Perceptual Cue Weighting Matters in Real-Time Integration of Acoustic Information during Spoken Word Recognition
Saved in:
| Title: | Perceptual Cue Weighting Matters in Real-Time Integration of Acoustic Information during Spoken Word Recognition |
|---|---|
| Language: | English |
| Authors: | Hyoju Kim (ORCID |
| Source: | Cognitive Science. 2024 48(12). |
| Availability: | Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us |
| Peer Reviewed: | Y |
| Page Count: | 38 |
| Publication Date: | 2024 |
| Document Type: | Journal Articles Reports - Research |
| Descriptors: | Foreign Countries, Asynchronous Communication, Cues, Auditory Stimuli, Phonetics, Language Rhythm, Distinctive Features (Language), Time Perspective, Acoustics, Word Recognition, Association (Psychology), Listening, Individual Development |
| Geographic Terms: | South Korea (Seoul) |
| DOI: | 10.1111/cogs.70026 |
| ISSN: | 0364-0213 1551-6709 |
| Abstract: | This study investigates whether listeners' cue weighting predicts their real-time use of asynchronous acoustic information in spoken word recognition at both group and individual levels. By focusing on the time course of cue integration, we seek to distinguish between two theoretical views: the "associated" view (cue weighting is linked to cue integration strategy) and the "independent" view (no such relationship). The current study examines Seoul Korean listeners' (n = 62) weighting of voice onset time (VOT, available earlier in time) and onset fundamental frequency of the following vowel (F0, available later in time) when perceiving Korean stop contrasts (Experiment 1: cue-weighting perception task) and the timing of VOT integration when recognizing Korean words that begin with a stop (Experiment 2: visual-world eye-tracking task). The group-level results reveal that the timing of the early cue (VOT) integration is delayed when the later cue (F0) serves as the primary cue to process the stop contrast, supporting a relationship between cue weighting and the timing of cue integration (the associated view). At the individual level, listeners with greater reliance on F0 than VOT exhibited a further delayed integration of VOT. These findings suggest that the real-time processing of asynchronously occurring acoustic cues for lexical activation is modulated by the weight that listeners assign to those cues, providing evidence for the associated view of cue integration. This study offers insights into the mechanisms of cue integration and spoken word recognition, and they shed light on variability in cue integration strategies among listeners. |
| Abstractor: | As Provided |
| Entry Date: | 2024 |
| Accession Number: | EJ1454985 |
| Database: | ERIC |
|
Full text is not displayed to guests.
Login for full access.
|
|
| FullText | Links: – Type: pdflink Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwG1FnAvMWlWQ8l5mL5GZgPfAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDPS9dJZEUF3fcKlL_AIBEICBm-qor5UuIgBl05YmZQpdQJzZENKRx5q9KRJnXFpaOWm8b85TRH0UPobL89KBaAQyuUYXwohWwolnI-ODrx4VLx5PPQihGYooJmDfZk_Kjod8OgcqQCv6P4EbVWVXEcjzvoR-24Tl5b1X8XPH6qKUtlyFLNO8q-r32v6Mi9_eRHSy7utfU2fnDMSvEQeJIK9Zxw6DjdSef6spRfpb Text: Availability: 1 Value: <anid>AN0181921441;cgn01dec.24;2024Dec31.04:29;v2.2.500</anid> <title id="AN0181921441-1">Perceptual Cue Weighting Matters in Real‐Time Integration of Acoustic Information During Spoken Word Recognition </title> <p>This study investigates whether listeners' cue weighting predicts their real‐time use of asynchronous acoustic information in spoken word recognition at both group and individual levels. By focusing on the time course of cue integration, we seek to distinguish between two theoretical views: the associated view (cue weighting is linked to cue integration strategy) and the independent view (no such relationship). The current study examines Seoul Korean listeners' (n = 62) weighting of voice onset time (VOT, available earlier in time) and onset fundamental frequency of the following vowel (F0, available later in time) when perceiving Korean stop contrasts (Experiment 1: cue‐weighting perception task) and the timing of VOT integration when recognizing Korean words that begin with a stop (Experiment 2: visual‐world eye‐tracking task). The group‐level results reveal that the timing of the early cue (VOT) integration is delayed when the later cue (F0) serves as the primary cue to process the stop contrast, supporting a relationship between cue weighting and the timing of cue integration (the associated view). At the individual level, listeners with greater reliance on F0 than VOT exhibited a further delayed integration of VOT. These findings suggest that the real‐time processing of asynchronously occurring acoustic cues for lexical activation is modulated by the weight that listeners assign to those cues, providing evidence for the associated view of cue integration. This study offers insights into the mechanisms of cue integration and spoken word recognition, and they shed light on variability in cue integration strategies among listeners.</p> <p>Keywords: Cue weighting; Cue integration; Visual World Paradigm; Spoken word recognition; Korean stop contrast</p> <hd id="AN0181921441-2">Introduction</hd> <p>When perceiving speech sounds, listeners map continuous and variable acoustic information onto sound categories. This categorization process and the phonological contrasts that result from it are not signaled by a single acoustic dimension but by multiple dimensions. For example, the English stop voicing contrast in word‐initial position can be signaled by a variety of acoustic cues, including voice onset time (VOT) (e.g., Liberman, Harris, Kinney, &amp; Lane, [<reflink idref="bib41" id="ref1">41</reflink>]), fundamental frequency (F0) at the onset of the following vowel (e.g., Haggard, Ambler, &amp; Callow, [<reflink idref="bib20" id="ref2">20</reflink>]), and the duration of the following vowel (e.g., Allen &amp; Miller, [<reflink idref="bib1" id="ref3">1</reflink>]; Summerfield, [<reflink idref="bib66" id="ref4">66</reflink>]). The acoustic cues involved in categorization, however, do not have equal perceptual importance in speech processing: While some acoustic cues contribute more reliably to signaling phonological contrasts (primary cues), other cues are more variable and play a lesser role (secondary cues) (e.g., Francis &amp; Nusbaum, [<reflink idref="bib17" id="ref5">17</reflink>]; Francis, Baldwin, &amp; Nusbaum, [<reflink idref="bib15" id="ref6">15</reflink>]; Holt &amp; Lotto, [<reflink idref="bib24" id="ref7">24</reflink>]; Zhang &amp; Francis, [<reflink idref="bib78" id="ref8">78</reflink>]). For instance, English listeners rely more on VOT than on onset F0 when identifying the stop voicing contrast in English (e.g., Francis, Kaganovich, &amp; Driscoll‐Huber, [<reflink idref="bib16" id="ref9">16</reflink>]).</p> <p>Crucially, an important characteristic of spoken language is that information from different cues often arrives <emph>asynchronously</emph> as the speech signal unfolds over time. For example, the time points when each cue signaling the English stop voicing contrast becomes available in the speech signal are different: VOT becomes available from the onset of the stop and accumulates over the VOT interval until the onset of the following vowel is reached; onset F0 is available right at the beginning of the vowel interval, as soon as voicing is heard; and vowel duration becomes available from the onset of the vowel and accumulates over the vowel interval until the end of the vowel is reached. How listeners integrate these asynchronous cues is a question that has only recently begun to receive some attention (Galle, Klein‐Packard, Schreiber, &amp; McMurray, [<reflink idref="bib18" id="ref10">18</reflink>]; McMurray, Clayards, Tanenhaus, &amp; Aslin, [<reflink idref="bib44" id="ref11">44</reflink>]; Ou, Yu, &amp; Xiang, [<reflink idref="bib58" id="ref12">58</reflink>]; Reinisch &amp; Sjerps, [<reflink idref="bib60" id="ref13">60</reflink>]).</p> <p>The most widespread view among various models of spoken word recognition is that information flows in a <emph>continuous cascaded</emph> manner, with each low‐level acoustic cue providing partial evidence for higher‐level units (such as phonemes or words) as soon as it arrives in the speech signal (e.g., McClelland &amp; Elman, [<reflink idref="bib43" id="ref14">43</reflink>]; McMurray et al., [<reflink idref="bib44" id="ref15">44</reflink>]; McMurray, Tanenhaus, &amp; Aslin, [<reflink idref="bib48" id="ref16">48</reflink>]; McQueen &amp; Dilley, [<reflink idref="bib50" id="ref17">50</reflink>]; Toscano &amp; McMurray, [[<reflink idref="bib68" id="ref18">68</reflink>]]; Weber &amp; Scharenborg, [<reflink idref="bib72" id="ref19">72</reflink>]). As a result, multiple lexical candidates are activated and updated as the remaining cues become available over time. Under this view, lexical access would be modulated by individual cues as soon as they become available in the speech signal (e.g., McMurray et al., [<reflink idref="bib44" id="ref20">44</reflink>]).</p> <p>In some contexts, however, it has been argued that the continuous integration of acoustic information may not apply and that the system instead takes a <emph>buffered integration approach</emph> (e.g., Galle et al., [<reflink idref="bib18" id="ref21">18</reflink>]; Gow, [<reflink idref="bib19" id="ref22">19</reflink>]; Oden &amp; Massaro, [<reflink idref="bib57" id="ref23">57</reflink>]; Schreiber &amp; McMurray, [<reflink idref="bib63" id="ref24">63</reflink>]), with listeners storing early low‐level acoustic cues in a temporary memory buffer until the remaining cues have arrived. A final decision on phonetic categories is then made once listeners have integrated all cues that are relevant to phonetic categorization. Under this view, lexical activation is sensitive to a phonological contrast and takes into account all relevant cues to the contrast as they arrive; as a result, access to higher‐level representations is delayed until sufficient acoustic information is available.</p> <p>Most existing empirical findings are in line with a <emph>continuous</emph> view of cue integration (e.g., consonantal contrasts: McMurray et al., [<reflink idref="bib44" id="ref25">44</reflink>]; McMurray et al., [<reflink idref="bib48" id="ref26">48</reflink>], vowel contrasts: Reinisch &amp; Sjerps, [<reflink idref="bib60" id="ref27">60</reflink>]). To illustrate, McMurray et al. ([<reflink idref="bib44" id="ref28">44</reflink>]) examined English listeners' real‐time processing of English voicing (e.g., /p/ vs. /b/) and manner (e.g., /w/ vs. /b/) contrasts that are cued earlier by information at word onset (VOT for voicing contrasts and formant transition cue for manner contrasts) and later by the length of the following vowel. Using the visual‐world eye‐tracking paradigm (VWP), McMurray and colleagues showed that the effect of the early cue (i.e., VOT and formant transition slope) preceded the effect of the later cue (i.e., vowel duration) in both contrasts, suggesting that the system does not appear to wait until both cues are available to make preliminary lexical commitments.</p> <p>Importantly, however, our understanding of the time course of cue integration is mainly limited to phonological contrasts where the earlier cue to the contrast is also the primary cue to the contrast. Prior studies have not examined cue integration where the later cue is the primary cue to the contrast, nor have they investigated the relationship between cue weighting and the time course of cue integration. Examining phonological contrasts where the cue available <emph>later</emph> brings greater perceptual weight is crucial, as it can provide evidence on whether the real‐time integration of acoustic information is contingent on the perceptual informativeness of this information.</p> <p>The present study does precisely that: Using VWP and cue‐weighting tasks, we investigate how listeners integrate asynchronous acoustic information as the speech signal unfolds over time and whether the real‐time processing of acoustic cues is associated with the perceptual weight of those cues. If there is an interplay between cue integration and cue weighting, we might expect listeners to make immediate use of an early cue for lexical activation when this cue is primary, and somewhat less immediate when the early cue is secondary. Addressing this hypothesis will advance our understanding of the mechanisms that underlie the <emph>moment‐by‐moment processing</emph> of fine‐grained acoustic information during spoken word recognition.</p> <p>We consider two potential relationships between cue weighting and cue integration. On the one hand, the real‐time processing of information could be associated with the informativeness of perceptual cues (hereafter, the <emph>associated view</emph>). This view predicts that listeners' real‐time integration of early acoustic information should be less continuous over time in lexical activation if the early cue is less informative than the later cue. Alternatively, the real‐time processing could be independent of the perceptual informativeness of acoustic information (hereafter, the <emph>independent view</emph>). Under this view, listeners should immediately start integrating <emph>any</emph> acoustic information as soon as it is available in the speech signal, even though the early cue is not the primary cue.</p> <p>It is critical to clarify that our concept of less continuous integration differs from the integration strategy outlined in buffered integration, which predicts an absence of lexical activation until other relevant cues are available. In contrast, less continuous integration reflects a case in which cues are integrated continuously, but with a delay, as listeners may take longer to integrate a cue, due to its insufficient reliability or informativeness at a given moment, particularly in the absence of an upcoming cue.</p> <p>This distinction becomes more explicit when considering potential variability in real‐time cue integration at the <emph>individual</emph> level. We now have empirical evidence for an extensive range of individual variability in cue weighting and categorization gradiency (e.g., Clayards, [<reflink idref="bib11" id="ref29">11</reflink>]; Hazan &amp; Rosen, [<reflink idref="bib23" id="ref30">23</reflink>]; Kapnoula, Edwards, &amp; McMurray, [<reflink idref="bib29" id="ref31">29</reflink>]; Kapnoula &amp; McMurray, [<reflink idref="bib30" id="ref32">30</reflink>]; Kapnoula, Winn, Kong, Edwards, &amp; McMurray, [<reflink idref="bib31" id="ref33">31</reflink>]; Kim, Clayards, &amp; Kong, [<reflink idref="bib32" id="ref34">32</reflink>]; Kong &amp; Edwards, [<reflink idref="bib37" id="ref35">37</reflink>]; Ou et al., [<reflink idref="bib58" id="ref36">58</reflink>]; Schertz, Cho, Lotto, &amp; Warner, [<reflink idref="bib62" id="ref37">62</reflink>]). However, it is unclear whether this variability is tied to variability in the time course of cue integration, though there have been some implications from prior studies.</p> <p>A recent study by Ou et al. ([<reflink idref="bib58" id="ref38">58</reflink>]) provides crucial evidence that individuals may exhibit varying degrees of cue use for lexical activation. They investigated the relationship between individual differences in categorization gradience and cue integration by measuring listeners' eye movements while categorizing stop voicing (/b/ vs. /p/) and tense‐lax vowel (/i/ vs. /ɪ/) contrasts in English where cues available earlier are the primary cues in both cases (VOT for voicing contrasts and spectral cues for vowel contrasts). Their eye‐movement results revealed that listeners with greater categorization gradience tended to adopt a less continuous integration strategy across different types of cues and contrasts, suggesting that these listeners were more likely to wait for secondary cues to arrive in the speech signal and thus integrate cues less continuously.</p> <p>This would suggest that categorization gradiency and cue weighting may be inherently linked to the different processing strategies adopted by individuals: Some individuals may make more immediate use of the primary cue for lexical activation, while others may adopt the primary cue less immediately. If that is the case, we might expect integration strategies to show consistent individual differences across cue integration and cue weighting. Thus, it is critical to incorporate individual differences into account for the link between cue weighting and cue integration.</p> <p>The present study addresses two crucial research questions. First, we examine whether acoustic cues are integrated continuously when the weight of the cue available later is greater than that of the cue available earlier. Answering this question is important for determining whether listeners start activating lexical items as soon as <emph>any</emph> perceptual cue arrives in the speech signal or whether lexical activation depends on cue informativeness. Second, we investigate whether an individual listener's cue weighting predicts their cue integration strategy to address the relationship between cue weighting and cue integration at the individual level. As a test case, we investigate how native Korean listeners weight and integrate acoustic cues to the Korean three‐way laryngeal stop contrasts in spoken word recognition.</p> <p>Korean has a typologically rare three‐way stop contrast among voiceless stops (the so‐called fortis, lenis, and aspirated stops), with VOT, F0 at vowel onset, closure duration, and H1‐H2 in the following vowel playing a role in distinguishing the three‐way contrast (e.g., Cho, Jun, &amp; Ladefoged, [<reflink idref="bib9" id="ref39">9</reflink>]). Traditionally, in the initial position of words produced in isolation, the fortis stop has a short VOT and high F0; the lenis stop has an intermediate VOT and low F0; and the aspirated stop has a long VOT and high F0 (Kim, [<reflink idref="bib34" id="ref40">34</reflink>]; Lisker &amp; Abramson, [<reflink idref="bib42" id="ref41">42</reflink>]).[<reflink idref="bib1" id="ref42">1</reflink>] Recently, in phrase‐initial or utterance‐initial position, the VOTs of word‐initial lenis and aspirated stops have gradually merged over time, whereas the VOT of fortis stops has not changed.[<reflink idref="bib2" id="ref43">2</reflink>] As a result, speakers are more likely to depend on the onset F0 of the vowel when distinguishing lenis stops from aspirated stops in speech perception (e.g., Lee, Politzer‐Ahles, &amp; Jongman, [<reflink idref="bib40" id="ref44">40</reflink>]; Schertz et al., [<reflink idref="bib62" id="ref45">62</reflink>]) as well as in speech production (e.g., Bang, Sonderegger, Kang, Clayards, &amp; Yoon, [<reflink idref="bib5" id="ref46">5</reflink>]; Choi, Kim, &amp; Cho, [<reflink idref="bib10" id="ref47">10</reflink>]; Kang &amp; Guion, [<reflink idref="bib27" id="ref48">27</reflink>]; Kang, [<reflink idref="bib28" id="ref49">28</reflink>]; Kim &amp; Jongman, [<reflink idref="bib33" id="ref50">33</reflink>]; Lee &amp; Jongman, [<reflink idref="bib39" id="ref51">39</reflink>]; Schertz et al., [<reflink idref="bib62" id="ref52">62</reflink>]; Silva, [<reflink idref="bib64" id="ref53">64</reflink>], but see Cho, [<reflink idref="bib8" id="ref54">8</reflink>] and Choi et al., [<reflink idref="bib10" id="ref55">10</reflink>], for a prosodic account in which high vs. low F0s are derived from post‐lexically assigned tones in the intonational phonology of Seoul Korean).</p> <p>In perception, Korean listeners perceive an intermediate VOT and lower F0 as a lenis stop, a long VOT and high F0 as an aspirated stop, and a short VOT and high F0 as a fortis stop (e.g., Lee et al., [<reflink idref="bib40" id="ref56">40</reflink>]; Schertz et al., [<reflink idref="bib62" id="ref57">62</reflink>]). These results indicate that VOT and F0 do not have the same weights for perceiving the three contrasts. Based on their cue weighting results, Schertz et al. ([<reflink idref="bib62" id="ref58">62</reflink>]) suggest that Korean listeners tend to use VOT as the primary cue and F0 as the secondary cue for the <emph>aspirated‐fortis</emph> contrast; VOT as the secondary cue and F0 as the primary cue for the <emph>aspirated‐lenis</emph> contrast; and both VOT and F0 as the primary cue for the <emph>lenis‐fortis</emph> contrast, with slightly greater use of F0 than VOT. These relative cue weightings make studying the time course of cue integration interesting because they raise the question of whether the cue integration that develops as the speech signal unfolds over time is contingent on the primary cue being heard earlier (VOT) or later (F0) in relation to the two stops that are being contrasted.</p> <p>To address this question, we must test the recognition of target words where VOT and F0 arrive at different points in time. For instance, the aspirated versus fortis contrast differs primarily by VOT and secondarily by F0. If listeners hear the fortis stop, the VOT and F0 cues will not be separated temporally in the speech signal because the VOT of fortis stops is very short. As a result, listeners will have little time to make use of VOT before F0 is heard, forcing listeners to use VOT and F0 around the same time. In contrast, if listeners hear the aspirated stop, as the VOT heard crosses the boundary between the fortis and aspirated stops, listeners may be able to conclude that the VOT matches that of an aspirated stop and not that of a fortis stop. As a result, listeners may use this information before the F0 cue is available in the speech signal. This means that words beginning with an aspirated stop are better candidates than words beginning with a fortis stop for testing the real‐time processing of cue integration.</p> <p>It is also important to consider the precise nature of the acoustic evidence needed for listeners to determine whether a particular cue belongs to one phonetic category or to a competing category. More specifically, given its temporal nature, VOT can provide evidence for the target word over the competitor word only if it is long enough to match the target word and <emph>longer</emph> than the assumed VOT of the competitor word. Thus, of any target‐competitor word pair, only the target word with the longer VOT can provide a good test case for the use of VOT before F0 becomes available in the speech signal. For example, the lenis stop has an intermediate VOT, and the aspirated stop has a long VOT. When listeners hear the lenis stop, they would not be able to integrate its VOT until the onset of F0, as the intermediate VOT is still compatible with the aspirated stop before the F0 cue arrives. On the other hand, when listeners hear the aspirated stop, they may be able to start determining that the VOT matches that of an aspirated stop and not that of a lenis stop as the VOT heard crosses the boundary between the lenis and aspirated stops. Listeners may then use this information before the F0 is available in the speech signal.</p> <p>Given these considerations, the stop contrasts that best serve as test cases for investigating the time course of cue integration are: (i) the aspirated versus fortis contrast (VOT as a primary cue and F0 as a secondary cue), with the auditory stimulus (target) containing the aspirated stop and its competitor beginning with the fortis stop; (ii) the aspirated versus lenis contrast (VOT as a secondary cue and F0 as a primary cue), with the auditory stimulus (target) containing the aspirated stop and its competitor beginning with the lenis stop; and (iii) the lenis versus fortis contrast (VOT and F0 as primary cues, with F0 being the primary cue), with the auditory stimulus (target) beginning with the lenis stop and its competitor beginning with the fortis stop. Examining when fixations to the target word diverge from those to the competitor word in these three conditions will shed important light on <emph>when</emph> listeners begin using VOT once it becomes available and before F0 is heard in the speech signal. In particular, the contrasts in (ii) and (iii) will be good candidates to test whether there is a relationship between cue weighting and cue integration: For the aspirated versus lenis contrast, the primary cue (F0) is available later than the secondary cue (VOT), and for the lenis versus fortis contrast, the cue with slightly greater perceptual weight (F0) is available later in the speech signal than the cue with slightly less perceptual weight (VOT). If there is an interplay between cue weighting and integration, listeners will be delayed in using VOT with (ii) and (iii) but not with (i).</p> <p>Crucially, this prediction is contingent on the assumption that listeners will use top‐down information from the visual display (i.e., the displayed words) to anticipate the cues that may be coming from the speech signal. In the presence of visual words that begin with an aspirated stop and a lenis stop (i.e., [ii]), listeners may wait for F0 to distinguish the target word from the competitor word and not use VOT when it becomes available, despite its potential usefulness. Similarly, in the presence of visual words that begin with a lenis stop and a fortis stop (i.e., [iii]), listeners may wait for F0 to distinguish the target and competitor words and not make immediate use of VOT. In contrast, in the presence of visual words that begin with an aspirated stop and fortis stop (i.e., [i]), listeners may immediately use VOT to distinguish the target word from the competitor since F0 is not informative enough to distinguish the two words. This top‐down information from the visual display is thus assumed to help listeners interpret the cues they hear from the speech signal and use them strategically to recognize spoken words.</p> <p>We test our hypotheses using two main experimental paradigms: cue‐weighting speech perception (Experiment 1) and VWP (Experiment 2). The cue‐weighting task quantifies listeners' reliance on each acoustic dimension of the Korean stop contrasts. The results of the cue‐weighting task will serve as a foundation to interpret the VWP results. Critically, the results will assert that listeners' relative reliance on a particular cue is contingent on the specific stop contrast being processed, which is crucial for examining the interplay between cue primacy and cue integration. The VWP examines the real‐time processing of acoustic information as the speech signal unfolds over time. In addition to testing the hypotheses at the group level, we examine whether individual listeners' cue weighting predicts their time course of cue integration. Like the findings of Ou et al. ([<reflink idref="bib58" id="ref59">58</reflink>]), we expect that individual listeners who rely more on a redundant (i.e., secondary) cue would be more likely to integrate acoustic information in a less continuous manner, as these listeners are more likely to wait for secondary cues to arrive in the speech signal. This would suggest that categorization gradiency and cue weighting may be inherently linked to the different processing strategies adopted by individuals. If that is the case, we might expect integration strategies to show individual differences that are consistent across cue integration and cue weighting.</p> <hd id="AN0181921441-3">Experiment 1</hd> <p></p> <hd id="AN0181921441-4">Methods</hd> <p></p> <hd id="AN0181921441-5">Participants</hd> <p>Participants included 62 college‐age native Seoul Korean speakers (mean age: 24.1, SD: 3.2, 38 female) with normal or corrected‐to‐normal vision and no history of speech and hearing disorders. All participants were recruited and tested at a university in Seoul and received monetary compensation for their participation ($10 per hour). They completed three tasks in the following order: (<reflink idref="bib1" id="ref60">1</reflink>) a language background questionnaire; (<reflink idref="bib2" id="ref61">2</reflink>) a VWP (Section 3); and (<reflink idref="bib3" id="ref62">3</reflink>) a cue‐weighting speech perception experiment. All participants gave informed consent, and the experimental research protocol was approved by the Institutional Review Board at a university in the United States.</p> <hd id="AN0181921441-6">Stimuli</hd> <p>The auditory stimuli for the cue‐weighting task were the Korean triplet, [p*ul] "horn," [pul] "fire," and [p<sups>h</sups>ul] "grass" (Lee et al., [<reflink idref="bib40" id="ref63">40</reflink>]). The triplet was recorded by a college‐age female Seoul Korean speaker with five repetitions. The recording was conducted in an anechoic chamber, using a microphone (Electro Voice N/D 767a) and a digital recorder (Marantz PMD 671) at a sampling rate of 22,050 Hz. The acoustic measurement and manipulations were performed in Praat (Boersma &amp; Weenink, [<reflink idref="bib7" id="ref64">7</reflink>]). The VOT of the word‐initial stop and the onset F0 were manipulated to yield seven‐step continua. The minimum and maximum values of the VOT continuum were 0 and 110 ms, and those of the F0 continuum were 180 and 290 Hz. These values were based on the acoustic analysis of the recorded tokens as well as the range of continuum used in previous perception studies (Lee et al., [<reflink idref="bib40" id="ref65">40</reflink>]; Schertz et al., [<reflink idref="bib62" id="ref66">62</reflink>]).</p> <p>After a visual inspection and acoustic analysis of the recorded tokens, one clear aspirated stop token (i.e., [p<sups>h</sups>ul] "grass"; with the longest VOT) and one clear fortis stop token (i.e., [p*ul] "horn"; with the shortest VOT) were selected as the base tokens for the stimuli manipulation. The VOT and F0 manipulation was done using a Praat script (Winn, [<reflink idref="bib74" id="ref67">74</reflink>]).[<reflink idref="bib3" id="ref68">3</reflink>] Since manipulating VOT yields different lengths of tokens within a triplet, the duration of the following vowel was lengthened or shortened to obtain the same length of tokens within a triplet. This was done so that the stimuli in this experiment would be parallel to those in the VWP (Experiment 2), where it is crucial to control the timing of disambiguation of the words. The manipulation procedures were as follows: First, for the triplet, the mean duration of the VOT‐manipulated tokens was calculated; second, for each token with a manipulated VOT, the difference between the mean duration and the duration of the token was added to or subtracted from the vowel duration of the token. For example, if a step 1 token had a VOT of 10 ms and a vowel duration of 190 ms, and the mean duration of the tokens across VOT steps was 210 ms, then the length of the vowel was increased to 200 ms so that the token would have a duration of 210 ms. A total of 49 auditory stimuli (7 steps of VOT × 7 steps of F0) were created using these procedures and then normalized to a mean amplitude of 70 dB using the <emph>Scale intensity</emph> function in Praat.</p> <hd id="AN0181921441-7">Procedure</hd> <p>The cue‐weighting task was built and presented using the Gorilla Experiment Builder (Anwyl‐Irvine, Massonnie, Flitton, Kirkham, &amp; Evershed, [<reflink idref="bib2" id="ref69">2</reflink>]). Participants heard an auditory stimulus over headphones and were instructed to identify what they heard by pressing either the left or the right arrow button on the keyboard. The left and right arrows represented the words on the left and the right, respectively. The word labels and corresponding buttons appeared on the computer monitor at the offset of each auditory stimulus, and the position of the word labels in relation to the word‐initial stop was counterbalanced across trials. Since the purpose of the experiment was to examine listeners' weighting of acoustic cues that distinguish one specific laryngeal category from another, each trial had two choices of response rather than three choices. As a result, there were three pairs of two alternatives (i.e., aspirated vs. lenis choice, aspirated vs. fortis choice, and lenis vs. fortis choice). The next trial began 1000 ms after the participant's response. Prior to the main session, there were 10 practice trials, consisting of a subset of actual stimuli in the main session. In the main session, a total of 441 trials (49 stimuli × 3 pairs of choices × 3 repetitions) were randomly presented in three blocks, with each stimulus being heard once per block. The task took approximately 15 min.</p> <hd id="AN0181921441-8">Data analysis and predictions</hd> <p>Mixed‐effects logistic regression models were conducted on participants' responses on each trial using the <emph>glmer</emph> function of the <emph>lme4</emph> package (Bates, Mächler, Bolker, &amp; Walker, [<reflink idref="bib6" id="ref70">6</reflink>]) of R (R Development Core Team, [<reflink idref="bib59" id="ref71">59</reflink>]). Separate analyses were run for each of the three contrast types (the fortis vs. aspirated contrast with listeners' aspirated responses coded as 1, the lenis vs. aspirated contrast with listeners' lenis responses coded as 1, and the fortis vs. lenis contrast with listeners' lenis responses coded as 1). For each model, the fixed effects included the two manipulated acoustic dimensions (VOT and F0, each centered) and their interaction. The model also included random intercepts for each item and participant. VOT and F0 were included as random slopes for each participant. For each analysis, the best model was selected by fitting the fixed effects backward using the log‐likelihood ratio test. When two cues showed a significant effect, their fixed‐effect coefficients were compared to determine whether one cue had a stronger effect than the other (Tremblay et al., [<reflink idref="bib70" id="ref72">70</reflink>]). For the analysis of the lenis versus fortis contrast, in addition to VOT and F0, VOT<sups>2</sups> was added to the model to examine the nonmonotonic (i.e., inverted U‐shaped) effect of VOT. Since previous studies have shown that there are more lenis stop responses at lower levels of VOT but fewer lenis stop responses at higher levels of VOT, the relationship between lenis response and VOT level is less likely to be linear (Lee et al., [<reflink idref="bib40" id="ref73">40</reflink>]). Adding VOT<sups>2</sups> as a predictor in the model allows us to examine this nonmonotonic effect.</p> <p>The results are predicted to be in line with previous perception studies (Lee et al., [<reflink idref="bib40" id="ref74">40</reflink>]; Schertz et al., [<reflink idref="bib62" id="ref75">62</reflink>]). First, when listeners distinguish aspirated from fortis stops, the likelihood of aspirated responses should increase as VOT increases, as evidenced by a significant positive coefficient of VOT. A comparison of the fixed‐effects coefficients for the model is expected to show a stronger effect of VOT compared to the effect of F0 (if any F0 effect is found). Second, when distinguishing aspirated from lenis stops, the likelihood of aspirated responses should increase as F0 increases. This should be evidenced by a significant positive coefficient of F0 in the aspirated versus lenis model. A comparison of the fixed‐effects coefficients for the model is expected to show a stronger effect of F0 compared to the effect of VOT. Third, listeners would be more likely to identify a token as a lenis stop for the lenis versus fortis contrast as VOT increases. A significant positive coefficient of VOT in the lenis versus fortis model should support this prediction. However, if lenis stop responses decrease at higher steps of VOT, the model should find a significant main effect of VOT<sups>2</sups>. Since both VOT and F0 play an important role in distinguishing lenis from fortis stops, there would also be a significant main effect of F0. A comparison of the fixed‐effects coefficients for the model is expected to reveal a somewhat stronger effect of F0 (Schertz et al., [<reflink idref="bib62" id="ref76">62</reflink>]).</p> <p>To quantify individual listeners' reliance on each acoustic dimension, each listener's logistic regression beta‐coefficients were extracted and used as a measure of perceptual cue weight (e.g., Clayards, [<reflink idref="bib11" id="ref77">11</reflink>]; Escudero, Benders, &amp; Lipski, [<reflink idref="bib13" id="ref78">13</reflink>]; Kong &amp; Edwards, [<reflink idref="bib37" id="ref79">37</reflink>]; Schertz et al., [<reflink idref="bib62" id="ref80">62</reflink>]; Tremblay et al., [<reflink idref="bib70" id="ref81">70</reflink>]) (see also Morrison, [<reflink idref="bib55" id="ref82">55</reflink>]; Morrison &amp; Kondaurova, [<reflink idref="bib56" id="ref83">56</reflink>], for the validation of using logistic regression coefficients as metrics for cue weights). The coefficients reflect how much a one‐step difference in one of the predictors (i.e., one of the acoustic dimensions) causes a change in the log‐odds of a listener's response. A logistic regression model was conducted for each listener and each contrast type with VOT and F0 (each centered) as predictors, but not with the interaction term, as the goal is to calculate a single quantification for each cue's weight.</p> <p>It should be noted that the coefficient for each predictor represents a listener's reliance on each acoustic dimension but not the relative weight of one cue in relation to the other. For example, a listener with a high sensitivity to VOT may also show a high sensitivity to F0, resulting in similar high coefficients for both cues from the model. In terms of cue weightings within a listener, however, similarly high coefficients mean that the relative weightings of the two cues are similar, and thus that both cues are used heavily to process the contrast. Following the method adopted by Escudero et al. ([<reflink idref="bib13" id="ref84">13</reflink>]), the relative weightings of the cues involved in the contrast were quantified by calculating the ratio of the coefficients of F0 (<emph>β</emph><subs>F0</subs>) to those of VOT (<emph>β</emph><subs>VOT</subs>) (i.e., relative cue weighting = <emph>β</emph><subs>F0</subs>/(<emph>β</emph><subs>VOT</subs> + <emph>β</emph><subs>F0</subs>)). If a listener weights both cues equally heavily, the relative cue weighting of the listener should be 0.5. If a listener weights the F0 cue more heavily, the relative cue weighting should be higher than 0.5; if the value is lower than 0.5, the VOT cue is weighted more heavily.</p> <hd id="AN0181921441-9">Results</hd> <p></p> <hd id="AN0181921441-10">Group results</hd> <p>Listeners' responses in the aspirated versus fortis contrast (Fig. 1A) showed that their aspirated responses are mainly modulated by the changes in VOT steps. The model of the aspirated versus fortis contrast (Table 1A) found a significant effect of VOT, with listeners' proportion of aspirated stop responses increasing as the VOT step increased. In contrast, listeners' proportion of aspirated stop responses did not increase significantly as the F0 step increased, indicating that F0 is not a vital cue to distinguish fortis stops from aspirated stops. The model also revealed a significant two‐way interaction between VOT and F0, with the effect of VOT in listeners' aspirated stop responses being greater at higher levels of F0 and with the effect of F0 reversing as VOT increases, suggesting that the likelihood of aspirated stop responses along the VOT continuum varies depending on F0.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CGN/01dec24/cogs70026-fig-0001.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="cogs70026-fig-0001.jpg" title="1 (A) Heat plot of listeners' aspirated responses on the aspirated versus fortis stop contrast, with listeners' aspirated stop responses coded as 1. The darker the cell, the greater the proportion of aspirated stop responses. (B) Heat plot of listeners' aspirated responses on the aspirated versus lenis stop contrast, with listeners' aspirated stop responses coded as 1. The darker the cell, the greater the proportion of aspirated stop responses. (C) Heat plot of listeners' lenis responses on the lenis versus fortis stop contrast, with listeners' lenis stop responses coded as 1. The darker the cell, the greater the proportion of lenis stop responses." /> </p> <p></p> <p>1 Table Summary of fixed‐effect coefficients in the mixed‐effects logistic regression model on listeners' response in the cue‐weighting task</p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th align="left"&gt;&lt;italic&gt;A. Model with listeners' response in the aspirated versus fortis contrast&lt;/italic&gt;&lt;/th&gt;&lt;/tr&gt;&lt;tr&gt;&lt;th align="left" /&gt;&lt;th align="center"&gt;Estimate&lt;/th&gt;&lt;th align="center"&gt;SE&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;z&lt;/italic&gt;&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;Pr&lt;/italic&gt;(&amp;#62;|z|)&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;(Intercept)&lt;/td&gt;&lt;td&gt;0.789&lt;/td&gt;&lt;td&gt;0.147&lt;/td&gt;&lt;td&gt;5.375&lt;/td&gt;&lt;td&gt;&amp;#60; 0.001&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;VOT&lt;/td&gt;&lt;td&gt;1.268&lt;/td&gt;&lt;td&gt;0.077&lt;/td&gt;&lt;td&gt;16.479&lt;/td&gt;&lt;td&gt;&amp;#60; 0.001&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;F0&lt;/td&gt;&lt;td&gt;0.112&lt;/td&gt;&lt;td&gt;0.066&lt;/td&gt;&lt;td&gt;1.697&lt;/td&gt;&lt;td&gt;0.089&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;VOT &amp;#215; F0&lt;/td&gt;&lt;td&gt;0.197&lt;/td&gt;&lt;td&gt;0.032&lt;/td&gt;&lt;td&gt;6.203&lt;/td&gt;&lt;td&gt;&amp;#60; 0.001&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p></p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th align="left"&gt;&lt;italic&gt;B. Model with listeners' response in the aspirated versus lenis contrast&lt;/italic&gt;&lt;/th&gt;&lt;/tr&gt;&lt;tr&gt;&lt;th align="left" /&gt;&lt;th align="center"&gt;Estimate&lt;/th&gt;&lt;th align="center"&gt;SE&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;z&lt;/italic&gt;&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;Pr&lt;/italic&gt;(&amp;#62;|z|)&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;(Intercept)&lt;/td&gt;&lt;td&gt;0.426&lt;/td&gt;&lt;td&gt;0.156&lt;/td&gt;&lt;td&gt;2.739&lt;/td&gt;&lt;td&gt;0.006&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;VOT&lt;/td&gt;&lt;td&gt;0.693&lt;/td&gt;&lt;td&gt;0.073&lt;/td&gt;&lt;td&gt;9.500&lt;/td&gt;&lt;td&gt;&amp;#60; 0.001&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;F0&lt;/td&gt;&lt;td&gt;0.959&lt;/td&gt;&lt;td&gt;0.071&lt;/td&gt;&lt;td&gt;13.531&lt;/td&gt;&lt;td&gt;&amp;#60; 0.001&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;VOT &amp;#215; F0&lt;/td&gt;&lt;td&gt;0.125&lt;/td&gt;&lt;td&gt;0.033&lt;/td&gt;&lt;td&gt;3.823&lt;/td&gt;&lt;td&gt;&amp;#60; 0.001&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p></p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th align="left"&gt;&lt;italic&gt;C. Model with listeners' response in the lenis versus fortis contrast&lt;/italic&gt;&lt;/th&gt;&lt;/tr&gt;&lt;tr&gt;&lt;th align="left" /&gt;&lt;th align="center"&gt;Estimate&lt;/th&gt;&lt;th align="center"&gt;SE&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;z&lt;/italic&gt;&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;Pr&lt;/italic&gt;(&amp;#62;|z|)&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;(Intercept)&lt;/td&gt;&lt;td&gt;1.185&lt;/td&gt;&lt;td&gt;0.196&lt;/td&gt;&lt;td&gt;6.043&lt;/td&gt;&lt;td&gt;&amp;#60; 0.001&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;VOT&lt;/td&gt;&lt;td&gt;0.541&lt;/td&gt;&lt;td&gt;0.066&lt;/td&gt;&lt;td&gt;8.260&lt;/td&gt;&lt;td&gt;&amp;#60; 0.001&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;F0&lt;/td&gt;&lt;td&gt;&amp;#8722;0.762&lt;/td&gt;&lt;td&gt;0.062&lt;/td&gt;&lt;td&gt;&amp;#8722;12.215&lt;/td&gt;&lt;td&gt;&amp;#60; 0.001&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;VOT&lt;sup&gt;2&lt;/sup&gt;&lt;/td&gt;&lt;td&gt;&amp;#8722;0.249&lt;/td&gt;&lt;td&gt;0.032&lt;/td&gt;&lt;td&gt;&amp;#8722;7.791&lt;/td&gt;&lt;td&gt;&amp;#60; 0.001&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;VOT &amp;#215; F0&lt;/td&gt;&lt;td&gt;0.075&lt;/td&gt;&lt;td&gt;0.028&lt;/td&gt;&lt;td&gt;2.677&lt;/td&gt;&lt;td&gt;0.007&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p>1 <emph>Notes</emph>. R code of the best model for the aspirated versus fortis and aspirated versus lenis contrast: <emph>glmer</emph>(Response ∼ VOT * F0 + (1 | item) + (VOT + F0 | subject)).</p> <p>2 R code of the best model for the lenis versus fortis contrast: <emph>glmer</emph>(Response ∼ VOT * F0 + VOT<sups>2</sups> + (1 | item) + (VOT + F0 | subject)).</p> <p>For the aspirated versus lenis contrast (Fig. 1B), listeners' aspirated responses were affected by both VOT and F0. The model of the aspirated versus lenis contrast (Table 1B) revealed significant main effects of VOT and F0, with listeners' proportion of aspirated stop responses increasing as the step number of each acoustic dimension increases. A comparison of the fixed‐effects coefficients found that the effect of F0 was stronger than that of VOT, <emph>t</emph>(9060) = 2.69, <emph>p</emph> &lt; .01. The model found a significant two‐way interaction between VOT and F0, with the effect of VOT in listeners' aspirated stop responses being greater at higher levels of F0 and with the effect of F0 being greater at higher levels of VOT.</p> <p>The model with listeners' responses in the lenis versus fortis contrast (Table 1C) revealed significant main effects of VOT and F0, with listeners' proportion of lenis stop responses increasing as the step number of VOT increases and that of F0 decreases. A comparison of the fixed‐effects coefficients found that the effect of F0 was stronger than that of VOT, <emph>t</emph>(9035) = −14.85, <emph>p</emph> &lt; .01. The model also found a significant main effect of VOT<sups>2</sups>, indicating that the effect of VOT is not monotonic. More specifically, lenis stop responses increase as VOT increases at low levels of VOT, and the responses decrease as VOT increases at high levels of VOT. Additionally, there was a significant two‐way interaction between VOT and F0, with the effect of VOT in listeners' lenis stop responses being greater at higher levels of F0 and with the effect of F0 being greater at lower levels of VOT. This suggests that the likelihood of lenis stop responses along the VOT continuum varies depending on F0.</p> <hd id="AN0181921441-12">Individual results</hd> <p>Fig. 2 plots the predictability (beta‐coefficients) of Korean listeners' reliance on each acoustic dimension measured from individual regression models. The beta‐coefficients of individual models were transformed into absolute values, as we compared individual listeners' <emph>reliance</emph> on each cue in processing each contrast. The beta‐coefficients among individual listeners were then square‐root transformed, as the distribution was moderately skewed. A numerically larger coefficient indicates a heavier reliance on a given acoustic dimension for listeners' processing of contrast (see Appendix A for the individual listeners' coefficients).</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CGN/01dec24/cogs70026-fig-0002.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="cogs70026-fig-0002.jpg" title="2 Predictability of listeners' reliance on each acoustic dimension. The white dots represent the mean values. The length of the violins indicates the range of values, and the width of the violins at a given y value represents the point density at that value." /> </p> <p></p> <p>For the aspirated versus fortis contrast (Fig. 2A), none of the listeners relied more on F0; VOT served as a primary cue for all 62 listeners. There was considerable variability among listeners for the other contrasts. For the aspirated versus lenis contrast (Fig. 2B), 43 listeners showed a stronger reliance on F0 than VOT. For the lenis versus fortis contrast (Fig. 2C), again, 43 listeners relied more on F0, though they were not completely the same listeners who showed stronger reliance on F0 in the aspirated versus lenis contrast. This quantified individual reliance on each cue will be used to predict individual listeners' time course of cue integration, which will be examined by the VWP (Experiment 2).</p> <hd id="AN0181921441-14">Discussion</hd> <p>The results of Experiment 1 largely replicated those of Lee et al. ([<reflink idref="bib40" id="ref85">40</reflink>]) and Schertz et al. ([<reflink idref="bib62" id="ref86">62</reflink>]). When distinguishing aspirated stops from fortis stops, listeners relied more heavily on VOT than F0. For the aspirated versus lenis contrast, listeners relied on both VOT and F0, with F0 being a perceptually more informative dimension. The results of the lenis versus fortis contrast showed that listeners relied on both VOT and F0, but F0 was again a perceptually more reliable dimension. Individual‐level analyses further clarified variability in listeners' reliance on acoustic dimensions. While the group pattern showed that listeners use F0 as a primary cue to the aspirated versus lenis or lenis versus fortis contrast, some individual listeners relied more on VOT than F0 to perceive the contrasts.</p> <p>Importantly, listeners' cue‐weighting patterns for each contrast type measured in Experiment 1 serve as a foundation for interpreting the time course of cue integration, as measured by a VWP experiment (Experiment 2) that seeks to determine whether the timing of cue integration is contingent on the perceptual primacy of cues.</p> <hd id="AN0181921441-15">Experiment 2</hd> <p></p> <hd id="AN0181921441-16">Method</hd> <p></p> <hd id="AN0181921441-17">Participants</hd> <p>The same native Korean listeners from Experiment 1 completed a VWP experiment.</p> <hd id="AN0181921441-18">Stimuli</hd> <p>Fifteen disyllabic Korean words (five disyllabic noun triplets) were used as the critical stimuli (Table B1). The word‐initial consonants of the critical triplets were bilabial stops. The lexical items within a triplet shared the same syllable structure and phonemes in the first syllable except for the laryngeal type of the word‐initial stop, such that there was a temporary lexical ambiguity contingent on the initial consonant. The critical words within a triplet further differed from one another at the onset of the second syllable for disambiguation. The critical words were controlled for (log‐transformed) token frequency (established from the Sejong Corpus) and number of letters.</p> <p>The words in the display were presented orthographically. An orthographic presentation was inevitable because some of the words are not easily imageable (for validation of the use of printed words in VWPs, see Huettig &amp; McQueen, [<reflink idref="bib25" id="ref87">25</reflink>]; McQueen &amp; Viebahn, [<reflink idref="bib51" id="ref88">51</reflink>]). Along with the critical stimuli, there were 15 filler stimuli (five disyllabic noun triplets) with a three‐way affricate contrast in word‐initial position (Table B2). Like the critical stimuli, the lexical items within a filler triplet shared the same syllable structure and phonemes in the first syllable except for the word‐initial affricate.</p> <p>The target and competitor words were presented in pairs, like in Experiment 1. For critical trials, there were six target‐competitor types (3 target types × 2 competitor types): the fortis target with the lenis competitor; the fortis target with the aspirated competitor; the lenis target with the fortis competitor; the lenis target with the aspirated competitor; the aspirated target with the fortis competitor; and the aspirated target with the lenis competitor. On each trial, a target‐competitor pair was grouped with one filler pair in the visual display, making a quadruplet (e.g., [p*alt*e] "straw" – [p<sups>h</sups>alʧ*i] "bracelet," [ʧ*okʧ*i] "note" – [ʧ<sups>h</sups>okk*am] "feel"); thus, the display consisted of one target word and one competitor word (a critical pair), and two distracter words (a filler pair).</p> <p>The same speaker who recorded items for Experiment 1 produced each word three times in the carrier sentence, [ʧiɡɨm] <emph>___</emph> [ɨl nulɨsejo] "Now click on ___." One clear recording of the carrier phrase was selected for the VWP. The target words were extracted from their original carrier sentence and reinserted into the selected carrier sentence after a 300‐ms silent pause after the first part of the carrier sentence (i.e., [ʧiɡɨm] "now"). After visual inspection and acoustic measurement of the recorded tokens, one lenis stop token was selected as the base token. Stimulus manipulation was done with the first syllables. Like Experiment 1, the VOT of the word‐initial stop and onset F0 of the following vowel were manipulated using a Praat script (Winn, [<reflink idref="bib74" id="ref89">74</reflink>]). However, unlike Experiment 1, the VOT of the word‐initial stop and onset F0 of the following vowel were manipulated to have designated VOT and F0 values (Table 2). These values were selected based on the results of the perception experiments from Kong and Lee ([<reflink idref="bib38" id="ref90">38</reflink>]), Lee et al. ([<reflink idref="bib40" id="ref91">40</reflink>]), and Schertz et al. ([<reflink idref="bib62" id="ref92">62</reflink>]). The selected VOT and F0 values were chosen to maximize the likelihood of each laryngeal type being perceived as the corresponding type, while also ensuring a clear differentiation between VOTs and F0s across the laryngeal types. In particular, the selected VOT values for each laryngeal type ensured that VOT and F0 arrive as <emph>asynchronously</emph> as possible when listeners hear words beginning with an aspirated stop or a lenis stop as targets.</p> <p>2 Table Designated VOT and F0 values of stimuli in the VWP</p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th&gt;Laryngeal type&lt;/th&gt;&lt;th align="center"&gt;VOT (ms)&lt;/th&gt;&lt;th&gt;F0 (Hz)&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;Fortis&lt;/td&gt;&lt;td&gt;0&lt;/td&gt;&lt;td&gt;250&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;Lenis&lt;/td&gt;&lt;td&gt;55&lt;/td&gt;&lt;td&gt;180&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;Aspirated&lt;/td&gt;&lt;td&gt;110&lt;/td&gt;&lt;td&gt;290&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p>Since manipulating VOT yielded different lengths of the first syllable within a triplet and resulted in longer VOTs, thus taking slightly more time to reach the point of disambiguation in the word (i.e., the first consonant of the second syllable), the duration of the following vowel was lengthened or shortened to ensure that the length of the first syllable would be the same across tokens within a triplet. The procedures of vowel duration manipulation were identical to those of Experiment 1. All stimuli were normalized to a mean amplitude of 70 dB.</p> <hd id="AN0181921441-19">Procedure</hd> <p>The task was built using Experiment Builder ([<reflink idref="bib14" id="ref93">14</reflink>]), and the eye movements of the participants were recorded with a head‐mounted SR EyeLink II eye‐tracker (2006), sampling at 250 Hz. The experiment began with a calibration of the eye tracker using the participants' right eye. The calibration was followed by four practice trials and by the main experiment. In each trial, participants saw a display with four printed Korean words (i.e., a target, a competitor, and two distracters) in Korean orthography, arrayed in an invisible 2 × 2 grid, for 2000 ms. This preview time allowed participants to briefly scan the location of each printed word on display, thus minimizing later eye movements due to visual search. The arrangements of the printed words in a grid were randomized on each trial. After the 2000 ms preview, the words disappeared, and a fixation cross appeared in the center of the screen for 500 ms. This was to ensure that the point of the gaze was at the center of the screen before the auditory stimuli onset (see Apfelbaum, Klein‐Packard, &amp; McMurray, [<reflink idref="bib4" id="ref94">4</reflink>]; Huettig &amp; McQueen, [<reflink idref="bib25" id="ref95">25</reflink>], for the role of preview in VWP). As soon as the fixation cross disappeared, the four printed words reappeared on display in the same location, and the auditory stimulus was heard over headphones. Participants were asked to click on the word corresponding to the auditory stimulus with the computer mouse as soon as they heard the target word within the carrier sentence, and their eye movements were measured from the onset of the target word. Each trial ended with the participant's response, and the subsequent trial began after a 1000 ms intertrial interval.</p> <p>There were 180 trials, including 90 critical trials (6 target‐competitor types × 5 triplets × 3 repetitions) and 90 filler trials (6 target‐competitor types × 5 triplets × 3 repetitions). These trials were presented in three blocks, each containing 60 trials. The order of the experimental and filler trials within a block and the order of blocks were randomized across participants. The testing session took approximately 40 min.</p> <hd id="AN0181921441-20">Data analysis and predictions</hd> <p>The proportions of fixations to the target, competitor, and distractors were extracted in 4 ms time bins from the onset to the offset of the target word. For each time bin, the difference between the empirical log‐transformed proportions of target and competitor fixations was calculated (i.e., target‐over‐competitor fixation advantage). Data analyses were conducted on the target‐competitor types that provide the best test cases for the timing of cue integration (i.e., the condition with an aspirated target and a fortis competitor, and the condition with an aspirated target and a lenis competitor, and the condition with a lenis target and a fortis competitor). For the statistical analyses, we adopted a generalized additive mixed model (GAMM; Wieling, [<reflink idref="bib73" id="ref96">73</reflink>]; Wood, [<reflink idref="bib75" id="ref97">75</reflink>]; Wood, [<reflink idref="bib77" id="ref98">77</reflink>]), using the <emph>bam</emph> function of the <emph>mgcv</emph> package in R (Wood, [<reflink idref="bib76" id="ref99">76</reflink>]; Wood, [<reflink idref="bib77" id="ref100">77</reflink>]). GAMM has an important advantage over other statistical analyses in that proportions of fixations do not need to be averaged over time or selected at a specific time point, and the residuals (i.e., the gap between the fitted value and the actual values) factor in autocorrelation, making it superior to other statistical approaches to time series data, such as Growth Curve Analysis (Mirman, Dixon, &amp; Magnuson, [<reflink idref="bib52" id="ref101">52</reflink>]). As such, this analysis allows us to assess how the specific nonlinear eye fixation patterns change as speech signal unfolds over time.</p> <p>For each target‐competitor type, the target‐over‐competitor fixation advantage was aligned with the VOT disambiguation point. The VOT disambiguation point was defined as the time point where listeners should start utilizing VOT to distinguish a target from a competitor in the speech signal. For the condition with an aspirated target and a fortis competitor, the VOT disambiguation point corresponds to the onset of VOT in the speech signal, as the VOT of fortis stops is short (typically less than 10 ms for bilabial stops). In this case, listeners would start using VOT as soon as it is available in the speech signal, as a VOT longer than 10 ms should distinguish the target aspirated stops from the competitor fortis stops. For the condition with an aspirated target and a lenis competitor, the VOT disambiguation point does not match with the onset of VOT because there would be approximately 55 ms of overlap in VOT between aspirated stops and lenis stops. As a result, listeners would be able to use VOT to distinguish the aspirated stops from the lenis stops competitor after the end of the overlapping interval—after 55 ms, making 55 ms the VOT disambiguation point. For the condition with a lenis target and a fortis competitor, the VOT disambiguation point corresponds to the onset of VOT, for the same reason mentioned for the condition with an aspirated target and a fortis competitor.</p> <p>Each of the GAMMs included the target‐over‐competitor fixation advantage from the VOT disambiguation point in the speech signal to the onset of the second syllable as the dependent variable, and the predictors included a smoothing pattern over time for the target‐competitor type and a binary smooth that models the constant difference between the nonlinear pattern of the fixation advantage of a target‐competitor type and 0 (i.e., no target‐over‐competitor advantage). Additionally, the model included each subject and lexical item as random intercepts, target‐competitor type as a random slope for each subject, and the nonlinear difference over time for each of the subjects and nonlinear difference over time for each of the lexical items as nonlinear random effects (see Appendix C for the full model specification). The time point of the onset of the VOT effect was quantified using the <emph>plot_diff</emph> function of <emph>itsadug</emph> package in R (van Rij et al., [<reflink idref="bib71" id="ref102">71</reflink>]). The function specifically calculated at which time point the target‐over‐competitor fixation advantage started showing a significant difference from 0 (null effect).</p> <p>Predictions for each target‐competitor type are summarized in Table 3. For the condition with an aspirated target and a fortis competitor, both the associated view and the independent view predict a similar output but for slightly different reasons. Under the associated view, listeners should start integrating VOT before the F0 cue is available, as VOT is a primary cue to the aspirated versus fortis contrast. Thus, we expect to see an immediate integration of VOT. Under the independent view, regardless of the relative cue weighting, listeners should start integrating VOT before the F0 cue is available, as VOT would be long enough for listeners to distinguish aspirated from fortis stops. Thus, we expect to see an immediate integration of VOT under the two views. Although this condition would not provide a good test of the relationship between cue integration and cue weighting, it would provide a reference point to interpret the timing of cue integration compared to the other two target‐competitor types.</p> <p>3 Table Summary of predictions for the timing of VOT integration</p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th&gt;Target&amp;#8208;competitor type&lt;/th&gt;&lt;th align="center"&gt;Cue weighting (Experiment 1)&lt;/th&gt;&lt;th align="center"&gt;Associated view&lt;/th&gt;&lt;th align="center"&gt;Independent view&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;Aspirated&amp;#8208;Fortis&lt;/td&gt;&lt;td&gt;VOT &amp;#62; F0&lt;/td&gt;&lt;td&gt;Immediate integration&lt;/td&gt;&lt;td&gt;Immediate integration&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;Aspirated&amp;#8208;Lenis&lt;/td&gt;&lt;td&gt;VOT &amp;#60; F0&lt;/td&gt;&lt;td&gt;Delayed integration&lt;/td&gt;&lt;td&gt;Immediate integration&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;Lenis&amp;#8208;Fortis&lt;/td&gt;&lt;td&gt;VOT &amp;#60; F0&lt;/td&gt;&lt;td&gt;Delayed integration&lt;/td&gt;&lt;td&gt;Immediate integration&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p>For the condition with an aspirated target and a lenis competitor, the associated view predicts that the integration of VOT should be delayed until the arrival of the F0 cue, as F0 serves as a perceptually more informative cue to distinguish aspirated stops from lenis stops. We should, therefore, see a less immediate use of VOT for lexical activation: The target‐over‐competitor fixation advantage should be significantly above zero (indicating a fixation divergence between target and competitor fixations) only when the F0 cue becomes available. Conversely, under the independent view, listeners would start integrating VOT before the F0 cue becomes available in the speech signal. Thus, we should find an immediate integration of VOT: Listeners should start looking at the target over the competitor before the F0 cue becomes available. In other words, the effect of VOT should kick in earlier than that of F0. Hence, this condition can tease apart the associated and independent views.</p> <p>For the condition with a lenis target and a fortis competitor, the associated view predicts that listeners should show less continuous integration of VOT until the arrival of the F0 cue, as F0 serves as the primary cue to distinguish lenis stops from fortis stops. Thus, we should see a less immediate use of VOT for lexical activation: The target‐over‐competitor fixation advantage should diverge only when the F0 cue kicks in. Under the independent view, however, listeners would start integrating VOT as soon as it is available in the speech signal. Thus, we should find an immediate integration of VOT: Listeners should immediately start looking at the target over the competitor before F0 arrives. Since F0 is a perceptually more informative cue to distinguish lenis stops from fortis stops, this condition would also be a good case to test the relationship between cue integration and relative cue weighting.</p> <p>In addition to the within‐condition analyses, GAMMs compared the target‐over‐competitor fixation advantage of the last two conditions to examine whether the onset of the VOT effect in one condition is significantly different from that in the other condition. For the cross‐condition analyses, we fitted the binary smooth specification that compares the smooths of the two target‐competitor types (Appendix C). The random effects specification was identical to the within‐condition analyses.</p> <p>For the individual‐level analysis, GAMM was conducted to quantify each listener's time point of VOT integration (henceforth referred to as the VOT integration index). For each model, the dependent variable was the target‐over‐competitor fixation advantage, and the predictors included nonlinear patterns over time and a binary smooth that models the constant difference between the nonlinear pattern of the fixation advantage of a target‐competitor type and 0 (i.e., no target‐over‐competitor advantage). Like the group‐level analysis, the time point of the onset of the VOT effect for each listener was calculated using the <emph>plot_diff</emph> function of the <emph>itsadug</emph> package in R. Individual listeners' VOT integration indices were then square‐root transformed. Pearson's product‐moment correlation analyses were conducted to examine the relationship between the individual's VOT integration index and the individual's weight of each acoustic dimension. The lower the VOT integration index, the earlier the integration of the VOT cue in spoken word recognition.</p> <p>If individual listeners' cue weighting is associated with their real‐time processing of acoustic information, listeners who rely more on VOT (available earlier) in perception would be more likely to show lower VOT integration indices (earlier VOT integration): These listeners would start looking at the target over the competitor as soon as the VOT effect kicks in. On the other hand, listeners who rely more on the F0 cue (available later) would be more likely to show higher VOT integration indices (delayed VOT integration): These listeners would be less likely to show a target‐over‐competitor fixation advantage until the F0 cue kicks in. These predictions would be evidenced by an inverse relationship between an individual's reliance on VOT in perception and an individual's VOT integration index in spoken word recognition, and by a significant positive relationship between the reliance on F0 in perception and the VOT integration index in spoken word recognition across the three target‐competitor types.</p> <p>Alternatively, listeners with greater sensitivity to secondary cues to the contrast could use all available acoustic information signaling the contrast (e.g., Ou et al., [<reflink idref="bib58" id="ref103">58</reflink>]), showing higher VOT integration indices. In this scenario, we expect to see results that go in the opposite direction for the condition with an aspirated stop target and a lenis stop competitor and for the condition with a lenis stop target and a fortis stop competitor, as VOT is the secondary cue for these conditions. That is, we should see a positive relationship between an individual's weighting of VOT and an individual's VOT integration index for those two target‐competitor types, which is different from the first set of predictions.</p> <p>Additionally, we examined whether listeners' timing of VOT integration is correlated across target‐competitor conditions. This cross‐condition comparison is to test whether the relationship between cue weighting and the time course of cue integration, if any, is further modulated by listeners' inherent processing strategy. For example, if we find that listeners who rely more on the cue available later to the contrast show a delayed integration of VOT, it is possible that listeners who rely more on the acoustic dimension that arrives later are individuals who inherently integrate cues more slowly than others. This cross‐condition comparison will thus clarify whether listeners use inherent processing strategies that affect their cue integration across the board and independently of cue weighting.</p> <hd id="AN0181921441-21">Results</hd> <p></p> <hd id="AN0181921441-22">Group results</hd> <p>The left column of Fig. 3 shows listeners' target‐over‐competitor fixation advantage for each target‐competitor type. For the condition with an aspirated target and a fortis competitor (Fig. 3A), a solid black vertical line on the 0 ms time point indicates the onset of the VOT as well as the VOT disambiguation point. Recall that the aspirated target had 110 ms of VOT, and (bilabial) fortis stops have a short lag of VOT. Thus, listeners should start disambiguating the target from the competitor as soon as the VOT interval begins in the speech signal, with the onset of VOT in the speech signal more or less coinciding with the onset of disambiguation based on VOT. We examined listeners' target‐over‐competitor fixation advantage with a delay of 200 ms from the VOT disambiguation point (dashed blue vertical line), as it takes approximately 200 ms for eye movements to reflect speech processing (Hallett, [<reflink idref="bib21" id="ref104">21</reflink>]; Salverda, Kleinschmidt, &amp; Tanenhaus, [<reflink idref="bib61" id="ref105">61</reflink>]). The corrected F0 onset is marked with an orange dashed vertical line at 310 ms from the onset of the VOT disambiguation point, which is 110 ms after the corrected VOT disambiguation point. Visual inspection showed that listeners looked more at the aspirated target than at the fortis competitor as soon as VOT became available in the speech signal.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CGN/01dec24/cogs70026-fig-0003.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="cogs70026-fig-0003.jpg" title="3 Listeners' target‐over‐competitor fixation advantage in (A) the condition with an aspirated target and a fortis competitor, (B) the condition with an aspirated target and a lenis competitor, and (C) the condition with a lenis target and a fortis competitor. Nonlinear smooths for (D) the condition with an aspirated target and a fortis competitor, (E) the condition with an aspirated target and a lenis competitor, and (F) the condition with a lenis target and a fortis competitor. For the plots showing each model, the point and interval where the shaded pointwise 95%‐confidence interval does not overlap with the null effect (horizontal black dotted line) are indicated by a vertical red dotted line and a solid red line on the x‐axis, respectively. For each plot, the vertical blue dashed line represents the corrected 200 ms oculomotor delay for the VOT disambiguation point. The vertical orange dashed line represents the corrected F0 onset. The vertical black dashed line represents the corrected onset of the second syllable." /> </p> <p></p> <p>Table 4A summarizes the GAMM conducted on the target‐over‐competitor proportions of fixation in the condition with an aspirated target and a fortis competitor. Since the main interest of the study is to examine whether listeners start integrating VOT as soon as it is available in the speech signal or delay integrating it until F0 arrives, the statistical analyses were limited to the time window ranging from the VOT disambiguation point in the speech signal (0 ms) to the corrected F0 onset; thus, the time window for analyzing the condition with an aspirated target and a fortis competitor was 0–310 ms. The first line of Table 4A, s(Time), represents the smooths of the target‐competitor condition. The <emph>edf</emph> (effective degrees of freedom) values of the fixed effects (first two lines) are indicative of the amount of nonlinearity of the smooths. If the value is close to 1, it means that the pattern is close to linear, and if it is close to 10, it indicates that the pattern is more complex (nonlinear). The <emph>p</emph>‐value associated with the smooth indicates whether the smooth is significantly different from 0. The second line of Table 4A, s(Time): Difference between the AF and 0, reveals that the nonlinear patterns of the target‐over‐competitor fixation advantage are significantly different from 0. In other words, before the F0 arrives in the speech signal (before 310 ms), listeners showed a significant degree of target‐over‐competitor fixation advantage, meaning that they started integrating VOT. The <emph>p</emph>‐value associated with the random effects (i.e., the last four rows of Table 4A) shows whether including random effects is necessary to improve the model fit. From the initial models, a by‐subject random intercept was excluded, as it did not significantly improve the model.</p> <p>4 Table Summary of smoothing function terms of the generalized additive mixed model</p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th align="left"&gt;A. The difference between target&amp;#8208;over&amp;#8208;competitor fixation proportions in the condition with an aspirated target and a fortis competitor (AF) and 0 (null effect).&lt;/th&gt;&lt;/tr&gt;&lt;tr&gt;&lt;th&gt;Smooth Functions&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;edf&lt;/italic&gt;&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;F&lt;/italic&gt;&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;p&lt;/italic&gt;&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;s(Time)&lt;/td&gt;&lt;td&gt;3.600&lt;/td&gt;&lt;td&gt;14.791&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time): Difference between the AF and 0&lt;/td&gt;&lt;td&gt;4.681&lt;/td&gt;&lt;td&gt;72.084&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Item)&lt;/td&gt;&lt;td&gt;1.971&lt;/td&gt;&lt;td&gt;0.983&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(TC Type, Participant)&lt;/td&gt;&lt;td&gt;117.548&lt;/td&gt;&lt;td&gt;26.730&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time, Participant)&lt;/td&gt;&lt;td&gt;264.523&lt;/td&gt;&lt;td&gt;4.544&lt;/td&gt;&lt;td&gt;.004***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time, Item)&lt;/td&gt;&lt;td&gt;19.355&lt;/td&gt;&lt;td&gt;61.012&lt;/td&gt;&lt;td&gt;.089&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p></p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th align="left"&gt;B. The difference between target&amp;#8208;over&amp;#8208;competitor fixation proportions in the condition with an aspirated target and a lenis competitor (AL) and 0 (null effect).&lt;/th&gt;&lt;/tr&gt;&lt;tr&gt;&lt;th&gt;Smooth Functions&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;edf&lt;/italic&gt;&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;F&lt;/italic&gt;&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;p&lt;/italic&gt;&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;s(Time)&lt;/td&gt;&lt;td&gt;3.234&lt;/td&gt;&lt;td&gt;4.799&lt;/td&gt;&lt;td&gt;.002**&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time): Difference between the AL and 0&lt;/td&gt;&lt;td&gt;3.879&lt;/td&gt;&lt;td&gt;8.993&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Item)&lt;/td&gt;&lt;td&gt;1.940&lt;/td&gt;&lt;td&gt;0.960&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(TC Type, Participant)&lt;/td&gt;&lt;td&gt;119.698&lt;/td&gt;&lt;td&gt;55.257&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time, Participant)&lt;/td&gt;&lt;td&gt;212.384&lt;/td&gt;&lt;td&gt;2.014&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time, Item)&lt;/td&gt;&lt;td&gt;16.216&lt;/td&gt;&lt;td&gt;16.856&lt;/td&gt;&lt;td&gt;.084&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p></p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th align="left"&gt;C. The difference between target&amp;#8208;over&amp;#8208;competitor fixation proportions in the condition with a lenis target and a fortis competitor (LF) and 0 (null effect).&lt;/th&gt;&lt;/tr&gt;&lt;tr&gt;&lt;th&gt;Smooth Functions&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;edf&lt;/italic&gt;&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;F&lt;/italic&gt;&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;p&lt;/italic&gt;&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;s(Time)&lt;/td&gt;&lt;td&gt;1.837&lt;/td&gt;&lt;td&gt;2.029&lt;/td&gt;&lt;td&gt;.121&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time): Difference between the LF and 0&lt;/td&gt;&lt;td&gt;3.040&lt;/td&gt;&lt;td&gt;4.055&lt;/td&gt;&lt;td&gt;.004**&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Item)&lt;/td&gt;&lt;td&gt;1.946&lt;/td&gt;&lt;td&gt;0.950&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(TC Type, Participant)&lt;/td&gt;&lt;td&gt;115.422&lt;/td&gt;&lt;td&gt;37.653&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time, Participant)&lt;/td&gt;&lt;td&gt;203.001&lt;/td&gt;&lt;td&gt;2.117&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time, Item)&lt;/td&gt;&lt;td&gt;18.384&lt;/td&gt;&lt;td&gt;19.500&lt;/td&gt;&lt;td&gt;.049*&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <ulist> <item>3 <emph>Notes</emph>. TC represents Target and Competitor. For each summary, the last four rows show the random effects structure of the model.</item> <item>4 *** &lt; .001, ** &lt; .01, * &lt; .05.</item> </ulist> <p>The right column of Fig. 3 visualizes the associated binary difference smooths, corroborating the model summary. Using the <emph>plot_diff</emph> function, the time point where the pointwise 95%‐confidence interval does not overlap with 0 (no target‐over‐competitor fixation advantage) was calculated. We took this time point as the onset of the VOT effect. The difference smooths revealed that the VOT effect emerged at 209.1 ms from the VOT onset/disambiguation time point. As shown in Fig. 3D, the onset of the VOT effect was close to the corrected VOT disambiguation point (200 ms), which was much earlier than the correct onset of F0 in the speech signal (310 ms).</p> <p>For the condition with an aspirated stop target and a lenis stop competitor, recall that the aspirated target had 110 ms of VOT, and lenis stops have approximately 55 ms of VOT. Thus, listeners would be able to use VOT to distinguish the aspirated stop target from the lenis competitor after approximately 55 ms of VOT. Hence, the VOT disambiguation point does not match the onset of VOT in the speech signal. For this condition (Fig. 3B), listeners' integration of VOT seems to be somewhat delayed until the F0 becomes available in the speech signal.</p> <p>As shown in Table 4B, s(Time): Difference between the AL and 0, the model revealed that the nonlinear patterns of the target‐over‐competitor fixation advantage shown in the condition with an aspirated target and a lenis competitor are significantly different from 0. In other words, before the F0 arrives in the speech signal (before 255 ms), listeners showed a significant target‐over‐competitor advantage, suggesting that they start integrating VOT before F0 is available. Notably, as shown in Fig. 3E, the difference smooths revealed that the VOT effect emerged at 247.5 ms from the VOT disambiguation time point, which is far from the corrected VOT disambiguation point (200 ms), and slightly earlier than the corrected onset of F0 in the speech signal (255 ms).</p> <p>Listeners' target‐over‐competitor fixation advantage in the condition with a lenis stop target and a fortis stop competitor is shown in Fig. 3C. Since the lenis target had 55 ms of VOT and fortis stops have a short lag of VOT, listeners should start disambiguating the target from the competitor as soon as VOT is available in the speech signal, with the onset of VOT in the speech signal corresponding to the VOT disambiguation time point. Like the condition with an aspirated target and lenis competitor, a visual inspection shows that listeners' use of VOT seems to be slightly delayed until the F0 arrives in the speech signal.</p> <p>Table 4C summarizes the results of the GAMM analysis for the condition with the aspirated target and fortis competitor, focusing on the time window from 0 ms (VOT disambiguation point in the speech signal) to 255 ms (corrected onset of F0). The second line of the table, s(Time): Difference between the AF and 0, shows significant nonlinear patterns of the target‐over‐competitor fixation advantage in the condition with a lenis target and a fortis competitor. In other words, listeners start distinguishing between the target and competitor as soon as VOT is available and before F0 arrives in the speech signal. Notably, the difference smooths (Fig. 3F) revealed that the VOT effect emerged at 245.5 ms from the VOT onset/disambiguation time point. The onset of the VOT effect was not close to the corrected VOT onset/disambiguation point (200 ms), but slightly earlier than the corrected onset of F0 in the speech signal (255 ms).</p> <p>We next examined whether the timing of the VOT effect observed in one condition is significantly different from that of the others. Table 5A summarizes the smoothing terms of the model comparing the condition with an aspirated target and lenis competitor and the condition with an aspirated target and fortis competitor, focusing on the time window from 0 to 255 ms. Notably, the second line, s(Time): Difference between AF and AL, reveals a significant difference between the aspirated target with a fortis competitor and the aspirated target with a lenis competitor. This means that, before the F0 arrives in the speech signal, listeners showed a greater target‐over‐competitor advantage when the aspirated stop target was paired with a fortis stop competitor than when it was paired with a lenis stop competitor.</p> <p>5 Table Summary of smoothing function terms of the generalized additive mixed model on the difference between target‐over‐competitor fixation proportions in two conditions</p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th align="left"&gt;A. Difference between the condition with an aspirated target and a lenis competitor (AL) and the condition with an aspirated target and a fortis competitor (AF).&lt;/th&gt;&lt;/tr&gt;&lt;tr&gt;&lt;th&gt;Smooth Functions (SFs)&lt;/th&gt;&lt;th align="center"&gt;edf&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;F&lt;/italic&gt;&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;p&lt;/italic&gt;&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;s(Time)&lt;/td&gt;&lt;td&gt;3.045&lt;/td&gt;&lt;td&gt;7.783&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time): Difference between AF and AL&lt;/td&gt;&lt;td&gt;2.252&lt;/td&gt;&lt;td&gt;5.642&lt;/td&gt;&lt;td&gt;.001**&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Item)&lt;/td&gt;&lt;td&gt;1.909&lt;/td&gt;&lt;td&gt;0.940&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(TC Type, Participant)&lt;/td&gt;&lt;td&gt;116.344&lt;/td&gt;&lt;td&gt;33.603&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time, Participant)&lt;/td&gt;&lt;td&gt;220.961&lt;/td&gt;&lt;td&gt;2.403&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time, Item)&lt;/td&gt;&lt;td&gt;11.378&lt;/td&gt;&lt;td&gt;6.975&lt;/td&gt;&lt;td&gt;.097&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p></p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th align="left"&gt;B. Difference between the condition with an aspirated target and a lenis competitor (AL) and the condition with a lenis target and a fortis competitor (LF).&lt;/th&gt;&lt;/tr&gt;&lt;tr&gt;&lt;th&gt;Smooth Functions (SFs)&lt;/th&gt;&lt;th align="center"&gt;edf&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;F&lt;/italic&gt;&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;p&lt;/italic&gt;&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;s(Time)&lt;/td&gt;&lt;td&gt;2.900&lt;/td&gt;&lt;td&gt;2.016&lt;/td&gt;&lt;td&gt;.146&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time): Difference between AL and LF&lt;/td&gt;&lt;td&gt;2.001&lt;/td&gt;&lt;td&gt;0.234&lt;/td&gt;&lt;td&gt;.791&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Item)&lt;/td&gt;&lt;td&gt;3.881&lt;/td&gt;&lt;td&gt;0.962&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(TC Type, Participant)&lt;/td&gt;&lt;td&gt;119.7&lt;/td&gt;&lt;td&gt;52.934&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time, Participant)&lt;/td&gt;&lt;td&gt;212.4&lt;/td&gt;&lt;td&gt;2.260&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time, Item)&lt;/td&gt;&lt;td&gt;35.68&lt;/td&gt;&lt;td&gt;19.110&lt;/td&gt;&lt;td&gt;.025*&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p></p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th align="left"&gt;C. Difference between the condition with a lenis target and a fortis competitor (LF) and the condition with an aspirated target and a fortis competitor (AF).&lt;/th&gt;&lt;/tr&gt;&lt;tr&gt;&lt;th&gt;Smooth Functions (SFs)&lt;/th&gt;&lt;th align="center"&gt;edf&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;F&lt;/italic&gt;&lt;/th&gt;&lt;th align="center"&gt;&lt;italic&gt;p&lt;/italic&gt;&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;s(Time)&lt;/td&gt;&lt;td&gt;2.423&lt;/td&gt;&lt;td&gt;4.136&lt;/td&gt;&lt;td&gt;.010**&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time): Difference between LF and AF&lt;/td&gt;&lt;td&gt;2.001&lt;/td&gt;&lt;td&gt;1.063&lt;/td&gt;&lt;td&gt;.346&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Item)&lt;/td&gt;&lt;td&gt;3.922&lt;/td&gt;&lt;td&gt;0.972&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(TC Type, Participant)&lt;/td&gt;&lt;td&gt;107.078&lt;/td&gt;&lt;td&gt;21.471&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time, Participant)&lt;/td&gt;&lt;td&gt;221.803&lt;/td&gt;&lt;td&gt;4.410&lt;/td&gt;&lt;td&gt;&amp;#60;&amp;#160;.001***&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;s(Time, Item)&lt;/td&gt;&lt;td&gt;34.347&lt;/td&gt;&lt;td&gt;35.781&lt;/td&gt;&lt;td&gt;.034*&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <ulist> <item>5 <emph>Notes</emph>. TC represents Target and Competitor. For each summary, the last four rows show the random effects structure of the model.</item> <item>6 *** &lt; .001, ** &lt; .01, * &lt; .05.</item> </ulist> <p>Table 5B summarizes the smoothing terms of the model comparing the condition with the aspirated target and lenis competitor and the condition with the lenis target and fortis competitor, focusing on the time window ranging from 0 to 255 ms. As shown in the second line of the model, listeners did not show different degrees of target‐over‐competitor advantage between the two target‐competitor types before the F0 arrived in the speech signal.</p> <p>Table 5C summarizes the smoothing terms of the model comparing the condition with an aspirated target and a fortis competitor with the condition with a lenis target and a fortis competitor, focusing on the time window from 0 to 255 ms. The second line of the model reveals that the difference between the two conditions is not significant. That is, before the onset of F0 in the speech signal, listeners showed comparable degrees of target‐over‐competitor advantage between the two target‐competitor types.</p> <hd id="AN0181921441-24">Individual results</hd> <p>We first examined whether the timing of VOT integration among individual listeners was consistent across the different target‐competitor types. The results showed that individual listeners' timing of VOT integration in one target‐competitor type is not correlated with that of other types (correlation between AL and AF: <emph>r</emph> = .16, <emph>p</emph> &gt; .1; between AL and LF: <emph>r</emph> = .04, <emph>p</emph> &gt; .1; between AF and LF: <emph>r</emph> = .12, <emph>p</emph> &gt; .1). Thus, it is less likely that some listeners consistently show an early or late integration of VOT across target‐competitor types. This cross‐type comparison suggests that whatever the relationship between cue weighting and cue integration looks like, it is unlikely to be due to listeners' use of an inherent processing strategy across target‐competitor types.</p> <p>We then examined whether individual listeners' cue integration is correlated with their cue reliance measured by the cue‐weighting task, which is a primary focus of our study (see Appendix D for the specific values of individual listeners' VOT integration index). Fig. 4A presents individual listeners' VOT integration index in the condition with an aspirated target and a fortis competitor as a function of their cue reliance for distinguishing the aspirated versus fortis contrast in the cue‐weighting task. Individual listeners' VOT integration index was positively correlated with their perceptual reliance on F0 (<emph>r</emph> = .26, <emph>p</emph> &lt; .05), but not with the reliance on VOT. This relationship suggests that listeners who rely more on F0 (secondary cue) to distinguish aspirated stops from fortis stops show a more delayed integration of VOT in lexical activation.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CGN/01dec24/cogs70026-fig-0004.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="cogs70026-fig-0004.jpg" title="4 Relationship between listeners' VOT integration indices and their cue reliance in the corresponding contrast type." /> </p> <p></p> <p>Fig. 4B shows individual listeners' VOT integration indices in the condition with an aspirated target with a lenis competitor as a function of their cue reliance to distinguish the aspirated versus lenis contrast in the cue‐weighting task. As shown in the left panel, individual listeners' VOT integration index was inversely correlated with their perceptual reliance on VOT (<emph>r</emph> = −.32, <emph>p</emph> &lt; .05), indicating that listeners who relied more on VOT (secondary cue) to the contrast showed an earlier integration of VOT in lexical activation. There was no significant correlation between the timing of the VOT effect and listeners' perceptual reliance on F0 (Fig. 4B, right panel).</p> <p>Individual listeners' VOT integration indices in the condition with a lenis target and a fortis competitor (Fig. 4C) were negatively correlated with their perceptual reliance on VOT (<emph>r</emph> = −.45, <emph>p</emph> &lt; .05) and were marginally positively correlated with their reliance on F0 (<emph>r</emph> = .25, <emph>p</emph> = .05). These results suggest that listeners who rely more on VOT to distinguish lenis stops from fortis stops tend to show a faster integration of VOT, and those who rely more on F0 to the corresponding contrast tend to show a more delayed integration of VOT.</p> <p>We subsequently looked at the relationship between individual listeners' relative cue weighting and the timing of VOT integration (Fig. 5). For the relationship between the timing of the VOT effect in the condition with an aspirated target and a fortis competitor and the relative cue weighting to the aspirated versus fortis contrast, there was a moderate positive correlation between the two variables (<emph>r</emph> = .25, <emph>p</emph> = .05), indicating that listeners who rely relatively more on F0 than VOT to perceive the aspirated versus fortis contrast tend to show a more delayed integration of VOT. The relationship between the timing of the VOT effect in the condition with an aspirated target and a lenis competitor and the relative cue weighting in the aspirated versus lenis contrast showed that there was a strong positive correlation between the two variables (<emph>r</emph> = .3, <emph>p</emph> &lt; .05), suggesting again that listeners who rely more on F0 than on VOT for the aspirated‐lenis contrast tended to delay their integration of VOT. The relationship between the timing of the VOT effect in the condition with a lenis target and a fortis competitor and the relative cue weighting in the lenis versus fortis contrast showed a similar pattern. Again, there was a significant positive correlation between the two variables (<emph>r</emph> = .5, <emph>p</emph> &lt; .05), indicating that listeners who rely relatively more on F0 than on VOT to the lenis versus fortis contrast tend to show a more delayed integration of VOT.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CGN/01dec24/cogs70026-fig-0005.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="cogs70026-fig-0005.jpg" title="5 Listeners' VOT integration indices as a function of their relative cue weighting. Relative cue weighting represents the ratio of the F0 coefficients to those of VOT. A value of 0.5 indicates equal reliance on both cues. Values above 0.5 show a preference for the F0 cue, while those below 0.5 indicate greater reliance on the VOT cue." /> </p> <p></p> <hd id="AN0181921441-27">Discussion</hd> <p>The main goal of Experiment 2 was to examine the time course of acoustic cue integration to the Korean stop contrasts and whether cue reliance predicts the timing of cue integration at the individual level. The main findings are summarized in Table 6.</p> <p>6 Table Summary of the main findings</p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th align="left" /&gt;&lt;th align="center"&gt;Target&amp;#8208;competitor type&lt;/th&gt;&lt;/tr&gt;&lt;tr&gt;&lt;th align="center"&gt;Aspirated&amp;#8208;Fortis&lt;/th&gt;&lt;th align="center"&gt;Aspirated&amp;#8208;Lenis&lt;/th&gt;&lt;th align="center"&gt;Lenis&amp;#8208;Fortis&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;Timing of VOT integration&lt;/td&gt;&lt;td&gt;209.1 ms&lt;/td&gt;&lt;td&gt;247.5 ms&lt;/td&gt;&lt;td&gt;245.5 ms&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;Relationship with VOT reliance&lt;/td&gt;&lt;td&gt;&amp;#8212;&lt;/td&gt;&lt;td&gt;Negative&lt;/td&gt;&lt;td&gt;Negative&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;Relationship with F0 reliance&lt;/td&gt;&lt;td&gt;Positive&lt;/td&gt;&lt;td&gt;&amp;#8212;&lt;/td&gt;&lt;td&gt;Positive&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;Relationship with the relative cue weighting (F0/VOT)&lt;/td&gt;&lt;td&gt;Positive&lt;/td&gt;&lt;td&gt;Positive&lt;/td&gt;&lt;td&gt;Positive&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p>For the condition with an aspirated target and a fortis competitor, where VOT is a primary cue to the corresponding contrast, listeners showed immediate integration of the cue available earlier (VOT) in the speech signal, as evidenced by an early increase of the target‐over‐competitor fixation advantage. Listeners immediately started distinguishing the target and competitor based solely on the early acoustic information (VOT), approximately 9 ms after the (corrected) onset of VOT disambiguation in the speech signal. Although this target‐competitor type is not a good case to test our hypotheses (the associated vs. independent view), it provides a baseline for understanding the timing of cue integration compared to the other target‐competitor types. Notably, this type of contrast demonstrates that the time course of cue integration is immediate and continuous, in line with previous findings on the time course of cue integration (e.g., McMurray et al., [<reflink idref="bib44" id="ref106">44</reflink>]; Reinisch &amp; Sjerps, [<reflink idref="bib60" id="ref107">60</reflink>]).</p> <p>More importantly, when processing the same aspirated target but with a lenis competitor, where F0 is a primary cue to the contrast, listeners showed a delayed integration of the early (VOT) cue. Listeners started distinguishing the target and competitor at approximately 48 ms after the (corrected) onset of VOT disambiguation, closer to the onset of F0 in the speech signal. This delayed VOT integration suggests that listeners' integration of asynchronous acoustic cues is associated with how they perceptually weight cues to the contrast, supporting the associated approach to the cue integration mechanism. It is important to note that this does not mean that listeners necessarily hold the early acoustic information in a buffer until sufficient information is available. Although the initiation of lexical activation is somewhat delayed, evidence from the GAMM model (Table 4B) indicates that the timing of VOT integration is nonetheless earlier than the onset of F0 in our group results. Thus, while further validation is needed due to the temporal proximity of these effects, which makes it challenging to draw firm conclusions, this timing difference suggests that early acoustic information may not be fully buffered. Instead, it is likely that listeners integrate acoustic cues continuously to activate lexical items, with the strength of activation mediated by how listeners weight these asynchronously available cues.</p> <p>Listeners' VOT integration in the condition with a lenis target and a fortis competitor was also in line with the prediction of the associated view. Listeners' integration of VOT was delayed until approximately 46 ms after the onset of VOT disambiguation in the speech signal. These results indicate that the time point of VOT integration is associated with the heavier perceptual weight of the cue available later in the speech signal. Overall, listeners clearly demonstrated that their integration of VOT is delayed when processing the target‐competitor type in which the cue available later plays a more critical role in distinguishing the stop contrasts, supporting the associated view of cue integration.</p> <p>Crucially, we found a significant relationship between cue integration and cue weighting at the individual level. Listeners' relative cue weighting was positively correlated with the timing of VOT integration across target‐competitor types, indicating that listeners who rely more on the cue available later tend to show more delayed integration of the early acoustic information. Interestingly, this relationship was associated with listeners' reliance on the secondary cue to the contrasts. For the condition with an aspirated target and a fortis competitor, listeners who relied more on F0 (secondary cue) showed a more delayed integration of VOT. For the other conditions, listeners who relied more on VOT (secondary cue) showed a more delayed integration of VOT. These results suggest that the timing of cue integration is strongly associated with individual listeners' use of the secondary cue, but not with the primary cue.</p> <p>The results of the condition with an aspirated target and a fortis competitor are in line with the findings of Ou et al. ([<reflink idref="bib58" id="ref108">58</reflink>]), who showed the link between listeners' categorization gradiency and cue integration strategy. Categorization gradiency is related to listeners' reliance on the secondary cue: More gradient listeners rely more on the secondary cue (Kapnoula et al., [<reflink idref="bib31" id="ref109">31</reflink>]; Kim et al., [<reflink idref="bib32" id="ref110">32</reflink>]; Ou et al., [<reflink idref="bib58" id="ref111">58</reflink>], but see also Kapnoula et al., [<reflink idref="bib29" id="ref112">29</reflink>]). While our study did not directly measure listeners' categorization gradiency, listeners who showed a stronger reliance on F0 to the aspirated versus fortis contrast could be more gradient listeners. However, it is uncertain whether this interpretation can be generalized to other conditions where the cue available earlier is the secondary cue, as the relationship between categorization gradiency and secondary cue has been examined in cases where the secondary cue is available later in the speech signal.</p> <hd id="AN0181921441-28">General discussion</hd> <p>We found evidence that the timing of cue integration by listeners, as cues become available sequentially, is associated with their perceptual weighting of those cues. When listeners process the condition with an aspirated stop target and a fortis competitor, they integrate VOT (a more reliable cue to the aspirated vs. fortis contrast that is available earlier) in lexical activation before they integrate F0 (a less reliable cue to the aspirated vs. fortis contrast that is available later). However, listeners did not adopt a similar integration strategy for other contrasts: For the condition with an aspirated stop target and a lenis competitor, they did not immediately integrate VOT (available first but less reliable for perceiving the aspirated vs. lenis contrast). A similar pattern of cue integration was found for the contrast where the lenis stop target competed with a fortis stop. In these two conditions, listeners did not integrate VOT (less reliable but available first) in lexical activation until F0 (more reliable but available later) became available in the speech signal.</p> <p>Under the assumption that acoustic information maps continuously onto lexical candidates, we hypothesized two views as to the role of perceptual cue primacy in the time course of cue integration. The associated view predicted that cue integration in real time would be adjusted by listeners' perceptual cue primacy, whereas the independent view predicted that the timing of cue integration would not be influenced by listeners' perceptual weighting. Our results support the associated view of cue integration, in that the real‐time integration of the cue available earlier in the speech signal was affected by the perceptual importance of the cue available later in the speech signal.</p> <p>From a theoretical perspective, our findings elaborate more on the nature of the cue integration mechanism used in spoken word recognition. Studies on cue integration have provided evidence in line with the immediate integration of acoustic information in stops, approximants (McMurray et al., [<reflink idref="bib44" id="ref113">44</reflink>]), and vowels (Reinisch &amp; Sjerps, [<reflink idref="bib60" id="ref114">60</reflink>]), suggesting that lexical activation does not wait for cue integration to complete, and the preliminary or partial information available in the signal immediately influences lexical activation. Our results for the condition where F0 is not as informative as VOT (an aspirated target and a fortis competitor) are in line with the continuous integration account, as listeners start mapping early acoustic information (VOT) onto lexical representations before the late acoustic information (F0) arrives in the speech signal. Notably, we demonstrated that this may not be the full story by showing that listeners' continuous incremental mapping of asynchronous acoustic information onto lexical items is further modulated by the informativeness of cues.</p> <p>While existing research has made significant contributions to our understanding of the <emph>outcome</emph> of speech categorization, only a few studies have examined the moment‐by‐moment <emph>process</emph> of categorization (McMurray et al., [<reflink idref="bib44" id="ref115">44</reflink>]; Ou et al., [<reflink idref="bib58" id="ref116">58</reflink>]; Reinisch &amp; Sjerps, [<reflink idref="bib60" id="ref117">60</reflink>]; Toscano &amp; McMurray, [<reflink idref="bib69" id="ref118">69</reflink>]) (see also Kingston, Levy, Rysling, &amp; Staub, [<reflink idref="bib36" id="ref119">36</reflink>]; Mitterer &amp; Reinisch, [<reflink idref="bib54" id="ref120">54</reflink>]). In most studies, listeners activate lexical items immediately and incrementally after hearing a single cue, updating lexical activation as additional information becomes available. Despite these findings, the role of cue weighting in the real‐time processing of speech remained unclear, as earlier studies had only examined cases where the early acoustic information was also relatively more informative for perceiving the contrast. Our findings fill this gap and provide an updated view of understanding cue integration mechanisms in spoken word recognition.</p> <p>The individual‐level results further clarify the relationship between cue weighting and the time course of cue integration. Individual listeners' timing of cue integration was significantly correlated with their relative weighting of acoustic cues, indicating that a listener who relies more on F0 than VOT shows a more delayed timing of VOT integration across contrast types. These findings suggest that there may be some listeners who take a less continuous integration strategy than others, but importantly, this strategy is unlikely to be attributed to the processing system itself, but rather to individuals' perceptual reliance on acoustic information.</p> <p>The current individual‐difference findings are consistent with those of Ou et al. ([<reflink idref="bib58" id="ref121">58</reflink>]) in which they found that individual differences in the time course of cue integration are associated with variability in categorization gradience, particularly when listeners process the tense versus lax vowel contrast in English. That is, listeners who paid more attention to the cue available later in time tended to integrate cues less continuously, suggesting that cue weighting reflects the processing strategy adopted by the individual. It is also worth noting that they did not find a significant relationship between categorical gradiency and the timing of cue integration for the stop‐voicing contrast, though the general pattern of results was consistent with that of the vowel contrast.</p> <p>This slightly different finding between their study and ours might be due to the different informativeness of F0 as a cue to the stop contrast between English and Korean. Among English listeners, some relied more heavily on F0 to distinguish voiced stops from voiceless stops than others. However, this does not necessarily mean that some individual listeners treated F0 as being more important than VOT for perceiving the contrast. In contrast, Korean listeners tended to place more weight on F0 than VOT when perceiving their L1 stop contrast, depending on the type of contrast they heard. As a result, F0 appeared to play a more crucial role in cueing the contrast for Korean listeners. This could explain why the relationship between cue weighting and the timing of cue integration at the individual level was more evident for Korean listeners than for English listeners.</p> <p>More broadly, our results add empirical evidence to the claim that cue integration may not be strictly incremental and immediate but show an idiosyncratic pattern depending on different types of phonological contrast (Galle et al., [<reflink idref="bib18" id="ref122">18</reflink>]; Schreiber &amp; McMurray, [<reflink idref="bib63" id="ref123">63</reflink>]) and individual listeners' categorization gradiency (Ou et al., [<reflink idref="bib58" id="ref124">58</reflink>]).[<reflink idref="bib4" id="ref125">4</reflink>] For instance, Galle et al. ([<reflink idref="bib18" id="ref126">18</reflink>]) examined the time course of cue integration in the processing of place contrasts in English sibilant fricatives, where the contrast is primarily signaled by the early spectral mean of the frication and secondarily by the later formant transitions (Jongman, Wayland, &amp; Wong, [<reflink idref="bib26" id="ref127">26</reflink>]). The results of a series of VWP consistently showed that lexical activation was delayed until the formant transition cue became available. Listeners used the frication spectrum simultaneously with the formant transitions, providing clear evidence for a buffered integration.</p> <p>Interestingly, the buffered integration in sibilant fricatives cannot be interpreted within the associated view. Since the cue available earlier in the speech signal (the frication spectrum) is a primary cue to the contrast, the associated view would predict immediate integration of the frication spectrum cue, which is not consistent with the empirical findings. One explanation for these results is that contextual factors such as talker and coarticulation influence the acoustic form in the fricative contrast to a greater extent than in the voicing contrast (Apfelbaum, Bullock‐Rest, Rhone, Jongman, &amp; McMurray, [<reflink idref="bib3" id="ref128">3</reflink>]; McMurray &amp; Jongman, [<reflink idref="bib46" id="ref129">46</reflink>]). The talker and vowel may not be clearly identified in the fricative, and the relevant information may only appear at the vocoid. Thus, a less continuous manner of cue integration may be needed until contextual information in the vocoid arrives (see, Galle et al., [<reflink idref="bib18" id="ref130">18</reflink>]; Schreiber &amp; McMurray, [<reflink idref="bib63" id="ref131">63</reflink>], for a longer discussion). This strategy might be useful in that it can help listeners avoid overcommitting and maintain competitor activation in case they were wrong and need to revise their initial interpretation (McMurray, Farris‐Trimble, &amp; Rigler, [<reflink idref="bib45" id="ref132">45</reflink>]; McMurray, Tanenhaus, &amp; Aslin, [<reflink idref="bib49" id="ref133">49</reflink>]) (see also Clopper &amp; Walker, [<reflink idref="bib12" id="ref134">12</reflink>]).</p> <p>Our study has placed a fairly narrow focus on the role of cue primacy in the time course of cue integration. Nevertheless, the current findings, in conjunction with prior research on buffered integration, have important theoretical implications for future research, as these collectively suggest that the immediate and continuous approach to cue integration may need to be qualified by an account of listeners' perceptual cue weighting to phonological contrasts. The dynamics of cue integration for higher‐level processes may exhibit more flexibility than commonly assumed (e.g., Ou et al., [<reflink idref="bib58" id="ref135">58</reflink>]). As McMurray et al. ([<reflink idref="bib44" id="ref136">44</reflink>]) proposed, buffered integration can contribute to mitigating the risk of committing to an incorrect lexical representation by integrating pertinent acoustic information at the sublexical level. In contrast, continuous integration allows for more expeditious access to lexical representation, but with an increased likelihood of preliminary commitments being less accurate.</p> <p>Although empirical evidence has largely ruled out the extreme version of buffered integration, a less dichotomous approach—one that postulates a sublexical unit that can selectively retain bottom‐up information before making preliminary commitments—may provide a more nuanced account of cue integration mechanisms. Our individual results corroborate this less dichotomous approach, suggesting that multiple sources of information contribute to the moment‐by‐moment processes that underlie spoken word recognition, albeit with potentially less continuous integration dynamics contingent upon the reliability of these sources.</p> <p>To broaden our understanding of cue integration mechanisms, future research could examine the real‐time processing of acoustic information for different types of contrast (e.g., stop‐place contrast). Additionally, given that the use of VOT is conditioned by the prosodic position of the stops (Kim &amp; Cho, [<reflink idref="bib35" id="ref137">35</reflink>]; Mitterer, Cho, &amp; Kim, [<reflink idref="bib53" id="ref138">53</reflink>]) and that the informativeness of F0 may derive from post‐lexically assigned tones in the language's intonational phonology (Cho, [<reflink idref="bib8" id="ref139">8</reflink>]; Choi et al., [<reflink idref="bib10" id="ref140">10</reflink>]), expanding the study to explore how cue integration interacts with prosodic structure, and how this interaction may further modulate cue weighting, could be beneficial (e.g., Steffman, Kim, Cho, &amp; Jun, [<reflink idref="bib65" id="ref141">65</reflink>]; see McQueen &amp; Dilley, [<reflink idref="bib50" id="ref142">50</reflink>], for related discussion). Finally, researchers should also expand the scope of test cases by examining phonological contrasts in languages other than English. This is particularly important given that the cue weighting of acoustic information and acoustic details involved in similar phonological contrasts vary across languages, as demonstrated by the current study.</p> <hd id="AN0181921441-29">Acknowledgments</hd> <p>Part of the data presented here is from the corresponding author's dissertation project and was presented at the <emph>20th International Congress of Phonetic Sciences</emph> in Prague, Czech Republic. The authors would like to thank Dr. Allard Jongman, Dr. Joan Sereno, Dr. Jie Zhang, and all members of LING 850 at the University of Kansas for their insightful feedback on this work. We would also like to thank all student members of the <emph>Hanyang Institute for Phonetics and Cognitive Science</emph> for their help with recruiting and testing participants in Korea.</p> <hd id="AN0181921441-30">A Appendix Beta‐coefficients of logistic regression models and relative cue weighting of indi...</hd> <p></p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th&gt;Subject&lt;/th&gt;&lt;th align="center"&gt;Aspirated versus Fortis&lt;/th&gt;&lt;th align="center"&gt;Aspirated versus Lenis&lt;/th&gt;&lt;th align="center"&gt;Lenis versus Fortis&lt;/th&gt;&lt;/tr&gt;&lt;tr&gt;&lt;th align="left" /&gt;&lt;th align="center"&gt;VOT&lt;/th&gt;&lt;th align="center"&gt;F0&lt;/th&gt;&lt;th align="center"&gt;RCW&lt;/th&gt;&lt;th align="center"&gt;VOT&lt;/th&gt;&lt;th align="center"&gt;F0&lt;/th&gt;&lt;th align="center"&gt;RCW&lt;/th&gt;&lt;th align="center"&gt;VOT&lt;/th&gt;&lt;th align="center"&gt;F0&lt;/th&gt;&lt;th align="center"&gt;RCW&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;p101&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.19&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.32&lt;/td&gt;&lt;td&gt;0.21&lt;/td&gt;&lt;td&gt;&lt;bold&gt;2.02&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;1.57&lt;/td&gt;&lt;td&gt;0.44&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.78&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.29&lt;/td&gt;&lt;td&gt;0.27&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p102&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.66&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.26&lt;/td&gt;&lt;td&gt;0.14&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.05&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.97&lt;/td&gt;&lt;td&gt;0.48&lt;/td&gt;&lt;td&gt;0.37&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.65&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.63&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p103&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.30&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.06&lt;/td&gt;&lt;td&gt;0.04&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.33&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;1.28&lt;/td&gt;&lt;td&gt;0.49&lt;/td&gt;&lt;td&gt;0.53&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.96&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.65&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p104&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.64&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.00&lt;/td&gt;&lt;td&gt;0.00&lt;/td&gt;&lt;td&gt;0.43&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.40&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.77&lt;/td&gt;&lt;td&gt;0.49&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.93&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.65&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p105&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.51&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.60&lt;/td&gt;&lt;td&gt;0.28&lt;/td&gt;&lt;td&gt;0.59&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.34&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.69&lt;/td&gt;&lt;td&gt;0.06&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.93&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.94&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p106&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.08&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.22&lt;/td&gt;&lt;td&gt;0.17&lt;/td&gt;&lt;td&gt;0.55&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.90&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.62&lt;/td&gt;&lt;td&gt;0.36&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.41&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.53&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p107&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.88&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.05&lt;/td&gt;&lt;td&gt;0.06&lt;/td&gt;&lt;td&gt;0.64&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.97&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.60&lt;/td&gt;&lt;td&gt;0.03&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.12&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.98&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p108&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.62&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.15&lt;/td&gt;&lt;td&gt;0.20&lt;/td&gt;&lt;td&gt;0.92&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.98&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.52&lt;/td&gt;&lt;td&gt;0.49&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.63&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.56&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p110&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.38&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.19&lt;/td&gt;&lt;td&gt;0.12&lt;/td&gt;&lt;td&gt;0.66&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.84&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.56&lt;/td&gt;&lt;td&gt;0.41&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.53&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.56&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p111&lt;/td&gt;&lt;td&gt;&lt;bold&gt;2.98&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.28&lt;/td&gt;&lt;td&gt;0.09&lt;/td&gt;&lt;td&gt;0.01&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.96&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.99&lt;/td&gt;&lt;td&gt;0.60&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.72&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.54&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p112&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.39&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.29&lt;/td&gt;&lt;td&gt;0.17&lt;/td&gt;&lt;td&gt;0.11&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.94&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.89&lt;/td&gt;&lt;td&gt;0.38&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.93&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.71&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p113&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.27&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.16&lt;/td&gt;&lt;td&gt;0.11&lt;/td&gt;&lt;td&gt;&lt;bold&gt;2.20&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;1.72&lt;/td&gt;&lt;td&gt;0.44&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.24&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.37&lt;/td&gt;&lt;td&gt;0.23&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p114&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.64&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.05&lt;/td&gt;&lt;td&gt;0.08&lt;/td&gt;&lt;td&gt;0.52&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.57&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.52&lt;/td&gt;&lt;td&gt;0.19&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.37&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.65&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p115&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.17&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.30&lt;/td&gt;&lt;td&gt;0.20&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.33&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.84&lt;/td&gt;&lt;td&gt;0.39&lt;/td&gt;&lt;td&gt;0.41&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.55&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.58&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p116&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.13&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.01&lt;/td&gt;&lt;td&gt;0.01&lt;/td&gt;&lt;td&gt;0.45&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.89&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.66&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.00&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.64&lt;/td&gt;&lt;td&gt;0.39&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p117&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.07&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.42&lt;/td&gt;&lt;td&gt;0.28&lt;/td&gt;&lt;td&gt;0.53&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.15&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.68&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.72&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.67&lt;/td&gt;&lt;td&gt;0.49&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p118&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.84&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.08&lt;/td&gt;&lt;td&gt;0.09&lt;/td&gt;&lt;td&gt;0.41&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.87&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.68&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.40&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.32&lt;/td&gt;&lt;td&gt;0.44&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p120&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.19&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.40&lt;/td&gt;&lt;td&gt;0.25&lt;/td&gt;&lt;td&gt;0.17&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.49&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.74&lt;/td&gt;&lt;td&gt;0.20&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.74&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.79&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p121&lt;/td&gt;&lt;td&gt;&lt;bold&gt;2.31&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.22&lt;/td&gt;&lt;td&gt;0.09&lt;/td&gt;&lt;td&gt;0.56&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.74&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.57&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.94&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.37&lt;/td&gt;&lt;td&gt;0.28&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p122&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.75&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.15&lt;/td&gt;&lt;td&gt;0.17&lt;/td&gt;&lt;td&gt;0.32&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.81&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.72&lt;/td&gt;&lt;td&gt;0.33&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.78&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.70&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p123&lt;/td&gt;&lt;td&gt;&lt;bold&gt;2.21&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.32&lt;/td&gt;&lt;td&gt;0.13&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.65&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.57&lt;/td&gt;&lt;td&gt;0.26&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.26&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.60&lt;/td&gt;&lt;td&gt;0.32&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p124&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.02&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.12&lt;/td&gt;&lt;td&gt;0.11&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.15&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;1.01&lt;/td&gt;&lt;td&gt;0.47&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.62&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.52&lt;/td&gt;&lt;td&gt;0.45&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p125&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.86&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.25&lt;/td&gt;&lt;td&gt;0.22&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.85&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.70&lt;/td&gt;&lt;td&gt;0.45&lt;/td&gt;&lt;td&gt;0.40&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.61&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.61&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p126&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.59&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.08&lt;/td&gt;&lt;td&gt;0.12&lt;/td&gt;&lt;td&gt;0.30&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.67&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.70&lt;/td&gt;&lt;td&gt;0.49&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.79&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.62&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p127&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.81&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.20&lt;/td&gt;&lt;td&gt;0.20&lt;/td&gt;&lt;td&gt;0.63&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.94&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.60&lt;/td&gt;&lt;td&gt;0.12&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.41&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.77&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p128&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.81&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.14&lt;/td&gt;&lt;td&gt;0.15&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.59&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.54&lt;/td&gt;&lt;td&gt;0.48&lt;/td&gt;&lt;td&gt;0.15&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.64&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.81&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p129&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.37&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.47&lt;/td&gt;&lt;td&gt;0.25&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.95&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.64&lt;/td&gt;&lt;td&gt;0.40&lt;/td&gt;&lt;td&gt;0.35&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.36&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.51&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p130&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.16&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.15&lt;/td&gt;&lt;td&gt;0.11&lt;/td&gt;&lt;td&gt;0.41&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.62&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.60&lt;/td&gt;&lt;td&gt;0.16&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.62&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.80&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p131&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.03&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.12&lt;/td&gt;&lt;td&gt;0.10&lt;/td&gt;&lt;td&gt;0.96&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.35&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.58&lt;/td&gt;&lt;td&gt;0.46&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.70&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.60&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p132&lt;/td&gt;&lt;td&gt;&lt;bold&gt;2.17&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.33&lt;/td&gt;&lt;td&gt;0.13&lt;/td&gt;&lt;td&gt;0.05&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.94&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.95&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.62&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.45&lt;/td&gt;&lt;td&gt;0.42&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p133&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.84&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.04&lt;/td&gt;&lt;td&gt;0.05&lt;/td&gt;&lt;td&gt;0.28&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.59&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.68&lt;/td&gt;&lt;td&gt;0.35&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.44&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.55&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p135&lt;/td&gt;&lt;td&gt;&lt;bold&gt;2.71&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.10&lt;/td&gt;&lt;td&gt;0.04&lt;/td&gt;&lt;td&gt;1.26&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.42&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.53&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.06&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.56&lt;/td&gt;&lt;td&gt;0.35&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p136&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.87&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.17&lt;/td&gt;&lt;td&gt;0.17&lt;/td&gt;&lt;td&gt;0.66&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.77&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.54&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.95&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.54&lt;/td&gt;&lt;td&gt;0.36&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p137&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.98&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.43&lt;/td&gt;&lt;td&gt;0.30&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.80&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.76&lt;/td&gt;&lt;td&gt;0.49&lt;/td&gt;&lt;td&gt;0.39&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.53&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.58&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p138&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.28&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.11&lt;/td&gt;&lt;td&gt;0.08&lt;/td&gt;&lt;td&gt;0.60&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.90&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.60&lt;/td&gt;&lt;td&gt;0.31&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.91&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.75&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p139&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.58&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.10&lt;/td&gt;&lt;td&gt;0.06&lt;/td&gt;&lt;td&gt;0.40&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.66&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.62&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.64&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.30&lt;/td&gt;&lt;td&gt;0.32&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p140&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.34&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.01&lt;/td&gt;&lt;td&gt;0.01&lt;/td&gt;&lt;td&gt;1.09&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.39&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.56&lt;/td&gt;&lt;td&gt;0.18&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.80&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.82&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p141&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.78&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.35&lt;/td&gt;&lt;td&gt;0.16&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.90&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.52&lt;/td&gt;&lt;td&gt;0.37&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.75&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.41&lt;/td&gt;&lt;td&gt;0.35&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p142&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.58&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.29&lt;/td&gt;&lt;td&gt;0.16&lt;/td&gt;&lt;td&gt;0.44&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.27&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.74&lt;/td&gt;&lt;td&gt;0.09&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.23&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.93&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p143&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.22&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.04&lt;/td&gt;&lt;td&gt;0.03&lt;/td&gt;&lt;td&gt;0.20&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.17&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.85&lt;/td&gt;&lt;td&gt;0.33&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.80&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.71&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p144&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.93&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.16&lt;/td&gt;&lt;td&gt;0.15&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.84&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.62&lt;/td&gt;&lt;td&gt;0.42&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.63&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.44&lt;/td&gt;&lt;td&gt;0.41&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p145&lt;/td&gt;&lt;td&gt;&lt;bold&gt;2.19&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.62&lt;/td&gt;&lt;td&gt;0.22&lt;/td&gt;&lt;td&gt;0.40&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.82&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.67&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.77&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.64&lt;/td&gt;&lt;td&gt;0.45&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p146&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.85&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.50&lt;/td&gt;&lt;td&gt;0.21&lt;/td&gt;&lt;td&gt;0.68&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.06&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.61&lt;/td&gt;&lt;td&gt;0.18&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.69&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.79&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p147&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.03&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.27&lt;/td&gt;&lt;td&gt;0.21&lt;/td&gt;&lt;td&gt;0.29&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.93&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.76&lt;/td&gt;&lt;td&gt;0.51&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.70&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.58&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p148&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.82&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.17&lt;/td&gt;&lt;td&gt;0.09&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.98&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.85&lt;/td&gt;&lt;td&gt;0.46&lt;/td&gt;&lt;td&gt;0.74&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.83&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.53&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p149&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.85&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.02&lt;/td&gt;&lt;td&gt;0.01&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.29&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.97&lt;/td&gt;&lt;td&gt;0.43&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.11&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.52&lt;/td&gt;&lt;td&gt;0.32&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p150&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.87&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.09&lt;/td&gt;&lt;td&gt;0.10&lt;/td&gt;&lt;td&gt;0.80&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.81&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.51&lt;/td&gt;&lt;td&gt;0.46&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.51&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.53&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p151&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.71&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.19&lt;/td&gt;&lt;td&gt;0.21&lt;/td&gt;&lt;td&gt;0.49&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.36&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.73&lt;/td&gt;&lt;td&gt;0.06&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.90&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.94&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p152&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.30&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.06&lt;/td&gt;&lt;td&gt;0.05&lt;/td&gt;&lt;td&gt;1.03&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.28&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.55&lt;/td&gt;&lt;td&gt;0.01&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.02&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.99&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p153&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.50&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.20&lt;/td&gt;&lt;td&gt;0.28&lt;/td&gt;&lt;td&gt;0.41&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.78&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.66&lt;/td&gt;&lt;td&gt;0.42&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.51&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.55&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p154&lt;/td&gt;&lt;td&gt;&lt;bold&gt;2.04&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.58&lt;/td&gt;&lt;td&gt;0.22&lt;/td&gt;&lt;td&gt;0.61&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.37&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.69&lt;/td&gt;&lt;td&gt;0.53&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.17&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.69&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p155&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.27&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.13&lt;/td&gt;&lt;td&gt;0.09&lt;/td&gt;&lt;td&gt;0.50&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.95&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.66&lt;/td&gt;&lt;td&gt;0.32&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.44&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.58&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p156&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.37&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.47&lt;/td&gt;&lt;td&gt;0.25&lt;/td&gt;&lt;td&gt;0.25&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.51&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.68&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.56&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.31&lt;/td&gt;&lt;td&gt;0.36&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p157&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.17&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.04&lt;/td&gt;&lt;td&gt;0.04&lt;/td&gt;&lt;td&gt;0.59&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.81&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.58&lt;/td&gt;&lt;td&gt;0.39&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.51&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.57&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p158&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.24&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.43&lt;/td&gt;&lt;td&gt;0.26&lt;/td&gt;&lt;td&gt;0.66&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.44&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.69&lt;/td&gt;&lt;td&gt;0.52&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.96&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.65&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p159&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.38&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.18&lt;/td&gt;&lt;td&gt;0.12&lt;/td&gt;&lt;td&gt;0.63&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.71&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.53&lt;/td&gt;&lt;td&gt;0.42&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.54&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.56&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p160&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.80&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.18&lt;/td&gt;&lt;td&gt;0.09&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.66&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.54&lt;/td&gt;&lt;td&gt;0.45&lt;/td&gt;&lt;td&gt;0.59&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.89&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.60&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p161&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.71&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.06&lt;/td&gt;&lt;td&gt;0.07&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.75&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.47&lt;/td&gt;&lt;td&gt;0.39&lt;/td&gt;&lt;td&gt;0.07&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.47&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.86&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p162&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.96&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.04&lt;/td&gt;&lt;td&gt;0.04&lt;/td&gt;&lt;td&gt;0.57&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.73&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.56&lt;/td&gt;&lt;td&gt;0.36&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.57&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.62&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p163&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.16&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.13&lt;/td&gt;&lt;td&gt;0.10&lt;/td&gt;&lt;td&gt;0.44&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.28&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.74&lt;/td&gt;&lt;td&gt;0.19&lt;/td&gt;&lt;td&gt;&lt;bold&gt;0.93&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.83&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p164&lt;/td&gt;&lt;td&gt;&lt;bold&gt;2.44&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.35&lt;/td&gt;&lt;td&gt;0.12&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.22&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;1.07&lt;/td&gt;&lt;td&gt;0.47&lt;/td&gt;&lt;td&gt;&lt;bold&gt;1.73&lt;/bold&gt;&lt;/td&gt;&lt;td&gt;0.71&lt;/td&gt;&lt;td&gt;0.29&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <hd id="AN0181921441-31">B Appendix Critical and filler stimuli used in VWP (Experiment 2)</hd> <p>B1 Table Critical stimuli (romanization, transcription, and English glosses)</p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th&gt;Fortis&lt;/th&gt;&lt;th&gt;Lenis&lt;/th&gt;&lt;th&gt;Aspirated&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;ppaldae&lt;/italic&gt; [p'alt'e]&lt;/p&gt;&lt;p&gt;"straw"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;balgul&lt;/italic&gt; [pal&amp;#609;ul]&lt;/p&gt;&lt;p&gt;"digging"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;paljji&lt;/italic&gt; [p&lt;sup&gt;h&lt;/sup&gt;al&amp;#679;'i]&lt;/p&gt;&lt;p&gt;"bracelet"&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;ppulte&lt;/italic&gt; [p'ult&lt;sup&gt;h&lt;/sup&gt;e]&lt;/p&gt;&lt;p&gt;"horn&amp;#8208;rims"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;bulpan&lt;/italic&gt; [pulp&lt;sup&gt;h&lt;/sup&gt;an]&lt;/p&gt;&lt;p&gt;"grills"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;pullip&lt;/italic&gt; [p&lt;sup&gt;h&lt;/sup&gt;ulnip]&lt;/p&gt;&lt;p&gt;"grass leaf"&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;ppalgang&lt;/italic&gt; [p'al&amp;#609;a&amp;#331;]&lt;/p&gt;&lt;p&gt;"red"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;baldeung&lt;/italic&gt; [palt'&amp;#616;&amp;#331;]&lt;/p&gt;&lt;p&gt;"instep"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;palja&lt;/italic&gt; [p&lt;sup&gt;h&lt;/sup&gt;al&amp;#679;'a]&lt;/p&gt;&lt;p&gt;"fate"&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;ppallae&lt;/italic&gt; [p'alle]&lt;/p&gt;&lt;p&gt;"laundry"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;balsa&lt;/italic&gt; [pals'a]&lt;/p&gt;&lt;p&gt;"launch"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;palmok&lt;/italic&gt; [p&lt;sup&gt;h&lt;/sup&gt;almok]&lt;/p&gt;&lt;p&gt;"wrist"&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;ppalpan&lt;/italic&gt; [p'alp&lt;sup&gt;h&lt;/sup&gt;an]&lt;/p&gt;&lt;p&gt;"suction tool"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;baljin&lt;/italic&gt; [pal&amp;#679;'in]&lt;/p&gt;&lt;p&gt;"takeoff"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;palttuk&lt;/italic&gt; [p&lt;sup&gt;h&lt;/sup&gt;alt'uk]&lt;/p&gt;&lt;p&gt;"forearm"&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p>B2 Table Filler stimuli (romanization, transcription, and English glosses)</p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th&gt;Fortis&lt;/th&gt;&lt;th&gt;Lenis&lt;/th&gt;&lt;th&gt;Aspirated&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;jjokji&lt;/italic&gt; [&amp;#679;'ok&amp;#679;'i]&lt;/p&gt;&lt;p&gt;"note"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;jokbo&lt;/italic&gt; [&amp;#679;okp'o]&lt;/p&gt;&lt;p&gt;"genealogy"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;chokgam&lt;/italic&gt; [&amp;#679;&lt;sup&gt;h&lt;/sup&gt;okk'am]&lt;/p&gt;&lt;p&gt;"feel"&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;jjimtong&lt;/italic&gt; [&amp;#679;'imt&lt;sup&gt;h&lt;/sup&gt;o&amp;#331;]&lt;/p&gt;&lt;p&gt;"steamer"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;jimseung&lt;/italic&gt; [&amp;#679;ims&amp;#616;&amp;#331;]&lt;/p&gt;&lt;p&gt;"beast"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;chimryak&lt;/italic&gt; [&amp;#679;&lt;sup&gt;h&lt;/sup&gt;imnjak]&lt;/p&gt;&lt;p&gt;"invasion"&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;jjajeung&lt;/italic&gt; [&amp;#679;'a&amp;#677;&amp;#616;&amp;#331;]&lt;/p&gt;&lt;p&gt;"annoyance"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;jagyeok&lt;/italic&gt; [&amp;#679;a&amp;#609;j&amp;#652;k]&lt;/p&gt;&lt;p&gt;"qualification"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;chabyeol&lt;/italic&gt; [&amp;#679;&lt;sup&gt;h&lt;/sup&gt;abj&amp;#652;l]&lt;/p&gt;&lt;p&gt;"discrimination"&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;jjamppong&lt;/italic&gt; [&amp;#679;'amp'o&amp;#331;]&lt;/p&gt;&lt;p&gt;"spicy noodle"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;jambok&lt;/italic&gt; [&amp;#679;ambok]&lt;/p&gt;&lt;p&gt;"ambush"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;chamgyeon&lt;/italic&gt; [&amp;#679;&lt;sup&gt;h&lt;/sup&gt;am&amp;#609;j&amp;#652;n]&lt;/p&gt;&lt;p&gt;"interference"&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;jjigae&lt;/italic&gt; [&amp;#679;'i&amp;#609;e]&lt;/p&gt;&lt;p&gt;"stew"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;jido&lt;/italic&gt; [&amp;#679;ido]&lt;/p&gt;&lt;p&gt;"map"&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;chima&lt;/italic&gt; [&amp;#679;&lt;sup&gt;h&lt;/sup&gt;ima]&lt;/p&gt;&lt;p&gt;"skirt"&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <hd id="AN0181921441-32">C Appendix R code for GAMM</hd> <p></p> <ulist> <item> (<reflink idref="bib1" id="ref143">1</reflink>) GAMM for the within‐condition analysis <ephtml> &lt;math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"&gt;&lt;semantics&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;Diff&lt;/mi&gt;&lt;mo&gt;%5f&lt;/mo&gt;&lt;mi&gt;fixation&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;&amp;#8764;&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced open="(" close=")"&gt;&lt;mi&gt;Time&lt;/mi&gt;&lt;/mfenced&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced separators="" open="(" close=")"&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;Time&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;by&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mrow&gt;&lt;mi&gt;binary&lt;/mi&gt;&lt;mo&gt;&amp;#95;&lt;/mo&gt;&lt;mi&gt;TC&lt;/mi&gt;&lt;mo&gt;&amp;#95;&lt;/mo&gt;&lt;mi&gt;Type&lt;/mi&gt;&lt;mo&gt;&amp;#95;&lt;/mo&gt;&lt;mn&gt;0&lt;/mn&gt;&lt;/mrow&gt;&lt;/mrow&gt;&lt;/mfenced&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced separators="" open="(" close=")"&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;subject&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;bs&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mrow&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;/mrow&gt;&lt;mi&gt;re&lt;/mi&gt;&lt;mo&gt;"&lt;/mo&gt;&lt;/mrow&gt;&lt;/mfenced&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced separators="" open="(" close=")"&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;item&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;bs&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mrow&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;/mrow&gt;&lt;mi&gt;re&lt;/mi&gt;&lt;mo&gt;"&lt;/mo&gt;&lt;/mrow&gt;&lt;/mfenced&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced separators="" open="(" close=")"&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;subject&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;TC&lt;/mi&gt;&lt;mo&gt;&amp;#95;&lt;/mo&gt;&lt;mi&gt;Type&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;bs&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mrow&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;/mrow&gt;&lt;mi&gt;re&lt;/mi&gt;&lt;mo&gt;"&lt;/mo&gt;&lt;/mrow&gt;&lt;/mfenced&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced separators="" open="(" close=")"&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;Time&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;subject&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;bs&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mrow&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;/mrow&gt;&lt;mi&gt;fs&lt;/mi&gt;&lt;mo&gt;"&lt;/mo&gt;&lt;/mrow&gt;&lt;/mfenced&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced separators="" open="(" close=")"&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;Time&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;item&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;bs&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mrow&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;/mrow&gt;&lt;mi&gt;fs&lt;/mi&gt;&lt;mo&gt;"&lt;/mo&gt;&lt;/mrow&gt;&lt;/mfenced&gt;&lt;/mrow&gt;&lt;annotation encoding="application/x-tex"&gt;$ {\mathrm{Diff\&amp;#95;fixation }}\sim {\mathrm{ s}}\left({{\mathrm{Time}}} \right) + {\mathrm{ s}}\left({{\mathrm{Time, by}} = {\mathrm{binary\&amp;#95;TC\&amp;#95;Type\&amp;#95;0}}} \right) + {\mathrm{ s}}\left({{\mathrm{subject, bs }} = {''}{\mathrm{re}}{"}} \right) + {\mathrm{s}}\left({{\mathrm{item, bs}} = {''}{\mathrm{re}}{"}} \right) + {\mathrm{s}}\left({{\mathrm{subject, TC\&amp;#95;Type, bs }} = {''}{\mathrm{re}}{"}} \right) + {\mathrm{s}}\left({{\mathrm{Time, subject, bs}} = {''}{\mathrm{fs}}{"}} \right) + {\mathrm{s}}\left({{\mathrm{Time, item, bs }} = {''}{\mathrm{fs}}{"}} \right)$&lt;/annotation&gt;&lt;/semantics&gt;&lt;/math&gt; </ephtml></item> <p></p> <item> (<reflink idref="bib2" id="ref144">2</reflink>) GAMM for the cross‐condition analysis <ephtml> &lt;math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"&gt;&lt;semantics&gt;&lt;mrow&gt;&lt;mi&gt;Diff&lt;/mi&gt;&lt;mo&gt;%5f&lt;/mo&gt;&lt;mi&gt;fixation&lt;/mi&gt;&lt;mo&gt;&amp;#8764;&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced open="(" close=")"&gt;&lt;mi&gt;Time&lt;/mi&gt;&lt;/mfenced&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced separators="" open="(" close=")"&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;Time&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;by&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mrow&gt;&lt;mi&gt;binary&lt;/mi&gt;&lt;mo&gt;&amp;#95;&lt;/mo&gt;&lt;mi&gt;TC&lt;/mi&gt;&lt;mo&gt;&amp;#95;&lt;/mo&gt;&lt;mi&gt;Type&lt;/mi&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;mo&gt;&amp;#95;&lt;/mo&gt;&lt;mi&gt;TC&lt;/mi&gt;&lt;mo&gt;&amp;#95;&lt;/mo&gt;&lt;mi&gt;Type&lt;/mi&gt;&lt;/mrow&gt;&lt;mn&gt;2&lt;/mn&gt;&lt;/mrow&gt;&lt;/mfenced&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced separators="" open="(" close=")"&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;subject&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;bs&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mrow&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;/mrow&gt;&lt;mi&gt;re&lt;/mi&gt;&lt;mo&gt;"&lt;/mo&gt;&lt;/mrow&gt;&lt;/mfenced&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced separators="" open="(" close=")"&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;item&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;bs&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mrow&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;/mrow&gt;&lt;mi&gt;re&lt;/mi&gt;&lt;mo&gt;"&lt;/mo&gt;&lt;/mrow&gt;&lt;/mfenced&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced separators="" open="(" close=")"&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;subject&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;TC&lt;/mi&gt;&lt;mo&gt;&amp;#95;&lt;/mo&gt;&lt;mi&gt;Type&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;bs&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mrow&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;/mrow&gt;&lt;mi&gt;re&lt;/mi&gt;&lt;mo&gt;"&lt;/mo&gt;&lt;/mrow&gt;&lt;/mfenced&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced separators="" open="(" close=")"&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;Time&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;subject&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;bs&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mrow&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;/mrow&gt;&lt;mi&gt;fs&lt;/mi&gt;&lt;mo&gt;"&lt;/mo&gt;&lt;/mrow&gt;&lt;/mfenced&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced separators="" open="(" close=")"&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;Time&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;item&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;bs&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mrow&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;/mrow&gt;&lt;mi&gt;fs&lt;/mi&gt;&lt;mo&gt;"&lt;/mo&gt;&lt;/mrow&gt;&lt;/mfenced&gt;&lt;/mrow&gt;&lt;annotation encoding="application/x-tex"&gt;$ {\mathrm{Diff}}\&amp;#95;{\mathrm{fixation}}\sim {\mathrm{s}}\left({{\mathrm{Time}}} \right) + {\mathrm{s}}\left({{\mathrm{Time, by}} = {\mathrm{binary\&amp;#95;TC\&amp;#95;Type1\&amp;#95;TC\&amp;#95;Type}}2} \right) + {\mathrm{s}}\left({{\mathrm{subject, bs}} = {''}{\mathrm{re}}{"}} \right) + {\mathrm{ s}}\left({{\mathrm{item, bs}} = {''}{\mathrm{re}}{"}} \right) + {\mathrm{s}}\left({{\mathrm{subject, TC\&amp;#95;Type, bs}} = {''}{\mathrm{re}}{"}} \right) + {\mathrm{ s}}\left({{\mathrm{Time, subject, bs}} = {''}{\mathrm{fs}}{"}} \right) + {\mathrm{ s}}\left({{\mathrm{Time, item, bs}} = {''}{\mathrm{fs}}{"}} \right) $&lt;/annotation&gt;&lt;/semantics&gt;&lt;/math&gt; </ephtml></item> <p></p> <item> (<reflink idref="bib3" id="ref145">3</reflink>) GAMM for the individual analysis <ephtml> &lt;math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"&gt;&lt;semantics&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;Diff&lt;/mi&gt;&lt;mo&gt;%5f&lt;/mo&gt;&lt;mi&gt;fixation&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;&amp;#8764;&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced open="(" close=")"&gt;&lt;mi&gt;Time&lt;/mi&gt;&lt;/mfenced&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced separators="" open="(" close=")"&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;Time&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;by&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mrow&gt;&lt;mi&gt;binary&lt;/mi&gt;&lt;mo&gt;&amp;#95;&lt;/mo&gt;&lt;mi&gt;TC&lt;/mi&gt;&lt;mo&gt;&amp;#95;&lt;/mo&gt;&lt;mi&gt;Type&lt;/mi&gt;&lt;mo&gt;&amp;#95;&lt;/mo&gt;&lt;/mrow&gt;&lt;mn&gt;0&lt;/mn&gt;&lt;/mrow&gt;&lt;/mfenced&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced separators="" open="(" close=")"&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;item&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;bs&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mrow&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;/mrow&gt;&lt;mi&gt;re&lt;/mi&gt;&lt;mo&gt;"&lt;/mo&gt;&lt;/mrow&gt;&lt;/mfenced&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;mi mathvariant="normal"&gt;s&lt;/mi&gt;&lt;mfenced separators="" open="(" close=")"&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi&gt;Time&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;item&lt;/mi&gt;&lt;mo&gt;,&lt;/mo&gt;&lt;mi&gt;bs&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mrow&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;mo&gt;'&lt;/mo&gt;&lt;/mrow&gt;&lt;mi&gt;fs&lt;/mi&gt;&lt;mo&gt;"&lt;/mo&gt;&lt;/mrow&gt;&lt;/mfenced&gt;&lt;/mrow&gt;&lt;annotation encoding="application/x-tex"&gt;${\mathrm{Diff\&amp;#95;fixation}}\sim {\mathrm{s}}\left({{\mathrm{Time}}} \right) + {\mathrm{s}}\left({{\mathrm{Time, by}} = {\mathrm{binary\&amp;#95;TC\&amp;#95;Type\&amp;#95;}}0} \right) + {\mathrm{ s}}\left({{\mathrm{item, bs}} = {''}{\mathrm{re}}{"}} \right) + {\mathrm{s}}\left({{\mathrm{Time, item, bs}} = {''}{\mathrm{fs}}{"}} \right)$&lt;/annotation&gt;&lt;/semantics&gt;&lt;/math&gt; </ephtml></item> </ulist> <hd id="AN0181921441-33">D Appendix Individual listeners' VOT integration index in milliseconds.</hd> <p></p> <p> <ephtml> &lt;table&gt;&lt;thead&gt;&lt;tr&gt;&lt;th&gt;Subject&lt;/th&gt;&lt;th&gt;&lt;p&gt;Timing of VOT effect in Aspirated target&lt;/p&gt;&lt;p&gt;&amp;#8722; Fortis competitor&lt;/p&gt;&lt;/th&gt;&lt;th&gt;&lt;p&gt;Timing of VOT effect in Aspirated target&lt;/p&gt;&lt;p&gt;&amp;#8722; Lenis competitor&lt;/p&gt;&lt;/th&gt;&lt;th&gt;&lt;p&gt;Timing of VOT effect in Lenis target&lt;/p&gt;&lt;p&gt;&amp;#8722; Fortis competitor&lt;/p&gt;&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;p101&lt;/td&gt;&lt;td&gt;104.65&lt;/td&gt;&lt;td&gt;191.92&lt;/td&gt;&lt;td&gt;150.51&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p102&lt;/td&gt;&lt;td&gt;189.90&lt;/td&gt;&lt;td&gt;212.12&lt;/td&gt;&lt;td&gt;398.99&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p103&lt;/td&gt;&lt;td&gt;173.74&lt;/td&gt;&lt;td&gt;202.00&lt;/td&gt;&lt;td&gt;221.21&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p104&lt;/td&gt;&lt;td&gt;206.06&lt;/td&gt;&lt;td&gt;237.37&lt;/td&gt;&lt;td&gt;251.21&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p105&lt;/td&gt;&lt;td&gt;311.11&lt;/td&gt;&lt;td&gt;242.42&lt;/td&gt;&lt;td&gt;404.04&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p106&lt;/td&gt;&lt;td&gt;210.00&lt;/td&gt;&lt;td&gt;318.18&lt;/td&gt;&lt;td&gt;212.12&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p107&lt;/td&gt;&lt;td&gt;106.97&lt;/td&gt;&lt;td&gt;252.53&lt;/td&gt;&lt;td&gt;250.20&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p108&lt;/td&gt;&lt;td&gt;121.21&lt;/td&gt;&lt;td&gt;282.83&lt;/td&gt;&lt;td&gt;334.34&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p109&lt;/td&gt;&lt;td&gt;165.66&lt;/td&gt;&lt;td&gt;272.53&lt;/td&gt;&lt;td&gt;186.87&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p110&lt;/td&gt;&lt;td&gt;210.10&lt;/td&gt;&lt;td&gt;277.78&lt;/td&gt;&lt;td&gt;348.48&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p111&lt;/td&gt;&lt;td&gt;185.86&lt;/td&gt;&lt;td&gt;328.18&lt;/td&gt;&lt;td&gt;378.79&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p112&lt;/td&gt;&lt;td&gt;250.51&lt;/td&gt;&lt;td&gt;318.18&lt;/td&gt;&lt;td&gt;247.47&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p113&lt;/td&gt;&lt;td&gt;210.10&lt;/td&gt;&lt;td&gt;257.58&lt;/td&gt;&lt;td&gt;NA&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p114&lt;/td&gt;&lt;td&gt;113.13&lt;/td&gt;&lt;td&gt;226.26&lt;/td&gt;&lt;td&gt;348.48&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p115&lt;/td&gt;&lt;td&gt;242.42&lt;/td&gt;&lt;td&gt;111.11&lt;/td&gt;&lt;td&gt;393.94&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p116&lt;/td&gt;&lt;td&gt;141.41&lt;/td&gt;&lt;td&gt;196.97&lt;/td&gt;&lt;td&gt;141.41&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p117&lt;/td&gt;&lt;td&gt;286.87&lt;/td&gt;&lt;td&gt;166.67&lt;/td&gt;&lt;td&gt;267.68&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p118&lt;/td&gt;&lt;td&gt;290.12&lt;/td&gt;&lt;td&gt;257.58&lt;/td&gt;&lt;td&gt;166.67&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p120&lt;/td&gt;&lt;td&gt;NA&lt;/td&gt;&lt;td&gt;261.62&lt;/td&gt;&lt;td&gt;333.33&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p121&lt;/td&gt;&lt;td&gt;121.21&lt;/td&gt;&lt;td&gt;207.58&lt;/td&gt;&lt;td&gt;230.30&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p122&lt;/td&gt;&lt;td&gt;127.27&lt;/td&gt;&lt;td&gt;333.33&lt;/td&gt;&lt;td&gt;328.28&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p123&lt;/td&gt;&lt;td&gt;296.97&lt;/td&gt;&lt;td&gt;252.53&lt;/td&gt;&lt;td&gt;297.98&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p124&lt;/td&gt;&lt;td&gt;212.12&lt;/td&gt;&lt;td&gt;252.53&lt;/td&gt;&lt;td&gt;272.73&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p125&lt;/td&gt;&lt;td&gt;163.64&lt;/td&gt;&lt;td&gt;227.27&lt;/td&gt;&lt;td&gt;265.66&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p126&lt;/td&gt;&lt;td&gt;333.33&lt;/td&gt;&lt;td&gt;424.24&lt;/td&gt;&lt;td&gt;237.37&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p127&lt;/td&gt;&lt;td&gt;351.52&lt;/td&gt;&lt;td&gt;287.88&lt;/td&gt;&lt;td&gt;257.58&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p128&lt;/td&gt;&lt;td&gt;321.21&lt;/td&gt;&lt;td&gt;282.83&lt;/td&gt;&lt;td&gt;251.52&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p129&lt;/td&gt;&lt;td&gt;157.58&lt;/td&gt;&lt;td&gt;232.32&lt;/td&gt;&lt;td&gt;196.97&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p130&lt;/td&gt;&lt;td&gt;187.88&lt;/td&gt;&lt;td&gt;326.26&lt;/td&gt;&lt;td&gt;242.42&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p131&lt;/td&gt;&lt;td&gt;212.12&lt;/td&gt;&lt;td&gt;217.17&lt;/td&gt;&lt;td&gt;252.53&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p132&lt;/td&gt;&lt;td&gt;181.82&lt;/td&gt;&lt;td&gt;252.53&lt;/td&gt;&lt;td&gt;242.42&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p133&lt;/td&gt;&lt;td&gt;178.79&lt;/td&gt;&lt;td&gt;252.53&lt;/td&gt;&lt;td&gt;240.40&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p135&lt;/td&gt;&lt;td&gt;187.88&lt;/td&gt;&lt;td&gt;308.08&lt;/td&gt;&lt;td&gt;222.22&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p136&lt;/td&gt;&lt;td&gt;206.06&lt;/td&gt;&lt;td&gt;222.22&lt;/td&gt;&lt;td&gt;161.62&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p137&lt;/td&gt;&lt;td&gt;315.15&lt;/td&gt;&lt;td&gt;181.82&lt;/td&gt;&lt;td&gt;242.42&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p138&lt;/td&gt;&lt;td&gt;181.82&lt;/td&gt;&lt;td&gt;338.38&lt;/td&gt;&lt;td&gt;378.79&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p139&lt;/td&gt;&lt;td&gt;175.76&lt;/td&gt;&lt;td&gt;262.53&lt;/td&gt;&lt;td&gt;222.22&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p140&lt;/td&gt;&lt;td&gt;206.06&lt;/td&gt;&lt;td&gt;NA&lt;/td&gt;&lt;td&gt;256.57&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p141&lt;/td&gt;&lt;td&gt;218.18&lt;/td&gt;&lt;td&gt;251.52&lt;/td&gt;&lt;td&gt;NA&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p142&lt;/td&gt;&lt;td&gt;172.73&lt;/td&gt;&lt;td&gt;211.11&lt;/td&gt;&lt;td&gt;256.57&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p143&lt;/td&gt;&lt;td&gt;260.61&lt;/td&gt;&lt;td&gt;368.69&lt;/td&gt;&lt;td&gt;217.17&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p144&lt;/td&gt;&lt;td&gt;236.36&lt;/td&gt;&lt;td&gt;232.32&lt;/td&gt;&lt;td&gt;247.47&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p145&lt;/td&gt;&lt;td&gt;151.52&lt;/td&gt;&lt;td&gt;328.38&lt;/td&gt;&lt;td&gt;303.03&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p146&lt;/td&gt;&lt;td&gt;363.64&lt;/td&gt;&lt;td&gt;287.88&lt;/td&gt;&lt;td&gt;271.72&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p147&lt;/td&gt;&lt;td&gt;200.00&lt;/td&gt;&lt;td&gt;409.09&lt;/td&gt;&lt;td&gt;186.87&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p148&lt;/td&gt;&lt;td&gt;363.64&lt;/td&gt;&lt;td&gt;328.28&lt;/td&gt;&lt;td&gt;252.42&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p149&lt;/td&gt;&lt;td&gt;175.76&lt;/td&gt;&lt;td&gt;257.58&lt;/td&gt;&lt;td&gt;161.62&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p150&lt;/td&gt;&lt;td&gt;254.55&lt;/td&gt;&lt;td&gt;232.32&lt;/td&gt;&lt;td&gt;NA&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p151&lt;/td&gt;&lt;td&gt;327.27&lt;/td&gt;&lt;td&gt;196.97&lt;/td&gt;&lt;td&gt;313.03&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p152&lt;/td&gt;&lt;td&gt;212.12&lt;/td&gt;&lt;td&gt;212.12&lt;/td&gt;&lt;td&gt;333.33&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p153&lt;/td&gt;&lt;td&gt;224.24&lt;/td&gt;&lt;td&gt;312.12&lt;/td&gt;&lt;td&gt;232.32&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p154&lt;/td&gt;&lt;td&gt;351.52&lt;/td&gt;&lt;td&gt;272.73&lt;/td&gt;&lt;td&gt;300.00&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p155&lt;/td&gt;&lt;td&gt;133.33&lt;/td&gt;&lt;td&gt;207.07&lt;/td&gt;&lt;td&gt;232.32&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p156&lt;/td&gt;&lt;td&gt;224.24&lt;/td&gt;&lt;td&gt;227.27&lt;/td&gt;&lt;td&gt;287.88&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p157&lt;/td&gt;&lt;td&gt;142.42&lt;/td&gt;&lt;td&gt;329.29&lt;/td&gt;&lt;td&gt;343.43&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p158&lt;/td&gt;&lt;td&gt;327.27&lt;/td&gt;&lt;td&gt;342.42&lt;/td&gt;&lt;td&gt;363.64&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p159&lt;/td&gt;&lt;td&gt;327.27&lt;/td&gt;&lt;td&gt;292.93&lt;/td&gt;&lt;td&gt;297.98&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p160&lt;/td&gt;&lt;td&gt;181.82&lt;/td&gt;&lt;td&gt;247.47&lt;/td&gt;&lt;td&gt;303.03&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p161&lt;/td&gt;&lt;td&gt;121.12&lt;/td&gt;&lt;td&gt;353.54&lt;/td&gt;&lt;td&gt;410.10&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p162&lt;/td&gt;&lt;td&gt;224.24&lt;/td&gt;&lt;td&gt;207.07&lt;/td&gt;&lt;td&gt;323.23&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p163&lt;/td&gt;&lt;td&gt;290.91&lt;/td&gt;&lt;td&gt;280.81&lt;/td&gt;&lt;td&gt;324.24&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;p164&lt;/td&gt;&lt;td&gt;375.76&lt;/td&gt;&lt;td&gt;303.03&lt;/td&gt;&lt;td&gt;202.02&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <ref id="AN0181921441-34"> <title> Footnotes </title> <blist> <bibl id="bib1" idref="ref3" type="bt">1</bibl> <bibtext> It should be noted that our descriptions of Korean are only limited to Seoul Korean, a Korean dialect spoken in Seoul and the Gyeonggi region of South Korea.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref43" type="bt">2</bibl> <bibtext> The acoustic realization of VOT and F0 is associated with prosodic structural factors, such as position and prominence. VOT remains a distinctive cue for the lenis‐aspirated distinction in phrase‐medial, non‐prominent (unfocused) contexts due to the voicing of lenis stops, while the role of F0 appears to be minimal, if not negligible (Choi, Kim, &amp; Cho, [10]).</bibtext> </blist> <blist> <bibl id="bib3" idref="ref62" type="bt">3</bibl> <bibtext> Available at: https://asa.scitation.org/doi/suppl/10.1121/10.0000692</bibtext> </blist> <blist> <bibl id="bib4" idref="ref94" type="bt">4</bibl> <bibtext> Our findings also provide some implications to the broader idea that the temporal processing of spoken words may not be strictly coupled to the temporal dynamics of the input (Apfelbaum, Bullock‐Rest, Rhone, Jongman, &amp; McMurray, [3]; Galle, Klein‐Packard, Schreiber, &amp; McMurray, [18]; Hannagan, Magnuson, &amp; Grainger, [22]; McMurray, Farris‐Trimble, &amp; Rigler, [45]; McMurray, Muegge, &amp; Kim, [47]; Schreiber &amp; McMurray, [63]; Toscano et al., [67]), though our study was not particularly designed to test theoretical accounts of temporal processing in speech.</bibtext> </blist> </ref> <ref id="AN0181921441-35"> <title> References </title> <blist> <bibtext> Allen, J. S., &amp; Miller, J. L. (1999). Effects of syllable‐initial voicing and speaking rate on the temporal characteristics of monosyllabic words. Journal of the Acoustical Society of America, 106 (4 Pt 1), 2031 – 2039. https://doi.org/10.1121/1.427949</bibtext> </blist> <blist> <bibtext> Anwyl‐Irvine, A. L., Massonnie, J., Flitton, A., Kirkham, N., &amp; Evershed, J. K. (2020). Gorilla in our midst: An online behavioral experiment builder. Behavior Research Methods, 52 (1), 388 – 407. https://doi.org/10.3758/s13428‐019‐01237‐x</bibtext> </blist> <blist> <bibtext> Apfelbaum, K. S., Bullock‐Rest, N., Rhone, A. E., Jongman, A., &amp; McMurray, B. (2014). Contingent categorization in speech perception. Language, Cognition and Neuroscience, 29 (9), 1070 – 1082. https://doi.org/10.1080/01690965.2013.824995</bibtext> </blist> <blist> <bibtext> Apfelbaum, K. S., Klein‐Packard, J., &amp; McMurray, B. (2021). The pictures who shall not be named: Empirical support for benefits of preview in the Visual World Paradigm. Journal of Memory and Language, 121, 104279. https://doi.org/10.1016/j.jml.2021.104279</bibtext> </blist> <blist> <bibl id="bib5" idref="ref46" type="bt">5</bibl> <bibtext> Bang, H.‐Y., Sonderegger, M., Kang, Y., Clayards, M., &amp; Yoon, T.‐J. (2018). The emergence, progress, and impact of sound change in progress in Seoul Korean: Implications for mechanisms of tonogenesis. Journal of Phonetics, 66, 120 – 144. https://doi.org/10.1016/j.wocn.2017.09.005</bibtext> </blist> <blist> <bibl id="bib6" idref="ref70" type="bt">6</bibl> <bibtext> Bates, D., Mächler, M., Bolker, B., &amp; Walker, S. (2015). Fitting linear mixed‐effects models using lme4. Journal of Statistical Software, 67 (1), 1 – 48. https://doi.org/10.18637/jss.v067.i01</bibtext> </blist> <blist> <bibl id="bib7" idref="ref64" type="bt">7</bibl> <bibtext> Boersma, P., &amp; Weenink, D. (2021). Praat: Doing phonetics by computer. Version 6.1.42. Retrieved from <ulink href="http://www.praat.org/">http://www.praat.org/</ulink></bibtext> </blist> <blist> <bibl id="bib8" idref="ref54" type="bt">8</bibl> <bibtext> Cho, T. (2022). The phonetics‐prosody interface and prosodic strengthening in Korean. In S. Cho &amp; J. Whitman (Eds.), Cambridge handbook of Korean linguistics (pp. 248 – 293). Cambridge University Press.</bibtext> </blist> <blist> <bibl id="bib9" idref="ref39" type="bt">9</bibl> <bibtext> Cho, T., Jun, S.‐A., &amp; Ladefoged, P. (2002). Acoustic and aerodynamic correlates of Korean stops and fricatives. Journal of Phonetics, 30 (2), 193 – 228. https://doi.org/10.1006/jpho.2001.0153</bibtext> </blist> <blist> <bibtext> Choi, J., Kim, S., &amp; Cho, T. (2020). An apparent‐time study of an ongoing sound change in Seoul Korean: A prosodic account. PLoS One, 15 (10), e0240682. https://doi.org/10.1371/journal.pone.0240682</bibtext> </blist> <blist> <bibtext> Clayards, M. (2018). Differences in cue weights for speech perception are correlated for individuals within and across contrasts. Journal of the Acoustical Society of America, 144 (3), EL172. https://doi.org/10.1121/1.5052025</bibtext> </blist> <blist> <bibtext> Clopper, C. G., &amp; Walker, A. (2017). Effects of lexical competition and dialect exposure on phonological priming. Language and Speech, 60 (1), 85 – 109. https://doi.org/10.1177/0023830916643737</bibtext> </blist> <blist> <bibtext> Escudero, P., Benders, T., &amp; Lipski, S. C. (2009). Native, non‐native and L2 perceptual cue weighting for Dutch vowels: The case of Dutch, German, and Spanish listeners. Journal of Phonetics, 37 (4), 452 – 465. https://doi.org/10.1016/j.wocn.2009.07.006</bibtext> </blist> <blist> <bibtext> Experiment Builder. (2015). SR Research Experiment Builder. SR Research.</bibtext> </blist> <blist> <bibtext> Francis, A. L., Baldwin, K., &amp; Nusbaum, H. C. (2000). Effects of training on attention to acoustic cues. Perception and Psychophysics, 62 (8), 1668 – 1680. https://doi.org/10.3758/bf03212164</bibtext> </blist> <blist> <bibtext> Francis, A. L., Kaganovich, N., &amp; Driscoll‐Huber, C. (2008). Cue‐specific effects of categorization training on the relative weighting of acoustic cues to consonant voicing in English. Journal of the Acoustical Society of America, 124 (2), 1234 – 1251. https://doi.org/10.1121/1.2945161</bibtext> </blist> <blist> <bibtext> Francis, A. L., &amp; Nusbaum, H. C. (2002). Selective attention and the acquisition of new phonetic categories. Journal of Experimental Psychology: Human Perception and Performance, 28 (2), 349 – 366. https://doi.org/10.1037//0096‐1523.28.2.349</bibtext> </blist> <blist> <bibtext> Galle, M. E., Klein‐Packard, J., Schreiber, K., &amp; McMurray, B. (2019). What are you waiting for? Real‐time integration of cues for fricatives suggests encapsulated auditory memory. Cognitive Science, 43 (1), e12700. https://doi.org/10.1111/cogs.12700</bibtext> </blist> <blist> <bibtext> Gow, D. W. (2003). Feature parsing: Feature cue mapping in spoken word recognition. Perception &amp; Psychophysics, 65 (4), 575 – 590.</bibtext> </blist> <blist> <bibtext> Haggard, M., Ambler, S., &amp; Callow, M. (1970). Pitch as a voicing cue. Journal of the Acoustical Society of America, 47 (2), 613 – 617. https://doi.org/10.1121/1.1911936</bibtext> </blist> <blist> <bibtext> Hallett, P. E. (1986). Eye movements. In K. R. Boff, L. Kaufman, &amp; J. P. Thomas (Eds.), Handbook of perception and human performance (Vol. 1). New York : Wiley.</bibtext> </blist> <blist> <bibtext> Hannagan, T., Magnuson, J. S., &amp; Grainger, J. (2013). Spoken word recognition without a TRACE. Frontiers in Psychology, 4, 563. https://doi.org/10.3389/fpsyg.2013.00563</bibtext> </blist> <blist> <bibtext> Hazan, V., &amp; Rosen, S. (1991). Individual variability in the perception of cues to place contrasts in initial stops. Perception &amp; Psychophysics, 49 (2), 187 – 200. https://doi.org/10.3758/BF03205038</bibtext> </blist> <blist> <bibtext> Holt, L. L., &amp; Lotto, A. J. (2006). Cue weighting in auditory categorization: Implications for first and second language acquisition. Journal of the Acoustical Society of America, 119 (5 Pt 1), 3059 – 3071. https://doi.org/10.1121/1.2188377</bibtext> </blist> <blist> <bibtext> Huettig, F., &amp; McQueen, J. M. (2007). The tug of war between phonological, semantic and shape information in language‐mediated visual search. Journal of Memory and Language, 57 (4), 460 – 482. https://doi.org/10.1016/j.jml.2007.02.001</bibtext> </blist> <blist> <bibtext> Jongman, A., Wayland, R., &amp; Wong, S. (2000). Acoustic characteristics of English fricatives. Journal of the Acoustical Society of America, 108 (3) (Pt 1), 1252 – 1263. https://doi.org/10.1121/1.1288413</bibtext> </blist> <blist> <bibtext> Kang, K. H., &amp; Guion, S. G. (2008). Clear speech production of Korean stops: Changing phonetic targets and enhancement strategies. Journal of the Acoustical Society of America, 124 (6), 3909 – 3917. https://doi.org/10.1121/1.2988292</bibtext> </blist> <blist> <bibtext> Kang, Y. (2014). Voice Onset Time merger and development of tonal contrast in Seoul Korean stops: A corpus study. Journal of Phonetics, 45, 76 – 90. https://doi.org/10.1016/j.wocn.2014.03.005</bibtext> </blist> <blist> <bibtext> Kapnoula, E. C., Edwards, J., &amp; McMurray, B. (2021). Gradient activation of speech categories facilitates listeners' recovery from lexical garden paths, but not perception of speech‐in‐noise. Journal of Experimental Psychology: Human Perception and Performance, 47 (4), 578 – 595. https://doi.org/10.1037/xhp0000900</bibtext> </blist> <blist> <bibtext> Kapnoula, E. C., &amp; McMurray, B. (2021). Idiosyncratic use of bottom‐up and top‐down information leads to differences in speech perception flexibility: Converging evidence from ERPs and eye‐tracking. Brain and Language, 223, 105031. https://doi.org/10.1016/j.bandl.2021.105031</bibtext> </blist> <blist> <bibtext> Kapnoula, E. C., Winn, M. B., Kong, E. J., Edwards, J., &amp; McMurray, B. (2017). Evaluating the sources and functions of gradiency in phoneme categorization: An individual differences approach. Journal of Experimental Psychology: Human Perception and Performance, 43 (9), 1594 – 1611. https://doi.org/10.1037/xhp0000410</bibtext> </blist> <blist> <bibtext> Kim, D., Clayards, M., &amp; Kong, E. J. (2020). Individual differences in perceptual adaptation to unfamiliar phonetic categories. Journal of Phonetics, 81, 100984. https://doi.org/10.1016/j.wocn.2020.100984</bibtext> </blist> <blist> <bibtext> Kim, H., &amp; Jongman, A. (2022). The influence of inter‐dialect contact on the Korean three‐way laryngeal distinction: An acoustic comparison among Seoul Korean speakers and Gyeongsang speakers with limited and extended residence in Seoul. Language and Speech, 65 (3), 531 – 553. https://doi.org/10.1177/00238309211037720</bibtext> </blist> <blist> <bibtext> Kim, M. R. C. (1994). Acoustic characteristics of Korean stops and perception of English stop consonants. University of Wisconsin‐Madison.</bibtext> </blist> <blist> <bibtext> Kim, S., &amp; Cho, T. (2013). Prosodic boundary information modulates phonetic categorization. Journal of the Acoustical Society of America, 134 (1), EL19 – 25. https://doi.org/10.1121/1.4807431</bibtext> </blist> <blist> <bibtext> Kingston, J., Levy, J., Rysling, A., &amp; Staub, A. (2016). Eye movement evidence for an immediate Ganong effect. Journal of Experimental Psychology: Human Perception and Performance, 42 (12), 1969 – 1988. https://doi.org/10.1037/xhp0000269</bibtext> </blist> <blist> <bibtext> Kong, E. J., &amp; Edwards, J. (2016). Individual differences in categorical perception of speech: Cue weighting and executive function. Journal of Phonetics, 59, 40 – 57. https://doi.org/10.1016/j.wocn.2016.08.006</bibtext> </blist> <blist> <bibtext> Kong, E. J., &amp; Lee, H. (2018). Attentional modulation and individual differences in explaining the changing role of fundamental frequency in Korean laryngeal stop perception. Language and Speech, 61 (3), 384 – 408. https://doi.org/10.1177/0023830917729840</bibtext> </blist> <blist> <bibtext> Lee, H., &amp; Jongman, A. (2012). Effects of tone on the three‐way laryngeal distinction in Korean: An acoustic and aerodynamic comparison of the Seoul and South Kyungsang dialects. Journal of the International Phonetic Association, 42 (2), 145 – 169. https://doi.org/10.1017/s0025100312000035</bibtext> </blist> <blist> <bibtext> Lee, H., Politzer‐Ahles, S., &amp; Jongman, A. (2013). Speakers of tonal and non‐tonal Korean dialects use different cue weightings in the perception of the three‐way laryngeal stop contrast. Journal of Phonetics, 41 (2), 117 – 132. https://doi.org/10.1016/j.wocn.2012.12.002</bibtext> </blist> <blist> <bibtext> Liberman, A. M., Harris, K. S., Kinney, J. A., &amp; Lane, H. (1961). The discrimination of relative onset‐time of the components of certain speech and nonspeech patterns. Journal of Experimental Psychology, 61, 379 – 388. https://doi.org/10.1037/h0049038</bibtext> </blist> <blist> <bibtext> Lisker, L., &amp; Abramson, A. S. (1964). A cross‐language study of voicing in initial stops: Acoustical measurements. WORD, 20 (3), 384 – 422. https://doi.org/10.1080/00437956.1964.11659830</bibtext> </blist> <blist> <bibtext> McClelland, J. L., &amp; Elman, J. L. (1986). The TRACE model of speech perception. Cognitive Psychology, 18 (1), 1 – 86. https://doi.org/10.1016/0010‐0285(86)90015‐0</bibtext> </blist> <blist> <bibtext> McMurray, B., Clayards, M. A., Tanenhaus, M. K., &amp; Aslin, R. N. (2008). Tracking the time course of phonetic cue integration during spoken word recognition. Psychonomic Bulletin and Review, 15 (6), 1064 – 1071. https://doi.org/10.3758/PBR.15.6.1064</bibtext> </blist> <blist> <bibtext> McMurray, B., Farris‐Trimble, A., &amp; Rigler, H. (2017). Waiting for lexical access: Cochlear implants or severely degraded input lead listeners to process speech less incrementally. Cognition, 169, 147 – 164. https://doi.org/10.1016/j.cognition.2017.08.013</bibtext> </blist> <blist> <bibtext> McMurray, B., &amp; Jongman, A. (2011). What information is necessary for speech categorization? Harnessing variability in the speech signal by integrating cues computed relative to expectations. Psychological Review, 118 (2), 219 – 246. https://doi.org/10.1037/a0022325</bibtext> </blist> <blist> <bibtext> McMurray, B., Muegge, J. B., &amp; Kim, H. (in revision). Decoupling language processing from time. https://doi.org/10.31234/osf.io/uwyfn</bibtext> </blist> <blist> <bibtext> McMurray, B., Tanenhaus, M. K., &amp; Aslin, R. N. (2002). Gradient effects of within‐category phonetic variation on lexical access. Cognition, 86 (2), B33 – 42. https://doi.org/10.1016/s0010‐0277(02)00157‐9</bibtext> </blist> <blist> <bibtext> McMurray, B., Tanenhaus, M. K., &amp; Aslin, R. N. (2009). Within‐category VOT affects recovery from "lexical" garden paths: Evidence against phoneme‐level inhibition. Journal of Memory and Language, 60 (1), 65 – 91. https://doi.org/10.1016/j.jml.2008.07.002</bibtext> </blist> <blist> <bibtext> McQueen, J. M., &amp; Dilley, L. (2021). Prosody and spoken‐word recognition. In C. Gussenhhoven &amp; A. Chen (Eds.), The Oxford handbook of language prosody (pp. 509 – 521). Oxford University Press. https://doi.org/10.1093/oxfordhb/9780198832232.013.33</bibtext> </blist> <blist> <bibtext> McQueen, J. M., &amp; Viebahn, M. C. (2007). Tracking recognition of spoken words by tracking looks to printed words. Quarterly Journal of Experimental Psychology (Hove), 60 (5), 661 – 671. https://doi.org/10.1080/17470210601183890</bibtext> </blist> <blist> <bibtext> Mirman, D., Dixon, J. A., &amp; Magnuson, J. S. (2008). Statistical and computational models of the visual world paradigm: Growth curves and individual differences. Journal of Memory and Language, 59 (4), 475 – 494. https://doi.org/10.1016/j.jml.2007.11.006</bibtext> </blist> <blist> <bibtext> Mitterer, H., Cho, T., &amp; Kim, S. (2016). How does prosody influence speech categorization? Journal of Phonetics, 54, 68 – 79. https://doi.org/10.1016/j.wocn.2015.09.002</bibtext> </blist> <blist> <bibtext> Mitterer, H., &amp; Reinisch, E. (2013). No delays in application of perceptual learning in speech recognition: Evidence from eye tracking. Journal of Memory and Language, 69 (4), 527 – 545. https://doi.org/10.1016/j.jml.2013.07.002</bibtext> </blist> <blist> <bibtext> Morrison, G. S. (2007). Logistic regression modelling for first and second language perception data. In M. J. Solé, P. Prieto, &amp; J. Mascaró (Eds.), Segmental and prosodic issues in Romance phonology (pp. 219 – 236). John Benjamins.</bibtext> </blist> <blist> <bibtext> Morrison, G. S., &amp; Kondaurova, M. V. (2009). Analysis of categorical response data: Use logistic regression rather than endpoint‐difference scores or discriminant analysis. Journal of the Acoustical Society of America, 126 (5), 2159 – 2162. https://doi.org/10.1121/1.3216917</bibtext> </blist> <blist> <bibtext> Oden, G. C., &amp; Massaro, D. W. (1978). Integration of featural information in speech perception. Psychological Review, 85 (3), 172 – 191. Retrieved from https://<ulink href="http://www.ncbi.nlm.nih.gov/pubmed/663005">www.ncbi.nlm.nih.gov/pubmed/663005</ulink></bibtext> </blist> <blist> <bibtext> Ou, J., Yu, A. C. L., &amp; Xiang, M. (2021). Individual differences in categorization gradience as predicted by online processing of phonetic cues during spoken word recognition: Evidence from eye movements. Cognitive Science, 45 (3), e12948. https://doi.org/10.1111/cogs.12948</bibtext> </blist> <blist> <bibtext> R Development Core Team. (2023). R: A language and environment for statistical computing. Version 4.3.2. R Foundation for Statistical Computing.</bibtext> </blist> <blist> <bibtext> Reinisch, E., &amp; Sjerps, M. J. (2013). The uptake of spectral and temporal cues in vowel perception is rapidly influenced by context. Journal of Phonetics, 41 (2), 101 – 116. https://doi.org/10.1016/j.wocn.2013.01.002</bibtext> </blist> <blist> <bibtext> Salverda, A. P., Kleinschmidt, D., &amp; Tanenhaus, M. K. (2014). Immediate effects of anticipatory coarticulation in spoken‐word recognition. Journal of Memory and Language, 71 (1), 145 – 163. https://doi.org/10.1016/j.jml.2013.11.002</bibtext> </blist> <blist> <bibtext> Schertz, J., Cho, T., Lotto, A., &amp; Warner, N. (2015). Individual differences in phonetic cue use in production and perception of a non‐native sound contrast. Journal of Phonetics, 52, 183 – 204. https://doi.org/10.1016/j.wocn.2015.07.003</bibtext> </blist> <blist> <bibtext> Schreiber, K. E., &amp; McMurray, B. (2019). Listeners can anticipate future segments before they identify the current one. Attention, Perception, &amp; Psychophysics, 81 (4), 1147 – 1166. https://doi.org/10.3758/s13414‐019‐01712‐9</bibtext> </blist> <blist> <bibtext> Silva, D. J. (2006). Acoustic evidence for the emergence of tonal contrast in contemporary Korean. Phonology, 23 (02), 287 – 308. https://doi.org/10.1017/s0952675706000911</bibtext> </blist> <blist> <bibtext> Steffman, J., Kim, S., Cho, T., &amp; Jun, S.‐A. (2022). Prosodic phrasing mediates listeners' perception of temporal cues: Evidence from the Korean Accentual Phrase. Journal of Phonetics, 94, 101156. https://doi.org/10.1016/j.wocn.2022.101156</bibtext> </blist> <blist> <bibtext> Summerfield, Q. (1981). Articulatory rate and perceptual constancy in phonetic perception. Journal of Experimental Psychology: Human Perception and Performance, 7 (5), 1074 – 1095. https://doi.org/10.1037//0096‐1523.7.5.1074</bibtext> </blist> <blist> <bibtext> Toscano, J. C., Anderson, N. D., &amp; McMurray, B. (2013). Reconsidering the role of temporal order in spoken word recognition. Psychon Bull Rev, 20 (5), 981 – 987. https://doi.org/10.3758/s13423‐013‐0417‐0.</bibtext> </blist> <blist> <bibtext> Toscano, J. C., &amp; McMurray, B. (2010). Cue integration with categories: Weighting acoustic cues in speech using unsupervised learning and distributional statistics. Cognitive Science, 34 (3), 434 – 464. https://doi.org/10.1111/j.1551‐6709.2009.01077.x</bibtext> </blist> <blist> <bibtext> Toscano, J. C., &amp; McMurray, B. (2012). Cue‐integration and context effects in speech: Evidence against speaking‐rate normalization. Attention, Perception, &amp; Psychophysics, 74 (6), 1284 – 1301. https://doi.org/10.3758/s13414‐012‐0306‐z</bibtext> </blist> <blist> <bibtext> Tremblay, A., Broersma, M., Zeng, Y., Kim, H., Lee, J., &amp; Shin, S. (2021). Dutch listeners' perception of English lexical stress: A cue‐weighting approach. Journal of the Acoustical Society of America, 149 (6), 3703. https://doi.org/10.1121/10.0005086</bibtext> </blist> <blist> <bibtext> van Rij, J., Wieling, M., Baayen, R., &amp; van Rijn, H. (2022). itsadug: Interpreting time series and autocorrelated data using GAMMs. Version 2.4.1 R package.</bibtext> </blist> <blist> <bibtext> Weber, A., &amp; Scharenborg, O. (2012). Models of spoken‐word recognition. Wiley Interdisciplinary Reviews: Cognitive Science, 3 (3), 387 – 401. https://doi.org/10.1002/wcs.1178</bibtext> </blist> <blist> <bibtext> Wieling, M. (2018). Analyzing dynamic phonetic data using generalized additive mixed modeling: A tutorial focusing on articulatory differences between L1 and L2 speakers of English. Journal of Phonetics, 70, 86 – 116. https://doi.org/10.1016/j.wocn.2018.03.002</bibtext> </blist> <blist> <bibtext> Winn, M. B. (2020). Manipulation of voice onset time in speech stimuli: A tutorial and flexible Praat script. Journal of the Acoustical Society of America, 147 (2), 852. https://doi.org/10.1121/10.0000692</bibtext> </blist> <blist> <bibtext> Wood, S. N. (2006). Generalized additive models: An introduction with R (1st ed.). CRC Press.</bibtext> </blist> <blist> <bibtext> Wood, S. N. (2011). Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. Journal of the Royal Statistical Society Series B: Statistical Methodology, 73 (1), 3 – 36.</bibtext> </blist> <blist> <bibtext> Wood, S. N. (2017). Generalized additive models: An introduction with R (2nd ed). CRC Press.</bibtext> </blist> <blist> <bibtext> Zhang, Y., &amp; Francis, A. (2010). The weighting of vowel quality in native and non‐native listeners' perception of English lexical stress. Journal of Phonetics, 38 (2), 260 – 271. https://doi.org/10.1016/j.wocn.2009.11.002</bibtext> </blist> </ref> <aug> <p>By Hyoju Kim; Annie Tremblay and Taehong Cho</p> <p>Reported by Author; Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib41" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib20" firstref="ref2"></nolink> <nolink nlid="nl3" bibid="bib66" firstref="ref4"></nolink> <nolink nlid="nl4" bibid="bib17" firstref="ref5"></nolink> <nolink nlid="nl5" bibid="bib15" firstref="ref6"></nolink> <nolink nlid="nl6" bibid="bib24" firstref="ref7"></nolink> <nolink nlid="nl7" bibid="bib78" firstref="ref8"></nolink> <nolink nlid="nl8" bibid="bib16" firstref="ref9"></nolink> <nolink nlid="nl9" bibid="bib18" firstref="ref10"></nolink> <nolink nlid="nl10" bibid="bib44" firstref="ref11"></nolink> <nolink nlid="nl11" bibid="bib58" firstref="ref12"></nolink> <nolink nlid="nl12" bibid="bib60" firstref="ref13"></nolink> <nolink nlid="nl13" bibid="bib43" firstref="ref14"></nolink> <nolink nlid="nl14" bibid="bib48" firstref="ref16"></nolink> <nolink nlid="nl15" bibid="bib50" firstref="ref17"></nolink> <nolink nlid="nl16" bibid="bib68" firstref="ref18"></nolink> <nolink nlid="nl17" bibid="bib72" firstref="ref19"></nolink> <nolink nlid="nl18" bibid="bib19" firstref="ref22"></nolink> <nolink nlid="nl19" bibid="bib57" firstref="ref23"></nolink> <nolink nlid="nl20" bibid="bib63" firstref="ref24"></nolink> <nolink nlid="nl21" bibid="bib11" firstref="ref29"></nolink> <nolink nlid="nl22" bibid="bib23" firstref="ref30"></nolink> <nolink nlid="nl23" bibid="bib29" firstref="ref31"></nolink> <nolink nlid="nl24" bibid="bib30" firstref="ref32"></nolink> <nolink nlid="nl25" bibid="bib31" firstref="ref33"></nolink> <nolink nlid="nl26" bibid="bib32" firstref="ref34"></nolink> <nolink nlid="nl27" bibid="bib37" firstref="ref35"></nolink> <nolink nlid="nl28" bibid="bib62" firstref="ref37"></nolink> <nolink nlid="nl29" bibid="bib34" firstref="ref40"></nolink> <nolink nlid="nl30" bibid="bib42" firstref="ref41"></nolink> <nolink nlid="nl31" bibid="bib40" firstref="ref44"></nolink> <nolink nlid="nl32" bibid="bib10" firstref="ref47"></nolink> <nolink nlid="nl33" bibid="bib27" firstref="ref48"></nolink> <nolink nlid="nl34" bibid="bib28" firstref="ref49"></nolink> <nolink nlid="nl35" bibid="bib33" firstref="ref50"></nolink> <nolink nlid="nl36" bibid="bib39" firstref="ref51"></nolink> <nolink nlid="nl37" bibid="bib64" firstref="ref53"></nolink> <nolink nlid="nl38" bibid="bib74" firstref="ref67"></nolink> <nolink nlid="nl39" bibid="bib59" firstref="ref71"></nolink> <nolink nlid="nl40" bibid="bib70" firstref="ref72"></nolink> <nolink nlid="nl41" bibid="bib13" firstref="ref78"></nolink> <nolink nlid="nl42" bibid="bib55" firstref="ref82"></nolink> <nolink nlid="nl43" bibid="bib56" firstref="ref83"></nolink> <nolink nlid="nl44" bibid="bib25" firstref="ref87"></nolink> <nolink nlid="nl45" bibid="bib51" firstref="ref88"></nolink> <nolink nlid="nl46" bibid="bib38" firstref="ref90"></nolink> <nolink nlid="nl47" bibid="bib14" firstref="ref93"></nolink> <nolink nlid="nl48" bibid="bib73" firstref="ref96"></nolink> <nolink nlid="nl49" bibid="bib75" firstref="ref97"></nolink> <nolink nlid="nl50" bibid="bib77" firstref="ref98"></nolink> <nolink nlid="nl51" bibid="bib76" firstref="ref99"></nolink> <nolink nlid="nl52" bibid="bib52" firstref="ref101"></nolink> <nolink nlid="nl53" bibid="bib71" firstref="ref102"></nolink> <nolink nlid="nl54" bibid="bib21" firstref="ref104"></nolink> <nolink nlid="nl55" bibid="bib61" firstref="ref105"></nolink> <nolink nlid="nl56" bibid="bib69" firstref="ref118"></nolink> <nolink nlid="nl57" bibid="bib36" firstref="ref119"></nolink> <nolink nlid="nl58" bibid="bib54" firstref="ref120"></nolink> <nolink nlid="nl59" bibid="bib26" firstref="ref127"></nolink> <nolink nlid="nl60" bibid="bib46" firstref="ref129"></nolink> <nolink nlid="nl61" bibid="bib45" firstref="ref132"></nolink> <nolink nlid="nl62" bibid="bib49" firstref="ref133"></nolink> <nolink nlid="nl63" bibid="bib12" firstref="ref134"></nolink> <nolink nlid="nl64" bibid="bib35" firstref="ref137"></nolink> <nolink nlid="nl65" bibid="bib53" firstref="ref138"></nolink> <nolink nlid="nl66" bibid="bib65" firstref="ref141"></nolink> |
|---|---|
| Header | DbId: eric DbLabel: ERIC An: EJ1454985 AccessLevel: 3 PubType: Academic Journal PubTypeId: academicJournal PreciseRelevancyScore: 0 |
| IllustrationInfo | |
| Items | – Name: Title Label: Title Group: Ti Data: Perceptual Cue Weighting Matters in Real-Time Integration of Acoustic Information during Spoken Word Recognition – Name: Language Label: Language Group: Lang Data: English – Name: Author Label: Authors Group: Au Data: <searchLink fieldCode="AR" term="%22Hyoju+Kim%22">Hyoju Kim</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-3357-4460">0000-0002-3357-4460</externalLink>)<br /><searchLink fieldCode="AR" term="%22Annie+Tremblay%22">Annie Tremblay</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-0748-172X">0000-0002-0748-172X</externalLink>)<br /><searchLink fieldCode="AR" term="%22Taehong+Cho%22">Taehong Cho</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-8148-745X">0000-0002-8148-745X</externalLink>) – Name: TitleSource Label: Source Group: Src Data: <searchLink fieldCode="SO" term="%22Cognitive+Science%22"><i>Cognitive Science</i></searchLink>. 2024 48(12). – Name: Avail Label: Availability Group: Avail Data: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us – Name: PeerReviewed Label: Peer Reviewed Group: SrcInfo Data: Y – Name: Pages Label: Page Count Group: Src Data: 38 – Name: DatePubCY Label: Publication Date Group: Date Data: 2024 – Name: TypeDocument Label: Document Type Group: TypDoc Data: Journal Articles<br />Reports - Research – Name: Subject Label: Descriptors Group: Su Data: <searchLink fieldCode="DE" term="%22Foreign+Countries%22">Foreign Countries</searchLink><br /><searchLink fieldCode="DE" term="%22Asynchronous+Communication%22">Asynchronous Communication</searchLink><br /><searchLink fieldCode="DE" term="%22Cues%22">Cues</searchLink><br /><searchLink fieldCode="DE" term="%22Auditory+Stimuli%22">Auditory Stimuli</searchLink><br /><searchLink fieldCode="DE" term="%22Phonetics%22">Phonetics</searchLink><br /><searchLink fieldCode="DE" term="%22Language+Rhythm%22">Language Rhythm</searchLink><br /><searchLink fieldCode="DE" term="%22Distinctive+Features+%28Language%29%22">Distinctive Features (Language)</searchLink><br /><searchLink fieldCode="DE" term="%22Time+Perspective%22">Time Perspective</searchLink><br /><searchLink fieldCode="DE" term="%22Acoustics%22">Acoustics</searchLink><br /><searchLink fieldCode="DE" term="%22Word+Recognition%22">Word Recognition</searchLink><br /><searchLink fieldCode="DE" term="%22Association+%28Psychology%29%22">Association (Psychology)</searchLink><br /><searchLink fieldCode="DE" term="%22Listening%22">Listening</searchLink><br /><searchLink fieldCode="DE" term="%22Individual+Development%22">Individual Development</searchLink> – Name: Subject Label: Geographic Terms Group: Su Data: <searchLink fieldCode="DE" term="%22South+Korea+%28Seoul%29%22">South Korea (Seoul)</searchLink> – Name: DOI Label: DOI Group: ID Data: 10.1111/cogs.70026 – Name: ISSN Label: ISSN Group: ISSN Data: 0364-0213<br />1551-6709 – Name: Abstract Label: Abstract Group: Ab Data: This study investigates whether listeners' cue weighting predicts their real-time use of asynchronous acoustic information in spoken word recognition at both group and individual levels. By focusing on the time course of cue integration, we seek to distinguish between two theoretical views: the "associated" view (cue weighting is linked to cue integration strategy) and the "independent" view (no such relationship). The current study examines Seoul Korean listeners' (n = 62) weighting of voice onset time (VOT, available earlier in time) and onset fundamental frequency of the following vowel (F0, available later in time) when perceiving Korean stop contrasts (Experiment 1: cue-weighting perception task) and the timing of VOT integration when recognizing Korean words that begin with a stop (Experiment 2: visual-world eye-tracking task). The group-level results reveal that the timing of the early cue (VOT) integration is delayed when the later cue (F0) serves as the primary cue to process the stop contrast, supporting a relationship between cue weighting and the timing of cue integration (the associated view). At the individual level, listeners with greater reliance on F0 than VOT exhibited a further delayed integration of VOT. These findings suggest that the real-time processing of asynchronously occurring acoustic cues for lexical activation is modulated by the weight that listeners assign to those cues, providing evidence for the associated view of cue integration. This study offers insights into the mechanisms of cue integration and spoken word recognition, and they shed light on variability in cue integration strategies among listeners. – Name: AbstractInfo Label: Abstractor Group: Ab Data: As Provided – Name: DateEntry Label: Entry Date Group: Date Data: 2024 – Name: AN Label: Accession Number Group: ID Data: EJ1454985 |
| PLink | https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1454985 |
| RecordInfo | BibRecord: BibEntity: Identifiers: – Type: doi Value: 10.1111/cogs.70026 Languages: – Text: English PhysicalDescription: Pagination: PageCount: 38 Subjects: – SubjectFull: Foreign Countries Type: general – SubjectFull: Asynchronous Communication Type: general – SubjectFull: Cues Type: general – SubjectFull: Auditory Stimuli Type: general – SubjectFull: Phonetics Type: general – SubjectFull: Language Rhythm Type: general – SubjectFull: Distinctive Features (Language) Type: general – SubjectFull: Time Perspective Type: general – SubjectFull: Acoustics Type: general – SubjectFull: Word Recognition Type: general – SubjectFull: Association (Psychology) Type: general – SubjectFull: Listening Type: general – SubjectFull: Individual Development Type: general – SubjectFull: South Korea (Seoul) Type: general Titles: – TitleFull: Perceptual Cue Weighting Matters in Real-Time Integration of Acoustic Information during Spoken Word Recognition Type: main BibRelationships: HasContributorRelationships: – PersonEntity: Name: NameFull: Hyoju Kim – PersonEntity: Name: NameFull: Annie Tremblay – PersonEntity: Name: NameFull: Taehong Cho IsPartOfRelationships: – BibEntity: Dates: – D: 01 M: 12 Type: published Y: 2024 Identifiers: – Type: issn-print Value: 0364-0213 – Type: issn-electronic Value: 1551-6709 Numbering: – Type: volume Value: 48 – Type: issue Value: 12 Titles: – TitleFull: Cognitive Science Type: main |
| ResultId | 1 |
