lib-MOCS-KMC364-20140106084141 221 BOOK REVIEWS Introduction to Information Science, Tefko Saracevic, ed. New York: Bowker 1970, 776 pp. $25.00 The editor has put together a large volume consisting of 776, 8~ x 11 pages and weighing almost 5 pounds. It comprises 66 different articles written by almost as many authors and covers the period from 1953 to 1970. Two-thirds of the articles were written during the period 1966-1969. In short, it is a collection of a large number of papers mostly from the last few years having to do in some way with information science or more properly, with information systems. The papers generally are good ones and in some cases have already become acknowledged classics. In a few cases I am a bit puzzled about their inclusion in a volume of this type. In the few months since I have had this book I have already found numerous occasions to consult several of the articles. Some of the other papers which I have not seen recently I have enjoyed reading again. The book is divided into four parts, which are further subdivided into thirteen chapters. The four parts are Basic Phenomena, Information Systems, Evaluation of Information Systems, and A Unifying Theory. Although the chapter headings are too numerous to list they include such topics as Notions of Information, Communication Processes, Behavior of Information Users, Concept of Relevance, Testing, as well as Economics and Growth. By virtue of the Parts, Chapters, and articles the editor has provided a type of classificalion system or structure for information science without attempting to define information science. Interspersed between each of the Parts and Chapters is up to a page of introductory and explanatory material provided by the editor. In a volume of this type it is important to recognize what the volume is and what it is not. As I have mentioned, it is a good anthology of important articles related to information. It is not, as the title implies, an introduction to information science. The papers are by and large unrelated to each other and the introductory comments by the editor do little to provide a unifying relationship. Furthermore, the overall scope of the articles is generally quite limited and, although the editor implies it is not so, tends to equate information science to information systems. The final paper in the volume by Professor William Goffmann is listed by the editor as Part Four-A Unifying Theory. The precise title of the chapter is somewhat less ambitious: namely, "A General Theory of Communication." The paper is an unpublished one (although similar papers by the author have been published elsewhere) and relates communication in a general sense to the theory of epidemic processes. Although the theory is an 222 Journal of Library Automation Vol. 4/4 December, 1971 interesting one, it would hardly qualify as a unifying theory for informa- tion science. It certainly does not provide the unifying relationships among the various articles included in the text. My guess would be that .other qualified individuals, in putting together a similar volume, would have included many different articles. This, how- ever, is the nature of the field at this time. By comparison note the recently published volume Key Papers in Inf01·mation Science, Arthur W. Elias editor. This book, although admittedly serving a somewhat different pur- pose, contains 19 papers with only a single paper in common with those of this particular volume. In summary, this is a good collection of relevant and useful articles in information science. It is probably desirable that they be included in a single volume. Serious students, educators, and research workers will find this volume to be of interest. As a reference book it will be quite useful. The book is not, however, an introduction to information science. The novice, the student, and the casual reader will probably be disappointed and confused, and in some cases might even be misled. Marshall C. Yovits Information Processing Letters. North-Holland Publishing Company, Am- sterdam. Vol. 1, No. 1, 1971. English. Bi-monthly. $25.00. This journal is published by a most reputable company and has a most impressive international list of editors and references. The affiliation of editors illuminates the orientation of the journal: six of them are from departments of mathematics, computer science or cybernetics and two are from IBM laboratories. Understandably, the journal is devoted basically to computer theory, software and applications, with a heavy accent on mathematically expressed theory related to the solution of computing problems, algorithms, etc. It is directed toward basic and applied scientists and not toward practitioners. People interested in library automation may, from time to time, find in it theoretical articles broadly related to their work, but they will have to do the "translating" themselves. This journal follows the tradition of "letters" journals in physics, biology and some other disciplines. The papers are short; publication is rapid; work reported generally tends to be very specific, preliminary to or a part of some larger research project; usually small items of knowledge are reported. The "letters" journals are received in the fields where they appear with mixed emotions. For instance, Ziman (Nature 224:318-324, 1969) questions very much the need for these publications. On the other hand, they are a useful outlet for authors who otherwise would not publish these often useful bits of specific knowledge. Recommended for research libraries related to computer science. T efko Saracevic Book Reviews 223 Handbook of Data Processing for Libraries. By Robert M. Hayes and Joseph Becker. New York: Becker & Hayes, Inc., 1970. 885 pages. $19.95. To write a universal handbook in a field so full of complex intellectual pro.blems and simultaneously satisfy every potential reader is an impossible assignment. Therefore the authors cannot be faulted for failing to satisfy everyone. They have succeeded in writing for a very important audience -administrators and decision makers. For this group, they have presented difficult technical material in a clear readable fashion-a reflection of ' their extensive teaching experience. For many library administrators, this handbook arrives five years too late. Had it been available earlier, a large number of current automation projects might never have been authorized by management, or at least might have been conducted on a sounder basis. Following a very con- servative approach, the authors generally remain within the limitations of the current state of the art, being careful to distinguish that which is feasible (i.e., practical ) from that which is possible. Over and over again, they warn librarians about the limitations of computers and caution against excessively high expectations. For administrators, the most useful material is in chapter 3, "Scientific Management of Libraries," and in chapter 8, "System Implementation." A reading of chapter 8 alone suffices to convey to the administrator the magnitude and complexity of even the most seemingly routine computer application in libraries. This chapter, the most important and useful in the entire book, covers planning, organization, staffing, hardware, site preparation, programming, data conversion, phase-over, staff orientation, and training. Each of these topics-deserving of complete chapters in themselves-is treated briefly, but in enough detail to communicate the complexity of each component in the long stream of system development activities, all of which must be completed to the last detail for success. There are three useful appendices: a glossary, an inventory of machine readable data bases, and a list of 115 sources for keeping up to date. Bibliographic footnotes abound and each chapter ends with a list of suggested readings. However, it is surprising how many references are five or more years old; in fact, there is a scarcity of current references. For example, Ballou's well-known Guide to Microreproduction Equipment, now entering its 5th edition, is cited in the first edition of 1959. The authors have been badly served by their proofreaders. The book is marred by an incredible number of spelling errors in text, tables, foot- notes and references, especially with personal names, plus incomplete citations. The index contains many entries too broad to be useful, such as: utilization of computer ( 1 entry ), time sharing ( 1 entry), hardware ( 3 entries), technical services ( 3 entries). Lacking from the index are name references to distinguished contributors to the literature, such as Avram, Cuadra, DeGennaro, Fasana, and others. Many of these names appear only in footnotes. 224 Journal of Library Automation Vol. 4/4 December, 1971 The book is rich in tabulated data and specifications for a variety of equipment. Unfortunately, much of this equipment is inapplicable to library use, or the tabulated data is in error. Table 12.25lists several defunct or never marketed equipments, such as IBM's Walnut and Eastman Kodak's Minicard, without indication of non-availability. In table 11.22 there are extensive listings of CRT terminals, most of which are unsuitable for library applications by reason of deficient character sets or excessive rentals. Nine of the units listed showed rentals of over $1,000 per month, and two of these were virtually at $5,000 per month, clearly beyond the reach of any library. Table 12.2 suggests the access time to one of 10,000 pages in microfiche is half a second, a figure that is off by an order of magnitude for mechanical equipment and by two orders of magnitude for manual systems. (More nearly correct figures are given in the text on page 396). Table 12.21 lists several microfilm cameras designed expressly for non-library applications and not adaptable to any library purpose. From a broader perspective, one misses several other features. Is a "handbook" for the practitioner? If so, this volume is too elementary. Can it be used as a textbook in a course in library automation and information science? The book contains no problems for students to attack, and except for references, no aids to the instructor. Possibly it can serve as supple- mentary reading, for it contains far too much tutorial material (yet only ten pages of nearly 900 are devoted to flow charting). One wishes for more specifics drawn from the real world. A hypo- thetical case study in chapter 11 is illustrative: a 5% error rate is assumed for input of a 300,000 record bibliographic data base to be converted to machine readable form. Not revealed in the example is that a relatively low error rate in keyboarding may result in a very high percentage of records which must be reprocessed to achieve a high quality data base. Each reprocessed record will consume computer resources: CPU time, core, disc I/0, tape reading and writing, etc. We know from MARC and RECON that the ratio of the total records processed to net yield is on the order of 3:2; i.e., each record must be processed on the average of one and a half times to get a "clean" record. The cost of this reprocessing is far beyond the 5% lost by faulty keyboarding. The Handbook will be a useful decision making tool for the generalist, a less helpful aid to the practitioner. It is hoped that a revised edition is in preparation, and particularly that the tabular material will be corrected and brought up to date. Chapter 8, the heart of the book, should be greatly expanded. For the next edition, some consideration might be given to a two-volume work: the first volume for the administrator, and the second containing much more technical detail for the practitioner. If the two volume pattern is followed, a loose-leaf format with regular updating would be most helpful for the second half. Allen B. V eaner Book Reviews 225 L~brary Automation: Experience, Methodology, and Technology of the Lzbrary as an Information System, by Edward W. Heiliger. New York: McGraw-Hill Book Co., 1971. xii, 333 pp. The need for a handbook and/or general introductory text on the topics of automation and systems analysis in libraries has been sorely felt for quite som.e time. During the past year, three have appeared (Chapman and St. Pierre, Library Systems Analysis Guidelines, Wiley, 1970; Hayes and Becker, Handbook of Data Processing for Libraries, Wiley, 1970 and the book here reviewed.) Unfortunately, none is completely satisfactory, for different reasons. A serious student wanting a reasonably comprehen- sive, systematic, and balanced treatment of these subjects will, I'm afraid, be forced to have to use all three of these titles and, even then, will have constant need to use supplementary materials for a number of aspects. The title being considered in this review by Heiliger and Henderson, if one judged only by the authors' intent as expressed in the Preface, would be exactly the kind of work that we've all felt the need for. As they state on page vii, the purpose "is to provide a perspective of the library func- tions that have been or might be mechanized or automated, an outline of the methodology of the systems approach, an overview of the technology available to the library, and a projection of the prospects for library auto- mation." And, indeed, if one looks at the table of contents there are four parts that closely parallel this statement of purpose. The parts themselves though, when inspected more closely, reveal not a systematic treatise or even an in-depth treatment of these topics, but rather a loosely connected series of essays, each on a fairly superficial level, discoursing on a variety of aspects associated with, or tangental to these topics. This indicates, at least to this reviewer, that the genesis of the book was a series of lectures presented and refined over a period of time by the authors. Although not in itself a bad thing, here it is unfortunate to some degree because not enough effort was expended in amplifying the material with additional data, library-oriented examples, and illustrations, nor in logically integrat- ing the various parts. Part I, entitled "Experience in Library Automation," begins by broadly citing a number of library automation projects mostly dating from the early 60's. The level is extremely superficial and the presentation not very enlightening, since only three or four projects are mentioned, and then only in passing. Immediately following are several excellent chapters describ- ing traditional library activities (e.g., acquisition, cataloging, reference,. etc. ) in functional terms. The approach, though extremely simple, is for the most part effective and is only marred by occasional, overly condescend- ing statements such as "Library filing is a very complicated matter" or "Reference librarians use serials literature extensively." Unfortunately, in the 104 pages of this section there is not one illustration. 226 Journal of Library Automation Vol. 4/4 December, 1971 Part II, "Methodology of Library Automation," attempts to describe the general approach and techniques of systems analysis. In a number of ways, this is the best part of the book. Unfortunately, the concepts that are so simply and succinctly described are only indifferently related to activities that will be familiar to librarians. As a brief essay on the objectives and concepts of systems analysis, it is quite adequate, but as a discussion of how they relate to library problems, it is totally inadequate and often misleading. Part III, "Technology for Library Automation," is probably the least informative part of the book, giving the reader virtually no practical in- formation. All of the important and obvious technological concepts are listed, but are dismissed with what oftentimes is little more than a brief definition. The one exception to this is Chapter 13, entitled for no apparent reason, "Concepts." This chapter is in fact an innovative and thought- provoking view of a library as a data-handling system. One wishes that this chapter had been amplified and treated more fully. Part IV, "Prospects for Library Automation," is the least effective part of the book, having in my mind only one merit: it doesn't tack on a Holly- wood-style happy ending. The authors' view of the 70's, as far as can be inferred from this too short section, is cautious and mundane. These will be, I'm convinced, the overriding characteristics of automation efforts for the next several years. I only wish that the authors had elaborated more fully on these points and presented their views more coherently. The book is augmented with a 61-page bibliography ( 1,029 citations), which, though reasonably current, is of dubious worth because it is neither annotated nor particularly well balanced. Certain classics, such as Bourne's Methods of Information Handling, or Information Storage and Retrieval by Becker and Hayes, and certain current, basic items, such as Cuadra's An- nual Review of Infonnation Science and Technology and the Journal of Library Automation, are not listed. Each chapter is accompanied by a "Suggested Reading List" wherein materials more or less pertinent to the subject of the chapter are listed. A glossary of terms in three parts (a total of 36 pages) is also included and, though difficult to use because it is in three alphabets and interspersed with the text, provides short but very adequate definitions. Unfortunately, several jargon terms used in the text itself are not included; one that was most irritating to this reviewer is the term "gigabyte" which to my knowl- edge has very little currency among the cognoscenti. On balance, Library Automation is a title that should be recommended for a wide range of readers. Though it will probably have little to offer experts in the field, it does have value as a text for library students or a general introduction for the average, non-technical librarian. Paul ]. Fasana Book Reviews 227 Sistema Colombiano de Informacion Cientifica y T echnica (SICOLDIC). A Colombian Network for Scientific Information, by Joseph Becker et al. Quirama, Colombia: May-June 1970. 59 p. Mimeo. The task of the study team which produced this report was to present "an implementation plan for strengthening the scientific communication process in Colombia by providing a permanent systematic mechanism to function in the context of Colombia's internal needs for scientific and tech- nical information in government, industry, and among the research activi- ties in higher education." More specifically, the expressed goal of such a mechanism is "to develop a network which will permit any scientific or technical researcher, in gov- ernment, industry, or university, to access the total information resources of the country without regard for his own physical location." The study was comple ted in two months (according to the cover dates) and comprised four areas of investigation, namely: 1) to elucidate the advantages of d eveloping a centrally administered national network in- cluding three levels of network nodes and a technical communications plan; 2) make an inventory of universities, institutes, telecommunications and computer facilities in Colombia; 3) recommend a mix of these factors to produce specific services, and 4) propose a seven-year budget. The Republic of Colombia is about the size of Texas and California combined, and its population is about 1 million less than New York State. Most scientific and technical workers are located in five major cities, and the country is divided into six administrative zones. Within these zones twenty universities and forty-four institutes were inventoried by the study team with respect to specialization, faculty, book collections and the like. From these universities and institutes, five primary and seven secondary nodes were named to be connected by means of a Telex communications system. The Telex connections are not to be computer-mediated in the forseeable future, but used for interlibrary loan and other messages. ( There were two teleprocessing systems operating in Colombia at the time of the study.) Basic recommendations are: that a governmental unit be established with responsibility for directing the development of SICOLDIC; that this unit, with a high echelon board of directors, should produce several directories, bibliographies and union lists, and publish a monthly catalog of govern- ment-sponsored scientific and technical research. In addition, a manual for use of the telecommunications system should be produced. The pro- posed budget is about $250,000. ( 4.5 million pesos ) for the first year, graduating to a 25-fold increase by 1976. In some aspects the SICOLDIC plan follows the pattern of some state library development plans being implemented in the U.S. The advantage of central control of information resources planning and fund control b y the SICOLDIC group, with fairly direct access to high governmental 228 Journal of Library Automation Vol. 4/4 December, 1971 authority, provides reasonable insurance for support of the plan, especially since these services contribute to the economic and scientific advance of Colombia. There is no indication of the acceptance of the plan by COLCIENCAS, the governmental unit which commissioned it. Of the sixty references in the bibliography, Spanish publications pre- dominate. Ronald Miller Cooperation Between Types of Libraries 1940-1968: An Annotated Biblio- graphy, by Ralph H. Stenstrom. Chicago: American Library Association, 1970. This bibliography is an effort to sift, organize and describe the literature of library cooperation produced during the period 1940-1968. Two criteria governed the selection of the 348 books and monographs listed: 1) they must deal with cooperative programs involving more than one type of library, and 2) they must describe programs in actual operation or likely to be implemented. Although most of the cooperative projects described are located in the United States, other countries are represented when the material about them is written in English. Cooperative programs in the audio-visual field are included. The annotations explain the nature of the cooperative projects and give the names of participating libraries. An appendix describes briefly about 35 recent cooperative ventures not yet reported in the literature, which the editor learned about through an appeal published in professional journals. Entries are arranged chronologically to facilitate direct access to the most recent developments and to permit tracing the evolution of a par- ticular project over a period of time. Three indexes provide approaches to the material by 1) name of author, cooperative project or library organization, 2) type of C