Classification, rhetoric, and the classificatory horizon.

ABSTRACT

BIBLIOGRAPHY PROVIDES A COMPELLING VANTAGE from which to study the interconnection of classification, rhetoric, and the making of knowledge. Bibliography, and the related activities of classification and retrieval, bears a direct relationship to textual studies and rhetoric. The paper examines this relationship by briefly tracing the development of bibliography forward into issues concomitant with the emergence of classification for retrieval. A striking similarity to problems raised in rhetoric and which spring from common concerns and intellectual sources is demonstrated around Gadamer's notion of intellectual horizon. Classification takes place within a horizon of material conditions and social constraints that are best viewed through a hermeneutic or deconstructive lens, termed the "classificatory horizon."

INTRODUCTION

Current scholarly work in classification and rhetoric has converged on a remarkably consistent set of ideas about the role of context, history, and material conditions in the dissemination of texts. Specifically, scholarship addressing classification for access bears a direct relationship to scholarship in material rhetoric. It is easiest to understand dais relationship by briefly tracing the development of bibliography and the concerns that development has carried with it into current scholarship in classification for access. These concerns, and scholarly conclusions about them, show a striking similarity to a particular set of problems raised in rhetoric and spring from common concerns and intellectual sources. This paper argues, in part, that classificationists should consider rhetoric a valuable part of what they do, and that rhetoricians should view classification as an underdeveloped part of rhetorical studies.

This paper will relate a set of ideas that have helped shape the related areas of classification and material rhetoric. It will refer to secondary scholarly sources where these ideas have been derived. Many of the ideas spring from common sources, while others represent similar conclusions reached by different paths. All, though, can be understood as grouped around a central idea, the idea of an intellectual horizon. This paper argues that the works in question, taken together, support the idea that classification takes place within a horizon of material conditions and social constraints that are best viewed through a hermeneutic or deconstructive lens. This application of hermeneutics or deconstruction to classification does not represent a grafting of alien ideas onto the study of information but rather the reconvergence of ideas that come from common sources. The idea of intellectual horizons has implications in two areas. First, it can offer us a theoretical understanding of classificatory practices, and second, it also can begin to suggest possible limits to our ability to fully represent texts through classification.

To describe the conclusions reached by these scholars and relate the ways in which they reinforce each other, this paper will take the following steps:

Trace a set of ideas originating in bibliographic scholarship that led to later work in classification and material rhetoric, described in two senses by Seizer and Crowley (1999) as the study of "the material aspects and groundings of language as rhetorical action" and "the rhetorical nature of material realities" (p. 9).

* Briefly describe deconstruction and postmodern hermeneutics and trace their relationship to classification and rhetoric.

* Discuss the recent relationship between classification and rhetoric, particularly in the area of material rhetoric, and posit the idea of a classificatory horizon.

* Discuss similar but independent ideas advanced in the contemporary study of information.

The paper will conclude with a brief summary of these relationships and discuss what they might offer for future scholarship.

EARLY BIBLIOGRAPHY

Besterman (1940) provides us with a concise history of early bibliography, and that history foreshadows contemporary critiques surrounding the implication of classification for access in the use and reception of texts. Besterman begins by distinguishing two basic types of bibliography, systematic and critical. He follows W. W. Greg in defining systematic bibliography as "the classification of individual books according to some guiding principle," or in his own terms, "enumeration and classification of books." Critical bibliography, in contrast, is "the comparative and historical study of their makeup" (Besterman, 1940, p. 1). Besterman argues for a definition that combines elements of the two, labeling a bibliography "a list of books arranged according to some permanent principle" (Besterman, 1940, p. 2). He acknowledges the difficulty inherent in any idea of permanent principles, and questions surrounding permanency and change in texts themselves and the way we classify them is an issue to which we will return.

For the moment, though, the focus should remain on the contrast drawn by Besterman. Systematic bibliography is closely related to classification for access, which arranges texts according to some principle meant to facilitate retrieval. Critical bibliography, in contrast, bears a strong resemblance to editing in its concern for the origin and make-up of texts. In turn, both editing and classification for retrieval eventually point to a particular strand of material rhetoric. Material rhetoric, as a whole, addresses a range of questions surrounding the materiality of discourse, including topics like the rhetoric of public monuments and the rhetoric of the body. The strand of material rhetoric that concerns us here, though, studies the range of accretions, from prefaces to classificatory marks, that are attached to texts and affect the way those texts are used and interpreted. Bibliographic scholars have shown concern for similar issues.

While Besterman does not use the term "rhetorical," he does show us that from its earliest inception, bibliography has been rhetorical in the sense of being an attempt at persuasion. It is important to this paper to revisit the emergence of bibliography and describe how the problems it first addressed persist into current scholarship.

Besterman argues that the earliest bibliographies, which predate printing, had clear ideological goals. Authenticity seems to have been prominent among them. As early as the second century, Galen felt compelled to mention works falsely attributed to him (Besterman, 1940, p. 3). St. Jerome took pains to "show those he regarded as heretics that the Church had produced many able writers" (Besterman, 1940, p. 4). Besterman argues that "Jerome ... looked upon his compilation as a piece of theological propaganda. He did not put out his bibliography to guide or instruct, but rather to convert." He also says more generally that the writers of early bibliographies "... usually, if not always, had some ulterior motive ..." (Besterman, 1940, p. 9). While later scholars like Johann Tritheim in the fifteenth century and Conrad Gesner in the sixteenth did begin to consider bibliographies practical works of reference, earlier scholars deployed bibliography in service of religion. Even compilation of a bibliography or reference work, though, assumes some organizational or retrieval principle. We are perhaps not accustomed to considering these principles ideological, but when we reach the discussion of material rhetoric and the involvement of library research in composition and rhetoric, we will have strong reason to do so. Any attempt to persuade, whether of authenticity or relevance for a particular purpose, constitutes an important part of the horizon within which classification occurs.

The religious orientation of early bibliographies raises questions of their connection with hermeneutics, which began as the study of correct biblical interpretation. While there is no evidence here to directly connect bibliography with hermeneutics, bibliography clearly shared with early hermeneutics a sense of propriety in what was to be admitted to the canon. This sense of propriety and hewing to a sense of original or authoritative speech constitute part of what Jacques Derrida (1976) has called the metaphysics of presence, the idea that original thought enjoys primacy over speech, and speech over writing. The next section will describe ways in which bibliographic scholarship contributed to the progression of ideas that played a role in the deconstructive critiques of that tradition.

MODERN BIBLIOGRAPHY

While Besterman looks backward to the history of bibliography, these questions surrounding the social milieu in which bibliography takes place persist in more recent scholarship. Brookes (1975/1973) discusses Jesse Shera and Margaret Egan's attempts to build a theory of bibliography. Brookes's discussion of their work ranges widely, but their concern with social epistemology is of particular importance here as it relates to the idea of an intellectual horizon. Brookes quotes two definitions of social epistemology from Shera:

   ... the study of those processes by which society as a whole seeks
   to achieve a perceptive or understanding relation to the total
   environment--physical, psychological and intellectual.... (Brookes,
   1975/1973, p. 68. Emphasis in original)

   ... the analysis of the production, distribution and utilization of
   intellectual products in much the same fashion as that in which the
   production, distribution and utilization of material products have
   long been investigated. Graphic communication provides objective
   evidence of the process. (Brookes, 1975/1973, p. 68. Emphasis mine)

Brookes, referring to the second quote, suggests the substitution of "bibliography" for "graphic communication," and "cognitive elements" or "ideas" for "intellectual products."

Shera's linking of bibliography and social epistemology appears again when Brookes cites him as arguing, "Even a cursory examination of the history of classification of the sciences emphasized the extent to which any attempt to organize knowledge is conditioned by the social epistemology of the age in which it was produced" (Brookes, 1975/1973, p. 69). The relationship Shera posits between bibliography and social epistemology does not flow in only one direction, though. Shera implicates bibliography in the making of knowledge, arguing that it is "by the grouping and regrouping of his data that the scholar discovers new relationships, new approaches to old problems, and new areas for exploitation" (Brookes, 1975/1973, p. 69). Shera also observes, on a distinctly ontological note, that, as knowledge becomes more complex, it becomes subject to fragmentation and centrifugal forces. Bibliography serves as a cohesive counterforce.

For Shera, the study of these processes is an objective, and often quantitative, process. We should not construe his work as postmodern or critical. But Shera's references to social epistemology and specialization prefigure the critiques that have emerged since then. While these critiques have tended to come from fields not directly concerned with the study of information, we will still find shared concerns informing those critiques. It is important to note that even modernist scholars like Shera feel compelled to locate bibliographic activities in a social and epistemological context in much the same way as scholars who show more overt concern for the term "horizon" in its various forms.

At about the same time as Shera's work, Tanselle (1971/1975) discusses the connections between bibliography, textual studies, and literary judgment. This line of thought eventually leads to a discussion of deconstruction's role in such scholarship and forms another connection to material rhetoric. Tanselle was not uniformly supportive of deconstruction, but some of the authors who followed him have been. Despite Tanselle's reservations, though, his work represents another important piece of scholarship for this paper. He draws connections to social milieu and epistemology that are similar to Shera's, but which move further in the direction of deconstructive and hermeneutic critiques. The next section will discuss how these strands of scholarship began to come together in the work of Tanselle.

DECONSTRUCTION AND CLASSIFICATION

Tanselle begins with the assertion that "Bibliographers have been reminded on many occasions that their field is not an exact science" (Tanselle, 1971/1975, p. 353). He argues that editorial and bibliographic work can be scientific only in the sense of being systematic, methodical, or scholarly. He uses a number of examples to point to the inherent uncertainty in any decisions about what a particular author has said in a particular text. The examples show the ambiguity that occurs even in distinguishing accidental changes like typographical errors from actual alterations during the editing and type-setting process. Throughout the article, he emphasizes the need for literary judgment. While he primarily shows concern for textual studies, he clearly includes bibliography in his discussion.

Tanselle's doubts about bibliography's scientific status do not amount to advocacy of careless or subjective work, and he praises efforts to make editing and bibliography systematic. His argument, though, acknowledges the indeterminacy that remains inherent in these decisions. Like Shera, he also shows a concern for epistemology. On a distinctly epistemological note, he admonishes textual scholars to avoid trying to make their work appear scientific, because such a stance can accord those works "a higher level of certainty than they deserve and can thus affect any further thinking based on those conclusions" (Tanselle, 1971/1975, p. 354). Shortly later he points to two perspectives from which we can examine the relationship between literary judgment and textual studies, arguing "one is the effect which the findings of bibliographical and textual research have on the ultimate meaning of the work of literature as evaluated by the critic; the other is the role which critical judgment plays in producing these findings in the first place" (Tanselle, 1971/1975, p. 355).

Taken together, these two perspectives bear a strong resemblance to the hermeneutic circle, the idea that all knowledge depends, in part, on fore-knowledge that alerts us to salient features of the environment. This new knowledge, in turn, becomes part of our foreknowledge for future situations. The idea of a hermeneutic circle is also not far from Hans-Georg Gadamer's idea of intellectual horizons, an idea echoed by other hermeneutic thinkers in various forms. Hermeneutics began as the study of correct biblical interpretation but has expanded to include the study of interpretation more generally. Recent hermeneutic scholarship like Gadamer's has often focused specifically on the role played by an interpreter's membership in a community of ideas and expectations and how that horizon affects the reception of a text. Gadamer (1975) argued that the "horizon is the range of vision that includes everything that can be seen from a particular vantage point" (p. 269). Horizon, though, does not serve solely as a limiting principle. Gadamer (1975) maintained that "To acquire a horizon means that one learns to look beyond what is close at hand--not in order to look away from it, but to see it better within a larger whole and in truer proportion" (p. 272). This sort of horizon is an idea that emerges in work that follows Tanselle.

Tanselle quotes D. F. McKenzie in arguing that "'scientific' refers simply to 'an honesty of method'" (Tanselle, 1971/1975, p. 355). He goes farther, though, than arguing that bibliography uses soft scientific methods and argues for the inevitable role of judgment: "... when bibliographic analysis provides anything less than demonstrable certainty, literary judgment must necessarily take over" (Tanselle, 1971/1975, p. 361).

Tanselle (1974/1979) echoes and reemphasizes these ideas. He compares analytical bibliography to historiography, defining analytic bibliography as "a form of historical investigation; its conclusions are on a lower plane of probability than the inductive generalizations of many sciences because of the impossibility in bibliography of repeating past events as experiments ..." (Tanselle, 1974/1979, p. 13).

He also expresses impatience with any continuing debate over the scientific status of bibliography, saying, "... the last word on the subject would seem to have been said, and said repeatedly" (Tanselle, 1974/1979, p. 19).

These concerns with the role of judgment and epistemology become more explicit in Tanselle (1990/1998), where he explores the application of deconstruction to textual criticism. Tanselle is in partial agreement with deconstructionists, but takes them to task for an insufficient "understanding of the medium of literature" and a failure to question "the logic of accepting any sequence of words as it comes to us" (Tanselle, 1990/1998, p. 204). Tanselle's objection stems from his distinction between texts and works. He defines a text as a "specific oral or written sequences of words and pauses (or marks of punctuation)" (Tanselle, 1990/1998, p. 206). Works, in contrast, are the referents to which texts imperfectly point. He prefigures Jerome McGann's concern with reading as a phenomenal event by distinguishing between the stationary and sequential arts. Stationary arts, like painting and sculpture, use tangible media. Sequential arts (like literature and music) use intangible media, and we preserve them through instructions for their repetition (Tanselle, 1990/1998, p. 235). These instructions, though, are imperfect, and the phenomenal event cannot be taken as a strict re-creation of an author's intent.

Tanselle's distinction between texts and works puts him very close to a deconstructionist position. The question of whether we can faithfully represent speech in writing, and whether speech or original thought possesses primacy over writing lies at the center of Derrida's critique of the metaphysics of presence in Of Orammatology. Chronologically, Tanselle's 1990 article comes after Derrida, but conceptually it comes prior to it. This cluster of concerns, the (im-) possibility of representation, the role of judgment, and the indeterminacy of reading, serves as an effective introduction to the issues central to deconstruction. Later authors like McGann take up Tanselle's ideas and ally themselves with deconstruction and postmodern rhetoric. Tanselle was perhaps right to take deconstructionists to task for too frequently ignoring the material circumstances of reading, but his own ideas about indeterminacy and the impossibility of perfect representation actually lie very close to the ideas of other scholars who are willing to go farther than he did in praising deconstruction and similar postmodern thought. They also lie close to the hermeneutic idea of horizons. Hermeneutic and deconstructive thinkers have broad, at times even antagonistic, differences, but they also have shared concerns. Both show interest in interpretive shifts and indeterminacy in texts as readers within different settings or horizons encounter them. All of the arguments about the role of judgment and epistemology in literary criticism and bibliography will have little bearing on classification for access, though, unless they remain connected to activities like the construction of retrieval systems and library catalogs. Tanselle (1977/1979) maintains this connection and argues that descriptive bibliography and library cataloging "have many points of contact and many elements in common. Their history has been intertwined in many respects." He refers to them as being "parts of a larger undertaking--the recording of intellectual products and their physical embodiments" (Tanselle, 1977/1979, p. 37).

Tanselle argues for three sets of paired concepts that elucidate this connection:

* Works vs. texts, a distinction we have already discussed.

* Reference bibliography (a guide to a set of works) vs. physical bibliography (a description of the physical makeup of books).

* Enumeration vs. detail (Tanselle, 1977/1979, pp. 48-50).

The first distinction is central, with the second also being important, but the third less so, because it pertains mostly to the specific use to which a bibliography or catalog is put. Tanselle also distinguishes between catalogs, which are concerned with the books that happen to be in a particular collection, and bibliographies, which are concerned with books that are related in some way, but are not concerned with specific copies of those books (Tanselle, 1977/1979, p. 40).

The particular distinctions are not the most important ideas here, although Tanselle writes at length about the way the distinctions play out in different kinds of activities. What is important about Tanselle (1977/1979) is that it points to literary criticism, bibliography, and cataloging as related activities. He seems to argue that we can make no perfect distinction between physical and intellectual retrieval. Library catalogs show traces of physical bibliography, and enumerative physical bibliographies are created for some (perhaps ideological) purpose, frequently either retrieval or intellectual collocation. All of these concerns remain bound up with our efforts to understand not just texts, but works as a product of an author's intent.

While Tanselle takes the first steps toward a postmodern understanding of the way in which we organize and retrieve information, Jerome McGann adopts an overtly postmodern approach that draws on hermeneutics and deconstruction. McGann (1991) calls for what he terms a materialist hermeneutics (p. 15). While he follows a general hermeneutic orientation, he argues that "Reading appears always and only as text ... [and] reading itself can only be understood when it has assumed specific material constitutions" (McGann, 1991, pp. 4-5). He shares some of the doubt that Tanselle harbors about deconstruction but applies it to hermeneutics in criticizing the modern hermeneutical tradition for paying too much attention to the reading of texts and not enough to their making. He goes much farther than Tanselle, though, in adopting a postmodern orientation.

McGann follows figures like Derrida in arguing that texts do not provide simple representations of reality and shares in the more radical idea that such representation is unattainable. He argues that "The textual condition's only immutable law is the law of change.... [T]exts do not simply vary over time. Texts vary from themselves (as it were) immediately, as soon as they engage with the readers they anticipate" (McGann, 1991, pp. 9-10).

This indeterminacy across groups has implications for the organization of resources, and we will see later that recent scholarship in classification has begun addressing it.

As with many postmodern thinkers, we should not view McGann's ideas as something alien being grafted onto the study of bibliography and classification. McGann follows Gerard Genette in referring to paratexts, which include items such as footnotes and prefaces that do not constitute part of the original text (Genette's work predates McGann [1991], but is available in English as Genette [1997]). McGann extends the idea of paratexts to include bibliographic elements and refers to text as "a laced network of linguistic and bibliographic codes" (McGann, 1991, p. 13). We should also remember that his work draws directly from that of Tanselle, who very clearly was concerned with bibliographic work in its various forms.

McGann's work is important for that connection to bibliography alone, but it leads to other salient issues, as well. He posits an editorial horizon, the social conditions that surround the production or reproduction of a text and constitute the context in which these actions occur. The works discussed so far have pointed, in various ways, to the idea of an intellectual horizon. Besterman points out the persuasive intent present in some bibliographies. Both Tanselle and McGann compare bibliography with the literary judgment involved in editing. Shera shows concern for epistemological context in the production of ideas. When we take these together, we are not far from positing a classificatory horizon, the material and social context within which classificatory decisions are made and in which they have efficacy in shaping discourse. McGann gives no concise definition of what he means by "horizon" but argues that "texts are produced and reproduced under specific social and institutional conditions, and hence ... every text, including those that may appear to be purely private, is a social text. This view entails a corollary understanding, that a 'text' is not a 'material thing' but a material event or set of events, a point in time (or a moment in space) where certain communicative interchanges are being practiced (McGann, 1991, p. 21).

If one substitutes "classification" or "classificatory act" for "text," they will have a workable definition of what is meant in this paper by "classificatory horizon."

Unless we wish to disregard the concern for social epistemology, cultural conditions, and the inevitably interpretive elements of bibliography raised by the authors discussed so far, the idea of a classificatory horizon is, if not inevitable, certainly worth considering. It is no stretch to say that classification for retrieval affects the utilization of material products referred to by Shera through its constitution of some of the paratexts referred to by scholars like McGann. These notions of horizon, i.e., McGann's and the one expressed here, are also not far from ideas of horizons, interpretive traditions, and shifts in textual meaning that figure prominently in hermeneutics and deconstruction. The idea of a classificatory horizon finds support in recent rhetorical scholarship, as well.

CLASSIFICATION AND MATERIAL RHETORIC

Similar concerns over horizons, communities, and other bounds placed on discourse by classification also show up, even though sometimes indirectly, in rhetorical scholarship. The connection is most clear in material rhetoric, but other rhetoricians show varying levels of concern with classification for access.

Bruffee (1997) focuses on collaborative learning in American colleges but touches on questions with implications for the study of classification and retrieval. Following a number of other textual scholars, Bruffee observes that "Writing is a technologically displaced form of conversation" (Bruffee, 1997, p. 400). Whether we call it displaced speech, displaced communication, or something else, this idea of displacement from one discourse community or horizon to another fits well with the concerns expressed previously. In being displaced and stored for later use, any text becomes a potentially classifiable object, and the way it is classified affects its later use. While referring specifically to the teaching of English, Bruffee's statement that such work "is one way of introducing students to the process by which communities of knowledgeable peers create referential connections between symbolic structures and 'reality'" (Bruffee, 1997, p. 410) could easily be a description of classification as used in academic libraries.

Bruffee (1997) focuses on the rise of collaborative learning as it grew in response to the influx of college students less well-prepared for academic life than their predecessors. Similar concerns have been voiced about learning bibliographic systems. Metcalfe (1976), in tracing the evolution of modern information retrieval from early bibliography and classification, argues that "One trouble in the past has been that bibliographies have differed widely in arrangement, each of which has had to be mastered before consultation can be relied on" (Metcalfe, 1976, p. 24). So when Bruffee argues that "collaborative learning inducts students into established knowledge communities and teaches them the normal discourse of those communities" (Bruffee, 1997, p. 409), we should remember that an important part of that discourse is bibliographic.

Other rhetoricians have referred more overtly to connections with libraries and archives. Berlin (1987) refers to the use of library tools such as periodical indexes for research papers that found their way into the current-traditional rhetoric curriculum. While Berlin assigned the research paper a relatively minor role in this development, he linked it to current-traditional views about the creation of meaning, arguing that "the research paper represented the insistence in current-traditional rhetoric on finding meaning outside the composing act, with writing itself serving as a simple transcription process" (Berlin, 1987, p. 70). Library research creates this connection to a larger context outside the text through classificatory systems that alternately highlight or obscure works. These processes are most directly addressed in strands of material rhetoric that address the physical production and dissemination of texts.

Scholars of material rhetoric have been the most insistent in linking the work of classificationists and archivists to that of rhetoricians. Sharer (1999) suggests, "fundamentally, that we expand our professional responsibilities into realms previously marked off as the territory of library and information science" (p. 136). She expresses great concern over the deterioration of physical resources in libraries and archives and argues for a greater voice for rhetoric scholars (and others) in deciding what is preserved and what is weeded out or allowed to deteriorate as an "accepted loss" (Sharer, 1999, p. 125).

Sharer's arguments are simple, but telling. She argues that writing a historical account constitutes an act of power that depends on "previous acts of power that configure the physical and material conditions of historical research" (Sharer, 1999, p. 120). She expresses concern with Louis Willard's idea that "The most reliable predictor of later use is past use" (Sharer, 1999, p. 126), pointing out that "Description and indexing practices ... establish and perpetuate cultural and social values by allowing only certain materials to become visible to researchers, while obscuring others" (Sharer, 1999, p. 128). Past use, then, is, in part, a product of the social values that guided classificatory decisions, and those values may haunt future research with their seeming objectivity. This argument bears strong similarity to McGann's assertion of the editorial horizon and to Shera's concern with social epistemology. It also serves as an example of how classificatory horizons might show their effects in practice. Not only does classification occur within an intellectual horizon, in a striking example of the hermeneutical circle, it also helps constitute that horizon.

Tollar Collins (1999) also draws on McGann's work to develop a methodology for material rhetoric, which she defines as "the theoretical investigation of discourse by examining how the rhetorical aims and functions of the initial text are changed by the processes of material production and distribution" (p. 547). As McGann refers to paratexts, Tollar Collins refers to rhetorical accretions, which accumulate in a "process of layering additional texts over and around the original text...." (Tollar Collins, 1999, p. 547).

In Tollar Collins's work, we find a familiar nexus of concerns. While she does not refer explicitly to classification, she alludes to bibliographical studies aimed at establishing authoritative versions of works (Tollar Collins, 1999, p. 549). She also talks about how the horizon of expectations can shape, and be shaped by, a work's rhetorical functions (Tollar Collins, 1999, p. 548). This notion of horizon tracks well with McGann's notion of the editorial horizon, the idea of a classificatory horizon posited in this paper, and hermeneutic notions of interpretive horizons.

The three traditions--bibliographic scholarship, rhetoric, and hermeneutics--have converged on a strikingly similar set of ideas about horizons and context. Another of this paper's contentions is that even this relatively brief review shows that these fields bear the marks of common wellsprings, and the tendency to consider them completely separate fields of endeavor is a mistake. Their respective ideas are strengthened by the presence of similar conclusions in the other fields under discussion. We also will see in the next section that scholars in classification have reached similar ideas about horizons by different paths.

THE STUDY OF INFORMATION

While this paper argues strongly for a set of shared intellectual origins, notably, contemporary scholars of information have come to similar conclusions by different means. Buckland (1991) posits information-as-thing--a broad term including data, documents, and objects taken to have informative value (p. 351). He calls on the work of European documentalists such as Paul Otlet and Suzanne Briet in defining a broad range of objects as potentially informational. He uses the example of antelope-as-document from Briet's work, arguing that an antelope in the wild is not a document, but an antelope in a zoo might well be considered one. However, Buckland argues against such an animal being discourse:

   Perhaps a better term for texts in the general sense of artifacts
   intended to represent some meaning would be "discourse".... However,
   we could hardly regard an antelope or a ship as being
   "discourse".... Their value as information or evidence derives from
   what they signify about themselves individually or, perhaps, about
   the class or classes of which they are members ... they could be
   viewed as representative. If an object is not representative of
   something, then it is not clear how far it can signify anything,
   i.e., be informative." (Buckland, 1991, p. 355, emphasis and quotes
   in original)

This raises questions of representation, but the disposition of those questions is not what should concern us here. What is important is the placement of artifacts at the center of discussion about information. Information, then, comes to us borne by a contextualized thing, and this constitutes a conclusion strikingly similar to the ones reached by Tanselle, Shera, McGann, Sharer, and Tollar Collins.

Buckland (1997) makes this similarity more apparent. He returns to Briet's definition of documents and infers four rules she used to determine when something is a document: the objects must be material; they must be intended as evidence or documentation; the objects must be processed; and there is a phenomenological position that the object is perceived as a document (Buckland, 1997, p. 806). He also refers briefly to a suggestion that Briet saw documentary status as stemming from indexicality, "the quality of having been placed in an organized, meaningful relationship with other evidence" (Buckland, 1997, p. 806).

Buckland makes an additional comment near the end of the paper, which bears even greater similarity to the thought of material rhetoricians and breaks with the modernist leanings of scholars like Otlet. He argues that "one difference between the views of the documentalists discussed above and contemporary views is the emphasis that would now be placed on the social construction of meaning, on the viewer's perception of the significance and evidential character of the documents" (Buckland, 1997, p. 807).

Social construction of meaning comes very close to the concept of horizon by calling attention to the role of context in meaning and communication, bringing Buckland even closer to scholars like McGann.

Albrechtsen and Jacob (1998) raise a similar set of concerns about scholarship in classification. They criticize both the one-size-fits-all universalism of rationalism, and empirical approaches that seek to compile large amountS of factual information about user actions (Albrechtsen and Jacob, 1998, p. 295). They call for a new approach based on social constructivism or historicism, which they argue can offer "a view of knowledge as a product of historical, cultural, and social factors, where the fundamental divisions and the fundamental concepts are products of the divisions of scientific/cultural/social labor in knowledge domains" (Albrechtsen and Jacob, 1998, p. 296).

They incorporate Star and Griesemer's idea of boundary objects, "objects that are both plastic enough to adapt to local needs and constraints of the several parties employing them, yet robust enough to maintain a common identity across sites. They are weakly structured in common use, and become strongly structured in individual-site use" (Albrechtsen and Jacob, 1998, p. 311).

Under this characterization of classification, classifiers become epistemic engineers who facilitate access across diverse user groups, rather than placing documents in a priori categories. This concern for the local rather than the purportedly universal, for the contingent rather than the permanent, bears strong similarity to deconstruction and contemporary hermeneutics in spirit if not in word and fits well with the previously discussed scholarship overtly concerned with horizons and meaning. It also answers Besterman's earlier concern over the position of permanent principles in the arrangement of texts. This very practical concern can find theoretical justification in these shared ideas.

Brown et al. (1996) employ the notion of boundary objects in a related criticism of the conduit metaphor. This metaphor posits information systems as transmission systems designed to impart information with a minimum of noise or distortion. In discussing documents as boundary objects, the authors employ language very close to the discussion of horizons, arguing "Documents that pass successfully between communities need to be able to engage (at least) two interpretive strategies and to survive where the recipients can no longer be assumed to share the interpretive assumptions of the members of the originating community" (Brown et al., 1996, p. 12).

In a different context, this statement could serve as an adequate recapitulation of Derrida's critique of the metaphysics of presence. It also corresponds well with McGann's ideas about textual change and lies not far from Gadamer's idea of merging horizons in human communications.

Unlike Albrechtsen and Jacob, Brown et al. (1996) do refer directly to rhetoricians, in this case pointing to Stephen Toulmin and Stanley Fish's writing on the social negotiation of meaning (Negotiating Meaning section, para. 2). Their focus falls on documents as defining properties of communities, and they address this issue in terms of the Web and other information technologies.

As with so much else in this paper, these scholars refer to each other's work, or take up problems that put them at a short remove from each other. These concerns over context and horizon began, in part, with the study of bibliography, and they now extend into our study of online information. These longstanding, shared concerns seem likely to remain.

CONCLUSION

From its beginnings, classification has been closely concerned with many of the same issues present in rhetoric. The early bibliographies which contributed to modern classification showed persuasive intent, and contemporary scholars of classification for access have developed critiques of their own field that track closely with the critiques put forward in rhetoric.

Rhetoricians, in turn, have shown both overt concern with the way classification affects discourse, as well as more subtle concern for the discursive environments into which documents enter and which they, in part, constitute. Certain material rhetoricians in particular have argued that the material conditions of textual production and dissemination shape the use and understanding of documents. This echoes the arguments by bibliographers that material conditions alter the ways in which we use texts to understand works.

This paper derives its importance from its attempt to begin unifying the trends represented by these scholarly questions, not in the way that individual authors attempt to resolve these questions. All of the authors refer to social context, and many do so in a way that at least obliquely connects them to the other fields. In the case of Tanselle and McGann, the connections are overt and easy to trace. McGann's idea of the editorial horizon has power here not just in similarity of word use, but in the fact that he derives his ideas, in part, directly from bibliography, hermeneutics, and textual studies. Some of the work in material rhetoric, in part, builds on McGann's work and also refers to classification.

A great deal of work in these related fields circles around the idea of horizon, expressed in various ways. This article has made a preliminary attempt to unify these notions under the term classificatory horizon and argues that the concept derives naturally from historical roots common to classification, rhetoric, and hermeneutics. The notion of a classificatory horizon is more than a simple metaphor. It represents a convergence of ideas from related fields that, taken together, can provide a theoretical framework for studying the rhetorical aspects of classification. This framework might eventually lead to a better understanding of the material and cultural limits that act on the representations in our classificatory systems.

We should pay less attention to the way the individual authors handle these questions and more attention to the fact that they all arrive at a similar set of concerns and conclusions that serve to strengthen each other. We should be surprised not by the similarities that we can find, but by the fact that these similarities don't receive more attention.

REFERENCES

Albrechtsen, H., & Jacob, E. K. (1998). The dynamics of classification systems as boundary objects for cooperation in the electronic library. Library Trends, 47(2), 293-312.

Berlin, J. A. (1987). Rhetoric and reality: Writing instruction in American colleges 1900-1985. Carbondale, IL: Southern Illinois University Press. For the Conference on College Composition and Communication.

Besterman, T. (revised 1940). The beginnings of systematic bibliography. 2nd ed. (1968 reprint) New York: Burt Franklin.

Brookes, B. C. (1975). Jesse Shera and the theory of bibliography. In V. J. Brenni (Ed.), Essays on bibliography (pp. 63-76). Metuchen, NJ: Scarecrow. (Reprinted from Journal of Librarianship, 5(4), pp. 233-245, 1973)

Brown, J. S., Duguid, P., & Dyson, E. (1996). The social life of documents. First Monday, 1(1). Retrieved September 27, 2003, from http://www.firstmonday.dk/issues/issue1/documents/index.html.

Bruffee, K. A. (1997). Collaborative learning and file "conversation of mankind." In V. Villanueva (Ed.), Cross-talk in comp theory: A reader: Urbana, IL: National Council of Teachers of English. (Reprinted from College English, 46, pp. 635-652, 1984)

Buckland, M. K. (1991). Information as thing. Journal of the American Society for Information Science, 42(5), 351-360.

Buckland, M. K. (1997). What is a "document"? Journal of the American Society for Information Science, 48(9), 804-809.

Derrida, J. (1976). Of grammatology. Baltimore: Johns Hopkins University Press.

Gadamer, H. G. (1975). Truth and method. New York: Seabury Press.

Gadamer, H. G. (1976). On the scope and function of hermeneutical reflection. In D. E. Linge (Trans. & Ed.), Philosophical hermeneutics. Berkeley: CA: University of California.

Genette, G. (1997). Paratexts: Thresholds of intepretation. New York: Cambridge.

McGann, J. J. (1991). The textual condition. Princeton, NJ: Princeton University Press.

Metcalfe, J. W. (1976). Information retrieval, British and American, 1876-1976. Metuchen, NJ: Scarecrow Press.

Selzer, J., & Crowley, S. (1999). Rhetorical bodies. Madison, WI: University of Wisconsin Press.

Sharer, W. B. (1999). Disintegrating bodies of knowledge: Historical material and revisionary histories of rhetoric. In J. Selzer, & S. Crowley (Eds.), Rhetorical bodies. Madison, WI: University of Wisconsin Press.

Tanselle, G. T. (1971). Textual study and literary judgment. Papers for the Bibliographic Society of America, 65, 109-122. (Reprinted in Essays on bibliography, pp. 353-364, by V. J. Brenni, Ed., 1975, Metuchen, NJ: Scarecrow)

Tanselle, G. T. (1974). Bibliography and science. Studies in Bibliography, 27, 55-89. (Reprinted in Selected studies in bibliography, pp. 1-35, by G. T. Tanselle, 1979, Charlottesville, VA: Bibliographic Society of the University of Virginia)

Tanselle, G. T. (1977). Descriptive bibliography and library cataloguing. Studies in Bibliography, 30, 1-56. (Reprinted in Selected studies in bibliography, pp. 37-92, by G. T. Tanselle, 1979, Charlottesville, VA: Bibliographic Society of the University of Virginia)

Tanselle, G. T. (1990). Textual criticism and deconstruction. Studies in Bibliography, 43, 1-33. (Reprinted in Literature and artifacts, pp. 203-235, by G. T. Tanselle, 1998, Charlottesville, VA: Bibliographic Society of the University of Virginia)

Tollar Collins, Vicki (1999). The speaker respoken: Material rhetoric as feminist methodology. College English, 61(5), 545-573.

Stephen Paling, Department of Library and Information Studies, School of Informatics, University of Buffalo, State University of New York, 534 Baldy Hall, Buffalo, NY 14260-1020