Wikisource:News/2019-09/Introducing WikiLaw (3)

Introduction
English Wikisource is the home of several law related WikiProjects. While the goals for each of these projects vary; in general, they seek to improve transcription processes for statutes, legislation, court cases, memoranda, or legal opinions whenever possible. It is the first of these items that presents the greatest challenge for Wikisource contributors. Unlike the 1911 Encyclopædia Britannica, legal codes change quite often with their use as a reference work greatly diminishing over time. Just like a physical compilation of a given legal code, the book becomes outdated the moment a new law is passed.

Now this is where the promise of WikiLaw (3) steps in. Instead of treating statutes as a transcriber would any other text within Wikisource's library, WikiLaw uses the capabilities of an ever changing database to compile legal text. Normally, for works hosted on Wikisource, transcription works through a page-by-page approach. However, the promise of WikiLaw is to be free of that constraint by treating statutes as a needing to be updated section-by-section instead. The end goal of course would be for Wikisource to retrieve these sections from WikiLaw to automatically compile a readable legal text for the library.

Cross-references
Not everything needs to be moved directly from Wikisource to WikiLaw. Court opinions are a particularly good fit as they are on Wikisource. The physical text never changes, so they are easily transcribed using the proofreading process. One may ask how WikiLaw would function without hosting case law then? The simple answer is cross-referencing.

Very little would need to change on Wikisource for how court opinions and memoranda are presented. Case law transcription projects would simply include, using interwiki links, a persistent link when citing statutes. The added benefit to this persistent link is that, though the statute may change, the citation would always point to the text of the statute as it was at the time. Most legal researchers are forced into very inconvenient processes in the hopes of understanding historical court opinions (generally rummaging through physical copies of books at legal libraries). In the future, their first point of reference may simply be Wikisource, the free library that anyone can improve.

Additionally, entries in the WikiLaw database would not be without references back to Wikisource. A given section would likely need a row for easily seeing which court cases have interpreted it. While this is the case for most physical versions of statutes, WikiLaw could simply give a convenient link to Wikisource; so readers would be able to read the court opinion for themselves. Of course, a short summary detailing the relevant effect ideally would be provided in lieu of having to read the entire court case; as this is the case for most compilations of statutes. Regardless, this builds a two-step cycle where pointers to WikiLaw are paid in kind with relevant cross-references back to Wikisource transcription projects and increasing the value of both.



Machine readability
To the left is a depiction of CGS 1-3 at the MJL Testing Wiki for WikiLaw (3). This example section entry for the possible WikiLaw database shows only a fraction of the possibilities. In particular, the entry should contain a minimum of (1) the text, (2) the component chapter, (3) the overarching legal code [for citations], and (4) the source for the entry.

The importance of maintaining a relational dataset for legal codes cannot be understated enough. Were sections ever to be compiled on Wikisource as intended an algorithm would need to be able to read the metadata for the section to understand what goes in what particular order. If a section has the been repealed, then it could be understood that it is not necessary to import that section unless another section erroneously references it.

Like most Wikimedia projects, verifiability remains a crucial aspect of Wikisource. While it once was considered sufficient for transcription projects to simply display a link to its source within the header, this practice has gradually been deprecated in favor of the proofreading process. While this is can be great method for transcribing pretty much any other work in the library, it is a needlessly time consuming process for individual legal sections. Wikidata-style references could be used instead on WikiLaw to greater effectiveness while also providing the same level of verifiability that proofreading has.

Legal library
While the use cases for a project such as WikiLaw are near endless, chief among them concerning Wikisource will likely be the construction of entire legal codes on Wikisource using WikiLaw data. Piece by piece, the data being built on WikiLaw can be called forth via API into a text form on Wikisource. From there, legal codes can be categorized and sorted like any other work on the project. Annotations could be configured to user preference and with each individual section linking back to its entry in the WikiLaw database for verification and improvement from any interested contributor.

The proper end goal of WikiLaw should not be to replace Wikisource. WikiLaw can, and always should, be a free legal repository within the greater free digital library. That is not to say WikiLaw is meant to be a child project of Wikisource. To put it simply, WikiLaw is an interlingual project like Commons and Wikidata. Not discussed here has been the capability of WikiLaw hosting translated statutes side-by-side native language versions much in the way it would historical editions.

Instead, WikiLaw should seek to have a similar relationship to Wikisource that the latter has with Wikipedia. While equals, Wikisource provides sourcing to give Wikipedia greater historiciocity. WikiLaw, in turn, will be able to transform Wikisource into even greater legal research resource through locally relevant knowledge.

Discussion Please feel welcome to share your thoughts on this proposal at WikiLaw (3). All feedback there is appreciated.