|
The morphosyntax of ghomala' verbs: focus on inherent complément verbs and serial verb constructionpar Corrine Minette FOKO MOKAM Université de Yaoundé 1 - Master 2020 |
ConclusionThe main task of this chapter was to provide the reader with background information on the grammatical aspects of Gh?maìlaì' based on previous works on the language. The sound system, noun morphology, the clause structure as well as functional categories have been discussed in a very sketchy fashion. The discussion reveals that Gh?maìlaì' is a tone and noun class Bantu grassfield language with SVO as unmarked word order. Tone is significant in the language since it contributes in distinguishing two lexical items and marks some functional categories as well as some constructions. The language is devoid of noun class prefixes and agreement is the best criterion used in noun classification. Noun modifiers vary depending on the noun class to which belongs the head noun. Subject verb agreement is not attested in the language and its inflectional domain isn't enough rich of functional heads. We have two negative particles in Gh?maìlaì'. The verb itself has not been addressed in this chapter; it will be deeply discussed in the following chapter. Chapter 3: Gh?maìlaì' verbs 58 IntroductionThe split of words into distinct categories or parts of speech is one of the well-known linguistic tradition. Some words are inflected for case whereas others are inflected for tense, aspect, mood and person. The verb refers to an activity or process performed or undergone and occurs as part of the predicate of a sentence. In the precedent chapter, functional categories related to verb have been addressed by exploring the TAM system of Gh?maìlaì'. The core issue of this chapter is to provide an overview of Gh?maìlaì verbs. The morphological, the syntactic and semantic properties of verbs are discussed. In doing so, a classification of verbs is done following three criteria namely the tonal, the syllabic and the morphological criterion. The basic verb structure is addressed as well as the derivative verbal affixes attested in the language. Moreover, the chapter also explores the issue of transitivity in the language and groups verbs on the basis of their argument structural characteristics. Semantic restrictions of some verbs are also analyzed. The chapter is structured as follows: 3.1. Classification of Gh?maìlaì' verbs As we said earlier, three criteria are used to classify verb roots in Gh?maìlaì'. Verbs can be grouped on the basis of their tone, their morphology and their syllable pattern. Adopting the tonal criterion, verbs roots are split into two groups: high tone verbs and low tone verbs. However, these tones can be subjected to some modifications due to TAM marking. (1) High tone verbs Low tone verbs
59 One of the relevant peculiarities of Gh?maìlaì' verbs is their form. Morphologically, verbs have two forms which are referred to as forme nue 7and forme alternative by Mba (1997). The first form is used for accomplished actions and in imperative construction whereas the second is used for unaccomplished actions. The difference between the first form and the second one relies on the initial consonants of the verb root. When the initial consonants of the first form are voiceless stops, those of the second form will be either voiced stops or pre-nasalized consonants as shown below: (2) First form Second form pé hé «take» t?ìm d?ìm «injure» ké ?ké «call/read» toì ntoì «braise» pf?ìp mpf?ìp «cover» Verbs beginning with voiced fricatives in their first form have their second form beginning either with the corresponding stops or affricates as illustrated below: (3) First form Second form Ì ?? m g?Ìm « speak» vù bvù « fall» 3w?Ìp d3w?Ìp «sing» Jû? d3û? «understand» wiÌm gwiÌm «possess» Verbs beginning with the lateral / l/ in their first form have their second form beginning with the voiced stop alveolar /d/ as shown below: (4) First form Second form lé dé «look» là daÌ «stick» lù dù «leave» l?ì d?ì «cry» 7 bare form and alternative form 60 Verbs beginning with nasals and voiceless fricatives in the first form do not undergo the process of first consonant alternation as illustrated below: (5) First form Second form s?Ì? s?Ì? «come» faÌ? faÌ? «work» hé hé «open» miÌ miÌ «swallow» ì n? né «cook» Looking at the syllable pattern of the verb roots, verb roots are generally monosyllabic. The onset of a syllable in Gh?maìlaì' is a consonant. This means that most of the words in this language begin with a consonant. However, we can find some words which are made up of a single sound that is a vowel such as personal pronoun. Verb roots in Gh?maìlaì' exhibit the following syllable types. ? The CV syllable type This is a combination of a consonant which is the onset and a vowel that is the nucleus and the tone bearing unit as in the following examples:
? The CVC syllable type According to Moguo (2016), this is the most common syllable pattern in the language. It is made up of a consonant which is the onset, followed by a vowel, the nucleus and the tone bearing unit, and another consonant that is the coda. All consonants, except the glottal ?, can figure at the onset position but only few can appear at the coda position namely p, k, m, ? and ? as shown below:
? The CGV syllable type 61 This syllable pattern is made up of a consonant followed by a glide and a vowel as in the following examples:
? The CGVC syllable type This is a combination of a consonant followed by a glide, a vowel and a consonant as shown below:
3.2. The structure of the verb in Gh?maìlaì' The structure of the verb in the language under study is slightly different from what is attested in other Bantu languages such as Tuki, BasaÌaì or Ewondo. In these languages, the basic verbal form is made up of a root and suffixes. The suffixes consist of extensive morphemes and the so called final vowel. Gh?maìlaì' is devoid of a final vowel (FV) and the verbal base can be preceded by the infinitive particle né- which can be literally translated as the English infinitive morpheme «to». 3.2.1. The infinitive form The infinitive form in Gh?maìlaì' is marked by n?ì- which, as I said earlier, precedes the verbal base and a floating low tone that is born by the last vowel of the first form of the verb. The floating tone is attached only to verb roots with high tones. In other words, when the tone borne by the root is a high tone, the floating tone attaches to it to form a falling tone as in the following
62 n?ì k?ìm Ì n?ìk?^m «to shew» n?ì lé Ì n?ìle^ «to look» Being the nominalized form of the verb, the infinitive form can occupy different structural position in the language as in the following sentences: (11) Subject position a. N?ì-n?^ ms?^ t?j?Ì Ba?ka^m INF-cook fufu corn be beyond Bakam «To cook fufu corn is beyond Bakam» b. N?ì-fa^? puÌ? (12) Complement position
Bucket INF-fletch water «The bucket for fletching water» In (11), the infinitive form is the subject of the sentences. It is relevant to indicate that the predicates of these sentences are not full lexical verbs but adjectives. In (12c), it is the complement of the noun within the noun phrase and it provides more information about the function of the bucket. Moreover, Mba (1997) observes that there is a restricted class of verbs that selects as complement an infinitive clause amongst which to learn, to know, to want, to be able, to help, etc. The verb base can be extended via the suffixation of extensive morphemes. 63 3.2.2. Derivative verbal affixes and their interpretations Two derivative verbal morphemes are attested in the language: -?à and -tà (Mba 1997). They encode different interpretations and can affect the argument structure of the verb. These suffixes are attached to the simple verbal bases to generate extended verbs. Moreover, these morphemes have no meaning on their own. The two suffixes cannot be used simultaneously in a verb. This implies that a verb root can select only one extensive morpheme. Amongst these two suffixes, the suffix -?? tends to affect the valency of the verb more than the other one. 3.2.2.1. The suffix -??Ì As observed by Mba (1997), it is really difficult to attribute a general meaning to this suffix. The emphasis is laid on the participants of the action described by the verb. This suffix expresses the notion of reflexivity and reciprocity. 3.2.2.1.1. Reciprocity The suffix -??Ì modifies the meaning of the verb by adding in some case the idea of reciprocity. Accordingly, Tamanji (2008) argues that the reciprocal meaning of a suffix expresses the idea of plurality of relations: A stands in a certain relationship with B and B also stands in exactly the same relation to A. When affixed to verb roots, the reciprocal suffix points out that the participants undergo the action inherent in the verb. The participants act as agents and recipients of the same action. The suffixation of -??Ì in this case has the effect of decreasing the valency of the verb as shown in the examples below: (13) a. Ta^laì ?kuì? Ba?ka^m Tala PRS.love Bakam «Tala loves Bakam»
In the data above, the verb ?kuì? «love», inherently, is a two-place predicate verb. It selects two arguments: the lover (agent) and the loved (patient). In (13b), when -??Ì is suffixed to this verb, it becomes a one-place predicate verb since the agents of the action also function as the patient of 64 the same action. The subject also becomes plural as two (13b) or more participants in the action are at the same time agent of their own action and patient. 3.2.2.1.2. Reflexivity The suffix -??Ì also changes the meaning of the verb by adding in some case the idea of reflexivity. The reflexive meaning of the suffix -??Ì expresses the idea of equality: A is equal to B. The reflexive suffix indicates that the agent undergoes the action described by the verb. The subject of the sentence blend with the object. When suffixed to the verb roots, this suffix has the effect of decreasing the valency of the verb as highlighted in the examples below: (14) a. Ta^la? t??Ì? luÌ?gaì Tala PRS1.tie bucket «Tala has tied the bucket»
Tala PRS.tie-REF In (14a), the verb t??Ì ? «tie» is a transitive verb, it selects two arguments: the agent Ta^laì and the theme luì?gaì «bucket». In (14b-c), when -??Ì is affixed to this verb, it becomes a one-place predicate verb. In other words, the extended verb no more allows a noun phrase to occur in the object position; this is the reason why (14c) is illegible. The extended verb selects one syntactic argument which is the subject and this one undergoes the action of hanging. 3.2.2.2 The suffix -t?Ì According to Mba (1997), the suffix -t?Ì alters the meaning of the verb root by assigning a general pluractional meaning to the verb. This pluractionalisation can be mirrored through the iterative, the attenuative and the distributive nature of an action as well as the increasing of number of participants and the propagation of an action within time and space. 3.2.2.2.1 Distributive The suffix -t?Ì modifies the meaning of the verb by expressing, in some case, the idea of distributive. The distributive suffix -t?Ì , as claimed by Moguo (2016), indicates that the action implied in the verb affects the same object several times. Let us consider the following examples: 65 (15) a. Waìp k?Ì t?aÌm paÌ? 3PL.NOM PST2 hit house «They hit the house»
Tala PST2 break-DIST door «Tala broke the door in several pieces» In the data above, the derived verb has a semantic link with the simple verb. The suffixation of the extensive morpheme -tà doesn't completely change the meaning of the verb; it rather adds some precisions on how the object is affected: it either undergoes the action in several spots (15b) or has been split into several pieces. Furthermore, this suffix doesn't affect the valency of the verb since its affixation neither decreases nor increases the number of arguments selected by the verb. 3.2.2.2.2 Iterative When affixed to the verb root, the suffix -tà indicates that the action is realized several times in some cases as in the examples below: (16) a. Ta^laì k?Ì v?Ì ?waÌ??? biì puì Tala PST2 write letter to 3PL.DAT «Tala wrote them a letter» b. Ta^laì k?Ì v?Ì-t?Ì
?waÌ??? biì puì c. Mu^ ts?ì ké Child PRS.lick dish «The child licks the dish» 66 d. Mû tsé-t?Ì ké Child PRS.lick-IT dish «The child licks the dish several times» As one can observe, the suffixation of the extensive morpheme -tà doesn't have an effect on the argument structure of the verbs. The extended verb keeps the meaning of its verb root. What has fundamentally changed in (16b and d) is the repetition of the action described by the verb. 3.2.2.2.3 Attenuative The suffixation of the extensive morpheme -tà to some verb roots, especially those that denote an action which requires a physical force, indicates that the intensity of the action on the object is weakened, only part of the object is affected by the action or the entire action lasts for a short time as in the following examples: (17) a. Ba^ka?m w?ì fiÌ? ?j? n?ì
nd?aÌp
Simo PRS4 scratch-ATT back father 3SG.1.POSS «Simo is slightly scratching the back of his father» 3.2.2.2.4 Plurality of participants When suffixed to some simple verbs, the extensive morpheme -tà has the overall effect of multiplying the participants in the action. More precisely, when suffixed to simple verbs, this suffix indicates that the action is either realized or undergone by several participants in the same location at the same time or in different locations at different times. Let us consider the example below: (18) a. Ta^laì k?Ì kaìp ka^f?ì Tala PST pick coffee «Tala picked coffee» b. Waìp k?Ì kaìp-t?Ì ka^f?ì 3PL.NOM PST2 pick- PL coffee (19) a. Oì g??? laìk m-kg 2PL.NOM FUT1collect dishes «You will collect dishes» b. Po g??? laìk-t?Ì m-kg 2PL.NOM FUT1 collect-PL PL-dish «You will collect a lot of dishes»
67 The above data are cases of pluralization of participants. In (18), the suffix -tà pluralizes the participant who is doing the action meanwhile in (19) it pluralizes the object. In (18b andd), the affixation of this morpheme indicates that the action is performed by several persons. This is the reason why the extended verbs select plural subjects. (19b and d) indicate that a good number of objects is referred to. 3.2.2.2.5 Extension of the action in space The suffixation of -tà indicates, in some cases, that the action is spread over a space as in the following examples: (20) a. Ta^laì k?Ì n?Ìk d?ap Tala PST2 spread vegetable «Tala spread vegetable seat»
68 «Bakam destroys the farm» d. Ba?ka^m sè-t?Ì neÌ Bakam PRS.destroy-PL farm «Bakam destroys a large surface of the farm.» The suffix -tth tends to have no effect on the argument structure of the extended verb unlike the morpheme -?th which decreases the valency of the derived verb. Furthermore, certain verbs do not admit these suffixes (see Mba 1997 for more details). Notions of instrumental, causative, applicative, separative and benefactive are expressed in some Bantu languages such as BasaÌaì, Tuki, Akoose via extensive morphemes. This is not the case in Gh?maìlaì' wherein these notions are expressed by prepositions which are positioned after the verb and causative is analytic. After having explore the structure of verb, the next section has do to with the argument structure in Gh?maìlaì'. 3.3. Argument structure in Gh?málá' This section looks at argument structure in the language under study. It aims at exploring what Gh?maìlaì' verbs require in a simple declarative sentence. In doing so, the issue of transitivity in Gh?maìlaì' is discussed and verbs are grouped on the basis of their argument structural characteristics. 3.1. Argument structure Grimshaw (1992) defines argument structure as the lexical representation of grammatical information about the predicate. The argument structure of a lexical item is thus part of its lexical entry. To provide an account of the regularities in the syntactic expression of arguments has been one of challenges of linguistic theory. The central question is how to map the arguments onto the syntactic structure. In this vein, two approaches to argument structure have been developed within the generative enterprise: the lexicalist approach (Gruber (1965), Jackendoff (1972), and Chomsky (1981, 1995, and 2000...etc) and the constructionist one (Hale and Keyser (1993), Borer (1994: 2005), and Harley (1995)). In the first approach, the lexical item (the verb) determines the argument structure of the clause; the verb has theta-roles and these are projected onto the syntax. In the second approach, the verb is seen as composed of smaller events: the initiation or causing subevent, the process subevent and the result subevent. The structure around the verb plays a major role in the argument structure. The lexicalist and the constructionist approaches are both important since both the information about the verb in the lexicon and the structure around the verb play a 69 role in determining the argument structure of a clause. More precisely, argument structure is determined by properties of verbs, in particular by the syntactic configurations in which they must appear. Two notions are essential for any discussion on argument structure: predicate and argument. A predicate is an expression that denotes an action, a state or a process in a sentence. In Gh?maìlaì', this notion is illustrated by the verb as in the examples below: (21) a. Ta^laì té ?k?deì Tala PRS.plant banana (action) b. LuÌ?gaÌ l?ì bucket
PRS.be clean (state) In the examples above, bold constituents are predicates. In some languages such as English, other lexical categories like adjectives can function as predicates but for the purposes of the present study, predicates are restricted to verbs only. Accordingly, Haegeman (1994) considers verbs as prototypical predicates since crosslinguistically verbs tend to be predicates. The argument, as claimed by Radford (2004), is an expression denoting a participant in the relevant activity or event. In other words, arguments are the participants that are required for a predicate to be expressed. In (21a) above, Ta^laì and ?k?dé «banana» are the arguments of té «plant». As far as arguments are concerned, Grimshaw (1979) and Pesetsky (1982) uses the dichotomy predicate's category-selection versus predicate's semantic-selection to distinguish semantic argument from syntactic argument. Korsah (2011) argues that s(emantic)-selection refers to the semantic constraints that a predicate puts on its arguments whereas c(ategory)-selection deals with the subcategorization frame; it talks about the fact that every predicate requires an argument of a particular XP, where X stands for a particular phrasal head. In (21a), for example, té «plant» s-selects two arguments, one which is able to realize the action of planting and another one that is capable of being planted. These arguments are DPs, that is, t?ì c-selects two DPs. Let us observe the data in (22) below: (22) a. Ta^laì k?Ì tsù p?ì Tala PST2 eat taro «Tala ate taro.»
70
In the data above, the c-selection of the verbs is satisfied in (22a-c) since they have two DPs as arguments. However, (22b) and (22c) are illegible because the semantic requirements of the verb are not satisfied. Unlike (22a), (22b) has an object that cannot be «eaten» as (22c) has a complement which is unable to be «planted». The ungrammaticality of (22d) is due to the fact that the predicate's c-selectional requirements are not met; (22d) has an AdjP instead of a DP. In fact, the idiosyncratic requirements of these verbs as specified in the lexicon are not met. The verbs tsé «eat» and té «plant» are projected into the syntax with the following properties:
Each argument bears one and only one theta-role, and each theta-role is assigned to one and only one argument. (Chomsky 1981: 335) The argument structure of a verb includes the number and type of theta-roles it has to assign and the corresponding number and syntactic type of arguments required to receive these theta-roles. In this vein, once a constituent is a semantic argument, it is automatically a syntactic argument. This observation does not hold in some languages such as English where a non-semantic but syntactic argument is attested with expletive constructions like: there are believed to have 71 occurred several riots where there is not semantically required but syntactically in order to satisfy the EPP requirement. Theta grids, apart from showing semantic roles, also illustrate the grammatical relation between arguments by indicating which argument should be the subject. To indicate that one of the arguments may become the grammatical subject, Williams (1981) underlines that argument and refers to it as the external argument as shown below: (25) a. ?ó «be sick»: < Experiencer > [___ DP] Ta^la? ?oì «Tala is sick» Tala PRS1.be sick # Expriencer
T?^ k?Ì haì ?kaìp biì Ta^laì «Your father gave money to Tala» Father.2SG.1.POSS PST2 give money to Tala # # # Agent Theme Goal Apart from indicating the structural position of arguments, the examples in (25) above suggest that an argument structure analysis relates to the categorization of verbs based on transitivity. 3.2. Gh?maìlaì' verbs and transitivity Naess (2003) defines transitivity as a type of grammatical relationship encoding the distinctness of participants in a situation described by the clause. It refers to the number and type of objects which appear in the clause and which are selected by the predicate's head. In terms of classification of verbs based on transitivity, three classes of verbs are attested in Gh?maìlaì': those 72 that select one participant, those which selects two participants and the other that allow three participants. From this observation, Gh?maìlaì' verbs can be grouped into intransitive, transitive and ditransitive verbs. This subsection aims at examining certain aspects of the syntax of the verbs involved. 3.2.1 Intransitive verbs These verbs in Gh?maìlaì' select one argument which is the subject. They occur without an object as in the examples below: (26)a. Ta^la? tiì Tala PRS1.sleep «Tala sleeps» b. Siìmo? pf?ì simo PRS1.die « Simo dies» The arguments are Ta^la? in (26a) and siìmo? in (26b). The theta-role of the only argument of intransitive verbs is determined by the semantics of the verb. In (26) above, the verbs assign the Theme theta-role to their only argument since they involve a meaning of a state of affairs in which the semantics of cause or control is not asserted. In (26) for example, Simo has no apparent control over pfû «die». Verbs that fall under the group of intransitive verbs in Gh?maìlaì' belong to the following semantic classes. 3.2.1.1 A semantic classification of Gh?maìlaì' intransitive verbs ? Stative verbs They designate states or conditions of particular entities. Gh?maìlaì' stative verbs tend to be translated as adjectives in English as shown in the examples below:
INF-beatiful INF-be mouldy b. bi^jé j??? z?Ì Groundnut this bitter «This groundnut is bitter» The z?Ì in (27b) describes the taste of the entity bi^j?ì «groundnut». ? Change of state verbs Verbs of this kind show a change in the physical condition of an entity to another condition. The following are examples: (28)a. INF-swell « to swell» n?ì pwa^ INF-be soft «to be soft» ì Ì n? mw? 73 n?ì kx?^ «to burn» n?ì pa^k «to spoil» INF-burn INF-spoil n?ì vo^ «to get old» n?ì pè « to ripe» INF-get old INF-ripe b. w?ìtsù j??? paìk Food this PRS1.spoil «This food has spoiled» The sentence in (28b) presupposes that the food was in an eatable state, but now it has become uneatable. There has been a change in the condition of the edibility. ? Verbs of weather condition Such verbs describe various conditions of the weather e.g raining, shining etc. These include luì «rain», tj?ì «shine», t??Ì ? «shake» (used for wind). (29) a. b?^? dû f?ìkdzû b. na?m tj?ì tj?ì??Ì ? Verbs for involuntary bodily processes These verbs designate some inherent body experiences of which such entities have very little or no control over. They include n?ì t?a^k??ì «to sneeze», n?ì ?kwe^ «to cough», n?ì s?^ ? «to shiver», n?ì ?weÌ??Ì «to breathe». (30)a. mû w?ì-?kweì Child PRS4-cough 74 «The child is coughing» b. Ta^laì kà w5-t?aìk?5 Tala PST2 PROG-sneeze «Tala was sneezing»
These are verbs that show the movement of an entity from one point or location to another. They include né s32 «to come», né 0 «to go», né lù «to leave», né kû? «to crawl» etc. (33)a. Ta^la? s?Ì? dj?? 75 b. Ba?kaÌm k?Ì y?Ì go? Bakam PST2 go farm «Bakam went to farm». In (33), Tala and Bakam moved respectively from an undisclosed location to djä «home» and go? «farm». They have volition and could be said to have control over what the verbs express. Intransitive verbs do not have internal arguments and they have either a theme or an agent theta role to assign and therefore require only one argument. Building on work by Hale and Keyser (1993), Chomsky (1995) argues that the external argument is hosted by Spec-?P, more precisely, the ?P-VP configuration expresses the agentive or causative role of the external argument. This implies that intransitive verbs lacking agents are simple VP structures meanwhile those having agents are complex ones. 3.2.1.2 Unergative- Unaccusative distinction in Gh?málá' The split of intransitive verbs into unaccusatives and unergatives roots from Perlmutter (1978). He has promulgated the unaccusative hypothesis within the Relational Grammar framework as follows «certain intransitive clauses have an initial 2 but no initial 1». This means that the only one argument of some intransitive verbs appears in the same underlying structural position as the object of transitive verbs. According to Essegbey (2010), it is the advancement of the single argument from 2 to 1 that Perlmutter calls unaccusative. Unergatives, as opposed to unaccusatives, only possess an initial 1 stratum. This implies that the argument of this class of intransitive verbs belong to the same level as arguments that appear in the subject position of transitive verbs. Within the Government and Binding theory, it is assumed that the only one argument of unaccusative verbs occurs as deep structure object whereas that of unergatives occurs as deep structure subject. Following Burzio (1986), it is argued that unaccusatives are not able to assign structural case (accusative case) to their single arguments so the latter are forced to move to subject position in order to receive case. This means that the subjects of unaccusative verbs do not originate as the subjects of their associated verbs but rather as their complements. The unergative/unaccusative distinction between intransitive clauses can also be determined by the semantics of the verbs that occur in them. In this vein, Perlmutter (1978) proposed that predicates describing willed or volitional acts as well as those describing certain involuntary bodily processes like `cough', `sneeze' and `weep', belong to the unergative class. On 76 the other hand, predicates expressed by adjectives in English, predicates whose initial argument is a semantic patient, predicates of existing and happening and predicates of non-voluntary emission of stimuli that impinge on the senses will be unaccusatives. Chomsky (1995:316) distinguishes unaccusatives from unergatives by claiming that unaccusatives are devoid of agents and, therefore, are simple VP structures meanwhile unergatives have agents and their external arguments are hosted by the specifier position of the light verb phrase within the Larsonian shell. This view is adopted by Radford (2006) when he argues that unergative verbs differ from unaccusatives in that the subject of an unergative verb has the theta-role of an Agent argument, whereas the subject of an unaccusative verb has the thematic property of being a Theme argument. To discuss the implications of the Unaccusativity Hypothesis for Gh?maìlaì', let us consider the following examples: (34) a. Jwâts?? p?ìptâ c. Jwâts?? paìk food PRS.spoil food PRS.spoil «The food is spoiled» «The food is spoiled» b. Ba^ka?m p?ìpt?ì jwâts?ì d. *Ba^ka?m paìk jwâts?ì Bakam PRS.spoil food Bakam PRS.spoil food «Bakam spoils the food (35) a. muì wâ ?kuì? child PROG crawl «The child is crawling» b.* Ba?ka^m wâ ?kuì? muì Bakam PROG crawl child It can be observed in (34) «spoil» is expressed by two verbs namely, p?ìpt?ì and paìk. It is relevant to note that paìk only selects «food» as argument whereas p?ìpt?ì can go with anything which can be spoiled. In (34a), the verb p?ìpt?ì «spoil» is intransitive with jw?ì ts?? «food», the only argument functioning as subject but in (34b) where p?ìpt?ì is used with two arguments (Ba^ka?m and jw?ì ts?ì ), the hitherto subject argument jw?ì ts?ì «food», acts as an object. Unlike p?ìpt?ì «spoil», paìk «spoil» fails to be transitivized (34d) as well as ?kuì? «crawl» in (35b). This is an indication that paìk and ?kuì? belong to the same class of verbs which is different from the one that p?ìpt?ì belongs to. Verbs that behave like p?ìpt?, i.e verbs that can undergo transitivization, are said to be unaccusatives. In (34a), the subject of p?ìpt?ì has a Theme theta-role and (34b) strengthens the Burzio's idea, according to which, the subjects of unaccusative verbs do not originate as the subjects of their associated verbs but rather as their complement. Verbs that function like paìk and ?kuì? are referred to as unergatives. They cannot occur within a transitive clause as their unaccusative colleagues. Using pre-minimalist terms, the single argument of an unergative is a subject at both D-structure and S-structure. The subject of the verb ?kuì? «crawl» in (35a) has an Agent thematic role and its classification as unergative verb is also motivated by the fact that it is a motion verb that implies volitional act using Perlmutter semantic criterion. Let us observe the following data: (36) a. Ba?ka^m ts?ì Bakam PRS.give birth «Bakam gives birth» (37)a. mû w?ì gwi? child PRS4 laugh «The child is laughing» b. Ba?ka^m ts?ì mû mb?^ Bakam PRS.give birth child boy «Bakam gives birth to a boy baby» b. mû w?ì gwiÌ Ba?ka^m «*The child is laughing Bakam» 77 In the data above, the intransitive verbs n?ì ts?^ «to give birth» and n?ì wiÌ «to laugh» undergo transitivization. They become two-place predicates, that is, they select two arguments (36b and 37a) instead of one argument as required by their c-selection properties. This strengthens the idea according to which, both the lexical information about the verb in the lexicon and the structure around the verb determine the argument structure of a clause. Considering Perlmutter and Burzio's criteria, ts?ì and wiÌ would neither be unergative nor unaccusative. Although they can be used in a two-place predicate clause, the subjects of these verbs do not originate as their complements. This means that a two-way distinction of verbs that occur in intransitive clause cannot hold in Gh?maìlaì'. On the basis of their structural patterns, intransitive verbs can be split into three groups. The first group is made up of verbs that fail to be transitivized and therefore, are strictly intransitive such as ?kuì? «crawl» in (35). The second group comprises verbs which can be transitivized but their subject argument in the intransitive clause become the object in the transitive construction e.g p?ìpt?ì «spoil» (34). The third group gathers together verbs that also can be transitivized but their subject arguments remain subjects in both clauses such as ts?ì «give birth» and gwiÌ «laugh» in (37). This three-way distinction is attested in other African languages such as Ewe (Essegbey 1999:93) and Ga (Korsah 2011:47). 78 3.2.2 Transitive verbs There are some verbs in Gh?maìlaì' that cannot be used within intransitive clause. They are two-place predicates or bivalents. To put things in another way, they select two arguments to which they should assign two theta-roles. Structurally, one argument occupies the subject position within the clause whereas the other argument is the object as in the examples below: (38)a. Ta^laì k?Ì- toÌ bap Tala PST2- roast meat «Tala roasted meat» b. Ba?ka^m k?Ì-w?ì nw?ì ?j? Bakam PST2-PROG drink water «Bakam was drinking water» In the examples above, Ta^laì and bap are the arguments of the verb toÌ «roast» in (38a) meanwhile in (38b) the verb nwé selects Ba?ka^m and ?j? as arguments. The thematic roles of these arguments are determined by the semantics of the verb. In (38) above, the verbs assign Agent theta-role to their external arguments and Theme theta-role to their internal argument. The verb involve a meaning of a state of affairs in which the semantics of cause or control is asserted as well as the semantics of affect. Transitive verbs in Gh?maìlaì fall within the following semantic groups: 3.2.2.1 Semantic classification of Gh?maìlaì' transitive verbs ? Verbs of consumption These are verbs that designate the ingestion of something into an entity. The consumed entity may not be a physical substance. Gh?maìlaì' verbs of consumption include né miÌ «to swallow», né tsè «to lick», né tsû «to eat», né nw?^ «to drink», né ?i^? «to learn»...etc. (39) a. pó k?Ì -w?ì- dz?ì
kuÌmkuÌm b. Ba?ka^m l? -?iì? Gh?maìlâ' Bakam PST3-learn Gh?maìlaì' «Bakam learned the Gh?maìlaì language» ? Verbs of affect 79 These verbs designate an action that causes some sort of effect between the entities involved, one participant inflicting the effect on the other. They include né tû m «to hurt», né 3wâ «to kill», né t?à «to beat», né ne «to cook», né tô «to burn», né ke «to fry»...etc. (40) a. Ta^laì w?ì-ntoì m-?e^ Tala PRS4-burn PL-grass «Tala is burning grasses» b.Ba?ka^m kà-n?ì kùmkùm Bakam PST2-cook fufu «Bakam cooked fufu» > Verbs of contact These verbs refer to the physical contact between two concrete entities. Gh?maìlaì' verbs of contact include né t?àm «to hit», né sî2 «to crush», né kàm «to squeeze»...etc. (41) a. Ba?ka^m kà- k?ìm d?aÌp Bakam PST2-squeeze vegetables «Bakam squeezed vegetables» b. Màtwa^ là t?aìm Ta^laì car PST3-hit Tala «The car hit Tala» > Verbs of change of position They describe the action that leads a situation where one participant causes the other participant to be relocated. They include né là «to take», né pîû 2 «to lift», né t??Ì «to drag»...etc. (42) Ta^laì kà- là ?kaìp Tala PST2- take money «Tala took the money» 3.2.2.2 Theta role assignment within transitive clause Transitive verbs in Gh?maìlaì' may denote an action, a feeling or a perception. When they denote action, they usually establish a relationship of cause and effect between their two participants. In this case, the two arguments receive Agent-Theme theta-role, with the Agent 80 argument (i.e. subject) producing some sort of effect on the Theme argument (i.e. object). Let us observe the following data:
b. f?Ìfa? kwaÌ th?ì pa? wind PRS 1.carry head house «The wind has removed the roof» The sentence in (44a) may be uttered in a situation where Ta^laì is working with a knife and it cut him accidentally. In (44b), the inanimate f?Ì fa? «the wind» causes the removing of the roof. Korsah (2011:51) refers to the theta-role assigned to such arguments as Agent-like. I believe this is due to the fact that these arguments are not causer of the action described by the verb in a stricto sensu. In (44a) for instance, the knife itself is unable to cut something unless there is an external force that pushes the knife to realize the act of cutting. In this vein, the knife is an instrument use to accomplish the act of cutting and therefore should receive the theta-role of Instrument (Fillmore 1968:10). 81 A transitive verb denoting perception or feeling may assign Experiencer-Theme theta roles to its arguments. The following data illustrate this: (45)a. Ta^laì l? j?ì m-?wa??? mj?ì Tala PST3 see PL-book 3SG.4.POSS «Tala saw his books» b. fiìpa? ??Ìm Ta^laì fever PRS 1.catch Tala «Tala has suffered from fever» In (45a), the Experiencer theta-role is assigned to Ta^laì, the entity that perceives while m?wa??? mjé «his books» which is the perceived objects receive the Theme theta-role. In (45b), the verb evokes a kind of feeling in its Experiencer argument Ta^laì. Adopting Chomsky (1995)'s proposal, according to which the ?P-VP configuration expresses the agentive or causative role of the external argument, I argue that external argument (EA) bearing the Agent theta-role is base-generated under Spec-?P meanwhile the one that has been assigned other thematic roles such as Theme, Experiencer or Instrument is purely merged in Spec-VP as illustrated below: (46a) TP
DP T' T ?P DP ?' EA ? VP
<Agent> V DP Internal argument In (46a), the external argument is pure merged under the specifier position of the light verb phrase. It is within this position that the Agent theta-role and it later moves to the specifier position of TP to satisfy the EPP requirement of the tense phrase. (46b) TP
DP T' T ?P DP ?' ? VP DP V' EA V DP <TH, EXP, INST> Internal argument In (46b), the external argument is base-generated in the specifier position of VP in accordance with the VP internal subject hypothesis as formulated in Koopman and Sportiche (1991). It raises later to the specifier position of ?P in order to be in the domain of Probe for holding of the operation Agree. It is in this position that the uninterpretable features of the external argument will be checked and valued. After this, it moves to Spec-TP to satisfy the EPP requirement of TP. 3.2.2.2.3 verb-complement semantic relation Looking at their semantics, it seems that the kind of complement that transitive verb in Gh?maìlaì' selects is affected by the conceptual semantic specificity that is attributed to the verb by the speakers. Most of the transitive verbs in Gh?maìlaì'ì can select a generic-meaning DP as object. They can occur with semantically vague complements such as jwé «thing» and moÌ «person», and still keep their core meaning as shown in the examples below: (47) a. Ta^laì k?Ì- pfâ mko?l?^siÌ Tala PST2-eat rice «Tala ate rice»
(48)a. Ba?ka^m k?Ì-t??ì poì Bakam PST2-beat PL.child «Bakam beat the children» b. Ba?ka^m k?Ì- t??ì pu?siÌ Bakam PST2-beat cat «Bakam beat the cat» c. Ba?ka^m k?Ì-t??ì moÌ 82 d. *Ta^laì k?Ì-pfâ d. *Ba?ka^m k?Ì-t??ì 83 Tala PST2-eat Bakam PST2-beat In (47 and 48), the verbs have a less variable meaning even when they occur with different objects. This shows how much these verbs are meaningfully independent. When the verb is used with the generic objects jwé (47c) and moÌ (48c), speakers subconsciously understand what the verb selects as complement. Moreover, (47d and 48d) show that such verbs cannot occur without their complements. This is an indication that the verbs involved are strictly transitive. There are some verbs which do not take generic objects. They rather select particular objects. Accordingly, Essegbey (1999:191) argues that verbs with more specific meaning occur with complements that have more general meaning while verbs which have less specific meaning occur with complements that have more specific meaning. Verbs requiring more specific complement are referred to as Inherent Complement verbs. These are some examples: (49) a. Ba?ka^m k?Ì-w?ì d3?ì
goì? b. Ba?ka^m k?Ì- w?ì d3?ì ?w?ì Bakam PST2-PROG see moon «Bakam was menstruating» In (49), the verb has a variable meaning when they occur with different objects. There is a semantic asymmetry between constructions in (47and 48) and those in (49). Verbs in (47and48) have their meaning specified in them whereas those in (49) have their meaning specified outside them. These verbs will be deeply discussed in the following chapter. Moreover, some transitive verbs have several forms, that is, different words are used to express the same reality. They are synonymous and are in complementary distribution. For instance, the act of eating in Gh?maìlaì' is expressed by two words tsé and pfáì. The latter is used for chewable and dry food while the first is used for wet and soft food that can be directly swallowed as shown below: (50)a. Ta^laì k?Ì-ts?ì kuÌmkuÌm c. Ta^laì k?Ì-pfâ bi^jé Tala PST2-eat fufu Tala PST2-eat groundnut «Tala ate fufu» «Tala ate groundnut» b.*Ta^laì k?Ì-ts?ì bi^j?ì d. *Ta^laì k?Ì-pfâ kùmkùm Tala PST2-eat groundnut Tala PST2-eat fufu «Tala ate groundnut» «Tala ate fufu» We observe that, in (50), each word selects a specific type of food. In the same vein, the act of harvesting is expressed by several words namely t?wà, kelp, kém and t?w5 2. t?wà is used for plants such as maize, banana tree, palm tree...etc, kelp goes with fruit trees and vegetables, kém is used for groundnut and t?w3 2 for bean and soya as illustrated in the following examples: (51)a. Ta^laì k?Ì-t?waÌ ?kàdeì Tala PST2-harvest banana «Tala harvested banana». b. *Ta^laì kà-t?waÌ pja
84 After having discuss intransitive and transitive verbs, the following section tackles ditransitive verbs in Gh?maìlaì'. 3.2.3 Ditransitive verbs Ditransitives can be defined in terms of theta-roles, namely as verbs that assign three semantic roles, usually one external (Agent) and two internal (Theme, Goal or Recipient). The grammatical functions of these arguments are subject, direct and indirect object. Such verbs are basically verbs of transfer as exemplified in (52) and may be used as in (53): (52) nâ t?jaÌ «to send» nâ la^?â «to teach» nâ ha^ «to give» nâ la^?tà «to show» nâ si^? «to say» nâ jô «to buy» ì n? và «to write» nâ ni^? «to put» (53) a. Ta^laì l?-và ?wa??? bi^ taìp e Tala PST3-write letter to father 3SG.1.POSS «Tala wrote a letter to his father» b. Ba?ka^m k?Ì-t?jaÌ ?kaìp bi^ pô
pjâ «Bakam sent money to her children» 85 c. Ta^laì w?ì-daì??ì ??maìlaì? bi^ po^ paìp Tala PROG-teach gh?maìlaì' to PL.child 3PL.2POSS «Tala is teaching their children Gh?maìlaì'» As it is shown in the data above, most of the ditransitive verbs in Gh?maìlaì' imply transfer but the kind of transfer involved may not obligatorily be a handing over of a tangible entity from one participant to another. In (53c) for instance, ??maìlaì?, an abstract DP is transferred from Ta^laì to po^ paìp «their children». Furthermore, the third argument to which verbs assign the Goal or Recipient theta-role is introduced by the preposition bi^ «to» in these examples. One can therefore infers that Gh?maìlaì' ditransitive clauses are instances of dative complementation since one of the two required by the verbs as complements is introduced by a preposition. In this vein, the Thematic Hierarchy is Agent >Theme > Goal/Recipient. This morphosyntactic feature distinguishes Gh?maìlaì' from other Bantu languages such as Tuki (Biloa 2013) wherein ditransitive clauses are cases of double object constructions; there is no element that occurs between the two complements since the language itself is devoid of prepositions but an applicative morpheme to mark applicative construction is attested in this language. Let us observe the data below: (54)a. Ta^laì k?Ì-siì? é nw? Tala PST2-tell 3SG.ACC affair «Tala told her something».
d.*Ta^laì l?-v?Ì taìp e ?wa??? Tala PST3-write father 3SG.1.POSS letter «Tala wrote his father a letter» e. *Ba?ka^m k?Ì-t?jaÌ po^ pj?Ì ?kaìp BakamPST2-send PL.child3SG.2.POSS money «Bakam sent her children money» The examples in (54) above suggest that there are two groups of ditransitive verbs in Gh?maìlaì'. The first group is made up of verbs that their third arguments are obligatorily introduced by a preposition, namely t?jaÌ «send», v?Ì «write», joì «buy»...etc. Verbs that their second objects may not necessarily be introduced by a preposition as in (54a-c) above constitute the second group. This distinction can be captured as follows: (55) a.1st Group t?jaÌ «send», v?Ì «write», joì «buy»...etc. b. 2nd Group siì? «tell/say», laì?t?Ì «show», laì??ì «teach»...etc. DP DP PP # # # Agent Theme Goal/Recipient Subject Direct Object Indirect Object DP DP DP # # # Subject Object Object 86 The structure in (55) has received several attentions in the literature since Larson (1988). Larson proposes a double VP structure that accommodates the two internal arguments of the verb. Within this structure, the higher VP head is empty (e) and the lower VP is head by the verb. The specifier position of the lower VP being filled by the direct object, the external argument occupies the specifier position of the higher VP. The lower V selects the PP and later moves to the higher V position. From this rationale, he proposes a derivation of double object constructions in English which is represented below: (56) a. VP Spec V' John V VP e DP V' a letter V PP < send > to Mary b. VP Spec V' John V VP 87 e DP V'
Larson argues that (56b) is derived from (56a). The case of the indirect object is absorbed (hence no to) as well as the theta-role of the direct object a letter (hence its adjunct status). Mary moves to the higher DP position and the verb send to the higher V in order to assign case to Mary. Case assignment to a letter is done by the reanalysis of the lower V' as a V which can then assign case to a letter. Baker (1997) provides an alternative analysis for (56b). He agrees with Larson that the indirect object moves to a higher position. However, For Baker, this higher position is the specifier of an inner Aspect Phrase which is projected between the two VP and the preposition incorporates into the verb. The incorporation of the preposition is not clearly justified as well as the case marking of the indirect object. Elly Van Gelderen (2013) building her analysis on Oehrle (1976), provides a minimalist analysis that accounts for the alternation in (57) below: (57) a. Ta^lá k?Ì-láp?ì Ba?ka^m ??málá' Tala PST2-teach Bakam Gh?málá' «Tala taught Bakam Gh?málá'» b. Ta^lá k?Ì-láp?ì ??málá' bi^ Ba?ka^m Tala PST2-teach Gh?málá' to Bakam «Tala taught Gh?málá' to Bakam» In (57), there is a meaning difference between the DP _DP (57a) and DP_ PP (57b) constructions. In (57a), Ba?ka^m actually learns some Gh?maìlaì' but she probably did not in (57b). This difference is in terms of the affectedness of the Goal. In (57a), Ba?ka^m is affected by the action denoted by the verb and therefore is an Experiencer. In terms of Thematic Hierarchy, the Experiencer is higher than the Theme whereas the Theme is higher than the Goal (Pesetsky 1995). The two structures are represented as follows: (58) a.
88 b.
Van Gelderen analysis of DP DP complement seems to be more straightforward and economic than early analysis (Larson 1988 and Baker 1997). |
|
