Joho the Blog
An Entry from the Archives

« At last, a practical use for your aura || Back to Blog | Congressional view of email »

July 15, 2005

IDs: Freeing us for messiness

Obvious thought of the day:

Messiness is disorder where place is the basis of order. E.g., if your top drawer is the place for socks, having socks in your living room floor is messy. Places order space by putting like next to like, a one-dimensional way of organizing.

Digital space, unlike physical space, does not demand that we pick one dominant trait — one way of being alike — over all others; data about two products that differ only in color may be stored in your database in non-contiguous RAM, and you don't care. In this environment, things do not need to be placed at all; they only need to be found. So give things unique IDs — e.g., the BBC's SMEF for content, Life Science IDs, CommonLanguage's telecom identifiers — and they will be radically messy (miscellaneous) but not disordered. Rather, they are ready to be ordered on demand, differently each time depending on the user's needs and assumptions.

Now I have to go pick up my socks. [Technorati tags: taxonomy EverythingIsMiscellaneous]

Posted by D. Weinberger at July 15, 2005 12:18 PM


Comments

A few comments, and maybe a question:

Place orders space more than just by like next to like. My sock drawer is in my bedroom because that's where I get dressed. It would make little sense for me to store my socks in a drawer in the kitchen. DVDs are stored near the televsion because that's where I would use them. It makes little sense to put them in the garage.

Second, and I'm not sure how much difference this makes, there are more "dimensions" to socks than just "sockness." Dirty socks go in the hamper, where they're mixed in with all the other dirty clothes. Socks that are no longer useful as socks go into the rag bag where they may be useful for something else. Which suggests that the "use" of an object is an additional dimension greater than simply its "likeness."

Digital space does rely on one dominant trait and that is location. I am not a computer scientist but I believe file systems must rely on one location that tells the OS where to begin looking on the disk for information about still other locations on the disk for information about where files are stored and where additional information may be stored (free blocks or sectors).

Additionally, if you're not willing to make inefficient use of your computer resources (i.e. "waste your time"), most search functions rely on indexes, an index being a specific location for information about locations.

It seems to me that you're relying on this abstraction to make a claim that, as a practical matter, from a "human" point of view, digital "objects" can be highly "disordered" and still be just as useful as though they were carefully "organized." I'm saying they are carefully organized, but that organization effort takes place outside the realm of ordinary human perception (yet created by humans), and in a manner that makes more sense for computers than for humans. Differently ordered, differently organized, but just as ordered, just as organized.


So, my question is, are you making a true statement? What insight are you trying to illuminate here that has some ramification for those of us who must put our socks on one at at time? (And if we've done our jobs "organizing" them, they'll both be the same color.)

Posted by: dave rogers | July 15, 2005 01:56 PM


Dave,

Total agreement that we use multiple likeness, and multiple types of likenesses, to order things in the physical world. E.g., what goes in your spice rack and what doesn't is determined by a complex set of "principles." (Why cloves but not chocolate sprinkles, or tabasco sauce, or salt?) And there is of course a reason why my sock drawer is in our bedroom, not garage. In fact, I'd just finished writing about this in the part of my book I'm working on. So, yup, I over-simplified. Nevertheless, those complex, intersecting principles all deal with assigning objects a single place. (Yes, there are exceptions.) Messiness in the physical world means being out of THE place.

File systems hide place from us. As you know, blocks of files on disk (and in RAM) are not necessarily contiguous, and we only care if disk fragmentation is slowing our computer down. (Wasn't there a type of disk encoding that placed logically contiguous blocks into discontinuous places on a disk that were synched with the head's read speed and disk rpm?) The OS shelters us from caring whether what looks like a single file is in fact lots of little fragments tucked away here and there. Yes, the order makes sense to the computer (as you say), but the actual order of that order is totally irrelevant to us humans. To us, the information can and should be (this is my claim) disordered until the moment we need a cut through it.

I don't know if this will help you get your socks on in the morning, Dave, but all I'm trying to say is that messiness is a sign of disorder in the real world (because each thing has a place) but is a virtue in the digital world (because late-binding of order results in orderings more useful to us sock-wearing humans).

Please re-read the first-line disclaimer of this post... :)

Posted by: David Weinberger | July 15, 2005 02:12 PM


I did read the disclaimer. I thought you were being "ironic." :^)

I'm saying that what you describe as "disorder" in the digital world is merely the result of computers having a better "memory" than carbon-based lifeforms, more so than any "virtue" in disorder. Data is just data, being able to have the data you need at the right time makes data useful. My apartment is a real mess, but I pretty much know where everything is. It's the old sitcom joke: "Don't clean my apartment! How will I ever find anything?!"

And, to go a little further, I would maintain that it is the "orderliness" of computer memory (indices) that is the digital analog (groan) of sock drawers.

So at some level, the greatest "virtue" ultimately does lie in some type of "order." Yes, the computer does allow us to remove ourselves a couple of levels of abstraction from the real "order," but it is present nonetheless, and present as a result of a human effort.

Just like putting my socks in a drawer in my bedroom. Or all those librarians putting the books back on the shelves.


Posted by: dave rogers | July 15, 2005 03:16 PM


There's another way at coming at this issue, and that is reconceptualizing what we mean by "place." For example, place in the physical sense is, more or less, obvious. In a dichotomy that divides physical from cyber, your assessment is (again, more or less) true: "placeness" is an attribute that gives "messiness" context, and therefore, meaning.

But consider "place" in the context of social sciences relating to social location. How does a more generalized (and I would suggest, politicized) understanding of "messiness" now fit in?

I'm not merely being pedantic (okay, not exclusively being pedantic) when I bring this up. The changes that I perceive in the world due to ubiquitous connectivity and pervasive proximity necessitate careful consideration of the implications of social location, for instance, in our conceptions of epistemology, and more generally, what is valued as knowledge. The politics of "placeness," I think, is crucial to your exploration of folksonomies, since it is folks that are so placed.

Posted by: Mark Federman | July 15, 2005 03:37 PM


Mark, can you expand your middle paragraph a bit?

Posted by: David Weinberger | July 15, 2005 03:46 PM


Post a comment

Guidelines for Commenting

Basically, you can say what you want. (Click here for the fine print.)

If you haven't left a comment here before, your comment may be put into a queue for me to approve. Sorry for the delay. Blame the damn spammers.