Out of the box

Abstract

An overview of the way in which computers and computing came to be applied in a legal context, and the ways in which they developed.

Notes

1 A drum.

2 Of 36 bit bytes.

3 At 1961 nominal values. A starting annual salary for an assistant lecturer in law was approximately the same as that hourly sum at around that time.

4 This can be dated from the seminal paper, A Turing ‘On computable numbers, with an application to the Entscheidungsproblem’ Proceedings of the London Mathematical Society Series 2–42, 17 November 1936.

5 In this country in the form of LEO (Lyons Electronic Office) with which the author worked as a student in 1956.

6 Which are essentially narrative indexes to their footnotes.

7 The most highly developed of which was the West Key Number system for the law of the United States.

8 ‘Automation in the legal World’ Symposium on the Mechanization of Thought Processes, National Physical Laboratory, 1958.

9 Later to become a Conseiller d'Etat.

10 Applying Boolean logic.

11 Like Mehl's.

12 Like Horty's.

13 Tools had however already been developed for use elsewhere, as documented in the UK by ASLIB (Association of Special Libraries and Information Bureaux).

14 The documentary unit was the section or clause, rather than the entire statute. This in turn was probably dictated by the pattern in the USA of codification of statutory materials.

15 On basis of the draftsman's slogan of ‘one meaning, one word; one word, one meaning’.

16 Precision ratio.

17 Recall ratio.

18 Statutory material is in fact more difficult to integrate into a working system, at least in England, owing to the phenomenon of repeal and amendment, and manipulation of when, and for how long, it is to be in force.

19 Supported by the Department of Scientific and Industrial Research (DSIR) to become a little later the Office for Scientific and Technical Information (OSTI).

20 Supported by the American Bar Association.

21 As early as the 1890s the West Publishing Company was using errors in its published work as an evidential basis for claims of infringement of copyright, and there are examples of the same document being entered more than once into West's Corpus Juris Secundum by different indexers who ascribed different key numbers to them.

22 In England the experiment extended also to comparisons at three levels, index terms (telegraphic abstracts) and narrative summaries (headnotes) as well as the full judgements; and to comparison between performance on a general database (the All England Reports) and a specialised one (Industrial Injury Tribunal Decisions). For fuller accounts see C Tapper ‘Feasibility study on retrieving legal information from two types of natural language text’ OSTI Research Report 5062, 1969; Computers and the Law (1973) Ch 6.

23 The research in the USA foundered because it required the assessors to use five different degrees of relevance, with the result that there were serious discrepancies.

24 By Professor Herbert Fiedler.

25 Norwegian Research Centre for Computers and Law, directed first by Professor Knut Selmer and then by Professor Jon Bing. This Centre has published a series of books and pamphlets on research into automated legal information retrieval as well as the law as it relates to computers.

26 For example the US Air Force implemented a system called FLITE (Federal Legal Information Through Electronics). At this time creative effort was extensively expended on the production of acronyms.

27 Both heavily influenced by Mehl.

28 OBAR (Ohio Bar Automated Research).

29 Mr Jerry Rubin.

30 By the Mead Corporation (a large paper and packaging company located in Ohio).

31 Under the auspices of Butterworth, the legal publisher.

32 First in Canada, which developed its own home-grown system, and then in Australia, where an English system, STATUS, developed independently at the Atomic Energy Research Establishment by Professor Bryan Niblett was used as the search engine for the first such system, as CLIRS (Commonwealth Legal Information Retrieval System).

33 In the USA, LEXIS and WESTLAW (a free text system eventually developed to replace the original key word system); in the UK, LEXIS and CONTEXT (a system based on the STATUS search engine).

34 CONTEXT had to be taken over by Butterworth.

35 In sharp contrast to the older methods of batch processing which still figured in some of the earliest implementations. At one meeting it was objected that complaints about slow response times were misguided as lawyers in one such jurisdiction were quite content to wait for up to two weeks for a response from the database.

36 Not entirely disinterestedly of course, because it was hoped that a habit of using developed as a student would be carried over into practice where it would be charged for at the full commercial rate.

37 It was supplemented by the application to it of a system of grouping documents known as cluster analysis.

38 See B Niblett and G Boreham ‘Cluster analysis in court’ Criminal Law Review p 175, 1976.

39 Some other systems have statistical techniques for achieving this, but they do not work very efficiently, and have never figured prominently in any successful commercial systems.

40 The theory is more fully explained in C Tapper ‘Citation patterns in legal information retrieval’ Datenverarbeitung im Recht Vol 5, No 3, p 249, 1976; and full experimental results reported in C Tapper ‘An experiment in the use of citation vectors in the area of legal data’ Complex Vol 36, No 9, 1982, p 1, Norwegian Research Centre for Computers and Law.