SENTENCE PATTERN AND USAGE IN NIGERIA ’ S 2015 PRESIDENTIAL DEBATE

This study of sentence pattern and usage in Nigeria’s 2015 Presidential Debate identifies and accounts for all occurring major and minor sentences, classifies the major sentences into simple, compound and complex subtypes, determines their typological and thematic distribution, and demonstrates how they were strategically employed to articulate each debater’s points. The data comprises 1876 sentences and was analysed using an improved version of the systemic grammatical model. Results show that major and minor sentences represent 92 and 8 per cent respectively and that while the simple sentence accounted for 77 per cent, compound and complex sentences make up 10 and 13 per cent respectively. Mean length of sentence was 13.6 words and clause density in respect of compound and complex sentences was 1.6 and 2.22 per sentence. Of the seventeen variants of the simple sentence isolated 21 per cent had their elements of structure realised by rankshifted clauses while 20 per cent were affected by multiplicity, mobility and inversion. Though the rest 59 per cent were kernel sentences of the basic SPCA structural pattern, it was not uncommon to find structurally complex groups as elements of clause structure. What determined which of alpha or beta was clause‐initially was the focus of the message conveyed. So thematic fronting is not limited to the single clause sentence. Sentence length and type, positioning of clauses or parts thereof, and decisions on conjunctions and finite or non-finite clauses, were greatly governed by theme and the speaker’s mediate goals and grammatical sophistication. These are proofs of the strategic use of the sentence by politicians.


INTRODUCTION
The initial motivation for this study came from a recently concluded one on the central pronouns in Nigeria's 2015 Presidential Debate (Adejare, 2018).While working on that paper it struck me that the twenty-three odd pronouns with a combined frequency of 2409 could not have been effective on their own in coding debaters' responses outside the wider context of the sentences in which they functioned.The focus then was on the strands, or, indeed, part thereof, that make up the whole.In the current study attention is paid to the whole itself, the syntactic unit sentence.A second factor is the revelation, following a critical examination of existing literature on political texts, of a paucity of studies specifically devoted to syntactic structure generally and the sentence particularly.What exist are occasional references that are excusably fragmentary and shallow because that is not their focus (cf.Ayeomoni, 2001;Opeibi, 2010;Bellova 2012;van Leeuwen 2012).This study is therefore holistic.It is concerned with how the largest grammatical unit syntactic patterns' formal and functional manifestations were employed to achieve debaters' goals.The sentence in political texts is not known to have been comprehensively analysed as envisaged in this study, whereas politicians have been known to exploit syntactic structures to a great personal advantage (van Dijk 1979).
This study therefore takes the grammatical analysis of political texts one step further by examining how the sentence and its formal and functional features were used to achieve debaters' goals in Nigeria's 2015 Presidential Debate.It aims specifically to identify and formally account for all occurring sentences in the corpus using the major/minor sentence dichotomy, classify all the major sentences into the structural subtypes of simple, compound and complex, determine the frequency of occurrence and thematic distribution of each structural type and subtype, and demonstrate how debaters strategically employed the different sentence types and patterns to articulate their policies and subsequently actualise their goals of convincing the electorate to vote for them.It is hoped that its outcome will fill some of the existing gaps in the syntax of political texts and stimulate further research on the subject.

REVIEW OF LITERATURE
Previous Studies Millar and Curie (1972) reveal that the normal SPCA pattern of the English clause structure can indeed be varied for the purpose of achieving rhyme in poetry and for particular effects in prose.On prose, they particularly demonstrate that the complement can be moved to initial position for emphasis, the subject and complement can be inverted in order to provide logical connection between two sentences or even paragraphs, the adjunct can be transferred from sentence end to execute both thematic and structural functions, and, where it is a single adverb, its positioning can alter the rhythm of the sentence with implications for meaning.Furthermore, they provide textual examples to show that alterations in the normal ordering of the elements of clause structure, including subject-predicator inversion, have a controlling effect on the reader/listener's attitude towards the rest of the clause.Although the current study is neither based on spoken prose nor prose as it is generally known, it will nevertheless benefit from these insights.
In his paper articulating the subject political discourse van Dijk (1979: 37) states that syntactic style can be politically manipulated but in more subtle and less obvious manner compared to lexical style and points out that variations of word order, the choice of specific syntactic categories, pronouns, clause embedding, and sentence complexity etc. are used to express underlying meanings in sentence structures.Wilson (2005:402) similarly highlights three ways in which syntax impact political texts as using specific syntactic transformations, employing relatively uncontroversial structural selections such as the use of nominalised clauses to cause attention to be shifted away from what is not desired, and the relative distribution of particular syntactic selections such as negation.Van Leeuwen (2012) adopts a stylistics approach to the analysis of political texts consisting of a check-list of four categories, namely lexical, grammatical, figures of speech, and context and cohesion.What is of interest here, which makes the present study all the more compelling, is his decision not to focus on the grammatical because it is the relatively infrequently analysed category in political texts.However, to demonstrate the importance of grammatical phenomena and show that complementation can add to the inventory of linguistic tools relevant to Critical Discourse Analysis (CDA), van Leeuwen compares Geert Wilder's 2007 speech contributing to the debate on 'Islamic Activism' with Vogelaar's.He reveals an average sentence length of 15.4 and 19.8 words per sentence (w/s) respectively and adds that where complementation constructions are present, they have rhetorical effects that enable the speaker to describe their viewpoint as their perspective, leaving room for discussion and negotiation, and that where complementation constructions are lacking, however, ideas are primarily presented as facts, leaving minimal room for negotiation and discussion.These go to show how syntactic simplicity or complexity impact political texts.Cienki (2009:8-9) compares two US vice presidential candidates' use of spoken language framing devices in the 2008 debate and reports that while Biden's argument structures conformed more to written discourse than to spoken language because of multiple embedding of clauses, Palin's syntax was characterised by simplicity as short sentences interlaced with repetition in a 'chant-like' manner to achieve logical connection between ideas.Bellova (2012) analyses political crisis speeches of ten American presidents at the syntactic, semantic and pragmatic levels and concludes that war discourse syntax is getting simpler with the sentences getting shorter and less complex.A mean length of sentence of 21, 16, and 20w/s was recorded for Obama, Bush, and Reagan as opposed to the 29, 39, and 35 for Lincoln, McKinley, and Wilson respectively.There was an even distribution of coordinated and subordinated sentences in Obama's speech whereas Reagan's was predominantly coordinated; and, while Truman's sentences lacked 'syntactic sophistication and complexity' due to the absence of inversion, finite clauses, and appositive phrases etc. which make discourse advanced, Roosevelt produced 'powerful' sentences with reversed word order.
The only syntax-related analysis in Astrero's (2017) CDA study aimed at identifying the personal pronouns, modals and mood of the sentences is the sparse information on the 89.47, 10.53, and 0 per cent proportion respectively of declarative, subjunctive and imperative mood in the corpus.Similarly, Abdel-Moety's (2015:5) observation that (Hillary) 'Clinton's language, is in general, not simple.Her sentences are long.Most of her sentences tend to be compound or complex rather than simple' is the only syntax-related statement in a work that fails to state the number and type of sentences produced possibly because the focus was on transitivity, modality and textual analysis.In Akinwotu's (2013) Speech Act analysis of acceptance of nomination speeches, no grammatical analysis of the eighteen extracts examined was undertaken.Finally, based on the title of Kenzhekanova's (2015) article one expects to find emerge from some form of analysis a list of linguistic features (including syntactic) that characterise political texts, but none can be pinpointed.
The foregoing reveals the absence of a rigorous analysis of the sentence and of syntactic structure in political texts aimed at their characterisation, which provides further justification for the current research effort.

Theoretical Framework
This work is predicated upon an enhanced systemic grammatical theory because it provides the best framework that can be improved to more accurately analyse the syntax of the Presidential Debate texts.An off-shoot of Firthian integrationist functionalism, Systemic Grammar (SG) is concerned with the discovery of the nature and structure of linguistic form and its functional properties.As a rule functionalists generally describe linguistic data in relation to its various functions at both the formal and textual levels.Systemic Grammar thus studies language as part of the social process and argues that it is the external relations between a language and the individual user or community that gives linguistic signs their shape and makes them recognisable in the first instance.It gives prominence to both meaning and context (Firth 1951;Halliday 1961Halliday , 1985;;Berry 1975;Halliday & Matthiessen 2004), and insists on an intrinsic integration of linguistic signs with their functions which generate formal meanings and therefore enables one to theoretically deal with language as a form of human behaviour (Harris 1990).Consequently, SG accounts for more elements of natural language than other models and is therefore able to account for more formal functions.
Systemic Grammar correctly recognises form and situation as the two primary elements of language structure, but following Saussure (1916) it claims that situation is not part of linguistics (cf.Halliday et.al., 1964:18).The absence of situation creates descriptive limitations, such as the lack of capacity to more accurately analyse the formal element of language beyond the clause.This study is based on the fact that form and situation are the primary elements of language structure and that both determine the structure of a text's form in language use.The primary function of language is text message coding, where the message of an encoder to a decoder determines a text's thesis, code (actual grammatical forms in a text), culture and date, all of which manifest in a text's actual form such as the Presidential Debate data here (cf.Adejare,0.,forthcoming).This work examines the occurrence of syntactic structure in a particular variety of language use; it is concerned with discovering the underlying grammatical form of the sentence used in political texts.An improved version of the systemic grammatical theory will therefore more adequately handle the data.

Grammatical Model
Because SG has the most accurate clause structure formula (S) P (C) (A) (Halliday (1961;Halliday et al. 1964), it is better suited to the problem of handling new data (Coulthard & Montgomery, 1985:7), theoretically fuller and more coherent (Butler, 1979:72), and therefore gives a clearer and fuller account of grammatical form and function.The newer model Systemic Functional Grammar (SFG) is theoretically formulated to solve the problem of 'texts' (Halliday 1966(Halliday , 1970(Halliday , 1985;;Morley 1985;Bloor & Bloor, 1990;Halliday & Matthiessen 2004;Eggins 2004), which is still being grappled with (Halliday &Webster 2008) and which lies outside the scope of this study.Not only does SFG not recognise the grammatical unit sentence, it has replaced it with 'clause complex' and consigned it to written language (Halliday, 1985:192-193).The notion of 'clause complex' cannot permit a full account to be made of all sentence types outside 'simple'.In his 1988 review of Halliday's An introduction to functional grammar, Huddleston notes that 'the shortcomings of rank-based constituency are indeed more apparent in the current functional grammar than in the original scale-and-category model' (p141), the 'multi-dimensional analysis' of the clause (textual, interpersonal and ideational) does not satisfactorily describe its grammatical properties (p157) and that the lack 'of clear grammatical criteria will make the analysis difficult to apply'.(p173).It is our strong conviction that the adequacy of a linguistic model lies not in its name or philosophical claims or age but in its ability to describe natural language most accurately and most efficaciously.The earlier version is therefore the grammatical model chosen.
Because of the limitation noted above, while SG's syntactic description is the most accurate among its rivals, it requires improvement to adequately account for the syntax of language use data.The largest syntactic unit is the sentence which is made up of one or more clauses.A clause is made up of one or more groups and a group is made up of one or more words.The word is made up of one or more morphemes.That a sentence is made up of one or more clauses generates three types of sentences -a simple sentence made up of one alpha (independent) clause, a compound sentence made up of more than one alpha clause, and a complex sentence made up of at least one alpha clause and one beta (dependent) clause.The clause is therefore the primary element of sentence structure.Whereas the alpha clause is obligatory, the beta clause is optional.The structure of the sentence is represented formulaically as S (β 0-n ) α 1-n (β 0-n ).Based on its function in texts, major and minor sentence types are also recognised.The major sentence is that clause with a predicator as an obligatory element of structure, while the minor sentence, also called verbless clause (Quirk et al. 1985: 996-997), is that clause that lacks a predicator (cf.Adejare & Adejare, 2006: 49-67;94-96).
Of all the existing clause structure formulas SG's four-element (S)P(C)(A) is the most accurate as earlier stated.But the situation functions of language (Adejare & Adejare, 2006: 86-92)  Thus, apart from recognising simple, compound and complex sentences as well as major/minor sentences, I will recognise multiplicity, mobility and inversion of some clause structure elements in the data coded by politicians.I will ignore the situation-induced moves among the moderators who asked the questions that each candidate answered because they have no major impact on the syntactic forms of the answers to the questions.

METHOD Data Base
The text of the transcript of the 2015 Presidential Debate is the data base for this study.It was downloaded using https:// youtu.be/MGwkZr3.

The Debaters
Twelve flag-bearers of twelve political parties participated in the two-session presidential debate, namely Tunde Anifowoshe Kelani (Action Alliance: AA), Alhaji Ganiyu O.

The Questions
The questions responded to covered eight themes as listed below (Each debater had three minutes to respond to each question and one to round off): Q1 (Motivation): 'Why do you want to be president of Nigeria?' Q2 (Power): What do you intend to do (.) to guarantee steady power supply in Nigeria?' Q3 (Oil): 'How do you intend to tackle the problem of (.) oil theft?' Q4 (Corruption): 'What will be your template for tackling corruption?' Q5 (Security):'What strategic policy direction would you put in place to guarantee the security or safety of life and property of Nigerians, to rehabilitate the victims of insurgency in the North East, reconstruct the zone and eventually bringing back the Chibok girls?' Q6 (Education): 'What are your immediate and long-term plans to restore dignity to our education?'Q7 (Agriculture): 'What specific plans do you have (.) to address the issue of agriculture in such a way that it will be useful to industries and also bring money into the pockets of our people?'Q8: Round off

Problems
The downloaded text of the transcript of the 2015 Presidential Debate was listened to several times to get familiar with its contents.It was discovered that some sections either ended abruptly or began midstream.Five questions were affected (cf.Table 4

The Corpus
The material was carefully transcribed orthographically and the number of words spoken by each debater in response to each question was manually counted.Questions 1-8 yielded 4876, 2490, 5168, 4549, 3180, 2298, 2123, and 785 words respectively.These totalled 25 469 words, which is the corpus size.

Data Collection
The 25 469-word corpus was carefully studied to identify and mark all the sentences therein.Again, these were manually counted and recorded for each debater and for each question.Altogether there occurred 1876 sentences, and this constitutes the data for the study.

Procedure for Data Analysis
The analysis of data was in two phases.Major sentences were identified and classified into the three structural types of simple, compound and complex and the frequency of each type was determined.Further analyses at more delicate levels were undertaken.For example, each identified simple sentence was critically examined to ascertain whether or not the optional elements of clause structure were realised by rankshifted clauses, or whether or not the situation-induced features of multiplicity, mobility and inversion that radically alter the structure of the English clause were at play.Seventeen variants were identified and the frequency of each one was determined according to themes and according to debaters.Next, the structurally compound sentence was examined to identify the number of alpha clauses therein.The same was done for the structurally complex sentence, with focus on whether the beta clause was finite or non-finite.As with the simple sentence the frequency of each variant of the compound and complex sentences was also determined.The minor sentences were similarly identified and dichotomised in terms of presence or absence of the verbal group (VBG).Since each compound and complex sentence minimally has two clauses, it became expedient to numerically ascertain their clause constituents and add same to those of simple sentence and minor sentence to generate the total number of clauses and also account for the data from the perspective of the clause.
The second phase of the analysis of data entailed a critical examination of the corpus to find out precisely how each identified sentence type or subtype was used to respond to the questions posed in terms of recognising the problem and justifying the thesis (R), specifying actions to be taken and justifying same (S), and concluding the submissions (C).To achieve this, each sentence type was first listed, followed by examples reflecting all the twelve debaters and all the eight themes as applicable.Samples of these are presented and discussed to demonstrate how they aided debaters to realise their goals.

Preliminary Remarks
The 1876 sentences comprise 1729 major and 147 minor types distributed as Table 1 displays.
From the facts displayed in Table 1 it can be seen that major sentences were dominant.But that is how far the comparison between major and minor sentences and indeed among the other variables can go.It cannot be validly stated, for instance, with the figures recorded for them, that UDP generated the highest number of sentences and ADC the least, because of the factors already hinted at above, which is more vividly captured by the empty cells in Table 4 below (Notice that only in Q1, Q3 and Q4 were all twelve debaters' responses fully recoverable).
To make meaningful comparison possible, the solution sought was to calculate the average length of sentence relative to the number of words generated by each debater and this yielded fruits.Since there were 1876 sentences in all and the corpus size is 25 469 words, the mean length of sentence was calculated as 13.6w/s.Five debaters recorded mean length of sentence above the group mean, six recorded below the group mean while one (PDP) posted the same mean sentence length as the group average.The highest was 20.03w/s while the lowest was 9.85w/s and these were recorded for ADC and HDP respectively.Table 2 exposes these and more facts.
These figures contrast sharply with what Bellova (2012) reported.Her subjects' sentences were by far longer, as only five debaters' mean sentence length is anywhere near the lowest.While difference in period may explain the disparity (Only Obama and Bush belong in the twenty-first century), a more logical explanation may lie in genre difference: The presidents' speeches are samples of spoken prose whereas the debaters' represent spontaneous speech.
All the 1729 major sentences were classified into types.As Table 3 below shows, the structurally simple sentence was dominant, accounting for 77 per cent.It was distantly followed by the structurally complex sentence.The compound sentence was the least recurring.
The 1729 major sentences bore 2103 major clauses made up of 1330 clauses from the same number of simple sentences, 261 from compound sentences, and 512 from complex sentences.This gives an average of 1.22 clauses per major sentence.However, it was 1.6 and 2.22 clauses per sentence respectively of compound and complex sentences.In essence, clause density was much higher with the complex sentence than with the compound sentence.Because the structurally simple sentence bears a single alpha clause only and the minor sentence is technically less sentence due to its partial realisation of elements of clause structure caused by natural language dynamism, the issue of clause density does not apply to these two categories.But, a slight difference was recorded when the 147 clauses from the same number of minor sentences were added to give 2250 clauses in all.The average number of clauses per sentence became 1.2.From the perspective of the data there thus occurred 2103 major clauses representing 93.5 per cent and 147 minor clauses constituting 6.53 per cent.The proportion of major clauses relative to minor was much higher than the 82 v18 reported in respect of spoken instructional texts (Adejare, 2013).At this juncture it is important to state that forty-three sentences representing 2.5 per cent of major sentence realisations were syntactically defective.They comprise six simple and thirty-seven complex types and are handled separately under collapsed sentences below.Further results are presented and discussed under the headings simple sentence, compound sentence, complex sentence, and minor sentence in that sequence.Features characterising the sentence in political texts are highlighted as concluding remarks.

The Simple Sentence
Table 4 below shows the distribution of the simple sentence according to themes and according to debaters.It gives a clear picture of the number of simple sentences produced and what each of the twelve debaters and eight themes actually contributed to the pool.
The empty cells earlier referred to notwithstanding, it is not difficult to see that Q3 had the highest number of simple sentences out of the three questions for which recovery of the text of the transcript was full and that the number of simple sentences not only varied according to theme but also differed according the individual debater.To determine each theme's average number of simple sentences the total recorded for them was divided by the number of debaters involved.The outcome shows that Q3 had the highest concentration of the structurally simple sentence (24) while Q8, where debaters were allowed only one minute as opposed to the three of Q1-Q7, record-ed the lowest (7.8).A similar computation done for the debaters reveals that HDP had the highest average (31.2) while ADC posted the lowest (13).The overall mean for debaters was 110 and for themes 166.3.These are clear indications of marked differences in the frequency of simple sentence occurrence occasioned by differences in themes and debaters.The absence of precedence makes comparison somewhat difficult, though.For instance, Cienki's (2009) report on Palin's syntax lacks supporting statistics.
Because of its little analytical value calculating length of sentence for the simple sentence was not pursued.Attention was consequently shifted to how each simple sentence faired syntactically and the outcome is detailed in Table 5.
As Table 5 displays, the 1330 structurally simple sentences were sorted into seventeen subtypes according to syntactic potentialities.Those not manifesting rankshifted clauses as primary elements of structure featured the most.Let us call them kernel sentences.They were distantly followed by those with rankshifted clauses as complement or as adjunct, adjuncts at initial position, subject-predicator inversion, and multiple adjuncts in that order.No instance of the simple sentence with the rankshifted clause as subject, complement and adjunct, or as subject and adjunct, was seen.
The seventeen syntactic variants were grouped into five based on realisation thus: the kernel sentence, the rankshifted clause as an element of structure, multiplicity of the complement and adjunct, mobility of the adjunct, and inversion of clausal elements.Indeed, with the notable exception of the kernel sentence, most others exhibited more than one feature.The pragmatic decision taken in such circumstances was to determine the syntactically more dominant feature and assign the given sentence accordingly.An instance is (14a) below, which could have easily fallen under the rankshifted clause as subject and adjunct but which was The presentation that follows takes each category at a time.The symbol or acronym adopted for each structural variant under consideration as displayed in Table 5 is placed after its appropriate heading for ease of recognition.Textual examples are provided and analysed, with the debater (e.g.,.,.,.AA), theme (e.g.,.,.,.Q1), and context or function (R, S and C for recognising, specifying, and concluding respectively) enclosed in brackets after each citation.Where necessary, the item under focus is italicised.Comments on how debaters' goals were achieved and on variations in meaning occasioned by changes in syntactic structure and context of use are made.This is done for compound, complex, minor, and collapsed sentences as well.
Eight structural variants of the kernel sentence are derivable from the (S)P(C)(A) clause structure formula, but || P|| and || P| A|| were not seen and || P|C|| and || P| C| A|| were rare outside the context of expressing gratitude as in (1a)-(1b).These are indications that imperative clauses expressing order as an underlying semantic function, as in (1c), were uncommon, which tends to confirm Astrero's (2017)    The foregoing paints a picture of syntactic simplicity because the nominal groups (NMGs), verbal groups (VBGs), adjectival groups (AJGs), adverbial groups (ADGs), and prepositional groups (PRGs) realising elements of clause structure are structurally H-Type and MH-Type.But this is only superficial: Most elements of clause structure were indeed realised by structurally HQ-Type and MHQ-Type groups, with some manifesting subordination at Q.The need to pack in as much information as possible into a single element of structure, which is symptomatic of semantic complexity, explains this phenomenon.Thus two sets of the kernel sentence, corresponding roughly to van Leeuwen's (2012) absence or presence of complementation constructions effects, are identifiable.Each element of structure of the second set is therefore examined more closely.(a)The Subject Some form of syntactic complexity was introduced into the kernel sentence when the NMG in subject position had a rankshifted clause or PRG as Q.In (7a)-(7f) rankshifted clauses as Q are illustrated.They provide further information on the noun headwords and help to distinguish them from similar ones.Though complex catenated VBGs functioned as predicators, none featured more than one layer of group subordination.There however occurred particles in secondary qualifier function as in (10e)-(10f) below.A catenated VBG indicates at least two actions, events, processes etc depending on its complexity.For instance, the H-element of (10a) expresses AD's future intention to undertake a process conveyed by the to-non-finite form at Q meaning purpose (Notice that in (9a)-(9b) above the predicators are structurally compound, particled MHQ-Type VBGs) The significance of A1 and A2, which respectively express time (presently) and location (in this country), would have been lost had they been postponed to the usual end position.So, their inversion is for greater effect as the speaker stresses the time as now, not past or future, and the location as Nigeria, not elsewhere, and appears to be saying matter-of-factly thus: I am addressing the issues that concern us as a nation at the moment.The PRG diagrammed below in order for its complexity to be fully appreciated has eight layers of group and clausal subordination.V .
V . .for an inte-po-sys-that will cer-give us the kind of ene-we ne-to po-our en-of growth grated wer tem tainly rgy ed wer gine

The rankshifted clause as an element of clause structure
Simple sentences manifesting rankshifted clauses as elements of clause structure represent a significant 21 per cent (281).
Their presence is indicative of syntactic complexity and grammatical sophistication, the absence of subject-complement-adjunct and subject-adjunct subtypes notwithstanding.Rankshifted clauses generally made more explicit debaters' intentions because of their capacity to contain large amount of information that would have probably taken several simple sentences to convey.Each variant is examined closely as follows.Below is one simple sentence characterised by grave internal syntactic complexity.Its adjunct, a twenty-nine-word long complex sentence in its own right, is split in two halves by a syntactically unconnected but textually related clause (bracketed) that affirms the factuality of A's content.Its effect would have been lost had it been delayed till the end.This shows how a debater's desire to achieve a mediate goal impacts syntactic structure.
S The last sentence exemplified here has two complements that assert the speaker's candidature and two adjuncts that identify the position sought (A 1 ) and why (A 2 ).Of interest are the layers of group subordination in the PRG at A I.

Mobility of the adjunct
Adjuncts ordinarily occupy the last slot and are the only syntactically mobile element of clause structure.Because they generally make reference to the circumstances of the verbal situation and comment on the propositional content of the clause (Quirk et al, 1985:730), moving adjuncts to other positions suggests the speaker intends to purposely highlight the specific meaning indicated for greater effect as this data shows.This is in agreement with Millar and Curie's (1972) report that adjuncts transferred from sentence end execute both thematic and structural functions.Mobility affected 10 per cent (131) of simple sentences with eighty-two adjuncts occurring clause-initially, thirty-two at median position, fifteen after the subject, and two before the complement.The only instance of initial complement truly exemplifying stylistic motivated inversion has identical H-Type AJGs at C, each immediately followed by a PRG at A, with both groups coordinated phonologically.Fronting the complement and repeating it with the accompanying adjunct was reiterative, serving as philosophical underpin for the position held by the speaker (cf.Millar & Curie, 1972).

The Compound Sentence
The 168 compound sentences occurred as displayed in Table 6 below, which clearly shows that the two-clause compound sentence was dominant and that none exceeding five clauses occurred.

The two-clause compound sentence
The two-clause compound sentence was more of declaratives than interrogatives and imperatives.The last two are shown in (32a)-( 32b) and (32c).The two-clause compound sentence marked for declara-tive indicative mood was distinguished according to whether or not there was an overt marker of coordination and according to the coordinating conjunction involved.Although no statistical count was taken, sentences with coordinators were evidently dominant and the coordinating conjunction was and meaning addition, which also served to provide more information and reinforce that previously given.An excerpt is taken from each theme thus:  ||and ||(α)we have citizens who share in the booty.|||(A-PAQ4R) Primarily contrastive but reinforced and justified each debater's position on the issues raised.For instance, the speaker in (35a) claimed he would be quicker at attaining his objectives than the respected sage, while that in (35b) contrasted two fundamentally different teacher-education institutions to underscore the need for policy reversal.The gap between policy formulation and implementation in three major sectors was highlighted using but thus: (35c) |||(α)The government has spent so much money on power,||but ||(α) the effort has been frustrated and sabotaged.|||(ACPNQ2 R) (35d)|||(α)So funding has been made||but ||(α)the equipment is not in place.|||(NCPQ6 R) (35e) |||(α)Argentina is one of the largest suppliers of beef in the world||but ||(α)livestock and cattle do not struggle for space with traffic or commuters in that country.|||(UPPQ7 R) A personal resolve to succeed against daunting odds and the superficial attitude toward social maladies like corruption are focussed on in (35f)-(35g).(35f)|||(α)It's going to take a lot of negotiation||but ||(α)I intend to give it a push.|||(ADCQ4 S) (35g)|||(α)Some people react to the symptoms||but ||(α) very few people know why corruption is described the way it is described.|||(UDPQ4 R)

The multiple-clause compound sentence
The only five-clause compound sentence linked by but occurred as UDP recalled a lecture on Excuses to Criminal Liability where 'Coker is my cousin', used as a contextual metaphor for nepotistic corruption in Nigeria, was cited as one.The bewildered students were told that a defendant confronted with the facts in (36a) had given it as response.

The Complex Sentence
A total 231 sentences comprising 512 clauses distributed as shown in Table 7 below were structurally complex, but thirty featured beta clauses only.The two-clause complex sentence was the most common, followed by the three-clause subtype.
There occurred 446 finite and sixty-six non-finite clauses.
Further observations on the complex sentence are based on the finiteness criterion.

The finite clause as beta clause
Complex sentences with the finite clause as beta clause were further distinguished according to the number of clauses therein.In the two-clause sentence the beta clause either came first or second as (38a)-( 38b) and ( 38c which we all know.|||(ADQ4 R) The foregoing shows there was no fixed order of occurrence of the elements of structure.What appeared to determine which of alpha or beta was clause-initially were the focus of the message conveyed and each debater's level of linguistic competence.In the corpus analysed the two-clause complex sentence with beta clause as initial was more prominent, attesting to the fact that thematic fronting is not limited to the single clause sentence.In (39a), for instance, the debater appears to emphasise artificial scarcity as the basis for unwholesome behaviour.Similarly, the speakers' intent at emphasising the content of the initial clause in (39b)-( 39c

The non-finite clause as beta clause
The complex sentence with the non-finite clause as beta clause primarily exhibited semantic attributes associated with each non-finite form.For instance, where the tonon-finite VBG clause with purpose as meaning was the first element of sentence structure, the debater was able to justifiably identify the direction in which he would go on a given issue.Similarly, the x+g suggested on-going actions while some form of completion was associated with the x+n form.These interpretations become clearer as the forms are encountered in context.One feature associated with the two-clause sentence was that where the initial clause was alpha, the beta clause was invariably x+g or x+n non-finite VBG clause as (41a)- While initial positioning for emphasis can be taken as given, the foregrounding of to-non-finite VBG clause meaning purpose provided additional motivation for listening.This interpretation should be understood against the fact that all sentences occurred in the context of specifying actions to be taken and listeners would want to know precisely what the speakers' intentions were.

The Minor Sentence
The minor sentence represents what Millar and Currie (1972:44)  As the examples show, the contextually inferable missing elements amplify the impact of the sentence since their absence does not subtract the meaning which is not already implicit (cf.Millar & Currie, 1972:44).Some more minor sentences are identified as follows: (42d) No more to hunger in Nigeria (AA Q1) Fellow Nigerians (UDP Q1; AA Q1) Very critical (NCP Q7) Kidnapping, insurgency and all of that (HDP Q5) The non-monetary (KOWA Q4) Opportunities for all (AA Q4) No (UPP Q1) Out of the ten local governments of Borno State, three remaining (PDP Q5)

Collapsed Sentences
It must be unambiguously stated that this brief consideration of collapsed sentences is not an analysis of errors; rather, it is an attempt to objectively account for all occurring sentences in the corpus.Forty-three sentences comprising thirty-seven complex and six simple subtypes collapsed.They consist of fifty-seven clauses including sixteen non-finite ones.Thirty syntactically understood complex sentences occurred as beta clauses only because the obligatory alpha clause was unrealised due to a combination of factors, including natural language dynamism.Although syntactically defective they nevertheless contributed toward achieving debaters' goals.For instance, (43a) collapsed but its textual function of justifying the action to be taken inferable from the preceding sentence (enclosed) remains largely unaffected.(43a)(And we said security votes must be tied to only security matters and must be accounted for.

Defining Features of the Sentence in Political Texts
Before now the sentence in political texts had not been described with this much detail as attempted here.So the question that logically follows is: What does this analysis of Nigeria's 2015 Presidential Debate say about the characterising features of the sentence in political texts?This question can only be answered summarily in this final section on data presentation and discussion of results.It must however be noted that speakers addressed an audience that was both small and immediate, wide and remote, at once; responded to questions on diverse subjects posed by moderators who are largely journalists; and, managed to impress their listeners and possibly woo them to their sides.To achieve their goals and adequately and unambiguously convey their ideas and feelings, debaters employed strategies that saw a 92 per cent dominance of the major sentence over its minor sentence counterpart out of a 1876-sentence corpus, for instance.That minor sentences occurred at all attests to the fact that the dynamics of natural language cannot be ignored even in very serious language use situations such as the 2015 Presidential Debate beamed life to a world audience.It is not known what the average length of sentence in political texts and in-deed everyday language use is.But, while the recorded mean of13.6 w/s may be indicative of both individual linguistic competence and oratorical style, its marked disparity with Bellova's (2012) findings suggests that even within political texts, period and genre differences may account for differences in sentence length.Some debaters' responses were characterised by very short sentences (9.85w/s) and others by relatively long ones (20.03w/s) and this was regardless of sentence type.How frequently structurally simple, compound and complex sentences occur in everyday language use situation, or, in a given text type, is, again, unknown.However, the revelation of a proportion of seventy-seven, ten and thirteen respectively makes the simple sentence dominant and the compound slightly less recurring than the complex subtype.Clause density was higher with the complex sentence than with the compound sentence at 2.22 and 1.6 clauses per sentence respectively.The fact that these varied according to speaker and according to thematic dimension definitely leads to the conclusion that subject matter, mediate goal such as specifying actions to be taken, speaker's linguistic repertoire, and style of discourse may jointly explain the patterns found.Thus it cannot be logically argued that debaters who used more complex and compound sentences were better able to articulate their points, or, presented same more succinctly, convincingly and effectively, than those for whom the use of the simple sentence may be said to represent an idiolect.There are two reasons: One, the combined frequency of compound and complex sentences was only 30 per cent of that of simple.Two, outside the syntactically plain kernel sentence which constitutes 59 per cent (Many were manifestly complex in their own right, with rankshifted groups and clauses as elements of group structure.),the rest exhibited diverse forms of internal syntactic complexity characterised by rankshifting of clauses to function as elements of clause structure and multiplicity, mobility and inversion of clause structure elements.Since it was not uncommon to find structurally complex groups as elements of clause structure it cannot be claimed that the texts were characterised by syntactic simplicity.It would seem that sentence types were selected in proportions that suited each speaker and the theme addressed.
Seventeen syntactic variants of the structurally simple sentence were isolated and only one lies within the eight generable by the (S)P(C)(A) clause structure formula handled as kernel sentences in this study.There was a total absence of the ||P|| and ||P||A|| and a paucity of ||P|C|| and || P||C||A|| structural variants, suggesting that, with the probable exception of the let-type imperative, clauses marked for imperative mood were a rarity because of the situational context.The other sixteen variants were almost evenly distributed between simple sentences with rankshifted clauses as elements of clause structure and those with their structures radically affected by situation-induced clausal processes of multiplicity, mobility and inversion.Implicit in these are syntactic complexity at another level of description and debaters' grammatical sophistication.An under-lying desire to place a greater emphasis on certain issues raised, and pack in as much information as are relevant and possible in order to drive home a point, was clearly evident in the choice and frequency of the different patterns of the simple sentence seen, including the positioning of adjuncts outside their normal end slot, which affected 10 per cent.These observations are equally true of even the structurally compound and complex sentences particularly in terms of clause density, clause positioning, and clause coordination.With the former, conjunctions were deliberately chosen to enable debaters connect related facts and highlight their attitude towards the contents of the proposition; and, in the case of the latter, the focus of attention determined which of the obligatory and optional elements of sentence structure was thematically fronted for maximum effect.

CONCLUSION
What this paper set out to do was to analyse sentence pattern and usage in the Presidential Debate held in Nigeria in 2015 as a study of syntactic structure in political texts.Employing an improved version of the systemic grammatical model as analytical tool, the study identified and accounted for all occurring sentences in the corpus and found the major sentence dominant over its minor sentence counterpart.It classified all the major sentences into simple, compound and complex subtypes, and revealed that not only was the structurally simple sentence over three times more frequently used than the structurally compound and complex sentences combined, the choice of sentence type or its syntactic variant, length of sentence and distribution, varied markedly according to the debates' eight themes, speakers' mediate goals, and debaters' grammatical sophistication.Finally, the study demonstrated how debaters strategically used the sentence to convince listeners to vote for them, noting that the positioning of clauses or clause structure elements, including which of alpha or beta was clause-initially, was ruled by the focus of the message conveyed, for instance.Since details of the outcome have been elaborately given above it only remains to consider some implications.
This analysis has proven that there are levels of simplicity and degrees of complexity in sentence structure and that an otherwise simple sentence can indeed manifest forms of complexity at its various elements of structure.This challenges us when we say we are studying or teaching the language that people actually use, but we employ a model that lacks the capacity to account for what people actually say.The systemic grammatical model chosen has its limitations as already seen, but our adoption of the upgraded (S)P(C)(A) clause structure formula, where situation-categories' motivated textual functions were added, together with the three primary sentence types of simple, compound and complex recognised, has enabled us to comprehensively account for all occurring sentences in the corpus.The choice of the sentence is underscored by the need to shift focus from the strands, which sometimes results in fragmentation, apart from the sentence being the primary unit of communication.This study has demonstrated that syntactic analysis at the level of the sentence is essential to a full and proper understanding of political texts (cf.van Leeuwen, 2012).If any new insights have been gained into the structure of the English sentence generally and the sentence in political texts particularly, following the analysis of Nigeria's 2015 Presidential Debate undertaken here, then the goals of the study would have been met and this should be consolidated.

Table 1 .
Typological Distribution of Sentences

Table 4 .
Thematic distribution of the simple sentence classified as subject only because of the dominance of the x+g non-finite VBG clause.

Table 5 .
Syntactic Variants of the Simple Sentence and their Distribution It has so far been shown that the subject and complement can be realised by NMGs with the rankshifted clause at Q, but the data also provides evidence of NMGs at both S and C exhibiting rankshifted clauses at Q.This is proof that the structurally simple sentence can otherwise be 'complex' in its own right.Evidence of structurally disrupted HQ-Type NMGs at C, occasioned by the need to bring the adjuncts into semantic prominence through syntactic prominence, is provided by (12a).
The finite clauses were mostly that-clauses, collocating with verbs like believe, hope, ensure, realise, and convince which stress debaters' conviction about the veracity of the statements made.||We | will ensure |[[that each state of the federation.establishes its own power plant]].||(ACPNQ2 S) (15c)||But | it| is| sad|[[that Nigerians can't even go to.good school without paying with an arm and a leg]].||function as adjuncts.The finite clauses featured overt markers of subordination and these were quite diverse formally and semantically, expressing time, place, and con- (15a)||We |believe|[[that we can create economy big enough to provide jobs for all]].||(AD Q1 S) (15b)even has the opportunity to steal what belongs to the public]].||(PDPQ4 S) (18e)||Our simple strategy| is|[[to review.the national agriculture system]].||(NCPQ7 S) (18f)||Our commitment | is |[[to make sure that it is opened up to private sector investment]].||(PDP Q3 S) (c)The Rankshifted Clause as Adjunct ([[A]]) Again, both finite and non-finite clauses were rankshift-ed to Remarkably the four-and five-adjunct clauses came from the same source.Sentence (24g) is a non-polar interrogative clause with four adjuncts realised by a PRG and three syntactically identical rankshifted finite clauses that are in semantic opposition temporally with the PRG.
(25j)||In terms of the import bills,| you | know | [[that we dropped by almost two-thirds.]].||(PDPQ7 S) (25k)||By virtue of the advancement of society | there |is |definitely | need | [[to be changes in the way things are done]].||(UDPQ7 R)' (25l)||Normally| there |will be| palliatives that will be made available to them.||(UDPQ5 S) blades are working there?Maybe four.What's the problem?What's wrong with the other twelve?Why are they not working?I mean what is stopping them from working? (AA Q2 R) (b) Predicator-Complement-Predicator Inversion (P-C-P)

Table 6 .
Manifestation of the Compound Sentence We have made fertiliser available,|| ||(α) people are improving on the yield of their farm||but ||(α) at the same time we have livestock.cattleeat up all of these things.|||(UPPQ7 R)

Table 7 .
Distribution of the complex sentence

)
|||(β)Because we know state governments in this country who use pliable state assemblies to appropriate for themselves security votes that are even more than their state recorded expenditure,|||(UPP Q4 S) The next sentence is syntactically interesting because, of its five beta clauses, only the first occurred with a marker of subordination.Its syntactic defect notwithstanding, the listener should be able to agree with the speaker that