Skip to navigation – Site map

The evolution of creaky voice use in read speech by native-French and native-English speakers in tandem: a pilot study

Claire Pillot-Loiseau, Céline Horgues, Sylwia Scheuer and Takeki Kamiyama

Abstracts

Creaky voice is a non-modal (non-neutral) speech feature occurring both at the segmental and suprasegmental levels of languages. It also performs paralinguistic, stylistic, sociolinguistic and extra-linguistic functions. It has been reported as prevalent in (especially American) English but it is rarely described in French.
During tandem exchanges between native-French speakers and native-English speakers, does voice quality found in each of the 2 languages spoken evolve in the course of the interactions? In this pilot study of read speech by 9 French/English tandem pairs (undergraduate Francophone and Anglophone students paired up), how do the number and duration of creaky occurrences vary according to: the language spoken (French or English), the language status (L1 or L2), the speaker, and the segmental or suprasegmental contexts?
Creaky occurrences were manually annotated. Their frequency and duration were calculated for the 9 native French female speakers and the 9 native English speakers (eight women, one man).
They were found to be more frequent among the Anglophones (in both languages) than their Francophone counterparts (transfer of creaky voice production from English to French for Anglophones), and more frequent when the Francophones spoke in English rather than French (after 3 months of interaction), with a high degree of inter-speaker variability. Moreover, the relative durations of creaky occurrences were similar in both languages spoken by the Francophones, while for the Anglophones, they were shorter in French than in English. The scope of creaky voice can be a segment, the end of stress groups, or whole intonational units (with longer occurrences). In French, glottalisation may also occur at a hiatus, especially when produced by Anglophones.
This pilot study is a first step in the research of how speech interaction affects voice quality. The results obtained from read speech will have to be compared to the analysis of spontaneous speech produced by these same speakers.

Top of page

Author's notes

Creaky voice is a non-modal (non-neutral) speech: “The modal register is so named because it includes the range of fundamental Phonatory settings frequencies that are normally used in speaking and singing” (Hollien 1972: 3).

Full text

1. Introduction

1Voice quality has a quasi-permanent character as it relates to all sounds being uttered by any speaker (Abercrombie 1967), and in that sense it can be defined globally. However, voice quality may change on some segments and not others, which gives it a local dimension as well. This is particularly the case of creaky voice.

1.1. Definition and terminology

2Creaky voice can be described in acoustical, physiological and perceptual terms. According to Hollien and Michel (1968: 600), it is “best described as a phonational register occurring at frequencies below those of the modal register (18Hz to 65Hz)”, and its fundamental frequency values are nearly the same for men and women (between 18Hz and 65Hz). Hollien and Michel (1968: 600) described vocal fry as existing as a “physiologically normal mode of laryngeal operation”. It consists in irregular glottal cycles and a longer closure than modal voice (Hollien et al. 1977). Perceptually, creak is perceived as a sudden decrease in pitch (Kreiman 1982).

3The terminology used to refer to this type of voice quality is relatively varied and depends on the field of investigation. Indeed, various terms have traditionally been used to describe creaky voice, the most common being: creakiness, creak, or fry.

4Researchers in pathological voice generally refer to vocal fry, pulse register, creaky voice or creak. For Hollien (1972), for example, the term pulse register was synonymous with vocal fry, glottal fry, creak, strohbass. For Monsen and Engebretson (1977), vocal fry and creaky voice are interchangeable, while in Imaizumi and Gauffin (1991) the term creak is only limited to refer to a special case of vocal fry (i.e. “Fry” only observed at final parts of the utterance).

5From a phonetic perspective, Abdelli-Beruh et al. (2014, 2016) define creaky voice as: irregular phonation, pulse phonation, glottalisation, or laryngealisation. In his book “The phonetic description of voice quality”, Laver (1980) differentiated creak from creaky voice defining creak as a simple phonation type and creaky voice as a compound phonation type (“blending modal voice and creak”, p.125).

6Since then, glottalisation and creak have been taken to be interchangeable (Redi and Shattuck-Hufnagel 2001). Gordon and Ladefoged (2001), who focused on phonation aspects, described creak and vocal fry as also partaking of the same phenomenon. In Gibson (2017) vocal fry, glottal fry or creaky voice are synonyms. Kolher (2000) specified that creak or creaky voice emerged as phrase-final types of laryngealisation.

7What glottalisation, breathiness and breathy voice have in common is that they all correspond to types of non-modal voice. The reason for such variety in the use of terminology can partly be explained by the fact that there are actually several perceptual and acoustic categories of creaky voice (as presented in Table 1 below, adapted from Redi and Shattuck-Hufnagel 2001 and Keating et al. 2015).

Table 1: Name (term used to describe the phenomenon), acoustic and perceptual correlates for different creaky voice categories (Redi and Shattuck-Hufnagel 2001; Keating et al. 2015).

Category

Acoustics (signal)

Perception

Aperiodicity (Redi and Shattuck-Hufnagel 2001), aperiodic voice (Keating et al. 2015)

Irregular periods

Fundamental frequency

F0 not detectable

‘Consonantal’ sound and/or low pitch (but audible)

Creak (Redi and Shattuck-Hufnagel 2001),

vocal fry (Keating et al. 2015)

F0 drop or constant low F0 level, damping of individual periods

Diplophonia (Redi and Shattuck-Hufnagel 2001), multiply pulsed voice (Keating et al. 2015)

Alternative shape of the signal, amplitude and duration of successive periods + sometimes presence of two prominent excitatory peaks in the signal

Rough voice + perception of a distinct pitch ≈ 1 octave <F0 for a nearby modal region (which correspond to a period doubling)

Glottal squeak (Redi and Shattuck-Hufnagel 2001)

High F0 + low amplitude

Instantaneous shift to a relatively high-pitched, low amplitude voice quality

Non constricted voice (Keating et al. 2015)

Low and irregular F0

Tense voice (Keating et al. 2015)

Mid or high F0 but regular

8In the present study, we will keep the generic term ‘creaky voice’.

9Let us now briefly present the main functions performed by creaky voice.

1.2. Linguistic functions of creaky voice

10Creaky voice in certain languages (some Chinese / Sinitic languages, English, German, Danish, among others; see Schuller and Batliner 2013) contributes to typical linguistic phenomena both on the segmental and suprasegmental levels.

1.2.1. Segmental

11At the segmental level, creaky voice may contribute to marking phonological contrast and/or allophonic variation (Dilley 1996; Ladefoged and Maddieson 1996). According to Laver (1980: 126), for example, "in Danish, two such words as hun "she" and hund "dog" are pronounced alike except for a difference of register, the second having creaky voice."

12In some languages, or certain specific language varieties, it can be assimilated to glottalisation, as shown by the following examples. More particularly, in most varieties of American English and British English, voiceless stops – most frequently /t/ – are often glottalized in syllable-coda position, e.g. jacket, a lot, where the glottalized segments function as allophones of final /t/ (coextensive glottalisation, glottal reinforcement or replacement). Garellek (2015: 822) specifies that in American English, t-glottalisation can be produced “as a glottal stop and/or with creaky voice”. In African American English (AAE), glottalisation has also been described as a form of enhancement or replacement of final /d/ in stressed syllables (e.g. Fasold 1981; Kohl and Anderson 2000). In some British English varieties (Cockney, Estuary English, basilects of East Anglia or Bristol), glottal stops may even occur in intervocalic word-medial contexts following an accented vowel (e.g. button, Saturday: Cruttenden 2001: 170).

13In North American English, allophony can also be performed through glottal insertion or a glottal stop [ʔ] in contexts of a vowel-initial word preceded by another vowel and when occurring in phrase-initial position (Dilley et al. 1996). Redi and Shattuck-Hufnagel (2001: 409) also reported a “high rate of word-initial-vowel glottalisation at a vowel-vowel hiatus” as in German (zu [ʔ] erhalten, Rodgers 1999: 222) where initial vowels can also canonically be realized with a glottal stop ([ʔiç] for ich).

14In Romance languages, however, glottalisation is relatively less present at the segmental level (see Kohler 2000 for Italian and French). Glottalisation is rarely found in word-initial position (except under emphatic stress, see in 1.2.2. below, or in hiatus, especially with identical or articulatorily similar vowels, as in "Sara a adoré" /saʀaaadoʀe/) and final stops are not reported to be glottalized.

1.2.2. Suprasegmental

15At the suprasegmental level, creaky voice can play the roles of:

16- a prosodic, demarcation marker (i.e. marking the right or left boundary of a prosodic unit). For example, it emerges in certain syntactic/prosodic positions in English, that is, there are more creaky voice occurrences at intonational phrase onsets and also at the end of utterances (Redi and Shattuck-Hufnagel 2001; phrase-final creak in Garellek 2015) or else at word-initial level (vowel initial words, Henton and Bladon 1988; Garellek 2015). In French, creaky voice can occur at phrase-final position, especially in case of interruption (self-correction or new speech turn by interlocutor, truncation glottalisation; Kohler 2000).

17- an accentual marker as it is observed in the environment of pitch accents (Henton and Bladon 1988). Garellek (2015: 822) also shows how lexical stress comes into play when he underlines that “word-initial glottalisation is the phenomenon by which stressed and/or accented vowels are creaky and/or preceded by glottal stops, e.g. “apple” pronounced as [ˈʔæpl]”.

18- a marker of focus or emphasis: for example Carton et al. (1976) described it as l’accent d’insistance in French, e.g. tu es vraiment incroyable!

19- a word type marker in some languages, like in German, where Rodgers (1999) showed that content words are more glottalized than function words (in spontaneous and read speech).

20Contrary to what happens at the segmental level, creaky voice or phrase-final creak (Garellek 2015) may span several syllables or words at the suprasegmental level: “STEVEN was the one who said the word button.”

1.3. The pragmatic, stylistic and attitudinal functions of creaky voice

21Creaky voice also relates to some pragmatic phenomena, such as marking hesitation or else managing speech turns. For example, Candea et al. (2005) showed that vowels in hesitation fillers are produced with a more unstable voice quality than intra-lexical vowels. As for the perceptual aspect of hesitation in synthesized speech, the presence of creaky voice occurrences allows a better perception of hesitation, but only in the first position within a phrase (Carlson et al. 2006).

22In the course of an interaction, creaky voice signals the end of a speaker’s turn (Laver 1980) and it therefore performs a ‘turn-yielding’ as Ogden (2001: 139) demonstrated for Finnish and English talk-in-interactions.

23Creaky voice is also sensitive to speaking style, register and speaking task effects. Ding et al. (2004, 2006) reported that, despite inter-subject differences, clear speech during careful reading increased the frequency of creaky voice at intervocalic hiatus in American English. In the same line, Rodgers (1999) found that glottal stops with or without glottalisation were more frequent in read than spontaneous speech while glottalisation “with no glottal reflex is more frequent in spontaneous than read speech” (Rodgers 1999: 194).

24Interactional effects pertaining to creaky voice alignment/accommodation/entrainment have also recently been reported showing that speakers use more creaky voice when interacting with an interlocutor who is a frequent creaker, as reported in Borrie and Delfino (2017) for instance. They showed that American women engaged in spoken dialogue “employed significantly more vocal fry when conversing with the partner who exhibited substantial vocal fry than when conversing with the partner who exhibited quantifiably less vocal fry” (p. 513).

25Finally, the attitudinal/emotional function of creaky voice was reported by Ishi et al. (2005), who showed that it was prevalent during the expression of surprise, admiration and suffering, and by Gobl and Ní Chasaide (2003), who described it as typical of sad, bored or relaxed voice.

1.4. The sociolinguistic and extralinguistic functions of creaky voice

26In certain cases, creaky voice may also be a marker of social identity. For example, Wolk et al. (2012) found that two thirds of female college students speaking Standard American English used creaky voice in their sentence reading, perhaps to match celebrities and popular figures. Their hypothesis is that some social settings – like undergraduate studies in the US – favour the extended use of creaky voice. According to Melvin and Clopper (2015: 1), “women use creaky voice more often than men because it is more salient in their speech and may thus more easily carry social meaning”.

27Creaky voice occurrence may depend on the speaker’s gender, although Redi and Shattuck-Hufnagel (2001: 409) point to the “conflicting results about whether males or females glottalise more frequently” emerging from various studies on different dialects, speakers or positions in the utterance. Creaky voice is considered “as a new type of female voice quality” (Yuasa 2010: 315).

28The speaker’s laryngeal disorders may also be associated with the production of creaky voice: Ylitalo and Hammarberg (2000) found that “low pitch, monotony and hyperfunction in combination with vocal fry may lead to voice problems” (Ylitalo and Hammarberg 2000: 563). Speech therapy can help to address these disorders.

29Finally, the production of creaky voice may depend on other extralinguistic contextual factors such as sporadic consumption of caffeinated drinks or speaking in an environment with background noise (Cantor-Cutiva et al. 2018).

1.5. Research questions

  • 1 On average, each pair interacted for about 25 minutes, with a language switch after approximatel (...)

30In light of the results obtained in the previous research presented above, we set out to carry out an innovative study about the production of creaky voice by speakers in tandem interaction, i.e. a collaborative set-up where two native speakers of two different languages collaborate to learn each other’s mother tongue by, in turn, using their first language but also by speaking in the foreign language they want to acquire. We were particularly interested in French/English tandem conversations as we aim to explore the interactional dynamics induced by this language contact situation (language switch after 10 mins approximately1), especially given the very different role played by creaky voice in English and French.

31Ultimately, the research questions we would like to address are the following: what is the evolution of creaky voice use, for a given language, in the context of two languages coming in contact (tandem interaction)? How does creaky voice use vary in the course of the interaction between the tandem partners?

32More specifically, in the limited framework of the present pilot study which is based on an English/French tandem reading task, we will address this specific question: how does creaky voice use vary according to: i) the language status / identity (the speaker’s first (L1) or foreign language (L2)), ii) the language spoken / language condition (French / English), iii) the speaker / interlocutor, iv) the segmental and the prosodic contexts?

2. Method

2.1. Corpus and participants

33The speech data used in this pilot study is the SITAF corpus (Horgues and Scheuer 2015), which corresponds to a very specific communicative setting: it contains spoken tandem interactions in French and in English produced by 21 ‘tandem pairs’. Each tandem pair was made up of a native-English student (coded A01 to A21) and a native-French student (coded F01 to F21) conversing informally in the two languages on a weekly basis in the course of the 3-month university term. Through their tandem exchanges, they aimed to learn more about each other’s native language and culture.

34The tandem pairs met 12 times on average, but video-recorded data with set speech tasks were collected on two occasions: session 1, which was organized a week after the tandem pairs met for the first time, and session 2, which took place at the end of their 3-month tandem exchange. The recordings were performed at the University’s recording studio (with 2 microphones positioned 30 cm above each speaker’s head, and 3 video cameras).

  • 2 The control semi-spontaneous tasks were performed at the start of session 1 to serve as a practi (...)

35The same speech tasks by the same tandem pairs of speakers (L1/L2) were recorded during the two recording sessions in the two languages (language switch after 10 mins): semi-spontaneous narrative and argumentative production tasks (see Horgues and Scheuer 2015 for more details), and a monitored reading task. Control interactions (L1/L1, with an interlocutor who was a fellow native speaker) were also collected2.

36Metadata were collected through written questionnaires upon registration for and completion of the tandem exchange. In the registration questionnaires, the participants gave information about their linguistic background (ages of acquisition of the various languages they spoke), and rated self-reported L2 proficiency scores (scores out of 10 for written and oral production skills but also written and oral comprehension in the L2). They also shared reasons and motivations for participating in the tandem program and preferred conversation topics. The two language groups (native-English and native-French speakers) self-rated similarly in their respective L2 overall proficiency score (6.9 and 7.2/10 respectively).

37In the post-exchange questionnaires, the participants gave introspective information about their tandem experience (perceived usefulness, self-reported progress and obstacles, etc.).

38In the framework of this pilot study, we focus solely on the analysis of the collaborative reading task in session 2 as it illustrated L2 reading competence after 3 month-experience of tandem learning and also because it contained both L1 and L2 readings of the same text (The North Wind and the Sun/ La bise et le soleil, see the Appendix). The general instructions the participants received before each recording session highlighted the need for solidarity and mutual assistance between tandem partners:

As for all tandem sessions, the aim of the activity is for you to learn something about your foreign language through interacting with your partner, but also for you to teach your partner something about your mother tongue, which will help him or her to learn it better. Solidarity and mutual assistance are therefore essential. You will need to make sure that both languages are used in a balanced way (50%-50%). When your partner speaks in English, let them do so as much as possible. However, feel free to help or correct them if they can’t find the right word or expression, or if you think what they are saying needs correcting.

39The instruction for the monitored reading task was as follows in session 1:

Read the following text twice:
- once with your tandem partner helping you especially if he/she does not understand what you are saying or if your reading is unclear
- and then a second time on your own (no interruption).

40In the second session, collaborative reading consisted in a simple reading (unmonitored, no second repetition) of the text by the two partners in their respective L2s first, and then in their respective L1s (control reading). Partners chose the order of languages randomly.

Read the text in your respective foreign languages first. Then read it again in your first languages.

41Out of the 21 tandem pairs in the corpus, 9 were randomly selected for analysis (pairs 01, 02, 03, 04, 06, 07, 08, 09, 10). The 18 participants were distributed as follows:
- 9 native-French students, all women, who studied English and/or Literature at University and whose mean age was 18.9 years old (standard deviation SD=1.5). English was the L2 for most of them (L3 for F03). They had studied it since the age of 10.4 years old on average. Their self-rated overall proficiency score in English was 7/10 (respectively 7 and 7.6/10 in oral and written comprehension; 6.7 and 6.7 in oral and written productions);
- 9 native-English students, 8 women and 1 man, learning French at Sorbonne Nouvelle University as part of the international mobility program. They came from a variety of English-speaking countries (2 from the UK, 1 from Ireland, 1 from Costa Rica/USA, 5 from the USA) and their mean age was 20.6 years old (SD=1.4). French was an L2 for all of them except A08, who studied it as an L4. They had studied it since the age of 14.4 years old on average. Their self-rated overall proficiency score in French was 6.6/10 (respectively 6.8 and 7.2/10 in oral and written comprehension; 5.9 and 6.4 in oral and written productions).

42The audio files of these 9 pairs’ reading tasks in session 2 were analyzed (i.e. L2 French, L2 English, L1 French, L1 English).

2.2. Manual speech segmentation and coding of creaky voice occurrences

43All the reading files under study were manually segmented in Praat (Boersma and Weenink 2018) and annotated in 8 tiers (see fig. 1) which were coded for:
1. The start and end boundaries of each creaky voice occurrence (C)
2. The acoustic category of each creaky voice occurrence
3. The prosodic position
4. Creaky voice segment (example: [i] at the end of the word “qui”: hiatus)
5. The syllables contained in the creaky voice occurrence
6. The word(s) covered in the creaky voice occurrence
7. The corresponding sentence
8. The speaker, language (French/English) and language status (L1/L2)

Figure 1: Example of a Praat signal, spectrogram, fundamental frequency curve and Textgrid of a sample of the text “qui arriverait” read in L2 French by native-English speaker A01, showing the details of the 8 annotation tiers.

Figure 1: Example of a Praat signal, spectrogram, fundamental frequency curve and Textgrid of a sample of the text “qui arriverait” read in L2 French by native-English speaker A01, showing the details of the 8 annotation tiers.

44The identification of creaky voice occurrences was performed through a combination of auditory inspection (1 rater) and acoustic visualization of the speech signal (the same annotator).

45Note that following the typology used in Redi and Shattuck-Hufnagel (2001) and Keating et al. (2015, see table 1 in part 1.1.), tier 2 is meant to be dedicated to the different acoustic categories of creaky voice. This part of the analysis will not be presented in the current study, as it still needs to be generalized systematically to all eighteen speakers.

46Data analysis was conducted following several dependent and independent variables.

47Dependent variables included: i) the relative proportion of creaky voice (percentage, i.e. the number of syllables with creaky occurrences divided by the total number of syllables in the utterance x 100), ii) the relative duration of the creaky voice occurrences (i.e. the duration of each creaky occurrence divided by the total reading time duration x 100, excluding inter-utterance pauses).

48Independent variables included: i) language status / identity (speaking one’s L1 or L2), ii) language condition (speech produced in French or English), iii) the two speaking partners interacting in the same pair, iv) the prosodic context, v) the segmental context (for the English data only).

49Our pilot study aims at describing the main creaky voice phenomena observed.

3. Results

3.1. Examples of creaky voice occurrences in the corpus

50Figures 2 and 3 illustrate examples of occurrences of creaky voice in a native-English speaker and in a native-French speaker reading the words "d’accord" (Figure 2) and "a ôté" (Figure 3): the signal and the spectrogram show that only the productions of the native-English subject show creaky voice occurrences where, for aperiodicity, the fundamental frequency (F0) is not detected, or, for the diplophonia, it is erroneously partially detected (F0 lowered by one octave).

Figure 2: Top to bottom: signal with periodicity pulses detected by Praat, spectrogram, fundamental frequency (F0) curve, and tiers of the word “d’accord” produced with modal voice by speaker F01 (L1, left) and with a diplophonia creaky voice by female speaker A01 (L2, right).

Figure 2: Top to bottom: signal with periodicity pulses detected by Praat, spectrogram, fundamental frequency (F0) curve, and tiers of the word “d’accord” produced with modal voice by speaker F01 (L1, left) and with a diplophonia creaky voice by female speaker A01 (L2, right).

Figure 3: Top to bottom: signal with periodicity pulses detected by Praat, spectrogram, fundamental frequency (F0) curve and tiers of the words (hiatus) “a ôté” produced with modal voice by speaker F01 (L1, left, except the end of “ôté”), and with a creak and then aperiodicity creaky voice by speaker A01 (L2, right).

Figure 3: Top to bottom: signal with periodicity pulses detected by Praat, spectrogram, fundamental frequency (F0) curve and tiers of the words (hiatus) “a ôté” produced with modal voice by speaker F01 (L1, left, except the end of “ôté”), and with a creak and then aperiodicity creaky voice by speaker A01 (L2, right).

3.2. Proportion of creaky voice occurrences

3.2.1. Overall values

51As can be seen from figures 4 and 5, the proportion of creaky voice occurrences is higher for the native-English than the native-French subjects, and it is higher in English than in French (except for F08, A01, A06; figure 5). This mean value is significantly higher in L1 English than in L1 French (Mann-Whitney unpaired test: z9=-2.6; p=0.009), and, for each native-French subject, it is significantly higher in English L2 than French L1 (paired Wilcoxon test: z9=-2.5; p=0.01; figure 4).

Figure 4: Mean and standard deviation of the proportion of creaky voice occurrences (%) for the native-French and the native-English subjects in function of language status (L1 or L2)

Figure 4: Mean and standard deviation of the proportion of creaky voice occurrences (%) for the native-French and the native-English subjects in function of language status (L1 or L2)

52Naturally, one can observe inter-subject variability (figure 5). For example, subjects F08 and A02 (an American woman) show minimal proportions of creaky voice occurrences, whereas subjects F09, A01 (an Irish woman) and A06 (a British woman) show maximal values. The American man (A09) does not show any peculiarities in terms of proportion of creaky voice occurrences.

Figure 5: Proportion of creaky voice occurrences (%) for each native-French (left) and native-English subject (right) in function of language status (L1 or L2). Boxed speakers show extreme values.

Figure 5: Proportion of creaky voice occurrences (%) for each native-French (left) and native-English subject (right) in function of language status (L1 or L2). Boxed speakers show extreme values.

3.2.2. Correlations between the proportions of creaky voice occurrences in L1 and L2 speech of the same speaker

53Figure 6 shows significant positive correlations between the proportion of creaky voice occurrences in L1 and L2 speech of the same speaker, especially for the native-French speakers: the higher the number of creaky voice occurrences in one language, the higher the number of occurrences in the other language spoken by a given subject.

Figure 6: Correlation and regression between the proportion of creaky voice occurrences (%) for L2 in function of L1 spoken by the same speaker: left: native-French speakers, right: native-English speakers.

Figure 6: Correlation and regression between the proportion of creaky voice occurrences (%) for L2 in function of L1 spoken by the same speaker: left: native-French speakers, right: native-English speakers.

3.2.3. Correlations between the proportions of creaky voice occurrences in the speech of the two tandem partners in the same language condition

54Figure 7 shows a negative correlation between the number of creaky voice occurrences in the same language condition for the native-French and the native-English speaker from the same tandem pair, especially in English: the more creaky voice occurrences in the L1 English speech of the native-English speaker, the fewer occurrences in the L2 English speech of the native-French partner. However, these correlations are not significant.

Figure 7: Correlation and regression between the proportion of creaky voice occurrences (%) for a speaker’s L2 in function of his/her partner’s L1 (within each tandem pair. Left: French language condition, right: English language condition.

Figure 7: Correlation and regression between the proportion of creaky voice occurrences (%) for a speaker’s L2 in function of his/her partner’s L1 (within each tandem pair. Left: French language condition, right: English language condition.

3.3. Relative duration of creaky voice occurrences

3.3.1. Native-French speakers

55Figure 8 shows the relative duration of creaky voice occurrences for the native-French speakers depending on the language spoken. We can see that the creaky voice occurrences are slightly longer in English (L2) than in French (L1) for 3 out of the 9 native-French speakers (F01, F04 and F08) but these differences are statistically non-significant. Inter-speaker variability is relatively low in this portion of the data. Speaker F09 shows significantly longer creaky voice occurrences in French than in English.

Figure 8: Boxplot showing interquartile range (box), median (horizontal line in box) and extreme values (whiskers) of the relative duration of creaky voice occurrences (ms) in a second for the native-French speakers depending on the language spoken.

Figure 8: Boxplot showing interquartile range (box), median (horizontal line in box) and extreme values (whiskers) of the relative duration of creaky voice occurrences (ms) in a second for the native-French speakers depending on the language spoken.

3.3.2. Native-English speakers

56As we can see in Figure 9, overall, the creaky voice occurrences are longer in the case of the native-English than the native-French speakers, especially when reading English. The occurrences are markedly longer in English (L1) than in French (L2) for 8 out of the 9 subjects. Inter-speaker variability is higher than in the case of the native-French speakers. There are great and significant differences between the L1 (English) and the L2 (French) speech of speakers A01, A03, A04, A06, A08 and A09 (male), in contrast to A10, whose creaky voice occurrences are non-significantly longer in L1 than in L2.

Figure 9: Boxplot showing interquartile range (box), median (horizontal line in box) and extreme values (whiskers) of the relative duration of creaky voice occurrences (ms) in a second for the native-English speakers in function of the language spoken.

Figure 9: Boxplot showing interquartile range (box), median (horizontal line in box) and extreme values (whiskers) of the relative duration of creaky voice occurrences (ms) in a second for the native-English speakers in function of the language spoken.

3.4. Prosodic context

3.4.1. English

57Figure 10 shows the percentage of creaky voice occurrences in English for the native-English and the native-French participants in function of the prosodic position. There are no major differences between the two groups of speakers in terms of the percentage of creaky voice occurrences they show in the various prosodic environments. The majority of creaky voice occurrences are found in stressed syllables (e.g. “was obliged”; “to confess”), tonic syllables (e.g. “fold his cloak around him”) and utterance-final position (e.g. “stronger of the two”), especially for the native-French speakers.

Figure 10: Percentage of creaky voice occurrences in English for the native-English and the native-French subjects in function of the prosodic position.

Figure 10: Percentage of creaky voice occurrences in English for the native-English and the native-French subjects in function of the prosodic position.

3.4.2. French

  • 3 The native-English speaker pronounced « une » instead of « un » for « un voyageur ».

58As can be seen in figure 11, there are more creaky voice occurrences at the end of intonation units in the speech of the native-French, compared to the native-English speakers (e.g. “La bise et le soleil se disputent, chacun assurant qu’il était le plus fort, quand ils ont vu une3 voyageur qui s’avançait, enveloppé dans son manteau.”). The native-French speakers show fewer creaky voice occurrences than their Anglophone counterparts at the beginning of words starting with a vowel (e.g.Alors le soleil a commencé à briller.”). On the other hand, the two populations produced virtually the same number of creaky voice occurrences in hiatus contexts (e.g. “Alors le soleil a commenc briller.”; “…le voyageur, réchauffé, a ôté son manteau.”).

Figure 11: Percentage of creaky voice occurrences in French for the native English (English NSs) and the native French subjects (French NSs) in function of the prosodic position. “Other”: end of a non-final word, after a consonant, grammatical word (e.g. “la bise a renoncé…”).

Figure 11: Percentage of creaky voice occurrences in French for the native English (English NSs) and the native French subjects (French NSs) in function of the prosodic position. “Other”: end of a non-final word, after a consonant, grammatical word (e.g. “la bise a renoncé…”).

3.5. Segmental context in English

59At this stage of our analysis, the segmental context has only been studied for the English language condition.

60Figure 12 shows the percentage of creaky voice occurrences in English for the native-English and the native-French participants in function of the segmental context. We observe similar tendencies for both groups of speakers in the various segmental contexts. The occurrences of creaky voice may concern only one segment, as is the case with the glottalisation of /t/ (e.g. "shone out warmly"), or with segments just before a sonorant consonant (e.g. "around" where the vowel preceding the /n/ is glottalized in this form of creaky voice). The context in which there are more creaky voice occurrences than anywhere else is before a pause, especially for the native-French speakers.

Figure 12: Percentage of creaky voice occurrences in English for the native-English and the native-French subjects in function of the segmental context.

Figure 12: Percentage of creaky voice occurrences in English for the native-English and the native-French subjects in function of the segmental context.

4. Discussion

61In the preceding section, we demonstrated how creaky voice use in our database of read speech varies as a function of the selected variables.

62The proportion of creaky voice occurrences depends on the speakers’ mother tongue (speaker group). It is higher in the speech of the native-English participants than the native-French participants. The former group also produces longer creaky voice occurrences than their French counterparts. Indeed, creaky voice does not appear in the same proportion in French and English (Esling and Wong 1983).

63The proportion of creaky voice occurrences depends less on the language status (L1 or L2) than the language condition: it is higher in English than in French, whether spoken as L1 or L2. There are also longer creaky voice occurrences in the English data, both L1 and L2, than in the French data. This ties in with the results obtained by Benoist-Lucy and Pillot-Loiseau (2013), who found that American speakers learning French showed significantly more and longer creaky voice occurrences in English than in French.

64In addition, we found a significant positive correlation between the L1 and the L2 speech of the same speaker (figure 6), which could point to either transfer of creaky voice from L1 to L2 (native-English speakers), or reverse transfer (L2 to L1; native-French speakers) after 3 months of interaction. In other words, the ‘creak’ habits developed in L2 – e.g. through accommodation to the tandem partner – might influence one’s voice quality in L1 (native-French speakers).

65What about the interaction between participants in a pair? We did not find any significant correlation between the French spoken by the native-French subjects, and the French spoken by the native-English subjects of the same pair. We found a non-significant negative correlation between the creaky voice behaviour of the two tandem partners in the English condition and in their respective L1 and L2 (figure 7). This is an unexpected but non-significant result, which runs counter to the findings reported by Borrie and Delfino (2017), regarding convergence and accommodation: people “align their behaviors to more closely match those of their conversational partner” (Borrie and Delfino 2017: 7). However, their analysis concerned measures of relative durations of creaky voice occurrences in a dialogue and reading tasks, presented together in their results, whereas we are dealing here with the proportion of creaky voice occurrences in a reading task only.

  • 4 However, we still have to quantify this number of hesitations by utterance and by participant, i (...)

66There is variation between speakers (as attested by e.g. Redi and Shattuck-Hufnagel 2001), in terms of both the relative proportion and the duration of creaky voice occurrences. The smallest proportion was produced by speakers F08 and A02 (who had been living in France for 7 months), and the highest proportion by F09, A01 (who studied French as an L3; their self-rated overall proficiency score in French was 5/10 in oral productions) and A06 (self-rated overall proficiency score in French: 5/10 in oral productions). These last two participants claimed to have a lower level of oral production in their foreign language than the other learners, and we found more hesitation in their French reading tasks4. Indeed, hesitation can precisely, among other things, cause the production of creaky voice occurrences (Candea et al. 2005; Carlson et al. 2006).

67The maximal durations are found in the speech of three American participants, A08 (who studied French as an L4 and had been in France for only one week at the time of the first recording), A09 (the male subject) and A10 (who had been in France for only five weeks at the time of the first recording). Overall, duration-wise, there is less inter-speaker variability in the case of the native-French than the native-English participants.

68The inter-subject variability that we obtained in our results can therefore be explained by certain aspects of the linguistic profile of our subjects, but also by other individual features/factors relating to their voice, such as the fundamental frequency. Figure 13 shows the correlations and regressions between the proportion of creaky voice occurrences and the mean fundamental frequency (F0, Hz) for each native-French and native-English speaker (NSs) in each language. There are negative but non-significant correlations: the more occurrences of creaky voice, the lower F0, except for English NSs in English. Indeed, “Creaky voice is associated with lowered fundamental frequency values (relative to modal phonation) in many languages, synchronically or diachronically” (Gordon and Ladefoged 2001: 400). This result suggests that, overall, creaky voice is much less frequent when speech is produced with greater effort and a higher fundamental frequency (Olson 2012).

Figure 13: Correlation and regression between the proportion of creaky voice occurrences (%) and the mean fundamental frequency (F0, Hz) for each native-French and native-English participant (NSs, the A09 man excluded) in each language.

Figure 13: Correlation and regression between the proportion of creaky voice occurrences (%) and the mean fundamental frequency (F0, Hz) for each native-French and native-English participant (NSs, the A09 man excluded) in each language.

69Furthermore, the segmental contexts which favour creaky voice occurrences the most are before a pause and, to a lesser extent, before voiceless consonants. This tallies with the data previously reported in the literature, e.g. Dilley et al. (1996). Garellek (2014) highlighted such glottalisations in the same phonetic contexts in American English.

70Concerning the prosodic context, in the English data, the majority of creaky voice occurrences are found in lexically stressed syllables, tonic syllables and utterance-final position (in accordance with Redi and Shattuck-Hufnagel 2001), especially in the productions of the native-French speakers. As for the French data, creaky voice occurred mostly at the end of intonation units and before pauses in the case of the native-French speakers, and at the beginning of words starting with a vowel for the native-English speakers. According to Henton and Bladon (1988), the demarcative intonational function of creaky voice is observed cross-linguistically.

71The two populations produced virtually the same number of creaky voice occurrences in a hiatus context, which was not expected for the native-French speakers: French speakers usually produce in their L1 hiatuses (Encrevé 1983), but their productions may be influenced by the Anglo-American model: this consists in individualizing the words by linking them less, with the intention of being heard to be well understood (Encrevé 1983). We assume that the reading task, more controlled than the spontaneous speech, favours these efforts to “detach” words.

72Our study has several limitations, such as the failure to take into account the specific regional origin (state/region/local accent) of our native-English subjects, which could be a factor of variation in the proportion of creaky voice occurrences (Redi and Shattuck-Hufnagel 2001).

73Another limitation is the lack of electroglottographic (EGG) recordings of the speech productions of our populations. Finally, we will need to refine our results using Analysis of Variance to determine whether there is an interaction between the factors “language”, “language status”, “speaker”, “segmental contexts” and “suprasegmental contexts”.

5. Conclusion

74In our pilot study, creaky voice use is more extensive in English than in French and varies little according to the status of these two languages (L1 or L2). Moreover, there is considerable inter-speaker variability that can be partially explained by each individual's linguistic profile, but also by other physiological characteristics such as the fundamental frequency. In addition, the significant positive correlation between the L1 and the L2 speech of the same speaker means that the ‘creak’ habits developed in L2 – perhaps through accommodation to the tandem partner – might influence one’s voice quality in L1, even in a controlled task as reading action. Finally, although varying in terms of segmental factors in French or English, the use of creaky voice is favoured by certain prosodic contexts such as in stressed and/or tonic syllables and by the final position in intonation units, which may make intonational creaky voice a candidate for a phonetic universal (Henton and Bladon 1988).

6. Perspectives

75The present study being limited to 9 pairs of tandem partners recorded in the SITAF corpus, a natural next step in our study will be to extend the analysis to the remaining 24 project participants. In addition to text reading, which we have reported on thus far, our aim is to ultimately analyze also the linguistic data obtained in the other two tasks, i.e. semi-spontaneous speech, while considering the same parameters and factors as with the read data.

76A comparison between the two recording sessions – i.e. session 1, which took place before the tandem exchanges properly started, and session 2 organized 3 months later – is bound to offer precious insights into how the frequency and the amount of creaky voice evolve in the L2 speech of the French learners of English, in the wake of an average of 12 weekly tandem meetings.

77Including other annotators of creaky voice occurrences will also enrich our study by rendering it more objective and comprehensive. The same goes for elaborating the acoustic measurements of the creaky voice occurrences, which we are planning on embarking upon in the future.

78We also have to take into account the effect of hesitation (Candea et al. 2005; Carlson et al. 2006) and other pragmatic, sociolinguistic and extralinguistic factors on creaky voice occurrences.

79Finally, our objective is also to examine the distribution of different acoustic patterns associated with the creaky voice occurrences produced by different subjects, such as aperiodicity, creak, diplophonia, or glottal squeak.

7. Acknowledgments

80This work is part of the programme "Investissements d’Avenir" overseen by the French National Research Agency, ANR-10-LABX-0083 (Labex EFL).

81The SITAF corpus was collected and finalized thanks to the SITAF research team, the SITAF participants, the studio technicians, the research laboratory Sesylia/Prismes (EA 4398), as well as the funding provided by the Commission de la Recherche (Université Sorbonne Nouvelle), and the Ortolang-Ircom programme.

Top of page

Bibliography

Abercrombie, David. Elements of General Phonetics. Edinburgh: Edinburgh University Press, 1967.

Abdelli-Beruh, Nassima B., Wolk, Lesley, and Dianne Slavin. "Prevalence of Vocal Fry in Young Adult Male American English Speakers." Journal of Voice 28.2 (2014): 185-190.

Abdelli-Beruh, Nassima B., Drugman, Thomas and R.H. Red Owl. "Occurrence Frequencies of Acoustic Patterns of Vocal Fry in American English Speakers." Journal of Voice 30.6 (2016): 759.e11-759.e20.

Benoist-Lucy, Agathe and Claire Pillot-Loiseau. "The Influence of language and speech task upon creaky voice use among six young American women learning French." Interspeech 2013. International Speech Communication Association, 2013: 2395-2399.

Boersma, Paul and David Weenink. "Praat: doing phonetics by computer" [Computer program]. Version 6.0.37, retrieved 14 March 2018 from http://www.praat.org/.

Borrie, Stephanie A. and Christine R. Delfino. "Conversational Entrainment of Vocal Fry in Young Adult Female American English Speakers." Journal of Voice 31.4 (2017): 513.e25-513.e32

Candea, Maria, Vasilescu, Ioana and Martine Adda-Decker. "Inter-and intra-language acoustic analysis of autonomous fillers." DISS 05, Disfluency in Spontaneous Speech Workshop. 2005: 47-52.

Cantor-Cutiva, Lady Catherine, Bottalico, Pasquale and Eric Hunter. "Factors associated with vocal fry among college students." Logopedics Phoniatrics Vocology 43.2 (2018): 73-79.

Carlson, Rolf, Gustafson, Kjell and Eva Strangert. "Cues for Hesitation in Speech Synthesis." Ninth International Conference on Spoken Language Processing, Interspeech. 2006: 1300-1303.

Carton, Fernand, Hirst, Daniel, Marchal, Alain and André Seguinot. L'accent d'insistance / Emphatic stress. Montréal: Didier, 1976.

Cruttenden, Alan. Gimson's Pronunciation of English. London: Arnold, 2001.

Dilley, Laura, Shattuck-Hufnagel, Stefanie and Mari Ostendorf. "Glottalization of word-initial vowels as a function of prosodic structure." Journal of phonetics 24.4 (1996): 423-444.

Ding, Hongwei, Jokisch, Oliver and Rüdiger Hofmann. "Glottalization in inventory construction: A cross-language study." Chinese Spoken Language Processing, 2004 International Symposium on. IEEE, (2004): 37-40.

Ding, Hongwei, Oliver Jokisch, and Rüdiger Hoffmann. "The Effect of Glottalization on Voice Preference." Proceedings of Speech Prosody. 2006: 1-4.

Encrevé, Pierre. « La liaison sans enchaînement. » Actes de la recherche en sciences sociales 46.1 (1983): 39-66.

Esling, John H. and Rita F. Wong. "Voice Quality Settings and the Teaching of Pronunciation." TESOL Quarterly 17.1 (1983): 89-95.

Fasold, Ralph W. "The Relation between Black and White Speech in the South." American Speech 56.3 (1981): 163-189.

Garellek, Marc. "Voice quality strengthening and glottalization." Journal of Phonetics 45 (2014): 106-113.

Garellek, Marc. "Perception of glottalization and phrase-final creak." The Journal of the Acoustical Society of America 137.2 (2015): 822-831.

Gibson, Todd A. "The Role of Lexical Stress on the Use of Vocal Fry in Young Adult Female Speakers." Journal of Voice 31.1 (2017): 62-66.

Gobl, Christer and Ailbhe Ní Chasaide. "The role of voice quality in communicating emotion, mood and attitude." Speech Communication 40.1-2 (2003): 189-212.

Gordon, Matthew and Peter Ladefoged. "Phonation types: a cross-linguistic overview." Journal of Phonetics 29.4 (2001): 383-406.

Henton, Caroline G. and Anthony W. Bladon. "Creak as a sociophonetic marker". Language, speech and mind: studies in honour of Victoria A. Fromkin. Ed. Hyman, Larry and Charles Li. Stanford: Routledge, 1988. 3-19.

Hollien, Harry and John F. Michel. "Vocal fry as a phonational register." Journal of Speech, Language, and Hearing Research 11.3 (1968): 600-604.

Hollien, Harry. "On Vocal Registers." Communication Sciences Laboratory Quarterly Report. 10.1 (1972): 1-33.

Hollien, Harry, Girard, Gary T. and Robert F. Coleman. "Vocal Fold Vibratory Patterns of Pulse Register Phonation." Folia Phoniatrica et Logopaedica 29.3 (1977): 200-205.

Horgues, Céline and Sylwia Scheuer. "Why Some Things Are Better Done in Tandem." Investigating English pronunciation: Current trends and directions. Ed. Mompeán Jose A. & Fouz-González Jonas. London: Palgrave Macmillan, 2015. 47-82.

Imaizumi, Satoshi and Jan Gauffin. "Acoustical perceptual characteristics of pathological voices: rough, creak, fry, and diplophonia." Ann. Bull. RILP 25 (1991): 109-119.

Ishi, Carlos Toshinori, Hiroshi, Ishiguro and Norihiro Hagita. "Proposal of Acoustic Measures for Automatic Detection of Vocal Fry." Ninth European Conference on Speech Communication and Technology, Interspeech. 2005: 481-484.

Keating, Patricia, Garellek, Marc and Jody Kreiman. "Acoustic properties of different kinds of creaky voice." Proceedings of the 18th International Congress of Phonetic Sciences. Glasgow, UK: the University of Glasgow. ISBN 978-0-85261-941-4. Paper number 0821.1-9 retrieved from http://www.internationalphoneticassociation.org/icphs-proceedings/ICPhS2015/Papers/ICPHS0821.pdf, 2015.

Kohl, Amanda and Bridget Anderson. "Glottalization as a sociolinguistic variable in Detroit." Proceedings of NWAV 29 (2000), 29, Michigan State University.

Kohler, Klaus J. "Linguistic and paralinguistic functions of non-modal voice in connected speech." Proceedings of the 5th Seminar on Speech Production: Models and Data. 2000: 89-92.

Kreiman, Jody. "Perception of sentence and paragraph boundaries in natural conversation." Journal of Phonetics 10.2 (1982): 163-175.

Ladefoged, Peter and Ian Maddieson. The Sounds of the World's Languages. Oxford: Blackwell, 1996.

Laver, John. "The phonetic description of voice quality." Cambridge: Cambridge University Press, 1980.

Melvin, Shannon and Cynthia G. Clopper. “Gender variation in creaky voice and fundamental frequency.” Proceedings of the 18th International Congress of Phonetic Sciences. Glasgow, UK: the University of Glasgow. ISBN 978-0-85261-941-4. Paper number 0320.1-9 retrieved from http://www.internationalphoneticassociation.org/icphs-proceedings/ICPhS2015/Papers/ICPHS0320.pdf, 2015.

Monsen, Randall B. and A. Maynard Engebretson. "Study of variations in the male and female glottal wave." The Journal of the Acoustical Society of America 62.4 (1977): 981-993.

Ogden, Richard. "Turn-holding, turn-yielding and laryngeal activity in Finnish talk-in-interaction." Journal of the International Phonetics Association 31.1 (2001): 139-152.

Olson, Daniel. "The phonetics of insertional code-switching: Suprasegmental analysis and a case for hyper-articulation." Linguistic Approaches to Bilingualism 2.4 (2012): 439-457.

Redi, Laura and Stefanie Shattuck-Hufnagel. "Variation in the realization of glottalization in normal speakers." Journal of Phonetics 29.4 (2001): 407-429.

Rodgers, Jonathan. "Three influences on glottalization in read and spontaneous German speech." Arbeitsberichte des Instituts für Phonetik und digitale Sprachverarbeitung der Universität Kiel (AIPUK) 34 (1999): 177-284.

Schuller, Björn and Anton Batliner. Computational paralinguistics: emotion, affect and personality in speech and language processing. Hoboken: John Wiley & Sons Ltd, 2013.

Wolk, Lesley, Abdelli-Beruh, Nassima B. and Dianne Slavin. "Habitual Use of Vocal Fry in Young Adult Female Speakers." Journal of Voice 26.3 (2012): e111-e116.

Ylitalo, Riitta, and Britta Hammarberg. "Voice Characteristics, Effects of Voice Therapy, and Long-term Follow-up of Contact Granuloma Patients." Journal of Voice 14.4 (2000): 557-566.

Yuasa, Ikuko Patricia. "Creaky voice: A new feminine voice quality for young urban-oriented upwardly mobile American women?" American Speech 85.3 (2010): 315-337.

Top of page

Annex

APPENDIX

The reading text

The North Wind and the Sun
The North Wind and the Sun were disputing which of them was stronger, when a traveller came along wrapped in a warm cloak*.
They agreed that the one who first succeeded in making the traveller take his cloak off should be considered stronger than the other.
Then the North Wind blew as hard as he could, but the more he blew, the more closely did the traveller fold his cloak around him; and at last the North Wind gave up the attempt.
Then the Sun shone out warmly, and immediately the traveller took off his cloak. And so the North Wind was obliged to confess that the Sun was the stronger of the two.
(* a cloak is a type of coat)

La bise* et le soleil
La bise et le soleil se disputaient, chacun assurant qu'il était le plus fort, quand ils ont vu un voyageur qui s'avançait, enveloppé dans son manteau.
Ils sont tombés d'accord, que celui qui arriverait le premier à faire ôter* son manteau au voyageur, serait regardé comme le plus fort.
Alors la bise s'est mise à souffler de toute sa force, mais plus elle soufflait, plus le voyageur serrait son manteau autour de lui; et à la fin, la bise a renoncé à le lui faire ôter.
Alors le soleil a commencé à briller et au bout d'un moment, le voyageur, réchauffé, a ôté son manteau.
Ainsi, la bise a dû reconnaître que le soleil était le plus fort des deux.
(* ici la bise : un vent très froid *ôter : retirer/enlever)

Top of page

Notes

1 On average, each pair interacted for about 25 minutes, with a language switch after approximately 10 minutes of conversation in one language (French or English). The conversation in the other language (another 10 minutes, approximately) was followed by the reading task in both languages, lasting about 5 minutes. In this particular case of read speech, there is a potentially interactive listener, who does intervene on occasions. In this sense, it could be argued that there is an interactant, which makes it a case of interaction.

2 The control semi-spontaneous tasks were performed at the start of session 1 to serve as a practice phase for all the participants to get used to the studio conditions and to integrate the instructions of the story-telling game and of the debating game. The control reading task was not performed until the end of session 2 to ensure that the native speaker’s reading of the text would not influence the L2 learner’s reading performance.

3 The native-English speaker pronounced « une » instead of « un » for « un voyageur ».

4 However, we still have to quantify this number of hesitations by utterance and by participant, in order to study a possible link between these hesitations and the creaky voice occurrences.

Top of page

List of illustrations

Title Figure 1: Example of a Praat signal, spectrogram, fundamental frequency curve and Textgrid of a sample of the text “qui arriverait” read in L2 French by native-English speaker A01, showing the details of the 8 annotation tiers.
URL http://journals.openedition.org/anglophonia/docannexe/image/2005/img-1.jpg
File image/jpeg, 196k
Title Figure 2: Top to bottom: signal with periodicity pulses detected by Praat, spectrogram, fundamental frequency (F0) curve, and tiers of the word “d’accord” produced with modal voice by speaker F01 (L1, left) and with a diplophonia creaky voice by female speaker A01 (L2, right).
URL http://journals.openedition.org/anglophonia/docannexe/image/2005/img-2.jpg
File image/jpeg, 676k
Title Figure 3: Top to bottom: signal with periodicity pulses detected by Praat, spectrogram, fundamental frequency (F0) curve and tiers of the words (hiatus) “a ôté” produced with modal voice by speaker F01 (L1, left, except the end of “ôté”), and with a creak and then aperiodicity creaky voice by speaker A01 (L2, right).
URL http://journals.openedition.org/anglophonia/docannexe/image/2005/img-3.jpg
File image/jpeg, 456k
Title Figure 4: Mean and standard deviation of the proportion of creaky voice occurrences (%) for the native-French and the native-English subjects in function of language status (L1 or L2)
URL http://journals.openedition.org/anglophonia/docannexe/image/2005/img-4.jpg
File image/jpeg, 80k
Title Figure 5: Proportion of creaky voice occurrences (%) for each native-French (left) and native-English subject (right) in function of language status (L1 or L2). Boxed speakers show extreme values.
URL http://journals.openedition.org/anglophonia/docannexe/image/2005/img-5.jpg
File image/jpeg, 224k
Title Figure 6: Correlation and regression between the proportion of creaky voice occurrences (%) for L2 in function of L1 spoken by the same speaker: left: native-French speakers, right: native-English speakers.
URL http://journals.openedition.org/anglophonia/docannexe/image/2005/img-6.jpg
File image/jpeg, 260k
Title Figure 7: Correlation and regression between the proportion of creaky voice occurrences (%) for a speaker’s L2 in function of his/her partner’s L1 (within each tandem pair. Left: French language condition, right: English language condition.
URL http://journals.openedition.org/anglophonia/docannexe/image/2005/img-7.jpg
File image/jpeg, 220k
Title Figure 8: Boxplot showing interquartile range (box), median (horizontal line in box) and extreme values (whiskers) of the relative duration of creaky voice occurrences (ms) in a second for the native-French speakers depending on the language spoken.
URL http://journals.openedition.org/anglophonia/docannexe/image/2005/img-8.jpg
File image/jpeg, 40k
Title Figure 9: Boxplot showing interquartile range (box), median (horizontal line in box) and extreme values (whiskers) of the relative duration of creaky voice occurrences (ms) in a second for the native-English speakers in function of the language spoken.
URL http://journals.openedition.org/anglophonia/docannexe/image/2005/img-9.jpg
File image/jpeg, 48k
Title Figure 10: Percentage of creaky voice occurrences in English for the native-English and the native-French subjects in function of the prosodic position.
URL http://journals.openedition.org/anglophonia/docannexe/image/2005/img-10.jpg
File image/jpeg, 252k
Title Figure 11: Percentage of creaky voice occurrences in French for the native English (English NSs) and the native French subjects (French NSs) in function of the prosodic position. “Other”: end of a non-final word, after a consonant, grammatical word (e.g. “la bise a renoncé…”).
URL http://journals.openedition.org/anglophonia/docannexe/image/2005/img-11.png
File image/png, 62k
Title Figure 12: Percentage of creaky voice occurrences in English for the native-English and the native-French subjects in function of the segmental context.
URL http://journals.openedition.org/anglophonia/docannexe/image/2005/img-12.jpg
File image/jpeg, 224k
Title Figure 13: Correlation and regression between the proportion of creaky voice occurrences (%) and the mean fundamental frequency (F0, Hz) for each native-French and native-English participant (NSs, the A09 man excluded) in each language.
URL http://journals.openedition.org/anglophonia/docannexe/image/2005/img-13.jpg
File image/jpeg, 349k
Top of page

References

Electronic reference

Claire Pillot-Loiseau, Céline Horgues, Sylwia Scheuer and Takeki Kamiyama, « The evolution of creaky voice use in read speech by native-French and native-English speakers in tandem: a pilot study », Anglophonia [Online], 27 | 2019, Online since 25 November 2019, connection on 20 January 2020. URL : http://journals.openedition.org/anglophonia/2005 ; DOI : 10.4000/anglophonia.2005

Top of page

About the authors

Claire Pillot-Loiseau

Université Sorbonne Nouvelle
UMR 7018, LPP
claire.pillot@sorbonne-nouvelle.fr

Céline Horgues

Université Sorbonne Nouvelle
EA 4398, SeSyliA, Prismes
celine.horgues@sorbonne-nouvelle.fr

By this author

Sylwia Scheuer

Université Sorbonne Nouvelle
EA 4398, SeSyliA, Prismes
sylwia.scheuer-samson@sorbonne-nouvelle.fr

Takeki Kamiyama

Université Paris 8
EA 1569, LeCSeL, TransCrit
Takeki.Kamiyama@univ-paris8.fr

Top of page

Copyright

Licence Creative Commons
Anglophonia – French Journal of English Linguistics est mis à disposition selon les termes de la licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Top of page
  • Logo Presses universitaires du Midi
  • Logo DOAJ - Directory of Open Access Journals
  • OpenEdition Journals