1. Introduction: Composition theory, linguistics, and improving student writing.

2. Review of approaches in composition studies.

2.1 The cognitivist approach.

2.2 The expressionist approach.

2.3 The social-epistemic approach.

2.4 Multiculturalism, the contact zone, and Bakhtin’s heterogeneity of voices

2.5 The context of written discourse: The rhetorical situation.

1. Exigence: The occasion and motivation for speaking/writing

2. Audience: Who are you speaking-writing to?

3. Speaker-writer: What are your roles, identities, voices, and attitudes?

2.6 Berman and the development of discourse.

2.7 Discourse stance, referential content, and global text organization.

1. Global text organization: Top-down (expository)/ bottom-up (narrative)

2. Categories of referential content.

3. Discourse stance: orientation, attitude, generality.

3. Methodology.

3.1 Creating a student writing guide: structure, aims, and intended use.

1. Ideologies of Writing in the SWG.

2. SWG, Introduction and recap.

3. SWG Parts 1-3, occasion, audience, and speaker-voice.

4. SWG Parts 4-5, information units, and structure/ overall organization.

1. Introduction: Composition theory, linguistics, and improving student writing

This thesis, in integrating theory from rhetoric/composition and linguistics, attempts to merge insights from two fields that, surprisingly, are not often intermingled. A field that is devoted to the study of written communication ought to be closely associated with a field that attempts to answer the most fundamental questions about the forms and functions of human language. Granted, it seems that the fields of linguistics and composition have been gradually benefiting mutually from one another, as evident by the creation of Linguistics, Composition and Rhetoric departments in some universities, but composition studies and linguistics have often been in “complementary distribution,” as it were, in the literature. This was not always the situation, Faigley (1989) points out, as “in earlier decades of the CCCC” (the major conference for composition studies), linguists had been seriously involved in composition studies, and that, “in the 1950s, linguists published articles frequently in CCC and held major offices in the organization” (1989: 241). “In the 1960s,” Faigley continues, “when rhetoric and composition blossomed as a discipline, [some] saw English teachers’ awareness of linguistics as the most important development of the first two decades of the CCCC, and [others] proposed linguistics as the basis of a modern theory of rhetoric” causing the 1960’s to be known as “the decade of language study and rhetorical theory” (1989: 241).

However, the 1970’s saw a gradual decline of linguists working in the field of composition and rhetoric that was attributable to a movement in the social sciences called the cognitive turn. The cognitive turn took shape in composition studies in the form of a new approach to writing—process theory—that re-envisioned writing as a structured process of discrete operations which could be described and analyzed rigorously. Thus, the 1970s became the decade of process theory so that by “1980 studies of writers’ processes had clearly become ascendant over studies of writers’ language” (1989: 241). Moreover, “language studies declined in importance” because of a linguistic turn in the field of linguistics, spawned by Chomsky’s theory of generative grammar, which diverged from the then dominant linguistic approaches in both aims and assumptions (1989: 241). Faigley recounts how the linguistic turn saw linguists re-prioritizing the formal aspects of language over and above the pedagogical, rhetorical and literary aspects of language:

“Chomsky’s theory of transformational-generative grammar influenced the study of language in North America as no other theory had in the past […] Language could be orderly only if it were idealized. If actual language was used as data, the orderliness of language predicted by generative grammar soon disintegrated. Chomsky insisted that language be viewed as abstract, formal, intuitive, and acontextual. His goal for a theory of language was describing a human’s innate capacity for language, not how people actually use language. When asked what relevance the study of linguistics had for education, Chomsky answered absolutely none. Gradually, those interested in studying discourse came to heed Chomsky’s warnings” (1989: 241-42).

Chomsky’s influence, Faigley suggests, was detrimental for those wishing to study discourse as it “accelerated the formation and growth of separate linguistics departments committed to the theoretical study of formal structure in language,” and up until now, theoretical linguists “have tended to dismiss as ‘uninteresting’ any applied questions about language, such as the educational implications of language theory,” asking only “‘interesting’ questions about abstract universals underlying language” (1989: 242). Faigley speculates that “had Chomsky’s influence been restricted to language theory, linguists in North America might have remained more active in the study of writing”—admitting, however, that “to blame Chomsky, […] for the decline of linguistics within compositions studies is not merely simplistic; it is wrong” (1989: 242).

Constructing a student writing guide grounded in findings on discourse development and built upon the foundation of principles of rhetoric/composition demonstrates how linguistics and composition theory can work harmoniously in concert. Linguistic findings on discourse and rhetoric/ composition theories can be mutually forged into a pedagogically-aimed tool to address the difficulties developing student-writers face when they sit down to write. To bring together insights from both of these fields in a useful way, constructing a guide for developing student-writers, is covertly aimed at satisfying this stark lacuna that demands filling.

Another aim underlying this thesis is to give writing-students a perspective and a grasp on their own written communication that traditional composition pedagogy has often denied them. Composition is a difficult exercise because it is “a highly abstract cognitive process” and an awkward communicative exercise because communication across the medium of paper seems to be heavily decontextualized from the rhetorical situation, stripped of the components that usually accompany real-life communication (Cheng & Steffensen 1996: 150). The challenge of composition may lie in the fact that discourse abstracted away form the contexts that gave it birth is less-readily meaningful  and coherent to its speaker-writers. A piece of notebook paper, the screen of a computer, or whatever other silent medium a writer happens to use, de-contextualizes the communicative moves that student writers are one the one hand used to deploying in real communicative (social) contexts, but on the other hand must now make in an artificial, abstracted dimension. The discourse of an academic paper is removed from the audible, real-life exigence (occasion) that usually gives it birth. The argumentation engaged in by able-students in public contexts, among colleagues, must, in composition, take place on a piece of paper. When “we hear naturally spoken language […] the music of prosody enacts some of the meaning so that we ‘hear’ it […] as though the meaning comes to us rather than us having to go after it” (Elbow 2006: 643). Audibility, which “tends to draw [an audience] into and through the words, increasing our experience of energy,” is often stripped away during the writing process (Elbow 2006: 643). This lack of audible, real-life context in writing diminishes the communicative energy that is so essential to verbal thought, creating roadblocks for students in the brainstorming and composing process. Despite the good intentions of many teachers of composition, traditional writing pedagogy seems often to stifle rather than to evoke the kind of dialogue that fosters a sense of the rhetorical situation used in writing. It seems that much of traditional writing pedagogy has not helped student-writers breathe communicative energy from real-life social contexts back into their writing.

Cheng & Steffensen (1996) aptly point out that many students of writing are often unable to consider a rhetorical audience in their writing because “most composition classrooms” fail “to develop into forums or discourse communities” which can “evoke a sense of audience in student writers” (1996: 152). Writing without consideration of the rhetorical situation results in a kind of artificial discourse produced in response to no other exigence, or set of motivating factors, than to produce a written product for the sake of producing a written product. This kind of artificial discourse is aimed at and motivated by satisfying the demands of a writing teacher. Bartholomae (1985) calls such discourse a “bastard discourse” produced by “students who are good at it and have learned to cope with academic tasks by developing a knowledge-telling strategy [that] undermines educational efforts to extend the variety of discourse schemata available to students” (1985: 633).

Part of the solution to this problem may lie in the use of the imagination to compensate for the absence of the audibility of real discourse contexts, to essentially re-imagine a rhetorical or discourse situation, to breathe context back into (re-inspire) student writing. The artificiality that characterizes many developing writers’ essays seems to be largely due to a lack of a sense of context that their written discourse ought to be couched in: the situation out of which their writing ought to flow and to which their writing ought to respond. Therefore, the final aim of this thesis it to produce and test a student writing guide to help students re-imagine a discourse (writing) situation and to deploy facets of well-formed discourse toward the (imagined) communicative ends in their (imagined) writing situations. After constructing the SWG by integrating insight from rhetoric/ composition studies and findings from empirical research on the development of well-formed discourse, the SWG will be assessed for its effect on drafts of argumentative essays that student-writers used to revise them.

In the following sections, I first provide a review of literature (section 2) in which I outline major trends in the field of composition studies (2.1-2.3), highlight relevant background material from literary and rhetorical studies (2.4-2.5), and then review findings from Berman’s cross-linguistic research about the development of discourse, focusing on three of Berman’s five facets of discourse development (2.6-2.7). The review informs a student writing guide (Appendix I) intended to aid freshman (level) composition students or advanced English language learners in the production of well-formed expository text (argumentative essays). In section 3, methodology, I discuss the structure and intended uses of the SWG and explain the reasons for the form it takes. Information about the subjects who participated in the study and the steps involved in assessing the efficacy of the SWG are also outlined. Finally, in section 4, I discuss my findings and, in section 5, I present my ideas for future research in this area.

2. Review of approaches in composition studies

The field of rhetoric/composition studies (or, simply composition studies) is represented by teachers and theorists who seek to study, contribute to, and enhance the theory and practice of writing pedagogy and whose articles often appear in journals such as College Composition and Communication and College English. Composition studies is a highly interdisciplinary and theoretically diverse field comprised of scholars who often represent views of writing theory and pedagogy that are based on varying epistemologies and ideologies. To characterize this, Berlin (1988) reviewed three major, differing approaches to composition and writing pedagogy in composition studies, pointing out that “a way of teaching is never innocent [and] pedagogy is imbricated in ideology, in a set of tacit assumptions about what is real, what is good, what is possible, and how power ought to be distributed” (492). For Berlin, no discourse or practice is free from ideology. Van Dijk’s socio-cognitive definition of ideology seems to be the most illustrative:

“[ideologies are] basic frameworks of social cognition, shared by members of social groups, constituted by relevant selections of sociocultural values, and organized by an ideological schema that represents the self-definition of a group [that sustain] the interests of groups [and] have the cognitive function of organizing the social representations (attitudes, knowledge) of the group, and thus indirectly monitor the group-related social practices, and hence also the text and talk of members” (van Dijk, 1995: 248).

If ideology is inscribed deeply on the way composition scholars/teachers approach their object of inquiry, then ways of teaching—including the text and talk used to deploy teaching methods—apparently are trapped, as it were, in an inescapable ideological circle. Certainly, even a cursory glance at the literature shows fierce, seemingly unresolvable debates in composition studies, which rage in public discourse; one particular instance of this is as a several-year long debate that took place about politics in the writing classroom that intensified to the point of professional friendships becoming “dissolved in the heat of the argument,” as Hariston put it (1993: 255). The underlying ideologies that composition teachers and theorists subscribe to are clearly a matter of great consequence.

What follows is a brief review of three major perspectives in composition studies—cognitive, expressionist, and social-epistemic approaches—that represent different ideologies and epistemologies. In addition to these perspectives, the concept of multi-voicedness, which has important applications in writing pedagogy, will be discussed. Then, I discuss literature pertaining to the “rhetorical situation,” which characterizes the context of written discourse. Finally, I review three of Berman’s (2008) five facets of discourse production, focusing on features of well-formed discourse which will later be integrated into the SWG.

2.1 The cognitivist approach

A cognitivist approach, recognized by many as the most traditional approach, is grounded in cognitive psychology and under-girded by an Aristotelian epistemology that views truth as objective, unambiguously knowable, and communicable from person to person (Berlin 1988: 488). The well-known “process over product” idea that urges writing pedagogy to address the writing process rather than the written product was promoted most vocally by (but did not originate with) the cognitivists, such as Flower & Hayes (1981), who viewed the writing process

“as a set of distinctive thinking processes which writers orchestrate or organize during the act of composing [and which] have a hierarchical, highly embedded organization in which any given process can be embedded within any other” (Flower & Hayes 1981: 366).

Cognitive rhetoric, therefore, brakes up the writing process into discrete, identifiable and logical steps in the sequence of pre-writing (planning), writing (translating), and rewriting (reviewing). A key assumption that cognitivists hold is that universal laws that underlie the writing process are discoverable; that the “features of composing” are analyzable “into discrete units” expressible in “linear, hierarchical terms;” and that a writer’s mind is like a computer processor that moves logically through operations in order to generate writing; and that each step of the writing process creates “a hierarchical network of goals [which] in turn guide the writing process” (Berlin 1988: 481-82,  Faigley 1986: 533).

Thus, writing pedagogy with a cognitivist bent would not only stress mastery of form, convention, and logical arrangement of the content of a written work, but would also stress competence in a sequenced writing process—prewriting, writing, and rewriting. Students might be measured against explicit, pre-set standards of each step in the writing process on, perhaps, a rubric or guide of writing protocols. Instructors may include brainstorming material and first drafts to be turned in along with final drafts in order to assess a student’s writing process.

2.2 The expressionist approach

An expressionist approach draws from a Platonic epistemology that sees truth and reality as knowable only subjectively; that is, in order to know, the thinker must know him- or herself. Since it is not possible to know objectively, knowledge is therefore not communicable from person to person. Berlin (1988) notes that the psychology of Rousseau, which considers the individual as inherently good, and the literary movement of romanticism, are major resources for expressionist thought, which recoils from “the urban horrors created by nineteenth century capitalism,” while promoting nature, emotionalism, and the imagination as preeminent over rationality and modern civilization (Berlin 1988: 484). He traces the roots of expressionist rhetoric to a reaction against the “elitist rhetoric of liberal culture” whose scheme argued “for writing as a gift of genius, an art accessible only to a few” (484). In expressionist rhetoric, writing is an inherent ability of all people. This proposition is at the heart of Elbow’s book of writing theory, entitled “Everyone can write” (2000). Berlin notes that expressionists like Elbow would view writing as a creative, artistic “paradigmatic instance” of “locating the individual’s authentic nature” or discovering one’s true self, and drawing upon the “reality of the material, the social, and the linguistic […] insofar as they serve the needs of the individual” (Berlin 1988: 484).

Expressionism has been no stranger to writing pedagogy. Berlin (1988) observes that an expressionist version of rhetoric “continues to thrive in high schools and at a number of colleges and universities” and is “openly opposed to establishment practices” that are hegemonic and homogenizing (1988: 487). In writing pedagogy with an expressionist bent, instructors might lay down a few guiding writing conventions, but largely get out of the way while students express themselves in their own unique creative processes. Teachers might encourage students to find their true, inner selves through writing. Indeed, the writing process is framed as a journey into the self in order to unearth one’s position on topics and to express one’s own self-aspirations. Convention and form, as well as error analysis, might well be secondary concerns in a more expressionistic writing class.

2.3 The social-epistemic approach

Social-Epistemic rhetoric holds that the material world, knowledge, communities, and even selves are all socially constructed. A wide range of scholars, such as Bruffee, Faigley, Trimbur and Bizzel are associated with social-epistemic rhetoric (or social constructionism). Among social-constructionists, Berlin points out, there are “as many conflicts […] as there are harmonies [but that they] share a notion of rhetoric as a political act involving a dialectical interaction engaging the material, the social, and the individual writer, with language as the agency of mediation” (1988: 488). Knowledge is a social construct, a product of a dialectical interaction between the individual, the discourse community, and the material conditions the individual exists in, which implies that knowledge and (versions of) reality are continually negotiable. Language constructs/distributes versions of external reality but is itself constructed by the language-using “observer, the discourse community, and the material conditions of existence [which are also] verbal constructs” (Berlin 1988: 488). Even, the individual (subject) is constructed socially, linguistically engendered by the linguistically-confined interplay between the “individual, the community, and the material world” implying that there are no coherent, unique individuals, but rather, all selves are products of historical and cultural moments and—although individuals have independent agency—their freedom is limited insofar as the ways of conformity and dissent are pre-circumscribed by socially constructed categories (Berlin 1988: 489). For social-constructionists, therefore, experience is prescribed by “socially-devised definitions [and] by the community in which the subject lives”—resulting in an all-enveloping “hermeneutic circle” that limits the possibilities of how we can even interpret the world so that no perception is free from self-interest and ideology (Berlin 1988: 489).

Naturally, social-epistemic rhetoric would be the most difficult to implement in the classroom because it “attempts to place the question of ideology at the center of the teaching of writing,” offering “both a detailed analysis of dehumanizing social experience and a self-critical and overtly historicized alternative based on democratic practices in the economic, social, political, and cultural spheres” (Berlin 1989: 491). In social-epistemic rhetoric, students “must be taught to identify the ways in which control over their own lives has been denied them, and denied in such a way that they have blamed themselves for their powerlessness” (Berlin 1989: 490). In such a classroom, students would be encouraged to engage in critical discourse about the established social structure and to explore the ways in which hegemonic powers has imposed itself on them and has shaped their identity and mindsets. Instructors would engage in liberator pedagogy, to invoke Frère, helping students free themselves from the socially-sanctioned ideologies that seek to imprison and atomize each of them in a pre-shaped epistemology that is aimed at sustaining its economic-industrial interests.

2.4 Multiculturalism, the contact zone, and Bakhtin’s heterogeneity of voices

As Edelstein (2005) suggests, social-constructionism has also led to a revolution in literary studies called multicultural studies—a set of critical theories that rethinks the “Other,” the marginalized, and the oppressed and that serves as a critique of assimilationist discourse. This literary movement (with its off-branching Postcolonial studies) came into being largely through the political movements of the 1950s and the subsequent civil rights movements of the late 1960s and early 70s which called for the end of the Vietnam War and the inclusion of Ethnic and Women’s studies in universities (Edelstein 2005: 19). In literary studies, multiculturalism promotes “an awareness of the otherness of the self to itself […] that we are all both someone’s other and ‘strangers to ourselves,’” an awareness that can “positively transform our relations to ‘others’” (Kristeva 1991, in Edelstein 2005: 35). A theory that seems to have had great influence in composition studies, and that emerged form multiculturalism, is Pratt’s (1991) notion of the contact zone.

The concept of a contact zone theoretically captures instances of contact between and among selves, communities, and cultures—and the conflicting values, assumptions, and ideology that language users bring to a text. Pratt (1991) coined the term contact zone in thinking about the “social spaces where cultures meet, clash, and grapple with each other, often in contexts of highly asymmetrical relations of power, such as colonialism, slavery, or their aftermaths as they are lived out in many parts of the world today” (1991: 33-40). Writing that comes out of contact zones, according to Pratt, can be both perilous, and artistic (1991: 4). In contact situations where one ideology is imposed on another, the perils of writing in the contact zone could be “miscomprehension, incomprehension, dead letters, unread masterpieces, [and] absolute heterogeneity of meaning,” whereas in situations where different cultures grapple but are equal in power, artistic production can result, such as “Autoethnography, transculturation, critique, collaboration, bilingualism, mediation, parody, denunciation, imaginary dialogue, [and] vernacular expression ” (Pratt 1991: 4).

The concept of the contact zone is also extendible to the individual consciousness, where the very self is the site of conflict. Multiculturalism’s revision of the other as being located within oneself means that the self can be a stranger to oneself. A person’s inner life could be seen as a contact zone between competing selves—perhaps a struggle between the three categories of selves from identity studies: the interactional, relational, and communal selves—or between nationalized or imagined selves. Following this view, the inner self becomes, in essence, “a heterogeneity of voices,” illuminating Bakhtin (1981) and Vygotsky’s (1962) views of the dialogic nature of the mind. Bakhtin (1981) holds that language in the speaker-writer’s mind is a polyphony of competing voices so that for “any individual consciousness living in it […] language is not an abstract system of normative forms but rather […] language for the speaker exists in the form of particular utterances from particular others”—that is, inner verbal thought is constituted by a “plurality of languages” (1981: 293, in Trimbur 1987: 219). Language is not “a unified abstraction [but] an ideological field, a social horizon, a struggle of voices” (1987: 219). For Bakhtin, “all words have the ‘taste’ of a profession, a genre, a tendency, a party, a particular work, a particular person, a generation, an age group, the day and hour” and each word “tastes of the context and contexts in which it has lived its socially charged life” (1981: 293, in Trimbur 1987: 219). Our choice of words in a composition, therefore, are never neutral in force, but are laden with meanings and associations that our inner voice ascribes to them and are, as Bakhtin says, “permeated with intentions” that carry the residual “aspirations and evaluations” of the others who spoke them (Todorov 1984: 202; cited in Trimbur 1987: 219). A single, unified authentic voice becomes so difficult to identify, that writing entails searching for the writer “within the social network these voices compose” and negotiating the “conflicting claims these voices make in the writer’s inner speech” (1987: 219). Producing written discourse, therefore, is not an individual effort, but an appropriation of all of the past and present discourses that make up a writer’s consciousness, and engaging with and responding to that world of discourse (Bakhtin 1981: 354-55).

Trimbur makes an application along these lines to composition, pointing out that “the language of inner speech condenses and internalizes the language of our conversations and relations with others” so that “the outer world of public discourse has already entered as a constitutive element into the inner world of verbal thought” (Trimbur 1987: 215). Verbal thought that generates writing is internalized social experience and constitutes “the language of inner speech” which is “saturated with sense [and] the sum of all the psychological events aroused in our consciousness by the word” (Vygotsky 1962: 146, in Trimbur 1987: 217). The thoughts that eventually get transferred to paper, therefore, are others’ voices that we have gleaned from social experience that coalesce and make up what seems to us as our own inner voice so that “the words we think of and experience as private thought are in fact constituted through the voices of others that echo in our verbal thought” (Trimbur, 1987: 217). Language is, therefore, already and simultaneously inside (the imagination) and outside its users, “so that there can be no distinction between private thought and public discourse” (1987: 219). Trimbur concludes that the goal of the writing process should no longer be considered merely as producing discourse (as cognitivists might say), for discourse already exists as the writer’s inner verbal thought. Internalized discourse is the raw material, as it were, out of which writers must “forge a voice and way of speaking from the conflicting social forces and the polyphony of voices that converge in their mental experience [and] individuate their verbal thought and expression” (1987: 220).

2.5 The context of written discourse: The rhetorical situation

Now that perspectives from composition and literary studies have been outlined, relevant background material on the situation or context of (written) discourse will be provided. The rhetorical situation is assumed to be the gene pool, as it were, for the written composition that grows out of it. It seems useful to think in terms of a communicative situation (context) in which communicative functions (ends) are achieved by means of linguistic forms (means). The emphasis on communicative function I take is positioned in the assumption that traditional writing pedagogy has largely stripped student writing of its natural context by emphasizing conventions and forms aimed at fabricating written products, while leaving out clear emphases on the communicative functions of those forms. Therefore, I will outline what I take to be the context of written discourse, without entertaining debates that have taken place about whether written discourse is driven more by the rhetorical situation or whether linguistic forms themselves engender discourse (Flower & Hayes: 1981).

Composition theorists seem to largely agree that Bitzer’s (1968) definition of “the rhetorical situation” is an appropriate departure point for discussion of the context of discourse. Bitzer’s definition of the rhetorical situation as the necessary and sufficient condition for discourse has been widely discussed in rhetoric and composition studies (Gorrell 1997: 395).  Bitzer wanted to define and characterize “the nature of those contexts in which speakers or writers create rhetorical discourse,” and elaborate on its components (1968: 1). It was his view that discourse does not give existence to its situation, but rather, that “it is the situation which calls discourse into existence,” discourse being a response to the situation (1968: 2). “Rhetorical situation” says Bitzer, is “the very ground of rhetorical activity,” “a natural context of persons, events, objects, relations, and an exigence which strongly invites utterance” and which amounts to “an imperative stimulus” (1968: 5). The “three constituents of any rhetorical situation” that Bitzer posited, are 1) an exigence, “an imperfection marked by urgency”—say, a problem needing addressing—that motivates discourse, 2) an audience, “persons who function as the mediators of [positive] change,” real or imagined, and 3) a set of constraints, “made up of persons, events, objects, and relations which are parts of the situation because they have the power to constrain decision and action needed  to modify the exigence,” say, for instance, the speaker-writer’s character, logical argument, and style (1968: 6-8).

My own model of the rhetorical situation takes Bitzer’s configuration as a departure point, and adopts some of the modifications that Grant-Davie (1997) makes, excluding “constraints” which “are the hardest of the rhetorical situation components to define neatly because they can include so many different things” (1997: 272). For me, therefore, the rhetorical situation includes 1) exigence, the motivating factor(s) of the discourse, 2) hearer/audience, imagined or real, intended or unintended receptors of the discourse, and 3) speaker-writer, the person(s) who, with authorial voice, presents the discourse to the audience in response to the exigence. I will describe each of these components in more detail, following Grant-Davie and, where appropriate, substantiating each component with perspectives from influential composition theorists.

The student writing guide (SWG) must help students see the significance of the context and situation that drives and motivates written discourse before plunging them into analysis and production of the linguistic features of a well-formed text. That is, writing needs to be re-characterized as communicative functions (ends) deployed by the use of linguistic forms (means). It is no longer sufficient to speak only in terms of the written product and the linguistic forms that comprise it. Much less is it appropriate to reduce the writing process to an act that merely meets the goals that a writing instructor prescribes. Rather, the act of producing written discourse is reconsidered as a reflex, a natural outflow of a writing situation. This means that a large part of students’ writing process involves re-imagining the writing situation of their text-in-process—not as an assignment that must be written for a teacher—but as a(n) (imagined) set of discourse moves that meets the ends of an imagined writing situation comprised of the exigence, the audience, and the voice of the speaker-writer.

1. Exigence: The occasion and motivation for speaking/writing

Exigence is “the matter and motivation of the discourse,” the speaker-writer’s sense “that a situation both calls for discourse and might be resolved by discourse” (Grant-Davie, 1997: 266). According to Grant-Davie, the exigence of a discourse is addressed by considering 1) what the discourse is about, 2) why the discourse is needed, and 3) what the discourse should accomplish. These three considerations represent, for Grant-Davie, a more comprehensive undertaking of discourse exigence than by solely asking why the discourse is needed, as Bitzer (1968) does.

a. What the discourse is about

Going beyond simply asking what the topic of the discourse is, Grant-Davie advocates a more comprehensive frame by asking what underlying fundamental issues the topic of the discourse represents—what “larger issues, values, or principles” are involved or at stake that “motivate people and can be invoked to lead audiences in certain directions on more specific topics” (267). Therefore, the exigence of discourse involves what more fundamental issues are at the heart of the specific topic being discussed—what implications the topic being discussed has for larger, fundamental issues, sets of social values, and so on. In other words, the exigence of discourse involves considering what deeper, underlying issue does the obvious, superficial topic instantiate.

b. Why the discourse is needed

Exigence also considers why the discourse is needed, what prompted the discourse at the time it was written/spoken; what caused it and why at that time, which evokes the notion of kairos, the optimal time-conditions or finest hour to speak/write. The optimal time conditions and the causes that prompt discourse give it exigence. A letter, for example, may be a reply to another letter or a response to a recent event. The question of the value of the discourse, or of why the speaking/writing matters, is also involved in exigence. In other words, how important are the issues and why do they need urgent addressing? The issues may be “intrinsically important, perhaps for moral reasons,” or the issues’ significance may lie in the implications they have for the situation, in what potential consequences may transpire if action is not taken (1997: 268).

c. What the discourse is trying to accomplish

The exigence of discourse also involves the goals or outcomes of the speaking/writing, e.g, how the audience should react to the discourse. Grant-Davie includes “objectives as part of the exigence for a discourse because resolving the exigence provides powerful motivation for the” speaker-writer (1997: 269). The objectives of a text may be aimed at persuading an audience about one of the many topics or components a larger issue may have, so that discourse can have multiple, hierarchical, objectives at once. Grant-Davie gives an example of a presidential candidate’s speech aimed (immediately) at rebutting the accusations of an opponent, while aimed (ultimately) at persuading an audience to cast votes in the candidate’s favor (1997: 269).

d. They say/ I say

A writer’s exigence provides an urgency, purpose and occasion for the discourse, as Graff and Birkenstein (2006) continually emphasize in their academic writing guide, They Say/ I Say: Moves that Matter in Academic Writing. In fact, Graff and Birkenstein base their entire tutorial of academic writing on the controlling idea that successful writers “state [their] ideas as a response to others,” that a statement/response dynamic is the core feature of good academic writing, just as “in the real world we don’t make arguments without being provoked.” (3). In saying this, they implicitly pick up on the importance of considering the exigence of the rhetorical situation, insisting that “what you are saying may be clear to your audience, but why you are saying it won’t be [because] it is what others are saying and thinking that motivates our writing and gives it reason for being” (Graff and Birkenstein, 2006: 4). What Graff and Birkenstein (2006) seem to be promoting, in essence, is a re-contextualization of written discourse as an outflow of the larger rhetorical situation. They highlight the fact that written discourse must not be produced as if it were merely a bundle of written form stripped of its context, but rather that, “the best academic writing has one underlying feature: it is deeply engaged in some way with other people’s views” and does not consist of “saying ‘true’ or ‘smart’ things in a vacuum, as if it were possible to argue effectively without being in conversation with someone else” (3).

2. Audience: Who are you speaking-writing to?

After the exigence for writing is firmly set in place, student writers should consider an audience to whom the exigence of the discourse is relevant. The audience of a rhetorical situation is comprised of real or imagined persons “with whom [speaker-writers] negotiate through discourse to achieve the rhetorical objectives” (Grant-Davie, 1997: 270). Grant-Davie takes audience to mean four different sets of people: 1) any real people who happen incidentally to hear or read the discourse, 2) the real people that the discourse was intended to be read by, 3) the audience that the speaker-writer has in mind, and 4) the audience that the discourse itself suggests (Grant-Davie, 1997: 270). The roles of speaker-writers and hearer/hearers, however, “Are dynamic and interdependent” so that “readers can play a variety of roles during the act of reading a discourse,” roles that are “negotiated with the [speaker-writer] through the discourse” which “may change during the process of reading” (Grant-Davie, 1997: 271). Overall, the role(s) of speaker-writers and hearer/readers may shift and morph as the discourse, the speaker, and the audience exert mutual influence upon one another to create new roles and accept new identities.

a. Imagining an audience

Bartholomae (1985) also discusses the importance of students writing “initially with a reader in mind,” imagining a reader, anticipating possible responses (1985: 627). For the student writer, the challenge is to imagine a rhetorical situation and to make communicative moves within a context that (often) does not exist. The “difficulty of this act of imagination and the burden of such conformity” to the aims of the writing pedagogy is “so much at the heart of the problem” (1985: 627). In fictionalizing their audience, writers “have to anticipate and acknowledge the readers’ assumptions and biases,” and in doing so, hopefully to detect their own assumptions and biases (Bartholomae 1985: 628).

Among the writer’s most important objectives is to imagine an audience that hears his or her response to an exigence, and to try on a variety of voices/identities in response to this rhetorical situation. Ong (1975) holds that, rather than responding to actual people in writing, a skillful writer will “construct in his imagination, clearly or vaguely, an audience cast in some sort of role, entertainment seekers, reflective sharers of research” (1975: 60). Ong elaborates that:

“The problem is not simply what to say but also whom to say it to. Say? The student is not talking. He is writing. No one is listening. There is no feedback. Where does he find his “audience”? He has to make his readers up, fictionalize them” (Ong 1975: 59).

In case one wants to make the intuitive proposal that the writing instructor is a sufficient audience, Ong asserts, “there is no conceivable setting in which [the student writer] could imagine telling his teacher how he spent his summer vacation other than in writing this paper, so that writing for the teacher does not solve his problems but only restates them” (59). Students instead must fictionalize both their attending audiences and their responding voices.

To aid in the fictionalization of audience, Ong (1975) points out that writing students can draw from “what [a] book felt like, how the voice in it addressed its readers, how the narrator hinted to his readers that they were related to him and he to them [and can] pick up that voice and, with it, its audience” (1975: 59-60). His point is that successful writers owe much of their coherence and vividness to the fictionalization of their audiences that otherwise would be impossible to invoke without the model of “earlier writers who were fictionalizing in their imagination audiences they learned to know in still earlier writers, and so on back to the dawn of written narrative” (Ong 1975: 60). Implicit in this view is something that Berman (2008) has found in her research: that previous literary education has a direct influence on students’ ability to produce well-formed academic discourse (2008). The fact that previous literary education is a major factor in whether students successfully participate in the rhetorical situation (in this case conceiving of audiences) is connected with Bakhtin’s contention that characters in novels embody a multiplicity of voices that access the reader’s participation and, subsequently, readers internalize and re-deploy these voices for later use in discourse production. Authorial voice, therefore, is an outgrowth of audience.

b. Connecting with an audience using meta-discourse markers

Having an audience in mind is related to several communicative components (functions) of successful writing. Cheng & Steffensen (1996) provide seven criteria for successful writing, out of which the “text-centered standard of coherence, and the user-centered standards of acceptability,  situationality, and informativity” are “clearly related to concepts of audience” (1996: 150-151). Audience is something that writing students “rarely have a clear sense of”—and when students do involve a rhetorical audience, “it is a real person who gives a perceptible response – a teacher who provides a grade, not someone with whom to create a dialogue” (1996: 152).

The standard of coherence refers to “the knowledge that provides the conceptual undergirding of a text” which “must be accessible to both the writer and the reader” and calls to attention the fact that “no text is completely explicit” and that readers must “share enough back-ground knowledge to be able to make successful inferences and fill gaps” (Cheng & Steffensen 1996: 151). The standard of acceptability “capture[s] the fact that […] the reader must accept the text as cohesive, coherent, and directed toward” the writer’s goals (1996: 151). “Situationality,” another user-centered standard for successful writing, “refers to the match between a text and the context for which it is intended;” and finally “informativity, […] captures the fact that “writers must be able to anticipate the amount of information shared by their readers” (1996: 152).

According to Cheng & Steffensen (1996), teaching students the explicit use of discourse markers (forms) helps them connect with an imagined audience better,  organize their writing more,  and consider the epistemological value of their propositions more realistically (functions). Before, touching upon these findings, I want to define discourse markers.

There is little agreement in the literature about what to call the connective words/phrases that are strewn throughout written discourse, which many English as a Second Language and remedial writing publications call transition words. Fraser (1996, 2005) seems to provide a useful classification of what he calls pragmatic markers, of which he posits four types (Fraser 2005). Although “discourse markers” appear as only one of the four types of Fraser’s pragmatic markers, I have adapted two of Fraser’s types of pragmatic markers for my own classification of connecting words relevant to academic writing, labeling all of them “discourse markers”—even those which Fraser calls “commentary pragmatic markers” (2005: 1), under the five headings below, which integrates many of the transitions that Graff & Birkenstein (2006) posit as commonly used in academic writing (2006: 232-234). The reason for only including two types of Fraser’s pragmatic markers in my own classification of discourse markers is that the two other excluded types seem to be used in conversation (or in informal writing, such as tweets or emails), and not in academic composition, which is the focus of this study. Discourse markers that are useful in academic composition, therefore, are considered in this study (with subheadings and examples) to be:

1) Elaborative discourse markers:  signaling an addition or refinement of a preceding statement.

Examples: and, above all, also, alternatively, besides, equally, for example, for instance, further(more), in addition, in other words, in particular, likewise, more accurately, more importantly, more precisely, more to the point, moreover, on that basis, on top of it all, or, otherwise, rather, similarly, that is (to say)

2) Contrastive discourse markers: signaling a denial, contrast, or unexpected development of a preceding statement

Examples:but,alternatively, although, contrary to expectations, conversely, despite (this/that), even so, however, in spite of (this/that), in comparison (with this/that), in contrast (to this/that), instead (of this/that), nevertheless, nonetheless, (this/that point), on the other hand, on the contrary, regardless (of this/that), still, though, whereas, yet

3) Inferential discourse markers: signaling a conclusion from a preceding statement.

Examples:so, after all, as a conclusion, as a conse­quence (of this/that), as a result (of this/that), because (of this/that), conse­quently, for this/that reason, hence, it follows that, accordingly, in this/that/any case, then, therefore, thus

4) Temporal discourse markers: signaling a shift in time or progression related to the discourse.

Examples:then, after, as soon as, before, eventually, finally, first, in the meantime, meanwhile, originally, second, subsequently

5) Meta-discourse markers: signaling a writer’s commentary to an audience about his/her discourse.

a. Assessment markers: signal a writer’s judgment/evaluation of the wider context (state of affairs) of the statement.

Examples: fortunately, sadly, strikingly, surprisingly, amazingly, astonishingly, hopefully, ideally, importantly, incredibly, inevitably, ironically, justifiably, oddly, predictably, prudently, refreshingly, regrettably, (un)fortunately, (un)reasonably, (un)remarkably

b. Manner-of-speaking markers: signaling the way a writer conveys the statement.

Examples: frankly, bluntly, briefly, candidly, confidentially, crudely, fairly, frankly, generally, honestly, ironically, metaphorically, objectively, personally, precisely, roughly, seriously, simply, strictly, truthfully, to speak candidly, roughly speaking, to be honest, and in all seriousness, worded plainly, stated quite simply, off the record, quite frankly, in the strictest confidence

c. Evidential markers: signaling a writer’s level of confidence of the truth of the statement.

Examples: conceivably, certainly, indeed, assuredly, decidedly, definitely, doubtless, evidently, incontestably, incontrovertibly, indisputably, indubitably, (most/ quite/ very) likely, obviously, patently, perhaps, possibly, presumably, seemingly, supposedly, surely, (un)arguably, undeniably, undoubtedly

d. Hearsay markers: signaling a writer’s comment about the source of information of the statement.

Examples: reportedly, allegedly, I have heard, it appears, it has been claimed, it is claimed, it is reported, it is rumored, it is said, one hears, purportedly, they allege, they say

e. Topic change markers: signaling writer’s digression or departure from a current topic.

Examples: back to my original point, before I forget, by the way, incidentally, just to update you, on a different note, parenthetically, put another way, returning to my point, speaking of X, that reminds me

One may notice that the above discourse markers are from four syntactic categories with significant overlap in (discourse) function: 1) coordinate conjunctions (and, but, for, yet, so), 2) subordinate conjunctions (while, since, although), 3) preposition phrases (above all, rather than, in fact), and 4) adverb phrases (anyway, furthermore, still, however) (Fraser, 2005: 6).

In an empirical study, Cheng & Steffensen (1996) showed how freshman composition students’ writing improved when they were instructed to use meta-discourse markers, meta-discourse being a device that “comments on the text itself or which directs comments to the reader” (150). Cheng & Steffensen reported that students “experienced a breakthrough in their consideration of the reader and an awareness of the rhetorical functions of composition” when the students learned about the concept of meta-discourse and meta-discourse features in terms of their rhetorical functions in an environmental setting” (179). Specifically, students trained to use meta-discourse markers wrote with more explicit structure and tone, an improved topical progression, an improved revising processes, and more attention paid to a realistic epistemology (Cheng & Steffensen 1996: 171-176).

As regards improvement of structure and tone, Cheng & Steffensen pointed out that meta-discourse makes “structure and tone explicit, by capturing them on paper” enabling writers to “consider what they are saying more easily and make appropriate changes and improvements” and leading them to revise with increased “attention [paid] to the reader’s perspective” (170).

In paying more attention to meta-discourse, students became more aware of the epistemological value of their statements, realizing that “what they wrote was being understood by their readers as their view of reality, and they became more ethical writers, paying more attention to the truth value of their propositional content and hedging it when necessary” (176).

3. Speaker-writer: What are your roles, identities, voices, and attitudes?

Speaker-writers are involved in the rhetorical situation as those, real or imagined, who are “responsible for the discourse and its authorial voice” (Grant-Davie 1997: 269). Though Bitzer left the speaker (rhetor) out in his definition of the rhetorical situation, the speaker-writer is just as much a constituent of the rhetorical situation as the audience, Grant-Davies argues. The role(s) of the speaker-writer “are partly predetermined but usually open to definition or redefinition” so that speaker-writers “need to consider who they are in a particular situation and be aware that their identity may vary from situation to situation” (Grant-Davie, 1997: 269). A Speaker-writer may be the originator of the content of the discourse, or just a designee chosen to deliver it, or neither, yet held accountable for the validity of the content (269). Authorial identity may shift, even, from whom the speaker-writer considers himself to be, to whom the hearer/readers infer the author’s identity or ethos to be. So, determining whom the speaker-writer is may not be simple, as exigence and audience exert influence on the roles, the identities, and the audience’s perception of the speaker-writer’s ethos. Interestingly, “new rhetorical situations change us and can lead us to add new roles to our repertoire,” which calls for “receptivity—the ability to adapt to new situations and not rigidly play the same role in every one” (1997: 270).

a. Trying on different voices in writing

The concept of voice in writing is widely used, discussed, and debated. Yet, according to Peter Elbow, there is no “critical consensus about voice” and the concept “leads us into theoretical brambles” (2000: 227). Elbow (2000) discusses the different ways in which voice has been characterized by different literary and composition theorists, for which space in this paper is not committed, highlighting the potential difficulty the concept could present for its analysis. Nevertheless, Elbow holds that “there is little reason to question voice as a solid critical term” because it “points to important qualities in texts” (2000: 205). To put it briefly, Elbow thinks of voice in terms of 1) audible voice, concerned with hearing the text; 2) dramatic voice, seeing every text as having an implied character behind the words; 3) authoritative voice, aimed at giving a more authoritative stance to overly deferent student writing; and 4) resonant voice, the idea of the trustworthy, genuine writer sounding present, behind the words, evoking ethos (2000: 226-227). A concise list follows of heuristically motivated questions that writing teachers may use to help students think about each notion of voice (adapted from Elbow, 2000: 184-221, 226):

1) Audible voice

Are you able to hear this section? How much of it can you hear?

If you are stuck, can you read the section out loud and hear the problem?

Look away from the paper and speak the idea out loud. How would you say it?

2) Dramatic voice

What kind of voice(s) do you imagine are speaking in this essay?

– a timid voice, an angry voice, an arrogant voice, a bureaucratic voice, a professorial voice, multiple voices

What sort of speaker or character are you speaking as here?

3) Authoritative voice

Could you rephrase this part to sound more assertive?

Speak your mind more in this section. Are you afraid of offending someone?

How would you say this if you were (some other speaker) the president?

4) Resonant voice

How much of yourself did you get behind these words?

Where is the real you in this section? Does it fit you?

Could you rephrase this (part) so sound more like yourself?

Could you be more sincere in this section?

Though each of these views of ‘voice’ is useful, I will focus on Elbow’s resonant voice to frame my understanding of voice presented here, a concept of voice that depends upon the analogy of one getting more of oneself behind his or her words; more than being authentic, but more so “sounding again” (2000: 208). The concept of voice, therefore, “points to the relationship between discourse and the unconscious” and “comes from getting more of ourselves behind the words,” getting “more of our unconscious into our discourse” (2000: 206, 207). Voices are tightly bound up with selves and identities, which tend to “evolve, change, take on new voices, and assimilate them (2000: 208). In a typically expressionistic way, Elbow holds that writing provides a crucial setting for “trying out parts of the self or unconscious that have been hidden or neglected or undeveloped—to experiment and try out ‘new subject positions’” (2000: 208).

Student writers need to explore and appropriate different voices that approximate the identities that fit within the rhetorical situation. The process, of approximating and re-approximating one’s own multiple, shifting, and fluid voices in writing is shaped by the also-dynamic process of audience fictionalization. It is this intentional, meta-awareness of one’s voice and audience that characterizes mature and skillful writers—the ability to deploy different voices to meet different communicative ends.

Bartholomae (1985) reflected that whenever a “student sits down to write, he has to invent the university for the occasion” by trying on different voices and interpretive frameworks. “The student has to appropriate a specialized discourse,” Bartholomae says, “and he has to do this as though he were easily and comfortably one with his audience” (624).

2.6 Berman and the development of discourse

From the perspective of 1^st Language Acquisition, Psycholinguistics, and Form/Function linguistics, Berman and her international team investigated how children develop the ability to construct discourse in French, Hebrew, Icelandic, Dutch, Spanish, Swedish, and American English in a large-scale project, known as the Spencer Project, or officially as “Developing literacy in different contexts and different languages,” which began in 1997[1].

The project involved the elicitation of written/spoken personal experience narratives and expository texts from 80 individuals from the seven countries represented by her team, evenly divided into four groups of 20 individuals each: Grade School (4^th graders, 9-10 years old), Junior High School (7^th graders, 12-13 years old), High School (11^th graders, 16-17 years old), and Adults (university graduate students in their 20’s and 30’s). As Berman and her team thoroughly analyzed the 320-text corpus for patterns of development across the independent variables of age-level, genre/modality, and language, dozens of studies were published on the different findings that emerged from the analysis. Berman and her colleagues wanted to describe “the linguistic, cognitive, and communicative resources that [speaker-writers] deploy in adapting their texts to different circumstances”—across modality and genre—“and to detect shared or different trends depending on the particular target language” (Berman & Katzenberger 2004: 62). These aims are summarized under the heading of measuring speaker-writer’s development of linguistic literacy. First outlined by Ravid & Tolchinsky (2002), the notion of “linguistic literacy” (not merely the “literacy,” of reading/writing) is characterized by speaker-writers possessing a varied repertoire of linguistic resources, the ability to access/deploy these resources to meet different communicative ends in different contexts, and a meta-linguistic awareness of the effect of using these different linguistic forms in different situations. Linguistic literacy develops toward its zenith of “rhetorical flexibility or adaptability,” which means being “rhetorically expressive” enough to “hold the attention of their addressee” with “interesting and varied linguistic output that is attuned to different addressees and communicative contexts” (Ravid & Tolchinsky 2002: 418). Key to the idea of rhetorical flexibility is the ability to intermingle features from different genres (narrative and expository) and to adapt these features to be as “rhetorically more powerful, convincing, and precise” as possible (2002: 435). Berman evokes and reiterates the same idea of linguistic literacy as the developmental target in the first Spencer Project article, referring to it as “the ability to readily and effectively access and deploy a wide range of written materials in a given target language” (Berman & Verhoeven 2002: 14).

To measure speaker-writer’s developmental stages of the achievement of linguistic literacy, or the production of well-formed discourse, Berman eventually developed a complex analytical framework made up of the different aspects of discourse production abilities (Berman 2008). Berman defined five aspects of discourse construction that were quantifiable across language, age, and genre/modality and which younger/older speaker-writers were found to deploy with gradual mastery at different stages of development. The five “functionally motivated” discourse features comprise an “integrative framework,” that proceeds from (1) global-level principles via (2) categories of referential content, to (3) clause-linking complex syntax, and (4) local linguistic expression, and ending with (5) overall discourse stance; these aspects “apply concurrently whenever a piece of discourse is produced” and are applied “in order to meet the challenge of connecting cognitive representation and linguistic knowledge in developing text construction” (Berman 2008: 737-740). Berman measured these five discourse features across the variables of genre (narrative/expository), modality (written/spoken), age, and language. These five facets of discourse production comprise an innovative, complex, and empirically grounded model for characterizing written and spoken discourse at the lowest and highest stages of development.

2.7 Discourse stance, referential content, and global text organization

Of these five facets, I will review three—discourse stance, categories of referential content, and global text organization—which directly pertain to the aims of the student writing guide, namely to help student-writers deploy forms of well-formed expository, argumentative discourse. The two excluded facets, clause-combining complex syntax and local linguistic expression were found to have low, even negative, correlations with overall global text quality. Some speaker-writers who used advanced vocabulary/syntax were found to construct “relatively flat or juvenile texts” while other speaker-writers who used simple forms of local-linguistic expression produced (globally) well-formed texts (Berman 2008: 757). Some of Berman’s subjects who scored high in local linguistic expression scored low on global text quality and vice versa. This means that lexico-syntactic complexity does not translate directly into well-organized global discourse structure, clarity of thinking, or interesting content. The use of advanced forms of lexical/syntactic expression is neither sufficient nor necessary for producing a well-formed or even beyond-well-formed text. Therefore, I will only review and rely on the facets of 1) global text organization, 2) categories of referential content, and 3) discourse stance to inform and help construct the student writing guide.

1. Global text organization: Top-down (expository)/ bottom-up (narrative)

The first facet of discourse text production, global text organization, brings into consideration the overall “structure and content of the text” and its “rhetorical effect” as based on type of genre (2008 p.740-741). First discussed in Berman & Nir-Sagiv (2007), global text organization depends on the notion of “mental representations of schema and category […] anchored in the shared cognitive ability to interrelate parts and wholes” (2007: 91). Speaker-writers demonstrate well-formed global text quality when they produce narratives that organize isolated events (bottom-up) within an action structure schema (top-down). Expository texts, on the other hand, are centered around one core proposition (top-down) that is elaborated upon by specific instances and details (bottom-up) (2008: 741). “Structural well-formedness” in the domain of global text quality is achieved when speaker-writers integrate “top-down and bottom-up principles of discourse organization” (Berman 2008: 741).

As different genres instantiate different cognitive representations and organizational principles, a text is considered well-formed that meets the genre-canonic global organizational structure. Younger speaker-writers typically rely only on one organizational schema: either on top-down for expository texts, or on bottom-up for narratives. Yet, a higher, more sophisticated level of text construction ability is demonstrated by the synthesis both of top-down and bottom-up organizational schemas so that a discourse contains both narrative- and expository-like organizational structure weaved together into “an ‘integrated re-representation’ of top-down and bottom-up processing” (2008: 741). In a study that led up to the notion of global text quality, Berman & Nir-Sagiv (2004) show that more mature speaker-writers are able to intersperse narrative elements in expository texts and to include generalized propositional material within narratives, which creates a more effective text.

It is the integration of genre-specific features that marks a speaker-writer’s text as “effective” and being beyond well-formed—a departure from genre-typical structure and content and the inclusion, for example, of bits of narrative-like event reference into an expository discourse or vice versa. Effective (global) text construction in expository texts involves the “explicit marking of logical relations between discourse segments or explicit meta-textual commentary,” the cognitive abilities of critical thinking and creative thinking—elaborating on a topic with reference to personal experience, and organizing information in new ways (respectively) (2008: 742).

a. Measuring global text quality

Based on the ideas of global text quality, the cognitive representation of narrative/expository, and the varying levels of well-formedness and effectiveness, Berman devised four levels of text construction abilities (Berman & Nir-Sagiv 2007). The levels flow along three dimensions, “Defining rank, Criterial Properties and Characteristic Features” (2008: 742). These dimensions, elaborated in detail in Berman & Nir-Sagiv (2007: 97-99) are analogous to “Representation and cognitive processing,” “Structure and content,” and “Discursive features.” The four (sequential, standard-driven) levels are 1) minimal representation (level I), 2) partial extension (level II), 3) structural well-formedness (level III), and 3) beyond well-formedness (level IV). This evaluative system is based on “rigorously defined measures of vocabulary and grammar” and is a more theoretically sound and explanatorily powerful measure of text quality than the largely subjective evaluative criteria used for determining the overall quality of a discourse (spoken/written) that is characteristic of official educational ratings (Berman & Nir-Sagiv 2007: 103, Berman 2008: 741-743).

Speaker-writers at level I show limited representation and cognitive processing abilities, as their texts stick to either bottom-up or top-down principles of discourse organization only and contain only basic components. The discursive features observed at this level are comprised of detached units, that is, isolated, unanchored events  arranged chronologically in narratives and uninvolved generalization using habitual present tense or generic future) in expository texts (Berman & Nir-Sagiv 2007: 97).

At level II, a speaker-writer’s representation and cognitive processing shows partial extension, with initial integration of bottom-up and top down organization becoming evident in their texts. In expository texts, speaker-writers at this level add further information beyond top-down generalities, while still maintaining prescriptive attitudes on topics. In terms of structure and content, texts show an initial reliance on genre-typical features. The discursive features of texts at this stage show initial anchoring so that, in narratives, speaker-writers might begin anchoring their past events to generalized states of affairs and, in expository texts, reference to causes or solutions and to past events may be demonstrated (Berman & Nir-Sagiv 2007: 98).

Speaker-writers make a developmental leap in level III, which is characterized by their texts demonstrating a full integration of top-down and bottom-up discourse organization principles in the dimension of representation and cognitive processing. Expository texts show “a top-down superordinate topic of conflict elaborated by specific bottom-up categories,” but the problem remains that such generalizations are “unanchored in specific references to past events, nor are there meta-cognitive or meta-textual comments to guide the reader” (2008: 744). The structure and content of texts at this stage is overt, yet still showing features that are schematic or categorical to the genre of the text. A typical level III discursive feature is when speaker-writers make a relation between the opening and ending, the introduction and conclusion, of the text. Narratives at this stage are genre-typical, showing schematic organization within a top-down action structure. Also, the first signs of evaluative interpretation of the events begin to emerge in level III. In expository texts, there begins to surface more alternation between the general and the specific and the inclusion of “culturally shared knowledge or […] past events illustrating the topic” as well as a clear conclusion (Berman & Nir-Sagiv 2007: 98).

At the most advanced stage of text construction abilities, level IV, the representation and cognitive processing of speaker-writers allows for the creative synthesis of parts into a whole. The structure and content dimension of the global text quality sees the inclusion of genre-external material: speaker-writers shifting from one discourse stance to another and alternate between subjective and objective stances toward reality (2008: 745). Also, meta-cognitive, inter-textual, and/or meta-textual commentary is characteristic of this stage’s discursive features involving the explicit marking of text segmentation with discourse markers. Here, narratives show more atemporal generalization, shifting from personal instances to detached commentary, with the inclusion of entire passages of evaluative/interpretive commentary. Expository texts show the alternation between “detached generic and subjective personalized discourse stance,” and integrate personal experiences and analytical, critical thinking, or even creatively synthesizing varying inferences/conclusions (Berman & Nir-Sagiv, 2007: 99)

b. Development of global-level content and structure

There is strong empirical support for the criteria Berman proposed for evaluating global text quality (Berman 2008: 745, Berman & Nir-Sagiv 2007). In middle childhood, speaker-writers show the presence of a narrative schema as most nine- to ten-year-old subjects reached beyond level I and all older subjects reached level IV. Yet, for expository texts, only adolescents demonstrated a grasp of the principles underlying global-level expository text construction; plus, there was a generally lower score across all age groups for expository than narrative global text quality (2008: 745). Berman & Verhoeven (2002) found differences specifically in the (English) linguistic forms used in narratives versus expository texts, including active/passive voice, verb structure, subject noun phrases, and modals (2008: 746). This rigid dichotomy between the linguistic devices used in the different genres was maintained until high school-age, so that only adolescents and adults included linguistic elements that were less genre-typical, such as the habitual present in narratives and specific past event reference in expository texts. The mixing of these features within narratives was observed only at 7^th grade and up, and even these were confined to story-linked evaluative codas. Yet, as age increased, there was a shift from dichotomy to divergence, as story-external generalizations increasingly appeared in places other than the openings and closings of the narratives (2008: 746). Discourse texture diversification increases with age, which is attributed to “general socio-cognitive development, on the one hand, combined with increased exposure to and experience with a range of different types of discourse, on the other” (Berman & Nir-Sagiv 2004: 375). Still, in expository texts, this diversification is much less prevalent and takes place later on than feature-diversification within narratives does. The global text qualities of well-formed narratives are achieved earlier than those of well-formed expository texts.

2. Categories of referential content

The second aspect of discourse production abilities in Berman’s analytical framework is a system of categories for the referential content within texts. After conducting an analysis of what features constitute informative vs. non-informative material in texts (Berman & Ravid 2008: 3), Berman pinpointed three categories of information units in narratives: Eventives (obligatory core eventive content), Descriptives, and Interperatives (elaborative) elements (2008: 747). These different types of information units, or “infus” were taken as the basic unit of text-informational analysis in Berman & Ravid (2008). With these topic-neutral, genre-dependent categories of referential content thus defined, Berman was able to “specify the ‘information density’ of a text” (2008: 747). Although the three information units described below were elaborated in a study on analyzing narrative information density (Berman & Ravid 2008), they are also relevant to expository text information units, in slightly different ways, which will be elaborated.

Obligatory core eventive elements function as plot advancers (in narratives) and as generalized propositions (in expository texts) that carry forth claims or arguments and that anchor descriptive/interpretive information units in a text. Berman & Ravid consider core obligatory eventive elements as analogous to Labov’s narrative (referential) clauses that encode sequential events (2008: 3).

Descriptive elements describe the eventive elements, specifying the physical characteristics of entities or states of affairs involved. They orient and anchor the core eventive elements, providing relevant background information for eventives. Berman & Ravid (2008) admit that delineating distinctions between eventive and descriptive infus is difficult and involves discussing the linguistic features and discourse functions involved in any given discourse (Berman & Ravid 2008: 5). For example, the question was raised as to whether background-setting events should be considered as eventives, descriptives, or as a subclass of eventives (2008: 5). Speech act verbs are considered to be eventives, while predicates of speech act verbs are counted as descriptive elements, findings that are rigorously attested to and substantiated by data (2008: 4-10).

Interpretive elements encode speaker-writers’ subjective, personal attitudes, judgments, and evaluative commentary about the core eventives or the descriptive background material. Interpretives are analogous to elements of narrative evaluation and are sometimes found in the “setting and coda of narratives” (Berman & Ravid 2008: 7).

a. Development of categories of referential content

Ravid & Berman (2006) found that, with age, speaker-writers show the inclusion of different types of information in the texts they construct. As the age of speaker-writers increased, interpretives in narratives increased and “the proportion of eventives decreased with age:” a shift that became more prevalent in the subjects in high school and beyond (2006: 132). Core generalized propositions (expressing socially-typical, prescriptive attitudes toward topics) in expository texts gradually became replaced by more individualized, cognitively influenced epistemic attitudes (2008:747).

3. Discourse stance: orientation, attitude, generality

Discourse stance is defined as “how speaker-writers use language to position themselves with respect to a piece of discourse under given circumstances” and instantiates the fact that “any state of affairs in the world can be described in multiple ways” and that “there is no ‘one way’ of talking or writing about a given topic, or about the same situation in the external world” (Berman & Verhoeven 2002, Berman 2005: 109). A large body of research is subsumed by the idea of discourse stance, namely, sociolinguistic narrative and conversational interaction analysis, psycholinguistic conversation analysis, studies comparing written vs. spoken discourse, and studies on children’s discourse development (Berman et al. 2002: 256).

Discourse stance is an “overall communicative framework shaping the course of text construction” and is expressed by speaker-writers via “a broad array of linguistic devices” (Berman 2002: 258, Berman 2008: 760). From a form/function perspective (which characterizes language as being an inventory of linguistic forms that speakers deploy to achieve certain communicative functions), Berman (2005: 107-110) outlined the three dimensions of discourse stance as being 1) orientation (either to Sender, Text, or Recipient), 2) attitude (either epistemic, deontic, or affective), and 3) generality of reference/quantification (how specific or general)—all three of which interact with one another and “can be alternated within a piece of discourse” (Berman et al. 2002: 258).

The speaker-writer’s orientation in a text is can be directed toward 1) sender (speaker-writer), 2) text (narrative/expository), or 3) recipient (hearer/reader). Discourse that is “deontically judgmental or affective in attitude, and specific in reference” and that has a subjective, (deictically) speaker-centered tone is sender-oriented. Recipient-oriented discourse seems to be addressing hearers/readers directly and is more “communicatively motivated,” evident by the use of discourse markers and use of the second person where a choice of other forms could have been made. Finally, discourse that takes a “conceptual or cognitive point of reference” toward the discourse topic and is characterized by “a totally distanced, impersonal and meta-textual level of orientation” is text-oriented (Berman et al. 2002: 259). The attitude of a text can be characterized as affective, prescriptive, or epistemic. Discourse expressing an epistemic attitude is characterized by the speaker-writer discussing a proposition “in terms of possibility, certainty, or the evidence for the individual’s belief that a given state of affairs is true (or false)” (Berman et al. 2002: 260). Discourse expressing a deontic attitude is marked by prescriptivism, a judgmental or evaluative stance toward the topic. Finally, discourse that expresses an affective attitude is demonstrated by the speaker-writer giving voice to his or her emotions with respect to a state of affairs. Each of these attitudes marks points along a continuum of objective-to-subjective stances, beginning with an epistemic, universal and objective attitude, through a socially sanctioned deontic attitude (by the speaker’s speech community), to a more subjective set of “reactions and personal feelings that an individual holds in relation to a given topic” (Berman et al. 2002: 260). In terms of generality of reference, speaker-writers can display varying degrees of generality and/or specificity with regard to the entities and states of affairs in their texts: specific, generic, and impersonal. Later, Berman enriched the dimension of generality of reference with the following distinctions: deictic or anaphoric, specific or generic, or concrete or abstract reference (2008: 758).

Overall, the three dimensions involved in discourse stance make up “a continuum of rhetorical means for moving from the personal to the general, from concrete to abstract, from specific to general, from immediate to distanced [and] from involved to detached (Berman et al. 2002: 263). These aspects of discourse stance (functions) are expressed via various linguistic constructions (forms), such as different kinds of pronominal reference, passive voice, and nominalizations for “abstract, distanced rendering of predicating content” (2008: 758). The discourse stance of a given text, for instance, can be more affective in attitude when interpretives are added to the descriptives of the state of affairs/propositions within the text (2008: 759). An expository text could display less-specific, categorical nominals, generic pronouns, and/or deontic/epistemic modals, which would indicate a more irrealis attitude or one that is cognitively distanced and epistemic (2008: 759). Such an impersonal, detached discourse stance is also often signaled by atemporal generic present statements or irrealis future-oriented contingencies, which is often in stark contrast to the subjectively-involved stance of (oral) personal experience narratives (2008: 760). Linguistic devices working in conjunction form these contrasts of discourse stance: 1) nominal reference (personal pronouns/concrete nouns vs. generic pronouns/abstract nominals), 2) temporal reference (specific past events vs. atemporal, detached generalizations, and 3) agent-orientation (passive vs. active voice). Speaker-writers begin to show knowledge of the use of these linguistic resources for the different discourse stances required by different genres only from high-school age and up (2008: 760).

a. Developmental findings for discourse stance

            Berman et al. (2002) made several tentative conclusions about the developmental trends of discourse stance. Distinctions in discourse stance were seen to be carried forth by a range of different, converging, linguistic forms, which in linguistic studies are often analyzed in isolation (Berman et al. 2002: 273, Berman 2004: 110). These linguistic devices, however, interact with one another to express (varying) discourse stance(s), and are considered to be a “confluence of [linguistic] cues” (Berman et al. 2002: 273).

With respect to age and schooling, Berman concluded that just as “the youngest (9-10 year old, grade-school level) children distinguish clearly between narrative and expository genre in both linguistic usage and thematic content,” they also “do so in the expression of discourse stance as well” (Berman et al. 2002: 280-281). This finding is a repeated instance of a prevalent theme in Berman’s findings: the development of usage of linguistic forms is depicted as proceeding from “dichotomy to divergence” (Berman 2004: 110). In narratives, younger speaker-writers show a high level of communicative, affective, and personalized orientation (dichotomy, monolithic, genre-canonic), whereas older speaker-writers show a more generalized stance, taking the form of “generalized evaluative commentary on the nature of [situations],” especially “in the introductory setting and the concluding coda segments” (divergence, rhetorical flexibility) (Berman et al. 2002: 281). In expository texts, a contrasting pattern to narrative text construction emerges. Younger speaker-writers take a deontic stance that is “almost entirely generalized, and so apparently abstract in nature,” and occasionally include “hypothetical examples” but usually lack “specific illustrative examples of relevant incidents in the past” and also lacking “specific, concrete solutions to the problems they were asked to discuss” (2002: 281). Older speaker-writers, on the other hand, combine an abstract, generalized stance toward the topic with “specific reference to illustrative incidents and concrete proposals for how to tackle the problem” along with “personal commentary on how the situation affects them as individuals,” though they still adopt a more general, abstract, and objective stance than younger speaker-writers overall (2002: 281).

With respect to the explicit marking of discourse stance, Berman found that the discourse stance shifts that mature speaker-writers make are marked deliberately, with explicit linguistic devices. Younger children’s shifting of discourse stance “will tend to be rhetorically inconsistent or communicatively inappropriate” and be characterized by a lack of overt marking of such shifts (Berman et al. 2002: 274, Berman 2004: 110). Rhetorical consistency, therefore, is the aim of this aspect of discourse stance development.

The dimension of orientation in discourse stance seems to develop concurrently with the general “factors of socio-cognitive development” (Berman et al. 2002: 275). Even though younger speaker-writers are expected to use specific and generic, personal and generalized, orientations, mostly the older speaker-writers will adopt a stance that is meta-textual in orientation.

With respect to the dimension of attitude, developing speaker-writers move from personalized attitudes, through socially-conditioned perspectives, ultimately to abstract, “academic-cognitive,” and “universalistic views on given states of affairs,” and from “affective or deontic to epistemic attitudes” in their texts (Berman 2004:111). That is, attitude development is consistent with “Piagetian and neo-Piagetian analyses of socio-cognitive and moral development” (Berman et al. 2002: 275). This trend is attributable to speaker-writers’ emerging ability to take multiple stances on a given topic (or event), enabling them to intermingle their own social values with abstract theorizing in a text (2002: 275).

3. Methodology

In this section, I will describe how I created the student writing guide and explain the reasons for its existing form, pointing out the various theoretical resources used to inform each of its five parts. Then, I will provide information on the subjects who participated in the study. Information about data collection and analysis will follow, concluding with a discussion of the findings and an outlook for future studies.

3.1 Creating a student writing guide: structure, aims, and intended use

The student writing guide entitled, Writing Well-formed Essays: Imagining and speaking (See Appendix I), is a concise, student-accessible writing tutorial designed for self-use. The SWG is broadly directed at helping developing student writers see the significance not only of the formal attributes of a well-formed text (as discussed by Berman’s facets of discourse production), but also to see the communicative functions those linguistic devices are deployed to meet and to imagine themselves speaking within the communicative contexts within which written discourse is deployed (as discussed by the concept of the rhetorical situation). The SWG is thus designed to enable students 1) to reconsider and re-contextualize their writing as something that flows out of a rhetorical situation, the context that gives rise to discourse, rather than considering writing as a bundle of forms expressed for its own sake, or worse yet, merely for a grade; and 2) to promote linguistic literacy and rhetorical flexibility within the (imagined) writing situation. The guide is designed to be a tool for students to improve the rhetorical effectiveness of their own written discourse and to foster linguistic literacy. As described above, linguistic literacy, the ability to use a varied repertoire of linguistic resources (forms) to meet different communicative ends (functions) in a variety of contexts (rhetorical situations), reaches its apex when it becomes rhetorical flexibility, the ability to hold an audience’s attention, intentionally intermingling narrative-text features with expository-text features in order to be precise, expressive, persuasive, and rhetorically effective. Therefore, the (linguistic) forms that the SWG promotes the use of are merely devices the writer/speaker deploys to meet the (communicative) functions of a writing situation. Parts 1) occasion, 2) audience, and 3) speaker-voice are devoted to characterizing the rhetorical situation, referred to in the SWG as the writing situation, and parts 4) information units, and 5) structure / overall organization are dedicated to showing more technical aspects of well-formed written discourse.

1. Ideologies of Writing in the SWG

Before explaining the theoretical resources informing the various parts of the writing guide, it is important, in view of all of the perspectives of composition outlined above, to clearly state what kinds of ideological resources underlie and inform the pedagogical approach taken in the student writing guide. The SWG does not assume one, coherent and sequential writing process, unlike a cognitivist approach would. Rather, the guide aims to provide functionally motivated criteria for well-formed writing that a student-user may appropriate in his or her own way. From the outset, the guide makes clear to users that the most important aspect of writing well-formed discourse is a communicative function executed by the use of linguistic devices in a communicative context. Therefore, communicative intent is ascendant over linguistic form used to convey the communicative purpose. The SWG, however, does posit a sequence of outgrowth from occasion, to audience, through speaker-voice, to the linguistic forms the speaker-uses to best adapt his/her communication to the rhetorical occasion he/she imagines. As regards expressionism, the SWG is unashamedly expressionistic in approach, relegating linguistic form to a role subservient role to the speaker-writer’s creative, identity-unearthing efforts. Also, much of the reason that the guide is written for students to use on their own, independently, without the aid of teachers or tutors, is underpinned by the expressionistic assumption that everyone can write—with or without the aid of writing instructors if the cognitive and communicative apparatuses—the imagination and voice—are put to use. Students are viewed as explorers of aspects of their own identities, creators of knowledge and keepers of values that do not necessarily have to defer to the identities, knowledge, and values of academia or of instructors. This student-centered approach is redolent of social-epistemic rhetoric, as it implicitly shifts power relations in the setting of writing pedagogy and critiques the versions of “good writing” espoused by so called experts of traditional composition pedagogy. The SWG explicitly points out to students that, although they are being graded on their writing in an established system, they should possess sets of personal and other goals for writing that extend beyond the classroom. Students are also encouraged to deploy the voices that are required by the contexts they write in, and not to be wary of the multi-voicedness of their own internal, verbal thought, but rather to use it as a resource for argumentative, written discourse. Maintaining the autonomy of the student writer and empowering him/her, and re-envisioning him/her as a possessor of the tools he/she already needs to write, therefore, are some of the most salient values underlying the SWG.

Student-accessibility is also an important value underlying the approach of the SWG, which is why an illustration in the form of a plant was created to graphically represent how the occasion, audience, speaker-voice—as well as the content the speaker produces—all grow out of one another and are mutually dependent on each other. Also an acronym comprised of the first letter of each section, OASIS, was designed to increase student-accessibility. Each of the five parts of the SWG is explained below in terms of how the background material reviewed above was incorporated and integrated to construct them.

2. SWG, Introduction and recap

The introduction of the SWG, entitled Imagining and Speaking, is designed to activate student-writer’s schemas about the difficulties they may have had with writing. There are frequent references to purpose and being real with regard to the activity of writing that are inherent in an expressionistic approach which considers the student-writer to be the locus of sufficient knowledge and experience to produce well-formed discourse (2.2). There is a heavy emphasis on imagination  and mentally conceiving context that is not actually present with the writer—to compensate with the imagination the absence of a social-communicative context in which the argument (of the essay) would take place. This confident leaning upon imagination the SWG assumes stems from the wealth of literature on the verbal nature of inner thought and consciousness, internalized dialogue, and the heterogeneity of voices briefly touched on above (2.4). The imagination, of course, is merely a mediatory device for picturing a writing situation, as the introduction and concluding section (recap) make clear. In both sections, there is a repeated emphasis on the importance of the writing situation as the necessary and sufficient context for written discourse, following Bitzer (1968) and Grant-Davie (1997) as reviewed above (2.5). To illustrate this, a graphically-augmented metaphor of a plant is put forth, in which written discourse is seen as the product of a speaker-voice (leaves) that grow out of the concept of audience (stems) which protrude up from a discourse occasion (roots). The acronym, OASIS, concludes the introduction and aims to provide readers with a conceptual road map of the forthcoming content.

3. SWG Parts 1-3, occasion, audience, and speaker-voice

Following Grant-Davie’s (1997) reformulation of Bitzer’s (1968) model of the rhetorical situation, I constructed three sections—occasion, audience, and voice—following the parts of the rhetorical situation reviewed above. The only major changes were to replace the term “exigence” with “occasion” in order to increase accessibility and to add “voice” onto Bitzer’s “rhetor”/Grant-Davie’s “speaker”—resulting in the convenient label “speaker-voice”—in order to retain the much used and appreciated concept of “voice” in writing pedagogy.

In the “occasion” section, the heuristic questions, “What am I really writing about,” “Why do people need my writing,” and “What should my writing do to/for others” is an adaptation of the comprehensive features of exigence that Grant-Davie provides—respectively, fact/definition, cause/value, and policy/procedure. I augmented this perspective by showing how an idea of the importance of exigence forms the entire foundation of Graff & Birkenstein’s (2006) tutorial on academic writing, specifically that skilled writers frame their statements as responses to prior statements which need responding to. This “They say/ I say” approach was supported with a brief example.

In the “audience” section, in addition to Bitzer (1968) and Grant-Davie (1997), a wealth of perspectives were drawn upon to support the efficacy of the all-important, long-attested concept of audience in well-formed writing. Ong (1975), a well-known theorist on imagined audience, is evoked in this section following his advice to writers to read literature to pick up author’s coherence-making strategies. Also, material following from Cheng & Steffensen’s (1996) study on how teaching students to use meta-discourse markers helped them connect with an audience in their writing is used. A list of meta-discourse markers, adapted from the list provided in section 2.5 above (following Fraser 1996, 2005), is provided for students who want an immediately-accessible inventory of meta-discourse devices at their disposal.

The speaker-voice section also is buttressed by a wealth of theory, particularly form literary theory’s “heterogeneity of voices” and Trimbur’s (1987) extension of it to the domain of composition. Specifically, if “no distinction between private thought and public discourse” exists, and if writers must “forge a voice and way of speaking from the conflicting social forces and the polyphony of voices that converge in their mental experience [and] individuate their verbal thought and expression,” then it follows that the practice of imagining and trying on different voices to respond to imagined social contexts—which the SWG proposes—is a well-grounded practice that reflects the reality of the writer’s mind (1987: 219-220). In this section, Berman’s observation of the nature and development of discourse stance—rephrased as discourse-viewpoint—is incorporated, along with a truncated list summarizing some different linguistic devices that writers might use to create more personal or more impersonal discourse stances.

4. SWG Parts 4-5, information units, and structure/ overall organization

The other two of Berman’s discourse facets reviewed above were incorporated in parts 4-5 of the SWG: categories of referential content, rephrased as “information units,” and “global text organization,” rephrased as “structure/ overall organization” (2.7.1-2). Part 4 briefly describes the three kinds of information units, gives examples of how they might be used together, and suggests that writers avoid flooding their drafts with core elements, according to Berman’s findings of how these infus are used by mature, skillful writers (2.7.2). Part 5 closes the guide with a definition of structure/ overall organization based on the genre-typical features of narrative and expository texts. A rendition of Berman’s four levels is provided in student-accessible language, and examples I composed approximating the global text quality of Berman’s own examples (Berman 2008: 742-745).

NOTES

[1] Berman and her colleagues had previously studied the different developmental and linguistic aspects of children’s narrative discourse, inspired by the work of Dan Slobin at Berkeley who conducted work on children’s construction of oral narratives (1996). Berman & Slobin’s (1994) “Relating events in narrative: a crosslinguistic developmental study,”known as the“Frog picture story book” studies, analyzed and compared the oral fictive narratives of 3-, 4-, 5-, 9-year olds and Adults in English, German, Hebrew, Turkish, and Spanish. Evident in this work, Berman expressed a view of language acquisition as an integrated, protracted process linguistic literacy, or, the ability to deploy flexible a variety of linguistic forms to meet various communicative functions in various contexts. Berman & Slobin assert that that “the mastery of language structure and language use cannot be explained in monolithic terms […] within the framework of any single linguistic, psycholinguistic, or sociolinguistic theory” (1994: 593). Berman asserted that children “bootstrap into knowledge of linguistic forms” through the confluence of different information sources—linguistic, cognitive, social-interactional, etc. (1994: 593). Though the youngest children in the study had mastered what she calls the “core grammar” of their languages, they were still “far from the ‘endstate’ of a proficient speaker or proficient narrator” (1994: 593). These notions of later-age first language acquisition remain as an important assumption throughout all Berman’s work.

Bibliography

Bakhtin, M. (1981). Discourse in the Novel. The Dialogic Imagination: Four Essays by M.M. Bakhtin. Trans. Caryl Emerson and Michael Holquist. Ed. Michael Holquist. Austin: U of Texas P: 259-422.

Bartholomae, D. (1985). Inventing the University. Cross-Talk in Composition Theory: A Reader. Ed. Villanueva, Victor. 2nd ed. Urbana, IL: NCTE, 2003: 623-653.

Berlin, J. (1988). Rhetoric and ideology in the writing class. College English 50, 5: 477–494.

Berman, R. A. (2005) Introduction: Developing discourse stance in different text types and languages. Journal of Pragmatics. 37, 2: 105-124.

Berman, R. A. (2008) The psycholinguistics of developing text construction.  Journal of Child Language, 35: 735-771.

Berman, R. A. & I. Katzenberger. (2004)  Form and function in introducing narrative and expository texts:  A developmental perspective. Discourse Processes, 38: 57-94.

Berman, R. A. & Nir, B. (2004) Linguistic indicators of inter-genre differentiation in later language development.  Journal of Child Language, 31: 339-380.

Berman, R. A. & B. Nir-Sagiv. (2007)  Comparing narrative and expository text construction   across adolescence:  A developmental paradox.  Discourse Processes, 43 (2): 79-120.

Berman, R. A., Ragnarsdóttir, H. & Strömqvist, S. (2002) Discourse stance. Written Languages and Literacy, Volume 5, 2: 255-290.

Berman, R. A. & Ravid, D. (2008) Analyzing narrative informativeness in speech and writing. Language in the Context of Use:  Cognitive Approaches to Language and Language Learning.  Eds.  Tyler, A. Kim, Y. & Takada, M. Mouton de Gruyter: The Hague [Cognitive Linguistics Research Series]: 79-101.

Berman, R. A. & Slobin, D. I. (1994). Relating events in narrative: A crosslinguistic

developmental study. Hillsdale, NJ: Lawrence Erlbaum.

Berman, R. A. & L. Verhoeven., eds.  (2002). Cross-linguistic perspectives on the development of text production abilities in speech and writing. Written Languages and Literacy, Vol. 5, Parts 1 and 2 (special issue).

Bitzer, L. F. (1968). The Rhetorical Situation. Philosophy & Rhetoric, Vol. 1, No. 1 (Jan., 1968), 1-14.

Cheng, X. & Steffensen, M. S. (1996). Metadiscourse: A Technique for Improving Student Writing. Research in the Teaching of English, Vol. 30, No. 2 (May, 1996): 149-181.

Edelstein, M. (2005). Multiculturalisms past, present, and future. College English, Vol. 68, No. 1, 14-41.

Elbow, P. (2000). Everyone Can Write: Essays Toward a Hopeful Theory of Writing and Teaching Writing. Oxford University Press.

Elbow, P. (2006). “The Music of Form: Rethinking Organization in Writing.” College Composition and Communication. 57, 4: 620-664.

Faigley, L. The study of writing and the study of language. Rhetoric Review 7, no. 2 (1989): 240–256.

Flower, L. & Hayes, J. R. (1981). A Cognitive Process Theory of Writing. College Composition and Communication, Vol. 32, No. 4 (Dec., 1981): 365-387

Fraser, B. (1996) Pragmatic markers. Pragmatics 6, 2: 167-90.

Fraser, B. (2005). Towards a theory of discourse markers. Approaches to Discourse Particles. Ed. Fischer, K. Elsevier Press.

Gorrell, D. (1997). The Rhetorical Situation Again: Linked Components in a Venn Diagram. Philosophy & Rhetoric, Vol. 30, No. 4 (1997): 395-412. Penn State University Press

Grant-Davie, K. (1997). Rhetorical Situations and Their Constituents. Rhetoric Review, Vol. 15, No. 2 (Spring, 1997): 264-279

Graff, G. & Birkenstein, C. (2006). ’They Say/I Say’: Moves That Matter in Academic Writing. W.W. Norton, 2006.

Ong, W. J. (1975). The Writer’s Audience is Always Fiction. Cross-Talk in Composition Theory: A Reader. Ed. Villanueva, Victor. 2nd edition. Urbana, IL: NCTE, 2003: 55-76.

Trimbur, J. (1987). Beyond Cognition: The Voices in Inner Speech. Rhetoric Review, Vol. 5, No. 2: 211-221.

Ravid, D. & Tolchinsky, L. (2002). Developing linguistic literacy: a comprehensive model. Journal of Child Language, 29: 417–447

Ravid, D. & Berman, R. A. (2006). Information density in the development of spoken and written narratives in English and Hebrew. Discourse Processes 41: 117-149.

Todorov, T. (1984). M.M. Bakhtin: The Dialogical Principle. Trans. Wlad Godzich. Minneapolis: U of Minnesota Press.

van Dijk, Teun. (1995) Discourse analysis as ideology analysis. Eds. Schäffner, C. & Wenden, A., Language and Peace. (pp. 17-33). Aldershot: Dartmouth Publishing.