First Monday

DiPP and eLanguage: Two cooperative models for open access by Cornelius Puschmann and Peter Reimer

This article describes the development, deployment and ongoing use of two collaborative platforms for the publication of peer–reviewed, open access, academic e–journals and the different organizational environments towards which each system is tailored. The two projects presented here, DiPP [1] and eLanguage [2], are examples of different approaches to electronic publishing: the first being a regional publishing cooperative, the second a disciplinary one [3]. Our description includes the technical specifications of the DiPP platform, which is an extension of the Plone [4] content management system and Fedora [5], and eLanguage, which is based on the Public Knowledge Project's Open Journal Systems (OJS) [6] and the blogging platform Wordpress [7]. The authors hope to convincingly illustrate how very specific factors — both technical and cultural — shape the requirements of those who work with different publishing products and that these factors must be carefully reviewed when building a publishing cooperative. We also wish to emphasize the importance of bringing together researchers, scholarly societies and libraries in a joint effort to create the ideal environment for open access. Scholarly societies in particular can act as influential stewards for the proliferation of open access due to their reach and the trust they enjoy, and therefor bringing them aboard is of critical importance [8].


Publishing cooperatives and open access
Two paths to open access



Publishing cooperatives and open access

Cooperatives in areas such as agriculture or manufacturing are created to more effectively utilize resources that can be shared among stakeholders with similar but not competing interests. The idea is simple and yet highly effective: expensive tools are shared to drive down costs, which is equally beneficial for everyone participating in the cooperative. At the same time goods can be sold at conditions that are more attractive to the producers because their combined production output is greater and they are thus able to negotiate better bargains. Apart from sharing tools and facilities, cooperatives are also a way of optimally utilizing knowledge and competence among stakeholders.

How does this relate to open access publishing?

While there are certainly countless differences between farmers sharing a plow and academics sharing a Web server, the key similarity is that both groups are in need of specialized tools, tools that can be shared without wearing out, but are at the same time costly to acquire and maintain. Although most researchers wanting to launch an electronic journal can theoretically do so with the help of the institution employing them, there are significant costs associated with digital publishing, not only financially but also it terms of time, energy and human resources.

A publishing cooperative addresses this issue by building a shared infrastructure that can be used by all participants. Digital infrastructures have the additional and unique advantage that they do not degrade and can be used at the same time by many people, so replicating them seems especially inefficient. Furthermore, what a researcher needs is a partner with specialized knowledge of topics such as how to manage a journal work flow, how to make content accessible to search engines, what formats are suitable for long–term preservation, how the impact of a journal can be measured and what business models exist for open access publications. While he can certainly acquire competence in those areas on his own — learning by doing — this is extremely inefficient considering that he already has an area of expertise that he should be allowed to focus on: his research. A publishing cooperative can take on those roles that a commercial publisher plays in a closed access setting, to the extent desired by editors and feasible for the cooperative.

Beyond benefits for journal editors, publishing cooperatives are also in the public interest, because they allow for better coordination of resources than can be accomplished through the funding and support of individual journals. A state–supported cooperative such as DiPP gives policy–makers an interface between research (that is in most cases publicly funded) and its effective output — in other words, an interface between the investment and its return. The degree of coordination that is possible in a regional cooperative is important, because it assures that aspects such as long–term preservation of content, the use of suitable data formats and legal issues of publishing can be centrally supported and where necessary also regulated, instead of leaving these pivotal questions exclusively for journal editors to solve, who may or may not have the expertise to make the best decisions.

Cooperatives — both disciplinary and regional — are thus an attractive way of making open access publishing sustainable and more effective. The two cooperatives subsequently presented in this article both have their signature qualities and are both being continually developed with careful consideration of their differences. It is our impression that, by supporting open access publishers as networks, we can optimize our support for them significantly.




The Digital Peer Publishing Initiative (DiPP) [9] was launched by the government of Germany’s largest state with the goal of fostering innovative forms of Web–based, scholarly communication. The focus of the initiative is on providing researchers with the technical, organisational and legal framework for founding and establishing open access electronic journals. In the context of the cooperative model outlined above, DiPP is a regional cooperative, though through its expansion into national and even international contexts it can perhaps more accurately characterized as having regional roots.

A toolbox for electronic publishing

DiPP started in the Spring of 2004 as an organization of both newly launched and pre–existing e–journals based in North Rhine Westphalia (NRW) [10]. Eight member journals belonged to the initial group that was publicly funded as part of the open access stimulation project that spearheaded the initiative. More partners subsequently became affiliated with DiPP, up to a total of 13 journals in 2007, with further admissions under way. One of the the founding members, Constructions, has now moved into a more international context — the eLanguage project, described in detail below. Today the DiPP repository contains over 600 peer–reviewed open access articles from a broad range of disciplines, partly combined with interactive and multi-media supplementary material.

The initiative is not focused on a single academic discipline, because to actively study the highly heterogeneous demands of different scientific communities and editorial groups is a vital part of the project. One of our goals is to determine in how far developed and sustained support structures can be generalized and transferred from one group of stakeholders to another. In order to foster generic technical development and ease effective knowledge transfer between individual e–journals, the University Library Center (HBZ) [11] has adopted the role of an aiding and integrating companion in the initiative.

Multiple levels of support

The services offered by DiPP aim to support editorial groups on multiple levels, specifically regarding technical, organizational and legal solutions.

  1. On the technical level, a comprehensive publication and work flow system is offered (see below). DiPP’s aim is to continually enhance its services according to the needs of its constituents and information management best practices.

  2. On the organizational level, editorial groups are independent and control their operations themselves. They are anchored at the university of the leading researcher in a (usually international) editorial group and involve partners of the local library who adopt various roles including metadata checking, indexing and formatting. Finding, testing and establishing novel practice scenarios and divisions of work is another goal of DiPP. The operation and business models are organized by the journals and universities themselves, though DiPP offers consulting services, access to the technical infrastructure and support.

  3. On the legal level, a modular set of open access content licenses (DPPL [12]) is provided that allows for flexible application in different demand profiles. For example, authors may allow changes to parts of a publication while fixating others [13]. The licenses were developed by legal specialists of the ifrOSS [14] and are both customized for academic use and compatible with national law.



Figure 1: Schematic view of the system architecture

Figure 1: Schematic view of the system architecture.


The DiPP platform was and is developed relying mostly on open source software. In order to provide maximum flexibility and to be able to adapt the system to the specific needs of an e–journal, the system is divided into three hierarchical levels: a presentation layer, a service layer and an application layer. The communication between the different layers is conducted with standardized protocols such as WSDL–SOAP [15] and XML–RPC. A schematic view of the architecture is given in Figure 1.

Service layer

The storage backbone of DiPP is the Fedora Repository System, developed at Cornell University. All publications and related files, such as images, video clips and other multimedia files are stored in the repository together with the accompanying qualified Dublin Core metadata. The content model has been adapted to represent the hierarchical structure of academic journals with volumes and issues, while the related metadata is exposed to harvesters via the OAI–PMH protocol [16]. The repository is also the basis for long-term storage.

Application logic

In this layer the the actual manuscript handling is done. In a first step, a persistent identifier is assigned to papers which are accepted for publication. This identifier is unique and is also used to create a URN [17] using a dedicated name space for the respective DiPP journal. In the next step the submitted papers are converted to DocBook XML, which was chosen as the DiPP storage format because it is open, free and vendor–independent. The conversion of source documents (which are usually uploaded in Rich Text Format [RTF]) is conducted with upCast [18], a closed source commercial application. DocBook XML also serves as the basis for further transformations, allowing conversion to the presentation formats XHTML and PDF. A word processor template is developed with the editorial staff of each e–journal, which is ideally used by authors to obtain a well–structured and semantic document, requiring as little manual post–formatting as possible. The names of the applied styles for paragraphs, headings, etc. are transformed to CSS classes in the final HTML file. It is thus possible to completely separate design and structure of the manuscript — an important aspect for future use of the document in different contexts [19]. The presentation layer, described in the next section, provides a WYSIWYG editor for final copy editing of the manuscript. Because documents are stored in XML, the copy editing process is directly integrated into the publishing work flow. This helps the editorial staff to keep track of the publishing progress of accepted papers and avoids needless back–and–forth handling of documents. Everyone involved in the process, from the editor to the author have dedicated roles and permissions in the system, shown in the work flow schema below.


Figure 2: Publishing work flow

Figure 2: Publishing work flow.


Presentation layer

The top level of the system is the presentation layer, which is implemented using the Plone content management system. Basis for Plone is the Web application server Zope [20], which is written in Python. Plone is a fully featured yet easy–to–use CMS with an impressive number of plug–ins, called “products” in Plone jargon. Among these are products for wikis, newsletters, blogs and many other services. We developed the Editorial Toolbox, which adds the following functionality to Plone:

External services

Due to the service oriented architecture, it is also possible to connect external services to the platform. An example is the GAP peer review work flow [22], which can use the user base of the DiPP platform for authentication via the LDAP protocol and inject accepted papers into the publication work flow via the OAI–PMH protocol.

Current state and outlook

In Germany, DiPP has become a technical standard and a publisher network at the same time, with activities beyond the geographical borders of the state and beyond national borders, if one considers eLanguage. Insights gained from working with an interdisciplinary family of journals were used to enhance and customize eLanguage and have at the same time made it possible for us to identify which areas of publishing support can be universally standardized. At the same time, DiPP will continue to grow by providing tools, support and technological infrastructure to researchers and by improving their access to scientific information.




The eLanguage project was launched to facilitate access to scholarly content in the academic discipline of linguists by giving those who create said content the right tools for the task and by supporting their efforts with a unified publishing platform [23]. This platform consists of both a technical and an organizational component, bringing together a system for publishing that is as well–adjusted to the researchers’ needs as possible and a group of people to manage the shared infrastructure. It was understood from the onset that both the organizational and technical structures that needed to be built had to fit the specific requirements of linguists and cater to their disciplinary needs. In contrast to DiPP, eLanguage targets only journals from a single field — linguistics — which meant that that there was less variability in the requirements of different participants. Another difference is that eLanguage is more international in scope, which has effected various issues, including choice of the publishing software and practices such as how member journals are admitted.

A scholarly society promotes open access

The initiator of the eLanguage project is the Linguistic Society of America (LSA) [24]. The LSA is the largest scholarly society in the field of linguistics in the world, with over 5.000 members in 63 countries. The society’s to–date experience in publishing stems from an 80–year involvement in Language [25], a highly valued and widely acknowledged print journal that has published many classical papers and reviews and is regarded as one of the most influential publications in the field.

While Language is highly regarded for its rigor and broad impact on the discipline, the limitations it is naturally under due to its paper format are increasingly recognized as problematic and therefore an additional electronic publication was discussed internally by the LSA in 2006. After considering several alternatives, a model was chosen that went well beyond simply replicating Language on the Web, a model that put the specific advantages of digital publishing — speed, low and fixed costs, no breadth limitations — to optimal use.

Similar to DiPP, eLanguage was devised as a hub or portal — a family of journals instead of a single publication. Each journal in eLanguage (called a co–journal) is fully independent and maintained by its own editors according to their wishes. This includes options such as hosting the journal Web site independently and choosing a separate domain name. Before being admitted into eLanguage, all co–journals have to be accredited (see below) and pledge themselves to the following points:

The need for these requirements was founded on technical and formal considerations (platform and style sheet), but also on what the society deemed to be beneficial for the discipline (open access) and the significance of quality control (peer review), especially in a new form of publishing still not fully trusted by many researchers. A decisive move was the society’s decision for open access, which came gradually and was informed by the mentality that something that benefited linguistics as such would also benefit the LSA and its members, and was thus worth an investment. Originally, the requirement that authors should be LSA members was included in the project outline, but this was dropped when it became increasingly clear that such a requirement would hamper the efforts of journals reliant on contributions from international scholars, many of whom were relatively unlikely to be LSA members. The choice for open access was thus also a completely practical one. Limiting access to eLanguage, even author access, would have effected quality and publication impact in a negative way, and this was a constraint that was deemed unacceptable both by journal editors and the society.

Organization and management

eLanguage is a joint effort of several different institutions. The Linguistic Society of America has initiated the project and acts as the publisher of the eLanguage Web site, while the ongoing management is in the hands of the editor–in–chief, who is elected by the LSA’s governing body, the executive committee (EC), for a period of seven years. Dieter Stein, chair of the department for English Language and Linguistics at the University of Düsseldorf, Germany, is the current and first serving editor–in–chief. He has overseen the formation of an editorial board in which renown linguists from different sub;disciplines are represented. This editorial board and the editor–in–chief are responsible for the day–to–day administrative management of the cooperative, while a technology provider manages the platform. The office of the editor–in–chief is in direct contact with both the co–journal editors and the technology provider that hosts the eLanguage site, while co–journal editors are also part of an advisory board that is consulted on issues that are relevant to all stakeholders. An advantage of such a structure is that the management team can act as a quasi buffer between the individual members and their needs and the service provider that may not be familiar with the field of linguistics or particular issues that are important to scholars.

IT support

Hosting and support for eLanguage is provided by the HBZ [26], the University Library Center, located in Cologne, Germany, which is also home to the DiPP project. The HBZ’s competence in electronic publishing and expertise in library and information services was the deciding factor for the collaboration, in addition to the close geographical proximity (Düsseldorf and Cologne both lie on the banks of the Rhine, roughly an hour apart). Because the technology provider assures both the accessibility and maintenance of the the tools that power eLanguage and optimizes access to the content itself (for example by assuring that article metadata for all member journals is available via an OAI–PMH interface [27]), it is essential that it retains both close communication with the editor–in–chief’s office and that it is a specialized, non–profit actor that has both the knowledge of and the commitment to the process of scientific publishing.


The first step in the development of eLanguage was the accreditation of co–journals, a process that continues and is among the core tasks of eLanguage’s editorial board. In order to bring a new co–journal into the cooperative, a researcher must submit a proposal which is then reviewed by the editorial board. A proposal should contain a detailed description of the area of research (e.g. formal syntax, computational linguistics, endangered languages) and nominations for an editorial board. If a proposal is accepted, the journal is given access to its own container inside of the journal publishing software (see below) and can then undertake its own formal organization.


The co–journals, which form the core of the eLanguage project, are fully independent entities with their own internal organizational structure (editorial board, advisory board, assistants et cetera) and full control over the publication process. They contribute the bulk of eLanguage’s content from a wide range of linguistic sub–disciplines which is then made available via the site’s central Web feed and on


Figure 3: The eLanguage main page (demo)

Figure 3: The eLanguage main page (demo).


All co–journal content is simultaneously available on the journal’s own site, which is linked to the eLanguage main page. Co–journal sites are not mere subsections of but have their own design and structure, within certain limits imposed by the publishing software. They can be managed by their owners via a user interface and accessed through a separate domain name [28], putting them entirely under the control of their editors. For example, while frequent publishing is encouraged, there are no fixed global schedules or time tables that editors must adhere to and the journal’s publication cycle is determined by its owners.

Legal aspects

One pronounced goal of the project is to increase access to scholarly works in linguistics, another to strengthen the rights of those who create such works. While intuitively one might assume that eLanguage as a organization or the LSA as its backer acts as the owner of the works that are published via the platform, this is not the case. Instead, the LSA acts merely as the mediator between authors and readers, assuring access and availability, but leaving the rights to the creation strictly in the hands of the creator.

All articles published in eLanguage are the property of their respective authors and made available under a Creative Commons license [29]. The specific Creative Commons variant (e.g. “no derivatives”, “share alike”) [30] is chosen by the author when he acknowledges the publication agreement.


When initially developing a technology framework of eLanguage, several factors motivated the choices made. Specifically, the following points were regarded as integral:

Considering the core mission of eLanguage — to facilitate access to scholarly information — the paradigm for the development of the platform was simplicity and practicality. Not building a new platform was the aim of the project, but building a community and a resource for academic content, and this motive informed our decisions. Apart from strictly technological criteria, aspects such as thorough English–language documentation and ease of use were of key importance [31].


The Public Knowledge Project’s Open Journals Systems (OJS) is an open source journal publishing software, based on PHP [32] and MySQL [33]. It is described as follows on the PKP Web site:

“OJS assists with every stage of the refereed publishing process, from submissions through to online publication and indexing. Through its management systems, its finely grained indexing of research, and the context it provides for research, OJS seeks to improve both the scholarly and public quality of referred research.” [34]

OJS was chosen as the backbone for eLanguage for several reasons. The impressive installation base of the product played a role [35], as did the extremely detailed documentation and the active and highly responsive developer community behind the code [36].


Figure 4: The OJS work flow

Figure 4: The OJS work flow.


Another perceived advantage of OJS was the close way in which its work flow emulates the procedures of print publishing, resulting in an extremely linear, even monolithic approach to journal management. This was seen as a benefit because it was clear to us from past experience that most of our editors would conceptualize their journals as the electronic equivalent of print journals, all advantages of Web technology not withstanding. This meant that especially those editors and writers with little technical background would prefer a linear and fairly narrow publishing work flow to an open, flexible and extensible content management system, because the freedom of such a system would be perceived as confusing and not beneficial. Furthermore, a very large array of journal–specific functions is integrated into OJS, many of which may not even be necessary in the strictest sense in a digital environment, but are understood as being part of what constitutes an electronic journal. Acknowledging the cultural frame of print publishing and finding a system that simulates it was thus a reason for picking OJS that took precedent over strictly technical aspects. The ideal tool, we discovered, is the one that wins the trust of its users. Many advantages that may seem important to the technologist are effectively invisible to the end–user and many seemingly superficial flaws in a user interface can deter from the adoption of a software despite well–written code. Purely technical deficiencies can be fixed later in the course of a project, if at a cost, but a deficit of trust will make success impossible from the very beginning.

Trust not withstanding, when working with OJS we learned that in terms of technology its greatest strength is also its greatest weakness. The highly formalized and sometimes inflexible work flow can obviously not support all modes of journal management equally well. For example, we found while working on DiPP that some linguistic journals preferred using their own internal system for managing submissions to OJS’ integrated peer review system. In such journals, the OJS interface would often be used merely to publish new articles after all prior steps of the work flow had already taken places outside the system.

Another issue, one that is now in the process of being addressed [37], is the way in which OJS treats articles in relation to the data files they are contained in. When submitting a piece, the author uploads a file to OJS which is then disseminated to reviewers. Later, when the article has been approved, so–called galley files, which can be in HTML or PDF format, are uploaded and subsequently published. The reason why what is published is not necessarily identical with what was submitted lies in the inadequacy of HTML and PDF as data storage formats. Both are not marked up semantically, but visually and therefor less well suited for digital archiving and editing than XML, the language which is usually considered ideal for these tasks.

It is important to note that while we found these points worthy of improvement, their selection is subjective and strongly reflects our specific needs (and they are partly in the process of being fixed). Obviously it is impossible to suit everyone’s needs equally well and, considering that, OJS has proven to be an excellent tool.


The second major building block of eLanguage is the popular blogging software Wordpress. Also built on PHP and MySQL, Wordpress is widely assumed to be the world’s most popular blogging platform [38]. While the function of OJS in the eLanguage mashup is to serve as the backbone on which the co–journals run, Wordpress powers the blogs that are used for the publication of “meta content” such as book notices (previously contained in Language), news from linguistics departments across the country and other relevant information that is important to linguists, but not strictly concerned with research. We chose Wordpress for its stability, ubiquity and ease of use, as well as for its appealing visual presentation. As mentioned above, we found that a clear user interface is not a negligible quality and Wordpress has a design that lends itself well to novice users.


Figure 5: Creating an eLanguage news item with Wordpress

Figure 5: Creating an eLanguage news item with Wordpress.


Current state and outlook

The official launch of eLanguage is anticipated for Spring of 2008. As of August 2007, a total of four co–journals have been accredited, while talks are being conducted with additional researchers interested in launching journals. The approach taken by the eLanguage editorial board and the editor in chief has been to be highly selective in admitting new journals up to this point, in order to counter any existing skepticism of the electronic format, a format which many still place less trust in than in established publications from major commercial publishers. Applying a cream of the crop approach is thus a strategy for overcoming such prejudices.

The four accredited journals at the time of writing are:

Talks are under way to admit further journals from other specialized areas of the discipline.



Two paths to open access

Our experience with DiPP and eLanguage shows that there are no universal solutions when it comes to creating an electronic publishing cooperative. Picking the best tools for the job is a matter of understanding the specific needs of the constituents. Community and technology need to be integrated as closely as possible, otherwise there is the imminent risk that the stakeholders will perceive the technology as something external that they are forced to use despite its shortcomings.

At the same time however, the pursuit of generally applicable solutions is both in the public interest and in the interest of scientific communities, because without standardization issues such as long–term preservation, availability via search and library catalogs and other essential aspects of publishing are bound to be neglected. The decisive question is: where should standardization be the goal and where is customization preferable?

Another look at the differences between DiPP and eLanguage is helpful in this context.


Table 1: DiPP and eLanguage side by side
regional/national cooperativedisciplinary cooperative
initiated by a public institutioninitiated by a scholarly association
“democratic approach”:
bring emerging and less affluent researchers to the table first
“elitist approach”:
bring established researchers to the table first
two–level approach:
journal + technical provider
three–level approach:
journal + management cooperative + technical provider
modular/flexible content management system (Plone)monolithic/digital replication of a print journal (OJS)


We have found that both models of collaboration — regional and disciplinary cooperative — have substantial advantages over purely punctual funding of individual e–journals. Cooperatives share publishing technology (software), IT infrastructure (hardware) and support (human resources) among several journals, thus making electronic publishing cheaper, more efficient and far more manageable for editors. At the same time journal editors often lack the technical expertise when it comes to problems such as finding the right publishing software, configuring it, hosting a journal Web site, making content accessibility to search engines, enriching articles with metadata and picking the right formats for the long–term archiving of content. This effectively reduces accessibility of articles and wastes precious resources that could otherwise be devoted to what researchers are most competent in: creating scientific content. The advantages of both models are dependent on the context they are used in: while regional (state or national) cooperatives are an attractive approach to open access for publicly funded initiatives, disciplinary cooperatives have potential especially if scholarly societies lead them. Both paths can be followed with equal success and we believe both lead to better access to knowledge. End of article


About the authors

Cornelius Puschmann is a doctoral candidate and lecturer at the Department of English Language and Linguistics at the University of Düsseldorf.
E–mail: cornelius [dot] puschmann [at] uni–duesseldorf [dot] de

Peter Reimer, University Library Center (HBZ)
E–mail: reimer [at] hbz–nrw [dot] de



1. See

2. See

3. We owe the term publishing cooperative to Raym Crow of SPARC, who held a presentation outlining the concept at the First PKP Scholarly Publishing Conference in Vancouver on 12 July 2007. The presentation slides are available at See also Raym Crow, 2006. “Publishing cooperatives: An alternative for non–profit publishers,” Scholarly Publishing and Academic Resource Coalition (SPARC) Discussion Paper, at; see also version in First Monday, volume 11, number 9 (September), at

4. See

5. See http://www.fedora– Note that this repository system is not to be confused with the Linux distribution of the same name.

6. See

7. See

8. See Veltrop (2003) and Waltham (2006) for a discussion of open access in relation to scholarly societies.

9. See


11. See

12. The DPPL licenses are both compatible with German law and internationally applicable, see e.g.

13. In comparison to CC?

14. See

15. See

16. See

17. DOI support is also possible.

18. See

19. For example, a semantic document structure enables use of documents on mobile devices and is conductive for long-term storage.

20. See

21. See

22. See

23. For details, see the eLanguage Agora at

24. See

25. See ttp://

26. See

27. See

28. See for example the Web site of Semantics & Pragmatics, at

29. See

30. See for an overview.

31. It is worth noting that the requirements for DiPP were different. The system’s interface is bilingual (English and German), while the documentation is exclusively in German.

32. See

33. See

34. Retrieved from on 24 August 2007.

35. The PKP Web site states that the system’s installation base exceeds 900 instances as of March 2007 (see

36. Especially via the PKP Support Forum (

37. We are referring to Lemon8 (see, a service that will make it possible to base work with articles on XML files, rather than having to rely on derivate presentation formats such as HTML and PDF.

38. As stated on “[...] the largest self–hosted blogging tool in the world, used on hundreds of thousands of sites and seen by tens of millions of people every day” (

39. Constructions is available at

40. Linguistic Issues in Language Technology can be accessed at

41. Semantics & Pragmatics can be accessed at



J. Velterop, 2003. “Should scholarly societies embrace open access (or is it the kiss of death)?” Learned Publishing, volume 16, number 3 (July), pp. 167–169.

M. Waltham, 2006. “Learned society business models and open access: Overview a recent JISC–funded study,” Learned Publishing volume 19, number 1 (January), pp. 15–30.



Contents Index

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License

DiPP and eLanguage: Two cooperative models for open access by Cornelius Puschmann and Peter Reimer
First Monday, Volume 12 Number 10 - 1 October 2007