The contributions and discussions from South African participants at the workshop on access to data convened by the Department of Science and Technology in late September raised some interesting perspectives on the practical and policy issues that would need to be addressed if the country were to have a coherent and powerful data management system at national level. I have blogged the findings and recommendations of the workshop in my Gray_Area blog, but here is further detail on South African contributions to the discussion.

Roy Page-Shipp, the consultant who undertook the drafting of a Position Paper on Access to Research Data in South Africa, reported that he had conducted around 50 interviews with stakeholders in to gather information on data management in South Africa. He found major data-sharing activities in some fields, notably spatially related data and bioinformatics, to global standards and globally networked.  However, data preservation in South Africa currently depends upon the passion and commitment of researchers, he argued. These researchers feel a sense of entitlement to their data and want the first go at analysing it, something that all too often leads to research data being locked up and unavailable to other researchers. There therefore needs to be an exclusivity period to allow for analysis and exploitation by the developer, the period differing according to received practice from discipline to discipline.

In South Africa a shortage of funding means that people will conform to the requirements of the funders and this can lead to conflicts as funders drive access requirements. There is still no major inducement for South African researchers to provide access and this is complicated by the diversity of sources for data.

In dealing with access, Page-Shipp warned that there might be unintended consequences arising from the Draft Bill on IPR for Publicly Funded Research and that a balance needed to be created between protection and access through the avoidance of sweeping rules and the provision of hierarchical access.

 A panel of local speakers addressed some of the blockages that might inhibit a vision for more open access to data.  Anthony Cooper of the CSIR Built Environment pointed out that the marginal costs of transmitting data in Africa are very high.  In particular, capacity and skills are needed.  Liz Gavin, from Statistics South Africa spelled out StatsSA’s open access approaches, but warned that bandwidth is a big problem, as well as cost. She said that charging for data has come to be seen as a mistake by StatsSA and more broadly by government. The major impediments are the culture around data and information management in government, where data management does not feature in government officials’ priorities. What is important, she said, is metadata, metadata, metadata ... cataloguing, discovery, utilisation. Michael Kahn of Knowledge Management at the HSRC, pointed to the publish and perish mentality of the South African academic reward system as an impediment, as it drove researchers to the production of journal articles rather than the development and curation of data. He said that because of a lack of adequate local research funding, much South African research was funded by international organisations, with the result that we became the data gatherers, but data analysis happened in the North. There are serious imbalances in global data systems and this needs to be addressed by more effective data development within the country. We are otherwise in danger of becoming the data collector for exploitation by the North, for example in clinical trials, he argued.

In discussion, speakers raised the question of Traditional Knowledge, which is being dealt with by the Department of Trade and Industry, with a Bill being drafted which, it was reported, would advocate perpetual copyright and state ownership. There seemed to be agreement that the question of Indigenous Knowledge Systems would be wrestled with for some time. What is critical, some argued, is to build and maintain a climate of trust with the owners if IKS. Beatriz Torres reported that experiences in Latin America are beginning to impact in this field, making data available in such a way as to prevent patent pirating. In Peru, people are working with this with indigenous people in the Amazon.

Daisy Selematsela reported on the progress made in the establishment of a Network for Data and Curation Centres (NeDICC) , which is being worked on under the auspices of the DST. Its mandate is the formation of partnerships to ensure and enable the accumulation of data. It aims to provide infrastructure and training to build capacity and provide experience; the identification of datasets and the encouragement of the reuse of data and the development of a user community that can rely on proper datasets and reliable information. Curation is needed for a growing volume of and curation infrastructure needs to be driven by domains rather than the institutions. Government support and funding will be needed as well as support from domain structures and institutions. A conference on African digital data curation, under the auspices of the DST and the Academy of Science of South Africa is planned for 13 and 13 February 2008.

In discussion of policy and legislative issues around the copyright in data in South Africa, there was general agreement that South Africa did not want a Data Directive of the kind that has been promulgated in the EU.  Copyright does not protect databases well, Paul Ulhir argued, from the perspective of the numerical data in databases and the numerical constituents of a database under copyright law are essentially in the public domain.  A protection that can be invoked is trade secrecy. If something is patentable, then that protection applies, but in the commercial sector it is trade secrecy that is more frequently used. In the public sector there are periods of protection to allow researchers to publish on the basis of the data. The period differs and is defined by large research programmes or by normative approaches in different disciplines.

Creative Commons licences use IP to allow for some protections – like attribution – and protect the minim rights that the developer want to retain by defining the use by others. Such licences work well, Uhlir suggested, but the problem with data is that copyright does not protect a lot of the numerical aspects and under copyright law this is essentially in the public domain. Since attribution is part of the scientific method and norms, there is in this regard a set of weak legal right with strong normative provisions.  This is not a legislative solution and the workshop agreed that legislation was not what was needed for the management of access to data. Rather, policy decisions had to be taken and an enabling policy environment created to create a normative context for data management. For example, the current reward system for research rewards publication and this needs to open out to include data. The ideal would be to provide financial rewards for the researcher, joked one participant, and give a medal to the university concerned.

There were concerns expressed throughout the workshop about global imbalances and ideological issues around knowledge ownership. Beatriz Torres argued that there need to be different approaches used to protect developing country knowledge. Technology has come to our rescue, she said, with more federated repositories. The national/international dividing line is now more fluid, she said and what comes into a federated database is recognised as South African if it comes from SA and the country gets the credit. What defines ownership?  Data comes with metadata and this identifies the author wherever the data goes to.

The conclusion of the workshop was that the government needs to develop policy for the development, management and curation of data; that the default position for access needs to be open access, particularly for publicly funded research, and that the importance of adequate funding for the development of data resources need to be acknowledged if South Africa is to take its place – as it wishes to – as a leading research country from the developing world.