The Dublin Irish Festival is on this weekend – Dublin, Ohio, that is. I notice that Moya Brennan is performing.
As some folks will know, Moya Brennan is an Irish singer who was a member of the well-known family group, Clannad. They emerged in the 70s, playing very much in a traditional irish music idiom. As they evolved, they developed a style that was influential in the emergence of the sort of new age, ‘celtic’ music that became popular and has some well known practitioners.
One of the most vivid memories I have of secondary school is when they put on a concert there (before they were very well known). We had never seen anything like it – songs in Irish, but with a double bass and a harp, long hair, and influences from other types of music.
At that time, Moya was known as Máire Ní Bhraonáin. And indeed, if you look at Worldcat identities page for ‘Moya Brennan’, you will see ‘Brennan, Máire’ and ‘Ní Bhraonáin, Máire’ listed as alternatives. (The latter puts the surname in its Irish form.)
‘Moya’ is an assumed name, perhaps sounding something like the Irish pronunciation of ‘Máire’. ‘Moya Brennan’ is presumably a more memorable name to have outside Ireland than ‘Máire Ní Bhraonáin’.
Worldcat Identities does a reasonable job with related Identities, which you can see and follow in the graphical form at the Worldcat Identities Network page. It pulls together the other members of Clannad – her brothers (Brennans) and uncles (Duggans) – and some others.
Interestingly, however, Identities, drawing on the underlying bib data, does not reveal what many people might feel is her most interesting related identity.
This is her more famous sister, who was briefly part of Clannad, and whose name is Eithne Ni Bhraonain. Eithne is somewhat better known as ……. Enya.
Now, as the relationships in Identities – mined from the bib records – are ‘bibliographic’, there is no reason that family or other relationships which are not somehow represented in the bibliographic data should be present. In fact, as my colleague Diane Vizine-Goetz pointed out there is a relationship in the bibliographic data, but at a very low threshold. See the names collocated in the statement of responsibility in this Danish record (the names are not consistently controlled in WC).
It seems to me that having the relationships made visible in Identities makes me want to see other important relationships – whether or not they are supported by bibliographic relationships. Not having them there seems to me to be an important limitation. This is because I am seeing things through a names or identity filter and the Identity becomes a subject of interest in itself. We are seeing more of this. Open Library, Amazon, Google Scholar, Trove and others all now present ‘author’ pages, because authors or creators are important topics of interest in themselves, as well as pivots for searching.
This prompts these brief observations about some inevitable futures if library practices in this area are to continue to be relevant:
- As I suggested the last time I spoke about variant forms of Irish names (Name authorities, crowdsourcing, and Máire Mhac an tSaoi), it would be nice to make suggestions to LC or the National Library of Ireland or the BL or … However, I have no way of easily doing that. Authorities work – and think NACO here – is a professional activity, hedged around by rules and procedures; it is after all ‘authorities’ work. However, it would seem sensible to open it up to suggestion and information.
- We now have access to several related online sources of data about people and names. Think of the related Wikipedia, Dbpedia and Freebase for example. It would be good to be able to programmatically mine them to enhance the data libraries manage. In Identities, we match names in Wikipedia and link to them when we are reasonably confident the link is correct, but it is not easy to extract relationship information.
- One could change authorities practice to record significant related people who are not supported by bibliographic relationships. This seems sensible to me, although it runs counter to the trend to reduce effort – unless a parallel mechanism is found to more widely share the task of creation. One probably would not want to record that Lorcan Dempsey is Laurence Dempsey’s son; but it would be interesting to know that Martin Amis is Kingsley Amis’s son, and not just let a general relationship emerge because they are both the subject of a book, for example. One might note here in passing that it would be nice to consistently type those relationships.
- Finally, the current model in library catalogs is one in which centralised creation of ‘knowledge organization’ systems (subject/name authorities in particular) is matched by decentralised application of those systems in local environments. There is some consistency, but in general local application don’t make best use of the data, there may be some variability in application, and links back to the fuller knowledge systems are not present. As more activities move to the network level, it is worth thinking more seriously about network level knowledge organization also, where those knowledge systems become network level resources which others link to. Think of how VIAF or the authorities the Library of Congress makes available at https://id.loc.gov/ might be used, for example, to provide richer context. Currently, a catalog may include, for example, a name in its controlled form: think of how the catalog could either link to, or actually incorporate, the richer data available in the authority file, or a system like Identities. In the context of this note, this would bring the design and construction of those knowledge organization systems more into focus and encourage some of the approaches mentioned above. This is alongside the work that happens in Wikipedia, or Amazon, or other sources. You can get a sense of this at the U Wisconsin Forward experimental system which pulls data from Freebase, with varying success. See for example this search for Conor Cruise O’Brien. It would be good to see library resources used to enrich results or local data. I will return to the topic of network level knowledge organization in a future post.
(Incidentally, ‘knowledge organization’ seems a rather antique term .. it would be nice to have a better one.)