Representing legislative Rules as Code: Reducing the problems of ‘scaling up’

Abstract

We propose an approach to analysing the nature of existing legal rules, particularly legislative rules, that regards legislation as fundamentally a set of related propositions. We propose a method for representing these rules, using a coding language developed for the project that uses a quasi-natural language representation. It enables the interpreter program for that coding language to process these rules so as to produce ‘consultations’ to determine the values of goals which the legislation is capable of determining, with dialogues and explanations generated ‘on the fly’.Progress that has been made in automating this coding process to create rules in that coding language directly from existing legislation, using a pre-processor program developed for the project. This can also be described as ‘scaling up’ the production of ‘Rules as Code’ or ‘Law as Code’. If successfully developed further this has potential to make a significant contribution toward realising the practical potential of Rules as Code.We conclude there is now evidence that these processes can be generalised (‘scaled up’) to deal with the conversion or production of large bodies of legislation, and that this has considerable value. The pre-processor software is evolving rapidly, in the variety of structural forms of legislation that it can convert into the coding language, and this work will continue.Based on this experience, we also demonstrate how the drafting of legislation could be changed so that appropriately drafted legislation is directly readable and understandable by humans and also directly usable by machines. To be effective, laws drafted in this way will need to be simultaneously authoritative legislative rules and code.