Is there anyone who knows how are made purpose clauses of Yoruba? I know that in typological literature they are classified as balanced (= the verb form of the purpose clause may also occur in a main declarative clause). However, I cannot find examples of this fact. Rowlands (1967) gives this example: sọ̀rọ̀ sókè dáadáa kí m baà lè gbọ́, but he doesn't provide glosses! Could someone analyze it morpheme by morpheme? Otherwise can someone provide relevant literature on the topic? Thank you a bunch :))



2 Answers 2


The broader semantic domain of purpose in Yoruba (as in any other language) can be realized by different kinds of expressions. In Yoruba this would range from clause complexes to single clauses containing verb (or VP) complexes, better known as serial verb constructions (SVCs). I’m gonna attempt an illustration of this range (or cline, if you will).

The kí dependent clause. The kí clause is the main construction for purposives. Starting with your example from Rowlands (1969),

sọ̀rọ̀ sókè dáadáa kí n ba lè gbọ́

(say-word to-top good-good thatIRR I PTCL can/may hear)

speak up good so I can hear

here you’ve got the independent clause sọ̀rọ̀ sókè and the dependent clause kí n ba lè gbọ́. The ‘independent’ version of the latter is

mo lè gbọ́

I can hear (n is the 1sg ‘syllabic nasal’ variant of mo in a kí clause)

Kí is a complementizer that marks what could all probably be subsumed under irrealis mood in Yoruba. For purposive clauses, it is equivalent to the subjunctive that in English found in conjunctions like so that or in order that.

Baà (mostly just ba in positive clauses) is a modal particle (PTCL) that co-occurs with the modal verb lè in purposive clauses, but in regular speech it can be chucked:

sọ̀rọ̀ sókè kí n lè gbọ́

speak up so I can hear

The modal verb lè can also be removed, but this gives the dependent clause less of a purposive and more of an imperative/hortative reading (still all irrealis):

sọ̀rọ̀ sókè kí n gbọ́(!)

speak up, let me hear(!)

Sometimes speakers code-mix like this:

sọ̀rọ̀ sókè so that mo á (lè) gbọ́

(Lit: speak up so that I FUT (be able to) hear)

In this form, future tense marking with is mandatory, in keeping with the general irrealis nature to purposives. I’m not sure the gloss comes out grammatical in English, but the code mix is perfectly fine for Yoruba speakers.

sọ̀rọ̀ sókè so that kí n (lè) gbọ́

is possible, but sounds comical as the listener will sense some tautology in having both the subjunctive that as well as kí in there.

The láti-VP construction. First off, láti- consists of a preposition ní and a VP-nominalizing prefix àti- (ní + àti- contracts to láti- following a couple of phonological processes). This makes láti-VP essentially a PP (ní + àti-VP).

Because láti-VP in the context of purposives translates to to VP (or in order to VP or so as to VP) in English, earlier works in the literature used to analyze láti- as a complementizer equivalent to the infinitival to. Problem though is that regarding verb forms, the very idea of finiteness or non-finiteness for Yoruba verbs remains in some debate. The orthography still retains the English influence and has láti- written as a word (láti), removing the prefix hyphen as it is easier on the eyes.

mo wá ibí láti wá(á) jẹun

(I come here [to [come eat]NOM])

I came here to eat

mo lọ ibẹ̀ láti lọ(ọ) jẹun

(I go there [to [go eat]NOM])

I went there to eat

Like the non-finite dependent clauses in English, láti-VP constructions are subjectless (or in theoretic terms, have covert subjects). The clause complex could thus be analyzed in terms of subject or object control.

To express the English purposive (in order) for NP to VP in Yoruba, you’d have to use the kí clause construction which affords you an overt subject. A literal translation of for NP to VP (which is fún NP láti VP) is just plain off:

English: You gotta speak up for me to hear

*O gbọ́dọ̀ sọ̀rọ̀ sókè fún mi láti gbọ́ (ungrammatical literal translation. Mi = (1sgACC))

?O gbọ́dọ̀ sọ̀rọ̀ sókè fún mi kí n ba lè gbọ́ (slightly better, basically you gotta speak up for me (so) that I may hear)

O gbọ́dọ̀ sọ̀rọ̀ sókè kí n ba lè gbọ́ (best, as mi and n are co-referential so you can chuck the fún NP: you gotta speak up that I may hear)

Surface SVCs. Generally speaking, when láti is omitted, you get more natural SVC-type utterances. We’re now arriving at the VP complexing zone of the cline:

mo wá ibí wá(á) jẹun

(I come here come eat)

I came here to eat

mo lọ ibẹ̀ lọ(ọ) jẹun

(I go there go eat)

I went there to eat

Note the optional extra vowel length (mora) on the second occurrence of the verbs wá and lọ, a feature that is specific to the purposive use of the pair of verbs.

Without the place expressions ibí and ibẹ̀, a speaker can just say

mo wá(á) jẹun (I came to eat)

mo lọ(ọ) jẹun (I went to eat)

In this example,

a ń rán ẹ lọ gbà ìwé yẹn fún John ( = 2sgACC)

(we IMPF send you go get book that for John)

we’re sending you to go get that book for John

attempting to shoehorn láti into these SVCs (like the infinitival to in the gloss) will make them sound less natural.

?a ń rán ẹ láti lọ gbà ìwé yẹn fún John

Again, a kí clause works better:

a ń rán ẹ kí o lọ gbà ìwé yẹn fún John

(we IMPF send you thatIRR you go get book that for John)

There are also apparent SVCs in Yoruba of the eager to please and easy to please ilk that sometimes realize purposive meanings. They feature VPs prefixed by láti-/àti-, or a quite interesting inter-VP high-tone syllable (HTS) marker which has had reams of paper thrown at it in the literature. I can’t do justice to all that here but they’re worth the shout-out.

Just to let you know, I encountered the balanced versus deranked concept from your post for the first time ever, and could barely understand it on first read nor immediately relate it to Yoruba. Suffice it to say that Yoruba is quite the analytic/isolating language so rather than speak of verb forms (there’s zero inflectional morphology there), I proceeded to briefly scan the system of VP-associated functional elements (e.g. TAM and other INFL particles). It was difficult for me off the top of my head to comb through the co-occurence of VPs with these particles within both independent and dependent clauses to identify situations (and work out the extent) of balance or deranking. I hope with my examples and glosses here I’ll have provided you with some food for thought towards that end.




I shall give it a go, but I invite native speakers to correct me:

Sọ̀rọ̀ sókè dáadáa kí m baà lè gbọ́.

speak up wellCOMPARATIVE () I meet/arrive can/may hear

The principal method for purpose clauses in modern Yoruba appears to be:

ki + SUBJECT + baà + VERB

... where ki is a conjunction and baà is a verb meaning "meet" or "arrive", used in serial verb constructions to suggest a sort of completion aspect.

Nonetheless, the conjunction ki + a verb seems to be sufficient to convey a subjunctive/optative meaning:

A (k)pè wọ́n ki o ṣe é.

we call them (ki) (s/he) do it

... which as Bowen (1858) states:

When the verb with 'ki' is preceded by an objective, it may often be rendered by an infinitive.

There is the use of another construction, mentioned in Crowther (1852)

lati + VERB

... where he states:

Ati, lati, “to,” “in order that,” “to the effect that,” is also very much used to express the infinitive, when an intention or object is had in view ; as, Emi nmura lati lọ, “I am preparing to go (or, to the intent of going);” Òjo npete ati rò, “The rain is about to fall;” Okko ṣetan lati sí, “The ship is ready to start.”

I have seen reference to a subjunctive-marking particle gàadà, but I think this is an error, as all references come from one source, and I cannot find a single instance of this word in actual Yoruba literature.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.