To achieve its objective, the Action will pursue the following main tasks:
establishment of a core of Open Scholarly Communities on the Web and of an appropriate academic and legal framework.
development of a distributed Web platform;
research on functional programming for complex interactive Web sites.
Starting from the model proposed by the pilot project, HyperNietzsche, the Scholarly Communities will provide the necessary specifications for constructing Hyper-Learning platforms that respond effectively to the specific needs of Humanities. Simultaneously, they will make it possible to test the systems in sizeable field experiments and to conduct hands-on, concrete evaluations of the scalability of a distributed architecture in at least a dozen different nodes. The creation of such a technological structure presupposes a profound understanding of the activities of teaching and research. Furthermore, it requires having available a critical mass of teachers and students working together from the beginning on a certain number of scholarly enterprises in order to test the scalability of the pilot project. Building Scholarly Communities in conjunction with the software platforms is also the only way to insure that these new technologies will have an impact on the actual practices of research and learning.
The Action plans to establish Open Scholarly Community on the
following authors: Nietzsche, Beckett, Eminescu, Konstantinov, Leibniz, Proust,
Schopenhauer, Wittgenstein, Woolf, Euripides, Puccini, Braudel. Each Scholarly Community
must attain a critical mass that will establish it as an indispensable research and educational
resource within its field of study. This will be possible if the Hyper-Learning platforms
succeed in integrating access to primary sources with the publication of top quality
scholarship, and if they are overseen by an editorial board of respected and established
scholars. The entire group of Scholarly Communities together must in turn attain a critical
mass relative to the humanities as a whole. In this way, and thanks also to the free distribution
of the software needed to activate a new node in the network, this group will act as a
multiplier and will have a structuring effect on the European Research Area.
The technology development component of the Hyper-Learning model will be implemented as:
A contextualized and highly accessible local data repository model for the storage of Research and Learning Objects (Hyper-Learning Platform).
A semi-decentralized peer-to-peer network of XML based web services and tools collaborating in a virtual, distributed and semantically structured Hyper-Learning
Open source, standard, software (Hyper-Learning Server) that allows anyone to easily install, configure and maintain a node within the Hyper-Learning Network.
In order to have a practical and effective impact on education and research, the Hyper- Learning model must be so conceived as to be easily generalisable for other authors or other fields of research. From a methodological point of view, this means taking particular care in order to guarantee simple installation, configuration and interoperability. Such a system would allow any researcher who has a computer, access to the internet, and a clear knowledge of the relevant subject matter to install and implement a Hyper-Learning Platform tailored to that particular field of research. No advanced technical competence would be required. The Action plans to use a technology designed to enable people with no technical training to install a Linux server. In addition to the Linux kernel and basic packages, the Hyper-Learning Server will come bundled with a certain number of other software programs, all of which are available at no charge and are freely distributed in accordance with the Open Source legal model. Once this software is installed, a module is launched which provides step-by-step configuration for adapting the general Hyper-Learning framework to the specific field of study chosen by the scholars.
To preserve the rigorous structure of a Hyper-Learning system while making it flexible enough to be adapted to many different objects of study and easy enough to use that it will be widely adopted by humanities researchers the various Hyper-Learning nodes must be developed in such a way that they can communicate easily with one another. For example, if Schopenhauer is cited in an essay on HyperNietzsche,the reader should be able to move from HyperNietzsche to HyperSchopenhauer with a simple click of the mouse, and so have immediate access to the original context of the passage from Schopenhauer, translations of the passage in different languages, and relevant commentaries from Schopenhauer specialists. This is referred to as crossing hypertexts.
Proceeding from theoretical elaboration and empirical verification, the Hyper-Learning
systems will experiment with a different model for organizing humanities scholarship and
communicating the results, founded on a new system for the production, evaluation, and
sharing of academic work. This system will be decentralized, cooperative, and cumulative, it
will be managed by the scholars themselves, and it will be open to anyone who is interested.
The Action will also promote research in computer science: the development of a functional programming language, that will be used first for the development of the web services of the Hyper-Learning Network. More generally this language, accompanied by its integrated programming environments, will permit the application of the functional approach to the development of all kinds of web applications. Functional programming is a programming paradigm that emphasizes the modular decomposition of a program into mathematical functions from arguments and initial state to results and final state. Unlike conventional imperative programming, which works by elementary modifications of the whole machine state, functional programming emphasizes high-level, abstract descriptions of the computations to be performed, and makes state modifications more coarse-grained and explicit (as functions from the old state to the new state), as well as more local (by explicit specification of the parts of the state relevant to the computation). Combined with the fact that functions are themselves first-class citizens and can be manipulated like any other data by other functions, these features of functional programming greatly enhance the modularity and compositionality of large programs, ensuring non-interference between unrelated program parts.
Functional programming is particularly well suited to the manipulation of complex, treeor graph-shaped data structures. Transformations on these data structures are expressed concisely and precisely as recursive traversals combined with high-level pattern-matching notations. Data structures are naturally immutable, implying that transformations do not modify the input structure in place, but reconstruct a fresh structure as result. Besides matching mathematical specifications more closely, this approach also supports safe sharing of data, which is crucial for complex cross-linked data structures.