Sekyi-Baidoo, Yaw

=The Akan Phrasal Verb as a Syntactic Manifestation=

Yaw Sekyi-Baidoo University of Education, Winneba

Abstract
 * The phrasal verb in Akan has not been given much attention in Akan linguistics. Whereas some authoritative research and publications ignore it, others confuse it with other verbal compounds in Akan’ or classify a great number of them as non-transparent or opaque serial verbs. This paper intends at drawing attention to the Akan phrasal verb as a syntactic manifestation distinguishing it from other verbal compounds. Having established its distinct identity, the paper discusses the phrasal verb in Akan looking at its basic constitution, types, and syntactic functions.

Introduction
The phenomenon of compound verbal forms is quite remarkable in Akan. In addition to phrasal verbs and ordinary verb + adposition (preposition/postposition) combinations, there are also compound verbs which combine otherwise independent lexical verbs, and also fixed verb and noun combinations which function basically as idioms. This paper, however, focuses on the phrasal verb in Akan, and from a basically syntactic point of view. It looks at its types and constitution, how it responds to the question of transitivity and also to its syntactic functions in Akan. This study is in response to an apparent dearth of information on the phenomenon of phrasal verbs in Akan, and it aims at drawing attention to certain peculiarities about the Akan phrasal verbs which enhance their identification. Even before we embark upon this study, it is necessary to caution that our use of the English phrasal verb as the framework for this analysis is not to subject the Akan phrasal verb forcibly to the parameters of the English phrasal verb. Where explicit comparisons occur, they are made with the aim of defining the Akan phrasal verb as an analogous but independent linguistic structure from the English forms. Again, the use of the English parameters is to establish the universal linguistic frame under which such verbs forms are classified in English as well as in Akan. That much study has not been conducted into the Akan phrasal verb is evident when one looks at authoritative researches on the Akan verb, word, or syntactic structure as by Christaller (1876/1964) Dolphyne (1969) Andoh-Kumi (1985 Akrofi (1958) Even where attention is drawn to the phenomenon as basically a verb and adposition element combination, as in Christaller (1876) and Akrofi (1958) not much has been done in terms of discussing it under such indispensable syntactic categories as transitivity. Christaller (1876), for instance studies phrasal verbs within the general framework of the use of prepositions and postpositions (adposition elements) in Akan, and specifically in the discussion of verb and adposition combinations.  He distinguishes explicitly between verbs with normal, optional use of adposition elements, prepositional verbs as well as phrasal verbs.  He also acknowledges the idiomaticity or opacity of the phrasal verb as a semantic entity, and other verbal compounds such as the serial verb. He draws attention to the differences between the surface meaning of the compound as a combination of semantically independent units and the compound as a single semantic entity, especially in the context of the phrasal verb. It is evident in this scenario that research done so far into the phenomenon of phrasal verbs in Akan come nowhere near establishing it as a distinctive linguistic manifestation. Consequently, only a limited number of phrasal verbs have been mentioned or documented. Additionally, the various syntactic manifestations of the phrasal verb have not been adequately studied. It is in response to this situation in Akan linguistic studies that this research has been undertaken, first, to identify the Akan phrasal verbal as separate and independent from the other verbal compounds, especially serial verbs and ordinary, non-idiomatic verb + adposition combinations; and second, to discover the various manifestations of the phrasal verb in respect of the principles of transitivity, complementation or collocation by which they are differentiated. It also looks at their syntactic functions. It is hoped also that this work would help to enhance the repertory of documented phrasal verb in Akan Linguistics.

Other Verbal Compounds in Akan
Before any analysis is carried into the Akan phrasal verb itself, it is necessary to differentiate it from other compound verbal structures in Akan. These others are serial verbs, verb and demonstrative compounds, verb and noun compounds, and the non-idiomatic verb and adposition combinations. The Akan serial verb is generally a ‘verb + verb’ combination which subtly captures otherwise distinct activities as one. Two of its primary manifestations are worth noting. These are the process and the intensity/direction types. The process type is exemplified by such verbs as “fabra” (bring) and “fato” (put). These verbs tend to capture the two component actions involved in a single endeavour. “Fabra” (bring) as an action comprises “fa” (take) and “bra” (come) lit. “take and come”. “Bringing” evidently involves taking something and moving towards a destination usually associated with the encoder. Similarly, “fa to” (put) captures the sequence of taking (“fa”) and putting (“to”). The intensity/direction compound verb form is associated with the use of “ba” (coming) and “kכ” (going) as a suffix to verb. When thus used, “kכ” especially may signify the intensity or direction of the activity, to which it is attached. For instance, “didikכ” and “kakכ” mean “go on eating” and “go on saying” respectively. It refers to the continuity, intensification or relentlessness of an action. On the other hand, “kכ” or “bra” may be used for adjunct purposes to signify the place or direction. Examples of compounds with such use of “bra” and “kכ” include “do bra” (weed approaching) “dכkכ” (weed way). The direction ones could still be modified by a place adjunct which specifies the destination of the movement. The adjunct here may be a prepositional adverb such “nkyεn” (aside) in “dכkכ nkyεn” (weed directing away); demonstrative “ha” (here) as in “dכbra ha” (weed approaching here); or nominal adjuncts such as “fie” (home), as in “dכbra fie” (weed approaching home). These are, however, different from phrasal verb basically because they do not use position elements as integral elements of the structure. Again, they do not have the idiomaticity associated with phrasal verbs.

Other researches also classify Akan serial verbs (and serial verbs of other languages) into the modifying or single-event and the linking or multiple-event types, (Bamgbose, 1974; Givon, 1991). Whatever the classification, one thing is clear and sufficient for our present discussion that serial verbs are verbal compounds consisting of two or more verbs (other than auxiliary forms or verbal particles) which occur within a single non-complex construction and which exhibit no overt signs of coordination. Some of them are idiomatic like the phrasal verbs as established by Bamgbose (1974).The Akan Teleteaching Course (2004) also establishes the idiomaticity of some phrasal verbs using the example of the two component verbs of the serial compound ‘gyedi’ ‘ - gye’ (receive/take) and ‘di’ (eat) -  which interpret as ‘believe:


 * In other cases, however, the composite meaning of the serial verb construction cannot be made transparent by considering the meanings of the individual verbs of which it is made up…the interpretation…. Is an idiomatic one, which cannot be derived in any obvious way by combining the meanings ‘receive’ and ‘eat’ which are usually assigned to the verbs when they occur by themselves.

Whereas such serial verbs share the idiomaticity of phrasal verbs, their constitution basically differs from the phrasal verb which necessarily has an adposition element. Verb + noun compounds in Akan are usually euphemistic and therefore share the idiomaticity of phrasal verbs. The constituents may be orthographically or phonetically separate, but semantically, they are conceived of as compounds which make a single meaning which is associated not with the individual elements but with the compounds as a whole. Examples are “kכ akuraa” (Lit. go to the village) – to die; and “ka baabi” (lit. to be locked up somewhere) – to die. They are also different from phrasal verbs because do not use the adposition elements associated with phrasal verbs. There are also compounds comprising verb and demonstrative. These, like phrasal verbs, are also idiomatic. Such compounds as “hyε hכ”, “wɔhכ”, (to be available or present); “ka hכ” (to be/become an incorrigible character trait) and “bכ hכ” (to exist in spite of any possible opposition or problem). Closest to the phrasal verb is the verb + adposition combination, and sometimes the structures for these two forms are the same. The difference is in the fact that the phrasal verb is a single lexicographical and semantic unit, and the constituent elements are therefore not seen to be of any clear semantic independence or contribution as far as their place in this idiomatic structure is concerned. Meaning is associated with the compound as a single unit of meaning and any attempt to change the constituent words changes the identity of the idiom and the meaning associated with the form. Such compounds include ‘kכ nkyɛn’ (Go aside/Go to the side of), ‘bra mu’ (come in) and ‘gyina NP so’ (Stand on NP). In such structures, each of the words makes a distinct semantic contribution to the meaning of the compound. They are, thus, not idiomatic.

Defining the Akan Phrasal Verb
Phrasal verbs, as indicated above, are basically verb and adposition element combinations which bear a single semantic significance which is different from the meaning of ordinary verb and adposition combination in which the meaning is to some extent the aggregate of the meanings of the constituent words. Phrasal verbs are, thus, basically idioms and they therefore have some degree of opacity. One other major characteristic of phrasal verbs is the stability or rigidity of their constitution. This means change of elements of the structure results in a change in meaning, semantic absurdity or in vaguity. Again, sometimes, the adposition element does not admit a complementation of the object unlike in the ordinary verb + position compound in which there is necessarily an object complementation. In such instances, position element  is referred to as a particle. For instance, ‘so’ (on) ‘ase’ (down) and ‘mu’ (in) in ‘gyina so’ (to be in top form); hwe ase (to fall down) and ‘ba mu’ (to come to pass/to manifest) respectively are particles, because they, as used here, are not supposed to be complemented. As stated above, the basic difference between the verb + adposition compound and the phrasal verb is the issue of transparency or idiomaticity respectively. There could be differences in specific phrasal verbs and their non-idiomatic forms in respect of the valency (of verb and adposition element) as well as in the nature of complementation. But these issues are peripheral as far as the basic classification of the two verbal compounds are concerned. We, thus, distinguish one use of a combination of verb and adposition element as a phrasal verb and the other as not, because that which we describe as a phrasal verb exhibits a basic meaning which deviates from the meaning which will ordinarily accrue to the compound if the meaning of its constituents is to be considered, and these are illustrated below. ‘Da so’ non-idiomatically means ‘lie on’. As a phrasal verb it means ‘to persist’ (or do something persistently) or ‘to manifest as expected’. Certainly, ‘da’ (to lie) and ‘’so’ (on) have nothing to do with these two meanings of the phrasal verb ‘da so’. Again, ‘te ase’ as a non-phrasal verb form means (to tear under) being a combination of ‘te’ (to tear) and ‘ase’ (under/below). As a phrasal verb, it interprets as ‘to live/be alive’ or ‘to understand’, and these meanings have nothing to do with the meaning of ‘te (tear) and ‘ase’ (under) put together.

The Adposition Element
It will be necessary, at this juncture, to discuss the nature and workings of the adposition element in the Akan. Adposition elements in Akan can be grouped into the ordinary and the verbal according to the grammatical form with which the meaning of the adposition element is associated; and also into preposition and postposition in respect of the placement of the adposition in relation to the noun to which it expresses relationship. It is possible to distinguish between adposition elements such as ‘so’ (on); ‘anim’ (in front of/before) and ‘nkyεn’ (beside/at the side of) which are ingressives (Christaller 1876) from others such as ‘kodu’/’kosi’ (to/up to) and ‘tra’ (beyond). Items of both groups, as we shall see below, are not primary semantic elements. Rather, they are secondary morphosyntactic units since they derive their meaning from senses usually associated with verbs, hence their description as verbal. The prepositions may be seen, for instance, as polysemic or metaphorical applications of the original meaning of verb forms. They capture situations of movement or dynamism, which is associated with verbs. This is seen in the fact that semantically, they are associated with the idea of movement to which the noun complementation serves as destination. ‘Kodu’ ‘twa mu’, ‘tra’ can, for instance, are ordinarily verbs which mean ‘reach’, ‘pass’, and ‘go beyond’ respectively. In ‘Etwa mu mfie mmienu’ (It is/goes beyond two years), the movement is associated with time which is pictured as going beyond the temporal destination of ‘two years’. ‘Kodu’ especially is even a serial verb made up of the verbs kɔ(go) and ‘du’ (reach). It presents one of the special cases in which a serial verb construction serves morphosyntactically as an adposition element. That adpositions relate to other parts of speech as we have intimated above has been acknowledged by a number of researchers. Payne (1997: 87) for instance, has this to say Adpositions derive historically from nouns or verbs. For some languages, 	particularly those that employ serial verbs extensively, it may be difficult to 	decide whether a given form is an adposition or a dependent verb.

Similarly, Cascad (1982) and Heine and Ren (1984) acknowledges that many adpositions derive from body-part nouns. For instance, the word for the body-part ‘back’ is used for locative preposition ‘behind’, and that for ‘face’ for the locative preposition ‘front’. The other categorization between adposition elements in Akan looks at them in terms of their place in relation to the noun which complements it. Unlike in English, the adposition element does not always come before the noun. Those that come after the nouns such as ‘akyi’ (behind) as in ‘Kwasi akyi’ and ‘mu’ (in/inside it), as in ‘nkwan mu’ (in the soup) are postpositions – i.e after the noun. Preposition, on the other hand, come before the noun complement. Examples are ‘kosi’ lit. ‘go stop at’ (to, up to) as in ‘kosi Kumasi’ (up to Kumasi); and ‘tra’ as in ‘tra afe’ (beyond a year). The use of pre (before) and post (after) indicates their positions in relation to the associated verbs. The issue of position here, unlike in English, is not a matter of choice or the arrangement of syntactic items in speech. Their position, before or after the objects, is permanent or fixed. The verbal and ordinary/non verbal distinction of adposition elements seem to coincide with the postposition/preposition dichotomy. Verbal adpositions are generally also the prepositions. These distinctions seem to tie in neatly with the semantic distinction of the referential and the directional. Referential positions are also the non-verbal ones. Here, the complement which comes before the adposition is used as a reference point from which the spatio/temporal location denoted by the adposition is determined. Examples are ‘akyi’ (behind) in ‘Kwasi’ akyi’, (behind Kwasi) and ‘anim’ (in front of) as in ‘Efie no anim’(in front o f the house. Referential adpositions have a nominal outlook since they are derived from nouns, and the complement which comes before them may be seen as a pre-modification. This relationship is sometimes captured by turning the phrase into a genitive: Kwasi’s back/behind		(Kwasi akyi) Kwasi POSS back

The house’s front		(Efie no anim) DEF house POSS front

Ama’s side			(Ama nkyεn) Ama POSS side The genitive role of the complementation is evident in the questions which focus on it: Hwan akyi? (whose Behind) POSS-INTERR back Hwan anim? (whose front?) POSS-INTERR front Hwan nkyεn? (Whose side) POSS-INTERR side and the referential position is signified by the adjunctive postpositions ‘akyi’, ‘anim’ and ‘nkyƐn’ as in the genitive forms above. Thus, the adjunct ‘behind’ (akyi) in front (‘anim’) is identified by reference to the noun, hence their name referential adpositions. Directional/destinational adpositions, on the other hand, are ‘‘the verbal’ and also ‘the prepositional’. The complement is seen as the destination or source of the movement associated with the adposition element, which is associated with some activity of movement. The adposition element is also seen as a link between one position (explicit or implicit) and the other which is the complement. ‘Ba fie’ (up to the house) and ‘Tra fie’ (beyond the house) both show some activity of movement associated with the adposition element and the idea of destination  or source which is associated with the complement. Having discussed the nature and workings of the adposition element in Akan, we would now go into the discussion of the Akan phrasal verb looking at its constitution, and especially its transitivity. Transitivity is being addressed here for three main reasons. First, it is sometimes used to distinguish between phrasal verbs and other verb + adpositions compounds. Second, the nature of complementation helps sometimes in differentiating between polysemous and homonymous phrasal verbs. Finally, discussion of transitivity helps us in analysing the form or structure of the phrasal verb and also in the description of the adposition element whether as a real adposition or as a particle. Transitivity is not being used here in the wide sense of the general system of participant relationships in a clause as seen by Halliday (1967) as the various kinds of processes and the participants who are associated with each other. Transitivity is being used here in the sense of the traditional transitive/intransitive or extensive/intensive verb opposition as advanced in the works of Quirk et al (1985). Cobuild (1990:137– 8) for instance has this distinction to make: If an action or event involves only one person or thing, you mention only the performer of the action (the subject) and the action (the verb)… clauses of this kind are called intransitive verb…  If the action or event involves another person or thing which the action affect, relates to; or produces, you put a noun group referring to them after the verb group. This is called the object of the verb or clause… Clauses which have direct objects are called transitive clauses, and the verbs which occur in transitive clauses are called transitive verbs.…a small number of transitive verbs also allow you to mention who benefits from n action or receive something as a result… verbs which can take an indirect object as well as a direct object re called di-transitive verbs.

Transitivity is thus used here in the sense of the capability or necessity of verb to take a object which according to Quirk et al (1985) is ‘the sufferer’ or the recipient of the verb. Better still, the object may be seen as a noun or nominal other than the subject whose involvement makes the action possible or realizable. This question of ‘involvement’ and ‘possibility’ has been adopted here as result of the semantic barrier sometimes associated with the use of ‘sufferer’ ad ‘recipient’. The Akan phrasal verb has three forms on the basis of transitivity, and these are the intransitive, the mono transitive, and the di-transitive. No complex transitive form has not yet been identified.

Intransitive Phrasal Verbs There are a number of intransitive phrasal verbs in Akan. Since they do not carry objects, we regard the position forms as adverbial particles. Intransitive phrasal verbs can be grouped into the copulative and the active. The copulative ones only describe the subject. Some of these are: ba mu		 (lit. come in)		to be manifest gyina so	(lit. stand on)		to be in top form kyere so	(lit. tie on)		to be hard ( condition/life) da so		(lit. lie on)		to be on course/persist hyε mu		(lit. pack/fill in/inside)	to be in appreciable form ba so		(lit. come on)		to be proud Notable with this group of intransitive verbs as with all copulative and perceptual verbs is that they cannot be used in imperatives. Others that are perceptual but not copulative are: brε ase		(lit. tire 	under)		to depreciate kכ so		(lit. go on)		to progress, to continue in existence, function ma so		(lit. lift/give on)		to rise (dough) ba so		(lit. come on)		to take place/to be in vogue mia mu		(lit. squeeze in/inside)	to intensify.

Other intransitive phrasal verbs are ergatives, or are used ergatively. This means that the actual object (the noun which other than the subject makes the action or its description possible) features syntactically as the subject; and the action involved is presented as a state, or being, or described as the subject. Ergative phrasal verbs in Akan include ‘firi ase/hyε ase’ (lit. come from below/be under) meaning ‘to start’ as in Adwuma no ahyε aseε/afiri aseε. (The work has started.) Work DEF PERF BE under/come from below. ‘twa mu’ (lit. cut between) meaning ‘to come to end/to be stopped) as in 	Mbra no atwa mu. (the law is no longer in force)       Law DEF PERF pass/cut between and ‘toa so’ (to continue) as in 	Nkommo no atoa so. (The conversation has continued.).       Conversation DEF PERF join on

It should be noted that these verbs also have non-ergative uses which are not descriptive of the subjects. Again, there are active intransitive phrasal verbs. These verbs can be realized in the imperative unlike those discussed before. Some of such verbs are:

tia so		(lit. step on) 		to speed off ka mu		(lit. to bite in/inside)	to scream ma so		(lit. give/lift on)		to dress elegantly hwe ase		(lit. flog/fall under)	to fall down

Mono-Transitive Phrasal Verbs The Akan phrasal verb has mono-transitive forms. As verbal compounds, they take single objects. There are two kinds of the mono-transitive. These are (i) those which take nouns as objects ad (ii) those which take verbs as objects. It not always very easy to determine whether the transitivity is associated with the verb item or the adposition item. Whether the object is nominal or verbal may determine its placement with the verb item or the adposition element. It must be understood however that since the phrasal verb is regarded as a single syntactic ad semantic unit, the question of whether the object is associated with the verb item or the adposition is not very significant. (i)	With nouns as objects Transitive phrasal verbs with nouns as object usually have the structure Verb + NP +adposition where the NP is a noun or nominal. Examples are: kכ NP so	(lit. go on NP) to attack NP, to continue with NP firi/hyε NP ase	(lit. come from/to begin NP) to work under NP bכ NP so	(lit. hit NP top/on) to make mention of NP, to give a help hand 	to NP

kyerε NP ase	(lit. show/teach NP under/below)	to explain to NP (ii)	With verbs as objects Phrasal verbs in this category may be also used in the previous category. The difference in terms of their constitution is seen in the fact that those in this category use verbs as objects, and again, the verbal object is placed after the adposition element. The structure for this group of phrasal verb is Verb +adposition + object (verbal) and the following are examples: hyε ase yε dwuma (no)		begin working/begin doing the work BE under DO work DEF firi ase sa			begin dancing COME FROM under DANCE kכ so didi			go on eating GO on EATING

It might be observed from the above examples that the use of gerund forms in the English translations of phrasal verbs with noun as in ‘dancing’ and ‘eating’ does not make the nominal/verbal object dichotomy in Akan vivid enough. In Akan itself, the use of the nominalising prefix ‘a’ makes the difference between the two forms quite explicit. Consider ‘didi’ (eat) – verb - and ‘adidie’ (eating) –nominal;  and ‘sa’ (dance) – verb - and ‘asa’ (dancing) – nominal. A proper understanding must therefore focus on the Akan forms instead on the English translation. One semantic difference associated with these two kinds of structures and objects is that the verb + NP (noun) + adposition structure can be used to make the object more explicit. This is because there is a great deal of variety associated with the nature of the object. Nouns associated with verb or activities (deverbal forms) such as ‘asa’ (dancing) and ‘adidie’ (eating) can be used. In addition, more specific form such as ‘aduane’ (food) and ‘nwoma’ (book) can also be used as illustrated below: Toa aduane no so	(continue with the food) JOIN food DEF on

Hyɛ mmo no ase	(start with the rice) BE rice DEF under

Kכ nnwoma no so	(continue with the book). GO book Def on

In these examples, the nature of noun which the activity in the verb affects is quite explicit. Even where nouns derived from verbs are used, as in ‘adidie’ (eating) and ‘asa’ (dancing), they could still be made explicit by modifying them with the name or description of the actual noun involved. For instance ‘asa’ and ‘adidie’ can be pre-modified to become ‘adowa asa’ (adowa dancing/dance adowa) as in Hyε adowa asa no ase. (Start the adowa dance/Start dancing Adowa) BE adowa dance DEF under On the other hand, the verbal objects as in the verb + adposition + object (verb) structure of the transitive phrasal verb makes the required action more explicit than what the verb + NP (noun) + position  structure does. Compare: Kכ nnwoma no so	Go on with the book GO book DEF on

Kכ so kenkan		Go on reading GO on READ

It is evident that (a) specifies the object involved, whereas (b) makes the required action explicit. The action in the verb + adposition + NP (verb) form may be made more explicit by realizing the object of the action or an adjunct/adverb, as in

Kכ so kenkan nwoma no. Go on reading the book. GO on READ book DEF Hyε ase di mmo no. Start eating the rice. BE under EAT rice DEF or Toa so didi ntεmtεm	     Continue eating quickly. JOIN on EAT fast The structures here are: verb + adposition + OBJ (verbal) +  NP Adv. P 1.  Kכ     	 so	      kenkan  	 nwoma no (Go on reading the book.) GO-IMP   on 	      READ	 book DEF 2.   Kɔ       	 so           	      didi	  	 ntεmtεm GO-IMP   on	      EAT		 fast

It is evident from the examples above that the verbal object this structure can be used intransitively, and can also be used with adverbs whether in its transitive or intransitive form. It must be understood that the object and the adverb involved here are not complementing the phrasal verbs but rather the verbal objects which complement the adposition elements. Their association with the phrasal verb is, thus, not direct. For explicit action, the object and the adverb could be put together giving such a structure as: Verb + prep   + verb      + 	NP (object)  + 	adverb kכ	so	kenkan		nwoma	 no	 ntεmtεm. GO       on        READ		book DEF (Go	on	reading		the book 	 fast.)

Reciprocal Phrasal Verbs Reciprocal verbs describe actions or processes which involve two groups of people or nouns. In English, some of these verbs are ‘argue’ ‘compete’ and ‘meet’. In English, one of these nouns can be brought in post verb position in active constructions as object as below where ‘Noah’ is the object, but in actual fact, both are subject and object at the same time: Jane married Noah. With Akan phrasal verbs, however, this transposition is not possible. Reciprocal phrasal verbs always take plural or compound subjects as with the following forms verbs: ‘(ka) bo mu’ (to unite) and‘hyia (mu)’, (to meet):

Kwasi ne  Kwame  bככ mu. Kwasi and Kwame united. (compound subject) Kwasi and Kwame hit  in

Yeatwa mu. We have separated. (plural subject) 1PL cut in/across

In ‘kabכ mu’ (to unite), however, we can realize both subject (compound or plural) and object where the subject of the action is different from the two nouns being united, and the reciprocity is associated not with the subject and the objects, but with the constituents of the compound object. . Here, ‘kabכ, (Lit. touch-hit) is seen as an idiomatic serial verb. The subject is placed before ‘ka’ whilst the united nouns, being objects, are placed between the two items of the serial verb, ‘ka’ and ‘bכ’, giving the structure: S 	+ verb 1  +	object	+	verb 2	+ particle Me	    kaa		wכn		bככ	   mu 	(I united them.) 1SING	   PAST-touch	3PL (OBJ)	PAST-hit	in

There is another use of ‘kabכ mu’ (to touch NP to hit NP) where it is not considered reciprocal. Here, there are two objects, one after each of the two items of the serial verb. The structure is: S  +   verb 1	+     object 1	+    verb 2  +	object 2 +   particle Me   kaa	        ne tiri	      bככ		pono no     mu. 1SING PAST-touch  3 POSS head    PAST-hit      table-DEF  in (I      pushed	        his head	       to hit	the table).

Where ‘ka bכ mu’ takes a singular object, it is also not considered reciprocal. It is associated with closing something noisily or slamming it. The structure involved is: S	+	verb 1	+	object	+	verb 2	+	particle Me		kaa		epono no	bככ		mu I SING		PAST touch     table DEF            PAST hit	in. (I slammed the door) Other phrasal verbs of this structure is Ka	NP	wa ano 	(to open door ajar) Touch  NP        suspend on

Ka	NP	hwe mu	(to slam door/window very hard) Touch  NP       beat/fall   in.

Reflexive Phrasal Verbs Reflexive phrasal verbs are transitive verbs in which the subject and the object are referentially identical. This idea then is that the subject did the action to itself. The objects of such verbs are reflexives such as ‘meho’ (myself) ‘neho’ (him/herself) and ‘woho’ (yourself). It is necessary to draw attention to the fact that due to the idiomaticity of phrasal verbs, one cannot actually see objects as the real recipients of action. This is particularly time of true reflexives. As indicated above, Akan phrasal verbs have two kinds of reflexives phrasal verbs – the true reflexive and the syntactic/functional reflexive. These reflexive phrasal verbs are always used with reflexive pronouns as objects, and they cannot change their objects without altering the meaning they carry. Those which use ‘ho’ include ‘Da..ho so’ (to be on the alert), ‘Hyε… ho so’ (to exercise restraint), ‘Brε…ho ase’ (to be humble), ‘Ma.. ho so’ (to be arrogant), ‘Pa… ho akyi’ (to forsake/renounce oneself), and ‘di…ho so’ (to be independent). As indicated above, to replace the reflexive pronouns here will cause meaning alteration. For instance, ‘da woho so’ means ‘to be on the alert’ when the object is a reflexive. When the reflexive object is replaced by a non-reflexive, it becomes a prepositional verb (lie) as in ‘Sekan to da poni no so’. (The knife is no the table.), or a phrasal verb ‘to depend on/to be one’s burden’ as in ‘Me mma adidie da me so’ (My children’s feeding is my burden). Again, the semantic deviation that accompanies ‘hyε so’ when the reflexive is changed is worth noting. ‘Hyε…so’ with reflexives as in ‘hyε ne ho so’ (to exercise restraint). On the other hand, when the reflexive is replaced as in ’Hyε ne so’, the meaning of ‘to force someone’ or ‘to be overbearing/domineering on someone’ is realised. There is another kind of true reflexive phrasal verbs which does not use ‘ho’ as the reflexive item but ‘mu’ giving such pronouns as ‘ne mu’ and ‘me mu’. Examples of such phrasal verbs realised in ‘Kwasi bככ ne mu ase.’ (Kwasi bowed.) and ‘Abכfra no maa ne mu so’ (The child rose). Another kind of phrasal verbs also uses parts of the body such as ‘nan’ (feet/legs) ‘nsa’ (arms/hands) ‘to’ (buttocks) abati (shoulders) as objects instead of ‘ho’ (self) as illustrated below: Ma wonan so		Hurry up -  in movement Ma wonsa so		Hurry 	  -  in some work/Give up or surrender

Functional reflexive phrasal verbs are, however, ordinary active transitive phrasal verbs which have been used reflexively. ‘Si…so’ (to repeat) and ‘hyε…ase’ (to forewarn) which are ordinary phrasal verbs are used reflexively in the following sentences where the subject benefits from its own action: Kofi    sii       ne ho so. (Kofi repeated the drink for himself.) Kofi Alight  3SING/POSS SELF  on Adu ahyε ne ho ase. (Adu has forewarned himself.) ForcePERF 3SING/POSS SELF under Reflexive phrasal verbs in Akan, as discussed above, are used with the following clausal structure; with the object accruing between the verb stem and the positional item: S    +     Verb stem  + NP(Refl)  +  Particle Kwasi	hyεε	       neho	 so. (Kwasi controlled/restrained himself.             force PAST   3SING/POSS self   	on.

Di-transitive Phrasal Verbs The transitive phrasal verbs we have so far discussed involve only one object. They are therefore mono-transitive. In addition to these, one can also identify di-transitive phrasal verbs. Di-transitive verbs in general exhibit a three-person relationship between the subject, a direct object whose involvement makes the action of the verb possible, and an indirect object who receives or benefits from the action. Di-transitive phrasal verbs realise in two clause structures in Akan. The first of these is: S	+  Verb	+   NP(Od)  +	NP(Oi)	+  Particle Kwame	   kyerεε    me		asεm	    ase. (Kwame explained the matter to me). It is understood that the direct object is associated neither with the verb stem nor the position element but the two as a whole constituting a single semantic unit as a di-transitive verb. Below is another example of this structure: S       +      vb	+	NP(od)	    +	NP(oi)	+  Post Abena		besi	Kwadwo  	nsa no	     so. (Abena will give Kwadwo another drink/glass of the drink.) Sometimes, however, a position element (preposition) is used to introduce the indirect object such that the preposition and the indirect object constitute a prepositional phrase. The structure here is:

S  +   vb	+      NP(od)	+	Post	+	pre	+ NP(oi) hyε	       adwuma no	ase		ma	       me te	       asεm no		ase		ma	       no

These two forms help us to distinguish between optional and obligatory form of the di-transitive phrasal verb patterns. The optional ones refer to those in which the recipient or the action (the indirect object) is not considered an inherent, indispensable part of the semantics of the verb and may therefore be added, being introduced by the prepositions ‘ma’ and ‘kyerε’ denoting ‘to’ and ‘for’ respectively. In the obligatory ones, the indirect object is an intrinsic realization of the clause structure. In Akan, only the obligatory di-transitives are generally associated with the S + vb + NP (indirect) + (direct) + post structure. They can, however be realised also in the S+ vb + NP (direct) + post + pre + NP (indirect) structure, which is the optional form. Whereas the obligatory class is generally restricted, the optional structure is used also for ordinary transitive phrasal verbs and the recipient which is the optional element is introduced by the prepositional phrase. Sometimes the use of the optional form for a structure which belongs to the obligatory form creates  different semantic implications: Si	 me	 asεm no	 so (Repeat the story to me.) Vb 	+ Oi +	Od +		 particle Bכ	 me	 asεm no	 mu (Tell me the story.) Vb 	+ Oi +	Od +		 particle. On the other hand, the di-transitive structure as in the sentences below: Si	asεm no	 so		 ma		 me		Bo	 asεm no so		ma		 me		(vb        + oi +	 particle            + pre	 +	od) indicates that the indirect object may not be the person to whom the story is being told or repeated. The implication is that the telling or repeating of the story (perhaps, to another audience) is important to the indirect object. He benefits not from the story itself, but from the fact that it is being told or repeated for the benefit of another whose being informed or informed again is beneficial to the indirect object.

Syntactic Functions of the Akan Phrasal Verbs Phrasal verbs function basically as verbs. As has been seen, they are both transitive and intransitive. As transitive, they, as seen earlier on, can be reflexives, reciprocals, mono and di-transitives. In addition to these functions, phrasal verbs can be seen also a nouns. As nouns, phrasal verbs can be divided into two. The first group is exemplified by ‘nteaseε’ (understanding), ‘ntoboaseε’ (patience) ‘ahyεase’ (beginning) and ‘nnianimu’ (forward). These nominal phrasal verbs are distinguished by the ‘n’ or ‘a’ nominalising prefix. This prefix is added to the uninverted stem + position form of the phrasal verb, giving the structure: ‘n’/‘a’	 +   vb stem 	+	 position (postpositive or particle) n	-	te	-	aseε 	(understanding) a	-	hyε	-	se	(beginning) n	-	bi	-	animu	(leadership) n	-	da	-	aseε	(thankgiving)

In the other type, there is an inversion of the position element and verb. Again, there is no nominalizing prefix, giving the structure: adposition	+	verb stem anim		-	die	(leadership, leading people) ase		-	hyε	(subservience, in subordinate position) ase		-	da	(thanks) ase		-	tena	(life, style of living) It is possible to make further distinction between the two groups of nominal phrasal verbs. This is very evident where two nominal phrasal forms are derived from the same phrasal verb. One of them, the uninverted, nominal-prefixed phrasal verb tends to be associated with general conceptions or abstractions, whilst the stem – positional inverted form is usually associated with specific instances. For example, whereas ‘ahyεase’ may refer to the beginning of things in general, ‘nhyεaseε’ is generally used in reference to the beginning of a specific event or activity. The distinction is further evident in the fact that whereas ‘nteaseε’ (understanding) is a abstract noun, especially as seen also in the use of the nominalizing morpheme, the inverted form is basically a verbal noun which can be seen as an equivalent of the English ‘to infinitive’ ‘to understand it’, or better still, the gerund, ‘understanding it. The verbal nature of this group emphasizes the action with which it is associated, and the specificity. Further, whereas the uninverted ones refers to the total phenomenon of say ‘beginning’, or ‘understanding’, the inverted or verbal forms capture the pain,  difficulty or excitement which is associated with the activity. Again, a look at the modification of these nominal phrasal verbs will be useful in distinguishing them, as evident below: A	Kwasi	aseteε 		(1)	Kwasi		nteaseε		(2) B	Kwame		animdie		(3)	Kwame		nnianimu	(4) C	Abena		Ntesoכ		(5)	Abena		soteε		(6) In sets A and B, it is evident that whereas as in (2) and (4) Kwasi and Kwame can be seen as genitives or possessive determines which pre-modify ‘inteaseε’ and ‘nnianimu’ respectively (Kwasi’s understanding/Kwame’s leadership. Kwasi and Kwame in (1) and (3) are the objects of the actions of ‘understanding’, and ‘leading’ or ‘directing’. This can be captured in the English forms: ‘understanding Kwasi’ and ‘Leading/directing	Kwame’. Sometimes the contrasting use of these nominal forms yields words of entirely different meanings. The differential use of the nominalising forms is to distinguish the nominal forms of homonymous phrasal verbs. The pairs below are illustrative of this distinction:	Asetena	 (way of life)  	Atenaseε	(sitting down/settling down)	Ahyεase (beginning)	Asehyε		(subservience)	Nkosoכ (progress)	Esokכ		(attacking/confronting NP)

It is useful to reiterate the phrasal verbs from which these nominal forms are derived: ‘asetena’ – from ‘tena ase’ ( to live), ‘atenaseε’ – from ‘ tena ase’ (to sit down or to settle); ‘ahyεaseε’ – from -‘hyε ase’ (begin), ‘asehyε’ from ‘hyε ase’ (to begin); and ‘nkosoכ’ – from ‘kכ so’ (to continue, progress), ‘εsokכ’ - from ‘kכ so’ (to attack). This distinction between nominal phrasal verbs has associations also with transitive and intransitive distinction of the phrasal verb. Transitive verbs can realise both as inverted and uninverted forms of the nominal phrasal verb as evident in ‘ahyεase’ (uninverted) and ‘asehyε’ (inverted) which are derived from the transitive phrasal verb ‘hyε NP ase’ (to begin NP) and ‘hyε NP ase’ (to be under NP) respectively. On the other hand, descriptive phrasal verbs, which function to describe the subject (with linking verb + complement implication), and which are intransitive, cannot be used to generate any of these two forms of nominal phrasal verbs. For instance, ‘gyina so’/’te so’(to be in top form) ‘wכ so’ (to be in vogue) and ‘bu so’ (to be abundant) cannot be realised as: ngyinasoכ*/esŏgyina* ntesoכ*/esoteε* nwosoכ*/εsowכ mbusoכ*/esobuo

A verb such as ‘ba so’ (to be proud), however, is realised as ‘mmasoכ’ in contravention of the generalization above. The special use of ‘aseda’ and ‘ndaase’ verbal nouns is also worth mention. It is believed that they must have started in conformity to such distinction but have conflated as a result, perhaps, of some misuse. However, deeply it can be seen that whereas ‘aseda’ is purely a noun meaning ‘thanks’, ‘ndaase’ refers generally or indirectly to ‘thanksgiving’ or the activity or process of giving thanks.

Conclusion In conclusion, it is evident that the Akan phrasal verb, by its structure ad transitivity, complementation and different functions cannot be ignored in Akan lexicography and grammar as has normally been the case. It is useful also that since the phrasal verb as idioms have a special semantic manifestation from the other verbs, there is also the need for further research into the semantic and discoursal significance of phrasal verbs in order to discover the relationship between the structure and syntax of the phrasal verb and its primary function of giving meaning.

REFERENCES Andoh – Kumi, K. 1981, Akan Kasa Mmra (An Akan Grammar). Unpublished Manuscipt, University of 	Cape Coast. ………………(no date) Akan Kasa Nnygyeε ne Mmra (Phonology and Grammar) Unpublished 	manuscript. Akan Teleteaching Course (2004) Unit 8.4: Notes on grammar. http:/www.akan.org/akan_cd/ALIAKAN/course/U8-Grammar-p10.html, 9/17/2004. Akrofi, C. A, 1958, Twi Kasa Mmra ,2nd Edition. London, Longmans. Balint Andvas, 1969, ‘Sector Analysis and Idioms’ Kivung, 1969. Vol. 2.. Cascad, E. H.1982. Cora Locations and Structure imagery. Unpublished PhD Dissertation University of 	California, San Diego. Bamgbose, A. 1974. ‘On Serial Verbs and Serial Status’. Journal of West African Linguistics. 9. 17-48. Chafe, W. L.,1968. ‘Idiomaticity as an Anomaly in the Chomkyan Paradigm,’ Foundations in 	Language, Vol 4. Christaller, J. G.1876. A Grammar of the Asante and Fante Language called Tshi. Reprinted by 	Gregg Press, 1964. Danso, R. O., no date, Twi Kasa Nhyehyee. Unpublished Manuscript. Dolphyne, F. A. 1969. ‘Classification of Akan Verb Systems’. Paper presented at 8th Congress of 	WALS, Abidjan. Cobuild, 1990. English Grammar. London: Collins. Givon, T. 1991. ‘Some Substantive Issues Concerning Verb Serialisation: Grammatical and Cognitive 	Packing’. In C. Lefebvre (ed) Serial Verbs: Grammatical, Comparative and Cognitive 	Approaches. Studies in the Science of Language Series 8. Amsterdam/Philadelphia: John 	Benjamins. Halliday, M. A. K. 1967. ‘Notes on transitivity and Theme in English’. Journal of Linguistics 3. Part 1: 	37-81. Part II: 199-244. Heine, B. and M. Reh. 1984. Grammaticalisation and Reanalysis of African Languages. Hamburg: Helmut Buske. Kari, E. E. 2003. Clitics in Degema: A meeting Point of Phonology, Morphology, and Syntax. Ph.D 	Dissertation. Tokyo University of Foreign Studies, Tokyo. Payne, T. E. 1997. Describing Morphosyntax: A Guide for Field Linguists. Cambridge: University Press. Quirk, R., S. Greenbaum, G. Leech & J. Svartick. 1985. A Comprehensive Grammar of the English Language. London: Longman. Sweet, H. 1891. A New English Grammar. Vol. 1. Oxford: Clarendon Press.