David Weinberger gave great comments on my Metadata and Vagueness post, and also posted more comments in his Meta conversation on metadata post on his blog. I will try to clarify things further here.
So, I basically took David's article to be saying that metadata efforts had brought us into a realm of new issues because of the nature of metadata. Based on David's subsequent comments, I now think he was saying something more general along the lines of: we are in a realm of issues that, with current metadata efforts, are becoming a lot bigger deal.
In part 2 of this post, Data and Desire, I will attempt to describe these issues in my own words, because I agree with David's post in this general sense.
But, here are some specific points of clarification on what I wrote in my previous post and David's comments on it.
I think of metadata as being something that exists in relationship to a system (e.g., conceptual set) of data, as opposed to be something that relates to data in general. The metadata of a specific system defines constraints in a specific data system. Outside of that system, the metadata may not be meta.
On the Internet / web, I do not believe in an ideal of a single system of data. So, to the degree that semantic webby things like namespaces could be thought of defining a single system of metadata, I see potentially deep conflicts with the viability of the web (which I would describe in this context as being unlimited points of view and ways of expressing them).
(With the iCite net, I am imagining a way to link diverse systems of data and metadata together more. And, to the degree that semantic webby things like namespaces could be thought of linking together things in this way, I see interesting possibilities.)
But, in any case, I think that the issues that David brings Are issues of both data and metadata. I see both data and metadata as being potentially "information about information".
While metadata always is data about data, and so always an "abstraction of an abstraction", data can be as well.
I think the major disagreement we may have maybe comes down to the word "about" actually!
For example, the URL of a web page, like www.example.com/webpage.html, is data. The web page is also data. The URL is data about web page data, at least the way I would use the word "about". But, I wouldn't say that the URL is metadata of the web page—though perhaps in some system it would be.
David, is his comments about my comparing date formats and FOAF, says:
I disagree with you about date formats vs. FOAF. Once we agree on the categories we need for FOAF then the discussion becomes like the date format discussion; it's the discussions before that that are interesting and important and fraught with impossibility.
With dates and date formats, decisions have been made about the categorization of time (which is itself a symbolic measure of observable occurences in the physical world). The decisions about categories of FOAF are only different in that they are decisions more about us than about our world. And, of course, at this point in time, we take something like dates and times as a given and its categorization generally concerns us less than the mapping of our friendships.
Also, one clarification of something I said in my original post. David, in his blog post, says:
We also disagree at least a little about what's driving the need for metadata. Jay thinks it's "the desire for increasing abstraction, rather than need for metadata itself."
I wasn't clear enough in my original post. What I meant was that the need for more abstraction in systems of information was not so much a need but an expression of the desire for abstraction. I didn't mean that the need for metadata was coming out of the desire for abstraction.
Finally, part of the reason I don't want to make a big distinction between data and metadata with regards to their degree of abstraction, is that I believe one person's data is another's metadata, and vice versa. Like, one person's grain of sand is another's world:
To see the world in a grain of sand,
And a heaven in a wild flower;
Hold infinity in the palm of your hand,
And eternity in an hour.
—William Blake
In general, I think the functional problem with data and metadata systems is that the systems don't usually allow you to suddenly see a world in a grain of sand, or vice versa. But, I think the web opens the possibility for systems to work more like this—and people are already using the web in this way with the current means at their disposal.
Note: All comments and trackbacks are moderated. Spam is deleted. Other comments are approved as promptly as possible.
Note: Older posts no longer accept new comments or trackbacks.
trackback from: the iCite net development blog
posted: Nov 11, 2003 3:58:09 PM
title: Semantic web, my last word?
Since I have been talking too much about this already, I thought I should declare my "last words" on the matter, and move on to some other topics. So, my last words are this: