At the time this article was written Matheiu Proulx was Director of Research and Parliamentary Procedure at the Quebec National Assembly

Over the past few years the Quebec National Assembly has set up a repository of information on parliamentary procedure named in honour of the first Clerk of Lower Canada's House of Assembly who was appointed in December 1792. This document explains briefly why a data bank was developed and what sort of information it contains.

Parliamentary institutions in Canada and Quebec are based on the British model. Our history in the new world bequeathed to us structures and principles which we have adapted over the years to suit our own needs. Even today we refer frequently to the rules of procedure in use at Westminster and to commentaries on those rules. We also look at the decisions and practices of other Canadian legislatures and from time to time at precedents established in other Commonwealth parliaments.

Contacts between Clerks of these parliaments are one way to obtain information about parliamentary procedure. There is also an incredible variety of books and periodicals available but the sheer abundance of the documentation can become burdensome to keep on library shelves or in filing cabinets. Furthermore it may not be consulted as much as it should if it is not suitably structured. This problem of records management led the National Assembly to initiate a pilot project to be carried out by its Procedural Research Branch.

Problems with storing and locating documentation had been identified: the traditional documentation chain and more precise indexing were no longer enough to keep up with the ever-expanding mountain of documents. Getting information to the people who needed it, often urgently, was slowed down by the time required to trace the relevant documentation. The pilot project proved to be a success, and the Data Bank was inaugurated by the President of the Assembly on April 28, 1994.

The choice of software for the data bank was made on the basis of a recommendation by an in-house task force. The choice of product and of computer environment was made very judiciously and only after considerable investigation of the possibilities. Because the data bank would have to be an effective documentation strategy for the future while solving the problems already identified, the computer support selected would have to be highly efficient at retrieving information, user friendly and adapted to performing within an overall documentation strategy for the Assembly. The software chosen was CDR, produced by a small Quebec firm called CEDROM Technologies Inc., which seems firmly committed to staying on the leading edge of a sector known for its rapid evolution.

CDR is a data-bank software allowing for whole-text searches, in other words a search may be made for any word in a text as well as on the basis of predetermined criteria. Documents entered in the bank can be in either French or English, it makes no difference. CDR comes with what is known as "hypertext-link capability", which allows the making of connections among documents in the data bank. Using hypertext links, a researcher can consult several documents simultaneously, as though he were sitting at a table in a library with books spread out around him. The software has two modules, CDA, which builds the data bank, and CDR, which searches the data bank. The amount stored can be large without slowing down the search time, and not only text but also images and audiovisual sequences can be stored. The support can be the hard drive of a PC or a compact disk (CD-ROM) for wider distribution. The research module is very easy to learn, using as it does a standard Windows presentation. Those, then, are the product's main technical characteristics.

It was using this software that the Samuel-Phillips Data Bank was created and installed. It should be noted that the whole process was carried out by employees of the Assembly. A development team of skilled and motivated people was set up to tailor and install the data bank, rather than having a ready-made system parachuted in by a supplier or a consultant. One of the attractive features of the chosen software was its adaptability: a variety of data-bank structures can be produced from it, depending on the documents and research that have to be accommodated. Having taken out a corporate license allowing for general use of the software, the National Assembly will be able to continue developing computerized records management within other administrative units.

The Samuel-Phillips Data Bank was structured with a view to simplifying searches for the records it contains. It offers 12 categories of document and eight information fields or research keys. The categories might be thought of as the shelves in a library, while the information fields or research keys are used when a researcher wants to track down a particular document. A list of the categories gives an excellent idea of the contents of the data bank. To make it easier for the reader, here is a brief description of the 12 categories:

Procedure and related statutes: provisions of the Assembly's Standing Orders and operating rules, earlier versions of the Standing Orders and a selection of the relevant statutes;

Procedure - English version: English versions of the Assembly's Standing Orders and rules and of certain statutes;

Documents prepared by procedural clerks: all documents prepared by experts with the Procedural Research Branch since its inception;

Doctrine: resumés of articles from selected periodicals and indices or tables of contents from works on procedure or bibliographies;

Decisions - Assembly: decisions handed down by the President of the National Assembly since 1984 (when the current Standing Orders were adopted), and a selection of decisions made before that date;

Decisions - committee: decisions handed down by committee chairmen since 1984 (when the current Standing Orders were adopted), and a selection of decisions made before that date;

House of Commons: Standing Orders of the House of Commons (English and French), Procedural Review and other relevant documents;

Other legislatures: selection of documents dealing with parliamentary procedure in other Canadian provinces and territories;

Other parliaments: selection of documents dealing with parliamentary procedure in other parliaments throughout the world;

Judicial rulings: selections of rulings from various courts having to do with parliamentary procedure;

Quebec legislature - misc.: documents of political or historical than rather than legal interest, documents from other administrative units of the Assembly, documents on parliamentary reform and summaries of and introductions to parliamentary procedure;

Documentary tools: tables of contents of the Standing Orders and operating rules, concordance tables, thesaurus and other technical documents, updated regularly.

Obviously we are dealing here with documents of great interest for anyone working in the field of parliamentary procedure at the National Assembly. If another legislature decided to set up a specialized data bank on parliamentary procedure, it is highly likely that virtually the same categories would serve. The documents have been grouped in this manner both to make it possible to distinguish among them when research is being done and because each document category is suited to a particular approach with respect to the information fields it contains. The more time spent analyzing and classifying at the start by the procedural experts who will be indexing the documentation before it is entered in the data bank, the less time researchers will have to devote to their searches. If the full texts of documents are simply fed in wholesale, it takes much more strategic research to find all - and only - the relevant documents.

The information fields constitute research "landmarks" or keys that can be used when interrogating the data bank. They were not predetermined by the software but were developed during the designing of the data bank to reflect the documents it would contain and special research needs. The Samuel-Phillips Data Bank uses eight research keys. Each of the documents in the bank has information fields added to it linked to the category to which it belongs. Most of the information fields are assigned by procedural experts at the time the documentation is indexed, before it is entered in the bank. These, briefly, are the eight research keys used.

Type of document: to scan one, or any combination of, the 12 categories described above;

Date: to limit the research to certain dates;

Enriched title: to scan words in the title, subtitle, marginal notes, key words and key articles;

Text: to scan all words in the text;

Authors: for a list of speakers, authors of doctrine and drafters of texts in the data bank;

Key words: for an alphabetical list of all key words identified for each document;

References: for file names and other bibliographical references.

A document's relevance is a very important element in documentary research. How does a researcher go about locating only those documents that are relevant to his research, out of the multitude of records in the data bank? It is true that relevance is ultimately judged by the user as he does his research, in relation to the specific question he is asking. But when the data bank was designed, tools were designed with it to ensure at least a minimum level of relevance in the information and responses generally sought and expected. The structure of a record requires that before it is entered in the bank it must have information fields, or research keys, assigned to it by procedural experts. In this way a relative importance is assigned to different words in a text. For example, an article cited in passing in a text would not be as important as an article considered a "key article", which would be a research key. The same applies to key words and so on for all the different research keys. When searching the bank, a researcher can target a question by using one or other of the research keys. In this way he will get a more precise answer than by combing the whole data bank without having made any distinctions. The immense quantity of records that a data bank can contain can become a real problem without a strategy like this.

A certain standardization in the rules of writing is also a factor that can facilitate a researcher's work, by reducing the number of variants that must be considered. A human being searching by concept can recognize the same concept under different guises, but a computer can only search for a string of characters identical to that in the question it has been asked. For this reason, when the procedural experts were indexing the records, they developed a thesaurus that makes it possible to standardize the data bank's vocabulary of key words.

The elements I have described, i.e. the classification of the documents into 12 categories, the development of eight information fields and the standardization of vocabulary using a thesaurus, are all characteristics of the Samuel-Phillips Data Bank. They make it possible to distinguish among all the records in the bank and thus ensure more relevant results when the data bank is being queried. It is still the researcher's responsibility, however, to implement an effective research strategy.

Introducing this technology in the workplace has meant taking a second look at traditional processes and procedures. Change cannot be avoided. Everyone concerned has to be ready to drop old habits and reflexes and adopt new ways of doing things. The new ways may be initially unsettling, but they quickly produce benefits for the organization and prove to be research tools of inestimable value for the researcher.

Copyright Issues

In the traditional documentation chain, with "paper support", we place documents on the library shelves as we obtain them. They are constantly available to people who want to consult them. The purchase of the document authorizes its future use by any interested person. If someone wishes to quote from a document in a text he is writing, the reference to the original document is carefully given. And everyone knows that photocopying the document or reproducing it in any other way is forbidden.

How should we proceed in the case of data banks? You will have noted that certain categories of document in the Samuel-Phillips Data Bank cover documents produced by other legislatures and parliaments. Much of the value of a data bank that is specialized in this way resides in the fact that documents from varied sources can be entered in it. At the start of this article I said that because Canadian and Quebec parliamentary procedures were based on the same principles as the British system, the latter has the advantage of offering a very wide pool of expertise in the area. However, this poses the problem of authorization for reproduction of documents in specialized data banks. We believe that to conform to the law we must request authorization to reproduce documents from the holders of their copyrights.

A large number of the documents that we want to include in the Samuel-Phillips Data Bank come from the federal government and from the legislatures of Canada's provinces. Others come from foreign parliaments, most frequently Great Britain and Australia, plus certain American legislatures. They are extremely interesting from the procedural standpoint. In other cases, the holder of the copyright is an individual or a publishing house. We should in all these cases ask permission before going ahead. We have reached that stage in our introduction of the system, and the next step will depend on the cooperation we receive from the copyright holders. Naturally we hope for favourable responses from them and the greatest possible openness to the enormous potential offered by these new technologies. The future looks promising for research into parliamentary procedure.