A Future for Very Large Primary Law Databases?

November 1, 1998

There are a number of very large primary law databases, of varying scope andcoverage, in the process of development in the UK at present. These include theGovernment’s own Statute Law Database project. By a very large primary lawdatabase I mean a project whose ultimate ambition may be to include all, or thegreater proportion of, the UK’s primary law. The materials embraced by theseprojects could be most of the UK statute book, with associated secondarylegislation and large amounts of precedent setting case law. These databases arelikely not merely to cover the law as it is now. They will also allow the userto see the corpus of primary law, as then in force, at any given date in thepast. Huge projects in their concept, these endeavours are capable of absorbinggreat resources in their creation.

The Commercial Imperative

The need for and utility of such large databases is clear, should they provedeliverable. The commercial law publishers have traded on their skills inacquiring, categorising, updating and presenting the raw primary material of lawfor many years. No one will rue the passing of cumbersome looseleaf texts iftheir digital replacement will be easier to access and search. The lawyer willbe thankful for greater certainty that the documents searched are complete andup to date. This is not the only advantage: for the first time the results of asearch will reflect the law in place at the time that the events the lawyeradvises on occurred.

None of the commercial law publishers can risk the loss of revenues whichwill follow if they are not in a position to respond to the database offeringsof their competitors, new and old. The Government will be persuaded that it hasa public duty to provide complete and simple access, at the least, to itsStatute Law Database, even if it was originally conceived as an internal toolfor Government use. Notions that the Government should charge the public formaterials that it has already paid to have created should be swept away. Thestrategic commitment of the necessary resources to create and exploit these hugedatabases is inescapable.

Issues that Remain to be Resolved

There are, however, some issues to be resolved about the scope of suchprojects. These may point to an over-ambition on the part of their creators.This over-ambition may result in flawed projects – at least of the huge scaleand universal application that may be contemplated by their sponsors. At theirsimplest these queries stem from the observation that little in the law isabsolute. Lawyers, after all, make their living from interpreting the law fortheir clients. Lawyers try to anticipate the reaction of the courts to aparticular statutory provision. They may even work hard to argue that theparticular facts of a case render previous decisions on a point of statutoryinterpretation irrelevant. Guidance for lawyers in making their interpretationof the raw primary materials is crucial to the success of analyticalpublications from the commercial law publishers.

When contemplating the design of a primary law database it is simple toidentify the raw materials to be included: statute, SI and judgment. Thismaterial is precise and identifiable. The data is publicly available, in someinstances at a price. It is also relatively simple to add the administrativemetadata (metadata is data about data, see Derek Sturdy, `Intranets: HowPublishers can Integrate their Data with Yours”,’ The Law Librarian,vol 28, no 3, p 129). The term includes such facts about the core item as itsprovenance, period in force, repeal or amendment date etc. This is allrelatively simple objective information capable of assembly and organisation bya skilled yet fairly mechanical editorial process. It is also reasonably simpleto extend the scope of the database by adding cross-references from one `gobbet’or item of primary law to other associated primary materials. For example, theSI that brought a particular section into effect, the case that interpreted it,the appeal that overturned or distinguished an earlier judgment etc. All thisinformation is objective. This additional information can be gathered andapplied to the core primary material in a neutral and dispassionate way.

This is not to underestimate the size of the task involved. The organisation,categorisation and application of this material into data fields is a realistictask provided sufficient resource is made available to achieve it.

Can a Government be Self-critical?

It may be at this point that any Government department will feel that itsproject is completed. The next level of material to be added to these very largedatabases will be subjective in nature. Concepts will be applied to the rawprimary material and its associated metadata. The data, with these additions,will not be purely objective. It may be open to Government to provide someexplanation of the thinking that went into the creation of a law. Lawyers maynow refer to Hansard for the debates that went before the enactment. TheGovernment makes texts of its working papers available for comment, prior tolegislation. It is not, however, easy for the state to be self-critical – inpublic at least. It is hard to imagine a Government-sponsored commentary on anitem of legislation which points out the flaws and contradictions that itcontains.

Commercial publishers are, however, free to apply value judgements to theprimary material that they present to their customers, in the form of criticalcommentary. The addition of such commentary and interpretation is a core lawpublishing skill whose roots go back to the medieval glosses.

Such commentary need not be entirely critical. The rubric of a law report,the catchphrases, the headnote, the analysis of the ratio of thedecision are all, in intent, objective and value neutral. Law reporters havedeveloped great skill in making such assessments of judgments, to the point thattheir analysis is accepted as objective. It is when the propositions that emergefrom such analysis are in their turn subjected to interpretation and prospectiveanalysis that problems may emerge. It is a great leap to move from saying that`Case X decided Y”‘ to saying `that because Case X decided Y it can besupposed that a future Case Xn will decide Yn.’ When this intellectual step istaken the links between gobbets of data become highly subjective.

Thus if a hierarchy of information types is identified it moves across aspectrum from the highly objective primary source at the root to highlysubjective argument at its peak (see figure 1).

Primary source > Administrative Metadata > Analysis > Commentary > Opinion > Argument

Most Objective

Most Subjective

Fig 1

Problems Raised by Adding Critical Analysis

Problems for very large database creators can easily be identified the moresubjective the point reached on the range. These problems may not be so evidentin constructing smaller, specialist databases from a particular standpoint. Ifthe database is designed to support a particular activity (such as maximisingclaims on behalf of PI plaintiffs) then the subjective slant is implicit in theproject and will not cause difficulty. However, interpretation of sections ofthe Housing Act may differ hugely between the landlord and the tenant. Thus toextend the universal database beyond straightforwardly objective elements maybegin to defeat the utility of the project by imposing insurmountable strains onthe data collectors or commentators. The needs and approach of every user cannotbe anticipated.

The nature of primary materials has the effect of guiding builders of verylarge databases down this gobbet-based path. It is an approach that may bebeguilingly similar to the experience of those who have constructed very largescientific databases. It offers a structure that will appeal to those who havethe practical task of organising legal data in a structured way amenable to theapplication of hyperlinks and search engines.

Further issues confront the builders of very large databases once the task ofupdating the database is considered. Provided the information contained in thevery large database is kept neutral and objective these need not beinsurmountable. The techniques used to map changes through citators andpublications such as Current Law or Is it in Force? can be appliedsuccessfully to the data. Routines will be developed to ensure that allassociated pieces of information are updated as a data gobbet is amended.

The complexity of this task grows exponentially if a (possibly large)proportion of the material to be amended is subjective opinion. The timescalesfor maintaining the data increase as changes require consideration as to theirimpact on the opinions contained in the database’s links – to a point wherethey may prove unmanageable.

Then there is the fact that the pace of change is not even over time. Thereare periods in the year relatively free of new legislation or case decisions.Other periods are quite frenetic in the amount of legislative and case activitythat could have an impact on a very large primary law database. The suppliers ofsuch databases will need to meet demanding `up to dateness’ targets if theiroffering is to prove fully useful. A management issue lurks in the need to copewith varying volumes of change at speeds that result in acceptable performancelevels for users.

The creators of such databases may presume that their users all requireinformation at the same level of complexity. This may not be the case. The needsof a non-lawyer may well differ from those of a lawyer. The needs of a studentmay differ from those of a senior barrister looking at the same section. Neitheris it true to suppose that all data items in such a large database are of equalimportance. The distraction can be, of course, that to offer total coverage ofeverything is a useful objective in itself. Thus the outcome may well be thateverything gets equal basic, objective, coverage in the database but that aneditorial assessment is then made as to which data items require greatercoverage. Some designers may even go so far as to structure their data in such away that different users may ask for and receive differing levels of commentary.

An Insurmountable Opportunity?

Up to now this article has concentrated on looking at the problems inherentin very large legal databases from the point of the data itself, from the gobbetof primary material that triggered the initial data entry. Users may notapproach the database in this way. Indeed they may not even go looking for aspecific gobbet at all. Many will want to find out what the law bearing on aparticular event or transaction was, is or, even, may be. Whilst the databasebuilder may be looking at particular gobbets of data and considering how to bestbuild them into an elaborate web of metadata and cross-referrals, the user maywant to find all the gobbets which apply to a particular problem. The wordscontained in the data gobbets may not provide appropriate links to the conceptsthat the user wishes to find discussed. Two illustrations may underline theproblem. Consider the difficulties one can have even using quite simple searchterms on the World Wide Web: search results may miss synonyms and includeresults based on alternate meanings for the words researched. Think also of thefrustration a user feels when realising that the indexer to a particular bookwas thinking in different concepts. It can be highly irritating to use an indexthat does not lead you, immediately, to a topic that you know has to becovered somewhere in the work.

It seems to me that these two main problem areas, that of the difficulty ofapplying subjective interpretation to a rapidly changing, vast volume of primarymaterial and anticipating the needs of the individual user, are insurmountableobstacles to the architects of these projects. For commercial informationproviders the creation of very large databases may thus turn out to beinsurmountable opportunities.

A Possible Outcome

What, then, will happen? First, I believe that the arguments now made to theGovernment by the SCL and others are so overwhelmingly persuasive that we willindeed see the creation of a freely available database of primary legislativematerials. The Government will keep this material fully up to date and current.The benefits to the Government itself in terms of internal economies are suchthat any investment in maintaining such a database will be internally offset bysavings in its own operations. Similarly, as has been observed elsewhere, theincreasing use of computers in the courts should lead to even widerdissemination of the raw texts of court judgments than is the case at present.The Internet will thus provide simple access for all to a series of coredatabases. These will contain the very same primary materials and informationthat some commercial publishers are seeking to develop for themselves. The needfor them to invest in this objective material is obsolescent.

Secondly, following from the first development, the commercial publisherswill begin to concentrate their resources on the potential market for primarymaterial by linking their commentaries to the core source. There should belittle opportunity for them to recoup the substantial investment required tobuild anew that which already exists freely. If this were to be the case, thetaxpayer would have justifiable cause for complaint that the Government hadfailed to apply its revenues effectively. Instead their efforts may be devotedto building interpretative and analytical databases to sit alongside the publicprimary data. Hyperlinking to the original text, maintenance of quality andup-to-dateness of their glosses, providing overt subjective opinions from aparticular approach: these are the tasks for which the commercial publishers arebest positioned.

In Where to Publish Law (Aestival Press, Cranbrook 1998) DavidSummers identified 60 law publishing entities active in the British Isles atpresent. The list is far from complete, especially when the WWW dimension isadded to the potential total of legal information providers. Rather thancommitting resources to competing with the Government, in an activity that theGovernment has overwhelming reasons to get correct for itself, thosecontemplating very large primary law databases will do better to devote theirenergies, instead, to developing ways of interacting with the Government’sproject. The potential for development of tools, supplied by law publishers andothers, which will fundamentally change the way that lawyers work is thenenormous.