9.2.1 Insights from theoretical perspectives

Ecology of communicative transactions

Kaufer and Carley's ecology of communicative transactions consists of agents exchanging communicative transactions within an evolving open systems ecology. This idea is an extraordinarily powerful one when discussing a rapidly changing field such as electronic scholarly publishing. Kaufer and Carley's analysis serves as a reminder not to assume that the obvious factors are driving the process, and to be aware of the complex inter-dependencies between technologies, stakeholders and the scholarly environment.

The existing print journal is an example of what Kaufer and Carley call a 'book agent', by which they mean anything that is both printed and mass-circulated. Associated with any form of written communication in their analysis are the concepts of fixity, reach and distance. Distance means the separation between writer and reader caused by any combination of space, time and culture. Communications technologies can extend an author's reach through asynchronicity, durability and multiplicity. Of these concepts, fixity and durability relate to the form of the scholarly journal. Reach, asynchronicity, and multiplicity relate to the functions of the scholarly journal (see 9.3: Transformations in the function of journals on page 172). How do these properties inform possible changes to the scholarly article and journal?

Fixity is the extent to which a communication technology enables the communication to be retransmitted without change. Print, of course, is an example of a medium that is inherently highly fixed. Electronic text is much less highly fixed (unless stored on a read-only device). It is quite easy to alter an electronic text accessed across a network, either on the server at source, or in transit between the reader and the client. The user may not be aware that any change had taken place or that what they are reading is not what was originally written by the author. This is particularly a problem if the electronic text is changing frequently without this being marked in some obvious way. In the world of print, articles are defined packages of content and books have editions to indicate significant change. Online publications may change on a daily basis either without any indicator or at best a version number. Some online journals (e.g. Public Access Computer Systems Review) now allow the author the option of updating a published article. In this environment, how can a reader know which version has been cited? Clearly, the current system of scholarly communication requires the fixity principle to be maintained in an online form.

Durability is defined in terms of how long the content of a communication is available for interaction [Kaufer and Carley, 1994, p. 34] and is a property of the medium (i.e. of the form) of the communicative artefact. Durable texts diffuse more widely and for longer and thus increase the author's reach. Durability is a problem for the transmission of texts across time as well as distance in the sense that it may be difficult to access an older electronic text because of technological change. Either the text may have to be migrated to a newer technology (violating the fixity principle by altering it) or it may be unable to access it (violating the durability principle by making it impossible to transmit it onwards). It is essential for electronic texts to be durable so that they can be accessed and cited into the foreseeable future.

In discussing the development of print, Kaufer and Carley point out that new media "coexist with, rather than replace, established media" [Kaufer and Carley, 1993, p. 6]. Rather than pitting print against the previous oral and written cultures they argue a more useful question is "what additional possibilities became available to speakers and writers once they could also rely on print?" [Kaufer and Carley, 1993, p. 6]. One might well extend this to ask what additional possibilities become available once writers can rely on hypermedia? These are ideas that some of the leading edge e-journals are just starting to explore.

On the subject of electronic print, they suggest a profitable research area might be to analyse how things like electronic mail and bulletin boards (and by extension electronic journals) allow the "diffusion of new ideas to the right people, over print alone" [Kaufer and Carley, 1993, p. 393]. Targeted mailing lists exist already, but no one seems to be suggesting targeted e-journals. However, it is possible to list requests for updates on topics of interest with a number of online services. The Amazon.Com online bookstore allows a user to list keywords and to be emailed when new books with these keywords become available. The Journal of Biological Chemistry Online allows a reader to indicate they wish to receive email when a specified article of interest is cited. It is quite possible that in the near future a scholar could have a series of alerting requests with a range of services providing a constant stream of messages about new discoveries and references to other scholar's (or their own) work. This is simply a logical extension of the old Strategic Dissemination of Information (SDI) services that libraries have been providing in conjunction with online databases for decades. The difference now is that it can be user initiated and operate on the full text of journals, not just databases of journal surrogates.

In their analysis of academe and the role of print journals Kaufer and Carley argue that in diffusing new ideas journals are simultaneously faster than book publication or face-to-face interaction (due to their frequency of issue and increased reach respectively), and slower than newspapers (due to the gatekeeper function of peer review). The consensus according to Kaufer and Carley is that many scientists regard the speed of journals as too slow, particularly in very fast-moving fields. This is clearly an area where e-journals can have a significant impact. E-journals can diffuse new ideas more quickly because of a range of factors:

• faster refereeing processes through use of email
• faster production processes through electronic submission
• faster appearance because there is no need to wait until a print issue is full before publishing
• faster delivery because there is no need to physically move anything to the reader

Kaufer and Carley argue for a non-deterministic view of the effects of new technologies, taking account of other sociocultural variables as well. They do, however, admit that the benefits of hindsight are much greater when looking at print than when considering the newer technologies. They make the explicit point that "researchers interested in (the newer technologies of communication) might learn something by analogy with our treatment of print" [Kaufer and Carley, 1993, p. 17]. This is a theme echoed by a number of other researchers in the field [Fjällbrant, 1997], [Rowland et al., 1995].

One of the things that such a researcher might learn is not to assume that a new communications technology implies a qualitative difference. Kaufer and Carley argue strongly for first considering the new on a common quantitative continuum with the old:

To ignore historical continuities across technologies is to reveal more than a historical ignorance of what came before. It is to remain primarily ignorant of the technologies we now take for granted. [Kaufer and Carley, 1993, p. 18].

These insights imply both that changes in journal form will not solely be driven by technology and that the history of the scholarly journal may well have much to say about its future.

Punctuated equilibrium and speciation

According to the punctuated equilibrium model, sudden change in species is driven by environmental change. More specifically, local environmental change on a sub-population of an organism will preadapt this sub-population for the expansion of the changed environment (Kaufer and Carley's analysis has modelled the process of communication as an ecology. Presumably, scholarly communication is a part of the overall communication ecology. Under this analysis, an individual journal title is analogous to a species, with changes in particular articles being the equivalent of changes at the sub-organism level.What then is the environmental change within this scholarly communication ecology that might drive speciation?

One excellent candidate is changes in the scholarly communication space caused by new technologies. The most important of these technologies for scholars since the middle of this century have been the communication and computing technologies. These have provided the potential for transformations of both professional practices and intermediary processes. The stakeholders who have been affected are publishers, scholars (as both consumers and producers of content), scholarly societies (as representatives of scholars and as publishing intermediaries in their own right) and librarians.

Most recently, the rise of the Internet and its increased possibilities for electronic publishing have provided a new ecological niche that is being colonized by all sorts of communications products, scholarly journals included. Is it possible that the move to the new electronic environment might involve the development of new species of scholarly communication artifacts?

The punctuated equilibrium model suggests that the equivalent of the sub-population adapting might be those journals now making the move into an online existence (with those who are making a break with paper altogether, rather than retaining parallel publishing, being best adapted). Within a relatively short period of time (far shorter in the world of publishing than in geological time) the remaining journals will need to either adapt and move to online or die out. This will appear in the equivalent of the fossil record (library serials holdings) as a quite quick transition from print only to electronic only. The consensus among discussions after electronic publishing conferences is that the transition should be effectively complete by 2010 - 2020.

Another insight provided by the punctuated equilibrium model is that the record of species in the fossil record manifests itself as long periods of stasis punctuated by rapid change or sudden jumps from one species to the next. The last of these punctuary jumps in scholarly communication was the spread of printing, followed by hundreds of years of relative stasis in the form (although not the numbers) of the scholarly journal. The current changes in communication technologies may well be the trigger for another punctuary jump followed by another (although probably shorter given the pace of technological change) period of stasis.

A final insight from the domain of palaeobiology is that patterns of speciation and extinction are chaotic, strongly influenced by the environment around them and very difficult to predict (as opposed to easy to infer with the benefit of hindsight). Stephen Jay Gould argues that if the history of life on Earth were to be replayed from the beginning, the likelihood of homo sapiens arising are vanishingly small [Gould, 1989]. The implications of this for any attempt to predict developments in electronic publishing technologies (and by implication the form of e-journals that rely on those technologies) are clear. These developments will also be hard to predict and influenced by the wider computing and communication environment.

One of the critical developments in the understanding of the effects of punctuated equilibrium on speciation and evolution is the fossil fauna of the Burgess Shale described by Stephen Jay Gould in his book Wonderful Life [Gould, 1989]. These organisms date back to 530 million years ago, just after the 'Cambrian explosion' - a time of dramatic diversification in evolution. The importance of the Burgess Shale is that it preserves many anatomical designs that existed for a while and then disappeared, never to be seen again. In accounting for this astonishing diversity, Gould points to three main factors:

• the first filling of the ecological barrel, providing empty ecological niches that life could rapidly colonize;
• a directional history for genetic systems, making significant change harder over time;
• early diversification and later locking-in as a property of all systems.

The Internet provides those empty ecological niches for new journals to colonise. The directional history of technology developments is evident in the ways that particular technology patterns (the dominance of Windows, the de facto use of Adobe's PDF as a page description standard) become rapidly widespread and resistant to change. The early diversification and later locking in are occurring around us in the e-journal domain at this very moment.

Genres and new media

Three of Agre's [Agre, 1995b] categories of analysis (communities, media and genres) provide a range of useful insights into changes in the form of hypermedia scholarly journals.

Different scholarly communities have always had quite different formatting requirements for print journals, traditions of using such journals and access to new technology. The Science, Technology and Medicine (STM) disciplines have always required more from their print publications in the way of the need for diagrams, equations, photographs and non-Roman symbols. There is some evidence ([Zuckerman&Merton1979], [Beyer1978]) that rejection rates in STM journals are significantly lower than in the social sciences and humanities. By the nature of their disciplines, the STM communities have tended to have earlier access to high technology of all sorts, including networked desktop computers. It is likely therefore that the STM journals will make the move to electronic delivery first and also use the more innovative delivery and presentation technologies. At present this seems to be the trend, with scholarly societies in the STM field driving the development of (mostly parallel delivery but soon electronic only) e-journals. Among the large commercial journal publishers, it is those with a large STM component (Elsevier, Blackwell Science) that are leading the move to parallel delivery.

With respect to media one of the most obvious differences between traditional print journal publishing and electronic journal publishing is the medium in which they are presented. Print has always been used for journals, monographs, conference proceedings, preprints, informal newsletters and the like. Off-line electronic media like CD-ROM and floppy disk have also been used for all of the above. The predominant electronic delivery mechanism at the moment (and into the future) seems to be some form of computer network, either the Internet or whatever it evolves into. Agre points out that the affordances of the medium govern how it is used. The affordances of the e-journal are clearly a concern to a number of readers, as indicated in their survey responses, and the design of new e-journals need to take account of these concerns and support existing usage patterns (where possible) or substitute improved alternatives (where not). The use of a live table of contents at the start of an e-journal as a (partial) replacement for being able to flip through the print pages of the paternal is one such example.

Finally, the scholarly journal as we perceive it today is a mature example of a genre that has evolved to be useful to members of a scholarly community. It is perhaps misleading to talk about the scholarly journal as a genre. In practice, there are a number of variant sub-genres including less formal newsletters, "letters'-type journals ( Physical Review Letters ), journals with an editor but no peer-review and anonymously peer-reviewed journals. Members of such sub-genres can be placed on continuums of refereeing rigour, type of content and speed of publication. Any replacement for the scholarly print journal will either need to reproduce the functions of this genre or replace them with new functions that are more useful.

Agre's idea of 'doing more' when designing new media is a central one. One of the critical issues is the question of whether hypermedia scholarly journals are a new genre or just an old genre in a new medium. In a real sense, this relates to the question of evolution versus revolution. At what point does an evolving new thing move from being quantitatively different to quantitatively new? Kaufer and Carley argue that researchers should not leap to declare something qualitatively different before first exploring the possibility of mere quantitative difference [Kaufer and Carley, 1993, p. 17].

Scholarly journals are almost always discipline-specific and are therefore intended for a particular scholarly community. It is instructive to consider what the answers to Agre's design questions might be in the context of hypermedia journal design for such a community.

Users: We probably have a fairly good idea of who our scholarly community is, although placing the journal on-line may well open the community up to interested outsiders. Do we want this? Are we prepared for the interactions that may result (particularly in areas like cosmology and evolution where people have strongly held views)?

Purpose: It is probably safe to assume that the purposes of hypermedia scholarly journals will remain the same as their print counterparts, at least for some time.

Activities: In an on-line environment, readers will have a wider range of potential interactions open to them than in print:

• moving directly to a particular known article (rather than reading the whole "issue')
• locating articles containing particular words
• locating articles containing particular media types (video/sound/simulation/image)
• tracking the ongoing peer commentary associated with an article
• using "forward-citation" links to identify all those articles that cite the one they are reading.

Usage: Do we keep the "look and feel" of the traditional journal (as many pilot projects are doing) or do we move beyond this? Will people's expectations of Web-based material (based on the other, increasingly commercial, Web sites they visit) influence what they expect to find in a Web-based e-journal? Is the appearance of the traditional article a means or an end? How might we respond to scientific articles that look like Encarta? What about scientific articles that look like a video game?

Medium: This relates to the 'do more' philosophy; what can we do that we can't already do in print?

Evolution: A Web site associated with a hypermedia journal would certainly add new content as new articles are submitted and as the richness of links in older articles perhaps increases. Considering this aspect of hypermedia journal design more broadly, it is possible to envisage a gradual evolution of the journal's presentation away from print traditions over time, in much the same way as print evolved away from simply copying the older manuscript traditions.

Awareness: For many journals, distribution is associated with membership in a scholarly society. E-journals linked to a print site-license would presumably be promoted internally within the site. If the journal wishes an expanded audience, either for increased subscriptions or to sell non-journal products (such as single articles or special supplements) it may need to advertise more widely.

Access: The increasing availability of technology has already been covered (see 4: Technology Developments on page 52). Over the next decade it is probably reasonable to assume that most university scholars in the developed world will gain access to the Web on desktop. The issues relating to the rest of the world, and to access away from the desktop are a real concern for some technological models, and a real drawback for e-journals relative to print.

The answers to these questions can only be determined through much more research with designers and users of e-journals.

Agre also points out that technology developments do not behave in the same way as developments in other fields because of what economists call path dependencies:

... media industries are powerfully path-dependent because of effects deriving from the compatibility of different commodities ...Once proprietary standards become entrenched in the marketplace, so that compatibility effects create ever-higher barriers to entry for potential competitors, their owners can start to extract rents from a variety of other parties. Moreover, network externalities (costs to individuals that derive from everyone else's choices) mean that dominance over a market tends to expand once it is established. [Agre, 1995b]

A good example of this in the domain of electronic publishing is the way in which Adobe's Acrobat technology is now the de facto standard for the distribution of 'electronic print'. As recently as three years ago there were a number of competitor technologies, some with superior technology offerings in some respects [Gruman, 1995]. Acrobat managed to gain a slight edge and that was enough to cause a swing in favour of its PDF technology to such an extent that the alternatives are now longer available on the market.