College and Research Libraries Open Systems for Open Minds: Building the Library without Walls John R. Sack As scholars are more frequently connected to electronic networks allowing access to research information and collegial interchange, the roles of the library and the computer center will need to shift from those of a central repository holding information and technology to that of sleuths and integrators of disparate information sources. The malleability of electronic information, the openness of systems containing it, and the expectations of scholars as they become less " patrons" of the library and more "users" of electronic information services will drive us to develop systems and organizations that readily facilitate the transmission and transformation of knowledge. o sharpen the point of my topic, I will begin by asking this ques- tion: when, how, and by whose hand will libraries disappear? There are several ways in which libraries might disappear, not all of them bad, and I suspect that many in RLG libraries are ac- tively working on effecting the disappear- ance. Of course, the most obvious way li- braries might disappear is through disuse, that is, through a shift of scholarly re- search away from libraries towards other places and other media. According to this scenario, libraries might disappear be- cause they lack significance in the aca- demic program-because they are notes- . sential. Many campus organizations are vulnerable in this way: one might suggest intercollegiate athletics; another might suggest government or faculty commit- tees. But a second way libraries might disap- pear is not so frightening: the library might disappear simply because it . blended so successfully into the back- ground of a scholar's activity that the scholar never needed to regard it explicitly as a place to go, or an interruption in an ideally seamless activity of research and reference. According to this second sce- nario, using a library becomes so effortless and natural an activity that scholars no longer have to think of it as a special (and time-consuming) component of their re- search. Thus, libraries disappear because they become invisible and because their location. is wherever you are: "without walls," if you will. More than a physical location, the library becomes a medium or ubiquitous utility, a service always ready at hand. Perhaps an example of such is the telephone or television . My last variation on library absconditus is a challenging one: the library may dis- appear by becoming something so differ- ent that patrons are tempted to call it by another name, as they treat it less as a storehouse than as a vehicle or conduit or service. If Wittgenstein is right, ':Yhen this John R. Sack is director of the Data Resources Group, Information Technology Seroices, Cypress Hall, Stan- ford, California 94305. This article was presented in its original form at a seminar sponsored by the Research Libraries Group in November 1985 at Conoco's Purple Sage Ranch in Ba ndera, Texas, th rough the hospitality of Conoco Inc. 535 536 College & Research Libraries happens we'll notice it by the way scholars begin speaking about libraries. Perhaps library will become an odd sort of verb (e.g., I libraried that topic and found new approaches) much as telephone be- came a verb soon after it was established as a noun. What I have in mind is an exten- sion in the scope of what library connotes, a change as substantial as the change from portraiture to photography. The impor- tant factor is not extension so much as ex- tensibility, in which the ability to change becomes fundamental to the medium. The new library might differ from the old one the way a television differs from a win- dow. By whose hand will libraries disappear? This question involves the control of destiny: the disappearance can come about at the libraries' direction or by the scholars' de- fection. Let me also assure you that similar questions of destiny weigh on the techno- logical professioDS, where there is far less of a ballast of tradition to rely on for safety. In fact the most fruitful transformation for each of these professions will come with the assistance of the other. EXPANDING THE VIEW OF LIBRARIES AND COMPUTER CENTERS Let me suggest some elements of a cri- tique that would encompass libraries and computer centers at the same time. Both libraries and computer centers have "high visibility" to scholars; they are modest hurdles, with immodest potential. Both institutions are hard to use, distant, rule-bound, inflexible; they aren't readily assimilated for the scholar's work-space and -time. Both sometimes appear to be devoted to a'' divide and complicate'' phi- losophy when the best scholarship tries to integrate and unify. The value of the library and the value of the computer seem to increase for the scholar as their nuisance factor-the inter- ruption of a line of inquiry to deal with the opacity of a foreign place, culture, and procedures-disappears. That is, as each institution becomes more malleable, open and translucent, it becomes more able to shape itself to the scholar, becoming less a specific place than a service and a near- transparent medium. The tool can then be November 1986 tailored the way one has a suit tailored, or the way one arranges an office. The key to achieving this malleability and placeless- ness is technology that can be shaped by imagination. The plight of libraries and of computer centers a decade or two ago was that they provided very small windows on a very large world. If one's view of ·the world were limited to what one could see out the window of one's home (particularly in downtown Palo Alto), then that view would be very limited indeed; the pano- ramic extensions to this view have come from technologies that let people be where they were not, such as the telephone and television. The telephone-answering ma- chine and the videocassette recorder even let these technologies act as our agents when we are not present, shifting time as other media shift space. While we may have personal antipathy for these technol- ogies, their unique effects on information access and distribution are undeniable. Libraries and computer centers have taken steps to enlarge their windows, in a sense. They have both connected them- selves to networks, for example. But they · have not really taken the steps to shift from windows to television, if you will. The current networks are still very lim- ited, rather like TV with only three broad- cast channels. Just as there is vastly more information available on television than there is out the window, there is vastly more information available to the fully networked scholar than in the utility-connected library. And yet t~e profession's capacity to handle in- formation has been increased only by add- ing more people as specialists in new or growing areas, rather than adopting a new strategy. The current strategy may not be sufficient to keep pace with arithmetic growth in information and ac- cess; it certainly isn't sufficient to handle exponential growth in the number of in- formation providers and the amount of in- formation accessible to individuals. We are, in a sense, running along as if we sim- ply had to flap our arms harder to take to the air. New approaches are needed, and a novel one has been suggested by Alan Kay, now an Apple Fellow, and inventor of the "dynabook" concept when he was at Xerox Palo Alto Research Center (P ARC). When Kay was at Atari he built a working model of a semiintelligent pro- cess, called an agent, which would scan various electronic news services at night and build a custom newspaper for you, based on its knowledge of what you would be interested in. Thus the headline might be that U.S. planes were bombing El Salvador or it might just as well be that your afternoon appointment was can- celled (which the agent learned by reading your electronic mail). Kay is now studying how one imprints agents with a character that allows them to recognize information of value to particular individuals. Society already has many models for such agents. The stock broker and real es- tate agent come readily to mind, and even the private eye is functionally similar. And SDI (selective dissemination of informa- tion) searches are a primitive, automated example from our own profession. At Stanford, as part of a project to study elec- tronic communication of research materi- als (Project Concourse), we will be allow- ing faculty and students to ' 1 characterize' I agents that will examine bibliographic and nonbibliographic databases and bulletin boards, retrieving new items of personal interest and placing them in electronic mail boxes. THE CAPACITIES OF THE RESEARCH LIBRARY The possibility that the library will dis- appear for the research scholar is also sug- gested by the Newman Report on Higher- Education Policy commissioned for the Carnegie Foundation for the Advance- ment of Teaching. The report states flatly that 'I the research community is moving beyond the capacity of the research li- brary'' (Chronicle of Higher Education, Sept. 18, 1985, p.17-29). Indeed the report's sec- tion on research libraries is worth a brief review here, since it argues that our pa- trons have supplanted the libraries- appropriately or not-by means of outside technologies and services available to them. The Newman report argues the need to find ways of defining the function of the library. Because of the cost of materials, Open Systems for Open Minds 537 interlibrary cooperation will clearly be necessary to provide researchers with the products of the "knowledge explosion." But by itself membership in a network will not be enough-not even membership in several networks. The fact is that despite patron confusion about widespread elec- tronic information, use of the upcoming tools will be easy enough to allow each person to be his or her own librarian. What is more, many of us believe elec- tronic access will be the way out of the par- adox that some materials are needed infre- quently but needed urgently and quickly when they are needed. Still there are sev- eral interesting problems with electronic access that the report emphasizes: • You first need to know where to look in order to find what you're looking for; this clearly penalizes patrons working outside their "home" discipline, for which they presumably know standard bibliographic sources and practices. • The refereeing and public criticism found among printed works is not part of the electronic journal article. • It is hard to establish an orderly histori- cal record online. • Funding problems arise because the li- brary traditionally discriminates in fa- vor of acquired as opposed to accessed ma- terial. Funding discriminates in favor of the haves as opposed to the have-nots. The last particularly affects librarians managing collection development and/or public service functions. For traditional acquired materials, the charge is levied only when the library gains ownership; but for electronic media, the charge is lev- ied when one gains access. The Newman report suggests that elec- tronic access will require a shift in library service outlook from "owning to sleuth- ing." (This electronic access should not be confused with automation of technical processing, of course.) The role of the sleuth "requires an educational and emo- . tional [philosophical] commitment to the shift in outlook required to change from owning, cataloguing, and lending, to be- coming electronic data sleuths ready to link a student or faculty member to some- one else's data bank." The Newman re- port's section on research libraries ends by calling for something with the unfortu- 538 College & Research Libraries nate name of "Scholarship Information Systems." In any case, whatever the name, this transformation would fit my third notion of the disappearance of the li- brary (qua warehouse) and the emergence of the library as the integrator of informa- tion. To begin this shift of emphasis, library staff will need many of the same tools that have started their faculty colleagues down the path of electronic access. Many re- searchers, for example, among science and engineering disciplines, frequently communicate electronically with their col- leagues, sharing text, experimental results, citations, and abstracts. The tools that have allowed some scientists to move away from the library are the same tools to which library staff need ready access. LOCAL AND REMOTE IN DECENTRALIZED INFORMATION NETWORKS I'd like to suggest a simple conceptual model that shows the topology of research information from the library's and the scholar's point of view. Let me call these the Ptolemaic and Copernican views (see figure 1: of course the library has to deal Scholar Scholar View of the Library: "Ptolemaic" November 1986 with thousands of scholars, and the dia- gram simplifies the effect by showing only one). I suspect that, in the scholar's view, the transformation of the library from the Pto- lemaic to the Copernican view is already an established fact. The library's adapta- tion to this transformation is not as well established; neither is the computer cen- ter's. Again, we may look at our own lan- guage as an indication of our attitudes: we refer to the extra library patrons as "re- mote users." But from the patron's point of view it is the physical library that is re- mote from his or her workplace. Similarly, in computer centers five years ago a printer was local if it was physically housed with the computer and remote it if was at a user's site. The opposite is now the case when we say those words: users now have local printers in their offices and think of the large printers at the computer center as remote. The library is a node in the scholar's in- formation web. But the library must take into account the scholar's entire research process and the variety of his or her sources and resources. The library must then comport itself as if it were a responsi- Library Acquisitions View of the Scholar: "Copernican" FIGURE 1 ble member of the scholar's ''information society.'' It must develop strong relation- ships with other information units in and out of the university. The scholar is sur- rounded by the resources of this society, some of which are facilitated by (not nec- essarily all "held by") the library. At Stanford about two years ago we came up with a similar critique of adminis- trative computing architecture (see figure 2). The analogy is not exact but the point is to identify the appropriate "center" for a service system and then to tailor the ser- vices to fit the entity at the center. In ad- ministrative computing, the center should be the individual department with its comprehensive service needs; in research, the center is of course the individual scholar. Perhaps it is easier to spot the philo- sophical transition when the shoe is being worn on the university's other foot. Those who have had university management re- sponsibilities can probably see that in the older architecture the burden of integra- tion of information falls squarely on the smaller unit, with the least talent and ex- pertise to handle it, instead of being shoul- dered by the larger organizations. The in- DeptC View of the Administrative Department: "Ptolemaic" Open Systems for Open Minds 539 tegration must be performed hundreds of times, being reinvented in each depart- ment. Scholars participate in many different information networks. In some of them the scholar acts as correspondent, in some as passive recipient, and in some as crea- tor or initiator. The intersection of these many networks would be too complex to draw, but you can readily imagine what it would be like: perhaps like a galaxy of so- lar systems. The drawing would quickly lose any sense of a center even if you tried to draw only a few scholars and a few in- formation providers (which might, of course, be other scholars). Rather than showing one center node intersecting with many lesser nodes or "satellites," the drawing would have to depict many equal units sharing information on an equal basis. Computer networks are ideally decen- tr'alized in the same manner as scholarship networks. And I am not using networks merely as a metaphor: here the medium is truly the message. A network, for exam- ple, is not centered on the warehouse of the mainframe, but on the medium or ser- vice of the wire. Figure 3 shows informa- Student Records View of the Academic Department: "Copernican" FIGURE 2 540 College & Research Libraries November 1986 Then: "Warehouse-centric" Now: "Wire-centric" FIGURE 3 tion exchange as it was before and after lo- cal computing and networks provided individuals and organizations with direct access to each other as providers and us- ers. The focus is on exchange, on communi- cation between equal partners. The indi- vidual members of the network should not have to perceive some central unit dis- pensing information (or obstructing infor- mation); rather the me.chanism or me- dium should itself be transparent and open so that all the individual members need perceive is the information itself. Networks don't own resources so much as make them accessible. Also, networks are typically "peer-to-peer" in the jargon, not hierarchical. This means that any member can communicate with any other member whether that member is an indi- vidual or institution. An individual scholar's workstation may ideally have the same access to major and minor infor- mation providers that RLG's mainframe has. DEFINING THE LIBRARY'S ROLE IN THE NETWORKS But how can libraries facilitate such a network, and how can they understand and prepare for their own part in the scheme? Two complementary sugges- tions come to mind. First, one can look outside the library to see where the university nonlibrary infor- mation society is tending over the next forty-eight months. Look at both "data flows" and" dollar flows" for information technology in academics and administra- tion, considering in addition both equip- ment and space. Look at new program ob- jectives in education and research (e.g., Stanford hopes to double the proportion of undergraduates doing honors work in . humanities and science). One should also appraise such factors as new faculty ap- pointments and title chaJ!ges on commit- tees overseeing technology. One might note especially new expenditures for net- working and fpr putting computer tools in the hands of faculty and students. One would certainly take a look at tools that the computer center is teaching and recom- mending. A second important approach is to look inside the library itself and try to refocus at- tention for a time on access to information, not material acquisition. For example, con- sider the following "thought-exper- iment'': imagine that library staff had only a micro, a communications line, and a phone, but no building or collection. What sort of services would they offer in order . to provide real added value with such minimal tools? One might also consider access in smaller research-oriented branch libraries; the staff in such branches often seem to understand intuitively the li- brary's place in the network of research in- formation communication, when it is situ- ated as one "service station" among many in a department or school. Perhaps the library should take on more responsibility for providing access to in- formation that it does not possess, order, and control; more· and more research in- formation will be of this sort. This parallels the transformation from the library being an owner of books to its being an ''integra- tor of · systems.'' The former is a limited and technical function, while the latter provides a professional service function well into the future. The "integrator" is just another version of the ''agent'' I de- scribed earlier. (A senior Stanford librar- ian has told me of the problem in defini- tion of the library profession; because technology has been having such a large impact, many librarians feel the future lies in becoming ''technologists'' in order to be able to build better systems. But there will always be systems of information for the scholar to use outside any one library or field, and this is why the role of the inte- grator of such systems may be the profes- sional high-ground.) When a library does buy or build sys- tems, it should make sure the system pro- vides the most general software and hard- ware possible. This facilitates the system's participation in the networks that scholars are already and will be using. One should assume that extra-library use of library systems will be equivalent in volume to in- library use over the next decade, and that extra-library use will replace only a small part of library use, especially when the system describes materials that are not in- side the library. A library ought not to buy anything that can't connect to a network unless it can afford to dispose of the equip- ment quickly rather than amortize its pur- chase price over several years. The library staff should have access to Open Systems for Open Minds 541 the same tools that faculty and students are using for electronic access. This will usually mean personal computers, mo- dems, and access to campus electronic mail systems. EXPECTATIONS OF PATRONS AND USERS What are the needs and expectations of such "remote" and networked users? Al- most all the special needs I can identify can be derived from a basic proposition: the scholar is more a computer "user" than a library "patron." His or her expec- tations will be derived largely from the culture of computer access and manipula- tion of information, not from library ac- cess to information. This attitude will arise if only because the user is not physically present in the library. For example, the current "online strat- egy'' patrons use with the online catalog in the library is largely a substitute for that used with the card catalog; the catalog (online or not) is seen as a locator or pointer to materials on a nearby shelf. But outside the library the catalog undergoes a metamorphosis into a research tool in its own right. For Stanford's online catalog, Socrates, one can compare patron suggestions that come from library terminals with those coming from users in their home, dorm or office: the latter often request nonlibrary information and services. (There are about twelve hundred patrons with access to So- crates outside the Stanford University Li- braries. Their use runs to about one thou- sand sessions per month.) For instance, we've frequently been asked why article abstracts aren't available to Socrates us- ers. We were even asked how to look up monetary exchange rates in Socrates. (We replied to the last suggestion that an on- campus travel agent and bank were good sources of information.) Incidentally, we noted that one community of test users who had problems with the early version of Socrates were computer science profes- sors. They found that Socrates' natural bent was to support a card catalog-style search strategy, and they wanted to ma- nipulate it like a.ny other set of databases they would use in their own work. (Later 542 College & Research Libraries versions of Socrates allow this'' database'' search strategy with greater flexibility.) The point of these anecdotes is that Soc- rates is already expected to be or perceived by its extra-library users as the integrator of diverse systems of information. This is partly due to a "critical mass" phenome- non in which any large and seemingly comprehensive source of information is expected to absorb subjects that are on the fringes of its mission: the size and sophis- tication of a medium attracts the attention of information users and providers who become progressively more interested (and demanding) as the size increases. The interest of users increases as the in- vestment of suppliers increases, and vice versa; and more use breeds even more use. The shift in expectations from those of the "patron" to those of the "user" is of course gradual and stratified. It will pre- sumably happen first in disciplines whose primary sources are already online. The computer science discipline was the first to shift; physics and engineering have largely done so by now. These cultures and others now have many of their sec- ondary sources online (e.g., works of scholarship including bibliographic and numeric databases, largely because of the critical need for timeliness in some fields). The shift will also come earliest in those disciplines, such as education and librari- anship, where electronic access is itself an object of study, or as with many under- graduate students, an object of pleasure or recreation. It will spread to some degree to most disciplines, no matter how "paper- oriented" a discipline may seem at the present moment. Incidentally, many scholars are going to expect some things to carry over from the "patron" culture. Undoubtedly many will expect these new forms of electronic access to be free of charge to the individual (as Socrates is at Stanford now). Certainly this expectation will be weaker for services available outside the library. I will note some of the characteristics of the "online culture," and I think we should particularly consider the points that run orthogonal to current and tradi- tional library procedures and/or patron November 1986 culture. The theme connecting most of them is increased immediacy of access. • Users focus on results, not procedures; the computer user usually doesn't care about how or why something is done (the "hacker" mentality is an extreme example of this). • Users demand speed, not deliberation. The computer user will often have cho- sen the computer because of its ability to provide instant results; spreadsheets and electronic mail are examples here. Users are impatient with any process that leaves their minds idle while they wait for something external to catch up. • A corollary to the above: most users pre- fer a fast but incomplete answer to a late and encyclopedic response. The scholar usually needs completeness eventually (particularly in central research areas); but in the short term a single citation may be enough to supply a missing fact or direct a search further; the user, of course, wants to make the final judg- ment of sufficiency/adequacy versus completeness. • Users demand two-way communica- tion, rather than passive acceptance of whatever comes down the wire. In So- crates, we receive between five and ten communications ("suggests") a day from users. We answer any suggestion that is signed, and this sometimes leads to a dialogue on a specific issue . Fre- quently we receive acquisition sugges- tions, rush processing requests, and, on occasion, a reference question; we've even received compliments about staff and complaints about bats and bathrooms-suggesting that this is an alternative communication tool for some. We respond as quickly as pqssi- ble (sometimes within a few minutes) and use electronic mail whenever possi- ble; two-way communication allows the human aspect of a service to be per- ceived, and reduces the isolation of "re- mote" users (who are perceived as re- mote by the library but naturally not by themselves). Such service must be re- sponsive and quick, if it is to be per- ceived as helpful at all. The remaining expectations worth not- ing derive from the principle practiced by many software and hardware vendors who realize that, to survive, their systems must connect with other systems. This is the general principle of "open" systems architecture, which finds specific expres- sion in the ISO/OSI system interconnec- tion standards, the MARC data 'inter- change standard and even the Macintosh clipboard. Fulfilling such expectations is easiest with the use of general-purpose hardware and software, particularly with respect to user interaction ("interface") and capabilities and data and network communications. • Users expect you to provide a relatively seamless integration of your system with whatever other systems they use. You must understand what other infor- mation systems scholars use and how those systems might influence expecta- tions for your system. At Stanford, So- crates users frequently send citations to colleagues via electronic mail or incor- porate citations into mainframe and mi- crocomputer documents and data- bases. • Users expect electronic information to be malleable, and expect the library's system to be flexible. They judge what computers can do from their own expe- rience with personal and departmental computers. So, for example, they will expect to be able to reformat citations to meet various publications' style re:.. quirements. After all, it is not the trans- mission, but the transformation of knowledge that occupies the attention of most scholars. • Users expect the library and its system to be ready to change to meet the ex- panded potential of electronic catalogs over manual ones. As one example, the office responsible for facilitating the use of Stanford facilities by the disabled asked us to enhance Socrates displays so that blind students can more conven- iently search and display records via a voice-output unit that is already familiar to them from other contexts. .- • Users have little sense of library tradi- tion and will not readily make a distinc- tion between owned-by-library and nonowned information, or between tra- ditional library materials and those not Open Systems for Open Minds 543 typically managed by libraries. In com- menting on this, one scholar remarked that his primary need was for a research tool, not an inventory system. It is sometimes more important to know of an item's existence than to know whether the library owns it or not. Inter- library loan has to some extent made "not owned by library" just another ci- tation status; the library already fills the "integrator" role ·here on a special- request basis. The principles of openness, intercon- nection, and extensibility were so impor- tant to the design of Socrates that more than half of the desiderata developed by the design team support them. I'll note those items that directly reflect the princi- ples: • The system must be accessible from any terminal device at any speed. • The system must be accessible from every campus network and beyond. • The system must support more than two hundred simultaneous users. • The system must provide the base for MARC and non-MARC data files, and for library and nonlibrary services and functions. • The system must provide for two-way communication between staff and pa- trons . • The system must be available twenty- four hours a day, seven days a week. Scheduled downtime is never accept- able. • The system must suggest to the user that it is not solely an online version of the card catalog by providing noncata- log services. • The system must be adaptable to use on microcomputers and electronic mail networks and must support formatted file transfer. • The user must not need any documen- tation except what appears on the screen. Other expectations will develop as the library meets current ones. I think one way to anticipate some of these is to ob- serve what today seem to be some of the more atypical or advanced uses of library information. Such uses can show how people define information by showing 544 College & Research Libraries what they do with it . I will mention a few from Stanford: • One of Stanford's foremost researchers in artificial intelligence frequently ap- pends bibliographies derived from Soc- rates to notices he posts on electronic bulletin boards. The latest was a "flame" (electronic mail heatedly ex- pressing an opinion) on South Africa. • A graduate student regularly searches Socrates before he attends a lecture by a visiting scholar to see what the lecturer has written. Another student looks up additional works by authors cited in journal articles as she reads at horne. • Members of a fraternity that was al- ready on probation for sexism began electronically sending Socrates citations on risque topics to other individuals, not realizing that the sender's name was displayed to the recipient. • An M.B.A. student reinvented copy cat- aloging when she started a project of us- ing Socrates citations to index her per- sonal library. • A radio announcer uses Socrates to find unusual works to play on his program. • The student newspaper carried an arti- cle entitled "Socrates Could Teach Them a Thing or Two at MIT,'' indica- ting students' pride of ownership in the system. • Frequently the first search a faculty member does outside the library is to check that the library has acquired all of his or her work. • Several people have described ''doing random searches for fun'' and using So- crates for "fishing expeditions" and ''whimsical browsing.'' • A student show this year features a skit in which the founders of the university get lost inside a Socrates terminal. • A staff member sent us a somber au- thority correction noting her father's re- cent death and asking that it be re- corded in the main entry for works he authored. November 1986 • It has been suggested that Socrates should ''contain everything,'' but most people would settle for retrieval of jour- nal articles, complete retrospective con- version, and access to other libraries' holdings (UC Berkeley and LC are men- tioned most often). · • One person suggested that Socrates note which items were available for pur- chase in the bookstore and another sug- gested online ordering of pizza. Perhaps the most pleasing report from a user was that "Socrates was the biggest li- brary improvement since open stacks." That comment certainly puts "openness" and access in perspective . A PARADOX FOR PROFESSIONALS I began this paper with something of a paradox, talking about building the library without walls-an open system readily fa- cilitating transmission and transformation of knowledge. The real revolution, if there is one, is not so much in the amount of information available but in the way individuals will adapt to this wealth using technological tools. That this technology will have an ef- fect as lasting as Gutenberg's technology is a commonplace, but I have chosen to fo- cus on the malleability of electronic infor- mation and the increasing openness of systems containing it as the distinctive characteristics to watch and respond to. These characteristics combine to encour- age the spread of information and ideas beyond the capacity or control-for better or for worse-of information specialists . If ready access to and demand for great quantities of information by individuals defines the next decade, then those of us in the information professions should de- fine ourselves in a positive relationship to the trend-as agents, as sleuths, as inte- grators of systems. A voiding for ourselves the fate of our own card catalogs, we will find that sharing information, not merely hold~~ IS the key to our own future.