Updated: June 18, 2016 (Initial publication: Nov. 8, 2015)



by Marie-Anne Frison-Roche

Etant donnés, Marcel Duchamp, 1946-1966

This Working paper will be used to support an  French written article to be published in a book, set in the media/assets/couvertures/couverture-internet-espace-dinterregulation.0.pdf">Regulations Series, Dalloz Publishing

This working paper was the basis for intervention in the symposium organized by the internet-espace-/">Journal of Regulation, Internet, Space of Interregulation.

View the slides used as support at the conference (in French).



After emphasizing that the concept of "data" is uncertain, the first perspective is to draw the regulatory consequences of the fact that what is often referred to as the "object" of the data item (the person, the company for financial data, the economy for rating data, etc.), is only its source, "underlying", the data which is manufactured by a company : the real objet of the data is  its purpose being the use for which the data is intended. The data is independent its underlying, is consolidated in the affected masses, takes an economic value based on the desires that have users, becomes available outside of time and space in the digital. This implies a specific interregulation.

But the data is also the Janus of digital because new black gold, pure financial instrument, immaterial by nature, the data also keep a record of people, the underlying that would protect, that we would like inseparable, or the structure that one would want legitimately to attack thank to the new mechanism of compliance. This double-sided of the data item leads to shocks of regulatory mechanisms in Internet

In addition, any Internet links back to the user, in whom we would gladly see "The Grand Interregulato" ". But is it so appropriate, legitimate and effective? The "consent" which refers this interregulation provided by the user himself raises doubts. However, as the displaced term of "right to be forgotten" hides a very effective weapon that can strike those who monopolize the data in a digital economy that seems to be in an ante-market mechanism. This regression pulverizes the marketself-regulation itself to replace the legal acts of exchange by connective legal acts, that for now Law and Regulation are struggling to understand, lack of legal qualifications to do so.


Law is a reconstruction of the world through definitions and categories, by words, to which connects the body of rules. There is always a share of invention in Law hand and a fidelity to the concrete world that Law transcribes to better govern this reality. . However, the term "data" is fairly imprecise.



Indeed, "data" are not precisely defined. Even if we take the term "personal data", for example in the French and European Legislation !footnote-208. If one refers to the article of the European directive, the definition refers to any information relating to an identified or identifiable natural person. But the decisional practice of the Regulators has produced a more accurate conception. Indeed, in casuistry developed by the case-Law elabored by Regulators that have always given relevance to the capacity by one who has any information to "recognize" a natural person, that is to say to make a link between this information and natural person he can find and distinguish this person  from other (identification process). More precisely,  Regulators now refer to "personal data" not so much in relation to the person to whom they relate but compared to the use that is made, or for which the "data" has been collected, processed and preserved!footnote-317.


1. The uncertainty of the concept of "data"

What is considered by the legal protection mechanisms, it is primarily the technical capacity to isolate a person in the mass of other people, coupled with the ability to extract information that may constitute even potentially a danger to herself or others. This is why the "Personal dimension", which is for the person that the 'substantial quality' is for the substance in the contract law, can only be understood in relation to the purpose for which the information on this person is collected.

A "data" can be defined more generally as factual elements that a person extracts from the real world, that is to say information!footnote-300.  Data is always a segment of the real, which is a mass. Taking the example of "personal data", it is opposed to the mass of the real, which is contrast "anonymous" because the world can never be reduced to one person. So when the neutralization of the harmfulness of personal data takes the form of anonymisation of data, this is only a return to what characterizes reality: being a magma.

But here we find the distinction difficult and often sought between the "given" and the "constructed" because we know that the information is itself a construction, since it does dut off a part of the reality only for isolate it (e.g. the address of a person). This division is usually operated by somebody (a business, a government, etc.) in order to be able to be juxtapose this part of reality to other realities selected by it (for example at other personal adresses) in a relevant link that will appear only by the use to be made (eg the creation of a list, to send relevant advertising, to monitor a group, etc.).

Science, which is subjective art to draw up information, is a construction of the world. Thus, what is called for under the new term "metadata" is the build of new information systems, and reports it in an old mechanism!footnote-318. One can take the example of the accounting mechanism of "balance", which juxtaposes accounting data. These, juxtaposed in accounting items, then by accounting restatements, end up producing “financial information!footnote-352, which can be said that it consists of metadata which the company and its stakeholders are the sources.

The data don't escape this non-immediacy between reality and the image that gives information. Without joking, it is some building into a « donnée »!footnote-316. This constructed nature of the data comes not only from the fact that information is a construction but also the fact that data has value because inserted in a "database", a "database", whose architecture is protected as a human creation. It is also because the data was influenced by the regulatory system of which it is subject.

your comment