Analyzing geospatial variation in articulation rate using crowdsourced speech data

Abstract

Most recent studies on the geographical distribution of acoustic features analyze comparatively few speakers and localities, both of which may be unrepresentative of the diversity found in larger or more spatially fragmented populations. In the present study we introduce a new paradigm that enables the crowdsourcing of acoustic features through smartphone devices. We used Dialäkt Äpp, a free iOS app that allows users to record themselves, to crowdsource audio data. Nearly 3,000 speakers from 452 localities in German-speaking Switzerland provided recordings; we measured articulation rates for these speakers using a metric based on duration intervals between consecutive vowel onsets. Results revealed distinct regional differences in articulation rate between major dialect regions and individual localities. The specification of 452 localities enabled analyses at an unprecedented spatial resolution. Results further revealed a robust effect of gender, with women articulating significantly more slowly than men. Both the geographical patterns and the effect of gender found in this study corroborate similar findings on Swiss German previously reported in a very limited set of localities, thus verifying the validity of the crowdsourcing framework. Given the application of this new framework, a large bulk of the discussion is devoted to discussing methodological caveats.

The Role of Acoustic Distance and Sociolinguistic Knowledge in Dialect Identification

“…He found GRG intonation contours to show a high pitch phrase-initially, a more marked declination, and to be less influenced by lexical stress than ZHG intonation. Analysis based on crowded-sourced data ( Leemann, 2016 ) further suggests that GRG is characterised by a slightly slower articulation rate in comparison to ZHG, confirming an earlier study ( Leemann, 2012 ).…”

“…However, since Zurich listeners were faster not only in the ambiguous, but across all three conditions, this interpretation does not seem plausible. A second, but arguably speculative explanation is that the between-group differences in RT result from cultural differences, given that articulation rate for the Zurich area was found to be faster as well ( Leemann, 2016 ; see Ebersbach et al, 2000 for differences between people from Berlin and Tyrol in gait velocity). A third and more plausible explanation for the difference in response latency may be different decision strategies.…”

Temporal variability in four Alemannic dialects and its influence on the respective varieties of Swiss Standard German

“…the number of syllables uttered in a given time frame without pauses [9]. Specifically, the results of these studies suggest that BE speakers enunciate more slowly (5.0 syllables/second = syl/s) than speakers from ZH and VS (5.8 syl/s), with speakers from GR lying somewhere in between [10,11,12]. [11] explain their results with BE speakers engaging in phrase-final lengthening, as vowels in prepausal position tend to be extended way more than in other dialects, leading to a lower average AR.…”

“…Multiple factors have been proven to influence it, amongst which are mood, speaking style, sex, utterance length, and age [12]. As for sex, women have been found to articulate more slowly, and thus, more carefully [13] in many Western languages, including British [14] and American English [15], Swedish [16], Dutch [17], German Standard German (GSG) [18], and even ALM [12,19]. It has been linked to women being more concerned about social status, which they linguistically convey by employing fewer non-standard and more prestigious features [20].…”

Articulation Rate in American English in a Corpus of YouTube Videos

“…Hahn and Siebenhaar (2016) analyzed articulation rates in German in recordings of reading passages by speakers in 67 localities in Germany, Austria, and Switzerland, and found that values generally increased from Northern Germany towards Southern Germany, Austria, and Switzerland. Leemann (2017) used a mobile telephone app to record 3000 Swiss German speakers from 452 Swiss localities speaking 16 words, then analyzed the variation in the length of time between vowel onsets in six disyllabic words ( Abend, Augen, fragen, Donnerstag, heben , and trinken ). He found a regional pattern that corresponds to some previous results from studies of speaking or articulation rate in Swiss German (Bern speakers articulate more slowly), but acknowledged that inferring conversational articulation rates from vowel onset times for isolated words spoken in a non-naturalistic context may not be reliable.…”