White Paper — Open Science in a Digital Republic|
Overview: “science in transition”
Massive validation of these findings
1For the first time in legislative history, the Digital Republic Bill was placed online, to enable it to be publicly discussed and to receive contributions from citizens, between 26 September and Sunday 18 October 2015.
2The initiative greatly mobilised Internet users: when the consultation closed, the counters showed:
3The result of the consultation on the article “Free access to scientific publications from public research” is as follows:
4When the public consultation on the Digital Republic Bill ended on 18 October, the former Article 9 of the draft bill devoted to “Free access to scientific publications from public research” had generated the most reactions among Internet users, with 3 334 votes and 108 proposals for changes, ahead of the articles dedicated to open data, or to the free reuse of data from industrial and commercial public services.
- 1 https://www.republique-numerique.fr/consultations/projet-de-loi-numerique/consultation/consultation (...)
5The proposal that received the most votes was the one by the CNRS Scientific and Technical Information Department entitled “A shorter embargo period, no hindrance to TDM (text and data mining) and no prohibition of commercial exploitation”,1 with 1 633 votes “for” out of 1 749.
CNRS - DIST - Renaud Fabre 30 September 2015 10:39
A shorter embargo period, no hindrance to TDM (text and data mining) and no prohibition of commercial exploitation
1 749 votes
Agree 1 633 (93.4%)
6Roberto Di Cosmo, Professor in Computer Science at Paris-Diderot University, posted the modification receiving the second largest number of votes, with 1 511 votes for his proposal to “Protect copyright on scientific articles, to enable free access to scientific research”, filed 10 days before the end of the consultation.
7Institutions and research organisations have also taken formal positions. These include the National Institute for Computer Science and Applied Mathematics (INRIA), the National Institute for Agricultural Research (INRA), the Couperin Consortium with the Association of Directors of University Libraries (ADBU), the National Union of Higher Education (SNESUP), Pierre and Marie Curie University (UPMC), the Institut National Polytechnique (INP) Toulouse, the Association of French Archivists (AAF), Cairn.info, etc. Their proposals were predominantly the following:
enable free access to the results of scientific research;
reduce embargo periods to 6 or 12 months, or even remove them in some cases, and for others make no distinction between the scientific fields;
enable articles to be deposited in open archives;
authorise text- and data-mining operations or data searches;
guarantee the possibility of exploiting the results generated by knowledge processing.
8The French Publishers’ Association (SNE) and the French Specialised Periodical Publishers’ Federation (FNPS) were favourable to the embargo periods of 12 and 24 months stipulated by the text; the French Professional Group for B-to-B Information and Knowledge (GFII) proposed that they be decided on the basis of impact assessments to be implemented.
9On 18 October 2015 a discussion forum (Gouv’Camp) was organised for the closure of the national contribution platform, to bring the different bill contributors together in working groups. A report presenting a consensual position was placed online by Alain Bensoussan, co-rapporteur with Grégory Colcanap of the “Article 9” group (which has become Article 17 in the latest version of the Bill). This report insists on the fact that:
“scientific texts must become common assets”;
the question of TDM must be addressed;
embargo periods before publications can be made freely accessible must be reduced to a maximum of 6 and 12 months;
depositing articles “in a long-term public open archive” must be encouraged;
the commercial exploitation of the contents of a scientific article must remain open.
10The needs expressed by the researchers and key witnesses interviewed in the context of this White Paper cover the same crucial points mentioned above by the contributors to the platform.
11Official summary. An official summary of the public consultation prefaced by Ms Axelle Lemaire lists the contributions that led to the amendment of the draft bill, including the contribution of the DIST. The government’s comment explains the scope of the amendments made in light of the proposal:
An almost general consensus emerged from the consultation regarding a clear demand to strengthen the rights of researchers to disseminate their work freely, when the work has been financed by public funds. Seeking a new balance between the positions of the different stakeholders in the digital age and the knowledge society, the government has developed the measure in the following way:
• the “embargo” periods, at the end of which authors of publications financed by public funds may, at the latest, make their texts freely available, have been reduced by half;
• if articles are made available by the online publisher free of charge, authors will be able to exercise their right immediately;
• the text also now states that the research data associated with these texts can be reused immediately, and that their circulation may not be impeded at the time the texts are published.
Among the requests made that were not followed up at this stage, it is important to mention the creation of an exception to copyright for the analysis of texts and data for research purposes (“text and data mining”, TDM), which is clearly supported by the scientific community. European law does not currently make it possible to create new exceptions, and the government hopes that this issue can be addressed in the framework of the European work in progress.
12Significant changes have emerged from this national contribution; some important points remain, however, which have not been taken into account in this Article 17 (the notion of publication of the author’s version/of the publisher’s version), or have been postponed to a future legislative text. Proposals for amendments to the text of the Bill are included in Part 2 of this White Paper (“The amendment to Article 17 of the Digital Republic Bill”).
13On 30 November 2015, the French Digital Council (CNNum) issued an Opinion on the Digital Republic Bill based on the outcome of the public consultation.
14The White Paper takes up here the Opinion of the CNNum on the provisions relating to Article 17 of the Bill and shares it entirely.
On the free access to scientific publications and data from public research (open access): Supplement the provisions of the Bill
The CNNum welcomes the limitation of exclusive transfer periods for scientific publications from public research to 6 months for the sciences, technology and medicine, and 12 months for the human and social sciences, by the recognition of a secondary right of exploitation for researchers.
In its report Ambition Numérique (Digital ambition), the Council recommended supplementing this provision with a requirement to make these publications accessible free of charge on an institutional website, in an open journal or on an open archive site. This could be applied to research organisations. The United States have taken this route by stipulating an embargo period equal to one year. The United Kingdom has decided to create incentives for free access by taking open scientific publications into account in the evaluation – and funding – of research.
Lastly, the CNNum welcomes the inclusion of data from public research activities made public legally under a regime of “commons”, within the meaning of Article 714 of the French Civil Code (chose commune).
15The French Digital Council also notes certain shortcomings in the Bill, in particular with regard to the circulation of data and knowledge and more particularly on text and data mining.
16The White Paper is also in line with the Opinion of the French Digital Council on this point and recommends an exception to copyright.
Authorise text and data mining
Text and data mining refers to a series of computer-processing operations that consist in extracting knowledge according to a criterion of novelty or similarity in texts or databases. For example, it enables searches to be conducted for “weak signals” that are difficult to grasp from cursory reading, identification or analysis of reports of failed experiments.
It has been regarded as highly promising for scientific discovery and the development of new knowledge. It should enable research to take advantage of progress in the analysis of Big Data, which is destined to become a major factor of international competitiveness. Ireland, the United Kingdom and also the United States and Japan allow it today.
- the automated searching of texts and data, as an information reading and extraction activity, is a practice that is not fundamentally different from the manual reading of information, which has always been carried out by research;
- copyright, which protects the form of expression and not the ideas, today makes it possible to read and reuse information or data included in a text for which a right of access has been obtained,
the CNNum considers that there is no legitimate reason to restrict this right in the framework of automated processing.
The major publishers that own the majority of scientific publications can today, by means of contracts, prohibit researchers from searching texts and data – in particular temporary copies, which are technically necessary in order to do this searching – even when the researchers have legal access to all of the scientific publications included in the databases searched. This ban is based mainly on the sui generis right concerning databases. This practice therefore requires the creation of an exception to copyright, on the basis of a reinterpretation of the exception for research, identical to the interpretation by the United Kingdom.
Taking into account the limits and constraints imposed by the contractual solutions, the CNNum therefore recommends establishing a real exception to copyright authorising text and data mining.
- 2 Council of State 3-12-2015 Opinion on a Digital Republic Bill No. 390741, page 5. “With regard to t (...)
17While the Council of State in its Opinion of 9 December regretted the lack of impact assessment, in particular on Article 17 of the Bill,2 on the same day the government proposed this impact assessment whose terms reinforce the present analysis.
18The assessment indeed insists that the objectives of such new legislation:
“[i]nvolve opening the possibility of free access dissemination of publicly funded scientific work, upon expiry of what is known as an ‘embargo’ period that preserves the exclusive rights of publishers. At the same time it involves legally securing existing practices in the scientific community that are well tolerated by publishers. The creation of this new right for the authors of the work requires the intervention of the legislator, in order for it to be imposed on all forthcoming publishing contracts. This provision creates new rights for the authors of publications and promotes a new balance in the relationship between researchers and their publishers”;
“[t]he measure also aims to promote and protect the free reuse of research data, from the time they are made public.”
19The impact of the law has been analysed at various levels:
the public authorities: it “promotes better regulation of the costs of scientific and technical information, today widely borne by the public authorities”;
economic and social: “the sharing of research data contributes to economic and social development”;
on research: “open access to publications and the free reuse of research data promote the sharing of knowledge and discoveries, earlier and recent, within the scientific community. It encourages cooperation and interdisciplinarity, limits the duplication of research efforts, and contributes to the overall improvement in the quality of work. It also paves the way for greater account to be taken of the expectations of civil society, promoting responsible research and innovation. Lastly, it will benefit companies looking to innovate, in particular small and medium-sized enterprises that do not have the capacity to invest in research and development”;
on the scientific publishing sector: the impact is very low:
foreign experience has shown “a limited decline in access via publishers’ websites”;
the risk of loss of subscribers is very low;
in French scientific publishing, the impact should be put in perspective since most of the turnover today comes from subsidies provided by the research establishments or laboratories. However, given the French publishing landscape, the government has decided to take into account the concerns expressed by many publishers and journal directors in the human and social sciences by defining a plan for the transition to free access for HSS journals;
on world scientific publishing: it “is today characterised by a high concentration, oligopolistic in nature, around a few international groups” and “scientific and technical information constitutes an exceptionally profitable activity”.
20The deputies have grasped the challenges for public research, not only by validating the principle of free access to scientific data but also by introducing a legal framework for text- and data-mining practices.
21A report on the impact of the principle of free access to scientific data on the scientific publishing market and on the circulation of ideas and scientific data will be submitted to Parliament by the government “no later than two years after the promulgation” of the Digital Republic Act.
- 3 Digital Republic Bill, text adopted by the National Assembly No. 663 on 26 January 2016, http://www (...)
22A new Article 18 bis has been introduced and creates an exception to copyright and an exception to the right of the database creator in favour of text and data mining.3
23A table in the Annex offers a comparison of the different versions of the text of the Bill, from the public consultation to the adoption of the Bill by the National Assembly.
2 Council of State 3-12-2015 Opinion on a Digital Republic Bill No. 390741, page 5. “With regard to the provision free of charge on the Internet of the results of publicly funded research, provided for by Article 14, the Council of State noted that the impact of such a measure on future contracts between publishers and authors is determined by its public policy nature, which can apply only on French territory, whereas dissemination on the Internet has a global effect. It considered that this inconsistency was an obstacle to the adoption of this measure. In addition, the Council of State was unable to retain this provision, nor the one that qualifies the data from publicly funded research as ‘commons’, within the meaning of Article 714 of the French Civil Code, and which allows the free reuse of these data once published, on the grounds that there was no real assessment available of the positive or negative impacts, either legal or economic, that can be expected.”
3 Digital Republic Bill, text adopted by the National Assembly No. 663 on 26 January 2016, http://www.assemblee-nationale.fr/14/ta/ta0663.asp
Table des illustrations
|Légende||Government, 26 September 2015|
|Légende||CNRS - DIST - Renaud Fabre 30 September 2015 10:39A shorter embargo period, no hindrance to TDM (text and data mining) and no prohibition of commercial exploitation1 749 votes59 argumentsAgree 1 633 (93.4%)|
© OpenEdition Press, 2016