The M Word: Metadata

May 29th, 2012

I hear you. “Metadata” has branding issues.

The components of the word are turnoffs in themselves, “meta” and “data”. Can you think of anything “meta” that you like? Metamorphosis perhaps? Metabolism? Both of those words have the same grounding in Greek, but growing up with them as we have, familiarity breeds…familiarity.

Bad enough that metadata has to do with “data”.  But as “data about data” it’s fatally flawed.

Wiktionary defines it as “data that describes other data, serving as an informative label.” Wikipedia manages to suffocate its explanation in a 7,300-word entry, leading off with the reassurance that “metadata is an ambiguous term which is used for two fundamentally different concepts (types).” Book publishing isn’t mentioned, although libraries are, but then so is “geospatial metadata” (describing geographic objects).

“I hate the word metadata: whenever I mention it, people’s eyes glaze over immediately,” said Fran Toolan of metadata service vendor Firebrand Technologies at the Metadata Perspectives conference last fall in Frankfurt. Reporter Andrew Wilkins soon popped the question: “Can’t we just choose another more sexy term?”

Oh, how I wish we could. I’ve tried many. The friendliest I can find is “bibliographic information” or “bibliographic record.” Those terms aren’t completely interchangeable with metadata, but they’re pretty darned close. Are they more comfortable for you?

The fact is that metadata for books is not data about data, unless you consider the physical (or digital) book itself and its content to be “data”. When I’m reading a 300-page 6″ x 9″ trade paperback and set it down for the evening I don’t say “I’m putting my data on the bedside table and going to sleep.”

So we turn instead to the “informative label” nature of metadata. At a minimum the informative label includes the title, author, page count, publication year, ISBN and price. That’s metadata. The book is the physical object or digital file and the words and images the medium conveys. So book metadata is not in fact data about data. It’s just information about a book.

Why all the fuss about metadata now? Because in a world gone digital sometimes we can’t find the there there and so we just have to describe it. When 300 sheets of paper turn into 300 kilobytes of data the labels are important. Sometimes all we have are the labels.

When books were kept in libraries and bookstores, even if we couldn’t find the listing in the online catalog, we could always walk over to the shelf and say: I see. Here it is.

Now that the book is just a digital file there’s no shelf to walk over to. How are you going to find that ebook file on your 100 gigabyte hard drive if you can’t remember the title? How are you going to buy it on Amazon if you also can’t remember the name of the author?

So, at a minimum, you’re going to need accurate (or relatively accurate) title and author metadata to find a book. Metadata for findability, let’s call it. And let’s make that the topic of my next blog entry on metadata.