JILT 1997 (2) - Philip Leith (1)
The Communication of Legislative Information in Ireland[1] Contents
AbstractThough computer-based information retrieval of legislation has been available for around three decades, it has been an area where little analysis of user need has been undertaken prior to implementation. This has meant that users frequently complain that access to law is difficult and not directed to their particular needs, or ways of accessing legal information. It is therefore clear that to ensure user-acceptance of legal information systems it is important to analyse need and meet this need. The research described in this article looked at user need of legal information in Ireland, a country which has had, to date, substantial problems in accessing its law. The research is tied to the implementation of a computerised Statute Book funded by the Office of the Attorney-General, which is to be produced, and was funded by that Office. The research is novel in that it attempts, before implementation of a system, to assess user need, but importantly, it also has more general lessons for implementors of legislative systems in other jurisdictions. The article looks to the various needs of users, their present access methods and the potential ways of supporting that need through computerised access methods. It suggests that the most important element in the design of a computerised Statute Book is that it should offer - and be linked to - a coherent view of the law.
This is a refereed article published on 30 June 1997. Citation: Leith P, 'The Communication of Legislative Information in Ireland, 1997 (2) The Journal of Information, Law and Technology (JILT). <http://elj.warwick.ac.uk/jilt/cal/97_2leit/>. New citation as at 1/1/04: <http://www2.warwick.ac.uk/fac/soc/law/elj/jilt/1997_2/leith1/> 1. IntroductionSince the computer moved away from its earliest role as a calculating machine, it has been seen as a useful tool in the handling of textual information. The earliest experiments in the handling of legislation (and which became the foundation for systems such as LEXIS) promised a world where there would be a levelling of access to legal information - every practice would have, on a metaphorical electronic shelf, a corpus of the law which was both complete and up-to-date. However, some twenty odd years later, the promise has not been fully reached. There have been many kinds of problems: the difficulty of getting a non-computer literate profession to use computer-based tools; the expense of computer-based research (in large part because of low take-up); the low profitability of providing access by computer has discouraged commercial firms; and, until recently the high costs of technology have persuaded governments not to undertake the task themselves. It has also been the case that there has been substantial lack of knowledge about just who wants access to law and in what manner should this access be arranged. This study has been directed towards the question of user need . It has been guided by the belief that while there surely are a number of purely technical reasons for lack of use of law via computer systems (cost, difficulty of use, etc.) there are most probably problems caused by the lack of understanding of the needs of users. We know, in fact, little about the information needs of the legal profession. This study was loosely linked to a proposed computerisation project to provide a computerised legislative database covering primary law in Ireland - that is, legislation. The original plan as conceived has been described in the following:
My attempt in this research has been to try to discover the views of potential users of such a computerised statute book. These users looked at have been, in particular, librarians who support the legal profession and also lawyers in practice. No attempt has been made to gauge public views on such a database, though this is referred to later in this report. Why is this area of interest? Firstly, when so little is known about the problem of what information is actually required by the profession, any conclusions are useful. Second, the problem of difficulty of access to legal information is widespread [3] , particularly amongst 'small' jurisdictions; by looking at the possibilities of electronic publishing in Ireland, broader conclusions can be arrived at which are useful in other jurisdictions. There has, anyway, been little information available until very recently on the economics of legal electronic publishing and it has been difficult for decision-makers to decide upon the utility of these kinds of approaches. Most have left the responsibility for electronic publishing in the commercial hands - and these commercial concerns have rarely been prepared to allow even the most basic of questions to be answered [4] . The main findings of this study generally are: (i) There is substantial enthusiasm for such a project from potential users. This covers all users investigated. All potential users believe a computerised statute book is an essential requirement to remedy a long standing problem. (ii) Such a project appears to offer substantial savings in research time over the current position - particularly with regard to Parliamentary need and government departments who presently have to compensate for the lack of easily available information. (iii) There is generally felt to be a very strong need for a mechanism for linking three types of legislative information: Acts of Parliament, statutory instruments and European Union Directives. (iv) There is a strongly expressed need for some indexing mechanism apart from that provided by searching for terms by pattern matching (as, for example, LEXIS type systems). Usually this was expressed to me as a requirement for an 'index' or thesaurus. (v) There is an expectation that the system will not necessarily be available free of charge - particularly if 'added value' is provided. (vi) There is no clear consensus on delivery mechanism (e.g. on-line, CD-ROM). (v) It is not envisaged by most potential users that the availability of such a system will immediately remove the need for paper-based versions. (vi) some users look beyond this legislative system to one which incorporates case law. In this report we will investigate the current problem, outline possible technical solutions and then look at the views of the various actors who might benefit from such a system. Finally, comments are made on potential computer-based mechanisms for providing legal information to the Irish legal community. No comments are made on the role of copyright in this proposed project, though this is a much discussed topic in other jurisdictions [5] . 2. The ProblemThere is clearly a problem in accessing primary legal information in Ireland. All those interviewed as part of this project were adamant that the current situation was untenable. Part of the problem has been that Ireland's legislative history has necessitated handling of legislation from several distinct political periods. These are pre-Irish Free State, during the Irish Free State, and the modern State. In many ways, though, this is not of major concern to most users of legal information who are more concerned with more recent law. However, there are a number of sources of law which are pre-1922 and which are still of relevance. The major problem has arisen because of the changing nature of the legislative process: •First, an Act of Parliament is no longer a self standing entity: it is frequently amended by other Acts rather than replaced in its entirety. •Second, it is often legislation which has been produced to accord with European law and thus needs to be read in the light of that European text. •Third, we have seen a considerable change in how government views legislation (and this is to be seen in the UK, too). Often, the legislation is seen as a general framework which will allow the relevant Minister to implement orders, rules and regulations as he sees fit. This usually happens with much less debate in Parliament than occurs with an Act and is thus both administratively and politically convenient. With such advantages these orders, rules and regulations can be expected to continue to be used widely. The difficulty of finding 'what the law is' in this mass of legislative information is large. It is made worse, as we see below, by the lack of reliable mechanisms for searching through this information. The general term for the problem appears to be 'information overload' a condition which has been much used in discussions on the handling of law. The computer has, for long, been seen as a tool to overcome this problem. It is capable of storing, searching and printing information cheaply and easily, once that information is in computer-readable format. There is now a long history of provision of legislation by computer - either privately [6] or, less frequently, by non-profit organisations [7] . These appear to have been successful implementations, though - as discussed below - there are sometimes problems when viewed from the perspective of the user. In the particular context of Irish legal information there are several fundamental problems: First, is that the statutes are not readily available unless one has access to a good law library. Bound volumes are available (comprising the previous year's Acts) and individual Acts can be purchased. Sweet & Maxwell's loose-leaf Irish Current Law statutes Annotated aids the problem of the lack of a good official index somewhat, is but is not seen as the total answer to the problem. Electronically LEXIS NEXIS does not provide files of Irish primary and secondary legislation. Second, is that the last consolidated keyword based index of the Acts only covers the date to 1985. O'Malley (1993) suggests that, 'There is supposed to be an separate index in pamphlet form for each year thereafter, but the most recent is for 1987'. This means that Acts which came onto the statute book after 1987 are not easily found. As on respondent said: 'For any statutes after this we have to go through the contents pages of the volumes, which is very time consuming.' Third, the present index to the Acts (a chronological table - rather than a keyword based index - published in 1996) is considered to be insufficiently detailed. Users looking for something which is not in the title of the statute will often not find it. Users may have to go and physically look at several statutes and try to figure out where the required Act is. Fourth, statutory instruments are difficult to either find or to link to Acts. This has recently been improved by the provision of a Stationary Office index, but - as an example of one problem with this - the time periods between publishing of indexes are seen as unhelpful. One librarian suggested:
Fifth, European legislation is not explicitly related to Acts. This means that users have difficulty in determining the European parent legislation. 3. The Importance of Existing Projects in Electronic Legislative ProvisionThere are a number of existing projects which should be considered in any discussion of the provision of an electronic statute book. These projects particularly concern the provision of legislation (and frequently case law and other pieces of legal information) in a cheap and accessible form over the Internet. Unfortunately, in the interviews undertaken for this research it was clear that most potential users had not interacted fully with such systems. They were certainly aware of these available systems, but few were able to suggest using such approaches in Ireland. The particular projects which are immediately relevant are those of the Legal Information Institute (LII) based at Cornell University in the US, the equivalent system jointly operated by The University of Technology,Sydney and The University of New South Wales in Australia (AustLII) and HMSO's on-line publishing in the UK. The first two seem to be directed principally by legal education's need for cheap and accessible legal information, whilst the latter is in response to pressures within the UK to provide public information without cost. The forerunner in the provision of legal information was the LII at Cornell. Legal information was published on the Internet and could be read by the first 'World Wide Web' browser ( Cello , which was the first of its kind, now superseded by commercially produced programs such as Netscape ). Of course, the information available to law schools and law students in the US is superior to that elsewhere: unlimited access to LEXIS and WESTLAW, are not the norm in Ireland. This has meant that LII has seen its task as not simply providing legal information in the gaps left by these organisations, but in providing legal information within a coherent structure to its users. As the LII states:
LII's formative work encouraged the development of other similar agents: for example, AustLII. In a field where there has been a substantial amount of secrecy due to the commercial nature of much of the work, it has been difficult for the academic to gain any real insights into electronic publishing in law. This has changed with the existence of AustLII [8] which through both the provision of a service on the Internet at www.austlii.edu.au and the dissemination of information about techniques used has provided much information on state of the art electronic publishing of law. Their aim has been to, "convince governments, courts, law reform bodies and other publicly-funded organisations to make legal materials they control available free on the Internet". Importantly, this task has been done through the provision of a large and useful practical system, rather than through a small-scale demonstration system. It is not the case that setting up AustLII has been without legal problem. Bannister refers to the difficulties caused by copyright in official legal materials (Bannister J, 1996) where some States have not allowed these to be made available. She reports that the existence of this problem, 'has necessitated extensive negotiations for the inclusion of materials on these sites.' There is some suggestion that the reason for the hesitancy in allowing use of legal materials is caused by the desire of legislatures to ensure accurate and authoritative publications. Picciotto has suggested several techniques to overcome this worry (Picciotto S, 1996) . Finally, the UK experience in HMSO has provided a model of governmental use of WWW in the provision of legislative materials. No attempt has been made to add 'legacy' material [9] to the HMSO site, but all Acts from 1996 are to be held on this system and all Statutory Instruments from 1997. The material is held in a formatted manner and can be printed off locally to appear like the original Act. 4. Access ModelsBy access model, we mean the way that the computer is used to handle legislation. This model of usage has a strong effect upon how useful any computer system will be. This is not surprising, given that any computerisation project involves substantial changes in the way that tasks are undertaken. It is impossible to insert a computer into a system without there being some radical change in that system. Sometimes the change is for the better, but unfortunately, if the model is not appropriate then the change can be for the worse - there are many systems which have been badly designed and have been switched off soon after implementation or not fully implemented at all [10] . Importantly, note that these models are not mutually exclusive. They can, if so designed, all be combined successfully to meet a variety of user's needs. 4.1 The Searching ModelThis is the earliest model of computer access to legislation. It is typified by the LEXIS-type approach, where there is a large database where each word in the database is indexed and the user can search, usually via 'Boolean techniques', for given words or phrases. A typical search might use one of the connectors: 'AND', 'OR', 'NOT', the pattern symbol '*', and a distance operator 'W' to produce a 'search query'. For example, to find a document which relates to the legal problem, 'a computer contract where negligence is at issue', the user must reformulate this problem into a query suitable for searching the database. This might be: computer w/4 contract and neglig* The system will then look for all documents where the word 'computer' and 'contract' appear within 4 words of each other, and where the document also contains a word beginning with 'neglig'. These can then be listed or browsed, on-screen. Despite the power of this technique - and it can be used to speedily search enormous numbers of documents - there are a number of problems which have been found when lawyers try to use these [11] . Some of them are: •The searching is not 'legal concept-based'; •Frequently too many documents are retrieved, and relevant documents are missed. •Lawyers generally find the systems difficult to use, particularly when they are not used regularly. To the academic, and to the legislative draftsman however, such systems are extremely powerful tools for legal research (FR Shapiro,1989) . A number of techniques can be used with these search tools which would be difficult to carry out in any other way: for example, searching for legal usage of words or phrases. Another example, based upon the fact that documents are usually split into 'segments', each segment containing one type of information - say title, or date of enactment or suchlike - means that the searcher can apply his query to only those segments which are relevant. Most users even if they use the search model, though, do not use these advanced techniques. [12] 4.2 The Publishing ModelThe publishing model is one which suggests that the primary role of the computer is to store and enable access to documents. Here, rather than in the search model where the user is perceived not to know which documents are required, the user does know (roughly at least) what he is searching for. The computer thus becomes a means of dissemination, rather like a book but with different facilities. In the earliest days of handling law by computer, there were two approaches: the full-text and the abstract approaches. As their names suggest, the full-text contained the entire text of the document whilst the abstract approach contained only a description of what was in he document. The former (evidenced by LEXIS) was believed to be superior by users over the abstract approach (used by WESTLAW) and the latter approach was dropped by WESTLAW in favour of a full-text approach. Why should this have happened? It is argued that for most lawyers, once a document has been considered to be relevant, it is essential that the document can be accessed easily. With the abstract approach, there is an implicit assumption that the lawyer has access to a library where the text can be got. In practice it was found that even if the lawyer had access to a sufficient library (and most do not) the required document might well be missing. The huge advantage of full-text computer databases is that the document is always available: there is not one copy which can go missing, but a facility to make any number of copies as required. 4.3 The Keyword ModelThe keyword model requires that each document is manually classified and that searching throughout the document space is carried out by these keywords. This technique is most akin to the book model: it is a kind of index of concepts rather than of words. Being 'legal concepts' these are adjudged to be particularly useful to lawyers. The disadvantages of this model are primarily the need for manual intervention, and this manual intervention has to be of a high standard: before a keyword can be attached to a document the contents must be fully understood. This means that legal expertise is required in applying keywords. Also, it is the case that minor sections in an Act may well require keywording (should they be provisions linked to other Acts). However, there is a traditional competence in carrying out this work and it might be expected that those indexes which already exist could be transferred across to a computer-based system. I do not include within this model the index which refers to Acts, cases or statutory instruments. These do not need manual indexing, since indexes to these can be automatically generated by computer-based techniques (see below, in the discussion of hypertext access). 4.4 The Practice Grouped ModelThis is an form of indexing which is an extension of the keyword model. Basically, the difference is that the 'keyword' becomes a higher level term which directs the user to a certain type of legislative document. For example, in a practice grouping related to intellectual property, one would have a list of several major pieces of legislation under the concept 'Patent Practice' (The Patents Act 1992, Industrial and Property Acts 1927-28, The European Patent Convention etc.) and also more minor legislation, SIs and relevant rules of patent examination etc.. There is an obvious notion of hierarchy here - the user is told which pieces of legislation are most important and which are of secondary import in defining the area legally. This model brings an ease of access to legislation. However, it may be that by concentrating upon the general requirements of a practice area, it is possible to omit sections of Acts which are not central to a specific practice area: to replace comprehensivity with understandability. The AustLII practice groupings (called 'practice books' in AustLII's terminology) are applied both by AustLII itself, but can also be prepared by other users of legislation who locate the groupings across different servers. This is, of course, a use of the hypertext model which is discussed next. (In fact, the practice books- though they index AustLII material - are provided as added value by Foundation Law .) 4.5 The Hypertext ModelThe models just discussed are relatively traditional. The hypertext model is one which has grown in importance over the past few years - particularly with the increase in graphical user interfaces (GUIs) such as 'Windows'. It basically is a model which is an extension of the publishing model, but which tries to integrate diverse documents together. This is done by means of links [13] which indicate a relationship between part of one document and another. The indexing is thus from within one document to another. The general term for this approach, where one moves along these links, is 'browsing'. It is the browsing factor which differentiates it from other more traditional indexing forms. A legislative example would be that when one stores an Act of Parliament, one also stores a link to the relevant statutory instruments which are related to that Act. The link is usually expressed on screen as a highlighted incidence of a word or phrase. When this is selected (by a mouse or other pointing device) the related document is accessed and presented to the user. The advantages which are claimed for this model are that it matches how lawyers do legal research (but there is doubt as to the truth of this claim) - moving from one document to another as the need arises. The major problem associated with this model is that, by following links, disorientation can come about. A number of technical approaches can be applied to reduce this disorientation on the user's part, but this usually requires some form of editing of links and thus an applied control which is - supposedly - the antithesis of the hypertext approach [14] . The hypertext model is now rampant. It appears in the on-line help systems for most computers and is the basis of document organisation on the 'World Wide Web', in which documents which are on computers in different countries can be linked together seamlessly. It offers a simple mechanism of finding and browsing documents which is intuitively easy to understand and which those with little computer literacy have few problems in interacting. There are techniques which allow added value to be attached to legislative material cheaply. These are primarily the automatic processing of legislation, looking for textual incidences which represent meaningful information. For example, the AustLII system contains some 440,000 pages of text (around 2.4 Gb of text) which has been processed automatically to produce some 12 million hypertext links. Handling these quantities of information manually would be hugely expensive, yet the use of computer-based techniques has meant that AustLII has been run as a very small scale operation. 5. The User Groups5.1 The Law LibrarianLaw librarians are to be found in the larger solicitor's practices, law schools, the bar library and also in the Judges Library [15] . They basically, in interview, describe similar problems. Users want legal information and usually expect librarians to be able to provide a package of materials which can be relied upon as being comprehensive and accurate. All interviewees complained of the difficulty they found in achieving this aim due to the fragmentary nature of legal information and the lack of a useful and up-to-date subject index. Given that they usually work under some pressure, they felt that much of their time was wasted in compensating for the lack of easy access to the statute book. Librarians are not usually aware of why material is wanted (it was interesting that those in solicitors practices couldn't tell me how the packages of materials they produced were actually used) but they are acutely conscious of the interrelationships between the various pieces of legislation and the lack of a coherent index to these materials. It was a clear requirement that the most important element of the proposed statute book would be a subject index. Without such a subject index, they suggested, the real problem they had in finding relevant sources would continue. The second requirement was that the Acts should be clearly related through some mechanism with the SIs and European Directives. In this they were simply reflecting the needs felt by their users (as set out below). Librarians generally were users of CD-ROMs and on-line sources. When computerised research was to be done, it is usually carried out by the law librarian. Despite their expertise in the use of these tools, there is a general antipathy towards the searching model of locating information. These librarians were knowledgeable about the problems of relevance and recall and were aware of how these theoretical concepts made their practical looking for material difficult. The existence of a law librarian usually implies the existence of a good law library. This means that the main problem is not one of lack of sources but one of being able to find the relevant sources of law in the library. This is why the index based upon a thesaurus was requested by all who were interviewed. Despite the existence of a good law library, though, all complained about books and material which went missing. Usually this was temporary and the material would eventually be found, but sources are generally required with urgency and this caused a considerable trouble for the librarians. The means of overcoming this at present is partly through the close contacts that librarians have with each other (e.g. faxing over a copy of an SI, perhaps) and partly through making use of the Law Society's service (below). The advantages of having material on-line and therefore printable were seen to be large. However, there was a firm belief that printed materials were not yet redundant: that lawyers preferred the original and that the courts would prefer receiving copies of Acts and other material in the original format, too. Only one law school librarian was interviewed, but it was interesting to hear that there was a broadening of interest in legal materials from non-lawyers. In part this was due to cross-disciplinary courses, but also due to statute law becoming much more important to other teaching areas (e.g. business studies). This suggests that there is actually a wider market for statutory materials than might at first be believed. Certainly more and more EU directives are concerned with areas which have traditionally been of only minor interest to lawyers: health and safety, manufacturing, agriculture etc.. It was also interesting that this interviewee clearly linked statutory material with case law. The two were seen - much like the barristerial view discussed below - to be part of one entity and thus should be - if a coherent computerised statute book was to be designed - considered together. A number of in-house computerised databases were being constructed by these librarians and most had plans for more broadly based systems. The existing computer resource sometimes dealt with SIs, unreported decisions (e.g. in the bar library), and other general 'know how'. This indicates that librarianship practice is making a more extensive use of computerised databases than simply pre-prepared CD-ROMs or on-line services. I did not fully pursue details of what was being kept, but it did appear that there was some overlapping of materials as the databases were being prepared as individual solutions to the same problem. 5.2 The SolicitorThere are a large number of practices in Ireland - some 1800 in total, of which around 800 or 900 are in the Dublin area. Over 50% of the these are sole practices, with only around 5 or 6 Dublin based ones being of any significant size. This obviously has implications for understanding the information need of the legal practitioner. There has been little or no research undertaken into the research practices of the Irish solicitor - what kind of research, how frequently etc.. However, there is some research from the UK (For example Campbell's research into Scottish solicitors from the 1970s: (Campbell C.M., 1976)) and there is reflected information from research into barristers ( Morison J and Leith P) which throws some light on this factor. In the academic legal sense, there seems to be little pure research. The pressures upon the small practitioner - constant phones ringing, multiple problems - mean that the lawyer rarely has the time or inclination (except when there is a vocational interest) to keep up to date with everything which is happening in the legal field. The need for legal information is usually 'practical' and what is required is a short article explaining a topic in a meaningful way, rather than access to esoteric cases. It would be sensible to believe this view to be true for the Irish small practitioner. However, does this mean that there is little need for access to legislation? Not necessarily, because there are frequently occasions when solicitors require copies of Acts and other materials. Their clients may well receive summons which cite a particular Act or they may require information found in financial statutory instruments, and this means that legislation must be accessed. There is evidence that this is the case: the Law Society library offers a service to practitioners where copies of legislation can be sent out on demand. The requested legislation can sometimes be specific (say, a section of an Act) or it may be for a package of information - Act and SIs which are related. There appears to be a constant demand for use of this service from the smaller practitioner, which indicates that though they may not (as academics suggest) carry out 'research', they do need access to law. And certainly, it is not always financially feasible to call upon counsel to carry out all tasks, so solicitors will do this themselves. However, the searching model was not viewed as being appropriate here - some kind of thesaurus or index is seen as essential for helping the solicitor to access the law. The need for this access to legislation led to the Law Society publishing bound volumes in the early 1980s. Some 200 sets were sold (at around £800) and these now fetch around £2000 at auction. The popularity of this service suggests to those in the Law Society library that solicitors are aware of the value of copies of legislation, and that they would prefer legislation above case reports, which are requested with less frequency than is legislation. There are also pressures on solicitors practices to gain recognition through ISO standards which include standards of documentation. These pressures, it was suggested, would lead to an increase in take-up of access to legislation - so long, of course, as they were not overpriced for the small practitioner. And it was also pointed out that small practices could tend not to purchase an up-date every year of a product, perhaps only doing this every second year. It has been noticed that over the past two years, small practices are becoming more and more computer literate. There is more use of modems, CD-ROMs and a general appreciation of the use of technology in practice. An indication of this is the Law Society's annual technology conference which clearly demonstrates that the smaller practitioner sees advantages in computerising. Usually, it was suggested to me, small practices are mainly interested in the text of the law in force at present. However, in some areas, such as financial legislation, it is necessary to have information concerning the law at a particular point in time. The Law Society has prepared an annotated version of this material which is currently being published. The situation with the larger practices is different. These fulfil - in the Irish context - the same functions as those met by the large London solicitors practices: advising clients on relatively complicated points of law and also acting as experts in given legal areas of statute law:
Another large practice lawyer indicated that access to legislation was a significant problem and that:
These lawyers echoed the views of the law librarians that access to the basic Act was not sufficient: too much relevant law was to be found in the SIs and in EU directives to be ignored and that this information was hard to find. This was particularly important when they were advising multi-million pound investments in Ireland, since the lawyers acting for these investors usually had much better access to legal sources than did the Irish representative:
Some even suggested that new legislation was difficult to access, due to an inconsistency in Government departmental approach:
The problem as perceived from the solicitor's view is thus a publishing one. Access to the statutes and related legislation is a primary cause of wasted time and effort. This problem of access is both the physical laying of hands onto the document, but also finding out what has relevance. The clear request was for some kind of coherent access to the statute book and related material through some kind of indexing/thesaurus mechanism. The lack of a subject index in the most recent version of the Chronological Table of the statutes was seen to be a major omission. When the problem seems too difficult to solve internally, the usual approach from the firm was to contact the relevant Department and hope that correct information could be gleaned from them. 5.3 The BarristerBarristers have received more research effort, perhaps, than the solicitor and my earlier research findings (Morison J and Leith P 1992) were found to mesh well with the information gleaned here. There is a spread of income which only partly reflects the ease of access to information: not all of those at the top of the profession would use computer-based research tools (such as LEXIS) but those at the bottom would for a variety of reasons (cost primarily, but also dependant upon caseload) rarely use such systems. The bar library allows - for those prepared to pay extra fees - access to a variety of computer-based materials. The bar library, though, has a wide ranging selection of materials and - most importantly - colleagues to whom one can turn for advice. There was an understanding that larger legal publishers had moved into the marketplace and were providing some materials. However, it was generally felt that this was only the 'creamier', more profitable areas (and barristers were aware that profits in Ireland were most probably low) and that this led to a fragmentation of legal information which was itself unhelpful. The desire for easy access to legislation was strongly stated here. The problems which others find in accessing legislative materials are enhanced because (i) they can be the bread and butter materials when running a case and are not just required for informational purposes; and (ii) they are linked to the generally unsatisfactory situation with respect to case reports. It was suggested to me by one youngish barrister that she spent 25% of her time in doing basic research, much of which was caused by lack of indexing mechanisms and most of which could not be charged to clients. The argument being put forward was that a coherent statute book would substantially increase the quality of service to clients because, when under pressure of time, one would ask a colleague for information but if no colleague could oblige, the information simply had to be done without. Whereas the general position had been that printed materials would still be required, one barrister suggested that this was not so: "You don't need the paper". A similar argument had been put to me by one law librarian that having access in the court-room to on-screen legislation would mean that paper could easily be done away with. The barrister believed that much of the reasoning behind the 'original document' argument was simply that users were unsure of change. Another pointed out that most judges didn't require a copy of the original Act - they were perfectly happy with a copy made from Sweet & Maxwell's annotated statutes. The general view of barristers appears to be that, like others interviewed: (i) an index to the legislation is essential; (ii) links to EU and SI materials are essential; However, they required further information (which appears to be required by the judiciary, too [16] ) which was related to the amending of legislation: (iii) the state of law on a particular date would be on the list of 'ideal facilities'; (iv) information on commencement (including amendments and SIs); (v) cross referencing to similar provisions in other Acts; These had been mentioned by law librarians as aspects of the 'ideal system', too. However their value was stressed much more by the barristers interviewed. Not only was this extra 'added value' seen to be important, but the barristers interviewed were inclined to extend the discussion to the need for easy access to case law. They did not take the view that a legislative system should be self-standing, rather that:
And hence that their needs included: (vi) linking to case reports; Such a view is one where the proposed computer system would be seen as a kind of Annotated Computer-based statute book and indeed it was pointed out that barristers like print-based publications which add value to statutory materials, and that they would also like the same tactic to be applied to the statute book on computer. In requesting a much more enhanced, 'value added product' one barrister was emphasising his view that you can't simply take the old content and place it in a new format, since the aim of developing the statute book is not simply to improve access, but to "improve legal knowledge, or improve certainty". His strongly held position was:
But he also suggested that:
However, the barrister who held this position recognised that he was unusual in being so positive towards the use of IT. He did, though, believe that younger members of the bar had a more rigorous attitude to legal research and practice, perhaps, than older members and that this would encourage usage of computer-based tools. The general view of the bar appears to be that while the ideal system is what they want, access to a more limited CD-ROM, say, of legislative materials would be welcome. 5.4 The Parliamentary LibrarianThe Oireachtas Library differs from those other users of legal information. Rather than requiring access to legislation for the solving of legal problems, the library services the needs of legislators. This means that a very high proportion of the working day is spent searching for and providing legislation to the Members of the Dáil and Seanad - I was given a figure of 40% of time spent in this task. It is also the case that this library is particularly well endowed with printed versions of Acts and statutory instruments. However, a recent investigation into the library by the Joint Services Committee in February 1994, suggested that it had been starved of funding for a 'very long period' and that there was an absence of computer-based tools to enable the librarians to provide an effective service. The conclusions of the report were that better facilities should be made available in the general interest of 'Members on a cross-party, confidential basis'. The librarians are on call to Members during the sitting of Parliament. This may mean working overnight and frequently in the evenings, responding to requests for information. Legislators are particularly concerned with the law as it currently is. Their historical interest is not so much in the history of any Act as in the debates which characterised the introduction of the Act. They also require information which is totally correct. This means that the information must be up-to-date. Usually the various government department who have responsibility for an Act are heavily relied upon to provide that correctness. For example, the difficulty of finding all relevant SIs is one which is met frequently (despite the library having a complete set of these) means that the library must go to the relevant department to ensure that the information to be given to a Member is totally up to date:
And EU legislation, too, is difficult to find, though computer-based tools can help:
This reliance upon the department means that sometimes - when that support is missing - there is a lack of confidence in the information being supplied to the Member. When correctness is in doubt, the Members must be told:
The information provided is usually in the form of a 'package' - Acts and ancillary material. This is photocopied and give to the Member. "If there is time and that's what they want and it is necessary" the librarians will provide a guide to the package. This guide may be something simple - a list or tree showing how everything relates to the other pieces. However, often that is supplied by the department rather than by the library itself. Though many of the Members are legally qualified, there is frequently a problem in discovering exactly what are their information needs. Often their knowledge of which Act is required for retrieval is fragmentary and it is not possible for the librarian to provide information without turning that knowledge into something more concrete. The librarians are strongly of the view that word searching is not appropriate for this kind of retrieval, but that some kind keyword or indexing mechanism is required. For example:
In particular it was suggested that:
The Joint Services Committee had suggested that the library should be provided with CD-ROM services. This had been done. However, the view was that this had not been a successful strategy and that the CD-ROMs were used rarely. There were several reasons: •Since current law was paramount, it was essential to ensure that information was up-to-date. CD-ROMs are not as current as on-line sources, so these on-line sources are preferred even though this was a more expensive approach. •They are difficult to learn to use and the library staff don't have the time to learn all the different search techniques involved. •There is still the need to provide the members with the information on paper, so wherever possible indexes to the paper versions are preferred. Although the library has the full text of all legislation and statutory instruments, like other libraries the information goes missing. I was told that a conference attended by Parliamentary librarians from around the world had suggested that their difficulty in losing texts was just as problematical as in practice. The view is that having text on-line, ready to print out when required would be ideal:
Though the customer of the Oireachtas Library is usually a Member, there are other requests made for information, particularly from other Parliaments. It was suggested that there would be an interest in the availability of Irish legislative information from other countries. 5.5 The PublicThe public need for access to legislation was not examined in the interviews undertaken for this report. However, there are some indicators which can be used to suggest that there is such a need. In the Irish context, King and Dennison (1996) have looked at the use of legal information in Public Libraries . They suggest:
However, there is no analysis in this article of the make-up of these users of legislation, but the fact that the Acts are used in a public arena is a clear indication of need. The authors also suggest that the budgets of branch libraries is not as high as those in larger libraries and that 'in some branches library staff have found that the introduction of computers has raised expectations even higher'. AustLII used the Internet Protocol address of users to keep statistics on usage [17] . Every computer which accesses their service can thus be monitored. In October 1996 there were some 80,000 accesses to the system (from at least 1,300 users per day). Generally the break-down [18] of these figures indicates that the general public does not heavily use this system - however some 15% of users are community organisations. It may well be that the general public are accessing the system via community-based IP addresses, due to not having easy access to the Internet as individuals. For comparison, the other indicated figures are:
6. Desired Commencement and CoverageThere was no clear view of when coverage might be required. It was pointed out that in many areas, legislation from prior to 1922 was frequently required:
No-one provided me with a clear view of when legislation in any computerised statute book should begin. There was a realisation that the pre-1922 legislation was sometimes important, but given the other pressures and needs (particularly access to Statutory Instruments) and lack of knowledge about any budget, no conclusions could be drawn. 7. CostThere was a general assumption that the statute book in computerised form would not be available without cost. Lawyers have become used to paying for their legal information and it was not assumed that this proposed project would be any different. Coming from a Law School environment where there is a growing expectation that legal information will be made available at no cost, I found this assumption to be slightly surprising. However, it clearly shows the difference in expectation between education and practice. On the other hand, AustLII has indicated that "The Law Foundation of NSW has made grants to AustLII exceeding $300,000 (about £150,000) for 1995-97 to enable AustLII to provide databases of primary legal materials from all Australian jurisdictions, and is our principal funding source for the provision of primary legal materials." Primary legal materials refers to both legislation and case reports. Other funding sources have been used to carry out research in the field. This relatively modest funding has allowed the service to be offered without cost to users. 8. Physical Publication MediaIt is clear from the information above that most users are particularly keen to see Irish legal information available following a general 'publishing model'. This model should also have some form of indexing available to users. There are two main physical methods of carrying this out: CD-ROM and on-line. Both have advantages and disadvantages, some of which were known to the potential users and some unknown. 8.1 CD-ROMCD-ROM is a well established means of transmitting legal information and is used by a number of publishers at present. It has a reasonable capacity (650 Mb, though this can be expected to increase) and equipment to read the contents are cheap and near-standard on new PCs. The major advantages which were cited by the interviewees were: (i) it is a well-known form of publishing; (ii) it has the psychological advantage of being a product rather than a service. This means that there is no 'clock' running when in use and users are free to access the contents at will. However, the disadvantages cited by interviewees are: (i) there is no standard format of interface to these products; (ii) the contents are usually limited to smallish areas of law, and necessitate multiple CD-ROM's and changing between disks. (iii) the standard of printing from these disks is often low and does not match the original printed page. (iv) the materials are often several months out of date. Generally, it seemed to me that there was an expectation that if the statute book project was undertaken, then it would be carried out using CD-ROM technology. But, it also appeared to me that this was assumed because of a general lack of usage of on-line sources. CD-ROM's have become popular as a form of publishing to publishers because they offer a product which can be controlled [19] and marketed in much the same way as can a book [20] . However, these are primarily advantages to the commercial publisher. To the non-commercial publisher, CD-ROM can be more problematic. The AustLII project suggests that:
It is a common-place that most of the enquires to the help-desks of on-line services are from users having difficulty connecting in to the service (this is particularly the case when the user is connecting modem to modem with the service provider). However, it is clear that providing CD-ROM products will not negate the need for customer support: Greener suggests that with each release of a CD-ROM product, the help-desk of the electronic publisher receives a substantially increased work-load (Greener R 1996) . 8.2 On-line SourcesThere was a distinct lack of expertise in using on-line information sources. For example, due to the absence of Irish legal information on service providers such as LEXIS, few used these resources. Some were aware of the JUSTIS link-up between their CD-ROM and the on-line service (to ensure up-to-date information is accessed) but most did not regularly access LEXIS or any of the Internet-based resources. Those who did comment upon commercial on-line services pointed to the costs involved (which were perceived to be high) and were aware that they should be used in a limited manner in order to reduce costs. The problem of difficult user interface meant that most on-line searches were carried out by library staff rather than lawyers. There was also, I noted, a substantial worry about security issues. While it is now commonplace for solicitors firms to install local area networks, few of these allow direct access from the fee-earner's desk to the outside world. The security threat is seen to be either from the introduction of viruses, or from unauthorised access from outside the firm. Certainly, it is the view of security experts that there is no certain way to 'firewall' a network from the outside world. It was commonly put to me that, should the statute book be available on-line, it was likely that access would be restricted to certain PCs within the firm to ensure a secure environment. One or two interviewees were aware of the Internet and the developments in this area in providing legal information. They were generally positive about this, but - given their expectations that the system would have a cost element - were unsure about using this approach. In many University environments, of course, the WWW interface has become a standard and has shown that users with a low level of computer literacy (e.g. law students) are easily enabled to access information. The advantage of on-line access (via a publishing model) is that the user has immediate access to the entire text of an Act. However, downloading from an on-line source can potentially be expensive (in connect time to the service if commercial) and it may be difficult to achieve speedy responses if providing for a large number of users due to bandwidth problems. It is also the case that the downloaded version may be textually accurate but may not physically resemble the official publication. We deal with this next. 9. The Relationship to a printed version of the ActInterviewees were asked about the need for a physical version of legislation. Most, but not all, emphasised that the existence of a computerised statute book would not necessarily remove the need for a physical copy of that Act. There were several primary reasons for this: (i) Lawyers are used to books and though younger members of the profession may be happy with good quality computer output, older members would not be; (ii) There may be a difference between the printed output from the computerised statute book and the original Act. What would be the status of the printed output in court as regards authenticity? (iii) Judges who requested a copy of an Act during court proceedings may not be prepared to wait for one to be printed out (though this assumes slow print speeds); (iv) It would be possible for users to edit ('accidentally') an Act and remove one or more words to change the meaning of the Act significantly. These problems seem to me to be significant and should be considered as fundamental to the success of the project. As an aside it should be noted that society has been in this position before. When the printing revolution came about, it was the case that early publishers tried to make their printed books look as much like manuscript versions as the they could. It was to be some years before the printed 'structure' as we know it now came into existence. Certainly, the appearance of printed legislation is something which is closely tied to the book. It can, of course, be easily handled by the computer: the page of an Act can be scanned in as an image and stored thus. However, in order to allow searching through the text of an Act a separate version in a searchable format would be required. This would indicate that some indexing between the searchable text and the imaged page is required. The alternative approach, taken by HMSO, is to format the searchable text to appear like the printed version (that is, with side notes etc.) There are radical solutions, of course. One would be to make the computerised statute book the authentic version of the text for publication purposes. So far in the provision of computerised legal information there have been few problems of this sort. There has always been an assumption that the computerised version is an added value product which is secondary to the printed version. It seems to me that IT in law has now developed significantly to the extent that questions of authority and authenticity of digital publications will have to be considered by all those publishing government information. 10. What Form of Indexing?Users were aware of the power of search engines which could look through a large body of text and discover words or phrases. However, they were also aware of the problems of finding irrelevant information and missing information which was relevant. All practitioners and librarians were therefore very keen to have some form of 'indexing' of the statutory material which would allow 'easier' access and access that was more appropriate to their needs. As has already been indicated there are several forms of 'indexing'. They methods discussed were: searching, keywords, practice grouping, and hypertext linking. 10.1 SearchingThis approach was well known amongst the interviewees and also the most disliked. In discussion the problem of finding relevant documents was mentioned: users want certainty that they will find relevant pieces of legislation and the general difficulties of using these search engines militates against this. However, it should be pointed out that the technology of searching databases with this technique is very well known, and also widely used. While interviewees were generally adverse to this kind of approach, they all used it when they used, for example, CD-ROM products. There are - to my knowledge - no databases in existence which do not allow some form of searching. Therefore, it would be expected that any statute book solution would incorporate this form of access as a standard option. 10.2 Keywords:All interviewees who specified that an index to the statutory information was essential, suggested that it should be done by some kind of keyword index. This index would preferably be a thesaurus which used a controlled vocabulary. Acts and other legislative instruments would be coded according to this thesaurus, which would allow all relevant legislation to be quickly found about a given topic. There is difficulty in this indexing process, because it is becoming common for provisions in one Act to affect a seemingly unrelated Act. It is also the case that amendments to an Act or Statutory Instruments would require careful indexing as well. However, it is a difficulty which can be overcome, since it is effectively found in many of the annotatory material which is linked to the statute book. This approach indicates that each document would require to be manually described and coded for future use. Manual effort equals expense, but when this was pointed out to the interviewees, all suggested that it was a necessary expense, since without this means of access, users would not have sufficient confidence that relevant legislation has been found. And since their enthusiasm for the project is related to the ability to access law with confidence, they felt that such a facility was essential. No-one suggested that the indexing process had to be carried out by the Attorney General's Office though it was perhaps implicit that some institutional body would be responsible for any fixed thesaurus. 10.3 Practice GroupingsThis approach, discussed above as being introduced by AustLII, was not mentioned by any of the interviewees. This may well be put down to their lack of information on the technique. Yet it does have potential advantages in the proposed project. For example: (i) It can be carried out as an 'added value' venture; (ii) It looks at legislation in the kinds of context which solicitors, in particular, require; (iii) It can be carried out by experts in a particular field of law; However, as noted above it has the disadvantage of putting clarity and importance of some pieces of legislation before total comprehensivity. This may however be an advantage for some classes of users - say solicitors. It may also be particularly useful when a user wants a broad brush overview of what legislation exists. 10.4 Hypertext LinkingMost potential users were aware of this type of indexing via links. A demonstration project had also been prepared [21] and some of the interviewees had access to this. There were generally positive comments upon the method, and the fact that it was very easy to use. However, that demonstration project was not necessarily viewed as being the best model to follow. The main reason given for the slightly sceptical viewpoint I met when discussing this demonstration project was that it seemed to be a project which focused on the technology more than it focused on information needs. It was pointed out that what librarians (and it was librarians who had seen this) wanted was legal content , not necessarily a fancy front-end to a limited number of Acts. Of course, the interviewees noted that it was only a demonstration system which looked at one small area of law and that their views might be changed if the system was to be of wider content. Certainly, the hypertext approach is used (along with other techniques) in AustLII and the general ease of use of 'WWW browsing' for users is now accepted as fact. It would be interesting to have received views on this system from interviewees, but most had not made use of it. 11. Legislative ReformSeveral of the interviewees were of the view that the statute book was in such a poor state that it would be better to begin from the beginning with a new consolidated statute book and then computerise it:
The history of the legislative reformer has been long, but unfortunately, relatively unsuccessful. From Bentham onwards to the logicians who today want to formalise law into squiggly symbols, there has been a distinct lack of improvement in coherence. If anything, the situation has become worse with the flood of subordinate legislation and European Directives having to be directly implemented in national legislation. This is not to say that no improvements can be made; my own view is indeed that computerisation will have a substantial effect upon how legislation is brought to fruition. However, that is not likely to happen in the immediate term, and it does not seem an appropriate tactic to await consolidation before computerisation: consolidation simply seems to last for a few short months before the process of amendment and change starts again. It is to this process of amendment and change which a properly designed computer-based system should be able to bring coherence. 12. ConclusionIt was clear to me that there was a great deal of enthusiasm for the proposed statute book computerisation project. The legal profession and parliament are demonstrably suffering from access to Irish Law. This suffering is primarily through wastage in time spent trying to ensure that relevant legislation is found and that nothing of importance is missed from the search. Much of the wasted effort also seems to be imposed upon government departments who - as the originators of legislation and experts in what exists - are called upon to provide information on a continual basis. Of course, the accusation also exists that these departments do not make the finding of legal information (particularly Statutory Instruments) very easy for the legal profession. The enthusiasm is particularly for some form of coherent statute book : one which has an imposed order. The preferred imposed order is that of a commonly agreed indexing thesaurus. However, the statute book is not seen as being an independent artifice. It is seen as standing in a closer relationship with European legislation and with other Irish subordinate material. This does not necessarily mean that users want the full text of all this material on-line (though they would like that as part of a 'Rolls Royce' solution) but that the Acts should have clearly indicated connection to these other essential elements of the law. Going beyond these two statements, however, the situation is less clear. There was generally a lack of firm view of the preferred possibilities which could be applied to the statute book to make it usable. For example, there seemed to be a general assumption that CD-ROM was an appropriate methodology but (apart from the 'clock running' reason) no-one posited any clear reasons why this should be used. There was also an unsureness about what might be available at what cost. If cost was to be a factor, it was quite clear that content was more important than form of presentation. And the problems of using paper based output from such a statute book did not seem to have been seriously considered by many of my interviewees. These are all examples demonstrating the need for 'further research'. However, these problems are not major obstacles to beginning the process of producing a computerised statute book. The expensive part of the process is in the producing of the basic machine readable text from the existing statute book. Given that this is done with an element of segmentation (e.g. with markers indicating title, section start etc.) much of the added value can be added retrospectively with very little additional costs. It will also be the case that as techniques develop, that the raw data can be re-used in new ways. And since all text is now produced on computer, future materials will be accessible cheaply without the processing costs involved in the presently proposed project. In conclusion, the availability of machine readable legislation is the basis upon which other developments can be made: without this first step, accessing Irish legal information will continue to be a difficult and time-consuming activity. Finally, it should be pointed out that the it is likely that this project, when implemented, will have implications for other aspects of legislation. In particular, it may well cause a reconsideration of the mechanisms of actually legislating itself. I mean by this that there may well build up pressure for a legislating system which demands a Statute Book which is incrementally consolidated - that is, that the legislating process is in part legislating and in part consolidating. I arrive at this belief because the main request from users is coherence , and even if all legislation is on-line or published on CD-ROM it is still not as coherent as the computer can make it. This is, of course, not a negative conclusion: the history of computing has been that when users find something useful they demand extensions and fuller facilities. The measure of the success of the Irish Statute Law project will be that users will become more sophisticated and more demanding. 13. ReferencesBannister, J (1996) 'Open Access to legal sources in Australasia', 3 The Journal of Information, Law and Technology (JILT). < http://www2.warwick.ac.uk/fac/soc/law/elj/jilt/1996_3/bannister/> Campbell C.M., (1976) "Lawyers and their Public", in Lawyers in their Social Setting, Greener R (1996) 'Arranging Deckchairs on the Titanic', in 3 The Journal of Information, Law and Technology. < http://www2.warwick.ac.uk/fac/soc/law/elj/jilt/1996_3/special/greener/> King, V & Dennison, M (1996) Provision of Legal Information in Public Libraries in the Republic of Ireland - Dublin Corporation Public Libraries and Munster', in The Law Librarian , Vol. 27, No 3. (This issue of the Law Librarian has a number of articles which are of particular interest in this context.) Lloyd M., (1986) Legal Databases in Europe, Elsevier Science Publishers, Amsterdam MacCormick D.N., (ed.) W.Green and Son, Edinburgh. Morison J & Leith P, The Barrister's World, Open University Press, 1992. O'Malley T, Sources of Law: an introduction to legal research and writing, Round Hall, Dublin, 1993. Picciotto S (1996) 'Towards Open Access to British Official Documents', 2 The Journal of Information, Law and Technology (JILT).< http://www2.warwick.ac.uk/fac/soc/law/elj/jilt/1996_2/picciotto/> Shapiro FR (1989) LEXIS - the complete user's guide, St Martins Press, New York. 14. Links•AustLII FAQ - http://www.austlii.edu.au/austlii/faq.html •Foundation Law - http://www.fl.asn.au/ Footnotes[1] Research funding was provided by the Office of the Attorney General but the views in this document are those of the author alone. [2] Consultancy tender document, 22/4/96, Office of the Attorney General. [3] See Leith, P The Computerised Lawyer, Springer-Verlag, 1990 for a general overview of the problem of legal information and the use of the computer in law. [4] For example, it has been impossible to find out who uses on-line information retrieval systems, how often, what they look for etc., because this has been seen as 'commercially sensitive' information. My suspicion is that if the systems were more successful with more users, we would certainly have been given more detailed information. [5] There has been, for example, much discussion in the UK over the role of HMSO's copyright in electronic publications. The situation has resolved itself with HMSO allowing access to its electronic material and encouraging 'added value' products to be based upon these. See the 'Dear Publisher' letter, CO(P) 48/1022, HMSO which explicitly requires that 'reproduction takes place within a value-added context: ie where the Official text has value added to it by compilation with other related text, analysis, commentary, annotation, indexing or cross-referencing ...' (§2.1.2(i)). See the comments from the electronic publisher, Context in Williamson R (1996) 'Revision of Crown Copyright licensing - comments from an electronic publisher', 2 Journal of Information, Law and Technology. [6] There are large US-based services (LEXIS, WESTLAW) or a number of smaller European systems (e.g. Context Ltd). [7] For example, all Norwegian, some US and much Australian legislation is so handled. [8] Though there are other such services - particularly in the US. These all appear to be based in Law Schools. [9] 'Legacy' refers to previously published information. [10] The computer press is full of such incidences. The general term for this difficulty in producing computer systems is called the 'software crisis'. [11] There are a number of US studies from the 1970s, but for the European experience see Lloyd M., (1986) Legal Databases in Europe, Elsevier Science Publishers, Amsterdam which I believe still holds true. [12] See, Lloyd (note 11) on the problems with computer-based legal information systems of this sort. [13] These links are pointers from one location in a document to another. There are a variety of types of links which need not be discussed here. [14] Because the user is meant - in the pure form of hypertext - to have control over which links are followed, not an editor. [15] The Law Librarian, Vol. 27, No3 carries a number of descriptive articles concerning the law librarian in Ireland. [16] But note that no judges were interviewed. Given though the common forms of communication between barrister and judge this common requirement is not surprising. [17] For example, my IP address (in mnemonic form) is qub.ac.uk - this tells the system that I am from the United Kingdom ('uk'), that I am an academic user ('ac'), and that I am based at Queen's University of Belfast ('qub'). [18] These figures are provided in great detail by AustLII. [19] Though with the rise in size of disk storage on PCs, it is now easy to illicitly copy the entire contents of a CD-ROM. [20] However, it has been noted that there are distinct differences between CD-ROM publishing projects and books: for example, CD-ROM products are expected to contain much more information than books, and for an established publisher can damage the market for the same information in book format. See the highly informative article from a representative of Sweet & Maxwell: Greener R (1996) 'Arranging Deckchairs on the Titanic', in 3 The Journal of Information, Law and Technology. [21] This was a project initiated by the Government Supplies Office and implemented by Deloitte & Touche, Dublin. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
