The UK Patent Office has published the Copyright and Rights in Databases Regulations, 1997. The consultation period ends on 30 September. The legislation will implement European Union Directive 96/9 on Databases, which is due to come into force in all EU countries on 1 January 1998. The Directive reserves the matter of "fair use" to national legislation.

The definition of "database" is very wide, and the implications are disturbing, especially for "full and open access" to data. There is a risk that the UK legislation will make NO "fair use" provisions unless scientists and educators argue for them - via ANY route they can think of to influence the DTI and Patent Office.

I have good reason to believe that the Patent Office has recently hardened its line on this matter, possibly at the behest of the European authorities who wish to promote the "data market" and appear to pay no heed to cultural questions. The Patent Office's "consultation paper" asks for quantified costs and benefits of having, or not having, exceptions to copyright. This request loads the dice in favour of "no exceptions", because the adverse effects will not be highly visible and so are hard to quantify.

It may not be difficult for the pro-copyright lobby to devise figures for the extra revenue that would accrue from having "no exceptions". It is much harder for teachers and scientists at the workplace to estimate the real cost of having to buy licences for each and every piece of data from proprietary databases. (It is easy to conceal £50's worth of work needed to process a payment for £10's worth of data. Accountancy does not solve every problem. Data managers know that "free of charge" is really the most cost-effective way of supplying data).

Seen from high up, the imposition of charges for data may seem inconsequential. Seen from the workplace, the effort and expense saps morale, slows the work, and accentuates the trend towards short-term rather than long-term research. The likely consequence is that good data will not be used, so research and teaching will suffer (as in the debacle of the privatised Landsat data, which ceased to be used because of high charges imposed by the contractor). The case for fair use must be strongly made.

I suggest the following objectives:

1. To ensure that past, present and future data, of kinds currently in the public domain, continue to be fully and openly available at minimum cost and formality. This applies especially to data created through public funds.
2. To press national governments for reasonable "fair use" provisions for scientific and educational use of data, that are sufficiently wide-ranging to allow the collation and assembly of data from diverse sources, in the way that is so important to science and education.