Cross-dialectal similarity of registers: The case of the sentence across Ghanaian and British newspaper editorials* George Kodie Frimpong, University of Ghana The claim of register theory is that there is cross-dialectal similarity in linguistic choices across texts from the same register because language used in a register serves a specific communicative purpose in a given situational context. In this empirical study, the sentence was investigated across a specialized corpus of editorials from Ghanaian and British newspapers with the hope of ascertaining the similarities or otherwise in this aspect of language use across native and nonnative dialects of English. In the light of Biber and Conrad’s (2009) model of register, we argue that quantitative dominance of specific sentence types and consistency in their distributional patterns across the two sociocultural contexts are indicative that those dominant sentence types are functional to the editorial register. The data reveal that in spite of a few distributional discrepancies, which reflect regional adjustments in situational and cultural contexts, sentence types are generally similarly distributed across editorials from the two socio- cultural contexts, with declarative and complex sentence types being consistently dominant across the two texts. Keywords: register theory, language and function, newspaper editorials, sentence types, hypotactic and rank shifted relations, the sociolinguistic profile of Ghana 1. Introduction This study is a corpus-based empirical study. It compares the use of sentence types across editorials from Ghanaian and British newspapers. The sentence, also conceptualized as a special type of clause (Greenbaum 1996; Bloor & Bloor 2004; Halliday & Matthiessen 2004; Downing & Lock 2006), has been argued in the literature as the highest grammatical structure in human language (c.f. Halliday & Matthiessen 2004: 9-10; Eggins 2004: 123; Fawcett 2000: 192). What this means, in the sense of grammatical rank scale or constituency (Fawcett 2000: 233, 237), is that the sentence is the most loaded grammatical unit (Frimpong 2015). It contains in it all the other units (clause, phrase, word, and morpheme) of grammar in a scalar order such that each higher unit embeds the unit immediately below it. Thus, a sentence is made up of at least one clause, a clause is made up of phrases, a phrase is made up of words and a word is made up of morphemes1 (Bloor & Bloor 2004). The attempt in this paper is to explore the usage dynamics and, thus, the functional motivation (Bloor & Bloor 2004) behind the distribution of sentence types across editorials of four selected Ghanaian and British newspapers (i.e. the Daily Graphic and Ghanaian Times from Ghana and The Daily Telegraph and The Times from Britain), taking the newspaper editorial as a unique register whose linguistic choices are influenced by the situational context of the editorial (Biber & Conrad 2009). Even though more than four newspapers from varied backgrounds could have been included in this study, we considered the four just adequate in enabling us to compare an aspect *This paper is based on my PhD project conducted at University of Ghana in 2015. I express my indebtedness to my supervisors, Professors J. F Wiredu, Emerita Kropp Dakubu (of blessed memory) and M. Huber and to ORID, University of Ghana for their financial support 1 These, though, are grammatical potentials since in reality each of these grammatical units may contain at least one of the units immediately below it. of the English used in Ghana with its historical stock,2 the British English. The essence of this study is not so much about the variation between a native and nonnative English since there is already ample literature in that regard (Kachru 1981; Schmied 1990; Young & Walsh 2010; Huber 2012; Chien 2014). The essence of this study is about the nature of the variation or otherwise. Especially since this is a study at the confluence of dialectology and register, one is not just interested in ascertaining how language manifests variability and/or uniformity but more importantly, one seeks to establish the nature of the variability or conformity in linguistic choices under the influence of these two external factors of linguistic variability. Thus, the study is premised on the following corresponding research questions: 1. How are functional and structural sentences distributed across editorials from Daily Graphic, Ghanaian Times, The Daily Telegraph, and The Times? 2. What functional motivations influence the distribution of functional and structural sentence types across editorials from the selected Ghanaian and British newspapers? 2. The newspaper editorial register Many newspapers have special sections for editorials which may be an institutional voice of the newspaper, an opinion of an individual and/or an advice of a virtual counselor (Biber 1988). This study is limited to the institutional editorial which is considered in this study as an independent subgenre of the newspaper macro-genre. This position is in line with arguments in the literature that newspaper editorials, unlike other newspaper subgenres, present opinions on factual information whose communicative purpose is essentially to influence the perception of their readership (van Djik 1989; Wiredu 2012; Frimpong 2017). Biber and Conrad (2009: 110) capture this communicative function of the editorial as follows: “[A]n editorial is meant to express an opinion overtly and persuade readers to that opinion.” Due to its explicit agenda of representing the ideological position of a particular media house and of persuading its audience in a particular way of thinking, the newspaper editorial (NE) is noted to use language strategically. Frimpong (2017), for instance, has observed that editorials from Ghanaian and British newspapers gravitate towards nominal clauses. This, he observes, is a linguistic strategy for communicative elaboration. Other studies on the NE register have identified specific linguistic features that mark them as a unique register. They are seen, for instance, as having similar generic structure (Ansary & Babaii 2009); presenting powerful ideological positions through careful use of language (Le 2009); having similar distribution of transitivity patterns (Shokouhi & Amin 2010), etc. This study aims to consolidate work on NE from grammatical perspective. Its cross- cultural focus is to affirm the register argument that because texts from the same register are inspired by the same communicative purpose, linguistic choices are largely similar even across different dialectal orientations. This is what Biber and Conrad argue as follows: Regardless of any dialect differences, speakers using the same register are doing similar communicative tasks; therefore in most basic respects the characteristic language features used in a given situation are similar across speakers from different dialects… 2 Over a century ago, the British planted the English language on the Ghanaian soil through colonialism. Ghana has continued to use English as an official language even though it gained freedom from British colonial rule in 1957. There are of course differences in pronunciation and word choice associated with different dialects, and there are even occasional differences in grammar. But these differences are minor when compared to the major linguistic differences among different registers, associated with different situations of use (Biber & Conrad 2009: 12 & 15). Thus, though the editorials studied in this paper are from different sociocultural contexts with variable dialectal orientations, they should share similar linguistic features since they share similar features in their situational context. The focus on editorials from Ghanaian and British newspapers is motivated by the desire to compare and describe aspects of the two varieties of English (the Ghanaian and British varieties of English, which coincidentally are historically related) from register perspective. Not only is the English used in Ghana a development of what was once introduced into that geographical context by the British colonial empire but also, the editorial tradition and the newspaper industry in general were introduced into the Gold Coast, now Ghana, by the British colonial empire (Asante 1996; Frimpong 2015). One would expect that this historical connection should influence both language use and newspaper styles of production. However, this is not the case. For instance, as argued below, Ghanaian newspapers have different classification, style of presentation, content coverage, etc. Besides, in terms of the language, the argument in the literature is that when a language migrates to a new location, it reflects the fauna and flora of its new location thereby undergoing a process of nativization (Schneider 2007). Thus, in a sense, this study forms part of the ongoing debate in the discipline that though Ghanaian English emerged from the British stock and that speakers of Ghanaian English are conscious of maintaining the British standard, there is ample evidence of variation in all aspects of the English used in Ghana (Sey 1973; Dako 2003; Huber 2012; 2014). This study extends the argument further to include register variation. Thus, given the argument above that Ghanaian English has significantly distanced itself linguistically from its historical stock, linguistic similarities in aspects of this variety may be safely explained from register perspective. 3. Audiences, ideologies and traditions of newspaper production and circulation across Ghana and Great Britain Great Britain boasts of one of the oldest surviving newspaper traditions in the world. The English press started in a handwritten pamphlet format in 1622 (Morison 1932). The two British newspapers, The Times and The Daily Telegraph, were founded in 1785 and 1855, respectively (Morison 1932; Frimpong 2015). British newspapers over the years have developed along two trajectories which aid their classification. On the one hand, newspapers in Britain have been classified along ideological lines, especially in terms of a newspaper’s political inclination. From this perspective, The Times and The Daily Telegraph are both pro conservative newspapers. This, of course, is a bias that could have been corrected by replacing one of them with a newspaper of an alternative political persuasion. This bias, however, is deemed not so significant to affect the use of grammatical structures. The focus of this paper is on quality newspapers and any of them qualifies to be selected. The second, and perhaps the most popular developmental trajectory, is along the lines of readership and the quality of language use. Based on a targeted class of readership and quality of language used, British newspapers are again classified into quality newspapers, midmarkets and the tabloids (Chan & Goldthorpe 2007). Newspapers from these categories target upper middle, lower middle and working classes, respectively. In terms of language use, whereas the quality newspapers, including The Times and The Daily Telegraph, are noted to use formal and serious mode of writing, devoid of sensationalism, the midmarkets and tabloids, however, use less formal and more sensational language (Quinn 2013). The Ghanaian modern newspaper industry is a relatively nascent one. Until 1950 when the oldest surviving newspaper, the Daily Graphic, was established, several attempts had been made unsuccessfully, first by the British colonial administration and then by some local Gold Coasters (Asante 1996). Prior to the advent of the Western-styled print media, the West African sub-region, including Ghana, used various indigenous forms of mass communication such as runners, talking drums, gong gong, fire, etc. to disseminate information to unenumerated audiences (Fosu 2014). It is noteworthy that the Gold Coasters who made the efforts at establishing and maintaining the newspaper industry were motivated primarily by political exigencies. It is not surprising, therefore, that the Ghanaian newspaper industry has developed along different paths. Thus, Ghanaian newspapers are classified into public or private newspapers according to which public newspapers are state funded and private newspapers are funded by private individuals. As state funded newspapers, public newspapers have wider circulation and topical coverage due to their ability to fund both quality production and national representations. The Daily Graphic and the Ghanaian Times, the two main public newspapers in Ghana, are by far the most voluminous newspapers covering more varied topical contents than their private counterparts. One key feature that differentiates British newspapers from Ghanaian newspapers is their readership audience. Though newspapers from the two sociocultural contexts are produced in English for English speaking readership, the sociolinguistic backgrounds of readers from the two contexts differ significantly. This is because, whereas readers of British newspapers are native speakers of English, readers of Ghanaian newspapers are non-native second language speakers. Similarly, Fosu (2014) notes that whereas British readers are predictable in their choice of newspaper type, Ghanaian readers are only predictable in their choice of a private newspaper. That is, sympathizers of the major political parties gravitate towards newspapers that promote their party. For public newspapers, however, readers are attracted to them based on the quality of language use. As Fosu argues, readers of public newspapers have variegated ideological backgrounds who are brought together by the common denominator of Standard English. This argument makes sense because in Ghana, Standard English is a mark of identity that brings together educated Ghanaians of all ideological and ethnic backgrounds. Because it is acquired in the classroom, only educated elites, regardless of political or ideological background use this language. Thus, Ghanaian public newspapers target this small constituency of the Ghanaian population as their core reading audience. 4. The sociolinguistic profile of Ghana and the language of Ghanaian newspapers In a study of this nature, an account on the sociolinguistic situation of Ghana is considered necessary. This is because, language use in Ghana is not as straightforward as it is in a native context. In mother-tongue English communities like the UK, the dominance of English has led to the weakening, even death of minority languages (Simo-Bobda 1997:88). The situation is different in Ghana. In Ghana, though the English language is the official monolingual language, just as in other West African countries, it co-exists with vibrant indigenous languages making the sociolinguistic situation of Ghana a very complex one (Dakubu 1988; Obeng 1997). As an officially monolingual country, (Dakubu 1988), Ghana uses only one language (English, an exolanguage inherited from British colonial administration (Guerini 2008)) in official domains (Simo-Bobda 1997:87). English is thus used as the language of education, in legal documents, in newspapers and magazines, in parliament and in all official transactions in Ghana (Huber 1999). As the language of education, it is not surprising that most published materials in Ghana are done in English, an observation made by Dakubu decades ago which she foresaw would not change in any foreseeable future (c.f. Dakubu 1988: 10). Thus, though monolingual in official domains, Ghana is highly multilingual (Dakubu 1988) in informal domains, home to over 80 indigenous languages (Ansah 2008) whose speakers live together and share social experiences as friends, neighbours, members of social groups and even as spouses, especially in urban contexts. The prestige given to English in Ghana by being selected above the 80 plus indigenous languages in all official domains and the social status and power English offers to its users in Ghana have created a kind of a diglossic situation with English as the high language and the local languages as the low and less prestigious languages (Guerini 2008). This prestige influences some parents in Ghana to introduce their children to English very early, even before school age. They rationalise that it is English that has economic power and can gain access for their children into prestigious schools, especially in urban and peri-urban centres. In spite of the prestige enjoyed by English in Ghana, Akan, one of the indigenous languages has managed to assume a lingua franca status, together with English and Hausa (Obeng 1997), spoken by more than 80% (40% native speakers and about 43% L2 speakers) of Ghana’s population (Guerini 2008). Ansah (2008) notes a number of practical and sociocultural reasons underlying the popularity and utility of Akan in Ghana. As she observes, much of private economic activities are engaged in Akan. So too are radio discussions and television programmes conducted mostly in Akan, to the extent that many radio and television stations across the length and breadth of Ghana are Akan-based (Ansah 2008). The prestige and popularity of Akan extends to religious and cultural ceremonies (such as naming, funeral, marriage, Christian worship, etc.) and political discussions. It is noteworthy that many speakers of Akan and of the other indigenous languages (both native and nonnative speakers) are not literate in these languages. This perhaps explains the popularity of Akan in informal domains only, especially in speech-based domains. This also explains why though just about 40% of Ghanaians can speak and read English, newspaper producers find it profitable to target this minority. It makes sense therefore that newspaper production in Ghana is exclusively done in English. Otherwise, it does not make economic sense to produce for the minority of Ghanaians (<40) who are able to read in English. 5. Language and function Function has been conceptualized in linguistics from several perspectives (Bloor & Bloor 2004; Payne 2011; Biber & Conrad 2009). It has been a common approach in traditional grammar to analyse grammatical structures from their formal and functional perspectives, based on the grammatical role a grammatical form is playing. Thus, a noun phrase may be a subject, object, complement, etc. (Quirk et al. 1985). Function in language has also been approached from a semantic perspective where a structure is assigned a participant role of an agent, goal, senser, locative, phenomenon, etc. (Payne 2011). From pragmatic perspective, however, function in language is associated with the speech acts underlying language use. This is what Bloor and Bloor (2004) consider as communicative function. From a more socio-functional perspective, Halliday and Matthiessen (2014) conceptualize function in language as multifaceted in their metafunctional theoretical model. By far, one of the most theorized relationships between language and social function, the metafunctional model argues that language is functional simultaneously in three layers. The first layer of language function is the ideational layer, while interpersonal and textual metafunctions constitute the last two layers (Halliday 2004; Eggins 2004). That is, language reflects realities of its situational context of use at the ideational level, reveals contact and social relations among participants in a communicative event at the interpersonal level and enables information to be organized to reflect a communicative purpose and context of use at the textual level (Eggins 2004). Though this study is not directly based within the systemic functional theoretical model, already one can imagine how newspaper editorial language can reveal aspects of ideational, interpersonal and textual meanings. For instance, the language of editorials reveals information that is packaged to persuade an unenumerated audience based on its argumentative style. Besides, it may be equally revelatory that the language is written rather than spoken by an unidentified author. These speculations are of course idealistic, taking the situational contexts within which newspaper editorials are written as similar across the two sociocultural contexts. On the contrary, however, the two contexts reflect different sociocultural orientations, which potentially induce linguistic variation in large measures. The orientations of the native/ nonnative context of language use are such that whereas in the former context English is used for both formal and informal activities, in the latter English is used mainly in formal domains. These variations in the use of the English language across the two contexts may reflect in levels of formality across editorials from the two contexts. As Schneider (2007) argues, the sociocultural factor is one crucial parameter that causes a language to be indigenized/nativised to reflect its sociocultural context. Simply put, a language is being used to express cultural and social realities of its new home, especially where the language has migrated as in the case of English in Ghana. Thus, though it is the same language planted on the new land by the native speaker, its novel sociocultural orientations enter the language to engender a new variety of the old language. And this accounts for the dialectal variation postulated in the literature of new Englishes. In this study, function is taken from the perspective of Biber and Conrad (2009), according to which a linguistic feature is deemed functional based on its relationship with the situational context of a particular register. Thus, a linguistic feature is functional if its quantitative representation is more prominent in a particular register than in another. From this functional perspective, users of language are seen as making linguistic choices based on a situational context which is made up of a number of situational features, including the participants of a discourse, the relationship between participants, the topic involved, the communicative purpose of a discourse, the setting of the discourse and the channel or medium of the discourse. For newspaper editorial, one estimates that the participants involved (an authoritative representative of a media house addressing an unidentified mass audience), the medium (written verbal communication), the setting of production and comprehension circumstances (availability of time for planning and editing the editorial before publication and the editor’s awareness that the audience have time to process the editorial for comprehension) and the communicative purpose (to persuade its readership to a certain ideological standpoint) have a high predisposition of influencing linguistic choices. Thus, these variables together with the variable sociocultural backgrounds of the newspapers explored in this study constitute the primary motivation of linguistic variability and/or similarity across the newspapers investigated 6. Sentence types The sentence has been classified in the literature based on its structural and functional properties. Structurally, a sentence may be either simple sentence (having only one main clause) as in sentence (1) below or non-simple/multiple sentence (having two or more clauses) as in sentence (2) to (4). Among the non-simple sentences, a differentiation is normally made among sentences which have a coordination of two or more independent clauses (compound sentences) as in sentence (2), sentences which have a subordination of one or more dependent clauses to a main clause (complex sentences) as in sentence (3) and sentences which have a coordination of two or more independent clauses and a subordination of one or more dependent clauses (Compound complex sentences) as in sentence (4) (Quirk et al. 1985; Wardhaugh 1995; Biber et al. 1999). (1) Simple sentence: President Sarkozy has not been arrested. (The Daily Telegraph) (2) Compound sentence: Ghana has scored high marks in its democratic march and the country is seen as the beacon of hope for Africa. (Daily Graphic) (3) Complex sentence: Of course, while the Times admits that a greater responsibility rests with Dr. Kwadwo Afari-Gyan and his team on the EC, it also holds the view that there is little the commission can do, without the cooperation of all Ghanaians. (Ghanaian Times) (4) Compound Complex sentence: The Canute-like frustration of politicians at their inability to control the tax affairs of the multinationals is out of proportion to the benefit they bring to the economy - or to the sums at stake which set against the country's vast indebtedness can only be symbolic. (The Daily Telegraph) Functionally, simple sentences are further classified in the literature into declarative, interrogative, imperative and exclamative sentences (c.f. Quirk et al. 1985; Biber et al. 1999). But this classification does not mean that non-simple sentences cannot be classified functionally; for, elsewhere Wiredu (2014) in a study of legal reports, classifies complex sentences, some of which have up to 20 clauses, into functional (predominantly declarative) types and his classification finds support in the literature. Huddleston’s (1999) argument about desententialized clauses offers grounds for classifying non-simple sentences into speech acts/ functional types. Huddleston (1999) has argued that when a clause (finite or nonfinite) is subordinated or coordinated to a main clause the subordinated or coordinated clause is desententialized. This means that the subordinated or coordinated clause has its illocutionary force depleted and therefore its speech act status is tied to the independent clause in the sentence. Based on this argumentation, functional sentences (simple or non-simple types) are classified in this work, in conformity with Wiredu’s (2014) classification, based on the speech act function of the main clause. Thus, a complex sentence, for instance, is declarative if its main clause is declarative, but interrogative if its main clause is interrogative – a principle that is used to classify the following sentences selected across the corpus. (5) Compound Declarative: The quality of our universities is one of Britain's great strengths and we already have 12 in the top 100, according to the Times Higher Education ranking. (The Times) (6) Simple Interrogative: Are the political parties going about their campaigns in a sensible and mature way? (Ghanaian Times) (7) Complex Imperative: To see where things are going wrong, consider the attempt to introduce elected mayors. (The Daily Telegraph) (8) Complex Exclamative: How cruel death is, to take away the lives of two MPs in a month in an election year! (Ghanaian Times) The four types of functional sentences constitute what (Sadock & Zwicky 1985: 159) refer to as a “mutually exclusive” system according to which membership of one type excludes a sentence from belonging to another type. Crystal and Davy (1973) note about these four types of sentence that in journalese writing, including editorials, it is the declarative type which is the most preferred. Similarly, Biber et al. (1999) have observed that complex sentences are preferred in the press. But given that the editorial constitutes an independent subgenre within the newspaper, as argued above, with unique situational features, it should be beneficial to our understanding of the linguistic versatility and diversity of the newspaper genre to uncover the usage dynamics of these sentence types in the subgenres of the newspaper such as editorials. In line with register theory (Biber & Conrad 2009), preference for a particular sentence-type is interpreted in this paper as functional. Additionally, this study places emphasis on consistency in the distribution of grammatical features across the two sociocultural contexts of the study. The argument is that if a prominent pattern is functional, it should be consistent in its prevalence across texts from both Ghanaian and British newspapers. 7. Methodology 7.1 The corpus This study is an empirical corpus-based study (Mcenery & Gabrielatos 2006). For the purpose of this study, a specialized corpus of editorials from selected Ghanaian and British newspapers was built. “…smaller, more specialized corpora…focusing on specific registers and genres” (Koester 2010: 66) are becoming increasingly necessary in special purpose research. A specialized corpus was deemed the most appropriate for this study because, as argued by Flowerdew (2004: 21) it is the most suitable method for investigating specific linguistic features and for exploring texts involving particular participants, mode and communicative purpose. For this study, a total of 144 editorial texts from 2 Ghanaian and 2 British newspapers (i.e. Daily Graphic and Ghanaian Times from Ghana and The Daily Telegraph and The Times from Great Britain) produced during 2012 were compiled. For ease of reference, these newspapers are henceforth referred to as DG, GT, DT and TT respectively for the Daily Graphic, the Ghanaian Times, The Daily Telegraph and The Times. The original motivation for conducting this research was to examine an aspect of the English used in Ghana from register perspective and the newspaper editorial came in handy in that in Ghana, it is among the most easily accessible English texts produced by Ghanaian speakers of English. Register studies, according to Biber and Conrad (2009), is better done comparatively and newspaper editorials of any national variety of English could have been compared with the Ghanaian editorials. However, the attempt to compare an aspect of Ghanaian English with a corresponding feature of British English forms part of an ongoing debate in the literature. In his pioneering work on English in Ghana, (Sey 1973) has made the observation that Ghanaian speakers of English are conscious of maintaining British English as an exonormative standard. Over the years, however, scholars have identified tendencies in the English used in Ghana that make it possible for one to argue in favour of a Ghanaian English (c.f. Dako & Huber 2004; Huber 2012). Huber (2012), for instance, argues that the English used in Ghana is unique not in the creation of new words but more importantly in restructuring features of Standard British English. In comparing an aspect of Ghanaian English with British English, one hopes to contribute to the debate on how similar or dissimilar the Ghanaian variety of English is from the British native variety. The selection of newspaper types was based on convenient sampling. Thus, though there are other equally influential quality newspapers, The Times and The Daily Telegraph were selected based on convenience. Though one could use the popularity of these two newspapers (c.f. Westin & Geisler 2002) as the motivation for selecting them, in this study they were chosen because they were accessible at the time of the study. On the other hand, the two Ghanaian newspapers were chosen because they are the two main public daily newspapers. The primary focus in selecting these newspapers was mainly comparability. Thus, one was keen at selecting Ghanaian newspapers that were somehow comparable with the British quality newspapers. Generally, the popularity and influence of the two Ghanaian newspapers, based on their wider circulation and assumed objective use of language, influenced their choice. Work on newspaper circulation in Ghana cites the Daily Graphic and Ghanaian Times as the most widely circulated (Yankson et al. 2010; Fosu 2014; Frimpong 2017). Fosu (2016: 6) argues that the two Ghanaian newspapers are among the “dominant, credible and influential quality publications in Ghana.” The notion of comparability in this paper is not to denigrate the contextual variability at the background of the newspapers explored. We are aware that people from different dialectal backgrounds use language differently. However, as the theoretical argumentation has it, language used in similar situational contexts reflects similar linguistic choices even across different dialects. It is in the hope of not comparing say tabloids with quality newspapers that we take these precautionary measures which limit our choice of newspapers from the Ghanaian context to the two public newspapers. Due to the large number and the varied topical scope of issues produced by the four newspapers during the period of the study and in the bid to control the variables compared, the selection of texts into the corpus was further limited to texts which have political content. These texts were not difficult to select from the British newspapers since their online archives have categorized texts into various thematic areas, including politics. The problem was with the Ghanaian newspapers which were uncategorized. One could observe, however, that most of the Ghanaian texts had political content in the sense that they almost always argued or commented on political debates, election processes and results, political campaigns and rallies, etc. Thus, in order to obtain adequate texts into the Ghanaian corpus, we focused on political editorials. Political editorials were operationalized in this study to mean texts which are about or comments on national elections and campaigns, national registration of voters, parliamentary and presidential debates, activities of government departments and agencies, speeches and activities by members of the four wings of government3, etc. These parameters equally influenced our selection of the types of political editorial texts that were compiled into the British corpus. The headlines of editorials for both British and Ghanaian newspapers selected for this study are listed in Appendix I. Texts selected from the four newspapers were the first three texts of each of the 12 months of 2012 whose contents suit the definition of political editorial above. A summary of the size of the corpus and the number of sentences and words are presented in Table 1 below. Table 1: Background information about the corpus DG GT DT TT TOTAL Number of Texts 36 36 36 36 144 Number of Sentences 700 534 774 1118 3126 Number of words 17666 14004 18027 21925 59022 Table 1 shows that though the same number of editorial texts was chosen from the four newspapers, the number of sentences attested varies from one newspaper to the other. This is not just because of stylistic variation but also because of the variable length of editorials from the four newspapers. One observes that the British newspapers typically have longer texts than the Ghanaian newspapers. Besides, there is internal variation in the length of editorials even from the same sociocultural context. For instance, The Times is about 3000 words longer than The Daily Telegraph and so is the Daily Graphic about 3000 words longer than the Ghanaian Times. To compensate for this, a normalization formula (c.f. Biber & Conrad 2009) was used to standardize the results (the detail of which is presented in section 7.2.2 below). 7.2 Analytical framework The analytical model adopted for this study is that of Biber and Conrad’s (2009) register model. According to their framework, a register-based study is a three-layered engagement, which requires a mixed methodological approach of quantitative and qualitative analysis. These include a description of the situational context, a description of the linguistic categories explored and an analysis of the functional relationship between the distributional patterns and the situational context. These tasks are engaged at the various sections of this study. For the purpose of analysis, the corpus of the study was tagged and the attestations normalized to facilitate comparability. 3 This operational definition is based on Oxford English Dictionary’s (OED) definition of political as “Of, belonging to, or concerned with the form, organization, and administration of a state, and with the regulation of its relations with other states.” 7.2.1 Tagging of corpus Since the editorial corpus was compiled purposely for this study, care was taken to annotate it with the grammatical categories studied in this work – the sentence. This task of annotation, which is very time-consuming (Mcenery & Gabrielatos 2006) because it is done manually, is very useful in facilitating the retrieval and counting of attestation. Afterwards, attested sentence types were retrieved using Microsoft Word search engine. In order to ensure that the attestations were not influenced by subjective biases, a national service personnel was trained as a checker who verified the tagging and the attested sentence patterns. 7.2.2 Normalizing attestations One principle that is crucial in a comparative register study is standardization of attested results either through percentage terms or normalization (Biber 1988; Biber & Conrad 2009). Standardization is essential because, usually, texts compared come in variable lengths, making the raw scores methodologically problematic. For instance, it is misleading to use the raw scores reported in Table 2 below. This is because, since the texts are of unequal lengths, it will be wrong to conclude based on these results that The Times, for example, used the highest number of declaratives. This is why it is necessary to standardize the scores to facilitate comparison. For the purpose of this study, normalization was used to standardize attested results. Table 2: Spread of functional and structural Sentence types in raw scores Functional DG GT DT TT Structural DG GT DT TT Declarative 690 520 751 1086 Simple 62 80 150 260 Interrogative 6 5 19 31 Complex 585 409 548 784 Imperative 4 8 2 1 Compound 18 15 26 34 Exclamative 0 1 2 0 Comp compl 35 30 50 40 Total 700 534 774 1118 Total 700 534 774 1118 Using Biber and Conrad’s (2009) framework, normed rates were computed out of the raw scores in Table 2 above with a fixed amount of text. For the purpose of this study, these raw counts were converted to rates of 100 sentences per newspaper. This is the rate at which a sentence type occurs in every 100 sentences. The formula for the conversion, according to Biber and Conrad (2009: 62), is: Normed rate = (raw count ÷ total sentence count) x the fixed amount of text Now using our fixed amount of 100 sentences per newspaper, the normed rate for the occurrence of declaratives in DG will be: (690 (sentences) ÷ 700 (total)) x 100 (sentences) = 98.57. This normalization formula above is applied to all the functional and structural sentence types and the normed rates of occurrence are reported in Tables 3 and 6 under sections 8.1 and 8.2 below. For purposes of comparability, a Chi-square test of independence is employed at alpha = 0.05. 8. Results and Discussions 8.1 Functional Sentence types The normed rates of occurrence for functional sentence types reported in Table 3 below present a very fair basis upon which comparison was based across the four newspapers investigated. To facilitate comparability, the Chi-square test was run on individual sentence types. This is because, the data variates the assumptions in the Chi-square distribution of having about 66.7% of the cells less than expected counts of 5. Table 3: Normed rates of occurrence for functional sentence types Sentence type DG GT DT TT  P-value Declarative 98.57 97.38 97.03 97.14 0.0308 0.9986 Interrogative 0.86 0.94 2.45 2.77 1.7416 0.6321 Imperative 0.57 1.50 0.26 0.09 1.9847 0.5756 Exclamative 0 0.18 0.26 0 0.4528 0.9291 Total 100 100 100 100 These results in Table 3 indicate that the functional sentence types are largely similarly distributed across the four newspapers – the declarative sentence being the dominant type (attesting >97), followed by the interrogative (hovering between 1 and 3) and the imperative (<2) consistently across the four newspapers. The least attested type (the exclamative sentence) attested almost nothing across the four newspapers. These patterns of distribution enable one to argue that choices of these sentence types are motivated by the situational context. That is, something about the newspaper editorial genre must have influenced the similarity in these linguistic choices. However, before we get to that argument, there is an observation that needs to be made. Though the Chi-square test run on all the four sentence types confirms that the distribution of the sentences does not depend on newspaper type since observable differences are statistically insignificant, one detects a pattern in the distribution of interrogative sentences which seems to distinguish Ghanaian editorials from British editorials. In the light of the register theory, results such as those for the interrogatives are deemed insignificant due to their low attestations. However, the regularity in their occurrence serves as a prod to investigate the interrogatives a little closely for dialectal reasons. A detailed analysis of the interrogatives reveals interesting observations. It is observable, from functional perspectives that the few attested interrogatives presented in Table 4 below are mainly the rhetorical type, especially for the Ghanaian editorials. This pattern of distribution of rhetorical interrogatives in the Ghanaian editorials seems to reflect the sociocultural context. It indicates a conscious effort of an editor to conform to norms of Standard English. That is, since Standard English is used in Ghanaian newspapers as an ideological plank to attract educated elite, as argued above, efforts are being made at carefully selecting grammatical patterns in conformity with Standard English. This variation is obvious when comparison is made with attestations in the British editorials. Observably, the British editorials used non-rhetorical types in combination with rhetorical interrogatives. This too may be explainable based on the sociocultural context. This use of interrogatives, perhaps, reflects interactivity, a phenomenon that may be attributed to the relatively more active presence of British newspapers online. During 2012 when the data were compiled, the two Ghanaian newspapers were not actively present online. The Times and The Daily Telegraph, however, were actively online and had created an interface for their readers to interact with the editors. And so the active presence of these newspapers online might have raised their consciousness of the need to draw closer to their audience through linguistic devices including interrogatives. Table 4: Spread of functional interrogative types across the four newspapers Functional DG % GT % DT % TT % Rhetorical Ques 6 100 5 100 12 75 24 75 Non-Rhetorical Ques 0 0 0 0 4 25 8 25 Total 6 100 5 100 16 100 32 100 A careful study of the corpus reveals that the dominant rhetorical questions seem to perform two communicative functions in the editorial data across the Ghanaian and British newspapers. They, first of all, raise polemical issues, which often serve as a plank to launch onto the next level of argumentation; a function that most of the attested interrogatives perform in the editorials. The set of interrogative sentences in the following excerpt from the data performs this function: (9) What is the ultimate impact of the crisis in the eurozone? How are we going to be able to afford all the necessities of an ageing population? And, looming above all these, how is Britain to equip its people with the resources to compete with economies that are already growing rapidly? These are the genuine leadership questions that Mr Cameron will face in Birmingham. (The Times) But this argumentative function performed by rhetorical questions in the data is quite different from the second function observed by Crystal and Davy (1973) according to which interrogatives are used to keep articles from dragging in newspaper reporting. The following example of interrogative sentence from the data is of the second type: (10) Had he, as the minister deciding on the purchase by News Corporation of the outstanding shares in BSkyB, acted inappropriately and favoured the company for political reasons? (The Times) In this rhetorical question, just like in the first above, the writer does not expect a direct answer from his/her readers; neither does he/she provide an answer himself/herself. The difference, however, is that the type of rhetorical question in sentence (10) is not used as a plank to launch onto an argument. Similar rhetorical questions found in the Ghanaian editorials include the following: (11) What do we seek to gain by preventing eligible persons from registering? (Daily Graphic) (12) Are the political parties going about their campaigns in a sensible and mature way? The answer lies in the incessant appeals to them to avoid acts which negate the gains so far made in our democratic pursuits. (Ghanaian Times) Equally noteworthy about the distribution of the interrogative sentences is the observation, captured in Table 5 below, that the most preferred structural patterns are the ones which do not just require YES or NO for an answer. They are the ones which have the potential of engaging their reading audience. Table 5: Spread of structural interrogative types across the four newspapers Structural DG % GT % DT % TT % Yes/No Ques 2 33.3 5 100 6 38 15 47 WH-Ques 4 66.7 0 0 10 62 17 53 Total 6 100 5 100 16 100 32 100 Though the usage patterns are not consistent across the four newspapers as seen in Table 5 (with GT as the odd one, conforming, perhaps, to a house style), it is clear that there is preference for WH-Questions. The following are examples of WH-Questions attested in the corpus. (13) What…would the protesters in Tahrir Square make of a nation in which more than 70 per cent of the upper chamber are political appointees, in which the legislature is "stuffed full of friends and colleagues of party leaders"? (The Daily Telegraph) (14) What happens to applicants who are unable to register at a centre whose equipment break down for three days? (Daily Graphic). Other equally important observations about the distribution of interrogative sentences, especially unique to the British newspapers, which have the highest representation of this type of sentence, include the following: • They do not seem to be evenly distributed. They tend to be used in successive order of two or three in particular texts whereas others do not have any interrogatives at all. In the example below, seven questions (Q1, Q2, Q3, Q4, Q5, Q6 and Q7) are asked one after the other in one text; and this is not an isolated pattern, especially among the British editorials. (15) (Q1) How about, for instance, appointing Boris Johnson as Party Chairman? (Q2) Does Mr Miliband want to follow a Blairite strategy of trying to keep Britain at the heart of decision-making, as implied by his party's relentless urging of David Cameron to ‘stay in the room’? (Q3) How does that fit with directing his party to vote for a position on the EU budget that would isolate the UK and diminish our influence? (Q4) And how does it affect Labour's policy towards the Liberal Democrats in the run-up to the next election? (Q5) Would Labour fight to repatriate powers from Brussels and if so, which? (Q6) What approach does he think Britain should take to a eurozone that is changing fast and in which a second division will soon be created: does he think the UK should be part of that second division? (Q7) If not, what role should it play? (The Times) • Interrogatives tend to appear as the last sentence of their texts. • Some interrogatives are premised upon preceding declarative sentences. As a result, they do not really make complete sense; neither do they appear to be complete sentences in isolation. For instance, in example (16) below, ‘But why?’ lacks both subject and predicate and is meaningless unless it is analyzed within its broader co-text. Similarly, ‘Why?’ in sentence (18) makes sense only when it is put after the preceding declarative. (16) But why? (The Daily Telegraph) (17) Yet it was overwhelmingly rejected. (The Daily Telegraph) (18) Why? (The Daily Telegraph) It is not surprising that all the examples of interrogatives that do not constitute complete sentences are from the British editorials. In the Ghanaian editorials, the grammatical rule of keeping sentences complete in order to avoid fragments seems to be among the guiding principles. This therefore seems to reflect the second language sociocultural context where English is used for formal purposes and the Standard English is an ideological link between the editor and his/her audience. Considering the patterns of distribution of the functional sentence types, the question that is outstanding at this point is what accounts for the regularities in the distribution of these grammatical structures in texts from different sociocultural contexts? In other words, what makes the declaratives similarly preferred across the four newspapers and the interrogatives, imperatives and exclamatives similarly dispreferred? These patterns of distribution are indicative of functionality in the framework of Biber and Conrad’s (2009) register model. They support the central claim of the register theory that linguistic features are similarly distributed across dialects of the same language. This pattern of distribution may be interpreted to be functional by arguing that declaratives relate with the situational context of the newspaper editorial register. In other words, the declarative sentence is dominant across the four newspapers because it is useful for the realization of the communicative function of editorials in propagating ideological positions. In relying predominantly on the declaratives, not only is the Ghanaian editor conforming to norms of Standard English. In doing so, the Ghanaian editor, just like the British editor, is making claims and statements about beliefs and opinions. Typically, it is the declarative sentence which is capable of performing a whole range of speech act functions, including expressing claims, hypotheses, descriptions, suggestions, promises, pledges, threats, assessment, facts, etc. (Finegan 2012). These are functions, one wants to argue, that are appropriate for the formulation of ideological positions in newspaper editorials. That is, the communicative functions of editorials to represent the ideological positions of their press houses are expressible through a variety of speech acts, which are realizable through several linguistic means, including the declarative sentence. The low representation of the non-declarative types (especially imperative and exclamative sentences), perhaps, suggests that they are not very essential linguistic strategies for formulating ideological positions. 8.2 Structural Sentence Types The structural sentence types – simple, complex, compound and compound complex sentences – are distributed in the editorials across the four newspapers in the patterns reported in Table 6 below; these are normed rates calculated using the formula under section 7.2.2 above. Here too, the Chi-square test was run on each of the sentence types. Table 6: Normed rates of occurrence for structural sentence types Sentence type DG GT DT TT  P-value Simple 8.86 14.98 19.38 23.25 6.8705 0.0765 Complex 83.57 76.59 70.80 70.13 1.5403 0.6730 Compound 2.57 2.81 3.36 3.04 0.1138 0.9901 Compound complex 5 5.62 6.46 3.58 0.8571 0.8358 100 100 100 100 Evidently, the results in Table 6 also reveal some level of consistency in distribution. Obviously, the complex sentence is consistently the dominant type across the four newspapers followed by the simple sentence, though after a wide margin. The compound complex and compound sentences are the least used structural types across the four newspapers. In spite of these patterns of consistency, one observes in the details that some variability exists in terms of how much of a particular structural sentence type was realized in a particular newspaper. For instance, though complex sentences in the two British newspapers are similarly distributed, the difference between attestations of the complex sentence for the two Ghanaian newspapers is too wide for one to argue for regional sway. The same dissimilarity exists in the distribution of the simple sentence across the four newspapers. However, despite these dissimilarities, the confirmatory Chi-square tests identify no statistically significant variation in the distribution of these sentence types across the four newspapers. In a sense, these distributional patterns support Hopper and Traugott’s (2003) observation: …to our knowledge human languages have had complex sentence structure available throughout recorded history. But reorganization of complex combinations is well evidenced…as is the association of certain complex sentence types with certain genres, especially of planned discourse” (Hopper & Traugott 2003: 177). This is one reason for the argument in this paper that the distribution of these sentence types is basically register-based, especially due to the level of consistency observed. In explanation, we argue that the communicative function and other situational features of newspaper editorials attract complex sentences among other linguistic devices as relevant structural tools In spite of these patterns of similarities, one observes strands of structural diversity across texts from the two sociocultural contexts. One observes, for instance, variability in the organization of the structural sentences across newspapers. It is observed that the Ghanaian editorials are typically texts whose paragraphs are predominantly made up of a single complex sentence. This is sharply contrasted by the British editorials whose paragraphs are made up of a number of complex sentences interspersed by one or two simple, compound or compound complex sentences. This may be argued to be indicative of linguistic versatility on the part of the native speaker at proficiently managing complex linguistic data, an assumption that is relevant to the sociocultural context. However, an observable variability in the situational context may be argued to be accountable for this variation in language use. One observes that though newspapers from the two contexts of this study may equally operate with some restrictions in the space allocated for their institutional editorials, the variation in the number of words (see Table 1 above) and space dimensions of editorials across Ghanaian and British newspapers are indicative of some of the forms of linguistic variation we have observed in this study. But structural variability is also evident in the way the immediate constituents of the dominant complex sentences are used. A close investigation of the complex sentences attested in the data reveals the following three patterns: • Pattern 1: Complex sentences in which a subordinate clause is an element of the sentence, functioning as a subject, object, complement or an adverbial. For instance, the underlined clauses in sentences (19) and (20) are adverbial and nominal clauses functioning as clause elements: (19) When the Budget speech is still leading the news three weeks after delivery, something has gone awry. (The Times) (20) But to surrender at this stage would send out the appalling message, which is precisely Labour's aim. (The Daily Telegraph) • Pattern 2: Complex sentences in which a subordinate clause is a post-modifier of a phrase, functioning as qualifier or phrase complement. The underlined clauses in excerpts (21) and (22) are a That-Relative clause functioning as a post-modifier (in excerpt (21)) and a That- nominal complement whose head is the noun ‘way’ (in excerpt (22)). (21) These can be called bye-laws or codes of ethics that serves as a guide for acceptable behavior or conduct. (Daily Graphic) (22) Of course, there should be some amount of humour to make proceedings in the House a bit interesting, but this should be done in such a way that discipline would not be undermined. (Ghanaian Times) • Pattern 3: Complex sentences which blend the first two. In sentence (23) below, we have, first, an adverbial clause ‘Although this arrangement is good’ which depends on the main clause which, itself, a matrix clause containing the complement clause ‘that…the entire exercise would end in confusion’ which complements (and therefore depends on) the adjectival head ‘obvious’: (23) Although the arrangement is good, it is obvious that without proper education, the entire exercise would end in confusion. Of these, Pattern 1 Complex sentences ((19) and (20) above) instantiate a hypotactic relationship. That is, clauses of unequal status (i.e. a main clause and a subordinate clause which depends on the main clause, rather than on a phrase) have been conjoined in one sentence (Quirk et al. 1985). Pattern 2 Complex sentences ((21) and (22) above) instantiate embedding or rank shifting. Our data reveal that there are more complex sentences with rank shifted/embedded subordination (Pattern 2 Complex sentences) than those with hypotactic ones across the four newspapers as Table 7 shows. Table 6: Normed representations of hypotactic and rank shifted relationships DG GT DT TT Hypotactic 42.1 38.1 46 42.9 Embedded 57.9 61.9 54 57.1 100 100 100 100 To arrive at these normed rates in Table 7, the raw attestations were converted to rates of 100 sentences per newspaper (c.f. 7.2.2 above). This is the rate at which a hypotactic or rank shifted relationship occurs in every 100 sentences. The evidence in Table 7 shows that the distribution of hypotactic and embedded/ rank shifted subordinate clauses in the editorials is independent of newspaper type. The consistent dominance of clause embedding across the four newspapers is perhaps an indication of the nature of the structural complexity of the newspaper editorial register. It seems rankshifting is among the linguistic means through which newspaper editorials achieve their communicative purpose. Put properly, rankshifting is more suitable to the situational context of the editorial register. However, the prevalence of rankshifted clauses is not as outstanding as that of the dominant declarative-complex sentence types. It is observable from Table 7 that even hypotactic clauses are appreciably attested across the four newspapers. This is, perhaps, an instructive indication that with ample data, new insights may emerge in the use of hypotactic and embedded structures within and across registers. 9. Implications of the findings to media discourse The findings of this study largely support the argument in the literature about the language of newspaper editorials. Until the advent of the World Wide Web, which allows newspapers to publish online, many aspects of the newspaper, including editorials, were deemed formal, elaborated and abstract (Biber 1988; Wiredu 2012; Fosu 2016; Frimpong 2017) not only because they were written texts but also because there was no direct interaction between the addresser and the addressee. And so complex declarative sentences were suitable for the situational context of newspaper editorials. During 2012 when the data were collected, the two Ghanaian newspapers studied in this paper were not actively online. Their British counterparts, however, were online with columns for their reading audience for correspondence. This crucial differentiation in the situational context of editorials from the two sociocultural contexts may be among the factors that have accounted for the disparity in the distribution of sentence types (especially simple and interrogative sentences). Thus, the higher representations of simple and interrogative sentences in the British newspapers, perhaps, are reflections of the consciousness of British editorialists of feedback from their reading audience. And this observation is important because as newspapers get closer to their audience, especially through the medium of the Internet and other social media platforms, adjustments in situational context is expected to reflect adjustment in situational context. 10. Limitations of the study It must be acknowledged that these sentence patterns have been compared in spite of variability in a number of contextual dynamics. And that is the audacity of the register theory which argues that in spite of these variations, choices of major linguistic elements in a register across different dialects of a language are normally similar (Biber & Conrad 2009). 11. Conclusion This paper has investigated the usage dynamics of the sentence in editorials from four newspapers across Ghanaian and British sociocultural contexts from register perspectives. It has been observed that the declarative and complex sentences are the dominant types across editorials from the four newspapers. This pattern of distribution, it has been argued, supports arguments in the register theory that since texts from the same register are produced in similar situational contexts, linguistic in texts from the same register are largely similar across dialects of a language. Though the Chi-square tests confirmed that any observable variation in distribution is statistically insignificant, we note that minor irregularities observed in the distribution of simple and interrogative sentences across the two sociocultural contexts reflect adjustments in the situational and sociocultural contexts of the editorials investigated. References Ansah, Gladys. 2008. Linguistic Diversity in the Modern World: Practicalities and Paradoxes. The International Journal of Language Society and Culture 26. Ansary, Hasan, & Babaii, Esmat. 2009. A Cross-cultural Analysis of English Newspaper Editorials: A Systemic-Functional View of Text for Contrastive Rhetoric Research. RELC, Vol. 40 (2), 211- 249. Asante, Clement E. 1996. The Press in Ghana: Problems and Prospects. New York: University Press of America. Biber, Douglas. 1988. Variation across speech and writing. Cambridge: Cambridge University Press. Biber, Douglas & Conrad, Susan. 2009. Register, Genre and Style. New York: Cambridge University Press. Biber, Douglas & Johanson, Stig & Leech, Geoffrey & Conrad, Susan & Finegan, Edward. 1999. The Longman grammar of spoken and written English. Cambridge: Cambridge University Press. Bloor, Thomas & Bloor, Meriel. 2004. The Functional Analysis of English. London: Hodder Education. Chan, Tak Wing & Goldthorpe, John. 2007. Social Status and Newspaper Readership. American Journal of Sociology, Vol 112 (No. 4), 1095 - 1134. Chien, Shouchun. 2014. Varieties of English: Taiwanese attitudes and perceptions. Newcastle and Northumbria Working Papers in Linguistics, Vol 20, 1-16 Crystal, David & Davy, Derek. 1973. Investigating English Style. London: Longman. Dako, Kari. 2003. Ghanaianisms - a Glossary. Accra. Ghana University Press. Dako, Kari & Huber, Magnus. 2004. Ghanaian English: morphology and syntax. In: Bernd Kortmann and Edgar W. Schneider (Eds), A Handbook of Varieties of English. A Multimedia Reference Tool. Vol. 2: Morphology and Syntax, 854 – 865. Berlin: Mouton de Gruyter. Daily Graphic. 2017. Graphic Communications Group Limited. http://www.graphic.com.org.gh. Dakubu, Mary Esther Kropp. 1988. The Languages of Ghana, London, Kegan Paul International. Denkabe, Aloysius & Gadzekpo, Audrey. 1996. WHAT IS FIT TO PRINT? Language of the Press in Ghana . Accra: FES. Downing, Angela & Lock, Philip. 2006. English Grammar: A University Guide (2nd Edition ed.). London: Routledge. Eggins, Suzanne. 2004. An Introduction to Systemic Functional Linguistics (2nd Edition ed.). London: Continuum International Publishing Group. Fawcett, Robin. 2000. A Theory of Syntax for Systemic Functional Linguistics. Amsterdam: John Benjamin. Finegan, Edward. 2012. Language: Its Structure and Use (6th ed.). United States: WADSWORTH CENGAGE. Flowerdew, Lynne. 2004. The Argument for Using English Specialized Corpora to Understand Academic and Professional Language. In U. Connor and T. Upton (eds) Discourse in the Professions: Perspectives from Corpus Linguistics. Amsterdam: John Benjamins, pp. 11–33. Fosu, Modestus. 2014. The Press and Political Participation: Newspapers and the Politics of Linguistic Exclusion and Inclusion in Ghana. The University of Leeds, Institute of Communication Studies. Unpublished PhD Thesis. Fosu, Modestus. 2016. A linguistic description of the language of Ghanaian newspapers: implications for the readability, comprehensibility and information function of the Ghanaian press. Ghana Journal of Linguistics 5.1: 1 - 36. Frimpong, Kodie George. 2015. A Comparative Register Analysis of Editorials from Ghanaian and British Newspapers. University of Ghana, Department of English. Accra: Unpublished PhD Thesis. Frimpong, Kodie George. 2017. Subordination Across Ghanaian and British Newspaper Editorials: A Register Perspective. Ghana Journal of Linguistics 6(1). 72 - 115. Frimpong, Kodie George. 2018. The complex sentence across written genres from native and nonnative contexts; a corpus-based study. SKASE Journal of Theoretical Linguistics 15. 3: 185 - 203 Ghanaian Times. 2017. New Times Corporation Ltd. http://www.ghanaiantimes.com.gh Greenbaum, Sydney. 1996. Oxford English Grammar. Oxford: Oxford University Press. Guerini, Federica. 2008. Multilingualism and language attitudes in Ghana: a preliminary survey. ETHNOREMA IV. 4: 1- 33 Halliday, Michael A.K. & Matthiessen, Christian. 2004. An Introduction to Functional Grammar (3rd Edition.). London: Hodder Education. Halliday, Michael A.K. & Matthiessen, Christian. 2014. Halliday's Introduction to Functional Grammar (4th Edition). London: Routledge. Hopper, Paul & Traugott, Elizabeth Closs. 2003. Grammaticalization (2nd Edition). Cambridge: Cambridge University Press. Huber, Magnus. 1999. Ghanaian Pidgin English in its West African Context. Amsterdam: Benjamins Huber, Magnus. 2012. Syntactic and variational complexity in British and Ghanaian English: Relative clause formation in the written parts of the International Corpus of English. In B. Kortmann, & B. Szmrecyani (Ed.), Linguistic Complexity: Second Language Acquisition, Indigenization, and Contact (pp. 218-242). Berlin: Walter de Gruyter. Huber, Magnus. 2014. Stylistic and sociolinguistic variation in Schneider's Nativization Phase: T- affrication and relativization in Ghanaian English. In S. Buschfeld, T. Hoffmann, & M. Huber (Ed.), The Evolution of Englishes: The Dynamic Model and Beyond (pp. 86-106). Amsterdam: John Benjamin. Huddleston, Rodney. 1999. Sentence Types and Clause Subordination. In K. Brown, & J. Miller (Ed.), Concise encyclopedia of grammatical categories (pp. 329- 343). Amsterdam: Elsevier. Kachru, Braj (1981). The Pragmatics of Non-Native Varieties of English. In Smith L. E. (ed.). English for Cross-Cultural Communication (15 - 39). London: Palgrave Macmillan. Koester, Almut. 2010. Building small specialised corpora. In A. O'Keefe & M. McCarthy (Eds.) The Routledge Handbook of Corpus Linguistics (pp. 66 - 79). Routledge: London and New York. Kuhi, Davud, Mojood, Manijheh. 2014. Metadiscourse in Newspaper Genre: a Cross-linguistic Study of English and Persian Editorials. Elsevier: Procedia - Social and Behavioral Sciences 98, 1046 – 1055. Le, Elisabeth. 2009. Editorials’ genre and media roles: Le Monde’s editorials from 1999 to 2001. Journal of Pragmatics 41 1727–1748. McEnery, Tony, & Gabrielatos, Costas. 2006. English Corpus Linguistics. In B. Aarts, & A. McMahon (Ed.), The Handbook of English Linguistics (pp. 33-71). Malden: Blackwell Publishing. Morison, Stanley. 1932. The English Newspaper, 1622 - 1932: An Account of the Physical Development of Journals Printed in London. Cambridge: Cambridge University Press. Obeng, Samuel. Gyasi. 1997. An Analysis of the Linguistic Situation in Ghana. In African Languages and Cultures, Vol. 10, No. 1 (1997), 63-81. Oxford University Press. 2015. Oxford English Dictionary. Oxford: Oxford University Press. Payne, Thomas. E. 2011. Understanding English Grammar: A Linguistic Introduction. Cambridge: Cambridge University Press. Quinn, Anthony. 2013. Magforum. Retrieved March 3, 2015, from Magforum Web site: http://www.magforum.com. Quirk, Randolf & Greenbaum, Sydney & Leech, Geoffrey & Svartvik, Jan. 1985. A comprehensive grammar of the English language. Edinburgh: Longman Group Limited. Sadock, Jerrold & Zwicky, Arnold. 1985. Speech Acts Distinctions in Syntax. In S. T. (Ed.), Language Typolohy and Syntactic Descrition. Vol. 1. Cambridge: Cambridge University Press. Sey, Kofi. 1973. Ghanaian English: An exploratory survey. London: Macmillam. Schmeid, Josef. 1990. Corpus linguistics and non-native varieties of English. World Englishes Vol 9:3, 255 - 266. Schneider, Edgar. W. 2007. Postcolonial English: Variations around the World. Cambridge: Cambridge University Press. Silverman, David. 2011. Interpreting Qualitative Data. London: SAGE. Simo-Bobda, Augustine. 1997. English in a multilingual society. In M. E. Kropp Dakubu (Ed.), English in Ghana. Accra: Ghana English Studies Association Shokouhi, Hossein. & Amin, Forough. 2010. A Systemist 'Verb Transitivity' Analysis of the Persian and English Newspaper Editorials: A Focus of Genre Familiarity on EFL Learner's Reading Comprehension. Journal of Language Teaching and Research, Vol. 1, No. 4, 387-396 The Daily Telegraph 2017. Telegraph Media Group Limited. www.telegraph.co.uk The Times. 2017. Times Newspapers Limited. http://www.thetimes.co.uk Van Dijk, Teun. 1989. Race, riots and the press: An analysis of editorials in the British press about the 1985 disorders. Gazette (43), 229 - 253. Wardhaugh, Ronald. 1995. Understanding English Grammar. Oxford: Blacwell. Westin, Ingrid & Geisler, Christer. 2002. A multi-dimensional study of diachronic variation in British newspaper editorials. ICAME Journal (No 26), 133 - 152. Wiredu, John Franklin. 2012. A Grammar of Newspaper Editorial language: the Complex Sentence. Legon Journal of Humanities 23. 75 - 96. Wiredu, John Franklin. 2014. The Complex Sentence in Legal English: A Study of Law Reports. UNILAG Journal of Humanities 2(02). 1-27. Yankson, Isaac Kofi & Browne, Edmund; Tagor, H; Donkor, Peter; Quansah, Robert; Asare, George Ernest; Mock, Charles & Ebel, Beth. 2010. Reporting on road traffic injury: Content analysis of injuries and prevention opportunities in Ghanaian newspapers. Injury Prevention. 16. 194- 197. Young, Johnstone Tony & Walsh, Steve. 2010. Which English? Whose English? An investigation of 'non-native' teachers' beliefs about target varieties. Language, Culture and Curriculum, Vol 23: 2, 123-137 APPENDIX 1: Editorial Headlines and other corpus information Newspaper No of words) Editorial Headline The Daily Telegraph DT01 443 Osborne bears good tidings to the East DT02 697 Rich or poor, we are all paying too much tax DT03 434 A shabby episode that Cameron may regret DT04 441 These Tory MPs are in tune with the country DT05 686 This is not the time for surrender on the NHS DT06 441 Faith must not be driven from Britain’s public life DT07 435 A tax that stalls the engines of recovery DT08 715 A pointless, politicized plan for Lords reform DT09 446 A budget for Britain, not for the Lib Dems DT10 442 Voters have no time for political manoeuvring DT11 434 We must stand up to the Strasbourg court DT12 714 Britain’s cities should say yes to mayors DT13 754 Re-elect Boris – then give him more powers DT14 719 There is still a route to a Conservative majority DT15 433 Stability has been the Coalition’s key success DT16 445 The baroness and a case of double standards DT17 440 The comfortable course is no longer an option DT18 421 Regional pay plan scuppered by politics DT19 711 The corrosive crisis of trust in our institution DT20 445 Tories must resist Clegg’s ridiculous blackmail DT21 448 Clegg’s foolish fantasy comes crashing down DT22 437 Open government is here to stay, like it or not DT23 448 Clegg shows his true colours on equal votes DT24 439 Clegg cares more about revenge than fairness DT25 437 A new team – but the problems don’t change DT26 441 More can still be done to get Britain growing DT27 434 This defence deal is fraught with danger DT28 447 Labour cannot be trusted to rebuild Britain DT29 593 So much for the Tories’ ‘secret weapons’ DT30 441 Sniping at the rich is futile and damaging DT31 436 Nick Clegg is on the wrong side of history DT32 710 Until voters feel involved, localism is a lost cause DT33 443 Vince’s mansion tax rises from the dead DT34 438 Osborne is still spending more than we can afford DT35 711 A lot of froth over Starbucks and tax DT36 456 Leaders’ debates are good for democracy The Times TT01 555 University Challenge TT02 620 Devolutionary Road TT03 561 Capitalism and Labour TT04 615 No Escape TT05 627 The Politics of Power TT06 558 Revising Legislation TT07 640 Court Out TT08 634 A Budget for Aspiration TT09 638 Special Effects TT10 545 Petrol Gauge TT11 633 Charity Case TT12 668 The Trouble Ahead TT13 559 The Choice for London TT14 637 Leading the Argument TT15 612 The Message TT16 628 The Hot Potato TT17 564 Cracking the Code TT18 562 In, Out, Shake It All About TT19 620 Politicians and Bankers TT20 643 House of Cards TT21 618 A Risible Delay TT22 571 Avoiding the Issue TT23 623 Judo Diplomacy TT24 652 Taxing the Rich TT25 617 Changing Places TT26 565 A Sorry Affair TT27 631 The Party is Over TT28 637 Vintage Labour TT29 650 Party Politics TT30 621 Defensive Merger TT31 620 Labour and Europe TT32 558 Working Together TT33 637 A Plane Mistake TT34 614 Aid and Abetting TT35 659 Home Truths TT36 543 The Difficult Decade Daily Graphic DG01 495 Everybody’s Peace DG02 561 Preparing for Election 2012 DG03 558 Strengthening internal Democracy DG04 544 Good Move but… DG05 624 PNC at the Crossroad DG06 577 Polarization Destroys Ghana DG07 489 Bright prospects await Ghana DG08 551 Guaranteeing continuity in change DG09 532 Working towards peaceful polls DG10 647 More Room for Improvement DG11 568 Sanity must Prevail DG12 496 Resolving Electoral Blues DG13 564 Well done, EC, but… DG14 569 Strengthening Ghana’s democratic system DG15 479 Another feather in Ghana’s democratic cap DG16 560 Towards peaceful transition DG17 502 EC, fix this anomaly DG18 536 Media shouldn’t declare election results DG19 505 Let’s die for Ghana, not for… DG20 577 A shot in the foot DG21 511 Ghana in tears DG22 566 Time to bond DG23 528 We are of age DG24 504 Towards credible polls DG25 481 Patronise voters register exhibition DG26 677 Peace before anything else DG27 612 Make room for jaw-jaw, not… DG28 527 Put Ghana first DG29 739 Upholding campaign promises DG30 557 Let’s play by the rules DG31 553 Deepening Ghana’s democracy DG32 497 Let’s learn from the US elections DG33 514 Let’s support the NCCE DG34 511 Let’s lead by example DG35 522 Resolving electoral disputes DG36 577 Politicians should stop funding macho men Ghanaian Times GT01 388 Protecting our infant democracy GT02 379 The President’s portraits and matters arising GT03 390 Lessons from the primaries GT04 442 Enough of the vain promises! GT05 366 Time for mature politics! GT06 377 Respect the president! GT07 647 Ensuring a credible voters register GT08 389 Getting all voters to register (1) GT09 378 Getting all voters registered (2) GT10 399 Making district assemblies efficient GT11 442 A good decision, Mr Minister! GT12 375 A worthy decision! GT13 398 Who should resign GT14 437 Let's have a united CPP GT15 343 Registration of prisoners GT16 373 Parliamentary fraud GT17 340 Ending the political mudslinging GT18 324 The dignity of the presidency GT19 422 By-elections in election year? GT20 350 Taking parliamentary business serious! GT21 342 Dialogue is the answer! GT22 350 Time to reflect on the economy GT23 318 The President must rest GT24 380 Ending the political mudslinging GT25 383 Towards a credible voters register GT26 340 Don't toy with our education! GT27 361 Politicians and promises GT28 384 Election 2012 and small arms proliferation GT29 353 Don’t sabotage our education! GT30 375 Rid the electoral process of jokers GT31 364 The issue of minors in election 2012 GT32 359 Keeping peace alive GT33 339 The heat is on! GT34 259 All must play by the rules GT35 447 All must accept outcome of election GT36 777 Election 2012; decision day today! George Kodie Frimpong Department of English P. O. Box LG129 University of Ghana Legon – Accra Ghana – W/A gfrimpong@ug.edu.gh frimpong.kodie@gmail.com In SKASE Journal of Theoretical Linguistics [online]. 2019, vol. 16, no. 2[cit. 2019-06-24].Available on web page http://www.skase.sk/Volumes/JTL40/pdf_doc/06.pdf. ISSN 1336-782X.