The Cutter Expansive Classification system has been called "the most scholarly and logical of American classification schemes" (Winke 123) and "the most philosophical, logical and closely co-ordinated system in existence" (anonymous, qtd. in Bowman 154). Its influence has been felt throughout American cataloging, with its schedules and notation used as the basis for the Library of Congress Classification (W. Cutter, 45; Miksa, Library Systematizer 61; Winke 123). Its system of book-marks and the alphabetic tables used to produce them have been incorporated into many other classification schemes, even being adopted by libraries using the Dewey Decimal System (Stump 44), once seen as the Expansive Classification's chief rival. The system is remarkable for the minuteness of its subdivisions, its brevity of marks, the integrity of symbols and its adaptiveness to local needs.
Although the Expansive Classification System was adopted by as many as fifty-seven American, Canadian and British libraries, only four libraries use the system as their primary classification scheme today (Winke 124). Francis Miksa writes that "His [Cutter's] classification scheme has for the most part become a footnote to classification history—an example of an outmoded nineteenth century exercise in subject enumeration" (Library Systematizer 13) and R. Conrad Winke suggests that the Expansive Classification System "remains barely a footnote in the history of modern classification" (123).
The system was initially criticized for its notation (Bowman 154), which mixed letters and numbers and was therefore thought to be confusing (Miksa, "Nineteenth-Century" 548), and for its minute subdivisions (Miksa, Library Systematizer 58), but most modern commentators blame the system's current obscurity on the its unfinished state at Cutter's death. Winke writes, "Cutter did not actively promote his classification and made no provisions for its continued revision and publication after his death" (123). In fact, Cutter promoted the scheme extensively and was greatly disappointed by his continued inability to procure the funds necessary to finish and publish the work (Miksa, Library Systematizer 25, 61). Certainly, some of these difficulties were due to Cutter's humble character, as contrasted with Dewey's blatant self-promotion, but it is my belief that the scheme's lack of popularity, and therefore the cause of Cutter's difficulties in procuring support for the project, was primarily due to Cutter's conception of the Expansive Classification as an adaptable system which could be implemented according to local needs. The resulting flexibility made the system difficult to implement, difficult to understand, and too often allowed libraries to use the system without benefitting from its many innovations. As Miksa writes:
Librarians, ever more pressed with growing collections, wanted a shelf classification system that was simple to understand and easy to apply. Despite testimonies to its superior enumeration of subjects, Cutter's system encountered increasing dissatisfaction because it was too adaptable. It had too many places where the classifier had to decide between alternatives or synthesize class-numbers especially adapted to their own library collections. They wanted, in other words, a scheme that entailed many fewer decisions in its application to any particular library—one that could be used with consummate technical efficiency. (Library Systematizer 64)
The Cutter Expansive Classification System is a collection of guidelines, suggested practices, and classification schedules meant to aid a librarian wishing to implement a shelf classification scheme suited to a particular collection without the trouble of creating such a system herself. Like other shelf classification schemes, the Expansive Classification assigns a notation to each book in a library's collection so that arranging the books according to an order on the notations collocates books with similar subjects or forms.
The Expansive Classification consists of seven separate sets of schedules, referred to collectively as the classifications. The first classification divides a library's collection into eight classes. It is sufficient only for the smallest libraries. Each successive classification introduces new classes and subclasses, carefully planned so that a growing library may reclassify its collection with a minimal amount of effort. In The First Six Classifications, Cutter explains:
In the ordinary methods of arranging and re-arranging libraries a small classification is made at first with classes and class-marks chosen somewhat at random, and with no provision whatever for growth, probably because there is no prevision that growth is to come. The consequence is that when a re-arrangement is to be made the old classification is thrown aside, an entirely new one is adopted, and every book in the library has to be reclassified, a process so long and costly that many libraries stagger on in the bonds of an utterly insufficient classification, because they dread to make a change, tho every year, by adding new books, renders the change more necessary and harder to make. In the classification here set forth, on the contrary, the classes chosen are parts of a carefully prepared whole, and the notation is such that other classes, which are sure to be needed in a library grown larger, can be intercalated without changing the classes already in use, except by taking some books out of them. (23)
Although Cutter gives several alternatives for the notation they all share a basic form in which the notation is divided into two halves: a class-mark, and a book-mark. The class-mark, which precedes the book-mark and is separated from it by punctuation, is used to group books on a common subject. The book-mark follows the class-mark and is used to arrange books within a class.
The notation used by the Expansive Classification includes letters, numbers, and punctuation. The system does not dictate a given notation, but gives the cataloger choices in assigning notations so as to serve local needs. In Cutter's preferred method, notations are compared from left to right, character by character, producing what a mathematician would call a lexicographic order. Numbers are compared to numbers and letters to letters in the natural way, two coming before three and B coming before C. When a letter and a number must be compared the number always comes first, and punctuation comes before both letters and numbers. Punctuation itself is left unordered—when ordering the notations the placement of punctuation is significant; the type of punctuation is not.
Inspired by Josephus N. Larned of the Buffalo Public Library, Cutter incorporated the idea of the integrity of symbols into the Expansive Classification. Cutter's original intention would have allowed a call number to be broken down into a number of notationally and semantically distinct units. The numbers 1 to 9 would be used only to indicate form and aspect, 5, for example, being used to indicate dictionaries, and 9 to indicate collections of works by several authors. The numbers 11 to 99 would be used only to denote geographical regions. The letters A to Z would be for classes and their subdivisions (Miksa, Library Systematizer 58). Despite these intentions, however, additional notations are used, especially in the larger classifications, and these uses sometimes interfere with the integrity of symbols. More than two digits are frequently used for a geographical subdivision and single digits and letters are sometimes used to for purposes other than those given above. These ambiguities are rarely a problem for one familiar with the schedules, but might be confusing for someone just becoming acquainted with the system.
Cutter recommended the use of four punctuation marks: a dot "•", a plus sign "+", two vertical lines "||" (C. Cutter, Expansive Classification 3), and the colon ":" (C. Cutter, Arts of Communication 40). The first three signs are placed between the class-mark and the book-mark, with "+" indicating quartos and folios, if they are to be shelved together but separately from the smaller books, and with "+" used for quartos and "||" for folios if the folios also require their own section as well (C. Cutter, Expansive Classification 3). The dot is also used to divide the parts of complicated book-marks (C. Cutter, Expansive Classification 144-145). Finally, the colon is used within the mark to insure the proper order of classes. Because punctuation comes before letters and numbers, the insertion of a colon into a mark can change a book's place on the shelves while leaving the mark's sequence of letters and numbers unchanged (C. Cutter, Arts of Communication 40; Forbes Lib., Revised schedules). (Cutter occasionally used the dot for this purpose, instead of the colon (C. Cutter, Expansive Classification 134-138; C. Cutter, Social Sciences 16), but I have not found a discussion of this usage in the schedules.)
Cutter's preferred notation for the book-mark generally consisted of a Cutter number indicating the author of the work. The Cutter number could be preceded by a single digit to indicate form or aspect. When more than one work by the same author fell within the same class the Cutter number would be followed by a sequence of letters known as the work-mark. If necessary, a number or date would follow to distinguish editions. Further notations added to the end of the book-mark could indicate translations, or even a commentary or dictionary that treats a particular work (C. Cutter, Expansive Classification 143-145).
Cutter also recommended an alternative book-mark for very prolific authors such as Shakespeare, Dante, and Goethe. These marks take into account the varying editions, translations, retellings, selections, etc. that the librarian may wish to distinguish and collocate in a large library's collection. They also provided subdivisions for doubtful and spurious works, derived works, criticism, bibliography, and other related materials (C. Cutter, Arts of Communication 49-74).
Class-marks, on the other hand were created from left to right according to a series of subdivisions. One or more letters, selected from the schedules, would indicate the main class, which could then, when appropriate, be further subdivided. Further subdivisions could be formed by taking the Cutter number of a key word or name, or by using the geographic notation. Both types of subdivision end the mark with numbers, and so additional letters would sometimes be used to further subdivide the class.
Special notations were available to indicate certain standard subdivisions. In addition to the local list, such subdivisions include those by historical period (C. Cutter, Expansive Classification 130-1), those by language (C. Cutter, Arts of Communication 6), and many notations specific to subdivisions in a given class.
"I have drawn out three notations for use with the present set of classifications," Cutter writes in The First Six Classifications, "as I know that some persons object to one feature of the one I prefer." (6). Here, Cutter is referring to the notations used for geographic subdivision, known as the local list. In addition Cutter presents four distinct notations for the book mark (12-3), two notations for chronological arrangement (130), three additional methods of geographic arrangement for libraries that wish to collocate all materials on a given country (C. Cutter, "Subject Divisions") and, throughout the work, provides more choices regarding notation than can be enumerated here.
Cutter also gives several alternate method of ordering the classes, relating to comparing letters and numbers after geographic subdivisions (C. Cutter, "Local List" 28) and comparing letters and numbers generally (C. Cutter, "Local List" 31). The method adopted does not affect the notation, but does have an effect on what materials will be brought together.
These alternatives in method and notation are few when compared with the decisions to be made when implementing the schedules. The purpose of the schedules is to enumerate existing classes and to provide a strict linear order both for existing classes and all new classes which may come to be introduced. Cutter was well aware that no such ordering could be perfect. "It is plain," he wrote, "that no shelf-classification can collect in one place all that the library possesses on each subject; because . . . some books treat of several subjects and yet can be put in only one place" (C. Cutter, Expansive Classification 5). Furthermore, as a result of his work in different libraries and his extensive correspondence Cutter knew well that different libraries had different needs when it came to collocation. A college library, for example, might wish to shelve all materials on a given country, including its history, geography, arts, and languages, in one place (C. Cutter, "Subject Divisions"). Libraries with special theatre collections might wish to shelve books on stage architecture and machinery with other books on theatre instead of with architecture, and might want to shelve them with books on scene-painting that would otherwise appear with the visual arts (C. Cutter, Expressive Arts 2). In such cases Cutter generally included options to achieve each end in his schedules. Miksa writes, "With regard to what kind of basic order was best, Cutter could make no final judgment," and goes on to quote Cutter, "You cannot say that any arrangement is the best. Everything depends on the end which you wish to reach" (Miska, "Nineteenth-Century" 540). The following remark, from the schedules of the seventh classification, is typical of Cutter's attitude:
Some works which from their title seem to be general histories of literature, treat only of the literature of Europe. Some persons may think it best to put all the histories of European literature both in general and in any of its periods or forms with the general histories of literature [as opposed to having a special class for European literature]. (C. Cutter, Arts of Communication 34)In other cases Cutter is less generous, marking a certain usage as preferred, or even marking some usages as "not recommended"—but he enumerates them nevertheless. Here Cutter recommends a particular usage:
This set of classes [HEA to HEZ] may be limited to the general works, in which case the local books (such as The iron industry of Pittsburg) would be put in class HE11-99, or both general and local works on a particular industry may come here, HE11 to HE99 being confined to works treating of several industries of a place. The latter course is the best, because HEA to HEZ can be subdivided locally, if necessary, but HE11 to HE99 cannot so well be divided by subject. (C. Cutter, Social Science 3)And again:
Each library must choose whether or not to put purely string quartets, quintets, etc., here [in VYC] or in VZS (strings), piano concertos here or in VZQ, music for the wood-wind instruments here or in VZWO. I prefer to use VZW, VZS and VZWO, confining VYC to collections of music for several combinations of instruments and to single pieces which do not come under either of those classes. (C. Cutter, Expressive Arts 13)In another example Cutter explicitly states that he does not recommend a usage:
An intermediate course is possible,—to put all the translations of the fiction of any literature immediately after all the originals, marking them G instead of F . . . I tried this in the Boston Athenaeum, and I cannot recommend it. (C. Cutter, Expansive Classification 38)Cutter's disapproval for the alternatives he supplies can be harsh:
Operas, Music dramas, Comic operas, Operettas, Opéra bouffe. Whether full-score, piano and voice, or voice with any other instrument [are put in the class YVO]. . . It is useless [emphasis added] to separate (by the mark VYOB) comic operas, etc, from grand operas. (C. Cutter, Expressive Arts 14)
Additional alternatives were given with respect to the single digit numbers used to indicate form or aspect. Though Cutter provided nine figures, he did not require their use and, in fact, considered several of them to be superfluous.
My experience has been that I very seldom have occasion to use 1; 2 and 3 I do not use at all, preferring to put all subject bibliography in ZW, and all biography in E; 8 is used for the Charters, Rules, Histories Proceedings, and many of the publications of societies, but it seems to me better to break up the sets of many societies, whose works have only the coherence of a name, and put the volumes under their respective subjects. (C. Cutter, Expansive Classification 130)
Librarians who wished to implement the Expansive Classification had many decisions to make regarding the alternatives described above. Many decisions about the schedules might be delayed at first, but questions of notation and the order of notations would have to be made before any classification could begin. The librarian would also have to decide which classification to use, for while the expansive nature of the Expansive Classification would aid the librarian who was forced to reclassify, reclassification would still be highly undesirable.
To understand the issues involved with implementing the classification it would be best to examine the practices of several libraries using the system now and in the past, as well as the observations of their catalogers. Here, however, I will only discuss the Expansive Classification at the Forbes Library, in Northampton Massachusetts.
Cutter published the first six classifications of the Expansive Classification from 1891 to 1893, and so it was natural, when he became Forbes Library's first librarian in 1894, for him to apply the Expansive Classification to this new collection. Since that time librarians at the Forbes have continually added new classes to the schedules, devised new notations and made decisions regarding usage. When examining the library's current practices it not always possible to determine when a given practice was adopted. Many decisions were doubtless made by Cutter himself; others were certainly made by later librarians.
To completely describe the current implementation of the Expansive Classification would involve reprinting the entire revised schedules. The description that follows simply highlights some areas where Forbes' practices differ from those recommended by Cutter in the original schedules and illustrates some of the issues the librarians have faced using the system over the last hundred years.
Forbes library uses a different method of ordering the marks than that initially suggested by Cutter. The marks are still compared character by character, from left to right, but the rule for comparing numbers to letters is different. Whereas numbers always come before letters in Cutter's preferred method, Forbes Library puts letters before numbers if the preceding character was a number and puts numbers before letters if the preceding character was a letter. At the beginning of a mark, or if the preceding character was punctuation, numbers are taken to come before letters (Forbes Lib., "Shelving Rules"). This method is easier to understand than it is to explain, and, as Cutter himself suggested, it has has important advantages when using the local list (C. Cutter, "Local List" 28). An example will illustrate the method and the reason for its use:
| Cutter's Preferred Method | Method Used at Forbes Library | ||
|---|---|---|---|
| F86 | Southern States -- History | F86 | Southern States -- History |
| F863 | Virginia -- History | F86Y | Southern States -- Social Conditions |
| F86Y | Southern States -- Social Conditions | F863 | Virginia -- History |
Forbes Library uses this method throughout the library with one exception: marks beginning "YF.9MS:" and "YF.9SF:" are put after the other marks beginning "YF.9", so that, for example "YF.9N423" comes before "YF.9MS:Ar18" even though N comes after M. These marks, which were not part of Cutter's schedules but introduced by a later cataloger at Forbes, are used to separate mystery and science fiction short story collections from more general collections of shorts stories. (A better choice of marks would have preserved both the method of comparing marks and the order of the books on the shelves. Marking mystery collections "YF.9:MS:" and the science fiction collections "YF.9:SF:" would, by introducing a second colon, place these collections together as intended, though before the general short story collections, rather than afterwards where they properly belong. A more drastic solution would be to mark the general short stories "YF9", incorporating the form division into the class mark, and the mystery and science fiction short stories "YF9MS" and "YF9SF". This would place short stories after the general fiction, and the mysteries and science fiction short stories after the general short stories.)
Books of fiction by a single author are marked with the class-mark alone, "YF", instead of with the book-mark alone as Cutter recommended (C. Cutter, Expansive Classification 17).
Forbes Library makes frequent use of the form-marks, with 9 used in nearly every class to indicate collections of works by multiple authors. In addition to Cutter's recommended form divisions Forbes makes occasional use of two more. Because Cutter had assigned a meaning to each of the nine digits an alternative notation had to be used. In both cases a two-digit number was adopted: 31 was introduced to located biographies of single lives near the collective biographies already indicated by 3; 06 was introduced to separate guidebooks for countries from guidebooks for "cities, states and groups of states" in the same class (Forbes Lib., Revised schedules). Though these notations resemble those from the local list, and thus conflict with the integrity of symbols, there is little confusion in practice.
Until recently Biscoe date letters were used to indicate time periods, as Cutter recommended (C. Cutter, Expansive Classification 130), however the library has now abandoned their use, preferring to indicate time periods with a date instead. (Cutter's table of Biscoe date letters did not allow for years past 1899, a fact which was a factor in this decision.)
Forbes library uses different classifications for different collections. In particular the main collection is classified according to the seventh classification, the children's collection primarily according to the sixth classification, and videos are classified primarily according to the fourth classification. In every case the seventh classification is used in such areas as history where the minuteness of its subdivisions have been considered desirable.
The schedules have been continually annotated with refinements, changed practices, and additional classes. The history of such changes is poorly documented, but evidence of the work that has been done can be seen by comparing the original schedules with those used today. Some handwritten and typewritten annotations to the schedules have also been preserved. In examining one section of the seventh classification, VY-VUUW, I found that the original thirty classes had been supplemented by an additional twenty-seven new classes. Seven classes from the original schedules are no longer used and at least two new classes were created only to be abandoned or assigned a new notation at a later time. One class was replaced altogether, its contents being moved into another and its notation used for a new purpose. Of the original thirty classes three required an explicit decision to be made between alternatives listed in the schedules and many more required additional clarification or refinement. In at least one case, the question of where to shelve biographies of the lives of single actors, an initial decision was made and later revised.
Cutter anticipated the need to revise the schedules to accommodate new knowledge, but he was certain that existing knowledge once well classified would remain well classified:
Buildings become too small, become antiquated, decay, are abandoned; but geografy [sic] does not become history; the natural sciences are not metamorphosed into the social sciences; mathematics will never be theology; fiction remains fiction, the drama the drama, poetry poetry, as long as literature and libraries last. (qtd. in Miksa Nineteenth-Century 542)
Experience however has shown that this is not so, as was illustrated when Forbes revised the schedules to move books on homosexuality from medicine to the social sciences (Forbes Lib., Revised Schedules).
It should be noted that in revising and updating the schedules Forbes Library has, naturally, chosen to ignore those alternatives not adopted locally. The consequence of this is that even though the Forbes Library has invested considerable time in maintaining the schedules, the revisions would be of only limited use to a library which had implemented the classification differently. Each revision would have to be examined and modified to conform with local practices.
Whether it was because of his belief that a classification should be tailored to local needs or simply out of politeness, Cutter was extremely hesitant to dictate a single and consistent practice in his classification. Perhaps it was because he so cherished the freedom he had enjoyed implementing his own scheme that he was blind to the fact that the abundance of choices he provided in the Expansive Classification would make it both difficult to implement and maintain.
Ironically, Cutter was an advocate for cooperation and uniformity of method between libraries. He wrote:
I take it many librarians would be very glad to have the book's places fixed for them without any trouble. And if, in the course of a generation or two, some such practice should become general, think what a saving,—to have one man do the whole classification for fifty, or a hundred or a thousand libraries, instead of fifty or a hundred or a thousand men doing each the same work independently. (qtd. in Miksa, "Nineteenth-Century" 532)And yet his own scheme hardly fixed the place of books without trouble and often seemed to encourage not uniformity of method but a diversity of practices. Surely this, more than any other, is the reason for the Expansive Classification's limited usage.
The largest flaw in the argument laid out above is my own lack of familiarity with other classification schemes. For the argument to hold it must be shown that the more popular classification schemes did not possess the same flexibility in practice as the Expansive Classification. Most importantly, a comparison should be made to the Dewey Decimal Classification, in its first proposals, its earliest schedules, and in its development since that time.
Further examination of contemporary reactions to the introduction of the Expansive Classification would be invaluable. Published accounts will be the easiest to find, but the reactions of the less prominent librarians who would have been infrequently published if at all would be more telling in this case. Inquiries should be made as to why specific libraries rejected or abandoned the system. Miksa's doctoral dissertation, "Charles Ammi Cutter: Nineteenth-Century Systematizer of Libraries," which I have not had time to read thoroughly, is sure to address many of these concerns.
A thorough analysis of the implementation and development of the classification at a variety of libraries might prove instructive. This would be very difficult work. Libraries do not frequently document the history of such work, regularly keeping little more than a description of current methods. In hindsight, my method of tracing the evolution of the system by comparing original and revised schedules was cumbersome and ineffective. A well thought out method should be devised before attempting a complete survey of this sort.
Bowman, J. H. "Classification in British Public Libraries: A Historical Perspective." Library History 21 (2005): 143-173. Library, Information Science & Technology Abstracts with Full Text. EBSCOHost. Simmons College Library GSLIS Collection. 2 Dec. 2007 <http://0-web.ebscohost.com.library.simmons.edu/>.
Cutter, C[harles] A[mmi]. Expansive Classification: Part 1: The First Six Classifications. Boston: C. A. Cutter, 1891-93. Various pag. Page numbers in citations refer to the first sequence.
[---]. Arts of Communication by Language. 2nd ed. n.p : n.p. : n.d. [From the schedules of the seventh classification.]
[---]. Expressive arts : synopsis ; mimetic art, theatre VT, VU ; music VV-VZ ; art W ; arts dependent on language X-Z. n.p. : n.p, n.d. [From the schedules of the seventh classification.]
[---]. "Local List." C. Cutter, Expansive Classification. Independent pag.
[---]. Social Sciences. n.p : n.p. : n.d. [From the schedules of the seventh classification.]
[---]. "Subject Divisions Under Countries." C. Cutter, Expansive Classification. Independent pag.
Cutter, William Parker. Charles Ammi Cutter. Chicago: American Library Association, 1931.
Forbes Library. Annotations to C. Cutter, Expressive arts. Aqusisition date and vendor number stamped on t.p. is Jun 25 1912 27. Forbes Lib., Northampton, MA.
---. "Shelving Rules." Circulation Desk Manual. Misc. ts. in binder. 2004. Continually updated. Forbes Lib., Northampton, MA. 4 Dec. 2007.
---. Revised schedules of the Expansive Classification. Unpublished. N. pag. Forbes Lib., Northampton, MA.
Miksa, Francis L., ed. Charles Ammi Cutter: Library Systematizer. Littleton, CO: Libraries Unlimited, 1977.
---. "Charles Ammi Cutter: Nineteenth-Century Systematizer of Libraries." Diss. U of Chicago, 1974. Dissertations & Theses: Full Text. ProQuest. Simmons College Library GSLIS Collection. 3 Dec. 2007 <http://0-proquest.umi.com.library.simmons.edu/>.
Stump, Sheryl, and Rick Torgerson. "The Basics of LC and Dewey." Mississippi Libraries 68.2 (2004): 43-5. Library Literature & Information Science Full Text. WilsonWeb. Simmons College Library GSLIS Collection. 3 Dec. 2007 <http://0-hwrelay.hwwilsonweb.com.library .simmons.edu/>.
Winke, R. Conrad. "The Contracting World of Cutter's Expansive Classification." Library Resources & Technical Services 48.2 (2004): 122-9. Library Literature & Information Science Full Text. WilsonWeb. Simmons College Library GSLIS Collection. 25 Sept. 2007 <http://0-hwrelay.hwwilsonweb.com.library.simmons.edu/>.