Language documentation in all areas of linguistics has salvaged many minority and endangered languages from near extinction or glutophagy. This mostly occurred in the era of language documentation in the last three to four decades. For example, Hoffman [1] was done on languages of the world to bring out the effect of language death on the language communities, the culture and its implications on academia.

Studies on Ghanaian languages such as Christaller [2], Westermann [3], Dakubu [4] and more recently, Dorvlo [5], Bobuafor [6], Asante [7] and Campbell [8] have all contributed immensely in documenting both majority and minority Ghanaian languages. Ahanta native and linguist Ntumy [9] and later on GILLBT [10], advanced Ahanta development by documenting for the first time the sound system and some vocabulary of the Ahanta language and the Ahanta Bible respectively. The Ghana Bible Society published the Ahanta Bible in 2015 while a radio station was established to broadcast sorely in Ahanta, in Agona Nkwanta (an Ahanta community) in order to aid generational transmission and revitalization. In the light of all these, this study was undertaken to add to Ahanta documentation by researching on an aspect of the core area of the phonology of Ahanta.

Background

Ahanta, pronounced Ahandaa belongs to the Volta-Comoe branch under Kwa language groups with related languages being Aowin (Esahie), Nzema and Akan. An Ahanta person is known as Ahandanii and the plural is Ahandamaa. According to a paper presented to parliament in 1821, Ahanta was a state or a country on its own, a rich coastal country now housed in the Western Region of Ghana.

The state of Ahanta had linguistic homogeneity with Ahanta as its language. Oral tradition indicates that like the Akan people, the Ahanta people moved from the Brong Ahafo area up north and settled in their present location down southern Ghana. It is believed that they first settled in Aboade before spreading out to occupy the various Ahanta towns of today. The name ‘Ahanta’ literally means “land of twins” while other sources claim the name was derived from the verb handa, meaning to ‘dry’ Others claim that Ahanta people are descendants of people who settled where they found a shark drying itself on the beach. This to them was a sign of peace because the shark did not bring any bad omen on them. Almost equal proportion of Ahanta people live in urban and rural areas. 

Classification of the Ahanta language

The Ahanta language is made up of three notable varieties. The variety spoken in the Sekondi-Takoradi metropolis up to areas around Apowa, Kwesimentsim and Beahu where the language is highly adulterated with Mfantse. This study categorized the Ahanta spoken here as ‘urban’ Ahanta. The second is ‘rural’ Ahanta spoken around Adwoa, Funko, Ewusie Joe, Aboade, Agona Nkwanta, Busua, Otopo etc. This is also considered by the Ahanta people as the unadulterated one. The third variety is Evaloe or Valoe spoken around Princess Town, Akatekyi and Cape Three Points. These areas are close to the border (River Ankobra) that is shared between Nzema and Ahanta. The variety found here is highly adulterated with Nzema and in recent times considered a variety of Nzema. Evaloe is also spoken beyond the Ankobra river among immediate Nzema communities. The difference among the three varieties resides in sound, tone and lexical items but are mutually intelligible. Data was collected in rural Ahanta in order to get a least adulterated Ahanta vocabulary. The study established the sound inventory of Ahanta, its distribution and discussed some of the phonological processes in Ahanta. The processes encompass both segments and suprasegments processes that occur within words, at word boundaries and within compounds.

Literature Review

There have been efforts in studies all over the world to safeguard the near extinction and death of minority languages. This has led to the documentation of aspects of such languages in linguistics, such as, phonetics and phonology, syntax, morphology, sociolinguistics, cultural studies and others. The monumental work of Christaller [2] is a lexicographical record of the Akan language. The study is a detailed record of almost all lexical items in all three written dialects (Akuapem Twi, Asante Twi and Mfantse) of Akan. Apart from the isolated lexemes, definitions are provided together with all available expressions associated with the headword. Hyponyms, synonyms and polysemous words are also provided. For example:

• tↄ-to buy, to fall, tↄ-to die

• Metaphorical use:               

• Tↄ famu -to fall down

• ano tↄ- to finish talking

• nsa tↄ so -to be conversant with doing something with the hand. 

• akoma tↄ yamu

• ani tↄ nsamu

• Christaller [2]

For the past hundred years, this dictionary has helped in studies related to the Akan language including knowledge in ethnography and anthropology.

Agbetsoamedo is a study on Sɛlɛɛ, a less studied Ghana Mountain-Togo (GMT) language spoken in Ghana’s Santrokofi, down across Togo. The project Aspects of the Grammar and Lexicon of Sɛlɛɛ, was based on five main areas of the Sɛlɛɛ, language, namely, noun classes in Sɛlɛɛ, the morphological encoding of diminution in Sɛlɛɛ, tense and aspect system of Sɛlɛɛ, standard negation in Sɛlɛɛ and lastly temperature terms in Sɛlɛɛ. After a thorough discussion of data collected from the Santrokofi area of the then Volta region of Ghana, where the language is spoken, the following conclusions were made concerning the syntactic structures. The project established that Sɛlɛɛ, has eight noun class prefixes in which four mark singular while the other four mark plural. It was noted that diminution is marked morphologically by the use of five different suffixes, -bi, -bii, -mii, -e or by the exclusive use of -nyi. Sometimes -nyi is employed in combination with noun class shift. The tense and aspect system shows that Sɛlɛɛ has three basic tenses being the present, hodiernal (pre-hodiernal and hodiernal) and some future tenses. Other aspects of the grammar established is negation and the grammatical constructions used for expressing temperature. Agbetsoamedo concluded that in terms of grammar, Sɛlɛɛ shares some similarities with other GTM languages.

An aspect of Logba (Ikpana), one of the fourteen GTM languages was studied by Dorvlo [5]. After the introduction comes a discussion on the phonological system of Logba which established that Logba has three open syllable types namely, peak only, simple onset and peak and lastly, an onset only. It also has two basic tones, Low and High which are found in most tone languages. Logba also has prefix classes, singular-plural pairings and agreement systems as its three interconnected noun class systems. The project also discusses the basic clause structure, verbs and verbal modifiers, clause types, ideophones, particles and interjections. In order to firm up the language base in the project, a Logba-English-Ewe and English-Logba wordlists are provided. In the same vein, Bobuafor [6], worked on the grammar of Tafi, a minority GTM language spoken in the then Volta region of Ghana. The study examines the phonological system and established five syllable structures in Tafi as, V/N (vowel or nasal), VC, CV, CVV and a CCV. Unlike Logba, Tafi has three-level tonemes, High, Mid and Low and also an ATR harmony. Adpositions were also found in Logba. Unlike other Kwa languages like Akan in which prepositions are non-existent, Tafi has both prepositions (locative and comitative) and postpositions. But interestingly, Tafi is similar to Akan in declarative and interrogative sentence types. Their structure and segmental forms are the same except in tone (pitch). The study concludes with the translations of riddles, proverbs and some methods of cooking which adds to the prevention of glutophagy. In a more recent study, Abunya [11] documented some aspects of the grammar of Kaakye based on the functional typological framework. This follows after studies like Dakubu [4] and Adonae [12] in the phonology of Kaakye. Kaakye is a minority and less studied Kwa language of the Niger-Congo family. The study focused on the noun class system, animacy distinctions, serialization, relativization and complementation. Establishing the noun class system of Kaakye, six classes were discovered based on the semantics of the nouns. Again, Kaakye was discovered to have five different complementizers which are employed in two different strategies, the relativized and nominalized complementation strategies.

Theoretical Framework

Autosegmental phonology (henceforth AP) is a framework of phonological analysis proposed by John Goldsmith. As a theory of phonological representation, AP developed as a formal account of ideas that had been sketched in earlier works by several linguists, notably, Firth [13] and Hockett [14]. Twelve years after the introduction of Generative phonology by Chomsky and Halle [15], Goldsmith [16] propounded AP. According to Goldsmith, Generative phonology was propounded to account for the things that Chomsky and Halle left out in Generative phonology. Goldsmith stated that:

• The general framework of this “Autosegmental” approach is generative phonology and we are, therefore concerned with phonological rules that relate phonological and phonetic as much as with providing a formal characterization of these two levels [17]

By the quote above, Goldsmith meant AP does not veer off completely from that of his predecessors but rather serves as an alternative framework for Generative phonology. This alternative provides a link whereby the phonetic level is linked with the phonological level. A further statement firms up the fact that AP can account for both phonetic and phonological levels and show the relationship between these two levels. Per this explanation, Goldsmith presents the phonetic level, that is the sound level which accounts for sounds that were involved in a change and the phonological level also accounting for how a change took place, turning one sound into another. 

In AP, phonological representations consist of more than one linear sequence of segments; each linear sequence constitutes a separate tier. The co-registration of elements (or autosegments) on one tier with those on another and these are linked by association lines. There is a close relationship between analysis of segments into distinctive features and an autosegmental analysis; each feature in a language appears on exactly one tier. The phonological processes discussed in this work are palatalization, nasalization and vowel harmony on the segmental tiers. The working hypothesis of Autosegmental analysis is that, a large part of phonological generalizations can be interpreted as a restructuring or reorganization of the autosegments in a representation. The motivation for its application in this study is that, Goldsmith [16] and Goldsmith [17] used it to analyse Ibo, an African and a tone language like Ahanta and was able to account for the phonological changes. Above all, he succeeded in using AP to account for both segments and suprasegments. After his study on Ibo, many studies such as Akanlig-Pare [18], Bota [19], Obeng and Ollennu [20] have been successful in using AP to explain phonological processes in the following Ghanaian languages, Buli, Bono, Nkami and Mfantse, respectively.

As already stated, information from the pre-data collection period indicates that Ahantaland can be divided into three such as Urban Ahanta, Rural Ahanta and Valoe. The indigenes consider Rural Ahanta speakers as those who speak the original and least adulterated variety. In order to elicit for the original language, data was collected in towns in rural Ahanta. The selected towns were Funko, Ewusie Joe and Agona Nkwanta. Fifteen participants were selected in each town. These were towns where generational transmission was relatively high, so the age of participants was fixed from ten years and above. Recording and notetaking were the research instruments employed, therefore spontaneous speeches on many topics such as food processing, storytelling sessions and proverb duels, farming, trading activities and jokes were recorded in homes, work places, market, public transport and playing grounds. Notetaking also helped to ask follow-up questions for participants to repeat certain interesting utterances that would enhance the studies.

The data was listened to many times to take note of the phonological changes that are not captured in the written Ahanta, then transcribed and needed portions translated into English.

The sound system of Ahanta

Ahanta has remained an unwritten language until the year 2001 when a Senior Language Advisor and a linguist in conjunction with the Ghana Institute of Linguistics and Literacy and Bible Translation documented it. It was a similar work that I.K Mensah produced years before, with the title Ahanta/English that Ntumy drew inspiration from to document the sounds of Ahanta. 

Vowels of Ahanta

Kenstowicz [21] in discussion of vowels said that the main difference between consonants and vowels is the non-obstruction of airstream in the production of vowels. The Ahanta vowels are divided into orthographic and phonemic sounds.

Ahanta vowel orthographic and phonemic levels: There are seven vowels in Ahanta.

From the Figure 1, three of the vowels /a e o/ have two variants, that is, each represents two sounds at the phonetic level. /a/ is divided into /a æ/, e divides into /e ɪ/ while /o/ maps unto /o ʋ/. The rest of the sounds /ɛ i ɔ u/ are single, that is, it is the same at both orthographic and phonetic levels. From the above, Ahanta has ten phonetic vowels. These vowels can be found in the following Ahanta vowel table (Table 1).

Figure 1: Ahanta Vowel Orthographic and Phonemic Levels

Table 1: Ahanta vowel table

Distinction of Vowels

Vowel sounds, just like consonants can be produced with the velum raised to block the airstream from flowing through the nasal cavity in order to produce oral vowels. On the other hand, it is also possible for the velum to be somehow raised such that the air flows through both the nasal and oral cavities to produce nasalized vowels. In Ahanta, vowels are nasalized when they occur with nasal consonants. Vowels can also be nasalized independent of nasal consonants. This phenomenon is also common in related languages like Akan, Ewe and Ga. The ten vowels of Ahanta are all oral.

Nasalized Vowels in Ahanta

Table 2 contains the the nasalised vowels in Ahanta. Table 3 shows vowels which are nasalized in Ahanta.

Table 2: The Nasalised Vowels in Ahanta

In column 1 of Table 3 are the three vowels which are not originally nasal but become nasalized when found in the environment of nasal consonants. Consequently, /e/ which is not a nasal is but found in-between two nasal consonants in [m̀mè͂ɲâ] become nasalized.

Table 3: Vowels Which are Nasalized in Ahanta

Based on tongue-root features, either an advanced or a retracted tongue root, Ahanta vowels are classified into two as illustrated below:

• Set 1: / i e o u æ/ - Advanced tongue root (+ATR) vowels: With this set, the root of the tongue retracts in their production

• Set 2: /ɪ ʋ ɛ ɔ a/: Unadvanced tongue root (-ATR) vowels: In the production of these, the root of the tongue is pushed forward

Ahanta Consonants

Consonants are the sounds produced with obstruction in the oral cavity. As shown in Table 4, Ahanta has 27 consonants.

Table 4: Ahanta Consonants

Ahanta consonant distribution indicates that there are no word final nasal consonants in Ahanta. The only one identified is /n/ as in [ɔ̀sʋ͂̀ǹ] elephant. This may be due to the strict CV syllable structure in Ahanta. A study of Ahanta words reveals a consistent CV syllable structure. This was determined when the data was screened to check the structure of the words elicited. (Explain further) But, it must be noted that the word ɔ̀sʋ͂̀ǹ is a borrowed word from Mfantse.

Phonological Processes

A phonological process is a process by which speech sounds undergo changes effected by adjacent sounds. These processes help one to understand the changes that occur when sounds come together in speech. The processes discussed here are nasalization, palatalization, assimilation, consonant mutation and vowel harmony.

Nasalization in Ahanta-within words

The process of nasalization occurs when an oral vowel acquires the nasality of an adjacent nasal consonant (Table 5).

Table 5: Adjacent Nasal Consonant

Figure 2: The Process of Nasalization

The exemplify the process of nasalization this can be expressed as in the Figure 2.

From the Figure 3, each sound has its own tier, therefore there are the consonant and the vowel tiers. The assimilatory process starts from the tables a) on the left and ends at table b) or c) as the case may be. The oral vowels in the environment of nasal consonants, copy the nasality of the nasal consonants and the vowels consequently become nasalized, resulting in the following drawn from example 1:

The nasality can spread to the left or right of the nasal consonant as can be seen in example 3. Once the vowel becomes nasalized, the [-nasal] tier is cut off and the nasalized vowel/s is/are associated to the nasal tier. Some phonologists argue that the feature spreading is always regressive and so towards the left and that the nasalized vowel found on the right immediately after the nasal consonant is an independent nasalized vowel. As long as the word order is concerned, it emphasized here that the spreading can occur on both sides of the nasal. The only condition that needs to be met is that the vowels must be next to the nasal consonant as seen in all the examples.

Palatalization

Palatalization is a phonological process in which a consonant acquires a palatal feature due to occurrence with high and mid front vowels, /ɪ i e ɛ/. These vowels occur with labial, alveolar and velar sounds such as /b d n l k/. There are two types of palatalization namely partial and complete palatalization. In the partial one, it is in the pronunciation or the spoken aspect that the palatalization is realized. In the complete one, the non-palatal consonant that occurs with the high front vowel is turned into a palatal sound. In Ahanta, palatalization occurs in the following words:

The above examples are an indication of partial palatalization in Ahanta because the palatalization occurs in the spoken aspect of the language. Even though it cannot be proved now that there is complete palatalization in Ahanta, traces of it can be found in the spellings of some words in which high and mid front vowels follow palatal consonants. For instance:

Place Assimilation: homorganic nasal assimilation

In place assimilation, consonants copy the place of articulation of nasals they co-occur with. The copying is known as homorganic nasal assimilation, where an initial nasal acting as a prefix is attached to a word whose initial segment is a consonant. In the process of attaching the prefix, the nasal which is argued to be placeless, copies the place of articulation of the initial segment such that they become the same in terms of place of articulation. This type of copying can occur at word initial, word medial and between phrases. This occurs common in Ahanta and the Table 6 exemplifies homorganic nasal assimilation at word initial and medial.

Table 6: Homorganic nasal assimilation at word initial and medial

Table 7: The process from homorganic assimilation to mutation

In column 1 of the table, the phoneme /n/ which is an alveolar nasal is realized as a velar nasal when it occurs before velar sounds like /k, g/. With the first word, the alveolar nasal in pↄnkↄ ‘horse’ is realized as a velar nasal because it occurs before a velar /k/ within the word. In column 2, when the same alveolar nasal occurs before alveolar consonants like /t, s, z/, it remains an alveolar because they have the same place of articulation. Consequently, /n/ in the words ba-n-kye cassava, n-do-n-do-ne ‘mosquito,’ kwanzerɛ ‘hawk’, ni-n-si-n-ii ‘herbalist’remains the same. The process of homorganic nasal assimilation continues in column three when the bilabial /m/ remains the same before bilabial sounds such as /b/ in atwom-b-ɛne ‘horn’. But changes to a labiodental nasal /ɱ/ when it precedes the labiodental fricative /v/ such that, kuro-m-via is realized as [kùrò-ɱ-víà] egg.

Consonant Mutation in Plural Preffixation

Consonant mutation is a phonological feature of several languages. There are different kinds of mutations. Stem-final consonant occurs in Dholuo, spoken in Kenya consonants. Consonant initial, medial and final mutations occur in Modern Hebrew while Japanese has word medial consonant mutation which co-occurs with voicing. There is evidence of initial consonant mutation type in Ahanta and also in Akan, a related language and the West African language Sere and Wolof spoken in Senegal and the Gambia. The type in Ahanta co-occurs with voicing, similar to Japanese. The initial consonant mutation occurs when the segment acting as a prefix copies the place of articulation and voicing features of the initial segment or phoneme. Mutation is a morphosyntactic phenomenon in that, the addition of the prefix to the word and where it occurs influence the change in the form of the word. The plural marker in Ahanta is a placeless nasal but changes according to the initial consonant of the word it is attached to. See the examples in Table 9 where the process from homorganic assimilation to mutation is shown.

The Table 7 has seven columns. The first contains the word in the autography followed by its gloss in English. Then, the plural marker in its bare form (considered placeless) and this also is followed by the transcribed or phonetic word. In the column named ‘homorganic’ is how the word stands when it undergoes homorganic assimilation (see section 1.6.4). In the next column, the mutated word is presented with its English gloss in the last column. For example, the word for ‘big’ in Ahanta is kezire and its plural prefix is /n/, a placeless plural marker. When n is added to kezire, the n according to the initial consonant assimilation rule copies the place of articulation of the phoneme /k/ which is a voiceless velar plosive. The alveolar nasal then becomes /ŋ/, the velar nasal. At this point, the assimilation is known as homorganic. But the process where the velar nasal moves on again to copy the exact place of articulation and voicing of the initial consonant so that that they up the same, is known as initial consonant mutation This description runs through all the examples.

Note: Example 10 involves a vowel as a prefix but qualifies as mutation when the voiceless labio-dental fricative as a prefix has been changed to voiced labio-dental fricative to agree in voicing with the segment /u/ in voicing. Some examples are expressed in the following Autosegmental Figure 4. I didn’t mark nasalization here.

Figure 3: Consonant and the Vowel Tiers

Figure 4: The segment /u/ in voicing

Figure 5: Feeding Rule

Mutation rule

There are two phonological rules involved in the mutation. The first is the voicing copying as seen in rule 1. The output of this rule becomes the input for rule 2, an example of a feeding rule. This in rule 3 is an expansion of rule 2. And both rule 1 and 2 when collapsed result in rule 4 (Figure 5a-c).

In Figure 5 example, only the part of the process, which is the voicing assimilation is achieved because the prefix which is the plural marker agrees in the feature, place [-cor] with the root.

Vowel Harmony

Vowel harmony is a type of vowel assimilation found across languages worldwide including African and Ghanaian languages and Ahanta. In this process only a class of vowels occur in a given word, across morphemes and phrase boundaries. In the literature, languages tend to classify their vowels into two sets and members in each set tend to occur exclusively in most words. Even when members from the two sets co-occur, some cavity tendencies change all the vowels into members of the same set. This cavity tendencies are called vowel harmony processes. However, it must be noted that there are unique cases where vowels of the two sets co-occur. The following words in example 10 show vowel harmony as it occurs within words (Figure 6).

Figure 6: Vowel Harmony

In Table 8, all the vowels in each of the ten words are +ATR vowels. But in the words in Table 9, all the words contain -ATR vowels only.

Table 8: Words with only +ATR vowels

The Table 9 also affirm the vowel harmony rule in Ahanta because only vowels from -ATR are in the words listed.

Table 9: Words contain -ATR vowels

Vowel Harmony Across morphemes 

In the Ahanta language, the future marker is kɔ and the verb eat is ri. In the construction of mikeri ‘I will eat’ which is a combination of three morphemes and in other words a sentence, the pronoun, the tense marker and the verb are written together as one word. Due to the +ATR feature of the vowel /i/ in the verb ‘eat’, the unadvanced vowel /[ɛ]/ in the future aspect marker and the unadvanced vowel in the first person pronoun me ‘I’ become advanced (+ATR). There is also an evidence of rounding harmony where the rounded vowel [ↄ] in kↄ loses its roundness to become a -Round vowel /e/. Therefore, mek]ri is derived as mekↄri. However, in example 5 wokeri ‘you will eat’ in table 12 shows that the rounding harmony only affect the vowel in the future marker but not the one in the pronoun. For the sake of this discussion, an inventory of the underlying forms of Ahanta pronouns have been presented (Table 10-11).

Figure 7: The Autosegmental Analysis of Rounding Harmony Across Phrases

Table 10: Ahanta Pronouns

Table 11:

-ATR harmony across morphemes. The verb here is ba ‘come’ used with the habitual positive to make a word in Ahanta but not in English. In all the examples in 13, the vowels assimilate the -ATR feature of the vowel in ba, the unadvanced central low vowel, in a regressive assimilation. It can also be deduced from the examples 11, 12 and 13 that the second, fourth and fifth pronouns change forms (Table 12).

Table 12:

Rounding Harmony 

Concerning vowels, Dolphyne [22], Abaka [23] confirm that there is also rounding harmony in Fante aside the usual ATR harmony. Obeng [24] have proved that the ATR harmony occurs together with the rounding harmony with the former occurring before the latter. In rounding harmony, all the vowels in a word apart from having +/-ATR features, may be only back vowels. In spite of a word having front vowels when it occurs with words that have back vowels, it copies the roundness of the vowels. This can occur in words, compounds, across word boundaries and phrases. The following are some examples in Ahanta (Table 13).

Table 13:

Vowel Harmony within words

The Table 14 shows rounding harmony across phrases. In the examples in 23, the phrasal verb tʋ ndwomi ‘sing’ has been used with the habitual tense. In Ahanta, this tense has no special marker as in some of the tenses in Ahanta. The tense is marked by tone. The rounding harmony spreads regressively from the back vowel of the first part of the phrasal verb, tʋ to the vowel in the pronoun. This is evident in the first person singular and third persons. However, in Mfantse, rounding harmony occurs with the first and second person singular only [24] (Table 14).  

Table 14: Examples of vowel harmony in Ahanta

Rounding harmony has been exemplified in the in the examples above. The first-person pronoun me has a -ATR vowel therefore vowel harmony is complete here. Firstly, the apical coronal consonants to which /t d/ belong are high sounds like the vowel high back vowel /ʋ/ therefore, /ʋ/ and /t/ are similar in height but not in voicing, Walker et al. [25]. There is also an occurrence of coronal harmony in /t d/ and the vowel /ʋ/, [26-30]. Highness, the common place of articulation for /ʋ/ and/t/ is also conducive for mutation. There is therefore an occurrence of partial mutation in that, the alveola voiceless plosive /t/ has copied the voicing feature regressively from the rounded high back vowel /ʋ/. This assertion is based on the fact that /t/ already occurs with a voiced sound /ʋ/ but remains voiceless when in isolation until attached to mʋ in mʋtʋ ndwomi tobecome mʋdʋ ndwomi. The voicing feature might not have come from the initial bilabial nasal /m/ because the nasality might have also docked on the /t/ to turn it into a nasal consonant. This would have made the phrase *mʋmʋ ndwomi. This is best described as partial mutation since the sounds involved are a vowel and a consonant.

This paper opens with the background of the Ahanta language and and its speakers. Due to a subsequent discussion on some phonological processes in Ahanta, the sound inventory and the pronouns of Ahanta are also discussed. Ahanta until recently was an unwritten language. One linguist, Ntumy [9] has managed to document it but very scantily. Ahanta consonants are different from Akan in only one sound, the velar fricative /x ɣ/. The vowels of Ahanta are the same as that of Akan. Some of the assimilatory processes discussed are palatalization, vowel harmony, rounding harmony and consonant mutation. The Autosegmental theory is advanced as examples from each assimilatory process is successfully analyzed in the framework without exceptions. Ahanta is also found to exhibit coronal harmony between vowels and consonant with the common place articulation feature of height. This is hoped to be explored further in subsequent studies. The paper terms the type of mutation that occurs between a vowel and consonant partial mutation in that, it is not the place of articulation, the nasal feature that are copied but it the place of articulation and voicing. In the usual initial consonant mutation, the sounds involved are nasals which already bear the feature voiced. As an understudied minority language, it is recommended that the established phonological processes be explored by other phonology researchers.

1. Hoffman, M. Endangered Languages, Linguistics, and Culture: Researching and Reviving the Unami Language of the Lenape. Bryn Mawr College, 2009.

2. Christaller, J.G. A Grammar of the Asante and Fante Languages Called Tshi. Basel Evangelical Missionary Society, 1875.

3. Westermann, D.H. Grammatik der Ewe-Sprache. Diedrich Reimer, 1909.

4. Dakubu, Mary Esther Kropp, Ed. The Languages of Ghana. Kegan Paul, 1988.

5. Dorvlo, K. A Grammar of Logba (Ikpani). PhD Dissertation, Leiden University, 2008.

6. Bobuafor, M. A Grammar of Tafi. PhD Dissertation, Leiden University, 2013.

7. Asante, Krobea Rogers. Nkami Language: Description and Analysis. PhD Dissertation, Tongji University, 2016.

8. Campbell, A. A Grammar of Ga. PhD Dissertation, Rice University, 2017.

9. Ntumy, S.K. Ahanta Without Tears. Ghana Institute of Linguistics, Literacy and Bible Translation, 2001.

10. GILLBT. Ahanta Bible. GILLBT Publication, 2014.

11. Abunya, L. Aspects of Kaakyi Grammar. University of Ghana, 2018.

12. Adonae, D. Kaakye Tonology. MPhil Thesis, University of Ghana, 2005.

13. Firth, J.R. "Sounds and Prosodies." Transactions of the Philological Society, March 1948, https://doi.org/10. 1111/j.1467-968X.1948.tb00556.x.

14. Hockett, C. A Manual of Phonology. Indiana University Publications in Anthropology and Linguistics, 1955.

15. Chomsky, N., and M. Halle. The Sound Pattern of English. Harper & Row, 1968.

16. Goldsmith, J. Autosegmental Phonology. PhD Dissertation, Massachusetts Institute of Technology, 1976.

17. Goldsmith, J. Autosegmental and Metrical Phonology. Basil Blackwell, 1990.

18. Akanlig-Pare, G. Buli Tonology. University of Ghana, 1996.

19. Bota, G. Some Aspects of Bono Phonology. University of Ghana, 2002.

20. Obeng, P. and Y. Ollennu. "Rounding Harmony in Mfantse: Evidence and Differences." 2020.

21. Kenstowicz, M. Phonology in Generative Grammar. Blackwell Publishing, 1994.

22. Dolphyne, F.A. The Akan (Twi-Fante) Language: Its Sound Systems and Tonal Structure. Ghana Universities Press, 1988.

23. Abaka, E.N. The Segmental and Tone Melodies of Akan. Norwegian University of Science and Technology, 2003.

24. Obeng, P. Fante Phonology: Dialectal Differences. University of Ghana, 2008.

25. Walker, R.D.B. et al. "An Articulatory View of Kinyarwanda Coronal Harmony." Phonology, vol. 25, 2008, pp. 499–535. Cambridge University Press, doi:10.1017/S0952675708001619.

26. Clements, G.N. "Representational Economy in Constraint-Based Phonology." Distinctive Feature Theory, edited by T. Alan Hall, Mouton de Gruyter, 2001, pp. 71–146.

27. Hansson, Gunnar Ólafur. Theoretical and Typological Issues in Consonant Harmony. PhD Dissertation, University of California, Berkeley, 2001.

28. Rose, S., and R. Walker. "A Typology of Consonant Agreement as Correspondence." Language, vol. 80, 2004, pp. 475–531.

29. McCarthy, J.J. "Consonant Harmony via Correspondence: Evidence from Chumash." In Leah Bateman, 2007.

30. Hume, Elizabeth. "Coronal Consonant, Front Vowel Parallels in Maltese." Natural Language & Linguistic Theory, vol. 14, no. 1, 1996, pp. 163–203. JSTOR, http:// www.jstor.org/stable/4047882. Accessed April 2021.