Link Search Menu Expand Document

3.2.5 Anticipating legal protection under data- and code-driven normativity

By Gianmarco Gori

As the ability to achieve a mutual understanding of the meaning of a text which unfolds in linguistic interactions between the members of the juristic community grounds text-driven normativity, it also affords to intelligibly translate law into the artificial language of code and data. The acquaintance with forms of expression mediated by text which distinguishes the jurists’ form of life extends to the different forms of inscription and reading afforded by new ICTs.

At the same time, not only code and data are distinguished by specific affordances which affect “what can be said” in such languages, as highlighted from the very first “speculations” on Artificial Intelligence and Law,1 but the very meaning and consequences of “saying something” changes once it is performed through a medium which differs from natural language. This translation implies a paradigmatic shift in the epistemologies and vocabularies which are necessary to make sense of law and which, ultimately, involves the emergence of different normative practices.

When approached from the perspective of the power of naming referred to in § 3.2.4., i.e., the power to establish “what counts as” law, the language of code and data stand out from the affordances which distinguish texts written in natural language. In code and data, indeed, the power of determining the effects of an inscription does not depend on the shared agreement and understanding shared by a plurality of actors involved into a common normative practice, but on the laws of the artificial languages which drive the behaviour of machines. This circumstance is likely to affect the mode of existence of positive law and of the Rule of law and, therefore, demands a rearticulation of mechanisms of legal protection afforded by the latter. In this respect, it is possible to raise a series of questions concerning the forms of normativity fostered by code and data:

  1. How do the artificial languages of code and data affect the form of dynamic protection that law derives from the affordances of natural language?

  2. How do the relations of conceptual determination which underlie what counts as law interact with the relations of causal determination that govern code and data driven systems?

  3. Which countervailing mechanisms can balance the power of determination which is exercised by inscribing law into code and data?

  4. Positive law and physical laws give rise to different forms of expectations: cognitive, i.e., concerning what it is likely to be, and normative, i.e., concerning what should be. The line between the two is, however, constantly redrawn in practice: in which way does the advent of computational technologies in legal practice affect the interplay between cognitive and normative expectations?


  1. B.G. Buchanan and T.E. Headrick, ‘Some Speculation about Artificial Intelligence and Legal Reasoning’ (1970) 23 Stanford Law Review 40, p. 46. 

This page was last updated on 16 March 2023.