Participant’s Roles and Cognitive Relations in Niger Delta Conflict News Discourse

Media studies on Niger Delta (ND) conflict discourse have largely utilized stylistic, pragmatic, and critical discourse analytical tools in exploring media representation of news actors and ideologies in news texts but have not accommodated such issues as participants’ roles and cognitive relations in the discourse. This paper analyses the contexts of ND conflict news reporting with a view to revealing not only the participant’s role relations involved, but also the lexico-semantic resources they are characterized by. Forty newspaper reports on ND conflicts (20 from four ND-based newspapers—The Tide, New Waves, The Pointer and Pioneer, and 20 from four national newspapers—The Punch, The Guardian, Vanguard and THISDAY), published between 2003 and 2009, were sampled and subjected to discourse analysis, with insights from van Dijk’s context models and aspects of relational semantics. Four types of role were identified, viz. interactional (embracing the participants in conflict), communicative (relating to the production roles), social (involving group membership), and instrumental (dealing with the entities utilized in actualizing specific goals). The cognitive foci of these roles are associated with participants’ goals and beliefs, and these inform the participants’ position and hence role in the conflict events. Linguistically, the interactional and social roles are marked by synonymous and converse lexical items, while the communicative and instrumental roles are indexed by homonymous and antonymous lexical features. The findings corroborate the fact that there is an interaction between participant roles and cognitive relations in the ND conflict events reported in Nigerian newspapers.


Introduction
The Niger Delta (ND), being in the South-south region of Nigeria covering some 70,000 kilometers (Rowell et al., 2005, p. 9), yields an estimated 2, 500,000 (402,000 m 3 ) barrels of crude oil daily. This accounts not only for over 95% of Nigeria's crude oil and gas resources, but also 92% of the country's foreign exchange earnings (Ogbogbo, 2005, p. 169;Oluniyi, 2017, p. 23). The massive environmental degradation of the area, which follows many years of oil exploration, is worsened by lack of jobs and infrastructure, industries, and employment, thereby exposing the people to severe health hazards and reduces their supply of food from fishing and agriculture (Chika, 2016;Kadafa, 2012).
Therefore, in order to address the peculiarities of the problems confronting their region, the ND peoples have embarked on a long and continued struggle for self-determination. This is aimed at controlling the resources from their "fatherland" (Darah, 2008). The struggle, according to Ononye and Osunbade (2015), has two primary motivations; namely, to arm-twist the Nigerian government into giving ND people a better economic and political posture in the country, and to attract international communities' attention to the injustice perpetrated in the region by successive Nigerian governments (p. 18). The physical armed conflicts that ensued involve two major participants or groups in the discourse; namely, the Federal Government of Nigeria (including the Joint Military Task Forces and other officials) and the ND (comprising the different militant groups and their sympathizers).
In corollary, the violence, including the vandalization of oil facilities, kidnapping of expatriate oil workers, and clashes with the governments' task forces, have consistently been reported in the newspapers and other media within and outside the region. These reports have produced an important sub-genre of ND discourse known as ND conflict media discourse (Ononye, 2014), which has attracted the attention of many scholars in Nigeria. In what follows shortly here, we reflect on what some earlier scholars have done on the ND discourse sub-genre with a view to situating the present attempt within the gaps in the earlier studies.

Situating the Research
Existing scholarship on ND discourse can be categorized into linguistic and non-linguistic studies. The latter category, constituting the largest body of works, spans the history/origins of ND struggle (Aboribo & Umukoro, 2008;Ebienfa, 2011), the economic and political implications of the conflicts (Abegunde, 2013;Ibaba & Ikelegbe, 2010), and the initiatives taken to resolve the crisis (Otite & Umukoro, 2011). The former category, which is more relevant to this study, shows that linguistic studies on ND conflicts have yielded research interests largely from pragmatic (e.g., Ayoola, 2008), critical linguistic (e.g., Agbedo, 2012;Ayoola, 2010;Chiluwa, 2011aChiluwa, , 2011bOnonye, 2015Ononye, , 2017, and stylistic (e.g., Chiluwa, 2007;Ononye & Osunbade, 2015) perspectives. Ayoola (2008), for example, uses pragmatic tools in exploring the discourse strategies employed by news participants in fostering specific ideologies in the ND discourse. With a pragma-linguistic framework, Ayoola (2008) examines the setting, topics and participants that were projected in the content and context of news reports on ND in selected Nigerian newspapers. The study reveals that "Niger-Delta discourse participants also resorted to several pragmatic and discourse strategies, such as the force of logic, the use of figures and percentages, the persuasion of science, interdiscursivity and intertextuality, rumour mongering, name calling, dysphemism, obfuscation and flattery" (Ayoola, 2008, p. 18).
Within the critical linguistic category, Ayoola (2010) employs CDA tools in presenting the ideological perspectives of news reporters in the news discourse. This differs from Ayoola's (2008) focus on the pragmatics of news actors' use of language. Chiluwa (2011b) applies CDA in examining the role of the media in manipulating public opinion and people's perception of the impact of the Government's Joint Military Task Force in the ND crisis. In a similar study, Chiluwa (2011a) examines the discursive construction of Nigeria's ND ethnic militias in the Nigerian press, but takes a corpus linguistic methodology. Both studies maintain that media representations, motivated by the Nigerian government's intention to attach a negative social attitude on the ND militia groups, are ideological. Ononye's (2015) study adopted a critical-stylistic framework through which he explores how the different register labels manifest in political news reports on ND crisis. He identifies five register labels-oil struggle, oil facilities, crisis location, news reporting, and participant-related registers-and suggests that the labels have specific ideologies, which can be undertaken by subsequent studies. For Ononye (2017), specific media ideologies by the media inform specific stylistic choices made in order for news texts to do such things as: framing participants, evaluating specific entities, and reducing the impact of the activities of specific news actors. Generally, these critical works can be distinguished from the present study particularly in terms of theoretical focus. Instead of media representations, the present study focuses on the different roles played by different discourse participants, and their cognitive relations in the conflict events reported.
The closest works to this study, however, are those of the stylistic category, where Chiluwa (2007) and Ononye and Osunbade (2015) come in. Here, Chiluwa (2007) investigates the peculiar discourse stylistic features used in news texts on ND. The study demonstrates how these features are utilized by news reporters as mediated means of representing the social actions and interactions in the conflict discourse. However, unlike the current study, he neither pays attention to ideological issues nor role relations in the texts. In a similar vein, Ononye and Osunbade (2015) explore naming strategies, privileging the stylistic strategies through which entities are named or described in ND conflict discourse. They identify metaphorization as one typical stylistic strategy which does not only allow news reporters to name specific entities in the conflict discourse, but also bring in their ideological evaluation of the entities. The study, however, does not accommodate how specific linguistic choices in news texts can betray news participants' roles and cognitive relations.
The studies reviewed have generally provided insights into the linguistic and critical discourse analytical dimensions to media reports on ND conflict discourse, which are only capable of revealing media representation of news actors and ideologies, but not necessarily the participant's role relations in the discourse, the understanding of which requires a theoretical context. The present study therefore goes further to explore such contextual elements as the roles played by various discourse participants (e.g., news actors, their environment, the instruments used, etc.), the cognitive bases of the roles, and the effect these have on the lexicosemantic forms deployed in the newspaper texts under investigation. Theoretically, the use of a combination of context models and aspects of lexical semantics, as the analysis will demonstrate, handles the representation of participant's role relations and their lexical indices better than previous studies' excessive use of CDA. Therefore, identifying the participant's roles and the way they cognitively relate in the discourse can provide the needed understanding of news texts (van Dijk, 2000, p. 22), specifically the proliferation of the conflictual events in the ND region. Hence, this informs the aim of the present study, which is to examine the participant's role relations that characterize the ND conflict discourse. This triggers off two research questions, namely: (a) what are the participant's roles and cognitive relations that manifest in ND conflict news discourse? And (b) how are they accounted for through the relational-semantic forms deployed in Nigerian newspapers? These questions have influenced the choice of the analytical framework below.

Theoretical Underpinnings
To pursue the objectives above, the present paper is anchored on two theories/concepts; namely, van Dijk's socio-cognitive model of context and aspects of the relational semantics idea of lexical relation.

Van Dijk's Socio-Cognitive Model of Context
The major thesis of van Dijk's ideas on text-context relation is that language users (both the news writer and reader, in this case) attend to issues in the discourse depending on their mental models of the communicative situation, which keep track of what language users find interesting or important to (for example) report (and even project in the front page or headlines) or read (and even get emotionally attached or influenced). It is in these reporting and reading models that news participants represent, among others, themselves, other participants and their relations with them, current time frames, location and direction, and social actions going on in the news (van Dijk, 2001, p. 19). It is, therefore, in this connection that van Dijk (1999) asserts that social situations cannot directly influence language use unless through a cognitive interface, which spells out how social situations are constructed, and hence interpreted by participants.
Like most of van Dijk's theoretical models, context models, or what he simply refers to as "contexts" (van Dijk, 1999(van Dijk, , 2001, are mentally represented in communicative events and social situations as they constrain the patterns discourse takes. One remarkable value of van Dijk's model of context, which differentiates it from most other theories of context, is its mentalist view of contexts, for example, as "not 'out there,' but 'in here..'.; [as] mental constructs of participants; [as] individually variable interpretations of the ongoing social situation" (van Dijk, 2001, p. 18). What this means is that context is conceptualized in the mind, which allows discourse participants to adequately contribute to ongoing discourse; to produce and understand speech acts; and to adapt topics, lexical items, style, and rhetoric to social events.
However, context models, according to van Dijk (2001) "are not static mental representations, but dynamic structures" (p. 18; emphasis original). In his view, context is constructed and reconstructed in real-time by each participant in an event, and it changes with the continuous changes in the goings-on and interpretations of situations. These changes allow participants to update their mental models (what they know) about the situations or events. Hence, as Mey (2000, p. 39) corroborates, context includes the conditions under which text production and consumption take place (cf. Maynard, 2006). van Dijk (2001) proposes a list of categories for the organization of context models as subjective representations of communicative events or situations, which he succinctly refers to as "structures of context" (p. 21). They are domain, setting, cognition, and participant roles. Domain relates to activity type, which manifests in the social actions going on in the discourse; setting includes the location (temporal and spacial dimensions) of events; cognition is defined in terms of the goals, knowledge, and other beliefs of the participants; and finally, participant roles relate to the various roles which affect the production and comprehension of discourse.
The four categories of context by van Dijk are all relevant but considering the objectives of this paper, more insights are drawn from participant roles and cognition. Cognition is the host of the semantic and pragmatic properties of discourse in that it establishes the basis for the interpretation of participants' actions and intentions. Participants, in any discourse, have different roles, and such roles may really affect, in van Dijk's (2000) terms, "the production and consumption of discourse" (p. 22). van Dijk (2001) identifies three basic types of roles in which participants generally engage in, namely: interactional, communicative, and social roles. Interactional roles account for the various participants involved in physical contact, such as, security officials, kidnap suspects, and so on; communicative roles account for the various production roles in institutional or authoritative situations (which come up after the interactional contact), such as, eye-witnesses/health or security authorities, newspaper correspondents, and so on; while social roles relate to group membership, which captures the ideological inclinations of participants. The relations between these roles involve a vast area of cognition, which controls virtually all levels of discourse, especially the lexical choices, in the instance of the present study.

The Concept of Lexical Relations
The consideration of the patterning and meaning-related properties of lexical items in texts gave rise to the concept of lexical semantics. In Cruse's (2006) view, lexical semantics was originally taken from the viewpoint of structural semantics to include sense relations and other lexical nuances in texts (pp. 163-164). The sense of a word is the sum of its sense relations with other words in the language. Considerations of word sense bifurcate into the Saussurean distinction between paradigmatism and syntagmatism (Ononye, 2015, p. 370). Paradigmatic relations are concerned with associations of similar categories, where the words involved stand in complementary distribution (Geeraerts, 2010, p. 58). These relations hold between items that can occupy the same position in a grammatical structure. This is akin to what Lyons (1968) calls "off-line similarity" (p. 431), which typically "involve[s] words belonging to the same syntactic category" (Cruse, 2000, p. 148). Paradigmatic sense relations commonly manifest in terms of synonymy, hyponymy, meronymy, and different kinds of opposition. Synonymy, according to Saeed (2004), is the lexical relation involving different phonological words with similar meaning, which are derived from a number of parameters, viz. different dialects (e.g., tap and faucet), different registers (e.g., wife and spouse), collocational restriction (e.g., boy and lad), and portraying positive/negative attitude of the user (e.g., activist and militant) (p. 65). Hyponymy (derived from Greek: hypomeaning "under") is the lexical relation of class-inclusion described in English by the phrase "kind/type/sort of." "A chain of hyponyms defines a hierarchy of elements" (Riemer, 2010, p. 142), where for example hibiscus, tulip, and rose are co-hyponyms of flower, which is their hypernym. Meronymy (Greek meros: "part") is the relation of part to whole, where the part (e.g., eye) is referred to as a meronym of face, while the whole (e.g., face) is known as the holonym of eye (Riemer, 2010). The notion of oppositeness embraces several different types of relation, the most common of which is antonymy or complementarity. Antonymy is characterized by a relationship of incompatibility between two items with respect to some given dimensions of contrast. Some words, for example, may be associated with more than one antonym with respect to the dimension of contrast involved (e.g., girl has both boy and woman depending on whether the dimension of contrast is sex or age; sweet has both bitter and sour: Murphy, 2003, p. 173). Generally, words that stand in paradigmatic relations should be of the same grammatical category, but sometimes they are not. For instance, there is no hypernym of which the following adjectives are hyponyms: round, square, oval, oblong, and triangular. However, they are all related in a hyponymlike way to the noun shape. Relations of this type are sometimes called 'quasi-relations'; the commonest of these is quasi-hyponymy.
Syntagmatic sense relations hold between items in the same grammatical structure. There is the possibility of a lexical element in a text to co-occur in larger wholes with other elements of the language in terms of, for example, compounds and derivations in the morphological realm, and constituents and sentences in syntax (Geeraerts, 2010, p. 57).
Here, relations between individual items are not usually given names on the lines of hyponymy, antonymy, and so forth, but certain effects of putting meanings together are recognized, such as anomaly (e.g., a light green illness), pleonasm (e.g., dental toothache), and meaning extension, such as metaphor (e.g., move mountain) and metonymy (e.g., nice wheels). The requirements for a "normal" combination are described as selectional restrictions or selectional preferences. For instance, it is by virtue of syntagmatic sense relations, in this case between verb and noun that Fred ran across the field is normal, whereas The field crawled across Fred is odd. As opposed to paradigmatic relations, syntagmatic relations constitute "on-line co-occurrence" (Lyons, 1968, p. 431). For Cruse (2000), paradigmatic sense relations are "an expression of coherence constraints" while syntagmatic relations are "an expression of such structuring" (p. 149). One relevant insight derived from this review is that paradigmatic and syntagmatic relations function concurrently in revealing the contexts of language use in ND newspaper texts, and this is where van Dijk's socio-cognitive notion of context models becomes relevant.

Methodology and Analytical Framework
The study is content analysis, which is a research tool used to determine the presence of certain words or concepts within a given text (Alenezi et al., 2018, pp. 205-251). This is chosen because it provides a qualitative understanding of participant's role relations in ND conflict discourse through the analysis of the contextual and lexico-semantic features deployed in reporting the conflict news events. Thus, based on the research questions guiding the study, the data for this study consist of print newspaper reports on ND conflicts, published between 2003 and 2009. To research ND conflicts here is not based on random pockets of violence in the ND region, but rather we focused on the social activities identified with the major ND conflict participants in an earlier broader study (see Ononye, 2014). These conflict activities have been categorized in Table 1 below. The four-year period selected captures President Olusegun Obasanjo's last tenure, which was when the ND conflicts escalated and got worse.
Because of the large volume of content, we chose 40 out of the 83 reports capturing the conflict activities over the period selected. These were purposively sampled, five each, from four ND-based newspapers (namely, The Tide, New Waves, The Pointer and Pioneer) and four national newspapers (namely, The Punch, The Guardian, Vanguard and THISDAY). For ease of reference in analysis, the ND-based newspapers are coded TEXTs 1 to 20 while the national ones are coded TEXTs 21 to 40. The two newspaper sets (national and ND-based) used were not only as a result of their comparative consistency in reporting the conflicts over the period selected, but also because of previous studies' undue focus on national newspapers alone. Thus, the study seeks to balance the views of national and ND-based newspaper reporters, especially on the issues of participant roles and cognitive relations in the conflict events reported.
As earlier pointed out, the theoretical insights drawn from van Dijk's (2001) model are limited to two categories of contexts, viz. participant roles and cognition. Therefore, with regard to our first research question, the three types of participant roles proposed by van Dijk (2001) have been modified, with a fourth one added; namely, instrumental role. Therefore, aside from the original interactional, communicative and social roles reviewed above, the supplementary instrumental roles account for the various entities (animate or inanimate) that are utilized by participants in the discourse for the actualization of specific goals. This corroborates Ononye's (2017) position that participant roles in discourse analysis go beyond animate participants to include other entities that may (in)directly enhance the realization of active participants' goals or beliefs (p. 18). With respect to our second research question, exploring the participant roles and cognitive relations in ND conflict news discourse is anchored on the analysis of the paradigmatic and syntagmatic lexical choices made in the data. Thus, establishing the participant's roles, cognitive relations, and their linguistic indices is expected to enhance the reader's understanding of the group-induced motivations for the conflict events. In the following section, therefore, we analyze the data and discuss the findings that follow.

Presentation of Findings
The discussion of findings is guided by the research questions earlier set. Hence, both questions 1 (identifying the participant's roles and cognitive relations) and 2 (accounting for lexico-semantic forms) are attended to in the course of the discussion.
Thus, the table below summarizes our findings on the participant's roles with respect to the respective social activities going on in the discourse. In the discussions that follow, we identify and describe the lexical choices and cognitive relations that are associated with the roles.
In an earlier study, Ononye (2014) identified four social activities going on in the conflict discourse; namely,  disruption of oil operation (DO), conflict control (CC), armed struggle (AS), and human-abduction (HA) (pp. 80-119). From these, 10 global actions were revealed as the actions that news participants are directly engaged in. For example, DO involves vandalization of oil facilities, and interruption of oil activities; CC has to do with amnesty for the ND militants, apprehension of perpetrators, and prosecution of offenders; AS includes ethnic, rescue, and reprisal clashes); and HA embraces kidnap and release of expatriates and government officials. (The italicized words, which are also represented in Table 1, are the focus of the global action described.) The categorization of the different participant roles in Table 1

Interactional Roles
In the DO activity, there is predominantly one interactional role found, that is, the ND insurgent groups that are involved in vandalizing oil facilities and interrupting oil services. This is about the only social activity in the discourse where only one group of participant roles is found. Of course, there are other participants here (e.g., the oil facilities that are vandalized, etc.), but they are categorized and discussed under "Instrumental Participants" below. There are observable instances of synonymy in terms of describing the different roles played by the major participants in the DO activity. For example, lexical items like "gunmen," "unknown gunmen," and "suspected gunmen"; "militants," "Nigerian militants," and "militant group"; "vandals," "suspected pipeline vandals," are all used in describing the same group of participants. These items are used irrespective of the global action engaged in by the participants. In the CC activity, two broad interactional roles are identified as cutting across the global actions of amnesty, apprehension, and prosecution; they are ND insurgent groups versus government officials. In the case of amnesty, it is between the ND groups and government (amnesty committee) officials; for apprehension, their roles respectively change to suspects (of various crimes including vandalization, kidnapping, human trafficking, etc.) and government law enforcement agents (police, army, Immigration, NAPTIP, etc.). With regard to prosecution, a set of participants is found, this time, accused persons versus government courts or judges, respectively. Lexicalization is reduced here in that the lexical choices are constrained to the conventional terms given to these participant roles in the domain of activity. Let us examine some of these interactional roles in the following excerpt: The activities in the text above belong to the CC category. Two interactional roles are captured, which key into a converse semantic opposition; namely, the "suspect" and the "SSS" officials, on the one hand, and the "accused" and the judge, on the other. However, while the suspect and the security officials relate to the global action of apprehension, the accused and the judge are of the prosecution category. While the suspect and accused are one and the same person, the security and the legal officials are not necessarily the same person, although they belong to the broader international role of government officials.
In the AS activity, the global action of ethnic clash demarcates the interactional roles into two groups of community youths that are engaged in tribal wars. The global actions of rescue and reprisal clashes, however, have similar interactional roles, which are "the ND insurgent groups" and "the security officials" (particularly, the Joint Military Task Force, JTF). While the ND groups are involved in carrying out violent attacks mainly to attract the government's and the multinational oil companies' attention to their (re)quest, the security officials engage themselves in repelling the attacks and rescuing the kidnapped victims. The Joint Task Force in the Niger Delta, Operation restore Hope . . . said that the number of its men missing . . . had risen to 18. (Midweek Pioneer, May 27, 2009) In Excerpt 2, the two sets of synonymous labels, "unknown gunmen" and "attackers," refer to the same ND insurgent group, while "The Joint Task Force in the Niger Delta" and "Operation restore Hope" (in Excerpt 3) relate to the security officials. These two groups of participants attract the synonymous labels because of the cognitive focus of their activities in the conflict discourse. The "unknown gunmen" are labeled "attackers" because of their goals to disrupt the oil operations thereby gaining attention of the government or the international community. The security officials, on the other hand, have the mandate and are believed to be keeping peace in the region; hence, they are dubbed "Operation Restore Hope," which is abstracted from the goal they are designed to achieve.
Finally, in the of HA activity, two interactional roles are identified in the global action of kidnap, namely: ND insurgents groups against the security officials. These roles are akin to the ones found with the AS activity. Here, while some of the ND groups engage in human abduction to attract the government's and the multinational oil companies' attention to certain issues, other groups do this for monetary purpose. Generally, the security officials are reported to engage themselves in quelling these activities.

Communicative Roles
Communicative roles, as applied here, cover participants who are directly or indirectly concerned with or affected by conflict activities, and (hence) provide information on the global actions of the interactional participants. The communicative roles bifurcate into the information-giving role and the information-seeking role. The information-giving role relates to participants who provide information on the actions taken by the interactional participants. Two divisions have further been observed with the information-giving participant roles, viz. immediate and distant informers. The former relates to such participants as eye-witnesses, correspondents on ground, etc., while the latter involves authoritative participants, who can either be adding information (e.g., Police Public Relations Officers; Police, Army, Immigration, spokesmen, MEND e-mail, etc.) or confirming the information (e.g., Police Commissioners, Governors, MEND e-mail, etc.). The information-seeking roles fit such participants as the various newsmen interested in the actions of the participants, and strive to get the information by, for example, calling or visiting the distant information-giving participants, and/or getting firsthand information from the immediate informers. Therefore, in the data and according to Table 1 Out of the persons confirmed dead yesterday, six were militants while the other four persons were policemen.
According to an eye witness account, the four policemen that were killed in the clash were made up of an inspector, a sergeant, a constable and one other whose rank could not be ascertained.
Our source disclosed that shooting started from Hotel Presidential . . . (The Pointer, January 2, 2008) In the text above, while such participants as "spokesman" of JTF, "security source," and "commissioner of Police" play a distant information-giving communicative role in the activities reported (in Excerpt 4), others like "eye witness" and "our source" (in Excerpt 5) belong to the immediate informationgiving role. One thing is quickly noticed here: all the distant information-giving communicative roles identified in the texts above can technically be placed in a co-homonymous relation with the broader (hypernymous) goal of authenticating certain global actions carried out by certain interactive participants. However, within the distant category, the "spokesman" of the JTF ("confirmed") is of the information-confirming subcategory, while "security source" and "commissioner of Police" ("said") relate to the information-adding sub-category.

Social (Ideological) Roles
Social roles, as earlier mentioned, are affected by group membership, which is largely defined on the bases of norms or standards of activity, ethnicity or minority, political or professional inclination, and so on, of members. Group membership (with its attendant social roles), according to Ogbogbo (2005), is constrained by the certain "deep" and "surface" issues in the ND conflict discourse (p. 170). The deep issues relate to the broader ideological basis of the different groups' involvement in the discourse, while the surface issues are expressed in the immediate social activities carried out in the discourse. From the deep perspective, therefore, two broad groups of participants have been found; they include: the FGN, its security agencies, supporters and the multinational oil companies, on the one hand, and the ND activists and their sympathizers, on the other. Here, the FGN group, with its capitalist or globalist role, is the dominant group, while the ND activist group is the minority group, with a self-determinist quest to resist exploitation by the dominant group. This is the single cognitive basis that virtually all the global actions involve these two groups in an unending struggle to push through their underlying (deep) ideological goals.
From the surface perspective, however, the social roles of these participants demonstrate a complementary oppositional change with respect to different social activities, which in turn affect different fragments of participant roles broadly grouped under FGN versus ND insurgents. In the DO activity, for example, the ND group becomes the "activists" against the "insensitive multinational oil companies" (FGN partners) whose facilities are being destroyed. In the CC activity, we have "repented militants" and "sincere FGN" (represented by the amnesty committees) with regard to the amnesty global action; we also have "criminals" against "sanitizers" and "punishers" (FGN security officials) in the global actions of apprehension and prosecution, respectively. For the AS activity, the social roles mainly change to ND "activists" versus "peace makers/keepers' (FGN JTF) across the global actions. In the HA activity, particularly with regard to kidnap, we still have 'activists/criminals' versus 'insensitive multinational oil companies and high-ranking Nigerian citizens,' on the one hand, and against 'hope restorers,' on the other. In global action of release, 'responsible activists,' and a 'reassuring government' are the social roles observed. Let us consider the examples that follow: Excerpt 6 According to the source, about 10.00 am . . . unidentified men attacked one of the sites of the company in the area and abducted two out of three its expatriates . . . (New Waves, July 27, 2008) Excerpt 7 . . . gunmen invaded the riverine settlement of Aleibiri and abducted the father of the state Deputy Governor, Chief Simon Ebebi, the Alei of Aleibiri Kingdom. (Vanguard, December 11, 2007) Excerpt 8 In what appears to be a major breakthrough for the Joint Taskforce against the militants and cult groups in Rivers State, yesterday, one of the cult leaders, Soboma George was reportedly killed in gun battle. (The Tide, August 17, 2007) In the texts above, the different social roles can be deduced from the roles played by the participants. First, from the FGN group's ideological point of view, such lexical items as "unidentified men" (used in describing the men who abducted the expatriates in Excerpt 6), "gunmen (as tagged by the Bayelsa State Deputy Governor in Excerpt 7), and 'militants and cult groups' (brought against the Joint Taskforce in Excerpt 8) are the hyponymous interactional roles given to the ND insurgent group, whose meaning is related to the broader social role given to the group as criminal. And this appears to have been the case in the news discourse with the usual reports that, for example, the 'sanitizers' (the FGN JTF) killed one of the 'criminals' that have been disturbing the oil workers and well-meaning Nigerian citizens. However, the negative social role attached to the lexical items will definitely be different if it were from the ND activist perspective.

Instrumental Roles
The instrumental role, as earlier hinted, is added to represent some entities, which are important for the full understanding the participant's role relations in the conflict discourse. The present study, in line with Gumperz's (1982) and Duranti and Goodwin's (1992) position that contexts include the "materials" in the environment, addresses the need to represent some relevant entities which are not accommodated in van Dijk's (2000van Dijk's ( , 2002 original context models. The instrumental roles, therefore, cover both animate and inanimate entities used by the interactional participants in the course of their global actions. The entities represented by this category roles bifurcate into entities used and entities acted upon. The used category includes the mobile entities utilized by interactional participants in the course of their actions (e.g., guns, speed boats, etc.), while the category acted upon involves the stationary entities utilized (e.g., police stations, pipelines, etc. . . . the policemen were killed between 10.00 pm on Friday night and early hours of yesterday when suspected militants razed two police stations . . . (The Tide, April 15, 2007) While such lexical items as "600 assorted riffles" and "800 ammunition" (in Excerpt 9) belong to the used entities category, other lexical items like "two police stations" (in Excerpt 10) relate to the category acted upon. Aside from playing part of the instrumental roles in the situation of amnesty observed in the text, "600 assorted riffles" and "800 ammunition" can be placed in a hyponymous relation with the larger concept of arms used by the ND activists in their struggle. These aspects of participants (and even the dressing or the inscriptions on the shirt of the militant leader, though not captured in the story) may contribute to the writer's production and reader's understanding of the global actions of amnesty for the ND militants and intentions behind them. Similarly-in Excerpt 10, the "police stations" burnt by militants, as well as everything inside and around them, belong to the stationary category of instrumental participants. The reader's imagination of the police stations and the general environment contributes to the knowledge they build about militant attacks, especially on policemen or their places of work.

Conclusion
This paper set out to investigate the participant's roles and cognitive relations reflected in Nigerian print media reports on ND conflicts. It also explored the lexico-semantic choices that index the role relations with respect to the conflictual events reported in Nigerian newspapers. After the analysis, four types of participant roles have been identified, viz. interactional (embracing the participants in conflict encounter), communicative (relating to the production roles), social (involving group membership), and instrumental (dealing with the entities utilized for the actualization of specific goals).
The cognitive foci of these roles are observably associated with the goals and beliefs of the opposing groups of participants involved, and these inform their positions and hence roles in the conflict events reported. For example, at different levels of the social activities that the participants are engaged, the two principal groups of participants in the conflict discourse, the ND youth activists and the FGN's JTF, have a perception of the communicative context (presence and elements of the actual situation: the objects, beliefs, goals and preferences of other participants) constraining the social activities reported. Hence, the JTF's knowledge of the fact that militants disrupt oil activities (through vandalization of oil facilities and interruption of oil operations) informs their interactional role of apprehension/prosecution of the militants on the one hand, and their communicative role of updating the newspaper correspondents, on the other. Conversely, the ND groups' belief (that their region is marginalized) and goal (to redress the problem) motivate them to maintain a range of opposing social roles against the FGN/JTF across the social activities observed in the data. For example, in DO activities, their major social role is described as "ND activists" in lexical opposition to the "FGN's multinational oil working communities," whose important operations are disrupted. In CC activities, the ND groups are variously seen as "ND repented criminals" and "ND criminals" against "FGN" and "FGN security officials." In another dimension, the ND groups' belief (that the FGN would not easily yield to their demands) and goal (to arm-twist the government and register their seriousness) have observably influenced the groups' engagement of certain instrumental roles which reveal the properties of objects ("explosives," "speed boats," "guns," etc.) used in the struggle. These findings satisfy the first research question raised about cognitive relations. In terms of lexical choices, it was demonstrated that synonymous and converse lexical items predominate the description of interactional and social roles, while the communicative and instrumental roles are marked by homonymous and antonymous lexical features. This part of the findings provides answers to the second research question raised in the study, which borders on the linguistic forms used in the texts. Here, the relational-semantic choices utilized in describing news participant's roles and cognitive relations are found to be constrained by the ideological views taken by news reporters.
Clearly, these findings have taken previous scholarship on ND conflict news discourse a step further in confirming that there is a link between the cognitive disposition of the different groups of participants to the conflicts and the actual roles they observably play in the conflict activities reported. Second, this is made possible through the lexicosemantic representation of the roles and cognitive relations of the participants in the conflict news discourse. This therefore makes it easy to have a clearer understanding of the group-induced motivation for the conflictual events in the ND region. Generally, the findings confirm the view established in previous scholarship that the (Nigerian) media have an ideological role to play in shaping people's opinion of the roles played by the different stakeholders in ND conflict. However, considering such limitations as not being able to accommodate other media platforms and/or explore them at various levels of linguistic analysis, a full understanding of media representation of participants in ND conflict may still pose some difficulty unless these areas are covered.
The findings above have opened up another context in which the age-long environmental problems in the ND region have continued to influence different participants' beliefs and goals in the struggle. The implication of this is that to fully understand participant's roles and cognitive relations in the ND region as presented by Nigerian media, the different groups of participants in the discourse have to be linked to their beliefs and goals of engagement in the conflict. Therefore, the paper has two recommendations. First, the FGN's policy on ensuring more structural development and particularly youth empowerment to the ND region should be taken more seriously to reduce the plight of community members, which can in turn positively change their ideological outlook in ND struggle. Second, the media should be more sensitive and objective in their representation of participant's roles and cognitive relations in the conflict discourse. This can be achieved through cautious use of emotive and subjective lexical items in the representation of entities in the news.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research and/or authorship of this article.