i
TABLE OF CONTENTS
Acknowledgements iv
Dedication v
Introduction vi
CHAPTER 1 KINANDI SYNTAX AND TRANSFORMATIONAL GRAMMAR
1.1 Introduction 1
1.2 Theoretical Framework 2
1.3 Kinandi Syntax 5
CHAPTER 2 MORPHOLOGICAL ANALYSIS OF KINANDI
2.0 Introduction 10
2.1 The Verb 10
2.2.1 Tense 11
2.2.2 Aspect 12
2.2.3 Habitual 13
2.2.4 The Infinitive 14
2.2.5 Mood 14
2.3 Transitivity 15
2.4 Reflexivization 16
2.5 The Copula 17
2.6 The Noun 18
2.6.1 Number 18
2.6.2 Possession 18
2.6.3 Demonstratives 20
2.6.4 Gender 21
2.7 Verbal Nouns 22
2.8 Pronouns 22
2.9 Adjectives 23
ii
CHAPTER 3 SENTENCE PATTERNS IN KINANDI
3.1.0 Introduction 25
3.1.1 Elementary Phrase Structure Rules of Kinandi 27
3.1.2.1 Introduction 27
3.1.2.2 The Sentence 28
3.1.2.3 The Predicate 28
CHAPTER 4 SYNTACTIC TRANSFORMATIONS 31
4.1 Introduction 31
4.2 Transformations in English 31
4.2.1 WH-MOVEMENT 31
4.2.2 NP-MOVEMENT 34
4.3 Transformations in Kinandi 35
4.3.1 NP-MOVEMENT 35
4.3.2 TOPICALISATION 42
4.3.3 WH-MOVEMENT 47
4.3.4 Further transformations 51
CHAPTER 5 CONSTRAINTS VERSUS TRANSFORMATIONS IN KINANDI 55
5.1.0 Introduction 55
5.1.1 NP MOVEMENT AND THE RELATIVE CLAUSE 56
5.1.2 OTHER MOVEMENTS AND THE RELATIVE CLAUSE 58
5.1.3 MOVEMENTS AND THE NOUN COMPLEMENT CLAUSE 60
5.1.4 COORDINATE STRUCTURE AND UNIT MOVEMENT CONSTRAINTS 61
5.1.5 WH-ISLAND CONSTRAINT 66
iii
5.1.5 MOVEMENTS AND SUBJACENCY IN KINANDI 68
5.1.7 SUBJACENCY AND RAISING 69
CHAPTER 6 SUMMARY AND CONCLUSION 72
6.0 Summary 72
6.1 Conclusion 73
References 73
iv
ACKNOWLEDGEMENTS
This book is a result of research to which many people contributed to its success both morally
and in kind. It is not possible to name all of them. To these unnamed „angels‟, I say Kongoi
mising „Thank you very much‟.
Special thanks go to my supervisors: Professors Karega Mutahi and Bureng Nyombe as well as
Dr. Judith Nyagah for all the helpful hints on analysis, reanalysis, rephrasing and the
reorganization of this study.
I also thank Professor Lucia Ndong‟a Omondi for having kindled in me an interest in syntax in
particular and African linguistics in general.
My colleagues, the „„Nairobiniks‟‟, Alemu Assefa, Khamete Magwaga, Charles Mboga,
Petronila Mebo, Emily Ogutu, Florence Owili and Mary Okebe deserve mention also for their
fruitful company, discussions and encouragement throughout.
However, I take responsibility for all the mistakes and errors that remain in this book.
v
DEDICATION
TO:
My father, the late Kipkurgatt arwaap Kimining‟,
My mother, Jepchoge Tapsiargaak,
And the entire family of Kiptororguut arwaap Simbolion, who inspired it
vi
Introduction
This book is a result of a study that tested the extent to which the transformational generative
framework can handle data and particularly the transformational processes in the Kinandi (a
dialect of the larger Kaleenjiin language). In particular, it basically describes a number of
syntactic transformations found in the language. This is then explained in relationship to the
constraints and conditions on syntactic transformations suggested by Chomsky (1971), Ross
(1967) and Schwartz (1972).
The resultant analyses revealed that there are various transformational rules that apply to the
language, namely; NP-MOVEMENT, WH-MOVEMENENT, RAISING, TOPICALISATION,
PP-PREPOSING and ADVERB PREPOSING. Further investigations which related the
transformational processes to the various constraints proposed by Chomsky, Ross and Schwartz,
found, for example, that none of the transformation processes could extract a constituent out of a
relative clause. Moreover, it was also found that the language is subject to the COMPLEX
NOUN PHRASE CONDITION which forbids extraction of constituents from adnominal clauses.
Further, data analyzed also showed that Kinandi is subject to the SUBJACENCY CONDITION
and that any attempt at applying transformations that are not cyclical, will lead to the formation
of ungrammatical structures.
It can be concluded that the framework that is used has both descriptive and explanatory
adequacy to handle data from the Nandi language.
1
CHAPTER 1
KINANDI SYNTAX AND TRANSFORMATIONAL GRAMMAR
1.0 Introduction
Kinandi („kutitab Nandi‟ (= the „tongue‟ of the Nandi)) is a dialect of the Kaleenjiin language
group who live on the highlands west of the Great Rift Valley of Kenya. They mainly inhabit the
counties of Nandi, Uasin Gishu and Trans Nzoia. The latest national census puts the population
of the Kaleenjiin as 4,967, 328. Though there are no statistics on the Nandi community itself, it
is the second most populous of the Kalenjin communities and therefore could safely be said to
currently number over 1.5 million.
Greenberg (1963) classifies Kaleenjiin as a Southern Nilotic language in the Eastern Section of
the Nilotic Branch of the larger Eastern Sudanic Language Family, which is composed of the
Western Nilotics (Luo, Langó, Acholi, Shilluk, Dinka, etc) and the Eastern Nilotics (Maasai,
Bari, Kakwa, Karamajong, Teso etc). The Nandi is part of this Kaleenjiin group, form the branch
of what is termed Highland Nilotes which comprise of the Pokot, Tugen, Keiyo, Marakwet,
Sabaot, Kipsigiis, Terik and Tatooga (spoken by a small group in northern Tanzania).
The original name of the Nandi is claimed to be „Il tieng’wal’ (in the Maasai language) meaning
„cattle raiders‟. The name, Nandi was used by the Swahili traders who had come in contact with
the group as a reference to what they perceived to be the long necks (like a cormorant‟s) of these
people and also to allude to what was also derogatorily considered as their voraciousness. The
word Kaleenjiin, on the other hand, was a political slogan coined in the 1940s. It means „I say‟ or
„I tell you‟ (since you can understand my language). It has since become a name to identify what
were originally called „Nandi-speaking people‟.
The dialects spoken by the Kaleenjiin continuum are basically the same except for minor
differences in pronunciation and vocabulary. The speakers can however, be mutually intelligible
when they want to. Some groups have higher mutual intelligibility. For example, Kipsigiis,
Nandi, Terik and Keiyo speakers understand most of the time compared to the Marakwet,
Sabaot, Tugen speakers who are likely to understand much of what the former are saying but not
2
vice versa. However, analysis of the syntax of Kinandi can be used as a basic reference to all the
Kaleenjiin dialects.
I wish to note here that I have used the Kiswahili prefix Ki- to refer to the language of the Nandi
people (thus Kinandi) as opposed to the people themselves (Nandi) because, in my view, it is
easier and less cumbersome to keep repeatedly using than the more anglicized term „Nandi
language‟. Moreover, it is more acceptable (in Kenya, at any rate) to now talk of the languages of
the various communities of Kenya in this way (e.g. We say that Luhya people speak Ki-luhya,
Kamba speak, Ki-kamba, Meru speak Ki-Meru, and Taita, Ki-taita, etc).
This book aims to describe some syntactic facts of Kinandi within a generative transformational
framework, that is, the Extended Standard Theory. Previous studies and literature of Kinandi was
done mainly by missionaries and anthropologists who mainly attempted to write grammars
thought to be “sufficiently full and detailed for anyone to learn the (Nandi) language adequately”
(Hollis 1969: xx). These grammars were written without any modern theoretical framework.
This book, guided by an explicit theoretical framework, gives us a richer perspective of Kinandi
than hitherto. This not only fills a gap in Kinandi language studies, but also becomes a
contribution to the study of African languages and General Linguistics.
1.2 Theoretical Framework
As mentioned in the previous section, the Extended Standard Theory underpins the analysis
of Kinandi syntax in this book. The theory is a further development of the Standard Theory
(„Aspects‟ model of 1965) so that the new theory recognized the function of the Semantic
Component. Before this, Chomsky (1957) had not taken into consideration certain semantic
abilities of the ideal speaker/hearer, such as:
a) detection of semantic ambiguity
b) detection of anomalous sentences.
c) detection of paraphrases
The „Aspects‟ model (1965), otherwise known as the Standard Theory, included a Semantic
Component, which together with the Phonological Component, were purely interpretive. For
example, Chomsky had asserted that semantic rules of the Semantic Component only applied
in the Deep Structure. This was subsequent to Katz and Postal‟s hypothesis (Katz and Postal
3
1963) that grammatical transformations should be structure-preserving. It was found that
some issues arose that that Standard theory could not handle, namely;
a) certain transformations reversed logical predicates
b) issue of analysis of anaphora
c) issues of presupposition and focus
These issues led to the extension of the Semantic Component to cover the Surface Structure. A
system of principles termed Universal Grammar evolved and contains which only allow a certain
number of grammars to be constructed. This grammar takes into consideration variations in
language and postulates a core grammar as well a Theory of Markedness. The theory of core
grammar and markedness is expected to cater for all languages a “a universal core of linguistic
principles which characterizes the full range of … grammatical phenomena found in natural
language” (Radford 1981:29). The core grammar contains unmarked structures whereas the
Theory of Markedness helps to explain marked structures i.e. those which show language
variation.
Thus, the resulting grammar is envisaged to consist of the Base Component which has the phrase
Structure (PS) rules, a lexicon with lexical entries, redundancy rules, lexical insertion rules, etc.
The output of the Base Component consists of Deep Structures. The transformational
Component, on the other hand, comprises of transformation and constraints on them. The output
of the transformations is the S-Structures. Case-marking rules and case filters make up the Case
Component with the output being case-marked S-structures which are fed into the Phonological
and Semantic Components which have their own rules. Phonological rules which consist of
deletion rules and surface structure filters give out filtered surface structures whereas the
semantic component gives out representations of meaning.
In summary, the theory explained above, claims that the relationship between the underlying
(Deep) structure and the surface is brought about by transformational (movement) rules. Since
our interest is the scope of the transformational component in Kinandi, it is fitting to illustrate
how this happens. The following two sentences will form the preliminary examples of how
movement will be analyzed here.
1. Kowo gaa Kiprono
4
„Tns-go home Kiprono‟
(Kiprono went home)
2. Kowo gaa Kiprono
„Tns- go Kiprono home‟
(Kiprono went home)
A transformational analysis (Fig1) will reveal that NP-MOVEMENT has taken place so that
sentence 1 becomes sentence 2 above.
S
INFL VP NP
Tns V NP2
pst V NP N eti
v N
Ko wo Kipronoi gaa
Fig 1 Transformational analysis of NP-MOVEMENT in Kinandi
Our interpretation here therefore is that NP-MOVEMENT has adjoined the subject NP to the
V-node as shown. Hence, we can show the relationship between sentence 1 (which may be
seen as a Deep Structure construction) and sentence 2 through the process of a
transformation (NP-MOVEMENT).
It is noteworthy to also explain here that the transformational component, in adition to
movement rules, has conditions on them, such as;
a) Relative Clauses are Islands, i.e. no movement can extract constituents out of a
relative clause
b) Subjacency Condition says that no constituent can be moved out of more than one
containing NP- or S-node in any single rule application.
5
We shall observe that in Figure 4 below, an object NP has been raised from a subordinate
relative clause to a higher clause. Assuming this is the correct analysis of Kinandi OBJECT
RAISING, then we may conclude that transformations in Kinandi are not subject to the
Relative Clause Island Constraint. If the analysis is correct, we will have, therefore, captured
a syntactic fact in Kinandi regarding the relationship of transformations (movements) to the
relative clauses.
1.3 Kinandi Syntax
Any theory of language has, as its ultimate aim, the description and explanation of the various
aspects of language, namely, phonology, morphology, syntax, semantics, pragmatics, etc.
This book‟s aim is specifically to analyze some syntactic facts pertaining to Kinandi, that is,
transformations. It will however refer to the deletion, semantic and phonological components
when they have been pertinent to the analysis of the syntax of Kinandi. Deletions are referred
to here, basically because they are considered as other types of transformations since they
may be seen as working on the output of transformations (S-structures) to derive surface
structures.
Basically, syntactic analysis aims to capture what has been termed the intrinsic competence
of the ideal (native) speaker/hearer of a language. Indeed, Chomsky points out that;
Syntax is the study of the principles and processes by which sentences are constructed
in particular languages (author‟s emphasis). Syntactic investigation of a given
language has as its goal the construction of a grammar that can be viewed as a device
of some sort for producing the sentences of the language under analysis.
(Chomsky 1957:11)
He reiterates further that „a grammar of a language purports to be a description of the ideal
speaker/hearer‟s intrinsic competence‟ (Chomsky 1963:4).
The task in this book, at a general level, is to account for this intrinsic competence of a
Kinandi speaker. The aim is two-fold;
a) To present some syntactic facts of Kinandi
6
b) To discover the suitability (the descriptive and explanatory power) of the theoretical
model chosen in handling transformations in Kinandi.
In doing this, claims of the Transformational Component of the Extended Standard Theory
will be tested. This component comprises of transformations (i.e. movement rules) like the
NP-MOVEMENT and WH-MOVEMENT including conditions (constraints) on these
movements like the UNIT MOVEMENT CONSTRAINT and SUBJACENCY CONDITION.
An analysis of Kinandi simple (declarative) sentences, embedded clauses and interrogatives
are made as a way of testing these claims. In essence a description of the language is made as
well as an account is given of the above component in view of that fact that the phenomenon
of movement occurs in Kinandi.
For example, Kinandi has what may be termed as NP-MOVEMENT when we analyze the
following sentence:
3. Kabir ineendet Joseph
„Tns- beat s/he Joseph‟
(Joseph beat him)
This sentence is sometimes paraphrased as:
4. Kabir Joseph ineendet
„Tns beat Joseph s/he‟
(Joseph beat him/her)
We can envisage an NP-MOVEMENT where the subject NP (Joseph) is adjoined to the V-
node as exemplified in the following phrase-marker (Fig 2).
7
S
INFL VP NP1
Tns *V NP2 eti
Pst NP
Pron
V N
Ka bir Josephi ineendet
Fig 2 Phrase Marker showing NP-MOVEMENT
We can also show a WH-MOVEMENT rule in the following sentence:
5. Ne kabir ng’oo inee?
„That tns-beat who?‟
(Who beat him/her?)
This can be analyzed in the deep structure as follows (Fig 3).
S1
COMP S
NP COMP INFL VP NP
Pron +wh tns V NP eti
Pst Pron
Ng‟o ne ka bir inee
Fig 3 Phrase Marker showing WH-MOVEMENT
8
Note that in this case, the object NP position has a pronoun which is dropped at the surface
structure. The pronoun ine (s/he) and all other personal pronouns get dropped in Knandi. This
will be further explained in Chapter 3.
Another example of NP- MOVEMENT may be seen as having taken place in sentences with
embedded clauses as in the following example.
6. Ka-a-nyooru chiito ne kobir lakweet
„Tns –I- found person who tns-beat child‟
(I found the person who the child beat)
As a first attempt, we can reckon that the object NP, chiito seems to have been moved from the
lower (embedded) clause to the higher clause as exemplified in the following simplified phrase
marker (Fig 4). Of course, this is just for illustration purposes but as we can see here, OBJECT
RAISING seems to have violated the RELATIVE CLAUSE COSTRAINT that forbids that a
constituent cannot be extracted out of a relative clause.
Thus, using knowledge and theoretical assumptions of the transformational component, we can
discover not only the nature of these movements but also the extent to which they accord with
the constraints imposed on transformations.
9
S1
COMP S
NP1
-WH Pron
ane
INFL VP
Agr V
Ne eti N V
Pro nyooru
tns NP2
Past N S1
Ka a chiito COMP S
-WH NP3
Ne N1
lakweet
INFL VP
Tns V NP4
pst bir etj
ko
Fig 4 Another possible example of NP- MOVEMENT in Kinandi
10
CHAPTER 2
MORPHOLOGICAL ANALYSIS OF KINANDI
2.1 Introduction
This chapter deals with a brief outline of the morphology of grammatical categories of
Kinandi. This is the kind of information that is normally found in traditional grammars with
the major categories introduced being the Verb, the Noun and Adjectives and their
inflectional categories. Included here is also a brief outline of sentence patterns and
elementary phrase structure rules.
2.2 The Verb
The base-form of the verb has three phonemes of the CVC (closed syllable) structure as in the
examples below.
gloss
1 a) sus „bite‟
b) ker „close‟
c) mach „want‟
d) bir „beat‟
There are also a restricted number of verbs whose base-form is of the VC or CVV structure
(open syllable structure) as in the following:
gloss
2. a) uch „tie‟
b) am „eat‟
c) tui „break‟ (into pieces)
d) mui „burry‟
11
All these verbs can be reduplicated to show that the action is being done many times as in the
examples below
3. a) susasus „bite many times‟
b) amaam „eat many times‟
c) birabir „beat repeatedly or many times‟
d) tuiatui „break into pieces‟
e) uchauch „tie many times‟
Furthermore, the verbs can be inflected by prefixes and suffixes to indicate number, tense,
person, and for the infinitive while aspect and mood are indicated by intonation and tone.
2.2.1 Tense
Kinandi has four morphemes which reflect tense. These are distant past ki-, immediate
past ko-, early today ka- and present –i as the following examples show.
4. a) King’eet
„S/he rose‟ (distant past)
b) Ko-ng’eet
„She rose‟ (immediate past)
c) Ka-ng’eet
„S/he rose‟ (early today)
d) Ng’eet-ei
„S/he rises‟ (present)
In addition, there are ten adverbials which act as tense markers. These are:
12
5. a) nguuno „now‟
b) nguuni „just a moment ago‟
c) atkai (olegai) „early today‟
d) amut/(olegonye) „yesterday‟
e) aiin „the day before yesterday‟
f) aiindok „sometime in the past/recently‟
g) koi „later today‟
h) mutaai „tomorrow‟
i) tuun ko aiin „the day after tomorrow‟
j) tuun „sometime in the future‟
These are usually added to the verb already inflected by tense to specify exactness. The
sentences below exemplify this.
6. (a) Weendii nguuno „S/he goes now‟
(b) Ka-weendii nguuni „S/he has just gone‟
(c) Ki-wo aiin „S/he went the day before yesterday‟
(d) Ko-wo amut „S/he went yesterday‟
(e) Tun weendii „S/he will go sometimes in the future‟
(f) Ko-wo atkai (olegai) „S/he went early today/not so long ago‟
Note that in olegai and olegonye in examples 5c, d and 6f, the voiceless velar plosive
/k/ undergoes a phonological process which changes it to a voiced velar fricative in an
intervocalic environment (k ɣ /v-v)
2.2.2 Aspect
Aspect in Kinandi is signaled by infixes. For instance, the present perfect (+perfect, +present) is
signaled by –a- for first person, –i- for second person, and –ko- for third person for singulars.
13
The plurals are –ke-, -o- and –ko- respectively (the vowel infixes are portmanteau morphemes
which indicate pronominal subject clitics – see chapter 4 below). Thus
7 a) Ka-a-siir oreet „I have passed the path‟
b) Ka-i-siir oreet „You have passed the path‟
c) Ka-go-siir oreet „S/he has passed the path‟
The plurals are:
8 a) Ka-ge-siir oreet „We have passed the path‟
b) Ka-o-siir oeret „You (plural) have passed the path‟
c) Ka-go-siir oreet „They have passed the path‟
The past perfective aspect (+past, +perfect) is indicated by an infix –ga- (which is really a
voiceless velar plosive / k / that, as explained in 2.2.1 above, has undergone change in an
intervocalic environment (e.g. k ɣ /v-v)). Examples are given thus;
9 a) Ka-ga-sir kitabuut „S/he had written a book‟
b) Ka-ga-inyooru lakweet „You had found the child‟
c) Ka-ga-gesich lakweet „The child had been born‟
The imperfective (-perfect) aspect is realized with an addition of a suffix –ei, -e or –ii on the
base-form as in;
10. a) Sir-ei kitabuut „S/he is writing a book‟
b) A-weend-ii gaa „I am going home‟
c) A-lu-e cheego „I am drinking milk‟
2.2.3 Habitual
The habitual in Kinandi is a free morpheme „cham’ which indicates the notion of habit and
occurs only in the present and past tenses in the perfective aspect as the following shows.
11 a) Cham koweendii logeet „He usually goes hunting‟
14
b) Ki cham aweendii gaa ‟I used to go home‟
In the imperfective aspect, the word emphasizes that the action of the object (not the subject) is
being done as in the examples below.
12. a) Cham kogeerei koweendii gaa Jeptanuui „S/he usually sees Jeptanui going home‟
b) Ko cham kekweriei tuuga koba ng’eeny „We usually drive the cows to the salt-lick‟
It is also worth noting that the word cham is differentiated from the same word which stands for
the verb „like‟ by tone, that is, cham (like) has a low tone whereas cham (for the habitual) has a
high tone.
2.2.4 The Infinitive
In Kinnandi, the infinitive is introduced by the morpheme ke (to) as in the following examples:
13 a) Ke-manda „To go out‟
b) Ke-sach „to split‟
c) Ke-tang’any ‘to stare in wonder‟
Unlike in English, the infinitive cannot be used with another verb after other verbs as in the
following examples;
13 a) Amache awo (not Amache ke-awo) „I want I go‟ (=I want to go)
b) Machei kosiris (not Machei ke-kosiris) „They want they write‟ (= They want to write)
2.2.5 Mood
Three moods are found in Kinandi, namely; indicative, imperative and subjunctive. Verbs in
Kinandi are realized with a high tone for the indicative mood as in the sentences below:
15 a) Íbú cihito kweenik
„Bring-tns somebody firewood‟
(Somebody is bringing firewood)
15
b) Chóbei amitwoogik
„Prepare -tns food‟
(S/he is preparing food)
The imperative mood on the other hand is realized with a mid-tone on the verb as below
16 a) Ibû chiito kweenik
„Bring person firewood‟
(Somebody, bring firewood!)
b) Chôb amitwoogik
„Prepare food‟
(You prepare food)
The subjunctive mood is realized with a high tone like the indicative mood but the verb is
prefixed with the morpheme Ki- as in the following
17 a) Ki-machei iwe suguul mutaai
„Wanted-aspect- you go school tomorrow‟
(= It is required that you go to school tomorrow)
2.3 Transitivity
Verbs in Kinandi may be subdivided into two sub-categoris namely; transitive and intransitive.
Some examples of transitive verbs are;
17 a) Ke-sich
„To bear or give birth‟
b) Ke-bar
„To kill‟
c) Ke-chot
16
„To mock or jeer‟
d) Ke-bir
„To beat‟
The above take direct objects as in the following sentences;
18. a) Ka-sich lakweet
„(She) bore (a) child‟
b) Ko-bar ng’etundo
„(He) killed (a) lion‟
e) Ki-chot chumbiindet
„(He) mocked (a) stranger‟
f) Ko-bir chorueet
„(S/he beat (a) friend‟
Intransitive verbs of course do not take direct objects as in the following sentences;
19 a) A-weendii
„I (am) going‟
b) Ru-ei
„(He) is sleeping‟
c) Ko-choot
„S/he died of old age‟ (Literally s/he or it melted)
2.4 Reflexivization
The reflexive in Kinandi is represented by the morpheme –kei. Though it may be said to be
potentially suffixed to any verb, its appearance in the language is actually limited by pragmatic
(contextual) grounds. Thus, it is grammatically appropriate to say:
20. a) Ka-toor-kei
17
„Tns-pierce-self‟
(S/he pierced himself/herself)
But not:
21. a) *Ka-sichkei
„Tns bear-self‟
(He bore [gave birth to] himself/herself)
b) *Ka-kutungchikei
„Tns kneel himself‟‟ (literally He knelt for himself)
2.5 The Copula
The copula (verb be) functions more or less like a verb in Kinandi. However, its realization (ko)
is restricted to a very narrow paradigm. But it most cases, it is not usually uttered at all. The
following is an example of a sentence which can be realized with or without the copula;
22. a) Ka-a-roo u notok
„Tns- I- see- like that‟
(= I see that it is like that)
This may be realized with a copula ko as:
23 b) Ka-a-roo ko u notok
„Tns- I- see-is- like that‟
(= I see that it is like that)
Otherwise, it can be said that Kinandi functions normally without the realization of the
copula as exemplified by the following additional sentences;
24 a) A miee
„I good‟
(= I am good)
b) I lakweet-aab chii
„You-child-of somebody‟
(= You are a child of somebody)
c) „A arwaap Kipkurgat’
18
„I son-of Kipkurgat‟
(I am the son of Kipkurgat)
d) Icheek kipsomaniinik
„They scholars‟
(= They are scholars)
2.6 The Noun
All nouns in Kinandi can generally be inflected. The inflectional categories are number,
possession, demonstrative and ( historically but now limited to only proper nouns), gender.
2.6.1 Number
Kinandi nouns have been said to display what has been termed as T-K characteristics (Tucker
1958) where the singular has some kind of „t-morpheme‟ which becomes a kind of „k-
morpheme‟ in the plural. This is generally true as the following examples show;
Singular gloss Plural gloss
25 a) Seeseet dog seeseenik dogs
b) Borieet battle borioosiek battles
c) Lugeet war lugoosiek wars
d) Kurgeet door kurgoosiek doors
In addition, most common nouns have indefinite counterparts. The examples below are
the indefinite counterparts of 25 above.
Singular Plural
26 a) seese seeseen
b) boorio boorioon or boorioot
c) lugo lugoon or lugoosio
d) kurga kurgoon or kurgoosia
These two renditions may be said to be analogous to „the‟ and „a‟ particles in English
which signal the value definiteness (+ definite) respectively.
2.6.2 Possession
19
Kinandi nouns can be inflected to indicate possession, that is, to show ownership or connection.
The morpheme which reflects possession may be suffixed to all nouns (including verbal nouns).
The following renditions of the possessive forms of teta „cow‟ shows how this happens.
27 a) teennyu ( teta + nyu)
„cow my‟ (= my cow)
b) teennyo (teta+ nyo)
„cow our‟ (= his/her cow)
c) teennyi (teta +nyi)
„cow his/hers‟
d) teennywa (teta + nywa)
„cow their‟ (= their cow)
e) teenng’ung (teta +ng’ung’ )
„cow your‟ (= your (singular) cow)
f) teenng’wong’ (teta + ng’wong’)
„cow your (pl) (= your (plural) cow)
Notice also that there is a historical process of vowel deletions in certain environments.
For example, in the singular possessives (27), we have a situation where the vowel /a/ in
teta is deleted. Thus, a →Ø. At the same time, the /t/ becomes a nasal in an environment
of a nasal, that is, t→n/- (nasal).
Possessive morphemes also agree in number with the nouns that inflect as exemplified
below (28).
28 a) tuuk-chuuk ( tuuga + chuuk)
„cows my‟
b) tuuk-kwook (tuuga + kwook)
„cows your(pl)‟
Again as in the singulars above, the process of deletion takes place so that; a→Ø, then,
ɣ→k/-[-voice].
20
2.6.3 Demonstratives
Demonstratives in Kinandi reflect three temporal and spatial dimensions. In the examples given
before for the word lakwet ‘child‟ (lagok is plural), note that 29(a) to (c) are distance
demonstratives while (d) to (f) are for time.
29 a (i) Lakwani (= laktweet + ni)
„Child + this‟ (= this child)
(ii) Lagoochu (lagook + chu)
„children these‟ (+ these chidren)
b (i) Lakwaanaa (= laktwet + noo)
„Child + that‟ (= that child)
ii) Lagoochoo (= lagok + choo)
„children those‟ (= those children)
c) (i) Lakwaaniin (= Lakwet + niin)
„child that (over there)‟ (= that child over there)
(ii) Lagoochuun (lagook + chuun)
„children those (over there)‟ (+ those children over there)
d) (i) Lakweetkai (lakweet + kai)
„child (whom we talked about today)‟
(ii) Lagoogai (Lagook + kai)
„children (whom we talked about today)‟
e) (i) Lakwet konye ( lakwet + konye)
„child (of yesterday)‟
(ii) Lagoogoche (lagook + kooche)
„children (of yesterday)‟
f) (i) Lakweetkiinye (lakweet + kiinye)
„child (of sometime in the past)‟
21
(ii) Lagoogiiche (lagook + kiiche)
„children + (of sometime in the past)‟
Specifiers may also be said to be some kid of demonstratives and are a closed class of
free morphemes as shown below.
30 a) i) kiibo ii) tuupo
„some (singular)‟ „some (plural)‟
b) i) age ii) alak
„other/another „ „others‟
c) (i) kitigiin ii) tutigiin
„small/little‟ „few /little (plural)‟
In Kinandi, example 30 (a) (i), (b) (i), (c) (i) modify countable nouns such as lakwa „child‟ (=
kiibo lakwa „some child‟), chiito „person‟ (= chiito age „another person‟) and boisieet ‘work‟ (=
boisieet kitigiin „a little piece of work), while 30 (a) (i) only modifies uncountable nouns such
as, bai „flour‟ (= tuupo bai), bei „water‟(= tuupo bei), ingui „vegetables‟ (= tuupo ingui) . Note
also that these latter nouns are indefinite [-definite] nouns which have definite [+definite]
counterparts, namely beek, bek and ingweek respectively. These can only be modified by only
the plural specifiers above. On the other hand, 30 (b)(ii), and (c)(ii) modify plural nouns such as
biik alak („people others‟ = other people), and kitabuusiek tutigiin ( „books few‟ = a few books).
2.6.4 Gender
Historically, Kinandi nouns were modified by gender which was signaled by the prefixes kip and
cheep. These gender prefixes functioned to show „opposites‟ so that „things‟ that were small (or
occupying a small space) would have cheep prefixed to it (for example, cheepkoigocheet
„tortoise‟, cheeptiirgichet „antelope‟ while big (or occupying a large space) would have kip
prefix (e.g.Kiptaaiyaat, leader, Kipsengwet „heavens‟. However, this is now indistinguishable
except in proper nouns (names) where male names start with Kip (Kiptororguut, Kipsaang’,
Kipruguut, etc.) while female names start with cheep (normally spelled with J (e.g. Jepletting,
Jepkorir, Jebiwoott, etc).
22
2.7 Verbal Nouns
Verbal nouns in Kinandi are derived from base forms of verbs and most of them are in pairs as
shown in 35 and 36 below. Like other verbs, they can also be classified as definite (+ definite)
(35) or indefinite (-definite) (36).
35 a) (i) ameet (ii) amisieet
„eating‟ „eating‟
b.) (i) taageet (ii) taagisieet
„welcome‟ „welcome‟
c.) (i) mageet (iii) magisieet
„want‟ „want‟
36 a) (i) amo (ii) amisio
„an eating‟ „an eating‟
b.) (i) taago (ii) taagisio
„a welcome‟ „a welcome‟
c.) (i) mago (ii) magisio
„a want‟ „a want‟
2.8 Pronouns
Kinandi personal pronouns are anee „I‟, inyee „you‟, inee „him/her/it‟ in the singular and acheek
„we‟, okweek „you‟ and icheek „they‟ in the plural. Reflexivisation usually suffixes the reflexive
morpheme- geei „self‟.
23
However the interesting thing is this suffixation does not create reflexive pronouns but adverbial
phrases e.g. ineegeei means „by himself/ herself/ itself‟.
Possessive pronouns are also found in Kinandi. They are usually suffixed to a noun or a wh-word
ne (which, who) as shown in 37.
37 a) (i) teennyu (=teeta+nyuu) (ii) teeta ne nyuu
„cow-my‟ „cow which my‟
(my cow) (the cow which is mine)
b) (i) Goonng ‘uung’ (=Goot +nguung’) (ii) Goot ne ng ‘uung’
„house your‟ „house which your‟
(your house) (the house which is yours)
Possessive pronouns also agree in number with the nouns they modify as shown below;
38 a) Tuukyuk (= tuuga+chuuk) (ii) tuuga che chuuk
„cows my‟ „cows which my‟
(my cows) (the cows which are mine)
(b) Goriikuuk (= goriik+guuk) (ii) goriik che guuk
„houses your ‟ „houses which your‟
(your houses) (the houses which are yours)
(NB: compare with 27 above)
2.9 Adjectives
In my view, it is not true, as Hollis claims (Hollis 1919: 180) that there are no “true” adjectives
in Kinandi (whatever he means). Perhaps his observation was influenced by his comparison of
the position of adjectives in an English sentence as opposed to Kinandi. Adjectives normally
follow the nouns they modify in Kinandi. For instance;
24
39 (a) Ka a roo ndaaret ko kooi
„I saw snake is long‟
(I saw a long snake)
(b) Ke al ngoriet ne tuui
„You bought cloth which black‟
(You bought a black cloth)
Sometimes however we find adjectives preceding the nouns they qualify as in 40 below;
40 (a) Leel sumei chichoo
„White hair man-that‟
(That is a white haired man)
(b) Koroom kirgoo
„Tough bull-that‟
(That is a tough bull)
(c) Mariich oreetaab kipsengwet
„Narrow road-of heaven‟
(Heaven‟s road is narrow)
25
CHAPTER 3
SENTENCE PATTERNS IN KINANDI
3.1.0 Introduction
Kinandi is basically regarded as a VSO language i.e. the verb phrase precedes the NPs. As
exemplified by declarative sentence (41) below.
41 Kabir Kipsaang’ lakweet
„Tns-beat Kipsaang‟ child‟
(Kipsaang‟ beat the child)
However, in the deep structure, the pattern is VOS (Verb, Object, and Subject) as shown
below (Fig 1).
S
INFL VP NP1
Tns v NP2 N
pst
Ka bir lakweet Kipsaang‟
Fig 1 VOS Pattern at Deep Structure
In chapter 4, we will see how NP-Movement adjoins the subject NP (Kipsang) to the verb.
Furthermore, the process of pronominal clitic placement (a morpheme derived in the
agreement node and usually transferred (cliticised) on the verb), may create a surface
structure SVO pattern as shown in 42 and Fig. 2 below. Note that the morpheme stands for
subject or object NPs when they (NPs) are dropped. (For more information refer to Chapter 4
Section 4.3.1)
42 (a) Ka-a-bir lakweet anee
26
„Tns-I-beat child I‟
(I beat the child)
S
INFL VP NP1
Tns Agr V NP2 Pron
Pst pro v N
Ka -a- bir lakweet anee
Fig 2 SVO Pattern of Sentence in KInandi
The Nandi language allows for the dropping of the subject pronoun (ane) in accordance with
the „AVOID THE PRONOUN‟ PRINCIPLE‟ which has been noted as allowing the NP
associated with (the clitic) to be dropped (Chomsky 1981:276). The clitic (-a-) in the
example above is generated in the AGR node and if it is a pronominal subject, it is adjoined
to the left of the verb but if it is a pronominal object it is adjoined to the right of the verb so
that it is realized in the surface structure as:
42 (b) Ka- bir-o inee
Tns- beat me he
(= He beat me)
The tree diagram below explains this process (Fig 3)
27
S
INFL VP NP1
Tns Agr *V NP2 Pron
Pst Pro v N Pron
Ka o bir e ane inee
Fig 3 SVO Pattern of Sentence with Dropped Subject Pronoun
It is also possible to drop the subject pronoun (ine). This gives us the sentence (43).
43 Ka-bir-o
Tns- beat- me
(= He beat me)
This is still accords with the AVOID THE PRONOUN PRINCIPLE because, though the
clitisisation does not take place in the third person sentences in Kinandi, we can envisage an
„abstract‟ clitic for the subject pronoun allowing its drop.
3.1.1 Elementary Phrase Structure Rules of Kinandi
3.1.2.1 Introduction
This section contains elementary phrase structure rules which generate simple sentences in
Kinandi.
28
3.1.2.2 The Sentence
The initial symbol (S1 is usually rewritten as COMP (complementizer) + S (sentence). The
complementizer may be empty (e) or it may have ne (that or who or which) (44)
44. a) S1 COMP + S
b) COMP + WH (ne)
Thus sentence (45) may be rewritten in bracket notation as shown in (46) below.
45 (Ne) Kaweendii gaa inee
„(That) tns-going home he‟
(He was going home)
46 S1 [COMP –WH (ne) S [Kaweendii gaa inee] S] S1
The sentence (S) may be rewritten as follows;
47 S INFLVP + NP
(INFL +VP +NP)
Using the same example (45) we can envisage (S) in bracket notation below.
48 S [INFL [Ka [weendii] V gaa] VP inee] S
3.1.2.3 The Predicate
The predicate (VP) may be written as;
49 VP NP
VP
cl- V (NP)
29
In bracket notation, we can have the schema below (50).
50 VP [ V [bir ] V NP [ lakweet] NP]VP
For the second alternative (cl-V) we refer to cases where cliticisation has taken place and the
pro-drop parameter has been effected as in sentence (51) below;
51 A-weendii gaa
I-going home
(I am going home)
This may be shown in a bracket notation as in (52) below
52 VP [V*[cl-weendii] V*]
The verb may be categorized as transitive or intransitive. For instance, wendi „go‟ is
intransitive. Hence it can be written as a complex symbol with the feature intransitive (i.e.
not needing an object NP)
53 Weendii: V, [ NP]
There are also certain simple sentences do not have overt VPs. These are the sentences
which have the form given in (54) below
54 Ka Jeruuto
„Tns Jeruuto‟
(It was Jeruuto)
The verb „be‟ ko is not usually realized on the surface but may be seen as present in the deep
structure as indicated below;
30
S
INFL VP NP
tns V N
pst
Ka (ko) Jeruuto
(cf. to 2.5 above)
31
CHAPTER 4
SYNTACTIC TRANSFORMATIONS
4.1 Introduction
In this chapter, we look at various transformations (moves α rules) in Kinandi. To do this, we are
going to analyse a corpus of carefully chosen sentences of Kinandi. It is necessary to have both
deep and surface structure illustrations in order to trace these transformations. We would like to
note that we will be analyzing NP-WH-MOVEMENTS, TOPICALISATION and PREPOSING
RULES (ADVERB-AND PP-PREPOSING) AND RISING.
As introduction we will illustrate various transformations (+movements) in English syntax. We
propose to do this as per the arguments of Radford (1981). In Radford (1981), a superficial
examination of certain sentences reveals apparent ungrammaticality, that is, there appears to be a
violation of certain facts of English grammar (at least at the surface level). These facts include
facts of subcategorisation in the lexicon and agreement facts. Our introductory examples,
however, will only deal with the aspect of subcategorisation in English grammar.
4.2 Transformations in English
4.2.1WH-MOVEMENT
In the English lexicon, certain verbs are subcategorized, that is, given certain sub-categorization
restrictions. For instance, certain verbs are subcategorized as requiring, among other restrictions
(which are not of interest to us here), object nouns. For example, the verbs give and put may be
subcategorized thus;
1. a) give: v, + [-NP]
b) put: v, +[-NP-PP+LOC]
(cf. Radford 1981:150)
However, in the interrogative sentences below, at first glance, seem to show contrary information
from what we would find in the lexicon as explained in 1 above because the verbs give and put
do not seem to have any phrases that are enumerated in 1.
2. (a) Which car will you put ? in the garage
32
(cf. Radford 1981:154)
(b) What will John give ?
As we can see, sentence 2 (a) does not seem to have an object NP before the locative PP (in the
garage). Similarly, the verb give in sentence 2 (b) seems to miss an object NP or PP. And yet
these two sentences are grammatical in English.
In order to analyze the two sentences above, it is necessary to point out that transformational
generative grammar envisages two levels of structure of analysis of sentences- the surface
structure and the deep structure (which we have alluded in sections 3.1.0 above), in which the
deep structure is linked to the surface structure by transformational rules. We therefore need to
look at the deep structure of sentences (2) in order to see which transformations (if any) have
taken place. Radford (1981:155-156) proposes the following (simplified) phrase markers for
analysis of such sentences.
Fig 1 (a)
S1
COMP S
+wh NP aux VP
e pron modal V NP PP
You will put det N P NP
which car in det N
the garage
Fig 1 (a) Deep Structure Phrase Marker for ‘Which car will you put in the garage?’
33
S1
COMP S
+wh NP aux VP
e N modal V NP
John will give Pron
what
Fig 1 (b) Deep Structure Phrase Marker for ‘What will John give?’
The two phrase-markers above show us that the two sentences (2a and 2b) have not violated the
sub-categorization restrictions because they do have the categories to which they had been
subcategorized. What has happened is that some movement rule(s) has or have transformed the
deep structure (1a and 1b) to the surface structure (2a and 2b) instead. This rule is what is
known as the WH-MOVEMENT. What has happened is that the movement has adjoined these
phrases (NPs) to the left of COMP in the manner shown below.
Fig 2 a
S1
COMP S
NPi COMP NP aux VP
det N +wh pron modal V NP PP
which car e you will put eti P NP
in det N
the garage
Fig 2a Deep Structure Phrase Marker showing WH-Movement
34
Fig 2 b
S1
COMP S
NP COMP NP aux VP
Pron +wh N Modal V NP
whati e John will give eti
Fig 2b Deep Structure Phrase Marker showing WH-Movement
In addition to WH-MOVEMENT in the manner stated above, there is also another
transformation called NP-AUX inversion. This transformation is necessary (in English Syntax)
in interrogative sentences. Thus, the NPs (John, you) are inverted with the modal auxiliary (will)
to give us the surface structures 2 (a) and (b) above..
4.2.2 NP MOVEMENT
The examples in the preceding section (4.2.1) can be said to be a type of NP-MOVEMENT, in
the sense that what was moved was a type of NP, in this case a WH-NP. There are instances
where a noun phrase per se is moved as illustrated in the following example from Radford
(1981:188).
3. John seems---------- to like Mary.
Radford suggests, correctly in our view, that the deep structure for this sentence can be
schematized as follows;
35
S1
COMP S2
WH NP VP
e e V S1
seems COMP S4
-wh NP V1
e N P V2
John to like NP
N
Mary
Fig 3 Deep Structure Phrase Marker of ‘John seems to like Mary’
This type of NP-MOVEMENT has also been termed (SUBJECT) RAISING (Radford 1981:188).
In this example, the NP (John) has been „raised‟ from the subordinate clause to become the
surface structure subject of the main (seems) clause.
4.3 Transformations in Kinandi
4.3.1 NP-MOVEMENT
We have seen certain movement rules namely; NP and WH-MOVEMENT as exemplified by
English grammar above. Our task now is to show types of transformations in Kinandi. We start
be showing how the following pair of sentences illustrates the NP-MOVEMENT phenomenon
in Kinandi.
5 a) Kabir lakweet Kipsaang’
„Tns-beat child Kipsaang‟‟
(= Kipsaang‟ beat the child)
36
b) Kabir Kipsaang’ lakweet
„Tns-beat Kipsaang‟ child‟
(= Kipsaang‟ beat the child)
The two sentences above are paraphrases in which the subject Kipsaang’ and the object (lakwet)
have interchanged positions on the surface structure. In order to illustrate this phenomenon as an
NP-MOVEMENT, let us look at the deep structure of this pair of sentences (4).
S
INFL VP NP1
Tns Agr V NP2 N
Pst
Ka e bir lakweet Kipsaang‟
Fig 4 Deep Structure Phrase Marker of ‘Kabir lakweet Kipsaang’’
We propose that in order for 4 (a) to be paraphrased into 4 (b), the subject NP Kipsaang’ is
adjoined to the V-node as shown in the phrase-marker below
S
INFL VP NP1
Tns Agr *V NP2 eti
Pst V NP N
N
Ka e bir Kipsaang‟i lakweet
Fig 5 Deep Structure Phrase Marker showing NP-Movement
37
This type of NP-MOVEMENT is an adjunction movement rather than a substitution movement
as shown in the English example in Fig.3 where a substitution movement moves an NP to an
empty node (we will deal with this subsequently).
One interesting thing we find here is that Kinandi can adjoin a subject NP to a verb. This is
nothing unusual in Kinandi because when Kinandi undergoes cliticisation there is a process in
which the pronominal clitic, which is generated in the agreement (AGR) node, is adjoined to the
V-node.
It is also worth noting here that Kinandi has a rich agreement and is a pro-drop language. Since
agreement generates clitics, these clitics serve as subjects or objects, and hence allows for a (pro)
noun to be dropped. In the following sentence (5), we can envisage a process of pronominal
object clitic placement where the clitic is adjoined to V before the object NP (which is usually a
pronoun) is dropped.
5. Ka-mach-in Kiptoo
„Tns-want-you Kiptoo‟
(= Kiptoo wanted you)
The following phrase-marker can be envisaged as deep as the deep structure of this sentence.
S
INFL VP NP1
Tns Agr V NP2 N
Pst Pro Pron
Ka -in mach inyee Kiptoo
Fig 6. Deep Structure of sentence ‘Ka machin Kiptoo’.
38
The pronominal object clitic is joined to the V-node in the same manner as the NP-movement
(Fig 5 above). The phrase maker below (Fig 7) illustrates this;
S
INFL VP NP1
Tns Agr *V NP2
Pst V N pron N
pro
Ka eti mach -ini inyee Kiptoo
Fig 7 Phrase marker showing clitic placement to the sentence ‘Ka machin Kiptoo’
Thus, with the onset of clitic placement, the object NP (pron) is then dropped so that we have the
surface structure as sentence 5 above.
Another instance of NP-MOVEMENT can be illustrated by the following pair of sentences.
6 a) (ne) miee chiito ni
„(who) good man this‟
(This is a good man)
b) Chiito ne miee ni
„man- who- good- this‟
(This is a good man)
Sentence 6 (b) can be envisaged as a paraphrase of sentence 6 (a). Here, we can say that the NP,
chiito has been fronted by an (NP) adjunction movement to the left of the COMP as shown in Fig
8 below.
39
S1
COMP S2
N COMP INFL VP N1
-wh tns V ADJ N2
Chiito ne prs (ko) miee N Det
eti ni
Fig 8 Phrase marker showing NP movement to the left of COMP
Here, NP-movement here has moved the noun, chiito only. It is also possible to envisage the
movement of the whole N2, chiito ni as shown in Fig 9 below.
S1
COMP S2
Ni COMP INFL VP N1
N Det -wh tns V ADJ N2
Chiito ni ne prs (ko) miee eti
Fig 9 Phrase marker showing an alternative NP movement to the left of COMP.
On the other hand, a substitution (NP) movement can be illustrated by sentence 7 bellow;
7 Uu Jebeet ne ka-wo
„seem Jebeet that tns-go‟
(= It seems that Jebeet went)
This sentence represents an example where the subject NP can be said to have been „raised‟ from
the subordinate clause to the super ordinate clause. The following deep structure shows this;
40
S1
COMP S2
-wh INFL VP NP1
e
ne tns V S1 S2
prs uu COMP INFL VP NP2
-wh tns
Pst V N
ne ka wo Jebeet
Fig 10 Phrase marker showing Subject Raising
It is also possible to envisage an adjunction to the higher V-node as shown in below.
S1
COMP S2 NP1
-wh INFL VP e
ne tns *V S3
V NP COMP S
prs uu N -wh INFL VP NP2
Jebeeti ne tns
pst V
ka wo eti
Fig 11 Phrase marker showing NP adjunction
41
This „raising‟ phenomenon is revisited in the next chapter when we deal with the
SUBJACENCY CONDITION
In addition, sentence 7 can also be interpreted as having an adjunction movement to COMP, that
is, the NP, Jebeet can be adjoined to COMP 2 as shown below
S1
COM P S2 NP1
-wh INFL VP e
Ne tns V S3
Prs uu COMP S
NP COMP INFL VP NP2
N -wh tns V
pst
Jebeeti ne ka wo eti
Fig 12 Phrase marker showing NP adjunction to COMP2
In addition, it is also possible for the same NP to be adjoined to COMP 1 to derive sentence 8
below.
8. Jebeet ne u ne ka wo
„Jebeet who seem who tns-go‟
( = The Jebeet who seems to have gone)
The phrase-marker below illustrates this;
42
S1
COMP S NP1
NP COMP INFL VP e
N -wh tns V S1
Jebeeti ne prs COMP S
u -wh INFL VP NP2
ne tns V
pst
ka wo eti
Fig 13 Phrase marker showing NP adjunction to COMP1.
The next section on TOPICALISATION, in addition to illustrating how focus constituents (NPs)
are fronted, also clarifies the differences between TOPICALISATION, NP-WH-MOVEMENT
and cleft constructions. The sentence above (8) for instance, may also be interpreted as having
undergone TOPICALISATION if we ignored the presence of the complementizer. But the next
section will show just how we can recognize the difference from both structural and
phonological angle.
4.3.2 TOPICALISATION
Topicalisation is the movement of an NP to the front of sentences (beyond COMP position. The
following English sentence (9) has undergone TOPICALISATION, that is, an NP has been made
a focus constituent.
9. This man, I like.
Here, we say that the object NP (this man) has been made a focus constituent, so to speak so that
it becomes a paraphrase of sentence 10 below;
10. I like this man.
43
It is also possible to envisage this movement (Topicalisation) on a phrase- marker as illustrated
below.
S1
TOPIC S2
COMP S
NP aux VP
e pron tns V NP
I prs like Det N
this man
Fig 14 Phrase marker showing topicalisation in English
In Kinandi syntax, it is a bit difficult to show that a sentence has undergone TOPICALISATION
unless we are acquainted with the international pattern of language. Otherwise on the surface
structure one cannot be able to tell (orthographically) whether NP-(WH-) MOVEMENT or
TOPICALISATION has taken place. It is even not easy to know whether the surface structure is
not a cleft construction. Our example 8 in section 4.3.1 is a clear example. One cannot be able to
tell whether this movement to the left most position (as we see orthographically) is an NP-
MOVEMENT or a TOPICALISATION and that of NP or WH-MOVEMENT, a task which we
attempt to do in this section, albeit, informally.
Before we do this, however, let us first look at an obvious case of TOPICALISATION. In a
simple declarative sentence like the following (11), it is possible to show TOPICALISATION.
11 a) Ki-a-yae ng’echeraani
„Tns-I-make chair this‟
(I made this chair)
44
b) Ng’echeraani, ki-a-yae
„Chair-this, tns-I-make‟
(This chair, I made)
TOPICALISATION
The following phrase-marker indicates the position of TOPICALISATION for this sentence 11
(notice also the pronominal subject clitic placement which, as we saw in 4.3.1, allows for
dropping of the subject NP pronoun (NP, (pro) anee).
S1
TOPIC S2 NP
COMP S pron
INFL VP anee
-wh tns Agr *V NP2
ne pst eti N V N Det
ki eti ai yae ng’echereet ni
Fig 15 Phrase marker showing topicalisation in Kinandi
However, on the surface structure it is not the diacritic, (comma) alone that will indicate
TOPICALISATION (even for the English example, 10). For Kinandi it is necessary to utter the
sentence 11 (b) with a high level intonational pattern for the syllables of the focus constituent
‘ngécheraani’ with a pause after the last syllable and followed by a low level intonational pattern
for the syllables of the last part of the sentence (Ki-a-yae). It is also necessary to delete the
COMP element ne (that/which) obligatorily. These are essentially what differentiate
TOPICALISATION from NP- or WH-MOVEMENT immediately to the left of COMP. For the
adjunction rules (NP or WH-) the sentence is uttered with a fairly high level intonational pattern
45
throughout. The sentence and the COMP element can be deleted optionally. In juxtaposition, the
two types of transformations may be informally illustrated as follows;
TOPICALISATION = [----- -----]
WH or NP MOVEMENT = [--------]
(NB: Each dash represents a syllable and its pitch height)
A cleft construction, on the other hand, can be recognized by its intonational pattern which is
different from those of TOPICALISATION and NP/WH-MOVEMENT. This is uttered with a
low level intonational pattern, again, informally presented here as [--------]. The Deep Structure
of a cleft construction is definitely different from those of TOPICALISATION and WH/ NP-
MOVEMENT. This is because, unlike the latter, a cleft construction has a subordinate clause as
illustrated in Fig 16 below.
S
NP VP
N Det V NP
Ng‟echereet ni (ko) pron S1
inee COMP S
-wh INFL VP
ne tns Agr *V NP
pst N V Pron
ki eti pro yae inee
-a-
Fig 16 Phrase marker showing a cleft construction in Kinandi
46
This sentence reads;
12. Ng’echereet ni ko ineEE ne ki-a-yae inee
„Chair this is it that tns-I- make it‟
(It is this chair that I made)
However since Kinandi is a pro-drop language, we can envisage sentence 12 as in 13
13. Ng’echereet ni ko ne kiayae
„Chair this is that tns-I-make‟
(=It is this chair that I made)
Another point we touched on in chapter 2 (2.5 and 2.1.5) is that Kinandi does not use the verb
„be‟ ko as an obligatory element. Hence, sentence 13 can as well be paraphrased as follows;
14 Ng’echereet ni ne ki-a-yae
„Chair this that tns-I make‟
(= It is this chair that I made)
Since this (sentence 14) resembles the surface structures of TOPICALISATION and WH/-NP-
MOVEMENT, it is then necessary to differentiate them by their intonational patterns.
In addition to their difference in intonational patterns, it is also possible to know the difference
between a cleft construction and a sentence that has undergone TOPICALISATION. After
TOPICALISATION, the complementizer, ne (who/ that) must be deleted obligatorily, whereas a
cleft-construction cannot undergo this deletion. This is because in Kinandi the complementizer
of a relative clause cannot be deleted.
47
It is difficult to differentiate between a cleft construction and a clause that has undergone
NP/WH-MOVEMENT because they both have complementizers. But their intonational patterns
will of course guide us. Moreover, it is possible to delete the complementizer for the latter
optionally.
Thus, the following surface structures help to differentiate between TOPICALISATION,
NP/WH-MOVEMENT and a cleft construction.
(a) Ng’echeraani, ki-a-yae
„Chair-this, tns-I-make‟ (= TOPICALISATION)
(This chair, I made)
(b) Ng’echeraani ne ki-a-yae
„Chair this, that tns-I-make‟ (= CLEFT-CONSTRUCTION)
(It is this chair that I made)
(c) Ng’echeraani (ne) ki-a-yae
„Chair-this (that) tns-I-make‟ (= WH/NP- MOVEMENT)
(This chair (that) I made)
We can, therefore, say that, orthographically, it is possible to know which movement has taken
place by the presence of a comma in TOPICALISATION (and a pause in speech) and the
optional deletion of ne (that/who) complementizer for WH- or NP-MOVEMENT. The obligatory
presence of ne complementizer in cleft-constructions is also pertinent. This is of course in
addition to the intonational patterns stated earlier on.
4.3.3 WH-MOVEMENT
WH-MOVEMENT is essentially an adjunction movement immediately in the left of COMP
(Radford 1981:169). Let us analyse the following sentence;
Ng’oo ne achame anee?
„Who that I love I ‟
(= Who do I love?)
48
Here, it is possible to envisage a Kinandi WH-MOVEMENT which adjunctions the WH-
phrase ng’oo to COMP in the manner shown below;
S1
COMP S
NP COMP INFL NP1
Pro +wh tns Agr VP Pron
Ng‟ooj ne prs eti *V NP2 anee
N V etj
Pro chame
ai
Fig 16 WH-MOVEMENT that adjoins WH-phrase in COMP
The subject pronoun can be dropped due to pronominal subject clitic‟s presence. The
example above (Fig 16) is at a clausal level. It also is possible to envisage a WH-phrase
which is moved by WH-MOVEMENT and then deleted in COMP. This happens in both
subordinate and higher clauses. Let us exemplify this by the examples below (16)
16. (a) Ka-a-teeb ngot ke-beendi gaa
„Tns-I-ask whether tns-going home‟
(I asked whether we are going home)
(b) Ka-leenjiin (ne) ka-chamei nee?
„Tns-tell-you (that) tns-love what?‟
(He told you (that) he loves what?)
49
Since the verb teeb „ask‟, is subcategorized in the lexicon as having an object NP, i.e.
teeb:v, +[-NP] then we can envisage the deep structure of sentence 16 (a) as hereunder:
S1
COMP S
-wh INFL VP NP
ne tns Agr *V NP pro
pst eti N V Pro S1 anee
ka pro teeb nee COMP S NP
-wh e
ngot INFL VP
Agr V NP
N
ke beendii gaa
Fig 17 WH-NP movement to higher COMP
What we see here in this deep structure (Fig 17) is that a WH-NP nee „what‟ can be WH-
MOVED to the higher COMP and then deleted by a WH-DELETION rule. The result is the
surface structure 16 (a) above. For sentence 16 (b), it is also possible to envisage a WH-
MOVEMENT to COMP in the subordinate clause before WH-DELETION as exemplified
by Fig 18 below.
50
S1 NP1
COMP S Pron
-wh INFL VP inee
ne tns Agr *V NP3
pst eti V N Pron S1 NPi2
ka leenj iin inyee COMP S Pron
-wh INFL VP inee
ne tns Agr *V NP4
pst etj V N Pron
ka cham pro
etj nee
Fig 18 Phrase Marker Possible WH-NP movement and WH-DELETION in COMP
Thus, the sentence after the transformations (i.e. WH-MOVEMENT and pronominal clitic
placement) looks like this;
17. (Ne) Kaleenjiin inyee inee nee ne chamei inee?
„(That) tns-tell-you -what -that -love- he -he‟
(=He told you that he loves what?)
The pro-drop process takes place to give the following:
18. (Ne) Kaleenjiin nee ne chamei?
(that) tns-tell-you what that loves
51
(= He told you that he loves what?)
In turn, WH-DELETION takes place.
19 Kaleenjiin chamei?
„he- tell-tns- you- love?‟
(=He told you that he loves what?)
Note that only WH-NPs may be deleted in COMP because an NP per se „has semantic
content‟ (Radford 1981:274).
20) Kalenjiin (ne) chamei Kiproono?
(he tell-tns you (that) he love-tns Kiproono?)
NP-MOVEMENT
Kaleenjiin Kiproono ne chamei-?
„tns-tell you Kiproono that love- he-‟
(He told you that he loves Kiproono? ≈
Did he tell you that he loves Kiproono?)
4.3.4 Further Transformations
Other transformations may be related to TOPICALISATION and ADVERB-PREPOSING
and PP (Prepositional phrase) PREPOSING. In Kinandi, it is easy to see that these two types
of transformations have a lot in common with TOPICALISATION in the sense that it is the
focus constituent that is fronted. Whereas TOPICALISATION involves the fronting of an
NP „beyond COMP‟, ADVERB -PREPOSING AND PP-PREPOSING involve the fronting
of ADVERB and PREPOSITIONAL PHRASE respectively, as illustrated in the following
sentences.
52
21 a) Ariibe tuuga raani ak mutaai
„I- look –cattle- today- and- tomorrow.‟
(= I look after the cattle today and tomorrow)
ADVERB PREPOSING
b) Raani, ariibe tuuga
„today -I- look- cattle‟
(= Today, I look after cattle)
It should be noted here that this is really a compound sentence which, in full can be interpreted
as Ariibe tuuga raani ak ariibe tuuga mutaai (I look after cattle today and I look after cattle
tomorrow). For purposes of ease of our analysis here, we ignore the second part of the
coordinate structure (and I look after cows tomorrow)
22 a) Kiriibe tuuga kobooto Kibeet
„We- look-cattle-with-Kibeet‟
(= We look after cattle with Kibeet)
PP- PREPOSING
b) Kobooto Kibeet, keriibe tuuga
„With- Kibeet- we- look- cattle‟
(= With Kibeet, we look after cattle)
The phrase-markers (Fig 19and 20) show the kind of deep structures for sentences 21 and 22.
53
S1
S1
COMP S NP
-wh INFL VP Pron
ne tns Agr * V NP anee
prs eti N V N ADVP
pro riibe tuuga raani
kii
Fig 19 Phrase marker showing ADVERB-PREPOSING
S1
S1
COMP S NP
-wh INFL VP Pron
ne tns Agr *V NP anee
Prs eti N V N PP
kei riibe tuuga P NP
kobooto Kibeet
Fig 20 Phrase marker showing PP-PREPOSING
54
It should be noted here that the COMP element in the two structures is optional but is
obligatorily deleted on the surface structure in order to differentiate the two movements from
TOPICALISATION. In the next chapter, we look at how some of these transformations relate to
the constraints propounded by Ross (1967) and consolidated by Chomsky (1971, 1977).
55
CHAPTER 5
CONSTRAINTS VERSUS TRANSFORMATIONS IN KINANDI
5.1.0 INTRODUCTION
In the preceding chapter, we dealt with movements (transformations). This chapter deals with the
analysis of transformations in Kinandi and their relationship to the constraints proposed by
Chomsky (1977), Ross (1967) and Schwartz (1972). Ross (1967) was the first to account for
constraints of transformations. His arguments were further consolidated by subsequent
scholarship, namely Chomsky (1972, 1977) and Schwartz (1972). These constraints are as
follows:
1. COMPLEX NOUN PHRASE CONSTRAINT (CNPC)
„No element can be extracted out of an adnominal clause‟
2. SENTENTIAL SUBJECT CONSTRAINT (SNSC)
„No element can be moved out of a clause which is the subject of another clause‟
3. WH-ISLAND CONSTRAINT (WHICH)
„No element can be moved out of a clause containing an overt WH-complementizer or
WH-phrase in COMP‟
4. COORDINATE STRUCTURE CONSTRAINT (CSC)
„No element can be moved out of a coordinate structure‟
5. A-OVER-A CONSTRAINT (AO AC)
No constituent of category A can be moved out of a larger containing constituent of
Category A‟
6. UNIT MOVEMENT CONSTRAINT (UMC)
„Only a string of elements which form a constituent can be moved together in any single
application of any movement rule‟
7. SUBJACENCY CONDITION (sc)
„No constituent can move across more than one bounding node in any single rule
application (bounding nodes = S& NP, and perhaps also s-bar‟)
8. COMPLEMENTIZER CONSTRAINT (CC)
„No constituent can be adjoined to a COMP which already contains a WH-constituent‟
56
9. TENSED S CONDITION
„No constituent of a tensed clause (=s-bar) not in COMP can be moved outside that
tensed clause, or construed with any element outside the tensed clause‟
10. SPECIFIED SUBJECT CONDITION (SPSC)
„No non-subject constituent of a clause or NP with specified subject can be moved (or
construed with any constituent) outside the clause or NP, unless it is in COMP‟
(Radford 1981:249)
Constraint 1, 2 and 4 were propounded by Ross while constraint 6 was proposed by Schwartz.
The rest were proposed by Chomsky (i.e. 3, 5, 7, 8, 9, and 10)
In this chapter we will mainly look at how WH-ISLAND CONSTRAINT: COORDINATE
STRUCTURE CONSTRAINT: UNIT MOVEMENT CONSTRAINT: SUBJACENCY
CONDITION and the COMPLEX NOUN PHRASE CONSTRAINT relate to various
movements that apply to Kinandi as given in Chapter 4, namely NP-MOVEMENT,WH
MOVEMENT, TOPICALISATION, ADVERB PROPOSING AND P.P-PROPOSING.
5.1.1 NP MOVEMENT AND THE RELATIVE CLAUSE
One constraint that Ross (1967) propounded was that we may term „RELATIVE CLAUSE
CONSTRAINT.‟ This constraint as the term suggests does not allow for extraction of any
constituent from a relative clause. This condition was adopted and further developed by
Chomsky with a view of generalizing it. Hence the more general constraints like WH-ISLAND
CONSTRAINT AND SUBJACENCY CONDITION.
We however look into this more specific constraint and its relationship to the relative clause and
its relationship to the relative clause and NP-MOVEMENT in Kinandi. We analyze the
following relative clause to see whether the NP-MOVEMENT can extract a constituent out of it.
1. Kirgit-noo ne ka al Kipruuto
„Bull-that that tns-buy Kipruuto‟
(That bull that Kipruuto bought)
57
If the subject NP (Kipruto) is extracted from the clause, we get an ungrammatical sentence as
below;
2. *Kipruto Kirgit noo ne ka al-
„Kipruto bull that tns-buy-‟
(Kipruto that bull that bought-)
In actual fact, the only movements that can take place here are NP and WH- MOVEMENTS. We
can envisage that underlying sentence 1, there is a WH-NP, ngoro (which) is moved to the left
of COMP and subsequently deleted. This is informally shown below as;
3. Kirgit noo S1[COMP ne S[VP[ka al ngoro]VP Kipruuto]s]S1
WH- MOVEMENT
Kirgit noo S1 [COMP ngoro ne S[VP[ka-al-]VP Kipruuto]S]S1
WH-DELETION
Kirgit noo ne ka al Kipruuto (= sentence 1 above).
On the other hand, NP-movement can be envisaged as an adjunction movement, in which
the subject NP, Kipruuto is adjoined to the V-node as shown hereunder.
NP
N Det S1
COMP S NP
Kirgit noo -wh INFL VP eti
ne tns *V NP2
pst V NP Pron
ka al N ngoro
Kipruutoi
Fig. 1 Example of NP-MOVEMENT as an adjunction movement within the clause.
58
From this analysis, we see that both these movements have taken place only within the clause.
5.1.2 OTHER MOVEMENTS AND THE RELATIVE CLAUSE
As we saw in the preceding section NP- movement cannot extract a constituent out of the relative
clause. We use the following example to find out whether WH-MOVEMENT can extract
elements from a relative clause.
4. Ko-me kirgit ne ko al Kipruuto
„Tns-die bull that tns-buy Kipruuto‟
(=The bull that Kipruuto bought died)
Underlying, this sentence is;
5. Kome kirgit ne ko al ngoro Kipruuto
„Tns-die bull that tns-buy which Kipruuto‟
(The bull died that Kipruuto bought which)
WH-MOVEMENT will give us the ungrammatical sentence below
6. *Ngoro kome Kirgit ne ko al Kipruuto
„Which tns-die bull that tns- buy Kipruuto‟
(=Which-WH the bull died that Kipruuto bought-)
TOPICALISATION also cannot extract elements out of a relative clause as exemplified by the
sentence repeated below as (7):
7. (a) Kome kirgit ne ko al ngoro Kipruuto
„The bull died that Kipruto bought which‟
= TOPICALISATION
(b) *Kipruuto, kome kirgit ne ko al ngoro
„Kipruuto, the bull died that bought which‟
Neither can ADVERB PREPOSING;
8. (a) Kome kirgit ne ko al Kipruuto amut
59
„Tns-die bull that tns-buy Kipruuto yesterday?‟‟
(The bull died that Kipruuto bought yesterday)
= Adverb-PREPOSING
(b) *Amut, kome kirgit ne ko-ale Kipruuto
Nor PP-PREPOSING;
9. (a) Kome kirigit ne ko al Kipruuto koboto kwaan
„Tns-die bull that tns-buy Kipruuto with father‟
(The bull died that Kipruuto bought with father)
= PP-PREPOSING
(b) *Koboto kwaan, kome kirgit ne ko al Kipruuto
„With father, died bull that buy-tns Kipruuto‟
(With father, the bull died that Kipruuto bought)
It is possible, however, to envisage the two preposing rules as taking place within the two super
ordinate clauses. Hence the two sentences may be envisaged as shown in 10 and 11 below. (Note
that the literal translation from Kinandi to English is only meant for exemplification purposes
only so that one can understand the constituents being discussed here. Clearly, the renditions in
English in example 11 are either „weird‟ or even ungrammatical, but that is beside the point.).
10. (a) Kome kirgit amut ne ko al Kipruuto
„Tns-die bull yesterday that tns-buy Kipruuto‟
(The bull died yesterday that Kipruuto bought)
= ADVERB PREPOSING
(b) Amut, kome kirgit ne ko al Kipruuto
„Ýesterday, tns die bull that tns-buy Kipruuto‟
(Yesterday, the bull died that Kipruto bought)
11. (a) Kome Kirgit kobooto kwaan ne ko al Kipruuto
„Tns-die bull with father that tns-buy Kipruuto‟
(The bull died with father that Kipruuto bought)
= PP- PREPOSING
60
(b) Kobooto kwaan, kome kirgit ne ko-ale Kipruuto
„With father, tns-bull that tns-buy Kipruuto‟
(With father the bull died that Kipruuto bought)
The point here is that in the latter (10 and 11), no constituent of the relative clause has been
extracted; hence their grammaticality after the two preposing rules. The elements (Adverb and
PP) in these two sentences are, so to speak, members of the super ordinate clause and not the
relative clause like the former (8 and 9).
At this point, therefore, we can generalize that no movement rule can move a constituent out of a
relative clause. Thus, it is possible for us to say that relative clauses are also islands in Kinandi
and that all movements in Kinandi are subject to the relative clause constraint.
5.1.3 MOVEMENTS AND THE NOUN COMPLEMENT CLAUSE
If relative clauses are islands in the Kinandi syntax, then it is necessary to know whether noun
complement clauses are also islands. This will provide us with the opportunity to propose the
more general constraint, that is, the COMPLEX NOUN PHRASE CONSTRAINT. We analyze
the following sentence with a complement clause with a view to having a general constraint on
adnominal clauses.
12. Ayanii lembecheek che kaamwachi Kipsaang’
„Tns-I-confess lie-pl- that tns-I tell Kipsaang‟‟
(I confess (the) lies that I told Kipsaang‟)
NP-MOVEMENT (i.e. to higher COMP) cannot extract an NP from the complement clause
above. Hence the sentence below is ungrammatical;
13. *Kipsaang ne ayanii lembecheek che ko-a-mwachi-
„Kipsaang‟ that tns-i-confess lei-pl that tns-I-tell‟
(Kipsaang‟ I confess the lies that I told-)
Nor can WH-MOVEMENT;
14. *Ng’oo ne a-yanii lembecheek che ko-a-mwachi
„Who tns-I-confess lie –pl that tns-l-tell‟
61
(Who I confess the lies that I told-)
Nor can TOPICALISATION;
15. *Kipsaang’, a-yanii lembecheek che ko-a-mwachi
(Kipsaang‟, I confess the lies that I told -)
Nor – ADVERB PREPOSING;
16. *Amut, a-yanii lembecheek che ko-a-mwachi Kipsaang’
(Yesterday, I confess the lies that I told Kipsaang‟)
Nor can PP-PREPOSING;
17. *Kobooto Jeruuto. a-yanii lembecheek che ko-mwachi Kipsaang’
(With Jeruuto, I confess the lies that I told Kipsaang‟)
Thus, it is plausible to propose that no movement can extract a constituent out of
adnominal clauses. Hence, Kinandi is subject to the COMPLEX NOUN PHRASE
CONSTRAINT.
5.1.4 COORDINATE STRUCTURE AND UNIT MOVEMENT CONSTRAINTS
Ross (1967) also formulated what is termed COORDINATE STRUCTURE CONSTRAINT
which prohibits movement of an element out of a coordinate structure. A coordinate like the
following example in English cannot be moved by any (movement) rule.
18. Did you go to school and where?
Movement of a subpart of this creates an ungrammatical sentence
19. *Where did you go to school and __________?
This constraint also applies to Kinandi. The following sentence shows how
TOPICALISATION creates an ungrammatical sentences (20 b and 21 b)
62
20. (a) A-chame Jeleel ak Jeuruto
(I love Jeleel and Jeruuto)
=TOPICALISATION
(b) *Jeruuto, achame Jeleel ak-
( Jeruuto, I-love Jelel and –)
21. So does WH-MOVEMENT (21 (b))
(a) (Ne) Ka-i-ro Jesaang’ ak ngóo?
„Tns- you-see Jesang and who‟
(You saw Jesang and who?)
=WH-MOVEMENT
(b) *Ng’oo ne ka-i-ro Jesaang’ ak-?
„Who that tns you see Jesang‟ and-‟
(Who did you see Jesaang‟ and -?
However, one interesting fact that came to light in this research is that some Kinandi speakers
feel that these sentences that have undergone WH-MOVEMENT are grammatical. After our
analysis, it is felt that what is really happening is that there is no clear-cut difference between a
coordinate structure and a structure that contains a PP (Prepositional Phrase) in Kinandi. Thus, a
Kinandi speaker will still find this sentence (20) is grammatical in that it is „analogically‟ likened
to a sentence that has undergone PP PREPOSING like the one below (22b)
22. (a) Kaarib tuuga (ane) ak Kipleel
„Tns-I-look cattle (I) with Kipleel‟
(I looked (after) cattle with Kipleel)
=PP PREPOSING=
(b) Ak Kipleel, kaarib tuuga
„with Kipleel tns-i- look cattle‟
(With Kipleel I looked after cattle)
This interesting phenomenon seems to have been extended to TOPICALISATION where the
right most part of the coordinate structure is TOPICALISED (21)
23. ? Ak Jeleel, achame Jeruuto
63
„And Jeleel, I love Jeruuto-‟
The same for WH-MOVEMENT, (24)
24. ? Ak ngóo ne ka-i-ro Jesaang’
„And who that you saw Jesaang‟
(With who (m) did you see Jesaang‟?)
The fact that these are accounted for as marginal sentences in Kinandi syntax is because they not
only violate the coordinate structure constraint but also that, by “borrowing” the syntax of PP-
PREPOSING, it breaks the UNIT MOVEMENT CONSTRAINT. The PP-PREPOSING
PROCESS below is very plausible because it does not violate this condition
25. S[ka VP[aV[riib]V]VP NP[tuuga]NP[anee]PP[P[ak][Kipleel]]PP]S
We can see that PP-PREPOSING does not violate the UNIT MOVEMENT CONSTRAINT
because a CONTINUOUS CONSTITUENT is PREPOSED. We can also move the NP (Kipleel)
and create a grammatical sentence which is marginal because unlike English for instance the NP
dominated by PP in Kinandi is not really a CONTINUOUS CONSTITUENT. Thus the sentence
below (26) is only marginally accepted.
26. ?Kipleel (ne) ka-a-rib tuuga ak
„Kipleel (that) tns-I –look cows with‟
(= Kipleel that I looked after cattle with)
TOPICALISATION and WH-MOVEMENT create marginal sentences by analogy (of sentences
which underwent PP PREPOSING) by violating the UNIT MOVEMENT CONSTRAINT in
moving non-continuous constituents together (i.e. a CONJUCTION and NP together)
27 a) …s[VP[V[achame]V[Jeruuto]NP]VP[ak]CONJ NP[Jeleel]NP]s
27 b) [COMP ne S [ka i roo Jesaang’ [ak] CONJ NP [ng’oo] NP] S] S1
64
(WH-MOVEMENT)
Thus, our conclusion is that Kinandi is Subject to both COORDINATE STRUCTURE
CONSTRAINT and the UNIT MOVEMENT CONSTRAINT. The marginality of the sentences
analysed are therefore results of the violation of one or the other of these constraints. This is also
in addition to the ambiguity of ak, which acts as both a conjunction and a preposition. However,
this ambiguity is syntactic, not pragmatic or even semantic, because the speakers somehow do
UNDERSTAND the utterance since intuitively, they know that both the conjunction and
preposition (signaled by ak) mean „in addition to‟. Indeed, the same phenomenon seems to
„afflict‟ ADVERB-PREPOSING as shown below
28. Ariibe tuuga raani ak mutaai
„I- look cattle today and tomorrow‟‟
(I look after cattle today and tomorrow)
S[VP[pro[a V[riib tns[e]tns]V pron[anee]]VP ADVP [ADV[raani]] ADVP CONJ[ak]
ADV[mutaai]]S
=TOPICALISATION
29. ? Ak mutaai, aribe tuuga raani –
(And tomorrow I- look (after cattle today –)
We see again that by analogy of PP-PREPOSING we have a marginally accepted syntactic
structure. Recall that if we had interpreted ak as a preposition in this case, we would not have
had a problem since in PP-PREPOSING, the preposition is moved along with the other unit (e.g.
NP, N, ADV, V, etc.) Thus, the following is correct grammatically in Kinandi,
30. a) Ariibe tuuga raani ak mutaai
(I look after cattle today with tomorrow)
S[VP[pro[a V[riib tns[e]tns]V pron[anee]]VP N[tuuga]N PP[P[raani]P]CONJ[ak] P[mutaai]PP]S
65
= TOPICALISATION
b) With tomorrow I- look (after cattle today-
If the synonym for ak, „with‟, that is, kobooto, is used in the sentence above, it then has to be
moved with the adverb for the sentence to be grammatical
31. a) Ariibe tuuga raani kobooto mutaai
(I look after cattle today with tomorrow)
S[VP[pro[a V[riib tns[e]tns]V pron[anee]]VP N[tuuga]N ADVP[ADV[raani]]P[kobooto]P
ADV[mutaai] ADVP]S
= TOPICALISATION
b) Kobooto mutai, ariibe tuuga raani___
(With tomorrow, I- look after cattle today )
ADVERB PREPOSING will not result in a grammatical sentence.
32. a) *Mutaai, ariibe tuuga raani kobooto
(Tomorrow, I- look after cattle today with)
(see also 29 above)
From this analysis, it seems to me that, for Kinandi, there is need for a reformulation of a more
general UNIT MOVEMENT CONSTRAINT that takes into consideration not only syntactic, but
also semantic aspects of sentences. This is because in Kinandi, the Preposition Phrase, apart from
being syntactically continuous, is also semantically continuous but the preposition is not
syntactically recognized without the Verb, Noun, Noun Phrase, Adverb, etc. This way, the
UNIT MOVEMENT CONSTRAINT will also include semantic units.
66
5.1.5 WH-ISLAND CONSTRAINT
The other constraint that Chomsky (1972) formulated was the WH-ISLAND constraint
which is already mentioned above. The rule argues that it is not possible to move a
constituent out of any clause containing WH-Phrase in COMP. Thus, a sentence in
English, for instance cannot undergo WH- or NP-MOVEMENT, or TOPICALISATION
as exemplified hereunder.
33. a) I don‟t know whether he is at home.
= TOPICALISATION
b) * Home, I don‟t know whether he is at .
Nor can NP-MOVEMENT
c) * I don‟t know home whether he is at .
These examples are simplified for heuristic purposes, but the point here is that some languages
like English which have a WH-phrase in COMP (e.g. whether in English) do not allow for the
extraction of any constituent as shown. However, in KInandi, it seems this constraint does not
hold. From the analysis of the following structures, we can see that Kinandi can undergo
TOPICALISATION, WH-MOVEMENT, ADVERB-PREPOSING, and PP-PREPOSING
despite the existence of a WH-phrase in COMP.
Let us start with TOPICALISATION;
34. a) Ma a ngen ngot kebendii acheek gaa
„Not I know whether tns-go-aspect we home‟
(= I don‟t know whether we are going home)
= TOPICALISATION
b) Gaa, ma a ngen ngot kebendii acheek
„Home, not-I-know whether tns-go-aspect we‟
(*Home, I don‟t know whether we are going ___
67
(= I don‟t know whether we are going home)
(*Note here that the English translation in parenthesis is ungrammatical, in English, because
English obeys the WH-ISLAND CONSTRAINT).
We can also see here that ADVERB PREPOSING and WH-MOVEMENT respectively, can
extract a constituent as shown by 35 and 36 below.
35 a) Ma a ngen ngot kebendii acheek gaa raani
„Not-I-know whether tns-go-aspect we home today‟
(= I don‟t know whether we are going home today)
= ADVERB PREPOSING
b) Raani, ma a ngen ngot kebendii acheek gaa____
„Today, not-I-know whether tns-go-aspect we home ‟
(= Today, I don‟t know whether we are going home)
36 a) Ma a ngen ngot kowendii ga ng’oo
„Not-I-know whether tns-go-aspect home who‟
(I don‟t know whether who is going home)
= WH-MOVEMENT
b) Ng’oo (ne) ma angen ngot kowendii gaa ____
„Who who not-I-know whether tns-go-aspect‟
(*Who I don‟t know whether ___ is going home = I don‟t know whether who
is going home [*Explanation after 34 b above also applies here]).
So can NP-MOVEMENT;
37 a) Ma a ngen ngot kowendii gaa Kipsaang’
„Not- I-know whether tns-go-aspect home Kipsaang‟‟
68
(= I don‟t know whether Kipsaang‟ ís going home)
= NP-MOVEMENT
b) Kipsaang‟ (ne) ma angen ngot kowendii gaa____
„Kipsaang‟ (who) not-I-know whether tns-go-aspect home‟
5.1.6 MOVEMENTS AND SUBJACENCY IN KINANDI
The SUBJACENCY CONDITION is a general constraint on movements which was
developed by Chomsky in the 1970s. This condition does not allow for movement of a
constituent out of more than one bounding node in any single application of a rule.
Before we look at how this applies to Kinandi we look at the following sentence in
English to help us understand how subjacency can be violated.
38 John is thought to want Mary to win.
A simplified underlying structure of this sentence would be;
COMP [S1 npe is thought COMP[S2 John to want COMP[S3 Mary to win]]]
(Radford 1981:232)
Now, if the NP, Mary is moved to the empty NP (npe) position, then there is a violation
of sujacency because the NP, Mary would move across two bounding nodes, that is, S3
and S2 when subjacency allows only movement across one bounding node.
Furthermore, even though subjacency allows successive cyclic movement of an NP, the
sentence above would not allow for the movement of Mary since S2 subject position is
already filled by John. Moreover, this movement, being a substitution rule, can only
allow movement of empty positions.
For Kinandi, we use the following sentence to illustrate NP MOVEMENT and its
relationship to subjacency;
69
40. Ka a bir Kibeet ne kalu chego
„Tns-I-beat Kibeet who tns-drink milk‟
(= I beat Kibeet who drank milk)
The underlying structure of this sentence is;
COMP1 ne S1[VP[Ka [a]tnsV[ bir[ inee]pro]VP NP1[anee]]COMP2 ne S2[Ka lu chego
Kibeet]]]
The subject NP, Kibeet has been adjoined to COMP by an adjunction rule to derive the
surface structure 40 above. This same NP, Kibeet can be adjoined to the higher COMP to
derive (informally);
41. Kibeeti COMP-wh ne ka-a bir ne kalu chego ti .
„Kibeet comp-wh who tns-I-beat who tns-drink milk__‟
(= The Kibeet who I beat drank milk)
This movement is allowed because it is an adjunction movement to COMP and it can apply
cyclically, that is, a COMP to COMP analysis of NP-movement can be envisaged for Kinandi.
5.1.7 SUBJACENCY AND RAISING
In Chapter 4 (4.3.1) we saw how „raising‟ was envisaged for the sentence below
42 Uu ne kawo Jebeet
„seem that tns- go Jebeet‟
(= It seems that Jebeet went)
The phrase marker to show (subject) raising is envisaged in Fig 2 below.
70
S1
COMP S NP
-wh INFL VP N
Ne tns V S2 Jebeeti
prs uu COMP S3
-wh INFL VP NP
ne tns V eti
pst wo
ka
Fig 2 Phrase marker showing subject raising in Kinandi
We find, therefore, that this „raising‟ does not violate sujacency because the NP (Jebeet), is
moved cross only one bounding, node, that is, S3. This also happens when it is adjoined to V on
the S2 cycle (cf Fig 11 of 4.3.1).
However, sentence 40 above cannot undergo an (object) NP raising because both subject and
object positions on the super ordinate clause are filled. Thus, there is no landing site for the NP.
The NP, Kibeet cannot be adjoined to the V node similar to Fig 11 (4.3.1) because it will not
only have violated subjacency by moving across two bounding nodes (S2 and NP1) but it will
have also violated the theta criterion and projection principles.
Similarly, in our heuristic example in Chapter 1 (Fig 4), raising cannot have taken place because
the theta criterion would be violated. Furthermore, NP2 contains this NP, chiito and its „co-
referent‟ object NP in the subordinate clause can be envisaged to have been as wh-word, ng’oo
71
which was moved to COMP and underwent a subsequent WH-DELETION rule as illustrated
below.
44 COMP-wh ne S2 [[ka a nyooru chiito] anee]S1[COMP –wh ne [ko bir ng’oo
lakweet] S3]
= WH-MOVEMENT
S1 [COMP-wh ne S2[[ka a nyooru chiito] S2 anee] S1 S1[comp ng’ooi comp
ne [ko bir ti lakweet] S3]S1
= WH-DELETION
= Ka a nyooru chiito anee ne ko bir lakweet.
This in turn leads to the subsequent pro-drop of the subject pronoun anee in the super ordinate
clause to give
45 Ka a nyooru chiito ne ko bir lakweet
„Tns-I-find man who tns-beat child‟
(= I found the man that the child beat)
From this analysis then, it can be claimed that movements in Kinandi are subject to
SUBJACENCY and if constituents have to cross more than one bounding node, then the
movements must apply cyclically. Since substitution rules allow only movement to empty nodes,
movement of an NP (such as exemplified by sentence 41 above) is blocked unless it is
TOPICALISATION or NP MOVEMENT to COMP.
72
CHAPTER 6
SUMMARY AND CONCLUSION
6.0 Summary
The main thrust of this book was to give an extensive description and explanation of the kind
of transformations that can be found in Kinandi syntax. Six types of transformations
(movements) were found in the language. These are NP-MOVEMENT, WH-MOVEMENT,
ADVERB PREPOSING, RAISING, TOPICALISATION AND PP-PREPOSING.
Subsequently, the various constraints proposed mainly by Chomsky, Ross and Schwartz were
used to show the extent to which they apply to Kinandi syntax. Here it was found, for
example, that none of these transformational rules could extract a constituent out of a relative
clause. Furthermore, it was also seen that this phenomenon applies to other adnominal
clauses. This led to the conclusion that Kinandi is subject to the COMPLEX NOUN
PHRASE CONDITION which forbids extraction of constituents from adnominal clauses.
Another finding is that movement to the left-most COMP can be envisaged to apply
cyclically. Hence, it can be claimed that Kinandi is subject to SUBJACENCY.
Further, it was shown that COORDINATE STRUCTURE AND UNIT MOVEMENT
CONSTRAINTS do apply to Kinandi. An examination of coordinate structures reveled that
no part of a coordinate structure can be extracted by any movement rule. However, it was
seen that, at certain times, the constraint is broken and this results in derivation of marginal
sentences. This was explained as a phenomenon that was borrowed from the syntax of the
Prepositional Phrase since, in most cases, it is not easy in Kinandi to differentiate the
preposition ak (with), with the conjunction ak (and).
In addition to this breaking of the coordinate structure constraint, it was also found that this
extraction also violated the UNIT MOVEMENT CONSTRAINT because the moved
elements were not continuous, that is, they were neither a conjunction nor a noun. Further, it
was shown also that PP-PREPOSING violated the UNIT MOVEMENT CONSTRAINT
73
when it (PP-PREPOSING) „borrowed‟ the syntax of the coordinate structure to move a sub-
part of a continuous constituent, that is, a PP (which in Kinandi is semantically continuous
and has to be extracted as a unit). It is proposed that semantic units should also be considered
in the interpretation of transformations, since a preposition in Kinandi cannot be left alone
when extracting and NP as exemplified in 5.1.4 (sentence 26).
5.1 Conclusion
The task that this research set out to do was to discover the extent to which the theory of
transformational generative grammar is able to handle the syntactic data of Kinandi. Of
particular interest was the extent to which the transformational processes in the language can
be described and explained.
It can be therefore, said that the framework chosen has been helpful in highlighting the
syntax of the language and especially the transformations. Thus, it can be claimed that the
framework has had both descriptive and explanatory adequacy to handle the data from
Kinandi.
In the rationale, it was pointed out that there is so far little literature on Kinandi and this is
even more so with respect to literature that uses a modern linguistic theory to describe and
explain the language. It is believed that this book will be an addition, albeit a humble one, to
modern description of Kinandi syntax as well as forming a corpus of information on African
linguistics in general. It is offered “not with any pretence of its being exhaustive, but as a
window on the diverse and far-reaching types of questions that concern theoretical generative
linguistics… (to) serve not so much as a repository of facts but as a spark, to urge readers to
explore further…” (Riemsdijk &Williams, 1986:342, author‟s emphasis).
References
Akmajian, A. and Heny, F. 1975. Introduction to the Principles of Transformational Syntax.
Cambridge, MA: MIT Press
Bach, E. 1974. Syntactic Theory. New York: Holt Rhinehart and Winston
74
Chomsky, N. 1957. Syntactic Structures. The Hague: Mouton
Chomsky, N. 1965. Aspects of the Theory of Syntax. Cambridge, MA: MIT Press
Chomsky, N.1966. Topics in the Theory of Generative Grammar. The Hague: Mouton &Co
Chomsky, N. 1981.Lectures on Government and Binding. Dordrecht: Foris Publications
Emonds, J.E. 1976. A Transformational Approach to English Syntax. New York: Academy Press
Greenberg, J. 1955. Studies in African Linguistics. Branford, Connecticut: Compass Publications
Greenberg, J. 1966. The Languages of Africa. The Hague: Mouton
Grinder, J.T and Elgin, S.H. 1973. Guide to Transformational Grammar. New York: Holt
Rhinehart and Winstion
Hollis, A.C. 1909. The Nandi: Their Language and Folklore: Oxford: Clarendon Press
Huddleston, R. 1976. An Introduction to English Transformational Syntax. London: Longman
Katz, J.J. and Fodor, J.A. 1964. The Structure of Language: Readings in the Philosophy of
Language. Englewood Cliffs, NJ: Prentice Hall Inc.
Kayne, R. 1975. French Syntax: The Transformational Cycle. Cambridge, Mass: MIT Press
Mumford, J. 1951. Nandi Studies: Fifty Lessons plus Vocabulary. Kapsabet:AIM
Omondi, L.N. 1982. The Major Syntactic Structures of Dholuo. Berlin: Dietrich Reimer Verlag
Onditti, T.L.S 1987. The Syntax of Complementation in Dholuo: A Government and Binding
Approach. Unpublished MA Theisis, University of Nairobi
Ross, J.R. 1967. Constraints on Variables in Syntax. Doctoral dissertation, MIT, Cambridge,
Massachusetts
Ross, J.R. 1986. Infinite Syntax! Norwood, NJ: Ablex.
75
Radford, A. 1981. Transformational Syntax. Cambridge: Cambridge University Press
Riemsdijk, H.V. and Williams, E. 1986. Introduction to the Teory of Grammar. Cambridge,
Mass: MIT Press
Stockwell, R.P. 1973. The Major Syntactic Structures of English. New York: Holt Rhinehart and
Winston
Stockwell, R.P. 1977. Foundations of Syntactic Theory. Englewood Cliffs, N. J, Prentice Hall.
Toweett, T 1975. Kalenjin Nouns and their Classifications with notes on phonology and a noun
list. Unpublished M.A Thesis. University of Nairobi.
Toweett, T. 1977. A Study of Kalenjin Linguistics. Nairobi: Kenya Literature Bureau
Tucker, A.N. 1956. The Non-Bantu Languages of North Eastern Africa. London: Oxford
University Press
Tucker, A.N and ole Mpaayei, J.T. 1954. A Maasai Grammar with Vocabulary. London:
Longman Green and Co.