The Categorial Status of the Yoruba Auxiliary Verb

 

The Categorial Status of the Yoruba Auxiliary Verbs

Credit: Prof L. O. Adewole

Yoruba for academic purpose

1.    Introduction[1]

English has been described by Zwicky (1986:1) as the first language of most linguistic theories in this century. A good example to show the correctness of Zwicky’s remark is the AUX hypothesis. For instance, twenty years or so after the “original AUX hypothesis” (McCawley 1985:849) of Chomsky, English has served as its main of exemplification. It was not until the very end of the last decade and the beginning of this that “the question of the extension of the Category AUX beyond English and the general issue of the bases for the application of the categorial label cross-linguistically” (Steele 1986:396) are addressed by Akmajian e.t al. (1979) and Steele et. al. (1981). In addition, in Heny and Richards (1983), whose major objective is to examine the hypothesis of Akmajian et. al. (1979) and Steele et. al. (1981), one of the two volumes is devoted entirely to the study of the English Auxiliary Verbs.

In contrast, there are only two studies of the Auxiliaries in Yorùbá, Ọ̀kẹ́ (1972) and Oyèláran (1982). Both clearly distinguish between the auxiliary and the verb and argue for the recognition of the Auxiliary as a distinct category. The main topic of this paper is to argue against the establishment of the category Aux in Yorùbá. We will first present the criteria for Aux-hood as defined by Steele et. al. (1981) and use these to review the two studies just mentioned. We will then present arguments to show that given Steele et. al.’s criteria for Aux-hood, the category Aux cannot be established for the language and that the items recognised by these two writers should be classified as subclasses of verb.

2. The Properties of AUX

In Steele et. al. (1981), AUX is defined as follows:

1.      Given a set of language-internal analyses, those constituents which may contain only a specified (i.e. fixed and small) set of elements, crucially containing elements marking tense and/or modality, will be identified as non-distinct. (p. 21).

Other properties of AUX noted in language-specific situation are:

2.      a.   Aux is a constituent.

b.   Aux occurs in first, second or final position of an S.

c.   For most choices of L, Aux may attach to some adjacent element.

d.   Aux contains a specified, i.e. fixed and small, class of elements.

e.   These elements occur in a fixed order.

f.    Aux MUST include elements marking tense and/or modality.

g.   Aux MAY also include elements indicating subject marking, subject agreement, questions, evidential, emphasis, aspect, object marking, object agreement, and negation. (pp. 155-156).

(1) and (2) above present the semantic and the syntactic criteria for Aux-hood. We shall regard these as constant and consider whether the category Aux can be established in Yorùbá.

3. Ọ̀kẹ́ (1972) and the Auxiliary Cluster

Ọ̀kẹ́ (1972) classifies as an “auxiliary cluster” any direct sequence of more than one auxiliary element. He argues for the rule in (3) as expressing the “normal order of syntactic precedence among the auxiliaries” (1972:150).

                        3  AUX ----> MODALS  INTENSIFIERS  PRE-EMPTIVES

The class labels in (3), according to him, are unimportant.

3.1 Comments on the Auxiliary Cluster

Ọ̀kẹ́’s work is undoubtedly “a very useful contribution” (Oyèláran 1982:1) to the study of Yorùbá because it is the first attempt to spell out the elements that make up the Auxiliary and discuss their semantic aspects in some detail. Despite this achievement, Ọ̀kẹ́’s work has not gone uncriticised.

According to Awóyalé (1974:18), most of the items Ọ̀kẹ́ claims cannot take prefixation to become nominalised items in fact do. He adds that some elements of the modals often occur nearer to the Full Verb than those of the pre-emptives.

Another linguist who has criticised Ọ̀kẹ́’s work is Oyèláran (1982:1-2), who makes the following points:

-          Ọ̀kẹ́ excludes without any explanation the negative marker (NEG) and the high tone syllable (HTS, following Awóbùlúyì’s (1975) usage). The latter occurs (without exception) after, and is assimilated to the final vowel of all non-pronominal subjects, followed by a Full Verb, as in:

                        Àwọn on túlẹ̀ ẹ́ gba ìsinmi

                        They students HTS take leave

                        “The students are going on holiday”

-          Ọ̀kẹ́ excludes without adequate justification formatives such as dédé, gbọ́ọ̀dọ̀ on the ground that they may be nominalised like verbs by the prefix “a-”.

-          The counter examples to the cooccurrence restrictions Ọ̀kẹ́ proposes for the formatives call for a deeper study of Aux in Yorùbá, given especially the tenuousness of the relationship between the accepted usage of the terms “modal” and “intensifier”, for example, and the syntactic and semantic content of the formatives assigned to them by Ọ̀kẹ́.

3.2 Further Comments

In support of these two scholars, we do not think nominalisation is the most appropriate criterion for distinguishing Auxiliary for Verb, because items which are not verbs would qualify as verbs. One case of note is that of the PP. From tí ilé “from home” and ti àárọ̀ “from morning”, both of which are PP, we can derive nominals such as àti ilé “from the house” and àti àárọ̀ “from/since the morning”. One can show that the derived items are nominals by topicalising or relativising them. This is relevant, because only nouns and nominals are relativised or topicalised in Yorùbá. Any element which is neither a noun nor a nominal would first undergo nominalisation before it is relativised or topicalised. (4) is the topicalisation of the above nominalised items, while (5) is the relativised version of the same:

                        4          Àti ilé ni mo ti ri i

                                    From house FOC I from see it

                                    “I saw it from home”

                        5          Àti ilé tí ó ti rí i . . .

                                    From house REL you () saw it . . .

                                    “The house from where you saw it . . .”

Furthermore, nominalisations such as àti-àárọ̀ can follow prepositions, as in (6).

                        6.         Mo dúró ní àti àárọ̀

                                    I wait in from morning

                                    “I have been waiting since morning”

In addition, not all verbs can be nominalised in the language, even when they constitute by themselves the VP in utterances. In (7b), the verb tì constitutes the VP, yet cannot be nominalised:

                        7          a   Ṣé ó lọ?

                                         O he go?

                                        “Did he go?”

                                    b   Ó tì

                                         He not

                                        “No”

We cannot have:

                        8          a  *àìtì “not not”

                                    b  *ìtì “the fact of not”

                                    c  *àti tì “the act of not”

Ọ̀kẹ́’s other argument which is quite unsatisfactory is his claim that “whenever the negator kò occurs in a clause, it always marks the onset of a VP” (Ọ̀kẹ́ 1969:96). We agree with Ọ̀kẹ́ that this rule applies to sentences such as (9a & b) where both kúkú “indeed” and lọ “go” are elements of the VP as defined by him, but the rule will also classify ti ilé “from home”, a PP, as the first element of VP in (9c).

                        9          a  Kò lọ

                                        Not go

                                        “He did not go”

                                    b  Kò kúkú lo

                                        Not indeed go

                                        “He didn’t go at all”

                                    c  Kò ti ilé lo

                                        Not from house go

                                        “He didn’t go from home”

4. Oyèláran (1982) and the Category Aux

After criticising Ọ̀kẹ́’s analysis of the Auxiliary Cluster, Oyèláran presents his own analysis of the Auxiliary. He regards the Auxiliary as a Category and he claims (personal communication) that Steele et. al.’s treatment of Category AUX supports his analysis.

4.1 Some Objections to Oyèláran’s (1982) Analysis

In this section, we shall show that what Oyèláran regards as Category Aux does not tally, or at best, only partially tallies with the definitional properties proposed by Steele et. al. or by any other linguist. For this reason, we are challenging the claim that the elements of his Category Aux can convincingly be established as a category in the language.

To start with, the only element which is crucial to the establishment of Aux as a language-specific category is not a constituent of Oyèláran’s Category Aux. According to him:

TNS is not a constituent of the AUX and is therefore not a grammatical category of the language ....(It) is not a term within any other auxiliary symbol in the Yorùbá Phrase Structure. It is therefore not a grammatical category of the language. (pp. 3 & 42)

Oyèláran does not claim here that location in time must be expressed grammatically. Rather, he is of the opinion that the periphrastic apparatus of modality and aspect is far more developed in the language, a claim which is lacking in other descriptions of Yorùbá, where verbal grammatical categories have been dominated by tense with aspect being treated as a conceptual waste basket.

Oyèláran’s view of Yorùbá as a tenseless language contradicts the various meanings given to Aux which McCawley (1985:849) summarises as follows:

                        Chomsky 1957

                        (Aux) meant a constituent consisting of a tense and all the Auxiliary                                               Verbs of the given clause.

                        Akmajian, Steele and Wasow 1979

                        (Aux) meant the tensed Auxiliary Verb.

                        Steele et al 1981

                        (Aux) means a constituent consisting of any number of temporal and/or                                          modal features.

Also commenting on Steele et. al.’s definition of AUX as a Category, Wekker (1985:472) states categorically that “Aux is claimed to mark tense and modality, but not aspect”. By this, he means that the presence of aspect is neither a necessary nor a sufficient condition for the establishment of a category Aux in a language.

In his own contribution to the debate on the universality of AUX, Righter (1986:358-359) quotes Reuland (1983) as asking “whether AUX should S-contain or P-contain a ‘tense-marker’. If AUX S-contain tense, every realizations of AUX .... From personal communication by Steele, it appears that S-contain was intended”.

The conclusion one can draw from the foregoing discussion is that if Yorùbá is a tenseless language, as Oyèláran claims, a category Aux cannot be convincingly established for the language.

Even if a category Aux is allowed in Yorùbá despite a clear absence of tense among its constituents, the Phrase Structure Rules proposed for the Yorùbá Predicate by Oyèláran (as shown in (10) are unsatisfactory.

                        10        a   PRED ---> AUX  PP  VP

                                    b   VP --->  V  NP  PP  VP

From (10), it is clear that AUX is a category in the structures proposed. One crucial point we would like the reader to note here is that Oyèláran’s grammar, at least as far as the analysis of Aux is concerned, is monostratal. The important points in (10) are: (i) category aux occurs between the NP and the preverbal PP and (ii) To make sure that the preverbal PP is always a PP[ti], it is stated that the PP should occur in the Predicate.

Note (11i-iii). Although the Auxiliary elements occur after the PP, the PP belongs to the NP.

                        11        i.          Baba ní ilé ni ó ń bá wí

                                                Father at home is you are referring to

                                                “It is to father at home that you are referring”

                                    ii.         Ìwé rẹ̀ sí mi kò tètè dé

                                                Letter is to me did not early arrive

                                                “His letter to me did not arrive in time”

                                    iii.        Enìkan ni Britein n fun wa ni ẹroplein

                                                “Someone in Britain is giving us an aeroplane”

                                                                        (Afolayan 1968:470-473)

The problem is that there are sentences which Oyèláran’s rules do not account for. For instance, there are many grammatical sentences in which the verb do occur before the PP[ti] in the Predicate. Examples are:

                        12        a.         Mo fẹ́ ẹ́ ti ilé ṣe é

                                                I want HTS INF from house do it

                                                “I want to do it from home”

                                    b.         Mo lọ ti ilé ṣe é

                                                I go from house to do it

                                                “I went to do it from home”

In (12a, b) the verbs fẹ́ “want” and lọ “go” occur before the PP. Oyèláran’s comment on this (personal communication) is that both sentences contain the “final construction marker”. This final construction marker is what Oyèláran, quoting Bámgbósé and others, refers to as the infinitive construction. Stated explicitly, according to him, the expression should take the form VP[_NMLPRED], that is, a verb taking a nominalised Predicate as complement. Thus, (12a, b) are the same as (13a, b):

                        13        a.         Mo fẹ́ í ti ilé ṣe é

                                                I want INF from house do it

                                                “I want to do it from home”

                                    b.         Mo lọ i ti ilé ṣe é

                                                I go INF to do it from home

                                                “I went home to do it”

tí ilé ṣe é (from home do it) “do it from home” in the two sentences is then taken to be a Predicate. In (12a), a retrogressive assimilation takes place and “i” changes to “e”. As for (12b) and (13b), Oyèláran concludes that lọ “go” is one of the verbs in the language in which the “final construction marker” is neutralised. He gives the following examples to show that all markers of case and other syntactic categories are optionally realized after lọ “go”:

                        14        a.         Mo lọ sí ilé

                                                (I go to home)

                                    b.         Mo lọ ọ́ lé

                                                (I go ( ) home)

                                    c.         Mo lọ ilé

                                                (I go home)

                                    d.         Mo lọ lé

                                                (I go home)

                                                “I went home”

                                                (the translations in brackets are ours)

While the “case” is marked in (14a) and the final construction in (14b), in (14c, d) lọ “go” has no markers.

Our objections to the claims made above have already been noted by Awóyalé (1983). What Oyèláran regards as the final construction and the VP[_NMLPRED] are regarded as the infinitive and the gerundive respectively by Awóyalé (1983). According to him, the PP in both (12a) and (12b) occur after the verbs, the reason being that despite the presence of the infinitive marker, the verb fẹ́ “want” in (12a) does not lose its verbal nature and lọ “go” in (12b) occurs in a serial verb construction, just like fò “jump” is functioning in (15):

                        15   Mo fò ti Ọ̀kẹ́ bọ́ sí ilẹ̀

                              I jump from up land to the floor

                              “I jumped down from the top”

Our next objection is to the examples in (14) given by Oyèláran in which attempts are made to show the “curious case of lọ ...” (personal communication). It is very difficult to see why lo “go” in the above examples is curious. Rather, the element which is “curious” is the preposition following lọ “go” which Oyèláran himself regards as a case marker, its semantic role is null.

For instance, the preposition in (14a) is one of the elements of which Abímbólá and Oyèláran (1975: 39, fn.) state earlier that

“... they are grammatical formatives because not only do they not have independent meaning, they also belong to a closed system in a non-trivial sense. Formatives performing similar functions like theirs may not normally be derived from any other formatives. And when they occur in utterances they either relate lexical items one to another or express an aspect of a lexical item belonging to a major category”.

(14b) can be treated as resulting from the elision of the “s” and the subsequent assimilation of the “i” of the preposition sí “to”. Examples of such elision and assimilation with elements other than lo “go” are (16a, b)

(16) (a) Èṣú bọ́ ọ́ dìí ìlì > Èṣù bọ́ sí ìdí ìlù

Èṣù (Subj. marker) falls to (the) base drum

‘Èṣù (Yorùbá so-called trickster god) set … to the drum’

(b) Wọ́n dà á ódì > Wọ́n dà á sí òdì

They pour it to opposite

’They rurned it into misunderstanding’

(Adapted with some modifications from Abimbọla and Oyelaran (1975: 4))

                       

Even if we agree that the structure of the predicates of (12a, b) is VP[_NMLPRED] and that the sentences can easily be accounted for by Oyèláran’s PS Rules, sentences such as (17a-f) pose another problem.

Mo ti ilé ń sọ òkò ọ̀rọ̀ sí i                                                                                                       

I from house ( ) throw stone word at him                                                                              

“From home, I sending unpleasant notes to him”

(17) (a) Ó mú ti oko bọ̀

He takes from farm come

“He brought (something) from the farm”

(b) Ó ń ti ilé lọ ń sọ̀rọ̀ sí i                                                                                                                   

He from house go ( ) talk to him                                                                                            

“He (always) goes from home to talk to him”

(c) Ó ké ti oko wá                                                                                                                  

He cry from farm come                                                                                                          

“He returned from the farm crying”

(d) Ó ti ilé lọ máa ń mú un                                                                                                     

He from house go continue ( ) take it

“He often goes from home to take it”

(e) Ó gbọ́n ti ilé wá                                                                                                                

He wise from house come                                                                                                      

“He is well taught from home”

An element of the Category Aux occurs after the PP in (17a) but occurs after the verb in (17c). Also in (17d), ké “cry” is a verb occurring before the PP. In (17e), máa and ń, both belonging to the established Category Aux, follow the PP. Finally in (17f), gbọ́n “to be wise”, a verb, precedes the PP.

With these examples, there is clear justification for suggesting that the only member of the established Category Aux that is fixed in the second position in S is the HTS, but, according to Thrane (1983:158), a Language Specific Category cannot be established on the basis of only one element.

What all this shows is not that one cannot account for the auxiliary verbs in Yorùbá but that one cannot do so by establishing a separate category for such verbs. The only criterion by which Oyèláran establishes the auxiliary as a separate category from the verb is that of the PP occurring after the auxiliary and before the verb. As we have shown above, this is quite unsatisfactory. As the grammar within which the auxiliary as a category is treated is monostratal, the PS Rules in (10) cannot adequately account for the cases:

(18i) Where the Verb occurs before the PP[ti] form as in (17b, d, f).

(ii) Where the Auxiliary occurs after the PP[ti] form as in (17a, e).

(iii) Where both the Auxiliary and the Verb occur after the PP as in (17c).

Problems of these types have led Pullum (1981:437) to give the following examples:

(19a) This is of course certainly causing problems                                                                 

(b) This of course is certainly causing problems                                                                    

(c) This of course certainly is causing problems

           

And concludes that ‘the finite auxiliary verb in English simply is not rigidly fixed in initial, second or final position in S like the Aus elements Steele identifies in various other languages’. From (17), we conclude that Oyelaran’s Auxiliary elements also do not occur in a rigidly fixed position.

5. Counting the Auxiliary

So far, what we have been discussing is the Language-Specific Syntactic aspects of Oyèláran’s criteria for Aux-hood. Now, turning to the notion of auxiliary verb as given by its advocates, we see that Oyèláran’s HTS, NEG, MOD and ASP qualify as auxiliary verb elements. There is no justification, however, for the inclusion of Oyèláran’s Intensifier and Specifier. Oyèláran (1982:17) is aware of this fact too when he states that:

Given that kuku and sese have different restrictions, (...) we suggest that they belong to different sub-categorization of the AUX. Let us call this INTENSIFIER (INT) for kuku and SPECIFIER (SPEC) for sese. Of particular interest in distinguishing these two is the observation that SPEC is verb-like in that it subcategorizes for NP for subject, while such subcategorization is non-relevant for INT.

Later, he states that:

More than any other label, INTENSIFIER is used for want of a better term. For one thing, we are not at all sure that their function is to “intensify”.  In what sense does bá, sáà, kàn, intensify for that instance? (p. 46 fn. 6)

The question then is, if these items (Oyèláran’s intensifiers and specifiers and Ọ̀kẹ́’s intensifier auxiliary) are not elements of the auxiliary verb and as Oyèláran has said, some of them are not intensifiers, what are they?

6. The Modifying Verbs

Awóbùlúyì (1973: 109) argues that the elements under discussion are neither auxiliaries nor verbs but that they are adverbs functioning as verb modifiers. In analysing these items as adverbs, Awóbùlúyì overlooks one of their principal characteristics, namely that, as Oyèláran (1976:18) rightly noted, they should be analysed “in relation to the verbs with which they must co-occur”. It is still controversial whether adjectives can easily be distinguished from Yorùbá adverbs hence, the modifier/qualifier in (20):

(20) O ká ìwé dáadáa                                                                                                             

He read book well                                                                                                      

(a) “He reads books well”                                                                                          

(b) “He reads good books”

may refer to the manner in which the book is read, (20a) being the appropriate gloss, or it may be the books that are being qualified, (20b) being the required gloss. The items under consideration i.e. Oyèláran’s INTENSIFIER and SPECIFIER and Ọ̀kẹ́’s  Intensifier Auxiliaries, unlike the adverb in (20), have no affinity to adjectives.

Still on the same items, Bámgbósé (1974:42-47) also argues that they should be analysed as modifying verbs. The differences between these items and the indisputable verbs often quoted, according to him, are:

They always occur in a modifying capacity to another verb.

(21) Tètè lọ                                                                                                                 

early go                                                                                                                      

“Go early”

They could occur in minimal sentences only in restricted context as in:

(22) Ẹ tètè                                                                                                                              

You (pl.) early                                                                                                                        

“Be early”

(22) still requires a verb to make a complete expression as in:

(23) Ẹ tètè lọ                                                                                                              

You (pl.) early go                                                                                                       

“Go early”

(24) is not acceptable in a restricted context as in (22).

(24) *Ẹ kúkú

You (pl.) even

On these examples, Bámgbósé asserts that if these differences are important for verb identification, non-controversial verbs would not occur in such contexts, yet examples of indisputable verbs occurring in the same contexts abound in the language. Some of the examples he gives to support his point are:

(25a) Ó fi ọwọ́ sí ẹnu                                                                                                             

He put money to mouth                                                                                             

“He puts a coin in his mouth”

(b) *Ó fi owó                                                                                                                         

He put coin

(26a) Aṣọ náà bẹ yòò                                                                                                 

dress the red bright                                                                                                    

“The dress is red bright”

(b) *Asọ náà bẹ                                                                                                                      

The dress is red

                       

On the ungrammatical sentences in (25) and (26), Bámgbósé states that the verb fi “put” in (25b) requires the post-verbal item sí “into”, hence, it cannot occur in a minimal sentence in any context. Also, the fact that bẹ “be red” in (26b) requires an adverb accounts for the unacceptable minimal sentence. He then concludes that there is nothing wrong with a verb having the features [+MOD] which is what he proposes for these items.

We agree with Bámgbósé that these items should be regarded as modifying verbs (V[+MOD]) because like all verbs, they can select their objects and subjects and do occur in verb infinitive phrase constructions. The major difference between them and the main verb is that they always carry an additional feature [+MOD] which is not a necessary feature of the main verb. 

In addition, if Oyèláran’s (1976) persuasive analysis of adverbs is taken, then, there is no way the modifying verbs could be classified as adverbs. Oyèláran says that adverbs always occur postverbally but as shown in (27), modifying verbs occur before the verbs they modify.

(27a) Ó ka ìwé dáadáaa

He read book good/well                                                                                                        

(i) “He reads well”                                                                                                                 

(ii) “He reads early”

(b) Ó tètè ka ìwé                                                                                                                    

He early read book                                                                                                     

“He reads early”         

Both dáadáa “well/good” and tètè “early” modify ka ìwé “read book” in different ways. The affinity between the adverb dáadáa “well/good” with the verb can be broken, as the glosses in (27ai, ii) show and already as pointed out in (20), but that of the modifying verb cannot be broken, hence, the single gloss. In (27) too, the adverb dáadáa “good/well”, in contradistinction to the modifying verb, occurs postverbally while the latter occurs in a preverbal position.

It should also be noted that the effect of both modal and aspect on these items is quite different from their effect on the adverb. Whereas both modal and aspect affect either the modifying verbs or the verbs the modifying verbs modify, both modal and aspect affect the verb and not the adverb that modifies it.

(28a) Ó máa ń tètè lọ

He often early go                   

“He often goes early”

(b) Ó lè tètè lọ                                                                                                                        

He may early go                                                                                                           

“He may go early”

(c) Ó ti lọ rí                                                                                                                

“He has gone before”

(d) Ó lè lọ rí

He may go before

“He may have gone before”

           

It is very difficult, in English translation, to show the effects the modal and aspect have on the verb and the modifying verb in (28a-d). The word for word glosses show how the examples are understood by the speakers of the language but these are not reflected in the literary translation.

7. How many Auxiliary Verbs in Yorùbá?

So far, we have been able to establish that Oyèláran’s INTENSIFIER and SPECIFIER are lexical verbs used to modify other verbs. We have also argued that the auxiliary should be regarded as a subcategory of verb in the language.

Having pointed out all these, we still need to address ourselves to the question of the number of the auxiliary verbs in the language. An answer to this question has actually been given in our discussion above. What can truly be regarded as the auxiliary verbs in the language are some of Ọ̀kẹ́’s Pre-emptive and Modal Auxiliaries and Oyèláran’s HTS, NEG and the items with which his MOD and ASP are realised.

Another question that needs be answered is the reason why none of the previous writers recognised only the items just noted as the auxiliary verbs in the language?

The fact is that the writers do implicitly recognise that these items are the only auxiliary verbs in the language but the problem is that they cannot adequately account for them in the framework of the grammars within which they work. For instance, in Ọ̀kẹ́’s treatment of the Auxiliary Cluster, only the Pre-emptive and the Modal Auxiliaries are treated at some length. The Intensifier is included, mainly to serve, along with the Modal and the Pre-emptive, as some sort of diagnostic context for his Full Verb. He does not devote as much space to it, at least to the semantic treatment, as he does to the Modal and the Pre-emptive Auxiliaries.

This also applies to Oyèláran. Without the inclusion of both the INTENSIFIER and the SPECIFIER, the delimitation of Category Aux with PP would not have been easy. It will be recalled that his establishment of Category Aux in the language is based mainly on the occurrence of the PP between the Aux and the verb in the Predicate. This would have been impossible if the elements considered did not include both the INTENSIFIER and the SPECIFIER. Even with these two constituents, we have been able to show in this work that a Category Aux cannot be convincingly established in the language.

Awóbùlúyì (1967:253-258) also recognises our auxiliary verb as a subclass on its own. What is surprising is that he regards them as a subclass of adverb.  Consider the following passage:

The Preverbs constitute four classes of co-occurring elements. The first and second comprise respectively the indefinite tense markers máa, yóò,  óò, and a on the one hand, and aspect markers ti, M, and maa on the other,  both classes of elements participating in the chronological system of the verbs.  The third class consists of the modal preverbs ibá, ibáà, lè and gbúùdọ̀ or gbọ́ọ̀dọ̀, while the fourth and the last class is made up of all the remaining manner preverbs. Only the first three are, in our opinion, of sufficient interest to warrant special treatment. (ibid. Pp. 257-258)

The classes that “warrant special treatment” are almost identical with our own auxiliary verbs. The differences are the non-inclusion of the NEG and the HTS, the latter of which he later recognises as a tense marker (Awóbùlúyì 1975:229). He also includes ba and ibaa which are the equivalent of the English “if” and “even if” respectively which are not generally included within the sphere of modality and differ syntactically from the other auxiliary verbs.

Finally, though Bámgbósé (1973:51) characterises the items under discussion, except n, NEG and HTS, as auxiliary verb, he does not treat them as a subclass on their own. Instead, he merges them with other items, classifying them as PREVERBS.

8. The Yorùbá Auxiliary Verbs

The following are the items recognised as the Auxiliary Verbs in this work:

                        29        a.         NEG

                                    b.         HTS

                                    c.         yóò “will” and its variants

                                    d.         ti “has” and its variant

                                    e.         gbọ́dọ̀ “must”

                                    f.          lè “can”

                                    g.         ń and its variant

                                    h.         máa ń and its variant

The functions of the elements listed in (29) will be discussed in another paper.

One of the reasons why we think that the auxiliary verbs should be clearly specified and properly discussed is that, despite the importance of the auxiliary verbs to any grammar (on this see Chomsky (1957:38)), many Yorùbá writers use the term “Auxiliary” as a sort of rag-bag into which to toss most verbal uses that do not seem to the author to be those of standard main verb.

Another reason is that many of the works in which the auxiliary verbs are mentioned focus on items other than those in (29).

9. Conclusion

To sum up, what we have done in the previous sections is to argue against the establishment of Category Aux, using Steele et. al.’s (1981) criteria of AUX-hood. We also criticise Ọ̀kẹ́ and Oyèláran, the former for excluding some items which are elements of the Auxiliary and for including others which are not, and the latter for including items which are not elements of the Auxiliary.

Our argument here supports the one earlier made by Thrane (1983:196), who states that AUX is not a universal category i.e. that it is not a “necessary, predicate-bound linguistic property, rather, it is a merely possible, significant, language-bound linguistic property”.

Following the auxiliary-as-verb hypothesis, we suggest that the items identified in (29) should be classified as a subclass of verb for the reasons we have discussed in this paper, to which the following could be added:

-          There are no rules that refer specifically to the auxiliary as there are in such languages as English where we have rules such as the subject auxiliary inversion (often called subject verb inversion).

-          Unlike some of the other categories in the languages, no subcategories of verb – main, auxiliary, modifying ... – can be coin joined. Hence, we cannot have (30a) whereas, we can have (30b).

-          (30ai) *Lọ àti bọ̀                                                                                                                    

-          Go and come (base verb)

-           

-          (ii) *Tètè àti kúkú                                                                                                       

-          Early and simply (modifying verb)

-           

-          (iii) *yóò àti máa                                                                                                                    

-          will and continue (auxiliary verb)

-           

-          (bi) Dúdú àti pupa                                                                                                      

-          black and red (nouns)

-           

-          (bii) ni ilé àti ni oko                                                                                                                     in house and in farm                                                                                                                  “At home and on the farm” (prep. phr.)

-          There are also some morphological phenomena that support an analysis of the Auxiliaries as verbs. For example, the morphology of both is the same: none of them takes inflection and they are all realised as lexical items especially if the lexical realisation of the HTS is compared with that of the third person pronominal object as in (31):

(31a) Olú ú lọ                                                                                                             

Olu HTS go                                                                                                    

“Olu went”

(b) Mo rí i                                                                                                                   

I see it/her/him                                                                                                           

“I saw it/her/him”

It will be recalled from our previous discussions that there are no features of the auxiliary verbs in the language which are not shared by some indisputable verbs. As this is the case, we think that a treatment which permits a generalisation of all verbs should be preferred in the language. The fact that the auxiliary verbs are highly grammaticalised does not mean that they should be regarded as items “which are used as substitute for inflections” (Poutsma (1926:15-17)). There are no inflections in the language and there is no point in saying that the auxiliary verbs are substituting for some. The analysis of the Yorùbá auxiliary verbs as Verbs is adequately captured by Safarewicz (1974:20) when he states that

If within the scope of a certain class of words that do not exhibit any explicit opposition as to their morphological structure, some function is characteristic of all these words while another function is characteristic of some of them only, then the function characteristic of all the words is their primary function.

It is in this light that we say that all verbs – may they be modifying, auxiliary, causative, etc. – should first be classified into their primary category – verb. It is after this that subsequent subclassification into functional classes which are peculiar to some of them could be made.

From the foregoing, it seems there is sufficient motivation both for classifying all auxiliary verbs in the language as verbs and for classifying them as a distinct subclass of verb just as the modifying verbs, the splitting verbs and a host of other subcategories of verbs in the language are classified. The shared properties of the auxiliary verbs and the main verbs are accounted for on the basis of the shared feature specification [+/-VERB], while the distinctive properties are accounted for on the basis of different specification for the feature [+/-AUXILIARY]. Thus the auxiliary verbs are [+VERB, +AUXILIARY] and the main verbs are [+VERB,-AUXILIARY]. As there are no auxiliaries that lack verbal properties, in the language, no lexical item will be specified [-VERB,+AUXILIARY] and all the auxiliaries will be allowed to take VP as complement just as most verbs do.

10. Summary

In this work, we have presented evidence to show that, given the criteria for Aux-hood as proposed by Akmajian et. al. (1979) and Steele et. al. (1981), the constituents identified by the previous writers in the language do not qualify as Aux. While some of the constituents so identified can be classified as auxiliary verbs, others should be classified as modifying verbs.

BIBLIOGRAPHY

Abimbola, W. And O.O. Oyèláran (1975), “Consonant Elision in Yoruba”, AFRICAN LANGUAGE STUDIES 16:37-60.

Afolayan, A. (1968), “The Linguistic Problems of Yoruba Learners and Users of English.”, Ph.D. Dissertation, University of London.

Afolayan, A. (1972), “The Predicative Adjectives as a Grammatical Category in Yoruba”, in YORUBA VERBS PHRASE, edited by Ayo Bamgbose, pp. 73-102. Ibadan: University Press.

Akmajian, A., S.M. Steele, and T.M. Wasow (1979), “The Category AUX in Universal Grammar”, LINGUISTIC INQUIRY 10: 1-64.

Awobuluyi, O. (1967), “Studies in the Syntax of the Standard Yoruba Verb.”,  Ph.D. Dissertation, Columbia University.

Awobuluyi, O. (1972), “Predicative Adjective in Yoruba: A Critique”, STUDIES IN AFRICAN LINGUISTICS 6, 3:215-238.

Awobuluyi, O. (1975), “Some Traditional Yoruba ‘Adverbs’ in True Perspectives”, JOURNAL OF WEST AFRICAN LANGUAGES 10, 1:28-54.

Awóyalé, ‘Yiwola, (1974), “Studies in the Syntax and Semantics of Yoruba Nominalization.”, Ph.D. Dissertation, University of Illinois, Urbana.

Awóyalé, ‘Yiwola, (1983), “On the Development of the Verb Infinitive Phrase in Yoruba”, STUDIES IN AFRICAN LINGUISTICS 14, 1:71-102

Bamgbose, Ayo (1963), “A Study of Structures and Classes in the Grammar of Modern Yoruba.”, Ph.D Dissertation, University of Edinburgh.

Bamgbose, Ayo (1966), A GRAMMAR OF YORUBA. Cambridge: University Press.

Bamgbose, Ayo (1973), “The Modifying Serial Construction”, STUDIES IN AFRICAN LINGUISTICS 4, 2: 207-217.

Bamgbose, Ayo (1974), “On Serial Verb and Verbal Status”, JOURNAL OF WEST AFRICAN LANGUAGES 9: 17-48.

Chomsky, N. (1957), SYNTACTIC STRUCTURE. Mouton: The Hague.

Ekundayo, S.A. and F. Niyi Akinnaso (1983), “Yoruba Serial Verb String Commutability Constraints”, LINGUA 60:115-133.

Gazdar, G. And G.K. Pullum (1982), GENERALIZED PHRASE STRUCTURE GRAMMAR: A THEORETICAL SYNOPSIS.  Bloomington, Indiana: Indiana University Linguistic Club.

Heny, F. and B. Richards (eds.) (1983), LINGUISTICS CATEGORIES: AUXILIARIES AND RELATED PUZZLES. VOL. 1: CATEGORIES. VOL 2: THE SCOPE, ORDER AND DISTRIBUTION OF ENGLISH AUXILIARY VERBS. Dordrecht: D. Reidel Publishing Company.

Lyons, J. (1966), “Towards a ‘Notional’ Theory of ‘Part of Speech’”, JOURNAL OF LINGUISTICS 2, 2: 209-236

Madugu, I.S.G. (1986), “Form and Substance in Verbs Serialization”, Paper delivered at the Institute of Culture, University of Ife, Nigeria, on the 20^th of January 1986.

McCawley, J.D. (1985), “Review of LINGUISTIC CATEGORIES: AUXILIARIES AND RELATED PUZZLES”, LANGUAGE 61, 4: 849-862.

Oyèláran, O.O. (1976), Eyanro-Ise Ede Yoruba (Yoruba Adverbs)”, Paper Delivered to the General Conference of the National Association of Teachers of Yoruba Language at St. Leo’s TTC, Abeokuta, Nigeria, 12-14 August, 1976.

Oyèláran, O.O. (1982), “The Category Aux in the Yoruba Phrase Structure”, Paper Presented at the 15^th West African Languages Conference, University of Port Harcourt, Nigeria, April 4-11, 1982.

Oyèláran, O.O. (1982a), “On the Scope of the Serial Verb Construction in Yoruba”. Mimeo.

Ọ̀kẹ́, D. O. (1969), “Grammatical Study of Yoruba Verb System.”, Ph.D. Dissertation,     University of  York.

Ọ̀kẹ́, D. O. (1972), “On the Construction and Semantic Interpretation of Auxiliary Cluster in Yoruba”, in YORUBA VERB PHRASE, edited by Ayo Bamgbose, pp. 135-162. Ibadan: University Press.

Poutsma, H. (1926), A GRAMMAR OF LATE MODERN ENGLISH, PART II, SECTION II. Groningen: Noordhoff.

Pullum, G.K. (1981), “Evidence against ‘Aux’ Node in Luiseno and English”, LINGUISTIC INQUIRY 12: 435-463

Pullum, G.K. (1982), “Syncategorematicity and the English Infinitival ‘to’”, GLOSSA       16: 181-215

Rigter, B. (1986), “Review of LINGUISTIC CATEGORIES: AUXILIARIES AND RELATED PUZZLES”, LINGUA 69, 4: 355-382.

Safarewicz, J. (1974), LINGUISTIC STUDIES. The Hague: Mouton.

Sereechareonsatit, Tansanee (1984), “Conjunct Verbs and Verbs in Series in Thai”. Ph.D. Dissertation, University of Illinois.

Stassen, Leone (1985), COMPASSION AND UNIVERSAL GRAMMAR. Oxford: Basil      Blackwell.

Steele, S. (1986), “Review of LINGUISTIC CATEGORIES: AUXILIARIES AND RELATED PUZZLES”, STUDIES IN LANGUAGE 9, 3: 395-407.

Steele, S., R. Akmajian, E. Demens, C. Jelinek, R. Oeherle Kitagawa and T. Wasow (1981), AN ENCYCLOPAEDIA OF AUX. Cambridge, Mass.: M. I. T. Press

Thrane, T. (1983), “On the Universality of AUX”, ACTA LINGUISTIC HAFNIENSIA      18: 154-200.

Wekker, H. (1985), “Review of Perkin’s MODAL EXPRESSIONS IN ENGLISH” JOURNAL OF LINGUISTICS 21: 520-521.

Zwicky, A.M. (1986), “Introduction to German Syntax in GPSG”, edited by Arnold M.   Zwicky, LINGUISTICS 24, 5:855-856.

---

[1] An earlier version of this paper was published as L.O. Adewole (1987), ‘The Categorial Status of the Yoruba Auxiliary Verbs’, Work in Progress 20: 7-25. Department of Linguistics, University of Edinburgh.