GBỌ́DỌ̀ ‘MUST: ANALYSIS OF A YORUBA MODAL VERB

 

GBỌ́DỌ̀ ‘MUST: ANALYSIS OF A YORUBA MODAL VERB

Credit: Prof L. O. Adewole

Yoruba for academic purpose

This paper examines the meanings of a Yoruba modal verb, gbọ́dọ̀ ‘must’. A clear distinction is made between gbọ́dọ̀ ‘must’ and ní láti- ‘have to’ which some scholars regard as its stylistic variant. There follows a systematic semantic analysis of gbọ́dọ̀ ‘must’. Examples are taken from a narrative text to avoid, as much as possible, a subjective interpretation of modal syntax and semantics.

Cette étude examine les sens d’un verbe modal Yoruba, gbọ́dọ̀ ‘falloir’. D’abord, une distinction nette est faite entre gbọ́dọ̀ ‘falloir’ et ní láti- ‘avoir à/devoir’ qui est considéré comme une variante stylistique par certains. Ensuite une analyse sémantique est proposée pour gbọ́dọ̀ ‘falloir’. Les exemples sont tirés d’un texte narratif afin d’éviter, autant que possible, une interprétation subjective de la syntaxe et la sémantique modales.

1.         INTRODUCTION

Modality is taken in this paper as a semantic label for notions like possibility, permission, obligation, etc., that alter, in some way, the neutral semantic value of a clause or sentence. Modals are auxiliary verbs which denote different kinds of modality. They are used to modify declarative and factual statements. Modal meanings have to do with the speaker’s attitude towards what he says or the degree of commitment to his utterances. A modal sentence or utterance, therefore, is one in which the truth of the prediction is subject to some contingency or modification. The modal verb we shall be concerned with here is gbọ́dọ̀ ‘must’. In our analysis, we shall make extensive use of a corpus of data taken from Atótó Arére, a fictional prose narrative by Ọládẹ̀jọ Òkédìjí based on some social shortcomings of the Nigerian society. Page numbers of the examples are enclosed in brackets.

2.         SOME COMMENTS ON ní láti- ‘HAVE TO’

2.1                   INTRODUCTION

It has been suggested that ní láti- ‘have to’ is a modal verb (Oyelaran 1982:17) and that it is a stylistic variant of gbọ́dọ̀ ‘must’ (Welmers 1973:341-42). In this section we shall take a critical look at these two suggestions and show, first, that the differences in meaning between the periphrastic ní láti- ‘have to’ and gbọ́dọ̀ ‘must’ are considerable and, second, that the periphrastic item is not a modal verb.

2.2                   THE PERIPHRASTIC ní láti- ‘HAVE TO’

Welmers quotes Elimelech (in a footnote) as saying that while ‘elderly people, whose contact with English has been minimal’ use gbọ́dọ̀ ‘must’, other Yoruba speakers use ní láti- ‘have to’. We understand from Elimelech’s claim that both gbọ́dọ̀ ‘must’ and ní láti- ‘have to’ have the same semantic interpretation, the only difference between the two being that ní láti- ‘have to’, in addition to sharing the basic meaning with gbọ́dọ̀ ‘must’, also reflects a quantity of formal education which the speaker/writer wants to show. To test Elimelech’s claim, we examine the use of ní láti- ‘have to’ in Atótó Arére. Taken from a prose narrative written by someone of considerable contact with English, one would expect ní láti- ‘have to’ to occur more than gbọ́dọ̀ ‘must’. We find the contrary, however, as shown below:

1.                  Total Sample  of ní láti- ‘have to’ and gbọ́dọ̀ ‘must’ = 117 (100%)

gbọ́dọ̀   ‘must’  = 92 (79%)

ní láti- ‘have to’ = 25 (21%)

An interview independently carried out by Oyetade (1989:9-11) supports this. He asked 43 undergraduate Yoruba students at Ọbáfẹ́mi Awólọ́wọ̀ University to translate the following sentences into Yoruba

You must go

You must not go

Forty-one students translated ‘must’ into gbọdọ̀, one student translated it as dọndọn ‘compulsory’ and only one as ní láti- . Since these students are undergraduates, if Elimelech’s claim is valid, one would expect more students to use ní láti- in their translation, but the reverse is the case.

In addition, we note that whereas gbọ́dọ̀ ‘must’ signals a commitment of the speaker to the truth of the proposition in which it occurs, the periphrastic item signals absence of such commitment. Thus, whereas gbọ́dọ̀ ‘must’ in 2 expresses the speaker’s commitment to the necessity or possibility of an action, ní láti- ‘have to’ in 3 occurs in a categorical statement to indicate that something is necessary or possible.

2.                    ó gbọ́dọ̀ ṣe é    ‘He is required/obliged to do it’

3.                    ó ní láti-ṣe é   ‘It is necessary/possible that he does it’

Witness that both ní láti- ‘have to’ and gbọ́dọ̀ ‘must’ occur in the same frame in 2 and 3. The importance of this is that any difference in meaning between the two sentences is taken as the difference between the meaning contributed by each of the two items to the sentence in which it occurs.

Using the same process of substitution, we also note that the meaning negation adds to gbọ́dọ̀ ‘must’ is quite different from that of ní láti- ‘have to’. Negation with gbọ́dọ̀ ‘must’ always bears on the semantic content of the proposition being negated. The meaning of gbọ́dọ̀ ‘must’ itself is not altered. This means that kò gbọ́dọ̀ lọ (NEG must go) is an ‘obligation not to go’ hence a ‘prohibition to go’. It also negates the existence of permission, for, if one is ‘forbidden to go’ one is at the same time ‘not permitted to go’. The negative of ní láti-lọ ie. kò ní láti-lọ (NEG have to go) is, on the other hand, not found in the areas of ‘obligation not to go’, rather, it expresses ‘absence of necessity to go’.

2.3       A LOOK AT ní láti- + VP.

Examining ní láti- ‘have to’ and the following VP, one may be able to suggest its true status. The composition of ní láti- ‘have to’ is quite different from the modals. It is made up of three elements - (i) a verb ní ‘have’, (ii) a preposition ní ‘in’ and (iii) a nominalizing prefix àti (see Fágbọrún 1985:14 n.3). The verb ní ‘have’ has a great deal in common with such verbs as fẹ́ ‘want’, to ‘enough’ and ṣòro ‘difficult’. This class of verbs is categorized by Awóyalé (1974:12-13) as factive verbs which often take ‘infinitive without overt complementizers’. For instance, taking the VP ṣe iṣé ‘do work’, the prefix àti can be used to nominalize the VP to become àti-ṣe-iṣẹ́ ‘manner/fact of doing work’. This nominal, like all nominals in the language, can in turn be preceded by the preposition ní ‘in’ as in ní àti-ṣe-iṣẹ́ ‘in the manner/fact of doing work’. Yusuf (1985:333) posits that this ní ‘in’ is ‘empty of any semantic content. Any meaning attributed to it is nothing inherent but derives from our knowledge of the NPs that follow it’. This PP or ‘ní-PP construction’ (to use Yusuf s terms) can serve as object to the class of factive verbs mentioned above. Hence we can have 4.

4a.     ó ní ní àti-ṣe-iṣẹ́ (ie. ó ní láti-ṣiṣẹ́)

he has in the manner/fact of working

‘He has to work’

b.         ó fé ní àti-ṣe-iṣẹ́ (ie. ó fẹ́ láti-ṣiṣẹ́)

he want in the manner/fact or working

‘He wants to work’

The two sentences are relativized in the same manner, eg.

5a.      iṣẹ́ tí ó ní ni àti-se (ie. iṣẹ́ tó ní láti-ṣe)

work REL he has in manner/fact do

‘The work he has to do’

b.      iṣẹ́ tí ó  fẹ́ ní àti-se (ie. iṣẹ́ tó fẹ́ láti-ṣe)

work REL he want in manner/fact do

‘The work he wants to do’

They are also negated in the same way.

6a.       kò ní ní àti-ṣe-iṣẹ́ (kò ní láti-siṣẹ́)

NEG has in manner/fact do work

‘He is not duty bound to do (a given) work’

b.         kò fẹ́ ní àti-ṣe-iṣẹ́

NEG want in manner/fact do work

‘He does not want to work’

It will be noted that it is the factive verbs that are being negated in 6.

2.4       WHERE THE CONFUSION LIES

Abraham (1958:438,7b) classifies ní ‘have’ as a verb but translates the combination of ní láti- as ‘must’. His example (numbering is ours) is:

7.         mo ní láti lọ

‘I must go’

Thus, it seems, from Abraham’s classification, that when ní ‘have’ precedes láti-, it ceases to be a verb. Abraham also classifies fẹ́ ‘want’ as a verb which it remains when it combines with láti-.

8.         ó fẹ́ láti lọ

‘He wants to go’          (Abraham 1958:207)

What is not clear to us is why Abraham gives different analyses to ní láti- and fẹ́ láti- when, as we have shown above, they have a great deal in common. It may be that Abraham believes that the first ní ‘have’ in ní láti- has developed from the lexical verb ní ‘have’ and that the construction ní láti- now behaves like a modal verb. The argument is possible because it looks as if the reading given to ní ‘have’ when it precedes ní láti- is quite distinct from the one given to it in such sentences as mo ní owó ‘I have (some) money’. But ní ‘have’ owes its different readings in ní owó ‘have money’ and ní láti- to the different lexical contexts in these two sentences showing that an argument along this line is not tenable. Hermeren (1978:70-71) has warned that the meaning of a given item should not be confused ‘with the meaning of the units surrounding it. In such a description one would have assigned... (to an item) as many meanings as there are contexts in which it occurs’ (our brackets).

Note that the meanings of  fẹ́ ‘want’ in fẹ́ ìyàwó ‘got married’,  fẹ́ Olú ‘loves/likes Olú’ and mo ń fẹ́ owó ‘I need some/lot of money’ are quite different from that of fẹ́ ‘want’ in examples 5, 6 and 8 yet fẹ́ ‘want’ is treated as the same verb in all the constructions. This means that Abraham gives  fẹ́ ‘want’  a unified treatment whereby the basic meaning of  fẹ́ ‘want’ is taken to be comprehensive, other meanings being derived from it. In the same manner we suggest that ní ‘have’ should also be given a unified treatment whereby all the derived meanings ie. to have, to occupy, to obtain, to possess, to get, to load (as a ship or canoe), etc. (CMS 1913:162) come from a single comprehensive meaning. While the semantic interpretation of ní ‘have’ as in mo ní ilé ‘I have a house’ can be regarded as one end of a continuum, ní ‘have’ in ní láti- can be taken as the other end.

In conclusion, as both ní láti- ‘have to’ and gbọ́dọ̀ ‘must’ do not share the same core meaning, we conclude that they are distinct items which should not be classified as variants. Having clearly distinguished between ní láti- ‘have to’ and gbọ́dọ̀ ‘must’, we shall now concentrate the discussion on the modal meanings of gbọ́dọ̀ ‘must’ which is our major concern in this paper.

3.         THE MODAL MEANINGS OF gbọ́dọ̀ ‘MUST

3.1       NECESSITY

This has the sense ‘the only possible conclusion from the evidence available is that...’, or, to use Chung and Timberlake’s terms (1985:242), ‘in all alternative worlds that one could imagine at this time, x is...’, eg.

9.         léèsì tó  wọ̀  sọ́rùn  ti  kóbó,  ojú  rẹ̀  ti  gbọ́dọ̀  rí  nǹkan (p.5)

lace that he wear in neck has fade, eye his has must see something

‘The lace garment he is wearing has faded, he must have suffered’

10.       ó  sá  ní  oye  ènìyàn  tí  òun  ganan  ti  fọwọ́  pa.

it have number people that he even has use hand kill.

bí kò  bá tó ogún,  yóò fẹ́rẹ̀ tó bẹ́ẹ̀. Ó tilẹ̀ gbọ́dọ̀ jù bẹ́ẹ̀ lọ  (p.222)

If NEG up to twenty, will almost up to that. It even must more that go

‘At least he has also killed a number of people. If the number of those killed by him is not up to 20, it will be nearing that. In fact, they must be more.’

In 9, the speaker’s confidence is overtly expressed and the reason for his confidence is given (the condition of the lace garment). In 10, the speaker starts with an unsubstantiated assertion yóò fẹ́rẹ̀ ‘they will almost’ but replaces this with a substantiated one, gbọ́dọ̀ jù bẹ́ẹ̀ ‘must be more’. The harmonic modifying verb tilẹ̀ ‘even’ overtly expresses the speaker’s confidence. The ‘necessity’ use of gbọ́dọ̀ ‘must’ applies to state and activity either in the past or in the present, but this does not mean that it cannot be used to indicate the speaker’s confidence in a future state or activity, eg.

11.       ẹ      ó  mọ̀    pé   ìyàtọ̀  gbọ́dọ̀ wà (p.14)

you(pl) will know say different must be

‘You will know that there must be some changes’

Here the main predication refers to the future. The harmonic phrase Ẹ ó mọ̀ pé ‘you will know that’ also indicates that the statement is subjective. 12, on the other hand, is an objective use of gbọ́dọ̀ ‘must’ based on pure logical necessity without any element of speaker-involvement.

12.       ó pẹ́ ní,  ó yá   ní,    ọmọọṣẹ́ gbọ́dọ̀ yaṣẹ́ lọ́dọ̀ ọ̀gá  (p.75)

            it late FOC, it early FOC, apprentice must separate work in place master

‘Sooner or later, an apprentice must complete his period of apprenticeship and start practising his craft.’

3.1.1                Syntactic Co-occurrence

Perfect marker: In 9, the perfect marker precedes gbọ́dọ̀ ‘must’ but as shown in 13, the perfect marker can follow gbọ́dọ̀ ‘must’. Gbọ́dọ̀ ‘must’, while occurring with the perfect marker, can refer to a past state or activity stretching up to the moment of speaking as in 9, or it may refer to a future state or activity contemplated from the present as in 13.

13.       kí     ó  tóó    dé,  owó  àwọn  gbọ́dọ̀ ti  búrẹ́kẹ́ (p.50)

COMP he before come, money their must have grow considerably

 ‘Before he comes back, their money must have increased considerably.’

Stative:  see examples 9 and 12.

Progressive aspect: There is no example of a progressive aspect alone co-occurring with gbọ́dọ̀ ‘must’ in the text. The invented one (14) includes the occurrence of the perfect marker and the sentence is given the reading in which the time he starts grinding precedes the time of speaking.

14.       ó   ti gbọ́dọ̀ máa   lọ̀  ọ́ lọ́wọ́

he has must PROG grind it in hand

‘He must have started grinding it’

3.1.2                Harmonic Combination

In 11 and 12, gbọ́dọ̀ ‘must’ expresses the same degree of modality with the harmonic clauses ẹ ò mọ̀ pé ‘you will know that’ and ó pẹ́ ni ó yá ni ‘sooner or later’ respectively. In such cases the harmonic forms are said to be ‘mutually reinforcing’ (Coates 1983:77). There are no examples of phrases with hedging quality co-occurring with gbọ́dọ̀ ‘must’ in the text but an invented example follows.

15.       mo   lérò   pé   ó gbọ́dọ̀ ti   dé

I have think say he must have come

‘I think he must have come’

15 is subjective in that it shows the speaker’s attitude to the proposition in the main predicate.

3.2       OBLIGATION OR COMPULSION

The use of gbọ́dọ̀ ‘must’ here involves the authority of the speaker. It is the speaker who states that an event must hold ‘in all subsequent worlds’ (Chung and Timberlake 1985:246), eg.

16.       ẹ         gbọ́dọ̀ ṣẹ́  náírà méjì kù (p.43)

you (pl) must break naira two left

‘You must leave a two naira balance’

In 16, the speaker is the one in authority telling the addressees that they are obliged to leave a two naira balance. There are cases where the speaker exerts power over himself. Leech (1981:72) refers to such cases as self-compulsion in which some power is exerted on oneself through a ‘sense of duty, through self-discipline, or merely through a sense of expediency’, eg.

17.       òní     tí  o  wọlé mi yí,  mo gbọ́dọ̀  wẹ̀    ẹ́   (p.53)

today that you enter house mine this, I must wash it

‘Today you enter my house, and I must celebrate it’

18.       mo gbọ́dọ̀ dé  ilé    Miloníà     kí n kí wọn níbẹ̀ (p.65)

I must reach house Millionaire let I greet them in there

‘I must go to Millionaire’s house to say hello to them there’

In 17 and 18, the reasons for the self-imposed compulsion are explicitly stated. The use of gbọ́dọ̀ ‘must’ for an obligation also has the following characteristics: (i) the subject is normally animate and (ii) it also excludes the occurrence of the perfect marker.

4.         EPISTEMIC AND DEONTIC MODES

Both philosophers and linguists have recognized two distinct kinds of modals, commonly called ‘epistemic’ and ‘deontic’. Chung and Timberlake (1985:242) define epistemic modality in terms of actual and alternative worlds. ‘Epistemic mode characterizes the event with respect to the actual world and its possible alternatives. If the event belongs to the actual world, it is actual, if it belongs to some alternative world (although not necessarily to the actual world), it is possible; and so on.’ They also define the deontic as characterizing ‘an event as non-actual by virtue of the fact that it is imposed on a given situation. Given the actual world at any point in time, there are a number of worlds that could conceivably develop out of that world. The deontic...restricts these subsequent worlds with respect to an event, such that the event has to belong to some or all of the subsequent worlds’ (Chung and Timberlake 1985:246). For example, ó gbọ́dọ̀ ní ìyàwó can mean

19.                  I (confidently) infer that he is married.

20.                  He is obliged to be married.

The first meaning is called the epistemic while the second is called the deontic (Lyons 1977:791). Note that the relationship between the deontic and epistemic modes is non-arbitrary. For example, the same modal used to convey a deontic meaning under ‘necessity’ in 19 is used to convey epistemic meaning under ‘obligation’ in 20. According to Lyons (1977:792-93), logicians would gloss 19 as 21 and 20 as 22.

21.       In the light of what is known, it is necessarily the case that he is married.

22.       I (hereby) oblige him to be married.

In addition, both deontic and epistemic modality can be given either a subjective or an objective interpretation. While a subjective modalized utterance is a statement of opinion, an objectively modalized utterance contains ‘an unqualified or categorical, I-say-so component’ (Lyons 1977:799). In the latter, the speaker is committed to the factuality of the information. For example, ẹ gbọ́dọ̀ lọ ‘you must go’ can be given an objective interpretation, meaning that the addressee is expected to go because this is what everybody does. What the speaker says to be the cause of an objectively modalized utterance can be denied or questioned, accepted or referred to by the complement of a factive predication (Lyons 1983:86).

23.       ẹ  gbọ́dọ̀ lọ

you(pl) must go

          ‘You must go’   

irọ́ ni  ‘That is not true’

lóòótọ́? ‘Is that trus?’

n ò gbà ‘I don’t agree’

mo gbọ́ ‘I agree’

mo mọ̀ ‘I know’

If the above utterance is given a subjective interpretation, it means that the speaker is subjectively qualifying his commitment to the possibility of the addressee going to the said place. An appropriate clause mo rò pé ‘I think that’ can be added to the utterance to show the speaker’s commitment. Subjectively modalized statements are not statements of fact, hence in reporting 23, we have (24).

24.       ó  ní  oun ro      pe  a gbọ́dọ̀ lọ

he say he think that we must go

‘He says he thinks we must go’

To report the objectively modalized version of 23, we have

25.       ó ní     a gbọ́dọ̀ lọ

he say we must go

‘He says we must go’

Given the parallelism just mentioned, we should note the differences between the two modal meanings. Two of the differences between deontic and epistemic modalities noted by Lyons (1977:823-25) are ‘that there is an intrinsic connection between deontic modality and futurity’ and that ‘deontic necessity typically proceeds, or derives, from some source or cause’. Another major difference is that while a deontic modal normally excludes the presence of the perfect marker, the epistemic modal goes well with it, eg.

26.       Obligation:       *ó  ti gbọ́dọ̀  lọ siré ní ìta

‘He has to have gone and played outside’

27.       Necessity:        Ó    ti gbọ́dọ̀ lọ siré ní ìta

‘He must have gone out to play’

5.         MODAL NEGATION

5.1       INTRODUCTION

Both the event and the modality (Palmer 1979:26) can be negated in a sentence containing a modal verb in Yoruba. When the event in mo lè bá a lọ ‘I can go with him’ or ‘I may go with him’ is negated as in 28, the scope of negation is said to be external, but when the modality is negated as in 29, the scope of negation is internal

28.       n kò     lè  bá        a lọ

I NEG can accompany him go

‘I can’t go with him’

29.       mo lè        má   ba               a lọ

I can/may NEG accompany him go

‘I may not go with him’

5.2       EPISTEMIC MODE

Epistemic gbọ́dọ̀ ‘must’ is outside the scope of both internal and external negation. That is, neither the event nor the modality can be negated by combining the negative verb with the epistemic gbọ́dọ̀ ‘must’. Hence, neither 31 nor 32 is the negation of 30. While 31, which is grammatical, is the negation of the deontic gbọ́dọ̀ ‘must’, 32 is ungrammatical and unacceptable. Even in verbal ‘crossing out’ (Halliday 1970:33) where a previous ‘must’ is specifically denied (and thus ‘mustn’t’ is possible in English), the use of the negative plus gbọ́dọ̀ ‘must’ is still not acceptable. So, whereas 33 is possible in English, 34 is unacceptable in Yoruba.

30.       ó  gbọ́dọ̀ wà nílé

he must be in house

‘He must be in’ (ie. I confidently infer that he is at home)

31.       kò gbọ́dọ̀ sí nílé

NEG must be in house

‘He must not be in’ (ie. it is necessary/important that he is not in)

32.       *ó gbọ́dọ̀ má sí nílé

he must NEG be in house

33.       He must be in - oh no, he mustn’t

34.       ó  gbọ́dọ̀ wà  nílé - *rárá, kò gbọ́dọ̀

he must be in house - No, NEG must

‘He must be in - No, he must not’

This is not to say that 30 cannot be negated. To negate it, we use lè ‘can/may’. This means that the difference between epistemic gbọ́dọ̀ ‘must’ and lè ‘can/may’ neutralizes in negation. Just as in 28 and 29, 30 can be given either an internal or external negation with the use of lè ‘can/may’ as in 35 and 36.

35.       kò  lè  sí  nílé

NEG can be in house

‘He cannot be in’

36.       ó  lè    má  sí  nílé

he may NEG be in house

‘He may not be in’

35 is the external negation of 28, and 36 is its internal negation.

5.3       DEONTIC MODE

Only the event can be negated in a sentence containing the deontic gbọ́dọ̀ ‘must’. O ò gbọ́dọ̀ ṣe é means ‘You are required not to do it’ or ‘I order you not to do it’. The meaning of deontic gbọ́dọ̀ ‘must’ is captured in the following conversation where X tells Y what he must not do.

37.

X:    o     ó     gbọ́dọ̀ wí nǹkan    kan  nípa  ọ̀gá  rẹ  fún ẹnì kan kan

you NEG must say thing one about master you for person one one

‘You must not say anything about your master to anyone’

Y:    kín ló     ṣe    ọ̀gá     mi sà?

Q   FOC it do master me sir

‘What happened to my master, sir?’

X:     o    ò      gbọ́dọ̀ sọ pe   ó    fara      pa

you NEG must    say say he take body cut

‘You mustn’t say he was wounded’

Y:    ọ̀gá      mi  kò   fara     pa sà

master me NEG take body cut sir

‘My master was not wounded, sir’

X:    o    ò      gbọ́dọ̀  sọ  pé  a  gbé  e wá    síhìn-ín lóru

you NEG must say say we carry him come to here late in the night

‘You must not say we brought him here late in the night’

Y:    ẹ   ẹ̀  gbé    e    wá      síhìn-ín     lóru      sà

You (pl) NEG carry him come to here late in the night sir

‘You didn’t bring him here late in the night sir’ (p.142)

6.         MODAL COMBINATION

It is possible to combine either gbọ́dọ̀ ‘must’ or yóò ‘will’ with lè ‘can’. When this happens, lè ‘can’ always follows the other modals, eg.

38.       ó gbọ́dọ̀ le ṣe é

he must can do it

‘He must be able to do it’

39.       n ó     (yóò)  lè ṣe é

I will         can do it

‘I will be able to do it’

7.         CONCLUSION

We have presented in this paper a straightforward semantic analysis of the modal verb gbọ́dọ̀ ‘must’.

The ‘internal complexity’ of modality, according to Chung and Timberlake (1985:241), is great. Because of this, examples have been taken from a narrative text to avoid, as much as possible, a subjective interpretation of the modal verb gbọ́dọ̀ ‘must’.

REFERENCES

Abraham, R.C.  1958. Dictionary of Modern Yoruba. London: Oxford University Press.

Adewole, L.O.    1987.    The Yoruba Language. Published Works and Doctoral Dissertations 1843-1986, ed. By Rottland, F. and R. Vossen.  Hamburg: Helmut Buske.

Awobuluyi, O.  1978. Essentials of the Yoruba Language. Ibadan: UPL.

Awoyale, ‘Yiwọla.    1983.    ‘On the Development of the Verb Infinitive Phrase in Yoruba’, Studies in African Linguistics 14.1:71-102.

Chung, Sandra and A. Timberlake. 1985. ‘Tense, Aspect and Mood’, in    Language Typology and Syntactic Description III: Grammatical Categories and the Lexicon, ed. by Timothy Shopen, pp.202-258.   Cambridge: University Press.

Coates, Jennifer.  1983. The Semantics of the Modal Auxiliaries. London: Croom Helm.

C.M.S.  1913. A Dictionary of the Yoruba Language.  Lagos, Nigeria: CMS Bookshop.

Fagborun, Gbenga.  1985.  ‘Asawo Agbekale Gbolohun Yoruba  (Foreign Structures in Yoruba Sentences)’,  Paper Presented at the Department of African Languages and Literatures, Obafemi Awolowo University, Ile-Ife, Nigeria, 3rd. June, 1985.

Haegman, Liliane M.V.    1983.    The Semantics of ‘will’ in Present-day British English: A Unified Account. Brussels: Paleis der Academien - Hertogsstraat 1.

Halliday, M.A.K.  1970.  ‘Functional Diversity in Language as Seen from a Consideration of Modality and Mood in English’, Foundations of Language 4:225-242.

Hermeren, Lars.    1978.   On Modality in English: A Study of the Semantics of the Modals.   Lund Studies in Linguistics 53.  Lund: GWK Gleerup.

Leech, G.  1981.  Meaning and the English Verb. London: Longman.

Lyons, J.  1977.  Semantics, Volumes 1 and 2.  Cambridge: University Press.

Okediji, Oladejo.  1983. Atótó Arére.  Lagos: Longman.

Oyelaran, O.O.   1982.   ‘The Category Aux in the Yoruba Phrase Structure’,   Paper Presented at the 15th West African Languages Congress, University of Port Harcourt, 4-11 April, 1982.

Oyetade, B.A.   1989.   ‘Iṣẹ́da Ohun Aarin Yoruba (The formation of Yoruba mid-tone)’,  Paper Presented at the Odujinrin Memorial Seminar, Dept. of Nigerian Languages and Literatures, Ogun State University, Ago-Iwoye, Nigeria, 5-8 April, 1989.

Palmer, F.R.  1979.  Modality and the English Modals. London: Longman.

Veluppillai, A.  1983. ‘Modality in Jafra Tamil’, Indian Linguistics 44:83-96.

Welmers, W.E.  1973. African Language Structures. Berkeley: University of California Press.

Yusuf, Ore.    1985.    ‘A Functional Explanation for the NI-NP Construction in Yoruba’,   Studies in African Linguistics, Supplement 9:329-334.