I agree there is value in this, if it could be done in a way that wasn’t otherwise so risky. I wonder how hard it would be to create a separate “sandboxing” function for taxon changes, which would display all the same information as “live” taxon changes, but could not itself be moved forward as a “real” change?
6 Likes
I’ve done a lot of this for Pyrrocoma, as you might remember, and found it useful for analyzing how many IDs would be pushed up the taxonomic tree in atlas overlap zones. In other cases where curators seem to use the draft stage this way, it seems like common practice to include a conspicuous header, like “HOLD OFF ON COMMITTING—MORE WORK TO BE DONE!” I think it’s ok to keep using the draft stage this way rather than requiring a new sandbox feature. That said, as soon as you draft a taxon change, the input taxa get the “heads up” banner on their taxon pages, which advertises the change as if it’s already been decided. It would be better if the “heads up” banner appeared once a change had been submitted for votes, instead.
10 Likes
Perhaps voting would more accurately reflect consensus if voting could be both for or against a proposal and only allowed to proceed with some measure of majority.
4 Likes
I had some contact with staff on earlier tweaks to the vascular plant taxon change process, and the impression I came away with is that they had two major concerns.
First was the scalability and general speed of the process. How do we push down on the numbers of flags (and perhaps deviations), ungrafted taxa, etc. in the plant taxonomic tree? Discussions have a way of stalling out with no one willing to make a yes or no decision on some taxonomic change. Hence the push for closer alignment with POWO, to try to nudge this closer to the way things work with vertebrate groups which are usually pretty tightly synced with an external authority. Basically, try to push curating vascular plants in a direction that doesn’t require as much specialist knowledge or energy.
This led to the second problem, which is that POWO itself is something of a work in progress, and is sometimes at variance with specialist knowledge (in some subset of taxonomy or geography). When someone commits a taxon change that’s out of step with that, staff gets requests to internally reverse the taxon change, which is a serious drain on time and resources.
I am not sure the “5 votes” mechanism was meant to be a broad mechanism for determining community consensus. I speculate that, like the waiting period that came before it, this was meant to be a mechanism by which specialists could get their eyes on taxon changes and either express confidence that this was fine or register an objection, before the change got committed.
Another part of the problem that’s maybe been under-discussed is that there is a lot of uncontroversial routine maintenance (the limited ability to tweak scientific names has helped keep some of this out of taxon changes). If all your routine maintenance tasks now require 5 upvotes, you and your colleagues will start to get very casual about upvotes, and then… Separating the potentially controversial from the uncontroversial is very difficult, however.
(I will happily accept correction if I’ve misunderstood parts of the conversation with staff–they were gracious and open-minded and I don’t claim to speak authoritatively for their opinions.)
8 Likes
Broadly related to the topic, I think that drafting a taxonomic change or submitting an intent to change (which I think should always be a preceder) should broadcast a notification to all observers and identifiers of that taxon, either automatically or at a push of a button, and a minimum time should elapse from that act, unless a staff member is asked to override it. Perhaps that idea might be refined based on some threshold but basically any plant I know about I would be interested in hearing about an upcoming change, whether or not I add anything to the discussion. This would also mean @-ing people is not required.
2 Likes
I agree with the general idea here, but, given the issues with overwhelming notifications already, I wouldn’t want to push notifications of taxonomic changes in any taxon as user has identified to all users. This would hit sooooo many users with lots of notifications for things that they have no interest in and rapidly overwhelm people, especially observers who used CV to ID their own observations.
1 Like
POWO will always be a problem but I don’t think the issues we face can be traced to POWO per se. If iNat switched to WFO tomorrow, we’d still have more-or-less the same issues. I think the root of the problem is the lack of staff involvement in taxonomy change management. If a dedicated staff member were assigned to oversee the existing cadre of volunteer curators, the change management process would immediately become more structured and predictable. Few (if any) splits would need to be reversed. Sure, there will always be people who are unhappy with change, but with staff at the helm, that change could be managed to optimize cost/benefit.
It might be helpful to compare the taxonomy to software. If iNat software development were managed like the taxonomy currently is, we’d certainly have chaos. On the other hand, if the taxonomy were managed more like an open source software project…well, I think that would be a move in the right direction.
3 Likes
The idea can be tweaked - for example it could be simple opt-in with taxon roots, or with a default of going out to those who assessment suggests are going to be highly interested (and then have both opt in and out, any notification providing an opt out link)… In my case for example I would opt in and choose plants and it would go out for any taxon I’ve posted or if possible also identified, so basically all the observations it’s going to impact can notify the id’ers who have the option set…
3 Likes
I’m going back to highlight this question
From what i’ve seen on many flags etc, attempts at engagement before “updates” often can fall on deaf ears until potentially some take offence that the new format doesn’t fit with their viewpoint. Then suddenly opinions come out. This can be because the change then affects the statements made by users in the past, and of course, users need time to digest what’s setup as new, perhaps to verify the literature themselves, or more broadly to discuss with trusted folks.
As i saw it, this discussion arose in part because an issue gained the required “five votes” then was executed, arguably without taking onboard broader discussion or concerns. I’m sorry, but whether a change is implemented quickly or slowly, there will be some implemented updates (executed swaps etc) that in hindsight or with greater input should (in some counter-view) not have happened. My concern here is that those then need staff input, and in such cases i’m increasingly scared that will no longer happen whilst in the past there was a viable path which volunteer curators had available. Mistakes happen. Some of those are of course subjective and opinion based. Regardless, there simply must be a route for rapid appeal for reversal in the few cases where the path forward is being revised in broader discussion.
On this, i’m really concerned that context here stems from the forum post linked below about “two changes to reduce bottlenecks for curators”. I highlight then it said “Currently, two of the largest curation bottlenecks that require staff input are”. I’m sorry, but the community of volunteer curators are doing the vast majority of taxonomic curation (as i see it), and we sometimes vitally need staff input. I get that staff may aim to reduce the burden you get there at that end, but for the maintenance you all are relying on a community who are at times attempting to speak to you, and yes that sometimes will be about desire to reverse a subset of curator actions. Above are many great ideas about how to improve the system for those who invest in trying to improve your functionality, I’m not keen on those which focus towards additional delays or hurdles. My main message though is that sometimes the reversal of actions still seems the most effective route, and I fear that’s increasingly unavailable.
[https://www.inaturalist.org/posts/117974-two-changes-to-reduce-bottlenecks-for-curators]
7 Likes
Point well taken about engagement on taxon changes often being reactive rather than proactive. Some of that, though, is down to insufficient effort up front to identify and tag more interested stakeholders during the discussion phase.
What I saw was (1) a change submitted for voting before it was ready for voting (atlases incomplete), (2) curators voting without (apparently) familiarizing themselves with the status of the atlases and other discussion points, and (3) the change then committed by a previously uninvolved curator without any communication first with the involved curators.
I’m curious what that previous viable path was that curators had available? To my knowledge, reverting a taxon change has always required staff intervention, and was never a power curators had. Are you thinking that staff is trying to disengage entirely from curatorial oversight or intervention? I think that is unlikely on many levels.
I think the “bottlenecks” referred to in the blog post are in fact an acknowledgement by staff that
and that, absent a full-time ober-curator on staff, there will always be a lag time for responses from staff with multiple other responsibilities. How successful the recent changes have been in reducing the need for staff attention is still an open question to me, but I do know that the things that went wrong this time - all fully preventable - still could and did go wrong with other past changes before the current curator voting step ever existed. Voting was never a foolproof fix for sure, but it’s possible it has helped reduce issues with other taxon changes - we only know about the one that went wrong.
5 Likes
Absolutely, which goes back to core of the initiation of the discussion, i.e “some consensus-building process should occur”. Here i think several of us see and agree on failure of the curatorial process/actions.
RE “As i saw it, this discussion”
Thanks for clarifying viewpoint on the actions here. Totally i hold my hands up of likely having misrepresented the actions, process etc in this case. Hence your (@jdmore) clarification of “What I saw was” adds value. Others of course may view that differently, but essence of it being enacted in a way than then led to kickback.That’s really down to the five who voted - and i see no way to look back to the who/why etc which is poor. The lack of communication at various stages seems like the key here. My worry though is that i expect we can easily pull up counter examples of stagnation when plurality of views lead to stagnation and inaction.
Else for “My concern here is that those then need staff input” and reply of “Are you thinking that staff is trying to disengage entirely from curatorial oversight or intervention? I think that is unlikely on many levels.”
Absolutely.- I increasingly feel that us volunteer curators are yet again talking amongst ourselves and it’s all falling on deaf ears at staff end - if any of them are reading then great, but i expect to hear nothing to reflect any of them have even done that. My view on “previous viable path” is just a biased reflection about how i (perhaps mistakenly) viewed staff engagement in my rose-tinted view of past.
Then replying to “and that, absent a full-time ober-curator on staff, there will always be a lag time for responses from staff with multiple other responsibilities”/
Of course, i see that, yet Is it so ludicrous to hope there can certain folk we can look tor for direction on taxonomy of various biological lineages? For this is see an imbalance of direction and those with capacity to be implementing.
[Note, edited the next lines] Of course voting was never foolproof, but i don’t want to see assessment of any revised scheme solely said in terms of burden to staff input. Of course staff have their own workload burdens, but for taxonomic changes, i’m seeing the bulk of the burden is instead being asked of volunteers - where i see adding extra steps and policies as likely increasing burden.
4 Likes