Variability of articulator positions and formants across nine English vowels

Abstract

Speech, though communicative, is quite variable both in articulation and acoustics, and it has often been claimed that articulation is more variable. Here we compared variability in articulation and acoustics for 32 speakers in the x-ray microbeam database (XRMB; Westbury, 1994). Variability in tongue, lip and jaw positions for nine English vowels (/u, ʊ, æ, ɑ, ʌ, ɔ, ε, ɪ, i/) was compared to that of the corresponding formant values. The domains were made comparable by creating three-dimensional spaces for each: the first three principal components from an analysis of a 14-dimensional space for articulation, and an F1xF2xF3 space for acoustics. More variability occurred in the articulation than the acoustics for half of the speakers, while the reverse was true for the other half. Individual tokens were further from the articulatory median than the acoustic median for 40-60% of tokens across speakers. A separate analysis of three non-low front vowels (/ε, ɪ, i/, for which the XRMB system provides the most direct articulatory evidence) did not differ from the omnibus analysis. Speakers tended to be either more or less variable consistently across vowels. Across speakers, there was a positive correlation between articulatory and acoustic variability, both for all vowels and for just the three non-low front vowels. Although the XRMB is an incomplete representation of articulation, it nonetheless provides data for direct comparisons between articulatory and acoustic variability that have not been reported previously. The results indicate that articulation is not more variable than acoustics, that speakers had relatively consistent variability across vowels, and that articulatory and acoustic variability were related for the vowels themselves.

Causal inference explains the stimulus-level relationship between the McGurk Effect and auditory speech perception

“…Speech production is known to be variable in both articulation and acoustics (Holmberg et al, 1994;Whalen et al, 2018). Although we did not quantify physical differences between stimuli, such as voice onset time (Wood, 1976) our findings show that variability in production leads to variability in perception that is similar for McGurk and everyday speech.…”

“…If McGurk and everyday speech are processed using common perceptual mechanisms, then variability across stimuli should create predictable changes in both McGurk and everyday perception. For instance, even for a single talker, there is substantial variability in repeated productions of the same speech token (Holmberg et al, 1994;Whalen et al, 2018). In CIMS, this variability is modeled with a representational space that collapses all auditory and visual speech features onto a one-dimensional line with /ba/, /da/, and /ga/ at neighboring locations.…”

Variability and Central Tendencies in Speech Production

“…Other acoustic or articulatory measures could be made, although the strongest predictions in the field have been about formant values. Measuring variability across the vowel system rather than for just two vowels would be useful (Whalen et al, 2018), although the number of tokens required becomes rather large. Finding word tokens that maintain the voicing of the final consonant would also be desirable.…”

“…Here, we needed many repetitions in order to examine the distributional characteristics of the productions, and so we, like many others, relied on the acoustic output to index the articulatory activity. Not only is the acoustic output reliably shaped by the articulation (Fant, 1960;Iskarous, 2010), there is also evidence that variability in the acoustic domain is highly related to the variability in the articulatory domain (Whalen et al, 2018). The use of acoustics therefore is a reasonable first step in analyzing production variability.…”

Interaction Mechanisms between Glottal Source and Vocal Tract in Pitch Glides

“…Furthermore, there is a large variability in both the articulator positions and the acoustic characteristics of vowels (see, e.g. Pisoni, 1980;Whalen et al, 2018), suggesting that the communication of information between a speaker and listener does not require stationary and precise matching of articulatory or acoustic targets. In fact, the prevalence of cardinal vowels [A, i, u] in most languages has been been explained using two theories: The dispersion theory is based on the argument that cardinal vowels are used due to the maximal discriminability between them (Liljencrants and Lindblom, 1972).…”