Name keys [082]


Use of the key= attribute on <persName> to uniquely identify individuals


The WWP uses a key= attribute on <persName> to uniquely identify individuals mentioned in the textbase, and to disambiguate individuals which share a name. The keys are stored in a database which links each key to the person to whom it refers, and to information which will distinguish that person from others sharing the same name. At present, we are not adding key= values in newly encoded texts, but we will resume using key= in the future.

The WWP uses keys only to identify names which originate outside the text: historical figures (including modern ones) and fictional and mythological figures which are part of the general culture. We do not currently key the names of fictional characters from within the text, though it is possible that we will move to add this encoding later on.

Mythical figures who are part of the culture and appear in many texts should be keyed, as should any fictional characters who are so well-known that they are mentioned repeatedly in numerous texts (for instance, Juliet, Falstaff, Gulliver). All people who had any historical existence, however minor, should also be keyed, including printers, publishers, WWP encoders, and the people who sign their initials to dedicatory poems.

The key is generated from the first initial (if there is one) and last (or only) name, plus two randomly generated characters to disambiguate keys and a checksum character to guarantee accuracy.

Hyphens are omitted. The keys are case insensitive, but for aesthetic reasons are usually entered with the first two letters upper case and the rest lower.

Keys can be globally updated, so that if research reveals that two people with different keys are in fact the same person, the keys can be changed to reflect this identity. Similarly, if we discover the real identity of someone whose initials appear in the textbase, the key can be changed to reflect the full name and the key can be updated accordingly. The keys thus serve to create permanent links between the identifiers that appear in the textbase and our current knowledge about who these people are.

The question of how to determine an individual’s true name is logically and practically distinct from the practice of using the key. As long as each individual is uniquely identified, the most important function of the key is served. Content research to determine the correct name to use in our documentation for a given individual also needs to be undertaken, but is less immediately crucial.

list all entries