The morphosyntax of adverbs in Shupamem
par Abass NGOUNGOUO YIAGNIGNI
Université de Yaoundé 1 - Master en Linguistique Générale 2016
This chapter aimed at presenting a grammatical sketch Shupamem. I presented the synthesis of some grammatical aspects of Shupamem. In the light of the previous studies made on the language, I presented the consonants, the vowels and the tones of Shupamem. In the same line, Ipresented the fifteen noun classes of Shupamem, the personal, the demonstrative and possessive pronouns, the adjectives, and the articles. Furthermore, I discussed verb tenses, aspects and moods. Finally, I discussed the basic sentence structure of Shupamem.
The aim of the previous chapter was to present the grammatical sketch of Shupamem, in order to familiarize the reader with the functioning of the language under study. As for this chapter, it presents the frameworks adopted for the study. These are the Minimalist Program of Chomsky (1993, 1995...) and the Cartography of Rizzi (1997). The main objective of this chapter is to identify and highlight relevant aspects of MP and Cartography which are adopted in the analysis of adverbs in Shupamem. Finally, I present silent works done on adverbs on various perspectives.
The Minimalist Program is a line of thought that has been developing in generative grammar since the early 1990s. It was initiated by Noam Chomsky and is presented by the latter as a program which aims at minimizing the mechanism of description of language phenomena. In fact, MP demands description and most importantly explanation, and it aims at achieving descriptive and explanatory adequacies. It renders simple the linguistic system, through economy both in derivation and representation.
Given that MP is dynamic and universal, it would really be interesting and necessary to test its assumptions against the data from all languages. In this perspective, I think that testing its assumption against the data from Shupamem will be a contribution to the development of the theory.
Furthermore, besides some works done within the generative approach, namely Ondoua (2004), Nchare (2005) and others, it is necessary to extend the research frontier and broaden the syntactic research by undertaking a minimalist study of the language.
As its name implies, the Minimalist Program is a linguistic theory that minimizes the mechanism of language description as much as possible. It seeks to achieve descriptive and explanatory adequacies and most importantly, to ease the language learnability. It comes in as the solution to the lapses of the previous theories that were used in linguistics. In fact, these theories laid much emphasis on language description with very little focus on the explanation of the language phenomena. In the same perspective, previous frameworks displayed an uncountable number of rules that, instead of easing the task, rather made language learnability much complicated. Therefore, some rules in the previous frameworks (in particular, Government and Binding Theory, the Principle and Parameters Theory) have undergone some reconstruction alongside various linguistic phenomena.
MP is centered on the Principle of Economy. In fact, it assumes that one should reduce unnecessary elements from the computational process so as to make the mechanism easy and to ease learnability. This goes in the same line with the Government and Binding Theory, from which it drew inspiration, though a radical change exists between them. In fact, MP advocates for Economy and Principle of Full Interpretation (PFI). The latter claims that no redundant elements, whether semantic, phonological or syntactic, should be included in a structure. Each element should be interpretable and play a given role.
The difference between MP and GB is appraised at the levels of grammatical representation that they display. In fact, GB has four different levels of representation, which are Deep Structure (DS), Surface structure (SS), Logical Form (LF) and Phonological Form (PF). At the level of DS, positions should be filled only if they are semantically active. As for SS, it is the level of representation in which the derivation splits, sending off one copy to PF for phonological interpretation, and another copy to LF for semantic interpretation.
As far as LF and PF are concerned, they are two interfaces which the sentence should satisfy in order to be grammatical. In other words, LF checks the grammaticality of the sentence at the semantic level, whereas PF does that at the phonological level. Within MP, the levels of representation have been reduced into two, (LF, PF), making easier the process of the Computation of Human Language, (CHL).
Within MP, CHL calls in a lexicon (lexical array) from which elements are selected to build the numeration. These elements merge externally the ones with the others to build the syntax, within which another merge operation, internal merge is applied. Internal merge is concerned with movements (copying, raising). From the syntax, we spell out the previously merged elements to the interfaces (LF, PF) for interpretation. This process is presented in the diagram below:
LEXICON NUMERATION SYNTAX (Spell out)
(Select) External Merge (Internal merge)
Fig5, the Computation of Human Language within MP
Out of the above listed changes, much has been brought into the linguistic analysis of the language by MP. Among others are the following principles:
a) Least Effort: also known as Last Resort principle, it stipulates that one should avoid movement throughout the computation. That is, there should be as few movements as possible.
b) Procrastinate: it stipulates that one should not move overtly, unless movement is imposed by some principle of Universal Grammar, (UG).
c) Greed: (do not move X unless X bears a feature that satisfies this movement). This strengthens the significance of Agreement within MP. In fact, for an element to undergo movement, its features should be checked, matched, valued and deleted.
d) Minimize chain movement: movement should be as shorter as possible. Here, long distance movements are to be avoided.
e) Relativized Minimality, RIZZI (1990: 7).
X á governs Y iff there is no Z such that:
(i) Z is a typical potential á governor for Y and
(ii) Z C-Commands Y and does not C-Command X.
Here, movement should be the nearest one to the landing site of the moved element; no identical element should be found between the probe and the goal, in order to avoid obstruction.
Other innovations brought in by MP concern representation. In fact, as mentioned earlier, the Principle of Economy is the guideline followed by MP. Thus, unlike the preceding frameworks wherein one could include traces in the structures, MP advocates for their exclusion. Given that traces are not present in the numeration, their presence in the syntax violates the Inclusiveness Condition.
Furthermore, within MP, phrase markers are binary branching, whereas the X-bar Theory could make use of unary branching or have as many branches as possible. In the same perspective, the privilege is given to the bottom-top merging fashion than to the top-bottom fashion, as was the case within the X-bar Theory.
In brief, the Minimalist Program has brought some amelioration to the previous frameworks, and seeks to achieve descriptive and explanatory adequacies. Most importantly, it aims at easing the learning process of human languages since it discards non-relevant elements and keeps only those that are relevant to the machinery of language analysis.
Cinque and Rizzi (2008) argue that «The cartography of syntactic structures is the line of research which addresses this topic: it is the attempt to draw maps as precise and detailed as possible of syntactic configurations. Broadly construed in this way, cartography is not an approach or a hypothesis: it is a research topic asking the question: what are the right structural maps for natural language syntax?»
According to them, this approach aims mostly at bringing out the right map of the syntactic elements in natural languages.
Quoting Shlongsky (2010), aspects of Cartography have been perceived in the works of Bernicaì (1988), Pollock (1989) with the split-IP Hypothesis, and Cinque (1990). But what can be considered as the first explicitly cartographic study is Rizzi (1997).
In fact, Rizzi (1997) studies the mapping of the elements above TP, that is, the elements of the left periphery. He proposes that fronted topics and foci are articulated as projections of Topic and Focus heads, as contrary to the traditional view wherein all fronted elements should be hosted by CP. Here, instead of allowing recursive CPs in multiple raising situations, the Cartographic approach advocates for the splitting of CP into many functional heads. The resulting functional projections are ForceP, FocP, TopP, AgrP, etc...
Out of the above listed works on Cartography, many other works have been done within the same approach. Among others are Cinque (1999, 2002), Beletti (2004), Rizzi (2004), Benincaì (2001, 2006), Benincaì and Poletto (2004) Cinque and Rizzi (2008), Biloa (2010) and others. The work of Cinque (1999) is mostly the one that concerns adverbs.
In Adverbs and Functional Heads: A Cross-linguistic Perspective, Cinque (1999) posits that adverbs occur in a fixed order in all the languages. He proposes that each adverb should occur at the specifier position of the various functional projections. These functional projections are the Mood (Mood-), the Modality, (Mod-), the Tense (T-), and the Aspect (Asp-). He proposes the following scheme for English adverbs to account for his view:
Frankly Mood-speech act>FortunatelyMood-evaluative>Allegedly Mood-evidential>Probably Mod-epistemic>Once T (Past) [Then T (Future)>Perhaps Mood-irrealis>Necessarily Mod-necessity>Possibly Mod-habitual>Again Asp-repetitive>Often Asp-frequentative>Intentionally Mod-volitional>Quickly Asp-celerative>Already T (anterior)>No longer Asp-terminative>Still Asp-continuative>Always Aspect-habitual>Just Asp-retrospective>Soon Asp-proximative>Briefly Asp-durative>Characteristically As-generic/progressive>Almost Asp-prospective>Completely Asp-Sg.Completive (I)>Tutto Asp P1Completive>Well Voice>Fast/early Asp-celerative (II)>Often Asp-frequentative (II)>Completely Asp-Sg.Completive (II)
Source: Cinque (1999:106)
Though the structure above is not systematically the same in all languages, Cinque's view is that all the languages have somehow a fixed hierarchy in which adverbs should appear.
There is tendency to consider the Cartographic Approach as a contradiction to the Minimalist Program. In fact, while MP seeks to minimize the language mechanism, the Cartographic Approach seeks to draw the maps of structures of the natural languages. It should be noted that this does not make Cartography an opposition or an alternative to Minimalism. On the contrary, as posited by Shlonsky (2010), the feature-driven approach to syntax, the reliance on simple operations such as Merge, Project and Search pave the way to the Cartographic enterprise whose goal is to draw up a precise inventory of features and discover their structural relations. In that same view, Cinque and Rizzi (2008) clearly argue:
«We believe that there is no contradiction between these two directions of research, and the tension, where real, is the sign of a fruitful division of labor. Minimalism focuses on the elementary mechanisms which are involved in syntactic computations (...) and cartography focuses on the fine details of the generated structures, two research topics which can be pursued in parallel in a fully consistent manner, and along lines which can fruitfully interact.»
According to them, MP is centered mostly on language computation which it seeks to make easier. As for the Cartographic enterprise, its aim is to draw the details of the structures of the languages. In clear, they apply to different domains.
Quoting Tabe (2015), adverbs have been treated as the least homogenous category to define in language because their analysis as a grammatical category remains peripheral to the basic argument structure of the sentence. Adverbs have been analysed as predicates (Roberts 1985; Rochette 1990), as arguments (McConell-Ginet 1982; Larson 1985), as modifiers (Sportiche 1988), and as operators. Several reasons account for this lack of clarification.
The first is attributed to the fact that adverbs do not present a homogeneous class. Givón (1993:71) sees adverbs as least homogeneous and the hardest to define. According to Payne (1997:69) any word with semantic content (other than grammatical particles) that is not clearly a noun, a verb, or an adjective is often put into the class of adverbs. In the same light, McCawley (1996:664) observes that the diversity of things that adverb has been applied to is in keeping with traditional definitions of it as modifier of a verb, an adjective, or another adverb, which in effect class as adverbs all modifiers other than adjectives. Adverbs cannot be declined and they are often grouped with prepositions and conjunctions as a subgroup of particles. This explains why they form a very heterogeneous group containing numerous overlapping with other grammatical categories.
Secondly, because adverbs demonstrate a correlation between syntactic and semantic structures, the behavior of adverbs has been analysed as inextricably bound to both syntactic and semantic phenomena (Tenny 2000:285-6). However, the analysis of what constitutes a syntactic or semantic underlying representation of adverbs in a sentence structure is unclear. In order to understand the nature of the interface between them, there is need to identify the syntactic or semantic elements necessary in explaining the distribution and properties of adverbs. Different approaches have been adopted for the classification of adverbs. One approach identifies them into distinct groups constrained by their syntactic and semantic properties.
Advocates of this line of thought (Jackendoff 1972; Travis 1988, etc.) posit that various types of adverbs may select for propositions, speech acts or events, each of which interacts with syntactic principles to produce different adverbial behaviours. The analysis supposes that the nature of the syntactic constituent that licenses the adverb determines its semantic interpretation. The latter is obtained given the semantic features associated with the adverb. In Jackendoff's (1972) analysis, adverbs are semantically classified into four groups. These comprise the speaker-oriented adverbs; subject-oriented adverbs; event-related adverbs and focus adverbs.
The speaker-oriented adverbs such as frankly, unfortunately among others carry information relating to the speaker. Subject-oriented adverbs (including clumsily, carefully...) introduce material relating to the subject of the clause. Event-oriented adverbs comprising manner, time and degree adverbs (like completely, frequently and eloquently...) introduce material relating to the event structure. Lastly, focus adverbs (including almost, merely, utterly...) introduce material which is discourse-oriented for scope purposes. The syntactic distribution of these adverbs relative to the hierarchical constituent structure shows that subject-oriented and speaker-oriented adverbs are sentence-level adverbs, while the event-related adverbs are verb phrase-level adverbs. Focus adverbs, in contrast, are hosted by the Aux. Head, a position dominated by the Aux. node. Travis (1988) fine-tunes the nature of the mapping between the semantic and syntactic composition of Jackedoff's adverbs by suggesting that the speaker-oriented adverbs take scope over CP, the sentence adverbs take scope over IP, the subject-oriented adverbs take scope over INFL, and the event-oriented adverbs take scope over the verb.
Another approach put forward to capture the cross-linguistic generalization on the distributional properties of adverbs is that of Cinque (1999). Given Cinque, there is no direct one-to-one correlation between the syntactic and the semantic composition of the adverbs. Thus the relation between the syntactic position occupied by an adverb and the semantic role discharged by the latter remains essentially non-compositional.Rather, emphasis is on teasing out the distinguishing syntactic properties of each adverb by showing associated positions of each with respect to a distinct functional projection. Recourse to the semantic contribution of adverbs on the syntax is captured indirectly. The adverbs types and their semantic properties are mirrored from an inventory into the various functional projections in the syntax.
Tenny (2000:290) adopts an approach that treads a middle ground between the views that have been projected above (that is, whether there is a direct mapping between semantic/syntactic composition or just a syntactic projection of functional heads with an indirect link to its semantic properties) in determining the distribution of adverbs. Tenny maintains that the semantic composition of the event is mediated in the syntax by a relatively small inventory of functional projections mirroring that composition. If one's observation is right, Tenny's treatment of adverbs is in consonant with that projected by Jackendoff (1972) and Travis (1988) earlier indicated. However, Tenny focuses more on elements lower down in the semantic composition of the clause. In particular, the event structure closer to the verb and internal to the event, rather than issues that appear at the higher level of the clause structure like speech acts, propositions, among others. As for the phrasal projection of adverbs, the literature supposes that adverbs can occupy adjoined positions (Ernst 1997), specifier positions (Laenzlinger 1993; Cinque 1999), can self-project into a maximal projection (Pollock 1989), and as being defective categories without a maximal projection (Travis 1988).