DISCOURSE ANALYSIS I: HALLIDAYAN APPROACHES
Since I welcome every available opportunity to pontificate on subjects that I know nothing about, a colloquium on discourse analysis provides a singularly apt occasion to display this rare skill.
— Moisés Silva
Discourse analysis is a burgeoning field and is one of the most exciting new areas of research related to Greek exegesis. While it has been a topic of discussion for decades within the wider world of linguistics (sometimes referred to as text linguistics),1 it is relatively new to New Testament studies and Greek linguistics. Discourse analysis is an interdisciplinary approach to understanding how units of text relate to one another in order to create the theme, message, and structure of a text. It is concerned to discover linguistic patterns in text, using grammatical and semantic criteria, such as cohesion, anaphora, and inter-sentence connectivity.2 Discourse analysts are often interested in “discourse markers,” which are elements of language that demarcate units of text.
The simplest way to think of discourse analysis is that it deals with text beyond the level of the sentence — the pericope, paragraph, wider units, and the text as a whole. In this way, it has obvious overlap with the interests of literary analysis (such as theme and structure) and also with the interests of rhetorical analysis (such as the way in which ideas are communicated). Discourse analysis does not attempt to replace these more traditional modes of exegesis, but is complementary to them. The most distinctive contribution that discourse analysis brings alongside literary and rhetorical analysis is its robustly linguistic nature. It generally moves from the grammar and syntax of a text out to these larger textual concerns, rather than starting with the big picture.
As we will see, discourse analysis will often yield results that are similar to traditional methods of Greek exegesis, and yet it will sometimes provide quite different outcomes. Nevertheless, even when the results of discourse analysis seem similar to traditional exegetical outcomes, the methods set out by such analysis provide a firmer foundation for results. Discourse analysis operates with a linguistically robust methodology that provides somewhat objective criteria by which to adjudicate exegetical issues. By contrast, traditional exegetical approaches often rely on the intuitive insights of the interpreter. While the intuitive conclusions of gifted interpreters will often be correct, it is difficult to assess such conclusions without a linguistic methodology in place. Discourse analysis provides a method (or, in fact, a variety of methods) by which to assess various conclusions about the text.
Because of the importance of discourse analysis and because of the variety of approaches within the field, we are dedicating two chapters to this topic. Each chapter will focus on one major “school” of discourse analysis, though there are many others. This chapter focuses on approaches to discourse analysis developed by M. A. K. Halliday, which have been applied by several scholars to the Greek New Testament. The following chapter focuses on the work of Stephen H. Levinsohn and Steven E. Runge. To put both of these chapters in context, we will begin by exploring four major schools of discourse analysis that are relevant to New Testament studies.
7.2 Four Major Schools of Discourse Analysis
Writing twenty years ago, Stanley E. Porter sketched out four major schools of discourse analysis in an essay belonging to the important volume, Discourse Analysis and Other Topics in Biblical Greek.3 Even then, the four schools that Porter addressed did not exhaust the various approaches to discourse analysis that existed in 1995, and today that is all the more the case. Nevertheless, Porter’s essay remains a useful introduction to the types of concerns found in discourse analysis, how methodologies differ across schools, and who some of the key exponents working in the field are — at least those who impact New Testament studies. It also offers mild critique of these schools. The following is summarized from Porter’s essay, with some argumentations.
7.2.1 Summer Institute of Linguistics (SIL)
Influenced by linguists such as Nida, Pike, and Lamb,4 SIL has predominantly been occupied with issues relating to Bible translation.5 Initially, the SIL school demonstrated interest in layers of language, proceeding from the smallest elements of language to increasingly larger structures (see §1.4.11).6 Kathleen Callow, for example, contributed a discussion of elements beyond the sentence level in her book, Discourse Considerations in Translating the Word of God.7 After Callow, however, Porter observes a return to sentence-level interest, citing works by Stephen H. Levinsohn (to be discussed in the next chapter) and David A. Black, who devote significant space to the sentence. While Porter applauds the contribution of SIL to Bible translation, he critiques the movement for not extending far beyond sentence-level analysis.8
According to Porter, M. A. K. Halliday and R. Hasan offer the most integrative approach of the four schools discussed.9 Rather than treating discourse as the extension of the sentence, Halliday and Hasan begin with discourse at the broad level, discussing the ideational, interpersonal, and textual features of text (see §§2.4.1 – 2).10 Since Halliday’s approach to discourse analysis is the topic of this chapter, more will be said later. For now, however, it is worth noting that this school of discourse analysis has not been “mapped” into Ancient Greek yet (having been developed in English), meaning that its correlation to the language of the New Testament must be considered by the practitioner. While there is much theoretical merit to the approach of this school, it requires a high level of commitment to be of use to New Testament studies. Fortunately, some scholars have already attempted to bridge the gap, such as Porter, Reed, Guthrie, and Westfall.11
This school is the least cohesive, demonstrating a number of influences and a focus on rhetoric. While there is a helpful concern with the macrostructure of texts in the work of scholars such as Hellholm, Schenk, and Johanson,12 Porter thinks the wide-ranging and interdisciplinary features of the Continental European school is so diverse that it becomes its own liability.13
J. P. Louw’s essay “Discourse Analysis and the Greek New Testament” first introduced discourse analysis to the wider New Testament world,14 and his later Semantics of New Testament Greek and A Semantic Discourse Analysis of Romans have been important influences on the field.15 Louw developed a method of colon analysis in which the text is analysed through its constituent cola. These cola are units formed around a nominative and predicate structure.16 The interconnectedness of the cola is diagrammed, demonstrating the relationships between the smallest units of meaning (the cola) and larger semantic units. While Porter regards Louw’s South African school of discourse analysis to be the best-coordinated school, having widespread influence, he identifies some weaknesses. The diagrammed structure of cola relationships is somewhat subjective in its creation, and in fact multiple structures may be offered. The question is how to judge which possible structure is best, and there is no clear methodology to address this problem.17 There are also questions raised about the nature of the colon itself and how syntax may affect analysis of cola.
As acknowledged above, the foregoing sketch of these four schools of discourse analysis does not include recent developments and is no doubt an oversimplification. Nevertheless, it remains useful insofar as it introduces the seminal currents of New Testament discourse analysis, and it helps us to understand contemporary developments in the context of the discussion. We turn now to look more closely at Halliday’s approaches to discourse analysis.
For Halliday, the central concern of discourse analysis is cohesion. His conception of cohesion was developed with Hasan in the seminal volume Cohesion in English,18 with a more recent expression found in Halliday and Matthiessen’s Halliday’s Introduction to Functional Grammar.19 Cohesion refers to the way in which a text hangs together: “it refers to relations of meaning that exist within the text, and that define it as a text.”20 According to Halliday and Matthiessen, language has “resources for managing the flow of discourse: for creating semantic links across sentences — or rather, semantic links which work equally well either within or across sentences.”21
The study of cohesion asks how speakers form texts into complete units. What makes a text a text rather than a collection of unrelated utterances? In fact, any text or speech will fit somewhere on a spectrum of cohesiveness that ranges from highly coherent through to incoherent. So, why are some texts more coherent than others? As Labov says, “The fundamental problem of discourse analysis is to show how one utterance follows another in a rational, rule-governed manner — in other words, how we understand coherent discourse.”22 Reed further elaborates:
At a very basic level, linguistic cohesiveness refers to the means by which an immediate linguistic context meaningfully relates to a preceding context and/or a context of situation (i.e. meaningful relationships between text, co-text and context). Linguistic cohesiveness provides speakers with the means to produce a “message” (i.e. theme) from individual and sometimes unrelated words and phrases.23
In answer to the question of what makes a text coherent or incoherent (or somewhere in between), Halliday and Hasan refer to two types of context. The first is the relationship between the text and its inner world — its co-text (what is normally called “context” or “literary context”).24 The second is the relationship between the text and its external world — its context of situation and culture (usually referred to as “historical context” or “background” within New Testament studies).25
The second type of context “refers to all those extra-linguistic factors which have some bearing on the text itself.”26 It aids cohesion in that the reader mentally links the text to a concept in the culture to which the text belongs.27 If a text’s cultural context is totally alien to the reader, its effective coherence would be significantly compromised. In this way, the coherence of a text is somewhat subjectively determined, depending on the reader’s ability to understand its external references.
The context of situation relates closely to register (see §6.4), since “the more specifically we can characterize the context of situation, the more specifically we can predict the properties of a text in that situation.”28 Cohesion and register work together to define a text: “A text is a passage of discourse which is coherent in these two regards: it is coherent with respect to the context of situation, and therefore consistent in register; and it is coherent with respect to itself, and therefore cohesive.”29
The first type of context does not depend on the world outside the text (external references), but is entirely internally focused. Halliday and Hasan refer to the interpretation of some element in the discourse that is dependent on that of another.30 The elements of a text are mutually defining and cannot be understood apart from each other. Cohesive relationships will be observed between words, phrases, sentences, and pericopes.
Both types of context are, of course, important for understanding any text. Our primary interest here, however, is the first. What are the ways in which a text creates cohesion within itself? It is to this topic we now turn.
Halliday and Matthiessen identify four ways in which cohesion is created in English: through conjunction, reference, ellipsis, and lexical cohesion.31 These four resources for cohesion are sketched out below.
Conjunction refers to markers that indicate that a clause relates to a previous one, thus creating continuity in a text. Words such as “but,” “meanwhile,” and “so” are obvious conjunctions in English.32 Subtypes of conjunction include elaboration, extension, and enhancement. Elaboration refers to the relationship between clauses in which one clause elaborates on another, either by way of apposition or clarification (epexegesis).33 Extension involves addition to a clause or a variation from it.34 Enhancement types include spatio-temporal (place, time), manner (how, why), causal-conditional (result, contingency), and matter (topic).35
Obviously, Greek conjunctions such as καί, ἀλλά, and ἵνα, and other connectives such as τότε and διὰ τοῦτο, may be understood as cohesive devices in the ways that Halliday and Hasan elucidate.
Reference creates cohesion by creating links between elements in a text. Reference can be exophoric, meaning a text points to things outside the text, or endophoric, meaning it points to something within the text itself.36 Repeated reference to the same things creates referential chains; such chains are natural bonds of cohesion. Repeated reference need not involve the same word, but there are a variety of means with which to refer to the same item: the ball. . . which he threw. . . it was red. Here the item (the ball) is referenced by means of the word “ball,” but also “which” and “it.” These three references to the same thing create a referential chain. Types of reference include personal (reference to the same person), demonstrative (reference using “this” or “that”), and comparative reference (relations of contrast, using words such as “different,” “bigger”).37 For Greek, we should pay special attention to the personal, reflexive, and relative pronouns ἔγω, ἐμαυτοῦ, and ὅς, as well as the demonstratives οὗτος and ἐκεῖνος.
Whereas reference refers to explicit markers that point to a common item (the referent), ellipsis leaves things out when they can be presumed from the context.38 The fact that the reader can supply the missing element/s from the wider context creates cohesion since the ellipsis itself depends on the text being read as a whole. Romans 3:27 provides a good example of this:39
Where then is boasting? It is excluded. By what kind of law? Of works? No, but a law of faith. (NASB)
The nonelliptical form would read:
Where then is boasting? It is excluded. By what kind of law [is boasting excluded]? [Is boasting excluded by a law] Of works? No [boasting is not excluded by a law of works], but [boasting is excluded by] a law of faith.
This kind of cohesion relies on the relationships between lexical items. Some types of lexical relationships include repetition, synonymy, hyponymy, meronymy, and collocation.40 Repetition refers, obviously, to the repetition of the same lexeme.41 An example of this is found in Romans 3, through the repetition of the word faith:
This righteousness is given through faith. . . . God presented Christ as a sacrifice of atonement, through the shedding of his blood — to be received by faith . . . he did it to demonstrate his righteousness at the present time, so as to be just and the one who justifies those who have faith in Jesus. . . . [Boasting is excluded] because of the law that requires faith. For we maintain that a person is justified by faith. (Romans 3:22, 25, 26, 27, 28, NIV)
Synonymy, however, creates lexical cohesion by using multiple words that overlap in meaning (they do not need to be perfectly overlapping to be described as synonymous).42 Hyponymy is about classification, so that one lexical item represents a class of thing and the second represents a (subordinate) subclass within the first (superordinate) class.43 Meronymy refers to lexical items that are each a part of something else, like flowers and fountains, which are meronyms of garden.44 Collocation recognizes a particular association between words that tend to occur together, such as flesh and blood, day and age, fish and chips.45
In the final chapter of Halliday and Hasan’s Cohesion in English, they suggest a method for the analysis of cohesion in a text. Key to the general principles of analysis of cohesion is the concept of the “tie,” which is a cohesive element that connects parts of a text together.46 In order to analyze any sentence, we should determine how many cohesive ties it contains, and “how many instances of a cohesive element that are not resolved by presupposition within the sentence. This shows the total extent of the demands it makes on the preceding (or rather the surrounding) text.”47 Next, it is important to specify what type of cohesion is involved (e.g., reference, substitution, et al.).48
Other pertinent questions relate to genre (differing genres “may exhibit a general tendency towards the use of certain features or modes rather than others”),49 idiolect (do particular individuals favor one type of cohesion over others?), variation of the density of cohesive ties at various points in a text,50 and the relation between cohesion and the division of a text into paragraphs.51
In order to explore further Halliday and Hasan’s notion of cohesive ties, we will draw on Jeffrey Reed’s summary of them since he relates them to the Greek of the New Testament.52
Organic ties are elements built into a language for connective purposes, such as markers of transition (e.g., γάρ, ἀλλά, δέ, καί), grammatical structures (e.g., genitive absolute using γίνομαι), and certain lexical items (e.g., λοιπόν).53 Organic ties consist of two functional systems, parataxis and hypotaxis. The former involves the relation between items of equal status, while the latter involves the relation between an independent item and an item that is dependent on it.54 Organic ties also set boundaries within a discourse; “they provide a means of organizing groups of componential ties [see below] into thematic sections/paragraphs.”55
Componential ties concern “the meaningful relationships between individual linguistic components in the discourse (e.g., repetition of words).”56 In other words, these are ties between words or phrases. There are three types of componential ties: co-reference, co-classification, and co-extension.
7.5.2.1 Co-reference. Co-reference refers to cohesive ties between items of the same identity (see §7.4.2). This might involve “person deixis,” which is the encoding of various participants into a discourse using pronouns, proper nouns, and verbal suffixes.57 Temporal deixis can create cohesion through references to the same temporal frame, indicated by adverbs such as τότε, νῦν, and μέχρι.58 Place deixis is closely related, tying a section together by the fact that events are unfolding in a single location.59
7.5.2.2 Co-classification. This second type of componential tie involves cohesive ties between linguistic items of the same class or genus. By substituting one item for another in the same class, this type of tie is created. Reed offers the following example: “I want the children to draw with crayons, and I want the teenagers to draw with pencils.” He says, “By substituting ‘teenagers’ for ‘children’ and ‘with pencils’ for ‘with crayons,’ the two sentences form a cohesive tie of co-classification with respect to who should do the drawing and how it should be done.”60
7.5.2.3 Co-extension. The third type of componential tie “refers to cohesive ties between linguistic items of the same semantic field, but not necessarily of the same class.”61 Reed elaborates:
In the sentences “John ate the pizza” and “Susie gobbled down the cake” the linguistic pairs “John” / “Susie,” “ate” / “gobbled down” and “pizza” / “cake” do not refer to the same entities nor do they refer to the same class (e.g. pizza is not a kind of cake). . . . By using words with similar senses speakers talk about similar things in similar ways.62
When cohesive ties are used throughout a text, they create different kinds of referential chains (see §7.4.2). “Identity chains” are expressed by co-classificational ties, while “similarity chains” are expressed by co-classificational and co-extensional ties.63 While the discernment of such referential chains form only a part of how a text achieves cohesion, they are nevertheless objectively and explicitly determined. That is, the discourse analyst can point to these explicit markers with a degree of ease in describing how a particular text hangs together.
The foregoing represents a summary of Halliday and Hasan’s theory of cohesion. Their approach has several strengths and useful applications for study of the Greek New Testament, some of which ought to be self-evident. Nevertheless, there are a few drawbacks, including criticisms of the theory and the lack of direct application to the language of Greek (having been developed in English).
Some of the critique of Halliday and Hasan is summarized by Annemieke Drummen:
Halliday and Hasan’s description of cohesion has been criticized (e.g. Carrell 1982, Sanders et al. 1992, Sanders and Pander Maat 2006) for treating the concept as a necessary condition for the connectedness of a discourse. In Carrell’s view (1982:486), cohesion is not the cause, but the effect of coherence. Similarly, Sanders et al. (1992:2 – 3) point out that cohesive elements are “important though not necessary features of discourse; they are linguistic markers, expressing the underlying conceptual relations that are of a cognitive nature.” These authors present coherence and cohesion as alternative approaches, but in fact (the students of) both concepts look at different phenomena, and can therefore also be seen as complementary.
Investigating cohesion means focusing on the linguistic reflections of coherence. Tanskanen (2006:7), for instance, adopts this milder view, assuming that cohesion contributes to coherence.64
Clearly there is some disagreement as to whether cohesion effects the coherence of a text or is itself the effect of its inherent coherence. Either way, students of the Greek New Testament will benefit from tracing the coherence of a text through the markers of cohesion elucidated by Halliday and Hasan.
An obvious drawback for the application of Halliday and Hasan’s theory of cohesion to the Greek of the New Testament, however, is that it is developed with respect to English and has not been “translated” into Greek, so to speak. There have been various attempts to use Hallidayan approaches for the study of the New Testament, and these offer helpful models for others who might wish to do likewise.65 That is not the same, however, as a comprehensive account of how Halliday “maps” onto Greek.
Perhaps the work that most closely approximates this goal is found outside New Testament Greek scholarship: Discourse Cohesion in Ancient Greek, edited by Stephanie Bakker and Gerry Wakker.66 This volume consists of papers presented at the Sixth International Colloquium on Ancient Greek Linguistics in June 2007 and acknowledges the foundational role that Halliday and Hasan occupy in discussions about cohesion. The two editors state:
there is no consensus on what exactly discourse cohesion is (nor on the question whether it really exists . . .). Most scholars working on discourse cohesion, however, will more or less accept the ideas of Halliday and Hasan (1976), the fathers [sic] of the concept of cohesion.67
The essays in this volume explore the cohesion functions of various parts of speech, such as third person pronouns,68 various particles and conjunctions,69 differing text types,70 and some verbs.71 The scope of literature discussed ranges through Homer, Lysias, Euripides, Hesiod, Greek tragedy, comedy, drama, and Classical and Ancient Greek in general.
This volume represents a significant addition to the literature that explores cohesion in Ancient Greek. Nevertheless, it is not the comprehensive “translation” of Halliday and Hasan into Greek that remains desirable. In fact, the volume exhibits a concentration on Greek particles, leaving much else of the language uncovered. This observation is not intended to be critical, since the volume is a collection of conference papers rather than a comprehensive treatment of the topic. But the fact remains that it is not a comprehensive treatment of the topic. Moreover, it obviously does not offer an application of Halliday and Hasan to the Greek of the New Testament.
Bakker, Stephanie and Gerry Wakker, eds. Discourse Cohesion in Ancient Greek. Amsterdam Studies in Classical Philology 16. Leiden: Brill, 2009.
Guthrie, George H. The Structure of Hebrews: A Text-Linguistic Analysis. NovTSup 73. Leiden: Brill, 1994.
Halliday, M. A. K., and R. Hasan. Cohesion in English. London: Routledge, 1976.
Halliday, M. A. K., and Christian M. I. M. Matthiessen. Halliday’s Introduction to Functional Grammar. Fourth revised edition. London: Routledge, 2014.
Porter, Stanley E. “Discourse Analysis and New Testament Studies: An Introductory Survey.” Pages 14 – 35 in Discourse Analysis and Other Topics in Biblical Greek. Edited by Stanley E. Porter and D. A. Carson. JSNTSup 133. Sheffield: Sheffield Academic, 1995.
Reed, Jeffrey T. “The Cohesiveness of Discourse: Towards a Model of Linguistic Criteria for Analyzing New Testament Discourse.” Pages 28 – 46 in Discourse Analysis and the New Testament: Approaches and Results. JSNTSup 170. Edited by Stanley E. Porter and Jeffrey T. Reed. Sheffield: Sheffield Academic, 1999.
Westfall, Cynthia Long. A Discourse Analysis of the Letter to the Hebrews: The Relationship between Form and Meaning. JSNTSup 297. New York: T&T Clark, 2005.
1. Some early contributions (other than those by members of SIL — see §7.2.1) include Wolfgang Dressler, Einführung in die Textlinguistik (Tübingen: Niemeyer, 1972); Joseph E. Grimes, The Thread of Discourse (The Hague: Mouton, 1975); Teun A. van Dijk, Text and Context: Explorations in the Semantics and Pragmatics of Discourse (London: Longman, 1977); Gillian Brown and George Yule, Discourse Analysis (Cambridge: Cambridge University Press, 1983); M. Coulthard, An Introduction to Discourse Analysis (2nd ed.; London: Longman, 1985); Timothy W. Crusius, Discourse: A Critique and Synthesis of Major Theories (New York: Modern Language Association, 1989).
2. Crystal, Linguistics and Phonetics, 148.
3. Stanley E. Porter and D. A. Carson, eds., Discourse Analysis and Other Topics in Biblical Greek (JSNTSup 113; Sheffield: Sheffield Academic Press, 1995). Porter’s essay in the volume is “Discourse Analysis and New Testament Studies: An Introductory Survey” (14 – 35).
4. E. A. Nida, Toward a Science of Translating: With Special Reference to Principles and Procedures Involved in Bible Translating (Leiden: Brill, 1964); Pike, Unified Theory; S. M. Lamb, Outline of Stratificational Grammar (Washington: Georgetown University Press, 1966).
5. It is possible to discern two distinct strands within SIL. American members such as Pike and Longacre put emphasis on extending the grammatical hierarchy to include paragraph and discourse levels. Kathleen Callow, John Callow, and John Beekman are more European (Firthian) in their approach, which is known as “Semantic Structure Analysis.”
6. Porter, “Discourse Analysis,” 25.
7. Kathleen Callow, Discourse Considerations in Translating the Word of God (Grand Rapids: Zondervan, 1974).
8. Porter, “Discourse Analysis,” 25 – 26. Porter overstates this criticism, since several of Levinsohn’s publications deal with issues well beyond the level of the sentence. See, e.g., Stephen H. Levinsohn, Textual Connections in Acts (Atlanta: SBL, 1987); “The Groupings and Classification of Events in Mark 14,” Notes on Translation 66 (1977): 19 – 28; “Preliminary Observations on the Use of the Historic Present in Mark,” Notes on Translation 65 (1977): 13 – 28.
9. Halliday and Matthiessen, Functional Grammar; Halliday and Hasan, Cohesion; idem, Language, Context, and Text.
10. Porter, “Discourse Analysis,” 28.
11. Stanley E. Porter, Idioms of the Greek New Testament (2nd ed.; Sheffield: Sheffield Academic Press, 1994), 298 – 307; J. T. Reed, “Cohesive Ties in 1 Timothy: In Defense of the Epistle’s Unity,” Neot 26 (1992): 131 – 47; George H. Guthrie, The Structure of Hebrews: A Text-Linguistic Analysis (NovTSup 73; Leiden: Brill, 1994), 45 – 75; Cynthia Long Westfall, A Discourse Analysis of the Letter to the Hebrews: The Relationship between Form and Meaning (JSNTSup 297; New York: T&T Clark, 2005).
12. D. Hellholm, Das Visionenbuch des Hermas als Apokalypse: Formgeschichtliche und texttheoretische Studien zu einer literarischen Gattung. I. Methodologische Vorüberlegungen und makrostrukturelle Textanalyse (ConBNT 13.1; Lund: Gleerup, 1980); W. Schenk, Der Philipperbrief des Paulus (Stuttgart: Kohlhammer, 1984); B. C. Johanson, To All the Brethren: A Text-Linguistic and Rhetorical Approach to 1 Thessalonians (ConBNT 16; Stockholm: Almqvist & Wiksell, 1987).
13. Porter, “Discourse Analysis,” 31.
14. J. P. Louw, “Discourse Analysis and the Greek New Testament,” BT 24 (1973): 101 – 18.
15. J. P. Louw, Semantics of New Testament Greek (Philadelphia: Fortress, 1982); idem, A Semantic Discourse Analysis of Romans (Pretoria: University of Pretoria, 1987).
16. Porter, “Discourse Analysis,” 33.
17. Ibid., 34.
18. Halliday and Hasan, Cohesion.
19. Halliday and Matthiessen, Functional Grammar, chapter 9.
20. Halliday and Hasan, Cohesion, 4.
21. Halliday and Matthiessen, Functional Grammar, 114.
22. William Labov, Sociolinguistic Patterns (Philadelphia: University of Philadelphia Press, 1972), 252.
23. Jeffrey T. Reed, “The Cohesiveness of Discourse: Towards a Model of Linguistic Criteria for Analyzing New Testament Discourse,” in Discourse Analysis and the New Testament, 29.
24. Halliday and Hasan, Cohesion, 6 – 19.
25. Ibid., 19 – 21.
26. Ibid., 21.
27. Reed, “Cohesiveness of Discourse,” 32.
28. Halliday and Hasan, Cohesion, 22.
29. Ibid., 23.
30. Ibid., 4.
31. Halliday and Matthiessen, Functional Grammar, 603. Their treatment summarizes the major discussions of these categories in Halliday and Hasan, Cohesion: conjunction is the burden of ch. 5 (pp. 226 – 73); reference, ch. 2 (pp. 31 – 87); ellipsis, ch. 4 (pp. 142 – 225; see also substitution in ch. 3); and lexical cohesion, ch. 6 (pp. 274 – 92).
32. Halliday and Matthiessen, Functional Grammar, 605.
33. Ibid., 615 – 16.
34. Ibid., 616 – 17.
35. Ibid., 617 – 20.
36. Ibid., 623 – 25.
37. Ibid., 626 – 34.
38. Ibid., 606.
39. The biblical examples in this section were developed by Daniel Waldschmidt for an Advanced Greek class at Trinity Evangelical Divinity School, Spring 2014.
40. Halliday and Matthiessen, Functional Grammar, 644 – 50.
41. Ibid., 644 – 45.
42. Ibid., 645 – 46.
43. Ibid., 646 – 47.
44. Ibid., 647 – 48.
45. Ibid., 648 – 50.
46. Halliday and Hasan, Cohesion, 329.
47. Ibid., 332.
48. Ibid.
49. Ibid.
50. Ibid.
51. Ibid., 333.
52. Reed, “Cohesiveness of Discourse,” 28 – 46.
53. Ibid., 32 – 33.
54. Ibid., 33.
55. Ibid., 36.
56. Ibid.
57. Ibid., 38.
58. Ibid., 39.
59. Ibid., 40.
60. Ibid., 40 – 41.
61. Ibid., 41.
62. Ibid.
63. Ibid., 43.
64. Annemieke Drummen, “Cohesion,” in Encyclopedia of Ancient Greek Language and Linguistics (ed. Georgios K. Giannakis et al.; Brill Online, 2014).
65. See Jeffrey T. Reed, Discourse Analysis of Philippians: Method and Rhetoric in the Debate over Literary Integrity (JSNTSup 136; Sheffield: Sheffield Academic, 1997); idem, “To Timothy or Not: A Discourse Analysis of 1 Timothy,” in Biblical Greek Language and Linguistics: Open Questions in Current Research (ed. Stanley E. Porter and D. A. Carson; JSNTSup 80; Sheffield: Sheffield Academic, 1993), 90 – 118; Guthrie, Hebrews; Westfall, Hebrews. There are also various points of confluence between Halliday and Levinsohn (cf., e.g., 7.5.1 and 8.4.1 – 2; 8.8).
66. Stephanie Bakker and Gerry Wakker, eds., Discourse Cohesion in Ancient Greek (Amsterdam Studies in Classical Philology 16; Leiden: Brill, 2009).
67. Ibid., xii.
68. Anna Bonifazi, “Discourse Cohesion through Third Person Pronouns: The Case of κεῖνος and αὐτός in Homer,” in ibid., 1 – 20.
69. Stéphanie J. Bakker, “On the Curious Combination of the Particles γάρ and οὖν,” in ibid., 41 – 62; Gerry C. Wakker, “Well I Will Now Present My Arguments”: Discourse Cohesion Marked by οὖν and τοίνυν in Lysias,” in ibid., 63 – 82; Antonio R. Revuelta Puigdollers, “The Particles αὖ and αὖτε in Ancient Greek as Topicalizing Devices,” in ibid., 83 – 110; A. Maria van Erp Taalman Kip, “Καὶ μήν, καὶ δή and ἤδη in Tragedy and Comedy,” in ibid., 111 – 34; Annemieke Drummen, “Discourse Cohesion in Dialogue. Turn-Initial ἀλλά in Greek Drama,” in ibid., 135 – 54; Coulter H. George, “Greek Particles: Just a Literary Phenomenon?,” in ibid., 155 – 70.
70. Rutger J. Allan, “Towards a Typology of the Narrative Modes in Ancient Greek. Text Type and Narrative Structure in Euripidean Messenger Speeches,” in ibid., 171 – 204; Albert Rijksbaron, “Discourse Cohesion in the Proem of Hesiod’s Theogony,” in ibid., 241 – 66.
71. Louis Basset, “The Use of the Imperfect to Express Completed States of Affairs. The Imperfect as a Marker of Narrative Cohesion,” in ibid., 205 – 20; Sander Orriens, “Involving the Past in the Present: The Classical Greek Perfect as a Situating Cohesion Device,” in ibid., 221 – 40.