In this paper I will catalog and describe the morphology of verbs in Luo as spoken in Kenya by our class informant Atieno Adaala. In addition to describing the morphology, I will attempt to account for the placement of those morphemes in Optimality Theory, which I believe will help me provide a coherent account of morpheme placement.

Let me begin by defining some terms that I will be using in my description and analysis. I will define a verb stem as that form that is used as the present infinitive. It is a form of the verb with no pronominal prefixes, and typically has a final {-o}. The verb root is the form that corresponds to the singular imperative. For those verbs that have a final {-o}, the verb root is lacking it. For those verbs which have no final {-o} in the present infinitive, the verb root and the verb stem are identical.

(1) Present Indicative

A. Present infinitive

"to write"             ndiko                                     "to study"          somo
"to come"             biro                                       "to sell"             uso

B. Present indicative paradigm

I write                  andiko                                    I study            asomo
you write              indiko                                    you study         isomo
he/she writes        ondiko                                   he/she studies   osomo
we write              wandiko                                 we study          wasomo
you (pl.)write       undiko                                    you (pl.) study  usomo
they write            gindiko                                   they study         gisomo

(I have included complete the paradigms of the verbs included in this paper in the appendix.)

As one can see from these examples, the present infinitive is formed from the verb stem (present infinitive) and a pronominal prefix. In sentences that have overt subjects, the pronominal prefix is omitted.

C. Overt subjects with present indicative

                       [ojo: uso]                   "Oyo sells"

The present indicative forms also give not only the English "he writes" meaning, but also "he is writing" meaning. There is no separate present progressive form.

(2) Plain Past

A. Plain past conjugation

I wrote                nandiko                     I studied               nasomo
you wrote            nindiko                      you studied          nisomo
he/she wrote        nondiko                     he/she studied      nosomo
we wrote             newandiko                we studied            newasomo
you (pl.) wrote     neundiko                   you (pl.) studied   neusomo
they wrote           negindiko                   they studied         negisomo

B. Plain past with an overt subject

                 [ojo: nouso]                  "Oyo sold"

The plain past forms similarly to the present indicative, and likewise encompasses the progressive aspect of English. It is composed of the verb stem preceded by the pronominal prefix. The past tense itself is marked by the addition of the prefix {n-} in the singular forms, and {ne-} in the plural. Unclear from just this paradigm is the status of the {ne-} form. Is the vowel epenthetic, or is it an additional overt plural marker? With overt subjects, unlike the present indicative, the pronominal prefix is maintained in the past.

(3) Additional tenses with one prefix

(For the entire paradigm, see appendix.)

A. Future tense

I will write                abondiko                   I will study              abosomo
we will write            wabondiko                we will study           wabosomo

B. Perfect tense/aspect

I have written           asendiko                    I have studied          asesomo
we have written       wasendiko                 we have studied       wasesomo

C. Present subjunctive

I should write          onegandik                  I should study          onegasom
we should write       onegowandik            we should study       onegowasom

D. Present optative

I may write              dipopandiko               I may study            dipopasomo
we may write          dipopwandiko            we may study         dipopwasomo

E. Conditional

I would write           dandiko                      I would study         dasomo
we would write        dewandiko                 we would study      dewasomo

F. Jussive

let me write!            andik!                          let me study!           asom!
let us write!             wandik!                       let us study!            wasom!

The tense marking of Luo, at first glance, begins to seem a bit irregular. Not only is it apparently unpredictable morphologically which forms use the verb root over the verb stem, the future tense is marked by a prefix {bo-} that comes between the pronominal prefix and the verb: an order that is contradictory to the morpheme placement of the plain past. The perfect tense is marked similarly with {se-} coming after the pronominal prefix. The other aspects given in C. through E. have morphemes that behave like the past marker and come prior to the pronominal prefix: the subjunctive is marked by {oneg-} or {onego-} in the plural; the optative by {dipop-} throughout; and the conditional by {d-} in the singular or {de-} in the plural. One additional thing of note is that of all these forms in A. through E., only the subjunctive takes the verb root rather than the verb stem. What happens when morphemes are concatenated and more than one must be used?

(4) Tenses/Aspects with Multiple Morphemes

A. Past Perfect/Pluperfect

I had written             nasendiko                     I had studied        nasesomo
we had written         newasendiko                we had studied     newasesomo

B. Past Subjunctive

I should have written      nonegandik             I should have sold       nonegaus
we should have written   nonegowandik       we should have sold     nonegowaus

These two paradigms suggest two things: first, that the tense morphemes need not occur together, as in the past perfect; second, that the [ne-] form of the past tense is probably the result of an epenthetic vowel and not a distinct plural form. The past subjunctive, where the past marker falls before a vowel-initial morpheme, the [e] is not present, and as we can see from the forms of the vowel-initial verb [uso] "sell", there is no particularly strong prohibition against vowel-vowel sequences. Also note that the past subjunctive, like the present subjunctive, is formed with the verb root.

How does the negative morpheme interact with the others?

(5) Negativization

A. Present negative

I don't write              okandik                     I don't study             okasom
we don't write           okwandik                 we don't study           okwasom

B. Future negative

I will not write           okabendiko              I will not sell               okabeuso
we will not write        okwabendiko           we will not sell            okwabeuso

C. Past negative

I didn't write            nokandiko                 I didn't sell              nokauso
we didn't write        nokwandiko               we didn't sell           nokwauso

D. Imperative/Jussive negative

don't let me write!             kikandik!

don't let us write!              kikwandik!

We can see from these forms that the negative marker is not the same everywhere. The imperative/jussive negative takes {kik-} rather than the {ok-} found elsewhere. In both cases, however, there is no epenthesis when consonants come together. Like the tenses in (4) that take two tense markers, there is no regular order of tense marker and negation marker.

(6) Synthesis

In all of the verb forms above, there is one consistency: the verb is always last in the string of morphemes. Whether the tense marker comes before the pronoun, or after, however, seems largely unpredictable. In the following table, I list all the verbal morphology and indicate where they come with respect to the pronoun in the verbal complex.

Precede Pronoun                 Follow Pronoun

n(e)-       past                      bo-              future
ok-         negative                se-               perfect
oneg(o)- subjunctive
dipop- optative
d(e)- conditional
kik- jus. negative

From this table, we can see that all tense/aspect markers that precede the pronoun are consonant-final. Furthermore, we can also see that those markers that have a final voiced consonant sometimes occur with epenthetic vowels. The nouns of Luo are often consonant-final, and undergo a regular devoicing process, but when followed by a vowel, as in plural-formation, the underlying voicing contrast may be revealed.

         tooth          lak            teeth          lake
         tortoise      opuk         tortoises    opuge

Based on this, one can argue that voicing is only contrastive in onset position. It is also the case that coda consonants are generally dispreferred in Luo. When possessive pronouns are added to noun stems that are consonant-final, there is often an epenthetic vowel added, particularly if the consonant is voiced underlyingly.

              my comb            ragonda           my soil              lopa
              their comb          ragond(w )gi      their soil           lobgi

What is plain from both the verbal and nominal examples is that, despite some previous claims (Harris 1998), coda consonants are not restricted to only word-internal situations. While some coda consonants may occur, they only occur at morpheme boundaries, butthere is, however, some restriction on what consonants can surface in codas. The movement of morphemes can be attributed to a desire, not only to preserve voicing contrasts, but also to create better syllables. There is no voicing contrast in nasals to be preserved since nasals do not run the risk of being devoiced syllable-finally. More likely, is the notion that while coda consonants are not forbidden, when possible, they are avoided. When a consonant-final morpheme can come before a vowel-initial pronoun, it can produce a well-formed CV syllable. When there is no advantage to it, then the morpheme is not moved from its syntactically preferred position. I have considered and rejected the notion that Luo is an isolating language, and that these are word-, as well as morpheme-boundaries. If coda consonants are permitted word-finally, there would be no justification for any epenthetic vowels, and less for movement driven by syllable structures.

(7) Analysis

In syntax, the verb phrase takes on the following structure, where S=Sentence; NP=NounPhrase; NegP=NegationPhrase; TP=TensePhrase; VP=VerbPhrase:

S

/\

NP VP

•  

  • /\

    NegP V' .

    • /\

      TP V'

      • /\

        V NP

If the morphology were to follow a straightforward concatenation process based on the syntactic structure, we would expect the tense markers all to fall inside the negative marker and the verb stem/root, and for the negative marker to fall inside the pronominal prefix and the verb. I do not know what is the preferred order for tense vs. aspect, so I will assume that it is either not relevant, or that aspect (AspP) would fall between NegP and TP. This syntactic account would give us the following constraints:

AlignTenseL: Align tense/aspect markers to left edge of the verb

AlignNegL: Align the negative marker to the left edge of the verb

AlignPronL: Align the pronominal prefix to the left edge of the verb

With the following harmonic ranking:

AlignTenseL >> AlignNegL >> AlignPronL

Ranked around these, are constraints that speak to syllable structure and faithfulness to the input, and certain output forms.

Onset: syllables must have an onset

NoCoda: syllables must not have codas

Ident [voi]: preserve the voicing feature of an input segment in its corresponding output segment

Dep: all output segments must have a corresponding input segment

Max: all input segments must have a corresponding output segment

*Lar & NoCoda: no voiced obstruents in coda position

OO-FaithV : all singular/plural verb forms should preserve the vowels in the 3rd person output form that agrees with it in number.

The final constraint-ranking, therefore, should be the following:

*Lar & NoCoda, Max, OO-Faith Ordering >> OO-Faith Order >> OO-FaithV, Ident [voi] >> AlignTenseL, NoCoda, Dep >> AlignNegL >> AlignPronL, Onset.

Tableau 1. Present indicative: {wa}+ {ndiko}

The AlignPronL constraint is violated because the pronoun is not to the immediate left of the verb. Now that we know this works in the simple case, what about a real test? ranking

Tableau 2. Plain past: {a} + {n} + {ndiko}

Tableau 3. Plain past: {gi} + {n} + {ndiko}ranking

In these tableaux, the AlignTenseL constraint is violated when the tense marker does not fall to the immediate left of the verb. Max is violated when a consonant in the input is not parsed into the output. NoCoda is violated when a consonant is parsed in a CVC syllable. Dep is violated when a vowel is inserted. Notice that it is the AlignPronL constraint that decides the winning candidate. This is a case of emergence of the unmarked. Again, I have shown that this ranking will work for a single morpheme that is consonant-final. What about two morphemes?  ranking

Tableau 4. Past perfect: {a} + {n} + {se} + {ndiko}

I have marked two violations for AlignTenseL in c. since there are two tense morphemes that are not aligned to the immediate left of the verb. {se-} also takes a third violation because there are two morphemes between it and the verb (in this tableau it is not necessary, however); one violation is incurred for each leftward movement away from the verb, as in the following tableaux.   ranking

Tableau 5. Past subjunctive: {gi} + {oneg} + {n} + {ndiko}

The constraint *Lar& NoCoda is violated in c. because there is a voiced obstruent parsed as a coda. Ident [voi] is violated in b. when the final consonant of {oneg-} is devoiced in coda position. As before in the simpler cases, it is AlignPronL that ultimately decides the winning candidate.  ranking

Tableau 6. Past negative: {wa} + {ok} + {n} + {ndiko}

Tableau 7. Past subjunctive: {wa} + {oneg} + {n} + {ndiko}   ranking

In Tableau 7, we see that the high-ranked output-output constraint forces paradigm regularity, by requiring the epenthetic vowel, and it is once again the AlignPronL constraint that exerts its influence.

One more constraint will be necessarily be added to force another kind of paradigm regularity. Another OO-Faith constraint that will require vowel-initial verbs to follow the morpheme-ordering of consonant-initial roots. This is certainly a justifiable constraint since consonant-initial verbs are far more the rule than the exception. Indeed, when I asked the informant to come up with some, she had a difficult time of it.

The final constraint-ranking, therefore, should be the following:

*Lar & NoCoda, Max, OO-Faith Ordering >> OO-Faith Order >> OO-FaithV, Ident [voi] >> AlignTenseL, NoCoda, Dep >> AlignNegL >> AlignPronL, Onset.

(8) Phonological behaviour

In my paradigms included in the paper, I have used only verbs whose vowels are [+ATR]. Luo undergoes a process of vowel harmony, whereby ATR features are spread from roots to affixes. I chose [+ATR] vowels to make my transcriptions easier for this paper, since my primary focus has been on ordering and not on the phonological behaviour of vowels. There also seems to be some other things going on. In some cases, the mid-vowels [e] and [o] appears to alternate with the high, [-ATR] vowels [w ] and [� ]. This is particularly true at the end of words and in the environment of other high vowels. I have also noted that there is some loss of contrast between [-ATR] and [+ATR] vowels before nasals and at the ends of words. What the details of the alternations are I have not completely determined yet, but they do impact on the vowels, particularly in my transcriptions of the pronominal prefixes in the conjugations of [ndiko], and with the final [-o]'s of some of the verbs. These alternations will have to be accounted for at another time. I will also note here, that there is a process of optional glottalization when a voiceless stop precedes a voiced one within a word. It may be a reflex of cutting on the voicing suddenly within a stop-stop sequence, particularly when the stops are of the same place of articulation. I have transcribed it in the paradigms in the appendix as an apostrophe.

The second-person pronouns will also behave like glides when it falls in front of a vowel-initial verb, especially one that is non-high. I have not included any of the relative paradigms here.

(9) Remote vs. Immediate Past

There is another tense distinction that is marked in Luo that is not determined by segmental morphology. The remote vs. immediate past distinction is marked by tone alone. The final {-o} of the verb stem is low in both cases, with the remote past carrying a high tone on the verb root and all preceding syllables, while the immediate past has a falling-rising tone that is spread backward over the verb and its affixes; the more syllables, the more rises and falls. This is also seemingly consistent between the plain past remote vs. immediate past distinction and the pluperfect remote vs. immediate past distinction.

(10) Additional forms for which I have incomplete paradigms

The data I have gathered so far is far from an exhaustive account of the verb forms in Luo. I have given here below some of the additional information I've gathered related to this topic.

A. Imperative (let G be the velar nasal)

write!                   ndik!                        look!              Gi!
write! (pl.)            ndiku:ru!                  look! (pl.)       Giu:ru!

B. Passive

The book was read.            buk nosesom.               The book was sold.     buk nous.

C. Past conditional (?)

I might have shouted                  dinagokoko

D. To be able

I can/could shout                       anyana gokoko

E. Transitive vs. intransitive

to eat (something)    chamo           to eat              chiemo

F. Repeated action

look repeatedly/a long time                GichaGicha
sit repeatedly                                      betabeta

G. Past possibility?

I might have sold the book;

it could be possible that I sold the book.    nyalo bet nnauso buk.

H. Present causative

I make him write        ami ondiko              I make him sell            ami ouso
we make him write     wami ondiko          we make him sell          wami ouso

I. Past causative negative

I don't make him write          okami ondik
we don't make him write       okwami ondik

(11) Course of future research

This account of verb morphology in Luo is far from complete, but it does yield some testable results. There are a number of other tenses and aspects to inquire about that will demonstrate the validity or falsehood of the claims I've made. Some of the additional tenses/aspects to ask about include: habitual aspect; tenses with the copula; past conditional; and negatives for many of the forms I have already collected. I will also need to complete the paradigms for which I have only partial information. I am also curious to know if infinitives in Luo can be marked for tense, and whether pronouns are always preserved in the third person constructions for which the morpheme is buried within the verbal complex, and how the behaviours of emphatic construction. I would also like to see if there is a regular correlation between vowel-alternations and transitive-intransitive pairs.

(12) Appendix

Below are the complete paradigms available for all the conjugations included in this paper.

A. Future

I will write               abondiko                I will study            abosomo
you will write           ibondiko                you will study         ibosomo
he/she will write       obondiko               he/she will study     obosomo
we will write            wabondiko             we will study          wabosomo
you (pl.) will write    ubondiko               you (pl.) will study   ubosomo
they will write          gibondiko               they will study         gibosomo

B. Perfect tense/aspect

I have written            asendiko                I have studied            asesomo
you have written        isendiko                 you have studied        isesomo
he/she has written      osendiko                he/she has studied     osesomo
we have written         wasendiko             we have studied         wasesomo
you (pl.) have written usendiko                you (pl.) have studied usesomo
they have written       gisendiko                they have written       gisesomo

C. Present subjunctive

I should write               onegandik            I should study              onegasom
you should write           onegindik            you should study           onegisom
he/she should write       onegondik           he/she should study       onegosom
we should write            onegowandik      we should study             onegowasom
you (pl.) should write    onegoundik         you (pl.) should study    onegousom
they should write          onegogindik         they should study          onegogisom

D. Present optative

I may write                   dipopandiko         I may study                  dipopasomo
you may write               dipopindiko          you may study             dipopisomo
he/she may write           dipopondiko         he/she may study         dipoposomo
we may write                dipopwandiko      we may study              dipopwasomo
you (pl.) may write        dipopundiko         you (pl.) may study      dipopusomo
they may write              dipopgindiko         they may write             dipopgisomo

E. Conditional

I would write                dandiko                 I would study                  dasomo
you would write           dindiko                  you would study               disomo
he/she would write       dondiko                 he/she would study           dosomo
we would write            dewandiko             we would study               dewasomo
you (pl.) would write    deundiko                you (pl.) would study      deusomo
they would write          degindiko                they would study            degisomo

F. Jussive

let me write!                  andik!                    let me study!                    asom!
write (?)                        (indik!)                  study! (?)                         (isom!)
let him/her write!            ondik!                   let him/her study!              osom!
let us write!                   wandik!                 let us study!                      wasom!
write! (pl.)                    undik!                    write! (pl.)                         usom!
let them write!               gindik!                   let them study!                   gisom!

G. Past Perfect/Pluperfect

I had written                   nasendiko             I had studied                      nasesomo
you had written               nisendiko             you had studied                  nisesomo
he/she had written           nosendiko            he/she had studied              nosesomo
we had written                newasendiko        we had studied                  newasesomo
you (pl.) had written        neusendiko          you (pl.) had studied          neusesomo
they had written              negindiko             they had studied                 negisomo

H. Past Subjunctive

I should have written              nonegandik           I should have sold             nonegaus
you should have written         nonegindik            you should have sold        nonegius
he/she should have written     nonegondik           he/she should have sold    nonegous
we should have written          nonegowandik      we should have sold          nonegowaus
you (pl.) should have written nonegoundik          you (pl.) should have sold nonegou:s
they should have written        nonegogindik         they should have sold       nonegogius

I. Present negative

I don't write                  okandik                      I don't study                     okasom
you don't write              okindik                       you don't study                 okisom
he/she doesn't write       okondik                     he/she doesn't study          okosom
we don't write               okwandik                  we don't study                   okwasom
you (pl.) don't write       okundik                     you (pl.) don't study           okusom
they don't write             ok'gindik                    they don't study                 ok'gisom

J. Future negative

I will not write                okabendiko                   I will not sell                 okabeuso
you will not write            okibendiko                   you will not sell              okibeuso
he/she will not write        okobendiko                  he/she will not sell          okobeuso
we will not write             okwabendiko               we will not sell                okwabeuso
you (pl.) will not write     okubendiko                  you (pl.) will not sell        okubeuso
they will not write           ok'gibendiko                 they will not write           ok'gibeuso

K. Past negative

I didn't write                    nokandiko                    I didn't sell                  nokauso
you didn't write               nokindiko                     you didn't sell               nokiuso
he/she didn't write           nokondiko                    he/she didn't sell           nokouso
we didn't write                nokwandiko                 we didn't sell                nokwauso
you (pl.) didn't write        nokundiko                    you (pl.) didn't sell        noku:so
they didn't write              nok'gindiko                   they didn't sell              nok'giuso

L. Imperative/Jussive negative

don't let me write!                 kikandik!
don't write!                           kikindik!
don't let him/her write!           kikondik!
don't let us write!                   kikwandik!
don't write! (pl.)                    kikundik!
don't let them write!              kik'gindik!

References:

•  
  • Biddulph, Joseph. 1985. An introduction to Luo, a Nilotic language of Kenya (with remarks on Acoli). Wales: Joseph Biddulph Publisher.

    Harris, John. 1998. The right edge of word and the left edge of syllables. Handout from paper given at NSF Workshop: Syllable Structure and Gesture Timing, LP98, Columbus, Ohio.

    Okoth-Okombo, Duncan. 1982. DhoLuo morphophonemics in a generative framework. Berlin: Reimer.

  •  
• CV

courses
teaching
writing
images
interests
the pewter gallery
giftedness 
research
quotes
school
travel
godless weddings
old web designs

Betsy McCall
Field Methods 2
29 March 1999