intertwingly

Well, darn it all Sam, there you go being all ‘connecting of the dots’ and all.

Party pooper

It’s that pesky controlled vocabulary problem: people are compelled to use the word “tag”, and then having no choice, they have to abuse it to mean different things in different contexts!

Actually, the other amusing thing is that each service itself uses the word tag to mean two different things too. Posting more about that on my blog shortly. . .

I’d asked Joshua and Stewart both in the past why they call 'em tags but they’re labels in Gmail and Technorati treats blog categories as equivalent, and LiveJournal calls them interests... but I guess it’s just being mean-spirited at some point. They’re demonstrating that names being ambiguous or having synonyms isn’t really a failing.

Paul Hammond : Sam Ruby: Irony - While we all use the word tag to describe what we have created, we each mean something completely different by the term...

Look at the RSS feed from delicious.  That tells you the information that is associated with a “tag”.  The act of associating that information is the act of tagging or of participating in folksonomy.  If  the web applications that associate that information share each others RSS feeds, then you will be “crosslinking your tags”.  The end result, if you do it right, will be that if i tag something at delicious it will end up being tagged the same way at furl ... etc.  If you need me to use the same username ("canyouhearme" in my case), then i will be more than happy to do so.  Now what is it about cross linking your tags that you still don’t understand?

While we all use the word tag to describe what we have created, we each mean something completely different by the term...

Paul, every time we use a word “we each mean something completely different”.  That is just the way words work.  There is a difference  between folksonomy and coding a program.  Folksonomy is just natural language.  Maybe that is why programmers are having troubles understanding why it works so well.

every time we use a word “we each mean something completely different”. That is just the way words work. There is a difference between folksonomy and coding a program. Folksonomy is just natural language. Maybe that is why programmers are having...

“Folksonomy is just natural language.  Maybe that is why programmers are having troubles understanding why it works so well. ”

On an individual basis, Seth. As Sam’s posting shows, when we try to interoperate, things break. That’s the whole point of taxonomies and ontologies and agreed on definitions and meaning: so that the data from multiple organizations can co-exist side by side.

I for one don’t see why ‘tags’ won’t work for any organization. But I also don’t see how they’ll work for many.

Shelly, i don’t understand you point in the light of my point.  Ask yourself, do natural language words work on an "individual basis"?  No they don’t, Shelly.  Now, i don’t know why words work,  but it certainly is not because we decide to agree that they mean the same thing to each other.  It seems to me that words work because the social behavior of ostension works.  Well my point is the folksonomy is that self same social behavior and that is why it works so well. 

Also i think that before people say that folksonomy doesn’t work should start doing some more tagging at places like delicious ... then come back and tell us it “don’t work” and why it don’t work. Fact is that my subscription to the common tag “folksonomy” is currently my very best source of current knowledge on that topic.  Now you can’t beat that with a stick!

Seth, several major aspects of current speech, including the rules against the double negative (I ain’t got no respect) arose from a Catholic Bishop who decided to apply logic to language, and it stuck.

Language has long been influenced by specific individuals and institutions, and resistent to manipulation and expansion by anarchistic individuals. True, if a word is used enough, such as ‘weblog’, it does enter the language pool. But much of the mechanics of our language and our form of communication have been codified by expert and agreed on by authority.

But I’m not being critical of tags. I think the flickr tags are fun, and I think we could do some interesting stuff with Technorati tags (such as my 'tagback'), and delicious is interesting if the term is finely defined enough. Wikipedia is the cat’s jamies.

If these individually are folksonomies, great.

But without some form of formalization and agreement, trying to mix data from each of the systems isn’t as successful. It does work with terms that have very narrowly defined meanings, but not with loosely defined terms, such as tag itself.

I’m saying folksonomies have failed; I’m just not saying folksonomies have succeeded.

Sorry, typo earlier:

I’m NOT saying folksonomies have failed; I’m just not saying folksonomies have succeeded.

... well i don’t think “much of the mechanics of our language and our form of communication have been codified by expert and agreed on by authority” is a true statement.  Language happens from the bottom up, it is not dictated by authority from the top down, nor does it work by agreement. Nobody asked me what the word “agreement” means. I made no agreement.  Dictionaries reflect usage, not codifications by authorities. 

There is no reason that the same process of ostension that happens when you point to a Internet resource and associate words with it at delicious cannot imply that those same associations are also at technorati. All it would take would be a little cooperation between those two enterprises.  Can you doubt that this would be good for the users?  Domain names in our folksonomy? ... we don’t need no stinkin domain names !

Seth Russell wrote: “There is no reason that the same process of ostension that happens when you point to a Internet resource and associate words with it at delicious cannot imply that those same associations are also at technorati.”...

There is a very big difference between a “tag” created by an author and inserted into content and a “tag” created by someone other then the author and stored externally. While the author is probably the best expert on the “subject” of the post, the author is also probably driven by a motivation to direct specific kinds of attention to the post. The result is that author-written tags often lack the “objectivity” of non-author, external tags. People who use del.icio.us are, presumably, creating tags for their own benefit and their influence on what other people see is of secondary importance. Thus, one can probably “trust” a del.icio.us tag much more than a tag which is created by the author and passed in the content. In-content, author-generated tags are much like “meta” tags in HTML. Once, meta-tags were useful, however, as soon as spammers and SEO people discovered their utility, they became useless. External tags may be easier to protect from spamming and SEO use.

bob wyman

Bob, you bring up a very good point; there is a big difference who tags an item.  But authors also tag their own items (at del.icio.us and elsewhere), and when those meta comments get repositioned, there really is no telling from whence they came.  Eventually a person’s identity will flow through the system and the veracity of their tagging can be ascertained by trust.  And, yes, factoring in whether a person is commenting on their own work will factor into that trust.

I guess your saying that Technorati is primarily tagging by authors and del.icio.us is primarily tagging by others.  But i have gotten in the habit of tagging all my authored items at del.icio.us, and i think others have too.  Also my tagging is not primarily for my own reference; rather it is to position what i have to say into the community. 

It seems to me that, for any given category, i could force the items in del.icio.us to also include all the items from Technorati in the same category; but they would be tagged from me instead of from the author.  But del.icio.us could internally program the same thing and make them come from the author.  Doing that would be very good for the users and the community.  This is the kind of cooperation that would give folksonomy a real bite.

Interesting.  I interpreted Sam’s post exactly opposite to how Shelley interpreted it.

I thought he was saying that these folks whom are so happy with their tagging mess (which succeeds at the network scale because of common natural classifications) seem to be missing the larger idea: the network effect makes the mess useful, and the larger the network, the more serendipitous classification agreement there will be.  There need not be total agreement between the users of the separate systems.  There will emerge enough common agreement that the integration would be useful.  Taxonomy be damned.

... One mitigating effect of this is that in the different tagging contexts, people don’t have visibility to related tags.  You might think this is a big deal.

When I first started using delicious, I would post something, then look at related tags for the stuff I tagged it with, and then add those tags to my own entry when it seemed useful.  But it occurred to me that 1) this was too much work, 2) others aren’t likely to be doing this, and 3) the system works well even so.

Sam, care to clarify which way you meant it?

Hi Sam,

I thought you might be interested that I’ve launched a second tagging / social bookmarking site: [link]

-Steve [link]

Jeremy: What I saw as ironic was for somebody who casually makes such connections to words created by others to preach caution when such connections are applied to words created by themselves.  I was not making any value judgments as to whether folksonomies are a failure or a success.

Sam,
  Sounds like we’re saying the same thing.  ;)