special subdivisions of subjects. The special student whose work is prosecuted in one of these "compartments" of the subject, as we may not inappropriately call it, finds a principal and very serviceable use of the general system of subject cataloging, as affecting his own studies, in the extent to which the various subject catalogs and subject lists serve to bring under his eye and to his immediate attention the existence of bibliographical helps in departments outside his own narrow specialty, indeed, but with a close bearing upon his own department. I will quote at this point from a letter so strongly confirmatory of the above statement, from the specialist's point of view, as to form an appropriate accompaniment to it. Dr. H. B. Adams, in charge of the Department of History and Politics at Johns Hopkins University, writes as follows in relation to the work of teachers and students at that university, and the aid furnished by subject catalogs and subject lists. He says: "From the standpoint of a student of history, I may say that without such aids teachers and pupils would often be as helpless as a traveller without a map or a guide-book. No specialist or any other seeker after historical truth can possibly be so familiar with the entire range of literature in a given field as to be above the necessity of consulting good bibliographies, published catalogs, etc." "In the practical workings of our department library in this university, the card catalogs, and the published catalogs of the Boston Athenæum, Brooklyn, and Boston Public Library, Poole's Index, etc., are found to be of the greatest service. Our own subject catalog is in constant use from morning till night. Graduate and undergraduate students run to it as men do to dictionaries, encyclopædias, historical atlases, and other works of reference. If they do not find what they want, they go to the Peabody Library, and consult the more elaborate subject catalogs of that institution" (card catalogs). "There are some things that a good student or a good library must know. First, what information is at hand upon a given subject; second, if nothing is at hand, where some thing can be found. Both the student and the library are driven to subject catalogs and subject lists, for a proper systematizing of their own collections." The second of the two objections, however, raises a question of no small importance,namely, that this work of subject cataloging is labor lost, because duplicating what is done elsewhere. The tendency of library work to-day in every department is emphatically against unnecessary duplicating, and as emphatically in favor of "doing a thing once for all," even in such external and mechanical details as registration forms and delivery systems. In these departments its importance is obvious, and it is none the less essential in the field of cataloging. There is perhaps no more urgent and perplexing problem demanding consideration in the management of what we may call the smaller libraries, those, for instance, of from 10,000 to 20,000 volumes, than just how far to carry the work of subject cataloging. Time and attention. have been devoted for years past to interesting these smaller libraries in this very direction. Now that there seems to be a more widespread tendency to engage in this work than ever before, and when we have abundant reason to be encouraged at these results, we are confronted with this new and very real cause of perplexity. Take it, for example, as related to the question of dealing with the contents of composite works, and of analytical entries in the catalog, covering such instances as volumes of essays, periodicals, proceedings of societies, reports of labor and other boards, and other similar instances. Now there never has been any question as to the desirableness of getting at the material hidden away in these composite or serial publications. The one practical question is the avoiding of waste or duplication of labor. One of the most important of these fields of analytical exposition has been very thoroughly laid open-and once for all - since the date at which Mr. Cutter and Mr. Noyes began the publication of their extraordinarily valuable catalogs. I mean, of course, the field of periodical literature in Poole's Index. In no subsequent catalog of an individual library, therefore, can it ever again be worth while entirely to duplicate this work on a separate scale. Of the almost equally indispensable material known under the general term of essays, we have not as yet an index, on the scale of Poole's Index. It can hardly be doubted, however, that the inevitable appearance of this "Index to General Literature" is only a question of time, and we are assured of this in Mr. Fletcher's very promising "reports of progress." For these (in many instances) and the "series" (like the Massachusetts Historical Society, Chaucer Society, etc.), we have the work done for us in catalogs such as those named above (the Boston Athenæum, etc.); and it seems, for the present at least, to be the part of wise economy, in the case of those libraries which are not blessed with so extensive funds as these larger ones, to make use of the clues thus generously furnished in these catalogs, rather than to construct duplicates of them for themselves; and, I will still further add, to make them more directly available by entering their own book-numbers on the margin. - and others there is no duplication of the [See also Poole's Index. 1st Supp.] [See Brooklyn Catal.] We may look, I think, for no abandonment of the subject catalog principle in the future. That principle is the inseparable accompaniment of the new and promising lines of library work in the direction of the specializing of reading for the general reader, which are becoming more common and more deeply rooted every year. But we may look, I think, for a wiser economy in the adaptation of the various means to this very desirable end. I have on an earlier occasion expressed my own conviction that one phase of the solution to this problem lies in the preparation of special subject lists, as occasion arises, as well as in the utilization of all available printed subject lists and catalogs; and that in the "bibliographical economy of the future we shall see two lines of library work advance side by side — namely, a general cataloging of the entire library, but within definitely prescribed limits as to fullness, and, on the other hand, the exhibition of the library's resources on particular topics, as occasion arises, with the utmost exhaustiveness possible at the time. And yet, long experience in the use of these catalogs, and of material similar to them in my own library, and the improbability that in that library we shall ever be able to make any very lavish expenditure for cataloging, have made me very ready to seize upon any means of still further incorporating their benefits into our catalogs, and of thus To this may now be added the suggestion, utilizing them to the fullest degree. As a already widely followed, as I am glad to find, result of the consideration given to this in many of the smaller libraries, of incorpomatter, I have adopted the following plan for rating this material, prepared when some a consolidated catalog, which I am expecting special occasion arises, into the library's to put very soon into the printer's hands. card catalog, and thus making it of permaIt is briefly as follows: In subject entries nent rather than ephemeral service. For the discussion on this paper, see PROCEEDINGS (Fifth session). A REPORT ON CLASSIFICATION. BY RICHARD BLISS, LIBRARIAN REDWOOD LIBRARY, NEWPORT, R. I. SAILOR in one of our seaport towns once explained his preference for attending a church where a responsive service was used rather than one of another denomination, on the ground that it was a satisfaction to be able to "jaw back" at the minister. Now the reporter on classification -a preacher for the time being has no wish that any of his fellow-members should refrain from "jawing back" in this service. On the contrary, in order that sufficient inducement to that innocent amusement may be afforded, he purposes making a few remarks, interspersed with criticism, on classification in general as preliminary to his report, which is, in itself, more of a commentary than a text. And as librarians are notoriously quite as sensitive on the subject of their bibliothical offspring as natural parents are in regard to their bodily issue, he doubts not that the privilege the sailor prized will be appreciated here also. So far as the arrangement of books is concerned, librarians may be divided into three groups the anti-classifiers, the pseudo-classifiers, and the classifiers, sometimes called close-classifiers. The anti-classifiers either arrange the books as they come in, without regard to juxtaposition of subject, or they adopt some sort of group-arrangement, perhaps on the mnemonic plan, which does duty for a classification. The pseudo-classifiers are primarily classifiers who have been dismayed by the difficulty of obtaining a satisfactory system, or who have been influenced by the animadversion of the anti-classifiers, and try to sit on two stools at once, with the usual result. As the former do not come within the scope of this paper, the reporter has nothing to report on them or their works. Coördination of knowledge must ever be a difficult matter, and a perfect systematic arrangement is perhaps impossible; but any one who will review the history of classification for the past twenty years will see how much progress has been made during that time in the systematization of knowledge. Past experience would seem to invalidate Mr. Fletcher's conclusion that "nothing better in the way of systems is to be hoped for than those we now have." One of the chief difficulties the classifier has to contend with is found in the nature of the subject itself. The interdigitation of certain branches of knowledge and their farreaching relationships seem to render any lineal gradation impossible. This is one of the points seized upon by the opponents of closeclassification as an argument for the uselessness of any attempt to classify at all. Folklore, for example, is a division which has many alliances, namely, with ethnology, religion, medicine, sociology, art and literature, and librarians may differ widely as to where to put it. But must we therefore conclude that there is no close relationship to be found in the subdivisions of science, philosophy, or the useful arts? The question how far classification shall be carried is one which, notwithstanding all that has been said upon it, is far from settled. Even the most ardent classifier must admit that there is a point beyond which classification cannot be extended to either with satisfaction or advantage; for, the farther one subdivides the less close is the relationship of the subdivisions. But it does not therefore follow that classification should stop with the main divisions of knowledge, and that it makes no difference whether aeronautics is put in cosmology, as in Mr. Perkins's “rational" classifi cation, or in arts, as in Mr. Cutter's Athenæum classification. In a small library it is not necessary to divide descriptive botany, for example, into subheads. To such it is of no sort of consequence whether Epilobium angustifolium belongs to the Onagraceæ, and the Onagraceæ to the Polypetalæ, or not, but it does make a difference to a botanical library like that at the Botanic Gardens in Cambridge whether or not suitable subdivisions are provided for the numerous specialties of which it is composed. Assuming, then, that some sort of a shelf classification is desirable, the point to be determined is what sort of one shall it be. On this point librarians are greatly at variance one with another. Much stress has recently been laid by certain of our members on the value of what they call a "rational " or "natural" classification as distinguished from a "logical" or "scientific " one, as if the system they advocated was alone rational and the others more or less artificial. The In a recent number of the Library journal Mr. Fletcher contrasts the analytic or synthetic with the logical or practical methods of classification by defining the former to be a treating of the whole realm of knowledge as a unit, and working down through subdivision to the minute subject; while the latter takes the individual book as a unit, and works up through aggregation to the entire library. This is the old argument translated into philosophic terms; but the difference is more apparent than real, since the book to be classified is the unit in both instances. The distinction is not unlike the difference between unheading a barrel of apples and consuming from the top down, and turning the barrel upside down, knocking the bottom out, and consuming from that end. Now, a "rational" classification must needs be one which is judicious or constructed in conformity to reason, and whether any given arrangement is judicious depends upon whether it adequately serves its purpose or not, a point which can only be determined upon trial. The appropriation of the term "rational" as a distinguishing designation for this or that arrangement is a petitio principii which is less valid than it is common. terms rational, natural, logical, and scientific, as applied to classificatory systems, apparently connotate quite diverse ideas in the minds of the individuals using them. Thus with some, "natural" seems limited to the idea of geographical or chronological progression, as shown in the arrangement of geography or history, ignoring the fact that there may be a natural evolutionary progress of other subjects corresponding to a natural mental progress of ideas. With others, " rational" is used as a contradistinctive term to close classification, which is thereby assumed to be irrational, another case of the petitio principii. A natural classification is one which follows some natural order, either subjective or objective, in the arrangement of its topics, and a logical classification one which follows definite principles, and conforms its succession of divisions to certain inferences according to the laws of thought. How a classification which is either or both of these can be other than rational, it is difficult to see. Much of mere dialectics would be avoided and time saved if the terms rational and logical, as applied to classification, were abandoned and syncretic and systematic substituted. The analytic or synthetic method Mr. Fletcher identifies with the system of close classification, whereof he is an earnest opponent, the disadvantages of which he affirms are, in effect: (1) the demand on the time and mental powers of the compiler; (2) the complicated notation involved; (3) the difficulty with which it is comprehended by the uninitiated; (4) its failure to show the resources of the library on a given subject, and (5) the tendency to encourage reference to the imperfect representation on the shelves instead of to the catalogue and to bibliographies. Of these objections, the second is the only really important one. Minute subdivision does necessarily entail long class marks, the objection to which lies quite as much in the difficulty of readily distinguishing the mark on the books, when closely arranged together on the shelves, as in the liability to error in callslips and charging. The latter objection is rendered invalid where such a system of charg ing is used as that in vogue in the Boston Athenæum, and it is not easy to see how any marks used to designate books in a large library can be entirely free from the former objection. Furthermore, Mr. Cutter's revised classification is tolerably minute, and the class marks are neither long nor complicated. Whether the demand on the time and patience of the compiler be a real objection or not, depends upon the relative worth of such a classi fication in itself. If a close classification serve a useful purpose, the time given to its construction is of little moment. To the third objection it may be replied that the librarian ought to be acquainted with all systems; if not, there is the Library School, designed to supply the defect, and that it is not necessary, though it may be profitable, that the public should know the meaning of the class marks. That a close classification on the shelves will not exhibit all the resources of the library on a given subject, is admitted; neither will any other system that has been or may be invented. The close classifier claims that this system shows more of the related subjects than do the others. Which, for instance, best exhibits the resources of a library, say on physics, Mr. Cutter's classification, where all the divisions of natural philosophy are grouped together under that head, or Mr. Schwartz's, where statics and dynamics are separated from optics by the whole of paleontology, and electrics from both by mineralogy, mathematics, and geology. Even in Mr. Fletcher's "rational" classification house sanitation is divorced from drainage and sewerage by carpentry, masonry, chemical technology, manufactures, mining, and bridge building. Lastly, the final cause of a library is to enable people to get what they want in the shortest possible time. If a person can do that more quickly from books than from cards, why refuse him the privilege by insisting that he shall take the slower way? As well urge a man not to use a bicycle in hurrying for a physician because that mode of progression is less safe and natural than the act of walking. The foregoing remarks may fitly serve as an introduction to a somewhat hasty review of the three or four new classifications which have appeared since the last regular meeting of the A. L. A., two of which have been constructed upon what has been called the scientific basis. Of these by far the most elaborate and thorough is Mr. Cutter's Revised Classification, which I have seen in manuscript. It was compiled originally for the Cary Library, at Lexington, but has been so arranged by its author as to be applicable either to a large or to a very small library without change. This has been accomplished by marking the divisions which will be needed for a small, or for a very small library, and leaving the rest, or as much of it as may be deemed necessary, for the larger library. One of the principal defects of most of the systems hitherto proposed has been a rigid inelasticity in the matter of enlargement or contraction. However complete the list of topics, space will surely be needed for those necessarily overlooked since no man is omniscient or for new subjects which the progress of human thought and activity calls into being. In the Dewey system, one of the most rigid of all systems, this is overcome rather than provided for by the intercalation of the new subject as near as possible to its congeners, with an additional figure to the previous class-mark, perhaps already too long. Others like Mr. Perkins leave certain numbers blank, to be filled up as needed. The objection to this latter method is the same as that to the fixed location in shelving; spaces left will sooner or later be filled up, and the rigidity of the system will necessitate a break in its logical sequence. In its perfect adaptability to varying conditions lies one of the chief excellences of the Cutter clas sification, such as is possible only in a systematic arrangement. It matters not how large the library grows to be, filling up is impossible. If applied to a very small library, only the principal subdivisions are used; if to a small library, as many of the subdivisions as may be necessary; while for a large one the arrangement permits an indefinite extension. So far as known to the reporter, Mr. Cutter's Revised Classification is the only system which allows unlimited contraction or expansion without rearrangement or an objectionable addition to the class-mark. In a note prefixed to the classification Mr. Cutter points out that the advantages to a small library using the shorter form are easier consultation of the table, requiring less knowledge and thought, and a consequent saving of time and labor. The disadvantage is that if the library grow rapidly some of the classes will have to be rearranged; i. e. broken up into smaller divisions. He thinks it better, therefore, for a small and growing library to use more of the divisions than he has selected. In his notation for the classes Mr. Cutter uses mainly letters, thereby avoiding the composite-looking marks employed in the Boston Athenæum classification, which are apt to suggest to the frivolous mind the idea of a combat between the alphabet and the multiplication table. Although the classes are thoroughly subdivided, so admirable is the arrangement of the letters that no long class-mark is used; |