Implement automatic delay for proposed taxon swaps

Platform(s), such as mobile, website, API, other: Website

URLs (aka web addresses) of any pages, if relevant:
https://forum.inaturalist.org/t/taxon-swap-lumping-recreation-of-a-lumped-name-under-a-different-taxon-id-and-potential-problems/38539
https://www.inaturalist.org/flags/590288

Description of need:
Committing of taxon swaps on the same day as they are proposed has created several (numerous?) problems when such swaps were inappropriate or arguable and require reversal. The above links offer just two examples from the realm of plants and moths.

Feature request details:
Institute an automatic waiting period of some substantial duration before any taxon swap can be committed. The waiting period should be long enough to garner the attention of experts (and anyone else) who might be interesting/involved with the subject taxa. Such a period would allow the opportunity to learn about, consider, and perhaps respond to a proposed swap if desired. As a starting point for discussion, I might suggest that a proposed taxon swap should be required to remain uncommitted for perhaps one to two weeks. In any event, no swap should be allowed to be committed within 24 to 48 hours of its proposal.

As someone who commits a lot of taxon changes with few to no observations, this would be a huge pain. Are you saying it needs to stay as a draft? Or that it would be committed but the changes don’t actually happen? Drafting a taxon change doesn’t notify anybody, although that might be nice. I’ve also found that frequently experts (if they even exist on iNat) don’t respond until a swap is committed, so asking input beforehand is frustratingly futile. Perhaps this could only apply to taxa with more than 100 or 1000 observations.

3 Likes

We already wait for months to have normal taxon changes to happen cause authors just leave them as drafts, imo if would be too big of a problem to wgat we already have, plus there’re drafts.

You could make this proportional to the number of obs. For example, if the swap affects no observations, it can be committed immediately. If it affects under 50 observations, make it have a 48h delay, if thousands, maybe two weeks, for example.

The proportional delay would not be hard to code, I think.

All in all, though, I agree with Marina. The problem of unprocessed or draft swaps is of a much bigger magnitude.

Personally, I think it would be far better if taxon swaps were more easily reversible. But in leu of that, I think a delay should be implemented as a simple precaution (I’m honestly not sure if it would help with my situation in the link, but it certainly sounds like it would have helped with yours). However, I think this should be modified as follows:

  1. The wait period should be proportional to the number of observations with a cap of a week or two. Taxa without any observations should go into effect immediately. (@mftasp looks like we came up with the idea independently!)
  2. We should distinguish between draft swaps, committed swaps, and committed swaps waiting to go into effect. Committed but waiting swaps would go into effect without any input after the delay. Input should take the form of a vote in addition to the comment period. In this case, a tied vote (one downvote under most circumstances) causes the swap to either return to draft form or enter a indefinite waiting period until a majority agree with the swap.
  3. The top 4-5 (maybe more?) identifiers for the taxon should receive a notification, not just the observers. I think this last one would be good regardless of whether a delay goes into effect or not. Identifiers are far more likely to care and have useful input than the observers in my opinion.
3 Likes

If we’re going to act on this, let’s also include a fix for something that I have seen occasionally: an inactive taxon remaining as an ID because there was no new taxon to swap it with.

2 Likes

Yes, I agree. However, just waiting is of little use if there is no discussion. Of course moderating a discussion is hard work.

Here’s an example of a proposed change that affects about 20K observations. Twenty users from the community were pulled into the conversation. That was the right thing to do, I think.

I understand that there are too few curators and too many flags, but when there’s an imbalance like that, mistakes happen more frequently. Some kind of change management is needed.

I agree with @gcwarbler

I probably didn’t think my Feature Request through as thoroughly as I might have.

@thomaseverest makes a good point about the lack of notification on draft taxon changes. I don’t know of a ready solution for that. Q: If one is subscribed to a taxon, do you receive notifications of proposed taxon swaps? I don’t think so. I’ve been subscribed to the lichen moth genus Cisthene for years and a couple of troublesome taxon swaps that happened in early 2022 eluded me; proposed and committed in one day. I don’t recall seeing a notification.

I tend to concur with the idea(s) of having a graduated wait time for swaps based on the number of extant observations, but I don’t know what those thresholds should be. I’m sure there’s no magic number. The issue, again, reverts back to notification and the chance for the community to have time to respond. I think in combination with some form of notifications, having a few set waiting periods for levels of # of observations might be sufficient. A caveat akin to the announcement at weddings might be attached to any notification: “If anyone objects to this [taxon swap], speak now or forever hold your peace!”

@Marina_Gorbunova I didn’t realize that there were large numbers of uncommitted taxon swaps. That is a problem, but it sounds more like a problem with the workloads or attention of the proposer. I hope there is some “sweet spot” compromise that avoids virtually instantaneous (same-day) swaps as well as those that linger unattended.

3 Likes

There should be in place some kind of “peer review” of proposed taxon swaps. I would be even more radical, and set a rule that the user creating the swap cannot be the same as the one committing it; but maybe there are not enough curators anyway.

3 Likes

I think a way to be automatically notified of pending taxon swaps would be helpful. It seems like I only know of them if someone pings me manually, or after they’re committed. Or maybe there’s already a notification system and I just haven’t found it…

Reversibility would also be a very good idea, as Nathan suggests.

1 Like

I think the big issue with taxon changes is that there is no notification to identifiers until after it happens. It would be great if that notification were sent out and there was a short waiting period before the name is changed with an option to stop or pause that change if discussion is needed. Bringing more people into the discussion or having to have votes or something could be very problematic though. There are a lot of people that just don’t like changes and they shouldn’t be able to stop a change just because of that. If they can supply good scientific evidence for why it shouldn’t change, that is fine though.

That said, if changes are just one to one, it doesn’t matter too much as it isn’t destructive to the identifications. I’m most concerned about lumpings. If lumpings happen that can’t be reversed, a huge amount of work that went into IDs could be lost.

3 Likes

It’s possible to make all changes to the taxonomy lossless and reversible. :-)

2 Likes

Making that happen should probably be one of the highest priorities.

3 Likes

I agree. I believe the limiting factor is that the powers that be are not willing to contemplate changes to how iNaturalist currently handles taxonomy and identifications.

2 Likes

Just made a feature request for this:
https://forum.inaturalist.org/t/notify-users-of-drafted-taxon-changes/38765

I also found another request similar to this thread:
https://forum.inaturalist.org/t/taxonomic-merge-swap-timer/581