The verb "warrant" means, according to WordNet 2.1, "show to be reasonable or provide adequate ground for" and "stand behind and guarantee the quality, accuracy, or condition of". 

In relation to knowledge organization (KO) has Clare Beghtol (1986) provided the following explanation of this concept:

"In general, the warrant of a classification system can be thought of as the authority a classification invokes first to justify and subsequently to verify decisions about what classes/concepts to include in the system, in what order classes/concepts should appear in the schedules, what units classes/concepts are divided into, how far subdivision should proceed, how much and where synthesis is available, whether citation order  are static or variable and similar questions. Warrant covers conscious or unconscious assumptions and decisions about what kinds and what units of analysis are appropriate to embody and to carry the meaning or use of a class to the classifier, who must interpret both the document and the classification system in order to classify the documents by means of syntactic devises. The semantic warrant of a system thus provides the principal authorization for supposing that some class or concept or notational device will be helpful and meaningful to classifiers and ultimately to users of documents" (Beghtol, 1986, 110-111).

The term "literary warrant" as well as the basic principle underlying this expression was introduced in the literature of Library and Information Science (LIS) by E. Wyndham Hulme (1911, p. 447). Hulme discusses whether, for example, the periodic system of chemistry should be used for book classification. He writes (p. 46-47):

"In Inorganic Chemistry what has philosophy to offer? [Philosophy here meaning science, which produced the periodical system]. Merely a classification by the names of the elements for which practically no literature in book form exists. No monograph, for instance, has yet been published on the Chemistry of Iron or Gold.

. . .

  Hence we must turn to our second alternative which bases definition upon a purely literary warrant. According to this principle definition is merely the result of an accurate survey and measurement of classes in literature. A class heading is warranted only when a literature in book form has been shown to exist, and the test of the validity of a heading is the degree of accuracy with which it describes the area of subject matter common to the class. Definition [of classes or subject headings], therefore, may be described as the plotting of areas pre-existing in literature. To this literary warrant a quantitative value can be assigned so soon as the bibliography of a subject has been definitely compiled. The real classifier of literature is the book-wright, the so-called book classifier is merely the recorder. " Hulme (1911, p. 46-47).

If Hulme's suggestion has some similarity with bibliometrics this is no accident. It was Hulme who in 1923 initiated the term "statistical bibliography" which was later changed to bibliometrics. Hulme used the term to describe the process of illuminating the history of science and technology by counting documents.

In LIS the term "literary warrant" thus means that that an indexer or classifier has to provide adequate ground for the indexing, classifying (as well as the definition of indexing terms and classes in classification systems) in the literature. Warrant is also the justification for the inclusion of a term or a class in a controlled vocabulary as well as its definition and relations to other terms.   

One thing is, however, to express this principle in abstract terms. Another thing is to specify how this should be done concretely. Beghtol (1986) explains how the principle of literary warrant has been interpreted differently. The Classification Research Group, for example, did not follow Hulme by basing warrant on book titles, but actually based it on the terminology of a subject field.

In the literature have other kinds of "warrant" been suggested, including "user warrant", "scientific warrant", "educational warrant" and "cultural warrant". (Beghtol (1986) uses the term "semantic warrant" as a generic term). Interestingly, Beghtol (1986) does not discuss user warrant, which is important in relation to user-oriented approaches to knowledge organization (seeUser and User Studies in KO).

NISO (1994) defines two kinds of warrant:

2.1. Literary warrant
"words and phrases drawn from the literature of the field should determine the formulation of descriptors. When two or more variants have literary warrant… most frequently used term should be selected as the descriptor."

2.2. User warrant
"justification for the representation of a concept in a [thesaurus] or for the selection of a preferred term because of frequent requests for information on the concept." (See also Request oriented indexing).

How is literary warrant done in practice? There seems to be no empirical studies on how this principle is interpreted in practice. Riesthuis (2005) writes that the third edition of the ASIS&T thesaurus is to a higher degree than previous editions based on literary warrant. It has no descriptor for all countries, only for the countries about which have been written in the publications on which it is based.

Bibliometric Knowledge Organization should be regarded as a a form of applying literary warrants, especially the method suggested by Schneider (2004), identifying candidate terms by the citations in full-text documents.

 

Scientific warrant.

Henry Bliss saw scientific warrant as consensus among scientists: "Bliss, then, believed that the fundamental authority that infused meaning into a bibliographic classification system was the best philosophical and scientific consensual thinking that was available to the classificationist and that only on this foundation could a classification system be created that would have relatively permanent validity and usefulness" (Beghtol, 1986, p. 115).

Educational warrant. In addition to Beghtol's (1986) description of the concept of educational warrant, attention should be turned to Wallerstein (1996) for a concrete example of how educational institutions may be responsible for creating the warrant of semantic relations:

"Thus, between 1850 and 1945, a series of disciplines came to be defined as constituting an area of knowledge to which the name “social science” was accorded. This was done by establishing in the principal universities first chairs, then departments offering courses leading to degrees in the discipline. The institutionalization of training was accompanied by the institutionalization of research: the creation of journals specialized in each of the disciplines; the construction of associations of scholars along disciplinary lines (first national, then international): the creation of library collections catalogued by disciplines”. (Wallerstein, 1996, p. 30).

Cultural and epistemological warrant. In Beghtol's paper (1986) cultural warrant is described as the insight that classifications and semantic relations are depending on the broader cultural context. In is not described, however, how the construction of knowledge organizing system (KOS) (or the indexing of specific documents) should take this insight and turn it into an approach to KO. If the collector of terms for a KOS is aware that different cultural and epistemological views are always at play in every field of knowledge, then at least some conscious action may be taken (although there are deep dilemmas involved in making decisions: both in declaring some views better than others and also in giving up any standpoint). The approach to KO that most explicitly deals with the issue of different cultural and epistemological views is the domain analytic approach to KO (see also: Culture and epistemology in KO).

The principle of literary warrant is widely recognized. For example, Mitchell (2001, p. 217) writes about Dewey Decimal Classification "Since the DDC is developed on the basis of literary warrant, associative relationships are often treated as equivalence or hierarchical relationships because that is how they are treated in the published literature".

The Library of Congress Subject Headings also applies this principle:

"The Library of Congress collections serve as the literary warrant (i.e., the literature on which the controlled vocabulary is based) for the Library of Congress subject headings system. The number and specificity of subject headings included in the Subject Authority File (the machine-readable database containing the master file of Library of Congress subject headings from which the printed list, the microform list, the CDMARC SUBJECTS, etc. are generated), are determined by the nature and scope of the Library of Congress collections. Subject headings are established as they are needed to catalog the materials being added to the collection or to establish links among existing headings. In recent years, headings contributed by libraries engaged in cooperative activities with the Library of Congress based on the needs of their collections have also been included.

3.2 References:

H 187 When To Establish A New Heading:

"Establish a subject heading for a topic that represents a discrete, identifiable concept when it is first encountered in a work being cataloged, rather than after several works on the topic have been published and cataloged."

H 198 Creating Authority Records For Headings Formerly Unprinted In LCSH (p.2):

"If, in establishing a new heading or in creating an authority record for a heading that previously had none, it is necessary to use, as a broader term in a 5XX field, another heading that lacks an authority record, prepare an authority worksheet also for the heading needed for use as a broader term...Continue upward in the hierarchy in this manner until no additional headings lacking authority records are encountered.""http://www.itsmarc.com/crs/shed0163.htm

It is important to realize that the principle of literary warrant (or other kinds of warrant) introduces an empirical principle in knowledge organization. What is classified? "Universe of knowledge" is an expression with a somewhat metaphysical sound without any empirical basis. The classes in classification systems are (mainly) based on the literature that they classify. In 1969 was published a Danish dissertation on autokinesis, which is a visual illusion. This dissertation was indexed in the Danish national bibliography and the concept was added to the index to the Danish Decimal Classification, DK5. There may be several hundred kinds of visual illusions which are not labeled in DK5, just as the generic class "visual illusion" is not a class in this specific system either (but may be formed if the monographic literature about this subject at a future point in time produce a warrant for it). The important implication is that classification systems based on literary warrant are not well-suited to classify other kinds of documents (or other samples of documents) than those they are based on - such as journal articles or literatures from other cultures. 

Criticism of literary warrant:

"Gnoli et al. concludes observing that "most KOS justify their disciplinary structure by the assumption that users, while searching for information, will follow the disciplinary organization they are familiar with. This may be an effective way to reproduce the literary warrant faithfully. However, the function of KO is not only to represent the existing literature, but also to suggest new paths of research through the discovery of relations in published knowledge. To the latter purpose, cross-disciplinary relations must be representable and made searchable. Projects like Szostak's and ILC go in this direction" [p 406]. " http://www.iskoi.org/ilc/leon.htm   (from: "The León manifesto")

Mooers (1972) may, perhaps, also be considered a critic of the principle of literary warrant:

"It was on these foundations that the descriptor retrieval systems were constructed: 1) The clear definition of the user community and its actual scope of interest; 2) the recognition of the substantial difference between the desired kind of retrieval clues based on ideas in comparison to narrative message terminology; and 3) the focus upon retrieval by invariant idea-elements rather than upon the literal verbal expression found in the documents."(Mooers, 1972).

Mooers does not directly say that the ideas are not to be found in the literature, but rather that the specific expressions found there should not be used.

Reference:

Beghtol, C. (1986). Semantic validity: Concepts of warrant in bibliographic classification systems. Library Resources & Technical Services, 30(2), 109-125.

Beghtol, C. (1995). Domain analysis, literary warrant, and consensus, the case of fiction studies. Journal of the American Society for Information Science, 46(1), 30-44.  

Haby, S. (2003). Literary warrant and specificity in ScoT: A discussion paper

Hulme, E. W. (1911). Principles of Book Classification. Library Association Record, 13:354-358, oct. 1911; 389-394, Nov. 1911 & 444-449, Dec. 1911. 

Mitchell, J. (2001). Relationships in the Dewey Decimal Classification System. IN: Bean, C. A. & Green, R. (Eds.). (2001). Relationships in the organization of knowledge. Dordrecht: Kluwer Academic Publishers. (Pp. 211-226).

Mooers, C. N. (1972). Descriptors. IN: Encyclopedia of Library and Information Science. Ed. by Allen Kent & Harold Lancour. Vol. 7. New York: Marcel Dekker. (Pp. 31-45).

NISO (1994). National Information Standards Organisation (1994). ANSI/NISO Z39.19-1993 Guidelines for the construction, format and management of monolingual thesauri. Bethesda, MD: NISO Press.

Ogg, N. J.; Sievert, M. E.; Li, Z.R. & Mitchell, J. A. (1994).  Construction of a medical informatics 

thesaurus. Journal of the American Medical Informatics Association, S, 900-904.  

Olson, H. A. (2004). Bacon, warrant and classification. ASIS&T SIG-CR WORKSHOP

SATURDAY, NOVEMBER 13, 2004.  

Riesthuis, G. (2005). Book review of: ASIS&T Thesaurus of Information Science, Technology, and Librarianship. 3rd ed. Medford, N.J.:

Information Today. Knowledge Organization, 32(4), 159-160.

Schneider, J. W. (2004). Verification of bibliometric methods' applicability for thesaurus construction.

Aalborg: Royal School of Library and Information Science. (PhD-dissertation).

Vizine-Goetz, D. & Beall, J. (2004). Using Literary Warrant to Define a Version of the DDC for Automated Classification Services. ISKO Conference, London, 13-16 July 2004.

Wallerstein, I. (1996). Open the Social Sciences, report of the Gulbenkian Commission on the Restructuring of the Social Sciences. Stanford, Calif.: Stanford University Press. 

WordNet 2.1