Exploring Sociolectal Identity Through Speech Rhythm in Philippine English

Abstract

This study explores rhythm metrics as a sociolinguistic marker in Philippine English (PhE), addressing gaps in understanding rhythmic variation in Southeast Asian Englishes. It aims to uncover how rhythmic patterns reflect sociolectal identities within a multilingual context. Using acoustic data from 30 participants in Manila, rhythm metrics (%V, ΔV, ΔC, nPVI, and rPVI) were analyzed to examine rhythmic tendencies. The findings reveal distinct patterns: PhE acrolect aligns with stress-timed rhythms of general American English, PhE basilect reflects syllable-timed features similar to Spanish, and PhE mesolect occupies a hybrid position blending elements of both. By emphasizing rhythm as a key identifier of sociolectal variation, this study advances the understanding of linguistic diversity in World Englishes and provides a novel framework for exploring identity in multilingual settings.

1. Introduction

World Englishes demonstrate the adaptability of English across diverse linguistic, cultural, and historical contexts, shaping distinct phonological and prosodic characteristics. In multilingual societies, language variation is not just structural but also a marker of identity, as speakers adjust their speech to different social contexts. Philippine English (PhE) has evolved within a multilingual society influenced by American English and over 170 local languages. As a nativized variety1, PhE reflects an interplay between global linguistic norms and indigenous languages, resulting in distinctive phonological, grammatical, and rhythmic features. Scholars such as Borlongan (2023) and Schneider (2003, 2023) recognize PhE as an independent English variety with its own linguistic norms, reinforcing its status within the World Englishes paradigm.

1.1. Sociolectal Variation in PhE

PhE is shaped by both historical factors—including American colonial rule, which institutionalized English as a medium of instruction and governance—and the contemporary multilingual reality, where English coexists with over 170 local languages. These sociolinguistic conditions have given rise to sociolectal variation, a phenomenon in which distinct linguistic features emerge within different social groups shaped by unequal access to English across education, media, and institutional settings (Tayao, 2004).

A sociolect refers to a linguistic variety spoken by a particular social group, distinguished by phonological, grammatical, or lexical differences that correlate with social variables such as class, education, occupation, and linguistic exposure. In the case of PhE, sociolectal variation is commonly described using three prototypical reference points—acrolect, mesolect, and basilect—rather than as rigid categories. These labels represent salient clusters of linguistic features, but speakers often occupy intermediate positions, reflecting a gradient continuum of variation rather than strictly bounded varieties (Magpale, 2024; Magpale & Hong, 2024). This study adopts Tayao’s (2004) framework, which conceptualizes PhE sociolectal stratification, but extends it by operationalizing sociolectal status through frequency of English use across social domains and self-rated English proficiency (see Section 2.1). These measurable factors provide a quantitative basis for distinguishing sociolects, ensuring consistency in classification.

PhE acrolect is spoken by individuals with high English proficiency, typically those who have received formal education in English-medium institutions and have greater exposure to English in academic and professional settings. PhE acrolect speakers exhibit phonological features aligned with general American English, including a full consonantal and vowel inventory, and characteristics such as vowel reduction in unstressed syllables (Lesho, 2017).

PhE mesolect represents a transitional sociolect, where speakers exhibit phonological features from both the PhE acrolect and PhE basilect due to bilingual interaction between English and Filipino. This variety is characterized by partial vowel reduction and epenthesis in consonant clusters, indicating phonotactic influences from both the English and Philippine languages (Tayao, 2008).

PhE basilect is the most localized sociolect, strongly influenced by Philippine languages and typically spoken by individuals with limited formal education in English and less frequent exposure to English-speaking environments. Speakers of this variety exhibit phonological substitutions (e.g., /f/ → /p/, /v/ → /b/), reduced vowel inventories, and a more syllable-timed rhythm, reflecting influences from local Philippine languages (Regala-Flores, 2014).

Beyond these segmental characteristics, each sociolect also exhibits distinct phonological processes that have implications for rhythmic variation. Vowel reduction, often observed in PhE acrolectal speech (Lesho, 2017; Magpale, 2024), involves the weakening or centralization of unstressed vowels, which contributes to increased variability in vocalic duration and lends a more stress-timed quality to speech. In contrast, epenthesis—the insertion of a vowel, typically /ɪ/, within consonant clusters—is more frequently found in PhE basilectal speech (Magpale & Hong, 2024), leading to more evenly timed syllables, increasing vocalic regularity and reducing temporal variation. Additionally, consonant cluster simplification, where complex onsets or codas are reduced (e.g., /str/ → / ɪs.tr/), is more common among PhE mesolectal and basilectal speakers (Tayao, 2004; Regala-Flores, 2014), reflecting both articulatory ease and the influence of Philippine languages. These processes not only shape phonotactic structure but also modulate speech timing, offering valuable insight into the prosodic distinctions across PhE sociolects.

The directionality of these variables follows a clear pattern: individuals with greater English exposure, formal education, and access to English-dominant domains tend to align with PhE acrolect, whereas those with lower exposure, limited formal education, and stronger local language influence tend to align with PhE basilect. PhE mesolect occupies a middle ground, reflecting a hybrid linguistic profile that adapts to varying communicative contexts.

Recent research has expanded our understanding of these sociolectal differences by modeling linguistic variation using probabilistic approaches. For instance, Magpale (2024) and Magpale and Hong (2024) employed constraint-based grammars to capture the gradient nature of sociolectal variation, demonstrating that linguistic features shift dynamically across the PhE acrolect–mesolect–basilect continuum. While these studies offer valuable insights into segmental and suprasegmental variation, they have not explored rhythmic aspects such as speech rhythm in depth. This study aims to bridge this gap by analyzing rhythm as a sociolinguistic marker of sociolectal identity, contributing to the broader discussion on prosody, phonological variation, and linguistic diversity in World Englishes.

1.2. Rhythm in Linguistic Typology and Quantitative Analysis

Speech rhythmic typology traditionally categorizes languages into stress-timed, syllable-timed, or mora-timed based on the distribution of rhythmic units. Stress-timed languages, such as English and German, have intervals between stressed syllables occurring at roughly regular intervals, resulting in variable syllable lengths. In contrast, syllable-timed languages, like Spanish and French, exhibit syllables of nearly equal duration regardless of stress. Mora-timed languages, such as Japanese, maintain uniform timing for morae, a smaller unit within syllables (Ladefoged & Johnson, 2011).

While these classifications have been widely used, early rhythm studies relied on perceptual observations, which often resulted in inconsistent language classifications (Dauer, 1983; Dasher & Bolinger, 1982). To address this limitation, researchers have developed quantitative rhythm metrics to systematically measure temporal patterns in speech and provide empirical validation for rhythm typologies.

One of the key developments in rhythm studies was the introduction of rhythm metrics by Ramus et al. (1999), which provided a standardized method for analyzing rhythmic variability. Their metrics—%V (proportion of vocalic intervals), ΔV (standard deviation of vowel durations), and ΔC (standard deviation of consonant durations)—are widely used to measure variability in vocalic and consonantal intervals.

The %V metric, which represents the proportion of an utterance occupied by vowels, is particularly useful in distinguishing between stress-timed and syllable-timed languages. Stress-timed languages (e.g., English, Dutch) tend to have lower %V values due to extensive vowel reduction and complex syllable structures, often featuring consonant clusters. In contrast, syllable-timed languages (e.g., Spanish, French) exhibit higher %V values, as vowels are more consistently realized, and syllable durations remain relatively stable across speech. Mora-timed languages (e.g., Japanese) tend to display %V values distinct from both categories, reflecting their unique rhythmic organization.

In addition to %V, ΔV (the standard deviation of vowel durations) and ΔC (the standard deviation of consonant durations) quantify rhythmic variability by measuring durational differences within vocalic and consonantal intervals. Ramus et al. (1999) found that %V and ΔC were strongly correlated, indicating that they captured similar aspects of syllable structure and vocalic-consonantal balance. In contrast, ΔV appeared to reflect a different dimension of rhythm, though the authors did not elaborate on its specific function. As such, ΔV remains less clearly interpreted, yet continues to be used as a complementary metric in rhythm analysis.

By providing quantifiable measures, these rhythm metrics have enabled a more systematic and comparative analysis of rhythmic typologies across languages, refining earlier perceptual classifications of rhythm. Taking a different approach to rhythm analysis, Low et al. (2000) introduced the Pairwise Variability Index (PVI), which quantifies the degree of durational variability between successive speech units. Unlike the metrics proposed by Ramus et al. (1999), which focused on global interval variability, the PVI emphasizes local timing contrasts, offering an alternative lens for examining rhythmic patterns. The PVI is calculated separately for vocalic (vowel) and consonantal intervals within an utterance, capturing differences in rhythmic timing patterns. The formula for the PVI is as follows:

$$\mathrm{P}\mathrm{V}\mathrm{I}=100\times \left({\displaystyle \sum _{k=1}^{m-1}} \left|\left(d_k-d_{k+1}\right)/\left(\left(d_k+d_{k+1}\right)/2\right)\right|\right)/(m-1)$$

(1)

where m represents the total number of measured intervals, and d_k and d_k+1 denote the durations of two successive intervals. The absolute difference between them is normalized by dividing by their mean duration.

Grabe and Low (2002) expanded on these metrics, applying both normalized Pairwise Variability Index (nPVI) and raw Pairwise Variability Index (rPVI) to a range of languages, including stress-timed (e.g., English, Dutch), syllable-timed (e.g., French, Spanish), and mora-timed (e.g., Japanese) languages. The nPVI normalizes for speech rate by dividing the absolute duration differences between successive vocalic intervals by their mean duration, ensuring comparability across utterances of varying speeds. However, it is not entirely immune to rate-related effects. Variations in articulation tempo, vowel reduction, or prosodic phrasing can still affect the relative timing of vowels, introducing rhythmic variability that may be reflected in nPVI values despite normalization (Arvaniti, 2009; White & Mattys, 2007). In contrast, rPVI remains unnormalized, directly measuring duration variability without accounting for speech rate. Their findings confirmed higher nPVI values in stress-timed languages, moderate values in rhythmically indeterminate languages (e.g., Catalan, Polish), and lower values in syllable-timed languages. Interestingly, they noted that Japanese, traditionally categorized as mora-timed, did not occupy a distinct rhythmic space, but rather overlapped with syllable-timed patterns in certain contexts, challenging traditional classifications.

Arvaniti (2009) highlights that %V, ΔV, ΔC, and nPVI can be influenced not only by speech rate and segmental properties, but also by prosodic phrasing, which complicates their use as strict typological indicators. However, subsequent rhythm studies have shown that some of these confounds can be mitigated—for instance, by marking the boundaries of intonational phrases (IPs), avoiding PVI calculation across phrase boundaries, and excluding final syllables of IPs, which tend to be lengthened (White & Mattys, 2007). While segmental variability is a natural part of what rhythm metrics aim to capture, there remains ongoing debate over whether rhythm exists independently of segmental characteristics. As Arvaniti and others note, such independence would be difficult to determine in languages like English that exhibit vowel reduction and phonemic length but may be more assessable in typologically distinct systems with lexical stress but fewer confounding segmental features. Nevertheless, these metrics remain widely applied in rhythm studies, particularly in research on World Englishes and L2 varieties, where rhythm is analyzed as a continuum rather than as a rigid classification (Grabe & Low, 2002; Mok & Lee, 2008). Since this study investigates sociolectal rhythm variation within PhE, rather than attempting to classify it into a binary rhythmic category, these measures remain valuable for capturing differences in phonotactic and durational variability across speaker groups.

Building upon these studies, White and Mattys (2007) conducted a comparative evaluation of rhythm metrics, including nPVI, rPVI, ΔV, ΔC, %V, and rate-normalized measures such as VarcoV (White & Mattys, 2007) and VarcoC (Dellwo, 2006).2 They found that while some interval measures, such as ΔV and ΔC, were highly influenced by speech rate, nPVI was particularly effective in distinguishing between stress-timed and syllable-timed languages. Their study reinforced the claim that nPVI is a reliable indicator of rhythmic differences, but they also cautioned against a rigid classification of languages into distinct rhythmic categories. Instead, they highlighted the role of gradient variation, where languages exhibit rhythmic tendencies rather than absolute classifications. Furthermore, their findings suggest that while stress-timed languages generally have higher nPVI values, certain factors, such as phonotactic constraints and segmental properties, can influence rhythmic outcomes, making it necessary to interpret rhythm metrics within the broader phonological context of each language. The adoption of rhythm metrics has not only enabled researchers to objectively classify languages along a rhythmic continuum, but has also highlighted the interplay of linguistic features underlying these patterns. For example, ΔV and ΔC emphasize the influence of syllable complexity and vowel reduction, which are key factors in distinguishing stress- and syllable-timed rhythms. Languages with complex syllable structures and significant vowel reduction, such as English, exhibit higher rhythmic variability, whereas languages with simpler syllable structures and minimal reduction, such as Spanish, align with syllable-timed characteristics.

In terms of localized English varieties, Tan and Low (2014) further explored rhythm metrics by analyzing Malaysian English (MalE) and Singapore English (SgE) using the Pairwise Variability Index (PVI) and VarcoV to quantify rhythmic differences. Their study provided empirical acoustic analysis, moving beyond previous impressionistic observations of MalE rhythm. The findings revealed that MalE aligns more strongly with syllable-timed rhythms than SgE. In both read and spontaneous speech, MalE speakers exhibited less vowel reduction, resulting in more stable syllable durations and a greater tendency toward syllable-timed rhythmic patterns. Conversely, SgE speakers exhibited higher vowel reduction rates, making their rhythm less strictly syllable-timed than MalE, though still distinct from stress-timed languages like British English.

Moreover, their syllable-based analysis of specific utterances confirmed that MalE speakers retained fuller vowels, whereas SgE speakers typically reduced them. This trend was observed consistently across both read and spontaneous speech, reinforcing the conclusion that MalE follows a stronger syllable-timed rhythmic pattern than SgE. While SgE incorporates stress-timed elements, it does so through phonological restructuring influenced by British English norms, particularly in formal education and media exposure. However, Tan and Low (2014) observed that the rhythmic differences between MalE and SgE are smaller than those between SgE and British English, suggesting that both varieties remain largely syllable-timed but exist on a rhythmic continuum rather than in discrete categories.

These findings underscore the effectiveness of quantitative rhythm metrics in capturing prosodic distinctions across nativized English varieties and highlight the role of phonological variation and vowel reduction patterns in shaping rhythmic characteristics.

1.3. Research Gap

Studies on rhythm in nativized English varieties, such as MalE and SgE, have employed rhythm metrics like %V, ΔV, ΔC, and the Pairwise Variability Index (PVI) to examine rhythmic tendencies across different linguistic contexts. These investigations reveal significant insights, such as MalE’s alignment with syllable-timed rhythms and SgE’s incorporation of stress-timed features, reflecting the influence of cultural, educational, and linguistic ecologies. However, these studies often treat MalE and SgE as homogeneous systems, without accounting for the internal variation that arises from sociolinguistic factors like socioeconomic status, education, and linguistic exposure.

In contrast, PhE operates within a more stratified sociolinguistic framework, shaped by the Philippines’ multilingual ecology and its unique sociolectal distinctions. As mentioned earlier, Tayao’s (2004) framework categorizes PhE into acrolect, mesolect, and basilect varieties, reflecting varying levels of proficiency and alignment with stress-timed and syllable-timed patterns. This stratification provides an opportunity to explore how rhythm varies across sociolects, offering a more granular understanding of PhE’s prosody. Despite this potential, research on PhE’s rhythmic properties remains sparse, and existing studies often overlook the impact of sociolectal variation on prosodic features like rhythm.

Furthermore, while rhythm metrics have proven effective in distinguishing between stress-timed and syllable-timed languages, their application to sociolectal variation within a single English variety remains underexplored. Existing studies have investigated rhythmic variation in dialects and contact varieties of English (e.g., Clopper & Smiljanic, 2015; Carter, 2005; Enzinna, 2016; Torgersen & Szakay, 2012), as well as in regional varieties of Spanish and French (e.g., O’Rourke, 2008; Kaminskaïa et al., 2015). These works provide useful precedents for examining rhythm beyond typological classification, yet few have focused on stratified sociolects within World Englishes. Tayao’s (2008) research on PhE has focused on segmental and suprasegmental features, such as phonological substitutions and vowel reduction, without systematically analyzing how rhythm interacts with social and educational factors across the PhE acrolect, mesolect, and basilect. Given that multilingual individuals often develop rhythmic flexibility to accommodate different linguistic contexts, the present study contributes to understanding how rhythmic adaptation serves as an index of sociolectal identity in a multilingual society. This gap in the literature limits our understanding of how PhE’s rhythm reflects its multilingual and sociocultural context, as well as its place within the broader continuum of stress- and syllable-timed rhythms observed in World Englishes.

By addressing these gaps, this study extends the application of rhythm metrics to analyze the sociolectal variations of PhE’s rhythm. Unlike studies on MalE and SgE, this research emphasizes the role of sociolectal stratification in shaping rhythmic tendencies, offering new insights into how linguistic and social diversity influence the prosody of nativized Englishes. Additionally, by situating speech rhythm within the broader discourse on linguistic identity and multilingualism, this study reveals how speakers navigate and express their social affiliations through rhythmic variation, thereby contributing to discussions on identity formation in multilingual contexts.

1.4. The Current Study

This study addresses the research gap by applying quantitative rhythm metrics to analyze the rhythmic features of PhE across its sociolects. Metrics such as %V (percentage of vocalic intervals), ΔV and ΔC (standard deviation of vocalic and consonantal intervals) from Ramus et al. (1999), along with nPVI (normalized Pairwise Variability Index) from Grabe and Low (2002) are employed to classify PhE’s rhythm.

These tools enable the investigation of the interaction between stress-timed and syllable-timed influences in PhE, which emerges from language contact between English and Filipino. English, a stress-timed language, exhibits uneven syllable durations, vowel reductions, and alternating patterns of stressed and unstressed syllables. In contrast, Filipino—the standardized form of Tagalog—is largely syllable-timed, characterized by consistent syllable durations, minimal vowel reduction, and a tendency to maintain clear segmental articulation regardless of stress. This classification has been empirically validated in a computational rhythm analysis by Guevara et al. (2010), which demonstrated that Filipino patterns closely with other syllable-timed languages such as French and Spanish. Given the distinct rhythmic properties of English and Filipino, the influence of these languages on PhE rhythm may vary across sociolects. PhE acrolect speakers with greater exposure to English are expected to exhibit stress-timed patterns, while PhE basilectal speakers may retain syllable-timed features influenced by Filipino.

1.4.1. Research Questions:

This study addresses the following research questions:

A. What are the rhythmic characteristics of the PhE acrolect, mesolect, and basilect sociolects?

• Examined through rhythm metrics (%V, ΔV, ΔC, nPVI, rPVI) derived from recorded speech samples.
• Analyzed using one-way ANOVA and Tukey’s HSD to determine statistically significant rhythmic differences across sociolects.

B. How do phonological processes such as vowel reduction, epenthesis, and consonant cluster simplification shape rhythmic variation across PhE sociolects?

• Investigated through descriptive analysis of rhythmic metrics in relation to observed phonological patterns.
• Interpreted qualitatively using representative phonetic realizations and prior sociophonetic findings (e.g., Tayao, 2008; Magpale & Hong, 2024).

C. How do the rhythmic properties of PhE sociolects compare to established stress-timed and syllable-timed languages?

• Compared using rhythm metric benchmarks from typological studies (Grabe & Low, 2002; Ramus et al., 1999).
• Evaluated through comparative tables and visual plots to determine each sociolect’s position along the stress–syllable timing continuum.

By addressing these questions, the study contributes to a more nuanced understanding of rhythm as a sociolinguistic marker in nativized Englishes and sheds light on how speakers navigate rhythmic variation across social groups in multilingual settings.

1.4.2. Hypotheses

Drawing on previous studies on rhythm typology and phonotactics (e.g., Grabe & Low, 2002; Ramus et al., 1999; Tayao, 2008), this study proposes the following hypotheses regarding the rhythmic characteristics of Philippine English (PhE) sociolects. For Research Question A, it is expected that the rhythmic patterns of PhE sociolects will reflect distinct timing tendencies. The PhE acrolect is hypothesized to exhibit lower %V, higher nPVI, and greater ΔV and ΔC values, reflecting a stress-timed rhythm influenced by vowel reduction and more complex syllable structures. In contrast, the PhE basilect is expected to show higher %V, lower nPVI, and reduced ΔV and ΔC values, consistent with a syllable-timed rhythm characterized by minimal vowel reduction and simpler phonotactics. Situated between these extremes, the PhE mesolect is anticipated to display intermediate values across the rhythm metrics, reflecting its hybrid phonological profile.

For Research Question B, certain phonological processes—namely, vowel reduction, epenthesis, and consonant cluster simplification—are expected to be associated with rhythmic tendencies across sociolects based on descriptive patterns rather than statistical correlations. Vowel reduction is expected to be most frequent in PhE acrolectal speech, contributing to shorter vocalic durations and greater variability, as reflected in the lower %V and higher nPVI. Epenthesis, or the insertion of vowels within consonant clusters, is anticipated to occur more often in PhE basilectal speech, leading to increased %V and more regular vowel timing. Meanwhile, consonant cluster simplification is expected to emerge more prominently in PhE mesolectal and basilectal speech, potentially contributing to lower ΔC values by reducing the variability of consonantal timing.

For Research Question C, the rhythmic characteristics of PhE sociolects are hypothesized to align with typological patterns observed in stress-timed and syllable-timed languages. PhE acrolect is expected to resemble stress-timed languages such as English, given its greater timing variability and phonotactic complexity. PhE basilect is predicted to align more closely with syllable-timed languages such as Spanish or Tagalog, due to its more stable syllable timing and limited vowel reduction. PhE mesolect, in turn, is anticipated to reflect a mix of rhythmic features, situating it along a continuum between the two prosodic types. These hypotheses offer a structured framework for interpreting the study’s findings and for situating rhythmic variation within broader models of sociophonetic identity in nativized Englishes.

2. Methods

This study employs a systematic approach to examine rhythmic variation across the sociolects of PhE. Combining participant profiling, controlled data collection, and quantitative acoustic analysis, the methodology ensures precise measurement of rhythmic patterns. The procedures described below are designed to provide robust and replicable insights into the sociolectal variation of PhE rhythm.

2.1. Participants

This study recruited 30 participants from Manila, evenly distributed among the three sociolects of Philippine English (PhE): acrolect, mesolect, and basilect. The participants, aged 18 to 35 years (M = 26.5, SD = 5.93), included 15 males and 15 females and were drawn from academic institutions, workplaces, and community networks to ensure diverse linguistic exposure.

To classify the participants, the study adopted Tayao’s (2004) framework, which categorizes sociolects based on linguistic exposure and usage patterns. Classification was determined using a self-administered demographic survey designed to assess three key sociolinguistic dimensions: (A) frequency of English use across five domains (home, workplace, restaurant, mall, and church); (B) language preference in the same domains on a scale from 100% English to 100% Filipino; and (C) self-rated English proficiency across four language skills (reading, writing, speaking, and listening). The full survey instrument is included in Appendix A.

Unlike Tayao’s original approach, which emphasized phonological features, this study operationalized sociolect classification through quantifiable linguistic and social indicators. Responses were numerically coded (1–5) and aggregated to identify each participant’s dominant sociolectal profile. Scores of 4 or higher reflected consistent acrolectal alignment, scores around 3 indicated mesolectal patterns, and scores near 1 suggested basilectal tendencies. Where responses straddled categories, the participants were classified based on majority trait alignment across the three survey dimensions using a prototypical model.

While the possibility of overlap across sociolect boundaries is acknowledged—an inherent feature of gradient sociolinguistic identity—systematic grouping is necessary for comparative rhythm analysis. This approach draws on prior descriptions of acrolectal and basilectal PhE varieties (Lesho, 2017; Regala-Flores, 2014) and maintains analytical clarity while reflecting the sociolinguistic continuum.

Table 1. Frequency of use of English across settings.

Setting	PhE Acrolect	PhE Mesolect	PhE Basilect
Home	4.9	2.8	1
Workplace	5	3	1.3
Restaurant	4.6	3	1.4
Mall	4.6	2.4	1.3
Church	5	3.2	1.4

Scale: 5 = always, 4 = often, 3 = sometimes, 2 = seldom, 1 = never.

,

Table 2. Language preference across settings.

Setting	PhE Acrolect	PhE Mesolect	PhE Basilect
Home	4.9	2	1.4
Workplace	5	3	1
Restaurant	4.5	2.4	1.1
Mall	4.3	2.4	1.4
Church	4.6	2.1	1.3

Scale: 5 = 100% English, 4 = 75% English and 25% Filipino, 3 = 50% English and 50% Filipino, 2 = 25% English and 75% Filipino, 1 = 100% Filipino.

and

Table 3. Self-assessment of English proficiency.

Setting	PhE Acrolect	PhE Mesolect	PhE Basilect
Reading	4.9	2.8	1.4
Writing	5	3.2	1
Speaking	4.6	3	1.2
Listening	4.6	2.4	1.4

Scale: 5 = excellent, 4 = good, 3 = average, 2 = below average, 1 = weak.

summarize the classification results. Table 1 shows that PhE acrolect speakers reported the highest frequency of English use, while PhE basilect speakers reported the lowest. Table 2 shows that acrolect speakers preferred English-dominant interactions, while basilect speakers favored Filipino. Table 3 confirms that acrolect speakers rated themselves highest in English proficiency across all skills, while basilect speakers rated themselves lowest.

Based on the composite scores derived from the demographic survey, the 30 participants were evenly classified into three groups: 10 PhE acrolect, 10 PhE mesolect, and 10 PhE basilect speakers. This balanced distribution allows for the systematic comparison of rhythm metrics across sociolectal categories using statistical analysis.

Justification for Manila as the Study Site

This study focuses on Manila to ensure consistency in analyzing sociolectal rhythm variation in PhE. As the country’s economic and educational center, Manila presents a clear stratification of PhE acrolect, mesolect, and basilect speakers, aligning with Tayao’s (2004) framework.

Conducting the study in a single metropolitan area helps avoid regional phonological differences that may arise from the varying influences of local languages outside the capital. Manila’s multilingual environment, where English coexists with Filipino, provides a controlled setting for examining rhythmic patterns without the added complexity of regional variation.

Additionally, limiting the study to Manila allows for greater consistency in data collection and analysis, ensuring comparability with previous research on Philippine English phonology and rhythm (e.g., Tayao, 2004; Lesho, 2017). Future research may explore regional variations in Philippine English rhythm to extend these findings beyond the Manila setting.

2.2. Ethical Considerations

The study adhered to ethical guidelines to ensure the protection of participants’ rights and well-being. All participants were fully informed about the purpose, scope, and procedures of the study through a detailed consent form. The voluntary nature of their participation was emphasized, and they were assured of their right to withdraw at any point without any consequences.

Informed consent was obtained from all participants prior to their involvement. Confidentiality and anonymity were strictly maintained by assigning pseudonyms and securely storing data in password-protected files accessible only to the research team.

The ethical approach followed established guidelines, ensuring respect for the participants’ autonomy and privacy while maintaining the integrity of the research process.

2.3. Materials

The controlled text, The North Wind and the Sun, served as the primary reading material for eliciting speech samples. This passage was selected for its phonological balance and comparability across speakers, a quality that makes it widely used in phonetic and rhythm research (e.g., Grabe & Low, 2002). The goal was not to generalize from the reading of this specific passage, but to use it as a neutral and consistent elicitation tool that could reveal group-level rhythmic tendencies.

The passage includes phonological environments relevant to PhE variation, such as coda cluster simplification (last → [las]), vowel reduction (traveller → [ˈtræv.lər] vs. [ˈtra.ve.ler]), and epenthesis in consonant clusters (stronger → [ɪs.troŋ.ger]). These features contribute to rhythmic variation, allowing the study to examine how timing patterns differ systematically across sociolects.

The participants were instructed to read the text at a natural pace to capture their typical rhythmic tendencies. A controlled reading task was chosen to ensure consistent elicitation conditions across speakers, minimizing lexical and syntactic variability while allowing rhythm metrics (%V, nPVI, rPVI) to reflect socially conditioned phonetic structuring. Although reading speech is more regular than spontaneous conversation, it provides a valid and replicable foundation for identifying sociolectal rhythm patterns in a comparative framework.

2.4. Recordings

Speech recordings were conducted in a quiet, controlled environment to ensure high-quality data collection. A microphone and a digital audio recorder with a sampling rate of 44.1 kHz were used to capture the recordings. The participants were given time to familiarize themselves with the text before the recording session and to complete one practice run prior to the final recording. All sessions were conducted under uniform conditions to ensure consistency across recordings.

To ensure data quality and consistency, all participants read the same controlled text (“The North Wind and the Sun”) at a natural pace in a quiet, controlled environment. While the quiet setting minimized external noise and ensured clean acoustic input, it did not eliminate inter- or intra-speaker variability, which is inherent in any sociophonetic study. Instead, consistency was promoted by using the same speech task across all participants, allowing for a clearer comparison of rhythmic metrics across sociolects.

2.5. Measurement

The recorded speech samples were analyzed acoustically using Praat. Each recording was manually segmented into vocalic and consonantal intervals for precise measurement of rhythm metrics. The study utilized several rhythm metrics to capture and analyze rhythmic variations across sociolects. Vocalic and consonantal intervals were manually segmented to accurately measure rhythm metrics like %V, ΔV, and ΔC. The segmentation process involved carefully identifying each vowel and consonant interval within the speech signal to ensure the accuracy of the rhythm analysis. These metrics and their descriptions are provided in

Table 4. Description of rhythm metrics ¹.

Metric	Description	Source
nPVI (Normalized Pairwise Variability Index)	Measures the variability of vowel durations across adjacent syllables, normalized for speech rate.	Grabe and Low (2002)
Intervocalic rPVI	Measures variability of consonantal (intervocalic) intervals without normalization.
Raw PVI (Pairwise Variability Index)	Directly measures variability between adjacent intervals (vocalic or consonantal).
%V (Proportion of Vocalic Intervals)	Percentage of total speech time occupied by vowels.	Ramus et al. (1999)
ΔV (Standard Deviation of Vocalic Intervals)	Standard deviation of vowel durations within a speech sample.
ΔC (Standard Deviation of Consonantal Intervals)	Standard deviation of consonantal intervals within a speech sample.

¹ The detailed steps for calculating rhythm metrics (%V, ΔV, ΔC, nPVI, and rPVI) are provided in Appendix B: Calculating Rhythm Metrics.

below.

2.6. Statistical Analysis

Statistical analyses were conducted using R software v.4.4.0. (R Core Team, n.d.) to examine rhythmic variation across sociolects. A one-way ANOVA tested for significant differences in rhythm metrics among the PhE acrolect, mesolect, and basilect groups. Tukey’s HSD post hoc tests were used to identify specific group differences. The dataset was relatively balanced (10 participants per group), and assumptions of normality and homogeneity of variance were met.

While mixed-effects modeling is often used to account for inter- and intra-speaker variability, this study prioritized the identification of systematic rhythmic differences across sociolectal groups rather than modeling speaker-specific random effects. A fixed-effects approach using a one-way ANOVA was therefore selected, providing a transparent and interpretable analytic framework. This decision aligns with White and Mattys (2007), who used rhythm metrics and ANOVA to examine rhythmic variation between L1 and L2 speaker groups. This study focuses on structured rhythmic variation across sociolects, not random variability within individuals.

3. Results

This section presents the findings of the study, focusing on rhythmic variation across the three sociolects of PhE: acrolect, mesolect, and basilect. Using quantitative rhythm metrics such as %V, ΔV, ΔC, nPVI, and rPVI, the analysis identifies distinct rhythmic patterns that reflect the sociolectal stratification of PhE. Comparisons among the three sociolects are presented first, ensuring a clear analysis of internal variation. Broader rhythmic typological comparisons (e.g., stress-timed and syllable-timed languages) are reserved for the discussion section. This structure allows for an in-depth understanding of intra-sociolectal differences before placing PhE within a typological framework, thereby enhancing the validity of the findings by ensuring that sociolectal variation is examined in its own context before being compared to external language models.

3.1. Descriptive Findings and Sociolectal Patterns

Table 5. Mean values of rhythm metrics (%V, ΔV, ΔC, nPVI, rPVI) for Philippine English sociolects. Standard deviations are shown in parentheses.

Sociolect	%V	ΔV	ΔC	nPVI	rPVI
PhE acrolect	40.49 (1.9)	3.81 (2.05)	5.24 (1.10)	54.1	66.3
PhE mesolect	42.90 (1.2)	4.03 (3.02)	4.64 (0.86)	44.2	49.1
PhE basilect	44.13 (1.86)	3.65 (1.27)	4.85 (1.76)	30.2	56.3

presents the mean values and standard deviations (SDs) for %V, ΔV, ΔC, nPVI, and rPVI across PhE acrolect, PhE mesolect, and PhE basilect, highlighting distinct rhythmic patterns. For %V and nPVI, a clear sociolectal gradient emerges, with PhE acrolect speakers exhibiting the lowest %V (40.49, SD = 1.9) and the highest nPVI (54.1), aligning with a stress-timed rhythm. Conversely, PhE basilect speakers show the highest %V (44.13, SD = 1.86) and lowest nPVI (30.2), characteristic of a syllable-timed rhythm. PhE mesolect speakers occupy an intermediate position (%V = 42.9, SD = 1.2; nPVI = 44.2), reinforcing the notion that PhE mesolect rhythm blends elements of both stress- and syllable-timed patterns.

For ΔV, while PhE mesolect (ΔV = 4.03, SD = 3.02) has a slightly higher mean than PhE acrolect (ΔV = 3.81, SD = 2.05) and PhE basilect (ΔV = 3.65, SD = 1.27), its notably larger SD suggests substantial intra-group variation. This variability indicates that some PhE mesolect speakers reduce vowels similarly to PhE acrolect, while others fully articulate them like PhE basilect speakers, contributing to its intermediate yet flexible rhythmic profile.

For ΔC, PhE acrolect (ΔC = 5.24, SD = 1.10) exhibits the highest consonantal variability due to the retention of complex clusters, while PhE basilect (ΔC = 4.85, SD = 1.76) has a slightly lower ΔC, reflecting its tendency to simplify clusters. PhE mesolect (ΔC = 4.64, SD = 0.86), however, presents the lowest variability, suggesting that consonantal timing in mesolect speech is more stable than in the other two sociolects. The low SD for PhE mesolect ΔC further implies that mesolect speakers tend to apply consistent phonotactic strategies when producing consonantal sequences.

For rPVI, while PhE acrolect (66.3) and PhE basilect (56.3) show distinct rhythmic patterns, PhE mesolect (49.1) presents a notably different timing structure. The mesolect’s intermediate rPVI value and variability suggest that speakers may adjust their consonantal timing depending on sociolinguistic context, leading to rhythmic flexibility rather than a strict acrolect-to-basilect continuum.

3.1.1. Vocalic Rhythm Metrics: %V, ΔV and ΔC

The percentage of vocalic intervals (%V) illustrates the rhythmic distinction between stress-timed and syllable-timed tendencies. A high %V, characteristic of syllable-timed rhythms, reflects fully pronounced vowels, even in unstressed syllables. For example, in the PhE basilect, all vowels are articulated, including the unstressed second syllable (e.g., [ve]- for /træ.və.lər/ ‘traveller’), resulting in higher %V values, as seen below in (1). Tayao (2008) notes the absence of schwa in the PhE basilect, further supporting this uniform vowel duration. Conversely, the PhE acrolect exhibits a low %V, reflecting reduced or omitted vowels in unstressed syllables, characteristic of stress-timed languages. For instance, the unstressed vowel /ə/ in /træ.və.lər/ ‘traveller’ is omitted entirely ([ˈtræv.lər]). PhE mesolect, as expected, exhibits an intermediate %V value, with speakers alternating between vowel reduction (as seen in PhE acrolect) and vowel retention (as observed in PhE basilect). Some PhE mesolectal speakers omit the unstressed vowel, while others retain it, leading to greater intra-group variability, as reflected in the standard deviation.

Similarly, ΔV values measure variability in vowel durations, with low ΔV indicating consistent vowel timing. The PhE basilect demonstrates uniform vowel durations, resulting in low ΔV, while the PhE acrolect, influenced by vowel reduction processes, exhibits greater variability and higher ΔV values. PhE mesolect falls between these two patterns, with some speakers reducing vowels in unstressed syllables (mirroring PhE acrolectal rhythm) and others maintaining full vowel articulation (closer to PhE basilectal rhythm). This variation results in a higher ΔV standard deviation for mesolectal speakers, indicating a flexible rhythmic structure that could depend on linguistic and social context.

(1)	Realizations for /træ.və.lər/ ‘traveller’
	PhE acrolect: [ˈtræv.lər]
	Phe mesolect: [ˈtræv.lər] ~[ˈtra.ve.ler]
	PhE basilect: [ˈtra.ve.ler]

Consonantal variability, measured by ΔC, highlights the role of syllable complexity. The PhE basilect simplifies coda clusters, such as reducing /st/ to /s/ (Tayao, 2004), resulting in a lower ΔC. In contrast, the PhE acrolect retains full coda clusters (e.g., [læst]), contributing to moderate consonantal variability, as reflected in its ΔC value of 5.24. While PhE acrolect exhibits more consonantal complexity than the PhE mesolect, it does not reach the same level of variability observed in the PhE basilect.

(2)	Realization for / læst/ ‘last’
	PhE acrolect: [læst]
	Phe mesolect: [læst] ~[las] (also occasionally [last] or [læs], reflecting variability)
	PhE basilect: [las]

3.1.2. Vocalic Rhythm Metrics: nPVI and rPVI

The normalized pairwise variability index (nPVI) measures vowel duration variability, with low values aligning with syllable-timed rhythms. In the PhE basilect, epenthesis processes, such as adding /ɪ/ to sC clusters (Magpale & Hong, 2024), as seen in (3), contribute to consistent vowel durations and low nPVI values. These findings mirror syllable-timed languages like Spanish and Tagalog, where vowel reduction is rare, and syllable structures are simpler. In contrast, the PhE acrolect, with its higher nPVI, reflects the irregular timing of stress-timed rhythms influenced by reduced vowels. Pike (1945) attributes these rhythmic patterns to the underlying phonological structure of stress-timed languages, which feature more complex syllables. On the other hand, PhE mesolect presents a mixed pattern, which results in nPVI values that do not fully align with either extreme.

(3)	Realizations for the /strɑŋ.gər/ ‘stronger’
	PhE acrolect: [strɑŋ.gər]
	Phe mesolect: [strɑŋ.gər] ~ [ɪs.troŋ.ger]
	PhE basilect: [ɪs.troŋ.ger]

Similarly, the raw pairwise variability index (rPVI) highlights consonantal variability. The PhE basilect simplifies consonantal clusters, such as reducing /rθ/ to /rt/ (e.g., [nɔrt]), resulting in smoother, more uniform timing and lower rPVI values. In contrast, the PhE acrolect retains full clusters (e.g., [nɔrtθ]), contributing to greater consonantal variability and a higher rPVI. PhE mesolect displays variable rPVI values, as some speakers simplify clusters, while others retain full clusters. This variation is evident in the standard deviation, reflecting the influence of both English and local phonotactic structures on mesolectal rhythm.

(4)	Realizations for / nɔrtθ/ ‘north’
	PhE acrolect: [nɔrtθ]
	Phe mesolect: [nɔrtθ] ~ [nɔrt]
	PhE basilect: [nɔrt]

The findings confirm the sociolectal stratification of PhE rhythm, with the PhE acrolect aligning closely with stress-timed norms and the PhE basilect reflecting syllable-timed influences from local Philippine languages. The PhE mesolect occupies a transitional position, blending features from both ends of the rhythmic spectrum. The findings also build on Tayao’s (2004) analysis of local phonology’s influence on PhE basilect rhythms, providing quantitative validation through higher %V and lower nPVI values. It also expands upon Tayao’s (2008) and Magpale and Hong’s (2024) work by confirming the quantitative impact of epenthesis and cluster simplification on PhE rhythm. Consistent with Dauer’s (1983) view that rhythm is shaped by structural influences rather than rigid classifications, this study positions PhE within a dynamic continuum, emphasizing the interplay between linguistic structure and sociolinguistic identity.

3.1.3. Results of Statistical Analysis

The previously discussed descriptive findings reveal a clear sociolectal gradient in PhE rhythm, with %V and nPVI distinguishing stress-timed and syllable-timed tendencies, while rPVI captures consonantal variability. To assess the statistical significance of these patterns, a one-way ANOVA was conducted, followed by Tukey HSD post hoc tests to identify specific pairwise differences among sociolects.

The ANOVA results (

Table 6. a. Summary of one-way ANOVA results for speech rhythm metrics across sociolects. b. Tukey post hoc comparisons of speech rhythm metrics across PhE sociolects ¹.

(a)
Metric	Sum of Squares (Between Groups)	df (Between Groups)	Mean Square (Between Groups)	F-Value	p-Value	Significance
%V	68.59	2	34.3	10.92	0.000335	***
ΔV	0.73	2	0.364	0.073	0.93	n.s.
ΔC	1.85	2	0.0927	0.551	0.583	n.s.
nPVI	2886.6	2	1443.3	178.2	2.78 × 10−16	*** (p < 0.001)
rPVI	304.3	2	746.03	52.4	5.06 × 10−6	***
(b)
Metric	Comparison		Difference	Lower 95% CI	Upper 95% CI	p-Value
%V	PhE acrolect vs. PhE basilect		3.641	1.675966	5.606034	0.00026
	PhE acrolect vs. PhE mesolect		2.409	0.443966	4.374034	0.014
	PhE mesolect vs. PhE basilect		−1.232	−3.19703	0.733034	0.282
nPVI	PhE acrolect vs. PhE basilect		−23.91	−27.0658	−20.7552	<0.001
	PhE acrolect vs. PhE mesolect		−9.9	−13.0558	−6.7442	<0.001
	PhE mesolect vs. PhE basilect		14.01	10.85419	17.16581	<0.001
rPVI	PhE acrolect vs. PhE basilect		17.2	13.3	21.2	<0.001
	PhE acrolect vs. PhE mesolect		10	6.1	13.9	<0.001
	PhE mesolect vs. PhE basilect		−7.2	3.3	−11	<0.001

¹ The primary objective of this study was to assess systematic differences across sociolectal groups rather than to account for participant-specific random variability. Since the sample was relatively balanced and met the assumptions of homogeneity of variance, a fixed-effects approach using ANOVA provided a clear and interpretable statistical framework for comparing mean differences across groups. While mixed-effects models offer advantages in accounting for within-speaker variability, they were not deemed necessary given the study’s focus on group-level phonetic variation.

a) indicate significant sociolectal differences in %V, nPVI, and rPVI, while ΔV and ΔC did not show significant variation across groups. High F-values and low p-values for %V (F(2, 27) = 10.92, p = 0.0003), nPVI (F(2, 27) = 178.2, p < 0.001), and rPVI (F(2, 27) = 52.4, p < 0.001) confirm that these rhythmic measures effectively capture systematic phonological differences across the PhE sociolects. In contrast, the non-significant results for ΔV (F = 0.073, p = 0.93) and ΔC (F = 0.551, p = 0.583) suggest that these metrics are less sensitive to sociolectal variation in rhythm.

The Tukey post hoc comparisons (Table 6b) further clarify the patterns of rhythmic differentiation. For %V, significant differences were observed between the PhE acrolect and PhE basilect (p < 0.001), and between the PhE acrolect and PhE mesolect (p = 0.014), reinforcing the previously noted sociolectal gradient: PhE acrolect speech exhibits more stress-timed characteristics, while PhE mesolect and PhE basilect speech trend more toward syllable timing. However, the difference between the PhE mesolect and the PhE basilect was not statistically significant (p = 0.282), suggesting a closer alignment in vocalic timing between these two groups.

The metric nPVI yielded significant differences across all three pairwise comparisons (p < 0.001), confirming its reliability as a rhythm indicator that captures variation in vowel interval durations. As expected, speakers from the PhE acrolect demonstrated the highest nPVI values, reflecting greater variability associated with vowel reduction and stress timing. Speakers from the PhE basilect, on the other hand, had the lowest nPVI values, consistent with more uniform vowel timing, while PhE mesolect speakers fell in between.

ΔV and ΔC did not show significant sociolectal differences, corroborating earlier findings that these measures are influenced more by phonotactic and prosodic structure than by rhythmic typology per se (Grabe & Low, 2002). As such, they appear to be less effective for distinguishing rhythm patterns in the PhE context.

The rPVI results also revealed significant differences across all three sociolects (p < 0.001). Speakers from the PhE acrolect produced the highest rPVI values (M = 66.3), followed by PhE basilect (M = 56.3) and PhE mesolect (M = 49.1). These results suggest that rPVI is a robust measure of consonantal timing in PhE, with PhE mesolect again occupying a transitional position between the more rhythmically divergent acrolect and basilect groups. The consistent distinctions across all three sociolects further highlight the sensitivity of rPVI to phonotactic variability and sociophonetic stratification.

Taken together, these results highlight %V, nPVI, and rPVI as the most robust indicators of rhythmic variation in PhE. While ΔV and ΔC did not reach statistical significance, their limitations are consistent with prior research (Grabe & Low, 2002; White & Mattys, 2007). These findings support the presence of a sociolectal continuum in PhE rhythmic structure, with the PhE mesolect exhibiting blended features from both ends of the spectrum. Rather than aiming for uniformity across all rhythm metrics, the study interprets these patterns as reflecting emergent rhythmic stratification—triangulated across multiple quantitative indicators and aligned with qualitative observations of sociolect-specific phonological processes.

4. Discussion

This section examines the rhythm metrics of PhE sociolects—acrolect, mesolect, and basilect—while drawing comparisons with broader rhythmic typologies. The findings highlight a rhythmic continuum shaped by varying degrees of stress- and syllable-timed features, aligning with the frameworks of Ramus et al. (1999) and Grabe and Low (2002). By situating PhE within established classifications, this analysis deepens our understanding of how sociolectal rhythm reflects broader phonological and sociolinguistic patterns, reinforcing PhE’s place within World Englishes.

4.1. Metrics Based on Ramus et al. (1999): %V, ΔV, ΔC

Table 7. The results of Ramus et al. (1999) combined with the results of the current study on PhE sociolects (arranged according to the %V value).

Languages/Sociolects	%V (STD)	∆V (STD)	∆V (STD)
English	40.1 (5.4)	4.64 (1.25)	5.35 (1.63)
PhE (acrolect)	40.49 (1.9)	3.81 (2.05)	5.24 (1.10)
Polish	41.0 (3.4)	2.51 (0.67)	5.14 (1.18)
Dutch	42.3 (4.2)	4.23 (0.93)	5.33 (1.50)
PhE (mesolect)	42.9 (1.2)	4.03 (3.02)	4.64 (0.86)
French	43.6 (4.5)	3.78 (1.21)	4.39 (0.74)
Spanish	43.8 (4.0)	3.32 (1.00)	4.74 (0.85)
PhE (basilect)	44.13 (1.86)	3.65 (1.27)	4.85 (1.76)
Italian	45.2 (3.9)	4.00 (1.05)	4.81 (0.89)
Catalan	45.6 (5.4)	3.68 (1.44)	4.52 (0.86)
Japanese	53.1 (3.4)	4.02 (0.58)	3.56 (0.74)

presents the rhythm metrics %V, ΔV, and ΔC for PhE sociolects, following Ramus et al. (1999). These results highlight the rhythmic gradient in PhE, reflecting a shift from stress-timed to syllable-timed properties across sociolects.

4.1.1. %V (Percentage of Vocalic Intervals)

PhE acrolect (40.49, SD = 1.9) closely aligns with English (40.1, SD = 5.4), reinforcing its stress-timed tendencies. PhE mesolect (42.9, SD = 1.2) occupies an intermediate position, resembling Dutch and French, suggesting a mix of stress- and syllable-timed features. PhE basilect (44.13, SD = 1.86) is closest to syllable-timed languages like Spanish (43.8) and Italian (45.2), confirming its increased vocalic duration and reduced stress timing.

4.1.2. ΔV (Standard Deviation of Vocalic Intervals)

PhE acrolect (3.81, SD = 2.05) exhibited lower vowel duration variability, aligning with English (4.64) and Dutch (4.23), consistent with stress-timed patterns. PhE mesolect (4.03, SD = 3.02) shows slightly greater variability, reinforcing its transitional status. PhE basilect (3.65, SD = 1.27) exhibits the lowest variability, reflecting more stable vowel durations typical of syllable-timed rhythms.

4.1.3. ΔC (Standard Deviation of Consonantal Intervals)

PhE acrolect (5.24, SD = 1.10) aligns with English (5.35) and Dutch (5.33), maintaining stress-timed consonantal variability. PhE mesolect (4.64, SD = 0.86) shows reduced variability, further supporting its mixed rhythmic profile. TPhE basilect (4.85, SD = 1.76) exhibits the highest consonantal variability among the three sociolects, likely due to phonological processes, such as epenthesis and cluster simplification.

To visualize the values in Table 7, Figure 1, Figure 2 and Figure 3 from Ramus et al. (1999) illustrate the rhythmic properties of various languages, now with added data points for PhE sociolects. Ramus et al. (1999) found that plotting languages along %V and ΔC provides the best distinction among rhythm categories. In Figure 1 (%V vs. ΔC), PhE acrolect clusters with stress-timed languages like English and Dutch, showing high consonantal variability (ΔC) and a lower vocalic proportion (%V). PhE mesolect, positioned near Italian and Catalan, displays a balance between consonantal and vocalic features. PhE basilect aligns with syllable-timed languages, such as Spanish and Italian, exhibiting lower ΔC and higher %V, consistent with its syllable-timed characteristics. In Figure 2 (%V vs. ΔV), PhE acrolect again aligns with stress-timed languages, with moderate vocalic variability (ΔV) and a lower %V. PhE mesolect, bridging the acrolect and basilect, exhibits slightly higher ΔV and an intermediate %V, reinforcing its hybrid rhythm. PhE basilect clusters with syllable-timed languages, showing higher %V and stable ΔV, reflecting more uniform vowel durations. In Figure 3 (ΔC vs. ΔV), PhE acrolect shows high ΔC and moderate ΔV, grouping with stress-timed languages. PhE mesolect has a lower ΔC and slightly higher ΔV than acrolect, reinforcing its transitional status. PhE basilect aligns with syllable-timed languages, exhibiting lower ΔC and stable ΔV, indicating reduced consonantal complexity and rhythmic regularity. Again, these patterns illustrate a gradual shift from stress-timed to syllable-timed rhythm across PhE sociolects, with mesolect acting as a transitional variety.

4.2. Metrics Based on Grabe and Low (2002): nPVI and rPVI

This section examines the rhythm metrics of PhE sociolects using the Pairwise Variability Index (PVI), contextualizing their rhythmic tendencies within Grabe and Low’s (2002) typological framework.

4.2.1. nPVI (Normalized Pairwise Variability Index)

As shown in

Table 8. The results of Grabe and Low (2002) combined with the results of the current study on PhE sociolects (arranged according to the nPVI values).

Languages/Sociolects	nPVI	rPVI
Thai	65.8	56.5
Dutch	65.5	57.4
PhE (acrolect)	54.1	66.3
German	59.7	55.3
British English	57.2	64.1
Tamil	55.8	70.2
Malay	53.6	63.3
Singapore English	52.3	68.2
PhE (mesolect)	44.2	49.1
Greek	48.7	59.6
Welsh	48.2	54.7
Rumanian	46.9	47.6
Polish	46.6	79.1
Estonian	45.4	40
Catalan	44.6	67.8
French	43.5	50.4
Japanese	40.9	62.5
PhE (basilect)	30.2	56.3
Luxembourg	37.7	55.4
Spanish	29.7	57.7
Mandarin	27	52

, the nPVI, which measures vocalic interval variability, highlights distinct rhythmic patterns across PhE sociolects. PhE acrolect (nPVI = 54.1) aligns with stress-timed languages such as British English (57.2) and German (59.7), indicating features like vowel reduction and complex syllable structures. PhE mesolect (nPVI = 44.2) falls between stress- and syllable-timed languages, similar to French (43.5), reflecting a mix of rhythmic influences. The PhE basilect (nPVI = 30.2) closely matches Spanish (29.7) and Mandarin (27.0), indicating more uniform syllable structures and reduced vocalic variability.

4.2.2. rPVI (Raw Pairwise Variability Index)

The rPVI, which measures consonantal interval variability, further distinguishes PhE sociolects. The PhE acrolect (rPVI = 66.3) exceeds British English (64.1) and is comparable to Singapore English (68.2), reflecting heightened consonantal complexity. This elevated rPVI may be attributed to several factors. First, certain phonetic realizations in PhE acrolectal, such as the affricated production of /θ/ as [tθ], contribute to increased cluster complexity (see l. 574). Additionally, PhE acrolectal speakers—especially in formal reading contexts—may hyperarticulate consonants due to prescriptive norms in Philippine education that emphasize “clear” English enunciation (Tayao, 2004). Unlike British English, where consonant lenition or elision may occur in spontaneous speech, these features are less common in controlled PhE readings. Finally, the use of a different stimulus text (The North Wind and the Sun) compared to Grabe and Low’s corpus introduces variation in phonotactic density, potentially affecting rPVI outcomes. Together, these factors may explain the higher consonantal variability in acrolectal PhE without necessarily implying greater inherent phonological complexity.

The PhE mesolect (rPVI = 49.1), positioned between the acrolect and basilect, shows reduced variability, reinforcing its transitional status. The PhE basilect (rPVI = 56.3) approaches values found in syllable-timed languages like Spanish (57.7) but retains some variability characteristic of stress-timed rhythm.

To further illustrate the values in Table 8, Figure 4, adapted from Grabe and Low (2002) with the inclusion of PhE sociolects, visualizes vowel duration variability using nPVI and intervocalic nPVI. The addition of intervocalic nPVI provides deeper insight into rhythmic differences by capturing variability between successive vocalic intervals. PhE acrolect clusters with stress-timed languages like British English (BE), exhibiting high nPVI values and greater rhythmic variability. PhE mesolect occupies an intermediate position, blending stress- and syllable-timed features. PhE basilect, with low nPVI values, aligns with syllable-timed languages such as Spanish, reflecting more uniform vowel durations. This continuum of rhythmic patterns reflects the sociolinguistic stratification within PhE, with PhE acrolect influenced by global English norms, PhE basilect shaped by local linguistic features, and PhE mesolect functioning as a transitional variety. These findings highlight the role of language contact and sociolinguistic factors in shaping speech rhythm in PhE.

5. Conclusions

This study investigated how rhythm metrics encode sociolectal identity in Philippine English (PhE), focusing on the acrolect, mesolect, and basilect. The findings reveal a continuum of rhythmic tendencies from stress-timed to syllable-timed, with the PhE acrolect aligning most closely with General American English norms, the basilect reflecting influences from syllable-timed Philippine languages, and the mesolect occupying an intermediary position that blends the features of both typologies.

Statistical analyses confirmed that %V and nPVI showed a clear sociolectal gradient, with significant differences between the acrolect and basilect (p < 0.001), supporting their utility as primary indicators of rhythmic variation in PhE. These findings are consistent with those of Grabe and Low (2002), who noted that such metrics reflect not only rhythmic classification but also phonotactic and prosodic structuring. Additionally, rPVI significantly differentiated the mesolect from both acrolect and basilect (p < 0.001), highlighting the mesolect’s unique durational variability in consonantal intervals.

While not all rhythm metrics yielded statistically significant differences, the observed patterns across %V, nPVI, and rPVI provide meaningful insights into how rhythmic variation is socially distributed in PhE. This partial but consistent patterning aligns with prior studies (e.g., White & Mattys, 2007; Nolan & Asu, 2009) showing that rhythm metrics often reveal complementary facets of speech timing and are most informative when interpreted in combination.

Overall, this study underscores the role of sociolectal variation in shaping prosodic rhythm in PhE. The rhythmic differences observed across lectal levels reflect not only the interaction of stress- and syllable-timed features but also the broader dynamics of L1–L2 phonological influence. In this regard, the basilect shows a stronger imprint from L1 syllable-timed languages, while the acrolect and parts of the mesolect preserve stress-based timing features characteristic of native English phonology. These findings contribute to our understanding of rhythmic identity in postcolonial Englishes and offer implications for L2 English acquisition in multilingual settings, echoing insights from White and Mattys (2007) and Tortel and Hirst (2010) on how L1 phonology shapes the prosody of L2 English.

5.1. Limitations and Recommendations

While this study provides important insights into rhythmic variation and sociolectal stratification in PhE, it is limited by its focus on Manila-based speakers, which may not fully capture the geographic and linguistic diversity of PhE. Future research should expand participant representation to include regional varieties and speakers with different levels of English exposure to better account for intra-Philippine rhythmic variation. Additionally, while rhythm serves as a strong sociolinguistic marker, further analysis of other prosodic features such as pitch, intonation, and tempo could provide a more comprehensive understanding of how PhE speakers construct identity through speech rhythm.

Another methodological limitation concerns the use of a controlled reading passage (The North Wind and the Sun), which, while phonologically balanced and widely used in rhythm research, does not reflect spontaneous speech. This choice was intentional and aimed at maintaining consistency across participants and minimizing lexical and syntactic variation. The emergence of sociolectally patterned rhythm metrics—despite uniform textual input—suggests that the variation observed was not a product of the text, but of speaker-specific rhythmic structuring tied to sociolinguistic identity. Nonetheless, future studies could incorporate extemporaneous or interactional speech samples to test the persistence of rhythmic patterns across speech styles and contexts.

The sociolectal classification used in this study was based on self-reported English proficiency and the frequency of English use across social domains. While this method is common in sociolinguistic research, it lacks direct phonological validation. Future studies should consider triangulating self-report data with phonological or acoustic measures to improve classification accuracy and better capture the linguistic behaviors of each sociolect group. Furthermore, while the current study focused on linguistic indicators of identity, future research could benefit from integrating sociolinguistic factors, such as education level, occupation, and socioeconomic status. These variables may reveal deeper interactions between social positioning and prosodic variation, refining our understanding of sociolectal distinction in PhE.

Future research could also explore L1 transfer in L2 English, particularly how phonotactic features from Filipino or regional languages influence rhythm in L2 English learners. As Sönning (2023) suggests, rhythm metrics can be used to assess language proficiency and accent, distinguishing native English speakers (e.g., PhE acrolect) from non-native speakers (e.g., PhE basilect or other L2 English learners). Investigating these interactions further would help clarify how rhythm functions as a marker of linguistic identity and proficiency in multilingual contexts.

5.2. Implications for Future Research and Applications

The findings reinforce rhythm as a key sociolinguistic variable, providing new insights into the prosody of Southeast Asian Englishes. Future studies could also examine the influence of digital communication on prosodic variation, particularly how social media, online discourse, and text-to-speech technology impact spoken English rhythms in younger speakers.

By positioning PhE within a global linguistic framework, this study highlights the evolving nature of English in Southeast Asia and its theoretical and practical significance in understanding rhythm, identity, and the diversity of World Englishes. Incorporating rhythm metrics into L2 acquisition research will enhance our understanding of L1–L2 rhythm interactions, sociolinguistic factors, and rhythm-based language assessment in multilingual environments. These findings pave the way for future research on both native and non-native English varieties, particularly within diverse multilingual societies like the Philippines.