How Discussion Lists are Transformed into Networks:
The H-NET, Humanities OnLine Experience
David Halsted and Melanie Shell, Michigan State University
The following section was written by Melanie Shell
H-Net, Humanities OnLine originally began as a collection of humanities discussion lists and has grown to encompass a large book and software review project, extensive WWW site and teaching center in addition to acting as the hub of a network of over 80 discussion lists which span a variety of disciplines and topics in the humanities and social sciences. Based on the principles of cooperation and a firm guiding belief that new opportunities for scholarship and teaching can stem from the rapidly developing computer and electronic technologies, H-Net has striven to maintain high scholarly and pedagogical standards while also remaining on the forefront of how these technologies are developing, coupled with a realistic assessment of what resources humanities and social science scholars, in a variety of disciplines and all around the world have at their disposal.
Motivating these endeavors has been the firm belief that by increasing the ability for scholars to communicate with each other, H-Net will increasingly democratize the way scholars communicate with each other. This has been accomplished as scholars within the United States and Europe are able to participate on a daily basis in scholarly discourse with international scholars in areas such as Africa, Asia and Latin America, who have long remained on the periphery of Anglo-European scholarship. By the same token, scholars who may have found themselves working isolation from others in their area or fields are able to use these discussion lists as a means of breaking their isolation and connecting with scholars from around the world. Likewise, these discussion lists have democratized scholarly discourse, by allowing university professors of varying levels, students, school teachers, journalists, librarians and policy makers to participate in what has been referred to as an "on-going and multi-subject seminar."
Still, maintaining this "democracy" while also maintaining pedagogical standards has required striking a careful balance, and so far H-Net has managed to walk the line between democracy and "the academy" quite successfully. H-Net Discussion Lists are run in a way that is comparable to print journals. Each of the discussion list editors is a professional scholar, teacher and/or librarian who works an average of 2 hours a day, editing, organizing and distributing the daily publications of his or her list. While the internet is a more immediate medium than print, and therefore carries a greater element of immediacy, the same high set of standards that is applied in reputable scholarly journals is applied to the post, essays and book reviews that are published on the H-Net lists. In the case of the book reviews, the editorial process is even more like that of print journals, but with a much faster turn-around time.
H-Net editors also determine the content of the webpages, to ensure that the material on the pages is valuable for teaching or research. Building on the strength of its discussion lists, the H-Net web pages contain syllabi, textbook and software reviews, discussions of scholarly and pedagogical issues, a database of academic events, updated job listings and more. Organized according to the same schematic as the discussion lists, the web pages include sites devoted to each of the discussion lists, as well a good deal of other information that is archived on these pages.
As most who have ever used the World Wide Web are aware, while there is a huge range of information to be found, much of it is shallow, chaotic or of little value. Too often links have changed and bring up dead ends, or the information itself is far too difficult to navigate and requires more time searching and finding little, that yielding useful information. Even more difficult is finding exactly what you are looking for, necessitating and strong and precise search engine as well as serious thought about how to organize the information in the website and to maintain it. Continual upgrading of hardware and storage capapcity as well as checking and rechecking of links and search engines are all necessary for creating a truly useful web resource.
Thanks to H-Net's editors as well as the technical staff at Michigan State University who work to ensure that the site is as useable and accessible as possible, H-Net continues to balance these concerns with much suceess. While new software developments have made tagging information much easier, designing the scholarly website itself that is both fast and useable is far more complicated and requires the committment of individuals who have an active interest and strong background in both technical and scholarly skills.
Because H-Net has remained sensitive to and committed to maintaining both an international membership and breaking the isolation of many scholars, this means that there are a number of accessibility issues that must be addressed. The reality that many academics are not online that that the costs of startup, as well as the requirement of good phone lines or ethernet connections, as well as machines that can support internet communications is an important factor that we have had to consider when designing the web site. To a certain extent, we can only address some of these concerns, as H-Net remains internet based, although we have worked (and are working) to train scholars on what to do with the hardware once they get it and also, by creating an online resource that is so valuable to scholars in the humanities and social sciences, promoting and furthering the belief that this is a resource scholars should and need to have access to. As many of you are aware, communicating via e-mail requires relatively minimal hardware support. One does not have to have a large machine with large memory capability or even a very speedy connection to be able to send or receive a piece of electronic mail. The Web, however, is a bit more complicated and requires a speedier connection and more memory in order to be able to support the software necessary to view the WWW itself. Those individuals who have access to Lynx, or other non-graphical browers can view textual information on the web, which loads quickly and can be used even with a modem connection of 9600 or higher. (It can also be done with even slower modems, although speed there decreases accordingly.) Still, to take true advantage of the multimedia capabilities of the web, more hardware and, ideally, an ethernet connection, is required. It is reality, however, that these resources are not available to all, an indeed a very large number, of scholars. H-Net has strived to make itself into a resource that remains cutting edge, then, while also meeting the needs of the majority of scholars who do not have "big, fancy machines." -- another important balance.
Further, in terms of content, the WWW simply contains greater capabilities than e-mail. As the WWW pages are heavily influenced by the discussion lists with editorial control falling largely into the hands of the H-Net Discussion List editors, this has remained an important concern. WWW pages allow graphics and multimedia which e-mail cannot support. Likewise, the WWW can allow searches, links and is based on hypertext -- in a sense, "live" text. While the accessibility and indeed very character of the WWW is fundamentally different than the discussion lists, it seems fair to derive then, that the nature of exchange on the discussion lists is also fundamentally different than that on the WWW.
Scholars use discussion lists to exchange information and research ideas and questions, develop teaching and research strategies, as well as reviewing current literature, and more. In this sense, the discussion lists could be aptly compared to scholarly conferences, which allow scholars to directly network with others in similar fields and research areas. The WWW, in a sense, is a more passive means of communication while at the same time, is a better archival source as information can be both searchable, maintained over an indefinite length of time, and can arrive in multimedia rather than purely textual form. To a certain extent, the WWW can be described as a combination of conference and journal. The information is collected and published in an archivable and ordered form, while remaining posted in a central location that scholars can refer back to, post comments on, and the like. Because it is a searchable means, and carries the benefits of multimedia, the WWW carries much potential as a teaching and research tool that e-mail, by its very nature, cannot. The WWW also requires the information be ordered in a way that is unlike any other medium.
The WWW consists of what is known as "hypertext," defined broadly as "a means of allowing widely differing material to coexist in a compter system [where] access is controlled by creating networks, links and branches, recognizing the spatial multidimensionality of written materials, their manifold interconnectedness." For puposes of this discussion, hypertext centralized the importance of audience and what purpose this information is to serve. And the answer to this question determines its most basic design.
The nature of hypertext and the history of H-Net itself have important implications for the design of the H-Net Web site. H-Net has grown and expanded from a network of discussion lists to an even larger collection of discussion lists as well as a large Website, while remaining committed to pioneering new technology to aid both scholarly research and teaching, we began to wonder if the "clientele" which we serve through the discussion lists are indeed the pool from which the majority of our web users are coming. In other words, is it list subscribers who predominantly use our web pages as well, or are the web users a different group of people altogether? And what does this imply for the sociology of academic information exchange in the internet as embodied in the H-Net project?
III. Web Site and E-Mail Lists: Two Different Audiences?
The relationship between H-Net's academic e-mail lists and its Web site is still evolving. The transition from e-mail list to Web is a transition from a two-dimensional world of textual discourse to a multidimensional hypertext environment. E-mail discussions are highly interactive, while Web pages are not necessarily interactive. Even where interactive Web pages have been put in place, they operate on a fundamentally different set of principles from a query for information sent out on an e-mail list. E-mail queries operate much as other traditional means of verbal information-seeking operate; the value and richness of responses is limited only by the imagination and knowledge of the group of people on a given list or set of lists. A Web query is limited by the range and scope of the data available, and its usefulness is limited by the skill of the user in formulating queries, by the ingenuity of the designer or designers of the page and by the richness of the dataset with which the page interacts.
The H-Net Web site represents a very large and complex effort to translate
the interactive discourse of H-Net's e-mail lists to the Web's multidimensional
world. In contrast to the H-Net e-mail lists, the H-Net Web site serves a number
of people who have not decided to join a discussion on a specific topic. In
order to find out what H-Net Web users and e-mail list participants expect from
the H-Net Web site, I designed a survey form for the Web and sent out a nearly
identical form to several of the H-Net lists. As of this writing, more than one
hundred individuals have responded to the two forms. The purpose of the survey
was to determine the extent to which the H-Net e-mail lists and the H-Net Web
site have two different audiences, to describe the character of those audiences
with some precision and to determine what kinds of expectations each audience
brings to our pages. The results, which I present here, will be used to help
guide the further development of the H-Net site. I believe that other academic
Web projects will encounter a similar range of audience and expectations, and
should take this into account during the site design process. Some of the
conclusions of this survey and others like it may even have implications for the
policies universities pursue with regard to the relationship between electronic
text and library collections, or the best ways to use the Internet to improve
IV. The Survey and its Results
The survey presented to the H-Net lists and to Web users consisted of 13 questions with several response options each. It also asked participants if they wished to be included in a monthly e-mail update of changes to the Web site and solicited general comments. The results of this survey were translated using a Perl script into text files which were then read into Microsoft Access and Excel files for analysis. The overall results of the survey can be summarized as follows:
In addition to these general conclusions, of course, there are specific
results about individual resources on our pages which will be very useful for us
as we deploy our labor force and creative energies in the near
V. The Two Audiences
In order to measure the difference between H-Net's e-mail list audience and the Web audience, we asked Web respondents to identify themselves as H-Net subscribers or non-subscribers. We also asked e-mail respondents to indicate whether they use the H-Net Web site or not. The results are summarized in the pie charts below.
|Fig. 1. H-Net List subscribers and non- subscribers in the Web sample (78 respondents)||Fig. 2. Web site users and non-users in the e-mail sample (24 respondents)|
The results demonstrate the difference between H-Net's two audiences and reveal something about the nature of response to the surveys. The H-Net Web site audience-or at least the part of it willing to answer our survey-consists mostly of individuals who do not subscribe to H-Net lists. On the other hand, the responses we received to the e-mail version of the survey reflect the expectations of that portion of the e-mail list subscriber base who use our Web site at least occasionally.
Our Web design must therefore take into account the expectations of two
rather different groups of users. Each H-Net page clearly must be regarded as a
potential publicity instrument for the H-Net subject lists themselves. For H-Net
e-mail lists, the Web site is a potentially powerful recruiting instrument,
helping the lists reach a broader audience in addition to providing topically
interesting information for list subscribers.
VI. The Character of the Audiences-Professional Status
In addition to their different status with respect to the H-Net lists
themselves, the two audiences show a very different professional status profile.
The original survey included six different categories-Professor/Teacher,
Student, Independent Scholar, Computer Professional, Librarian, and Other. Of
these, only Professor/Teacher, Student, and Independent Scholar received
significant responses ( more than 10%) on one or both surveys.
|Fig. 3. Professional Status of E-Mail
|Fig. 4. Professional Status of
Although there were six response options for our question, only these three received significant response on on or both surveys. This confirms that the Web audience, like the e-mail list subscriber base, comes overwhelmingly from the world of education; the respondents' lack of interest in materials aimed at high school students and teachers, which will discuss later, indicates the predominance of college and university faculty, staff and students. While the respondents' pool from the e-mail survey consisted primarily of professors and teachers, a majority of the respondents from the Web sample are still students. This conforms to the overall rule that Web users are relatively young, and reinforces the point that the Web site can bring new subscribers into the H-Net family.
The professional profile of academic Web users has direct implications for academic Web sites and their future. Even if, to some degree, the Web is a marginal phenomenon in the "core culture" of history and social science departments today, inasmuch as that core culture still revolves around printed material and paper publication, the situation is likely to change relatively quickly. As today's students become tomorrow's young faculty, it will be expected that the Web will be a useful scholarly resource. The patterns, standards and models created by this first generation of creators of academic Web resources will inevitably have a profound effect on the way the next generation perceives the potential and limitations of the Web as a venue for academic communication and investigation. It is the obligation of this generation to provide standards that are workable and flexible, to establish a high level of quality for academic Web work and to demonstrate that computing has the potential to move both teaching and research forward in ways that will benefit traditional academic culture.
Indeed, as a new, increasingly Web-literate generation comes into its own, the separate cultures of the Web and of the traditional academy are likely to converge. The Web culture will change as collaborative institutions like H-Net replace the "brilliant loner" model-the skilled, inspired but overworked professor who establishes a site in his or her field or fields of interest, because quality Web work requires collaboration and a high degree of competence in several fields at once. To this extent, the culture that will build the Web will be different from the academic research model of today, which, in the Humanities at least, has traditionally emhasized single-author work.
As might be expected, our different audiences also use different resources. Our survey provided for seven possible responses, of which only four were operational at the time the survey was taken: the individual sites associated with H-Net lists, the H-Net Job Guide, and our Links and Events databases. Distribution of results here reflect the different interests of the e-mail and Web audiences:
|Fig. 5. H-Net Web resources used (e-mail sample)||Fig. 6. H-Net Web resources used (Web sample)|
The two groups expressed a roughly similar pattern of interest in
currently-existing H-Net resources. To be sure, the e-mail audience was clearer
in its response; a full 77% of e-mail respondents said they used the H-Net list
sites, compared to 46% of Web respondents. A roughly similar fraction of e-mail
respondents and Web respondents used the H-Net Job Guide (38% of e-mail
respondents, 33% of Web respondents). The primary difference comes in the
relationship between the use of the Events and Links resources; more e-mail
respondents were interested in links, while more Web respondents were interested
in events. I think the difference can be explained readily enough by the
difference in professional status referred to above. The younger Web audience
can use the Web to replace or supplement the informal networks more experienced
scholars have developed in other media and by other means. By the same token,
the younger users already know how to look around for links in their areas of
interest, while the older users want pages like those H-Net provides to help
introduce them to the world of the Web. Each audience looks to the Web site to
help supplement currently available resources in areas where it feels a need for
VIII. Suggested Resources
In some ways, the Suggested Resources category is the most crucial of all, because it provides information about what people expect to find on a site like H-Net's and lets us know where our audiences want us to go in the future. Here again we see significant differences between the Web and e-mail audiences within a generally similar response pattern. For representations of the data, see Figures 7 and 8 (attached). The most important difference comes in the expressed desire for teaching materials. The Web audience was barely interested in finding teaching materials (12%), while a full two-thirds of the e-mail audience indicated that they would like to see more teaching materials on our site. This is consistent with a younger Web audience. It probably also reflects the character of the lists to which the surveys were sent. Teaching material was the most sought-after category of resource by e-mail respondents, while professional material took the lead for Web respondents-probably indicating that both groups are most interested in finding answers to their most pressing professional needs. In both cases, primary research material came in a close second, followed by secondary material and, for the e-mail respondents, professional material. About a quarter of e-mail correspondents wanted more information about H-Net list discussions; about a fifth of Web respondents did. Surprisingly few Web respondents were interested in multimedia files, while nearly a third of e-mail respondents were. This probably indicates that users view multimedia primarily as a teaching tool, corresponding to the Web users' lack of interest in teaching materials in general. Consistent for both groups was a marked disinterest in material for a larger public-for high school students and teachers and for the general public. It appears from this survey that both H-Net subscribers and non-subscribers come to the H-Net Web site for information relevant to those active in higher education, not for material relevant to K-12 or "public humanities." We may surmise that the vast majority of respondents in the Professor/Teacher category are employed at the post-secondary level, and that H-Net's primary audience for both e-mail and Web offerings comes from the world of higher education. This is consistent with what we know about our list subscribers.
The implications of these results for the future of the H-Net site-and for
academic Web sites in general-are complex. The interests users bring to the
sites cover the range of interest categories we see in the academy more
generally. Teaching, research, and professional concerns all vie for first
place, and it is probable that a graduate student for whom professional material
is the uppermost concern today will be most interested in teaching materials a
few years hence. One of the curious things about the prominence of the primary
research materials category in both samples is that there is hardly any primary
research material on the site as it stands. Depending on the section of the
site, there are some useful resources in secondary research materials broadly
defined-bibliographies, tables of contents from recent journals, working papers,
and the like-and there are some good teaching resources (though never enough).
The interest in primary research materials is one area in which the current
state of the site and the expectations of its users seem to be most at odds. It
is also a difficult area to develop, as useful primary materials must generally
be put into digital form before posting to the Web, and of course there are
serious and unresolved copyright issues to contend with. Still, this survey
indicates significant demand in the academic community at all levels for primary
research material on the Web.
In addition to questions with pre-established response options, a Comments section provided users with a chance to offer general comments, both good and bad, about the H-Net site. This format provided respondents with a chance to suggest new directions for the site and for academic Web sites generally. Interesting, too, were the range of evaluative comments, from "I did not find the site very useful"" to "Thank you for providing an invaluable service to teachers and scholars everywhere." Apparently, quality and coverage expectations for academic sites are still in the process of being formed.
The comments we received fall into two large categories: suggestions for new
features in site design and suggestions for new resources. In the site design
category, a number of respondents suggested improvements in search capabilities.
Some of these had to do with the interface for our search engine. The engine
itself is deliberately simple in design, allowing it to run on the relatively
modest hardware at H-Net's disposal. However, these comments point to a desire
for easier orientation within the site and quicker routes to specific resources.
Other suggestions have to do with improving the searchability of the logs H-Net
lists generate, of the Job Guide or of our Book Reviews. All were in the works
at the time of the survey and are now operational. In contrast, the suggestions
for new resources will be harder for H-Net to tackle. Suggestions for future
additions range from richer syllabus banks to map collections, a resume bank,
reportage of political events, more thorough bibliographies, and the inclusion
of historical dictionaries.
Combined with the results from the response-option section of the survey, the comments section helps give a sense for the expectations users bring to an academic Web site of the type H-Net provides. Users expect to find teaching and research materials, and they expect to be able to move rapidly and easily between and through sections of the site in their search for them. They are not as interested in finding secondary research material as might have been expected; for example, nobody mentioned e-journals or electronic versions of print journals as a kind of resource they would be glad to see. Instead, they are looking for basic tools for orienting themselves in periods and places they either do not know or tools for communicating with others (teaching), and for materials that wil help them pursue serious research in their own fields.
The pattern I see in the data analyzed here leads me to believe that most
users want the Web to supplement resources they have available in their
university or college libraries, rather than to replace them. The Web is not
asked to replace the traditional print form of the scholarly journal, even
though it could do so in very interesting ways. It is not asked to make recent
scholarly monographs available online. I believe the prominence of primary over
secondary research material in both e-mail and Web samples is consistenet with
this interpretation of the data collected here. Archival material is difficult
and expensive to study, requiring the physical presence of the scholar; it can
also be extremely difficult to search efficiently. Putting archival material on
the Web would be of benefit not only to research scholars but also to teachers,
thus fulfilling one of hypertext's greatest promises-the possibility of
encompassing a number of different levels of expertise within a single hypertext
suite, creating a resource within which researchers, teachers and students can
all find something useful. Further studies of this type will be needed to
determine more precisely what expectations academics have already formed for the
Web. Once those expectations are better met, the Web is sure to take an
important place in the academic world, but it will probably not replace existing
institutions like the campus library. Instead, it will supplement those
institutions, in ways that will act to the benefit of scholars, teachers, and
students alike. Site designers can contribute to this process by taking into
account both audience expectations and the vast potential of the Internet for
the creation of new kinds of resources which might generate more expectations,
even as we all work within the constraints imposed by the costs of resource
The provision of more materials on the Web will contribute to the democratization of academic culture discussed at the beginning of this paper. As more research and teaching materials become available in electronic (and hopefully free or very low-cost) form, one aspect of the research advantage enjoyed by rich universities in rich economies will begin to erode, though of course other factors are likely to remain unchanged. Scholars will be playing on an increasingly level field. The H-Net Web site, if we continue to build and develop it, can have the same kind of impact the H-Net lists have had. In fact, the impact of the Web site could be much more widely felt than that of the lists, because the potential audience for the site is so much larger than the subscriber base for the lists. If the H-Net site can help acclimatize traditional scholars in a newly electronic world, it will have served a valuable purpose indeed. If it can offer primary research materials, it will help contribute to the democratization of the academic research culture dominant in higher education. The Events database can help new scholars break into the information networks that established scholars already have formed. Finally, the site could serve as a crucial interface between the academy and the outside world. As of this writing, it remains to be seen whether this enormous potential can be realized; some of it is already happening.
Fig. 7. Suggested Resources-E-Mail Sample
Fig. 8. Suggested Resources-Web
Please Note: This paper must not be distributed
or reprinted in this form without express permission from the authors.