Folktales, Folksonomies, and Swallowing the Sea.

I’m a sucker for  grand explain everything attempts to categorize it all.  I find them extremely amusing. Here for example is a five node template that attempts to encompass all folktales. (You can also use this for presentations to your angel investor.)

There is a much more complex template that aspires to the same goal, see here.

The awe inspiring attempt to categorize the universe of folktales is the six volume work by Stith Thompson (who’s name is suspiciously similar to Sith); the Motif-index of Folktales.

The motif-index to distills out the ingredients of all folk tales. “Identity test based on [glass] slipper fitting.” Folktale ingredients is a finite set; at least they are if you limit your self to those Stith could read in a lifetime. Reading some sections makes it clear that’s a fools errand. For example, kinds of fools: “Stupid Wife,” “Extravagant Wife,” “Stupid Husband,” “Stupid Village,” … “Foolish Brahmin,” … “Foolish city dwellers,” “cowardly fools,” … “bungling fools.” Clearly this set is innumerable. 

Many entries in the motif-index are sufficient to let you visualize an entire story. “Fool sent to acquire two 15 year old slaves returns with on 30 year old slave.”

Some of the short comings are quite telling. For example, there are only a handful of entries on slander. Recall Karl Rove’s motto: recovering from slander is like recovering water spilt on dry earth. Other holes arise because Thompson didn’t have access to all the folktales out there. So, others have tried to fill in the gaps. For example: “Sparks of burning cannibal woman become mosquitoes“, has a Native American origin.

Like accounting codes (Horses 10101506), or the Dewey Decimal system (Composting 631.875), the entries in this grand classification have identifiers (F952.1 – Blindness cured by tears).

Systems like these always have an owner, the authority. Thompson’s system evolved from a system designed by Aarne. Aarne’s system was an index to stories; where Thompson’s system finer grain – story elements. I’m not sure I have the history straight, but it appears that some effort was made to update Aarne’s system creating the Aarne-Thompson index of story lines (AT 762 – Woman with three hundred and sixty-five children.).

Thompson failed to establish a clear line of succession for who’d own his system when he died. So if you want to update it there’s nobody to talk too. Which is a problem. As we know, the semantic web wishes to encompass all these systems. (D???. magic helper who can swallow sea).

One might imagine going to the IETF and proposing a new protocol – folktale. (folktale:/AT/1384 – Man sets forth to find three others as foolish as his wife.). You might complain that such a thing wouldn’t have an obvious protocol. But then I’d point out that are already approved URI “protocols” that don’t have what most practitioners would think of as protocols (domain names, ports, etc.) tel:, info:, and tag: are three examples.

The puzzle with these kinds of identifiers is how to bridge from the carefully designed household which is their home into less rule bound global space of URIs. The tel protocol (tel:+1-416-395-5400 Dial-a-story at the Toronto Library) is a exemplar of that process; since there is a highly regulated and extremely complex universe of phone numbers. Phone numbers have high stakes politics, commercial players, property rights, etc. etc.

The tag URI is almost all the way over to the other side of things. Though all tags URI’s have an authority, in this case a domain name or an email address, they are primarily designed to be extremely light weight to create. I like that the tag RFC uses the term mint for the act of creating a tag; many of these identifiers have all the properties of property rights or currency. While I guess there isn’t anything to prevent an authority from announcing that he is governing his tags with great pomp and circumstance that’s not the typical behavior.

Info urls come closest to being a generalized scheme for bridging from the URI identifier space into the identifier spaces ruled by others. For example the info URI: info:ofi/enc:UTF-8 denotes something in the OFI world, i.e. the National Information Standards Organization’s OpenURL Framework. Messy, but necessary, these bridges.

While I still think my favorite entry in the motif index entry is “Fool mistakes pumpkin for asses egg.” I now know that the UPSPC code for asses is 10101509 and that pumpkins don’t have a code.

5 thoughts on “Folktales, Folksonomies, and Swallowing the Sea.

  1. mtraven

    I think the interesting bits of this kind of ontology are at the upper (abstract) end, rather than classifying magic pumpkins differently from tomatoes. Like the template you show, or the work of Vladimir Propp who is the narratologist I’m more familiar with (btw your link under the picture is broken?) But then I’m an abstract kind of guy, someone has to collect the raw data I suppose…evolution is interesting but I can’t see spending my life collecting and classifying barnacles like Darwin did.

    The thing that drives me crazy about folksonomy is you have no way to know which level of the ontological hierarchy you ought to be tagging. But I guess that’s part of the point, there is no hierarchy, or if there is it emerges sometime later (Web 3.0).

  2. bhyde

    Fixed the broken link, thanks – and amusingly it’s to a garish web page about Vladimir Propp’s work.

    I like both the motif’s and the story grammer stuff. I tried drawing a picture of the 30+ Propp story template; that didn’t work. Then I tried writting an LR grammar for it – which was fun and sort of reveals it’s short commings.

    I admire that Thompson managed to find a way to spend his life reading fairytails; and get paid doing it. And, I think we need to bring back the cabnet of curiosities.

  3. Pingback: Ascription is an Anathema to any Enthusiasm » Blog Archive » Magic Helpers

  4. Pingback: links for 2006-07-15 at blog every brand new day

Leave a Reply

Your email address will not be published. Required fields are marked *