a million dollars and a Nobel prize
Classical music metadata.
Part of the reason it has been so long between entries is because my work on classical music metadata has been so myriad, intense, and well frustrating, to be honest, that even thinking about writing about it has been a task in itself. Am I being hyperbolic when I say that it is the problem of our age? Probably. But what I mean to convey by making such a statement is that it is a problem without a solution. Also, it is a problem that many don't realize is a problem until you ask them certain probing questions…but that's something I will return to later in the post.
This metadata project came about in a very practical fashion: NPR's in-house music database has a legacy file naming convention for its art music back from when it was digitizing LPs and moving from a physical collection to an online one. I won't go too much into why the system exists as it does but what's important to know is that it is as anachronistic as possible. There is very little connection between it and any other "standard" and makes it nearly impossible to discover anything as the search is exact rather than flexible. So being good librarians, we want to fix it. Making that statement was the easy part.
What followed was a very intense meeting in which my supervisor and I went through the pros and cons of various metadata & cataloging systems (our in-house database, iTunes, and others). There were far more cons than pros. It gave us a lot to consider and some things we could put in place but still left an uneasy feeling. So I took it to the internet and posed a question:
As I watched the votes come in, it was clear that there was no unified answer. Lots of people started tweeting me, or rather, tweeting at me, asking themselves rhetorical questions & finding more questions that answers. Why is it that there is no simple way to handle this one part of classical music metadata? What does this mean for the rest of the process? What exactly do I do in my own library? It was a metaphorical rabbit warren of epic proportions. It was through this conversation, both on Twitter and Facebook, that people realized that there is a problem and that they were at least ignoring it and at most making severe concessions. So just why is there such a problem?
Classical music is not pop music. This is probably an overstatement but it is the heart of the problem. The information needed for popular music metadata is fairly straightforward & rarely with differing parts: the track artist is probably the album artist and if that's not the case, it's most likely not different for every track. There is agreement on the song title (and what language it should be in), track composer information is not necessary but when added doesn't create severe problems. For (Western) classical art music, there are just too many variables, as evinced by the image above. However, if it were solely a problem of variables, perhaps it would be fixed by now. Classical music listeners are, generally, collectors. Libraries of their own making and how they curate those libraries are deeply related to how they listen. And as everyone curates and listens differently, it makes it difficult to come up with any kind of standard. I'd like to think that my library & file naming conventions are ideal but I know many (well not many, it is very good) who would be willing to contest parts of it. And at the end of the day, there is no financial motivation for record labels or streaming services to invest labor into coming up with a solution.
In my conversations with the taxonomist for Amazon, she relayed that she is often drafted to help with this problem because that's a better solution than devoting resources to a problem that yields very little financial gain. If you've ever tried to find classical music on Spotify, you've probably hated every minute of it (I suggest trying to find the Cleveland/Szell/Fleisher Brahms Piano Concerto recording, go on, give it a try) and that's because the metadata provided just is not complete. But there's no reason to be concerned with discovery: this Chicago Tribune article doesn't even mention classical music when talking about what genres do and don't benefit from streaming services. And speaking of articles: if you google "classical music metadata spotify," you will see a litany of posts & articles about how streaming services get classical music metadata wrong & how it is the one unsolvable problem left to tackle (see the first paragraph). So we collectively recognize that it is a problem but have resolved to do nothing about it and/or assume that nothing can be done. So where does that leave me?
I think with current systems in place it is impossible. They would have to be retrofit to accomplish something they were never tasked to do in the first place. It's time to build something from the ground up. I think companies like iDagio have the right idea in building a robust search function that accounts for all of the different variables mentioned earlier. I think there has to be a push to make classical music a priority, for record labels, for content providers, and for listeners. I have experienced how much classical music listeners tend to bristle at the idea of listening digitally (something that resonates with me on certain levels) but as I've said before, the way we listen is changing. And if we want to make sure that we can hear classical music in the best way possible, we have to start here. I'm not advocating for any specific way, I'm advocating for any way that is innovative and not reactionary. And god knows I can't do it on my own, as much as I'd like to.