JAC Volume 16 Issue 3

In a 1979 article, "On Audience and Composition," Lisa Ede articulated "the need for a more audience-centered approach to writing"(291), acknowledging that few compositionists seemed aware of the centrality of audience to the writing process. Ede's plea for renewed attention to audience has not gone unanswered. Since the early 1980s, there has been a flood of critical commentary and research on audience. An on-line search of ERIC abstracts from 1982 to March 1995 yields 4,159 entries focusing on the keyword "audience," and the MLA Bibliography records 3,999 audience-related entries. Researchers and theorists in various disciplines—not only composition and rhetoric, but also speech communication, technical communication, literary studies, education, cognitive psychology, and linguistics—have acknowledged the importance of audience to the communicative process, a concept seen as crucial to the writer's invention process, formulation of purpose, arrangement, and stylistic and linguistic choices. Since Aristotle, audience has been seen as a key component of the rhetorical situation: "Of the three elements in speech-making—speaker, subject, and person addressed—it is the last one, the hearer, that determines the speech's end and object" (Rhetoric 1.3). And, recently, audience has gained renewed attention in light of social, dialogic views of discourse. Without an audience, there is no dialogue, or as M.M. Bakhtin notes, "without addressivity the utterance does not and cannot exist," for "from the very beginning, the utterance is constructed while taking into account possible responsive reactions, for whose sake, in essence, it is actually created" (94).

While theorists and researchers seem to agree upon the importance of audience to the communicative process, there is no agreed upon meaning of audience. "Audience" is an unstable referent, a floating signifier. The term can refer to a construct in the writer's mind—the "imagined," "intended," or "invented" audience. Or it can refer to a textual presence—the audience "implied" by textual cues or "inscribed" in the text. Audiences can also refer to "real" people—the actual readers who exist either apart from and prior to the text (as "addressed" readers) or those who exist as part of the community in which the text is produced.

These multiple meanings of audience have been explored in numerous articles published in the late 1970s and early 1980s. Central to these early attempts to mediate between the multiple meanings of audience in written discourse is Douglas Park's description of the range of audience definitions in his article, "The Meanings of Audience" (1982). According to Park, there are "two general directions of meaning of 'audience'—outside the text and back into the text" (250). This dichotomy of readers outside the text versus readers created within the text is the subject of a key and oft-cited article that appeared in a special audience issue of College Composition and Communication (Volume 35, May 1984). In their article entitled "Audience Addressed/Audience Invoked: The Role of Audience in Composition Theory and Pedagogy," Lisa Ede and Andrea Lunsford distinguish between "audience-addressed," the audience that exists outside the text and receives messages, and "audience-invoked," the audience that writers imagine and cast by means of textual cues into a role that real readers are then invited to adopt in responding to the text. However, after setting up an opposition between these two central perspectives on audience, Ede and Lunsford argue for a "synthesis" of the two perspectives, one that would "balance the creativity of the writer with the different, but equally important, creativity of the reader" (169) and that would acknowledge the importance of all elements of the rhetorical situation that interact with audience and shape meaning.

While Ede and Lunsford's article is cited in virtually all discussions of audience since 1984 and has created ongoing discussion about "invoked" versus "addressed" audiences, their definitions of audience as well as their descriptions of the interaction of these two perspectives of audience can be further complicated by recent social theories of communication and by attention to the notion of "multiple audiences," a social view of audience developing out of professional writing research. In the decade since Ede and Lunsford's article was published, several social theories have evolved, theories that challenge their descriptions of audience, but more importantly, extend and clarify their perspectives. In this paper, I will examine the ways that these social theories of audience inform the perspectives of "invoked" and "addressed" readers and will simultaneously critique these theories, particularly for their failure to adequately acknowledge the multiplicity of real readers. Finally, drawing on audience research in professional and technical communication, I will argue that compositionists should embrace a more comprehensive social model of audience—one that encompasses the multiplicity and complexity of readers.

Rereading the "Invoked" Audience

Ede and Lunsford define the invoked audience as "a construction of the writer" (160) or what Long calls a "created fiction" (225). This mental sketch of audience—the writer's cognitive construct—aids in invention and directs the selection of textual features that, in turn, embody this sense of audience. While studies of the "invoked" reader are diverse, most focus on the writer's internal image or mental sketch of audience and examine how this inner "sense" of audience affects the writer's rhetorical, organizational and stylistic choices during both the invention and revision stages (Barry Kroll 1978; Russell Long 1980; Carol Berkenkotter 1981; Bennett Rafoth 1985, 1989; Robert Roth 1987). While those who research the "invoked" audience do not deny the existence of real readers, they assume that writers cannot know these readers and therefore must "create" or "invent" them and must employ textual and rhetorical devices to "inscribe" these readers in the text. From this perspective, audience has little to do with the identities of the actual people who later read the text—such as teachers, editors, managers, supervisors—real readers who may not coincide with the "invoked" readers.

If audiences are "invoked" by writers, where do such mental representations of audience come from? Various responses to this question have been supplied: from a writer's stored plans, from schemas, from interpretive frameworks, from the writer's intuition. Most of these early explanations regarding the source of a writer's awareness of audience are "inner-directed" and are explained by turning to innate processes and mental structures. Social theories, however, can illuminate what otherwise seems a mysterious process—"invoking" a reader, or building a mental representation of readers—by more accurately and concretely acknowledging the source of the set of conceptions or awareness of audience in the writer's consciousness, an awareness that shapes the discourse.

Following Patricia Bizzell's (1982) call for "outer-directed" as well as "inner-directed" approaches to studying writing and her attention to the social context of language use, Douglas Park (1986) proposed a socio-cognitive view of audience that examined how one's "inner" awareness (or what Park calls a "sense of audience") is affected by "outer" forces—by the social function of the discourse or the social institution that the discourse serves. Park sees social context as crucial for understanding the identity of an audience and argues that when responding to the question, "Who is my audience?" writers always respond in terms of the social institutions and social function that the discourse serves. According to Park, "Unspoken but always present in any such simple identification of an audience is the whole complex of the social situation that has brought that audience into being" (483).

The writer's knowledge of the social situation that has brought the audience into being shapes the "invoked" audience, guiding the writer's rhetorical, organizational, linguistic and stylistic decisions. The writer can also use the social institution or social relationships within which the audience exists to determine the previous knowledge of the audience and to shape the tone and purpose accordingly. For example, a writer addressing an audience of composition scholars will be unable to clearly "invoke" readers without clarification of the social function. Will the paper be published in a journal? Which one? Or will the paper be presented at a conference? Will it be a workshop or a large panel discussion? Certainly the writer's decisions regarding not only the audience's expectations and preconceptions but also topic, length, formality, etc. will be determined by her answers to these questions. Envisioning an audience in the context of a conference workshop, perhaps a more specialized audience compared to the nationwide reading audience of a journal, will likely result in a paper that deals with issues more narrowly and specifically and is less likely to give a great deal of background information. A presentation for a workshop, because the audience members are seen as co-participants, might also be much less formal than a paper presented in a panel presentation. Thus, the more the writer knows about the social function of the discourse, the clearer that her sense of audience becomes, in turn clarifying her rhetorical choices. Such a perspective extends Ede and Lunsford's discussion of how the rhetorical situation affects a writer's "invoked" audience by demonstrating that the social situation can equally inform a writer's representation of audience and her rhetorical choices based on this audience representation.

Park's social theory of audience, however, accounts very little for the multiplicity of audience members in a social setting. While Park does acknowledge that a piece of discourse may serve "more than one function and audience," most of his discussion focuses on how the social context for the discourse defines the audience as a collectivity(485). According to Park, "[B]ecause a written discourse always exists within some larger social setting and reaches its dispersed readers through a given physical means of distribution for an accepted social function, readers of prose are very much part of a collectivity" (483). This perspective portrays the audience as a homogeneous entity and fails to recognize the dissensus and conflict created by the multiple audiences in a social group. Those compositionists attending the conference workshop, to return to the above example, are likely to occupy various positions of power and to bring different value systems and ideological perspectives to their participation in the workshop. For instance, if the workshop happened to be focused on the topic of audience, those who value cognitive perspectives may judge the speaker's presentation on its ability to answer questions regarding the writer's mental representation of readers, while those with an interest in linguistics might respond negatively to a presentation that ignores the textually inscribed or "encoded" reader. The multiple subjectivities of the audience members—their differing disciplinary training and their adherence to differing theoretical perspectives and methodologies—can complicate the speaker's sense of audience; however, awareness of these competing ideologies may enable both speakers and writers to approach language and concepts from multiple angles and to filter their ideas through differing ideological stances.

In addition to its failure to acknowledge the differing ideologies of multiple audiences in a given social setting or institution, Park's social view of audience stops short of acknowledging that readers themselves are socially constructed and are equal participants in the making of meaning. In Park's analysis, it is the writer who defines situation, using conventions of written discourse to "evoke the shadow of a situation" (487). Park presents a view of audience that locates cognition in context. The writer's "sense of audience" is "informed by a grasp of the social function of the prose" (487). However, absent from his discussion is a sense of the actual readers who exist as participants in a social group or institution and, more importantly, the readers who act as co-creators in the production of discourse.

Park's socio-cognitive view of audience—a view of how social context influences a writer's "invoked" audience—has been revised by more recent social perspectives developing out of social constructionist and poststructuralist theories. Theorists such as Bennett Rafoth (1988) and James Porter (1992) propose a community approach to audience which examines how the writer's "invoked" audience is shaped by participation in the discourse community. A discourse-community perspective of audience redirects the focus of a writer who would "invoke" an audience. A writer posing the question, "Who is my audience?" or "Who do I want my audience to be?" would, instead of looking inward, direct her attention to the communities and contexts that give rise to a text, including the range of conventions and constraints that shape different kinds of writing in various communities.

Porter's work is significant in that it clarifies and extends the view of "audience-invoked" by examining how communal expectations and conventions work to shape a writer's awareness of audience and the writer's textual representation of this audience. However, his community approach to audience is predicated on Michel Foucault's concept of community as a "discursive formation," a perspective which envisions a community as defined by its textual practices. As a result, Porter defines community as a "textual system" and focuses on the audience that is constituted textually rather than socially. Ironically, while Porter acknowledges the socially constructed nature of meaning and knowledge, his perspective on audience is primarily text-based and encompasses only to a very limited extent the actual readers who negotiate textual meanings with writers. His notion of audience as a "textual property," as "a set of attitudes or conventions located in sets of texts" does not seem too far removed from an audience "invoked" by stylistic and semantic cues (xi).

Overall, while both Porter and Park advance social perspectives that help clarify the source of the "invoked" audience, they focus primarily on the writer's representation of audience and/or the textual representation of audience. Both are important studies that clarify the social contexts and communal expectations and knowledge that shape a writer's awareness of audience, thus extending and clarifying Ede and Lunsford's rhetorical perspective. However, their perspectives lack a focus on "real" readers, the readers that actually exist as concrete realities, readers that Ede and Lunsford designate as the "addressed" audience.

Rereading the "Addressed" Audience

Ede and Lunsford define the addressed audience as "the actual or intended readers of a discourse [who] exist outside the text" (167). The "addressed" audience is based on the classical rhetorical view of audience as a concrete, physical reality existing outside the text—a fairly static entity that can be known through analysis of background, beliefs, values, attitudes and other demographic variables. While a notion of readers existing as a concrete reality would seem to lend itself to a more social perspective, Ede and Lunsford assume that knowledge of this audience is possible only via observation and analysis of the writer, thus distinguishing it little from the cognitive perspective of writers who "invoke" their audiences. In fact, their use of the term "intended" readers in their above definition of "audience addressed" suggests this cognitive angle. The addressed readers may seem like actual readers, but they are actual only in the sense that they are "actualized" by the writer. While Ede and Lunsford allude to actual readers in their study—the editors and colleagues who responded to their article—and define addressed readers as physically present readers existing outside the text, they qualify their definition of the "real" reader by noting that "writers embody or give life to their conception of this reader" (167) through their use of textual cues and resources of language.

Porter's conception of "real" readers corresponds to Ede and Lunsford's conception of "addressed" readers—those passive recipients of meaning who exist apart from and prior to the text. The "real" readers that he refers to throughout his work are fictions, imaginative constructs of the writer. Porter does not acknowledge the existence of actual readers who participate in the discourse. He states, "For the writer in the act of composing, the real reader may not exist or, even if it does, it may be irrelevant to the composing act" (5). The reader, for Porter, exists primarily as a discursive formation, "a structure embodied in the sets of texts that define a given discourse community" (83).

Porter attributes his lack of attention to "the behaviors of actual readers" to his rhetorical focus (13). While one would expect a social view of audience—a view of audience as "community"—to include those real readers who exist socially and participate in the community, Porter defines discourse community as a "textual system" rather than a social or sociological system and defines it in terms of its rhetorical features. He also concedes that "discourse community" is an abstract and "fuzzy" concept, a limitation that Joseph Harris notes in his 1989 article on "The Idea of Community in the Study of Writing." Harris points out that "discourse community" often functions as a metaphor, bearing little relation to physical groupings of writers and readers. In addition, despite Porter's attempt to define discourse communities as dynamic and flexible entities, the readers who exist communally are portrayed as homogeneous and monolithic. A main shortcoming that Harris notes in his critique of "discourse community" is that it fails to account for conflict and dissonance. Porter responds to this critique by acknowledging, at least on an abstract level, the complex network and heterogeneous layers that constitute a discourse community. However, on a concrete level, the "infinite variations of particular audiences" that such a definition of discourse community would entail leads him to "rhetorical despair" (107). He thus turns to "forums," more stable, concrete locations of discourse production and reception. Porter defines forum analysis as "a kind of audience analysis—but not focused on real readers" (112). And as his sample "Forum Analysis Heuristic" demonstrates, with its questions regarding the audience's background, values, beliefs and attitudes, a rather static, monolithic audience is assumed. Bennett Rafoth (1988), another leading proponent of the discourse community view of audience, concedes that a community approach to audience has its limitations, namely the failure to account for the "multiple allegiances" of the actual participants in the community and the "conflict" that can result (143). Due to their focus on the communally constituted audience, both Rafoth and Porter fail to adequately account for the multiplicity of real readers that constitute the community and that participate in the discourse.

Marilyn Cooper has been most vocal in exposing the limitations of audience perspectives that ignore the real readers. To counter such perspectives, including Ede and Lunsford's perspective on "addressed" and "invoked" audiences, Cooper offers a social perspective of audience based on the readers writers know through real social encounters and receive actual feedback from—"readers as real social beings" (372). Cooper transforms the writer-based constructs of "invoked" or "addressed" audiences (audiences that writers invent or analyze) into "real readers"—friends, colleagues, and roommates who read and respond to drafts, editors who comment on and critique submitted articles, bosses who comment on a report or memo written by an employee, etc. These readers are not only "addressed," but they contribute to the construction of meaning, bringing to their negotiation of textual meaning various attitudes, values, predispositions, epistemologies, and ideologies.

Although Cooper focuses on the actual feedback provided by real readers who dynamically interact with the writer and text she stops short of acknowledging the conflict and tension that can result from the varying epistemologies and ideologies that real readers—socially and culturally situated readers—bring to texts. These tensions between writers and actual readers demonstrate that actual feedback from real readers is not always useful or constructive feedback. In a recent article, Barbara Tomlinson examines the review process of scholarly articles in composition and describes the powerful—and often negative—effects that real readers (in this case, reviewers) have on writers. These real readers, who bring their own personal, social, political and intellectual agendas with them, may create conflicts, particularly if those agendas vary from the writer's agendas. In addition, reviewers may occupy different discourse communities within the larger community of composition studies and may frustrate writers by refusing to become the "implied" reader and by challenging the writer's reasons for making particular rhetorical decisions. However, while Tomlinson critiques the destructive feedback of readers who respond by exerting their authority and power over the writer, she argues that the feedback of various real readers has "enormous" benefits: "The explicit or implicit critique of other frameworks, other rhetorical strategies can work heuristically, opening out new world views, new categories, new takes on what we do and how we talk about it" (88).

The multiple readers of this article—including various colleagues with differing disciplinary interests and, eventually, anonymous reviewers for JAC—demonstrate the conflict that results from the multiple subjectivities of real readers. One colleague who responded to my article professed to have more of a background in literary studies and noted that she "stumbled" upon the introduction of the concept of "discourse community" in the paper and was "not sufficiently educated on the concept." Another reader, who values the pedagogical angle, noted that the paper "skirts" the issue of how one might actually teach a concept of multiple audiences. And once the article was sent out for review, feedback was given by readers whose responses reflected the issues and stylistic strategies that they value. In fact, the entire review process demonstrates the tensions among writers and real readers—often useful tensions—and the negotiation that must follow. The responses I received from my multiple readers, while conflicting, helped me clarify my stance, reevaluate my rhetorical decisions and better understand the various ideological stances within my own and others' disciplines.

Comparing writers to musical composers and their multiple readers (reviewers) to musicians who play the musical piece, Tomlinson explains that the participation of real readers in the text can either validate the writer's meaning or challenge that meaning:

Beyond the Invoked/Addressed Dichotomy: Embracing a Model of Multiple Audiences

Jack Selzer, in his recent article entitled "More Meanings of Audience" (1992), problematizes the constructs of both invoked and addressed audiences by "call[ing] attention to the multiplicity of readers inside and outside the text" (163). Following Selzer's lead, I propose that compositionists embrace a social model of audience that accounts for the multiple and shifting roles of readers as they participate in social groups. Such a model exists in studies of workplace writing. Professional writing scholars who have examined the workings of audience in actual discourse communities—such as workplace settings—have detected "multiple audiences," a term used in technical communication to distinguish the various readers of documents in complex social organizations. J.C. Mathes and Dwight Stevenson—in their book, Designing Technical Reports: Writing for Audiences in Organizations—provide an extensive discussion of the many audience levels involved in technical communication and provide one of the earliest glimpses of the social model of audience as community.

The contextual model of writing in the workplace is multilevel and interactive. Mathes and Stevenson describe the various "communication routes" of a technical communication document and explain how multiple audiences exist, audiences shaped by their position and role in the social organization. Communicative events in organizations are layered, with audiences existing at various levels, described by Mathes and Stevenson as horizontal, vertical, and external audiences. "Horizontal audiences" consist of people in the same office or project group or persons in adjacent groups. While more distant, "vertical audiences" are still within the same organization but may consist of distant management such as department heads. Finally, "external audiences" consist of persons outside of the organization (15-16). This concept of multiple, layered and potentially conflicting audiences presents a view of audience as socially constituted and organized, thus extending the textual or individual perspectives of audience.

Multiple audiences are not only defined by their different places in organizations but are also defined in terms of how they will use the document. The "primary" audience is the audience most directly impacted—the audience who makes decisions or acts on the basis of the information presented, while the "secondary" audience is affected by the decisions and actions. Finally, the "immediate" audience consists of those who route the document, such as the writer's supervisor or another middle management person (Mathes and Stevenson 21). According to Mathes and Stevenson, these audience designations may overlap, with the immediate audience becoming part of the primary audience or the secondary audience.

An example should help clarify these various audience levels. For instance, suppose that a graphic designer in a marketing department of an advertising firm believes that the department is losing some important clients due to the use of outdated equipment and decides to write up his recommendations. His "immediate" audience is his boss, a middle-level manager who cannot authorize large expenditures but can recommend ideas to her boss, the "primary" audience who will make the decision on whether or not to purchase the equipment. "Secondary" audiences in this situation would be others affected by such a decision, such as the employees in the purchasing department who would also receive the proposal, which would likely detail the cost to purchase, maintain and operate the new equipment.

While the above example may imply a rather linear and hierarchical process, a multiple-audience perspective is much more chaotic and fluid, with writers simultaneously negotiating textual meanings with these various readers and continuously receiving feedback from these audiences as they publicly draft their documents. Aletha S. Hendrickson, in her study of CPA writing in its rhetorical contexts, demonstrates how multiple audiences often do not layer into a hierarchy of audience segments and instead may overlap. Drawing on CPA writing, Hendrickson shows how the client-audience is simultaneously primary, secondary and immediate—primary, because they make financial decisions for their company based on the report; secondary, because they are affected by decisions of others such as bankers who grant or deny loans based on the report; and immediate, because they transmit or publish the report (323). This example provides a more fully contextualized, dynamic, and polyphonic model of audience, demonstrating that not only do writers negotiate the multiple layers of audience that exist in social organizations but that these layers often overlap. However, such a complex perspective of multiple audiences is obscured by the focus in our field on the "invoked" and "addressed" dichotomy and by the discussions in our textbooks of a uniform and static audience that can be determined through application of audience analysis heuristics.

A more enriched view of audience is provided by those who continually expand our notion of the multiplicity of readers. To the three levels of audience identified by Mathes and Stevenson, Kitty O. Locker, in Business and Administrative Communication, adds a "gatekeeping audience" that has the power not only to route the document but to stop it from leaving the writer's organization. And Vincent Brown has recently added a "watchdog audience," consisting of external industry reviewers. A watchdog audience observes and evaluates the writer's performance and success in satisfying the primary audience and "has substantial political, social, or economic power over the authors" (71). While writers focus mainly on the primary audience, Brown's study found that they frequently consider the watchdog audience as they write and revise.

Selzer also extends the definition of multiple readers, noting that these various readers are not just defined by their place or role in the organization, but by "the multiple tasks or multiple reading behaviors that various readers may require of a text" (170). For example, many major corporations publish annual reports, texts that highlight the company's major projects and which also function as a financial report or balance sheet. This text may have various "real" readers, readers with multiple reading behaviors: employees, who may read out of interest in other company projects that they weren't involved in, or to gain a sense of accomplishment from projects they did participate in; clients, who read it more as an informational piece, to see what they can learn about the company; and owners/shareholders, who read more critically, to see if their investment is paying off or to determine whether they will buy or sell stock in the company. These multiple audiences may or may not be "invoked" by the text, and some of these reading roles may be brought to the text "from outside" by real readers, readers who—unlike Ede and Lunsford's "addressed" audience or even Porter's "discourse-community" audience—may assume various and shifting roles in relation to writer and/or text.

In addition, these multiple and shifting roles of readers may cause conflict and dissensus among writers and their multiple readers. The writer of the annual report in the above example may wish to focus on the "external" readers of the document, thus envisioning the report as a conceptual piece and writing it in a manner that it is interesting to outsiders. However, the readers from various departments or subdivisions within the corporation will likely be more concerned with the portrayal of their contributions rather than the "conceptual" quality of the document. In this case, the writer is faced with negotiating the corporate politics of multiple readers—superiors who can censor the text and various department heads who "play politics" to suppress negative angles or to highlight positive angles. While the writer must struggle o find a middle ground for all of these competing factions, these tensions can help the writer shape the text. Out of the writer's negotiation of these tensions and conflicts emerges a document that is open to multiple angles and considers the various viewpoints of the multiple readers, while perhaps privileging the audience at the top of the power structure. Tomlinson encourages writers to "consider all these multiple, layered, conflicting audiences when it seems heuristic or helpful," but to "abandon" some layers of audience when necessary.

Overall, a social model that accounts for the multiple and shifting roles of readers better captures the complex and dynamic interaction of readers, writers, and texts in the communicative process. Such a perspective enlarges and complicates our understanding of audience. By focusing attention on how a writer routes a document through a social organization, receiving feedback from multiple and often conflicting audiences, such a perspective not only takes into account both the production and reception of discourse in a social group (thus encompassing a more social rather than writer-based or text-based perspective), but also acknowledges the various overlapping and conflicting roles of those who receive the discourse.

Pedagogical Implications of a Multiple-Audience Perspective

A social perspective of audience that encompasses the multiple and shifting roles of reader has implications for the teaching of audience. Many of those who study writing in the workplace acknowledge the disparity between nonacademic and academic settings. Brown notes that "social context is one big difference between academic and nonacademic discourse" (73). And Mathes and Stevenson note that "when students write in their roles as students to individuals in their roles as professors, this is not true communication" (4). However, others have argued convincingly for the integration of these two worlds in undergraduate writing instruction (Odell and Goswami, 1985; Matalene, 1989). In "Workplace and Classroom: Principles for Designing Writing Courses," Lauerman, et. al. argue that their study of writing in the workplace "made [them] look at writing in new ways" and "made [them] especially aware of the complexity of writing situations" (450). Certainly the workplace perspective of multiple audiences demonstrates the complexity of multilevel writing situations and, in turn, clarifies the complexity of audiences in the classroom as well.

Just as horizontal, vertical and external audiences exist in workplace settings, these layers can also exist in academic settings. The horizontal readers of the workplace (those readers who share an equal status with writers) are paralleled by a writer's classroom peers, those who review the writer's work and collaborate with the writer on class projects and papers. The student's peer audience in composition courses bears striking similarities to the horizontal audience described by Mathes and Stevenson—readers who may have little in common with the writer beyond working for the same organization (or taking the same class), having the same rank, or perhaps having the same level of education although varied educational experiences (12). The differences between writer and reader become more magnified with vertical audiences. The vertical readers described by Mathes and Stevenson (more distant audiences holding positions above the writer) coincide with the composition teacher, who, while perhaps seeking to decenter his/her authority, still maintains such authority with the power to assign grades. Finally, external audiences (audiences that exist beyond the organization) often come into play in composition classrooms as well, as teachers create "real-world" contexts and ask students to address readers of Time magazine or their state legislators. If teachers of composition make "external" audiences part of the assignment, they become a "watchdog" audience of sorts, one who observes and evaluates the writer's performance and success in satisfying the assigned audience. And, of course, all teachers assume the role of the "gatekeeping audience" that Locker describes. But while the gatekeeping audience in technical communication has the power to stop the document from leaving the author's organization, the teacher as "gatekeeper" can fail not only the paper but the student as well. Peter Elbow (1983), who has explored the gatekeeping function of the instructor as it conflicts with other nurturing roles, has made a case for "embracing contraries" in our teaching; however, we might also acknowledge the conflictual roles of audience, for students—whether conscious of it or not—are continuously negotiating multiple audiences.

The question becomes whether or not teachers should make students aware of these multiple audiences—whether they should "embrace" these various and potentially oppositional audiences. Selzer aligns himself with Ede and Lunsford in his attempt "to subdue the meanings of audience for the sake of teaching" (163-64). However, such an attitude obscures the complex variety of audiences that constrain or influence a writer's rhetorical choices. Robert Roth (1990) notes that having students envision multiple reading roles is "a way of opening up the possibilities of the text" (182). In Roth's example of a student, Johanna, who wrote for multiple audiences (which he identifies as self, immediate readers and extended readers), such an openness to a wide range of potential readers expanded her reflection, exploration and development of ideas. Thus, while juggling multiple notions of audience may complicate the communicative act for students, such an approach allows more flexibility than following a rigid definition of audience and responding to a set of heuristics designed to describe this monolithic audience through analysis of character traits and demographic variables. The multiple-audience situation is much more dynamic and fluid than prevailing audience adaptation models, models which portray readers as static and homogeneous. As a result, students stand to gain much from the process of negotiating multiple audiences, perhaps becoming more aware of how the process of meaning construction is truly a social negotiation, as well as a potentially conflictual negotiation.

An engineering student in an advanced composition course I taught this past semester illustrates how such negotiation of multiple audiences takes place in the writing classroom. Because the advanced expository writing course allowed students to write in their disciplines, Denise chose to write her papers in the field of engineering. The papers she wrote—on such topics as "High Speed Civil Transport Design" and "Laser Light Detection of Forest Fires"—were addressed to her fellow engineering students and engineering instructors but also accommodated her composition classmates and writing instructor whose reading of the text differed radically in terms of level of background knowledge and knowledge of language and discourse conventions. A typical description of audience (as part of a writing situation analysis) demonstrates Denise's attempts at negotiating these conflicting audiences:

Teachers in technical and business writing courses have recently begun to incorporate the concept of multiple audiences in their pedagogical approaches. L. Lee Forsberg notes the importance of acknowledging the existence of multiple audiences in the business-writing classroom:

An awareness that most texts are written for more than one audience is applicable not only to technical and business writing but to writing classes in general. When making a writing assignment, composition instructors should encourage students not to "consider their audience" but to "consider their various audiences." As Vincent Bron suggests, student writers should ask, "Who will be reading the text?" followed by the question, "who else?" (73). Such an approach to audience will not only open up the possibilities of the text by acknowledging a wide range of potential readers, but it will also force writing teachers to be explicit in their acknowledgment that while the assignment may identify an audience of peers, they will also be reading and evaluating the paper. Such a recognition of multiple audiences in the classroom would allow teachers of writing to recognize the conflict that students often feel (but that is not always acknowledged)—a conflict between the assigned audience, an audience that does not always coincide with the actual readers, and the teacher as reader. While audiences that are identified as part of the assigned topic usually form the primary audience for students, they cannot afford to overlook the "watchdog" audience—the teacher who is evaluating the paper and observing the writer's success in meeting the expectations of the primary audience. Instead of trying to ignore these multiple audiences or trying to reconcile them by pretending to "become" the editor of the local newspaper or the congresswoman that students are addressing in their papers, teachers might begin to acknowledge the existence of multiple readers and reading roles and to be upfront with students regarding the difficult task of negotiating these audiences. In this way, teachers do not have to deny the fact that an actual audience exists—the teacher as reader—a reader who belongs to the academic community and evaluates texts according to the conventions of this community. At the same time, they do not have to ignore the benefits of giving students practice in writing for "real" situations—for a variety of socio-rhetorical contexts that move outside the classroom and that involve "external" audiences. Finally, such a dynamic view of audience would also place value on the responses of the other real readers, the student's peers, whose feedback is not just read and tossed aside but becomes part of the negotiation of textual meaning—another stop on the "communication route" of a text.

Ultimately, instead of overlooking the various layers of audience that exist, writing teachers can acknowledge these multiple and conflicting audiences, perhaps using them to their benefit in the classroom to demonstrate the difficulty of defining discrete interacting elements of discourse or to account more comprehensively for the social nature of communication. Such an approach accounts more fully for the truly complex and social nature of communication and enlarges and enriches both our own and our student's understanding of audience as a dynamic social concept.

Conclusion: From Synthesis to Fragmentation?

Ede and Lunsford, arguing for a representation of audience that allows for interaction among all the elements of rhetoric, claim to have achieved a synthesis of invoked and addressed audiences, one that emphasizes the "dynamic duality" and "integrated, interdependent nature" of the process of reading and writing (169). However, this "dynamic duality" is less dynamic given their view that the addressed readers, the so-called "real" readers, are merely a "stimulus" to writers and that "it is the writer who chooses what to accept or reject" (166). Ultimately, there is nothing to synthesize. Despite the contrasts that have been drawn between audience-addressed and audience-invoked, both show the authority the writer has in influencing the audience, whether it be a concrete, physically present audience the writer knows or an audience the writer creates. In addition, their notion of so-called "real" readers—the "audience addressed"—is based on traditional rhetorical assumptions of audience as a "knowable" and thus static entity, a perspective that fails to acknowledge the dynamic interactions of writers with multiple readers.

While the recent discourse-community view of audience comes closer to achieving this synthesis of the writer who creates the reader and the real reader who (re)constructs his/her role through textual cues, such a view is also inadequate for describing the complex interactions of readers and writers in social settings. Porter argues that "audience as discourse community" provides a middle ground between the notion of invoked readers and the notion of real readers and that "the roles of writer and audience become blurred [and] coalesce in the notion of discourse community" (116). From this social perspective, the synthesis that Ede and Lunsford wanted is seemingly reached: "the creative, dynamic duality of the process of reading and writing, whereby writers create readers and readers create writers" (169). However, to suggest, as Porter does, that the roles of writer and audience neatly "coalesce" in the discourse community is to ignore the potential conflicts that a writer may encounter in shaping a text to meet the expectations of audiences with differing levels of experience and expertise and with various reasons for reading the text. Writers may draw on their knowledge of real readers and their knowledge of community values, beliefs and conventions in order to construct an audience, and, in turn, the real readers may use their knowledge of the community and its discourse conventions to help them navigate a text. However, this blurring of imagined and real readers and, by extension, the blurring of writer and reader roles, is complicated by the multiplicity of real readers who may participate in the discourse to varying degrees and who may bring to the discourse various reading roles—some of whom may even exist outside the community.

Not only the false distinction that has been drawn between readers inscribed by the text and the persons actually reading it, but also the various attempts to synthesize these two views of audience, has obscured the complex variety of audiences that shape and constrain the production of texts in social groups. Such attempts at synthesis seek to stabilize audience. However, a social perspective of audience that envisions the multiple, layered, conflicting audiences that exist more comprehensively accounts for the dynamic, social nature of communication. Thus, rather than trying "to subdue the meanings of audience," we should try to enlarge and complicate our understanding of this concept, to contribute to "the fully elaborated view of audience" that Ede and Lunsford sought nearly a decade ago. We can accomplish this by looking beyond theories of audience that invoke the "imagined" versus "real" dichotomy—a prominent debate in our field that has perhaps obscured other relevant views; by acknowledging the important work done on writing in nonacademic settings and its relevance to our understanding of how audience operates socially, in multiple and complex ways; and by taking a critical perspective toward textbooks on writing that try to obscure the "several audiences lurking behind the apparent audience" (Brown 74). In addition to complicating our view of audiences by recognizing their multiplicity, we should also encourage students to see that writing often entails a negotiation among various and multiple readers.

Overall, while "the need for a more audience-centered approach to writing" (Ede 1979) has long been met, a recognition of the multiple, heterogeneous audiences that exist in social processes of communication—while readily acknowledged in professional and technical communication—has yet to be fully acknowledged in composition studies.