GL: Access and Authentication

Authenticity is a concern in courts, where any document offered as evidence must be free of tampering. Inasmuch as GL might be evidence, it would also need to be authenticated in some way. Vincent and Dominic have pointed out that GL has been cited as authority, sometimes with some degree of necessity (where no legal treatment of a concept exists), and sometimes rather bizarrely. Of course, if we include government documents in GL, I’ll bet there is a fairly consistent appearance of this class of material in the courts in the forms of royal commission reports, law reform reports ( See Michel-Adrien’s post), and other authoritative sources of legal thinking.

Authority in the legal sense and in the popular sense includes the idea of accuracy, correctness or reliability of information. Without authenticity, which no document can be accepted as an authority. Thus we have heard (in the past, now) about judges who do not accept computer-generated texts as authorities. Insisting on photocopies from known reporters and bound statutes is an effective, if less than expedient, method of eliminating some opportunities for tampering.

I’d like to sketch an imaginary courtroom, at least partly in the spirit of some others we’ve seen on Slaw, where a judge is determining a question of libel. The complainant, the Canadian Judicial System, is suing The Media because “the public now thinks we are inaccessible and intimidating, when in fact we are available, and rather attractive.” The Court needs to determine whether or not there is a crisis in public confidence in the Canadian judicial system. As it happens, there is a document that speaks directly to this question, produced by the Canadian Forum on Civil Justice.Mary Stratton and Diana Lowe. Public Confidence and the Civil Justice System: What Do We Know About the Issues? (Edmonton: Canadian Forum on Civil Justice, 2006), p.7. Both parties have found the document with ease on the website of the producing body, and have downloaded it for free, However, they have submitted differing versions of the crucial text:

From the Complainant:

3. Recommendations for action
Our review strongly suggests that we have clear and reliable information about the low level of public confidence in both our criminal and civil justice systems.

From the Respondent:

3. Recommendations for action
Our review strongly suggests that we lack clear and reliable information about the level of public confidence in either our criminal or civil justice system.

After quizzing the librarian of the Forum extensively, it was determined that, under pressure from the research director, he had posted a version of the report that was in fact, a draft version, and that later he corrected this, and posted a final version. The research director confirmed that there were two versions, and that, as the research progressed the passage in question was altered to reflect the emerging results of the research. She apologized for unnecessarily pressuring the librarian, and promised to be more respectful in future.

Despite this, the Judge determined that regardless of the opinions of the Forum employees there was nothing particularly ‘final’ about the version they now claimed was final, and that a new version of the report could be published tomorrow with altogether different findings.

“If only there was some way to determine which version of this report truly was the final version,” he mused. “If we were in Ontario, we might be using the Sedona Principles by now, which would at least provide evidence of the stages of the information life cycle that the document had passed through, as outlined on page v of Overview of Changes in the 2005 Version of The Sedona Guidelines on the Management of Electronic Information, available at Still, I would be relying on evidence provided by the creating body, rather than from an independent source. I wonder what the InterPARES project would have to say on this issue, particularly the Authenticity Task Force report and Appendix 2–requirements for a presumption of authenticity found here . What to do?”


  1. Michael,

    Thanks for the entertaining image of the courtroom – complete with judicial musing! And thanks also for opening the door for the InterPARES authenticity requirements.

    Creating documents is a process or results from a process, and we all know there can be multiple versions of a document. This seems particularly true in the electronic environment where any jotting must be frequently “saved”. The InterPARES authenticity requirements Michael references (available at – final version in case anybody wonders) requires, among other things, that records be identified and that where multiple copies exist that the authoritative copy forms part of the identification. What this would do is attach to the document the information which the judge eventually learned by questioning the librarian and the research director. Whether the existing version was truly the “final” version would also be evident, because only the final version would have the authoritative indicator. (InterPARES 2 defines the term authoritative copy as “The instantiation of a record that is considered by the creator to be its official record and is usually subject to procedural controls that are not required for other instantiations.”)

    Micheal’s post also raises two similar but different concepts: authenticity and authentication. Evidence law presumes the authenticity of records if “if it is established that the electronic record was recorded or stored in the usual and ordinary course of business by a person who is not a party to the proceedings and who did not record or store it under the control of the party seeking to introduce the record.” (quoted from a version of the Uniform Electronic Evidence Act, s.5(c))

    Defining the ‘usual and ordinary course of business’ is a critical element to support the presumption of authenticity. What becomes clear in the example is that the usual and ordinary course of business involved multiple drafts of reports. The usual and ordinary course of business may also involve making the current “state of thinking” publicly available. Knowing the usual and ordinary course of business is critical to presuming authenticity. This is part of a record’s identity.

    Authentication of a record is a different process. Perhaps, as in Michael’s example, authenticity cannot be presumed, or at least for any other documents from that source. Authenticating a record means establishing, at a given point in time, that a record is authentic (or not). An authentication exercise for the two versions in Michael’s example might authenticate one as the preliminary version and the other as the current or final version, again based on the activity that had brought the record into being. Authentication might be inconclusive if that activity is ill-defined or can no longer be determined. Current authentication technologies, e.g., digital signatures, assess the file encoding at a bit-level, measuring the result against a threshold. If the result is correct it will open the document or announce that it has not been modified, but it won’t tell you whether you were looking at a draft or final, authoritative document.

    Btw: My apologies for the delayed response as I was unavoidably detained by meetings this morning.