Doctoral Student Presentations
Yu-Fu Chien, Robert Fiorentino, Xiao Yang, and Joan Sereno. Poster presented at the 5th Joint Meeting of the Acoustical Society of America and Acoustical Society of Japan: Honolulu, HI.
Katrina Connell, Simone Huls, Maria Teresa Martínez-García, Zhen Qin, Seulgi Shin, Hanbo Yan, and Annie Tremblay. Paper presented at the 8th International Conference on Second-Language Speech (New Sounds 2016): Aarhus, Denmark.
Katrina Connell, Annie Tremblay, and Jie Zhang. Poster presented at the 2016 Tonal Aspects of Language Conference: Buffalo, NY.
Katrina Connell, Simone Huls, Maria Teresa Martínez-García, Zhen Qin, Seulgi Shin, Hanbo Yan, and Annie Tremblay. Poster presented at the 5th Joint meeting of the Acoustical Society of America and the Acoustical Society of Japan: Honolulu, HI.
This study investigates whether the presence of lexical stress in the native language (L1) determines second-language (L2) learners’ ability to use stress in L2 lexical access. It focuses on (standard Mandarin) Chinese and (Seoul) Korean listeners’ (and native English listeners’) use of segmental and suprasegmental cues to stress in English word recognition. Stress placement in English is signaled by segmental (vowelreduction) and suprasegmental (fundamental frequency, duration, and intensity) cues. Chinese has full-full and full-reduced words that differ in stress placement, with segmental and suprasegmental cues signaling stress. By contrast, Korean does not have lexical stress. Participants completed an eye-tracking experiment. They heard stimuli containing a target word with initial stress (parrot), and saw four orthographic words in the display, including the target and one of two competitors (stress match: parish; stress mismatch: parade). The first syllable of the target and stress-mismatch competitor differed in both segmental and suprasegmental information (parrot-parade) or only in suprasegmental information (mystic-mistake). Growth-curve analyses on fixations revealed that only Chinese and English listeners used stress to recognize English words, confirming L1 effects on the use of stress in L2 lexical access. Furthermore, only English listeners made greater use of stress in the presence of vowel reduction.
Lauren Covey and Xiao Yang. Joint presentation given at 2016 KU Cognitive Brain Science Brownbag Talk series: Lawrence, KS.
Lauren Covey, Alison Gabriele, and Robert Fiorentino. Poster presented at 30th Annual CUNY Conference on Human Sentence Processing. MIT.
This study investigates the nature of island constraints by comparing the processing of licit and illicit wh-dependencies using EEG and including cognitive measures targeting processing abilities. An N400 emerged only within licit dependencies, and was correlated with Stroop reaction times; implications for grammatical and processing accounts of islands will be discussed.
Lauren Covey, Robert Fiorentino, and Alison Gabriele. Poster presented at 29th Annual CUNY Conference on Human Sentence Processing. University of Florida.
Recent studies have shown variability in the processing of pronominal reference in both ambiguous contexts (David shot at John as he...vs. David shot at Linda as he) and in contexts of ‘referential failure’ in which there is no gender-matching antecedent within the sentence (The boy thought that she/he....) (Osterhout and Mobley, 1995; Nieuwland and Van Berkum, 2006; Van Berkum and Nieuwland, 2008; Nieuwland, 2014). These contexts have been shown to give rise to a sustained, frontal negative shift (Nref) or a P600 (or both ERP components) depending on both the accompanying behavioral task and individual differences in working memory (e.g. Nieuwland, 2014). For cases of ‘referential failure,’ it has been proposed that the presence of an explicit acceptability judgment task may give rise to P600s, suggesting attempts at co- reference despite the mismatch in the gender of the pronouns (Osterhout and Mobley, 1995; Nieuwland, 2014). In both ambiguous contexts and contexts of ‘referential failure,’ high working memory has been associated with Nref, a component that indexes the inability to assign a unique referent, as opposed to P600 (Nieuwland and Van Beckum, 2006; Nieuwland, 2014. The present study further examines individual differences in pronominal processing, testing both kinds of pronominal contexts, and including a wider range of cognitive measures. Experiment 1 targeted ambiguous pronouns (One/Two referent: Linda/Tyler grabbed Eric because he was falling down the stairs) and Experiment 2 targeted referential failure (One/No Referent: Nicole believe Steven/Alice because he was a very genuine person). Participants included native English speakers (n=33) who completed tests of working memory (counting span, reading span), attentional control (Stroop), and an offline measure assessing the ambiguity of the experimental sentences. In the ERP experiment, participants read 240 sentences (160 targets, 80 fillers) and were asked to respond to fill-in-the-blank recall questions following one third of the stimulus sentences. Results of Experiment 1 for all participants (n=33) showed no significant effect of ambiguity. However, follow-up analyses following Van Berkum and Nieuwland (2008), revealed two distinct groups of participants: Individuals who showed a positivity in the posterior region in the 500-1400 time window (n=18) also showed a significant positivity in the anterior region, revealing a broad positivity which may index revision of the initial reference assignment at the pronoun or analyzing the gender of the pronoun as a grammatical violation. Individuals who did not show a positivity in the posterior region (n=15) showed a sustained frontal negativity in the anterior region (Nref). The size of the Nref was significantly correlated with performance on the counting span task, suggesting in line with previous research that the Nref is larger in individuals with high working memory. Results of Experiment 2 showed a significant positivity in the posterior region in the 500-900 time window, consistent with the P600, suggesting that participants may attempt co-reference even in the absence of an explicit acceptability judgment task. The P600 effect size was significantly positively correlated with the reading span task suggesting that P600 may not necessarily be associated with low working memory (cf. Nieuwland, 2014). Furthermore, additional analyses revealed an emerging trend in which the group differences in Experiment 1 impact the processing of referential failure in Experiment 2: an anterior negativity (Nref) emerges only in the group which yielded an Nref in Experiment 1 suggesting a possible commonality in the processing of pronominal reference across contexts that require a kind of complex inferencing in response to referential difficulty.
Lauren Covey, Nicholas Feroce, Alison Gabriele, and Robert Fiorentino. Poster presented at Georgetown University Round Table on Languages and Linguistics. Georgetown University.
We examine variability in the processing of pronominal ambiguity in English and Spanish using ERP. Results revealed two distinct groups with one showing an Nref, which indexes referential ambiguity, and the other, a broad positivity. The size of the Nref in both languages was positively correlated with working memory.
Philip T. Duncan. Invited talk presented at the Spring 2017 KU Department of Linguistics Colloquy: Lawrence, KS.
The subject of ergativity has drawn the attention of language scholars for decades, and it remains a vibrant domain of linguistic inquiry. One well-known and intriguing correlate of ergativity is that virtually no language isconsistently ergative (Moravcsik 1978; Dixon 1979). Instead, ergative languages commonly showcase context-dependent "splits" where the ergative pattern is lost. Such splits are so pervasive that some scholars prefer to speak of ergative "patterns" or "systems" rather than ergative languages (Coon & Preminger to appear). In this talk, I present evidence from Meꞌphaa, an understudied Otomanguean language from Guerrero, Mexico, whose brand of ergativity is unique on account of its complex patterns of verbal agreement. I maintain that Meꞌphaa is consistently ergative in terms of manifesting “ergativity properties” (Deal 2015), and apply recent insights into the syntax of argument structure to account for how it gets that way.
Philip T. Duncan and Harold Torrence. Invited talk presented at the 91st Annual Meeting of the Linguistic Society of America: Austin, TX.
Philip T. Duncan. Invited talk presented at the Center for Latin American & Caribbean Stduies Fall Meriendas: Lawrence, KS.
Philip T. Duncan. Invited talk presented at the Form and Analysis in Mayan Linguistics 4, Saki' (Valladolid): Yucatán, México.
Philip T. Duncan. Paper presented at the Proceedings of the 42nd Annual Meeting of the Berkeley Linguistics Society: Berkeley, CA.
Philip T. Duncan, Travis Major, and Mfon Udoinyang. Paper presented at the Proceedings of the 47th Annual Conference on African Linguistics. Berkeley, CA.
Masashi Harada. Poster presented at the Buckeye East Asian Linguistics Forum 2: Columbus, OH.
Masashi Harada. Poster presented at The 12th Workshop on Altaic Formal Linguistics: New Britain, CT.
Masashi Harada. Poster presented at the Linguistic Society of America 2016 Annual Meetings: Washington D.C..
Although Japanese sluicing-like constructions (SLC) are generally analyzed as deletion of the presupposition CP (e.g. Kizu 1997) unlike English ‘genuine’ sluicing (GS), i.e. TP deletion following wh-movement, Takita (2009) claims that Japanese also has GS when the maximal verb taking the remnant phrase is a control verb. In this paper, however, I argue against his supporting evidence, and then propose a Small Clause-based analysis, which is not only free from the problems his analysis suffers but can also explain why ni sure ‘do,’ which is semantically vacuous just like the copula da, can appear instead of da in such SLC.
Mingxing Li. Oral talk presented at the Spring 2017 Kansas Linguistics Colloquy Series: Lawrence, KS.
Studies in language typology and sound changes have shown that languages generally prefer more distinct contrasts and that vowel contexts influence the perceptual distinctiveness of consonant contrasts. In a two-part study, I focus on the influence of vowel contexts on the perceptual distinction of dental vs. palatal sibilants. In the first part, a survey across 170 Chinese dialects showed a tendency to avoid contrastive dental vs. palatal sibilants, in the [_i] context and a perceptual experiment confirmed that dental vs. palatal sibilants are perceptually less distinct in the [_i] context than in other vowel contexts. In the second part, an examination of the historical sound changes in Mandarin showed that contrastive dental vs. palatal sibilants in the [_i] context have been avoided across different stages and a phonetic study of Xiangtan Chinese supported the contention that such contrasts were unstable and tend to be avoided in a sound system.
Mingxing Li. Poster presented at The 91st Annul Meeting of the Linguistic Society of America: Austin, TX.
Crosslinguistically, sibilant place contrasts before [_i] are often avoided. In this paper, we examine the realization of /si tsi tsʰi/ in Xiangtan Chinese, which are reported to contrast with /ɕi tɕi tɕʰi/. Eleven speakers’ production of the sibilants was recorded. In a discriminant analysis, we first trained a classifier to distinguish dental and palatal sibilants using acoustic measurements of canonical dentals and palatals, then applied the classifier to the onsets of /si tsi tsʰi/. The results showed both within- and across-speaker variation of /s ts tsʰ/, which were sometimes realized as palatals, indicating a merger in progress.
Kotoko Nakata. Talk presented at The 2017 Annual Spring Conference of American Association of Japanese Teacher: Toronto, ON.
Mimetics are commonly used by Japanese native speakers to express the manner of actions and sensations. However, they are often not taught explicitly in many Japanese language classrooms. The current study will test a novel teaching methodology to help English-speaking learners of Japanese learn Japanese mimetics. Learners will be explicitly taught three phonological/morphological rules during learning. The three rules are: (i) voicing, (ii) gemination, and (iii) reduplication. In Japanese mimetics, these phonological/morphological factors systematically affect the meaning of mimetics. Mimetics with voiced sounds often express largeness, heaviness, roughness, and aggressiveness, whereas mimetics with voiceless sounds often express smallness, lightness, smoothness, and quickness. Gemination in mimetics often indicates that the movement/action is quick and instantaneous. Reduplication usually indicates that the movement/action is repeated consecutively. The current study examines whether explicitly teaching these three rules helps English-speaking learners of Japanese who vary in Japanese proficiency acquire mimetics as well as helps them generalize these rules to newly encountered mimetics. The procedure uses a Pretest-Learning-Posttest design. The Experimental group will explicitly learn the three phonological/morphological rules while the Control group will not. In the Learning Session, all participants will be taught 32 mimetic words with a verbal description (a static picture along with a sentence that contains the mimetic word). All learners will then participate in a Posttest and a Delayed posttest (one month later) to assess their retention of the mimetic vocabulary. The current study will assess (i) whether this novel teaching methodology can aid in learning mimetics using verbal descriptions and, (ii) whether learners’ proficiency affects learning mimetics, and (iii) most importantly, whether explicitly knowing the three phonological/morphological rules plays a role in acquiring mimetics as well as retaining the lexical knowledge.
Kotoko Nakata. Poster presented at the 2nd Buckeye East Asian Lingusitcs Forum: Columbus, OH.
Zhen Qin, Jie Zhang, and Annie Tremblay. Paper presented at the 173rd meeting Acoustical Society of America and 8th Forum Acusticum: Boston, MA.
This study investigates whether within-category gradience in lexical tones influences native and non-native Chinese listeners’ word recognition. Previous offline research found that Chinese listeners have a more categori- cal perception of lexical tones, and thus show less sensitivity to within-cate- gory variability in tones, than non-native listeners. However, it is unclear whether native and non-native listeners have sensitivity to within-category gradience during online word recognition. Native Chinese listeners and pro- ficient adult English-speaking Chinese learners were tested in a visual-world eye-tracking experiment. The target was a level tone and the competitor was a high-rising tone, or vice versa. The auditory stimuli were manipulated such that the target tone was either canonical in the standard condition, pho- netically more distant from the competitor in the distant condition, or pho- netically closer to the competitor in the close condition. Growth curve analysis on fixations suggested that native listeners showed a gradient pat- tern of lexical competition, with less competition in the distant condition and more competition in the close condition than in the standard condition; learners, on the other hand, showed increased competition in both the distant and close conditions than in the standard condition. The native and non- native listeners’ difference suggested the influence of their language backgrounds.
Zhen Qin, Allard Jongman, and Joan Sereno. Paper presented at the 5th Joint meeting of the Acoustical Society of America and the Acoustical Society of Japan: Honolulu, HI.
Previous studies suggest that Chinese listeners may be more sensitive to pitch direction while American listeners primarily attend to pitch height. The present study sought to establish JNDs for pitch discrimination using a three-interval, forced-choice procedure with a two-down, one-up staircase design (cf. Liu, JASA 2013). We manipulated a high rising and a high falling Mandarin tone produced by a female speaker. Specifically, for pitch height, a standard contour rising from 200 Hz at tone onset to 230 Hz at tone offset was paired with deviants that were parallel to the standard but started (and ended) at lower and higher f0 values (onsets ranging from 185 to 215 Hz in 1 Hz steps). Likewise, a standard contour falling from 200 Hz at tone onset to 170 Hz at tone offset was paired with parallel deviants whose onsets ranged from 215 to 185 Hz. For pitch direction, standard rising and falling contours were paired with deviants with either steeper or shallower slopes, again varying in 1 Hz steps. Results indicate that, overall, Chinese listeners are more sensitive to pitchdirection and American listeners to pitch height. However, these effects are modulated by both the direction (falling/rising) and steepness of the contours.
Zhen Qin. Invited talk presented at the KU Linguistics Colloquy and Child Language Proseminar: Lawrence, KS.
Zhen Qin. Invited talk presented at the Lab meeting of Center for Brain and Cognitive Science: Peking University, Beijing.
Languages use different types of information to differentiate words. For example, Mandarin Chinese uses lexical tones to contrast word meanings whereas English does not. These cross-linguistic differences may pose word recognition difficulties for second language (L2) learners. The project aims to examine how native speakers and L2 learners of Chinese use tonal information in the recognition of Chinese words as the speech signal unfolds. Specially, two visual-world eye-tracking experiments in Chinese were conducted with two groups of participants: 36 native Chinese listeners and 36 English-speaking L2 learners of Chinese, to test the following questions: whether or not native and non-native listeners make similar use of early pitch height as well as fine-grained within-category pitch information in the recognition of words that differ in lexical tones, and how their use of tonal information changes at different points of the pitch contour (i.e., as the pitch contour is gradually heard). Preliminary results showed that the time course of use of tonal information differed for native and non-native Chinese listeners, with L2 learners potentially showing earlier use of and more sensitivity to fine-grained pitch height information than native listeners but ultimately having more difficulty in integrating this information in the word recognition processes. The findings help elucidate how tonal information can be incorporated into models of native and L2 word recognition, and have important implications for the teaching of lexical tones in the classroom.
Charles Redmon, Keith Leung, Yue Wang, Allard Jongman, and Joan Sereno. Poster presented at the 5th Joint Meeting of the ASA and ASJ: Honolulu, HI.
Studies on acoustic and visual characteristics of English tense and lax vowels show consistent enhancement of tensity contrasts in clear speech. However, the degree to which listeners utilize these enhancements in speech perception remains unclear. The present study addresses this issue by testing speech style effects on tense and lax vowel perception by 23 native English and 30 non-native Mandarin-Chinese listeners in audio-only (AO), audio-visual (AV), and visual-only (VO) stimulus modes. English and Chinese listeners showed similar relative differences in performance by mode (VO < AO < AV) and style (plain < clear). However, the two groups differed in the nature of the interaction between tensity, style, and stimulus mode. English listeners showed advantages for clear speech for both tense and lax vowels in all but VO stimuli, whereas Chinese listeners showed a clear speech advantage only for tense vowels, while clear lax vowels showed no improvement in AO and reduced accuracy in AV and VO. While temporal and spectral acoustic cues may coordinate to preserve or improve tense-lax category identity in clear speech, non-native listeners may not be attending to both dimensions. Further, Chinese listeners' greater reliance on visual information may account for their less accurate lax vowel identification.
Charles Redmon and Phongshak Phom. Poster presented at the 22nd Himalayan Languages Symposium: Guwahati, India.
Corpus measurements of tonal ambiguity in written Phom as a function of word size, N-gram context, and frequency are reported. A significant inverse relation between the number of syllables in a word and the tonal contrast size was found, while frequency of the preceding context was shown to be positively correlated with the size of the contrast comprising the ambiguity. N-gram context was also inversely related to the probability of disambiguation, but measures of relative entropy revealed this effect to be nonlinear; i.e. gains in information at the bigram were significantly greater than further gains in the trigram context.
Charles Redmon. Poster presented at the 171st Meeting of the Acoustical Society of America: Salt Lake City, UT
Previous work on the acoustic discrimination of fricativecontrasts has commonly studied categories which are not held to vary phonemically; instead the focus is placed on accounting for variation in the salience and stability of the acoustic cues mapping onto those categories (cf. Forrest et al., 1988; Jongman et al., 2000; McMurray and Jongman, 2011). The present study, like Jannedy et al. (2015), addresses a set of contrasts—the voiceless posterior obstruents /x, h, kh/ in Assamese—where /h/-voicing and /kh/-lenition processes interact with the discriminability of the velar fricativedifferentially according to positional (CV, VCV, and VC) and speaker (gender, dialect) variables. Results of principal components logistic regression (PCLR) models applied at both fixed and cumulative time windows over consonant and transition intervals are presented. Overall, /x/-/h/ discrimination was significantly more accurate in intervocalic position (96%) relative to word-initial position (84%). Data from intervocalic productions further suggest a significant dialectal difference in the /x/-/kh/ contrast, with discrimination poorer in Nalbari (68%) than in Jorhat (86%) data, in line with Sarma’s (2012) analysis. Implications for the integration of lexical information by means of phonotactic priors derived from a corpus (Baker et al., 2002) are also discussed.
Charles Redmon and Allard Jongman. Poster presented at the 5th Joint Meeting of the ASA and ASJ: Honolulu, HI.
Present models of voiceless fricative acoustics assume excitation of the vocal cavity anterior to a narrow constriction by an aperiodic source generated from turbulence within that constriction (or in the case of dipole sources from interaction of the jet with an obstacle downstream). This model has been applied successfully to anterior places of articulation (labial to palatal) but for dorsal fricatives, i.e. velar /x/ and uvular /χ/, there is evidence to suggest the assumption of a line source may not hold. Rather, preliminary data from our lab and from studies such as Zeroual (ICPhS, 2003) and Shosted & Chikovani (JIPA, 2006) indicates intermittent presence of a mixed source due to passive vibration of the uvula, yet no study to date has directly addressed this element of the sound source in dorsal fricative production. Acoustic and aerodynamic data on dorsal fricatives produced by 4 speakers each of Arabic, Persian, and Spanish are presented to validate the nature of the dorsal fricative source component and the degree of cross-linguistic variation in production of these sounds. Measures of periodicity duration and cycle amplitude/shape are employed to estimate the oscillatory pattern of the uvula and model its contribution to the radiated acoustic signal.
Xiao Yang, Robert Fiorentino, and Utako Minai. Invited presentation at the Child Language Proseminar: Lawrence, KS.
Xiao Yang, Robert Fiorentino, and Utako Minai. Poster presented at the 30th Annual CUNY Conference on Human Sentence Processing: Cambridge, MA.
Xiao Yang. Presentation given at 2016 KU Graduate Research Competition: Lawrence, KS.