Languages vary from place to place, social group to social group, and circumstance to circumstance. According to Labov [1], variability is an integral part of a linguistic system.  Language is not homogeneous, which means that no matter how closely two people are related, they cannot speak identically. Thus, the study of language variation can lead to significant language study insights. This means that it can assist us in observing the evolution of languages over time. This is the reason why Wardhaugh and Fuller [2] assert that "you cannot pronounce words anyway you please, inflect or not inflect words such as nouns and verbs arbitrarily or make drastic alteration in word order in sentences as the mood suits you”.

Sociophonetics is where phonetics and sociolinguistics intersect. Thus, it is a branch of linguistics that combines sociolinguistics and phonetics in a broad sense. Sociophonetics is defined by Hay and Drager [3] as the study of socially conditioned phonetic variation in speech. This field of sociophonetics and sociolinguistics began in the 1960s and 1970s with the work of William Labov, who discovered statistical correlations between the use of particular pronunciations and membership in particular social categories. Labov [4] examined the frequency and intensity of the variable “/r/” a post-vocalic “/r/” in words such as “fourth floor” in New York. The results revealed that the use of “/r/” was more prevalent among people of higher social class, indicating that the variable's usage was socially stratified.

According to Baronowski [5], this field of study focuses on differences in speech production, the social significance of specific pronunciations, perceptions and perceivability of sociophonetic patterns, and the role of sociocultural factors in phonetic models of production. Foulkes, Scobbie, and Watt [6] are of the view that sociophonetics is especially interested in the causes of speech variation, with numerous studies focusing on pronunciation differences across regions, social classes, races and ethnicities, gender, and age.

Botha [7] distinguishes among three approaches to a sociophonetic study. These approaches are the distributionist or quantitative approach, ethnographic or qualitative approach and the constructionist approach. According to her, the most popular approach to studying variation is through quantitative method using survey and this method builds on the groundwork study by Labov [4]. The second approach to studying variation is qualitatively by investigating speech communities and the relationship between members of those communities. The last approach has to do with how speakers use linguistic and other cultural resources in the on-going construction and reconstruction of personal and group identity.       

Mclaughlin [8] writes that due to the influx of migrants speaking minority languages coming from within the country, from other regions of Africa and from abroad, African cities are more multilingual now than at any other point in their history. This has brought so much variation and change in languages spoken in parts of Africa and West Africa is no exception. This variation and change is taking place in all aspects of language, thus phonetics, phonology, morphology, syntax and other aspects and that is this work tends to review. The objective of the review is to present some issues of sociophonetic variation of lingua franca in West Africa and what has been done concerning those issues.

Related Works on Akan as a Lingua Franca

Yankson [9] examined language contact and change in linguistically diverse urban communities by concentrating on Akan in Accra. Due to the ethnic and linguistic diversity of African cities, she argued that in many instances one or more urban vernaculars have emerged as the vernacular language of the city. The observations made by Yankson [9] regarding African cities in general are applicable to Accra, the capital of Ghana. Accra is the most populous city in Ghana in terms of employment and other means of subsistence, and the city has attracted migrants from all over the country and the world. Akan is not an indigenous language of Accra; rather, it is a migrant language with numerous imported dialects.

The focus of the study was on particular lexical, morphosyntactic, and phonological variables. It primarily discussed the structure of the Akan dialect spoken by Akan migrants in Accra within the context of new dialect formation. Yankson also investigated interlanguage, codeswitching, and borrowing due to the presence of other languages, including English and Ga.

The primary objective of her project was to compare the Akan dialect spoken by Akan migrants in Accra to the dialect spoken in the rural areas where the language was originally spoken in order to identify any possible changes. The research was guided by several research questions, but the current review focuses on the lexical, grammatical, and phonological differences between the Akan variety spoken in Accra and that spoken in the rural areas of origin.

Yankson [9] employed three types of data to explicate her research questions in her methodology. First, a questionnaire survey was administered to Akan migrants in Accra to collect data. Through semi-structured formal and informal conversational interviews with participants in Accra and rural Akan communities, the second and third types of data were collected. The study's research area included the Accra Metropolitan area, the heart of Accra, as well as the Asante and Kwawu rural communities. In addition, the study employed two sampling techniques, namely convenience sampling and judgement sampling. In total, 168 individuals participated in the study. To investigate the change and possibly the emergence of a new dialect, participants from both rural and urban areas were recruited. In summary, 48 individuals participated in both semi-structured and free, informal conversational interviews with second-generation migrants in Accra and participants from rural communities. In contrast, 129 Akan migrants in Accra responded to the questionnaire.

Sociolinguistic variables are important in a sociophonetic study. The linguistic variables selected for Yankson [9] investigation were some lexical, morphosyntactic and phonological variables. However, for the purpose of this review attention would be given to the phonological variable. The phonological variables investigated were: 

• The application of the vowel harmony rule 

• The 3rd person plural pronoun won 

• The progressive marker re 

• The use of /k/ and /ky/ allophones in some selected lexical words

According to the vowel harmony rule in Akan, all the vowels in the word should belong to the same category. So, the use of vowel harmony rule among the Akan second generation in Accra was investigated to find out if there could be possible changes because of contact with languages like Ga without ±ATR harmony rule. Also, /k/ and /ky/ variants are allophonic variants in Akan. They complement each other in their distribution and that they have different pronunciation but of the same basic consonant. For example, sakra, sakyera “repent”. Nevertheless, by the observation of Yankson [9] there are a few words in which the use of /k/ and /ky/ has become dialect specific and these words include Akan equivalent of ‘sugar’, ‘repent’ and ‘tongue’.

According to the findings of the study, the Akan variety spoken by Akan migrants in Accra differs from that spoken in the original rural areas. For the selected lexical variables, the variants used by second-generation migrants were significantly different from those used by participants from rural communities. The variants that the second-generation migrants used for the lexical variables reflected their contact with various Akan dialects. Extreme variability and a phonological shift distinguished the dialects used by second-generation migrants. In conclusion, the second-generation immigrant's Akan dialect is undergoing the process of new dialect formation. This is a result of Accra's ethnic and linguistic diversity; however, there are additional variants in this dialect as a result of its contact with other languages in this environment.

Although Yankson's [9] work is not a sociophonetic study in its entirety, it is worthy of discussion. Because it supports the argument of this review that language evolves over time. Akan is one of the most widely spoken languages in West Africa, and the urban variety has undergone changes in phonology, morphosyntax, and vocabulary.

In a similar study on the changes taking place in Akan, Boakye [10] also investigated the phonological variable /r/ in Akan. He observed that in Akan (l), (r) and (d) occur intervocalically and after some consonants in Asante, Fante and other Akan dialects and they are in free variation. This means Akan speakers are noted for interchanging one variant for the other in speech. He argued that that a closer look at the use of the variants show that it can be constrained by non-linguistic factors and there now seems to be a strong stigma attached to the use of some of the free variants in Akan. 

Unlike Yankson [9], Boakye employed the Variationist sociolinguistic framework for his study. The linguistic variable chosen for the study was the phonological variable (r) and its variants (r), (l) and (d). The data collection instruments used for the study were interviews, pictures and questionnaire. Interviews and pictures were used for the production study whilst the questionnaire through the matched guised technique was used for the perception study. A quantitative statistical analysis was used for the data analysis.

The study revealed that the (r) variant is regarded as prestigious in speech, resulting in the stigmatisation of the other variants, particularly the (d) variant, which was the least frequent. It was discovered that Asante speakers are more likely than Fantes to use (d) in speech. As educated Akan speakers stigmatise the (d) and (l) variants and favor the prestigious form (r), the first finding reveals that there is a greater variation in the use of the variants based on educational attainment. Second, the study demonstrates that female Akan speakers in both dialects are more likely to use the prestigious (r) variant, whereas males are more likely to use the less prestigious (l) and (d) variants. Men are less influenced by social stigma, whereas women prefer variants with obvious prestige. Listeners attribute certain group characteristics to speakers based on the choice of (r), (l), and (d) in speech, according to the results of a perception study of variants. Confirming the production data on the stigmatization of the (d) and (l) variants, Akan speakers consistently evaluate the (r) variant positively across all social categories studied.

The works discussed above have compared the arguments in studies in language variation and change that social variables can influence the choice of language features. It also supports the argument that in a typical sociolinguistic setting, women prefer variants with overt prestige while men are likely to use the less prestigious variant.

The work of Boakye [10] could have considered the pronunciation of the /r/ sound by examining other dialects with a significant number of speakers, such as Akuapem, Bono, and other common dialects. Equally, data could have been collected from both rural and urban areas to examine the correlation between /r/, /d/, and /l/ alternation. In addition to the pronunciation of /r/, there are quite a few sounds in the Akan inventory with multiple realizations that could have been taken into account in the study.

Sociophonetic Works on West African English

There are other sociophonetics works that focus on the pronunciation of sounds in English by West Africans. An example is Okyere [11], who investigated the variation in the use of affricates and fricatives in English spoken in Ghana by Ghanaian students and teachers at the primary, Junior and Senior High Schools. The aim of his study first looks at the alternative pronunciations of the affricates and the fricatives by the respondents from the various educational levels. Also, he compared the realization of the sounds under consideration by the various educational levels to determine whether there is any evidence of differences or similarities which are brought about as a result of the differences in their educational levels. Wells’ Synchronic Analysis Approach was used to examine the accents as they are. The alternative realizations by the students at each level were compared to find out whether the individual’s educational level was responsible for the variations and for that matter the selection of a particular variant. Again, Okyere [11] compared the realizations of these students at each level to that of their respective teachers to determine whether the variations exhibited in the students’ speeches were influence by their teachers.

In that case, the data generated from the students were compared to those of their respective teachers to determine whether the teachers had any influence on the way their students speak English. All the student respondents, and for that matter speakers of English language at all the levels studied, realized both the voiceless and the voiced affricates /ʧ/ and /ʤ/ reasonably correctly. The voiceless palate-alveolar affricate /ʧ/ was found to have three alternative realizations of the same phoneme /ʧ/ at all the levels and by all the teachers. They vary between (ʧ), (ʃ) and (t). The Received Pronunciation /ʧ/ has the highest rate of occurrence among all the variants. For example, the following represent the percentages of [ʧ] 68.3%, (t) 20.3%, (ʃ) 9.7%, (t) or (ʃ) and (ʧ) or (ʃ) used by the same respondent. 

The voiced palate-alveolar affricate /ʤ/ also had three alternative variants of the same phoneme /ʤ/. These are (ʤ), (d) and (g), with the RP variant (ʤ) having the highest frequency of 42.2%. This is followed by (d) 32.5% and (g) 25.3%. It was also identified that whereas (ʤ) occurs in words such as joy, large, lodge, and so on (g) surfaced in digit, digital etc. with (d) occurring in ‘educate’, ‘graduate’ etc. The researchers concluded that these variations occurred at all the levels with the teachers exhibiting similar traits. Therefore, in his study it appears as if orthography plays a part in the determination of the variant that was selected.

Moreover, the researcher stated that the fricatives were realized by all the respondents at all levels. The voiced and voiceless labiodental fricatives /f, v/ as well as the glottal fricative /h/ did not show any alternative realizations in their declaration by all the respondents but the inter-dental fricatives /θ/ and /ð/ were also realized differently by all the respondents at all the levels. The voiceless interdental fricative /θ/ had three alternative variations (θ), (t) and (f) with (θ) having the highest rate of occurrence. The voiced interdental fricative /ð/ on the other hand, had five alternative variants (θ), (t), (f), (d) and (ð) with (d) being the highest variant in the respondents’ speeches among all the variants. The percentages among the five variants were (d) 50.8%, (t) 31.7%, (θ) 11.1%, (ð) 6.1% and (f) 0.3%. 

According to the researchers the voiced and the voiceless alveolar fricatives /s/ and /z/ were realized as it is in Received Pronunciation in certain words by all the respondents. Nonetheless, it was realized by the researchers that there were consistent and conscious switches from either (s) to (z) and vice versa in the realization of certain words. For example, words such as visit, assume, basic, etc. The voiced alveo-palatal fricative /ӡ/ had three alternative variants (ӡ), (ʃ) and (ʤ) with (ʃ) being the highest among all the variants in the speeches of all the respondents. Apart from (ʃ) in ‘mission’ which was realized as (ӡ) by two respondents, all the other respondents realized it as it is in RP and therefore did not show any ostensible variation in the speeches of over 90% of all the respondents. While words such as measure, and pleasure were all realized with (ӡ), that in occasion, conclusion, decision, etc. were all realized with (ʃ) and this was consistent with all the realizations. 

At the end of his research, the findings revealed that all the three levels used alternative realizations identified in the study but differences were shown in percentages and preferences. The analysis again revealed that what goes on in the various educational levels determines how certain sounds are made by students in Ghana. How English is taught and learnt has greater effect on speech variation than educational level, though it must be admitted that one’s level of education to a large extent determines one’s level of competence in grammar and semantics but not in pronunciation.

The choice of a particular variant by the primary, JHS and SHS respondents were not arbitrary but were influenced by teacher motivations as shown by the study. How the teachers pronounce words and speak English in each of the school communities studied affect the way students pronounce words and speak English. 

It is good to our argument on how the target language is taught and learnt having greater effect on the variation but I am of the view that learners and teachers’ attitude and that of the society in general towards the language also come to bear. The moment one tries to speak or emulate the native speaker that individual is seen as being too knowing. This is because the English language is the language of the elite. More so, the English language is not revered by learners as they respect and see foreign languages like French and Spanish because no particular attention is paid to Received Pronunciation. Again, most of these phonemes are not found in the local languages and so when it comes to producing them, we tend to use the alternatives found in our local languages or substitute them with sounds that are closer to the said sounds.

Moreover, Oladipupo and Akinjobi [12] examined the variable use of r-liaison and boundary consonant deletion processes in the speech of young Nigerian speakers of English. This was with a view of confirming the hypothesis that continuous speech processes (CSPs) can be socially differentiated in a speech community. It was stated in the study that although social class is much more pronounced amongst the native speakers of English, it cannot be completely disapproved in the L2 setting. The reason is that class difference is inherent in every society, though at varying levels and in terms of different factors ranging from economic to cultural and political and Nigeria cannot be an exception. As stated by the researchers, Nigerian society is essentially polarized into haves and have-nots on the basis of economic and political advantages that a class has over the other and this determines to a large extent the quality of education, level of social exposure and opportunities available to members of each class, which had somehow influence their speech. Base on the scenario therefore, the researchers saw how imperative it is to begin to pay attention to class as a social variable in the L2 setting. So, they attempt to examine the use of r-liaison and boundary consonant deletion processes in the speech of young Nigerian speakers of English in order to establish possible correlations of these features with gender and social class in Nigerian English.

Two broad questions were asked by the researchers in which they sought to receive an outcome at the end of the study. The questions were as follows:

• Do gender and class variations exist in young Nigerian English speakers' use of the r-liaison process

• Does the boundary consonant deletion process correlate with gender and social class of young Nigerian English speakers

A sample of 180 young educated Nigerian speakers (between 18 and 35 years) was drawn, through stratified and purposive techniques, from the six geo-political zones in Nigeria (North West, North East, North Central, South West, South East and South-South). This was with a view to selecting participants who are representative of young speakers of English across the entire country.

The participants for the study comprised students of public and private higher institutions as well as members of the National Youth Service Corps who were fresh graduates observing the one-year mandatory service to the nation in those parts of the country. The choice of students and graduates was motivated by the need to sample educated speakers for the research. The young speakers so selected from all the geo-political zones were stratified into gender and social class without consideration for their ethnic backgrounds (male-low class: 45 speakers; male-high class: 45 speakers; female-low class: 45 speakers; female-high class: 45 speakers).

The researchers and his trained research assistants visited the participants in their various institutions and places of primary assignment and administered to them, on one-on-one basis, 180 copies of a structured questionnaire and a test comprising 19 utterances and a short passage, both containing potential r-liaison and boundary consonant deletion sites. The questionnaire was meant to verify and identify their social backgrounds (whether male or female; low or high social class). Some examples of the relevant items, extracted from tests include;

Item                                      Liaison type

Peter at                                 Linking /r/

More of him                                         “

After a while                        ,,

Boundary Consonant deletion

Doesn’t she            won’t do it     kept quiet                 exact colour                          

Test drive               don't buy it         jumped well     equipped with

The scores were analyzed statistically, using the Analysis of Variance (ANOVA), based on gender and class of speakers. The two dependent variables (r-liaison and boundary consonant deletion) were analyzed separately and their results were reported before discussion.

The result suggested that only a little difference existed between male and female speakers. While the males had a mean score of 0.844, the females' score was slightly higher (0.933). The same trend was found between the two social classes. The high-class speakers had only a little higher mean score (0.922) than the low class (0.856). In order to determine the significance of the results, therefore, a univariate Analysis of Variance (ANOVA) was carried out, with participants' scores as the dependent variable and gender and class as independent variables, using the IBM SPSS statistics 20 packages.

The ANOVA results in the study confirmed absence of gender variation, class variation, and group interaction. This implies that no variation was found amongst the participants in r-liaison usage.

The mean scores for the independent variables (gender and class) were calculated from the individual scores in boundary consonant deletion process. The findings reveal a considerable difference between male and female participants in boundary consonant deletion. Aggregated mean scores of 9.51 for males and 8.56 for females suggest that the incidence of boundary consonant deletion was higher amongst male participants than female participants. On the other hand, the class difference was smaller. While low class speakers had an aggregated mean score of 8.80 the high class had 9.27.

The results of ANOVA analysis in the study show that, at 0.05 significant level, there was a significant variation between the mean scores of male and female speakers relative to boundary consonant deletion, (F(1, 176) = 6.24, p= .013); whereas, no significant class difference was found, (F(1, 176) = 1.49, p = .224). 

This implies that male speakers significantly deleted consonants at word and morpheme boundaries more than female speakers, while both the low and the high-class participants had equal tendency towards deletion. However, the table also reveals gender-class interaction effect, F(1, 176) = 5.13, p = .025), which means that social class differed significantly between both sexes: male high, with a mean score of 10.178, used boundary consonant deletion more than female high with 8.356. This shows that the gender variation was due to differences between male and female high-class speakers.

In the light of these findings vis-à-vis related submissions in sociophonetic research therefore, this study has not sufficiently demonstrated variability in the speech pattern of young Nigerian speakers in relation to r-liaison and consonant deletion. To start with, it is usually claimed in the sociophonetic tradition that female speakers use more standard or prestigious speech variants and fewer low-status forms than males [13-14], and that the speech of members of higher social classes correlates with the standard forms, while vernacular forms are most prevalent among the lower social classes [4].

In light of this, a greater usage of r-liaison (as a prestige variant) would have been anticipated not only from female speakers but also from high-class participants. However, this was not the case, as neither gender nor class variation was observed in this CSP, indicating that r-liaison usage among the participants was of equal status regardless of gender and class. The gender difference can be traced to the need for male speakers to achieve articulatory economy. Elision is considered a phonetically motivated process that is typical of connected speech because it facilitates articulation [15]. Overall, the study demonstrates that Kerswill's [16-17] assertion that CSPs may be socially differentiated in a speech community is not fully supported in Nigeria, particularly in terms of the variable use of r-liaison and consonant deletion in connected speech by young speakers. This is due to the fact that there was very little variation in the data.

Folajimi, Zeyu, and Sin [18] investigated the realisation of the voiceless dental fricative / and its variation among educated Nigerian English speakers, with respect to the variation of English spoken in West Africa. They noted that previous research on / variation revealed that / is frequently pronounced as /t/ in Nigerian English. However, a systematic analysis of the / variation and investigation into its potential correlation with social factors are lacking. 33 educated English-speaking Nigerians were chosen for the study. They were chosen from the International Corpus of English (ICE Nigeria), with gender, ethnicity (Hausa, Igbo, and Yoruba), and age group (20-29, 30-39, 40-29, and 50 or older) serving as social factors for the study. There were 729 tokens contributed by 18 male and 15 female speakers. The transcription was performed phonetically by repeatedly listening to the recordings. The tokens were categorized into three main phonemes: /θ/, /t/ and /s/.

Using a mixed effects logistic regression model, the data were analysed. While there are three major variants of /θ/ realisations in educated Nigerian English, speakers appear to prefer to articulate the voiceless dental fricative as /t/ and /θ/. The analysis of 729 tokens revealed that the majority of speakers preferred to realise the variable as /t/ (356 tokens), /θ/ (344 tokens), and /s/ (29 tokens). There was a narrow distinction between /θ/ and /t/. This demonstrates that /θ/ has become an integral part of the phonology of educated Nigerian English speakers.

The findings also indicated that /θ/ variation in Nigerian English is sometimes affected by sociolinguistic factors. Female speakers significantly favor /θ/ more than /t/ while male speakers use more /t/ than the female counterparts. This shows that educated females tend to favour the use of the target like variety than educated men. It was found that the realization of the /θ/ variable was affected significantly by the speakers’ ethnic group. Yoruba speakers used significantly more /θ/ than speakers from other ethnic group whereas Igbo speakers used significantly more /t/ than Yoruba and Hausa people. Age factor did not have much impact on the speakers’ performance apart from speakers in their 30s who recorded significantly less /θ/ than the other age groups. There were no indications that the production of the TH variants was strongly influenced by linguistic factors.

Other Related Works on Sociophonetics

Inusah, Amuzu and Akanlig-Pare [19] is another sociophonetic study that confirms the connection between social factors and linguistic variants. They observed that the standard Dagbani language is losing the pronunciation of [l] in a number of feminine names, including Labi, Lukaya, and Luba. They contend that [r] does not occur in a word initial position in Dagbani, yet it is prominent in the current pronunciation of Dagbani names. Dagbani is primarily spoken by Dagbamba people in Ghana's Northern Region. Dagbani has three primary dialects: Nayahili, which is spoken in Yendi and its neighbouring villages; Tomosili, which is spoken in and around Tamale; and Nanuli, which is spoken in Bimbila and its environs.

The purpose of the study was to test the hypothesis that the proportion of [r] would be prevalent in the speech of the younger generation, given that [r] is an innovation in the Dagbani speech community due to the influence of age and gender, whereas [l] would be prevalent in the speech of the older generation.

The data for the study was derived from the Nayahili dialect spoken in Yendi and the surrounding villages. The majority of the data used was collected through in-depth interviews with key informants based on the researcher's intuitions as a native speaker of Dagbani. Rapid, anonymous surveys and sociolinguistic interviews were employed as interview techniques. In addition, the researcher relied on his own intuition to supplement the data obtained from the unstructured interview. The participants used in the studies were young and elderly male and female natives who lacked formal education and rarely travelled outside of the city. A total of 160 participants were purposefully and randomly selected based on their age and gender. Examples 1 and 2 present the alternation of [l] and [r].

Rabi tʃan Tamale                                                     

Rabi go PAST Tamale

‘Rabi went to tamale’

Labi tʃan Tamale

Labi goPAST Tamale

‘Labi went to Tamale’

According to the findings, 71 percent of participants used [l], making it the predominant form. The results on gender use of [r] confirm that [r] is the dominant variant in the speech of males, who account for 73.8% of [r] usage, and females, who account for 15% of [r] usage. In addition, the data revealed that social factors are capable of spreading [r], although age is more prevalent than gender. The differences between Age's stages are indicative of networking that will impact the spread of [r]. Thus, [r] was more prevalent in the speech of the younger generation, whereas [l] is more prevalent in the speech of the older generation.

Korsah [20] presented a quantitative and qualitative analysis of the linguistic preference of a speech community, the University of Ghana, with reference to the use of the phono-lexical variants [lègↄn] and [légↄn] and how this helps members and non-members of this community understand the sociolinguistic dynamics of the community.

The objective of Korsah's work is to demonstrate quantitatively, how the use of the word "Legon" /legↄn/ by the University of Ghana speech community in reference to the university or its location predicts a situation analogous to the studies already cited, namely the significance of the phonological variants [lègↄn] and [légↄn]. This study demonstrated how the use of these can reveal the linguistic and social orientation of the Legon speech community's members.

The researchers observed meticulously, the tone differences in the first vowel which is /é/ in [légↄn] and /è/ in [lègↄn]. Korsah [20] made it clear that the tone in the second vowel is almost always the same as in [lègↄ‵n] and [légↄn] but occationally [lègȏn]. This study as the researchers stipulated did not pursue the tone on the second syllable. So, Korsah’s working hypothesis was that [lègↄn] is the more dorminant variant while [légↄn] was used more by female members than males. 

Korsah [20] utilised a questionnaire and participant observation to collect data for determining the outcome of phonological variants. In total, he interviewed and sampled fifty individuals from two groups of fifty. In addition, the researchers targeted subjects who could respond to the interview questionnaire. First, the researchers considered twenty-five males and twenty-five females from all corners of the community of speakers of the said language. Utilized were thirty students and twenty non-students. The non-students included four lecturers, four departmental staff members, one security officer, two campus vendors, a photocopier operator, a cleaner, a porter, and a banker.

The majority of the interview was conducted in English, with the exception of the responses, which subjects were encouraged to provide in their native tongue, and almost all of them chose Akan. The interview was recorded with an audio recorder and supplemented with orthographic recordings made with a pencil on a separate sheet, on which all social and linguistic variables were tabulated and verified with the interviewer. The second engagement consisted of at least two minutes of casual conversation with the interviewer before the actual interview questions were asked. The second group consisted of fifty students only, with twenty-five males and twenty-five females.

From the available data, according to the researchers, both hypotheses were confirmed in the first conversational interview. About 60% of the respondent consistently used [lègↄn] and never used [légↄn]. About 18% (9 out of 50) alternated between [lègↄn] and [légↄn]. This was a confirmation of hypothesis one and also confirmed hypothesis two. About 32% (8 out of 25) of the female respondents as against 12% (3 out of 25) of the males used [légↄn] alone.

It was then inferred from the data that the use of [légↄn] is somewhat evaluated as more prestigious than the use of [lègↄn]. In addition, more female respondents used [légↄn] in the casual speech (conversation) interview than the males. Since prestige is crucial in deciding on the standard variety of a particular speech form, it was deduced that the Legon community perhaps covertly takes [légon] to be the standard variant. A conclusion was drawn by Korsah [20] that if the premise is acceptable, then the crosslinguistic commonality that “In stable sociolinguistic stratification, men use a higher frequency of non-standard forms than women” [14] is attested per the results of Korsah’s investigation.

Reviewers’ opinion of the subject in the sub region.

Though sociophonetics is a wide area in Linguistics, much study has not been done in Africa since it is still a growing field in this part of the world. It is important that linguists embrace the fact that languages change over time and innovation in linguistic system is inevitable. 

Generally, the pattern of the works reviewed confirms the observations and findings from other works done in other parts of the world. This shows that if attention is given to the field in West Africa, there will be development in the field of study. In addition, since Africa has many languages, contact among the languages will reveal many variation and change taking place in the languages if studied. Yankson [9] and Boakye [10] has proven variation and change in Akan lexically, grammatically and phonologically. The works also show that social factors, such as age, sex, ethnicity, and class as well as others have a relationship with language.

For example, the results from most of the studies confirm linguistic differentiations between men and women have proven to be clearest and most consistent in sociolinguistic research. Firstly, Boakye argued that in a stable sociolinguistic stratification, men use a higher frequency of non-standard forms than women. Secondly, in the majority of linguistic changes, women use a higher frequency of the incoming forms than men. The review has shown that in West Africa gender plays a major role in the pronunciation of sounds and words in both local languages and English. The phonological changes in Akan as reported in Boakye [10] are and evidence of gender related variation.

Moreover, the review indicates that education as a social variable has shown to be a major factor in sociophonetic variation in West Africa especially on English language. Several studies have drawn a distinction between languages used by the educated and the less or uneducated speakers as well as people at different levels of education for example Okyere [11], Folajimi et al. [18] and Oladipupo and Akinjobi [12]. Furthermore, there are several ethnic groups in West Africa and each of them has a language which is distinct phonologically, morphologically and syntactically. Yankson posits that Akan as a Lingua Franca is going through a change in Accra where it is in contact with other languages leading to a new dialect formation.

Sociophonetic studies is a growing field in West Africa. The papers used were drawn from two West African countries state of Ghana and Nigeria. The objective of the review is to present some issues of sociophonetic variation of lingua franca in West Africa and what has been done concerning those issues. The review has shown that Lingua Franca (Akan variety) spoken by the second generation migrant is undergoing the process of new dialect formation due to the ethnic and linguistic diversity of Urban areas. In addition, the review shows that the English spoken in West Africa is no exception to the variation and change taking place lexically, grammatically and phonologically.

1. Labov, W. The Social Stratification of English in New York City. 2nd ed., Cambridge University Press, 2006.

2. Wardhaugh, R. and Fuller, M. J. An Introduction to Sociolinguistics. 7th ed., John Wiley & Sons, 2015.

3. Hay, J. and Drager, K. “Sociophonetics.” Annual Review of Anthropology, vol. 36, no. 1, 2007, pp. 89–103.

4. Labov, W. The Social Stratification of English in New York City. Center of Applied Linguistics, 1966.

5. Baronowski, M. “Sociophonetics.” The Oxford Handbook of Sociolinguistics, edited by Robert Bayley, Richard Cameron and Ceil Lucas, Oxford University Press, 2013.

6. Foulkes, P et al. “Sociophonetics.” Handbook of Phonetic Sciences, edited by William Hardcastle, John Laver and Fiona Gibbon, 2nd ed., Wiley-Blackwell, 2010.

7. Botha, W. Dimensions in Variationist Sociolinguistics: A Sociolinguistic Investigation of Language Variation in Macau. M.A. thesis, University of South Africa, 2011.

8. McLaughlin, F. “Introduction to the Languages of Urban Africa.” The Languages of Urban Africa, edited by F. McLaughlin, Continuum International Publishing Group, 2009, pp. 1–18.

9. Yankson, S. A. Language Contact and Change in Linguistically Heterogeneous Urban Communities: A Case of Akan in Accra. PhD dissertation, Radboud University Nijmegen, 2018.

10. Boakye, B. A Sociolinguistic Analysis of the Phonological Variable [r] in Akan. M.Phil. thesis, University of Education, Winneba, Ghana, 2018.

11. Okyere, J. G. A Sociolinguistic Study of Language Variation in the English Spoken in Ghana: A Case Study of Some Selected Consonants. M.Phil. thesis, University of Ghana, 2013.

12. Oladipupo, R. O. and Akinjobi, A. A. “Sociophonetic Study of Young Nigerian English Speakers.” Ghana Journal of Linguistics, vol. 4, no. 2, 2015, pp. 92–107.

13. Hudson, R. Sociolinguistics: The Quantitative Study of Speech. Cambridge University Press, 1996.

14. Labov, W. “The Intersection of Sex and Social Class in the Course of Linguistic Change.” Language Variation and Change, vol. 2, no. 2, 1990, pp. 205–254.

15. Abiero, D. A et al. Pedagogical Challenges in the Teaching and Learning of Oral Skills in English in Secondary Schools in Kenya, 2019.

16. Kerswill, P. E. “A Sociophonetic Study of Connected Speech Processes in Cambridge English: An Outline and Some Results.” Cambridge Papers in Phonetics and Experimental Linguistics, vol. 4, 1987, pp. 25–49.

17. Kerswill, P. E. “Levels of Linguistic Variation in Durham1.” Journal of Linguistics, vol. 23, no. 1, 1987, pp. 25–49.

18. Folajimi, O. et al Sociophonetic Study on TH Variation in Educated Nigerian English. University of Munster, Germany, 2019

19. Inusah, A.R. et al “Sociolinguistic Variation of [ɾ] in Dagbani.” Sociology International Journal, vol. 1, no. 3, 2017, pp. 1–13.

20. Korsah, S. Whether Lègↄn or Légↄn, We’re in Legon. University of Ghana, 2012.

21. Koba, E. L. K. “A Critical Discourse Analysis of Jacob Zuma’s First Term (2009–2014) Inaugural and State-of-the-Nation Addresses through a Transitivity Focus.” Revue des Arts, Linguistique, Littérature & Civilisations (RA2LC), no. 1, 2020, pp. 79–94.

22. Wells, J. C. Accents of English. 3 vols. Cambridge University Press, 2000.