I don't want to leave false and possibly harmful statements unchallenged.

I think one effect of such a label would be to make quite a few people less vigilant when the label isn't present. So the net effect might be the opposite of what one might wish (especially since the label will certainly be missing from even obvious trollings initially).

I'm quite sure that another effect of such a label would be a lot of nodes being posted about how that label got applied to something that wasn't trolling (and a lot of those nodes would be authored by trolls, sure). It would also result in more consideration of nodes to try to get the "troll" labeled applied when whatever scheme failed to do so.

And anything that results in more attention being given tends to encourage trolling, not discourage it.

Given the recent case of several senior monks posting quite a few nodes in response to trolls where the subject of their own nodes contained "do not feed the trolls" and several exclamation points, I have little faith that such a label would even serve as much of a discouragement.

What do you think? ... Is it worth the effort?

I'd say it's not worth the effort. IMHO, everyone simply practicing for themselves to resist the temptation (i.e. just don't reply — even though it seems hard at times) would be a more effective measure against trolling...  By far most of the Monastery's residents are clueful people who can figure for themselves what is trolling and what isn't.

I think XP and trolling aren't closely related

You could please elaborate?

When I take a look at Worst Nodes of the Month or Year, a large part is what I consider trolling. For me that counts as "closely related".

You could please elaborate?

Usually, if someone makes a valid opinion which goes against the opinions of some of the monastery residents, she will get down voted. More often than not, if said someone's opinion is valid, she will get up-voted and those tend to balance themselves, but this tendency isn't definite, so a post that reaches low negative votes might be suspected to be trolling, but it also might not.

For instance, messily-formatted posts tend to get heavily downvoted to a point of double-figure negative reputation. Also, when a debate becomes engaging enough, or appealing enough for many people, the reputation tends to be high in both directions.

One way to handle this would be to add a description to the vote, slashdot-style. You could then look out for posts that are marked as flame or troll, questions is, whether such a system could be implemented here...

Stop saying 'script'. Stop saying 'line-noise'.
We have nothing to lose but our metaphors.

I think the hardest question is how to mark a node as being a troll post. A rather simplistic approach would be to set an XP threshold, for example -5, and consider everything below that trolling. Or we could reuse the consideration mechanism. Or have a different method altogether.

I personally believe that troll tagging would be nice. But basing it on node reputation would not be appropriate. In fact nodes by some reputable monks sometimes happen to be heavily downvoted, due to controversial opinions; but they're not trolls, by any means. (Incidentally, IMHO Win is not a real troll either, rather an annoying crackpot exhibiting trollish behaviour every now and then.) The consideration mechanism would be ok for this task.