Change wording used by the system when downgrading an observation to an higher level taxa

Actually you are not, you are saying (even if it is not your intention), this is any possible member of the genus except the one you identified it as.

2 Likes

The whole point of this (by now very long) thread is that the system doesn’t make any distinction between these two, and in fact explicitly says that insufficient evidence for the species (i.e. what @mftasp is doing) is a disagreeing ID.

Personally, I add a comment saying whether it could be or definitely is not the species given as the previous ID, but it would be nice if the iNat had the distinction built in.

2 Likes

Yes, combining that suggestion (which has been made elsewhere in the forum) with the existing “green button” allows the observer to keep the observation listed as the species they believe they saw while blocking RG. A good choice if the evidence neither confirms nor denies the species-level ID.

A better third option – Michele, would this fit with what you are suggesting? – would be to add a button that doesn’t change the ID but does vote against RG at species level.

The comment box would then say: “Your ID might be right but I don’t think this should become RG unless you can explain how _______ can be ruled out.” If the identifier fills in the blank with at least one taxon, an extra vote will now be needed to reach RG.

Or, the observer respond and convince the identifier to remove the block. For example, in Charlie’s story:

Charlie or someone else could reply that the ID had been confirmed by [the experts]. Or, if the observation showed a bird, the observer might reply “Look, I know it’s not a great photo, but I got a good look at the bird and it had a flycatcher beak that was bi-colored and showed the distinctive pattern on the tailfeathers illustrated in ___.”

I would withdraw my objections in a heartbeat based on either of those replies.

3 Likes

When someone is only offering a genus level, or less specific level, ID, after a species level ID is already assigned to an observation, iNaturalist asks this sometimes misleading question: “Is the evidence provided enough to confirm this is (the prior genus and species ID given, maybe ‘Homo sapiens’)?” Then iNaturalist has a green box labelled "I don’t know, but I’m sure this is a (the potentially disagreeing identifier’s less specific ID, maybe Homo), or a red box labelled “No, but I’m sure it is a member of (his less specific level ID, maybe Homo)”

Some people, especially those who have extensively studied the group in question, interpret the first question as indicating that the person, or people, offering the earlier, more specific ID, need to have shown enough “evidence” to “confirm” (or prove) the ID, then they check the red box, with a comment telling anyone that made the earlier ID that they can’t make their species level ID without doing a genital dissection, or maybe doing a microscopic examination of a given feature, or otherwise using the technique they learned was necessary to identify it by, in some key, or after reading some paper.

By checking the red box they have vetoed the prior ID, without either offering a conflicting ID, or without saying they know the first ID is wrong. I’ve seen more than one of these people going through the ID’s of these, often difficult to distinguish, taxa, in a group they have studied, clicking the red box, vetoing the previous ID without saying they know the ID is wrong, but only saying the other people can’t identify that taxon to species without using the technique they claim is required. This can get the people who made the earlier ID very irritated, when their ID was vetoed, without the person vetoing it being able to dispute their ID.

I would urge iNaturalist to altogether drop the first question “Is the evidence provided enough to confirm this is (the prior genus and species ID given, maybe ‘Homo sapiens’)?” that could imply that anyone with the earlier ID needs to have “provided enough evidence” to “confirm” the ID.

The green box could then read: “I know it is (the less specific ID, eg. genus that new identifier is offering, eg Homo), but I don’t know that it is (the more specific ID, maybe Homo sapiens)”

The red box could then read “I am sure it is not (eg Homo sapiens), but I agree that it is a member of (eg Homo)”.

1 Like

I noticed the same problem with Human Papillomavirus. There are several genera with that common name, but on iNat, using that common name makes it Genus Alphapapillomavirus. So then, inevitably someone has to send it back to Family Firstpapillomavirinae, which I doubt any layperson would have thought of.

I disagree with this idea. As was discussed earlier with the sedges, maybe the previous people know a way to identify it that the dissenter does not know. Why should someone who lacks a key piece of knowledge have that kind of power over people who have it?

Fine if that was really how it worked. But as currently done, it does in fact go as “I can’t, so you can’t either.”

Yes. In the new thread that just got merged with this one, that is the crux of the problem.

charlie would still have the option of opting out of community ID, which would keep it at species level. Or, if he gives more context in a comment then there’s a good chance identifiers will withdraw their IDs.

I think regular identifers should know that charlie, for example, is experienced with plant ID and probably knows information about the plant that isn’t presented in the observation and would pass over the observation or add a “green button disagreement”.

However, if an identifier sees an observation identified as the same species, but with the computer vision symbol and from a user with no profile description and only a few observations, they can probably assume that the observer has no idea how to identify sedges and just chose the CV suggestion. There’s a wide grey area between and there’s been lots of discussion on how to deal with that.

There’s a major bias for newer observers and even more experienced ones to identify to species level when that’s often not appropriate. The orange button is the best way to address that false precision, but identifiers have a responsibility to follow up if it turns out the species ID was realistic.

4 Likes

It annoys me the other way round. If I identify a species within someone else’s suggestion of a genus or family, etc., why does it say I “disagree” when I am actually agreeing!

do you have any examples of this?

1 Like

It shouldn’t asking if community taxon is higher than what you are trying to add.

You will be on a different branch of the taxonomy for this to happen. Open the species page (in a seperate browser maybe) and at the top of the page click the three dots in the middle of the taxonomy branch list, and see if the other identifiers genus or family is an ancestor of your species level ID… If it is, then please do give a screenshot or link so we can check it out!

The title of this Feature Request fits what I’ve been thinking about lately. I’m been noticing observations where a person has disagreed with a species ID of a flowering plant by adding an ID of “Plants.” The person probably did not intend to disagree that the organism is a flowering plant, because the wording that results is to the effect that there was a disagreement with the species. This seems to then require 2 more identifiers to get the community taxon back to “flowering plants.” So looking through this topic, I see wording in @loarie’s June 2019 post, under “Branch Disagreement” which (in an example involving lady beetles) says “I disagree that this is Asian Lady Beetle and all taxa on the branch between Asian Lady Beetle and the taxon I have proposed,” which seems to me much more accurate. This topic is still open, so I’m here to ask whether there is a chance of this type of language being adopted on the identifications. I have to admit I didn’t re-read this entire discussion.

3 Likes

In PNW USA Bombus, we have two veery similar species, Bombus vosnesenskii and Bombus caliginosus. In photos you can tell them apart only if you can compare the malar space and/or see some yellow hairs underneath the abdomen. In our Washington Native Bee Society Project we decided to move any B. vos back to Subgenus Pyrobombus that did not show underneath the abdomen. Then one of our members made a fantastic Pyrbombus Observation Field which includes that pair and many other Pyrobombus which can mimic one another. As B. vos is oue most spotted species (1338 counted by Nov 2021) it has been a very big pain to write to each person to explain that yes their ID is probably correct by because of the very similar B. caliginosus we would appreciate them using the Observation Field and the Pyrobombus Subgenus. We hope that eventually the Computer Vision will be able to tell the difference and people have been great about taking this one - but it is a real pain to write over and over and over and over…

2 Likes

I’d argue maybe we shouldn’t be doing this at all. If the species can’t be told apart is it worth disrupting the database and causing all these problems? You’ll never be able to keep a scientifically valid bombus vos on iNat since the species can barely ever be identified in here.

:man_shrugging:

Of course we shouldn’t just live species out because someone can’t id it? It is an identifiable species, how is it “can’t be told apart”? By that logic we shouldn’t check any Bombus sp. ids at all as most of them require many shots of many parts and cv always tells you it’s terrestis if it’s anything like subg. Bombus. Problem is in cv use, not taxa. We have complexes on iNat right for that, and subg. id is a good one.

1 Like

hmm, that isn’t what i said at all. i said maybe we shouldn’t be downgrading them if “we” can’t tell, like someone should only downgrade if they think the ID is wrong, not if they don’t know. I just feel like it breaks a bunch of the site without offering any clear benefit ig the species can barely ever be gold apart. But, it’s fine to disagree, per usual my opinion is probably not the majority one. Just putting it out there. I am not speaking to any official iNat policy as a moderator.
(I also think the observer should be probably knocking back to subgenus after being presented with this evidence, if they aren’t sure)

If user saw features not seen on photos, then they should write them down so others will see them, or write after that id, right? It doesn’t break anything to clear up mess, bumblebee species are not some microspecies and are identifiable, it’s people who make one shot of insect and think it’s enough for species id, it’s perfectly okay if it will stay at higher level.
People don’t change ids even when you tell them there’s no way to id species from that single pic of wrong for id angle, they don’t care, e.g. Metellina segmentata observations with dorsal photos when you can’t id species from the back, people use cv, I bet they didn’t know genus name before that, thus don’t know how to id the species, or they’d do better pics! Same with tons of other species, e.g. Flammulina species are idiable only with spores, but general public knows about one species, you can’t guess species from general appearance, so nobody knows what this specimen is, so it should be at genus level, or Alchemilla sp. where cv knows only one, but species id requires flowers?
It’s iNat design flaw, ofc people shouldn’t use disagreement that way, but users usually too new to know or care about how ids should be done, so what’s the difference if there’s a disagreement + “no it can’t be better” = RG genus or believing observer will answer you one day? I see such observations all the time, months or years after comment of ider, it’s not working.

4 Likes

I guess i don’t see either approach as working very well, it is true that people often don’t actively curate their observations, and i don’t know if that can be improved. But conversely i see people going on sprees of knocking things back to higher taxonomic levels and just sweep through taxa or through all the observations of an individual user, knocking back to casual or higher taxonomic level things that many other people could identify and also disenchanging the community. Then some of them reject community ID and it’s all for nothing anyway. And, as with other stuff, i don’t feel like iNat is hitting the balance as well as it could be, but of course we don’t all agree on where the balance should lie.

1 Like

i guess i said it better in this first comment, and feel about the same about it, but the situation hasn’t changed much.

I agree there’re better ways to spend time, but if you’re actually working on particular taxon, it makes sense to check all the observations of it, I also against making things casual as some iders do, saying casual are observations too, when actually they’re just gone for iNat after that, so there always should be a golden middle.

1 Like

i actually think it would be neat to find a way to track how many times something was reviewed, and if like 20 people have reviewed something and hit R with no ID, bumping it into a different category “difficult, requires expert or else can not be ndentified” that could be treated differently. At least have another step before we automatically throw them into the casual or coasre ID pit. I recently had a case where one person identified stuff to species that should really be to genus, and another user went through a ton of their observations and used the ‘can not be identified further’ button plus a very coarse ID to ‘vascular plants’ to knock a bunch of stuff into casual status. Most of them were identifiable to genus or at least family. I went through and identified them as such but if i hadn’t specifically noticed the issue they’d probably be lost forever. That’s the thing. Yes, sometimes photos are way insufficient for an ID an we shouldn’t just leave them in Needs Id to get milled over by multiple people wasting everyone’s time. But i also think people get very overactive with saying things aren’t identifiable. To be blunt, i am really good at identifying plants by texture and slight color variations to the point lots of people have commented on it. (my ‘plant synesthesia thing’ noted in the neurodiversity thread even works somewhat for photos). So i find all kinds of stuff knocked back to high taxonomic levels that I am totally confident in what it is. That doesn’t mean i don’t make mistakes too of course… but i just feel like people are too overactive with this stuff and it results in a big loss of good data. Maybe bees were a bad place for me to try to make this point. I’ve got no bee synesthesia :sweat_smile:

4 Likes