Academia.eduAcademia.edu

Figure 2 – uploaded by T. Florian Jaeger

See full PDFdownloadDownload figure
Table 2. PND models for (log-transformed) speech onset latencies, (log-transformed) word duration, and vowel dispersion.

Table 2 PND models for (log-transformed) speech onset latencies, (log-transformed) word duration, and vowel dispersion.

Related Figures (9)

Figure 1. Possible relations between production ease and articulation. Panel (a) Radical production ease predicts that any systematic (i. e. non-random) variation in articulation is caused by production ease. Panel (b) Moderate production ease accounts predict that pro- duction ease is one of several factors that cause systematic variation in articulation. Panel (c) It is also possible that articulation is not affected by production ease.
Figure 1. Possible relations between production ease and articulation. Panel (a) Radical production ease predicts that any systematic (i. e. non-random) variation in articulation is caused by production ease. Panel (b) Moderate production ease accounts predict that pro- duction ease is one of several factors that cause systematic variation in articulation. Panel (c) It is also possible that articulation is not affected by production ease.
Figure 2. Some possible influences of PND on lexical planning and articulation under production ease accounts. In all of (a)—(c), pro- duction ease affects articulation. The panels differ, however, in the extent to which production ease explains effects of PND on articula- tion. Panel (a) shows the prediction of radical production ease accounts: the effect of PND on articulation is fully mediated by its effect on planning (e.g. if PND reduces lexical planning time it also reduces pronunciation detail). Panels (b) and (c) show the predictions of two different types of moderate production ease accounts. In (b), production ease explains some of the effect of PND on articulation: there are both an independent and mediated effects of PND on articulation. In (c), production ease affects articulation, but the effect of PND is independent of that.
Figure 2. Some possible influences of PND on lexical planning and articulation under production ease accounts. In all of (a)—(c), pro- duction ease affects articulation. The panels differ, however, in the extent to which production ease explains effects of PND on articula- tion. Panel (a) shows the prediction of radical production ease accounts: the effect of PND on articulation is fully mediated by its effect on planning (e.g. if PND reduces lexical planning time it also reduces pronunciation detail). Panels (b) and (c) show the predictions of two different types of moderate production ease accounts. In (b), production ease explains some of the effect of PND on articulation: there are both an independent and mediated effects of PND on articulation. In (c), production ease affects articulation, but the effect of PND is independent of that.
Table 1. Mean values and standard errors of control variables by high vs. low log-frequency-weighted PND. In addition to the 40 minimal pair pictures, there were 60 filler pictures. Fillers were pictures whose dominant labels were mono (9), bi-syllabic (42), tri-syllabic (7), or quadri-syllabic (2). Finally, eight pictures served as
Table 1. Mean values and standard errors of control variables by high vs. low log-frequency-weighted PND. In addition to the 40 minimal pair pictures, there were 60 filler pictures. Fillers were pictures whose dominant labels were mono (9), bi-syllabic (42), tri-syllabic (7), or quadri-syllabic (2). Finally, eight pictures served as
Figure 3. Schematic illustration of trial structure (a) and trial ordering within a minimal pair (b) for the picture naming experiment. Our minimal pair design holds constant syllable, coda, and onset complexity, all of which could influence word durations and vowel dispersion. However, differences in consonant contexts are known to affect vowel duration, specifically voicing and manner (House, 1961). Chi- squared tests of independence show that, across the density groups, our stimuli pairs did not significantly differ in manner (plosive, nasal, fricative, lateral), place (bilabial, labial, labio-dental, alveolar, velar), or voicing (p's >.1). In addition, we balanced (log-transformed) fre- quency, average biphone log probability, number of alternative picture labels for paired pictures, and pro- portion of usage of the dominant label (no difference assessed by paired t-tests). The mean (and standard devi- ation) of these measures by high vs. low log-frequency- weighted PND condition are provided in Table 1. We report log-frequency-weighted PND based on IPhOD2, a lexical database of 54,000 tokens of English from the SUBTLEXus corpus (Brysbaert & New, 2009). All results reported below replicated robustly when log-fre- quency-weighted PND was calculated based on CELEX (Baayen, Piepenbrock, & van Rijn, 1993), as in, for example, Scarborough (2010, 2012). Dies co cel cd Sao ce a ee ol ee Cw fe practice trials, following instructions and preceding the main session. Fillers and practice labels were chosen as to not be phonological neighbours with any of the criti- cal or filler items.
Figure 3. Schematic illustration of trial structure (a) and trial ordering within a minimal pair (b) for the picture naming experiment. Our minimal pair design holds constant syllable, coda, and onset complexity, all of which could influence word durations and vowel dispersion. However, differences in consonant contexts are known to affect vowel duration, specifically voicing and manner (House, 1961). Chi- squared tests of independence show that, across the density groups, our stimuli pairs did not significantly differ in manner (plosive, nasal, fricative, lateral), place (bilabial, labial, labio-dental, alveolar, velar), or voicing (p's >.1). In addition, we balanced (log-transformed) fre- quency, average biphone log probability, number of alternative picture labels for paired pictures, and pro- portion of usage of the dominant label (no difference assessed by paired t-tests). The mean (and standard devi- ation) of these measures by high vs. low log-frequency- weighted PND condition are provided in Table 1. We report log-frequency-weighted PND based on IPhOD2, a lexical database of 54,000 tokens of English from the SUBTLEXus corpus (Brysbaert & New, 2009). All results reported below replicated robustly when log-fre- quency-weighted PND was calculated based on CELEX (Baayen, Piepenbrock, & van Rijn, 1993), as in, for example, Scarborough (2010, 2012). Dies co cel cd Sao ce a ee ol ee Cw fe practice trials, following instructions and preceding the main session. Fillers and practice labels were chosen as to not be phonological neighbours with any of the criti- cal or filler items.
Table 3. Descriptive statistics of speech onset latencies, word durations, and vowel dispersion across conditions. The results are summarised in Table 4 and visualised in Figure 5. Speech onset latencies had a significant
Table 3. Descriptive statistics of speech onset latencies, word durations, and vowel dispersion across conditions. The results are summarised in Table 4 and visualised in Figure 5. Speech onset latencies had a significant
Figure 4. Model estimates of influence of log-frequency-weighted PND on planning and articulatory measures. Error bar represents 95% bootstrapped confidence interval of group difference.
Figure 4. Model estimates of influence of log-frequency-weighted PND on planning and articulatory measures. Error bar represents 95% bootstrapped confidence interval of group difference.
Table 4. Effect of (log) speech onset latencies and PND on (log) word durations. A comparison of the effects of PND on word duration and vowel dispersion, depending on whether speech onset latencies were included in the model (Table 4) or not (Table 2, Columns 2 and 3), further shows that the effects of PND on articulation are orthogonal to the
Table 4. Effect of (log) speech onset latencies and PND on (log) word durations. A comparison of the effects of PND on word duration and vowel dispersion, depending on whether speech onset latencies were included in the model (Table 4) or not (Table 2, Columns 2 and 3), further shows that the effects of PND on articulation are orthogonal to the
Figure 5. Raw data and model estimates of influence of log-transformed speech onset latency and log-frequency-weighted PND or articulatory measures. our results suggest that the effect of PND on articulation is orthogonal to whatever effect lexical planning has on articulation (supporting the architecture illustrated in Figure 2(c)). correlation between speech onset latencies and articula- tion (both the estimated effect sizes of PND and their standard errors remained virtually unchanged). Indeed, the fixed effect correlations between speech onset latency and both word duration and vowel dispersion were very low (fixed effect r<0.084 and r< 0.086, respectively). This is unexpected under any form of partial or full mediation.
Figure 5. Raw data and model estimates of influence of log-transformed speech onset latency and log-frequency-weighted PND or articulatory measures. our results suggest that the effect of PND on articulation is orthogonal to whatever effect lexical planning has on articulation (supporting the architecture illustrated in Figure 2(c)). correlation between speech onset latencies and articula- tion (both the estimated effect sizes of PND and their standard errors remained virtually unchanged). Indeed, the fixed effect correlations between speech onset latency and both word duration and vowel dispersion were very low (fixed effect r<0.084 and r< 0.086, respectively). This is unexpected under any form of partial or full mediation.
Related topics:
Cognitive PsychologyCognitive ScienceCommunicationLanguages and LinguisticsPsycholinguisticsCognitive LinguisticsLanguage ProductionPhonetics and PronunciationPhonological NeighborhoodsExemplar-based Models of Language
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved