[TSC-public] The role of email, attachments, slack, github, etc, in FHISO's work

Thomas Wetmore ttw4 at verizon.net
Fri Aug 29 03:25:43 CDT 2014


I think there is confusion about the terms "source" and "citation" that leads to disagreements, but because the assumed definitions are different, not because there is true disagreement.

To some people a "citation" is a specially formatted string that refers to (and might actually even contain!) a specific item of information from a specific source. Because there are many types of sources, and many items of information that can be found in sources, rules for how to compose these specially formatted strings can come in overwhelming numbers and complexity. Of course, EE is often trundled out as an example. When some of us use the term "citation" it is this universe that we imagine around us.

To others a "citation" is a more abstract concept; it is "a place in a source" that certainly needs to be defined as a set of appropriate properties (say page number, or chapter title, or volume issue number, or household reference number, etc). In this universe, this more abstract concept of a citation is converted to a specially formatted string, as a separate operation, say when preparing a work for publication, by applying an approved "template" to the abstract citation, to produce the formatted string.

I believe that formatted citations do not belong in genealogical databases; databases should have the abstract form of citations only.

The DeadEnds model does not use the word citation. In that model there are sources, a typical bundle of properties (usually called meta-data in the bibliographic/archival world), and source references, which are bundles of properties that establish places in sources. The underlying assumption is that the generation of citations (that is, the generation of the formatted strings) is done by applying sources and source references to a software based template matcher that will produce the strings on demand.

Enno's concerns, I believe, primarily concern internationalization. Certainly, if we were to use the "formatted string" definition of citation, and legislate the formats that those strings would have, "Enno's continent" would have every right, maybe the obligation, to rise up in revolt. If we take the "bundle of properties" approach, then we can establish, one dearly hopes, an internationally acceptable vocabulary for those properties.

On Aug 28, 2014, at 5:24 PM, Enno Borgsteede wrote:
> I think that if scientists can use a controlled vocabulary to classify all creatures on earth, serious genealogist should be able to do the same for sources, which have much less diversity, when you take a good look at those. That is my main objection against EE based templates, which are extremely unbalanced, not open, and impossible to translate, because of the sheer number of terms. And my experience shows that most users and developers in English speaking countries are completely blind to that, and even people that I consider to be quite smart seem to reject what I write about that, saying that citation is an art not a science, and can therefore not fit in a reasonably sized controlled vocabulary. Such reasoning looks very much like crap to me.
> 
> When people are serious about being balanced, reaching consensus, getting international support, etc., an objection like mine, which might be shared by many developers on my continent, will automatically lead to EE and its likes not to become accepted as a standard, which would not mean that I would not accept citation templates as a part of it. What I write here applies to vocabularies that affect the way users all over the world see data, not only for citations, but also for event types, and a couple of other things.




More information about the TSC-public mailing list