Friday, March 20, 2009

Ruminations on Metadata

If metadata is merely data about data, then I believe that our class discussion has thus far not fully explored the broader concepts of this idea. While David Weinberger in Everything is Miscellaneous uses again and again the example of the library in organizing, quite well in fact, I feel that with the ability to tag and identify pieces with descriptions and keywords much longer than the original piece itself makes for some interesting analogies on how the web is essentially all data and all metadata at the same time. For my discussion here, I am using a somewhat looser definition of metadata than Weinberger, for instead of thinking solely of cataloguing purposes, I also am including material that can be used for a greater understanding of an item on pretty much any level.

This first began to crystallize with me as I read Jonathan Boyarin & Martin Land’s Time and Human Language Now, an epistolary examination of time relating to language that was much more complicated due to the quantum physics than I had originally realized. The two friend email each other extended discussions about the nature of their respective fields of research, at one point mentioning that these files were being sent not as emails themselves, but instead as Word attachments to a sort of cover email explaining the contents. While the epistolary form that makes up the book has an easy to identify audience, Boyarin when Land is writing and vice versa, it also is geared towards being readable by strangers to the lives and disciplines of the two men. But these cover letters have only one audience since they were not included in the book, and I began to think that reading this book digitally would allow the cover letters to be included as a sort of metadata: they are information about the information. I began to ponder.

When President Obama released his NCAA tournament picks earlier this week, I was a bit surprised. One would hope that with all the other issues on his plate, this is a minor one, but he is a basketball fan and I suppose it is a god way for him to show that he is ‘just like us’ to the average guys who bet on the tournament (guys like me, go Memphis). Then a reader of Andrew Sullivan’s Daily Dish blog wrote in analyzing the president’s bracket from a political stance. The president only selected three underdogs, and of the three favored teams picked against, two are from California (
a Democratic lock in 2012), while the other was from (possibly unwinnable) Arizona. Two of the underdogs picked were from political swing states Pennsylvania and Virginia.

I wondered how I might store all this information in the perfect electronic system. If someone’s research led them to President Obama’s selections, it would be interesting to have the add commentary of people like Sullivan’s
reader. Basically, metadata for the primary data of the brackets. But perhaps someone is instead researching how politicians in power court votes, and they find this comment about the president’s picks. Then the actual brackets themselves would be the metadata.

Perhaps this isn’t anything revolutionary, but I started to think about all the information contained on the internet and how it could be applied in these sorts of scenarios. If I am reading an article about the issues in Gaza as it relates to the Israeli/Lebanese War in 1982, I could easily click over to an article that would help me put that war into context since it happened when I was only three. In fact, with the wealth of information available, practically every single thing on the web is simultaneously data and metadata. One has hundreds of places to look for information about the information they are reading.

Even this post is a perfect example. It is a response to books by Weinberger and Boyarin & Land, it comm
ents not only on President Obama’s NCAA picks but also on a response to those picks, and it serves as an example of the sort of posts that go up on this site. Yet it stands on its own as data too. Many of these connections will likely never be used, but the point is that they can be not that they will be, and I would imagine that the future will see ideas like these played out.

1 comment:

Deb said...

Alas, I had hoped that our discussion about meta data would go into greater depth or inspire more insights into possibilities.

As you describe, metadata can be really powerful, and mashups might be the most current manifestation of some of what you describe, at least for geographically-related data. Searching the ever-emerging sea of metadata will be another problem. I already get thousands of hits for most items I search for on the web, for example.

When I was a technical communicator (until 2001), the going advice was that for every team of technical communicators, one of them should have a particular expertise in indexing. I would guess now they would say that ever person on the team should have some expertise in indexing and I would recommend that at least two of them focus on indexing, particularly with metadata.