on Mastodon, where there's a polling function, so far the tally is 61% favor Canada, 39% Mexico, 0% USA! (that's from a total 18 votes counted from a nonrandom sample not representative of anything. so you know that it's reliable!)

as I was typing, USA pulled into contention! 5% favor, of now 20 votes counted. USA! USA!

my complaint isn't fundamentally about ad views. and i don't claim indexers should ignore specimen links. i claim indexers should understand them. 1/

if an index is interested in surfacing "high quality" information, it should take into account the referrers explicit disendorsement of the quality of the information linked. /fin

What if we could distinguish, when we link or quote, between citation and specimen? 1/

The default that hypertext is built around is citation: We are engaged together a collaborative exercise to construct some approximation of truth. 2/

As humans, we frequently err. So much of our conversation is necessarily, and constructively, critique. Critique can be dry and civil. It can be cutting, bitter, hilarious. We may be naughty, we may be nice. But we understand ourselves as speaking to one another, critic, critiqued, audience. 3/

But sometimes that presumed relationship is just not accurate, is not what we intend. Sometimes we are not in conversation at all. Sometimes a piece of text is a mere artifact, a specimen we are conversing about but not at all with. 4/

Even bitter critique implies a modicum of good faith on the part of author critiqued. There is a mind which, however biased by virtue of position or commitments, has given the matter some thought, and believes what it has written. 5/

If we think that behind the document we are addressing there is no such good faith, citation — inclusion in our collaborative project of truth production — is not the appropriate relationship. 6/

If a document is pure propaganda, if it has been tailored instrumentally to affect or manipulate, represents no coauthor's imperfect but sincere yearning towards an edifice we might productively settle upon as truth, then we should not cite it. 7/

But we might still wish to refer to it, to converse about (rather than with) it. We should be able to quote or link it in a way that makes the specimen relationship explicit, and imposes informative friction (e.g. some interstitial) to people who might naively follow it as citation. 8/

Our scheme should prevent naive indexers (e.g. "page rank") from following such links as citations. (Indexers sophisticated enough to work around the block would have an opportunity to choose how they want to interpret such very distinct links.) 9/

Screenshotting but not linking a source is the closest approximation of this in current practice, I think. There's also HTML's rel="nofollow" attribute. Neither sufficiently expresses and fully enables what we should want of a specimen link. I think there's some scope for innovation here! /fin