College and Research Libraries Electronic Information and Research Library Technical Services Peter S. Graham The relation of libraries to the electronic information explosion has been a focus of discussion for several years, but the impact of this explosion on the technical seroices function within libraries has not been adequately explored. In what follows, my contention is that technical seroices are not solely dependent on decisions their libraries make regarding electronic information. They should be a driving agent as well. I. ISSUES AND CONTEXTS Some of the new contexts in which our research libraries technicaL services de- partments operate include the following: 1. First, the electronic library is coming into being within the context of the uni- versity: information in electronic form is increasing in importance. For a full devel- opment of this context, see the excellent arti- cle by David Lewis, "Inventing the Elec- tronic University."1 Larger and larger quantities of informa- tion are appearing electronically, al- though they are still small in proportion to the quantities we expect in a few years. Online databases are now numbered in the ~housands with hundreds more ap- pearmg each year. Local academic files of electronic information are being created at a rate that no one can guess; and CD- ROMs versions of over two hundred data- bases have been published. 2 Most CD- R OMs are indexes such as Current Contents, InfoTrac, PsycLit, and Medline, but textual and statistical data are also be- ing published in this form; for example census data, the Oxford English Diction- ary, the Medieval and Early Modern· Data Bank, and the Thesaurus Linguae Graecae. Then there is e-mail: Bitnet, Arpanet, Internet, NSFnet, NYSERnet, ALAnet, AppleLink, Compuserve and The Source are all in heavy use among scholars and many other groups. It is hard to know how much data goes over these lines in addition to correspondence. We do know that Kermit, the communications program developed at Columbia University, is enormously popular among academics, and one of its virtues is its file-transfer ca- pability.3 There is greatly increased use of elec- tronic information at some of our universi- ties. This is especially true at campuses where the computing services have facili- tated its availability by providing com- munication networks, e.g., Michigan, Stanford, Carnegie-Mellon. 2. There is broad OPAC availability. There is, however, wide variation in how these are made available to patrons: in- building only, library star networks, dial- up access, limited LAN availability, and campus-wide network access. Wide cam- pus OP AC availability is generally viewed as a desideratum. 3. The automation of technical services has largely been accomplished and related is- sues are fully understood. There are few Peter S. Graham is Associate Vice President for Information Seruices and Associate University Librarian for Technical and Automated Services, Rutgers University Libraries, New Brunswick, New Jersey 08903. 241 242 College & Research Libraries libraries in the country that have not been automated to some extent. Meanwhile, it is also evident that the present level of au- tomation cannot keep up with traditional tasks. One example is the British Library's recent announcement that it would cut by one half the fullness of much of its catalog- ing in order to process the remainder more promptly. ''Our traditional staffing definitions are more and more open to ques- tion.'' 4. Our traditional staffing definitions are more and more open to question. There is difficulty now in finding technical services staff, especially catalogers, who have the bibliographic skills, the judgment, and the managerial skills we need. We also need staff who have skills and experience with electronic technology or have the self- starting ability to obtain them quickly. When we do find good staff they are of- ten very good indeed; but the supply is limited. 5. Friendly interfaces: The concept of the integrated information system is fading. Historically, our goal in technical services has been to provide the user with a single place to search and a single tool. But just as we began to want an integrated mono- graphic acquisitions, cataloging and circu- lation system, the idea of linking abstract- and-index databases to serials holdings raised our expectations. We now know that we will never have all necessary infor- mation accessible through a single file. We need to turn our attention to user-friendly interfaces. 6. MRDFs, or computer files: Most re- search libraries are cataloging machine- readable data files. Some have only cata- loged a few. Some libraries are providing bibliographic access to large quantities of commercial and private databases, both for those owned by the library and for those not owned, with very considerable implications for change in what libraries are and how we do things. 7. Electronic publishing: There isn't May1990 much yet but it will grow. Libraries will presumably have to provide access. 8. On-site commercial databases: Many campuses are now mounting major com- mercial databases locally to provide on- site searching. The universities include Carnegie-Mellon, University of Pennsyl- vania, Columbia, Georgia Tech, Stanford, and Rutgers. These databases are often linked to other types of library informa- tion. Access to these databases can be through the OP AC and other library offer- ings. 9. Local information files: Job files, fellow- ship announcements, campus even.ts, news, advice sources, research in prog- ress, housing files, community informa- tion, vertical file-like files are increasingly being made electronically available on campus, often through the library system or one adjacent to it. These may be linked to the OPAC. 10. Full-text in electronic form is becom- ing more common. Special libraries have long sought to provide full-text informa- tion. It is technically possible to have library information systems that provide full elec- tronic texts. Facsimile transmission is an early manifestation of this possibility; so is the development of broadband image transmission from media centers. The li- braries at Rutgers and Princeton have been funded by the National Endowment for the Humanities, (NEH) to plan a Cen- ter for Machine-Readable Texts in the Hu- manities. Various forms of optical disk (CD-Rom, WORM) and high-density magnetic storage, and high-bandwidth transmission systems will increase the technological attractiveness of full-text provision. 11. Information immediacy is increasing: Libraries must provide more timely access to information. The timeliness of access can be arrayed on a continuum from the immediate to the archival. (On the one hand we have provided quick access to re- cent data, and on the other we have pro- vided slower access to historical knowl- edge.) Typically, our bibliographic access tools have been effective in inverse pro- portion to the timeliness requirement of the information. Electronic information places a premium on immediacy of access to information, but our techniques for cre- ating bibliographic tools are best equipped for the archival. This must change if li- braries are to be properly effective. II. TECHNICAL SERVICES DEFINITIONS Technical services of libraries are pri- marily responsible for the acquisition of materials and their organization. For prag- matic reasons both circulation systems and preservation activities have some- times been included in the technical ser- vices responsibility. But in general the purview of technical services has been rec- ognized as the buying and receiving of se- lected materials, and their organization for use: acquisitions, gifts and exchange, serials control, cataloging: from selection to shelf, mark it and park it. Up until very recently, technology has not changed that mission. · There are three distinctive but implicit assumptions in the traditional technical services role that can no longer be taken for granted. Prediction of use: The traditional role calls for information provision predictive of use. The great bulk of the information pro- vided by libraries has been acquired in ad- vance of expected need on behalf of users who did not yet know of their need. Pre- diction of use has been both the pride and the bane of collection development policies-one need only mention the Pitts- burgh study-and it is the driving force behind cooperative programs at all levels. The best interests of our users require us to predict their needs as best we can; how- ever, ~ovision of information on demand can be expected to rise with the availability of electronic information, requiring much more responsive acquisition and delivery systems. Ownership: The traditional technical ser- vices role assumes library ownership of in- formation. A library acquires information, then organizes it, makes it available, and preserves it. However, we increasingly recognize that there is information not owned by the library that we must make available to our users. We have done this in the past for serials we haven't owned, through our printed journal indexes; we Electronic Information 243 are increasingly making this happen through what we call our catalogs. This in- cludes catalog information about the full range of GPO documents, the databases in the computer center, or the holdings of the Center for Research Libraries. Hard copy: Our traditional technical ser- vice role has assumed hard copy. We have traditionally bought and cataloged books, journals, films, tapes, records, floppy disks and CD-ROMs. We now must con- front library and user needs for access to machine-readable files that have ex- tremely itinerant lives. At Rutgers, for ex- ample, is being cataloged in the library in- structional software owned by the computer center. Kinko's and IBM issue these programs on individual floppy disks; but in the Rutgers pilot program, the computer center is maintaining them all on hard disks for examination, and is is- suing loan copies to faculty who wish to evaluate them. Immediately thrown out was any idea of added copy cataloging; but that was the easy decision. What hap- pens with software upgrades? What hap- pens now that Kinko is no longer cooper- ating with Apple? Where is this stuff going to be located next year, or next week? And with other materials-which copy of the ICPSR survey research data should we be cataloging? What kind of ac- cess record do we provide to the Medieval Data Bank information? Mark it and park it? Mark what, and park it where? III. NEW TASKS FOR TECHNICAL SERVICES New contexts and the changing of tradi- tional assumptions imply a wide range of new understandings and activities for technical services. Most of us recognize this, and discussion and change has been widespread in research libraries. Informa- tion interfacing and intellectual preserva- tion are two of the most interesting issues being discussed in research libraries. 1. Linking access systems. The goal of the integrated bibliographic system, that seamless database allowing access to all resources is fading. We cannot provide II one-stop shopping'' (and it should be re- membered that the term's usage is from 244 College & Research Libraries stores that wish to persuade us that what they happen to have is all we need). It is increasingly necessary to envision linked information systems rather than inte- grated ones. 4 Consider the multiplicity of information sources we must provide to patrons: the OP AC itself, acquisitions files, serials con- trol systems; government document files, technical reports, vertical files, media cen- ters; remote search services, OCLC, spe- cial databases in RLIN, local databases; CD-ROMs, ICPSR files at computer cen- ters, and campus information systems. Research libraries are the institutions to provide patron access to all this informa- tion, and we can never seamlessly present it all. We will have to determine the principles by which we intelligently present this ar- ray of information to users, and then we must find ways to implement them. To do this we will require librarians equipped with both technological expertise and bib- liographic understanding. These librari- ans will continue to embody the service goals our profession has always held high. They will find challenge in knotty tech- nical problems (the kind that used to be found in formulating and implementing cataloging codes and the MARC record), and they will find satisfaction through the resolution of these problems in favor of patron service. For these librarians, the technical problems will simply have dif- ferent characteristics. In this future will be found our response to the reported dearth of librarians interested in cataloging and technical services. 5 An example of a knotty problem for linked information systems is authority control. One of the most important intel- lectual contributions librarians have made to information access is the concept of the controlled thesaurus, or authorized vo- cabulary. Providing syndetic structures and authority files for great catalogs of millions of entries require enormous ef- fort, and we are just getting the issues un- der control in our automated cataloging systems and OPACs. Consider now the problems facing a pa- tron who is moving his or her search from May 1990 the government documents catalog to the OPAC, and perhaps to the ERIC database or to Chemical Abstracts. The search movement should be immediate and the interfaces should assist rather than assault the user. But is it also possible to address the shifts in controlled vocabularies as the patron moves from one database to an- other? The challenge for serious and tech- nically minded librarians is great. 2. Intellectual Preservation. The second new task for technical services is intellec- tual preservation. Intellectual preserva- tion is different from preservation of a physical artifact. Intellectual preservation addresses the transience, the evanes- cence, the fundamental ephemerality of electronic information. The Bodleian Li- brary at Oxford maintains the John John- son Collection of Printed Ephemera, but this is for materials explicitly intended for short lives. All our research libraries are about to include collections of electronic ephemera: material intended to live, but in a form conducive to destruction. Ed Brownrigg, Gordon Neavill, and others have commented on the intellec- tual risks involved in committing our dis- course to the fragile yet fertile medium of electronic recording. 6' 7 It is obvious to all of us from personal experience that elec- tronic information can be easily lost. The proliferation of versions of texts is another problem. Even in print form, it has long troubled scholars, bibliogra- phers, and librarians. Ithiel de Sola Pool has commented on the new possibilities in an electronic age, as easily reproducible texts and versions of texts are transmitted back and forth across academe. 8 The con- cept of a canonical text-already in trouble in printed forms-could cease to exist alto- gether. At the personal level, many of us have already had the experience of distrib- uting memos in draft form over e-mail to various colleagues for response, with vari- ations in the drafts following hard upon one another as the comments come in. De Sola Pool is concerned about this is- sue in terms of redactions of literary forms, both for texts of the past (consider for example the current scandal over the text of Joyce's Ulysses)9 and also for those Electronic Information 245 Electronic Information-Malleability and Responsibility The malleability of information that is one of the major advantages of compute_r- based electronic systems has as its corollary the potential transience of information. . . . Proponents of computer-based electronic systems have not addressed the issue of the long-term survival of information. . . . The survival of information in an elec- tronic environment becomes an intellectual and technological problem in its own right .... So long as intellectual works are recorded in tangible form, the primary responsibility for defining and shaping a society's stock of knowledge rests with its librarians. -Gordon B. Neavill, "Electronic Publishing, Libraries, and the Survival of Information", LRTS 28:76-89 (Jan. 1984) [They] keep talking up ffiM' s . . . still unannounced document-management soft- ware. Running on the mainframe, the RMS software automatically rids an Office Vi- sion system of documents no one wants anymore. Users can designate individual documents to be held indefinitely, but otherwise, the system relies upon user- defined time frames or the system's own default values to regularly prune docu- ments. Not only does RMS reduce the need to buy additional storage as the years roll by ... it also helps make possible the elusive paperless office! Don't tell the printer people. -Spencer F. Katt, "Rumor Central," PC Week July 24, 1989, p.112. I thought it important to note that whereas the advent of the printing press had once stabilized our texts by removing the pitfalls inherent in manual copying, the com- puter now renders the modification of electronic materials by the ordinary user al- most child's play. I thought it important also to stress that unless continuous efforts are made to control the integrity of TLG materials now in circulation, a multiplicity of electronic text versions will soon be inundating the field, and before long an apparatus criticus to electronic texts will become a necessity .... Most electronic data users ... do not wish to deal with such issues and problems; in fact, many of them do not understand (or-worse yet-do not care about) them. Someone must, lest re- cent progress in Classics be stifled, or even reversed. -Theodore F. Brunner, Director, Thesaurus Linguae Graecae, Newsletter No. 15 (July 1989), p.1-2. In her inaugural address . . . Berger talked about information access leading the pack of issues identified as critical to ALA members .... ''In a democracy, informa- tion access requires an information base secure from intrusion, distortion, and de- struction; one proteCted from both physical and technological deterioration. . . . '' She talked about preserving the _printed word, but also census tapes, Presidential tapes and other non-print information, and above all, computer security. -"Patricia Wilson Berger Inaugurated as ALA's Presi- dent," Library Hotline V.18 No.27, p.1 (July 10, 1989). 246 College & Research Libraries of the future. I have heard otherwise re- sponsible scholars speak of the desirability of changing established texts to suit imme- diate needs or to ''improve'' them for spe- cific purposes. De Sola Pool speaks of such textual proliferation as overwhelm- ing a true identification of what is, in fact, the world's literature. Will this be the case? Will scholarship stand for it? Will libraries? Libraries and li- brarians should help to see that it not hap- pen. And we should do more than just stick a finger in the dyke. Someone must provide mechanisms to establish and maintain firm electronic texts. Librarians are in fact well placed, for historical and functional reasons, to assure that a body of knowledge can be identified, and to provide the means to assure that society's intellectual growth is not based on shifting sands. The great asset of information in elec- tronic form is also its great liability. Gor- don Neavill speaks of the "malleability" of electronic information; that is, its ability to be easily transformed and manipu- lated. 10 Electronic manipulation, while de- sirable, is a two-edged sword: information that is easily changed is easily lost. "Neavill identifies three ways in which information can be endan- gered through its existence in elec- tronic form. 11 Neavill identifies three ways in which information can be endangered through its existence in electronic form: 11 First, formal contributions to scholarly literature that are apparently in low de- mand may intentionally be purged. Gerald Salton of Cornell, among others, has written in favor of this for less interest- ing items and those less wanted. 12 There are cost-effectiveness issues at play here; in a print collection, purging unused ma- terials is an expensive proposition to iden- tify, physically remove, and delete from the various records. The same is true for electronic information. The space to store print materials doesn't cost much in to- May 1990 day's library accounting schemes, but is a very specific and budgeted cost for elec- tronic systems. Second, unpublished electronic writ- ings that could have secondary, later his- torical use are more likely than paper to be purged as of no further primary use. These are the kinds of ephemeral writings that in the past have eventually found their way into archives or special collec- tions: drafts of novels, pre-contractual ne- gotiation papers, and even certain classes of White House memoranda. 13 The third way in which information can be intentionally endangered is through the updating of noncurrent information. Examples include publisher lists, annual data handbooks, short-term economic data, and outdated versions of computer software (especially where it is only dis- tributed electronically, as for example Ker- mit). As we reach pure electronic publish- ing, where hard-copy is not produced, such cases will become more and more fre- quent. (A recent bellwether example is the difficulty the Museum of the Moving Im- age has just found in locating video games less than ten years old.)14 So far these examples on{y address the ''keep or lose'' possibility: either we keep the particular data set or we lose it in its entirety. There is a more insidious danger, and that is the danger of data modifica- tion. Such modification may be legitimate and public, or it may be illicit and fraudu- lent. Both cases present libraries with problems. A legitimate modification of data might exist, for example, as Books in Print is updated to reflect changing inven- tories. If-when-the database exists only electronically, what is the possibility of maintaining an archival record of the pub- lishing industry? And whose responsibil- ity will it be? For scholars, lawyers, business people, and bureaucrats we have until now had confidence that in a printed work my text will be the same as your text, and that my page reference will guide me to the same information as yours. Now, however, the transience of data prevents the assump- tion that my copy is the same as your copy. This will present one set of prob- lems for business and government, and a similar but more long-range set of prob- lems for scholarship. The case will also exist of fraudulent and opportunistic modification of data. Con- sider an electronic publication as a result of which the author is sued for libel; and consider the possibility of the author then modifying the offending text. What hap- pens to the lawsuit? What happens to the intellectual integrity of the text and its dis- tributed copies? We know that integrity has been want- ing recently at almost every level of gov- ernment and in many business institu- tions, so we know we can anticipate conscious manipulation of data from these sources. Academia is not pure either, as the pressures of ambition and tenure have already led to the falsification of experi- mental data and reports. We need tools to assure the integrity of electronic information; not only to pre- serve versions of texts, but to preserve them in unmodified form. (Let me sketch- ily suggest the exploration of document checksums, calculated using letter and spacing values and maintained as part of every document, according to a standard adhered to by major word-processing vendors that would manage to ignore the effect of purely formatting information.) If such tools are not provided, although librarians are well placed to provide them, then scholarship and intellectual life will be damaged in two ways. Scholars and students will become cynical and skepti- cal; and as a consequence society's scien- tific and intellectual growth will be slowed. As Neavill says, electronic publishing proponets so far ''have not addressed the issue of the long-term survival of informa- tion .... The survival of information in an electronic environment becomes an intel- lectual and technological problem in its own right." If we want to assure perma- nence of the intellectual record that is pub- lished electronically, he urges, then it will be necessary consciously to design and build mechanisms within electronic sys- tems in order to do so. 15 Physical preservation of media is one thing. Intellectual preservation is another. In electronic publishing, how will ~evi- Electronic Information 247 sions be made (and how will libraries re- cord them)?16 Who in research libraries should be responsible for monitoring the volatility of electronic information? In "circulation" of electronic information, for what safeguards is the library respon- sible? Is data manipulation and electronic fraud a library concern? If so, is it a techni- cal services concern? If not, whose is it? IV. OPTIONS FOR RESEARCH LIBRARIES IN AN ERA OF ELECTRONIC INFORMATION There are four potential models for tech- nical service operations in large libraries in a technological era: 1. Do nothing, or little: In this model, a library continues traditional acquisition and information provision patterns. Vir- tually all research libraries will do more than this. For a library that follows this path, how- ever, the consequence for technical ser- vices will be stagnation. Patrons will go elsewhere for much of their information. Technical services will remain highly sig- nificant within a less significant institu- tion. 2. Expand access to electronic materials: Continue to purchase hard copy, but also acquire electronic data and provide ex- panded access to off-site information. Most research libraries are taking on this task. Some regard it as the end, not the be- ginning, of what they do, at least for tech- nical services. Y. T. Feng' s prediction of its limited role is that ''technical services will continue to keep somewhat its traditional bibliographic orientation, but with more and more dependence on electronic tech- nology to improve efficiency and to facili- tate cooperation.' 117 In this option, the hard-copy acquisition and cataloging activities will become pro- portionally smaller than at present, infor- mation access is considered an issue pri- marily left to the public service librarians, and the importance of traditional technical services functions will proportionally de- crease. If the traditional functions are all that are maintained, technical services will be- come a kind of personnel backwater as it 248 College & Research Libraries ''If the traditional functions are all that are maintained, technical ser- vices will become a kind of personnel backwater as it congeals into a rule- driven decadence." congeals into a rule-driven decadence. Traditional technical services will never disappear, for hard-copy needs are fore- seeable indefinitely; and certain standard technological requirements will continue, as for example the provision of OPACs. But as is already happening, technical functions are becoming routinized as they become better understood and auto- mated. 3. Develop interfaces: If technical services take on the task of providing intelligent in- terfaces between access systems, then this necessary patron service will provide a natural growth pattern for our trained and future staffs. The skills we regard highly now will be supplemented by further technological skills. By their nature these new skills will require effective communi- cation and transmission throughout the li- brary, reinforcing and enhancing the tra- ditional role of technical services at the center of information provision to librari- ans and patrons. As John Sack has sug- gested, ''the role of the integrator of such systems may be the professional high ground. " 18 There is a danger that a caste system will develop: a few professional experts could become counterposed to journeyman cat- alogers. (We see something like this now where separate library systems offices ex- ist.) It is up to all our professionals and to management to see that this does not oc- cur; not only would it be unnecessarily di- visive, but it would inhibit the provision of the most effective access to information. 4. Assert institutional responsibility for the preservation of knowledge: This is the most difficult model of the four, and the most speculative. For libraries to take on the preservation of knowledge, and not just of artifacts, may seem simply to be an extension of our existing rol~ in the classic paradigm (ac- May1990 quire, organize, make available, and pre- serve). I believe however that it will be dif- ferent and more difficult. The obstacles include the financial, the technological, and the social. Because in preserving elec- tronic information we may be attempting to go upstream against the current of tech- nologically and socially established throw-away patterns, the task may be im- possible. Archiving of information solely in elec- tronic form is a recent development, but will grow. Much U.S. census information will be preserved only in this way. The di- rector of corporate technology of Xerox says "Paper is portable, disposable, com- fortable to use, and cheap. It will, how- ever, lose its role as an archiving medium. The 'truth', the up-to-date information, will not be on paper but in electronic form. " 19 An Educom task force has initi- ated discussion of serious aspects of the intellectual property issue with the intent of expanding the concepts both of author- ship and of intellectual access beyond the constraints imposed by the history of print. 20 In these developments the preser- vation of electronic information from taint is not addressed, though some scholars and librarians are beginning to call atten- tion to the problem (see sidebar). Librarians' acts of preservation have al- ways been upon objects we have owned. To protect information we will continue to require some title over it. We need to ap- proach the intellectual property issue with the following understanding: society must grant libraries some form of trustee- ship over knowledge if society wishes knowledge to be preserved. Libraries-and our parent institutions- must be prepared to pay in part for this privilege and responsibility. It will not be easy to confront the information industry with this concept; it will be desirable rather to work out a common understand- ing of rights and responsibilities if libraries are to take over a preservative function while the return on investment for infor- mation creators remains appropriately in their hands. It will require the intelligent political action of libraries, universities, civic groups and public bodies, and ven- dors to work out an acceptable form of knowledge trusteeship. Another way to preserve information and to maintain trusteeship over it is, of · course, to create it in the first place. For ex- amp~e, libraries could actively become publishers, or at least distributors of elec- tronic forms of certain data. The Research Libr~ies Gro~p is in some respects taking on this role with the Eighteenth Century Short Title Catalog, the Avery Index, and the ~.EMDB. The National Library of Medicme has long done so with Medline. Universities and other public agencies re- sponsible for libraries need to take seri- ously the idea of information ownership. There is a confluence of interests on this is- sue and the serials pricing crisis. For cost rea.sons. t~is has also led to proposals for uruversibes to assert trusteeship over the information they create. 21 The Association of Research Libraries has suggested that university presses s~ould ~gain consider taking over publica- bon of JO.urnals .. Ac~demic libraries might well begm considermg the publication of elec~ronic journals, thereby actively be- commg part of the electronic information distribution process. Aside from the ad- va~tage of ~ectly maintaining trustee- ship over the information, this would al- low libraries to learn the pitfalls and the opportunities of the process, and to be part of .the dial~gue that is changing the face of information transfer. Libraries are in a good position to provide the technical facilities that are necessary for electronic publication: we have ready to hand the in- formatiol_l itself, the computing services, and the hterate and numerate editors in- formation compilers and database ~an­ ~gers. As }am7s Thompson has suggested man acute article on serials pricing, it may be time for "the idea of the academy retak- ing control of the bulk of scholarly pub- lishing."22 Electronic Information 249 V.SUMMARY AND CONCLUSION I have suggested several possibilities for t7chnical services roles in the coming pe- nod,. broadly. comprising stagnation, ex- pansiOn, and mnovation. In the latter two models, the role of technical services is to become more technologically adept. There is a ne~d .for a c~ngruence of sys- t~ms .and o.f b~bhographtc skills and orga- ~a~IOn w1thm libraries. This is happen- mg m many larger libraries, as the same people have become responsible for both sets of skills. Unders~anding and implementing the preservation of electronic data will lead to increasing sophistication within library staffs about systems, hardware and soft- ~are, copy~g techniques, backup mecha- rusms, optical technology, networking, and data transfer. But it will also lead to in- creasing sophistication about legal issues, tech1_1ology transfer, intellectual property, and mform~tion provisi.on. These knotty problems will actually rud us in attracting the staff we need who will want to rise to these challenges. For technical services the medium becomes the message a~ never before. If libraries are to become active agents in the .electronic ~evolution, once again as- sertmg cur atonal responsibility over infor- mation in all its forms, then the technical service function must transform itself. Electronic information is going to thrust change upon technical services. As a con- sequence, technical services units will ei- ther stagnate or transform themselves along with their libraries; they are unlikely to remain similar to what they have been in the past decade. REFERENCES AND NOTES 1. ~~i~:S)L1:r~ ."I~~nting the Ele~tron~c University," College & Research Libraries 49:291-304 . Y . · . . hon nnportant readings mclude: Caroline Arms, ed., Campus Networkin Strate- gtes (n. p .. Dtgttal Press: 1988); and the Research Libraries Group Local Systems Plannin ! ·t tee, Agenda for October 7, 1988 Meeting; includes returns of ''RUN Needs Assessments g 0 ~s - esp. Colorado State, Stanford, Yale. urvey. ee 2. Susan~· M?tley, "Optical Disc Technology and Libraries: A Review of the 1988 Literatur " CD ROM Ltbranan 4:12 (May 1989). e, - 250 College & Research Libraries May 1990 3. A harbinger for the future of Kermit and similar programs is attested to by plans to include it in the read-only memory of 250,000 personal computers to be "manufactured and delivered to Soviet secondary schools as part of the current five-year plan .... " Info-Kermit Digest 9 (13 June 1989), s. v. Kermit Conference Report (this reference is to an electronically distributed journal which, as it happens, lacks pagination). 4. I'm talking here about a public access system, not about an integrated processing system, still a desideratum for handling traditional acquisitions, cataloging, serials, and circulation functions. 5. For a discussion both of the scholars' need and the kind of activity that will satisfy it, see Donald Langenberg, "Supporting the Global Scholar," Academic Computing 3:12-16, esp. p.15. 6. Edwin Brownrigg, Clifford Lynch, and Mary Engle, "Technical Services in the Age of Electronic - Publishing," Library Resources & Technical Services 28:59-67 Oan. 1984). 7. Gordon B. Neavill, "Electronic Publishing, Libraries, and the Survival of Information," Library Resources & Technical Services 28:76-89 Oan. 1984). 8. Ithiel de Sola Pool, "The Culture of Electronic Print," Daedalus 111:27-28 (Fall1982). 9. For a summary at one stage of the ongoing dispute, see Charles Roman, "The New 'Ulysses': The Hidden Controversy," New York Review of Books, December 8, 1988. 10. Neavill, p.77. Professor Harvey Wheeler, of the University of Southern California, in a keynote speech at the October 1988 UTA conference, spoke of the "fungibility" of information. I think this is an incorrect use of the term, as it implies that one piece of information is just as good as another; and I think its use reflects Wheeler's lack of concern for firm texts, and explicit enthusiasm for the riot of change that becomes possible. 11. Neavill, p.81. 12. Gerald Salton, Dynamic Information and Library Processing (Englewood Cliffs, N.J.: Prentice-Hall, 1975), p.16. 13. I'm referring of course to the attempted deletions that took place during the Iran-Contra scandal. See also Roberto Suro, "Tapes on Alaska's Oil Spill Erased by Exxon Technician," The New York Times (July 2, 1989), p.14. 14. Glenn Collins, "An Archeological Hunt for 'Old' Video Games," The New York Times (June 19, 1989), p.C11,C16. 15. Neavill, p.78. 16. For a knowledgeable summary of the process and problems in keeping track of authorial and edi- torial revisions, see Thomas W. Smith, "Desktop Publishing in the University: Current Progress, Future Visions," Academic Computing 3:26-36 (May 1989). 17. Yen-Tsai Feng, "Technical and Reader Services for the Research Library-The Challenge for the Next Decade,'' Journal of Library and Information Science 13:54 (April 1987). 18. John R. Sack, "Open Systems for Open Minds: Building the Library without Walls," College & Research Libraries 47:541 (Nov. 1986). 19. Robert Spinrad, "Dynamic'Documents," Information Technology Quarterly (Harvard) 7:18 (Spring 1988). 20. Dana Cartwright, Steven W. Gilbert, and Peter Lyman, "Beyond Intellectual Property," Educom Review 24:7-8 (Summer 1989). 21. Ann Okerson, Of Making Many Books There Is No End: Report on Serial Prices for the Association of Research Libraries (Eastchester, N.Y.: April19, 1989). 22. James Thompson, ''Guest Editorial: Journal Costs: Perception and Reality in the Dialogue,'' Col- lege & Research Libraries 49:482 (Nov. 1988).