Asian J. Arts Cult. 2026; 26(2): 77

English has become a global lingua franca, serving as the primary medium of communication among speakers of different native languages. Adherents of the World Englishes paradigm advocate for the acceptance of diverse ways in which English can be used—such as patterns and structures used by speakers from various linguistic backgrounds—and reject the idealisation of native English speakers from the Inner Circle (Kachru, 1989). Several studies support the ideology of World Englishes that English should be learnt for successful communication and a variety of Englishes should be accepted (Boonsuk, 2025; Jeong & Lindemann, 2025; Pathanasin, 2025. Under the principle of World Englishes, the goal of pronunciation learning should be for clear communication, rather than to be native-like. Nevertheless, intelligibility is crucial for effective communication, and English pronunciation plays a vital role in achieving this. Mispronunciation can alter the intended meaning or make the message unintelligible. Poor pronunciation not only causes misunderstanding, but also creates anxiety and lowers confidence among language learners (Nguyen, 2023; Singhanuwananon, 2018).

English pronunciation poses a significant challenge for learners of English as a Foreign Language (EFL) worldwide. Due to differences between the phonological systems of their first Language (L1) and the target language, learners often struggle with pronunciation more than with other linguistic aspects, such as vocabulary and grammar (Yule, 2006). According to Yule (2006), L1 interferes with phonological performance in a speaker in three aspects. The first aspect involves differences in phonemic distinctions for sounds that are not present in both the L1 and English, as many EFL learners struggle to differentiate these sounds. The second aspect concerns differences in phonological rules between L1 and English. Learners tend to apply their L1 phonological rules to English pronunciation. Finally, L1 transfer becomes evident when learners impose the phonotactic constraints of their L1.

Thai is the only national language in Thailand, while English is the most important foreign language. Thai and English belong to different language families, as do their phonological systems (Timyam, 2015). The distinct phonemic inventories are seen to account for English pronunciation problems in Thai learners. Several studies have investigated the causes of pronunciation difficulties among Thai EFL learners (Atthaphonphiphat, 2017; Kanokpermpoon, 2007; Peerachachayanee, 2022; Suntornwawet, 2019). Results of previous studies have identified L1 interference and phonological differences between the two languages as the primary causes of the pronunciation problems in Thai learners (e.g., Jaiprasong & Pongpairoj, 2020; Sahatsathatsana, 2017; Thongsongsee & Watanapokakul, 2023). Although much research has examined pronunciation issues among Thai EFL learners, more focused studies on learners from different linguistic backgrounds could provide educators with deeper insights into their challenges and help design more effective lessons to enhance their performance. The present study specifically targets a group of learners with a linguistic background distinct from that of EFL learners in other regions of Thailand. It is hypothesised that while the participants share common pronunciation errors with other Thai EFL learners, their linguistic background contributes to unique consonant variations not observed in previous studies. This study aims to examine the pronunciation difficulties related to word-final consonant sounds among university students in southern Thailand who use the Southern Thai dialect in daily communication. Specifically, this study addresses two research questions:

1. What are the patterns and challenges associated with English word-final consonant pronunciation among Southern Thai EFL learners?

2. What factors contribute to the pronunciation difficulties of Thai learners with the Southern Thai dialect in their linguistic background?

Literature review

This section provides the background knowledge underpinning the research design, with a focus on phonological distinctions between English and Thai, as well as on the sound systems of Central and Southern Thai. The phonemes or abstract sounds are represented in slant brackets / /, and allophones or the actual pronunciation are shown in square brackets [ ].

Phonological distinctions between English and Thai

Several linguists have examined differences between the English and Thai phonological systems. Since this study focuses on the pronunciation of English word-final consonants, this section highlights the key distinctions between the two languages. English and Thai differ notably in their consonant inventory sizes. English has 24 consonant phonemes, while Thai has 21. The phonological inventories of English and Thai are presented in Table 1, based on Raktham (2018); Ronnakiat and Chitwiriyanon (2022); and Tuaycharoen (1990). The nine English consonant phonemes not found in Thai are highlighted in bold in Table 1.

Based on the manner of articulation, the English consonant sounds are categorised into: stops /p, b, t, d, k, ɡ/, nasals /m, n, ŋ/, fricatives /f, v, θ, ð, s, z, ʃ, ʒ, h/, affricates /tʃ, ʤ/, approximants /r, w, j/, and lateral /l/. Among these sounds, there are nine voiceless consonants: /p, t, k, f, θ, s, ʃ, h, tʃ/ and fifteen voiced consonant phonemes: /b, d, ɡ, m, n, ŋ, v, ð, z, ʒ, ʤ, l, r, w, j/.

However, the Thai’s consonant system features a smaller number of voiced and voiceless distinctions, relying more heavily on aspiration for contrast. Thai has nine stop consonants, including unaspirated voiceless stops: /p, t, k/, aspirated voiceless stops: /pʰ, tʰ, kʰ/; voiced stops: /b, d/; and a voiceless glottal stop /ʔ/. There are only three fricatives, and all of them are voiceless: /f, s, h/, and two voiceless affricates: /ts, tsʰ/. The remaining phonemes are all voiced sounds—nasals: /m, n, ŋ/; approximants /w, j/, trill /r/, and lateral /l/.

We can see that one of the key differences between English and Thai consonant inventories is the aspiration in Thai. Thai has a more obvious distinction between aspirated and unaspirated stops as well as affricates, which are not prominent in English. On the contrary, English makes use of a larger variety of voiced/voiceless contrasts which distinguish between the eight pairs of phonemes: p/b; t/d; k/ɡ; f/v; s/z; θ/ð; ʃ/ʒ; and tʃ/ʤ. Furthermore, nine English consonant phonemes are not found in Thai: /ɡ, v, θ, ð, z, ʃ, ʒ, tʃ, ʤ/. The phoneme /r/ is realised differently in English and Thai. To be specific, it is pronounced as an alveolar approximant [ɹ] in English, whereas in Thai, it corresponds to an alveolar trill [r]. According to Ashby (2005), in rhotic accents (e.g., American English), the /r/ is pronounced in all positions of words. In contrast, in non-rhotic varieties (e.g., British English), the final /r/ is not pronounced unless followed by a vowel.

Table 1 Comparison of English and Thai consonant inventories

Final consonant sounds in English and Thai

Almost all English consonant phonemes can occur in word-final position, except for /h/, /w/, and /j/. In contrast, only nine Thai consonant sounds can appear at the end of words: stops /p, t, k, ʔ/, nasals /m, n, ŋ/, and approximants /w, j/.The phonological differences between English and Thai tend to present significant challenges for Thai EFL learners. As noted by Kanokpermpoon (2007), some English phonemes are absent from Thai or occur in different distributions. The restrictive nature of the Thai syllable structure also contributes to pronunciation difficulties. While English allows a broad range of final consonant sounds, Thai permits only nine (Kanokpermpoon, 2007; Peerachachayanee, 2022). This limitation often results in common errors for Thai learners, such as substitutions, omissions, and simplifications when Thai speakers encounter unfamiliar final sounds in English (Narksompong, 2007).

Another common issue among Thai EFL learners is the devoicing of final voiced stops /b, d, ɡ/. These are frequently replaced with their voiceless counterparts /p, t, k/, which can appear as final sounds in Thai. In addition to devoiced sounds, Thai learners often pronounce these sounds as unreleased. For instance, the word good is often pronounced as [ɡʊt ], and big as [bɪk ] (Narksompong, 2007; Peerachachayanee, 2022). Similarly, English final fricatives—completely absent in Thai—also pose a major challenge. Learners often substitute them with stops or delete them. Words, such as leave and breathe are commonly pronounced as [liːp ] and [briːt ], while please becomes [pliːt ]. These patterns suggest that Thai learners tend to approximate unfamiliar phonemes with the closest sounds available in their L1 (Atthaphonphiphat, 2017; Kanokpermpoon, 2007; Suntornsawet, 2019).

In addition to substitutions, Thai speakers often omit final consonants, especially fricatives and affricates which are absent in Thai phonology. The lack of final affricates aggravates this issue, leading to simplification strategies where they tend to replace them with stop sounds. For instance, church may be realized as [cʰɤːt ], and judge as [ʤʌt ] (Kanokpermpoon, 2007; Peerachachayanee, 2022; Suntornsawet, 2019). These phonological changes significantly affect intelligibility, as English listeners often rely on final sounds to distinguish words.

Moreover, Thai learners frequently modify or omit final /l/ and /r/ when speaking English. The final /l/ is often substituted with /n/, or /w/, producing words like call→[kʰɔːn] and hotel→[hoːtʰew] (Kanokpermpoon, 2007; Peerachachayanee, 2022). Likewise, the final /r/ is frequently dropped, resulting in car being realised as [k^haː] (Atthaphonphiphat, 2017; Suntornsawet, 2019).

Phonological variation between Central Thai and Southern Thai

Thailand has several major regional dialects, mainly northern, northeastern, southern and suburban dialects in the central area. These regional variations have also highlighted pronunciation differences among Thai EFL learners. Standard Central Thai serves as the official national language. Southern Thai is spoken south of Chumphon Province and extends into adjacent areas in northern Malaysia (Hartmann & Hudak, 2000). Several researchers have documented the phonological features of Central Thai and Southern Thai (e.g., Abramson & Tingsabadh, 1999; Awirutthiyothin, 2010; Hartmann & Hudak, 2000; Kamalanavin, 2005; Phonyarit, 2025; Rose, 1997), as summarised in Table 2.

Table 2 Comparison of Central Thai and Southern Thai phonological features

Although final consonants in Standard Thai are typically pronounced without an audible release (Abramson & Tingsabadh, 1999) and limited to nine sounds: /p, t, k, m, n, ŋ, j, w, /, the pronunciation of final consonants in Southern Thai is significantly different (Kamalanavin, 2005). These variations, which are observed in consonant deletion, modified articulation, and clarity, affect the intelligibility across dialects in spite of the fact that the consonant inventory, as a whole, is similar. Southern Thai speakers tend to drop final stops and fricatives, creating open syllables whereas Standard Thai speakers often retain these sounds, producing clearer final consonant articulation. Additionally, the two varieties are further differentiated by articulation differences. Final /p, t, k/ in Standard Thai are normally unreleased, thus producing a perceptible closure, but in Southern Thai, they are normally softened or omitted (Kamalanavin, 2005). This tendency to use the open syllable structure may also lead to the difficulty in pronouncing English words with a closed syllable structure. Even final nasals /m, n, ŋ / are also more likely to be longer in Southern Thai, affecting speech rhythm and overall prosody. Such differences do not merely represent dialectal differences, but also provide insight into the L1 transfer patterns in Southern Thai EFL learners.

As presented in Table 2, other phonological features, including vowels and tones, also show minor differences. Generally, the vowel sounds of Central and Southern Thai are identical, however the tonal systems are slightly different. While Central Thai has only five tones, Southern Thai uses seven tones. While a recent study by Phonyarit (2025) identified nine major regional dialects within Southern Thai, each with minor phonetic variations, the study also confirmed previous findings that the Southern Thai tonal system consists of seven contrastive tones. These differences in tone use, though not directly linked to final consonant pronunciation, contribute to the unique rhythm and prosody. This prosodic variation may have an indirect effect on English pronunciation, especially word stress placement and word-final articulation. Hence, although the vowel and tone systems are relatively comparable across dialects, the most significant variations lie in consonant production, particularly in word-final position. This may significantly contribute to the pronunciation challenges encountered by Southern Thai learners.

Challenges in English pronunciation among Thai EFL learners

Several studies have examined pronunciation difficulties among Thai EFL learners, often attributing errors to phonological differences between English and Thai. For instance, Makamthong and Hesmatantya (2022) found that students substituted Thai sounds when pronouncing the fricatives /θ/ and /ð/, and the affricate /ʤ/, with substitutions varying by word position. Consistent with prior findings, Plailek and Essien (2021) also found that Thai undergraduates struggled with the pronunciation of word-final consonants, including /θ/, /ð/, /v/, /f/, /s/, and /l/. Similarly, Muangphruek (2017) observed that learners frequently omitted or substituted final sounds, especially those absent in Thai, such as /θ/, and struggled with consonant clusters like /st/ and /rt/. Chakma (2014) also reported that Thai high school students substituted Thai sounds for problematic English consonants in both initial and final positions. In addition, Ruengkul (2020) explored the pronunciation of -ed endings with a group of 33 university students. In her study, the participants read 15 words that ended in -ed sounds, pronounced as /Id/, /d/, and /t/. It was found that they performed the best when words ended in /Id/ sounds, whereas /d/ and /t/ ranked second and third, respectively.

Given such persistent challenges, Sridahnyarat (2017) further investigated the acquisition of English fricatives in Thai undergraduates grouped by English proficiency (high, intermediate, low) using three speaking tasks: word list, sentence reading, and interviews. The findings indicated that advanced learners could accurately produce /s/, /f/ and /ʃ/ word-finally and word-initially. During word list reading, intermediate learners had difficulty with final /ʃ/, while low-proficiency learners were unable to pronounce /s/, /f/, and /ʃ/ in final positions in either word list or sentence reading tasks. These findings reinforce the role of phonological interference in Thai EFL learners.

Beyond phonological influences, several studies have explored factors accounting for pronunciation accuracy. For example, Sahatsathatsana’s (2017) study not only L1 interference but also teaching methods, and learner motivation as key issues. Furthermore, Plailek and Essien (2021) found that the foundational knowledge of English pronunciation was the most significant factor. Teacher instruction quality ranked as the second, followed by the frequency of English pronunciation practice. Additionally, the phonological differences between Thai and English sound systems were also highlighted as a major source of difficulty.

Complementing these findings, comparative phonological studies (Atthaphonphiphat, 2017; Kanokpermpoon, 2007; Makamthong & Hesmatantya, 2022; Narksompong, 2007; Peerachachayanee, 2022; Suntornsawet, 2019) consistently indicate that L1 interference is a key factor contributing to pronunciation errors among Thai EFL learners. These errors are primarily attributed to structural differences in phonotactics, syllable structure, and consonant distribution between English and Thai.

Collectively, these studies reveal not only the consistent impact of L1 transfer on the pronunciation of English consonant sounds but also the role of learner proficiency, establishing a foundation for further exploration into how different linguistic background may affect English pronunciation patterns of Thai learners.

In terms of English pronunciation challenges, Jehma and Phoocharoensil (2014) analysed pronunciation difficulties in Southern Thai EFL learners. Their participants are fifth and sixth grade Pattani-Malay students from Yala Province. The results showed that the participants struggled more with English fricatives than with stops across all word positions, with errors occurring in 32.90% of fricatives, compared to only 4.50% of stops in final positions. Similarly, focusing on final sounds, Atthaphonphiphat (2017) conducted a study with 350 university students in a southern Thai province, and reported that learners pronounced English final consonants correctly in less than 50% of cases. Among the target sounds, /s/ showed the highest accuracy rate (47.40%), while /ð/ was the least accurately pronounced (0.70%). The findings also supported other previous studies that phonological transfer further influences the realisation of English word-final consonants.

In summary, previous research affirms that Thai EFL learners have difficulty with pronunciation due to L1 transfer and instructional limitations. While most studies have examined Thai learners in central regions (Muangphruek, 2017; Peerachachayanee, 2022; Plailek & Essien, 2021; Ruengkul, 2020), limited research has focused specifically on pronunciation difficulties among Southern Thai EFL learners who speak Southern Thai dialect. This study addresses that gap by analysing their pronunciation patterns, offering insights for educators and curriculum developers to design targeted instruction that reflects these learners’ distinctive linguistic backgrounds.

Methodology

Participants

The study involved 27 second-year undergraduate students who were enrolled in an English pronunciation course (19 males and 8 females, aged 19-22), from a university in southern Thailand. Their majors are Environmental Technology and Management, and their English proficiency were primarily at the A2 CEFR level. A purposive sampling method was employed based on two criteria: (1) participants had lived and attended school in provinces across southern Thailand, and (2) they regularly spoke Southern Thai in daily life. Although most of them also communicated in Central Thai, Southern Thai remained their primary spoken variety.

Research instruments

This study employed two research tools: a pronunciation analysis task and a semi-structured interview. For the pronunciation analysis, the researchers compiled a list of 21 words, each ending in a distinct English consonant sound that can occur in the word-final position. These words were selected from a first-year English textbook, assuming participants’ prior exposure and familiarity.

The interviews adapted questions from Thongsongsee and Watanapokakul’s (2023) study on word stress patterns. Using a semi-structured format, the interview combined both structured and open-ended questions to explore the learners’ pronunciation challenges. The interview focused on four key aspects: pronunciation awareness, previous instruction, learners’ approaches to unfamiliar word pronunciation, and strategies used for pronunciation improvement.

To ensure content validity and reliability, three English lecturers—each with over five years of teaching experience—evaluated the interview questions using the Index of Item-Objective Congruence (IOC). The evaluation yielded a high agreement score of 0.89. Based on their feedback, necessary revisions were made. Prior to data collection, the study received ethical approval to verify its adherence to the principles outlined in the Declaration of Helsinki and the Belmont Report, ensuring participants’ rights, research transparency, and ethical integrity.

Data collection

Participation in the study was voluntary, and all participants were fully informed about the study’s objectives, procedures, and data confidentiality protocols. They were assured that participation would not affect their course grades or studies. After addressing any concerns, all participants signed a written consent.

For the pronunciation analysis, the first researcher, a native English speaker and instructor of the pronunciation course, gave a list of 21 words to the participants and instructed them individually to read aloud the words to him. If the participants wanted to repeat any words, they were allowed to do so, and the last version they produced would be used for data analysis. After reading, their pronunciation was transcribed using the International Phonetic Alphabet (IPA). Cases of omission were marked with the symbol ø, and additional phonetic notes were recorded as needed. To enhance transcription reliability, all sessions were video-recorded, allowing for later verification. Each session lasted 2-3 minutes. Immediately following the pronunciation task, participants were individually interviewed in Thai by the second researcher in semi-structured sessions lasting 5-10 minutes. All interviews were audio-recorded for subsequent analysis.

Data analysis

The pronunciation data were transcribed in real time and then reviewed using video recordings for transcription accuracy. The third researcher, trained in English phonetics, independently verified the transcriptions. The final data set were entered into Microsoft Excel spreadsheet for statistical analysis. Accuracy rates and common substitutions were calculated, and frequent substitution patterns were identified.

Interview data were analysed thematically, following Braun and Clarke’s (2006) guidelines. Audio recordings were transcribed verbatim, translated into English, and thoroughly reviewed for accuracy. The transcripts were then systematically coded and organised into themes. To enhance the reliability of interpretation, an expert in qualitative research assessed the data for accuracy of the coding to ensure consistency and thematic validity.

Results

To investigate the pronunciation challenges of English word-final consonant sounds among Southern Thai EFL learners and the factors influencing these difficulties, the data were analysed both quantitatively and qualitatively. The findings are presented in two sections: pronunciation accuracy and contributing factors.

Figure 1 illustrates the significant variation in pronunciation accuracy across phonemes, with learners exhibiting greater difficulty with voiced consonants. The nasal consonants /m/, /n/, and /ŋ/ were articulated with 100% accuracy, followed by the voiceless velar stop /k/ (96.3%) and the voiceless alveolar fricative /s/ (92.6%), suggesting these sounds were produced with relative ease. In contrast, the voiced sounds—bilabial stop /b/, velar stop /ɡ/, interdental fricative /ð/, and post-alveolar fricative /ʒ/—all recorded 0% accuracy, indicating considerable challenges in their production. Similarly, the alveolar stop /d/, labiodental fricative /v/, alveolar fricative /z/, and post-alveolar affricate /ʤ/ showed very low accuracy (3.7%). Notably, the approximant /r/ in word-final position was often omitted, which may be influenced by British English pronunciation patterns.

The data in Table 3 highlight the relative difficulty levels of word-final consonants in English, with a specific focus on sound substitutions. The cases of omission are marked with the symbol ø. It should be noted that the data excluded mispronunciations that resulted in unintelligible words. The following section categorises the results according to manner of articulation to better illustrate phonological patterns in learner errors.

Table 3 Sound substitutions of English word-final consonants by order of difficulty

Stops

Among the six stop sounds, voiced stops exhibited more difficulties. No learners accurately pronounced the final /b/, substituting it primarily with the voiceless counterpart [p] (81.5%) and [f] (7.4%), while 11.1% of learners omitted the [b] entirely. Similarly, for the final /ɡ/, no learner produced it as a voiced sound; all substituted [ɡ] in drug with the voiceless [k]. The final /d/ also proved particularly challenging, with only one learner (3.7%) producing it accurately. Most learners omitted the [d] (70.4%), while some substituted it with [t] (22.2%) or [s] (3.7%).

In contrast, voiceless stops were generally produced with relatively high accuracy in word-final positions, compared to other consonant types. Among these, /t/ had an accuracy rate of 63.0%, with omissions in 25.9% of cases and substitutions with [s] (11.1%). The final /p/ was the second most accurately articulated by 77.8% of learners; inaccuracies involved omission (11.1%) and substitutions with [t] (7.4%) and [f] (3.7%). The most accurately articulated stop was /k/, with a correct production rate of 96.3%, and one learner substituted it with [t] (3.7%).

Nasals

Nasal sounds /m, n, ŋ/ did not pose any difficulty for the Southern Thai EFL learners. All participants pronounced them accurately, with no instances of substitution or omission.

Fricatives

Among the eight final fricatives analysed, the voiced sounds /ð/, /ʒ/, /v/, and /z/ appeared to be more problematic for the learners than the voiceless phonemes /ʃ/, /θ/, and /f/. The voiceless alveolar fricative /s/ was the least challenging, with 92.6% of learners producing it accurately.

For the final /ð/, most learners (77.8%) substituted it with [t], followed by [s] (14.8%), while two learners (7.4%) replaced it with the voiceless counterpart [θ]. Regarding the final /ʒ/, most learners substituted it with [tʃ] (77.8%), and the rest with [t] (22.2%).

The next most challenging sound, /v/, showed a very low accuracy rate of 3.7%. A majority of learners (63.0%) substituted it with its voiceless counterpart [f], while 33.3% used [p] as a replacement. Similarly, the final /z/ posed significant difficulty, with only 3.7% pronouncing it correctly. Most substituted it with the voiceless [s] (81.5%), while others used [t] and [tʃ].

The subsequent three problematic sounds were /ʃ/, /θ/, and /f/. For /ʃ/, only 14.8% of learners produced it correctly. Common substitutions included [tʃ] (66.7%), [t] (11.1%), and [s] (7.4%), indicating a tendency to replace fricatives with affricates or stops. The voiceless interdental fricative /θ/ was accurately produced by 18.5% of learners, while 66.7% substituted it with [t] and others with [s] or [ʃ].

The voiceless /f/ in tough was the least challenging of the three, with 22.2% of learners producing it accurately. Some substituted it with [p] (7.4%) or [t] (3.7%). A total of 66.7% mispronounced the word tough, producing errors, such as tongue, dough, touch, too, talk, and taught. Regarding the final /s/, this least problematic final fricative was produced accurately by 92.6%of learners. The incorrect cases involved substitution with [t] (7.4%).

Affricates

Between the two affricates, the voiced /ʤ/ was significantly more problematic than its voiceless counterpart /tʃ/, which has a close equivalent in Thai. Only 3.7% of learners pronounced /ʤ/ correctly. The majority substituted it with [tʃ] (63.0%), while others used [t] (18.5%). In contrast, /tʃ/ was accurately produced by 85.2% of learners. Some errors occurred, with 7.4% substituting it with [t], and a few with [ʃ] or [s].

Approximant

The approximant /r/, as in discover, was not pronounced in any instance. Some learners (29.6%) followed the British non-rhotic pronunciation, producing a schwa [ə], while the remaining 70.4% realised it as a long vowel [ɜː].

Lateral

The lateral /l/ in natural was accurately produced by 40.7% of learners. The majority (59.3%) substituted it with the approximant [w], which cannot occur word-finally in English.

Overall, the most problematic word-final sounds for the learners were /b/, /ɡ/, /ð/, and /ʒ/, which were not produced accurately by any participants. These were followed by /d/, /v/, /z/, and /ʤ/, all of which showed minimal accuracy. The data indicate that voiced stops, voiced fricatives, and the voiced affricate posed the greatest difficulty, while voiceless stops and nasal consonants were produced with greater ease.

The findings in Table 4 illustrate the patterns of errors in the pronunciation of English word-final consonants among Southern Thai EFL learners. These errors are categorised into three primary types: substitution, devoicing, and omission.

The following sections provide a detailed analysis of each error type identified in Table 4.

Table 4 Patterns of English word-final consonants errors in Southern Thai EFL learners