by the range of possible values for the credence levels in the truth of the highlighted alternative. Figure 2.20 shows that the speaker’s initial commitment is expressed by an interval ranging from

Figure 2.21: The special discourse effects of incredulous declarative questions expressed by the range of possible credence levels in the truth of the highlighted alternative. commitment, that is, at least at first glance, they seem to be in an accessibility relation. Figure 2.21 shows that the initial commitment of the speaker ranges from 0 to below 0.5, while the contextual evidence expresses an interval ranging from above 0.5 to 1. The extreme value of 1 is included here because e has the potential to trigger commitment in the discourse participants. How- ever, in the case of incredulous declarative questions, this commitment is not made. As mentioned before, whether a piece of evidence counts as compelling or not depends on the speaker, on how reli- able they consider the source of evidence. The speaker’s resulting commitment is the same as the initial commitment, that is, at least at first glance, they seem to be in an accessibility relation.

Figure 2.22: The special discourse effects of surprise declarative questions expressed by the range of possible values for credence levels in the truth of the highlighted alternative. Recall that the diagnostic tests in (33) revealed that a surprise declarative question can be felici- tously replaced both by questions and by assertions, both of which can convey surprise. This property is due to the resulting commitment expressed by this type of declarative question, full commitment to a, which is also what assertions express by their basic conventional discourse effects, see (82).

ranges from above 0.5 to 1. The resulting commitment to a is a reduction of the initial commitment set The fact that reiterative declarative questions, but not surprise declarative questions, express a reduction of the initial commitment set, captures the difference in their meaning, namely that reiterative declarative questions are not surprising, despite being mirative utterances. Expressing that the resulting commitment is unexpected is what surprise declarative questions do, which is captured by the fact that the resulting commitment came about by a non-monotonic belief revision.

The special discourse effects represented as probabilities are given in Table 2.7. it, the result is thus an accessible resulting commitment set. Before concluding this section, I return to the prosodic experiments on Russian and Hungariar 2° Additionally, the belief revision of such declarative questions may happen in the speaker’s beliefs about the addressee’s beliefs (i.e., the speaker’s version of the common ground). This form of revision is not modeled here as it only models revision regarding a itself. \s such, they signal a non-monotonic update on the speaker’s commitment set.

declarative questions. In light of the current proposal, the following form-meaning relations can bi established based on Tables 2.5 and 2.6. Table 2.8: Form-meaning relations in Russian declarative questions based on (Munteanu and Kiss, in prep). Abbreviations: ISQ = information-seeking question, DQ = declarative question, o = stressed syllable, V = verb. declarative questions. In light of the current proposal, the following form-meaning relations can be

As for Hungarian declarative questions, the following form-meaning relations can be proposed: Questionhood or inquisitiveness is signaled by the high intermediate phrase H-, representing the penultimate high tone syllable in questioning utterances. The presence of non-trivial contextual evi- dence is signaled by the multiple realization of the given question contour (whether it is L*H-L% or H*H-L%), and the H« tone seems to correlate with commitment. no) evidence, and utterances low in PC1 signal non-trivial evidence. And PC4 is now taken to signal penultimate high tone syllable in questioning utterances. The presence of non-trivial contextual evi-

of the highlighted alternative to the TaBLE (Farkas and Roelofsen, 2017).

Figure 3.4: Two ways of representing the meaning of Who sings?

Table 3.1: Summary of the pitch and duration results for information-seeking questions (ISQ), empty set rhetorical questions (RQ—) and non-empty set rhetorical questions (RQ+) (Lo and Kiss, 2020) set rhetorical questions was in between the two. There were two areas in the utterance that showed a distinction in terms of pitch. The pitch (measured as FO) of the wh-word and the sentence-final parti cles again showed a three-way distinction. Wh-words in information-seeking questions had the lowest pitch, and the wh-words of empty set rhetorical questions had the highest pitch, with non-empty set rhetorical questions in between the two. As for the pitch of the sentence-final particles, the highest pitch was found in information-seeking questions, and the lowest, in empty set rhetorical questions with non-empty set rhetorical questions in between. Lastly, the duration of the sentence-final particle also showed a three-way distinction. Namely, sentence-final particles in information-seeking ques tions had the longest duration, and they were shortest in empty set rhetorical questions. Non-empty set rhetorical questions were yet again in between the two. A brief summary is shown in Table 3.1, and the mean curves for each reading are visualized in Figure 3.6. the mean curves for each reading are visualized in Figure 3.6.

Figure 3.5: Proportion of creaky sentence-final particles (SFPs) across the three question types: information-seeking questions (ISQ), non-empty set rhetorical questions (RQ+) and empty set rhetori- cal questions (RQ—) (Lo and Kiss, 2020).

Table 3.2: The pitch and duration of the sentence-final particles aa! and aa’ in Lo et al. (2019) rising tone, see the summary below.

Kiss et al. (to appear) tested whether a three-way contrast among the three question types woulc also be found in a perception experiment. Participants were presented utterances of wh-interrogative: with a prosodically manipulated sentence-final particle. The utterances that formed the basis of the stimuli were produced by a female native speaker of Cantonese (also a linguist), who pronounced then in the information-seeking question reading. The interrogatives had the sentence-final particle aa3 which got manipulated orthogonally by the duration and pitch values found in Lo et al. (2019): with < duration of 250 ms, 310 ms, 370 ms, 430 ms, 490 ms, and 550 ms, and with a pitch contour realized wit a rise of 0 Hz, 25 Hz, 50 Hz, or 75 Hz. See some of the combinations in Figure 3.7. Figure 3.7: Schematic representation of some manipulations (Kiss and Lo, 2021) a rise of 0 Hz, 25 Hz, 50 Hz, or 75 Hz. See some of the combinations in Figure 3.7.

Figure 3.8: Response proportions of information-seeking questions (ISQ), empty set rhetorical questions (RQ—) and non-empty set rhetorical questions (RQ+) as a function of manipulated duration (Kiss et al., to appear) non-empty set rhetorical questions were not clearly associated with any pattern, see Figure 3.8.

Figure 3.9: Response proportions of information-seeking questions (ISQ), empty set rhetorical questions (RQ-—), and non-empty set rhetorical questions (RQ+) as a function of manipulated pitch rise (Kiss et al., to appear)

Figure 4.1: Tsai’s (2008) proposal for causal/denial how

Figure 4.2: Pak’s (2017) analysis of propositional how

Figure 4.3: The base-position and landing site of a NwH-word as proposed by (Cheung, 2008, p. 39)

Figure 4.4: The post-subject position of a NwH-word

12 participants were recorded in this pilot session, everyone produced 8 strings of since when- questions under two readings: as information-seeking questions and as adversative questions. The target sentences had the same structure, they all consisted of midta ‘how long’/‘since when’, and either a lexical verb and an argument, or a copula and an adjective. No phonological properties were controlled for, as the purpose of the recordings was to gather impressions on which part of the utterance is marked. Some example contours are shown in Figure 4.5. for, as the purpose of the recordings was to gather impressions on which part of the utterance is marked. What can be observed overall is that participants produced the question word midta ‘since when’

Table 4.1: Summary of wh-expressions in adversative questions based on the elicitation data described in the Appendix. (Abbreviations: NAm = North American, N = Northern, S = Southern, SW = South- western.)

Figure 4.6: The clause type marking of an adversative question explains why it is possible that such questions contribute a discourse referent just like other expressions with a sentence radical (falling declaratives, declarative questions, and polar questions) and unlike wh- interrogatives, which do not have a highlighted alternative but merely a highlighted property. The fact that the response particles (yes and no) cannot refer back to the discourse referent adversative questions contribute is arguably due to the fact that such particles may require a highlighted alternative which adversative questions do not have. explains why it is possible that such questions contribute a discourse referent just like other expressions

Table 4.2: Summary of wh-phrases in Hungarian adversative questions (W acceptable, “cond acceptable in the conditional mood only, “mod preferred with a modal, (W) acceptable with certain predicates only). A summary of the question words available in Hungarian adversative questions is shown in Table I ne An example of a summary table of wh-words in adversative questions is shown in Table B.: 1 Hungarian, an adversative question with hol ‘where’ can only express incredulity, while question ith midta ‘since when’ and hogy ‘how’ can express a challenge or surprise in addition to increduli lote that in Hungarian, this question word requires conditional mood or modal verbs, and few of tl inguages surveyed here did not require modals for ‘how’-questions. The question word hogyhos low come’ is reserved for surprise questions. The question word mit ‘what’ is used in what-as-wl uestions and share with other adversative questions that there must be a contextually salient event ( 1at licenses such a question. The adversativity of incredulous, challenging and surprise adversatir uestions arises from the contrast between the initial commitment and the piece of contextual eviden: ); in the case of what-as-why questions, an additional source of adversativity is the contrast betwee (or the resulting commitment) and the speaker’s wish of how things should be. A summary of the question words available in Hungarian adversative questions is shown in Tab

The summary of wh-phrases in Yoruba adversative questions is shown in Table B.2

Table B.5: Summary of wh-phrases in Mauritian Creole adversative questions (acceptable, “mod ac- ceptable with a modal verb) The summary of wh-phrases in Mauritian Creole adversative questions is given in Table B.5.

Table B.12: Summary of wh-phrases in Ukranian adversative questions (% acceptable by some speakers, “acceptable, /(mod) acceptable, status of modal verb unclear) The summary of wh-phrases in Ukranian adversative questions is shown in Table B.12.

The summary of wh-phrases in Serbian adversative questions is shown in Table B.13

Table B.14: Summary of wh-phrases in Dutch adversative questions (W acceptable) The summary of wh-phrases in Dutch adversative questions is shown in Table B.14

The summary of wh-phrases in Mexican Spanish adversative questions is shown in Table ??. Table B.20: Summary of wh-phrases in Mexican Spanish adversative questions (W acceptable, (W) ac- ceptable with predicates in a certain tense, @ attested but not used by the consultant)

The summary of wh-phrases in Hungarian adversative questions is shown in B.21. Table B.21: Summary of wh-phrases in Hungarian adversative questions (Macceptable, /cond accept- able in the conditional mood only, “mod preferred with a modal, (W) acceptable with certain predicates only)

The summary of wh-phrases in Turkish adversative questions is shown in Table B.22 Table B.22: Summary of wh-phrases in Turkish adversative questions (acceptable)

Table B.25: Summary of wh-phrases in Japanese adversative questions (W acceptable, “cond acceptable in the conditional mood) The summary of wh-phrases in Japanese adversative questions is shown in Table B.25

The summary of wh-phrases in Shanghainese adversative questions is shown in Table B.28.

Table B.33: Summary of wh-phrases in Taishanese adversative questions (Wacceptable, (W) acceptable with certain predicates, (mod) status of modal verb unclear) The summary of wh-phrases in Taishanese adversative questions is shown in Table B.33

Related topics:

Speech Prosody Pragmatics Semantics Formal Semantics/Pragmatics Formal Semantics Kurdish Language Russian linguistics Hungarian language Belief Revision Mandarin Chinese

Connect with 287M+ leading minds in your field

Discover breakthrough research and expand your academic network

Explore
Papers
Topics

Features
Mentions
Analytics
PDF Packages
Advanced Search
Search Alerts

Journals
Academia.edu Journals
My submissions
Reviewer Hub
Why publish with us
Testimonials

Company
About
Careers
Press
Help Center
Terms
Privacy
Copyright
Content Policy

580 California St., Suite 400

San Francisco, CA, 94104

© 2025 Academia. All rights reserved