Emphatic variation of the labio-velar /w/ in two Jordanian Arabic dialects

Abstract

Much work on emphatic segments in Arabic dialects has focused on primary emphasis. However, secondary emphasis has been less of a target of study. Our research investigates the emphatic variation of the secondarily emphatic labio-velar /w/ between males and females in two Jordanian Arabic sub-dialects: Rural Jordanian Arabic and Urban Jordanian Arabic. Twenty-four native speakers from the two dialects (equally stratified according to their gender and dialect) were asked to read a carrier phrase that included fifteen tri-syllabic words having the sound /w/ in medial position. Our research confirmed that there is a variation in the degree of emphasis based on gender and dialect. More particularly, males produced stronger emphatic segments than females did, and Urban speakers produced weaker empathic segments than Rural speakers did. Results also revealed that the secondarily emphatic /w/ caused the neighboring vowels to have lowered F2 and raised F1 and F3. These findings suggest that emphasis, whether primary or secondary, tend to have the same acoustic correlates. In closing, the theoretical implications of these finding are discussed.

acoustic correlates, gender, emphasis, emphatics, Jordanian Arabic, sociophonetics, pharyngealization, linguistic variation

1. Introduction

Semitic languages in general (e.g., Standard and Spoken Arabic and Modern Hebrew) and Jordanian Arabic (JA) sub-dialects in particular are characterized by the pharyngealization feature known as emphasis (Al-Masri and Jongman, 2004). Emphasis refers to segments that are produced with a primary constriction in the dental/alveolar oral tract and a secondary constriction in the posterior vocal tract (Jongman et al., 2011). Emphasis is an articulatory feature associated with a constriction near the uvula caused by retraction in tongue root [+RTR]. It is also considered a kind of coarticulation (i.e., involving temporal and spatial overlap between the articulatory gestures). This phenomenon has been referred to by several, yet relevant, terms in the literature. Arab grammarians usually refer to this phenomenon as ‘ʔitbɑ̄q’ or ‘ʔistiʕlaaʔ’. It is also referred to as pharyngealization (Wahba, 1993; Davis, 1995; Zawaydeh, 1999), uvularization (McCarthy, 1994), velarization (Obrecht, 1968); Norlin, 1978); Hetzron, 1998), and tongue retraction (Lehn, 1963). As for auditory interpretations, emphasis is referred to as ‘tafkhiim’ or what Almbark (2008) calls “the listeners' deep impression of emphasis” (p. 3). In this study, we use the term emphasis to refer to this phenomenon in Urban Jordanian Arabic (UJA) and Rural Jordanian Arabic (RJA). The phonetic feature of emphasis is considered as the most common feature that is shared among all Arabic varieties.

Emphasis is a cover term that entails two major types: primary and secondary emphasis. Primary emphasis contrasts the emphatic /tˤ/, sˤ/, /dˤ/, /, /ðˤ/, /zˤ/ with their plain counterpart /t/, /d/, /s/, /ð/, /z/ as illustrated in (1).

(1)

a.	ti:n	‘figs’	vs.	tˤi:n	‘mud’
b.	se:f	‘sword’	vs.	sˤe:f	‘summer’
c.	dala:l	‘pamper’	vs.	dˤala:l	‘aberrance’
d.	ðal:	‘humiliated’	vs.	ðˤal:	‘He stayed’

Therefore, the distinction between emphatic and non-emphatic segments is contrastive/phonemic, as the presence/absence of this feature may change the meaning. Primary emphasis contrasts minimal pairs based on the presence/absence of the emphasis feature as illustrated in the examples in (1), and, therefore, it is unpredictable and always in contrastive distribution.

By contrast, secondary emphasis can be further divided into two sub-types: allophonic emphasis and free-variational emphasis. The former either surfaces as a result of assimilation (affected by a neighboring primarily emphatic consonant) as in (2a-2b) or conditioned as it is bound to certain environments. For example, plain /r/ and /l/ can become emphatic in certain environments (conditioned) and thus realized as allophones because they are predictable and in complementary distribution as shown in examples (2c-2e).

(2)

a.	ba:lu	‘his mind’	vs.	ba:ˤlˤa	‘thrift’
b.	tal	‘hill’	vs.	tˤaˤlˤ	‘showed up’
c.	fard	‘individual’	vs.	faˤrˤ	‘ran away’
d.	d͡ʒard	‘inventory’	vs.	d͡ʒaˤrˤ	‘dragged’
e.	ħa:ris	‘guard’	vs.	ħa:ˤrˤ	‘hot’

With regards to the other type of secondary emphasis, that is, free variational, it surfaces as result of dialectal and/or gender variations. For example, the plain phoneme /w/ can be realized as /wˤ/ in certain dialects and by different genders. However, this realization is not conditioned by a certain environment nor is it in contrastive distribution with its plain counterpart /w/. Hence, it is neither a distinct phoneme nor a predictable allophone as shown in the following examples in (3).

(3)

a.	bawadi	‘deserts’	vs.	bˤaˤwˤaˤdi	‘deserts’
b.	d͡ʒaw	‘weather’	vs.	d͡ʒˤaˤwˤ	‘weather’
c.	wadi	‘valley’	vs.	wˤaˤdi	‘valley’

One striking similarity between the two types of emphasis is the directionality of the spreading of emphasis feature to neighboring segments. Primary and secondary (allophonic and free variational) emphasis can be active in one direction or more, and this depends on the emphatic segment position in the word (word-initial, word-medial or word-final), and this spreading is triggered by the low front vowel /a/. Therefore, the directionality can be leftward, rightward, or unrestrictedly bidirectional as long as there is no high vowel (e.g., /i/ or /u/) that blocks the emphasis spreading.

The current study provides answers to the research questions below:

• 1.Do males produce more emphatic /w/ than females do in RJA and UJA?
• 2.Do rural speakers produce stronger emphatic correlates than their urban counterparts do?
• 3.Is vowel duration significantly affected by gender and dialect?

The paper is organized as follows. Section 2 provides an overview on emphasis and pertinent related literature. Section 3 describes the dialects under investigation, namely UJA and RJA. Section four focuses on the sociophonetics of emphasis. Section 5 presents detailed description of the sample, data collection and data analysis procedures. Section 6 lays out the results of the study. Section 7 discusses the results and integrates them with the wider existing literature. Section 8 provides the concluding remarks and an overall summary of the results.

2. Literature review

2.1. Background

The study of emphasis in Arabic dialects in general and in JA (sub-dialects) in particular has recently gained momentum though it is limited to the scrutiny of primary emphasis, where there is a phonemic contrast between emphatic and plain segments (McCarthy, 1994; Zawaydeh, 1999; de Jong and Zawaydeh, 2002, Zawaydeh and De Jong, 2011; Jongman et al., 2011; Davis, 1995; Watson, 1999; Al-Masri and Jongman, 2004; Khattab et al., 2006; Abudalbuh, 2010; Al-Deaibes, 2016; Jaber et al., 2019, inter alia). The other type of emphasis, secondary emphasis, has, nevertheless, been nearly completely understudied, which leaves a gap in the literature on emphasis. The present study, however, goes further to unravel the secondary emphasis by exploring free variational emphasis as produced by speakers of different genders (males and females) and different dialects (RJA and UJA).

2.2. Emphasis in Arabic

Acoustically speaking, emphatic segments have a strong bearing on the F2 of their neighboring vowels. Several studies have reported that the neighboring vowels of emphatic segments have a lowered F2 compared with the F2 of the same vowels when not occurring in an environment of emphatic phonemes (Card, 1983 for Palestinian Arabic; Wahba, 1993 for Alexandrian Egyptian Arabic; Al-Masri and Jongman, 2004; Jongman et al., 2007, 2011; Omari and Jaber, 2019 for JA).

Card (1983) investigated the acoustic correlates of emphasis in Palestinian Arabic. The results reported that there were lower values of F2 for emphatic vowels than those of plain vowels. Also, vowels that were close to emphatic consonants received lower F2 values than those of neighboring vowels. Similarly, Wahba (1993) examined the acoustic correlates of emphasis as used in Alexandrian Egyptian Arabic. The results revealed that there was no significant difference between emphatic and plain vowels in terms of F1 values. However, F2 lowering was for both the onset and the midpoint transition of the emphatic vowels compared to the F2 values of plain vowels. Providing an articulatory explanation related to vowels’ positions in the phonetic space, Wahba (1993) found that emphasis is best expressed in terms of F2 lowering on low central vowels.

Exploring the acoustic study of two Syrian Arabic varieties: Damascene and Aleppian, Almbark (2008) reported that there was no significant difference in the friction duration between emphatic and plain fricatives. She further found that there is no significant difference between emphatic and plain environments.

2.3. Emphasis in Jordanian Arabic

It has been reported in the literature that F2 of vowels adjacent to an emphatic segment (whether in the same or in a proximate syllable) is significantly lowered (Al-Masri and Jongman, 2004). Similarly, Jongman et al. (2007), who studied emphasis in JA, reported that vowels adjacent to emphatic segments tend to have lowered F2 and raised F1 and F3.

Khattab et al. (2006), who investigated the acoustic differences between /t/ and /tˁ/ in JA in the contexts of the high vowel /i/ and the low vowel /æ/, have also confirmed the same patterns: raising F1 and lowering F2 at the onset of the target vowels. They also reported that /tˁ/ has a shorter VOT than /t/.

In his socio-phonetic study, Abudalbuh (2010) found that vowels adjacent to emphatic consonants have significantly higher F1 and F3 and lower F2 than those in the context of plain segments. Moreover, he also reported that the emphatic voiceless stops tend to have shorter a shorter VOT than that of plain voiceless stops.

Investigating the acoustic correlates of emphasis in JA as spoken in the northern urban dialect, Jongman et al. (2011) revealed that emphatic vowels, regardless of their length, were found to have higher F1 and F3 and lower F2 compared to plain vowels. However, emphasis was best manifested on short vowels. In their study, the F2 value was found lower in the vowel /æ/ than that in the vowels /i/ and/u/. They concluded that emphatic stops have a lower spectral mean than the plain ones.

2.4. The spread of emphasis

Emphatic articulation, whether in secondary or primary emphasis, affects the neighboring segments through a process called ‘emphasis spread’. The spreading of the emphasis feature can affect the preceding and/or the following segments within a unit as small as a syllable, and sometimes can even spread further to a neighboring word with plain consonants under a prosodic condition; both words must form one prosodic unit, as large as a prosodic word. Therefore, the spreading of the emphasis feature can be regressive, progressive, or bidirectional, and the scope of spreading can be direct or maximal.

It has been found that emphasis spread can be rightward and/or leftward, but in some dialects, this spread is either blocked or decreased by consonants and/or vowels. The high vowel /i/ was found to block the rightward emphasis spread (Card, 1983; Davis, 1995; Watson, 2002). Davis (1995) found that in addition to the aforementioned high vowels, the rightward emphasis spread was also blocked by /j/ and /ʃ/ in the southern Palestinian dialect while Hassan and Esling, 2007 concluded that the presence of the sounds [i] and [ʃ] block or weaken the rightward emphasis spread.

Jongman et al. (2011) reported that emphasis spreads to the non-target stop consonants in UJA. They also noticed that the formant frequencies of vowels occurring close to emphatic consonants were highly affected, suggesting that the spread of emphasis is gradient. In other words, the spread depends on the neighboring vowel and its occurrence. Al-Masri and Jongman (2004) posited that emphasis spreads in both directions to the adjacent syllables but the degree of F2 lowering was lesser than that of vowels occurring in the target syllables. Also, having the vowels /i/ and /u/ in the target syllable was found to block the emphasis spread to the other vowels occurring in the right-adjacent syllables. As for the gradient emphasis spread, Almbark (2008) reported that rightward emphasis spread in Syrian Arabic was decreased and was not weakened by the vowels /u/ and /i/.

In a nutshell, a review of the previous studies unveils that there is much research on emphasis in Arabic dialects in general and JA sub-dialects in particular. Yet, such studies focus mainly on primary rather than secondary emphasis. Therefore, the present study is mainly concerned with emphatic variation of the labio-velar /w/ between males and females in two JA sub-dialects, namely, UJA and RJA. In addition, another motivation for the present study is that, unlike the secondarily emphatic consonants /r/ and /l/, /w/ is not conditioned, and, thus, it is neither a phoneme nor an allophone. It is an example of free variation that is produced by different speakers based on dialect and gender. Furthermore, one of the goals of this study is to scrutinize emphasis spread by examining the formant frequencies and duration of the vowels adjacent to the potentially emphatic consonant /w/. The formant tracks of the vowels adjacent to empathic consonants gives a clear manifestation of emphasis; therefore, studying emphasis spread on vowels will provide an indication of the extent of emphasis spread (and directionality) within the phonological word.

3. Brief overview of Jordanian Arabic sub-dialects

This diversity in Arabic dialects is also clear in the Hashemite Kingdom of Jordan, where different regional dialects with unique phonological and lexical preferences exist. Specifically, JA is divided into three main regional categories: rural JA, Bedouin JA and, urban JA (Al-Sughayer, 1990; Sakarna, 1999; Al-Deaibes, 2021a; Al-Deaibes, 2021b). As mentioned by Abd-El-Jawad (1987), the urban dialect (prestigious dialect) is primarily used by city residents who migrated to Jordan from neighboring urban centers such as Syrians, Palestinians, and Lebanese, and who live mainly in Amman, Zarqa, and Irbid. It is mainly spoken by educated and upper (middle) class people. The Bedouin dialect, on the other hand, is primarily used by speakers of Bedouin tribes who previously lived a nomadic life. There are five main tribes who speak the Bedouin dialect, namely, Bani Hasan, Al-Abbadi, Al-Huweitat, Bani Sakher, and Al-Ajarma (Sakarna, 1999). The rural dialect is spoken by villagers or village-born city residents in areas like Al-Rumtha, Bani Ebeid, Bani Kenana, and the city of Der'aa and its outskirts in Syria (Al-Deaibes and Rosen, 2019).

Each one of these three sub-dialects has its own distinct phonetic, morphological and syntactic structures. At the phonetic level, the main concern of the present paper, these three dialects differ considerably in the pronunciation of different consonants and vowels ‘rather than being restricted to only one consonant or one vowel’ (Jarbou and Al-Share, 2012, p. 6–7). It should be emphasized at this point that RJA and BJA have almost similar variants of vowels and consonants; yet, the same variants used in UJA are ‘distinguishably different’ from RJA and BJA (Jarbou and Al-Share, 2012). Such distinctions between RJA and BJA, on the one hand, and UJA, on the other hand, constitutes the main reason for choosing RJA and UJA in this study to compare and contrast the emphatic variation of the labio-velar /w/ and its correlation with the effect of gender on the variant realization or choice.

4. The sociophonetics of emphasis

Gender has been reported as one of the social variables that evoked the linguistic variation with its consistent influence; women have always been the ones who favor the prestigious variety than men (Al-Wer, 1997; Assiri, 2014; Omari and Van Herk, 2016; Al-Shawashreh, 2016). This influence was attested acoustically in Arabic (Kahn, 1975; Ahmad, 1979; Royal, 1985; Wahba, 1993; Khattab et al., 2006; Abudalbuh, 2010).

Women have been reported to have higher tendency to avoid stigmatized variants and favor prestigious ones than men do (Labov, 2001). It should be noticed that the prestigious variety can be correlated with gender (the social stereotype of women's /men's speech), region (urban, rural and Bedouin) and social class (lower class, middle class, and upper class). As far as emphasis variation is concerned, Royal (1985) reported that there are two variants for emphasis; weak variant (the prestigious variant) that are usually attached with femininity and urbanity, and strong variant (the non-prestigious variant) that are frequently attached with masculinity and rural regions. Laradi (1983: 312) concluded that “the less pharyngealization is produced, the higher the socio-economic status of these women.” In other words, this means that the higher the social class, the higher the tendency to use the weak emphatic variants (i.e., less pharyngealization).

Kahn (1975) examined the production of plain /t/ and /s/ and their emphatic counterparts /tˁ/ and /sˁ/ occurring in the environment of the vowels /i/ and /u/ in Cairene Arabic. In her study, F1 raising and F2 lowering in the vowels adjacent to the emphatic consonants were observed in the speech of both males and females. However, Kahn reported that the acoustic cues of emphasis were stronger in males' speech than in females'. She attributed her results to the compared portion of change from plain vowels to emphatic vowels in females'/males' speech. The portion of F2 lowering was greater in males' speech than that in females’ speech.

Similar findings were reached by Ahmad (1979) and Wahba (1993). Investigating the production of emphasis between /d/ and its emphatic counterpart /dˁ/ occurring in the environment of the vowels /i/, /u/ and /æ/ in Cairene Arabic, Ahmad (1979) found that the acoustic cues of emphasis in men's speech were greater than those in women's speech. She ascribed her results to the degree of F2 lowering in both men's and women's speech while Wahba (1993) ascribed his result to the regression slopes of plain and emphatic vowels for females and males. He found that the plain and emphatic slopes of females were closer to each other compared to men's (i.e., females did not show greater degree of emphasis than men did).

The overall sample from the lower-middle class in Royal (1985) study showed that there was no significant difference in terms of F2 transition between both male and female unless it was correlated with age and social class. In other words, the differences in women's and men's speech were manifested when age and social class came into play. Royal found that women produced lesser degree of emphasis than that of men in the upper class. Similar tendency was observed in low-middle class; Royal also found that only younger women from low-middle class tend to produce weak emphasis variant whereas the acoustic cues of emphasis of older females from low-middle class were greater than those of their male counterparts.

Depending on F2 lowering and F1–F2 approximation of vowels following emphatics, Almbark (2008) found that the aforementioned cues were the reliable acoustic parameters in depicting the differences of gender and region between plain and emphatic consonants of Syrian Arabic. Her findings were also influenced by the linguistic environment. That is to say, the acoustic cues (F2 lowering and F1–F2 approximation) following the emphatic stops were greater in women's speech compared to those in men's speech, whereas the cues that followed the emphatic fricatives were greater in the speech of males compared to females. It should be noted that Almbark examined VOT as a potential cue that reflected these differences, but her result did not show any significant differences. However, VOT was reported to be reliable in Heselwood (1996) and Bellem (2007). Heselwood (1996) found that while producing emphatics, the glottal state resulted in significant differences between Cairene and Baghdadi speakers. These differences were evident in the longer VOT values of /tˁ/ for Cairene speakers than those for Baghdadi speakers. Bellem (2007) reported that Bedouin dialects showed greater differences between plain and emphatic consonants, regarding VOT, compared to urban dialects.

In JA, Al-Masri and Jongman (2004), relying on the degree of F2 lowering, concluded that females realized emphatics more than men did. However, Abudalbuh (2010) ascribed his results to the significant difference in F1 raising and F2 lowering at onset and midpoint of the vowel for male speakers. He reported that males realized emphasis more than females. On the other hand, Khattab et al. (2006) argued that having firm conclusions about the correlation between gender and emphasis in JA was not easy to be drawn because of the contradiction that resulted from the inter-dialectical variation for the female data. However, they suggested that the findings of Royal (1985) in terms of the correlations between emphasis, gender and social class may be applied on JA as well. Unlike Al-Masri and Jongman (2004), Khattab et al. (2006) found that men realized emphasis in the production of /tˁ/ more than women did. Like Royal (1985) and Abudalbuh (2010), Omari and Jaber (2019) found that strong cues of emphasis were produced by low-middle class and male speakers respectively.

5. Methods

5.1. Participants

Participants of the study were 24 native male and female speakers of JA, equally distributed according to their gender (12 males and 12 females) and dialect (12 UJA and 12 RJA speakers). The total number of tokens was 360. Participants of the study had reported, before the experiment was conducted, that they all had no history of speech or hearing impairment at the time of recordings and they all spoke JA as their L1 and English as their L2. They were of nearly the same age group, 20–26 years old as shown in Table (1).

Table 1.

Participants profiles.

Participant code name	Sex	Age	Dialect spoken	Languages spoken
AL	Female	23	Rural	Arabic and English
AS	Female	24	Urban	Arabic and English
HG	Female	21	Urban	Arabic and English, Hebrew
LH	Female	23	Rural	Arabic and English
HS	Female	20	Rural	Arabic and English
LK	Female	25	Urban	Arabic and English
MA	Female	21	Urban	Arabic and English
MT	Female	24	Rural	Arabic and English
NH	Female	22	Urban	Arabic and English
NR	Female	23	Rural	Arabic and English
SB	Female	23	Urban	Arabic and English
SK	Female	25	Rural	Arabic and English
AF	Male	21	Rural	Arabic, English, Spanish
AO	Male	26	Rural	Arabic and English
BA	Male	22	Urban	Arabic and English
GT	Male	21	Urban	Arabic and English
HD	Male	24	Rural	Arabic and English
JL	Male	24	Urban	Arabic and English
MD	Male	22	Rural	Arabic and English, French
MH	Male	21	Urban	Arabic and English
MO	Male	22	Urban	Arabic and English
MN	Male	20	Rural	Arabic and English
RJ	Male	25	Urban	Arabic and English
SD	Male	26	Rural	Arabic and English

In order to comply with research ethics and protect the confidentiality of the participants, a consent form was handed to the participants prior to data collection to seek their permission to record the stimuli and to explain the purpose of the study.

5.2. Data

To ensure that this study complies with all research ethics regulations, we obtained the IRB approval (Protocol #: H20-008) from the Research Ethics Committee at Khalifa University to conduct the experiment. Upon receiving the approval, the participants were provided with a written list of 15 trisyllabic words (see appendix A) that included the sound /w/ word-medially. The purpose of choosing the segment /w/ word-medially is to check whether the emphasis spreading directionality is rightward, leftward, or symmetrical within the same word. The target words of the stimulus were recorded in the carrier phrase [ˈʔiħki___ˈmær:æh] (“Say____once”). The importance of using a carrier sentence is to make the production of the tokens resemble that of natural speech. The tokens were recorded and acoustically analyzed in Praat (version 6.1.04) (Boersma and Weenink, 2019). The stimuli were printed in Arabic script on a sheet supplemented with diacritic markings. It is worth pointing out in this regard that none of the tokens included an inherently emphatic phoneme. The reason behind this is to exclude the effect of any inherently emphatic consonants on the inherently non-emphatic /w/.

Participants were instructed to read the stimuli the way they would use them in daily life communications and at a normal pace. The participants were not informed about the specific purpose of the study to make the production of phrases more natural and unbiased. The recording was performed with a Marantz PMD-660 solid state recorder and an Audix OM 2 microphone in a quiet room at the University of Jordan. The recordings were done as a mono sound, and the sampling rate was set to 44100 Hz.

5.3. Measurements

The measurements included the duration of the short front low vowel /a/ preceding /w/ and the front low long vowel /a:/ following /w/. Further, F1, F2, as well as F3 of vowels preceding and following /w/ were measured.

Vowels were manually transcribed and segmented in Praat (Boersma and Weenink, 2019) by the authors. Since the data were collected from both male and female speakers and because there are physiological and anatomical differences between them when it comes to the formant frequencies, the data have been normalized by uploading the waveforms and transcriptions to FAVE (Rosenfelder et al., 2011) which measures formant values at one-third of vowel duration. The formant values were also normalized under FAVE's default Lobanov (1971) normalization method. The following Figure (1) is an example of how the segmentation was manually marked in Praat.

Figure 1.

Illustration of the manual segmentation of the word bˤaˤwˤa:ˤdi ‘deserts’.

5.4. Statistical analysis

The resulting output was exported to R (R Core Team, 2015) which was used to conduct the statistical analysis, including mixed effects linear modelling using the lme4 package (Bates et al., 2015) for R. For this purpose, speaker profiles were coded with demographic information including speaker dialect (Rural and Urban) and gender (male and female). More specifically, we used package lme4 (Bates et al., 2015) to perform a linear mixed effects analysis of the relationship between (F1, F2, F3, vowel duration for previous and following vowels) and gender and dialect. As fixed effects, we entered (F1, F2, F3, vowel duration for previous and following vowels) and gender and dialect (without interaction term) into the model. As random effects, we had intercepts for speaker and by-speaker random slopes for the effect of acoustic correlates. Residuals did not reveal any noticeable deviation from normality upon vsual inspection. The following is a sample code used in R to run a statistical analysis test for vowel F3 preceding the consonants /w/: ‘F3Postmodel = lmer(F3Post ∼ Gender∗Dialect+(1|Speaker), data = labiovelar)’

6. Results

In this section, we lay out the results of the acoustic measurements of the first three formant frequencies as well as the duration of the vowels preceding and following the emphaticized consonant /w/.

6.1. Vowel /a/ preceding /w/

6.1.1. F1, F2, F3, and vowel duration

As mentioned in section (4), the most reliable acoustic correlates that characterize emphatic consonants are the first three formant frequencies and duration of vowels adjacent to them. The mixed effects statistical analysis results, as shown in Table (2), indicate that mean frequency of F1 of the vowel preceding the emphaticized /w/ was higher in males by 21 Hz, which is not statistically significant and thus indicates that F1 was not affected by gender as illustrated in Figure (2). However, it is apparent from Table (2) that dialect is a significant factor indicating that urban speakers produced lower F1 than rural speakers by 71 Hz.

Table 2.

F1 of the vowel /a/ preceding /w/.

	Estimate	Std. Error	df	t value	Pr (>|t|)
(Intercept)	507.079	17.203	46.469	29.476	<2e-16 ∗∗∗1
GenderMale	20.888	25.193	28.454	0.829	0.41393
DialectUrban	-70.782	23.210	119.523	-3.050	0.00282 ∗∗
GenderMale:DialectUrban	1.156	34.983	36.677	0.033	0.97381

¹These are the significant codes that will be used as keys for the p-value for all statistical results in this study: 0 ‘∗∗∗’ 0.001 ‘∗∗’ 0.01 ‘∗’ 0.05 ‘.’ 0.1 ‘’ 1.

Figure 2.

Gender and dialect effect on F1 of vowel preceding /w/.

The boxplots in Figure (2) 1and the spectrograms in Figures (3) and (4) show how F1 is more lowered by female speakers on the one hand and urban speakers on the other.

Figure 3.

The word /tawa:li/ as produced by a Rural female speaker.

Figure 4.

The word /tawa:li/ as produced by an Urban female speaker.

Regarding F2 of the vowel preceding the emphaticized /w/, the mixed effects statistical results, as given in Table (3), show that gender is a significant factor predicting that male speakers in general produced lower F2 by 195 Hz than female speakers (as shown in Figure (4). The tests also reveal that dialect has a significant effect on F2 predicting that urban speakers produced significantly higher F2 by 109 Hz than rural speakers as shown in Table (3) and Figure (5). The significant interaction between gender and dialect indicates that the gender effect is significantly different between urban and rural datasets. Combining the visual results of F1, we can conclude that gender is a more significant predictor in rural speakers.

Table 3.

F2 of the vowel /a/ preceding /w/.

	Estimate	Std. Error	df	t value	Pr (>|t|)
(Intercept)	1209.21	23.79	42.10	50.827	<2e-16 ∗∗∗
GenderMale	-195.51	40.03	26.27	-4.885	4.44e-05 ∗∗∗
DialectUrban	196.47	24.21	306.93	8.114	1.19e-14 ∗∗∗
GenderMale:DialectUrban	-107.01	50.55	39.54	-2.117	0.0406 ∗

Figure 5.

Gender and dialect effect on F2 of vowel preceding /w/.

The boxplots in Figure (5) show how F2 is more lowered by male speakers on the one hand and rural speakers on the other.

The mixed effects test for F3 in the vowel /a/ preceding the emphaticized /w/ reveals that gender has the strongest effect on F3 values, and that males in general have higher F3 values than females by 193 Hz as given in Table (4) and Figure (6). Dialect is also significant in that urban speakers produced the vowel /a/ with higher F3 values by 159 Hz than rural speakers. The significant interaction between gender and dialect indicates that the gender effect is significantly different between urban and rural datasets. Combining the visual results of F3, we can conclude that gender is a more significant predictor in rural speakers.

Table 4.

F3 of the vowel /a/ preceding /w/.

	Estimate	Std. Error	df	t value	Pr (>|t|)
(Intercept)	2330.30	34.52	30.83	67.503	<2e-16 ∗∗∗
GenderMale	193.01	48.84	22.81	3.951	0.000643 ∗∗∗
DialectUrban	158.68	43.73	186.50	3.256	0.001340 ∗∗
GenderMale:DialectUrban	-155.48	69.01	43.57	-2.253	0.029349 ∗

Figure 6.

Gender and dialect effect on F3 of vowel preceding /w/.

The boxplots in Figure (6) show how F3 is more lowered by female speakers on the one hand and rural speakers on the other.

As far as vowel duration is concerned, the results of the mixed effects statistical test, as shown in Table (5), uncovered that gender was a non-significant factor in both Urban and Rural datasets. However, dialect is a significant factor indicating that urban speakers have shorter vowel duration by 18 ms than rural speakers.

Table 5.

Duration of the vowel /a/ preceding /w/.

	Estimate	Std. Error	df	t value	Pr (>|t|)
(Intercept)	0.184743	0.003860	32.705060	47.858	<2e-16 ∗∗∗
GenderMale	9.664	5.305	2.413	0.182	0.857
DialectUrban	-0.018740	0.002308	275.545	-8.121	1.56e-14 ∗∗∗
GenderMale:DialectUrban	-0.471	6.476	3.688	-0.702	0.487

The following boxplots in Figure (7) visualize shorter mean vowel duration before the emphaticized consonant /w/ as produced by speakers of both genders and both dialects.

Figure 7.

Gender and dialect effect on duration of vowel preceding /w/.

6.2. Vowel /a:/ following /w/

As with the vowel /a/ preceding /w/, the long vowel /a:/ was also examined in terms of the first formant frequencies and duration.

6.2.1. F1, F2, F3, and vowel duration

The mixed effects statistical analysis results, as shown in Table (6), indicate that mean frequency of F1 of the vowel following the emphaticized /w/ was higher in males by 19 Hz, which is not statistically significant and thus indicates that F1 was not affected by gender as illustrated in Figure (8). However, dialect is a significant factor indicating that urban speakers have lower F1 by 102 than rural speakers. The test also indicates that there is a significant interaction between gender and dialect. The effect of gender on F1 is more significant in urban dataset as compared to rural dataset.

Table 6.

F1 of the vowel /a:/ following /w/.

	Estimate	Std. Error	df	t value	Pr (>|t|)
(Intercept)	516.71	14.32	42.19	36.088	<2e-16 ∗∗∗
GenderMale	19.20	23.15	25.00	0.829	0.4149
DialectUrban	-101.88	16.13	252.64	-6.317	1.19e-09 ∗∗∗
GenderMale:DialectUrban	74.37	30.20	36.51	2.463	0.0186 ∗

Figure 8.

Gender and dialect effect on F1 of vowel following /w/.

The boxplots in Figure (8) show how F1 of the vowel /a:/ following the emphaticized /w/ is more lowered by female speakers on the one hand and Urban speakers on the other.

As for the F2 of the vowel /a:/ following the emphaticized /w/, the mixed effects test show that dialect has the strongest effect in this model. Urban speakers produce significantly higher F2 by 266 Hz than rural speakers as shown in Table (7). The results also show that gender is significant; males in general have lower F2 by 194 Hz than females. The results also reveal that there is a significant interaction between gender and dialect, indicating that the effect of gender on F2 is more significant in urban dataset as compared to rural dataset.

Table 7.

F2 of the vowel /a:/ following /w/.

	Estimate	Std. Error	df	t value	Pr (>|t|)
(Intercept)	1136.09	21.00	45.79	54.104	<2e-16 ∗∗∗
GenderMale	-193.81	34.07	27.35	-5.689	4.6e-06 ∗∗∗
DialectUrban	265.64	23.47	264.23	11.317	<2e-16 ∗∗∗
GenderMale:DialectUrban	-147.82	44.32	39.94	-3.335	0.00185 ∗∗

The boxplots in Figure (9) show how F2 of the vowel /a:/ following the emphaticized /w/ is more raised by female speakers on the one hand and Urban speakers on the other.

Figure 9.

Gender and dialect effect on F2 of vowel following /w/.

The mixed effects test showed that F3 of the vowel following the emphaticized /w/ is significantly higher in urban speakers by 103 Hz than rural speakers as shown in Table (8). More importantly, the strongest effect in this model is gender: males have higher F3 by 244 Hz than females. However, there is also an interaction between gender and dialect indicating that gender has different degree of effects on F3 depending on the dialect. The plot result and statistical tests on sub-datasets suggest that female speakers and male speakers are significantly different in the dataset of rural speakers as opposed to urban speakers.

Table 8.

F3 of the vowel /a:/ following /w/.

	Estimate	Std. Error	df	t value	Pr (>|t|)
(Intercept)	2366.53	34.33	354.20	68.942	<2e-16 ∗∗∗
GenderMale	243.63	48.55	336.90	5.019	8.24e-07 ∗∗∗
DialectUrban	102.11	49.40	339.75	2.067	0.0395 ∗
GenderMale:DialectUrban	-164.87	68.65	356.00	-2.402	0.0168 ∗

The boxplots in Figure (10) show how F3 of the vowel /a:/ following the emphaticized /w/ is more raised by male speakers on the one hand and rural speakers on the other.

Figure 10.

Gender and dialect effect on F3 of vowel following /w/.

As far as vowel duration is concerned, the mixed effects statistical test showed that, as shown in Table (9), gender was a non-significant factor in neither the urban nor rural datasets. However, dialect is a significant factor indicating that urban speakers have shorter vowel duration by 20 ms than rural speakers.

Table 9.

Duration of the vowel /a:/ following /w/.

	Estimate	Std. Error	df	t value	Pr (>|t|)
(Intercept)	0.204305	0.0018	34.453	158.7	<2e-16 ∗∗∗
GenderMale	8.306	2.699	2.733	0.308	0.761
DialectUrban	-0.020549	0.0018	98.8797	-12.7	<2e-16 ∗∗∗
GenderMale:DialectUrban	-3.158	3.607	3.832	-0.875	0.387

The following boxplots in Figure (11) visualize shorter mean vowel duration after the emphaticized consonant /w/ as produced by speakers of both genders and both dialects.

Figure 11.

Gender and dialect effect on duration of vowel following /w/.

7. Discussion

After having presented the results of F1, F2, F3 and vowel duration of the vowels preceding and following the labio-velar /w/, we turn now to answer the research questions posited at the outset of this paper and integrate our results within the wider existing literature.

The results of this study show that the acoustic correlates of secondarily emphatic sounds are similar to the ones reported in the previous studies on primarily emphatic sounds (Al-Masri and Jongman, 2004; Khattab et al., 2006; Jongman et al., 2007, 2011; Omari and Jaber, 2019), raised F1 and F3 and lowered F2 in the adjacent vowels of the emphatic sound.

Regarding the first research question on whether gender affects emphatic variation of /w/, the results show that gender plays a significant role in predicting the acoustic correlates of /w/ and the adjacent vowels preceding and following it. More specifically, female speakers produced weaker emphatic segments when compared with males. This result is in line with the one reported by Kahn (1975), Ahmad (1979), Abudalbuh (2010) and Omari and Jaber (2019) with males producing stronger emphatic sounds and females producing weaker emphatic sounds. However, this pattern is not confirmed in Lehn (1963), who finds that cues of emphasis in women's speech were greater than those in the speech of their male counterparts in Cairene Arabic. Similarly, relying on the degree of F2 lowering, Al-Masri and Jongman (2004) find that females realized emphatics more than men did. It should be mentioned here that these contradicting results might be a result of the lack of data normalization for gender effect (Kahn, 1975; Royal, 1985; Al-Masri and Jongman, 2004; Almbark, 2008). In other words, these studies did not pay attention to the frequency counts when they compared the results of the gender factor in different data or corpora. Notwithstanding these different results, most of the studies on emphatic sounds in JA (Khattab et al., 2006; Abudalbuh, 2010; Omari and Jaber, 2019) confirm that the acoustic cues of emphasis are evident in males' speech more than in the females' speech.

As for the second research question concerning the effect of dialect on the production of the emphatic labio-velar /w/, the results show that dialect has a significant role in predicting the acoustic correlates of the emphatic sound and its adjacent vowels. The results demonstrate that UJA speakers (both males and females) produced weaker emphatic consonants than RJA speakers. This result converges with the ones reached by Heselwood (1996) and Bellem (2007), who find significant differences between Cairene and Baghdadi speakers and Bedouin and urban dialects respectively in their production of emphatic sounds.

Taking the first two research questions together, it can be noticed that women generally tend to distinguish themselves linguistically from men through producing forms that are different from those produced by men. In case there are no alternative forms, women usually produce the same form in a way that distinguishes them from their male counterparts. In terms of secondary emphatic sounds (labio-velar /w/ in our study), women distinguish themselves from men by producing a weaker emphatic /w/ than that produced by men. Royal (1985) reports that strong pharyngealization is a characteristic of men's speech, and that the use/appearance of stronger pharyngealization sounds in the speech of women is considered unfeminine (Kriba, 2009). This means that strong emphatic sounds in the speech of women are stigmatized in Arabic dialects. The case of JA is no exception where it is stigmatized for a woman to speak using strong emphatic sounds. That is why women in Jordan avoid producing strong emphatic sounds in order for their speech not to be described as unfeminine. Thus, they opt for weaker emphatic sounds to show 'softness' and 'femininity'. This pattern is in line with the one reported in previous studies on Arabic (Al-Khatib, 1988; Khattab et al., 2006; Habib, 2010; Al-Wer and Herin, 2011) as well as other languages (Labov, 2001). Labov (2001, p.266) concludes that “women show a lower rate of stigmatized variants and a higher rate of prestige variants than men.” Added to this is the association of weak emphatic sounds and urbanity which is associated with mobility and openness (Al-Wer and Herin, 2011) that make urban variety the supra-local one (Abd El-Jawad, 1987). Unlike men, women are reported to favor prestigious urban variants that happen to be weak emphatic sounds in JA. This conclusion can be articulated as follows:

With the perceptual realization of emphatic sounds as heavy and dark (Walter, 2006), and the tendency for females to talk more delicately and softly than males (Crawford, 1995), the woman tendency [….] to produce fewer cues to emphasis may be accounted for in terms of their tendency to keep their speech urban and, at the same time, avoid male-dominated linguistic norms should alternative norms exist. In doing so, women use fine acoustic details to project their distinct social identity. This tendency may provide an example of extending feminism as a political, economical and cultural movement into a linguistic movement where women opt to distinguish themselves linguistically from men. This account attempts, by no means, to rule out the role of prestige in driving women's linguistic choices. It just attempts to provide new insights in the gender-related variation in the production of emphasis (Abudalbuh 2010, p.70).

Concerning the third research question about the effects of gender and dialect on vowel duration, the results show that while gender does not have a significant effect on the duration of the adjacent vowels, dialect significantly affects vowel duration in the emphatic context. The results show that RJA speakers (males and females) produce longer empathic vowel duration across the board, which shows that vowel duration may be a robust acoustic correlate when investigating secondarily emphatic segments as opposed to primarily emphatic segments. This result collides with the ones reported in the previous literature (Heselwood, 1996; Khattab et al., 2006; Bellem, 2007; Abudalbuh, 2010; Omari and Jaber, 2019). Heselwood (1996) finds significant differences in vowel duration between Cairene and Baghdadi speakers. These differences are evident in the longer VOT values of /tˁ/ for Cairene speakers than those for Baghdadi speakers. Similarly, Bellem (2007) points out that Bedouin dialects show greater differences between plain and emphatic consonants, regarding VOT, compared to urban dialects. Khattab, Al-Tamimi, and Heselwood (2006) also conclude that the VOT of /tˁ/; the voiceless emphatic stop, is shorter than that of its plain counterpart /t/. For his part, Abudalbuh (2010) finds that the VOT of emphatic voiceless stops is shorter than the VOT of plain voiceless stops in JA.

8. Conclusion

In this paper, we investigated the emphatic variation of the labio-velar /w/ in two JA sub-dialects: RJA and UJA More specifically, we studied the impact of gender and dialect on the production of secondary emphasis. The results revealed that emphasis, whether primary or secondary, has the same acoustic correlates: lowered F2 and raised F1 and F3. It has also been shown that there is a variation in the degree of emphasis based on gender and dialect, where males produced stronger emphatic segments than females did, and Urban speakers produced weaker empathic segments than Rural speakers did. Vowel duration was found to be a salient acoustic correlate when investigating emphatic segments as opposed to primarily emphatic ones. With regards to emphasis spread directionality, the results showed that it is bidirectional, ranging from proximal to long-distance segments.

Declarations

Author contribution statement

Mutasim Al-Deaibes, Ekab Al-Shawashreh, Marwan Jarrah: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Funding statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability statement

The data that has been used is confidential.

Declaration of interests statement

The authors declare no conflict of interest.

Additional information

No additional information is available for this paper.

Appendix A. Stimuli.

#	Token	Meaning
1	tawa:li	remnants
2	dawa:li	grape leaves
3	θaw:ani	seconds
4	hawa:d͡ʒis	premonitions
5	nawa:fil	secondaries ‘not primary’
6	sawa:lif	chitchats
7	kawa:li:s	scenes
8	xawa:li	olden
9	zawa:jid	extras
10	ʃawa:ħin	chargers
11	ʔawa:ʕi	clothes
12	lawa:zim	necessities
13	ʕawa:nis	spinsters
14	bawa:di	deserts
15	ħawa:d͡ʒiz	barricades

Appendix B. –F1, F2, and F3 across genders and dialects.

Speaker	Gender	Dialect	F1Pre	F2Pre	F3Pre	F1Post	F2Post	F3Post
AL	Female	Rural	424	1198	2618	484	1108	2618
AL	Female	Rural	426	1192	2287	486	1112	2287
AL	Female	Rural	449	1233	2550	459	1133	2550
AL	Female	Rural	457	1217	2597	467	1117	2697
AL	Female	Rural	459	1222	2486	479	1122	2586
AL	Female	Rural	460	1393	2185	446	1148	2163
AL	Female	Rural	479	1353	2259	489	1153	2259
AL	Female	Rural	481	1321	2192	471	1321	2192
AL	Female	Rural	497	1315	2281	487	1305	2301
AL	Female	Rural	498	1205	2685	498	1105	2785
AL	Female	Rural	508	1373	2530	518	1073	2530
AL	Female	Rural	513	1370	2402	510	1270	2411
AL	Female	Rural	557	1123	2446	557	1023	2546
AL	Female	Rural	565	1373	2009	575	1063	2011
AL	Female	Rural	565	1168	2029	585	1128	2039
AS	Female	Urban	251	1558	2352	261	1570	2321
AS	Female	Urban	253	1427	2185	263	1474	2137
AS	Female	Urban	255	1332	2376	255	1332	2376
AS	Female	Urban	258	1547	2724	278	1593	2744
AS	Female	Urban	320	1602	2885	330	1649	2745
AS	Female	Urban	328	1545	2417	338	1578	2429
AS	Female	Urban	342	1291	2523	342	1291	2523
AS	Female	Urban	349	1370	2391	359	1270	2388
AS	Female	Urban	450	1372	2662	360	1362	2678
AS	Female	Urban	465	1206	2644	465	1206	2644
AS	Female	Urban	465	1190	2188	355	1190	2188
AS	Female	Urban	483	1323	2252	493	1323	2252
AS	Female	Urban	508	1203	2381	408	1203	2381
AS	Female	Urban	530	1662	2028	530	1654	2028
AS	Female	Urban	556	1579	2179	566	1545	2179
HG	Female	Urban	255	1272	2530	315	1262	2530
HG	Female	Urban	256	1687	2748	226	1668	2648
HG	Female	Urban	257	1667	2000	257	1699	2000
HG	Female	Urban	259	1261	2568	269	1261	2568
HG	Female	Urban	318	1685	2881	328	1688	2871
HG	Female	Urban	337	1306	2377	337	1206	2377
HG	Female	Urban	346	1054	2777	346	1054	2777
HG	Female	Urban	349	1388	2004	359	1365	3011
HG	Female	Urban	457	1325	2075	347	1325	2075
HG	Female	Urban	462	1086	2286	472	1086	2286
HG	Female	Urban	463	1449	2378	473	1474	2364
HG	Female	Urban	475	1688	2154	465	1588	2167
HG	Female	Urban	482	1673	2477	382	1678	2477
HG	Female	Urban	490	1382	2455	490	1382	2455
HG	Female	Urban	531	1307	2108	541	1207	2108
HS	Female	Rural	358	1304	2446	368	1312	2477
HS	Female	Rural	376	975	2135	396	965	2235
HS	Female	Rural	421	947	2111	451	937	2211
HS	Female	Rural	432	1175	2793	492	1135	2893
HS	Female	Rural	442	1136	2728	482	1036	2828
HS	Female	Rural	449	959	2511	459	949	2611
HS	Female	Rural	451	1381	2553	451	1281	2553
HS	Female	Rural	465	916	2195	475	916	2295
HS	Female	Rural	470	1232	1990	490	1132	1990
HS	Female	Rural	474	1219	2647	484	1119	2647
HS	Female	Rural	492	1361	2189	482	1261	2189
HS	Female	Rural	493	1152	2331	493	1122	2431
HS	Female	Rural	551	1138	2222	571	1038	2322
HS	Female	Rural	557	1293	2485	567	1293	2485
HS	Female	Rural	575	1189	1857	585	1159	1857
LH	Female	Rural	479	1309	2093	489	1309	2093
LH	Female	Rural	487	1095	2221	487	1095	2221
LH	Female	Rural	496	1311	2148	496	1311	2148
LH	Female	Rural	499	1391	1765	489	1077	1695
LH	Female	Rural	560	1187	1918	560	1177	1918
LH	Female	Rural	614	1049	1967	614	1019	1967
LH	Female	Rural	618	1240	1804	628	1140	1804
LH	Female	Rural	621	1103	1973	621	1003	1973
LH	Female	Rural	644	1219	1849	645	1119	1949
LH	Female	Rural	680	1044	1974	690	1044	1974
LH	Female	Rural	716	1176	1940	716	1166	1940
LH	Female	Rural	502	1245	2738	532	1245	2838
LH	Female	Rural	570	1009	2920	580	1009	2920
LH	Female	Rural	629	1256	2618	639	1116	2818
LH	Female	Rural	606	1640	2617	286	1625	2917
LK	Female	Urban	423	1338	2430	533	1338	2430
LK	Female	Urban	456	1345	2268	366	1345	2268
LK	Female	Urban	361	1345	2344	471	1345	2364
LK	Female	Urban	411	906	2204	441	906	2304
LK	Female	Urban	254	1383	2477	254	1383	2477
LK	Female	Urban	259	1320	2558	279	1320	2758
LK	Female	Urban	261	1432	2018	281	1483	3018
LK	Female	Urban	343	1599	2561	333	1576	2573
LK	Female	Urban	348	1295	2765	348	1285	2965
LK	Female	Urban	394	1260	2941	384	1160	2981
LK	Female	Urban	404	1298	2796	404	1298	2996
LK	Female	Urban	443	1238	2870	453	1138	2870
LK	Female	Urban	453	1209	3007	463	1109	3007
LK	Female	Urban	458	1351	2895	358	1311	2885
LK	Female	Urban	489	1327	2080	479	1327	3080
MA	Female	Urban	253	1537	2173	233	1598	2189
MA	Female	Urban	254	1324	2676	254	1324	2676
MA	Female	Urban	255	1393	2409	210	1374	2478
MA	Female	Urban	257	1321	2411	257	1321	2411
MA	Female	Urban	263	1658	2673	289	1610	2783
MA	Female	Urban	304	1229	2510	304	1229	2510
MA	Female	Urban	316	1661	3116	416	1658	3116
MA	Female	Urban	347	1608	2017	357	1666	2231
MA	Female	Urban	441	1479	2435	431	1485	2878
MA	Female	Urban	458	1326	1867	458	1326	1867
MA	Female	Urban	482	1510	2810	462	1578	2833
MA	Female	Urban	482	1091	2051	382	1091	3051
MA	Female	Urban	485	1319	2364	485	1319	2364
MA	Female	Urban	502	1304	2345	412	1204	2345
MA	Female	Urban	513	1117	2048	423	1117	2048
MT	Female	Rural	449	1269	1880	459	1169	1890
MT	Female	Rural	482	905	2325	492	905	2425
MT	Female	Rural	517	1387	2334	527	1297	2365
MT	Female	Rural	530	1130	2172	570	1030	2272
MT	Female	Rural	622	1317	1911	612	1217	1921
MT	Female	Rural	641	1356	1969	631	1326	1989
MT	Female	Rural	650	1395	2198	640	1074	2189
MT	Female	Rural	657	1092	2116	657	1062	2116
MT	Female	Rural	743	1113	1996	743	1013	1996
MT	Female	Rural	363	1187	3074	393	1137	3074
MT	Female	Rural	438	1389	2592	427	1076	2602
MT	Female	Rural	460	1185	2702	470	1165	2702
MT	Female	Rural	527	1277	2938	537	1237	2938
MT	Female	Rural	537	1255	2682	547	1155	2882
MT	Female	Rural	586	1196	2897	596	1126	2897
NH	Female	Urban	338	1319	2129	338	1319	2129
NH	Female	Urban	355	1594	2199	479	1578	2121
NH	Female	Urban	429	1215	2163	429	1115	2263
NH	Female	Urban	430	1284	1963	420	1264	1963
NH	Female	Urban	470	1531	2397	480	1547	2365
NH	Female	Urban	475	1677	2309	488	1666	2377
NH	Female	Urban	481	1249	2150	491	1149	2150
NH	Female	Urban	486	1277	2368	486	1287	2368
NH	Female	Urban	512	1201	2263	522	1201	2263
NH	Female	Urban	527	1572	2210	527	1533	2330
NH	Female	Urban	536	1396	1908	526	1324	1949
NH	Female	Urban	546	1645	2246	576	1699	2346
NH	Female	Urban	573	1394	1903	563	1178	1999
NH	Female	Urban	606	1366	1833	606	1066	1877
NH	Female	Urban	698	1231	1979	698	1131	1979
NR	Female	Rural	322	1517	2490	332	1554	2455
NR	Female	Rural	400	1427	2693	401	1498	2674
NR	Female	Rural	422	1197	2693	462	1107	2693
NR	Female	Rural	425	1173	2903	485	1753	2903
NR	Female	Rural	460	1295	2631	470	1295	2631
NR	Female	Rural	468	1311	2973	578	1311	2973
NR	Female	Rural	474	1389	2138	364	1384	3018
NR	Female	Rural	474	1546	2496	487	1538	2421
NR	Female	Rural	490	1595	2798	380	1577	2808
NR	Female	Rural	493	1363	2773	493	1063	2773
NR	Female	Rural	530	1683	2066	570	1629	3004
NR	Female	Rural	541	1206	2854	561	1206	2854
NR	Female	Rural	563	1287	3014	563	1287	3014
NR	Female	Rural	585	1077	2651	585	1037	2751
NR	Female	Rural	608	1247	2706	618	1147	2706
SB	Female	Urban	303	1274	2570	263	1264	2670
SB	Female	Urban	324	1104	2627	324	1104	2627
SB	Female	Urban	355	1260	2535	365	1160	2635
SB	Female	Urban	364	1262	2476	360	1278	2486
SB	Female	Urban	453	1220	2596	453	1120	2596
SB	Female	Urban	458	1194	2623	488	1134	2623
SB	Female	Urban	458	1689	2234	465	1577	2454
SB	Female	Urban	460	1200	2277	490	1100	2477
SB	Female	Urban	461	1324	2637	461	1324	2637
SB	Female	Urban	468	1352	2377	468	1322	2357
SB	Female	Urban	492	1340	2610	492	1340	2670
SB	Female	Urban	519	1296	2482	419	1396	2482
SB	Female	Urban	533	1674	2737	547	1675	2697
SB	Female	Urban	553	1327	2536	553	1327	2536
SB	Female	Urban	573	1308	2657	573	1308	2657
SK	Female	Rural	259	1381	2279	269	1381	2279
SK	Female	Rural	352	1305	2212	352	1305	2212
SK	Female	Rural	374	1244	1852	394	1144	1852
SK	Female	Rural	406	1288	2195	406	1288	2195
SK	Female	Rural	459	1275	1943	449	1175	1993
SK	Female	Rural	463	1101	1809	453	1101	1909
SK	Female	Rural	472	1553	2213	462	1549	2343
SK	Female	Rural	474	1316	1889	464	1226	1897
SK	Female	Rural	503	1217	2010	523	1117	2110
SK	Female	Rural	516	1354	1905	516	1154	1905
SK	Female	Rural	519	1349	2320	526	1319	2330
SK	Female	Rural	559	1264	2075	569	1194	2085
SK	Female	Rural	566	1628	2230	556	1685	2234
SK	Female	Rural	586	1191	2153	586	1121	2153
SK	Female	Rural	596	993	2181	596	973	2281
AF	Male	Rural	396	1140	2617	386	1010	2927
AF	Male	Rural	403	774	2770	443	754	2970
AF	Male	Rural	420	1002	2885	430	1012	2985
AF	Male	Rural	446	1054	2777	476	1024	2877
AF	Male	Rural	458	951	2895	488	841	2995
AF	Male	Rural	465	1006	2444	485	1005	2944
AF	Male	Rural	469	1020	2758	479	1017	2858
AF	Male	Rural	482	810	2810	492	703	2910
AF	Male	Rural	516	1161	2116	526	1011	3116
AF	Male	Rural	518	985	2781	528	885	2881
AF	Male	Rural	582	1091	2951	592	1091	3091
AF	Male	Rural	590	1195	2798	580	1055	2698
AF	Male	Rural	600	927	2693	610	817	2993
AF	Male	Rural	626	787	2748	626	767	2948
AF	Male	Rural	630	1083	2066	620	1073	3166
AO	Male	Rural	447	958	2317	467	848	2417
AO	Male	Rural	452	1205	2212	462	1105	2322
AO	Male	Rural	455	1094	2499	475	1094	2699
AO	Male	Rural	456	945	2268	466	835	2368
AO	Male	Rural	458	726	1867	498	726	1967
AO	Male	Rural	458	799	2234	488	779	2434
AO	Male	Rural	473	827	2185	493	717	2385
AO	Male	Rural	482	1073	2477	472	1063	2487
AO	Male	Rural	483	1223	2252	473	1023	2452
AO	Male	Rural	522	1117	2490	542	1010	2500
AO	Male	Rural	527	1172	2210	537	1022	2410
AO	Male	Rural	530	962	2028	560	852	2128
AO	Male	Rural	586	1028	2230	666	1018	2330
AO	Male	Rural	606	988	2195	626	878	2195
AO	Male	Rural	665	1190	2488	635	1010	2699
BA	Male	Urban	404	1131	3303	514	1121	3027
BA	Male	Urban	412	1140	2350	632	1130	2550
BA	Male	Urban	422	1117	2350	632	1017	2350
BA	Male	Urban	442	1103	3303	642	1103	3027
BA	Male	Urban	452	1125	2399	642	1025	2499
BA	Male	Urban	458	1088	2523	458	1088	2523
BA	Male	Urban	458	1140	2536	558	1130	2636
BA	Male	Urban	461	1091	2644	471	1081	2644
BA	Male	Urban	474	1095	2676	584	1095	2676
BA	Male	Urban	497	1111	2777	497	1111	2777
BA	Male	Urban	504	1086	2482	504	1086	2482
BA	Male	Urban	522	1117	2350	642	1017	2350
BA	Male	Urban	522	1138	2350	522	1128	2450
BA	Male	Urban	616	1104	3303	736	1104	3027
BA	Male	Urban	616	1127	2350	726	1117	2450
GT	Male	Urban	383	1091	2646	383	1091	2646
GT	Male	Urban	422	1083	2434	422	1073	2434
GT	Male	Urban	425	958	2844	445	958	2844
GT	Male	Urban	430	988	2739	420	978	2739
GT	Male	Urban	434	972	2394	424	952	2394
GT	Male	Urban	440	955	2360	550	945	2360
GT	Male	Urban	451	962	2238	571	952	2438
GT	Male	Urban	504	906	2099	514	906	2299
GT	Male	Urban	505	1009	1766	625	1009	1626
GT	Male	Urban	511	1019	3133	521	1019	3133
GT	Male	Urban	578	1006	2205	668	1006	2205
GT	Male	Urban	578	1395	1792	688	1264	1792
GT	Male	Urban	580	1002	2125	670	1002	2225
GT	Male	Urban	616	1008	3061	726	1008	3061
GT	Male	Urban	642	919	2399	752	909	2499
HD	Male	Rural	405	1193	2409	409	1103	2609
HD	Male	Rural	538	919	2429	548	811	2929
HD	Male	Rural	557	1125	2075	577	1105	2185
HD	Male	Rural	574	1089	2138	584	1009	3238
HD	Male	Rural	575	888	2754	585	878	2954
HD	Male	Rural	587	895	2221	597	785	2421
HD	Male	Rural	589	1007	2980	599	1006	3080
HD	Male	Rural	604	1198	2896	614	1058	2996
HD	Male	Rural	619	1196	2482	629	1056	2582
HD	Male	Rural	619	1149	2320	619	1039	2400
HD	Male	Rural	631	1107	2108	641	1007	2308
HD	Male	Rural	648	1095	2865	648	1095	2995
HD	Male	Rural	668	1111	2973	678	1010	2983
HD	Male	Rural	708	1103	2381	708	1003	2491
HD	Male	Rural	753	1027	2536	273	1017	2636
JL	Male	Urban	264	974	2244	284	954	2444
JL	Male	Urban	301	824	2357	423	714	2457
JL	Male	Urban	368	921	2932	398	921	2932
JL	Male	Urban	422	896	2790	422	866	2890
JL	Male	Urban	426	888	1879	426	878	1879
JL	Male	Urban	431	877	2476	421	877	2476
JL	Male	Urban	443	827	2702	463	854	2802
JL	Male	Urban	448	887	1640	458	877	1640
JL	Male	Urban	449	879	2394	449	869	2494
JL	Male	Urban	461	899	2737	571	879	2837
JL	Male	Urban	462	810	2398	472	703	2498
JL	Male	Urban	469	804	2530	489	803	2630
JL	Male	Urban	469	829	2496	579	829	2496
JL	Male	Urban	490	927	2350	490	917	2350
JL	Male	Urban	490	926	2283	480	916	2283
MD	Male	Rural	404	829	2510	414	719	2710
MD	Male	Rural	434	983	2477	444	873	2477
MD	Male	Rural	434	824	2476	454	714	2976
MD	Male	Rural	449	1088	2004	499	1080	3104
MD	Male	Rural	463	1049	2378	473	1029	2578
MD	Male	Rural	491	1032	2018	481	1032	3118
MD	Male	Rural	537	906	2377	537	803	2577
MD	Male	Rural	543	899	2161	533	789	2661
MD	Male	Rural	545	1032	2376	545	1022	2576
MD	Male	Rural	546	955	2246	566	845	2346
MD	Male	Rural	547	921	2411	567	820	2611
MD	Male	Rural	549	1170	2391	559	1030	2591
MD	Male	Rural	556	1079	2179	576	1059	2289
MD	Male	Rural	575	1077	2309	585	1067	2429
MD	Male	Rural	662	1086	2286	672	1006	2486
MH	Male	Urban	257	1295	2350	377	1195	2450
MH	Male	Urban	258	1281	2965	388	1181	2965
MH	Male	Urban	308	1299	1879	428	1199	1979
MH	Male	Urban	313	1279	2885	463	1179	2985
MH	Male	Urban	355	1305	1792	365	1205	1892
MH	Male	Urban	417	1174	2881	437	1074	2881
MH	Male	Urban	426	1170	2798	446	1160	2898
MH	Male	Urban	449	1298	2790	469	1198	2890
MH	Male	Urban	457	1290	2973	587	1190	2973
MH	Male	Urban	459	1296	2739	589	1196	2839
MH	Male	Urban	461	1161	2496	481	1161	2496
MH	Male	Urban	474	1149	2758	484	1119	2758
MH	Male	Urban	474	1167	2510	484	1117	2510
MH	Male	Urban	493	1272	2817	493	1172	2817
MH	Male	Urban	504	1152	2350	514	1112	2350
MN	Male	Rural	322	1028	2357	422	1018	2457
MN	Male	Rural	422	1020	2357	432	1010	2457
MN	Male	Rural	449	985	2357	439	985	2457
MN	Male	Rural	462	882	2357	492	872	2357
MN	Male	Rural	504	1095	2350	514	1095	2350
MN	Male	Rural	642	945	2357	782	945	2457
MN	Male	Rural	351	1311	2646	351	1111	2746
MN	Male	Rural	353	1327	2758	373	1227	2858
MN	Male	Rural	364	1293	2996	384	1193	2996
MN	Male	Rural	424	1389	2798	434	1287	2858
MN	Male	Rural	438	1323	2737	448	1223	2837
MN	Male	Rural	449	1307	3303	439	1207	3027
MN	Male	Rural	457	1337	2777	477	1237	2877
MN	Male	Rural	490	1345	2357	490	1245	2457
MN	Male	Rural	629	1394	2817	739	1162	2817
MO	Male	Rural	442	991	2523	452	881	2523
MO	Male	Rural	450	972	2462	470	862	2962
MO	Male	Rural	460	995	2631	470	893	2731
MO	Male	Rural	470	1131	2397	480	1011	2497
MO	Male	Rural	474	896	2496	494	776	2696
MO	Male	Rural	485	1019	2264	495	1012	2664
MO	Male	Rural	486	877	2368	496	867	2568
MO	Male	Rural	488	927	2724	498	817	2924
MO	Male	Rural	493	1058	2673	593	1038	2973
MO	Male	Rural	527	967	2500	547	957	2600
MO	Male	Rural	579	1009	2693	589	1006	2693
MO	Male	Rural	592	1140	2910	588	1020	2690
MO	Male	Rural	602	804	2345	622	702	2445
MO	Male	Rural	623	938	2430	633	828	2630
MO	Male	Rural	672	1053	2213	682	1033	2413
RJ	Male	Urban	402	1032	1900	622	1022	1900
RJ	Male	Urban	404	1058	2350	504	1058	2350
RJ	Male	Urban	432	1072	2350	542	1072	2350
RJ	Male	Urban	458	1058	2693	468	1058	2693
RJ	Male	Urban	458	1283	2490	478	1183	2590
RJ	Male	Urban	463	1053	2652	473	1053	2652
RJ	Male	Urban	466	1051	2093	476	1051	2093
RJ	Male	Urban	481	1032	2701	491	1022	2901
RJ	Male	Urban	504	1073	3303	524	1063	3027
RJ	Male	Urban	504	1261	2844	624	1161	2844
RJ	Male	Urban	513	1049	3303	633	1039	3027
RJ	Male	Urban	522	1045	2503	732	1035	2603
RJ	Male	Urban	566	1054	2722	666	1054	2722
RJ	Male	Urban	574	1027	2986	684	1017	2986
RJ	Male	Urban	616	1077	2350	716	1067	2350
SD	Male	Rural	415	1072	2530	420	1052	2680
SD	Male	Rural	424	1104	2527	494	1104	2827
SD	Male	Rural	428	945	2417	438	835	2617
SD	Male	Rural	461	845	2344	471	735	2444
SD	Male	Rural	468	1152	2377	478	1022	2877
SD	Male	Rural	469	1181	2279	490	1020	2500
SD	Male	Rural	479	1061	2568	499	1051	2668
SD	Male	Rural	496	1111	2148	497	1011	2348
SD	Male	Rural	513	1017	2348	523	1011	2348
SD	Male	Rural	523	1037	2173	533	1007	2273
SD	Male	Rural	533	974	2437	553	964	2837
SD	Male	Rural	573	908	2457	583	801	2957
SD	Male	Rural	590	882	2355	598	772	2655
SD	Male	Rural	641	779	2835	651	759	2935
SD	Male	Rural	651	1058	2352	661	1038	2452