Mispronunciation of High Front and Low Hausa Vowels among the Yorùbá Speakers

_____________________________________________________________________________________________________ Pronunciation in second language learning is sometimes challenging, especially the vowels. Vowels such as [i] and [a] are found both in Hausa and Yorùbá but [i:] and [a:] are peculiar to Hausa alone. While Hausa has short and long vowels, Yorùbá has only oral and nasal vowels in their vowel inventories. Such phonemic differences constitute learning challenges, especially for the Yorùbá native speakers. This is a cross-sectional study design using mixed methods to examines the production of high front vowels: [i], and [i:], as well as low: [a], and [a:] Hausa vowels by the Yorùbá speakers to identify which group perform better between group 1 (Yorùbá native speakers who learned Hausa in the secondary school before going to the college of education), and group 2 (Yorùbá native speakers who learned Hausa informally before going to the college of education). The study also seeks to find out vowel substitutions that occur in the pronunciation tasks using 80 participants from 18 years old and above from the College of Education system in Nigeria who were selected based on purposive sampling. The findings were discussed in line with Flege & Bohn’s (2020) ‘Revised Speech Learning Model’. 8 stimuli were audio-recorded, transcribed, and rated by two independent raters, in addition to participant observation techniques adapted. The results of the Mann-Whitney test revealed that group 2 performed better than group 1. The study discovered also that the short [a] in the first and second syllables had the highest frequency of substitution compared to [i], [i:] and [a:] vowels. Such problems have pedagogical implications for learning Hausa as a second language.

, the standard Yorùbá on the other hand, has 7 oral vowel phonemes: [i], [ (Eme & Uba, 2016;Maikanti, Thai, Burkhardt, Fung, Husain, & Olúwadọro, 2021). Hausa and Yorùbá shared five (5) oral monophthongs phones: [i], [e], [a], [o], [u]. While Yorùbá has two extra vowels which are [ɛ] and [ɔ], Hausa on the other hand, has two diphthongs, [ai] and [au], which Yorùbá does not possess. Moreover, Hausa has a phonemic distinction between short versus long among these five vowels whereas vowel length differences are only phonetic in Yorùbá. The latter has five nasal vowel phonemes [ĩ], [ɛ], [ã], [ɔ], [ũ], which Hausa does not possess. The phonemic contrasts between Hausa and Yorùbá made it difficult for the Yorùbá native speaker to produce the Hausa vowels correctly. The purpose of this research was to look into the production of [i], [i:], [a], and [a:] Hausa vowels in the pronunciation of disyllabic Hausa words in the first and second syllables among the Yorùbá native speakers. The study compares the performance of group 1-Yorùbá native speakers who learned Hausa formally in the secondary school before going to the college of education, and group 2-Yorùbá native speakers who learned Hausa informally before going to the college of education to identify which group perform better in pronunciation.
Although there are numerous theories on second language learning (e.g., Lado's (1957) Contrastive Analysis Hypothesis; Corder's (1967) Error Analysis Model; Best's (1994Best's ( , 1995; Flege's (1995); Best & Tyler's (2007) Perceptual Assimilation Model for Second Language; and Flege & Bohn's (2020) Revised Speech Learning Model), the postulation of Flege & Bohn's (2020) 'SLM-r' best predicts the outcome of the present research. While theories such as 'PAM'- Best (1994Best ( , 1995, 'PAM-L2'- Best & Tyler (2007) centre on the perception of native and non-native speech sounds, they also discuss the types and causes of errors in L2 learning. The 'SLM-r' therefore, argues that second language learners have more problems with the sounds that are shared between the L1 and L2, while the unshared sounds are easy for the L2 learners to produce. The fact that the present study is on production task, the postulations of 'SLM-r' on how some participants could pronounce the front high and low Hausa vowels correctly, while some participants mispronounce the vowels, appeared to be the best option for this research. The findings of this study contributes to the formal and informal teaching and learning of Hausa as a second language, particularly in the Nigerian school system. The non-Hausa speakers (such as Nupe, Gwari, Fulfulde, and Yorùbá) who share boundaries with Hausa land would also be encouraged to learn Hausa as a second language. This is in addition to assisting the learners to understand how to read, write, and speak Hausa fluently for effective communication.

Methods
Since the present study is a cross-sectional design whereby one of the researchers measure the outcome and the exposure of the participants using mixed methods of data analysis, Babbie (1989) regarded this type of research as a 'one-shot' study. One of the aims of this type of design is to find out issues relating to the prevalence of a phenomenon, situation, problem, attitude or issue, by taking a cross-section of the population at one time (89). With this, Almalki (2016), Sileyew (2019) add that this type of approach allows the researcher to make detailed explanation concerning second language learning phenomena. 80 participants (40 students from each group) aged 18 years old and above who were learning from colleges education in the South-West of Nigeria were recruited based on purposive sampling. This figure represents the whole population of the students in the five schools (FCE Osiele-Abeokuta, ACE Ondo, COE Ikere-Ekiti, FCE (Special) Oyo, and COE Oro) involved in the present research, which Creswell (2014) also suggests the use of all the population as a sample size, especially when the population in research is so small. Holton & Bernett (1997) equally add that the ability to use a small number of participants to make inferences about larger groups is one of the advantages of the quantitative method of research. As such, all the participants were also screened to ensure that only the right participants (Yorùbá native speakers) were recruited. The study location consists of the following states in the southwest of Nigeria: Ekiti, Ondo, Oyo, Ogun, and the Kwara States. Western education in Nigeria has a long history. However, the 6-3-3-4 system of education in Nigeria which started in 1983 required the students to attend the following steps of education: 6-year primary education, 3-year junior as well as 3-year senior secondary education, and finally, the 4-year tertiary education. In all three stages of schooling, provisions were made for second language learning (Igbokwe, 2015;Uwaifo & Uddin, 2009). Students who attend colleges of education were expected to among others, spend 3 years to acquire the minimum teaching qualification of National Certificate in Education (NCE) to qualify them to teach in either primary or secondary school, in the absence of L1 Hausa teachers (Nigeria Certificate in Education Minimum Standards for Languages, 2020; National Policy on Education, 2004). In this study, participants in group 1 were categorised as students who learned Hausa in secondary school but admitted to the college of education to further learn Hausa as a second language. Participants in group 2 were categorised as students who only acquired Hausa on the street, and they were also admitted to the college of education to study Hausa as a second language. Ethical approval was granted by the relevant authority before the data collection in line with Leedy & Ormrod (2005). Since participation in the present study was voluntary, participants were informed in writing that the data they provided was for research, and they had the right to withdraw from the study at any time without any implication. This is in addition to the assurance given to them that all their identities, including the information they provided, would be kept confidential. They were also informed that data collected from them would be kept in retrieval storage. Those who agreed to participate in the study were made to sign a consent form before they participated in the research.
Eight wordlists comprised of target and non-target vowels prepared in carrier phrases were used as research instruments for this study. The stimuli used in the present research with CV.CVV and CVV.CV syllable patterns account for each vowel in the first and second syllables. For instance, while 'gídá' had a short [i] in the first syllable, 'gádì' had a similar short [i] in the second syllable. This is similar to where 'bítà' had a long [i:] in the first syllable, 'fârí' had a similar long [i:] in the second syllable.) Since the study investigates the mispronunciation of vowels in the first and second syllables of Hausa words, only disyllabic Hausa words were selected for this study. The reason for putting the wordlist in carrier phrases was to let the participants to read the stimuli as freely as they wanted in their normal pronunciation without recognising the exact items being studied by the researchers. Thus, they didn't have to make an extra effort with the pronunciation. Each vowel was also assigned a tone for easy identification while reading the stimuli, and the participants were given enough time to complete the pronunciation task. In line with Stella (1985), Tailor (1992) who recommended that speech recordings in research be conducted in a sound-proof to eliminate background noises, the speech sounds of the participants were audio-recorded by one of the researchers in the language laboratories. Scoring were done independently by the two Hausa native speakers and linguists, who awarded '1' mark each for the correct pronunciation, and '0' mark for the wrong pronunciation. The inter-rater reliability was carried out using Pearson Correlation d to determine the level of agreement between the two raters. One of the researchers transcribed the pronunciation tasks performed by the participants to determine the substitution that took place. The fact that the data was not normally distributed in line with Skewness and Kurtosis K-values for not falling within the acceptable range of ± 1 to ± 2 (George & Mallery, 2003;Rovain, Baker, and Ponton, 2013), a Mann-Whitney U test was conducted to compare the performance of the two groups in the pronunciation tasks.

Pronunciation of [i] and [i:] in the first syllable
To determine whether there is a significant difference in pronunciation performance between group 1 -Yorùbá native speakers who learned Hausa formally (in secondary school before attending a college of education) and group 2 -Yorùbá native speakers who learned Hausa informally (before attending a college of education), it was decided to use the Mann-Whitney U test. Participants' proficiency in pronouncing the [i] vowel was examined between groups 1 and 2. The findings demonstrated that there was no statistically significant difference in performance between the two groups when it came to producing short [i] in the first syllable (U = 740; p = .079). The mean rank indicated that the Yorùbá native speakers who learned Hausa informally before going to the college of education performed slightly better than the Yorùbá native speakers who learned Hausa formally before going to the college www.msocialsciences.com of education to learn Hausa as a second language. A Mann-Whitney U test indicated that there was a significant difference between the two groups compared in terms of the production of long [i:] vowel in the first syllable (U = 280; p < .001). The mean rank shows that the Yorùbá native speakers who learned Hausa informally before going to the college of education performed much better than their counterparts who learned Hausa in the secondary school before going to the college of education. The results presented in table 1 indicated that the mean ranks for the pronunciation of [i] vowel by participants in group 2 in the first syllable was better compared to the performance of group 1. Similarly, the mean rank for the pronunciation of [i:] vowel by participants in the first syllable by group 2 was also better than the performance of participants in group 1.

Pronunciation of [a] and [a:] in the first syllable
According to the results of the Mann-Whitney U test, there is no significant difference between the two groups' performance in producing the short [a] in the first syllable (U = 740; p = .502). The mean rank indicates that the Yorùbá native speakers who learned Hausa informally before going to the college of education performed slightly better than the Yorùbá native speakers who learned Hausa in the secondary school before going to the college of education. The Mann-Whitney U test was also used to evaluate the two groups' performance, and the results demonstrate that there is a significant difference in the pronunciation of the long [a:] in the first syllable (U = 620; p < .002). The mean rank reveals that the Yorùbá native speakers who learned Hausa informally before going to the college of education performed better than the Yorùbá native speakers who learned Hausa in secondary school before going to the college of education. The summary of results of the two groups based on the pronunciation of [a] and [a:] vowels in the first syllable are presented in Table 2 below: The results in table 2 revealed that the mean ranks for the production of [a] vowel by participants in group 2 in the first syllable was better than the performance of group 1. Similarly, the mean rank for the production of [a:] vowel by participants in the first syllable by group 2 was also better than the performance of participants in group 1.

Pronunciation of [i] and [i:] in the second syllable
The Mann-Whitney U test indicates that pronouncing the short [i] in the second syllable differs statistically and significantly (U = 540; p < .002) between the two groups. The mean rank revealed that Yorùbá native speakers who learned Hausa informally before going to the college of education performed significantly better than the Yorùbá native speakers who learned Hausa in the secondary school before going to the college of education. According to the Mann-Whitney U test, there is no statistically significant difference between the two groups' performance in pronouncing the long [i:] in the second syllable (U = 760; p < .308). However, the mean rank reveals that the Yorùbá native speakers who learned Hausa informally before going to the college of education performed slightly better than the Yorùbá native speakers who learned Hausa in the secondary school before going to the college of education.   Table 3 shows that the mean ranks for the pronunciation of [i] in the second syllable by participants in group 2 were higher than the performance of group 1. Similarly, the mean ranks for participants in group 2 also in the pronunciation of [i:] vowel in the second syllable was higher than that of group 1.

Pronunciation of [a] and [a:] in the second syllable
The Mann-Whitney U test demonstrates a significant difference in the pronunciation of short [a] in the second syllable between the two groups (U = 620; p < .044). The mean rank shows that Yorùbá native speakers who learnt Hausa informally before attending a college of education did much better than Yorùbá native speakers who learned Hausa in the secondary school before attending a college of education. The Mann-Whitney U test found that there is a statistically significant difference between the two groups in the pronunciation of long [a:] in the second syllable (U = 720; p < .041). The mean rank also shows that Yorùbá native speakers who learnt Hausa informally before attending a college of education scored much better than the Yorùbá native speakers who learned Hausa in secondary school before attending a college of education. Table 4 summarised the findings of the two groups in the production of the [a] and [a:] vowels in the second syllable:  Table 4 showed that the mean rank for the pronunciation of short [a] in the second syllable by participants in group 2 was higher than the performance of group 1. Similarly, the mean rank for participants in group 2 in the pronunciation of long [a:] in the second syllable was also higher than that of participants in group 1. All of these are examples of the two groups' pronunciation errors.  The fact that the substitution is a sound replacement process in which one sound replaces another in the same environment, it can also be linked to word mispronunciation in second language learning. Adekunle (2014) adds that when sounds are realised as different phonemes in pronunciation, such sounds manifest some kinds of changes at the segmental level. The substitutions that took place resulted to mispronouncing of some disyllabic Hausa words by the Yorùbá native speakers, taking into account the high front and low vowels in the first and second syllables. The substitutions made by the participants in this study are shown in Tables 6 to 9. Any substitution less than 4 is considered trivial and treated as a performance error, whereas substitution from 4 and above is recognised as a competence error, especially when done by the Yorùbá participant. Note. While the Hausa vowels were underlined, and the Yorùbá vowels being italicised, the shared vowels were therefore bolden, and a dash signifying no substitution took place.
The error counts in Table 6 were converted into percentages, with the most frequent occurrence at the top and the least frequent occurrence at the bottom of the ranking. Other substitution cases recorded in Table 6 range between 1 and 2 also point to performance errors due to their infrequent occurrence. Such infrequent substitutions that occurred in the present study are related to group 1, especially in the first syllable, which includes:        Table 8 serial number 1-4 above). The erroneous vowel substitution of L1 sounds by L2 speakers usually have to do with differences in vowel inventory in different co-occurrence restrictions of vowels and different types of tones and tonal patterns between L1 and L2. However, Table 10 illustrates erroneous vowel replacements with their numbers based on negative inference from L1 on L2 (i.e. Yorùbá to Hausa) pronunciation. [a] * * * 3

Vowel i i: e e: a a: o o: u u: ai au ɛ ɔ ĩ ɛ̃ ã ɔ̃ ũ
Table 10 indicated vowels that were substituted, and the ones substituted with, in the first and second syllables. Meanwhile, the asterisk * denoted that no substitution had taken place. It also showed that Hausa phonemes were correctly pronounced depending on the syllable. Thus, certain disyllabic Hausa words were mispronounced, and have their meaning changed according to learner pronunciation. Table 11 and 12 below illustrate how some Hausa words were realised:  Tables 11 and 12 identified vowels that were replaced and the mispronunciation of Hausa words that occurred due to substitutions in the first and second syllables. Please note, * shows vowel was not substituted, and the meaning did not change. ** indicates that vowel was substituted, and yet words retained their meaning.  Table 11 serial number 1). In this case, the entire meaning of such Hausa word is lost due to wrong pronunciation. The lengthening of the second syllable vowel can be explained not only due to vowel length, but also due to vowel quality/quantity from a front high to a mid-low vowel (with no change in tonal height). According to Yorùbá convention, the pronunciation of vowels at the word-final position is relatively long regardless of tone (Eme & Uba, 2016). As observed from the Yorùbá dictionary, the áCó (HH) (VCV) sequence has the lowest frequency in terms of word counts, compared to áCo (HM) and finally áCò ( (2010) considers as a case of reinterpretation whereby a phoneme is replaced with a source language phoneme for learning to progress. English word such as [ɡɪˈtɑː(r)] was mispronounced as [gìtá] 'guitar'. Since vowel length is merely phonemic in Yorùbá, length variations in the pronunciation of vowels occur. This is because vowel initial nouns are never high-tone initial in Yorùbá (Akinlabi, 2007;Babarinde, 2015), and that Yorùbá vowels are sometimes pronounced neither short nor long, but somewhere in-between which could be regarded as mid-long pronunciation regardless of tone. The results in this study are in line with Samson, Abdullahi, & Olagunju, (2014)  According to the study, it was a case of reinterpretation sounds where a phoneme is replaced with a source language phoneme which is readily available in the learners' mother tongue for a smooth learning process.

Implication
The implication of continuous learning the pronunciation in error, the standard of Hausa will not only fall but will continue to be learnt in error. Therefore, this article recommends that pronunciations should be given priority attention in schools by making it a compulsory subject for all language-related courses starting from primary to tertiary level. The teaching of Hausa subjects in schools should be handled by only qualified Hausa language teachers. More emphasis should also be given to the teaching and learning of vowels. Well-equipped modern language laboratories should also be provided for effective second language learning in schools. In taking all these measures, the Yorùbá speakers learning Hausa as a second language will attain a high level of proficiency in the Hausa language.

Conclusion
Experts in language and linguistics have since recognised that learners' performance in the target language is invariably marked by faults, both in speech and in writing, as vowel production is a key component of second language learning. As a result, the discrepancy between L1 and L2 causes negative transfer errors in learning a second language. While the results of the quantitative analysis for the two groups were statistically not the same, the performance of individual participants also differs according to syllables. However, certain non-linguistic factors such as age, learners' attitude and individual differences, as well as a learning environment (Maikanti, Thai, Burkhardt, Fung, Husain, and Olúwadọro, 2021) all play significant roles in determining the performance of the two groups measured. As evidenced from the data, while the performance of the two groups in the pronunciation of The findings of the present study align with Nhem (2019) who compared the performance of young learners and adolescent learners on language learning strategies where the results of the two groups are statistically not the same with their p-value less than (p=.033). The findings also concur with the study in Chilkiewicz (2015) who also compared the mean scores of young learners (M=18.75) with the adolescent learners (M=17.46) in terms of language learning with the p-value less than p ≤ .050. While the general outcome of the present study support the predictions of Flege & Bohn (2020) that the shared sounds between L1 and L2 are difficult for the Yorùbá speakers, learners of Hausa, the unshared sounds between the mother tongue and the target language are easier for the L2 learners to produce. Accordingly, the availability of particular sounds in the learner's mother tongue makes his brain to be less active in the processing of new sounds in the target language and vice versa. The fact that one of the researchers is a native speaker of Hausa, instances of mispronunciation were personally observed from the pronunciation of some non-Hausa speakers such as the case of pronouncing [bà:ƙó:] 'stranger', vs [Bà:ƙó] 'personal name, as well as [gàdá:] 'antelope' vs /Gàdá/ [Gàdá] 'name of a town'. The findings of the present study has been supported by the predictions of 'Revised Speech Learning Model' (SLM-r) which says that the challenges faced among the Yorùbá native speakers in terms of pronouncing some disyllabic Hausa words. Such challenges were due to the substitution of certain sounds with other sounds during pronunciation. Other factors responsible for the difficulties include vowel shortening, vowel lengthening, as well as over generalization of language rules.