À la Recherche des Travaux Préparatoires:
An Approach to Researching the Drafting History of International Agreements*
By Jonathan Pratter
Jonathan Pratter has been the Foreign and International Law Librarian at Tarlton Law Library, Jamail Center for Legal Research of the University of Texas at Austin since 1985. He holds a law degree from the University of Nebraska and a Masters in Library and Information Science from the University of Illinois. When not working, teaching, or writing as a law librarian, he can be found studying a new foreign language or taking a long walk.
Published December 2005
Table of Contents
There are at least two good reasons why one would go in search of the travaux préparatoires to an international agreement (and/or ask for the assistance of a law librarian in doing so). Before we go into those reasons, what exactly are travaux préparatoires?
The phrase is of course French and translates literally as “preparatory works.” Synonymous phrases in English are “negotiating history” or “drafting history.” It is better to avoid using the phrase “legislative history” as a synonym. While they bear similarities, treaty interpretation differs significantly from statutory construction (especially as the latter is done in the U.S.). As well, there is another use for the phrase “legislative history” – as a synonym for the ratification history of an international agreement as that takes place in domestic law.
Two definitions from leading texts are:
An omnibus expression which is used rather loosely to indicate all the documents, such as memoranda, minutes of conferences, and drafts of the treaty under negotiation, for the purpose of interpreting the treaty.
[T]he record of the negotiations preceding the conclusion of a treaty, the minutes of the plenary meetings and of committees of the Conference which adopted a treaty, and so on ….
Two further requirements, unstated but implied in these definitions, are made explicit in the award in the Young Loan Arbitration:
It must first be stressed that the term [travaux préparatoires] must normally be restricted to material set down in writing – and thereby actually available at a later date. … A further prerequisite if material is to be considered as a component of travaux préparatoires is that it was actually accessible and known to all the original parties.
The first reason for seeking out travaux préparatoires can be called the interpretive reason. There is doubt or disagreement about the meaning of an international agreement. Those charged with interpreting the agreement -- it could be a court, or an arbitral tribunal, or anybody who is interested in the meaning of the agreement, including scholars -- will want to consult the travaux préparatoires for insight into the “common intentions and agreed definitions” of the negotiators.
The Vienna Convention on the Law of Treaties has a specific rule:
Recourse may be had to supplementary means of interpretation, including the preparatory work of the treaty and the circumstances of its conclusion, in order to confirm the meaning resulting from the application of article 31, or to determine the meaning when the interpretation according to article 31:
a) leaves the meaning ambiguous or obscure; or
b) leads to a result which is manifestly absurd or unreasonable.
So the Vienna Convention assigns a limited, supplementary role to travaux préparatoires. They can be consulted to confirm the meaning, to overcome an ambiguity, or to avoid an absurdity in the plain text. On the other hand, recognized treatises have ascribed a larger roll to travaux préparatoires. Jennings and Watts point out that
“[T]he International Court of Justice and its predecessor have frequently affirmed the usefulness of recourse to travaux préparatoires. … [W]here a treaty has been negotiated with thorough preparation and full deliberation, and an efficient and complete record … has been kept, the value of the travaux préparatoires may be great ….
Daillier and Pellet note that “an evolution can be traced for some years now tending to accord greater weight to travaux préparatoires.”
While 105 states are parties to the Vienna Convention, the United States is not. (Neither is France.) In fact, courts in the United States take a more liberal view of the use of travaux préparatoires than the Vienna Convention does. A recent case in the United States Court of Appeals for the Ninth Circuit, Gonzalez v. Gutierrez, demonstrates this. The international agreement before the court was the Hague Convention on the Civil Aspects of International Child Abduction. The Gonzalez court said: “Although in interpreting a treaty we begin with the text, we may look beyond the written words to other factors for interpretive guidance. Appropriate sources to consult include the purposes of the treaty, its drafting history, and its post-ratification understanding.” The United States Supreme Court has expressed itself even more strongly. “Because a treaty ratified by the United States is not only the law of this land … but also an agreement among sovereign powers, we have traditionally considered as aids to its interpretation the negotiating and drafting history (travaux préparatoires) ….” There is nothing in these quotations about a merely secondary or supplementary role for travaux préparatoires.
There is another reason for consulting travaux préparatoires that has little to do with interpretation as a matter of law. We can call this other reason the genetic reason. There may be absolutely no doubt about the meaning of the treaty text; it is clear to every reader, even to a lawyer. Yet, we may take great interest in how the text of the agreement evolved into its final form. In other words, the evolution of the text has intrinsic historical interest. The examples are of course limitless. One brief illustration will have to serve.
The very first right specified in the International Covenant on Civil and Political Rights is in Article 1(1), first sentence: “All peoples have the right of self-determination.” The researcher will learn from the travaux préparatoires that the proposal to include peoples’ right of self-determination in the Covenant sparked a sharp controversy. Was this a political principle or a legal right? Three contending schools of thought formed. The debate covered a span of years in the drafting of the Covenant. A working party to deal with the question was convened. It made its report, recommending the inclusion of the right of self-determination. In the end, we know what happened, but if it were not for consulting the travaux préparatoires the researcher would have no inkling of the struggle that preceded the adoption of Article 1(1), first sentence of the Covenant.
The dual significance of travaux préparatoires having been established, the question naturally arises (especially for law librarians), “How do you find these things?” The answer to this question is mutifaceted. It turns out that travaux préparatoires come in a variety of guises. The search for travaux préparatoires can range from quite easy at one end of the scale to impossible at the other end. The World Wide Web has made a remarkable difference, but the Web is not a universal antidote to the difficulties of tracking down travaux préparatoires.
I have produced a set of models of publication of travaux préparatoires that helps to clarify the situation for anyone who is about to embark on a search for them:
Model I – Unavailable (nonexistent or inaccessible)
Model II – Collected under one title
Model III – Treaty-specific conference records
Model IV – Integrated with standard international organization documents
First, however, I want to mention two kinds of resources that can immeasurably facilitate research in this area. The first is a guide to the travaux préparatoires that tracks the agreement article-by-article, with references to the relevant places in the travaux préparatoires where each article (or the text that preceded it) is discussed.
This kind of guide is immensely useful. Travaux préparatoires are usually published in the chronological order of their production and do not correlate well with the order of the final text of the agreement. They can amount to a large and complex body of documentation. If somebody takes the trouble to analyze the travaux préparatoires by collating them with the final text of the agreement, an invaluable research resource is created. Given the amount of effort required to produce such guides, not many of them exist. However, the researcher should always check first before launching into her own hunt.
There are few enough of these guides that they deserve to be noted here:
The second kind of resource is an article-by-article commentary on a particular agreement. In the nature of things such a commentary refers to the travaux préparatoires, either for interpretive guidance or historical context. Two outstanding recent examples are Schlechtriem’s commentary on the CISG and Schreuer’s commentary on the ICSID Convention.
I start with the model of unavailability not because I want to make the process of research seem terribly difficult or frustrating, but rather because I want to inject a note of realism and common sense into the endeavor. If it is understood that unavailability is one model of (non)access to travaux préparatoires, it will save the law librarian or other researcher the disappointment of unreasonable expectations.
Many, many bilateral agreements are entered into without a great deal of formal negotiation. What formal negotiation does take place is not recorded, and any record that does exist is probably in the form of notes taken by the negotiators, and certainly is not intended for publication. The obvious conclusion is that in the case of many bilateral agreements, the researcher should not expect to find travaux préparatoires.
A significant multilateral agreement that has presented much difficulty is the North American Free Trade Agreement (NAFTA). For a decade after it went into force all documents relating to the drafting history of NAFTA were inaccessible. That included both the various draft texts and the minutes of the meetings of the various negotiating committees. It is not that they formally were declared confidential; you just couldn’t see them. Then finally the NAFTA Free Trade Commission in the Joint Statement (www.dfait-maeci.gc.ca/nafta-alena/JS-SanAntonio-en.asp) on its July 2004 meeting in San Antonio had this to say:
We are committed to transparency in trade negotiations. The negotiating texts of the NAFTA are documents of historical value and we recognize the level of public interest in them. We asked our officials to compile the NAFTA negotiating texts, bearing in mind the time necessary to complete this. We began the process with Chapter 11 [Investment] and are pleased to announce that Chapter 11 texts will be available through our websites.
Sure enough, you can now find the NAFTA Chapter 11 negotiating texts on the websites of the Canadian Ministry of Foreign Affairs and International Trade and of the U.S. Trade Representative. Now there are only 21 chapters of NAFTA left to go.
Two cases of confidentiality deserve mention. For 35 years the travaux préparatoires of the Treaty of Rome establishing the European Economic Community were declared to be inaccessible. Then, in 1994 they were declared to be open. They are now available in 415 dossiers on 815 microfiches in the Historical Archives of the European Union in Florence, Italy. There is an excellent online inventory of the collection on the website of the Historical Archives. (Click Fonds presentation and then locate CM3/NEGO).
The travaux préparatoires of the European Convention on Human Rights were published in eight volumes between 1975 and 1985. However, the travaux préparatoires relating to all of the other international agreements sponsored by the Council of Europe, and there are about 195 of these, remain confidential. In 1965 the Consultative Assembly recommended to the Committee of Ministers that it should authorize the publication of the travaux préparatoires of Council of Europe conventions. The Council of Ministers refused, but agreed to put in place an alternative. This is the publication of an official explanatory report that accompanies each convention sponsored by the Council of Europe. Not all conventions or protocols have an explanatory report, but most do.
These are easy to find. At the main Council of Europe conventions website, the researcher needs only to click on Full list, and then on the name of the convention of interest. A page “What do you want to know about this treaty?” appears. The researcher clicks on the link to the Explanatory Report, which comes up in full in printer-friendly HTML format. In hardcopy the texts of the Council of Europe conventions together with their Explanatory Reports are available from the Council of Europe Bookshop Online.
The Hague Conference on Private International Law is the best example of Model II. The first Hague conference was convened in 1893. In 1951the Hague Conference became a permanent intergovernmental organization, founded on its Statute, which is an international agreement. The Conference has 64 member states, but 122 nations from all parts of the world are parties to one or more Hague conventions.
The purpose of the Hague Conference is “to work for the progressive unification of the rules of private international law.” The principal method for achieving this goal is the drafting and negotiation of multilateral international agreements. To date the Hague Conference has sponsored 35 conventions on a variety of subjects in private international law.
The working method of the Hague Conference can be summarized as follows. The Permanent Bureau (secretariat) undertakes preparatory work. Then preliminary drafts of a convention are drawn up by special commissions made up of governmental experts. The draft convention is then considered, debated, perhaps amended, and adopted at a plenary session of the Hague Conference, which convenes every four years.
What kind of documentation is produced during the preparation of a Hague convention? We can identify at least the following kinds:
} In one category, the documents of the preliminary work, which include such items as:
o A report drawn up by the Permanent Bureau on the advisability of proceeding to draft a convention on the subject in question;
o A checklist drawn up by the Permanent Bureau of provisions that might be included in the convention;
o The preliminary conclusions of the special commission appointed to draft the convention;
o The preliminary draft convention adopted by the special commission;
o An explanatory commentary on the draft convention drawn up by one of the expert members (the rapporteur) of the special commission.
} In another category, the so-called “Acts” of the session of the Hague Conference that considered the convention, including:
o Working documents of the commission appointed by the plenary to consider the draft convention, including proposals to amend the draft text;
o The minutes of the meetings of this commission;
o Working documents of the plenary session;
o The minutes of the plenary session.
} In a third category, the final text of the convention as adopted and the official explanatory report drawn up by a recognized expert delegate, usually the person who had previously been appointed as rapporteur during the drafting phase.
All of this documentation is gathered up and published in a series known popularly as “Actes et Documents.” If these volumes are cataloged under one title and classed together, as in my view they most certainly should be, the full title will be Actes et Documents de la … Session. The corporate author will of course be the Hague Conference on Private International Law.
Now, the Hague Conference maintains an excellent website. This is a fairly recent development. The question is, what effect has the website had on the availability of Hague Conference travaux préparatoires? The answer is that there has been a highly favorable, but not yet revolutionary, impact. The reason is that the full text of the Actes et Documents series is not available on the website. If we take as an example the most recent convention sponsored by the Hague Conference, the Convention on Choice of Court Agreements (2005) we find on the website an extensive collection of the preliminary documents but not the final explanatory report, or the procès verbaux of the meetings of the sessions of the Conference. For other conventions, only the official explanatory report is available. For some conventions, there are Hague Conference documents about the practical operation of conventions after they have gone into force. My conclusion is that researchers should make full use of the website when researching the travaux préparatoires to Hague Conference conventions, but that all libraries with a substantial collection of international legal materials should definitely have the hardcopy set of Actes et Documents.
The other example of Model II is the International Institute for the Unification of Private Law (UNIDROIT). UNIDROIT also is an intergovernmental organization whose purpose is to study the needs and to propose methods for modernizing, harmonizing and coordinating private law, in particular commercial law. UNIDROIT’s main working method is to draft legal instruments, including international agreements, that harmonize the law in particular fields. UNIDROIT emphasizes that its focus is on rules of substantive law, not rules of private international law, which is what distinguishes UNIDROIT from the Hague Conference.
Much of UNIDROIT’s official work is published in an annual series under one title. The series is known by both its French and English title, Actes et Documents or Proceedings and Papers. This series publishes official documents of UNIDROIT, including the final texts of adopted uniform laws or international agreements, as well as preliminary study documents and draft instruments that qualify for the description of travaux préparatoires. The UNIDROIT series is not as complete as the Actes et Documents of the Hague Conference, because it does not contain the verbatim record of meetings.
The World Wide Web has had a beneficial effect on researching the documentation of UNIDROIT. The documents that compose the annual volumes of Proceedings and Papers are available in PDF on the UNIDROIT website from 1997 to the present. The documentation for 1996 is incomplete.
This is both the most common model and the most difficult to research. When a significant multilateral agreement is to be negotiated, a sponsoring body – it could be a state, say Switzerland; or it could be an international organization, say the United Nations – convenes an international conference to which as many states as possible will be invited to send representatives. The conference elects a president and organizes itself into working committees, usually including a drafting committee. The conference works through and prepares the text of the agreement, which is then adopted in the plenary session of the conference in a document called the “final act.” In the course of its work, the conference produces such documents as procès verbaux (verbatim records of discussion in committee and plenary sessions), summary records of discussion, and working drafts of the agreement. These documents are gathered together and published as the record, that is, the travaux préparatoires, of the conference.
From the perspective of research, the difficulty of working with Model III is that there is no necessary link between the name of an international agreement and the travaux préparatoires of the conference that produced it. The cataloging record of the conference publication does not have to, and frequently does not, contain a tracing or added entry for the name of the agreement. Therefore, a search using the name of the agreement, which is usually what the researcher has in hand, often will not turn up the cataloging record for the travaux préparatoires.
An example illustrates both the difficulty and the method. The famous international agreements negotiated in 1949 for the protection of war victims, are called almost universally in the legal literature by their popular name, the Geneva Conventions. This is the information that the researcher has to start with.
A search on OCLC WorldCat for keywords “Geneva Conventions” returns over a thousand records. A title-keyword search returns a little over 300, which is more manageable. However, the record for the travaux préparatoires is not among them. What does appear in many of the records is the name of the conference, as officially established in the cataloging: Diplomatic Conference for the Establishment of International Conventions for the Protection of Victims of War (1949: Geneva, Switzerland). A search on the terms in the name of the conference quickly reveals the record for the travaux préparatoires under their formal title, Final Record of the Diplomatic Conference of Geneva of 1949. This method of searching for Model III documents may not be ideal, but it does work. The caveat that I would enter here is that for best results the researcher should be a law librarian.
There is a complicating factor that deserves mention. It can happen that the record of the negotiating conference does not represent the entirety of the travaux préparatoires. A good example is the Vienna Convention on the Law of Treaties. The travaux préparatoires of the conference have indeed been published. However, in the decade preceding the decision to convene a conference (1955-1966), the International Law Commission carried out an immense amount of preparatory work on the law of treaties, which led directly to the General Assembly’s decision to convene the conference. Fortunately, this work is thoroughly documented on the website of the International Law Commission. (Click Analytical guide). This pattern obtains for several agreements for which the International Law Commission did preparatory work. The Commission’s website has complete information.
Working with Model IV requires knowledge of the international organization in question and its documentation. Each international organization is unique and has to be taken on its own terms. Of course, we are concerned here with international organizations that carry out law-making functions through the adoption of international agreements. There are several of these. We take the International Labour Organization (ILO) as an example, partly because the ILO, through its outstanding website, has revolutionized research into its travaux préparatoires.
Founded in 1919, the ILO became the first specialized agency of the UN in 1946. The structure of the ILO has three main components. The International Labour Conference is composed of the member states of the ILO. The Conference meets in Geneva in June of each year. The Governing Body is the executive council of the ILO. One of its significant functions for our purposes is to propose the agenda of subjects to be considered by the Conference. The International Labour Office is the permanent secretariat of the ILO in Geneva.
One of the major functions of the ILO is standards-setting on a broad array of questions in the general field of labor and labor rights. One of the chief mechanisms of the ILO’s standards-setting activity is the adoption of international agreements, called conventions. As of 2005, the ILO has adopted 185 conventions dealing with all aspects of labor standards. One of the most significant recent conventions, which we take as an example, is the Worst Forms of Child Labour Convention of 1999.
Once the International Labour Conference accepts a standards-setting issue for its agenda, it asks the International Labour Office to prepare reports for submission to the next session of the Conference. The Conference then assigns consideration of the agenda item to a committee of the Conference. The committee makes its own report, which is considered by the plenary of the Conference. If the proposal for enacting a standard in the form of a convention is approved, the plenary session of the Conference adopts the text of the proposed convention. It should be noted here that the consideration of proposed conventions usually takes place during two consecutive sessions of the Conference.
In hardcopy, the work of the International Labour Conference is published in the annual serial, Record of Proceedings. On the other hand, the reports that the International Labour Office submits to the Conference are published as discrete bibliographic entities. Each has a title and even an ISBN. A header on the title page referring to the session of the Conference, along with a reference to the agenda item, indicate that the document is a report made to the Conference.
I want to propose that today the ideal method for researching the travaux préparatoires of ILO conventions is online. The ILO has an outstanding website with all kinds of information. One of the central components of the website is the database of ILO-sponsored labor standards, called ILOLEX. Of course, a major element of ILOLEX is the collection of ILO conventions. The full text of every ILO convention is easily accessible, along with links to ratification information. However, there is no link from the text of a particular convention to the relevant travaux préparatoires, even though they are available elsewhere on the ILO website.
To get to the travaux préparatoires you have to go to the link on the ILO homepage to the International Labour Conference. With the knowledge that the Worst Forms of Child Labour Convention was adopted in 1999, the researcher selects the link for the 87th Session, which took place in June 1999. There the researcher finds links to a full range of sessional documents. For our purposes, the significant documents are:
} “Reports Submitted to the Conference” (by the International Labour Office). These are organized by agenda item with an indication of the subject, so it is easy to find the reports titled “Child Labour.”
} “Reports of Conference Committees and discussion in Plenary.” These too are organized with an indication of the subject.
Knowing that proposed conventions are discussed at two consecutive sessions of the Conference, the researcher then goes to the link for the 86th Session (1998) and repeats the process. Here I have to enter a caveat. This kind of research using the ILO website is possible only as far back as the 86th Session. However, from 1998 going forward, the website has revolutionized research on the travaux préparatoires of ILO conventions. Already the website is posting documents for the 95th Session of the Conference in 2006.
Having worked through all four models, the observant reader will have noted a small difficulty. When the United Nations General Assembly convenes an international conference for the purpose of negotiating an international agreement, the travaux préparatoires are normally published as stand-alone bibliographic entities, which puts them in Model III. However, at the same time these documents are issued in the A/CONF./[number] series of United Nations documents, which puts them in Model IV. This dissolves the distinction between Models III and IV. In response to this small difficulty, I do not claim that these models represent airtight categories. I do maintain that my four models of travaux préparatoires represent a useful working framework in which to approach the task of researching them.
*This article derives from a presentation delivered at the AALL 2005 Annual Meeting in San Antonio for the program, Navigating the Maze of U.S. Treaty Research. Some references to specifically U.S. situations have been retained.
Lord McNair, The Law of Treaties 410 (1961).
Robert Jennings and Arthur Watts, 1 Oppenheim’s International Law 1277 (9th ed. 1992).
59 I.L.R. 495 (1980).
Id. at 544.
Id. at 545.
Vienna Convention on the Law of Treaties, May 23, 1969, art. 32, 1155 U.N.T.S. 331 (emphasis added). Article 31(1) of the Convention, which contains the general rule of interpretation, reads: “A treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose.”
Sir Robert Jennings, Sir Arthur Watts, Oppenheim’s International Law 1277 (9th ed. 1992).
Patrick Daillier, Allain Pellet, Droit International Public 262 (7th ed 2002) (citing several cases in the International Court of Justice).
This fact has not prevented U.S. courts from citing the Vienna Convention in a variety of contexts. A Westlaw search in the ALLFEDS database returns 73 cases that have cited the Convention. Chubb & Son, Inc. v. Asiana Airlines, 214 F.3d 301 (2d Cir. 2000), has a good discussion of the court’s view of the Vienna Convention’s place. The court said in conclusion that “[w]e therefore treat the Vienna Convention as an authoritative guide to the customary international law of treaties.” See also, Maria Frankowska, “The Vienna Convention on the Law of Treaties before the United States Courts,” 28 Va. J. Int’l L. 281 (1988); Evan Criddle, The Vienna Convention on the Law of Treaties in U.S. Treaty Interpretation, 44 Va. J. Int’l L. 431 (2004).
311 F.3d 942 (9th Cir. 2002).
Gonazalez at 948 (emphasis added).
Zicherman v. Korean Air Lines Co., 516 U.S. 217, 226 (1996) (emphasis added).
Naturally, I did not develop this example from original research. Rather, I consulted Bossuyt’s unsurpassable guide to the travaux préparatoires of the Covenant, mentioned below.
This source is a monograph, but it contains detailed references to the travaux préparatoires on several aspects of the Covenant.
Peter Schlechtriem, Ingeborg Schwenzer, Commentary on the UN Convention on the International Sale of Goods (CISG) (2d (English) ed. 2005).
Christoph Schreuer, The ICSID Convention: A Commentary (2001).
See the extract of the procès verbal of the meeting of the Committee of Permanent Representatives, April 3, 1959 (on file with the author).
Collected Edition of the “Travaux Préparatoires.” The Hague: M. Nijhoff, 1975-1985. The formal name of the convention is Convention for the Protection of Human Rights and Fundamental Freedoms. The convention is frequently cited to 213 U.N.T.S. 221. However, as the convention has been amended several times, the best place for a reliable current text of the convention is the Council of Europe’s conventions website. (http://conventions.coe.int/Treaty/en/Treaties/Html/005.htm)
Recommendation 417 (1965) (on file with the author).
Conclusion (65) 144 (on file with the author).
Statute of the Hague Conference on Private International Law, art. 1, Oct. 9, 1951, 15 U.S.T. 2228, 220 U.N.T.S. 12.
This is obviously a much-abbreviated summary. For a fuller account see Robert Jennings & Arthur Watts, Oppenheim’s International Law 1183-1187 (9th ed. 1992) and Conferences and Congresses, International in the Encyclopedia of Public International Law (1992).
To be fair to catalogers, I should say that when the travaux préparatoires include the text of the agreement, an astute cataloger will take note of that fact by making an added subject entry for the name of the agreement. That will establish the link the researcher hopes for.
A Westlaw search on the phrase “Geneva Conventions” in the JLR database returns over 3000 documents. Searches on other variations of the name return far fewer.
This caveat is brought home by the observation that a Google search on “Geneva Conventions” has over 750,000 hits, rendering it useless.
United Nations Conference on the Law of Treaties: Official Records. New York: United Nations, 1969-1971.
38 I.L.M. 1207 (1999). The full name is Convention (C182) Concerning the Prohibition and Immediate Action for the Elimination of the Worst Forms of Child Labour. However, the preamble to the convention says that it may be cited as the Worst Forms of Child Labour Convention. Here we have another example of the name difficulty discussed above.
For a fuller discussion of the structure and working methods of the ILO, see Ebere Okieke, Constitutional Law and Practice in the International Labour Organisation (1985).
The ILO website also has a section devoted to its International Programme on the Elimination of Child Labour: IPEC. (http://www.ilo.org/public/english/standards/ipec/index.htm)