APPROACHES TO DISCOURSE ANALYSIS

1.1 An Introductory Note
This paper makes use of the relevant findings of various approaches to discourse analysis, envisaged as a means for a purposive end: as a problem-solving apparatus. In order to justify the adoption of certain discourse analysis models rather than others in testing the validity of the hypotheses of any discourse analysis study, a general theoretical account of the various influential approaches to coherence in text linguistics has to be presented first, and critically approached next to get at their range of operationality and usefulness as a means to an end.
1.2 Discourse / Text
Though these two terms are the central theoretical entities in the domain of discourse analysis, their status and dimensions have remained in dispute and have fluctuated with each influential linguistic trend. One problem in this respect is that the term text is commonly used in a variety of ways. Van Peer (1989) cites the following uses of this term: (a) a record (b) a literary work of art (c) a composition (d) a (set of) meaningful utterance(s) (e) a linguistic structure of two (or more) sentences (ibid. 275). While the last two definitions allow every kind of language-use to be termed text, the first and third definitions suggest that texts must be written. Moreover, three of the above definitions (a, d, and e) do not distinguish texts from other kinds of language-use, while the second definition excludes non-literary texts.
Relevant to this issue is the overlap between the notion of text and discourse. Although these two terms are not synonymous, some linguists see very little difference between them and use them, more´-or-less, interchangeably. Others draw strict demarcation lines between them by restricting the term discourse to spoken utterances and reserving the term text to written pieces of language use. Still others see either one of these two terms as the theoretical construct verbally realised by the other (see the following table).
This overlap is complicated by the fact that research in discourse analysis/test linguistics is not restricted to the domain of linguistics alone, but have been carried out also by philosophers, psychologists, sociologists, sociolinguists, conversational analysts, anthropologists, literary critics . . . etc.
Besides the absence of a general terminological consensus among text linguists, the fact is that there is no one generally accepted theory of discourse analysis that undertakes to provide THE complete analysis of texts (Prince, 1992: 295). While all text analysts acknowledge the fact that a text has structure, coherence, --function--, organization, character and development, their approaches differ as to how each of these properties is realised and mutually related to other properties (cf. Halliday, 1982: 225). Hence the advantages of the eclectic approach which provides for the necessary step of integrating a variety of compatible systems of discourse analysis whenever these are found useful and adaptable to the requirements of each study. Such an approach, while lessening the problems of indeterminacy and partiality, remains just one model yielding one specific interpretation. However, variation in interpretations are resorvable and can ultimately be made definitive given a text and the same vital background information and approache(s) (Jordan, 1992: 196).
In view of the above facts, such terminological niceties in the use of the two terms “discourse” and “text” will be bypassed in this study by using them interchangeably. Both terms are also used with the zero article to denote generic reference. The use of the (in)definite article(s) before them indicates specific reference.
1.3 Approaches to Aspects of Text Coherence
As a distinctive linguistic unit of organization, TEXT is a highly complex phenomenon of communicative interaction representing a world with relevant social and institutional contexts (Halliday: 1985, xvi). The complexity of text can be shown by the fact that the overall meaning of a text is not simply the sum-total of the cumulative meanings signalled by its individual sentences, clauses, phrases, words, and morphemes. This is because these operational units of textual realization obtain additional signalling values and --function--s by virtue of: (i) becoming parts of a higher level of organization, (ii) contracting a web of structural and semantic (inter)dependencies manifesting a specific texture, (iii) the --function--ing of the whole of the text as an appropriate unit of communication in its environment. Therefore, a communicative message does not consist of units, but of a goal-oriented units-in-relation (Callow & Callow, 1992: 8).
A text cannot be adequately comprehended unless its constitutive units cohere´-or-hang together. The human mind can only grasp what relates coherently both to our existing knowledge and to the rest of the text (ibid.). Therefore COHERENCE is the fundamental property of textuality. A discourse is coherent whenever its sentences are easily related to one another. Coherent passages enable us to identify the general thread of discourse and the way the individual sentences fit together to achieve this purpose (Carroll, 1986: 213). So, coherence characterizes all the appropriate forms of language use by defining its unity. However, the characteristics of this unity are not straightforward in fact they are complex and multidimensional. The problem of defining the true nature of coherence and its constitutive elements is highlighted by Chomsky (1968):
Just what “appropriateness” and “coherence” may consist in we cannot say in any clear´-or-definite way, but there is no doubt that these are meaningful concepts. We can distinguish normal use of language from the ravings of a maniac´-or-the output of a computer with a random element.
(ibid. 11)
This quotation shows that while the presence of discourse coherence is uncontroversial, it is not always easy to determine its constitutive elements, especially if we understand coherence in its wider all-inclusive sense as THE constructive standard of textuality with its multi-level: phonological, lexical, syntactic, semantic, stylistic, rhetorical, and pragmatic aspects. Different approaches have been taken in the study to determine the constituents of coherence as a text-formation concept, all having their own terminological distinctions. In this section some of the influential linguistic approaches will be briefly sketched by grouping relatable trends together on the basis of their general affinity rather than their chronological order.
1.3.1 Textual Cohesion
Halliday and Hasan (1976) use the term COHESION to denote the general text-forming set of relations having a suprasentential cohesive --function--. According to their aforementioned model, a text has texture (= coherence) which is the property that distinguishes it from non-text (ibid. 2). Texture is the combination of semantic conformations of two different types: register and cohesion. Register refers to the variety of language which is appropriate for the situation of the speech event with its three contextual features of FIELD (what is happening, TENOR (who are taking part), MODE (What part the language is playing) (ibid. 22). While register relates the text to the external aspects of textuality, cohesion relates the text to itself. Cohesion is provided by certain cohesive relations termed: COHESIVE TIES that allow analysing a text in terms of its cohesive properties. Accordingly, the relation of cohesion is set up where the interpretation of some element in one of the sentences of the text is dependent on that of another, simply because the one PRESUPPOSES the other, and both are thereby integrated into the text (ibid. 4). The co-authors offer the following example from the instructions in a cookery book.
Example (1)
Wash and core six cooking apples. Put them into a fireproof dish.
(ibid. 2)
Here, the anaphoric --function-- of them in the second sentence-referring back to the six cooking apples—gives cohesion to the two sentences, and that is why we interpret them as forming part of the same text. Five categories of textual cohesive ties are recognized: reference (=co-reference), substitution, ellipsis, conjunction, and lexical cohesion (reiteration and collocation) (ibid. 9).
This briefly outlined model of cohesion as an indent of measuring textual coherence has been subjected to serious criticism. First, it has been argued that the concept of cohesive ties is inadequate in that it does not guarantee “textness”. Enkvist (1978), among others, gives the following example where the reiteration of lexical themes does not turn this grouping of contiguous sentences into a coherent text:
Example (2)
I bought a Ford. A car in which President Wilson rode down the Champs Elysees was black. Black English has been widely discussed. The discussions between the presidents ended last week. A week has seven days. Every day I feed my cat. Cats have four legs. The cat is on the mat. Mat has three letters. (ibid. 110)
Second, it has been argued that Halliday and Hasan’s concept of cohesion fails to take into account the readers’ contributions to text coherence. Empirical evidence shows that the lack of connectives does not seriously damage comprehension because the readers are usually able to make bridging inferences (Freebody & Anderson: 1981, 19). Third, it has been pointed out that text coherence is a matter of context which happens to have linguistic consequences, not the vice versa. In other words, Halliday and Hasan’s categories of cohesive ties are the effect of content coherence, not its cause (Morgan and Sellner:1980, 179). Such pieces of evidence have induced Brown and Yule (1983) to conclude that formal cohesion will neither guarantee the identification of textuality nor textual coherence (ibid. 197).
However, as Brown and Yule are quick to acknowledge themselves, it is only fair to point out that Halliday and Hasan are not “concerned to produce a de--script--ion which accounts for how texts are understood”, and that “their examination of linguistic resources is rich, interesting, and insightful” (ibid.:204). Moreover, Halliday and Hasan themselves have widened the scope of their classical model of cohesion in their later and more elaborate work: “Language, Text, and Context: Aspects of language in a Socio-Semiotic Perspective” (1985). In this book, cohesion is divided into “structural” and “nonstructural” cohesive devices. Structural cohesion is effected by Parallelism, Theme-Rheme Development, and Given and New Organization. Non-structural cohesion is set up by two kinds of relations: (i) componential and, (ii) organic. Componential relations are realised by grammatical and lexical cohesive devices, while organic relations are brought about by conjunctives, adjacency pairs, and continuatives (ibid. 82).
1.3.2 Lexical Cohesion
In his work Patterns of Lexis in Text (1991), Hoey introduces a new and detailed model for investigating how cohesive features combine to effect the organization of text. In this model – essentially based on the previous one – lexical repetition is taken as the governing principle that forms cohesive links. Here, text is viewed as being made up of interrelated, yet separate packages of information. The interrelation among these packages (i.e. sentences) is established by repetition (ibid. 12). Separate sentences linked by repetition are more closely related than adjacent sentences not so linked. Consequently, sentences that are linked to a greater number of other sentences are termed central sentences whereas sentences that are linked to a few of other sentences,´-or-to none, are called marginal sentences. Central sentences are further classified, in terms of their --function--, as either topic initiators,´-or-topic concluders. Topic initiators are those sentences to which the subsequent ones are linked, while topic concluders are the ones linked to the preceding sentences (ibid. 43).
The significance of the central sentences lies in the necessary information they provide the readers about the main themes and motifs in the text. In contrast, marginal sentences neither build lexically upon what has gone before, nor provide the lexis for the subsequent statements (ibid.).
The model also introduces the key concept of bonding. If two sentences share three´-or-more links, they form a bond (ibid. 265). Each set of interconnecting bonds among the sentences of the text is called a net (ibid. 267). In effect, the final interconnection among all the nets in a text will result in the text s cohesion.
Hoey states that coherence is not synonymous with cohesion in that coherence could only be determined by the addressee s evaluation, whereas cohesion is a property of text. In other words, cohesion is an objective feature inherent in text, while coherence is a relative and subjective feature that depends on the addressee s interpretation of the text and as such is subject to the external factor of the addressee s response. Consequently, there could be only one type of cohesion in a text, but potentially many varying kinds of coherence (ibid. 12).
1.3.3 Textual Coherence
Among the earlier views that draw a distinction between coherence and cohesion are those of Kane and Peters (1966) and Cummings et al. (1970).
Kane and Peters differentiate between two kinds of discourse unity: unity of thought (coherence) and unity of form (cohesion) (ibid. 141). Coherence is actuated by: (i) relevance, (ii) proper order, and (iii) inclusiveness. Cohesion—termed “flow”—is realised through sentential ties, and transitional links such as repetition, connectives, pronouns, demonstratives, and syntactic patterns (ibid. 145).
A similar, but a wider, outlook is expressed by Cummings et al. Who maintain that discourse is characterized by four standards of “rhetorical form” (i.e., textuality): appropriateness, coherence, unity, and continuation (ibid. 195). Of these, appropriateness is considered to be the most basic standard whose requirements impose constraints on all the other standards. Appropriateness involves the complex interrelationship of the writer, reader, and subject matter all infused with the particular pragmatic purpose motivating the issuance of a particular discourse. While appropriateness is the pragmatic feature, coherence, unity, and continuity are syntactic-semantic features. Unity of discourse arises from explicit´-or-implicit links among the sentences. Such links also require some carry-over of the sameness of content from one sentence to another. A more complex kind of unity is based on the semantic relationship between discourse sentences and some overacting topic that may be made explicit´-or-implicit (ibid.:196). Coherence depends not on the interrelatedness of the discourse sentences, but upon their interdependence such as the need for a fixed sequence. The authors give this example to illustrate their point:
Example (3)
1. The car is blue. The car is fast.
2. The car is blue. It is also fast.
The first two sentences, though unified, are not coherent because of the absence of any interdependency, a fact which allows either of them to come first. In the second pair of sentences, the semantic dependence of the second sentence on the first one makes the two sentences coherent: i.e., having fixed sequence.
While unity and coherence grow out of controlled sameness, continuation grows out of a sense of the controlled difference in the introduction of new information into the discourse. The carryover of sameness by tying the new information with the “old” ones adds control on the difference´-or-the newness of rhemic material (ibid.196).
A related and still a wider approach has been adopted by Beaugrande and Dressler (1981) who see texts as vehicles of purposeful interaction, and judge them in terms of their effects upon the receivers (ibid. 14). Their model of discourse analysis endeavours to answer the central questions of how discoverable structures are built through the operations of decision and selection, and what applications these operations have for the purpose of communicative interaction (ibid. 15). The co-authors combine several elements from different text notions to attain a wider inclusiveness. Their eclectic definition of text is that it is a communicative occurrence which satisfies seven standards of textuality termed: cohesion, coherence, intentionality, acceptability, informativity, situationality, and intertextuality (ibid. 1-11). Accordingly, the lack of any one of these defining text-formation standards will reveal a non-text. The first two standards of cohesion and coherence are text-centered, while the other five are user-centered bearing on the contextual --function--s of the textual communication.
By cohesion is meant the mutual sequential connectedness among the components of the textual surface structures´-or-the text’s actual words. The co-authors agree that mere cohesion is not decisive in the making of communication without its interaction with the other standards of textuality. They state that cohesion is the result of stability within a text as a system of grammatical dependencies both with, and beyond the sentence boundaries (ibid. 50). Cohesion structures subsume repetitions, parallelisms, paraphrase, junctives, proforms, ellipsis, as well as recurrences in grammatical subsystems such as tense, aspect, and articles.
Coherence concerns the ways in which the configuration of concepts and relations underlying the surface text are made mutually accessible and relevant. By “concept” is meant the cognitive content which can be recovered´-or-activated with more´-or-less unity and consistency in the mind. By the “relations” are meant the links that hold between the concepts appearing together in the textual world. Relations may be explicit´-or-implicit. In the latter case, text receivers normally infer the needful relations to make sense out of the text as it stands. Among the concepts of coherence are the relations of time, cause, reason, enablement, and purpose. In this respect, the co-authors conclude:
Coherence already illustrates the nature of a science of texts s human activities. A text does not make sense by itself, but rather by the interaction of text-presented knowledge with people’s stored knowledge of the world.
(ibid. 6)
The other five standards of textuality are related to the degree of the instrumentality of the text’s cohesion and coherence in: (i) fulfilling the intended goal (intentionality) (ii) providing use´-or-relevance (acceptance) (iii) showing relevance to the situation (situationality) (iv) imparting the expected´-or-unexpected knowledge (informativity) (v) having relevant connection(s) with the previously encountered texts (intertextuality). In addition, the co-authors stipulate the existence of at least three regulative principles in texts: (i) efficiency, (ii) effectiveness, and (iii) appropriateness (ibid. 11).
The broad inclusiveness of this model reveals the merits of the synthesis of compatible elements, an approach which will be adopted in this study.
Beaugrande and Dressler’s differentiation between cohesion and coherence has its precedence in the works of Widdowson (1975, 1979a, 1980) who assigns cohesion to the propositional relations between parts of a discourse, and coherence to the illocutionary relations holding between them (1979a: 29). To him, cohesion rests upon the overt structural links between the sentences, while coherence is established by the link between the communicative acts that the sentences are used to perform. In order to relate cohesion and coherence Widdowson (ibid. 147) suggests regarding the interactive acts as instruments of the propositional development mediating between the two “procedures”. He expresses these relations in the following diagram (ibid. 148):
procedures cohesion coherence
interactivity
By the interactivity (or the interactive acts) is meant the way in which the propositions expressed in a discourse are organized and managed,´-or-how the interaction is negotiated between the participants (ibid. 138). Widdowson states in the same place that the difference between the interactive acts and the illocutionary acts is that the latter sort of acts are social activities which relate to the world outside discourse, whereas the interactive acts are essentially ways of organizing the discourse itself and are defined by their internal --function--s.
The approach of relating coherence to speech actions by moving outwards from the discourse to the meanings of linguistic forms is related to the works in ethnomethodology, sociolinguistics, and pragmatics (see, for example: Dundes et al., 1972 Labov, 1972b Sacks, 1972 Cicourel, 1973 Turner, 1974 Sinclair and Coulthard, 1975 and Katz, 1977 . . . etc.). Following Austin’s Speech Act Theory (1962), it has been noticed that both monological and dialogical discourses can be defined at a pragmatic level as being sequences of speech acts (ways of doing) as well as sequences of sentences (ways of saying). In order for these acts to --function-- as coherent assertions, promises, threats . . . etc., they must satisfy certain conditions definable in terms of their appropriateness in a given context. This approach of interpreting pieces of discourse merely in terms of actions is extremely problematic simply because there are no one-to-one relations between actions and utterances. Labov (1972b) states that sequencing rules (i.e. coherence) do not operate between the actual utterances, but between the actions performed with those utterances due to the fact that there are usually no connections between successive sentences at all (ibid. 254). Therefore, Labov concludes that, as receivers at pieces of discourse, we rely upon our intuitions to distinguish between coherent and incoherent discourse, and others the following “extreme” example (borrowed from Laffal, 1965: 85) from a conversation between a doctor and a schizophrenic patient:
Example (4)
A: What is your name?
B: Well, let’s say you might have thought you had something from before, but you haven’t got it any more.
A: I’m going to call you Dean.
(ibid. 252)
An explanation to this kind of immediate experience is attempted by Tsui (1991) in the form of a definitional term. Conversation, like discourse in general, is said to be governed by the Coherence Principle:
. . . in order for an utterance to form a coherent sequence with the preceding utterance, it must either fulfill the illocutionary intention of the latter,´-or-address its pragmatic presuppositions.
(ibid. 120)
Tsui also affirms that violating the rules governing coherent sequences can be accounted for since this violation results in an incoherent discourse which is noticed and attended to by the interlocutors (ibid. 1111). However, the assignment of coherence to the addressor’s mental intentions´-or-presuppositions hardly explains anything in linguistic terms. It merely shifts the office of coherence from the discourse itself (as a self-sufficient, analysable linguistic product) to the minds of the interlocutors themselves.
In an earlier and a similar vein, Brown and Yule (1983) assume coherence to be what the interlocutors bring to the interpretation of linguistic messages. Accordingly, coherence will produce one particular interpretation in which the elements of the message are seen to be connected, with´-or-without overt linguistic connections (i.e. cohesive ties) between those elements (ibid. 224). Still, they also perceive text interpretation to be based on coherence, analogy, local interpretation, general features of context, structural regularities, and the reader’s conventional socio-cultural knowledge (ibid. 225).
Tannen (1984) views coherence to be the underlying organization structure which renders the sequences of words and sentences into a unified discourse having cultural significance for those who create and comprehend it. According to her, coherence subsumes cohesion, the latter being but one out of many factors all contributing to coherence including the relationships to other texts (intertextuality), relationships to the producer’s intentions, and relationships to the world outside the text (ibid. xiv).
A more recent approach which differentiates between coherence and cohesion is that of Van Peer (1989) who integrates Halliday and Hasan’s approach (1976, 1985) with that of Brown and Yule (1983) and Tannen (1984) by assigning cohesion to the specific linguistic devices such as anaphoric expressions, conjunctions, enumeration, and adverbials that can link up the various parts of a text and ascribing coherence to some (real´-or-imagined) world that is coherent. Van Peer adds that this coherence can be brought about by verbal repetition, co-reference, determinate and deictic expressions, personal pronouns, proper names and the like (1989: 279). However, he also points out that both text-internal cohesion and text-external coherence are necessarily incomplete in almost all texts since there is hardly any text which explicitly mentions all text-internal and external relations. It is because of this that the use of texts calls for complex mental activities on the part of the addressee so that both cohesion and coherence can be constructed as adequately as possible in the course of text-processing (ibid. 280).
Similarly, Wright and Hope (1996) consider cohesion as a surface feature that involves connecting the sections of a text by formal linguistic links which can be listed, pointed at, and classified. They acknowledge coherence to be more difficult to define´-or-analyse than cohesion since the former refers to the ways we know a text gets together through thematic continuity, cause and effect, and so on (ibid. 176). They also point out the fact that in some literary works coherence may emerge “slowly” in a text since by delaying the readers’ realization of the coherence of a passage the writer can make that realization all the more powerful (ibid. 177).
Some text analysts do not keep cohesion and coherence terminologically distinct. Van Dijk (1981) uses the two terms synonymously. In his discussion of the structures of sentences in sequences, he affirms that these structures are not only characterized by the use of various kinds of interclausal connectives (e.g. and, but, because, although), but also by other connectives which may only appear at the beginning of new sentences (i.e. however, so, consequently, yet) (ibid. 4). Furthermore, he points out that he takes connection as a particular case of coherence, viz. The semantic relation between whole sentences´-or-propositions as they follow each other in texts (ibid. 22).
The works of van Dijk (1972, 1977, 1980, 1981, 1984, 1985, 1987) and van Dijk & Kintsch (1977, 1978, 1983) offer one of the most comprehensive approaches to the study of the nature and elements of discourse coherence. Van Dijk (1985) affirms that discourses are not just sets of sentences, they are also organized sequences having conventional constraints on the possible meaningful orderings representing certain fact-structures. These sequences of sentences express sequences of propositions (ibid. 107). Generally speaking, the proposition sequence underlying an acceptable discourse must satisfy various conditions of semantic, pragmatic, and stylistic coherence (van Dijk, 1981: 5 van Dijk & Kintsch, 1983: 149).
Semantic coherence is both local and global. Local semantic coherence accounts for the microstructural level of the relations between the sentences´-or-propositions of the discourse. It establishes the meaningful intra/intersentential connections signalled by word-order, sentence-order, the use of connectives, sentential adverbs, verb tenses, and pronouns (van Dijk, 1985: 108). Besides the linear ordering of the propositons, there is the additional hierarchical organization of the underlying semantic structures: their spatial, temporal, and conditional connections.
Global coherence is concerned with the meaning of larger discourse chunks´-or-whole discourses which cannot be defined in terms of local coherence conditions mentioned above. This is because global coherence is concerned with the topic´-or-theme of the whole discourse (i.e. its “macrostructure”) (ibid. 115). A macrostructure is a theoretical reconstruction of the higher level global meaning that is derived from the propositional sequences of the text by a number of “macrorules” (1981: 4). In other words, the macrostructure is the semantic information that gives the discourse its overall unity. Frequently, such underlying macrostructures are given by the text itself in the titles, summaries,´-or-announcements . . . Macrorules are recursive operations that select, reduce, generalize, and reconstruct discourse propositions into fewer higher level´-or-more abstract propositions. For example, these macrorules can derive the macrostructure: “Peter took the train from:
Example (5)
Peter went to the station.
Peter bought a ticket.
Peter went to the platform . . . etc.
(van Dijk, 1981: 4)
As mentioned earlier, macrostructures are semantic abstractions conditioned by the more concrete cognitive operations and representations (van Dijk & Kintsch, 1983: 150). Since the individuals’ world-knowledge, beliefs, opinions, attitudes, interests, and goals of communication are not identical, differences in assigning global meanings to the same discourse may occur, stemming from different evaluations about what is relevant´-or-important information for the discourse. However, these subjective variations often manifest enough overlap to guarantee successful communication and interaction (van Dijk, 1985: 117).
Stylistic coherence is related to the --function--s and relations between all kinds of variations in language. Stylistic variations can be grammatical, schematic,´-or-rhetorical (van Dijk and Kintsch, 1983: 94). In principle, stylistic variation correlates alternative ways of expression with an underlying identity´-or-similarity of theme´-or-semantic representation. This variation has highly complex coherence agencies such as signalling the speaker-hearer and discourse-context relations,´-or-regulating ease of decoding and understanding (ibid. 18). Metrical´-or-prosodic patterns in literary´-or-ritual texts such as meter, rhyme, alliteration, repetition, and figures of speech represent additional organizational patterns characterising whole discourses. As such, they are superstructures for text forms which consist of conventional categories—often hierarchically organized—that assign further structures and overall organization to discourse and facilitate the remembrance and reproduction of textual macrostructures (ibid. 16f).
Correspond to this view is Mann et al.’s (1992) --function--al approach to discourse coherence. They precept that the recognition of text is contingent on its coherence. To them, coherence rests in the hierarchical organization of --function--ally significant parts each contributing to create a sense of an overall organic unity. They assign textual unity to the agency of the imputed --function--. A text is perceived as coherent when all its component parts contribute to a single purpose of the writer (ibid. 43). Similar to the general approach of van Dijk (and many other discourse analysts such as Grimes (1975), for example) they recognize an asymmetry of textual relations whereby one text span is more central (i.e. the nucleus) and the other is peripheral (i.e. the satellite) (ibid. 44).
A basically similar --function--al approach is adopted by Callow and Callow (1992) who assert that messages cannot be conceived unless they coherently relate both to our existing knowledge and to the rest of the message. The relations which bind a message together are termed coherence relations (ibid. 8). These relations may hold between the message units at any level, but are best illustrated at the propositional level where they can be realized by forms such as: because, after, and, for example . . . Relations between the propositions are expressed as follows:
Related propositions constitute a configuration. The clause Mary blushed and Simon laughed, do not, as they stand, realise a configuration, because their relationship is not clear. But Mary blushed because Simon laughed, does realise a configuration. When Mary blushed, Simon laughed, realises a different configuration: the units are the same but the coherence relation is different.
(ibid. 9)
The Callows’ convincing argument that text structure has to do with the coherence relations --function--ing among its component parts has been highlighted since the late sixties of the last century by the rhetorical approach to text-linguistics (cf. Winter, 1968). The origins of this approach can be traced back to Sweet’s analysis of the try-succeed binary pair (1891) and to Jesperson (1914 - 49). According to this relational approach, text coherence is sought in the use of a small set of highly recurrent binary/logico-semantic relations holding between potentially any parts of a text, whether paragraphs, sentences, clauses,´-or-even nominal groups´-or-single words (Jordan, 1992: 180). Rhetoricians have put forward a number of terms to denote such relations, among these are: Clause Relations Sentence Relations Propositional Relations and Rhetorical Structures (Beekman:1970 Beekman and Callow:1994, Callow:1974 Callow and Callow:1970, 1992 Hoey:1974, 1983a, 1983b, 1993 Hoey and Winter:1986 Jordan:1978, 1980, 1984, 1985, 1988, 1990, 1992 Mann and Thompson:1986, 1987, 1989 Mann et al. 1992 Winter:1968, 1971, 1974, 1977, 1982, 1992, 1996). Though the number of these relations is very stable and culturally-specific, it is also open to modification and addition (Mann et al. 1992: 46). Among these relations are: Cause-Effect Problem-Solution, Argument-Evidence, Condition-Consequence, Error-Correction, Assessment-Basis, Generalization-Exception . . . etc. Such pair-wise logical relations can either be lexically signalled (by subordinators, conjuncts, and lexical items)´-or-implicitly indicated:
Example (6)
Because she was in trouble, he helped her.
(Explicit: Reason/Result)
She was in trouble. He helped her.
(Implicit: Reason/Result)
These relations connect elements of texts at all levels by definable meanings. Beekman and Callow (1974: 317) offer the following example to this effect:
Example (7)
1. Christ has now reconciled you (to God)
2. by means of dying physically Means of 1.
3. in order that you will be holy. Purpose of 1.
where the first clause enters into a relationship of Means with the second (i.e., providing the means of reconciliation) and of Purpose with the third clause (i.e. providing the purpose of the reconciliation).
As shown above, the members of a rhetorical structure may contract additional chain relations with the other elements of the text,´-or---function-- as a nucleus´-or-a satellite for still other rhetorical relations spanning larger portions of the text. The following are the categories of one taxonomy of some combinations of these relations:
Motivation, Concession, Background
Circumstance, Solutionhood, Elaboration
Purpose, Non-Volitional Result, Means, Restatement
Sequence, Contrast, Joint (Mann et al. 1992: 70)
Similar complex patterns representing a combination of metastructures have also been specified and discussed by Winter (1982), Hoey (1983), and Jordan (1984). One basic pattern is that of: Situation-Problem-Solution-Evaluation. Hoey cites the following “constructed” example:
Example (8)
I was on sentry duty. I saw the enemy approaching.
I opened fire. I beat off the attack.
(Hoey, 1983: 35)
Here, the first sentence sets up the situation, the second indicates the problem, the third offers the solution, and the fourth presents the evaluation. However, not every member of the pattern need be mentioned one´-or-more members may be implied´-or-deleted as deemed contextually suitable.
The theory of Rhetorical Structures provides a general and effective tool for the de--script--ion of the abstract coherence relations/whether explicitly stated´-or-inferred) among the organizational elements in a text. Moreover, it serves to account for the --function--al potential of texts and enables a unified de--script--ion of the coherence-structure of all text-components regardless of its genre (Mann et al. 1992: 68). Finally, this approach lays a foundation for studies in contrastive rhetoric. One example in this respect is Cui’s analysis of Mandarin and English essays (Cui, 1985). This study will build on such a foundation in the contrastive study of rhetorical pattern in the Arabic and English poems in Chapter (4) where more details are offered.
1.4 Coherence in Poetry
This section investigates in broad lines some of the specific poetic means of achieving structural coherence in English. It will also touch upon the theoretical and practical relevance of modern linguistic notions of parallelism, prominence, deviation, textual information organization. Though some of these important aspects and notions of coherence lie outside the procedural scope of this study, their general stylistic bearings on the problem discussed here provide an additional support to the hypotheses of the study, and are, therefore, worth outlining.
Literary texts in general and poems in particular are highly coherent pieces of language-use where many levels of content and form unite realizing a complex tissue of interrelations and interdependencies spanning the whole of the text to achieve one and the same end: overall unity. Poems have aspects of internal phonological, lexical, syntactic, and semantic patterning of reiteration and/or contrasts far exceeding those of non-literary texts (Traugott and Pratt, 1980: 21, 38). The extra poetic structure brought about by meter, rhyme, refrains, alliteration, assonance, consonance, enjambment, regular stanzaic forms, and other poetic devices adds extra meaning to poetry (Fowler, 1984: 41).
Phonological patterning alone can be the source of coherence in a poem as manifested in some nonsensical nursery rhymes,´-or-bilingual pieces of poetry such as this fun-invoking Latin-English poem by Swift:
Example (9)
Mollis Abuti
Has an accuti
no Lasso finis
Molli divis
(Horvell, 1958: 335)
Similarly, the possibilities of the relations between phonology and morphology can be systematically deployed in constructing poems where the innovative sound-sequences acquire unique meanings unrecognizable outside their new systematic sequencing:
Example (10)
‘Twas bryllyg and slythy toves
Did gyre and gymble in the wabe
All mimsy were the borogroves
And the moms raths outgrabe.
(Lewis Carroll, 1965: 118)
Besides neologism, the poetic play on homonomy (pun) and reduplication can be a rich source of coherence set up by the inventive sonorous patterning at levels beyond the logic of the sentence:
Jakobson summarises the general principle of coherence in poetry as follows: “The poetic --function-- projects the principle of equivalence from the axis of selection into the axis of combination” (Jakobson, 1960: 358). By this statement Jakobson means that, in poetry, partial´-or-full equivalent phonological, morphological, syntactic, and semantic patterns are systematically combined in meaningful sequential order. In other words, unlike other forms of language-use, syntagmatic relations in poetry are interpreted as if they were paradigmatic. That is, the horizontal relations of language have to be considered as if they were vertical ones. Therefore, the addressee considers the relationships between the words of a poem in a non-linear differential way as if they existed together in one coherent temporal plant (Newton, 1988: 118-9). Jakobson cites Caesar’s famous proclamation: veni, vidi, vici (I came, I saw, I conquered) as an example where the setting up of multilayer levels of equivalent patternings leads to the creation of unique sequential meanings. Here, the three words are symmetrical in their grammatical category, tense, inflection, number of syllables, rhyme, stress pattern, and alliteration. At the semantic level, the equation between the acts of “coming” and “seeming” with that of “conquering” gives the impression that the last act was as easy and natural as the first two. Jakobson terms such aspects of structural and symantic symmetry: parallelism (Jakobson, 1960: 368). The English poet, Gerald Manley Hopkins goes as far as to claim that “the artifice of poetry reduces itself to the principle of parallelism” (House and Storey, 1959: 84). By parallelism he means any form of the apportionment of invariants and variables (Jakobson, 1966: 423). However, Shapiro (1967) believes that the basis of poetic coherence lies in asymmetry (contrast) not exact symmetry.
References
Althusser, L. (1971) Lenin and Philosophy New York, Monthly Review Press
Bourdieu, P. (1971) ‘Intellectual Field and Creative Project’, in M.Young (ed.) Knowledge and Control,.( London: Collier-MacMillan)
Cole, B. 2001 ‘Managing to Include’ in Calvert M., Harvey J.A., and Cole B. (eds.) Managing Change. Sheffield: Philip Armstrong Publishing.
Fairclough N. 1989 Language and Power. London: Longman.
Fairclough, N 1992 Critical Language Awareness. London: Longman
Fairclough, N. 1993 ‘Critical discourse analysis and the marketisation of public discourse: the universities,’ Discourse and Society, 4/2, pp133-68.
Fairclough N. 1995 Critical Discourse Analysis, London: Longman
Fairclough N 1996 ‘A reply to Henry Widdowson’s “Discourse analysis: a critical view.”’ Language and Literature 5/1, 49-56.
Fairclough N. 2003 Analysing Discourse: Text Analysis for Social Research, London: Routledge.
Fairclough, N. and Wodak, R. (1997). ‘Critical discourse analysis’. In T. A. van Dijk (ed.), Discourse Studies: A Multidisciplinary Introduction, Vol. 2. London: Sage.
Fish, S. 1981 ‘What is stylistics and why are they saying such terrible things about it?’ in D. Freeman (ed.) Essays in Modern Stylistics. London: Methuen, pp.53-78.
Foucault, M. 1980 Power/Knowledge: selected interviews and other writings 1972-1977. (Colin Gordon ed.) London: Harvester.
Fowler et al., (1979) R. Fowler, R. Hodge, G. Kress and T. Trew, Language and control, Routledge and Kegan Paul, London
Freire, P. 1972 Pedagogy of the Oppressed. Harmondsworth: Penguin.
Gee,P.(1990) Social Linguistics and Literacies (2nd edition 1996) London: Falmer Press
Gee, P. (2000) ‘Discourse and Sociocultural Studies in Reading’ .
Gillborn D (1998) ‘Racism and the politics of qualitative research: learning from controversy and critique’ in Connolly P and Troyna B (eds.) Researching Racism in Education: politics, theory and practice. Buckingham: Open University Press.
Gramsci, A. (1971). Selections from the Prison Notebooks. London: Lawrence and Wishart.
Halliday, M. A. K. & Hasan, R. (1976) Cohesion in English. London: Longman
Hoey, M. (2000) ‘Persuasive rhetoric in linguistics: a stylistic study of some features of the language of Noam Chomsky’. In: S. Hunston and G. Thompson, Editors, Evaluation in text. Authorial stance and the construction of discourse, Oxford University Press, Oxford
Hyatt, D. (1999) ‘Language, Philosophy, Journalistic Positionality and the Globalisation of the Mass Media: the masquerade of neutrality in political interviewer discourse’ in Actas 2, 6th International Symposium of Social Communication, Centre for Applied Linguistics, Ministry of Science, Technology and the Environment, Santiago de Cuba.
Hyatt, D. 2003 ‘Towards a Critical Literacy Frame: a textual analysis of the adversarial political interview’. Unpublished PhD Thesis. University of Sheffield.
Hyatt, D. 2005 ‘A Critical Literacy Frame for UK secondary education contexts’. English in Education, Vol. 39/1, pp. 43 – 59.
Hyatt, D. 2007 ‘Applying A Critical Systemic---function--al Literacy Frame to UK secondary education contexts’ in McCabe, A., O’Donnell, M., and Whittaker, R. (eds.) Advances in Language & Education. London: Continuum.
Ivanič-;-, R. (1997) Writing and Identity. Amersterdam: John Benjamins
Jewitt, C. and Kress, G. (2003) ‘A Multimodal Approach to Research in Education’ in Language, Literacy and Education: A Reader. Goodman, S. Lillis, T, Maybin, J. and Mercer, N. (eds.) London: Trentham Books/ The Open University.
Kress, G. and Van Leeuwen, T. (2001) Multimodal Discourse: The modes and media of contemporary communication. London: Arnold
Laclau, E and Mouffe C (1985) Hegemony and Socialist Strategy. London: Verso.
Luke, A. and Freebody, P. (1997) Chapter 1: ‘Critical Literacy and the question of Normativity: An Introduction’. In Muspratt, S. Luke, A & Freebody, P. (1997) Constructing Critical Literacies: Teaching and learning textual practice. St Leonards, NSW: Allen and Unwin
Maley, A. 1994 Review of Fairclough, N. (ed.)‘Critical Language Awareness’. Applied Linguistics. 15/3.
Mills, S. (1997) Discourse. London: Routledge.
McCarthy, M. and Carter, R. (1994) Language as Discourse: Perspectives for Language Teaching. London and New York: Longman
Norris, S. (2004) Analyzing Multimodal Interaction: A Methodological Framework. London: Routledge.
Rogers, R. (2004) An Introduction to Critical Discourse Analysis in Education . Mahwah, NJ: Lawrence Erlbaum.
Sinclair J.& Coultard J (1975) Towards an analysis of discourse : the English used by teachers and pupils. London : Oxford University Press
van Dijk, T.A. (1998a). Critical discourse analysis. Available:
http://www.hum.uva.nl/teun/cda.htm. (1/25/2011)
van Dijk, T.A. (1998b). Opinions and Ideologies in the Press. In Bell, Allan and Peter
Garrett (Eds.). Approaches to Media Discourse. Oxford: Blackwell
van Dijk, T.A. (1996). Discourse, Opinions and Ideologies. In Christina Schaffner &
Helen Kelly-Holmes (eds.) Discourse and Ideologies. Clevedon: Multilingual matters
Ltd, 1996. 7-37
van Dijk, T.A. (1995). Discourse Analysis as Ideology Analysis. In Christiina Schaffner and Anita L. Wenden (eds.). Language and Peace. Dartmouth: Aldershot. Pp.17-33
van Dijk, T.A. (1993). Elite Discourse and Racism. London: Sage Publications.
Pp.242-282
van Dijk, T.A. (1991). Racism and the Press. London: Routledge
van Dijk, T.A. (1988a). News Analysis: Case Studies of International and National
News in the Press. Hillsdale, N.J.: Lawrence Erlbaum Associates
Van Dijk (1988b). News as Discourse. Hillside, NJ: Erlbaum
Widdowson H.G. (1995) ‘Discourse analysis: a critical view.’ Language and Literature 4/3, 157-72.
Widdowson, H.G. (1998) The Theory and Practice of Critical Discourse Analysis.
Applied Linguistics, 19/1: 136-151

Rate the article
Bad	1	2	3	4	5	6	7	8	9	10	Very good
Result : 55%							Participated in the vote : 4