TENSE AND ASPECT IN SESOTHO
' Balillo ncrolocg
B. A,, uriversity of Botswana Lesotho and Swaziland
A THESIS SUBMITTED IN PARTIAL PULP ILLXENT OF
THE REQUTRERFNTS FOR THE DEGREE OF
HASTER OF ARTS
in the Departmlnt
of Languages, Literatures and Linguistics
STHON PRASER UNIVERSITY
All riqhts reserved. This thesis may not be
reproduced in who12 cr in part, by photocopy or other means, without permissionof the author. APPROVAL
Name : Malillo Morolong
Degree : Master of Arts
Title of Thesis: Tense and Aspect in Sesotho
Examining Coounittee:
Chairman: R. C. DeArmond
B. E. Newton Senior Supervisor
4( ]w Knowles
P. Stiser External Examiner Associate Professor Simon Fraser University
Date Approved: Sept. 11, 1978 PARTIAL COPYRIGHT UCENSE
I hereby grant to Simon Fraser University the right to lend my thesis or dissertation (the title of which is shown below) to users of the Simon Fraser University Library, and to make partial or single copies only for such users or in response to a request from the library of any other university, or other educational institution, on its own behalf or for one of its users. I further agree that permission for multiple copying of this thesis for scholarly purposes may be granted by me or the Dean of Graduate Studies. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.
Author !
(signature )
(name )
(date) ABSTRACT
TENSE AND ASPECT IN SESOTHO
In this thesis ar. analysis of Tens2 and Aspect ir. Sesotho is undertaken, The framework adopted is tbe one introduced by
Reicfienbach (1947) In g&gmsnt~ of Svmbo1ic_ Loqic- with subsaqu~nt modifications and elaborations in Hornstein 11 377) and Smith (1978).
chaptes 1 is i~troductory. It contains background information about Sesothc morphology; scope of the study and definitions.
Chapter 2 presents an aralysis of the ~tgg+ile. It is suqgested that this fcrm should be reqardsd as an immediate p3st tense instead of 'perfect ', This reclassification provid2s a
solution to the problem of interaction with present and past time adverbials.
chapter 3 deals with ths analysis of the hletsteg. This tense is regardea as a retgcte tens?. Cooccurrence restrictions involving this tense and the imm2diate tenses are outliznd,
Chaptcr 4 contains an analysis of the ne+stem tense, The claim made is that this is the past tense correspondirg tc the
present tenss. Distinctioas are made between this tense and iii other tenses,
Chapter 5 examines coGccurrence r~strictions between aspectual markers and different tenses, Tests are proposed in order to determine aspectual implications inherant in different tenses.
The final chapter examines the a2pficability of the
Reichenbach/Hornstein analysis to Sesotho, The conclusion made
is that the imediate/remote distiction plays an important role in the Sesotho tense system and that this must be incorporated into any theory of tense which may adequately account for
Sesotho tenses,
The thesis cculd form the basis for encoding into appropriate formalisms the semantic and syntactic facts of Sesotho tenses. ACKNOHLEDGEMENTS
I wLsh to express my gratitude to ?he members of my supervisory Coinmi ttee, my senior supervisor, Professor Brian
Neut~n for suqgesting ths topic and providing the biblioqraphy,end for many hours of discussior and patiect suparvision. I am thankful to Professor John Kaowles fcr his comments and ancouraqement.
Ply qratitude also goes to Professor Alfredo Hurtado for his insiqhtful comments and sugqestions.
I am thankful to all my friends for their encouragement. I atn mostly indebted to Ma tseliso Moletsane, Victor Ketso,and
Thabang Ramotete for their invaluable assistance in checking my sentences.
1 am greatly indebted to th? National University of Lesotho
for granting me study lzave for the duration of my study, and tha Canadiao International Dedvelopment Aqency for financFal support.
Finally I would lika to say "thanks" to Matseliso for typing the thesis and EC Chondoma for his brains behind FMT. TABLE OP COYTBNTS
Abstract ...... iii ~cknouledgements...... v List of Tables ...... me...... m..ix CHhPTER 1 ...... a..*oI
INTRODUCTION ~~~~~~~~~m~~~~~~*~*m*~~~~~m*~~e*~~~~~e~~~~~~~~1 1.0 Background Information ...... 1 1.1 Orthography ...... 1 1.2.0 8orphology ...... 5 1.2.1 Tense- aspect- modal narkers ...... 6 1.2.2 Tense ...... 7 1.2.3 aspect ...... 8 1.2.4 Modal ...... 8 1.2.5 Derivational Suffixes ...... 9 1.2.6 Inflexional Suffix ...... 10 1.3 Problem and Purpose ...... 10 1.4 Choice of Framework ...... 10 1.5 Scope of the Study ...... 13 1.6.0 Definitions ...... 14 1.6.1 Notations used with SRE ...... 15 1.6.2 SRE Hodifications ...... 15 1-7 Tenporal adverbs ...... 17 Tense And Aspect ...... 20
CHAPTER 2 ...... o.ooo23 2.0 Introduction ...... *...... *.23 2.1 Previous Studies on perfect in sesotho ...... -24 vi 2.2 Comparison betweer, -ila acd non -if9 pressnt ...... 27 2.3 The Use O•’ -ile with Tezporal Adverbs ...... 31 2.3.1 -il~as Grammaticalizsd Sta tive ...... 33 2.3.2 -i10 As Perfectivs ...... 35 2.3.4.1 -il~as Immediate Past ...... 36 2.3.4.2 Distirction between tsca and -il2 tenses ...... 39 CHAPTER 3 ...... 0.49 3.1 Semote Past Tense: Iletstem ...... 49 3.2.0 SCme Cooccurrence Rsstric %ions Betweer S tem+ile And Ile+sten Tenses ...... 56 3.2.1 The Proqressive Sa Still ...... -56 3.2.2 Immediate And Relnote Tecsss In Compleloent Seotences -58 CHAPTE2 4 ...... 62 4.1.0 Ne+verb Tense ...... 62 4.1.1 Is Ne Asp~ctualOr Tense Warker? ...... -64 4.1.2 Some Correspondences Between the Present Tense and Ne+stem Tense ...... 67 4.1.3 Differences/similarities Betvaen Ne+stem And Other Past Tenses ...... 71 4.2.0 Ns+past Tenses ...... 77 4.2.1 Netpast In elation To Haaning And Cooccurrence
Restrictions Of Simple Past Tensos ...... 79 . CHAPTER 5 ...... 86 5.1.0 Aspect ...... 86 5.1 .1 Perfective/'Imperfect Ive Oppcsit icn ...... 87 5.1.2 Durativt Vs Nondura tive Oppositian ...... 90 vii 5.1.3 Stative Vs Nonstative Opposition ...... 95 5.2.0 Aspectual narkers ...... 33 5.2.1 Sa still ...... I00 5.2.2. Sa+tsoa+stem ...... 100 5.2.3 Sa+stem+ile ...... 100 5.2.4 Sa+ile+stem ...... 101 5.2.5.1 Ne+sa+stem ...... 102 5.2.5.2 Ne+sa+tsoa+stern ...... 102 5.2.5.3 Ne+sa+stem+ile ...... 102 5.2.5.4 *ne+sa+ile+stem ...... 103 5.3.1 Se Already ...... 103 5.3.2 *se+tsoa+stem ...... 704 . 5.3.3. Se+stem+ile ...... lo4
CHAPTER 6 ...... 6.1.0 Gn The Applicability of Reichenbach/Hcrnst+in Analysis tc Sesotho ...... I08 6.1.2 On the Interaction Retween Tense and Temporal Adverbs
6.1.3 On Aspect ...... 115 6.2 Conclusion ...... 116 BIBLIOGRAPHY ...... *...... 117 viii List 6f Tables able I .Sesothc Phcnemes ...... 2 able I1 - Tenses in relation to aspectual features ...... 106 Sesotho is a Southern Bantu language spoksn by over 3 nillion people in Lesotho and parts of South Africa. In literature the language is often referred to as Southerc Sotho in order to distinguish it from Pedi {Horthtrn Sctho) s~oken in Northern ~ransvaal and Tsvana (Western Sotho) spoken in
Botswana,
The present thesis is an examination of Xense and Aspect in spoken Sesotho. Our data does not depend on written materials such as samples from newspapers, novels, plays etc., by Sesotho writers, The analysis may, however, be applied to the written language.
1.1 Qsthoqravhy
Since the present thesis will be mostiy useful to language learners rather than linguists, practical orthography, as aqainst phonetic transcription, will be used. Following is 'the table of Sesotho segments in phonetic transcription acd a short list of the orthographic correspondences, CONSONANTS :
STOPS
NASALS
FRICATIVES
NON- FRICATIVES
AFRICATES
FRICTION-
CONTINUENTS
TABLE I - Sesotho Phonemes TABLE I - Sesotho Phonemes (cont,)
gw.enta [dl belongs to [l] phoneme. It oocurs when [I] is followed by high vowels [i ] and [ u], and are written ny and ng respectively. It must be noted that nasals become syllabic if not folloved by a vowel. A sequence of two nasals is written as apostrophe nasal (e,g. Ipp for ppq) .
is orthographic representation of s
n Dz n ,* n tsh
4 n C
88 ch n x
88 t.b
I tbh Q I8 C
•÷h II (3h
e It Y
0 09 Y
There are severdl complex consonant alterations in the languaqs
(see Doka E nofokeng 1957, Gnilia 197 1 etc,)
Sesotho is prefixing and suffixing, Althouqh a cumber ef words are usually uninflected, a greater number are dzrived by iaorpho1oqical processes such as prefixation, suffixation, reduplication and cornpounding (for a detailtd study of these processes see Doke and nofokeng 1957, Guma 1971 etc.).
Houn class prefixes play an important part in detarmininq agreement with the verb, adjectiva, possessive and relative stems. Verbs may incorporate a number of prefixes and suffixes. norphophonemic changes connected with prefixation and suffixation are beyond the scope of the present study and thus not presented. The relevant affixes of verbal unit may be summarized as follows:
1. A - (a) - (OF) - verb stem - (IS) - (DS) - (IS) - a where
A stands for subject verb agreement H for tense-aspsct -modal forms
OP for cbject pronoun DS for derivational suffix (es)
IS for inflexioxial suffix
-a for the verb final vowel which may be replaced by either a negative, subjurctive, imperative vith OP etc.
The subject verb agreeaent is of no immediate concern since it presents no problem for our analysis. Tne tensa-aspsct nodal forms deserve an explanation since we will come across them quite often.
1.2.1 Tense- aspect- godal garkers
Ye assume that tense aspect and modals are constituents of the verbal unit in view of the strong degres of cohesion between these markers and the verb . Thus while it is possible in English to have an adverb between the first auxiliary and the VP as follows:
2 nary will, probably, have spoken to John It is not possible to have any lexical item between the tense- aspect-modal markers and the verb. e.g.
3. *Haria o-ile, hantle-ntle, a-bua 1s-Jchanne
Hary A past in fact speak vith John
*&ary did, in fact, speak to John.'
Uhile in other Bantu Languages the ver~alunit is written as one word, the currently accept~dsystem of writing Sesctho is disjunctive e .g.
4 (af ke ne ke sa ja I past I still eat
* I was still eating', Compare Zulu:
(b) Bengisadla m 1 still eat
'1 was still eating,'
Hyphens will be used ia this study to indicatt bound forms,
7-2.2 Tense The following forns are used to express tense.
5 ti) ---tsoa 'past', This form which is also used as a verb 'come fromt is used to express irarnadiate past.
(ii) ile- (not to be confused with suffix ile) is the past
of gp 'go'. It expresses remote past.1
(iii) is also used to express past, Compound tenses may be
formed by using this form with 2202 and j&,
(iv) _tl_h 'come9 is used to express future,
Tense category fills two slots the first cf which is filled
by ne and the second by , fig and uq. We do not imply that these are definitely deep structure positiors. What we have been
mostly concerned with is determining the correct positicn cf tense, aspect and modafs in surface structure. The followinq is surface characterization of the verbal unit, tsoa batla il e anela
tl a
The presence of the first tense position is supported by constructions such as the following:
7, Ba-ne ba-ile ba-b ua le-eena
they past A past speak with him
'They had spoken to him.' in which ne is used with i&.
The follovinq forms may be used to express aspectual distinctions: *stills
8 already '
' habitual*
--ke loccasional* These will be dealt with in Chapter 5. This category together with tense and aspect, has boec dealt with under the heading 'Deficient verbsg (see Doke acd
~ofokenq, Guma etc,). The term deficient is a syntactic notion vhich means that such verbs require a complementary vert to comple te their predication, ue have separated modals from tense/aspect markers since modals may be used with tense/aspect markers. There may, however, be some cooccurrence restrictions between tense and certain modals depending on the semantic features inherent in such modals, e,g. Ma'always' may occur with gg but not fig, Follouing is a list of some modal verbs in Sesotho: 9 . &&&a 'want ' h&h 'always* ---lala 'act t hrough the night' ---atisa *do frequently ' ---anela 'manages ---t&anela 'do of necessity' exaeple of use:
10, Barutuoa ba-ne ba-thanela ho-bua sakhooa
Students A past do of necessity to speak English 'The students were supposed to speak English',
There is abundant evidence vhich suggests that modals form a separate constituent frcm ordinary verbs. This falls beyond the scopa cf present study, and will not be dealt with here. 1,2*5 De~l~2i~~qhSuf fiqes The verb stam Bay take one or more derivationd suffix3s . The32 include: 11. Benefactive -el- Causative -is-
Passive -0 -, -uo- Reciprocal -an-
These suffixes present an interesting aspect of interaction betwasn derivational morphology and syntax, yet to be studied. In so far as tense and aspect is concerned, these suffixes may be accounted for by the analysis used for sliuple sentences,
The suffix which may follow a stem or stem 4 derivational suffix, is used to express immediate past tense. This suffix may be realized as -*,-3, -~e,-=a, --, -its€, -ere,- and -ntse depending on phonologica!. processes which do cot concern us here.
1.3 Probl~~A@ PuruSSg
In many studies on Sesotho graremar tense and tem~oral adverbs have been treated as se~arateentities. No mention has been aade about possible interaction between different tenses and temporal adverbs-2 the main purpose of the study is to assess applicability of Hornstein/Reichenbach analysis to sesotho with special reference to such interaction.
The method of analysis followzd in the present study is the
one introduced by Reichenbach (1947) in blame^& of a&~ljc ~oqicwith subsequent modifications and alaboration in Hcrnsteir!
(1977) 'Touards a Theory of Tense*, and Smith (1978) 'The Syntax
and Interpretation of Temporal EXF~~SS~O~Sin English*.
The existence of regular cooccurrence between tense-forms and te~poral adverbs has long been noted. In a footnote to m~ts(1965: 216) Chonsky notes the suyqestion by Barbara Hall which proposes that " .,, Eacb element of the auxiliary has associated with it certain characteristic adverbials that may
(or in thecase of present tense, must) cooccur with this auxiliary element .."
Kiparsky (1968) proposes that tense and adverb be treated as a single category in underlying structure (see also HcCawley , 1971 Gallaqher, l97O), This analysis regards present
adverbials as source for present tense and Fast adverbials as
source for past tense, The probletn arises with future sentencfs
which have future adverk with present tense. The future
adverbial cannot be regarded as source of present tense in thess
cases. The analysis encounters a similar problem in tho case of sesotho tsoa*stem and ile+ster tenses, As will be noted latsr,
these tenses may cccur with either present or past advarh. It 11 uill be difficult to decide which adverb is the source of tense in such cases,
Huddleston*s analysis (1969) which appeals to deep ters~ would encounter more ol: less the same problam. In classifyiny the verbs according to restrictions they impose on the tense selection &f would have to be classified as takicg both presant and past tenses.
An analysis which divides tense into past and non-past distinctions is likely to run into problam in dealing with stea+ilg tense as it occurs with both past and present advarbs (for this analysis see Hirtle (1975) and references therein)) .
Our study is essentially seaantically based. A semantic approach to the subject bas been found neczssary in view of the fact that different syntactic ex~ressions may have the same semantic realization. Consider the case of present used with future advertial, and the auxiliary tkg used to express future in the f ollouing sentences:
Sello o-chakela liasern hosasa
Sello A visit naseru toaorrow 'Sello visits Haseru tomorrowa Sello o-tla-chakela naseru
Sello A M visit naseru 'Sallo will visit tlaseru '
the above sentences both gla and h~sasgindicate that 12 the event referred to takes place at a tims later than th+ aolltent of speaking, Eut th is syxitacticaliy an auxiliary uhils bosasa is an adverb. This shows that future is net realized on the surface with the same syntactic means. We hope that semantic approach aaay provids generalizations about some as~ectsof different temporal constructions.
As the title of the thesis indicates, we are concerned with the semantics of Tense and Aspect with particular refzrence to their interaction with temporal adverbials. Tfie study is on
Sesotho. Reference will, however, be made to English and other languages where deemed necessary,
Although our study is semantically based, we nay resor+ to syntactic aspects where these Frovide better undzrstandinq of semantic behaviour.
It vill be noticed that our analysis is limited to the indicative mood though reference has bee3 made in literature to sutjunctive mood. Similarly, the imperative and the so called potential uood will be excluded since they ara restricted in the range of tenses used,
vhile the analysis of past tsnsz alone would prcvlde swfficient material for a thesis, it has become necessary to examine the present tense in view of thz discussion on It will also be necessary to examine cooccurrence restrictions between the different tenses.
As indicated in 1.2.5, there is no difterence in tsngoral functions b~tueen the simple forms on the one hand and the passive, benefactive, reciprocal causative otc. Oc the cther hand. Ve are, therefore, going to draw exanrples frcm any of these constructions,
Reichenbach's temporal specification involves three theoretical entities which define kasic tenses in terms of linear combinations:
(a) S-point of speech- is the time at which a given sentence is uttered, This is the keystone of the system
in that the other times are orierltdd to it, The
relations of sequence and si~ultaneityare basic tc the system. This results in tenses anterior, simultaneous
and posterior to S, The various ~ositionsin relation
to S correspond to the traditional labels, past, present and future,
( b) E-point of event- refers to the time at which an event or state occurs,
(C) R-point of reference- is the time indicated by a
sentence, sometimes specified by an overt adverbial sornetitnes derivable from the context.
sentences such as:
13. Re-ile ra-fuoa chelete maobane we past we given money yesterday 'We were giv~nmoney yesterday8 the point of event and the point cf reference are anterior to the moment of speech as follows: E, %---So The santanca is thus past, Yhile in sentence 13 the point of event and tha point of reference are simultaneous, it is possible to havz sentences in which they are not e,g.
14. Ke-ne ke-se ke-buile le-eena maobaue
I past I already spoke with hie yesterday
*Ibad already spcken to him yesterday'
In sentence 14 the reference point is maobang 'yesterday' and the mornrnt at which I spoke to him nay have Deen the day before yesterday, The tense representation in such cases is E--- B---S .
1.6,1 lotations used qith
The following interpretations connected with our notations
~ustbe coted. Points separated by a comma are inter~retedas contemporaneous, those separated by a line are interpreted as further in the future or the past depsndinq on whether they are to*the right or left respectively. A diacritic above E is interpreted as *progressivei, Various views have been expressed in ccnnection with the use of sentences with diffsrent adverbials, Reichenbach claims
that when a time determination such as 'ncwa or 'yesterday' is used it is referred to the reference point and not the event. An
adverbial may, however, refer to the event ir? cases where points of reference and of event coincide,
Hornstein (1977) suggests that time adverbials are mapped onto R or E and that this ray result in shifting of R or E
towards or away from S. The main motivation for HornsteFnyn
rules mapping adverbials ontc R or E instsad of B only appears
to be need for accounting for amkiguity in English sentences such as the following:
15. The secretary had eaten at 3 p,~.
Which is tuo-ways aabigious as follows:
16(a) the time that the secretary actually ate was 3 p.m.
(b) the secretary had already eaten by the tial 3 p,m. rolled around.
Horstein further postulates constraints on SBE movement as follows:
17. Rules shifting SRE to yield derived Tense structure
(TS) must maintain basic TS,
The constraint in 17 laads to the following stipulations:
18 (a) none of SRE are associated in derived TS that wers not 16
associated in basic TS,
(b) the li~earorder cf SBE io basic TS is iaairtained In derived TS.
(c) x associates with y iff x is separated from y by only a
comiaa,
Smith (1978) regards the combination of tense aud advarbiai as responsible for establishing the ~ointof reference. She proceeds to divide adverbs into those which establish reference
time and those which do not, She mentions that combinations of
tense and adverbial that establish RT have compatible relaticnal
values, whereas combinations that do not, have contradictory relational values,
Reference will often be made t5 thess visws in our discussion about interaction between tense and adv2rbs in
Sesotho.
The vor king definition of temporal adver b/adverbial
accepted in the present study includes all ncrds and phrases
which could be elicited as possible answers to the question nag
? *when?* Adverbs vill then be divided into thrte classes according to whether they indicate simultaneity, anteriority or posteriority with moment of speech. Let us call thess indexical
adverbs:
19 (a) Anterior: saobang yesterday1, Qekegq 2 &tile= 'iast week*, hhoelina g=$&&sg 'last mocthg s*lemmgg sg=
fa-q 'fast year', garfi_e= a-fctL&= last wintsrq
etc. Such adverbials will sometimes be referred tc as past adverbs in our discussion.
(b) SinuLtaneous: hona joal-e 'right now* &ii& lnow'. These adverbs will be referred to as present adverbs, (c) Posterior: -hosasa3 'tomorrow' ka-aoso 'day after tomorrow' b-beke- g-tlgeq 'in the coning weak'
&&gL&ig e-tlag 'coning month ' seJesonq se-tlanq
' coaing year 8.
In additicn to the above three classes, there are several non-indexical adverbs which may indicate a period that might precede or follow the moment of speech, These include hc~p~q
'morning* ~gtseare *day time' *afternoons ka=la bobeAi 'on Tuesday' b='sa 'in April* gariha 'in wint(3r1
&&g~g g=&&g- 'the following week* khoeligg e-lqtdqg 'the following month* etc.
There are also certain indexical adverbs such as kaieno
'today' Dekenq 'this week' b2&bgeaa 'this aonth' etc., which indicate a sFan of time which includes th? moment cf speech.
The last group of adverbs may be ussd with pres~nt, past or future sentences as follows:
20 (1) Ke-chakela Baseru kajeno
I visit Naseru today 'Today 1 visit Haseru' (ii) Ke- ile ka-chakela flaseru kajeao
I past visit Haseru
'Today I visited Haserul (iii) Ke-tia chakela ilaseru kajeno
I Put visit Maseru
'Today I will visit Maseru*
It must be noted that while one sentznce may have tuo
adverbs (anindexical one belonging to 19 (a) and the other nonindexical belonging to those adverbs uhxh may indicatz a
tine that precedes or follows the moment cf speech) , it is not possible to have two adverbs from the different sections in 19, e. g.
21. Lisebo o-pheh-ile nama maobane hoseng
Lisebo A cock past meat yesterday incrning 'Lisebo cooked meat yesterday morning ' but not
22 *Lisebo o-pheh-ile nama hona joafe maobane
Lisebo A cock past meat right now yesterday *Lisebo cooked aeat right now yesterday ' . Although it is possible to use -ile tense with either h~aa qoale or maobane alone, the sentence in 22 is unacceptable. This
unacceptability can be accounted for on the basis of constraints which are definable in terms of seinantic classificaticn of temporal adverbials. 19
The sentence in 21 is acceptable on the basis or hierarchic system of 'units of time8 (see Leech 1973:114-118). Such adverbials are regarded as constitutinq one t5mporai adverbial,
1.8.0 Tense
It has been indicated above (1.6.O)thar tense relates the tiae of the event to the moaent of speaking, Tenses which relate the tiare of event tc the present moment are often rsferred to as absolute tenses (Comrie 1976:2). It is also possible to have the point of event related to the tiae of some other event; in such cases we speak of relative tiae reference.
Aspect has been defined as "different ways of viewing the internal temporal constituency of a situaticnM(Cornria 1976: 5) . ~Bisdefinition nay be illustrated by the foliowing sente~ces from Sesotho:
23(a) K8-ile ka-ngola lengolo
I past write letter '1 wrote a letter'
(b) Ke-ne ke-sa-nqola lengolo
1 past still write letter
'1 was still writing a letter* 20 in uttering 23(a) the speakel: views ths act of writing a letrer in its totality. No reference is made to the iztercal constitusncy of the event (e, 5. Vritinq the addrsss, date, graetings etc.), The sentence in 23 (b) indicates that some phasaa of writing the letter had hean uniisrtaksn,
The definitions of tense and aspect given above may give the reader an itnpression that there is no connection betwesn tense and aspect, It aust be noted, however, that the two categories are concerned with tim~,though in different nays. mhile tense locates an event to the moment of speaking, aspect relates the time of the event with the internal constituents of the event, These time distinctions have bean referred to as situation-external- time (tense) and situation-internal- time (aspect) (see Comrie 1976) .
The relationship between tense and aspect may alsc be ncted in cases where certain aspectual distinctions are restricted to one or more tenses. This is ao indication that there are ways in which tense and aspect impinge on one another. This will be dealt with in chapter 5. 2 1
Notes Tg CA~ter1
1. There is a localist view expressed by Triiqott (1875:215) that locative deictics squivafent to 'comeq and 'go8 exprass past and future respectively (see also ~ivor,
1973). It is interesting to note how tense interacts
with some locative deictics in Sesotho. Thus while zq
*gog may express future, tha past tense of this v~rb --ile expresses past tense. Jacottet (1968: 19) mentions that ~erfect' in Sesctho
Ray be folloved by an adjunct of time. He does nct,
however, say what makes this possible.
Sometimes ~QU or aasaaf *tomorrow8 . CHAPTER 2
rm PEBPECT:SX~+L_LE
The question of whether 'perfect* (as it is conventionally used in accounts of tense systems) shculd ke regarded as an aspect or tense has been dealt with in great detail in literature (see Conrie 1976: 52 and references therein) . It has keen noted that perfect is different from aspect in that it tells us nothing about the internal structure of tha event, but rather expresses a relation between two tima-points namely, the time of the state resulting from a prior event and the timo of the event itself, This relation can easlly be expressed in terms of S, R and E within the franework adopted here, and thus qualify as a tense like other tenses,
It is important to note that 'perfect' is often used in purely formal sense for structures with 'have '+participle so that what has been referred to as *perfecta in qramnar books may have different meanings in different languages, In soma lanquages it covers both perfect and past tense meaning, In others it has taken over the past tense meaning, Different languages may thus show different behaviour as regards the use of time specif icaticn with the perfect form. Although reference has been made to rparfecta in S~sotho grammars no explanation has been given for its icllosyncratic behaviour. Reference has often been maue to idiomatic us% whenever explanation cannot be found. Jacottet (1965: 18) notes that: nthe perfect has exactly the same meaui~q as English; it expresses an action which is finished at the present time and has been done in the past, or a state which has been entered into in the past .,.n
The above grammarian further notes that: . . In contradistincticn to English qrammar, it may 'De followed by an adjunct of [email protected]. &S huuf le-ntate mobane ,
*I spoke with my father yesterday.*
Doke and tlofokenq (1957: 176) have the following to say about the perfect stem: nwith ordinary verbs in Sotho the perfect stem is usad in the formation of past tenses e.g. He-13-m-ex maobarie (1 saw thee yesterday); but with stative verbs 4i.e. verbs which indicat? a state already completed and still persisting)=, the perfect stea is used in present tense formation also, e.9. Ke luts-e I am sitting) ho-fu&&metse (it is warm) ". The two grammarians later note that:
'fcertain non-stative verbs may be used idiomatically uith present perfect signification e.q. Be-j--l; g (we have eaten; we have had cur meal)
f;a-tl-if e (they have cone and are here)
te-fihk-ile (I have arrived and ate here)
it is also noted that:
neach of these verbs nay, however, be used with ordinary past
significance. Note the f cllouinq:
ke-it?&% &g&hg (I ate bread)
,- I
I C Gu~a(1971: 173) lists the uses of perfect as follows: "the perfect is used as one of the tenses cf the indicative and
conditional mood, With inchoative radicals i.s. Radicals which indicate entry into a state, the perfect tense indicates an
-?action that has been coapleted, resulting in a stat2 that exists './ now, at the tiw cf speakinq e,q. EpLgt~g=e_ (I am seated) c_=
agetso=g (she is married) WgdLf (I am hungry ) pqoana i;= gobets-g (the child is asleep),
With most radicals the perfect tense indicates an action that took place in the immediate past - a few hours aqo or
yesterday, e,q. kg= me-bows hosq~(I saw him this morning) p=
&i=g&js-itse maoQa ( he sold them yesterday).
The uses of perfect mentioned by th2 above grammariaxis
reveal the amount of doubt that has existed about the exact use of this form,
Jacottet merely mention the similarity between Eny lish an3 sesotho perfect but doas nct show ways in which these meanings are the same. The exact use of perfect in Eaglish is still dekatable. It is thus difficult to deduce, from a statemect such as above what the exact meaning of perfect in Sesotho is. The present study does not a-ttem~ttc coapara the usti of perfact in sesotho and English. Beference will, however, be made in the course of our discussion to differences and/cr similarities when deemed necessary.
Jacottet further mentions that it is possible tc uss perfect with adverbs such as *yesterday8. Ha doesn't, houever, say what makes this combination possible in Sesotho but not
English. The use of *perfectg with teaporal advarbs will be considered later in this chapter.
Doks and Hofokeng have noted three uses of the LLg suffix: (i) formation of present tenses (ii) formaticn cf past tenses (iii) idiomatic uses with res sen t perfect signification .
The use of the saae suffix to indicate both present and
past tenses has never, as far as I know, been reccrdod in lanquages although languages that have tsnse appear to treat the distinction of past and non-past {present and future) as being of basic importance. Although use of in the formation cf
present and past could suggest that there is a broad conce~tof present which steps into the past or vice versa, it is doubtful that this could be the case, since the present cannot in genlral occur with adverbials which refer to the Fast, ncr the past with those referring to the present. E.g. 24(a) *Barutuoa ba-reka libuka selfmong se-fetiisng
students a buy books year A past **students buy books last year* (b) *Barutuoa ba-ils la-reka libuka noaa joale
students A past buy books fight now **students bought books right now.'
An atteapt will be made in the following section to cornpar2 the and non-ile present forms. Similarly com~arisonwill h+ made between past tense and other past tense forms.
Doke and Bofokeng make reference to idiomatic use of ths perfect. There doesn't appear to be anything idicaatic about tha use of perfect in sentences quoted above. This ccnfusicn derives from the! fact that these graamarians regard -ile as the prfser,% and past tense formative. Any meaning which falls outside these two tenses had to be regarded as idiomatic. The choice of th3 tera #idicmaticl instead of 'exception* is significant in that it shows that the two grammarians were aware that the meaniny referred to is quite regular.
The Sesotho graraarians referred to above have menticned that with inchoative verbs (Guma) or stative verbs (Doke and
Bofokeng) =j& mean 'present'. This vieu aerives from tha fact that verbs such as lutse 'sit' indicate that the person referr?d to is still seated. It will be found hcuever, that even verbs such as pheh-ile 'cookeds bes-itse 'grill, make fire8 which do 2 7 not qualify as statives according to definizicns given by these grammarians may indicate that the action is still continuing.
Further examination of Sesotho verbs may suqgzst that durativ2 situations (in the sense defined by Lycns, 1977:711) may indicate a co~tinuingstate cr process,*
-2.2 Co~~aris~&between Rnd Ecn-ile Present
In assigning the same tense and aspect interpretation to = ---ile and non-ile forms we commit ourselves tc regarding such sentenes as semantically equivalent. The question that ona asks oneself is whether such sentences can ba ussa int+rchangeably, The following pairs of sentences are vorth considering: 25(a) atosali o-lula setulonq
woman A sit on chair
Ithe woman sits on the chair1 (b] aaosali a-luts-e setulonq
woaan A sit past on chair 'the woiaan is sitting on the chair1
26 (a) nqoanana o-ema nonyako girl A stand door Ith~girl stands by the door'
(b) Ngoanana o-erne monyako
gir1 A s+and perfect door
'the girl is standing by the doorg the (a) and (b) sentences in 25 and 26 can never be used interchangeably, consider a case of a husband suspecting that 28 the wifs intends to sit on his chair and ord~ringher not to do so. The appropriate answer frc~ the wife who has no such intention will be:
27 (a) ha ke-lule setulonq ke lula fatse
neg I sit on chair I sit down
I an not sitting on the chair I am sitting on the floor* but not:
(b) ha kt-as-lula setulong ke lula fatse
neg I sit on chair I sit down
'I am not sittinq on the chair I sit down' consider again the case of a doctor who may have instructed his patient to sit on the chair askinq:
28. Na u-ntse u-lutse se tulong? Question you still you sit perfect on chair
'are you still sitting on the chair?' for which the answer from the loyal patient will be: 29 (a) E, ke-ntse ke-luts-e setulong
yes I still I sit past on chair
'yes, I am still sitting on the chair' but not (b) E, ke-ntse ke-lula setulcnq
yes f still sit cn chair
'yes 1 still sit on the chair' if tha patient uttered 29(b) the doctor might ask the question again suspecting that his ~atientdid not undzrstand the first question. The doctor could also suspect that the pationt did not follow the instructions as given ~revicuslyic which &s aay hava to instruct his patient to sit dokn.
The reader may wonder if there is, in fact, any difference between the use of -ile and non-ile present since the two may be used as an answer asked in the present (see s~ntence 27(a) fcr ---non-jJ_e answer6 and 29 (a) for -ile answar) . Further distinction between &J& and a-2 present must be aadi. Let us consider the following sentences: Ha ke-a-bua le-Haria hobane o-bona bana hona jcale
Neg I speak vith Hary because sne ses children right now '1 have not spoken to nary because she is going to see
the children nome Ha ke-a-bua le-Haria hcbane o-bon-2 bana hona jcala
Neg I speak with Rasp because she see past childran
right now
81 have not spoken to nary because she has seen th?
children1
The sentence in 30(a) depicts 4aryts seeing the children As simultaneous vith the moment of speaking. Thus her seeing the children causes my failure to speak to her. Tho sentence ir
30 (b), on the other hand depicts Mary's seelng the children as prior t~ the moment of speech. Hy decisiou not to speak to her follows froa the fact that she has seen the children. Pha sentences in 30 characterize non-ile form as a 'cause8 and tha = --ile form as a 'result*,
There is an additional difference of meaning between rile and non-ils forms which must be noted. Tha sant~uces such as
25(a) and 26 (a) may have habitual' aeaning while 25 (b) and 26 (k) sentences do not, The habitual reading in (a) sentences makes the use of adverbial expressicn of frequency with these ser;tecces possible. E, g.
3 1 Hosali 0-lula setulong kaaehla
Uonan A sit on chair everytime
*A woaan sits on the chair everytine' Adverbial expressions of frequency cannot be used with =iJ.. forms as may be noted in the following example:
3 2. *#osali o-luts-e setulonq kaaehla
Uonan A sit past on chair everytime 'A womn is sitting on the chair everytime'
----Summary Host Sesotho qrammarians regard with stative verbs as present. The discussion above indicates that there are at least three ways in which non-11% and -ile 'pr~sent*sentences diff?r.
Firstly, the non-ile sentences imply that the state referred +c by the verb is coming into being, while 2141 sentences indicate that the state has already come into being. (sae sentence 27 (a) ) . secondly, non-jJg forms may be used to indicate a 'cause' while rile fcrms indicate a result (see sentences 30 (a) and 3 1
(b)). Thirdly, non-ile sentences nay have habitual meaning while
---ilg ser,teaces can never (see sentences 31 and 32 ).
2.3 Thg use of -ile with Aem~oraLadverbs
Although Sesotho grammarians have attributed present meaning specifically to the stative verbs used vith -i12, ths use of temporal adverbials with forms suggest that most verbs may be used vith -ile to refer to the present time. Lst us consider the following sentences:
3 3 (a)i Banna ba-f ihl-ile hona joale
nen A arrive past right now The men have arrived this aomentr ii Lisebo obon-e bana hona joale
Lisebo A see past children right now 'Lisebo has seen the children now* iii BaEana ba-pheh-ile naaa hona joale
qirls A cook past meat right now 'The girls have cooked nowr
(b) i Banna ba- f ihl-ile maobane
Ren A arrive past yesterday 'The men arrived yesterdayr ii Lisebo A-bon-e bana maobane
Lisebo A see past children yesterday 'Lisebo saw the children yesterday' iii banana ba-pheh-ile aama maobane
Girls A cook past meat yesterday 'The girls cooked meat yesterday* (c)i *Banca ba-f ihl-ile Ocsasa
Hen A arrive past tomorrcw *'The men arrived tomorrow* . . i +Lisebo o-bon-a bana hosasa
Lisebo A see past children to~orrow **Lisabo saw the children tomorrow* iii *Banana ba-pheh-ile nama hosasa
Girls A cook past meat tcmorrow
*'T he girls cooked meat tonorrow' the cooccurrence restrictions linking =& f crms with tpm~oral expressions referring to past, present and future pose a number of prcbleas. The acceptability of u~thhona joglg 'right now1 suggests that sgis present, while the acceptability of this for@ with nraobau 'yesterday* sugqests that zjlg is past.
The unacceptability of -2 with ~Q~SBgtomorrowB rules out the possibility of regarding this form as present since it is possible to use present with adverbs referring tc future time as follows:
34. Band ba-fihla hosasa
Hen A arrive tomorrow taen arrive tomorrow * The unacceptability of =i& sentences used with hosasp
@ tomorrowt supports the view expressed earlier that the event expressed by =&he is evoked as pricr to the present and thus implied to be in the past, If =& expresses past, hcw can w+ 3 3 account for the acce~tability of this forn with presect adver Dials?
A number of analyses which account for the cooccurrence restrictions between tense and adverbs in languages such as
English are likely to fall short of adequat? descri~ticn cf
Sesotho facts. An analysis which views tense from the point of view of past vs non-past distinction uculd predict that sentences in 33 (a) are unacceptable. Even if the Sesotho =AJg forms could be regarded as corras~ondingto English perfect the problem remains unsolved since the English perfect is incompatible with adverbs such as yesterday,
There are a nunber of possibilities for accounting for thz sentences in 33. Ue could, for instance, bypothesize that -ile is a form of gramaaticafized stative. In languages such as Zulu stees distinguish themselves as statives from the ones that use the same suffix but are not stative in thsir negatives (see
Vestphal 1950: 126; Beauchat 1966: 75) e .q.
35 (a) Ba-lamb-ile
They hunger state
'They are hungry'
(b) A-ba-lamb-ile Neg they hunger state
'They are not hunqry' q&* =L They walk past
@ 'They walked' 3. .. (b) A-ba-ha aba-anga
Neg they valk neg they did not walka
The negative test used for Zulu cannot be ap~liedin Sesotho since there are no distinct forms for stative and non-stativa
negative forms E.Q* 37(a) Ba-lap-ile
They hunger 'They are hungrya
(b) Ha- ba-a-la pa
Neg they hunger 'They are not hungry*
38 (a) Ba-tsaaa-ile
They walk
'They left*
(b) Ha-ba-a-tsamaea Neg they walk
'They did not leave'
T he hypothesis that -Fk is grammatical ized stative might,
perhaps, be valid if we could find a group of nonstativ2 verbs
which do act occur with it, Ve might then assume that with lexically stative verbs it is a redundant speciiicati.cn of
stativity. It must be noted, however, that occurs with both stative and nonstative verbs.
If arguraer,ts could te found to support the visw that -;1e is grammaticalized stative a number of probiems would easily kc solved. The question of perfect signiflcance which Dcke and nofokenq have noted could be accounted for. For as Lyons (1977:415) notes, ".., Perfect is commcnly and perhaps always ir, origin, a statfve aspect with the more specific feature that it is used to represent the state that it denotes as teinq consequent upon the completicn of the process which the verb denotes.
se could also explain the use of past advarks as due tc the tendency of aspect to develop into past tenses or tenses referring to the recent past as noted hy Anderscn (1973: 42).
2.3.2 -ile g~ Perfective
A second possibility would be reqardinq wile as a fcrm of present perfective (in the sense defined by Conrrie 1976:16)8.
This analysis would provide a balanced system for present and past tenses. We could view the simple form of present as imperfective --being also habitual, Progressivity would be expressed by sg 'still8 and perfectivity by ----ile. Correspondingly the iaperfective past would also b~ habitual; progressivity would be expressed by a *stillsand perfectivity by h&+stem (an account of this tense wili be given later). 39 is schematic representation of such an analysis: f t ---1 -----I I 1 I I 1 1 4 IMPERFECTIVE f PROGRESSIYEIPERFECTIVE 1 I 4 I I i 1 1 J 1 1 --t -4 1 1 1 I 1 \ 1 Preseat 1 stem ) sa+stem ]steni+iir 4 1 1 i L 1 1 1 )------.f- 1 a -+ 1 1 1 1 1 4 Past I ne+stea i ne+sa+ste~ntile+stea 1 1 1 1 1 I I f 1 1 1
while the above analysis may be plausibla in that it establishes
relationship between stem+i& on the one hand, and ila'steg on the other hand, it fails to account for the occurence of li&s
with past adverbials,
2.3.4.1 -ile ps immediate pas
The claim made in this section is that ~tsa+ileis an
ilsmediate past tense, The distinction between remote and
im~ediatetense has been ncted for other Bantu languages, Dok~
and aofokeng (1957:187) give the following axample from Lanba- a central Bantu Language:
'1 did (before today) (b) Nacita '1 did (today)
the same immediate/remote distinction has keen noted for Zulu by
Doke (1927:169) Westphal (1950: 132) and Beauchat (1966:76) e.q. , 41 (a) Sithenge izingubu kuthangi
@We bouqbt clothes the day befors yestsrda y *
(b) Abelusi basenge izingomo izolo
'The herdboys milked the cows yesterday' iiestphal furthar notes . . AIthouyh sindmbil~[we walked - recently) is usually imlnediate it can also ~e used with adverbs
such as 'fast year8 while saahamba (we walked-long ago) is
rarely used with recent adverbs ... IV
The exact definition cf immediate tense does nct aFpaar
clear cut. As Beauchat notes "what constitutes an iaraediate past
varies to a great extent fro& one speaker to another. fcr macy people an action which tock place more than a week ago is expressed in the remote past, whereas for others, the tias limit
between these tenses is as lnuch as six mcnths ...11
Doke and Bof okeng (1957: 187) observe that: '*the Southern
Sotho tense division according to time is not so clear and balanced as that of many other Bantu Languaqes."
We claim that Sesotho, like other Bantu Laguages, has
remote and immediate tenses; and that the bcundary ketuesc remote and immediate tenses is not clear cut so that the earlier grammarians concluded that there is no such distinction.
One possible source of confusion may De the fact that -iLe
may be used with adverbs which range from tcday tc remote Fast.
Pollowinq are exaa~lssof such uses: hosenq they arrive past maobana maoba bekeng e-fetiileng
seleaong se-fetilenq they arrived yesterday day before yesterday last week, month, year.
---wile can extend as far back as a century. Tire follovinq historical fact may be expressed by using -ile 43. Borena Boshoeshoe o-fall-etse Thaba Eosiu ka-1824
Chief lYoshoeshoe a move to past Thaba-Bcsiu in 1824
'Chief moshoeshce moved to Thaba-Bosiu in 1824'
There are other constructions in Sesotho which exgress immediacy, It must be noted that such constructions may be used with prssent and past adverbials as follows:
4 4. Johanne o-tsoa-fi hla hona- joale
John A past arrive right now 'Johc has just arrived now'
45. Johanne c- tsoa-fihla maobane mona John A past arrive yesterday hare 'John arrived this mere yesterday8
The &spa+stem construction above cannot be used with removed nun-specific adverbs such as khale *a long time ago1 e.9. 46. *Johanne o-tsoa-fihla khale John A past arrive long time ago
*'John just arrived a long time agcl
the ggpa+ste~ constructicn may, however, be used with advtrbs
which refer to the day before yesterday, last week, last month
etc. A perscc u~ccorsiders a year too short a time fcr a itoman
to have divorced her husband could easily say: 47, .. . Eapa o-tsoa-n yaloa selemong se-f ttilanj sena but she past be-married gear ~assedthis '., , But she was narried this mere last yeari Similarly an aunt who has not been home for five years may
express surprise at the rate at which her nepnew has grcun:
48, Ha a-ts'o9-hlaha ka- 1973 ha ke-le aoo?
leg he past appear in 1973 when I be here
'Wasn't he born in 1973 when I was hera?
The sentences above indicate that iamediacy is a subjective term which may extend as far as the speaker wishes. This subjectivity
has led to relaxation on the degree of recentness required fcr
& so that 1 has tended to be used intdrchangeably with reniote past tenses.10
2-3-4.2 Distincti~qbetreenn tsoa and -ile tenses
The clai~made in 2,3,4,1 is that both tsoa and tenses
are immediate past tenses. Be have noted that the two tenses may
be used with present and past adverbs, It was further notsd that
the two tenses may be used with adverbs such as fast year and
five years aqo, depecding on the speaker's attituds tovards ths 40 time mentioned. We may alsc note that the two tenses cannot b2 used with future adverbials as noted in the foilowfrg sectences:
49(a) *Banana ba-pheh-ile naina hosasa
girls A cook past meat tomorrow
**Thz girls cooked meat tomorro#'
(b) Banana ba-tsoa-pheha nama hosasa
girls A past cook meat tomorrow
'The girls have just cooked meat tomorrowg
The reader map wonder whether there is any distinctior, between the two forms a and tsoq since they seem to appear in the same environmenks. There is same evidence to sugqtst that the two tenses may not be used interchangeably,
In an attempt to find the distinction between tho tuo tenses a nueber of pairs of sentences were qiven tc nativz speakors so that they might choose that sentence of each pair which th3y found More natural,
50fa) Ke-tsoa-ja ~etsotsonge-mehlano e-r'etileng" past eat minutes five passed $1 have just eaten five minutes ago'
(b) Ke- j-ele aetsotsonq e-mehlano e-fetileng
I eat past minute five passed '1 ate five minutes agog
51 (a) o-tsoa-•’eta metsotsong e-mashoma a-mararo e-fetil~ng
He past pass minutes tens three passed *He has just gone past thirty ninutas ago' 4 1
o-fet-iie metsotsong e-mashame a-maraxo e-fetileng
He pass past miruter tans three pased
'He went past thirty minutes ago* Ba-tsoa-pheha nama horeng tse-peli tse-fetilenq
They past cook aeat hour two passed
'They have just ccoked meat two hours ago'
Ba-pheh-ile nama horeng tse-peli tse-f e tileny They cook past meat hour two passed
'They cooked meat two hours agc'
uhile all the speakers regarded the (a) and (b) sentences acceptable, there uas a strong indication fcr the prefertncs of one against the other. Por instance, when asked which of the sentences in 50 they would use if invited to eat but they cannct eat since they have eaten five minutes ago, they stronqly indicated that they would use the (a) sentence.
Similarly, when asked which of the sentences in 51 they would use as an answer to the question about the time at which soaeone went past, they emphatically said they would use the sentence in (a).
When asked which of the se~tencesin 52 they would use as an answer to the question about the time at which some people cooked meat, they said they would use the szntence in (b).
The choice of (a) seritences in 50 and 51 as against tha
choice of (b) sentence in 52 suggests that tsoa tense is much more immediate than the zifg tense so that whenever the tims referred ro is less thar? thirty minutes prefere~ce is fcr the tsoa -tense,
The distinction betweeo the two immedia te tenses becomes evident in constructions involving the co~~lemsrtof coercive verbs (when not reduced to subjunctivc or infinitive fcrms) . In such constructions there seem to be a restriction that the immediate tense in the complement verb must be the same as the one in the aain sentence, e,g. Ke-no-laets-e hore a-tsamaee 'ms o-tsama-ilt
I him order past that he go and he go past
'1 ordered him to go and he wat,* but not
(b) *Ke no-laets-e hore a- tsaaaee 'me o-sa- tsoa tsamaeal2 I him order past that he go and he Fast qo
'I ordered him to go and he has just gone'
Ke-sa-tsoa-so-laela hore a-tsaaaee me o-sa tsca- tsamaea
past him order that he gc and be past gc
'I have just ordered him to go and he has just left'
but not
(b) *ke-sa-tsoa-no-laela hore a-tsamaee 'me o-tsama-ile
I past him order that he go and he go past
'1 have just orderad him to gc and he left'
ke-mo-qobeletse ho-ja, 'me o- jele I him force past to eat and he lat past
'1 fsrced his to sat arid he ate.' but not
(b) *keno- qobeletse ho- ja 'me o-sa tsoa- ja
1 him force past to eat and he past eat
'I forced him to eat and he has just r3atzn'
56 (a) ke-sa-tsoa-no-qobella ho-ja, 'as o-sa- tsoa ja
I past him force to eat and he past eat
'1 have just forced him to eat and he has jus+ eatec'
but not:
(b) *ke-sa-tsoa-mo qokella hc- ja 'me o- jela
I past him force to eat and he eat past
*I have just forced him to eat and he ate'
The (O) sentences in 53-56 akove are unacczptable becauss different immediate tenses have been used in ths maic sentences and subordinate sentences. This suggests that there is need to
distinguish the two immediate tenses,
2.3.5 aiug advantages of analvshq -iIe (1s imaediate past tense
The analysis of =- as immediate past explains why it behaves differently from the English perfect thcuqh It has oftsn
been compared with it. Thus the characteristic aspectual nature of perfect which is manifested by the restriction that govern coabination of perfect and past time adverkials is not foucd in
Sesotho e.q.
57. English: *he has left yesterday Sesotho: o-tsama- ile iaaokane
Be valk past yesterday *He left yesterday*
The present analysis further explains why p~rsistenc~of the result is not a rscessary condition for the use of -ile. Thus we may have
58, Johanne o-f ihl-ile
John A arrive past 'John has arrived indicating that John is still here, It is also possible, however, to specify that John is no longer bere:
59. Johanne o-f ihl-ile ka-lakobeli a-tsoha a- tsamaea
John A arrive past on Tuesday he wakes he valk ka-3. aboraro on Uednesday
**John has arrived on Tufsday and left on i?ednesday8
Tha use of -ire as immediate past bridqes the gap between past and present and thus explains the lack cf habitual mearicg in sentences such as 251b) and 26(b) (repeated her2 for convenience as 60 and 61),
60. flosali o-lutse setulong
Woaan A sit past cn chair 'The voaan is sitting on the chair*
61. Ngoanana c-eae monyako
girl A stand past door
*The girl is standing by the door* +-
Given that =i&g is im~ediatepast, it AS possibie to rejara
it as present depending on how past the event or state referred
to is. This accounts for the resent use of the perfect which Doke and Hofokenq make reference to (see 2.1. Above). 1. See 1.2,6 for different realization of -ile 2. The definition of stativity qiven here apprsars
contradictory, It is difficult to think of a state
lco~pltAed' and 'stiii persisting ghat these
qrammarians want to Express is that tha state has been
entered into in the past and is still pessisting.
Doke and Hofekenqls account is quite obscure. One miqht
think that perfect significance in the sentences q'i v8n
is similar to what Comrie (1976:56) refers to as
'perfect of result* in that 'we have eaten' implies
that we are satisfied and do not iaaediately need food;
and *they have core* imply that they are still here.
This does not seem to be the case since &g j-e& Mobs
'I ate bread* could iaply that I don't immediately need
bread,
Professor Newton suggests that *sit1and *cook1 might
both be terminative tbe rule being that with
terminatives &am+ile indicates continuation of terminal state, The follcvinq frame fcr terminatives
suggests that &beha *cook* but not sit* is a
terminative e.g. re-nk-ile hora tse-peli ho-pheha nama
we take past hour two to cook msat
'ue took two hcurs to cook meatg ii) *Re-nk-ile hora ts5-peli ho-lula setulong we take past hour two to sit on chair
'uz took two hours to sit on the cnair*
The -a- in this ccnstruction is often referred to as
indefinite. It appears #hen no adjunct in the fcrm of
object or adverb is used with the verD, e.g.
I)a. Ngoanaaa o-ema monyako 1 'The girl stands by the doory
b. Nqoanana 0-a-ema 'The girl is standing* it is also used with the neqative forlss of -ile suffix
6. The question/prohibition may be somethinq like:
u etsalng haeka u-lula setulong sa-ka tje?
You do what as if you sit on chair of mine like this
'why are you sitting on my chair?'
The term *habitual1 will be used in a broad sense to 3
related set of terms such as customary, frequent, regular ,usual, normal etc. "perfectivity indicates the view of a situation as a
single whcle, without distinction of the various
separate phases that make up that situation."
9. This is a contracted form of tsoetno 'Neg coms+to".
10. Scae historical facts such as those expresses by
sentence 43 are probably not sub j~ctto subjectivity.
It may be interesting to make a survey of
conversational style and determine which of the two
past tenses ile+s&sgand stemtill. is commonly used. It 48
is possible that a fcra that is fr2quently used may
sasily replace thf oce that is not.
11, The verbal force of 'come from9 is strongly felt in --tsoa- tenses, For this reascn thi progressive form sg is ccmmonly used to distinguish the i~mtdiate tens?
from the 'ccme f rcm meaning*.
12. The corresponding English sentence for this sentonc3 is acceptable, It seems possible that in Sssotho an action
follcwing from the order nust be undertaken within ths
saae time sFan as the order. The sentenca would b=
acceptabls if the connective used indicate somethinq
contrary to expectation. Thus: ke-no-ldetgg a- tsamaee s- qr_sg-tsoa -pea '1 ordered him to go but
he has just leftg imply that the speaker expected the
other perscn to have left earlier, CHAPTER 3
3-1 REBOTE PAST TENSE: /LE+JTZi!!
The remote tense consists of the subject prtf ix+ile+sukject prefix folloued by a verb, e,q, 62, Ke - ile ka - reka I past I buy '1 boughtr
There seem to be similarity between this form and the immediate past suffix 1 which could suggest some
tranformational derivation of one from the cther. It must be noted, however, that this would be an isolated case in ths
grammar of Sesothc Language. There are no other processes which involve either the suffix fronting or verbai pcst~osinq.
The difference between -ile and ug= past tenses has nevsr been clearly defined. Doke and Uofokeng (1957:202 fn ii) regard
the latter tense as mcre emphatic than the fcrmer. We caonot accept emphasis as the primary distinctive feature betwezn the tuo tenses since there is no practical means of proving that one
form is more emphatic than the other,
ahen asked about the difference between the two tecses mcst
speakers mention that one nay use one or ths other depending on I. how the question is framed. Yhere the question is framed on
i&e+stqg the acsuer will be framed ir the same way; and rihere
the question is framed on stee+ile the answer will afsc bc
framed on stem+i&. It will be noticed, however, that there seem
to be riothing absurd in answering one with the other as may bs
noticed in the f ollouing sentences:
63. Na u - ile ua - rzka libuka ? Question you past buy books
'Did you buy books?'
for which the answer nay be: 64(a) E, ke - ile ka - li reka Yes 1 past them buy
'Yes I bought these*
(b) E, ke - li - rekile Yes I them buy paet
*Yes, I bought them '
It has been suggested in the precedinq chapter that the
difference between the and -ile tenses is cne of remcteness
and proxiaity to the present, This suggestion derives from she
fact that one may be used with adverbs which refer to present
while the other cannot. E.9,
64 (a) Johanne o-f ihl-ile hona joale
John A arrive past right now
'John has arrived now'
(b) *Johanna o-ile a-fihla hcna joale John A past arrive right now 'John arrived now* Banaca ba-pheh-ile nama hona joale girls A past cook meat ncw 'The giris ccoked aeat nova
*Banana ba-ile ba-pheha naaa hona joale
qirls A past cook aeat right nou 'The girls cooked meat nowJ
1n an attempt tc further Fursue the proximity/remo teness distinction the following sentences were given to sesotho speakers so that they could choosc ths one they fcurid most natural. 67fa) Ke - 'mon - e hautsoanyane taona I hia see past shortwhile here 'I saw him a shortwhile agoa {b) Ke - ile ka-'mona hautsoanyane mona I past see shortwhile here
'I saw him a shortwhile ago8
68 (a) Re-rok -ile libnka hautscanyane ntona
We buy past books shortwhile here 'We bouqht books a shortwhile ago*
(b) Be- ile ra- reka libuka hautsoanyane Dona we past buy books shortwhile here
'We bought books a shortwhile ago'
The (a) and (b) sentences in 67 and 68 represent immediate and remote tenses respectively. The two tenses are ussd with a phrase wkich refers to a shortrhile ago, All the speakers consulted ch~se the (a) sertences (i;ilmeCiate past) as thz most natural, This choice supports our view that ile+~temis a remote tense.
It has bee& noted ir, the preceding chapter (2.3.4.1) that it is possible to use immediate tense with adverbs which refsr to yesterday, day before yesterday, last weak month, yaar etc.
It may be aentioned that it is also possikle tc use the remcte tense with adverbs which refer to this morning as may be noted in the following sentences.
69, Re-ile ra 'mona hoseng tjena
we past hie see mcrning this *Re saw hie this morninq'
7 0. 0-ile a -1i- fu~flana hoseng t jena he past then find aorning this *He found them this morning' One nay view these sentences as a counterargumsnt to our claim, but this does not seen to Le the case. It has aiready been noted that prcxienity is a subjective term, so that it Bay be influenced by the speaker's attitude, Although different speakers may use one or the other tense in similar situations dependinq on the speakersg attitude, there seam to be reascn to believe that the distincticn between immediate and remote tenses exists in Sesctho, This is supported by the fact that when native speakers are given sentences with bcth tensss and different adverbs, cne recent, the other remote, most speakers pref2r the sentences in which an immediate tense is used uith a secent adverb, The following sentences were qiven to ndtive speakers.
71 (a) Ka jeno Haria 0-sebelitse ho-•’eta kamoo
today Hary A work past to pass in there a-ileng a- sebetsa khoeling e-fetileng
past work month passed
'Today Hary worked more much more than she did last month' (b) Kajeno Waria o-ile a-sebetsa ho-feta ka-moo Today Hary past work to pass in there a-sebelitseng khoeling e-fetileng.
She work past month passed
'Today Eary worked much more than she did last month*
72 (a) Selelaong se-fetileng re f umane mskotla e-mengata
year past we find past bags @any ho-feta eo re-ileng ra-e-f uaana lilemong tse-tharo tse-
f etileng
to pass that we past we it find year three passed *Last year we got more baqs than we did three years ago '
(b) Selemong se-fetileng re-ile ra-fumana tuekotla e-manqata year passed we pazt find bags Many
hofeta eo re- e-fumaneng lilemcng tse-tharo tse-
f etilscg
to pass that we thee find years three passed 54
'Last year we got more bags than we did three years ago ' non ghali Ntsane c-tsoile boticnereng uaobane atiie nof umahali fir Ntsane B leave past teaching yesterday while wife oa-hae o-ile a-tsoa lilenong tse-hlano tse-fetilenq of his past leave years fiva passed
*Hr Ntsane retired froa teaching yesterday, while his wife retired five years ago1 nonghali Ntsane o-ile atsoa botichersnq macbane, athe
ar Ntsane A past leave teachinq, yesterday while
iaofumahali oa-bae o-tscile Lileaonq tse-hlano tse-
f etileng wife of his left Fast years five passed
'Br Ntsane retired from teaching yest2rday uhile his
wife retired five years ago'
Raeba palo ea-batho kaileng ba- phekoloa selemcnq se-
f etileng if number of people past cure-passivs year passed
e-tse joa, ho-bobebe hc-f u~ana palo ea-batho Da- phekotsoeng
know-passive it easy to find num~erof people A cure- passive past bekeng &-fetileng week passed
'If the numker of people cured last year is known,it is easy to find the number of people cured last week* 55
(b) Haeba pa10 ea batho ha-ptekotsoeng selamcng se-f .i tileng
if nugber cf pec~lecure-passiva past year passed
o-tse joa ho-bobebe hc-f ueana palo ea-batho baileng ba- phekoloa
know passive it is easy to find number of people past cure-passive bekang e-fetifeng week passed
'If the number of people cured last year is known, it
is easy to find the number of peopie cured last vesk*
For all the sentences given in 71-74, the native speakers preferred the (a) sentences, As may be noted, the (a) sentence in 71 is used with adverbs ka-ienl 'todays and IcbQelinq e-feticq
'last month'. The first adverb is used with immdiate tense and the latter with reaote tense, In the (b) sentence the adverb -kadeno --- ' today* is used with remote tense and Ichgdi~ge-fetilgrg 'last month8 is used with immediate tense. The sentences in 72 and 73 follow the saae pattern i.e immediate tense+recent adverb, reraote tense+re@ate adverb for (a): immediate tense+remota adverb for (b), In 74 the order is sightly chanqed. Ye thus have reaote tense+reaote adverb, immediate tense +recent adverb for {a) and immediate tense+remote adverb, remote tense+recent adverb for (b) ,
The response qivan by the native speakzrs for the ss~tencos in 71-74 is consistent with cur hypothesis that gtem+ile tense is an immediate tense while ile+stg&is a remote tense.
We may note. at this point, that a number of stuai9s dealing with Bantu multiple$ past systems divide tenses icto today, yesterday, and before yesterday tenses {seo Givcn 1972,
Yald 1976). We have avoided the use of such divisions since the degree of pastness in a--ile and j&= tenses does nct coincide with then, The terms immediate and remote have been found approsiate in that they do not denote a fixed length of the frcm the acment of speech.2
It has been noted in 3.1 above that stem+ile may be used with an adverb which refer to tbe present while ile+stem cannot.
This will not be repeated in this section,
3.2.1 The ~l;pq+essivggg *still8
Although the analysis of aspect will be dealt with in chapter 5, it is necessary, at this point, to consider the occurrence of sa with stem+ile and its ncn-cccurrencs with &Je+ste$. consider the following sentences: 75(a) Thabo o - sa3 - qal-ile rxlosebetsi ka-5 Thabo A still begin ~astwork at 5 * Thabo has still begun vork at 5' Thabo A still past begin work at 5
* Thabo still beqan work at S8 76(a) BanaIia ba - sa pheh-ile nama girls A still cook meat * The girls have still ccoked meat* (b) Banana ba - sa - ile ba- heh ha nataa
girls A still past cook meat
The girls still cooked neat1
77 (a) Le-khoeling e-fetileng barutuoa ba-sa-reki1.2 libuka
tse- liog
with month passed students A still buy past books
others
Even last month students bought additional booksv
jb) *Le-khoeling e- fetileng barutuoa ba-sa-iie ba-rek? libuka tse- ling
with month ~assed students A still past buy kooks
others * Even last month students bought additional books* One may be tempted to regard the distinction between 41~and = --ile tsnses as cne of perfective vs imperfective respsctively. ~hecooccurrence restricticn against gg and jktense will thus be attributed to perfectivity, while occurrence with =- tense will be attributed to imperfectivity, This could be in accordance with an observation sade by Horissey (1973:65) that
'still* is related to progressivity. We will notice in 5.1.1 that both tenses are ptrfective.
I C f The cooccurrence restriction between sg and ile- tense i i appear to support our view that this tense is remcte whih -ils I tense is ax imrsediate tense, Givon (1969:172-213) has fully
damonstrated in the case of ChiBemba, a Bantu Language, that
there is an increase of markedness* with the increase of remoteness from the moment of speech, Givcn nctes that a more marked tense/aspect imposes sore restricticns on its envircment
than a correspondir;q less sarked one, Rhile the analysis of the
markadness hypothesis will not be provided in present study, it
seems plausible to conclude (using facts from another Bantu
Language) that the coocc urrence restriction between and ilgz
tense is due to its reaoteness while its occurrence with -ile is
due to immediacy of this tense. This view sdy be suppcrted by
the fact that the other immediate past tense ssoa- may be used
with Sa e.q, 78. Be-sa-tsoa - 'aona kekenq e-f~tileng we still past him see week passed
* Ws have just seen him last ueek'
3.2.2 gmaediate aqg pote tenses iq ccrplern~gtsenterces
There is a grouF of verbs vhich function in twc differect
syntactic capacities: as regular verbs not requiring cbmpleiitentation and as modal verbs taking sentzntial
complements, with the corplement verb taking tensed form. A number of thesf constructions show restrictions cn the
P occurrence of the reaote tense with the cu~plsinentvzrb, %a thus E have the following acceptable (a) sentences but not (b) : $ 79 {a) Re -f ihl-ile a-rokets-e
we arrive past he sieep past
He found him asleep8
but not (b) * Re -fihl- ils a-ile a-rohala ve arrive past he past sleep
we found him asleep'
80(a) Re -funan-e a-pheh-ile calsa
we find past he cook past aeat
Ue found him cooking meat' but not
(b) * Re funan-e a-ile a-pheha we find past he past cook meat We found him having cooked aeat* Ona may think that the akove restriction is due tc th2
difference of form of the tenses used with the modal verb and
that with the complement verb. It will ke noticed, however, that this restriction does not hold rhen other immadiate tenses ars
used as may be noted by the acceptability of the folfowing $sg& tense:
81. Re -turnan-e a-sa-tsoa-pheha nana
we find past he still cock aeat
We found that he had just cooked meatg Be - fihl-ile a-s a- t soa rsbala we arrive past hf still Fast sleep * We arrived just after he slept1 The above examples sugqest that the statema nt of bl ockirig
conventions which bar certain tenses frcm cccuring with
complement verbs will make reference to remoteness or immediacy of tenses.
--Sumearp
The distinction sad~between g&em+ile and ile+ste~is one
of proximity and remcteness respectively. It has been noted that
an immediate tense aay be used with adverbs which refer to the
present and the past while the remote t~nsecan cnly be ustd
with adverbs referrinq to the past. We have also r.otfd that
although immediate past tense nay be used with rencte adverbs, and remote past tense with recent adverbs, nios? speakers prefer
to use reaote tenses vith remote adverbs and immediate tenses
with recent adverbs. We have also noted that the i~msdiate past tense may occur with progressive a 'still while the remote
past cannot, This has been attributed to the markedness on ths part of remote tenses, ----Notes to c>g& 3
1, The term is borrowed fro~Maid (1976). It refers to a
tense system expressing nore than cne degree of
pastness.
2 The meaning of these terms may be compared with that cf implicit ccaparatives such as 'many* which assert a
comparison between some observed instance and an
average value (see John Robert Ross A note on
implicit comparatives1 in Linguistic Inquiry 1, 3, 1970: 363- 366) . 3 This should not he confused uith contracted form cf ~g
*alreadye +a, 4 . The use of this term in Given's analysis does not soem to imply that the meaning of rhe unmarked category
enccmpasses that cf its marked counterpart. Chapter 4
Constructions with =+verb have been referred to as
*iraperfect tense* by Jacottet (1965:118); and as 'past continuous tense4 by Guaa (1971:175), Doke ard Nofokenq
{1957:252) refer to _ne as *a p~rfectforrs used in Fast time*,
The different tenses formed with pg are described as (a) past continuous for awesent (Ic) Fast perfect for ~g+~exfectand
(c) pluperfect for ne+ile+ stela the three tens+s ir. a-c are illustrated by the following sentences:
83. Sello o-ne a-reka libuka
Sello A continuous buy books
'Sello was buying booksg
84. Sello o-ne a- rek- ile libuka
Sello A past buy past books 'Sello had bought books'
85. Sello o-ne a-ile a-reka lihuka
Sellc A pluperfect past tuy books 'Sello had bought books*
The ~eaning of the tersa 'imperf~ct'as used by Jacottet cannot be deduced from his analysis. In dealing with 'present inperfect' (ae+present) Jacottet say it means that the acticn cf i the verb was being dcne at a certain tine. Although Jaccttet
uses tha tera 'tensey with im~erfect (xhich we assume refors to
imperfective in the sense defined by Ccmrie (1976: 16) his u~eof
'being done8 sugqests that he regards gg as an aspectual marker.
The meaning becomes even acre coaplica ted in the aralysis of (imperfect of the perfect' (rie+stem+ile), Jacottet says it
Beans that the action cf the verb had already been done at a
certain time. One would expect that such constructions would be
reqarded as perfective, but the presence of tho aspectual marker --ne persuades this qraaaarian to regard such constructions as inperfect,
Guma uses the term 'past continuous*, the term whoss
definition is nut apparent in his analysis, As he states, this
tenss indicates that an action was in process at some time in
the past, It is not clear whether the term 'coctinuousl has the
same meaning as 'proyressive' in English cr whether it has a
different meaning, Continucusness is def inea by Conrie (1976: 33)
as imperfectivity that is not occasioned by habituality. It will
be noted later in our discussion that the gje= tense, though unmarked8 for habituality, may have a habitual meaning, idhatever
the meaning of continuous as used by Guma may be, it sesms possible that he regards ne as an aspectual marker,
Doke and Bofokeng's analysis appears to be as confusing as
the analyses above. It is not clear in the first place what they
mean by 'a perfect fora used in the past'. The term '~ertect' 64 has often been used interchangeably with *perfective8 in traditlcnal grammars. Bssumirg that this term is usad tc maan
*perfective* by Doke and nofokeng, it would be contradictory to regard perfective form as used in the formation of continucus tense.
One could also assume that is a *perfectg form fcr t~hich there is a corresponding non-perfect pp (i.e. Stsm withcut - ile). This could be possible since a number of verbs ending in --na take -e as perfect (our inmediate past) e.g. bona >ton-e (see* fana > fan-e 'give' kena > ken-e *enter8. One will nctice, however, that there is nc corresponding tansa/aspect in for which qs may be regarded as perfect. We thus have:
8 6, Ke-ne ke-reka
I past I buy
*I was bu yinq*
but not 87, * ke- na ke- reka
I present I buy for '1 am buying' An inconsistency similar to that found in Jaccttetas analysis is also found in Doke acd lYofckeng*s analysis, Thus &g is regardad as aspectual marker before the simple form (presect), and as tense marker before past tenses.
4,l.l zg gg agpectual pg &qsg marker? The question that the reader may ask hiaself is whether the difference between 2s and cther past tenses is one of aspect or tense. It is clear from the abcve discussion that the problem has not been resolved, It is doubtful that this can be easily resolved since there is nc clear-cut boundary between tense and aspect. Aspect is often regarded as more bdsic %ban tense in that children whose native lauguage has both master aspect more quickly than they do tense (see Lyons l977:705). The case of &g appears to be different in that it may express i~perfectivity and tense. It is thus unclear to me whether in cases where one form has both aspectual and tense meaning the childr3n will nastsr aspectual meaning first, sc that we can regard aspectual meaning as more basic. This problem will nct apply tc Sesotho only, but other languages in which there are different past tense forms which may express certain asp~ctualdistinctions.
The claim made in the discussion which follows is that gg is a past tense fcrm, In as far as tense relates the time of the event to the moment of speakinq the ne form, like the past tenses examiced before, refers to the tiiua prior to the moment of speech and thus qualifies as a past tense. Thus the sentence:
88. Bana ba-ne ba-ja lebese
children A past eat milk 'The children were eating ilk' refsrs to an event which took place before the moment of speakinq, Compare the following past tense:
89. Bana ba-ile ba- ja lebese children A past eat milk 'The childreri ate iuilki
The translation given for the sentence in 88 suggests that petstem tense is ~rcgressive, It must be noted that the meaniog of progressive in languages where it has been studied is so
extended that it is impossible to speak of basic progressive meaning. Scheffer (117 5: 17-40] has noted (and thoroughly criticized) the following meanings attributsd to progressive in
English: the progressive expresses duration; denotes acticn in
progress; forms a time- frane; it is used descriptively and
subjectively ; has an iaperfective character; sxprasses temporary validity or heighten teaporary relevance etc. Criticisms on thcl traditional deficitions cf progressive dre also in Ccmrie
( 1976: 33-40] iioisetschlaeger f1976:28-35) and Lyons (1977:713-
19); all of which point to the need for a general meaning rhich may include progressivity and various other aeanings.
In regarding tense as basic to ge forms we Bay list the various aeanings found in this tense. This classification is consistent with the analysis cf the past tenses examFned
earlier. There does not appear to be any sound reasoc for giving
the semantic label 'progressiveq tc forms but not
#perfectiveq to ile- form exaained earlier, Ye will indicate ir,
the next chapter the tenses which have perfective and imperfective meanings. There are some semantic and syntactic correspondences between the present tense and the past tense in ns which suggest that the terse in 4s is the Fast tense ccrresponding tc the prdsant tenso. In addition to the charactaristic Frenent/past tense meaning, the two tense forms are unmarked fcr habirual meaning. Thus although the sentences in 90 and 91 have no overt habitual marksr, they may have hatitnal meaninq:
90. nalimo a-bolaea batho
cannibals A kill gec~le 'Cannibals kill p~ople'
9 1 Plaliro a-ne a-bolaea batho
cannibals A past kill people 'Cannibals used to kill ~eople' The two tenses nay be used with habitual marker as follows: 92. Halimo a-ee a-bolaee batho
cacnibals A habitually kill pecple 'Canaibals used tc kill people'
93. Ealino a-ne a-ee a-bolaee batho
cannibals A past habitually kill people 'Cannibals used to kill people' idhen the unmarked sentences in 90 and 91 are used with indexical adverbs such as Bqna hale 'right now' and maobana 'yesterday' respactively, the habitual meaning is neutralized. This is understandable since habitual has no specific time reference. 68
~utthe used of indexical adverbs with the sent~ncesmarked for nabituality results in unacceptable sentences as follows:
94. *Balimo a-ee a-bolaee batho hona joale cannibals habitually kill pecple this ecment
+*Cannibals habitually kill people this momsnt'
95. *Ealiao a-ne a-ee a-kolaee batho macbane
cannibals past hakituall y kill people yesterday *'Cannibals used to kill people yesterday*
It is important, at this point, to note ths meaning difference the aarked and the unmarked habitual sentences above
(repeated here as 96 and 97).
96. flaliao a-ne a- bolaea batho cannibals past kill people 'Cannibals used to kill people8
97. Baliao a-ne a-ee a-bolaee batho cannibals past habitually kill people *Cannibals used to kill people8
The sentences in 96 and 97 cannot be used interchangeably.
Imagine s case of a child vho has often heard about the cannibals but has never been able to find the exact meaning of the word 'cannibals1 since the word is often used to refer to people uho like meat to such an extent that they may fight over it. The implication is that these people like meat to such an extent that in the absence of meat they may kill and sat other people. The child may ask the iaother what the cannibals are, for which the mother will answer:
98. nalimo ke-batho ba-oeng ka-ja nama ea-batho cannibals it is ~eo~lepast eat meat of people
'cannibals are people who ate human flash' but not:
99. Halimo ke-batho ba-neng ka-ee ba-j& nama ea-bathe
cannibals it is people past habituailp eat meat of people 'Cannibals are pecple who used to eat huiuan f lesh8
The sentence in 98 is appropriate in that it characterizss cannibals as 'people eaters* whereas sentence in 99 refer to frequency of human-flesh eatinq.2
Consider again a case of a child who has noticed that a number of een in his hcme-village go to rork in the mines and asking his mother about his dead father:
100. 'Hg, na ntate o-ne a-ee a-ee gauteng?
Mother question father past habitually go to the gcld
'rlother, uas my father used to going to the mines?
For which the nother's answer will be:
101. C he, ngoanaka o-ne a-ruta sekolong se-p hahaman y
no my child past teach school high
'No, my child, he was teaching at nigh school' but not:
102. Che, ngoanaka c-ne a-ee a-rut? sekolong se-phahameng
No my child past habitually teach school high
'No ay child, he used to teach at high schocl* 70
The sentsnce in 101 suggests that the childrs father was a professional teacher wheraas 102 tells us that he frequently taught at high schocl thouqh be sight havs been a qualified doctor. Thus to 102 but not 101 one may add: 103. .. . Laho ja a-ne a-sena lenyolo la- botichere although past Neg have letter of teaching . . Although he had nc teachers certificate' A similar 'characterizingr feature has been ncted by Uoitshlaeger 11976:36) in the case of ~aglish non- proqressive forms. uhile the meaninq cf ~roqressivemay differ in dif fsrent languages, ue could sap that the examples such as above icdicat~ that ge+stem sentences are not kasically progressive.
He may note at this pcint that both the present and th? -----ne+stem sentences may be used with progressive forms as fcllows: 104. Bana ba-sa- ja lebese
children A still eat milk *The children as still eating milk*
105. Bana ba-ne ba-sa- ja lebese
children A past still eat milk
'The children were still eating milkl The use of progressive a 'still* with @-ten= appears to bs countar arquaent against the view that ne is in itself a progressive marker. It may be argued, however, that 32 is redundant in gg sentences, It will be noticed, however, that there is a difference in meaning betueen the sentences with acd those without progressive m. Consider the f cllowing sentencss: D 106. Ha ke-fihla bana ta-ne ba- ja lebese E when I arrive children ~asteat miik
'when I arrived the children were eating milk8
107, Ha ke-f ihla bana .ka-ne ba-sa- ja lebese
when 1 arrive children past stiil eat milk
'ghen I arrived, the children were still eating milk'
The difference between the tuo sentences abcvs is that sontencz
107 presuppose that the children were eating milk before I cams
whereas sentence 106 has nc such res supposition.
4.1.3 ~~fferences/similarities between netsteg 2nd cther past --t en s2~
He have already mentioned that the ne+stem like other past
tenses, refers to the event that took place before the moment of
speaking, Like other past tenses ne+stem may be found in conplement sentences as follows:
Ba- utlo-ile hore morena o-ne a-kula
they hear past that chief past be-ill
'They heard that the chief was ill9 Ba-utlo-ile hore morena G-ife a-kula
they hear past that chief past be-ill *They heard that the chief was illg Ke kholoa hcre o- ile a-t saaaea
I think that past qo
'I think that he left* Ke kholoa hcre c-ne a-tsanaea I think that past go
'1 think that he uas leaving* The iueaning difference between the senteces with ne+- above is rather difficult to characterize, though every speaker of the language knows when to use one and not the other. The difference will for the moment be characterized in tarms of Grice8s conversatioral maxims, (Gxice 1975: 46) quality being ths specific maxis in this case. It is possible that an account which characterize the difference in terms cf tense and aspect will be found later.
The maxi@ required to account for the meaning difference between (a) and (b) sentences in 109 reads: "Eo not say that for which yon lack adequate evidence". In uttering 109(a) the speaker does not commit himself to having witnessed tha actual
'going away8, while 109 (b) could not be uttered unless the speaker has seen x leave. This view is sup~crted by the fact that it is possible to add the phrase abefcre I came* tc 109(a) but not (b) as follows:
110 (a) Ke kholoa hore o-ile a-tsamaea pele ke-fihla
I think that he past go before I arrive
'1 think that he left before I arrived*
(b) *Ke kholoa hore o-ne a-tramaea pelt ke-fihla
1 think that he past go Cefore I arrive
*'I think that he was leaving before I arrived' (Habitual reading pcssible) 7 3
The type of complexity which may involve the use of Grice's maxiins should not be underestiriat~d. The sentences in 109 have been easier to handle since they involve ccnversation between A and B about a third Ferson C. Sentences such as 108 which involve conversation between A and B about what a third person C has heard about a fourth person D may not be as easy to characterize.
The sentences in 108 and 109 have the ~estggand ile+stein tenses as compleutents. It is alsc possible tc have these tenses in the main sentences as fcllows:
111. Re-ne ke-bona hore o-halef ile leha a-ne
I past I see that he be-angry althcugh he past a-leka ho-iphapanya
he try to pretend
**I was seeing that he is anqry although he pretended not to beq 112. Ke- ile ka-bona hore whalefile leha a-ne
I past I see that he be-angry althcugh he-past
a-leka ho- iphapan ya
he try to pretend
'1 noticed that he was angry although he pretended no$ to be.'
It is difficult to decide whether unacceptability of the English translaticn for sentences such as 111 is due to incompatibility of proqressivity with the verbs of inert cognition (Leech 1971:21) or due to a restriction that seritences which arE pmjressiv~ cannot occur as ~ain sentences in complement sentences, If the latter is the case it may be an indicaticn that Sesctho ne+E&g i~ not basically proqressive.
It must also be noted that ne+stsm like other past tenses may ae used with y43g clauses as follows:
1 l3(a) Hoshoeshoe o-ne a-fallela ~haba-Bosiu ha ntataie
noshoeshoe A past move tc ThakaBosiu when his father
a-bo laoa ke-malimc be-killed by cannibals
'Boshceshoe was ntcving tc Thaba-Bosiu when his father
was killed by the cannibals'
(b) Hoshoeshoe o-ile a-f allela Thaba-Bosiu ha ntatage
Roshoeshoe A past move tc Thaba-Bosiu when his father a-bolaoa ke-aaliao
be-killed by cannibals
fioshceshoe soved to Thaha-Bosiu when his father was
kill~dby the cannibalsg
Ths (a) sentence in 113 suggests that moshoeshoe's father was killed by the cannibals during his movement to 'Ihaba-Bosiu. The sentence in (b) suggest that the moveme~,t to Thaba-Eosju followed the death of Moshoesho~'~fat her.
Leech (1971: 17) gives a similar contrast for the English prbgressive and non-progressive sentences. Hs observes that progressive aspect has the effect of surrounding a particular event by a 't~syoralframe*. This account, though accuratz. has been criticized by Woisetscfilasger (1976:30) who man tioris that examples such as those mentioned do not primarily illustrate a point about neiqhbouring tense forms, but rather, something about the semantic properties of the te~pordlconnective whes.
Until a study of Sescthc temporal ccnnectives has been undertaken, we regard ~verlapand succession as characteristic meaning of -+stem and ile+stem tenses respectively.
Reference has already been made to occurrence cf qg and proqrassive sg. nention has alsc keen made of cccurrence cf ~q with immediate past tenses, and its non-occurrence with ile+st2m tense. Bcre discussion on this will be found in the next
We have already noted that ne_ may appsar with habitual 22 while ths other past tenses dc not. This question too will be considered in the next chapter.
The use of prcqressive ~a with ge+stem groups this tense with the iaaediate tenses in tsca- and rile cn the one side and ile+stes on the opposite side. The use of ~e+steguith 'do habitually* separates ne+stea from the rest of past tenses. he cooccurrence restriction with temporal adverbiais referring to the present groups ge+stem tense with the tense in ile+stem.
Wo'te the folloninq examples: Il4(a) *i3anana ba-ne ba- heh ha nama hona joale girls A past cook meat right ncu 'The girls were ccoking zeat this doinent' 114(b) *Banana ba-ile ha-ph~ha namaa hona jcale
girls A past cock meat right now
'The girls cooked meat right now* The use of &e+stem ccnstructions with tenporal adverbs referring to future time must be noted, Consider the following examples: 115. Jobanne o-ne a-fihla hosasa hoseng empa
John A past arrive tomcrrow morninq but joale ntata'e o-itse a-tle kamoso
now his father said he cone day after tomorrow
'Joho had planned to come toacrrcw morning but his father has ordered him to come day after tomorrowo 116, Be-ne re-ea Haseru hosasa, joale rakhali
we past 90 naseru tcmorrcw nou aunt o-se a-tla koano
she already coare here
'We intended going to Baseru but now our aunt is ccmicg over * Sentances such as 115 and 116 express ~lans,intentions, wishes, hopes etc., and should not be confused with the temporal uss cf
---ne+&,eg in which reference is ntade to an event that took place. The occurence of ge+s$s with future adv~rbials should not he regarded as an argu~entagainst the use of pg+steiu as past tense mark3r. The term 'Past Perfect' which has been used by earlier sesotho grammarians appear to be inap~ropriatssince it derives
froa the change of tense of the 'perfect form' for languages
such as English i.e, has > had, The reader rill nctice that the
thzes tsnses &ggp+gtgg, stgul. and ile+st=g may be used with
the 42 morpheme so that it iz difficult to find a term which
covsrs three of thee. Pallcving are examplas in which the three tenses are used with gs
1 l7(a) Ha letsatsi le-likela re-ne re-tsoa tlcha Maseru
when sur A disappers we past leave Baseru
*At sunset we had just left Maseru8
(b) Ha l~tsatsile-likela re-ne re-tloh-ile Hasoru
when sun A disappear we Fast we leave past Haseru
*At sunset we had left Maseru*
(c) Ha lztsatsi le-likela re-ne re-se re-ile
when sun A disappear we past we already we past ra-tloha naseru
we leave Maseru
'At sunset we had already left Has~ru~
118 (a) Ba-no ba-sa-tsoa- utlca here Lerata o-tsoeros they past they past hear that Lerata be-impriscned
'They had just heard that Lerata was imprisoneda (b) Ba-ne ba-utlc-ile hore Lerata o-tsoeroe
they past they hear past that Lerata bs-impriscned
'They had heard that Lerata is imprisoned8 78
For reascw not yat clear to me i&+ietfg tefise requires ac cbligatory use of ~g *already8 as may be noted in sentence 117 (c) . one Ptay think that the optional use of Se with &gs+i&g is due to redundant use of 'alreadya with 'perfectJ as ncted by
Traugott and Haterhouse (1969t298) in the case of English (sea also Hirtle 1977). We could assuae for tha tims beir.g that this is the case since we have noted some similarities betwro,n the imaediate tense in stemfib acd the ~nglishprfsct (2.3.5).
The tegpe indicates that the event referred tc tcok place before another event ir the past. Thus in santences such as 117, the 'departure' tock ~lacfbefore sunset. That the event referred to by ge+gast precedes the time of another event in ths past may be indicated by its incompatib~lity with tem~oral connectives which indicate time after, and its compatibility with those that indicate time befcre. E.g,
119 {a) Letsatsi le-ne le-sa-tsoa-likela pele re-tloha
sun A past it past disappear before we leave Plaseru Haseru
'The sun had just set tffore we left maserul
(b) Letsatsi le-ne le-liketse pele re- tloha Maseru
sun A past it disappear Fast before ve leave Haseru
*The sun had set kefore we left Maseru*
(c) Letsatsi le-ne le-ile la-likela pele re-tloha Haseru
sun A past it past it disappear before we leave Plaseru 'The sun had set tefore we left Maseru' compare
*Letsatsi le-ne le-sa-tsca-li ksla kaaora re- tlcha Maseru after we leave naseru
'The sun had set aftzr we left Plaseru* * Letsatsi le-re le-likstse kainora ra-tloha Maseru
sun A past it disa~parpast after we leava Ha~eru * The sun had set after we left Maseru* * Letsatsi le-ne le-ile la-likela kamora re-tloha Maseru
sun A past it past it disappear after we leave Uaseru
'The sun had set after we left Haseru4
The sense of immediacy is still ielt even when the
immediate tenses are used with qg tense. Thus the ne+%+qp sentences such as 117(a) are felt to be nearer to the reference point than it is the case with ~;e+stem+jlesentence in ll7(t) ..
Although immediacy Hay not be as strongly felt with 222 sten+&&e tense as it is with ne+tsoa tense, most speakers prefer th& use of the immediate tense compounds with recent advertials
and remote tense compounds with remote adverbials. Thus given 80 sentence 121 in which a recent adverblal is used with ne+immzdiate tense and remote adverbial witn ne+remote tense; and sentence 122 in which recent adverbial is used uith ne+remote tense and remcte adverbial with ne+immediate tense most speakers prefer 121, 121 . Seleaong se-fetileng re-ne re-fuman-e mekotla e-menqata year passed we past we find past bags many ho-feta eo re-neng re-ile ra-e-fumana
to pass that past we past them find
lilemong tse- tharc tse-f etileng
we past them find years three passed
'Last year we had found aore bags than we had thr~s years ago9
122. selemoog se-fetileng re-ns re-ile ra-fumana mekotla
year passed past past we find bags
e-aengata ho feta eo re-neng re-e-tumane
many to pass that we past them find
lilemong tse- tharc tse-f etileng years three passed
'Last year ve had found more bags than ue had three
years ago8 arasressise Sn IstU Ne-compound tenses whoso simple ccunterparts can be used uith sa map themselves be used with sq toc; and tho ones whose sim~le tehses cannot be used uith gg cannot e.g.
123. Re-nf re-sa-tsoa feta flaseru ve past Fast pass Maseru 'Oe had just passed ?laserua (b) Re-ne re-sa-fet-ile Plaseru we past pass past Maseru
'He had only passed aaserus (c) *Re-ne re-sa-ife ra-feta flaseru
we past past pass Basseru
*i?e still gassed naseru*
X~g~ralg dver bs
Although immediate tenses may, on thair own, be used with an adverb which refer to the present, ne+imaediate tecses cannot be used with such adverbs, e.g.
124 (a) *0-ne a-tsoa-bua le-rona hona joale
he past past speak with us right now
'He had just spoken to us this moment*
(b) *O-ne a-buile le-rona hona joale
he past speak past with us right ccw
'He had spoken with us this moment'
The use of pi?+-t with temporal adverbs deserves further note.
Hornstein (1977:530) notes amkiguity in the f cllcwing tense-cum- adverb English sentence:
125. The secretary had eaten at 3 p.m, which may be ~araphrasedinto:
(a) The time that the secretary actually ate was 3 p.m.
(b) Tbe secretary had already eaten by the time 3 p.m.
rolled around, 82
There seem to be no aetiquity wlth neL~sgg+ile and -ce+ile+gteg tenses in Sesotho, Let us consider the follcuing sentences: 126. Theko o-ne a-no-ele likala tse-tharo ka-3
Theko A past drink Fast cans three at 3
*Theko had drunk three cans of beer at three*
127. Thekc o-ne a-ile a-noa likala tse-tharo ka-3
Theko A past past drink cans three at 3 *Theko had drunk three cans of beer at three*
The sentences in 126 and 127 tell us that the time at which
Theko drank beer was 3 08clock. In order to express the meacinq that the drinking took place before three we have tc use zg * alreadyt as follows: 128. Theko o-ne a-se a-no-ele likala tss-tharo ka-3
Theko A past already drink past cans three at 3
*Theko had already drunk three cans of beer at 3'
129. Theko o-ne a-se a-ile a-noa likala tse- tharo ka 3 Theko past already past drink cans three at 3
'Theko had already three cans of beer at three'
The sentences such as 120 and 127 should not be regarded as a counterargument tc our earlier view that netnast indicatss that the event referred to took place before another event in the past. Senter,cos such as these are seraantically dependent on other sentences (see Smith 1978:51 for the analysis of corresponding English sentences). Let us consider a sentence il? which the time specified by the adverbial is the time at which the other past event took lace:
730. *tIa re-fihla ka-3, Theko o-ue a-iie a-noa
when we arrive at 3 Thekc past past drink likala tse-tharo cans three
'At three vhen we arrived Thako haa drunk three cans of
compare
131, Ha re-fihla ka-3, Theko c-ne a-ss a-ile a-noa
when we arrive at 3 Thekc past alrzady past drink likala tse- tharo cans three
'At three when ue arrived Theko had already drunk thr~e cans of beers
The sentence in 130 is unacceptable because it ideritiflos tho time of the other past event as the saae as that of tense, When the overt marker 'already* is used, the sentence is acceptable,
-----Summary
--Ne is regarded as past tense marker instead of aspectupl marker as earlier graamarians suggest, Like ile+s&g& tense, ------ne+stem cooccurs with adverbs which refer to the past time but not the ones referring to the present time. Unlike ether Fast tsnses, ne+stew Bay have habitual meaning in isolation, and at the same tine cooccur with habitual marker. In 'when( sentenc~s 84 -----ne+stem expresses time overlap while the cther tenses ex~ress succession, Ehen &g is used uith cther past tenses it indicates that the event referred tc by pwqgt took place before ancther event in the past. As a result, this tense cannot be used dith temporai connectives which refer to time after. 1 . Unmarked here aeans absence of overt marking. 2 . A sentence such as this may be appropriate, for instance, where somecne knows that during Lifaqane some
people (say Zulu) frequently ate human flesh; and in
speakers opini.cn they are not tc be characterized as ' people eatingg. CHAPTER 5
He have qiven a list of some aspectual markers in 1.2.3. Bsfore examining some cooccurrence restrictions between these markers and past tenses we have to dstermfne soms secondary aspectual i~plications inherent in different tenses. Our analysis will be based on the oppositions established by Priedrich (1974: 35) . Priedrich suggests that the three most basic oppositions in terms of which a variety of as~ectual systens can be analyzed are durative vs nondurative, completive vs noncompletive, and stative vs nonstative. In order tc be consistent with our terainology we will us3 perfective/itaperfective instead of completiv~/noncompletive opposition (we ignore the distinction saae batwe9n ccapletive and perfective aspect (see Yoisetshlaeger 1976: 28) .
A number of works on aspect have made it clear that the examination of aspect must be carried out in terms cf intra- sentential relations not by looking at the verb in isolaticn
(see Verkuyl 1972, Friedrich 1974). A Bore detailed treatment of aspect vould require us to go intc the question of interdependence that holds between aspect on the one hasd and nuaber, countability, specificity of reference and negaticn cn tfie other hand. Our treatment will be selective rather than comprehensive; we will examine a relativ5ly small number of points than it would be case with an elaborate study on aspect as such.
An exact definition of ~erfective/impezfective opposition
IS difficult tc find. Coetrie (1976: 16) notes that ~erfectivity indicates the view of a situation as a single whole without distinction of the various separate phases that make up that situation, while iaperfectivity rays essential attention to the internal structure of the situation. As Macaulay (1978: 4 17) correctly observes, the above distinction is rather cpaque.
Comrie 1ater expands the def initicn of perfectivity to includa
'situations that are internally ccmplex) such as those that last for a considerable pericd of time, or include a nuniber of distinct icternal phases provided that the whole situaticn is subsumed as a single whole. Ue will regard corapletion as ac essential chdracterizati.cn of perfectivity and assume that a tense which nay be used with the form qetplla 'end by' is perfective while the one that cannot be used with this form is imperfective, Let us consider the following sentences:
132(a) * Ke qeteletse ke sa-tsca- pheha
f ended I past cock
11 have ended up ccoking'
(b) *Ke-qeteletse ke sa-tsoa-tsamaea
1 ended I past walk
*I ended going* 133 (a) Ke-geteletse ke- pheh-ile
* 1 ended I cook Fast
I ended ccokinq*
ke qeteletse ke-tsama-ile
I ended I walk past
*I ended qcingl
134 (a) Ke-qeteletse ke-ile ka-pheha
I ended I past cock
I ended cookingi
Ke-qe teletse ke-ile ka- t samaea
1 ended 1 walk
I ended going* 135 (a) * Ke qeteletse ke-nts ke-pheha
I ended I past cock
* I ended cookinq, * Ke-qeteletse ke-ne Ice- tsamaea
I ended I past walk
*I ended going* 136 (a) * Ke-qeteletse ke-ne ke- sa- tsoa-pheha I ended I past still cook
1 ended just cocking' * Ke- qeteletse ke-ne ke-sa-tsoa-tsamaea
f ended I past I still walk
'1 ended just gcing*
'1.37 (a) *Ke-qeteletse ke-ne ke-pbeh-ile
I ended I past I cook past I ended having cooked* * Ke-q~teietse ke-na ke-tsaaaile
I ended I past walk past
* I ended having gcne' *Ke-qeteletse ke-ne ke-ile ka-phehd
I ended I past I past cock
* I ended having cooked' * Ke-qeteletse ke-ne ke-ile ka-tsamaea
I eniied I past past walk '1 ended having gcne'
The sentences in 132 are examples of tsoa past tense; 133 ------stem+ile past tense; 134 ile+sten past tense ; and sentecces 135 examples of ~+stempast tense. Sentences 136-138 are exaaples of agfpgsf, 2-2s . We may conclude froa the acceptability cf sentences 133 and 134 that stem+i&g and i1.2+s_tg tenses are perfective; while unacceptability of sentences 132, 135, 136, 137 and 138 suggest that &soa+stgr, =+stem , and qg+~&& tef,s= are imperfective.
The perfective/i~perfective distincticn noted abcve may be further denonstrated by the use cf the fram* ' he was ,. . hut did not finish'. Tenses which do not fit intc this frame will be regarded as perfective.
139(s) 0-tsoa-fiela 'ae ha a qeta 1
he past sweep and Neq finish * Ha has just swept and has not finished'
(b) * 0-fietse 'ne ha a qeta he sweep past and Neg ficish 'He swept ard did rct finish* * 0-ile a-fiela 'me ha a qeta
he past sweep and Nsg finish * HS swept and did not finish1
0- ne a-fiela (ae ha a qeta
he past sweep and Neg finish * He was sueepinq and did not finis
0-ne a-tsoa-fiela 'me a =a qeta
he past past sweep and Neg finish
'He had just swept aod had not finisheda
0-ne a-fietse 'me a-sa- qeta
he past sueep past and Neg finish
* He had swept and had nct finisheds 0- ne a-ile a-fiela *me a-sa-qera
he past Fast sueep and Beg finish
He had swept and had nct finishsd'
As may be noted from the sentences in 139, tsoa+stent , ne+stem and qe+east are imperfective while ~tein+ileand ile+ste& are pzrfective.
5.1.2 pqrative 12 ncndurativg ~ppgpitis
The distinction between durative and nondurativa aspect examined in this secticn primarily concerns the possibility of sentences in different tenses to contain durational adverbials such as mkp e-telglg 'f cr along time1 Jetsggsi lohla 'for the
I whole day' etc. We will indicate that there is a -^Lelaticn cf cohesion between certain tenses and durational advertials
similar to that between tense and te~poraladverkials. We dc nct
deny that durational features may be inherent in lexicai itenis
as such rather than their gra~uiaticalform. Note th? follcui~q examples: 140 * Johanne o-fihl-ile nakc e- telele John arrive past time long John arrived for a long time8 141 * Sello o-qal- ile mosebetsi ho-fihlela nantsiboaa* Sello begin past work until afternocn * Sello began work until afternoon' compare
142 Thabo o-bu-ile nako e-telels Thabo speak past time long
* Thabo spoke for a lcng time* 143 Fekisi o-luts-e moo ho-f ihlela mantsiboea Pekisi sit pst there until afternoon
*Pekisi sat there until afternoon9
The unacceptability cf sentexes 140 and 141 cannot be attributed to cooccurrence restrictions bztween ~iem+iletense
and duratinal adverbial nafo *for d long time' and $21
fihlelq g&otdboea 'until afternoon* since the acceptable
sentences 142 and 143 in g&ggz&letense are also used with the
same durational adverbials. The unacceptability of 140 and 141 92 reflect lack of durational featur9s in verbs such as fihla 'arrive* uala 'begin' ~g~fh1break8 atc.
We are aware that durational adver~ialshave a selectional relationsh~pto a configuration cf categories rather than cne lexical iter as proposed by Lakoff and Ross (see Verkuyl
1971:18) ,
we are also aware that verbs such as chcatla 'break' may be durational or nondurational depending on whether the object NP is singular or plural. We thus have:
144. Haria o-choatl-ile lii~cneletsatsi lohle nary break past mirrcrs day whole
*PIary broke the mirrors for the whcle dayo
but not
145, *flaria o-choatl-ile seipone letsatsi lokle Mary break past mirrcr day whole 'nary broke a mirror for a long time1 The sertence in 144 does not necessarily express frequency. It may have a single event readinq. lhus if nary went on a rampag2 against mirrors she eight decide to go cn breaking until all ar* brokerr.
There are several other factors which may affect duration. one of these is the difference between the ccnstituents occurrinq as subject. E1gl 146 (a) *~oetio-f ihl-ile hof ihlela mantsihcza visitor arrive past until afternoon 'The visitcr arrived until af ternocnn
It) Baeti ba-f ihl-iie hof ihiela mantsi~oea visitors arrive past until afternoon 'Visitors arrived until a•’ternoong
As the reades aay notice, we are left with a rather liaited
choice of 'neutral' lexical iteas. An attenapt will be mads to find sentences which may not be strongly affected by thz factors
mentioned above, so that it should be possible to claim that
tense is the only contributing factor to the cocccurrence restriction between the verb and the durational adverb. *Banana ba-sa-tsoa pheha nama nako e-telele
girls A past cook meat time long
'The girls just ccoked itifat for a long time' *Banna ba-sa-tsoa sebetsa natc o-telele
.en A past work tine long
'The men just vorked for a long times Banana ba-pheh-ile nama nako e-telele
girls A cook past meat time lcng
'The girls cooked meat for a long tiaa9 Banna ba-sebel-itse nako e-telelo
men A work past time long 'The men worked for a long time' Banana ba-ile ba-pheha nama nab e-telele
girls A past cook meat time lcnq 'The girls cooked meat for a long time'
Banna ba-ile ba-sebetsa nako e-telele men A past work time long
'Hen worked for a long tias' *naobane banana ba-ne ba-pheha nama nako e-telele yesterday girls past cook meat rims long
8Yesterday the qirls cooked meat for a long time8 *Maobace banna ba-ne La-sebetsa nakc e- telele yesterday men past work time long 'Yesterday the men wcrked for a long time' Banana ba-ne ba-sa- tsoa heh ha naaa nako e- telele girls A past past cook meat time lcng 'The girls had just cooked meat for a long tiae* Banna ba-ne ba-sa-tsoa-sebetsa nakc e-tflele men Fast past work tine long 'Hen had just worked for a long tias8 Banana ba-ne ba-pheh-ile nama nako e-Lalzle girls past cook past seat time lcnq
'The girls had cocked aeat for a long time*
Banna ba-ne ba-sf kel-itse nako e- telele men past wcrk past tine long
'The men had worked for a long time' banana ba-ne ba-ile ba-pheha nama nako e telele girls past past cook aeat time lcnq
'The girls had ccoked meat for a long time8
Banna ba-ne ba-ile ba-pheha nama nako e-telele laen ~astpast work time long
'The men had worked for a long time' i
95
The sentences in 147 are examples of tsoa+stem past tense used
vith Zurativ~adverbial. As may be noted from unaccaptabiiity of
these sentences, this tense is nondurative, The s_tem+ile and ------ile+stere tecses on the other hand ara duratlvs. Ye have used ar, adverb paobane 'yesterday' with sentences In 150 in crder tc
exclude habitual reading, The ge+lfea like tsoa+stew tense is
noodurative. The ge++ tense sentences in 151-153 suqgest that
these tenses are durative,
It is interesting to note that nost tecses regarded as psrfective (133-136) may also be durative (148-149); and most
tenses which are imperfective (134 and 137) are also fiondurative (147 and 150) , This suggests a possible objection to the 'either
oru opposition between durative/nondurativs and
conpletive/nonconpletive suggested by Friedrich (1974: 35) 3rd
aacaulay 1971 (see flacaulay l978:4 19).
The sentences vith gigast tense seze tc follow a different pattern from the other past tenses. These tenses are bcth iraper fective (138- 140) and durative (151-153) .
The distinction between stative and nonstative verks in Spsotho is not clear, Although certain vsrts have been referred
to as stative, the critericn used for such classification has
never been stated (see Doke and nofokeng 1957, Guma 1971). Some
grammarians have doubted the need for distinction betwasc 96 stative and nonstative verbs (Usstphal 1950). Most grazmariacs have, however, noted an overt stative tsarxer -eh-, ijote ih- followinq examples,
154, Rata 'like' > ratfha 'br likeable'
ruta 't~ach' > ruteha 'be educated' roba 'break' > robifha 'be brokena lebala 'forget' > lebaleha *be forgotten*
These constructions are different from passive constructions in that the agent is only implied but can nevez be specified.
We will regard incompatibility of -eh- with any verbs as an indication of stativity, Sciae of the verbs which do not take - eh- are the following:
155. Plakala ' be surprised'
hlonama 'sulk' tsoana 'resemble' inaea 'stcop' phahama 'rise up'
khothala 'be cheerful' khathala 'get tired*
shoa 'die*
The above examples may suggest sow relationship hetween stativity and intransitivity. 97
Let us now consider the use of both lexlcai and derived stative verbs with different tenses:
156 (a) *Tebcho o-sa-tsoa rcteha ho-•’eta tnaka tsa-hae
Teboho A past educat~dtc pass age-mate of his
*Teboho has been much mote educated than his age matas8
(b) *ke-sa-tsoa makala hc utloa taba tseo
I past be surprised to hear news those
'1 have been surprised tc hear that news*
157 (a) Teboho o-ruteh-ile ho-f eta thaka tsa-hae
Teboho A educated past to pass age-mates of his
*Teboho is much more educated than his age mates*
(b) Ke maketse ho utlca taka tseo
I be surprised to hear nevs those
*I am surprised to hear that nevs*
158 (a) Teboho o-ilo a-ruteha ho-fe ta thaka tsa-hae Teboho past educated to pass age-mates cf his *Teboho was much more educated than his age mates*
(b) Ke-ile ka- makala hc-utloa litaba tseo
I past surprised to hear news those *I was surprised to hear that news*
159 (a) *Teboho 0-ne a-ru teha hofeta thaka tsa- hae
Taboho past educated to pass age mates of his
'Teboho was much more educated than his age mates8
(b) *Ke-ne ke-eakala ho-utloa litaba tsco
I past 1 surprised tc hear news those
'I was surprised to hear that nevs* Teboho o-ne a-sa- tsoa-rutsha ho-f eta thaka tsa-hae
Tebo ho past past educated to pass dqi mates of his
'Teboh o had just b~eneducated much mere than his matesg Ke-ne ka-sa rsoa makala ho-utloa taba tseo
I past past surprised to hear news those 'I had been just surprised to hvar that nawsl
Teboho o-ne a-ruteh-ila ho-fe ta thaka tsa-hae
Teboho past educated past to pass age-mates of his
'Teboho was euch more educated than his aqe mates1 Re-ne ke-maketse ho utloa taba tseo
I past I surprised to hear news that 'I uas surprised to hear that news* Teboho o-ne a-ile a-ruteha ho-feta thaka tsa-hae Teboho past past educated to Fass aga-mates of his
'Teboho had been much more educatea than his friends* Ke-ne ke-ile ka-makala hc-utloa tam tseo 1 past past surprised tc hear news thos~ '1 had been surprised to hear that news1
As may be noticed from the above sentences, tenses which resist use of stative verbs are &=+sten! and ne+stea. These tenses may thzrefore, be regarded as nonsiative. The g&galg ile+~tsanh ne+~asttenses on the other hand are stative.
For reasons not yet clear the stative verb ratehp #be likeable' shows a different behaviour from other stative verbs.
With &em+ils and jle+stem tenses the sentences with yatsha are 99 unacceptakle while with ug+ggteha the sertence is acceptable.
The following list of aspectual markers was givsn in 1-2:
sa *stillr
Se 'already*
Of these markers only sa *still( and sg 'already* may he ussd with a nuaber of past tenses. Iiakitual markers may De used with -ne+s&eg tense only. We have ncted use of gg with this teasa in 4.1.2. The other habitual markers behave in the same way as ee and may (like ) be characterized by the fact that they cannot specify a particular moment in time. This characteristic feature is also found with & 'do occasionally'.
Analysis of Sa 'still * and _se 'already' is of particular interest since observations have been made about the restrictions of lstill* and 'already* to certain tenses and aspects in languages such as English (sle Trungott and Haterhouse (1 969), ncrrisey 11973) , Hirtle, 1977) . The label 'progressive' used in grammars for sa+stem will be retained. The form sa_ a still1 gives to tne v&rb the notior, of continuity. Thus a sentence such as K%rsas z&p 'we are still buying1 presuppose that the event referred tc 'rekag has b~~n going cn before the momect of speech and is bzing continued.
The heaning of progressive given above requires that the verb used be durative (e. last for a certain period cf time).
It is possible, however, to use Sa with punctual verb (not lasting in tiae) such as fhhl4 *arrivev, tA~hq'leave8 otc. In such cases the meaning expressed may be ' just8 or 'onlyg. The sentence Q-sa-fihlq mean @he has just arrived*.
The aeaning of ~g is redundant in this tense. The prograssive form is frequently used in order to neutralize tho verbal force of tsoa 'come froag.
The fofm sg with gtea 2 expresses the meaning $againg e. g.
163. 3 anana ba-sa-re k- ile libuka
girls A still buy past books
*The girls have bcught books again8 The 'again' nsaning with the s&gg 2 tense appear tc derive from the fact that thi~tense is perLectiva so that the continuous meaning cf the progressive forces a repetitive reading onto the verb.
Although it is pozsible to use the simple fcrm of stative verbs such as &g&p 'know8 with m, the past tense of this verb cannot be used with a, Yhile we map say, [of a bey whc learnt
Sesotho and went to a place where he could not use the language for some tiae) :
164. Teboho o-sa-tseba Sesotho Teboho A still kncw Sesotho 'Teboho still knows Sesotho8
we cannot say:
165. *Teboho o-sa-tseb-ile Sesotho
Teboho A still knew past Sesotho (Teboho still knew Sesothol
An explanation for this apFears to be that it is not possible tc
express the 'aqaint meaning ncted for stemfile. We normally speak of someone 'learning again* but not 'knowing again*.
The 54 fora cannot be used with ile+st~qtens?. Several
reasons for this cooccurrence restricticn come to mind. One
aight assume that expresses imperfecti vity while ile+st?m
express perfectiity, This cannot be a valid reason for this 102 rlstriction since we have indicated in 5.1.1 that szgi + ilg is a perfective tecse. In fact the naeaninq 'ayairi* expressed by Sa+ ------stem+ile appears tc sup~ort the view that this tense is perfective. There seem to be no reason vhy one perfective tenss should occur uith sa and the cthor shoul3 nct.
One could also suggest that syntactically and j& fill the same position sc that once one has been used the cther cannot. This seems plausible since the gg fcrm cannct be regarded as filling the same psition as i& for the two nay appear in the same conztxuction. It is doubtful, however, that this could bs a valid claim since tsoa (uhich we assume fill the same position as a)may occur with a. The questicn cf markedness expressed in 3.2.1 remains the valid claim.
The meaninq of sa in this tense is the saae as that mentioned in 5.2.1.
The seguence above exFzesses the meaning 'had just'. As ma.y be noted, the eaning of sa is still redundant.
5.2.5.3 ne+sa+stem+ile
The aeaning expressed by this ssquence is 'had only'. Unlike giq sill, Se 'already' does not presuppose existence of the event, cooccurring with it-, before the moaer,t of speech, out rather presuppose its nonsxlste~ce bff ore the moment of speech and its 'corning into being',
It must be noted that while a sentence with ~p im~lissthat another event follcws that expressed by sa+ste~a sentence with --se implies that another event preceded the cne expressed with -se. Thus the sentence: 166 (a) Re-sa-reka libuka
we still buy books 'We are still buying books' may be completed thus: (b) ... Re tla reka liphahlo ha re-qeta we Put buy clcthes when we finish '.. . We will buy clothes thereafter' while the sentence:
167 (a) Re-se re-reka libuka
we already kuy books
'We are already buyinq bcoks*
nay be prgceded by: (b) Re rek-ile liphahlo joale .. . . We buy past cloth,- now .... 'We bcught clothes ncw we are buying books1
The only reading we get with this tense is 'already cone from9. nith verbs that are iaccmpatible with 'comingt e.g, Sloha
'lsavsr the sentence is unacceptable, 168. *Ba-se ka-sa-tsoa-tloha they already past leave * @Theyhave already just left*
ahen used with temporal adverb ~gksgem+ile express change of plan expectation etc, Thus the sentence:
169, Ba-se ba-tsama-ile ~aobane they already walk Fast yesterday @They already left yesterday' imply that the people referred tc were expected to qo at a tims different from the one expressed ty adverb.
This seqnencE conveys the same meaning as that i~
5.3-2 above. The meaning in this sequerce is similar to that mectioned in 5.3.1,
5.3.4.2 *N&+se+tsoa+stem
This sequence is unacceptable.
5.3.4.3 ne+se+stem+ile
The naaning ex~ressedis 'had already*
5.3.4.4 ne+sa+ile+stem
The meaning is the same as in 5.3.4.3 above.
The folloving chart represent the various tenses in relation to aspectual features and cooccurrence restrictions with different as~ectualmarkers. t 1 1 1 111 ]PERFECTIVE JDURATIVE ISTATIVE jHABITUA1 jsaj sel
TABLE XI -Tenses in Relaticn to Arpectual Features ---Notes &p -2s 5 1 Scme speakers prefer the use of verb qala 'begin' with this sentence, e.9,
0-tsoa-qala ho fiela 'me ha a geta
*Be has just started sweeping and has not finished1
Note that ceqative ccnstruction with the same adverbial
is possible, e.g, Sells q-kg g qqlq ~osebetsigq -----fihlela ------raantsiboea 'Seflc did not begin work uatll afternoon', In the preceding chapters we have given the description of
Tense and Aspect in Sesotbo. Ye have demonstrated that the sten+ile form which has been referred to as 'perfzct* in the literature should be regarded as an immediate past tense.
Comparison between this tense and the immediate past tense in -----tsoa+s teg is made. We ha~~also indicated some similarities/dif ferences between imaediate and remote past tenses. We have suqqested that geistem shculd be regarded as
past tense instead of aspectual distinction. Finally we have indicated some as~ectual features inherant in different past tenses,
In this chapter we set out tc indicate the extent to which
~eichenbach/Hornstein analysis can acccunt for tha facts mentioned above, Ye cffer no solution to thz probleres that the theory based on this analysis is likely to encounter in dealing with Tense and Aspect in Sesotho. We nerely vish to pcint out some of the facts that must be considered in the formulaticn of a tense theory which may adequately describe th? facts and perhaps make a~propriatepredicticns. One of the main tasks which any thacry of tsnse based on ~eicbenbach~sanalysis will encounter in dealing with Sesctho tenses is to draw the distincti.cn betwaen immediate and remote tanses, We have already noted that as far as the relaiicnshi.~ between the point of event ( E ) and tha raornznt of speech I S ) is concerned, E is anterior to S in both immediate and re~ote tenses, Thus in the sentences: 170 (a) Johanne c-tsoa-fihla
John A past arrive John baa just arrived*
(b) Johanne o-f ihl-ile
John A arrive past John arrived8
(c) Jofialine o-ile a-fihla
John A past arrive John arrived8 John's arrival is viewed as prior to thd moment of speakinq. without time specification one map be inclined to thirk that the point of reference in immediate tenses snould be viewsd as simultaneous with the mcment cf speech while with the remo!e tense it should be regarded as sirultaneous with the point of event. We would then have the following 171(a) as tense representation for sentences 170 (a) and (b) and 171 Ib) as tense representation for sentence 170 (c)
171 (a) E ------SfR (b) E,R ------S
The representation of 170 (a) and (b) as 171 (a) would be supported by the fact that these tenses may be used with temporal adverb h~mjcglg 'right ncug while sentence 170 (c) cannot. It nust be noted that the tense structure in 171(a) corresponds to the ~nglish perfect, whcse differences with
Sesotho immediate tenses are noted in 2.3.5, It should also be noted that the three sentences in 170 may bs used with adverbs referring to the past as follows:
172 (a) Johanne c-tsoa f ihla raobane
John A past arrive yesterday 'John just arrived yesterday
(b) Johanne o-f ihl-ile ~aobane
John A arrioe past yesterday 'John arrived yesterdayg
(c) Johanne o-ile a-fihla aaobane
John A past arrive yesterday 'John arrived yesterday8
In all these sentences the point of event is simultaneous with the point of reference, The tense representation for the three tenses thus become 172 fb) .
~hileone could probably manipulate the three theoretical entities S R and E and finally ccme out with some scluticn, there akpears to be another problem closely bound up with the question of immediacy and remoteness which may be difficult tc incorporate within SEE. This is the question of subjectivity. 11 1
Thus although in all the sentences in 172, yesterday is
identified as the pcint cf reference, ths speaker may chccse cne sectence instead of the other depending on how he views the occurrence of the event in relation to his ex~ectations. Thus sentences 172 (a) and (b) may be an appropriate answer where th~ speaker had expected that John shculd have arrivsd earlier.
Several other interrelationships between the actual time cf occurrence of events with the invclvement ok the speaker, hearer
and other pasticipants in the events have been noted by Lakcff (1970) . It is necessary that such facts be considered in the foraulaticn of tense theory,
One of the ways in which Hornstein related the quasi- ~eichenbachfraaework he proposed to languaqe acquisition, was
that it would provide the child with a rather system with vhich to work in deciphering the tense phenomena and their
corresponding temporal interpretation. Cne may view exclusicn of
the distinctions such as iamediate/remot~ as an attempt at
providing such a tight system, It is doubtful, however, that exclusion of certain facts abcut a lanquage would Frovide a unified account, One may note in this connection that the claim
made about degree of markedness in immediate/remote tecses suggests a possibility of acquisition of scmf tensss before
others.
It nay be arqued that the distinction between immediate and
reniote tenses should be regarded as aspcctual, One could expand 112 the definitioc of aspect given in 1,8.1 to include external constituency of event (e.g. Hanner, repetition etc.) . In regarding the distinction expressed by &soa+stem (immediate)
lle+steg (remote) as as~ectual,ue claim that ts~pand jlfg forms
fill the same position as aspectual markers. We have ncted in
the previous chapter that t=+stea and iJe+st+q may be used
with such aspactual markers as sp 'stillg and ' already* This
indicate that &pg and j& fill a different positicn from that
filled by aspectual markers. Ye cculd also mention that while it
is possible for aspectual markers to be used with several
tenses, and forntr can never be used with other tenses.
We thus, never find any of the following sequences:
173 (a) *Johanne o-tsoa tsama-ile
John A past walk Fast for 'John just left'
(b) *Johanne o-tsoa ile a-tsamaea
Jchn A past past walk for 'John just left* but we have the following sequence: 174. Johanne o-se a-ile a-tsamaea
John A already past walk
'John has already left'
Reichenbach notes that there may be certain deviatiocs frcm
the original meaning as a result of combination with certain 11 3 adverbs "... In the sentence 'Nou I shall go' the simple future has the msa~inq 5, B----- E, .. . How~ver,in tne sentenca ' I shall go toaorrow' the future tense has the meaning S----- H,E.
It is out of this cbservation that Hornst2i.n developed rules cf adverbs which account for acceptability and unacceptability of certain adverb/tense combinations. Let us consider the follcwinq sentences:
Thabo o-tsoa fumana lengclo hona joale
Thabo A past find letter right now
'Thabo has just got a letter now1 Thabo o-fumane lengolo hcna joale
Thabo A find past letter right now 'Thabo got a letter nowc *Thabo o-ile a-fumana lengolo hona joale
Thabo A past ficd letter right now
**Thai10 got a letter nows ThaCo o-sa tsoa f umana lengolo maobane
Thabo A past find letter yesterday 'Thabo just got a letter yesterday' Thabo o-f uman-e lengolo maobane
Thabo A find past letter Yesterday 'Thabo got a letter yesterdayt Thabo o-ile a-f umana lengolo maobane
Thabo A past find letter, yesterday 'Thabo got a letter yesterdayQ *Thabo o-tsoa fumana lengolo hcsasa Thabo A past find letter tomorrow
**Thabo just got a letter toinorrow' *Thabo o-f ulaan-e lenqolo hosasa
Thabo A find past letter tomorrow *Thabo got a letter tomorrow'
*Thabo o-ile a-f umana lengolo hosasd Thabo past find letter tcinorrcw **Thabo qot a letter tcmcrrow'
have mentioned earlier that tense representaticn for past tenses is E, R---- S,, According to Hornstain's analysis all the sentences in 175 should be unacceptable since the combination of tense and advert violates the principle cn associativity vhicb disaflows the association of S R and E in derived structures if they were nct associated in basic tense structure, The use of adverb boa jqgdg 'now8 with E, R---- S changes this structure into E---- S,R. Be note however, that although t he principle on a~sociativity is violated the sentences in 175(a) and (b) are acceptable. If sentence 175 (ct was acceptable too, it would probably be necessary to ignore tho above principle.
Another possibility would be to assume that all the past tenses have the structure E---- S,R and that past adverbs are mapped into E (and R where they are simultaneous) where E is anterior to S; while present adverbs are mapped intc R cr E when they are sirauitaneous with S, The acceptability cf (a) and (b) 115 sentences in 175 would then be due to associaticn of the adverb hona jg& with Z; and the acceptability of sentences in 176 due
to the association of w=gg 'yesterday' uith E, While this may appear to be a plausible sclution, it does nct acccunt for the unacceptability of sentence 175 (c) ,
He are nct able to come out witn any solution to the
problem of the interaction between adverbs and Tense within
Reichenbach8s analysis. It is possible that the analysis cannct
be adapt~dto relative tenses, or different constraints may be required in order to account for ths above facts. Until such time that a solution is found one cannot understand what is taking place in compound tenses,
The only aspectual feature which has been inccr~oratedinto Reichenbach8s analysis is progressivitg. Progressivity is but
one of the semantic noticns of aspect, Others are stativity,
duration, completion, habituality etc, Smith (1978: 69) has indicated a possibility of incorpcrating ha~itualsentences into
Beichenbach's scheme for teeporal specificaticn. While I dc not
in principle accept Smith's analysis, it suqgests sonte possible modifications on Reichenbach's analysis.
It is possible that a number of aspectual featurss can
easily te accounted for by redundancy rules, We rotice in table
11 for instance, that all tenses that are durative are alsc 116 stative. Binnick (1976: 44) notes an unpublished analysis, by
PlcGilvray based on aeichsnbach which suggasts that th-> 'adsf ini-tionaltt function of asserting the (aspectual) relatcnshi~
ketwezn E and R pertain tc all verbs within a sentence. This is
understood to suggest a possibility of eliminating even more aspectual features by more redundancy rules.
Our analysis has indicated the irportance of ~aintaininq the distinction between imsediate and remote tenses in Sesotho;
and the need to consider necessarj constraints to be intergrated into an explanatory and uell- constrained theory of Tense.
Our presentation is ky nc means definitive, Thsre are a
number of topics which have not been examined becaus* of lac^ cf time and space, soae of these are tenporal co~nectivesand their interaction with different tenses; serial vsrb constructiocs
etc. Ye hope that this study will servs as a stimulus for research into syntax and semantics of Sesothc sincs nc detailed description in any topic has been carried out. Allen, "n I.. 1966. The vert system of present day Americac Englirh. The Hague: Houton,
Anderson, J, 1973, An essay concerning aspect, The Hagua: Mouton,
Beauchat, P-D, 1966. The verb in Zulu, Jonannesbury: Witua tersrand University Press.
Binnick, R, I. 1976, Row aspect languages get tense. Chicagc Linguistic Society 12 40-50.
Bull, P. E. 1960, Time, tense and the verb, Berkeley: university of California Press,
Chomsky, B. 1965. Aspects of the theory of syntax. cambridge: HIT Press.
Comrie, B, 1976, Aspect. Cambridge: ~ambridcnunivsrsity Press.
Crystal, Dm 1966, Specification and English tenses. Journal of Linguistics 2 1-34,
Dillon, G. L. 1973, Perfect and other aspects in a case grammar of English, Journal of Linguistics 9 271- 279,
Diver, u. 1963. Tbe chronological system of the English verk. Pord 19 141-181,
Doke, C. ll, 1927, Textbook of Znlu graamar, Cape Town: Longmans.
Doke, C. He and aofokeng S, PI, 1957, Textbook of Southern Sotho gramsar, Cape Town: Locgmans.
Dowty, D, R, 1977, Toward a semantic analysis of verb aspect and the English *Iniperfectivsl progressive, Linguistics and Philosophy 1 45-77,
Friedrich. P. 1974. On aspect and Homeric aspect. Internstional Journal of Aaerican Linguistics 40 fYemc'ir 28,
~ivon,T, 1972, Forward imQications, kackward presuppcsiticns, and the time-axis of the verbs, In Kimbal, 3. Syntaxs and semantics 1.
------me- 1972. Studios ir, ChiBeaba and Bantu qraemar. Studies in ~frican Linguistics su~plement 3. Chapter 4, ------I 973. The time-axis ph~nomenon. Lanquage 49 890-925,
Grico, P. 1975. Logic and conversation. In Cole, i?/ HargaL 3. Syntaxs and semantics 3.
Guma, S. P!, 1971, An outline structure of Southern Sotho. Pietermaritzburg: Shutfr and Shooter.
Hirtle, W. H. 1977. Already, still and yet, Archivum Linguisticua 8 28-45. ------1967, The si~pleand progressive forms: an analytical a~proach. Quebec: Lss Pressfs d~ 1'Universite Laval. ------_-----1975, Tiae, aspect and the verb. Quebec: Les Presses de 1*~niversiteLaval.
Hornstein, N, 1977, Toward a theory of tense. Linguistic Inquiry 8 521-557,
Huddloston, R, 1969. Some observations on tense and deixis. Language 45 777-806,
Jacottet, E, 1965, A practical method to learn Sesutc, nori jar Sesuto Book Depot.
Kirpasky, E. 1968, Tense and mood in Indo-European syntax. Foundations of Language 4 30-57.
Kucera, H, and Trnka, K, 1975, Time in Language: tamporal advertial ccnstructions in Czech, Russian and English, Ann Arbor: University of Hichigan.
Lakoff, R, 1969, Tense and its relation to participacts. Language 46 838-49.
Leach, G. 1. 1971 meaning and the Engf ish verb. London: Lonqman.
Lyons, J. 1977. Semantics 2. Londcn: Cambridge Prass.
Bacaulay, B. K. S. 1978. Review cf Coaris B, Aspect, and Friedrich P. On aspect theory and Homeric aspect, Language 54 416-420. accawley, J. D. 1971. Tense and time refarence in English. I3 Fillmore and Langandon (eds), Studies in linguistic semantics. florissey, M, D, 1973. The English perfective and still/a nymore. Journal of Linguistics 9 65-b9. Mufwene, S, 1978. A recon.siderati.cn of Linyaia temporal inf iections. Studies i~ African Li~quistics 9 91-105.
Palmer, F. fi. 1974. The English verb. London: Lonqman.
Beichenbach, H, 1947. Elements of symbclic logic. New Ycrk: Free Press.
Schef fer, J. 1975. The progressive in ~nylish,Amsterdam: Ncrth Holland. smith, C. 1976. Present curiosities. Chicayc tinquistic Scciety 12 supplement 568-58 1. ------1978. The syntax and hterpretati.cn of te~poral ex~ressions in English. Linguistics and Philosophy 2 43-99.
Taylor, B. 1977. Tense and continuity. ~inguisticsand Philosophy 1 199-220.
Trauqott, E. C. 1975. Spatial expressions of tense and temporal sequency: a contribution to the study cf semantic fields, Seniotica 15 207-230.
Traugott, E, C. and Raterhouse, J. 1969. 'Areadya and *yetw: a sup~letiveset of aspect markers? Journal of Linguistics 5 287-304.
Vendler, 2. 1967. Verbs and times, In Vendler, Linguistics in philosophy. Ithaca: Cornell University Press.
Vsrkuyl, H. J. 1972. On the compositional nature of aspects. Dordrecht: Reidel.
Yald, B. 1976, comparative notes cn past tenses in Kenyan Northeast Bantu Languages, Studies ir African linguistics sup~leaent6 267-2E1.
Yestphal, E. 1950. The stative conjugation in Zulu, Sotho and Venda. African Studies 9 125-137.
Woisetschlaeger, E. F. 1976. A semantic thecry of the Enqlish auxiliary system. Eloomington: Indiana University Linguistics Club.