College and Research Libraries By M O R T I M E R T A U B E The Theory of Book Selection Mortimer Taube is head of the Order Department of Duke University Library. P a r t I TH E T H E O R Y of book selection is a branch of the general theory of value since it is concerned with problems of choice and discrimination between com- peting values. Like ethics, aesthetics, or any other branch of value theory, it has two aspects, the descriptive and the norma- tive. T h e descriptive part of the theory attempts to present a discussion of the bases of choice that are actually operative in practice, of the considerations that de- termine the selection of this book rather than that. T h e normative part is con- cerned with what ought to be the basis of decision in the light of general deci- sions concerning the objectives of libraries, the purpose of reading, or the aims of education and scholarship. T h e first part of this paper will be re- stricted to a discussion of the descriptive aspect of the theory, with a view of identi- fying and describing a number of criteria of value employed in book selection. In the second part there will be presented certain reasons why these criteria ought or ought not to be used, and the proper subordination of one criterion to another will also be suggested. M o s t librarians who employ these cri- teria do so without being conscious of them and in the day to day practice of book selection it is proper and fitting t h a t this should be so. W e cannot ask a li- brarian to defer commitment and to state the theoretical grounds for his judgment every time he contemplates making a pur- chase. Nevertheless, unless he has some realization of what these grounds are, in the long run his policy and practice are apt to be misdirected, confused, and waste- ful. I t is unfortunately the case that value standards may conflict, that a book which has great value if measured by one criterion may be valueless if measured by another. For a college or university li- brary this possibility of conflicting stand- ards presents a serious problem. In such libraries the responsibility for book selec- tion is divided between the librarian and the diverse groups which make up the faculty. Unless there is reciprocal under- standing of the criteria to be employed, results may be disastrous. Moreover, uni- versity libraries are composed of many dif- ferent collections serving many different interests. T h e proper service of one or another of these interests may require an exclusive regard for one or another of the criteria. I t is possible to distinguish five rela- tively independent criteria of value that determine policies of book selection, namely, the additive, reference, critical, documentary, and monetary. Since each criterion defines a scale of values, it fol- lows that books have additive, reference, critical, documentary, and monetary value. Any book may have a position on all five JUNE, 1941 221 scales, but its position on any one scale is not determined by its position on any other or by any combination of the others. In mathematical terms, the criteria are inde- pendent variables. W e may discover em- pirical correlations between the different values, that is to say, certain books may have high values on each scale or books high on one scale may be low on another; but we cannot deduce the position on one scale from the position on another. In a normative discussion it might be held that some of these criteria ought not to be independent; specifically it might be held that the monetary value should be a func- tion of critical value or that additive value should be a function of documentary value, etc. But since each of these criteria operates independently in the actual prac- tice of book selection, a description of the practice should consider them as inde- pendent variables. Additive Value T h e term "additive" does not repre- sent an altogether happy choice. W e b - ster gives the following definition: "proper to be added . . . capable of being joined so as to cause quantitative increase; in- volving or characterized by addition." Books, being physical objects, can be added, and libraries publish annual state- ments of their "quantitative increase." A book has additive value when its addition to a library increases the size of that li- brary. T h e relevance of additive value to the practice of book selection is best illustrated by the fact that libraries are ranked by the sizes of their collections. O t h e r cri- teria may be employed, but I know of no attempt to rule out additive value as a standard for estimating the value of li- brary collections. T h e r e is a sense in which the additive value of all books is equivalent. Adding a book increases the size of the collection, regardless of the nature of the book. But a book may have different additive values for different libraries. I n the first place, even though duplicate copies are usually counted in determining the size of a li- brary, the additive value is usually esti- mated in terms of new titles or books not previously owned by the library. Hence a book will have additive value for one library and not for another. Secondly, the addition of the same num- ber of titles by several libraries may repre- sent different proportions of increase. Large libraries, in order to maintain numerical superiority and to insure an- nual increases equal in proportion to the increases of smaller libraries, are forced to subordinate all other criteria to the additive standard. T h e r e are some li- brarians whose chief concern in book selec- tion seems to be to discover titles not al- ready in their libraries. N o censure is intended in this statement. Aside from consideration of size, li- braries have another pseudo-quantitative goal for which additive value functions as the criterion of achievement: the at- tempts to attain complete collections in special subjects. Before a book can have additive value for special collections, it must have a certain imprint, treat a par- ticular topic, be written at a particular time, or be published in a particular place. Once these prior discriminations are made, collectors concern themselves only with additive value. T h i s can be seen in the actual statements of objectives issued by libraries with certain special collections. One collects "all available material relat- ing to M a i n e ; " another "aims to collect everything printed in America before 222 COLLEGE AND RESEARCH LIBRARIES 1820;" a third "aims to assemble every- thing relating to the literature and his- tory, particularly local history, of Italy since 1870;" a fourth "has attempted to secure everything published in the United States on education;" a fifth seeks "all available material relating to the history of the South." Such examples could be listed indefinitely. T h e fact is that almost every important library in the country is collecting " a l l " of some specified type of material; and since the specification of type includes no reference to any criterion of value, the unqualified " a l l " indicates the ubiquity of the additive standard of value. Every book has some conceivable ref- erence value, but the independent char- acter of this criterion is best illustrated by considering the types of books which function almost exclusively as reference materials, namely, telephone books, di- rectories, W h o ' s Who's, dictionaries, bib- liographies, periodical indexes, etc. Such books are ( 1 ) intended to be consulted for specific information rather than to be read, and ( 2 ) are guides to the use of other materials. T h e degree to which they serve these purposes determines their reference value and their selection by li- brarians. Critical Value T h e obvious character of reference value makes any lengthy discussion of its relevance unnecessary; the same would be true of critical value except that here, as in the discussion of additive value, the lack of an adequate terminology seems to necessitate the use of an ambiguous term.1 By "critical value" is meant that type of 1 I t is one of the aims of this p a p e r to c o n t r i b u t e to t h e e s t a b l i s h m e n t of an a d e q u a t e t e r m i n o l o g y . T h i s makes it n e c e s s a r y to point o u t distinctions a n d to n a m e what is t h u s d i s t i n g u i s h e d . N o brief is held f o r the choice of n a m e s . value usually intended in judgments that critics, reviewers, or librarians make con- cerning the average book. I n fact, lay- men generally suppose, and many li- brarians share the supposition, that critical value is the only criterion genuinely ger- mane to the concerns of a library. But it is only in relatively small libraries that considerations of critical value are para- mount. In spite of widespread opinions to the contrary, the determination of critical value is not a difficult task. T h e dis- agreements of critics concerning the values of books are accorded much more pub- licity than their importance warrants. In general there is widespread agreement concerning the values of books among competent authorities and it is only in regard to borderline cases that disagree- ment is apt to occur. T h e r e may also be a difference of opinion concerning relative rank. Selective Lists T h e general agreement concerning criti- cal value has made possible the publica- tion of several lists to which a librarian may refer in selecting books. Such com- pilations as the Shaw list, the A.L.A. Catalog, the Sonnenschein list, etc., are not presented as products of subjective choice but as the results of a critical consensus. Hence the librarian who is reasonably well informed will have little difficulty utilizing the criterion of critical value. It was observed above that the ques- tions of critical value seem to be of im- portance primarily in small libraries. T h i s statement must now be qualified. Books of high critical value are of major importance to all libraries, but their small number relative to the existing volume JUNE, 1941 223 of publications makes it possible for large libraries to possess or buy all of them and the librarian is not confronted with the necessity of choice. N o college or uni- versity library of average size is forced to choose between M i l t o n or Shakespeare, Tolstoi or Dostoevsky, Dickens or Scott, Frost or Jeffries. And every first class college or university library has an annual budget appropriation sufficient to purchase the new books of outstanding critical value in fields represented in its curriculum. M a n y will object to this observation. I t is sufficient to remark that librarians will sometimes spend for a rare imprint or first edition enough money to buy the complete works of a dozen m a j o r English poets. Documentary Value T h e exhibits of a great art museum or the repertoire of a great concert orchestra reflect the taste and judgment of genera- tions of critics, but the shelves of a great research library do not reflect a parallel discrimination. Some of our research li- braries seem to collect the bad book, the cheap novel, the pompous genealogy, the insipid poem, the lying history, the dull report, the stupid diary, the ephemeral tract, etc., just as they collect the works of established critical value. It must be that in flaunting the decisions of the critics and the accumulated judgment of thou- sands of readers, the research library is appealing to some other standard of val- ues. T h i s is indeed the case; the research library serves the scholar who may be primarily concerned with documentary value.2 " I n E u r o p e t h e r e is a g e n e r a l a p p r e c i a t i o n of the distinction b e t w e e n books of critical a n d c u l t u r a l v a l u e a n d d o c u m e n t s "collected f o r p u r p o s e s of evidence, verification, or s t u d y , " a l t h o u g h t h e r e is no a g r e e m e n t as to w h e t h e r t h e collection of docu- m e n t a r y m a t e r i a l is a f u n c t i o n of l i b r a r i e s or a task Any type of printed matter or manu- script has documentary value if it can conceivably be used by the literary, politi- cal, or social historian. T h e historian uses these materials as the anthropologist or archeologist uses artifacts, namely, to gain an understanding of the past that shall be as complete as possible. Consid- ered as historical evidence, the trashiest novel may be as significant as a literary masterpiece. And this fact has given rise to the paradox of book selection. T h e books of critical value of any historical era will, in general, be reasonable in price and common; but the dime novel or the penny broadside will be expensive and rare. Hence, a librarian of a research institution may be asked to devote an un- due proportion of his funds and energy to the purchase of material with docu- mentary value but of no critical value. T h e question must arise as to whether there is any limitation which applies to the class of materials. which have docu- mentary value. It is generally assumed that everything has such value and if this is so, it follows that additive and docu- mentary value tend to coincide. It may be objected that the t w o criteria are distinguished by the fact that the addi- tive value of all books is identical, whereas the documentary values of books may vary. Unfortunately, current practice gives us no satisfactory indication of how relative documentary value is to be determined. T h e documentary value of any type of material may rise or fall with the chang- ing fashions and problems of scholarship and may vary in different institutions and f o r s e p a r a t e i n s t i t u t i o n s . I n A m e r i c a t h e t e r m " d o c u m e n t " is u s u a l l y r e s t r i c t e d to publications of g o v e r n m e n t agencies. A s used h e r e i n , this restric- tion does not apply. F o r a discussion of d o c u m e n t a t i o n , see " L i b r a r i e s a n d D o c u m e n t a t i o n " by Marcel Godet in t h e Library Quarterly 9:185-92, A p r . 1939. 224 COLLEGE AND RESEARCH LIBRARIES regions depending upon the interests of local scholars. Most librarians feel quite justly that they cannot build balanced col- lections upon the shifting sands of cur- rent interest. But lacking any criterion that can measure the significance of schol- ars' interests or any foreknowledge of what may interest the next generation of scholars, libraries with sufficient funds tend to identify documentary and additive value. Libraries that cannot afford this identification must play a guessing game and bury their bad guesses in the stacks. It is a fair judgment that about half of the books in the Library of Congress will never be used from now to the end of time, but no scholar or librarian, or even a committee composed of many scholars and librarians, would venture to decide "which half." Monetary Value If it were possible to consider all the reasons and considerations which deter- mine the book selections of libraries it would not be necessary to treat monetary value as an independent variable. Mone- tary value would become a function of all other values and would be determined by the critical value, the scholarly value, etc. However, the considerations which deter- mine book selection are unlimited. T h e four criteria already discussed are the most significant but many more could be pre- sented. Libraries value books because they were published before a certain date, be- cause they were written by local authors, because they are bound in certain ways and are printed with a certain type font, be- cause they were once possessed by famous men, because they are numbered copies, because they are autographed, because they are first editions, because they are fake first editions, because they are scarce, etc. It should be apparent that it is hopeless to attempt to discuss all the minor criteria of value used by libraries. Rather, I have chosen to regard monetary value as their common denominator. All of them taken together determine the price of books (other than trade books) to a greater ex- tent than do any of the criteria that have been discussed. Hence, so f a r as these four criteria are concerned, monetary value is an independent variable. If any- one wishes to attempt an exhaustive ac- count of the reasons why books are con- sidered valuable, I wish him joy in the venture and promise to deny the inde- pendence of monetary value as soon as he announces his success.3 ( T o be continued) 3 I t is acceptable scientific practice to t r e a t a v a r i a b l e as i n d e p e n d e n t so long as it is impossible to specify the conditions upon which it d e p e n d s , even if t h e r e is r e a s o n to believe in t h e existence of such c o n d i t i o n s . JUNE, 1941 225