This article was originally published as: "E-Mail for Development: An Exploratory Study of the CDS-ISIS Distribution List"

Email for Development: An Exploratory Study of the CDS-ISIS Discussion List
by Christopher Brown-Syed and Kerstin Witzke.

KEYWORDS:

CDS/ISIS; Internet; Electronic Mailing Lists; Unesco; Information in development.

ABSTRACT

During the Spring and early Summer of 1996, a study was conducted to examine the role and usage patterns of the CDS-ISIS user forum, a Bitnet mailing list (Listserv) which constitutes a multilingual and multinational forum for librarians and documentation specialists using Unesco's CDS/ISIS software. Between April and July 1996, more than 600 electronic mail messages, representing the complete traffic in the Bitnet list, CDS-ISIS were downloaded and examined, together with summary lists of subscribers provided by the Listserv software.

A content analysis of these postings allowed the creation of a snapshot of the list's user demographics, functions, and use patterns. As well, an ancillary survey was sent to all current subscribers of the list, to elicit supporting information. While data obtained using this second strategy cannot be regarded as statistically relevant, since the survey's response rate (8%) was by no means representative, its responses lend additional credence to the content analysis of posted messages.

CDS/ISIS is a scalable cataloguing system based upon the ISO 2709 (MARC) format, developed by Unesco with the aim of providing low-cost automation for libraries and documentation centres in the developing world. The ISIS electronic mailing list was selected for study because of its geographic scope, its relation to the concerns of library practitioners, and its use as a means of requesting and obtaining support in instances where direct support by the vendor must be supplemented by support from other practitioners, or where the geographic dispersion of installations effectively prohibits traditional software user group activities.

The study concludes that the list is being used effectively, both to disseminate information originating in the “central” or “developed” nations, and among users working in the “less developed” nations of the technological “periphery”. It is a highly appreciated support tool for the CDS-ISIS user community worldwide, and an effective supplement to traditional vendor-initiated technical support. The study suggests (but does not clearly demonstrate) a direct and positive relationship between users’ expertise and their posting and reading of software-specific mailing lists. Comparative studies should be undertaken using other software-specific mailing lists to ascertain the potential effectiveness of electronic mail in meeting the technical support needs of librarians in the developing nations.

BACKGROUND

During the debate over the necessity for a New World Information and Communications Order (NWICO), representatives of the developing nations, and some scholars in the industrialized countries, argued that information was not only a precursor to economic development, but also a fundamental human right. Much of this debate took place on the premises of the United Nations, especially Unesco, (whose 1984 report: Many Voices One World, was the topic of heated debate), and at Non-Aligned Movement conferences, during the era of super-power confrontation, but it has received a wider audience through the works of Smith, Chomsky, and others.

The NWICO debate aroused ire in the West, particularly because its proponents seemed to be arguing for censorship of the press when they asked for local editorial control, or when they raised concerns over the concentration of press ownership. While this may have been true in some instances, the NWICO argument was more subtle and far-reaching. Its proponents saw information, particularly that passed on by the mass media, as originating primarily in the wealthy “central” nations of the North, and being consumed in the poorer “peripheral” nations of the South. Along with First World movies and newscasts, they argued, went First World concepts about wealth, success, and society.

Even the terminology used in the Northern media was being imposed upon Southern consumers, they claimed. For instance, analogies like “The Botswana meteorite carved out a crater the size of a football field”, clear to home audiences in the USA or Canada, might be misleading to people in nations which played soccer. Worse yet, such countries might find themselves calling their national sport by its American designation “soccer” instead of “football”, as the big networks took over responsibility for sports feeds. Even events occurring in the periphery were being repackaged for audiences in the centre, and there was some fear that as language went, so would go values and concepts.

In essence, both left and centrist commentators agreed that the media were becoming increasingly centralized, and that communication and information necessary for development appeared to be originating in the central nations, or being repackaged there, and distributed to those on the periphery. Liberal optimists like Ithiel de Sola Pool, saw the remedy in technology. Communications satellites, he argued, could become true “technologies of freedom”, because they would obviate the problems posed by landlines and allow small countries to benefit from direct communication with powerful ones. But critics argued that this sort of communication would still be largely one way. Thus, in the 1970s and 1980s, we find doctoral dissertations being written about the need for Islamic bibliography, Hebrew documentation centres, medical abstracts in Urdu, or library catalogues in Swahili.

The development of the contemporary Internet, coupled with the dramatic increase in computer power and decrease in computing costs, presents an entirely new set of possibilities for information delivery round the world. Are librarians and computer site managers in the developing nations making use of the tools now available – Bitnet Listservs, Usenet Newsgroups, World Wide Web sites? If so, is communication truly bi-directional, or does the majority of traffic flow from the centre to the periphery?

ROLES OF SOFTWARE-ORIENTED ELECTRONIC MAILING LISTS

In widespread use since the mid 1980s, Bitnet and Internet electronic mailing lists (Listservs) have broad application in the library and information sphere. Listservs can be regarded as tools for providing remote reference services as well as support for professional subgroups such as cataloguers, members of particular associations, or users of software products from specific vendors. For example, the list LIBREF-L promotes communication on reference topics. JESSE provides a forum for library educators, LIBJOBS a global recruiting venue. Some lists cater to users of particular suites of library software. Directories of library-related lists are available at many Internet and World Wide Web sites, of which the BUBL service in the United Kingdom (http://bubl.ac.uk), and the Library of Congress site in the United States (telnet://locis.loc.gov) will serve as examples.

A Listserv is a "mail reflector" residing upon a particular Internet computer, which manages one or more topical electronic mailing lists. (Krol, 1994:138). By sending a single copy of an email message to a list server, a computer user can distribute that message effectively to hundreds of other users worldwide. While the lists are maintained by purpose-designed programs, subscribers to lists need no special software, other than one of the commonly used Internet mailers such as Pine, Eudora, Pegasus, or smail. Listservs originated in the Bitnet and Arpanet subnetworks of today's global Internet, while an alternative forum for discussion, the Usenet Newsgroups, were concurrently being developed in the Usenet (UNIX) community (see news.announce.newusers).

As the information community becomes increasingly globalized, Listservs and Newsgroups may become commensurately vital tools for establishing and maintaining contact among members of disparate groups, who might realistically never be able to establish such contact by other means. Traditional methods such as regional user group meetings, annual conferences, printed journals, and vendor initiated training and briefing sessions, are by contrast expensive, time sensitive, and logistically problematic at a global level.

For librarians and other users alike, electronic mailing lists and newsgroups offer financially attractive alternatives to direct person-to-person contact. End users of technology, particularly the "user groups" comprised of clients of particular software and hardware vendors, have long benefited from electronic discussion. The Advance-L group, which is composed of users of Geac, and groups associated with Innovative Interfaces, NOTIS, and Dynix software clients serve as examples. The usage patterns, and the relative effectiveness of these end-user discussion forums, may be of potential interest to librarians seeking to participate in them, or to provide similar services to groups of library patrons who share the same academic or professional concerns.

Libraries with modest means and geographically disparate clients, have been able to employ electronic mail in reference service throughout the 1990s. By the late 1980s, Poyne and Edison (1988) had explored the effectiveness of email in reference situations. Nor was email reference confined to larger institutions. Australia's Orange Agricultural College, whose enrollment was only about one thousand, was using email to provide reference services to its off-campus students as early as 1992 (Brown-Syed, 1996).

Electronic discussions would seem particularly attractive to groups of scholars or practitioners

whose members are geographically dispersed, especially if these groups contain many

representatives from the developing nations. However, the extent to which they could effectively reduce or even eliminate the need for direct contact, conferences, or meetings with computing vendors and developers, requires further research.

REVIEW OF THE LITERATURE

Information has long been suggested as a precursor to economic and social development. (Smith, 1980; Pool, 1983; Unesco, 1984). However, the problems of unequal access to information between the more and less developed areas of the world are exacerbated by the financial problems posed by traditional means of communication.

The Computerized Documentation Service/Integrated Set of Information Systems (CDS/ISIS), has been the subject of various treatments in the literature of librarianship and documentation over the past decade. Articles about ISIS tend to fall into various categories. Beginning in about 1989,

the literature has included:

(A) field reports about the implementation and use of ISIS for specific library and documentation applications,
(B) accounts of the features of new software releases,
(C) discussions of user-written utility software and enhancements to the software suite,
(D) accounts of experiments and pilot projects using ISIS,
(E) articles written by developers and system administrators to provide technical tips and assistance for other ISIS users, and
(F) analyses of the ISIS user communities in various regions or nations.

As a selection of recent articles demonstrates, ISIS is being used in a variety of settings and for managing bibliographic data on a wide range of topics. Wongkoltoot and Indee describe the use of ISIS at the Thailand Development Research Institute Library (Wongkoltoot, 1992). Bokos (1993), describes the system's use at the National Library of Greece, Santiago (1992) discusses the use of OSIRIS, a companion serials-management system, while Ravi and others (1992) discuss production of an Indian scientific catalogue. Treloar (1991) writes of an Australian project to catalogue medicinal plants using the software, while Chowdhury (1995), Chaudhry (1990), and Wijetunge (1996), discuss the use of ISIS in vocabulary control, indexing, and authority work.

Examples of software developed by ISIS users include utility programs for record conversion, and even complete "front ends" or user interface programs. At one extreme, utility programs have

been written by users to convert record formats, enhance the appearance of records, add diacritics, and other relatively simple tasks. The program "DB3ISO", allows users to convert records from the format used by popular database management programs such as dBASE III, and similar DBMS systems, to the ISO 2709 format used by ISIS. (Goyal, 1990). A similar utility program, "ASCII", allows conversion of records with fields having standard prefixes, (AU for author, TI for title, etc.) used by systems such as Dialog and InMagic, to ISIS format. These programs are often exchanged among ISIS users and sometimes inserted for convenience in the "CONVERT" directory supplied with the ISIS suite. Novice users, who are unfamiliar with the ISIS community, may be unaware of their existence, since they are not described in some Unesco-supplied manuals, being written in some cases by users themselves. At the other extreme, user interface software, and even entire integrated library systems have been written by ISIS users to supplement the Unesco software suite. These include SANJAY (Bhargava, 1993), and Eurisco (Nowicki, 1992).

Hugo Besemer, moderator of the ISIS Listserv, has provided technical information on a fairly regular basis through "User to User" columns appearing in the pages of the Quarterly Bulletin of the International Association of Agricultural Information Specialists (Besemer, 1989--). Hopkinson, (1989--), is the author of a similar column appearing from time to time in Information Development. The journal Program frequently carries ISIS related articles.

Rodriguez, (1995), presents a statistical analysis of ISIS use in Latin America and the Caribbean, while Frantao (1994) offers a similar analysis of African ISIS users. More recently, Smet (1995) presents an evaluation of a particular ISIS application in Belgium. In 1996, a group of students of Library and Information Studies at the University of Stuttgart reported on a study to obtain more information about the CDS/ISIS applications. (ftp://ftp.unistuttgart.de/pub/doc/isis/survey96.txt). Their findings indicated that "CDS/ISIS is mostly used on stand-alone PCs and in few cases on local area networks." The Stuttgart study also identified a growing interest in UNIX versions of ISIS. The current study suggests an increasing trend toward multi-user applications and toward other computing environments such as Windows and UNIX.

To our knowledge, there have been no studies of the CDS-ISIS mailing list itself, aimed at categorizing the types of questions posted to the list, and the role of email in providing assistance with technical or procedural questions posed by software users. We believe that the ISIS software and its users, along with the CDS-ISIS mailing list, provide a unique opportunity to study the pervasiveness and effectiveness of email in facilitating library automation projects across geographical and cultural barriers.

Rudy, writing in 1996, provides a comprehensive review of scholarly research into the use of electronic mail. The paper analyzes research studies according to a fourfold schema: media choice, media effects, email research methodologies, and neglected avenues of research (Rudy, 1996: 199).

While Rudy notes considerable progress in the formation of theories to model information richness, the notion of meaning in email communication, and the communication styles of users, he notes among the neglected areas a lack of material about the user's perspective of email utility. Our survey, coupled with our examination of the question and answer patterns evident in the list, are attempts to address this issue. (Rudy, 1996:208).

As well, Rudy notes that past studies of email had been unable to relate evidence derived from experiment to actual workplace contexts, since many studies involved students rather than bona fide managers and workers, and because their dependence upon statistics would seem to suggest that context-independent rules for media use could be discerned. Noting in particular that the investigation of discussion "threads" seemed to required more investigation, Rudy suggested using qualitative methods, such as network analysis and ethnography, in further research.

In this endeavour, Rudy suggests, approaches centering around the notion of "information richness" (cf.. Daft and Lengel, 1986), offers considerable promise. The information richness model subsumes notions of feedback, language variety, opportunities for personal expression, and the presence of multiple cues in a medium.

Electronic mail lacks some of the characteristics of a "rich" medium so defined, since it limits cues to written symbols, eliminating the possibilities for tone of voice, inflection, body language, gesture, and other symbols available in person to person communication. In the case of an international interactive email forum such as a Listserv, problems are exacerbated by the multiplicity of users' workplace contexts, by the possibility of their being unfamiliar with the

primary language of communication, and with the medium itself. In the case of the ISIS list, users may bring to bear various degrees of familiarity with the technical content of messages as well. Moreover, while English is the primary language of the list, French and Spanish are also considered acceptable, and messages have been posted in other languages too. For these reasons, the ISIS list brings to into sharp focus many problems identified by communications researchers.

Unesco's CDS/ISIS SOFTWARE AND ITS USERS

The current study examines the world-wide discussion list concerned with the suite of software known as the Computerized Documentation Service/Integrated Set of Information Systems (CDS-ISIS). The ISIS mailing list is especially appropriate for such a study, since the software was itself developed in an attempt to remedy global information inequalities.Produced by the United Nations Educational Cultural and Scientific Organization (Unesco), and developed in France, the Netherlands, Canada, and other nations, the ISIS software was envisaged as a cataloguing and documentation support system suitable for the automation of documentation centres and library technical services in developing nations. A mini-computer version of ISIS, known as MinISIS, has long been supported by the Canadian International Development Research Council (IDRC). A version suitable for operation on DOS-based microcomputers (Micro CDS/ISIS), is currently distributed free of charge to librarians and documentalists in developing nations, and is available for a handling-fee to individuals and not-for-profit institutions worldwide.

Micro CDS/ISIS version 1.0 was released in December 1985, almost exactly one year after the United States' withdrawal from the Unesco.. By 1996, more than 15.000 implementations in the world had been reported (gopher.bib.wau.nl: April 1996). Because Unesco chose to deliver and support Micro CDS/ISIS through third-party distributors in various countries, rather than through direct support provided by the developers in Paris or Amsterdam, information about new releases, upgrades, troubleshooting of existing systems, or the implementation and configuration of local systems has been problematic for end users. For instance, users in the United States and Canada may order ISIS from a distribution group located in Ottawa, Ontario, but would typically never meet programmers face to face, nor receive training directly from the development group. The problem may be exacerbated in more remote places, such as Sub-Saharan Africa, Pakistan, or other emerging areas of the globe. Moreover, Unesco's regional or national distribution teams, as well as other developers, have been encouraged to enhance, and to develop new interfaces for the core suite of programs.

ISIS uses the International Standards Organization's international data communications format ISO 2709 - the basic convention used for Machine Readable Cataloguing (MARC), facilitating import and export of records to and from other library automation systems. As well, Unesco provides with the software suite a modified version of the Pascal programming language, which allows users and local developers to implement new subsystems or to otherwise enhance the basic software suite at individual ISIS sites. While this strategy allows continued product growth and customization, it may result in confusion for new users -- it is sometimes difficult for librarians to tell whether the software utility they are using was created by Unesco or by a third party.

THE LIST

A fluctuating number of ISIS users -- about 500 world wide -- participate in a moderated electronic mailing list (Listserv), known as CDS-ISIS, which is currently administered from a site in the Netherlands. Like many electronic forums, the ISIS list was established with certain stated aims: "The aim of this discussion group is to provide a forum where users can ask questions and announce new developments and co-programs." (Besemer, CDS-ISIS posting: 20 June, 1996).

At the beginning of the current study, a total number of 369 subscribers was obtained from the Listserv's membership report (17 April 1996). By 20 June 1996, the number reported was 623. However, the numbers fluctuate in both directions. On 5 August, 1996, when the survey which accompanies this study was posted, there were 517 subscribers. At that time, subscribers hailed from approximately 60 countries and from every continent.

Because of its world-wide scope, its consequent multilingual nature, and its focus upon a specific suite of library automation software, the ISIS list would seem to provide an excellent testbed for an examination of the uses of the medium to exchange mission-critical reference information among a group of geographically disparate professionals. The aim of this exploratory study is to determine whether and how the electronic mailing list can obviate some of the problems posed by traditional documentation and customer support strategies.

AN EXPLORATORY STUDY

The current paper presents the results of 1) a content analysis of Listserv traffic downloaded and archived during the summer of 1996, and 2) the evidence of responses to a supporting survey sent to all known users during the same period. It categorizes user needs, and the topics discussed on the list, and attempts to assess usefulness of the medium in providing understandable and timely information and support.

While the list is demonstrably effective in addressing a range of social and technical concerns, our analysis of its traffic suggests that developing and developed countries seem fairly equally represented in the its user community. Listservs eliminate some of the costs of in- person meetings and distribution of printed documention, and allow for the timely global dispersion of information.It is recognized that consistent availability of telecommunications services may be problematic in particular areas of the developing nations. It is likewise understood that an underlying assumption is at work here -- the premise that Unesco's software is itself beneficial.

METHODOLOGY AND DATA COLLECTION

Between April and July 1996, all posted messages to the list (CDS-ISIS@HEARN.BITNET) were studied. Approximately 60 to 70 messages were captured and archived each fortnight during the capture period. List traffic was obtained in order to determine:

1. who was using the Listserv;
2. the types of questions being asked - as manifestations of users' levels of familiarity and facility with the ISIS software; and
3. the stylistic spects of list messages, as indications of users' levels of comfort with medium itself.

The purpose of the ancillary survey was to determine:

1. basic demographic information about the users' residence ;
2. setting in which CDS-ISIS is being used ;
3. reasons for referring to the list.

We were also interested in learning what problems users encounter during their information seeking experiences, and how they view themselves with respect to the global CDS-ISIS community.

Within hours of the survey's issue, the first answers were being received. Unfortunately, the flow soon ceased. By the end of the capture period, only 34 of the reported 517 recipients had replied. A second appeal yielded no further returns. While this number of responses is too small to be statistically relevant, the responses received were of assistance in checking the validity of assumptions made during the content content-analytical process, and shed additional light on the demographics of the list community. Of particular interest were sections which asked users to report their frequencies of use and reasons for subscribing to the list.

There may be several reasons for such a low response rate. Users may have lacked the technical understanding to file and complete the form and to return it electronically. Users may have felt inundated with surveys, and simply declined to fill out another questionnaire. It is also possible that the time of year was wrong. August is the high summer season in the northern hemisphere, and may simply be a bad time for conducting surveys, especially ones involving academics. Future research which includes electronic mail surveys should include instructions for filing and completing the forms. As well, questionnaires should be distributed at times of the year which take into account academic schedules for both hemispheres.

CONTENTS OF LIST MESSAGES -- GENERAL OBSERVATIONS

The first step in this analysis consisted of catgorizing archived messages from the list, and identifying various patterns of form and content. This analysis follows a form suggested by Bogdan and Biklen (1982), “interpreting, organizing and synthesizing data, searching for patterns, and discerning important features”. The intent was to understand how the Listserv functioned in its particular setting.

CDS-ISIS is a moderated list. Each message intended for the list is first posted to the list owner, who acts as an editor. Theoretically, this process ensures that only relevant information should reach the list, making it a highly application-focused and somewhat technical forum. In practice, the occasional off-topic posting reaches the list. However, the moderation appears to be largely effective.

Postings are accepted in three "official" languages, namely English, Spanish, and French. While Besemer suggests English and French as the languages of choice for the user group, (CDS-ISIS posting, 20 June 1996), Spanish is also used on ISIS related Web sites and in numerous CDS-ISIS postings. Nevertheless, users are not prevented from posting in other languages which can be represented using the ISO Latin character set.

However, both the postings and our survey suggest that many users do not realize that there are any "official languages" at all. Posters use the language(s) in which they feel most fluent, or copy the language used in the previous message in a conversation thread. In other words, attempts are made to answer posted questions using the languages of the posters, often with apologies for the respondents' lack of facility with them.

The most frequently used language is English, followed by Spanish, but postings are occasionally submitted in Portuguese, French, Polish, Dutch, or Italian. From this, we can discern some necessity for users to possess at least a basic knowledge of English or Spanish. While many users must find this a hindrance, it is also possible that the forced use of a common language might foster cohesion across the list. We were able to discern a high degree of tolerance for different levels of language facility among its users. Interestingly, regardless of the languages used in the message bodies, the subject lines of message headers were almost exclusively given in English.

MESSAGE STYLE AND NETIQUETTE

In a 1988 study, Roysdon and Elliott presented one of the first significant discussions of e-mail library reference services. They suggested that "electronic reference encourages a more thoughtful, leisured, and coherent approach to question answering." (Roysdon & Elliott, 1988). They also contended that email reference appeared to be more accurate than telephone reference since transcription errors could be avoided. These remarks apply in part to the forum under observation here. List users tend to adopt a casual, colloquial, but nonetheless businesslike style of address. This informality may serve the function of reducing potential communication barriers, since it contributes to a friendly atmosphere which may in turn reduce anxiety.

Although very moderate in their tone, almost all respondent tend to be outspoken. Misconceptions, technical errors, or omissions in messages are noted by respondents in conversation threads, politely but with certitude. Since such clarifications tend to be posted to all list members, more accurate information and clarifying comments are shared to the benefit of the Listserv community as a whole.

For example, during the study period, a particular user asked how to convert a "text" file into an "ASCII" file. This request drew several unequivocal and perhaps pointed replies, to the effect that "ASCII" and "text" files were one and the same. While this behavior could be construed as network "flaming", we suggest that it is exactly in cases like this that the list seems the appropriate forum. In this instance, a useful detail about the output of the system was furnished by a few technically sophisticated ISIS users, to the benefit of others less familiar with underlying computing concepts.

Compared to the list of subscribers, the number of active posters is small. That is, the number of unique personal names appearing in message headers is far less than the number of registered subscribers. In network parlance, registered users of a list who refrain from posting are termed "lurkers". Future research should be conducted to determine whether the ratio of "lurkers" to "active participants" holds true for similar technical lists.

Various reasons for this ratio suggest themselves. Users may feel reticent or embarrassed to post questions which they judge "mundane" or "too simple". However, given the friendly and supportive atmosphere of this particular list, we suggest that other mechanisms are more likely, since even questions posed by novices are seldom "flamed". Moreover, even seemingly trivial problems may have immediate workplace consequences, and the users of this list appear to recognize this fact. This concern may stem from the shared experience of list users, who are evidently dependent upon the software suite in their own work settings. What is more, active posters seem to be both highly competent and dedicated users and approachable individuals. Rarely was an answer to a question condescending or embarrassing in its wording.

From the user's perspective, convenience would seem to be a primary benefit of email and mailing lists. Questions can be asked as problems occur on the job, and rectified quickly provided that timely answers are received. As well, if questions and responses are retained or archived, users can build their own repositories of technical tips, embellishing software manuals with real workplace examples. We suggest that the ISIS list works well in this fashion.

CATEGORIES OF POSTINGS

As one might expect the majority of messages to refer to technical matters. From among the many individual postings, certain recurrent themes are evident. These tend to be esoteric and pragmatic - they relate to the specific software in use, and to immediate workplace concerns.

Many postings concern procedures of obtaining passwords and/or license numbers for particular CDS-ISIS products. Since distribution is conducted by independent regional representatives, there seems to be no standard policy for these procedures, or if there is, it is ambiguously communicated to the user community Many postings contained questions about software, releases, and about the most current versions of particular products.

Many users wish to know whether their version is the most current, or where information about new products is available. This prevalence suggests that product information reaches the users only sporadically. While there is at least one Website devoted to CDS-ISIS products, (http://unesco.org/ch-extern/db/micro-cds-isis/doc.html), users may either be unaware of the site, may not have reliable access to the Web, or may prefer personal question and answer as an information delivery mechanism. As well, the information available at this site may not address their specific questions. For instance, we found a long-outdated mailing address and phone number for one ISIS regional distributor on the Unesco site.

In only one posting did we observe a representative of a development team (at Unesco, Paris) discussing the features of a new version of ISIS, and the procedures for obtaining it. (CDS-ISIS posting, 15 March, 1996). Since this is certainly a "frequently asked question, we conclude that the list is being used to solicit information from other users to supplement scanty new product and distribution information from the developer and its regional representatives.

Many postings addressed specific differences between individual versions of the software, or versions designed to work on different computing platforms. For instance, in addition to the officially released versions written for Microsoft DOS platforms, versions written for the MS Windows environment and the UNIX family of operating systems also exits. Several users are unclear about the options available and the operational procedures that apply when ISIS is utilized on these different platforms.

With the increasing use of local area networks, many users seem to be grappling with technical conundrums ancillary to the ISIS software itself. These include networking and shared printing concerns, and problems posed by concurrent the use of ISIS databases at several workstations. How does one assign different user privileges within a network setting? How can one transfer data from ISIS to a word processor? What do particular error messages mean in various contexts?

Users occasionally sought advice about specific pieces of Pascal programming code. Although the majority of posters appear to use CDS-ISIS in "turnkey" fashion, as delivered at installation time, a major attraction of the ISIS suite lies in the fact that it comes with a Pascal compiler. ISIS can be tailored to the specific needs of an organization. This can be accomplished using a variety of built-in design and configuration options, and ultimately through the addition of locally written Pascal language programs.

ISIS has traditionally been distributed with no user-friendly "front end" or presentation layer. Until Release 3.0, the distribution disks did not include an installation program. As a result, the software has been conducive to an experimental or "hacker" approach. Based upon posting content, we conclude that this spirit still prevails. "Learning by doing" seems the encouraged approach by developers and users alike.

Ernesto Spinak, an active list member, an ISIS user and developer, and the author of a printed user manual for Release 3.07 suggests: "...there is no better way to learn to program than to write your own programs and the best way to start is by copying and analyzing somebody else's programs..." (Spinak, 1994). However, different users evidently experience unequal success with this method, and the list is one forum in which failures or unsuccessful approaches can be reported, in the hope of obtaining assistance. Occasionally, entire Pascal programs are posted. In this, we find a clear demonstration of expertise sharing and cooperation among list members.

Questions were frequently detailed, stating the nature and workplace contexts of problems or projects, sometimes including exact error messages, sample field definitions, or comments of code proposed as solutions. Many of these factual questions seemed not to require instant answers, however, the average response time lies between two and four days. As well, subscribers with the same, or similar problems, replied to questions stating that they had encountered similar symptoms and were also hoping for support. Postings tend to be concise, yet explanatory enough to supply background information in addition to a problem. This detail promotes understanding by parties who may not have experienced similar situations.

Another issue then current, related to ISIS databases on the Internet. Users wondered exactly how such databases should be interfaced, and frequently asked for the addresses of ISIS Websites. A year after the initial study, we find that these queries can claim the status of “frequently asked questions” (FAQs).

While ISIS development is a Unesco project, the agency does not support the product in the field directly, but rather, makes use of third party distributors. Consequently, in recent years, we have seen the foundation of a series of user groups whose major idea is to enable, develop, and support the exchange of information and experience among ISIS users from different parts of the world, both globally and regionally. Examples would be Thailand, Malaysia, Poland, and Belgium, to name just a few. Some list discussions question Unesco’s role with relation to the user community.

The Listserv is used to announce regional training sessions, providing pertinent information for potential participants, and to inform users about current ISIS products. The development of additional programs, (sometimes referred to as “co-programs”), based on Unesco’s software, plays an important role. Ernesto Spinak, from the Pan-American Health Organization, for example, has developed more than 150 programs, which are available free of charge. In addition there are commercial programs, e.g. UKER, from Lorre Engineering, which supports the conversion, maintenance, and import/export of data, or, the Electronic Library Manager (ELM), a fully integrated database that can be tailored to fit various library requirements. (Further information is available from Emil MonKman in Slovakia: monKman@center.fmph.uniba.sk). As noted earlier, the majority of users do not develop their own customized software, but use the version provided by Unesco.

ISIS terminology is itself a frequent topic of discussion. Many users seem not to find sufficient explanations for acronyms or programming vocabulary in the reference manuals, and appeal to the list for clarification. The glossary in the 2.3 manual, for example, is not very elaborate, and the same seems true of more recent versions. With the migration of ISIS from the Microsoft DOS to the UNIX, Linux, Windows NT, and similar environments, terms related more properly to these operating systems must also be acquired and digested.

It would be difficult for Unesco itself to provide a glossary which anticipated terms originating elsewhere in the industry. However, to a user in trouble, the origin of a term is moot. On March 22, 96 a user posted this plea:

"HELP. When I joined the CDS-ISIS group a year ago, the discussion (mostly) made sense to me. Everybody were discussing CDS-ISIS problems. But lately the messages seem to be more and more cryptic. What is NetBSD (with SCO System V compatibility)? Am I missing something essential?"

USER CATEGORIES

It might be useful to place list users into three broad categories, based upon content and stylistic aspects of list postings, viz:

1. Newcomers. Often self-identified, these subscribers are new to the ISIS software and tend to seek general assistance.

2. Long term users. With at least a year’s experience, these users are interested in performing adjustments or enhancements to the system, or express specific concerns about the system’s functions or its internals. Such users tend to treat the list as an aid in “troubleshooting”, or as a quick, effective source for specific information.

3. Experts. With several years’ experience, these highly knowledgeable users frequently provide the answers to all kinds of questions posted to the list. Moreover, their names appear to be familiar to many list users. They appear to be interested in long-term or technically esoteric topics, such as ways to improve cooperation between developers and users, in developing new user interface software, or in “brainstorming” with other experts over the list

SURVEY RESULTS

The preceding discussion derives from analysis of list contents. At this point, we present some of the survey findings by way of corroboration. As of 12 August, 1996, 517 people from 54 countries subscribed to the CDS/ISIS list. All of these subscribers were sent copies of the survey. The survey consisted of multiple choice questions, and participants were encouraged to check all applicable answers, or to supply other information as necessary. Respondents were asked to identify their institution types, the primary uses of the software, their length of time on the list, their frequency of posting, and their reasons for posting to or for reading the list. Additional questions sought to determine respondents' degrees of technical expertise, and their estimates of the effectiveness of the list as a support mechanism.

In all, 34 subscribers (8% of the known population) from 21 countries (including 18 developing nations), responded to the questionnaire. Interestingly, many survey respondents were familiar to us, since they were among the frequent users of the list. However, respondents' self reported frequency of contribution to the list was not commensurately high. Several users reported using a language other than their national tongues in postings. All respondents had been using ISIS for more than one year, and most for more than two years. Six respondents identified themselves as ISIS distributors.

Respondents' workplace settings varied widely. They included independent research institutions, public libraries, and records management operations. Respondents were asked to explain their choice of ISIS over another system. A factor frequently mentioned was "reasonable cost". This is somewhat misleading, since ISIS is distributed free of charge in developing nations.

The survey returns suggest that this list is mainly perceived as a tool for obtaining quick reference answers to technical questions, and information about current developments within the ISIS community. While few list users reported that they relied entirely upon CDS-ISIS for technical support, thirteen out of 34 respondents claimed never to have consulted with Unesco directly, and only 1 user reported being satisfied with Unesco's user support.

As well, several respondents provided additional prose comments, which were themselves highly suggestive. Some saw North American unfamiliarity with ISIS, and the consequent paucity of articles about ISIS in the professional literature, as detrimental. Others mentioned a desire for more information about integrating ISIS with other systems and packages.

Why do people use the ISIS mailing list? Respondents implied that the rapidity and accuracy of responses to questions are factors that motivate use. It seems evident that these are interrelated phenomena. As noted earlier, immediacy also entails being able to ask a question as soon as it occurs, rather than having to wait for an "appropriate moment" or the for the next user group meeting. As well, while not allowing complete anonymity, email makes it easier to ask a seemingly trivial question with minimal embarrassment.

The assumptions about why users posted to and read the list is borne out by the responses to the survey. Those who responded to the survey tended to be long term ISIS users. Of the 34 respondents, 26 stated that they had been using the ISIS software for more than two years, while 8 more reported having run ISIS for between 1 and 2 years. No one reported having used the software less than a year. Not surprisingly, while users reported working with several versions of ISIS, the MS DOS product was indicated in 20 responses. At the time of this survey, UNIX versions were still considered to be unofficial, and MS Windows versions, while in use in some countries, had not been distributed world wide.

Perhaps most tellingly, 24 respondents reported having designed new ISIS databases, and 12 had written programs in the version of Pascal distributed with the software. Asked whether the manuals distributed with the product were adequate, 25 respondents ranked them moderately useful, compared with four who gave the documentation "very useful" ratings. Since 23 users indicated flexibility and the ability to design and develop formats and programs as a major reason for choosing ISIS, and 21 reported receiving no technical support from Unesco's regional distributors, we can conclude that respondents in this sample were for the most part sophisticated users, capable of modifying and enhancing the existing software. This supposition is confirmed by the fact that 21 users considered the list a time saver, but not essential to their use of the software. In short, while this group appreciates technical manuals, it seems to be composed largely of people who are comfortable with experiment as a means of acquiring knowledge.

One third of the survey's respondents had used the list for between 1 and 6 months, while 10 had been subscribers for longer than 2 years. Most users check the list for less than 30 minutes at a time, and at least twice a week. A substantial number (19) reported reading it daily.

Responses to questions about users' reasons for posting to the list and for reading its contents support our contentions that email can contribute to the dissemination of technical and professional information otherwise available, and can contribute to a sense of community and concern. Nearly as many people reported a desire to share knowledge with others as they did the desire to receive technical tips.

We believe that this Listserv provides a good example of grass-roots mutual support -- a third of the respondents had never consulted Unesco for technical support, whereas requests for new product information and for help with specific technical issues were the most frequently cited reasons for subscribing to the list.

While the low number of responses rendered the survey statistically inconclusive, these responses corroborate some of the inferences drawn from the content analysis of list traffic, and point the way to future studies. In particular, we believe that it would be useful to investigate the relationships among technical competence, list reading, and list posting. A more methodical survey could be constructed to compare the frequency and duration of email interactions with software experience, controlling for geographic dispersion and the availability of other means of support.

It should be noted that any quantitative evaluation of Listservs would have to resolve various sampling problems. First, it is difficult to determine the population of list users because the number of subscribers to a list may fluctuate widely over a period of weeks or months, and secondly, because many Bitnet lists are gated to Usenet News, and are thereby accessible to a far greater (and indeterminate) number of casual readers who can select and deselect newsgroups at will, and for whom no adequate method of accounting is available. For these reasons, we advocate downloading current lists of subscribers, and mailing surveys to random samples, for instance, to one in ten users. Attempts to control for geographic dispersion might be hindered in such instances, by the fact that upper level Internet domain names (com, org, gov, ca for Canada, de for Deutschland, etc.) are used inconsistently around the Net.

DISCUSSION

It seems evident that many questions are posted with two aims. Users hope to obtain quick solutions to problems at hand, and to share information with others in the user community. By nature, factual “ready reference” questions require comparatively little “question negotiation”, and seem most suited for email correspondence. However, because users may seek clarification through followup postings, it is also possible to pose lengthier or more esoteric research questions, and to obtain responses from other users. When this happens, numerous messages are posted, either among a few subscribers who try to clarify questions, by means of an email discussion “thread”, or by many respondents who post individual replies in an uncoordinated fashion, each basing his response upon his own interpretation of the question.

The 1995 Stuttgart survey identified concern over UNIX versions of the software, and suggested the formation of a special UNIX/ISIS interest group. However, we suggest that the use of a common list fosters community among different users, and allows them to extrapolate from problems resolved in different settings as well. An alternative to establishing a separate list would be to encourage the use of more consistent subject lines in message headers. The subject lines used in the list do not always allow for a clear assessment of message contents. The adoption of conventions, such as indicating the software release of platform in the headers, might allow users to select messages more readily, while still allowing them to browse.

Another option would be to archive list messages on separate topics. This could be done either by individual users, or at some central repository. Increased voluntary standardization of message headers, especially subject lines, could facilitate and perhaps help automate this activity -- but the attendant problem of making newcomers aware of the conventions then presents itself. We are aware that some Internet and Web search engines, (e.g. DejaNews, or www.liszt.com), offer full text searching and retrieval, however, we believe that better indexing could reduce dependence upon such machine-intensive processes.

This study focuses on examining particular patterns in the email traffic of the CDS-ISIS list. In the course of the study, it became clear that the list facilitated the sharing of ideas, discoveries, problem descriptions, and problem solutions, among ISIS users. It provides a vehicle for spreading knowledge about different ISIS products, and about the specific settings in which these are being used. Moreover, the list is self-organizing. Its active members determine the topics and conventions of the list, and those users come from countries on the “periphery” of modernity, the “less developed” nations, as well as from the technological “centre” in the North or West.

SUGGESTIONS FOR FURTHER RESEARCH

Librarians and other scholars and practitioners concerned with world development have much to learn by examining the use of specialized mailing lists. Electronic reference may well become a key component of our profession's vision for coming digital age. This list represents a different, but related phenomenon – peer counseling by members of a community of interest. In our consideration of email as a medium for the provision of professional reference and research services, we would do well to examine how members of such communities are already making use of the medium.

It may be possible for us to anticipate user needs through judiciously targeted qualitative and quantitative examination of email lists and newsgroup traffic. However, though is technically feasibile to conduct statistical and content analyses of archived network traffic, we suggest that several comparative analyses would be necessary to obviate the problems identified by Rudy (1996). We also suggest that comparative studies of the lists or newsgroups of other user communities be undertaken. This particular list is an exemplar of cooperation, politeness, and willingness to help one’s peers. Since ISIS is primarily used by not-for-profit or non-governmental organizations, by regional developers, or by librarians in government institutions, the list’s members may already share a tendency toward cooperation. How true would this be for lists or newsgroups maintained by user groups whose members were not concerned with development issues?

Although we know how this particular list is being used, we know comparatively little about the people who post the questions. The current study provides some insight into their information seeking patterns, and how they relate to pragmatic information needs. It may be seen as a first step toward obtaining a picture of the software applications themselves, of the success of a particular user-run software support effort, and of role of email as a medium for such activities. A research agenda building upon these themes could contribute to Unesco's and IFLA's overall project of rectifying the information disparities between the developed and developing nations.

References

Besemer, H. (1991--)"CDS/ISIS user to user". Quarterly Bulletin of the International Association of Agricultural Information Specialists 36 (3): 193-5; 37(3): 164-6; 37(4): 227-8; 38(1): 34; 38(2-3): 162-3; etc.

Besemer, H. (1996).CDS/ISIS listserv posting, 06/20/96.

Bhargava, J.K.; Srivastava, R.K.; Murthy, S.S. (1993). "SANJAY: an Indian library automation package based on CDS/ISIS".

Bogdan, R., & Biklen, S. K. (1982). Qualitative research for education: An introduction to theory and methods. Boston: Greenwood. (p.145)

Bokos, G.D. (1993). "UNIMARC, CDS/ISIS and conversion of records in the National Library of Greece". Program 27:135-48.

Brown-Syed, C. (1996). From CLANN to UNILINC: an automated library consortium from a soft systems perspective. [Thesis, Ph.D.]. University of Toronto Faculty of Information Studies.

Buxton; Hopkinson; Radwanski, Aleksander. (1995). “The CDS/ISIS handbook” Program. 29(3): 356.

Chaudhry, A. A. (1989).“File inversion in CDS/ISIS; Fastsort helps to do it faster “.Program 23: 295-305.

Chaudhry, A. A. (1989).“Interfacing CDS/ISIS with WordStar; ICIMOD's experience.” Program 23: 65-72 July 1989

Chaudhry, A. A. (1990).“ Adaptinc CDS/ISIS to index hyphenated and numeric words and incorporating case sensitive storage and retrieval”. Program 24: 371-9.

Chowdhury, G.G.; Neelameghan, A.; Chowdhury, S. (1995). "VOCON: vocabulary control online in MicroISIS databases." Knowledge Organization 22(10: 18-22.

Daft, R.L.; Lengel, R.H. (1986). "Organizational information requirements, media richness, and structural design". Management Science 32: 554-71.

Deshmukh, S. (1993). “Book Acquisition System using CDS/ISIS Version 2.3.” Library science with a slant to documentation. 30(4): 160.

Frantao, S. (1994). "L'utilisation de CDS/ISIS dans les réseaux d'information: les perspectives pour l'Afrique." Quarterly Bulletin of the International Association of Agricultural Information Specialists 39(1-2): 85-7.

Gallina, P. L. and others. (1989).“Micro CDS/ISIS for scientific and technical documentation”. Canadian Library Journal 46: 35-7 February 1989

Gimilio, R. (1988). “Le logiciel CDS/ISIS/UNESCO version mini-micro”. Documentaliste 25:148-52

Goyal, R.K.; Kumar, S. (1990). "dBASE to CDS/ISIS: a program to convert data from dBASE/FoxBASE to CDS/ISIS". Program 24: 181-6.

Hopkinson, A. (1989--). “CDS/ISIS information”. Information Development 5:135-7, 5:200-3, 6:4-7... [July 1989]...

Hopkinson, A. (1990).“A MARC database on your desk with CDS/ISIS”. Cat Index 96: 5-7.

Hopkinson, A. (1996). “CDS/ISIS Information”. Information development. 12(4):203.

Krol, E. (1994). The whole Internet: user's guide and catalog. Sebastopol, CA: O'Reilly & Associates.

Maiido, H. (1992).“ Use of CDS/ISIS for a survey of computer systems in medical libraries”. Program 27: 183-9 .

Marchi, %D .; Kumar, S. (1989).“The documentation centre for Polimoda” Art Libraries 18(4)4: 28-9 1989. [Presented at the 1989 IFLA conference, Paris].

Naber, G. (1996). “Drain: a CDS/ISIS library management system”. Information Development. 12(1): 26.

Nieuwenhuysen, P. (1991).“Computerised storage and retrieval of structured text information: CDS/ISIS Version 2.3”. Program 25: 1-18.

Nowicki, Z. (1992). "First meeting of CDS/ISIS distributors in Europe and the Mediterranean area, Viareggio, Italy, 16-18 March, 1992". Electronic Library 10: 179-80.

Plassard, M.-F. (ed.). (1994). UNIMARC & CDS/ISIS. New Providence : K. G. Saur.

Pool, I. de Sola. (1983). Technologies of freedom. Cambridge, MA: Belknap Press.

Preston, W.; Herman, E.S.; Schiller, H.I. (1989). Hope and folly: the United States and Unesco, 1945-1985. Minneapolis, MN: University of Minnesota Press.

Puntodewo, A. (1991).“Creation of a map catalogue using CDS/ISIS [at the University of New South Wales]”. LASIE 21: 88-105 January/February 1991

Ravi, A. S. and others. (1992). “Production of a union catalogue using CDS/ISIS and Ventura: some experiences in India [DTP catalog of conference proceedings in Indian scientific and technical libraries]”. Program 26: 183-90.

Rodriguez, K. (1995). "CDS/ISIS: a statistical analysis of usage in Latin America and the Caribbean." International Information & Library Review 27: 225-35.

Roysdon, C.; Elliott, L. (1988). "Electronic integration of library services through a campuswide network". RQ 28: 82-93.

Rudy, I. A. (1996). “A critical review of research on electronic mail”. European journal of information systems : an official journal of the Operational Research Society. 4(4): 198.

Santiago, S. (1992). "OSIRIS, a microcomputer-based online serials information registration and inquiry system". IFLA Journal 18(4): 333-8.

Shailendra, K. and others. (1989).“Use of CDS/ISIS for creating India's database of serials [at the Indian National Scientific Documentation Centre]”. Program 23: 287-94.

Smet, E. de. (1995). "Evaluation of a computerised community information system through transaction analysis and suer survey." Libri 45:36-44.

Smith, A. (1980). The geopolitics of information: how Western culture dominates the world. London: New York: Oxford University Press.

Spinak, E. (1994). Manual of Programming CDS/ISIS Pascal: (Isispas): Mini-micro CDS/ISIS (version 3.07), Madrid. [from: CDS-ISIS listserv/sender: E Spinak /date: 06/2/96 /subject: Manual de Programacion ISISPascal]

Sreenivasa, R.A.; Hariharan, A.; Sadananda Rao, B. (1992). "Production of a union catalogue using CDS/ISIS and Ventura: some experiences in India". Program 26:183-90.

Stuttgart University Survey (1996).URL: ftp://ftp.uni-stuttgart.de/pub/doc/isis/survey96.txt

Treloar, A. (1991). "Plant facts by example: a novel application of CDS/ISIS for accessing information about medicinal plants." LASIE 22: 28-40.

Unesco (1984). Many voices, one world: Communication and society, today and tomorrow: the MacBride Report. Paris: Unesco.

Unesco (1989). Mini-Micro CDS/ISIS reference manual. Paris: Unesco. Office of Information Programmes and Services. Division of Software Development and Applications.

Wijetunge, P. (1996). "An experiment to edit spelling mistakes of CDS/ISIS records online at the University of Colombo, Sri Lanka". Program 30: 51-9.

Wongkoltoot, P.; Indee, S. (1992). "Mini-Micro CDS/ISIS in the Thailand Development Research Institute Library". Information Development 8: 147-58.

Listserv® is a registered trademark of L-Soft. Copyright © Christopher Brown-Syed 1995-2001. Disclaimers.