The focus and distinctiveness of CRLP
Core focus
The core focus of Jersey’s Computer-Readable Legislation Project is – formalising and digitising the logical structures of modern Commonwealth legislative drafting.
Jersey’s Legislative Drafting Office has other ongoing work on technology, and CRLP’s volunteers have projects of their own outside CRLP. To expand a bit on the focus statement, what distinguishes the core focus of CRLP is that we are looking for drafter-friendly non-proprietary ways of formalising and digitising the logical structures that legislative drafters create in their drafts in Commonwealth jurisdictions.
Current priorities
This core leads to more specific current priorities during our unfunded phase.
- We look into what can be learned, from the core work, to promote non-tech improvements in legislative drafting.
- We look into how to make the digitised logical structures usable by humans and **computers –
- By human readers, using tools like the L4 visualiser. This is particularly for lawyers and advisers (CAB or equivalent) but ideally also by citizens and businesses.
- By computer reasoning systems and AI chatbots/agents. We keep an open mind about whether AI does really “reason” better when fed with more structured material, and whether AI works reliably enough as an interface with a reliable symbolic reasoner.
- We look into how to publish the coded/marked-up digital structure of the legislation (via an API and with unofficial status, or perhaps in some other way that Sprind might suggest through OECD). That includes looking at whether it is practicable for drafting offices to publish these more basic digital versions of all new legislation while govt depts choose which items of legislation to turn into more sophisticated encodings (published separately but linked), and others produce fuller encodings tailor-made for particular uses (like integrating into a bank’s onboarding software, or matching with relevant case-law, etc.).
- We look into how to make best use of AI to help with, but not replace, these core elements. Many other people are going to be looking at AI’s application to legislation, so CRLP’s interest is relatively specific.
Distinctive perspective of legislative drafting
In each of these, what will remain distinctive about CRLP’s contribution is that –
- we bring the perspective of Commonwealth legislative drafters, and
- we aim to make our findings clear and useful to that same audience.
This perspective feeds into our rationale for not taking on some of the more technical work that is better done by others in this field.
- We are not holding ourselves out as developers or trying to produce any working software, let alone the possibly mythical one single best software solution. At our most technical, we are producing illustrative demonstrations, to show rather than tell, so that drafters can picture what we are talking about.
- We are not linking ourselves to any particular technology initiative, whether software or AI. We have worked with L4, but also with DataLex, and we aim to maintain relations with Blawx, Rulemapping, Logical English and others. This links to the idea that drafting offices would produce open source encodings that are published for re-use by multiple others in multiple systems, where the more sophisticated technology builds on the simpler output of the drafting office.
What we are not doing
Our priorities also mean we are not doing various other things that other people are doing in “Rules as Code”, computational law, and related fields.
- Drafts of new legislation only – so we are not looking (as a priority) for ways in which the logical structures of already enacted legislation can be digitised. Previously enacted legislation will inevitably contain ambiguities in its logical structure which can only be resolved by courts (or by amendment by new legislation, which can be encoded while in draft). But we will engage with people who seek to digitise past legislation.
- Legislation only – so we are not looking for ways to digitise more than what the legislation says on its face (when read with other digitised legislation and with the construction canons on which the drafter relied). We are not trying to digitise all of the law that applies to a given scenario, by digitising other types of law that affect the impact of the legislation, such as case-law, customary law, constitutions, human rights, international law (& EU law for MSs, & UK law for BOTs & CDs), and general principles. So we are interested in helping people to see what the legislation itself says on its face when read with other legislation read the same way.
- Logical structures only – so we are not looking for ways to digitise the meaning of undefined non-logical terms and the application of those terms to particular fact scenarios. We do seek to digitise the logical structure of definitions (including those applied by other Laws/Acts such as the Interpretation Law/Act), but we do so while being fully aware that all definitions introduce additional undefined terms. Again the idea is to help people understand the legislation itself, as a starting point before the separate task of working out whether it does really mean what it says when applied to a particular person’s circumstances.
For how CRLP fits into the wider Rules as Code and computational law worlds, see our notes on RaC.
If you are interested in volunteering some spare time to work with us on these issues, please get in touch with our team.