edited by Birger Hjørland and Claudio Gnoli


Library classification
Part 1: Introduction and premodern classification

by Thomas M. Dousa

Table of contents:
1. Introduction
    1.1 Semantic and structural features
    1.2 Notation
    1.3 Function
    1.4 Structure and scope of this article
2. Premodern library classification
    2.1 The Ancient World: Egypt and the Ancient Near East
    2.2 The Ancient World: Greece and Rome
    2.3 The Ancient World: China
    2.4 The Middle Ages: Europe: 2.4.1 Medieval classifications of knowledge; 2.4.2 Medieval library classifications
    2.5 The Middle Ages: the Islamic world
    2.6 From the Renaissance to the late 19th Century: early Modernity: 2.6.1 Tradition and innovation: Conrad Gesner’s bibliographical classification; 2.6.2 The faculty system; 2.6.3 The early modern French tradition from Garnier to Brunet; 2.6.4 The Baconian tradition

Current systems of library classification are tributary to a long and complex process of historical development that extends back to the earliest literate civilizations. By considering the longue durée of library classification as a practice, one can achieve a better understanding of the forces that have shaped, and continue to shape, the landscape of library classification today. This two-part article considers the different historical phases of development in library classifications, adopting a three-stage historiographical framework that distinguishes between “pre-modern” classifications created and used before the rise of modern library classification theory (Third Millennium BC up to ca. 1875 AD); “modern” classifications created in the period between the rise of librarianship as a profession and the development of a discourse of library classification theory in the final quarter of the nineteenth century and the advent of large-scale use of digital computing in libraries in the 1980s; and “post-modern classification”, which encompasses both practical tendencies associated with the shift to an online environment in libraries and the development of new theoretical trends within the discourse of library classification since the 1980s. The first part of this article surveys premodern library classifications from the Ancient Near East, Ancient China, the Greco-Roman Mediterranean world, Medieval Europe, the Medieval Islamic world, and Early Modern Europe and North America.

1. Introduction

A → classification is “a series or system of classes arranged in some order according to some principle or conception, purpose or interest, or some combination of such” (Bliss 1929, 143). Enunciated by a theorist of classification whose writings have exerted a decisive influence on the field of → knowledge organization (KO), this definition of classifications posits that they issue from two processes. On one hand, the creation of a classification requires the generation of a set of classes into which a given universe of things can be partitioned in such a way that things sharing certain characteristics deemed salient are brought together into a class and distinguished from those things not sharing those characteristics (Broughton 2015, 7; Hulme 1950 [1911–1912], 3; Sayers 1955, 3; Svenonius 2000, 10). On the other, it involves the ordering of these classes in a series, the sequence of which is determined according to some principle of organization (cf. Broughton 2015, 8–9; Kaiser 1911, § 99, s.v. “to classify”). The resultant system of classes constitutes a classification scheme (cf. Broughton 2015, 378, s.v. classification scheme). Schemes of classification are the products of intellectual operations and so are conceptual in nature (Sayers 1955, 8; cf. Hjørland 2017, 99). In literate societies, they are often given stable and communicable semiotic form as written texts, as diagrams, and, perhaps most prototypically, as ordered lists of terms denoting classes, known as tables or schedules (Sayers 1955, 11). Such graphic representations of classifications function as guides to the conceptual structure of their respective classification schemes and as aids for those persons engaged in assigning objects to classes within the framework of a given scheme — an activity known as classing (Bliss 1929, 143; Kaiser 1911, § 104, s.v. “Classing”). Classifications may also find expression in the physical arrangement of the objects being classed, though this is not, by any means, obligatory.

Classifications can be created with respect to any domain of objects (Hjørland 2017, 98). As its name implies, library classification takes its domain to be the kinds of objects collected and made available in libraries. The conceptualization of this domain has shifted somewhat over time. Historically, the book was the prototypical kind of object kept in a library [1]: thus it is unsurprising that, in the past, writers frequently referred to library classification as “book classification” (e.g., Brown 1898, 39; Hulme 1950 [1911–1912], Richardson 1901, 153; Sayers 1955, 41) [2], even though, in addition to books, library collections typically included other kinds of “graphic records”, such as “manuscripts, documents, maps, [and] prints” (Sayers 1955, 2) [3]. Over the course of the Twentieth and early Twenty-first centuries, many other kinds of objects, both physical and virtual, such as audiovisual and/or computer-mediated materials, have been incorporated into library collections, so that today it is common to think of libraries as offering access to “resources” or “information resources” (Joudrey, Taylor and Miller 2015, 4; 991, s.v. “Resource”) rather than simply to books or → documents. Library classifications thus cannot be regarded tout court as classifications of books. Nevertheless, because so much of the historical discourse about library classification has centered around books and because they continue to be the core element in most library collections, they shall serve as the primary point of reference in the following discussion.

[top of entry]

1.1 Semantic and structural features

As noted earlier, classifications are composed of classes based on certain salient characteristics of the objects being classified. In the case of books — and, mutandis mutandis, other information resources — , there is no lack of characteristics that can serve as the basis for the definition of classes. To cite but a few examples, classifications of books can, in principle, be constructed around their physical size; the nature of their binding; their date of publication; their place of publication; the names of their authors or, should these be lacking, their titles; the subjects whereof they treat; the genres of writing that they represent; the languages in which they are written; the audiences for which they are written; or their provenance (cf. Bliss 1939, 23; Cutter 1904, 15; Hulme 1950 [1911–1912], 4–5, with Dousa 2017; Richardson 1901, 60–65). Historically, however, library classifications have tended to be based on a limited set of bibliographic attributes — namely, the → subject content and/or literary form of books (Bliss 1939, 24; Cutter 1904, 16; Hulme 1950 [1911–1912], 5; Mills 1968, 3). For this reason, they are sometimes referred to as “subject-classification[s]” (Bliss 1939, 25) or “subject classification schemes” (e.g., IFLA Working Group on Guidelines for National Bibliographic Agencies 2012, 22; Hider 2012, 162–170). Insofar as the subjects of books constitute objects of knowledge, some commentators have also characterized them as “knowledge organization classification systems” (Beghtol 2010). Yet if subject content and literary form have tended to be the primary foundations upon which library classifications have been built, one should not overlook the fact that, very often, some other attributes, such as the names of authors and date of publication, have served as principles for the sub-arrangement of books and other items within classes (cf. Sayers 1955, 14) and that physical features, such as size, have often governed the actual arrangement of books on shelves (cf. Brown 1898, 102–103; Mills 1968, 2). Very few, if any, library classifications have been, in practice, pure subject classifications.

Although there is considerable variability across library classifications with regard to the choice and definition of classes, as well as to the articulation of the relations between them, they share certain common features that are worth noting at the outset. At a minimum, every such classification consists of an array of very broad classes that, in principle, cover the entire universe of subjects to which it pertains. In the limiting case, this single array of classes constitutes the sole set of classes within the classification: such schemes are known as constitutive classifications (Dousa 2014). Most often, however, the classes belonging to this initial array are subdivided into arrays of subclasses which, in turn, may be divided yet further: they thus constitute the main classes (Sayers 1955, 11) of a hierarchically organized classification scheme. Different library classifications may have different degrees of hierarchical depth, some confining themselves to relatively shallow hierarchies with general subjects while others have deep hierarchies yielding classes for relatively precise subjects at their bottommost subdivisions [4]. The hierarchical relationships obtaining among the classes of library classifications of library classifications typically include genus-species, or generic relationships; kind-instance, or instantive relationships; whole-part, or partonomic, relationships; perspectival, or aspect, relationships; and topic-subtopic, or conventional relationships (Svenonius 1989, 39–40; 2000, 151–152) [5]. Some commentators have claimed that the genus-species relationship should be seen as the prototype for hierarchical relationships in library classifications (e.g., Sayers 1955, 26–32): however, for reasons that will become apparent in the second part of this article, it is, in fact, perspectival and topic-subtopic relationships that have tended to be especially prominent in many library classification schemes.

Another salient feature of library classifications — one which flows from the nature of classification itself — is the ordering of classes in sequence. Theorists of library classification have identified a host of different principles by which classes may be arranged, ranging from various semantically motivated orderings to strictly practical and conventional ones (Broughton 2015, 45–46; Ranganathan 1967, 183–197; Richardson 1901, 65–67). Semantically motivated orderings are considered especially appropriate and desirable for subject classification schemes. In particular, the sequence of main classes within a given classification often is based, explicitly or implicitly, upon the theory of knowledge held by the person designing it (Sayers 1955, 17) or, at the very least, upon an understanding of how the universe of knowledge is structured (Miksa 1992). Yet library classifications are rarely, if ever, theoretically pure in the sequential organizations of classes that they use: in practice, a single classification scheme is likely to use different principles of ordering at different locations within its hierarchical structure (Richardson 1901, 65–67).

[top of entry]

1.2 Notation

Whatever the principles of order chosen within a given classification, designers of library classifications have often sought to indicate the sequence of classes within their schemes in a semiotically efficient and perspicacious manner. The device used for this purpose is → notation — “a series of symbols, or shorthand signs … for the terms in [a] classification” (Sayers 1955, 53; for full discussion, see Gnoli 2018). Generally, the notations of library classifications have tended to consist of alphabetical, numerical, or mixed alphanumerical signs assigned to the (names of) classes. The position of a notational symbol in the sequence of alphabetical, numerical or alphanumerical characters to which it belongs corresponds to the position of the class to which it refers in the sequence of classes within the classification scheme in question. Such a symbol constitutes the classmark (Broughton 2015, 378, s.v. “classmark”; Sayers 1995, 53) or class number (Mann 1930, 114). This, in turn, may be further subdivided by an additional code to indicate the specific book falling into the class in question: this is known as the book number (Broughton 2015, 378, s.v. “book number”; Mann 1930, 115). Combining the classmark with the book number yields the call number which serves as the full notational designation for the book in question (Mann 1930, 115).

Theorists of library classification have identified various desired qualities of notation, such as, for example, simplicity, brevity, expressivity (i.e., the capacity to indicate hierarchical and other relations of classes in the structure of classmarks), mnemonicity (i.e., the capacity to correlate directly a classmark with its subject in a manner that makes it easy to remember), hospitality (i.e., the capacity to allow for the intercalation of new classmarks within the sequence of classmarks as needed), and, if possible, pronounceability (Beghtol 2010, 1056; Gnoli 2018, §4; Mills 1964, 39-48; Sayers 1955, 55–63). Needless to say, some of these desired qualities may conflict with one another — for example, to achieve expressivity in a hierarchically deep classification, one may have to forgo brevity — and so designers of library classifications must decide which of these qualities to foreground in the construction of notational schemes. Generally, it is agreed that hospitality is of especial importance if a library classification is to be able to accommodate new subjects over time (Sayers 1955, 58) and much ingenuity has been expended in elaborating methods of making schemes notationally hospitable (e.g., Mills 1961, 41–47; Ranganathan 1967, 309–323).

[top of entry]

1.3 Function

Historically, library classification schemes have been put to two primary uses. First, they have served as the basis for the physical arrangement of books and other information resources on library shelves — that is to say, as mechanisms for what is sometimes designated shelf, book, or bibliothecal classification (Satija 2000, 222; Sayers 1955, 5). Second, they have functioned as a means of organizing entries for works within library catalogs — that is to say, as a foundation for what is sometimes called catalogue classification or bibliographical classification (Satija 2000, 222; Sayers 1955, 5) [6]. The use of library classification schemes for shelf classification has imposed certain constraints upon the ways in which they have been conceptualized. Most notably, the fact that books are typically arrayed in sequence upon shelves has encouraged the understanding of library classifications used as shelf classification as lines composed of points in which “each particular book represents a specific point in [the] line” (Shera 1965, 97; cf. Sayers 1955, 5–6; 19). Corresponding to this assumption has been another according to which “a library classification is a sequence (a one-dimensional, linear order) of terms [denoting classes — TMD]” (Mills 1964, 41, italics his) in which each of the subject or form classes composing a given library classification scheme will occupy a unique position within the sequence of classes (Shera 1965, 98). Linearity and uniqueness may be termed postulates of positional rigidity and affect the classing of books and other resources within a collection: as one commentator has observed, “[s]helf classification can show a book in one place, at one subject, in one relation, and that only” (Sayers 1955, 19, italics in original; cf. Shera 1965, 99–100). The use of library classification schemes in classified catalogues is not as strongly subject to these postulates since a catalog entry for a single book can, in principle, be placed under multiple class headings: moreover, conceptually and formally distinct parts of a single book (such as chapters of a monograph or component essays in an anthology) can receive their own entries in a classified catalogue (Richardson 1901, 83; Sayers 1955, 6; 195–196), a practice known as analytical entry (Mann 1930, 152, s.v. “Analytical entry”). Both modes of classification constitute, each in its own way, important mechanisms for facilitating retrieval of, and hence access to, books and other resources in libraries: shelf classifications allow for the physical collocation of books on like subjects and so render possible browsing at library shelves, while catalogue classifications collocate catalog descriptions of bibliographical items — not only books but parts of books — that treat of the same subject and so pave the way for a more comprehensive collection of resources pertinent to a topic.

[top of entry]

1.4 Structure and scope of this article

These, then, are some of the basic structural and functional features of library classifications. The study of such classifications has long been a subject of great interest within KO: indeed, it is not too much to say that the tradition of sustained theoretical discourse about library classification that has developed since the last quarter of the 19th century has exercised a decisive influence upon the theoretical discourse in KO (cf. Hjørland 2008; 2016). It is thus fully justifiable to treat library classifications within the framework of an encyclopedia of KO. Now there are various ways in which one can approach the subject of library classification: generally speaking, one can examine the essential features of library classifications and the different kinds of theories underpinning their design (i.e., a systematic approach) (e.g., Beghtol 2010) or one can trace the development of library classification over time (i.e., a historical approach) (e.g., Miksa 1994). This introduction has sought to provide a summary sketch of library classifications as systems: the approach that shall be employed in subsequent sections will be historical, on the grounds that it is only through consideration of the longue durée of library classification as a practice that one can come to understand some of the forces that have shaped, and continue to shape, the landscape of library classification today.

The remainder of this essay is articulated as follows. First, we shall take a bird’s-eye view of the development of library classification from ancient times until the rise of modern library classifications and classification theory in the last quarter of the 19th century: this vast expanse of time constitutes what may be called the premodern period of library classification. Then, we shall proceed to consider some of the high points of library classification from the last quarter of the nineteenth century until, roughly, the rise of online library catalogs and systems in the 1980s: this period constitutes what may be called the era of modern classification [7]. Finally, we shall consider some of the trends that have arisen in the wake of potent new technological and cultural forces from the 1980s to the present day, an age that has witnessed a partial turn toward what may be termed postmodern classification. This triadic periodization of the history of library classification may well seem simplistic to some readers and it doubtless does not do justice to the complex historical developments within each of the three periods, much less to the continuities perduring across them. However, it draws upon a convention that has deep roots in general historiography (cf. Breisach 1994, 139–142; 181; 210–213; 272–274); it imposes upon an otherwise chaotic mass of historical data a chronologically-bounded high-level order that is open to further articulation and refinement; and the rough chronological boundaries it draws appear to correspond grosso modo with sea-changes in the social, cultural, and technological conditions within which library classifications were designed and used. It is thus a blunt but serviceable tool with which to get an initial purchase on our subject [8].

One more limitation must be acknowledged at the outset regarding the scope of coverage. Ideally, a complete history of library classification would cover the developments in this field in all the major cultures of the world from their beginnings to the present day, taking into account not only the internal history of library classification as a field in all its various dimensions but also its relations to the particular history of each different society in which it has been practiced. Such a “total history” probably lies beyond the capacity of any single individual to write [9]: it most definitely cannot be captured within the compass of an encyclopedia article and, accordingly, it is necessary to be selective here. This article will train its focus primarily upon general library classifications as they have developed in Western Europe and North America, while casting brief sideway glances at a few other major cultural areas for comparison [10]. The reason for this choice is that North American and Western European library classifications have had the greatest effect upon the development of theoretical discourse about library classification in KO and this discourse cannot be properly understood without reference to them. Even with this circumscription, the following discussion will perforce have to be evocative and allusive, touching upon its subjects with the broadest of brush strokes and confining itself to the cursory treatment of a handful of representative examples. It is hoped that, even with these limitations, it will prove useful as a point of entry for considering the history of library classification and its implications for the present and future.

[top of entry]

2. Premodern library classification

For our purposes, the premodern era can be divided into three periods: Antiquity, extending from the beginning of literate civilization to about the end of the 5th century AD; the Middle Ages, spanning the millennium approximately from the 6th through middle of the 15th century; and Early Modernity, covering the span from the later 15th century to the final quarter of the 19th century [11]. Significant developments in library classification occurred within each of these periods. In the Ancient World, the earliest attested library and bibliographical classifications developed certain features, such as a tendency towards classification by subjects and literary form, that have continued to characterize library classifications to this day (see sections 2.1–2.3 below). In China, some of the canonical forms of bibliographical classification, based on a distinct philosophical perspective, were laid down in Antiquity (see section 2.3 below). In the Mediterranean world, by contrast, there is little evidence to suggest that Greek and Roman compilers of bibliographic works and book lists drew upon the philosophical and pedagogical classifications of knowledge that were current in Hellenistic and Roman times (see section 2.2 below). The Middle Ages mark an important stage in the history of library classification. On one hand, classifications of knowledge, largely derived from Greco-Roman philosophical and pedagogical models, began to be incorporated, albeit generally in simplified form, into the structure of library classifications (see sections 2.4.1–2.4.2 below); on the other, some of the technical features of classification-making, such as the use of notation, are first attested in the latter half of this period (see section 2.4.2 below). The Early Modern period witnessed both the consolidation of certain traditional classificatory patterns that had first taken form in the Middle Ages (see, e.g., sections 2.6.1–2.6.2 below) and the creation of new classificatory patterns that reflected, in lesser or greater measure, developments in contemporary classifications of knowledge (see, e.g., sections 2.6.3–2.6.4 below). These latter, in turn, provided the background of thought and practice within which modern library classification would emerge.

[top of entry]

2.1. The Ancient World: Egypt and the Ancient Near East

Over the last century and a half, the patient labors of archaeologists and philologists have uncovered a considerable body of evidence for the existence of archives and libraries in ancient Mesopotamia, Syria, Anatolia (e.g., Hout 2002; Max 1994, esp. pp. 26–29; Potts 2002; Pedersén 1998; Robson 2013), and Egypt (e.g., Burkard 1980; Webb 2013). In these early civilizations, libraries were associated primarily with temples and royal palaces — that is to say, religious and administrative centers — , though it is clear that members of the small proportion of the population that was literate kept private collections of writings for their own use (cf. Blumenthal 2011; Parpola 1983, 8–10; Webb 2013, 27–29). Numerous examples of the books kept in these ancient libraries — papyrus and leather book-rolls in Egypt; clay tablets and wax-covered wooden boards in Mesopotamia — have survived but very little is known regarding the principles of classification that were used to organize them (cf. Šamurin 1964–1967, 1/6–12; Weitemeyer 1956, 231–232). There are, however, a few scattered pieces of evidence that offer some light, however faint and uncertain, on this subject.

The best preserved of the exiguous Egyptian sources for library classification is a list of books carved on the walls of the room serving as the library of the temple of Horus in Edfu [12]. Dating to the reign of Ptolemy VIII (145–116 BC) and forming part of the textual component of a wall decoration depicting ritual scenes, this list is quite short [13] and most likely represents the contents of a select reference collection of works consulted for purposes of temple administration and for carrying out temple rituals (Kurth 1994, 141; Wessetzky 1973, 55). The list is divided into two sections, each associated with a different ritual scene: one lists books relating to temple organization and administration, apotropaic rituals (i.e., rituals to ward off evil), astronomy, and religious geography, while the other enumerates books containing the texts of various apotropaic rituals performed as part of the temple cult (Grimm 1989, 159–161; Wessetzky 1984, 86). Although there is some overlap between the two lists, it is apparent that each has a distinctive emphasis: one focuses primarily on handbooks for the different kinds of knowledge — administrative, astronomical, and theological — necessary for keeping the temple and its rituals in operation, whereas the other is devoted primarily to texts of a ritual nature designed to keep evil at bay (Wessetzky 1958, 159). The bipartite list thus seems to have been organized — at least implicitly — on the basis of the function and, to some degree, the subject content of the texts enumerated. Some Egyptologists refer to the list as a “catalog” (e.g., Grimm 1989; Wessetzky 1958, 158–160; 1974, 55): however, it is not at all obvious that it actually had this function or that its structure reflects the physical organization of the temple library’s collection. Despite the many uncertainties that hedge interpretation of the Edfu book list, its form reflects an underlying classificatory structure — one that appears to have been suited to the needs of a working temple library. Whether this classificatory structure governed the organization of other Egyptian temple libraries is unclear. The fragments of a comparable monumental book list found at the Ptolemaic-period temple at Tod suggests that both the content of the collection and its mode of organization differed substantially from those of Edfu (Grimm 1989; Thiers 2004): most likely, the particular requirements of local cults governed the classification of books in individual temple libraries.

Evidence for library classification is also sparse for ancient Mesopotamia and the Levantine world, though intriguing hints can be gleaned from various finds. There is some evidence that, at certain times, the shape of the clay tablets on which texts were written were correlated with different genres of text. For example, the palace archive-cum-library at the northern Syrian site of Ebla (present-day Tell Mardikh), which dates to the mid-third millennium BC, contained both round and oblong tablets, with the former used to record texts relating to economic and administrative matters and the latter, all other genres of texts (Wellisch 1981, 491). Interestingly, these two kinds of tablets were also kept apart in the storerooms in which they were housed on wooden shelves: the round tablets were kept on shelves close to the floor, while the oblong ones were located on higher shelves, with tablets containing reference works, such as dictionaries, glossaries, and cuneiform sign syllabaries, kept in a location distinct from other genres (p. 495). On a somewhat different front, an analysis of Sumerian-language catalogs found at various locations and dating to the second half of the second millennium and the first half of the second millennium BC has revealed that, whereas some catalogs appear to have followed no discernible order in their enumeration of works, others seem to have adopted principles of organization reflecting various aspects of the contents of their works — most notably, four hymn-lists organized the works to which they referred according to musical form and the identity of the god addressed — or their location within the collection in which they were kept (Dalby 1986, esp. 478–480; 482–483). Taken together, the evidence from Ebla and the Sumerian catalogs reveals that Mesopotamian catalogue classification involved classification by genre and subject and, that, in some cases, catalogue classification mirrored shelf classification.

Perhaps the most extensive piece of evidence for library classification in Mesopotamia is found in three cuneiform tablets from the great palace library of the Assyrian king Assurbanipal (668–627 BC) at Nineveh recording the acquisition of new books for the library in the wake of that king’s repression of a rebellion in Babylon in 647 BC (Parpola 1983, 9–10; Potts 2002, 25). These acquisitions included both literary works and compositions embodying Mesopotamian scientific and religious lore [14]. In the library’s records, the new books were identified and categorized either by title (in the case of literary texts) or by the kind of lore that they contained (in the case of the scientific and religious texts), which including such categories as “Exorcists’ lore”, “Astrological omens”, “Teratological omens”, “Terrestrial omens”, “Medical recipes”, “Dream omens”, and “Haruspical omens” (Parpola 1983, 6; Potts 2002, 24). When these text categories were mentioned together in the acquisition records, they were often listed in the following order — Astrological omens, Haruspical omens, Terrestrial omens, Teratological omens, Exorcists’ lore, Medical recipes, and Dream omens — though this sequence was by no means obligatory (Parpola 1983, 6, with n. 15). Each of these text categories, as well as the individual literary works, constituted a series of cuneiform tablets and wooden boards: as one commentator has noted, such an organization of tablets into series, a practice that can be traced back at least to the late second millennium BC, represented an attempt on the part of Mesopotamian scribes (and librarians) to collocate texts belonging to one composition (in the case of literary works) or to one subject area (in the case of the scientific and religious works) (Weitemeyer 1956, 227). The library records from the palace library of Assurbanipal indicate that at least some books in that library were assigned to categories on the basis of an interlacing of considerations of subject, genre, and function. Yet many questions remain: the rationale for the ordering of these categories in the records is not immediately obvious [15] and it is unclear how the (varying) sequence of these categories might have related to the physical arrangement of books in the library.

[top of entry]

2.2. The Ancient World: Greece and Rome

When we turn from the ancient Near Eastern world to that of Greece and Rome, we find that evidence for library classification remains scarce and fragmentary. Emblematic in this regard is the most renowned bibliographical work of antiquity, the Tables of those who were eminent in all learning and what they wrote (Pinakes tōn en pasēi paideiai dialampsantōn, kai hōn sunegrapsan), compiled by the Alexandrian scholar and poet Callimachus of Cyrene (ca. 305-240 BC) in the first half of the 3rd century BC (Fraser 1972, Volume 1, 452). Generally known to classical scholars simply as the Pinakes [16], this was a systematic bio-bibliography of all writers in the Greek language, based on, but probably not limited to, the collections of the great library associated with the Mouseion at Alexandria (Bagnall 2002, 356; Barnes 2002, 69; Blum 1991, 238-239; Witty 1958, 132). Of this monumental work, which once covered one hundred and twenty papyrus rolls, only some twenty-five fragments survive [17]. From analysis of these scraps of text, scattered as obiter dicta in the writings of later authors, philologists have inferred that Callimachus arranged his Pinakes by classes of authors, defined by the genres they wrote in or the subjects they wrote about (a class of “orators” is attested; classes for historians, philosophers, medical writers, and different kinds of poets are presumed to have been present) and genres of texts (classes of “laws” and “oratory” are attested as is, interestingly, a class of “miscellaneous prose writings”); that these classes were divided into subclasses; and that the authors whose works were enumerated within class were listed alphabetically (Blum 1991, 153-155; Witty 1958, 136). Many basic features of the classification, such as the total number of classes that it contained and their sequence, remain unknown, though some historians believe, on the basis of patterns common to later Greek literary historians, that the basic division was between poetic and prose writers, with the former category preceding the latter (e.g., Blum 1991, 154; cf. Müller 2011, 121). It is also unclear to what extent the organization of classes in the Pinakes reflected the physical organization of papyrus and parchment rolls in the library of the Mouseion: many modern commentators assume that there was a correspondence between the two (e.g., Blum 1991, 226; Olesen-Bagneux 2014, 5–6).

The Pinakes were the result of an extraordinary scholarly project associated with the collection of one of the largest libraries in the Hellenistic world [18] and so the question naturally arises how representative its mode of classification was. The tables of books compiled for the great Attalid library at Pergamum in Asia Minor in the 2nd century BC appear to have followed the structure of the Pinakes at Alexandria (Blum 1991, 182; Olesen-Bagneux 2015, 284): however, the Pergamene library, like the Alexandrian library, was a royal foundation possessing a sizeable collection and so may have been no more representative than the latter. A fragmentary inscription from the beginning of the 1st century BC listing the authors of works on politics and rhetoric held by a library associated with a gymnasion (i.e., a finishing school for young men) on the island of Rhodes has been interpreted by some commentators as evidence for classification based on subjects (e.g., Marrou 1964, 259; Segre 1935, 221–222), though too little is known about the scope of the collection to draw firm conclusions on the matter (cf. Blum 1991, 185–187, esp. 187). Papyri from the time of the Roman empire containing the inventories of smaller collections of book-rolls give only “limited information on how the compilers conceptualized or organized their collections” (Houston 2014, 44): nevertheless, “[p]oetry is often, but not always, kept distinct from prose, and in several lists there are indications of an organization by genre or subject matter”. Such scattered pieces of evidence suggest that the classification by genre and/or subject matter employed by Callimachus in the Pinakes does indeed reflect a more general pattern in the practice of bibliographical — and presumably bibliothecal — organization in Hellenistic and Roman times. However, it is likely that other principles of classification were utilized as well in the organization of library collections: for example, a handful of literary and epigraphic sources indicate that at least some libraries in late Republican and early Imperial Rome distinguished between Greek- and Latin-language sections in their collections, though little is known about how this linguistic distinction affected the physical arrangement of their holdings (Hall 2015, 22; Horsfall 1993; Nicholls 2010).

The Greek and Roman world witnessed the development of several significant classifications of knowledge outside of the bibliothecal realm that require mention. In the 4th century BC, the Greek philosopher Aristotle (384-322 BC) set forth the idea that knowledge is either theoretical, practical, or productive: in this schema, theoretical knowledge comprised physics, mathematics, and “first philosophy” or “theology” (Aristotle’s name for metaphysics) [19]; practical knowledge provided the basis for conduct (e.g., ethics and politics); while productive knowledge guided the making of things and covered the useful and productive arts (Berti 2014, 17–20; Guthrie 1981, 131; Ross 1964, 20; 62; 187; 276; Santoro 2015, § 4; Weisheipl 1976, 467–468) [20]. In the 3rd century BC, members of the Stoic school of philosophy developed a schema according to which philosophy, or, more precisely, philosophical discourse, was to be divided into three parts: logic, physics, and ethics (Long & Sedley 1987, vol. 2, 163–165; Weisheipl 1976, 469) [21]. By the first century BC, there had developed a tradition that certain arts formed the basis for a “rounded”, or general, education (enkyklios paideia) (Marrou 1964, 243–244; 1976, 51–52). Although, there was some variation in the number of these foundational arts recognized by different writers, seven formed the basic core — namely, grammar, logic, rhetoric, arithmetic, geometry, astronomy, and music (Marrou 1976, 53–57). In the Roman context, these arts came to be identified as the artes liberales, or liberal arts — those subjects that a free man (hence, liberales) had to master if he were to be truly educated (Jacobs 2002, 10). In the early Medieval period, the seven liberal arts would be codified into the Trivium, comprising the linguistic subjects of grammar, logic, and rhetoric, and the Quadrivium, encompassing the mathematical subjects of arithmetic, geometry, astronomy, and music [22]. These Aristotelian, Stoic, and educational classifications of knowledge do not appear to have informed library classification in antiquity: however, the concepts and conceptual distinctions that they introduced would have a profound influence on library classifications of later periods.

[top of entry]

2.3. The Ancient World: China

Outside of the Mediterranean basin, the best-documented tradition of bibliographical classification in the ancient world is that of China. The earliest known Chinese bibliographic classification is the Seven Epitomes (Qi lüe), compiled by the scholar Liu Xin (53 BC–23 AD) in the final years of the 1st century BC (Jiang 2007, 4–5; Lee 2008, 276; Lee and Lan 2011, 26–28; Tsien 1952, 309). The Seven Epitomes arose from a project, initiated by the Han Dynasty emperor Cheng (ruled 32–7 BC) in 26 BC, to collate and organize the writings kept at the imperial library. Liu Xin’s father Liu Xiang (79–8 BC) was assigned the task of overseeing the edition, description, and summarization of the texts; after his death, his son brought the work to completion with a catalog of the works collated and described by his father. The Seven Epitomes may well have served as a basis for shelf arrangement in the imperial library, though this remains largely a matter of speculation (Lee and Lan 2009, 223). Although the original text of the Seven Epitomes is no longer extant, its structure has been reconstructed by scholars from the epitome of the work embedded in a chapter of the History of the Former Han Dynasty compiled by the historian and poet Ban Gu (32–92 AD) (Jiang 2007, 5; Lee 2008, 276; 2012a, 380): paradoxically, it offers modern readers the fullest picture of an ancient bibliographic classification system.

Despite its name, the Seven Epitomes contained six main classes, each of which was divided into individual subclasses [23]. The six main classes were: “The Six Arts”, the “Masters”, “Lyrics and Rhapsodies”, “Military Texts”, “Divination and Numbers”, and “Formulae and Techniques” (Lee 2008, 277; Lee and Lan 2011, 27; 29) [24]. The epitome of “Six Arts” encompassed six canonical works of “classicist”, or Confucian thought and commentaries; that of the “Masters” covered works from important philosophical and literary schools; that of “Lyrics and Rhapsodies” dealt primarily with poetry and related genres; that of “Military Texts” enumerated works on the military arts; that of “Divination and Numbers” listed texts on the divinatory arts; and that of “Formulae and Techniques” was concerned with literature related to health, including topics such as medicine, sexology, and techniques of longevity (Lee 2012b, 72). Analysis of the sequence of main classes reveals that several principles appear to have informed the general structure of the classification. One of these was the dichotomy between the concepts of dao (“the Way”) and qi (“the vessels”), which, mutatis mutandis, were analogous to the Greek concepts of theoretical and practical knowledge, respectively (Lee 2012a, 386). Closely related to this was the division between works attributed to sages drawing upon the teachings of sage kings and those attributed to writers knowledgeable in technical matters (Lee 2012a, 389; Lee and Lan 2011, 35). The first three classes were works illuminating the dao written by sages: of these, the first class (i.e., that of the “Six Arts”) dealt with works associated with the sage par excellence, Kung Fu (i.e., Confucius), while the latter two dealt with works of other propounders of the dao, the second (i.e., that of “The Masters”) consisting of primarily prose texts and the third (i.e., that of Lyrics and Rhapsodies), of poetic texts (Lee 2012a, 389; Lee 2012b, 74; Lee and Lan 2011, 34–35). The final three classes, on the other hand, mirrored the domain of qi as it was organized in the royal household — namely, into departments of the military arts, divination, and health services (Lee 2012b, 72). Since the authority of the sage kings and the realm of dao was deemed to be greater than that of writers on qi, the philosophical and literary classes preceded the technical ones, with the most authoritative writings of all — the “Six Arts” — accorded pride of place as the first class. In general, then, the Liu Xin’s classification was based on a combination of specific works of canonical literature, genre, and subject as criteria for the formation of classes.

The Seven Epitomes served as a template for bibliographical classification in China for a little over two centuries until 235, when the Wei imperial librarian Xun Xu (265-317), basing himself on Zheng Mo’s (213–280) catalog Central Collection of Classics, developed a different classification in his New List of the Central Collection of Classics (Jiang 2007, 6; cf. Campbell 2017, 15). This classification encompassed only four main classes: the “Classics”, or Confucian literature and commentaries thereto; “Philosophy”, which encompassed not only the writings of the Masters but also military, divinatory, and medical literature; “History”; and “Belles-Lettres”, consisting primarily of poetic literature but also covering maps (Dudbridge 2017, 150; Campbell 2015, 50; Tsien 1952, 312) [25]. As modern commentators have noted, this new quadripartite classification was based, in large measure, on the fission or fusion of classes originally established in the Seven Epitomes: for example, material from the Six Arts was divided between “Classics” and “History”; the classes of Masters, Military Texts, Divination and Numbers, and Formulae and Techniques were all merged into the class of “Philosophy”, while Lyrics and Rhapsodies were taken up into the category of “Belles-Lettres” (Jiang 2007, 7, Figure 1; Lee 2008, 276; Tsien 1952, 312). The fourfold classification introduced in the New List of the Central Collection of Classics continued to be elaborated by scholars and versions of it were adopted by the compilers of official imperial bibliographies in subsequent dynasties: most notably, Li Chong (died ca. 362), inverted the order of the second and third classes to yield the classificatory system of “Classics”, “History”, “Philosophy”, and “Belles Lettres” (Jiang 2007, 7–8, with Figure 2). Although other systems of classification were developed and brought into use, one variation or another of the quadripartite classification dominated bibliographical practice in China from the third until the early twentieth century, when modern library classifications, such as the Dewey Decimal Classification, where introduced into Chinese libraries (Tsien 1952, 311–320).

[top of entry]

2.4. The Middle Ages: Europe

The Middle Ages have left fuller material evidence for library classification practices than that available for the Ancient World. This period was marked by new social and cultural forces that reshaped European civilization and thereby affected the organization of libraries. The cardinal force was Christianity, which not only permeated the worldview of much of the population but also provided the primary institutional settings for education and the cultivation of intellectual life. From the Sixth through the Twelfth centuries, education took place primarily in schools associated with monasteries and cathedrals (Grant 2001, 25–29; Riché and Verger 2011, 19–150) and, accordingly, the study of the canonical literature of Christianity and Christian theology became a part of the curriculum, alongside the traditional liberal arts bequeathed from antiquity. In the late Twelfth and early Thirteenth centuries, a new kind of educational institution emerged, the curricular structure of which would come to have considerable influence on later classifications of knowledge: the university (Grant 1996, 33-53; 2001, 98–114; Riché and Verger 2011, pp. 150–251). It is unsurprising, then, that the evidence for library classification in medieval Europe comes primarily from the book inventories of monasteries and university colleges, as well as booklists of the private libraries of scholarly individuals living within those milieux.

[top of entry]

2.4.1 Medieval classifications of knowledge

In considering European library classification of the Middle Ages, one must keep in mind contemporary developments in the classification of knowledge. For the most part, medieval classifications of knowledge built upon classical precedents, which were brought into relation with Christian learning and culture. An example of a somewhat rudimentary and pedagogically oriented classificatory schema of this sort is embodied in the Institutes of Divine and Human Readings (Institutiones divinarum et humanarum lectionum), a bibliographical guide compiled by the Italian statesman and scholar Cassiodorus (ca. 485-ca. 580) for the edification of the monks of Vivarium, a monastery that he founded in southern Italy to serve as a center for the copying, editing, and study of texts of Christian literature (Jones 1969 [1946], 27–32, 42-44; cf. Crippa 2015, esp. 88-89, 99-102). The Institutes were divided into two parts, the first of which was devoted primarily to discussion of “Sacred Scripture, exegesis, hagiography, and religious discipline” and the second, to “a summary of the seven liberal arts” (Weisheipl 1965, 62). Here, the primary principle of division was the distinction between religious and secular learning, in conformity with Cassiodorus’ intention that “the unbroken line of the Divine Scriptures and the compendious knowledge of secular letters […] might be related” (Institutiones, Preface 1, in Cassiodorus Senator 1969 [1946], 68). Interestingly, this distinction was not absolutely maintained: the final sections of the first part include brief discussions of literature on cosmography, gardening, and medicine (Jones 1969 [1946], 33, 35–36), secular areas of study and endeavor that did not fit within the framework of the liberal arts.

Medieval scholars frequently drew upon the classical philosophical classifications of knowledge. One highly influential classification of science was propounded by a contemporary of Cassiodorus, the philosopher Boethius (ca. 480–525/526), who simplified Aristotle’s tripartite division of scientific knowledge into theoretical, practical, and productive knowledge into a bipartite division between speculative (i.e., theoretical) and practical knowledge (Besson 1980, 6; Covington 2005, 50; Meriant 1901, 65–71; Weisheipl 1965, 63; 1976, 471). As it had been in Aristotle’s classification, speculative knowledge was further subdivided into physics, mathematics, and metaphysics, or theology, while practical knowledge was divided into ethics pertaining to the individual, ethics pertaining to the household, and ethics pertaining to society at large. Significantly, Boethius went on to identify the mathematical sciences with the Quadrivium (i.e., arithmetic, geometry, music, and astronomy), thus engrafting the classification of the liberal arts onto his simplified Aristotelian scheme. Another system that enjoyed considerable influence was an elaboration of the old Stoic classification of philosophy developed by the scholar and archbishop, Saint Isidore of Seville (ca. 560–636). According to St. Isidore, the three divisions of philosophy — logic, ethics, and physics — were to be further subdivided in the following manner: logic into rhetoric and dialectics; ethics into prudence, justice, fortitude, and temperance; and physics into arithmetic, geometry, music, and astronomy (Weisheipl 1965, 64; 1976, 473). In essence, this classification identified logic with the Trivium, ethics with the four cardinal virtues, and physics with the Quadrivium, thus combining the Stoic scheme with the liberal arts.

Perhaps the most comprehensive medieval classification of sciences was the one propounded by the French theologian and scholar Hugo of St. Victor (ca. 1095–1141) in his Didascalicon, a study guide for students attending the school at the Abbey of St. Victor in Paris. According to Hugo, philosophy was divided into four parts that, taken together, “contain all knowledge” (Didascalicon II.I; Hugo von Sankt Viktor 1997, 156): speculative knowledge (speculativa), practical knowledge (practica), mechanical knowledge (mechanica), and the knowledge of discourse (sermocinalis) (Besson 1980, 8–18; Covington 2005, 51; 65–66, Weisheipl 1965; 1976, 473). Speculative knowledge was divided into theology (including metaphysics), mathematics, and physics, with mathematics further subdivided into the Quadrivium of arithmetic, music, geometry and analysis, and practical knowledge was divided into ethics of the individual, ethics of the household, and public ethics. In this, Hugo was faithfully following the standard Boethian classification. Mechanical knowledge, on the other hand, represented a new category covering seven arts dealing with “human works” by which natural objects are transformed (Didascalicon II.I, II.20, in Hugo von Sankt Viktor 1997, 158; 192): fabric-making, weapons-making, commerce, agriculture, hunting, medicine, and theatrics. A concept that had received its first formulation in the early Middle Ages (Jacobs 2002, 11–13), the category of mechanical sciences clearly corresponded to the Aristotelian category of productive knowledge and were the precursors to what would later come to be termed the useful arts. Finally, the category of knowledge relating to discourse encompassed the Trivium of grammar, dialectics (i.e., logic), and rhetoric, thus elevating these fields of study from the status of propaedeutic arts to full-fledged sciences in their own right. Hugo’s classification thus both restored the productive arts into the truncated Aristotelian scheme set forth by Boethius and enlarged it by the addition of the arts of discourse, which had previously stood outside the ambit of scientific knowledge proper in Aristotelian classifications of the sciences [26]. The mechanical, or useful, arts and the science of logic would both, mutatis mutandis, feature in later medieval classifications of knowledge (Besson 1980, 19–23; Covington 2005, 52–53).

[top of entry]

2.4.2. Medieval library classifications

Although examples of medieval classifications of knowledge could be multiplied, the foregoing examples suffice to give an indication of the intellectual background within which the compilers of medieval library classifications worked. As already noted, most examples of library classification from the Middle Ages occurred in booklists of monastic and, later, university college libraries. In considering the structure of these lists, it is important to keep a few points in mind. First, most medieval library collections were comparatively small, ranging from a few dozen to a few hundred volumes: only rarely did the number of books in library exceed more than a thousand books (Besson 1980, 34 & 87–88, n. 1 to Chapter 4; Guthrie 1992, 94). Second, the booklists often functioned less as catalogs than as inventories listing the property of a given monastic or educational institution (Besson 1980, 35; Courtenay 1988, 283; Guthrie 1992, 95) [27]. It is thus often difficult to discern an order in the books listed (Besson 1980, 35; Guthrie 1992, 96). Even when books were enumerated in a classified order, it was sometimes the case that the names of the classes were not explicitly stated in the list: in such cases, they must be reconstructed through an analysis of the titles listed (cf., e.g., Norris 1939, 18–19).

Each monastic or university college library organized its own holdings and so there was great variability across library classifications. Nevertheless, some standard patterns can be discerned. One very common ordering of entries began with the Bible, parts of the Bible and commentaries thereto, proceeded to “the writings of the Church Fathers in varying sequences” and those of medieval theologians, and concluded with the works of ancient (i.e., pagan) authors, and works on liberal arts, philosophy, or sciences (Guthrie 1992, 97) [28]. This mode of arrangement, which divided religious and theological literature from secular subjects and placed the former before the latter, is strongly reminiscent of the organization of subjects in Cassiodorus’ Institutes. A variation on this pattern inverted the order, so that secular arts and sciences preceded the religious and theological literature [29]. In both of these orders, the basis for classes included specific canonical works (e.g., the Bible), specific authors (e.g., writings of individual Church Fathers and theologians), subject domains (e.g., the different disciplines of the liberal arts), genres both sacred (e.g., sermons) and secular (e.g., poetry), and, sometimes, language [30]. Yet not all medieval booklists were organized by the content, form, or author of the books recorded. Some inventories were arranged by the location of books within the institution holding them [31], while others crosscut subject, genre, and author criteria for classes with indications whether the volumes in question were chained to library tables or had been checked out to students, as did the catalog of Peterhouse at Cambridge compiled 1418 (Norris 1939, 87–89) [32]. Yet, on the whole, when booklists show some sort of arrangement, the preferred principle of organization appears to have been a mélange of subject domain and genre. Hierarchical relations were, as a rule, not expressed in medieval booklists: rather, classes were enumerated as members of a single array of main classes, even if they could be grouped into yet broader, higher-level classes.

Classifications of knowledge informed the organization of books in medieval libraries. Often, this influence was of a very general sort. We have already noted the prevalence of the Cassiodorian distinction between religious and secular literature. Likewise common, within the classification of secular literature, was the use of the liberal arts as a structuring principle (Besson 1980, 24), though there was considerable variation on this score and few classifications seem to have incorporated all seven arts of the Trivium and Quadrivium into their sequence of classes [33]. Occasionally, however, classifications of knowledge shaped medieval library classifications in more elaborate ways. A remarkable example of this is the classification outlined in a mid-13 century text entitled Biblionomia, composed by Richard de Fournival (1201-ca. 1260), a scholar, physician, poet, and chancellor of the church at Amiens. In this work, de Fournival set forth a classification for 162 volumes of his personal library, which, by his own account, he made available to interested fellow-townsmen in Amiens (Delisle 1874, 520–521) [34]. The classification consisted of three major fields (areole), each of which was divided into tables (tabulae) representing subjects or groups of subjects (Besson 1980, 25; Delisle 1874, 519). The first and largest of the fields contained “philosophical books” and its tables encompassed the following classes: grammar, dialectic (i.e., logic), rhetoric, geometry and arithmetic, music and astrology (i.e., astronomy), physics and metaphysics, metaphysics and ethics, “various (lit., wandering) books of philosophers (libri vagi philosophorum)”, poetic works, large-format books, as well as “secret books”, access to which was restricted. If one sets aside the classes of “various books of philosophers”, which functioned as a “miscellaneous” category for philosophical works that didn’t readily fit into any of the other classes [35]; large-format books; and the enigmatic secret books [36], then Fournival’s classification of philosophical books represents a synthesis of two traditions: the seven liberal arts with the standard Aristotelian classification of knowledge including theoretical sciences (physics and metaphysics), practical sciences (ethics), and a productive art (poetry). The second field was dedicated to books on the “lucrative sciences” of medicine and law (Delisle 1874, 521; 523), so-called because they were plied for fees, while the third field was devoted to theology. The class of lucrative sciences bears some analogy to the class of the mechanical arts in Hugo of St. Victor’s classification, in that it refers to fields of study and endeavor standing in contrast to the liberal, theoretical, and practical arts, which were considered to be the appanage of the “free man” who did not — in theory, at least — cultivate them to make a living. However, the parallelism is inexact, for, unlike the mechanical arts, which reflected the Aristotelian idea of the productive arts, the lucrative sciences were defined by their status as pathways to wealth. Fournival’s detachment of medicine, law, and theology from philosophy also reflects another classification of knowledge that was becoming increasingly pervasive in his day — namely, the division of faculties in universities, where students typically undertook a course of study in the faculty of arts before proceeding, if they wished, to one of the three higher faculties of theology, medicine, or law (Burke 2000, 91–92; Colish 1997, 268; Grant 1996, 42–49; 2001, 101). Taken as a whole, the Biblionomia nicely illustrates the strength of the influence that classifications of knowledge could exert upon the classification of books in the medieval world.

Another noteworthy feature of Fournival’s classification is that it made use of a notation to identify the location of the books in the tables composing the subdivisions of the three main areas (Besson 1980, 26–32). The notational base was rudimentary, consisting solely of letters of the alphabet, one of which was assigned to each book. Fournival devised ingenious methods to expand it: the color of ink in which the letters were written served to designate the major field within which the book belonged (gold was used for theology; silver for the lucrative sciences, and blue, violet, and red for philosophy) and the use of different calligraphic forms of a single letter allowed for multiple assignment of a single letter within a given major field (Glorieux 1963, 205). The system of notation was idiosyncratic and the principles underlying its application are not immediately apparent, since the allocation of notation did not coincide with the arrangement of classes within the major fields (Besson 1980, 29). Yet, even if much remains unclear about Fournival’s notational system, it is a significant early witness to an emerging practice in later medieval library classification. The use of press marks (i.e., call numbers) to indicate the location of books on the library shelves is well-attested, if not common, in catalogs of the 14th and 15th centuries (e.g., Bresson 1980, 44–47; 65–69; Norris 1939, 47–63; 74; 78–87; 93–96; 99–106; 110–112; 114–117; 122–123). Varying from library to library, notational systems for press marks typically made use of alphabetical characters, numerical characters, or a mixture of the two. Press marks were indexed to locations on the shelves, though such locations were often correlated to the subject content of books. This mode of using notation for designating fixed location would continue in use until the late 19th century.

[top of entry]

2.5. The Middle Ages: the Islamic world

As we have seen, medieval European library classifications were rooted in an intellectual tradition that combined the heritage of classical philosophical thought with the lore and literature of a culturally dominant religion. Similar conditions obtained in another major cultural area during the Middle Ages: the Islamic world. As is well known, Islam emerged within an Arabian setting and so was intimately tied to Arabic language and culture, even as it spread to other lands and language communities. From the Eighteenth century on, a large number of Greek philosophical, scientific, and medical texts were translated into the Arabic language and, combined with streams of tradition derived from the pre-Islamic Near East and India, exerted a powerful influence on the development of these fields in Islamic culture (Al-Hassan and Hill 1992, 24–28; Fakhry 1983, 1–18; Lindberg 1978, 55-58; Wellisch 1987, 3–4) [37]. Consciousness of these different sources of the Islamic intellectual tradition shaped the conceptual contours of Islamic classifications of knowledge and, through them, Islamic bibliographical classifications.

Classifications of knowledge in the Islamic world tended to follow two classificatory structures. One of these was a traditional Islamic classification that divided knowledge into three major areas: sciences of Islamic origin, or Muslim sciences (e.g., the study of the Qur’an, the Hadith [i.e., authoritative traditions about the Prophet Muhammad], and Islamic law), together with the philological and historical sciences necessary to interpret these writings; science cultivated by the Arabs prior to the advent of Islam (e.g., Arabic poetry and oratorical literature); and the sciences translated into Arabic from foreign sources (e.g., philosophy, astronomy, mathematics, geometry, and medicine) (Aman 1968, 13–14; Wellisch 1987, 32). The principles of division here were twofold: a religious distinction between Muslim and non-Muslim knowledge and an ethno-cultural distinction between Arabic and non-Arabic knowledge [38]. The other common classificatory structure was the Aristotelian division of the sciences in its dyadic form as a distinction between theoretical and practical sciences, which was favored by a number of leading Muslim philosophers [39]: as in Europe, the further subdivision of the theoretical sciences into mathematics, physics, and metaphysics, or theology was generally maintained. These two frameworks — the traditional Islamic and the Aristotelian — were sometimes brought into relation with each other in a single classificatory system [40]. However, most often, one or the other of the structures predominated in a given classification: for example, many philosophers tended to downplay the Islamic and Arabic sciences in favor of the Aristotelian schema of theoretical and practical sciences within their classifications of knowledge.

When we turn to bibliographical classifications, we find that, much as in Europe, they made selective use of patterns of organization derived from classifications of knowledge. The first and most influential of Islamic bibliographic classifications was that of the Fihrist al-‘Ulūm (Index, or Catalog, of the Sciences), a monumental classified bio-bibliography of authors who had written in Arabic or whose works were translated into that language that was compiled in Baghdad by the scholar Al-Nadim (ca. 935-ca. 990). The classification of the Fihrist consisted of ten main classes: language, scripts, and Qur’an; grammarians and language scholars; historians, biographers, and genealogists; poetry and poets; theology; Islamic law; philosophy and the sciences; stories, fables, and magic; schools of thought and doctrines of non-Muslims; and alchemy (Al Najjar 1996, 75; Wellisch 1987, 13–28). The unspoken basis for this sequence of classes was the traditional distinction between Arabic/Muslim sciences and ones derived from foreign sources. The first six classes all belonged to the categories of Muslim or Arabic sciences: writing, philology, and history and biography were studies ancillary to the study of the Qur’an; poetry was an Arabic science; and theology and Islamic law were manifestly Muslim sciences. The final four classes, by contrast, all dealt with subjects that were considered to be of foreign origin (philosophy and the sciences; the sciences and alchemy; and stories, fables, and magic) [41] or involved non-Islamic thought (schools of thought and doctrines of non-Muslims). Each of the classes was further divided into subclasses. Interestingly, the section on philosophy and sciences betrays virtually no trace of the Aristotelian classification of philosophy that was so popular among writers on this subject [42]: evidently, there were limits to what Al-Nadim was willing to take over from philosophical classifications of knowledge.

Libraries found a place in a wide range of institutional settings in the medieval Islamic world: there is evidence for the existence of court libraries, public libraries, mosque-affiliated libraries, libraries associated with schools of higher learning, and private libraries (Elhayyan 1990; Wilkins 1996). Surprisingly little is known about the systems of classification employed at these institutions. Many modern commentators believe that Al-Nadim’s classification served as a model for Islamic libraries and bibliographies in the later Middle Ages (Aman 1968, 51; Elayyan 1990, 130; Wellisch 1987, 33–34). One may well wonder whether, in practice, his classification was followed in all of its details: nevertheless, it is highly likely that the basic distinction between Islamic/Arabic and non-Islamic sciences on which it was based provided the conceptual framework within which librarians organized their collections (Wilkins 1996, 306). The pull of tradition proved to be strong in the Islamic world: as was the case in China, traditional library and bibliographical classifications continued in use into the Twentieth century (Aman 1968, 54).

[top of entry]

2.6 From the Renaissance to the late 19th Century: early Modernity

The period between the latter half of the 15th century and the last quarter of the 19th century was one of social, economic, technical, and cultural transformation in the Western world. At the beginning of this period, the recent discovery of the Americas and intensifying naval exploration of other previously unknown parts of the world expanded the horizons of the European imagination and led to the rise of far-flung trading empires (Brotton 2006, 79–97; Weaver 2015, 40–61). Renaissance humanism brought a renewed focus on the cultural traditions of pagan antiquity and fostered novel understandings of philology and new approaches to the arts (Brotton 2006, 38–57; Turner 2014, 33–44). The European discovery of movable-type printing led to an expansion of the ready availability of texts beyond the monastery, the school and university, and the palace (Brotton 2006, 48–51; Eisenstein 1983): information overload would henceforth become a common complaint among scholars (Blair 2010; Burke 2000, 103). One of the first movements to capitalize on this new technology was the Reformation, which, in the first half of the 16th century, shattered the unity of Christianity in Europe (Eisenstein 1983, 148–186; Marshall 2009, 17). The late 16th and 17th centuries saw important developments in the intellectual sphere: on one hand, a scientific revolution that resulted in a mechanistic and mathematical approach to explaining scientific phenomena (Okasha 2002, 2–9; Principe 2011) and, on the other, the emergence of new philosophies that, in contrast to the Scholasticism that had dominated the thought of the Middle Ages, took the epistemological question of the role of the human qua knower in the constitution of knowledge as a central question of philosophy (Deely 1994, 16) [43]. The 18th century ushered in the age of Enlightenment, characterized by increased confidence in the powers of human reason to arrive at a comprehensive and epistemologically secure knowledge of the world, a corresponding eschewal of the intellectual bases of traditional political and religious authority, and the optimistic belief that, through the use of reason, humanity would, over time, inevitably progress in all spheres of life — social, political, and cultural (Copleston 1994, 33–35; Grenz 1996, 60–81). The same century also witnessed the beginnings of an industrial revolution that applied new technologies grounded in the sciences to the production of goods and the rise of the market economy as a dominant framework for the exchange of goods — trends that would only intensify in the 19th century (Allen 2017).

Needless to say, all of the foregoing developments had far-reaching effects on both the conceptual articulation and institutional organization of knowledge. Within the universities, fields of study such as natural philosophy (i.e., the natural sciences), cosmography (i.e., geography) and history emerged from the positions subordinate to philosophy and theology that they had held in the intellectual mappa mundi of the Middle Ages and became autonomous subjects in their own right (Burke 2000, 99–102). Over time, these departments of knowledge underwent further differentiation — for example, natural philosophy was further partitioned into fields such as natural history, physics, chemistry, and geology in the 17th and 18th centuries — , a process that led to a proliferation of new, ever more specialized fields of inquiry, especially from the mid-19th century on (Burke 2012, 168–169; 2015, 19). The designers of bibliographical and library classifications in the early modern period were aware of these changes and generally sought to incorporate them into their classification schemes. Nevertheless, it is striking that, for the most part, bibliographic and library classifications tended to follow certain well-established patterns established in earlier classifications of knowledge.

[top of entry]

2.6.1 Tradition and innovation: Conrad Gesner’s bibliographical classification

The interplay of conservatism and innovation that marked the bibliographic and library classifications of this period is already apparent in the universal bibliography compiled by the Swiss polymath and bibliographer Conrad Gesner (1516–1565) between 1545 and 1555, which consisted of both an alphabetical and a classified section (Araújo, Sabba and Crippa 2016). Gesner’s classification, most of which was encompassed in a volume of the bibliography entitled the Pandectae [44], divided philosophy into preparatory (Preparantes) and substantive (Substantiales) sciences; the preparatory sciences were further subdivided into necessary (Necessariae) and decorative (Ornantes) sciences, and the necessary sciences, in turn, into discursive (Sermocinales) and mathematical (Mathematicae) sciences (Maltby 1975, 115–116; Šamurin 1964–1967, 1/118–119; Serrai 1977, 74–75). The twenty-one primary “partitions”, or classes, of the classification were distributed across these superclasses. First came the discursive sciences of grammar, dialectic, and rhetoric, joined by poetry, which Gesner originally classed among the discursive sciences but came to treat as a decorative science (Serrai 1977, 74). Then followed the mathematical sciences of arithmetic, geometry, music, astronomy, and astrology and, after them, the decorative sciences of history, geography, divination and magic, and mechanical and other arts useful to human life. Finally came the sequence of substantial sciences, comprising physics, metaphysics and non-Christian theology, ethics, economics, politics, jurisprudence, medicine, and Christian theology.

Even a cursory glance at Gesner’s scheme indicates that the discursive and mathematical arts represented the medieval Trivium and Quadrivium, respectively, while the substantial arts reflected the old Aristotelian scheme of theoretical (i.e. physics, metaphysics/non-Christian theology) and practical (i.e., ethics, economics, politics) sciences, as well as the three higher university faculties (law, medicine, theology) (Živný 1920, 55). Departing somewhat from medieval tradition was the category of decorative sciences, which brought together the emergent fields of history and geography with the mechanic arts and the occult science of divination and magic. This structuring of the primary classes, each of which was further subdivided into subclasses [45], neatly demonstrates the degree to which traditional patterns in the classification of knowledge continued to have a hold on bibliographic classification in the 16th century.

[top of entry]

2.6.2 The faculty system

Though traditional frameworks for classifications continued in use, there were also some significant shifts in emphasis. Perhaps the most far-reaching of these in an era when university libraries were growing in size and importance was the elevation of the universities’ faculty structure — i.e., their administrative division into faculties of arts, law, medicine, and theology — into a master template for the structuring of classifications (Šamurin 1964–1967, 1/139–141). The ways in which the faculty system was adapted to bibliographical and library classifications varied in complexity and sophistication. At its simplest, the fourfold division of the faculties was transposed directly to the library catalog or the shelves. Such was, for example, the structure of the first catalog of the Bodleian Library at Oxford University, a classified catalog issued in 1605 that consisted of four classes: theology, medicine, law, and the arts (Norris 1939, 143–144; 147) [46]. In most classifications, however, the subjects treated in the faculty of arts were specified in greater detail after the obligatory triad of theology, law, and medicine: to cite but three examples, in the late 16th and early 17th centuries, the library of Leiden University divided its collections into seven main classes — theology, law, medicine, literature, philosophy, mathematics, and history (Burke 2000, 104–105) [47]; in the first years of the 17th century, Václav Brežan (ca. 1568–ca. 1680), archivist and librarian of the large library of the noble Rožmberk family in Třeboň, prepared a catalog of its holdings consisting of five main classes — theology; law; medicine, chemistry and allied subjects; history, cosmography, geography, antiquities, and arts; and philosophy, poetry, and music (Živný 1920, 55); and in 1624, the German librarian Georg Draud proposed a classification of seven classes — theological books; law books; medical books; historical, geographical, and political books; books of the humane and non-humane sciences; poetical books; and books of music (Serrai 1977, 98–99). The faculty system also featured in proposals for ideal library classifications set forth by two eminent polymaths: the French physician and librarian Gabriel Naudé (1600–1653), who in his Advis pour dresser une bibliothèque (1627), recommended a sequence of main classes consisting of theology, medicine, law, history, philosophy, mathematics, humanities, etc. (Naudé 1627, 100; Šamurin 1964–1967, 1/134–135; Serrai 1977, 133) [48], and the German philosopher, mathematician, jurist, historian, and librarian Gottfried Wilhelm Liebniz (1646–1716), who, in a manuscript written in the final years of the 17th century, set forth a classification comprising the main classes of theology, law, medicine, intellectual philosophy, mathematics, physics, philological sciences, civil history, literary history and bibliography, and general and miscellaneous books [49], each of which, save the last, was further subdivided (Šamurin 1964–1967, 1/142–144; Schulte-Albert 1971, 141–142; Serrai 1977, 181–183) [50]. Widely used in the 17th century, the faculty system proved to have historical durability: vestiges of it are still apparent in the classification used at the British Library in the late 19th century [51].

Although the faculty system enjoyed widespread popularity, it was not the only conceptual framework used to constructing library classifications. As had been the case in the Middle Ages, each individual library created its own classification scheme or adopted and adapted that of some other library: thus, the period between the beginning of the 17th century and the middle of the 19th century saw the proliferation of many different classificatory schemes rooted in various classifications of knowledge. Amidst this variety, one can discern two streams of tradition that proved to be especially influential, one of which developed in France and another whose origins can be traced to England.

[top of entry]

2.6.3 The early Modern French tradition from Garnier to Brunet

The early modern French tradition of library classification can be viewed as a development away from the faculty model of classification. One of the first and, theoretically, the most significant representatives of it is the classification of the library at the Jesuit College in Paris established by one of its keepers, Jean Garnier (1612–1681), an account of which was published in 1678 (Šamurin 1964–1967, 1/155–159; Serrai 1977, 135–141). Garnier grounded the framework of his classification in epistemological considerations, more specifically on the “learning which is contained in books (doctrina, quae libris comprehenditur)” (Systema Bibliothecae Collegii Parisiensis Societatis Jesu 1678, 10) [52]. He identified four “powers of the human soul capable of learning” (animi vires capaces doctrinae)” and correlated them to different areas of human knowledge: the “superior reason (ratio superior)” capable of “divine learning (doctrina divina)”; the “inferior reason (ratio inferior)” capable of “human learning ([doctrina] humana)”; the “power of memory (vis reminiscendi)” capable of “the learning of times (i.e., temporal events)” (doctrina temporum)”; and the “power of entering into society with one another (vis societatem cum aliis ineundi)”, capable of “the learning of law (doctrina juris)” (p. 10). These correlations formed the basis for a library classification consisting of four main classes — theology, philosophy, history, and law — each of which was subdivided into numerous subclasses (p. 11) [53]. Garnier set out principles for different kinds of subarrangement of the latter, which were to be applied at appropriate points of the classification (pp. 12–13) [54]: one modern commentator has described the resultant classification as one that balanced theoretical principle with pragmatic flexibility (Serrai 1977, 141).

Garnier’s epistemologically-based rationale for the sequence of main classes marked a significant intellectual departure from the faculty system, even if all of its constituent classes were derived from the latter. Contemporary and later French librarians and bibliographers followed this path in practice, elaborating the set of main classes in his system and rearranging them in various combinations, though detaching them from the philosophical underpinnings that he had used to justify them. The astronomer and librarian Ismael Boulliau (1605–1694) set forth a similar classification in his catalog of the de Thou family’s library issued in 1679, where the main classes were theology, law, history, philosophy, and humane letters (i.e., philology and literature) (LaMontagne 1961, 126; Petzholdt 1866, 27): here, philology and literature, which Garnier had treated as subdivisions of philosophy (Systema Bibliothecae Collegii Parisiensis Societatis Jesu 1678, 8, 41–48) were elevated to the status of a main class and the order of the main classes reconfigured in a pattern closer to that of the faculty system.

Boulliau’s sequence of classes came to enjoy considerable popularity among members of the French bookselling community. In the next century, the catalogs of the influential French bookseller Gabriel Martin (1678–1761) featured the main classes of theology, law, sciences and arts, belles lettres, and history (Šamurin 1964–1967, 1/364–365, n. 15; Serrai 1977, 208), as did the classification used by the bookseller Guillaume François de Bure in his Bibliothèque instructive of 1763 (Živný 1920, 63; cf. Petzholdt 1866, 34–35) [55]. This booksellers’ tradition found its culmination in the classification of books set forth by the bibliographer Jean Charles Brunet (1780–1867) in his Manuel du libraire et de l’amateur des livres. First published in 1810, Brunet’s system featured the same main classes as Martin’s and de Bure’s classifications, which were articulated into an extensive series of further divisions and subdivisions (LaMontagne 1961, 127–129; Šamurin 1964–1967, 1/242-248; Serrai 1977, 280; Živn&yac 1920, 69-70). According to Brunet, this classification scheme was meant not to represent any principled “encyclopedic” classification of knowledge but to serve as a system of organizing books that was practically useful because it conformed to usage [56]. Brunet’s classification proved to be immensely influential not only in France but in many other countries: as one commentator has noted, “most library classifications, which were drawn up in the first three quarters of the 19th century in France, England, Italy, the United States of America, and partially other countries are either a simple imitation or a modification of Brunet’s schema” (Šamurin 1964–1967, 1/241).

[top of entry]

2.6.4 The Baconian tradition

The other major tradition of library classification was rooted in a classification of knowledge first propounded by the British statesman and philosopher of science Francis Bacon (1561–1626) in his treatise The Advancement of Learning (1605) and later refined and expanded in De Augmentis Scientiarum (1623). As Garnier would do over half a century later, Bacon adopted an epistemological rationale for his classification. Dividing the field of knowledge into “human learning” and “divine learning”, Bacon (1996 [1605], 175) held that “[t]he parts of human learning have reference to the three parts of Man’s Understanding: History to his Memory, Poesy to his Imagination, and Philosophy to his Reason”. According the first place in his sequence of main classes to history, he subdivided it into natural history and political history; poesy, he divided into epic, dramatic, and allegorical-didactic poetry; while the third and final main class of philosophy encompassed divine philosophy, natural philosophy (itself further subdivided into theoretical philosophy, the primary divisions of which were physics and mathematics and practical philosophy, the primary subclasses of which were mechanics, and magic), and the philosophy of man, concerning either man as an individual being possessing a body and soul or man as a social being (LaMontagne 1961, 36–37; Šamurin 1964–1967, 1/159–165; Serrai 1977, 86–93). Bacon (1996 [1605], 189) held that the three faculties of understanding were applicable to divine learning, or theology, as well, but did not elaborate this part of his classification in any detail, for his attention was trained almost exclusively on the articulation of human learning, which he worked out in great detail. Bacon elaborated this classification within a general vision of the unity of knowledge, according to which “the distributions and partitions of knowledge […] are like branches of a tree that meet in stem, which hath a dimension and quantity of entireness and continuance” (Bacon 1996 [1605], 189). This vision of human knowledge as a unified whole — a universe of knowledge, if one will — would become a leading assumption underlying library classification in the modern period.

Bacon’s classification of the sciences had far-reaching influence on later classifications of knowledge and on library classifications (Šamurin 1964–1967, 1/175; Shera 1965, 79–80). The French Enlightenment thinkers Jean le Rond d’Alembert (1717–1783) and Denis Diderot (1713–1784) used it as the basis for the classification of sciences in the Encylopédie ou Dictionnaire raisonée des sciences, des arts et des métiers that they published between 1751 and 1780, albeit not without many changes, the most significant of which was the reordering of the sequence of the three faculties of the human intellect into memory, reason, and imagination, and the consequent rearrangement of the three main divisions to history, philosophy, and poesy (Éncyclopedie 1986, 1/187-200; Šamurin 1964–1967, 1/187–200; LaMontagne 1961, 37–38; Serrai 1977, 201–208). Thomas Jefferson (1743–1826), the eminent North American statesman, intellectual, and third president of the United States of America, adopted and further modified D’Alembert and Diderot’s system into a classification for his personal library, the three primary divisions of which were history, philosophy, and the fine arts, to which he added a final class of polygraphic works (Gilreath and Wilson 1989, ix; 2–3; LaMontagne 1961, 38–39; Miksa 1984, 5–7). When Jefferson sold his library to the United States in 1815, this modified Baconian system was adopted at the newly established Library of Congress, where it continued in use as a catalog classification until 1864 and, with numerous modifications, as a shelf classification until the end of the 19th century (Miksa 1984, 13).

Likewise building on the Encyclopedists’ version of Bacon’s classification was the American librarian Edward William Johnston (1799–1867), who adopted the sequence of history, philosophy, and poetry, followed by polygraphic writings, in classed catalogs that he prepared for the libraries of the College of South Carolina (1836), the Mercantile Library Association in New York (1837), and the St. Louis Mercantile Association (1867) (LaMontagne 1961, 156–163). The last of these three catalogs, in turn, inspired the philosopher and educator William Torrey Harris (1835–1909) to develop a new library classification for the library of the St. Louis Public School Library (Harris 1870a, xiii, n. *; cf. LaMontagne 1961, 175). Unlike Johnston, Harris did not take up D’Alembert’s version of Bacon’s classification but inverted the original order of main classes so that his classification followed the sequence of science (≈ Bacon’s philosophy), art (≈ Bacon’s poesy), and history, with a class for miscellaneous works such as encyclopedias and periodicals added at its end as an appendix (Harris 1870a, xv-xvi; 1870b, 119; 122–125) — a reconfiguration that brought it into line with Harris’s own Hegelian views [57]. First published in 1870, Harris’s rearticulation of the Baconian scheme would be cited as a source of inspiration by the creator of one of the most prominent modern library classifications some three years later and thus can be seen as serving as a bridge between the premodern and modern eras of library classification.

[top of entry]


I wish to express my gratitude to the two anonymous reviewers for their careful reading and trenchant critiques of the original draft of this article, which pushed me to improve it. I also wish to thank Claudio Gnoli for his careful editorial work and for drawing my attention to some useful bibliographic references. Above all, I wish to thank Birger Hjørland for his editorial perspicacity and tact, his generous sharing of references, and, last but no means, his exquisite patience with an all-too-dilatory author. Any errors of fact or interpretation or failures of historiographical or theoretical perspective are my responsibility alone.

[top of entry]


1. Etymologically, the English library — “place to contain books accessible for reading” — is derived from the Latin word libraria (taberna) “(shop) pertaining to books” (i.e., “bookseller’s shop”) (Onions, Friedrichsen and Burchfield, 1966, 526a, s.v. “library”), while the analogous word in many European languages — e.g., bibliothèque in French; biblioteca in Italian, Portuguese, and Spanish; and Bibliothek in German — derives from bibliotheca, the Latin form of the Greek word bibliothēkē “book repository” (Lewis and Short 1987, 235 a, s.v. bibliotheca). In both cases, the root word — liber and biblos, respectively — signifies, in its primary meaning, “book”. It goes without saying that, over the course of history, books — or, at any rate, text-bearing objects comparable to them — have taken many different forms, ranging from clay cuneiform tablets, wooden boards, and papyrus rolls in the ancient world to codex volumes, manuscript or printed, in late antique, medieval, and modern times, and, most recently, various forms of electronic books (Kilgour 1998).

2. On this view, all library classifications would be book classifications. However, the converse does not hold true, for not all book classifications are library classifications. Bibliographers use classifications for compiling lists of books without indexing them to particular libraries and booksellers make use of classifications to organize their stocks and to structure their catalogs. Library classification should thus be seen as forming a branch of the broader sphere of activity of bibliographical classification.

3. A striking witness to the prototypicality of the book as bibliographic entity is the vocabulary used by Paul Otlet, bibliographer and the founder of the Documentation movement in the late 19th and early 20th centuries. Although Otlet considered documents — understood lato sensu as including not only graphic records but all real-world objects treated as sources of information (cf. Buckland 1991, 586; 1997, 805) — to be the central objects of bibliographical concerns, he used the word book as a shorthand term to refer to the kinds of objects that he defined as documents (Day 2001, 10; 17; Wright 2014, 102).

4. Library classifications with fairly general classes and shallow hierarchies are sometimes known as broad classifications (Mann 1930, 45; 47; Satija 2000, 222–223), which some commentators contrast with close classifications, that is to say, library classifications with “[a]n arrangement of subjects in minute divisions” (Mann 1930, 47). It should be noted, though, that the opposition between “broad” and “close” classification is often used to refer to the use of a given classification system in a specific context rather than its structure. When making use of a hierarchically rich classification, a given library can choose to make use of only the general, topmost levels of the hierarchy or it can choose to adopt the full spectrum of hierarchical levels, either generally or for some preferred subject areas: in such cases, the former constitutes an example of broad classification, while the latter represents an example of close classification (Broughton 2015, 377, s.v. “broad classification”; 378, s.v. “close classification”; Sayers and Maltby 1975, 263–266).

5. According to the typology of relationships elaborated by Svenonius, these different kinds of hierarchical relationships can be characterized as follows:

  • Genus-species, or generic, relationships are “logically or analytically true” (Svenonius 1989, 39), with all members of the species subclass inheriting all the properties of the genus superclass and so included in it. An example is the relationship between the genus “Mammals” and the species “Dogs”, in that all dogs display all the properties of being a mammal and so form logical subclass of the superclass of mammals. Such relationships are best exemplified by relationships between “natural kinds”.
  • Kind-instance, or instantiative, relationships are relationships between a class embodying a concept of a kind of entity and the individuals comprising that class: for example, the relationship between the class “Library classificationists” and the individual Melvil Dewey is a kind-instance relationship, in that Melvil Dewey is a member of (i.e., instantiates) the class of library classificationists.
  • Whole-part, or partonomic, or meronomic, relationships encompass the relationships between a whole and a part: for example, the relationship between the history of the Czech lands and the history of Europe is a partonomic relationship, in that the former is a part of the later.
  • Perspectival, or aspect, relationships are relationships between a superclass defined by attributes that may or may not be possessed by the kinds of entities constituting its subclass(es). An example is the relationship between the superclass of “Pets” and the subclass of “Dogs”, in that some dogs are domesticated animals that are kept in a household for the purpose of companionship (i.e., pets) and others are not. In this and other such relationships, the superclass represents, so to speak, a perspective that captures a certain aspect, by no means a necessary one, of the classes of entities to which it applies. For slightly different accounts of these relationships, see Austin 1984, 79–80, who calls them “quasi-generic” relationships; Broughton 2015, 27-29, who prefers to call them “syntactic relationships.”
  • Topic-subtopic, or conventional, relationships are relationships between a given field or discipline, and the subtopics that are conventionally treated within the that given field or discipline. An example of such a relationship is that between Mathematics and the various topics whereof it treats, such as Arithmetic, Algebra, Geometry, and Calculus, or, again, between Calculus and differential equations, infinitesimals, and so on.

6. The terms shelf classification and catalogue classification shall be preferentially used here on the grounds that they are semantically more precise and transparent than their alternatives: moreover, bibliographical classification and bibliothecal classification are often used to refer to library classifications tout court without regard to their use within the library.

7. This periodization of the history of classification has drawn inspiration from the historiographical schemata presented in Miksa (1994; 1998). More specifically, it has sought to accommodate the distinction between “modern” and “postmodern” library classification implicit in the latter work into the broader chronological framework of the former.

8. As regards the utility and limits of periodization in history, it is useful to keep the following remarks from a renowned French historian in mind: “Periodization, as the work of human minds, is at once artificial and provisional. In this respect its usefulness is twofold: it allows us to make better sense of the past, in the light of the most recent research, while at the same time reminding us of the imperfections of this instrument of knowledge, we call history” (Le Goff 2017 [2014], 17).

9. Closest to a total history of bibliographical classification is the monumental and richly documented two-volume work Ocherky po istorii bibliotechno-bibliograficheskoĭ klassifikatsii (1955–1959) written by the Soviet librarian Evgenii I. Shamurin (Šamurin) (1889-1962) and available in German translation as Geschichte der bibliothekarisch-bibliographischen Klassifikation (Šamurin 1964-1967), which, despite its age and dependence on problematic ideological assumptions, continues to be an invaluable source of detailed information about a number of historical library classifications that have received little or no attention from other writers on the history of library classification. Yet even this veritable opus magnum does not cover all aspects of the subject.

10. The principle of selection governing the choice of non-Western cultural regions to discuss here — Ancient Near Eastern, Chinese, and Islamic — has been almost entirely pragmatic in nature. I have covered only those areas for which I could find substantial and reliable studies about the history of library classification written in the languages — exclusively European — that I can read. The absence of some regions from this survey does not, of course, imply that their premodern library classifications are of lesser historical significance than those that have been included: it simply indicates that a critical mass of information about them is not readily available to non-specialist researchers unfamiliar with their cultural histories. It is earnestly to be hoped that scholars of KO from other regions not represented here, such as India, Japan, and Korea, will undertake the exposition and interpretation of their premodern library classification systems for the benefit of a global audience. [Editor's note: Information about premodern Indian library classification has been collected for this entry and a section on it will be added in the next version.]

11. Readers will doubtless find it a contradiction in terms to call the third and final phase of the premodern period of library classification “early modern” and, indeed, it is. This terminological inconcinnity arises from the fact that, in practice, historians of library classification have tended to set the boundary separating “modern” from “premodern” library classification chronologically much later (ca. 1875) than the currently regnant periodization of general history, according to which the boundary between the “modern” and the “premodern” is generally set either at the end of the Middle Ages (ca. 1450–1500) or, at latest, at the time of the Enlightenment (ca. 1700), with the span between these points generally known as the Early Modern period.

12. The best modern translation of this list is found in Kurth 1994, 144–145; 146–147.

13. Since some of the items in the list are susceptible to different interpretations, the precise number of books in the list varies across different authors. Grimm (1989, 161) counts 41 titles; Wessetzky (1958, 159) gives 32 titles, Šamurin (1964–1967, 1/8–10) lists 37 titles, while I count 33 in the recent translation of Kurth (1994).

14. For a conspectus of the kinds of texts archaeologically attested in the palace library, see Potts 2002, 23; Robson 2013, 43.

15. The editor of the library records notes that the order in which the text categories is listed is not determined by the number of tablets acquired nor by the extent of the series and suggests that “[it] may accordingly reflect the popularity the texts enjoyed” (Parpola 1984, 6). He does not, however, specify how “popularity” would be gauged — perhaps by the frequency of consultation by palace experts in religious and scientific lore?

16. The word pinax (pl. pinakes) originally referred to wooden tablets and, in later usage, took on the additional meaning of lists or tables of titles and authors written upon such tablets: for discussion and further references, see Witty 1958, 132. Blum (1991, 151-152) argues that the full title of Callimachus’s work, which is attested only in Byzantine-period sources, was probably not the original one given by its author but a later elaboration.

17. The fragments are conveniently collected and translated in Witty (1958).

18. Byzantine-period sources claim that, at its peak, the library at Alexandria, which was situated in two locations within the city, held upwards of 532,800 papyrus rolls, while a Roman-period writer of the 2nd century AD claimed that it held over 700,000 rolls: present-day scholars are highly skeptical of these figures (Bagnall 2002, 351–356; Blum 1991, 99, 106–108).

19. These three subdivisions were based on the philosophical consideration that physics has to do with physical objects subject to motion, mathematics deals with objects abstracted from matter and without motion, while metaphysics, or theology, considers objects that are inherently abstract, motionless, and eternal (Metaphysika, 1025b19–1026a23, in Aristotelés 2008, 156–157). For further discussion, see Berti 2014, 17-18; Mariétan 1901, 25–38; Ross 1964, 62.

20. Interestingly, Aristotle did not consider logic, or “dialectics”, to fall within the realm of knowledge but to be propaideutic to it (Guthrie 1981, 135; Ross 1964, 20; Serrai 1977, 4; contra Weisheipl 1976, 468, who identifies it as a productive art).

21. According to Marrou (1976, 46), this schema is to be traced back to the traditions of the Academic (i.e., Platonic) school of philosophy: at any rate, some later writers, such as St. Augustine of Hippo, attributed it to Plato himself.

22. The term quadrivium (lit., “the four ways”) was coined by the early medieval philosopher Boethius (ca. 480–525/526) in the early 6th century AD (Chadwick 1981, 73; Jacobs 2002, 10), while the term trivium (lit., “the three ways”) appears to have been first used in Carolinigian times (8th–9th century AD) (Rajna 1928).

23. According to some modern commentators, the first epitome, or division, of the Seven Epitomes most likely comprised a general introduction to the compilation (Lee 2012a, 380; 2012b, 64; Lee and Lan 2011, 28; Tsien 1952, 310), while others characterize it as a set of “editorial notes” on Liu Xiang’s collations of texts (Jiang 2007, 5).

24. The names of these classes are rendered differently in English by different translators. For example, Jiang (2007, 5) translates the names of the six classes, respectively, as “six scriptures”, “various masters”, “odes”, “military writings”, “divination and calculation”, and “medicine and formulas”, while Campbell (2017, 49) renders them as “The Canon”, “The Philosophers”, “Poetry”, “Military Works”, “Numerology and Techniques”, and “Medicine”. Both of these authors characterize the Seven Epitomes as a bibliography rather than a catalog, though Jiang’s (2007, 5) description of them as “the first bibliography of an imperial collection” amounts to identifying them as a catalog.

25. It should be noted that, as Jiang (2007, 6-7) points out, Xun Xu did not label these classes with subject names but simply referred to them in ordinal fashion as the “First”, “Second, “Third”, and Fourth” classes. The class names “Classics”, “Philosophy”, “History”, and “Poetry” were given to the classes over three centuries later in a bibliographic work known as the Treatises on Classics and Books (pp. 7-8).

26. Hugo’s classification ramified into yet further subdivisions, which are briefly laid out in Besson 1980, 10–14; for further discussion, see Olson 2010, 128–134.

27. One indice of the inventorial nature of monastic book-lists is that they not infrequently included non-bibliographical items among the objects enumerated. For example, the inventory of the library of the Augustinian monastery in the Bohemian town of Třeboň compiled in 1415, which enumerated saints’ relics, objects used in church services, and ritual vestments as well as the books themselves (Hlaváček 2005, 174; 201–207): the books are listed before the ritual objects and before the vestments.

28. For good examples of such an arrangement, see the catalog of the library at the Sorbonne from 1338 (Besson 1980, 38–40); the catalog of the monastic library of the Austin (i.e., Augustinian) Friars at York compiled in 1372 (Norris 1939, 47–52); the catalog of St. Martin’s Priory in Dover from 1382 (pp. 52–57); and the catalogs of the Durham Cathedral Library from 1391 and 1395 (pp. 57–62).

29. See, for example, the late 14th-century book inventory of Oriel College in Oxford, which was organized into the following sequence of classes: grammar, logic, philosophy, civil law, canon law, rhetoric, arithmetic, geometry, astronomy, and theology, with a reversion to a mixture of philosophical and liberal arts books at the very end of the list (Courtenay 1988, 284–288). The 1321 catalog of the Sorbonne college provides another, much more extensive example of this kind of arrangement (Besson 1980, 44–47).

30. With regard to language, almost all books kept in monastic and university libraries were written in Latin. Thus, one occasionally finds classes for books written in vernacular language. For example, the otherwise quite rudimentary catalog of books from the library of Durham Cathedral included a category of “English (i.e., Old English) books (libri anglici)” (Norris 1939, 14, 16), while the final class in the catalog of books in the Sorbonne library from 1338 comprised “Books in French (libri in Gallico)” (Besson 1980, 40).

31. For example, the list of books belonging to the Cistercian Abbey at Meaux, or Melsa, in Yorkshire dating to 1396 divides them into the following categories: “in front of the great altar”, “in the choir”, “in the chapel of the infirmary”, “in the common cupboard in the church”, “in other cupboards of the office of the cantor in church”; “in the common cupboard of the cloister”, “in the highest shelf above the door”, “in the highest shelf opposite”, and “in the same press in other shelves distinguished by the alphabet” (Norris 1939, 68–70).

32. In this catalog, one finds such classes as “chained books of theology (libri theologie cathenati)” and “books of theology assigned to students (libri theologie assignati sociis)”; “chained books of poetry (libri poetrie cathenati)” and “books of poetry and grammar assigned to students (libri poetrie et gramatice assignati sociis)”; “chained books of civil law (libri iuris civilis cathenati)” and “books of civil law divided among students (libri iuris civilis divisi inter socios)”; and, again “chained books of canon law (libri iuris canonici cathenati)” and “libri iuris canonici dividendi inter socios (libri iuris canonici dividendi inter socios)”.

33. For example, a fragmentary late 12th-century list of books from Christchurch in Canterbury included a sequence of classes on secular subjects — grammar, rhetoric, music, philosophy, poetry, astronomy, miscellaneous, and logic, dialectic, and law — in which grammar, rhetoric, music, astronomy, and logic and dialectic are all components of the traditional liberal arts, though the latter has been mingled with the subject of law (Norris 1939, 18–19). Or consider the late 14th-century book inventory from Oriel College, which began with the following classes: grammar, logic, philosophy, civil law, canon law, rhetoric, arithmetic, geometry, and astronomy (Courtenay 1988, 284–285): if one removes the intrusive classes of philosophy, civil law, and canon law, then the classification offers six of the seven canonical liberal arts in their classical order.

34. There has been scholarly debate over whether Fournival was, indeed, describing a real or an idealized library collection: for example, Birkenmajer (1922, esp. pp. 6–9) and Glorieux (1963, esp. pp. 206–212) argue for the former view, while Haye (2010) advances arguments for the latter. On balance, the extant evidence appears to favor the thesis of authenticity.

35. De Fournival describes the category of philosophiae vagae thus: “There are certain volumes that everywhere concern themselves with so many parts of philosophy that they do not make an unmixed statement for any of them [sci. the parts of philosophy] nor can they be rightly placed under any of them” (Delisle 1874, 521). The list of books placed under this category is heterogeneous in nature, including, inter alia, selections from the Hermetic writings, Macrobius’s Commentary on the Dream of Scipio, Aulus Gellius’s (“Agellius’”) collection of philological miscellanies known as the Attic Nights, Pliny the Elder’s Natural History, excerpts from the architectural books of Vitruvius, Martianus Capella’s disquisition on the liberal arts entitled the Marriage of Mercury and Philology, Boethius’s Consolation of Philosophy and his works on philosophical theology, as well as chronicles and cosmographies (pp. 530–531). The one unifying feature of all of these books was that they could not be easily accommodated under any of the liberal arts, physics, metaphysics, or ethics. For further discussion of this category, see Glorieux 1963, 227; Haye 2010, 216; 219.

36. The nature of the subject matter of these secret books, “the profundity of which is not worthy to be set forth to public eyes” (Delisle 1874, 521), is unknown. Some commentators have assumed that the volumes in question contained works on alchemy or astrology (e.g., Birkenmajer 1922, 86), though some incline to the view that the works in question were esoteric philosophical works (e.g., Haye 2010, 217).

37. Although Islamic philosophy and natural science built upon Greek, Indian, and other traditions, it should not be viewed as entirely derivative of the latter: rather, it represents a creative reinterpretation and re-elaboration of these earlier traditions into new forms augmented by original research and observation (Haq 2009, 37-39; Efron 2009, 83-84). Islamic philosophy and natural science, in turn, would have a tremendous impact upon other cultural regions, most notably Europe, in the high and later Middle Ages.

38. This general framework is especially salient in the classifications of knowledge constructed by the mathematician, astronomer, and geographer, Al-Khwārizmī (780–ca. 858), the theologian and philosopher Al-Ghazālī (1058-1111) in his middle period, and the historian and philosopher of history Ibn-Khaldūn (1332–1406) (Aman 1968, 31-38, 40–51; Charaf 2004, 217–219).

39. This distinction can be found in the classifications of science propounded by the philosophers Al-Kindī (ca. 800-ca. 873), Ibn-Sīnā (or Avicenna) (980-1037), and, at least in his youth, Al-Ghazālī (1058-1111) (Al Najjar 1996, 65-68; Aman 1968, 16–17, 23; Charaf 2004, 217). Some philosophers did not utilize this distinction directly but it was implicit in their work. Thus, for example, in Al-Fārābī’s (870-950) classification of knowledge into eight primary classes — language sciences, logic, mathematics, physics, metaphysics, politics, jurisprudence, and dialectics — , it is apparent that mathematics, physics, and metaphysics correspond to the theoretical sciences, while politics and, perhaps, jurisprudence, correspond to the practical ones (Aman 1968, 20–21; cf. Charaf 2004, 214–216, who excludes jurisprudence and dialectics from the classification).

40. Perhaps the most complete synthesis of the traditional Islamic and Aristotelian structures is to be found in the classification of al-Khwārizmī, who divided the sciences into “Sciences of religious laws (‘ulūm al-Shari‘ah)” and “Foreign sciences (‘ulūm al-‘agam al-falsafa)”. The Sciences of religious laws he subdivided into jurisprudence, dogmatic theology, grammar, scribal subjects, poetry and prosody, and history, while the Foreign sciences were subdivided into theoretical and practical sciences, with the former encompassing natural sciences, mathematics, and theology and the latter, ethics, domestic economy, and politics (Aman 1968, 34–37).

41. On the prevalence of foreign materials in the section on stories, fables, and magic, see Wellisch 1987, 24-25.

42. The subdivisions of this class were: “the philosophers of the natural sciences and of logic”, “men of learning who were geometricians, arithmeticians, musicians, calculators, astrologers, makers of instruments, and persons interested in mechanics and dynamics”, and “ancient and recent physicians” (Wellisch 1987, 21–24). The first two classes can, perhaps, be assimilated to the classes of “physics” and “mathematics”, respectively, within the theoretical sciences of the Aristotelian classification. However, the presence of logic in the first class and that of “persons interested in mechanics and dynamics” indicate that, even if the latter was the ultimate source of the distinction between the two classes, the classes were in process of redefinition; moreover, their separation from the subject of theology and their association with medicine clearly set them outside of a distinctively Aristotelian context.

43. Commentators have tended to attribute the decisive philosophical move in the turn to epistemology to the French philosopher René Descartes (1596–1650); see, e.g., Copleston 1994, 150–152; Deely 1994, 15; Dorter 1972, esp. 2–4; Grenz 1996, 63–65; Hill 2016, 129. According to a narrative of the historical development of early modern philosophy current since the early 19th century, Descartes inaugurated the tradition of rationalist philosophy, whose proponents valorized innate ideas as the ultimate basis for knowledge, in reaction to which an alternative empiricist view of knowledge was developed by philosophers such as John Locke (1632-1704) and David Hume (1711-1776), who argued that experience was the sole source of knowledge: these two opposing accounts of knowledge were partially reconciled by Immanuel Kant (1724-1804), who held that all knowledge derives from human experience of the world, which, however, is conformed to categories of understanding constituted by the mind. At stake in the debate between rationalist and empiricist philosophers was the question of the source and (possible) scope of human knowledge.´ Although latter-day historians of philosophy consider this received account to be vastly oversimplified and dubious as a historiographical schema (e.g., Vanzo, 2016), it has at least the merit of highlighting the shift in emphasis towards epistemology within philosophical discourse over the course of the 17th and 18th centuries.

44. The full title proper was Pandectarum sive partitionum universalium Conradi Gesneri Tigurini, medici & philosophiae professoris, libri XXI. As its wording makes clear, this second volume of the bibliography sought to establish “universal divisions” (partitiones universales) that would offer a comprehensive system (pandectae) for the ordering of literature. It should be noted that the Pandectae only contained the first nineteen of the projected twenty-one divisions. The final twenty-first division was published separately as the Partitiones theologicae, while the twentieth division, allotted to works on medicine, was never written. On all this, see Araújo, Sabba and Crippa 2016, 60–61.

45. The sole exception to this rule was the class of medicine, the subdivisions of which were never published. For a partial listing of the more important subdivisions in the first nineteen classes, see Šamurin 1964–1967, 1/118–123.

46. The main classes were subdivided by authors in alphabetical order, while, on the shelves, books were organized by size and subject. This inaugural catalog would remain in service for only fifteen years. In 1620, a new catalog was issued in which classified order was abandoned in favor of alphabetical listing by author’s surname. A third catalog, also in alphabetical order, was published in 1674, as were the fourth and fifth published in 1738 and 1843, respectively. See Norris 1939, 147-155.

47. Interestingly, an engraved representation of the Leiden University Library from 1610 depicts how these seven subjects were disposed on the library’s shelves; for a reproduction, see Burke 2000, 104. The engraving shows two columns of book presses facing the viewer, which are separated by a central aisle. There are eleven presses in each column, with each press consisting of a single shelf in to which the individual books, which apparently had to be read in situ, are attached. The name of the subject class of books held by each press is engraved on it. Beginning with the presses furthest from the viewer, the column on the left side consisted of six presses of books by theologians (theologici), two presses of books by literary authors (literatores), two presses of books by philosophers (philosophi), and one press of books by mathematicians (mathematici), while the column on the right has five presses of books by scholars of law (iurisconsulti), two presses of books by medical writers (medici), and four presses of books by historians (historici). Three points are worth noting here. First, presses pertaining to the three higher faculties of the university — theology, law, and medicine — predominate, with theology and law alone accounting for half of the collection. Second, the classes are named not according by subject but by kind of author characterized by subject — a pattern that recalls Callimachus’s Pinakes. Finally, the ordering of the classes — theology, literature, philosophy, mathematics, law, medicine, and history — does not cluster the higher faculty subjects together, as one might expect, but intersperses them among other the others.

48. It is worth noting that, in practice, Naudé did not follow this ideal order. The sequence of classes that he devised for the catalog of the Bibliotheca Cordesiana published in 1643 was, in greatly simplified form: Bible and theology; libraries; chronology, geography, and history; law; philosophy, mathematics, and medicine; politics; literature. The classes in this catalog, which were named after the kind of author characterized by subject, were further subdivided by size of book. See Serrai 1977, 134–135; Živný 1920, 56.

49. Leibniz’s inclusion of a class for “general and miscellaneous” works reflects an early modern development in library classification that would become more pronounced in the 18th and 19th century. See, for example, the eighteenth class of “philologists and polyhistors” in the ideal classification of the Jesuit scholar Claude Clement (1596-1643) in the Musei sive Bibliothecae tam privatae quam publicae extructio, instructio, cura, usus, a handbook of library economy published in Lyon in 1635 (Serrai 1977, 120–121; cf. Živný 1920, 57) who gives it as the ninth class of the classification and characterizes it as “polygraphy”); the final two classes of “miscellanea” and “manuscripts” in the classification of the Clementinum library at Charles University from 1747 (Živný 1920, 58); and the fifth main class of “polygraphy” and “dictionaries” given in Prosper Marchand’s classed catalog of Joachim Faultrier’s library, which appeared in 1709 (p. 63). Occurring only sporadically in the early modern period, Generalia classes would become a regular feature of modern library classifications.

50. Leibniz actually wrote out two separate outlines for an ideal library classification, a long one and a short one, which differed substantively from one another (Serrai 1977, 181). It is the shorter one that is mentioned here.

51. As outlined in Garnett (1877), the main classes of this classification were: theology, law, natural history and medicine, philosophy, history, geography, biography, belles lettres, and philosophy. For a convenient tabular representation of the classification, see Brown 1898, 45–48.

52. Garnier’s classification seems to have been the first library classification explicitly based on a theory of the faculties of the mind. It was not, however, the first classification of knowledge to take such an epistemological theory as its basis. The 16th-century Spanish humanist scholar Juan Huarte de San Juan (1529–1588/1589) had, in his tractate Examen de ingenios para las ciencias (1575) proposed a classification of sciences based on three basic powers of the soul — memory, intellect, and imagination (Jacobs 2002, 53–55; Serrai 1977, 77–78). Some thirty years later, the English philosopher and statesman Francis Bacon (1561–1626) proposed classification of sciences based on the same three faculties of mind in his treatise The Advancement of Science (1605), later expanded into the De Augmentis Scientiarum (1623) (see below). Šamurin (1964–1967, 1/156) has claimed that Bacon’s use of epistemological theory as a basis for the classification of science influenced Garnier in his choice of a rationale for his library classification: however, it may well be that Garnier drew inspiration from Huarte’s work, which, in some respects, seems closer to his. This is a question that calls for further investigation.

53. In addition to these subject-based classes, Garnier appended two special classes of books stored in the “museum rooms” of the library: prohibited heretical books and a collection of manuscripts and realia (Systema Bibliothecae Collegii Parisiensis Societatis Jesu 1678, 7; 9–10; 105–118).

54. Grenier identified six different principles of ordering — ordering by nature, ordering by doctrine, ordering by languages, ordering by time, ordering by social rank/status, and ordering by societies (i.e., schools of thought) — each of which had its specific rules: for example, according to natural order, genera should precede species, species should precede individuals, and the whole should precede its parts; according to the order of doctrines, doctrines should be divided by the human faculty to which they pertain and then arranged according to natural order; in the order of languages, “learned” languages (i.e., Hebrew, Greek, and Latin) should precede “common” languages; in ordering by time, the commonly agreed-upon and uncontentious dates should be used; in ordering by social rank/status, common accepted systems of social ranking should serve as a basis for arrangement; and in ordering by societies, or schools of thought, those with the most members or those occurring chronologically earlier than others should precede those with smaller numbers of members or occurring later in time. Grenier was well aware that different sections of the classification would require application of different principles of subarrangement: for example, he noted, arrangement by societies, or schools of thought would be largely restricted to the classes of theology and philosophy (Systema Bibliothecae Collegii Parisiensis Societatis Jesu 1678, 13).

55. Somewhat different in its articulation was the classified catalog of the library of Joachim Faultrier compiled by the bibliographer Prosper Marchand (1678–1756), the main classes of which were bibliography; philosophy, or, human science; theology, or, divine science; history, or, the science of events; polygraphic writings, and dictionaries (Živný 1920, 63; cf. Petzholdt 1866, 34–35). Noteworthy here is the addition of the form classes for bibliography and polygraphic works, the absence of law as a main class, and the position of philosophy before theology.

56. Cf. Brunet 1865, xv: “A certain way of avoiding […] inconveniences would be to renounce these encyclopedic sequences, [which are] so seductive at first glance, but the emptiness of which is revealed by practice. For, when all is said and done, in the classing of books, it is much less the natural — or the so-called natural — sequence of the sciences that it is necessary to consider than the real relation (le rapport réel) which they conserve among themselves in the usage that one makes of them, be it in the great and learned professions to which they are applied, or in the practice of life. We shall thus […] hold fast to the system of the bookshops of Paris […] not only because it is more generally known than all the others are, […] but especially because it is easily adapted to the nature of the books which are most ordinarily held by libraries great and small” [translation by TMD].

57. Harris (1870b, 119) gave the following explanation for inverting Bacon’s sequence of classes: “Science should come first on account of its furnishing the method and principles for what follows:

  1. SCIENCE gives the department of books in which conscious system prevails.
  2. ART (Aesthetics) gives the department in which “organic unity” or unconscious system prevails.
  3. HISTORY gives the department in which the system is determined by accidental relations, such as time and place.”

As modern commentators have noted, Harris was a Hegelian in outlook and this rationale for the rearrangement of Bacon’s classification of the sciences aligned its three principal classes with Hegel’s categories of Begriff (“concept”) (≈ science), Wesen (“essence”) (≈ art), and Sein (“being”) (≈ history), each of which represented a different moment in the dynamic ontological structure of universal becoming (Graziano 1959, 48-51; Sales and Pires 2017, 8). Further discussion of the relation between the Hegelian and Baconian elements of Harris’s classification can be found in Ferreira and Sales 2018.

[top of entry]


Al-Hassan, Ahmad Y. and Donald R. Hill. 1992. Islamic Technology: An Illustrated History. Cambridge: Cambridge University Press; Paris: Unesco.

Al Najjar, Abdul Majid. 1996. “Classification of Sciences in Islamic Thought: Between Imitation and Originality”. The American Journal of Islamic Social Sciences, 13, no. 1: 59–87.

Allen, Robert C. 2017. The Industrial Revolution: A Very Short Introduction. Oxford: Oxford University Press.

Aman, Muhammad Muhammad. 1968. Analysis of Terminology, Form and Structure of Subject Headings in Arabic Literature and Formulation of Rules for Arabic Subject Headings. Pittsburgh: University of Pittsburgh. (Ph.D. Dissertation)

Araújo, Andre Vieira de, Fiammetta Sabba and Giulia Crippa. 2016. “Semantic Order in the 16th Century: An Introductory Discussion of Conrad Gesner’s Pandectae.” In Knowledge Organization for a Sustainable World: Proceedings of the Fourteenth International ISKO Conference, 27-29 September 2016, Rio de Janeiro, Brazil, ed. José Augusto Chaves Guimarães, Suellen Oliveiri Milani, Vera Dodebei. Advances in Knowledge Organization 15. Würzburg: Ergon Verlag, 59–66.

Aristotelés. 2008. Metafyzika. Translated and annotated by Antonín Kříž. Prague: Rezek.

Austin, Derek. 1984. PRECIS: A Manual of Concept Analysis and Subject Indexing. Second edition. London: The British Library, Bibliographic Services Division.

Bacon, Francis. 1996. “The Advancement of Learning”. In Francis Bacon: The Major Works. Edited by Brian Vickers. Oxford: Oxford University Press, 120–299. [Originally published in 1605].

Bagnall, Roger S. 2002. “Alexandria: Library of Dreams”. Proceedings of the American Philosophical Society, 146, no. 4: 348-362.

Beghtol, Clare. 2010. “Classification Theory”. In Encyclopedia of library and information sciences, third edition. Edited by Marcia J. Bates and Mary Niles Maack. Boca Raton, FL: CRC Press, vol. 2, 1045-1060.

Berti, Enrico. 2014. “La classificazione aristotelica delle scienze in Pietro d'Abano”. Trans - Form - Ação, 37, n. 3: 17-32. Available online at: http://www.scielo.br/pdf/trans/v37n3/04.pdf.

Besson, Alain. 1980. Medieval Classification and Cataloguing: Classification Practices and Cataloguing Methods in France from the 12th. to 15th. Centuries. Library History Series, No. 4. Biggleswade, Beds: Clover Publications.

Birkenmajer, Aleksander. 1922. Bibljoteka Ryszarda de Fournival poety i uczonego francuskiego z poczatku XIII-go wieku i jej pózniejsze losy. Polska Akademja Umiejętności, Wydział Filologiczny. Rozprawy, T. LX, Nr. 4. W Krakowie: Nakładem Polskiej Akademji Umiejętności.

Blair, Ann M. 2010. Too Much to Know: Managing Scholarly Information before the Modern Age. New Haven/London: Yale University Press.

Bliss, Henry Evelyn. 1929. The Organization of Knowledge and the System of the Sciences. New York: Henry Holt and Company.

Bliss, Henry Evelyn. 1939. The Organization of Knowledge in Libraries. New York: The H.W. Wilson Company.

Blum, Rudolf. 1991. Kallimachos: The Alexandrian Library and the Origins of Bibliography. Madison, Wisconsin/London, England: The University of Wisconsin Press.

Blumenthal, Elke. 2011. “Privater Buchbesitz im pharaonischen Ägypten“. In Elke Blumenthal & Wolfgang Schmitz (eds.), Bibliotheken im Altertum. Wolfenbütteler Schriften zur Geschichte des Buchwesens, 45. Wiesbaden: Harrassowitz Verlag, 51–85.

Breisach, Ernst. 1994. Historiography: Ancient, Medieval, & Modern. Second edition. Chicago/London: The University of Chicago Press.

Brotton, Jerry. 2006. The Renaissance: A Very Short Introduction. Oxford: Oxford University Press.

Broughton, Vanda. 2015. Essential Classification, Second edition. Chicago: Neal-Schuman.

Brown, James D. 1898. Manual of Library Classification and Shelf Arrangement. London: Library Supply Company.

Brunet, Jacques-Charles. 1865. Manuel du libraire et de l'amateur de livres. Tome Sixième. Cinquième édition. Paris: Librairie de Firmin Didot Frères.

Buckland, Michael K. 1992. “Information Retrieval of More than Text”. Journal of the American Society for Information Science 42, no. 8: 586–588.

Buckland, Michael K. 1997. “What is a ‘Document’?” Journal of the American Soceity for Information Science 48, no. 9: 804–809.

Burkard, Günter. 1980. „Bibliotheken im alten Ägypten“. Bibliothek: Forschung und Praxis 4, no. 2: 79–115.

Burke, Peter. 2000. A Social History of Knowledge: From Gutenberg to Diderot. Cambridge: Polity Press.

Burke, Peter. 2012. A Social History of Knowledge, Volume II: From the Encyclopédie to Wikipedia. Cambridge, UK: Polity Press.

Burke, Peter. 2015. What is the History of Knowledge? What is History? Cambridge, UK/Malden, MA: Polity Press.

Campbell, Duncan. 2017. “Bibliography”. In Michael Dillon (ed.), Encyclopedia of Chinese History. London/New York: Routledge, 49-50.

Cassiodorus Senior. 1969. An Introduction to Divine and Human Readings. Translated with Introduction and Notes by Leslie Webber Jones. New York: W.W. Norton & Company, 1–64. [Originally published by Columbia University Press, 1946]

Chadwick, Henry. 1981. Boethius: The Consolations of Music, Logic, Theology, and Philosophy. Oxford: Oxford University Press.

Charaf, Inaam. 2004. “Conceptualisation and Organization of Knowledge between the 10th and 14th Centuries in Arabic Culture”. Knowledge Organization, 31, no. 4: 213–221.

Colish, Marcia L. 1997. Medieval Foundations of the Western Intellectual Tradition 400–1400. New Haven/London: Yale University Press.

Copleston, Frederick. 1994. A History of Philosophy, Volume 4: Modern Philosophy: From Descartes to Leibniz. New York: Image Books.

Courtenay, William J. 1988. “The Fourteenth-Century Booklist of the Oriel College Library”. Viator 19: 283–290.

Covington, Michael A. 2005. “Scientia Sermocinalis. Grammar in Medieval Classifications of the Sciences”. In Nicola McLelland, Andrew Lin (Eds.). Flores Grammaticae: Essays in Memory of Vivien Law. Münster: Nodus Publikationen, 49–54.

Crippa, Giulia. 2015. “Cassiodoro e as Institutiones Divinarum Litterarum como fonte histórica para a discussão sobre práticas bibliográficas e organização do conhecimento”. Informação & Informação 20, no. 2: 86–117. Available online at: http://www.uel.br/revistas/uel/index.php/informacao/article/view/23126

Cutter, Charles A. 1904. Rules for a Dictionary Catalog. Fourth Edition, Rewritten. U.S. Bureau of Education: Special Report on Public Libraries — Part II. Washington: Government Printing Office.

Dalby, Andrew. 1986. “The Sumerian Catalogs”. Journal of Library History, 21, no. 3: 475-487.

Day, Ronald E. 2001. The Modern Information of Information: Discourse, History, and Power. Carbondale/Edwardsville: Southern Illinois University Press.

Deely, John. 1994. New Beginnings: Early Modern Philosophy and Postmodern Thought. Toronto Studies in Semiotics. Toronto/Buffalo/New York: University of Toronto Press.

Delisle, Léopold. 1874. Le Cabinet des Manuscrits de la Bibliothèque Nationale: études sur la formation de la dépôt comprenant les éléments d’une histoire de la calligraphie de la miniature, de la reliure, et du commerce des livres à Paris avant l’invention de l’imprimerie. Tome II. Paris: Imprimerie Nationale.

Dorter, Kenneth. 1972. “First Philosophy: Metaphysics or Epistemology?” Dialogue: Canadian Philosophical Revue = Revue Canadienne de Philosophie 11, no. 1: 1–22.

Dousa, Thomas M. 2014. “Classificatory Structure and the Evaluation of Document Classifications: The Case of Constitutive Classification”. In Knowledge Organization in the 21st Century: Between Historical Patterns and Future Prospects: Proceedings of the Thirteenth International ISKO Conference, 19–22 May 2014, Kraków, Poland. Edited by Wieslaw Babik. Advances in Knowledge organization 14. Würzburg: Ergon Verlag, 208–215.

Dousa, Thomas M. 2017. “E. Wyndham Hulme’s Classification of the Attributes of Books: On an Early Model of a Core Bibliographical Entity”. Knowledge Organization 44, no. 8: 592–604.

Dudbridge, Glen. 2017. “Libraries, Book Catalogues, Lost Writings”. In Wiebke Denecke, Wai-Yee Li, & Xiaiofei Tian (eds.), The Oxford Handbook of Classical Chinese Literature (1000 BCE–900 CE). Oxford: Oxford University Press, 147–162.

Eisenstein, Elizabeth L. 1983. The Printing Revolution in Early Modern Europe. Cambridge, UK: Cambridge University Press.

Efron, Noah J. 2009. “Myth 9: That Christianity Gave Birth to Modern Science”. In Ronald L. Numbers (ed.), Galileo Goes to Jail and Other Myths about Science and Religion. Cambridge, Mass./London: Harvard University Press, 79–89.

Elayyan, Ribhi Mustafa. 1990. “The History of the Arabic-Islamic Libraries: 7th to 14th Centuries”. International Library Review 22: 119–135.

Encyclopédie, ou, Dictionnaire raisonné des sciences des arats et des métiers (articles choisis). 1986. 2 vols. Chronology, introduction, and bibliography by Alain Pons. Paris: Flammarion.

Fakhry, Majid. 1983. A History of Islamic Philosophy. 2nd edition. Studies in Asian Culture, 5. New York: Columbia University Press.

Ferreira, Verónica de Sá, & Sales, Rodrigo de. 2018. “Confluências filosóficas para os estudos de classificação das artes: aproximações de Bacon, Hegel e Harris”. Scire, 24, no. 1: 55–66. Available online at: https://www.ibersid.eu/ojs/index.php/scire/article/viewFile/4522/3906

Fraser, P. M. 1972. Ptolemaic Alexandria (2 volumes). Oxford: The Clarendon Press.

Garnett, Richard. 1877. “On the System of Classifying Books on the Shelves Followed at the British Museum”. Library Journal 2, nos. 3–4: 194–200.

Gilreath, James and Dougals L. Wilson (eds.). 1989. Thomas Jefferson’s Library: A Catalog with the Entries in his Own Order. Washington: Library of Congress.

Glorieux, P. 1963. “Études sur la ‘Biblionomia’ de Richard de Fournival”. Recherches de théologie ancienne et médiévale 30: 205–231.

Gnoli, Claudio. 2018. “Notation”. In Birger Hjørland (ed.), ISKO Encyclopedia of Knowledge Organization. Available online at: http://www.isko.org/cyclo/notation

Grant, Edward. 1996. The Foundations of Modern Science in the Middle Ages. Cambridge: Cambridge University Press.

Grant, Edward. 2001. God and Reason in the Middle Ages. Cambridge: Cambridge University Press.

Graziano, Eugene. 1959. “Hegel’s Philosophy as Basis for the Dewey Classification Schedule”. Libri 9, no. 1: 45-52.

Grenz, Stanley J. 1996. A Primer on Postmodernism. Grand Rapids, Michigan/Cambridge, UK: William B. Eerdmans Publishing Company.

Grimm, Alfred. 1989. „Altägyptische Tempelliteratur: Zur Gliederung und Funktion der Bücherkatalog von Edfu und et-Tôd“. In Sylvia Schoske (ed.), Akten des vierten internationalen Ägyptologen Kongresses München 1985, Volume 3. BSAK 3. Hamburg: Helmut Buske Verlag, 159–169.

Guthrie, Lawrence S., II. 1992. “An Overview of Medieval Library Cataloging”. Cataloging & Classification Quarterly, 15, no. 3: 93-100.

Guthrie, W.K.C. 1981. A History of Greek Philosophy. Volume 6: Aristotle: An Encounter. Cambridge, UK: Cambridge University Press.

Hall, Edith. 2015. “Adventures in Ancient Greek and Roman Libraries.” In Alice Crawford (ed.), The Meaning of the Library: A Cultural History. Princeton: Princeton University Press, 1-30.

Haq, Syed Nomanul. 2009. “Myth 4: That Medieval Islamic Culture was Inhospitable to Science.” In Ronald L. Numbers (ed.), Galileo Goes to Jail and Other Myths about Science and Religion. Cambridge, Mass./London: Harvard University Press, 35–42.

Harris, William Torrey. 1870a. “Essay on the System of Classification”. In Catalogue, Classified and Alphabetical, of the Books of the St. Louis Public School Library, Including, also, the Collections of the St. Louis Academy of Science, and St. Louis Law School. Prepared by John Jay Bailey. St. Louis: Missouri Democrat Book and Job Printing House, ix-xvi.

Harris, William Torrey. 1870b. “Book Classification”. Journal of Speculative Philosophy 4, no. 2: 114–129.

Haye, Thomas. 2010. “Canon ou Catalogue? Perspectives historico-littéraires dans la Biblionomia de Richard de Fournival”. Romania 128, nos. 509-510: 213-233.

Hider, Philip. 2012. Information Resource Description: Creating and Managing Metadata. London: Facet Publishing.

Hill, John Lawrence. 2016. After the Natural Law: How the Classical Worldview Supports Our Modern Moral and Political Values. San Francisco: Ignatius Press.

Hjørland, B. 2008. “What is Knowledge Organization (KO)?” Knowledge Organization 35: 86–101.

Hjørland, Birger. 2016. “Knowledge Organization (KO)”. Knowledge Organization 43, no. 6: 475–484. Available online at: http://www.isko.org/cyclo/knowledge_organization.

Hjørland, Birger. 2017. “Classification”. Knowledge Organization 44, no. 2: 97–128. Available online at: http://www.isko.org/cyclo/classification.

Hlaváček, Ivan. 2005. Knihy a knihovny v českém středověku: studie k jejich dějinám do husitství. Prague: Univerzita Karlova v Praze, Nakladatelství Karolinum.

Horsfall, Nicholas. 1993. “Empty Shelves on the Palatine”. Greece & Rome 40, no. 1: 58–67.

Houston, George W. 2014. Inside Roman Libraries: Book Collections and Their Management in Antiquity. Chapel Hill: The University of North Carolina Press.

Hout, Theo P. J. van den. 2002. Miles of Clay: Information Management in the Ancient Near Eastern Hittite Empire. Published as electronic article at the University of Chicago Library, Fathom Archive. Available online at: http://fathom.lib.uchicago.edu/1/777777190247/.

Hugo von Sankt Viktor. 1997. Didascalicon de Studio Legendi = Studienbuch. Übersetzt und Eingeleitet von Thilo Offergeld. Fontes Christiani, 27. Freiburg: Herder.

Hulme, E. Wyndham. 1950 [1911-1912]. Principles of Book Classification. A.A.L. Reprints No. 1. London: The Association of Assistant Librarians. [Originally published in: Library Association Record 13, 1911: 354-358, 389-394, 449; 14, 1912: 39-44, 174-181, 216-221]

IFLA Working Group on Guidelines for Subject Access by National Bibliographic Agencies. 2012. Guidelines for Subject Access in National Bibliographies. Edited by Yvonne Jahns. IFLA Series on Bibliographic Control. Berlin/Boston: Walter de Gruyter.

Jacobs, Helmut C. 2002. Divisiones Philosophiae: Clasificaciones Españoles de las Artes y las Ciencias en la Edad Media y el Siglo de Oro. Translated by Beatriz Galán Echevarría. Madrid: Iberoamericana.

Jiang, Shuyong. 2007. “Into the Source and History of Chinese Culture: Knowledge Classification in Ancient China”. Libraries & the Cultural Record 42, no. 1: 1-20.

Jones, Leslie Webber. 1969. “Introduction”. In Cassiodorus Senator, An Introduction to Divine and Human Readings. Translated with Introduction and Notes by Leslie Webber Jones. New York: W.W. Norton & Company, 1–64. [Originally published by Columbia University Press, 1946].

Joudrey, Daniel N., Arlene G Taylor and David P. Miller. 2015. Introduction to Cataloging and Classification. Eleventh edition. Santa Barbara/Denver: Libraries Unlimited.

Kaiser, J[ulius]. 1911. Systematic Indexing. The Card System Series, Vol II. London: Sir Isaac Pitman & Sons, Ltd.

Kilgour, Frederick G. 1998. The Evolution of the Book. New York/Oxford: Oxford University Press.

Kurth, Dieter (trans.). 1994. Treffpunkt der Götter: Inschriften aus dem Tempel des Horus von Edfu. Zürich/München: Artemis Verlag.

LaMontagne, Leo E. 1961. American Library Classification with Special Reference to the Library of Congress. Hamden, Connecticut: The Shoe String Press, Inc.

Le Goff, Jacques. 2017. Must We Divide History into Periods? Translated by Malcolm DeBevoise. European Perspectives. New York: Columbia University Press. [Originally published as Faut-il vraiment découper l’histoire en tranches? Paris: Éditions du Seuil, 2014].

Lee, Hur-Li. 2008. “Origins of the Main Classes in the First Chinese Bibliographic Classification.” In Culture and Identity in Knowledge Organization: Proceedings of the Tenth International ISKO Conference, 5-8 August 2008, Montréal, Canada. Edited by Clément Arsenault and Joseph Tennis. Advances in Knowledge organization 11. Würzburg: Ergon Verlag, 275–281.

Lee, Hur-Li. 2012a. “Epistemic Foundation of Bibliographic Classification in Early China: A Ru Classicist Perspective”. Journal of Documentation 68, no. 3: 378–301.

Lee, Hur-Li. 2012b. “Praxes of Knowledge Organization in the First Chinese Library, the Seven Epitomes”. In Richard P. Smiraglia and Hur-Li Lee (eds.), Cultural Frames of Reference. Würzburg: Ergon Verlag, 63–77.

Lee, Hur-Li and Wen-Chin Lan. 2009. “Purposes and Objectives of a Pioneer Library Catalog in China.” Library Quarterly 79, no. 2: 205–231.

Lee, Hur-Li and Wen-Chin Lan. 2011. “Proclaiming Intellectual Authority through Classification: The Case of the Seven Epitomes”. Knowledge Organization 38, no. 1: 25–42.

Lewis, C.T. and C. Short. 1987. A Latin Dictionary founded on Andrews’ edition of Freund’s Latin Dictionary. Oxford: at the Clarendon Press. [First published in 1879].

Lindberg, David C. 1978. “The Transmission of Greek and Arabic Learning to the West”. In David C. Lindberg (ed.), Science in the Middle Ages. The Chicago History of Science and Medicine. Chicago: The University of Chicago Press.

Long, A. A. and D. N. Sedley. 1987. The Hellenistic Philosophers (2 vols.). Cambridge, UK: Cambridge University Press.

Mann, Margaret. 1930. Introduction to Cataloging and the Classification of Books. Chicago: American Library Association.

Mariétan, Joseph. 1901. Problème de la classification des sciences d’Aristote a St-Thomas. Paris: Felix Alcan; St. Maurice: Imprimerie St.-Augustin.

Marrou, H[enri]-I[renée]. 1964. A History of Education in Antiquity. New York: Mentor Books.

Marrou, Henri-Irenée. 1976. Patristique et Humanisme: Mélanges. Patristica Sorbonensia, 9. Paris: Éditions du Seuil.

Marshall, Peter. 2009. The Reformation: A Very Short Introduction. Oxford: Oxford University Press.

Max, Gerald E. 1994. “Ancient Near East”. In Wayne A. Wiegand and Donald J. Davis, Jr. (eds.), Encyclopedia of Library History (pp. 23–31). New York/London: Garland Publishing.

Miksa, Francis [L.]. 1983. The Development of Classification at the Library of Congress. University of Illinois Graduate School of Library and Information Science Occasional Papers, Number 164. Champaign: University of Illinois Graduate School of Library and Information Science.

Miksa, Francis L. 1992. “The Concept of the Universe of Knowledge and the Purpose of LIS Classification”. In N.J. Williamson and M. Hudon (eds.), Classification Research for Knowledge Representation and Organization: Proceedings of the 5th International Study Conference on Classification Research, Toronto, Canada, June 24-28, 1991. Amsterdam: Elsevier, 101–126.

Miksa, Francis L. 1994. “Classification”. In Encyclopedia of Library History. Edited by Wayne A. Wiegand and Donald G. Davis, Jr. New York/London: Garland Publishing, Inc.,

Miksa, Francis L. 1998. The DDC, The Universe of Knowledge, and the Post-Modern Library. Albany, New York: Forest Press.

Mills, Jack. 1968. A modern outline of library classification (6th impr.). London: Chapman & Hall.

Müller, Carl Werner. 2011. „Griechische Büchersammlungen und Bibliotheken vom sechstren Jahrhundert v. Chr. bis in hellenistische Zeit“. In Elke Blumenthal & Wolfgang Schmitz (eds.), Bibliotheken im Altertum. Wolfenbütteler Schriften zur Geschichte des Buchwesens, 45. Wiesbaden: Harrassowitz Verlag, 101–122.

Naudé, Gabriel. 1627. Advis pour dresser une bibliothèque. A Paris: chez François Targa.

Nicholls, Matthew. 2010. “Bibliotheca Latina Graecaque: on the Possible Division of Roman Public Libraries by Language.” In Yves Perrin (éd.), Neronia VIII: Bibliothèques, livres et culture écrite dans l’empire romain de César à Hadrien: Actes du VIIIe Colloque international de la SIEN (Paris, 2–4 october 2008). Collection Latomus, Vol. 327. Bruxelles: Éditions Latomus, 11–21.

Norris, Dorothy May. 1939. A History of Cataloguing and Cataloguing Methods, 1100-1850: With an Introductory Survey of Ancient Times. London: Grafton & Co.

Okasha, Samir. 2002. Philosophy of Science: A Very Short Introduction. Oxford: Oxford University Press.

Olesen-Bagneux, Ole. 2014. “The Memory Library: How the Library in Hellenistic Alexandria Worked.” Knowledge Organization 41, no. 1: 3–13.

Olesen-Bagneux, Ole. 2015. “Mnemonics in the Mouseion: Considerations on Spatial Mnemonics as a Tool for Classification and Retrieval”. Journal of Documentation 71, no. 2: 278–293.

Olson, Hope A. 2010. “Earthly Order and the Oneness of Mysticism: Hugh of Saint Victor and Medieval Classification of Wisdom”. Knowledge Organization 37, no. 2: 121–138.

Onions, C. T., G. W. S Friedrichsen and R. W. Burchfield (eds.). 1966. The Oxford Dictionary of English Etymology. New York/Oxford: Oxford University Press.

Parpola, Simo. 1983. “Assyrian Library Records”. Journal of Near Eastern Studies 42, no. 1: 1–29.

Pedersén, Olof. 1998. Archives and Libraries in the Ancient Near East. Bethesda: CDL Press.

Petzhold, Julius. 1866. Bibliotheca Bibliographica: Kritisches Verzeichniss der das Gesammtgebiet der Bibliographie Betreffenden Litteratur des In- und Auslandes in systematischer Ordnung. Leipzig: Verlag von Wilhelm Engelmann.

Potts, D. T. 2002. “Before Alexandria: Libraries in the Ancient Near East”. In The Library of Alexandria: Centre of Learning in the Ancient World, edited by Roy MacLeod (pages 19–34). London/New York: I.B. Tauris.

Principe, Lawrence M. 2011. The Scientific Revolution: A Very Short Introduction. Oxford: Oxford University Press.

Rajna, Pio. 1928. “Le denominazioni Trivium e Quadrivium”. Studi medievali 1: 4–36.

Ranganathan, S. R. 1967a. Prolegomena to library classification. Third edition, assisted by M. A. Gopinath. New York: Asia Publishing House.

Richardson, Ernest Cushing. 1901. Classification: Theoretical and Practical. New York: Charles Scribner’s Sons.

Riché, Pierre and Jacques Verger. 2011. Učitelé a žáci ve středovĕku. Praha: Vyšehard.

Ross, David. 1964. Aristotle. University Paperbacks, vol. 65. London/New York: Methuen.

Sales, Rodrigo de and Thiago Blanch Pires. 2017. “The classification of Harris: Influences of Bacon and Hegel in the Universe of Library Classification”. In Program of the North American Symposium on Knowledge Organization, Champaign, Illinois, June 15-16, 2017. Available online at: http://www.iskocus.org/NASKO2017papers/NASKO2017_paper_5.pdf.

Šamurin, E. I. 1964–1967. Geschichte der bibliothekarisch-bibliographischen Klassifikation. Zwei Bande. Leipzing: VEB Bibliographisches Institut.

Santoro, Michele. 2015. “Per una storia delle classificazioni bibliografiche. Parte seconda: la Grecia antica”. Bibliotime, 18, no. 2. Available online at: http://www.aib.it/aib/sezioni/emr/bibtime/num-xviii-2/santoro.htm.

Satija, Mohinder P. 2000. “Library Classification: An Essay in Terminology”. Knowledge Organization 27, no. 4: 221–229.

Sayers, W. C. Berwick. 1955. An Introduction to Library Classification: Theoretical, Historical and Practical with Readings, Exercises and Examination Papers. 9th edition. London: Grafton & Co.

Sayers, W. C. Berwick and Arthur Maltby. 1975. Sayers' Manual of Classification for Librarians. 5th ed. London: André Deutsch.

Schulte-Albert, Hans. 1971. “Gottfried Wilhelm Leibniz and Library Classification”. Journal of Library History 6, no. 2: 133-152.

Segre, Mario. 1935. “Epigraphica”. Rivista di Filologia e d’Istruzione Classica, nuova serie 13 (= 53): 214-225.

Serrai, Alfredo. 1977. Le classificazioni: idee e materiali per una teoria e per una storia. Biblioteconomia e bibliografia, 10. Firenze: Leo S. Olschki Editore.

Shera, Jesse. 1965. Libraries and the Organization of Knowledge. Hamden, Connecticut: Archon Books.

Svenonius, Elaine. 1989. “An Ideal Classification for an On-line Catalog”. In Classification Theory in the Computer Age: Conversations Across the Disciplines. Proceedings from the Conference, November 18–19, 1988, Albany, New York. Albany, New York: Nelson A. Rockefeller College of Public Affairs and Policy, University at Albany, State University of New York, 35–43.

Svenonius, Elaine. 2000. The Intellectual Foundation of Information Organization. Cambridge, Massachusetts/London, England: The MIT Press.

Systema Bibliothecae Collegii Parisiensis Societatis Jesu. 1678. Parisiis: Excudebat Sebastianus Mabre-Cramoisy.

Thiers, Christophe. 2004. “Fragments de théologies thébaines: la Bibliothèque du temple de Tôd”. Bulletin de l’Institut Français de l’Archéologie Orientale 104: 553–572.

Tsien, Tsuen-Hsuen. 1952. “A History of Bibliographic Classification in China”. The Library Quarterly 22, no. 4: 307–324.

Turner, James. 2014. Philology: The Forgotten Origins of the Modern Humanities. Princeton/Oxford: Princeton University Press.

Vanzo, Alberto. 2016. “Empiricism and Rationalism in Nineteenth-Century Histories of Philosophy”. Journal of the History of Ideas 77, no. 2: 253–282.

Weaver, Stewart A. 2015. Exploration: A Very Short Introduction. Oxford: Oxford University Press.

Webb, Kerry. 2013. “’The House of Books’: Libraries and Archives in Ancient Egypt”. Libri 63, no. 1: 21–32.

Weisheipl, James. 1965. “Classification of the Sciences in Medieval Thought”. Mediaeval Studies 27: 54–90.

Weisheipl, James. 1976. “The Nature, Scope, and Classification of the Sciences”. In David C. Lindberg, Science in the Middle Ages. Chicago/London: The University of Chicago Press, 461–82.

Weitemayer, Mogens. 1956. “Archive and Library Technique in Ancient Mesopotamia”. Libri 6, no. 3: 217–238.

Wellisch, Hans H. 1981. “Ebla: The World’s Oldest Library”. The Journal of Library History 16, no. 3: 488–500.

Wellisch, Hans H. 1986. The First Arab Bibliography: Fihrist al-‘Ulum. Occasional Papers, 175. Champaign: University of Illinois Graduate School of Library and Information Science. Available online at: https://www.ideals.illinois.edu/bitstream/handle/2142/3825/gslisoccasionalpv00000i00175.pdf.

Wessetzky, Vilmos. 1958. “Az Óegyptomi Könyvtár”. Antik Tanulmanyok 5: 149–159.

Wessetzky, Vilmos. 1973. „Die ägyptische Tempelbibliothek: Der Schlüssel der Lösung liegt doch in der Bibliothek des Osymandyas?“ Zeitschrift für ägyptische Sprache und Altertumskunde 100: 54–59.

Wessetzky, Vilmos. 1984. „Die Bücherliste des Tempels von Edfu und Imhotep“. Göttinger Miszellen 83: 85–89.

Wilkins, M. Lesley. 1994. “Islamic Libraries to 1920”. In Wayne A. Wiegand & Donald J. Davis, Jr. (eds.), Encyclopedia of Library History. New York/London: Garland Publishing, 296–306.

Witty, Francis J. 1958. “The Pinakes of Callimachus”. The Library Quarterly 28, no. 2: 132-136.

Wright, Alex. 2014. Cataloging the World: Paul Otlet and the Birth of the Information Age. Oxford: Oxford University Press.

Živný, Ladislav J. 1920. Třídění bibliologické: Úvod do theorie a praxe třídění bibliotečního a bibliografického. V Praze: Stanislav Minařík.

[top of entry]


Visited Hit Counter by Digits times.

Version 1.0; published 2018-11-27, last edited 2021-12-21
Article category: History of classification and KO

©2018 ISKO. All rights reserved.