Hello! I’m new to the forums, and an iNat observer for ~3 years. I started out with essentially no identification knowledge, and I’m still fairly clueless, but learning a thing or two through the help of other users has been a very enjoyable process. I also admin a project documenting biodiversity of a local forest.
I’m usually somewhat bold with the initial ID I place on my own observations (but timid when I attempt to ID for others). I’ve gotten plenty of corrections when my initial ID is outright wrong, and I try to withdraw quickly. However, there have surely been cases where my initial ID is possible but overly optimistic: e.g. I didn’t know there were lookalikes, the photo’s too blurry for a species ID, etc. Rarely have I gotten an ID where someone bumps it up to a higher taxon, and even more rarely is the “ID is as good as it can be” option selected. I think I understand why this is from the identifier side, since that sort of ID work is probably especially challenging, ambiguous, and thankless.
My questions are:
As an observer, is there anything I should I do if I suspect I may have been overly bold in my initial ID? I imagine at least that some identifiers have chosen to pass over it rather than offer a “buzzkill” ID. Tag someone and ask if they think there’s sufficient info for the ID?
As admin of a project that I would like to improve/curate, is there anything I should do about the bold initial IDs of others? There are a lot of species level IDs made using the CV, with photos that I suspect have insufficient information.
PS after drafting this I saw that there has been some discussion along these lines in the “Identification Etiquette” wiki, but I think my questions are still distinct.
I would suggest that the answer to both of your questions is to add a new ID to the observation that aligns with your current confidence level.
In both cases, add a comment with your reasoning whenever possible, to help others understand either why you changed your mind (on your own observations) or why you disagreed (on someone else’s ID)…
There will inevitably be some who are put off by the change, especially if it moves the community taxon to a higher rank and/or out of Research Grade. If you have done your best to explain your reasoning, then that’s on them, not on you. iNaturalist is all about suggesting IDs that eventually lead to a community consensus, even if there are disagreements along the way.
In other cases they may respond with reasoning of their own that turns out to be more convincing to you. In either case, the learning process for all involved is at least as valuable as the ultimate ID, and multiplies through the community to improve many more observations.
Backtrack to where you are confident, and make ‘overly bold’ a comment.
As (admin) identifier - push the ID back to where you are confident.
But be aware of hard disagreement triggering your UNintended Ancestor Disagreement. (skesau says it is not That plant but it is A Plant, and then skesau disagrees with ALL your plant IDs without even looking at them)
if I make an ID (on my own observations) that I am not too confident about, I usually add a question mark in the comment field or leave a comment about not being 100% sure about it. That way, hopefully users who have some expertise in the taxon might look at it and find it easier to agree or disagree.
For observations by others, I often don’t identify to species if I am a bit unsure and especially when I am the first one to leave an ID. The more I know, the less confident I become in my identifications …
Good luck with your project! Hopefully you can find someone with knowledge of your areas who can help with the IDs? I would find it a bit daunting to determine what is a ‘bold’ ID unless I can get a sense of the user making it.
if it’s not too much trouble, a comment left in the description or on the ID itself makes things much clearer. it also indicates that you are likely to be interested in an explanation!
I don’t spend much time explaining my process to randos because 9 times out of 10 they aren’t interested in it or have already moved on. but if I know someone is curious, I am eager to help out!
I personally don’t mind an overly-specific ID so long as you keep up with your notifications and are realistic about the extent of your knowledge. It’s easier to knock back an incorrect species ID than find something that isn’t labeled. maybe step back to Genus, Tribe, or Family.
I am also a happy observer that doesn’t ID much because I lack the knowledge, but I’d like to share my experience. I must say that I find myself often opting for a slightly bolder ID than I am confident with on my own pics, because I think to myself that it is more likely that someone will see it and confirm/fix it. Usually this ID I provide is backed by the CV’s recommendation, and either higher than species level (therefore not reaching RG easily), or if it is species level I try to do a quick search to make sure there aren’t similar looking species in the same region.
My fear is that if I genuinely ID all my pics to the true level of my confidence, they would mostly be at class/subclass level so less likely to be seen by an expert, and it feels like making more work for the community, when for the most part these guesses are not too far off (and when they are I try to withdraw my ID quickly). It’s also hard to be truly confident even with genera/species that I have seen a lot, because the knowledge I have is rarely from any formal education but rather through watching people ID my pics over time and a little poking around.
I do this less when I really know nothing, eg. with fungi, and I know that blurred bird/feather pics will be IDed relatively quickly even if I put Aves if they are IDable at all.
I am happy to hear if the community disagrees with my ways, and for example I am curious as to whether I am right about the fact that a non-species level ID is relatively harmless to guesstimate (because it won’t reach RG “easily”).
Sometimes I wish there was a way for people like me to provide a good first guess, while indicating that it’s just that, so that it is perhaps not included towards RG. I do get quite uncomfortable when my pictures reach RG purely with the help of only one additional ID. Or another thing I have thought about is that perhaps when the CV has a certain level of confidence, its top suggestion (with the poster’s confirmation) should be used as a guide to place images into an approximative “sorting basket”, without it counting towards RG either.
I think the process you describe sounds ok. I agree that a “bold” genus or higher level ID is less worrisome than a species level one, though I don’t think that bold species level IDs are necessarily a problem either - as long as one uses judgment and follows up with notifications! I often use the genus level CV ID when I lack expertise and can at least verify that the genus level ID is reasonable. I will also usually comment something like “ID based on iNat CV” to let others know my ID process. Genus level IDs are generally sufficient to get an observation in front of experts.
I sometimes make bold IDs on my own observations and add a comment, and, if I think a user may have confirmed my bold ID without expertise or with a blind agree, I will either bump my own ID back to genus, or use the Yes vote on “can the Community Taxon still be confirmed or improved?” to keep it in the Needs ID pool.
This is my philosophy. I’m more likely to get a correction from a species than I am genus, tribe, or family. Heck, I’m more likely to get anyone to look at it if I go to species. Not that I incorrectly ID things on purpose, but if I feel confident or confidently optimistic, I’ll go ahead and go to species.
And to your first point, I think you do corrections well. As an enthusiast, I usually defer to people who know more than me, and if I have a question as to why they disagreed, I’ll ask. Otherwise, I withdraw my ID and move along!
I definitely see that, too. All thanks to Cunningham’s Law, I assume. I.e., “the best way to get the right answer on the internet is not to ask a question; it’s to post the wrong answer.” If I keep my ID to something safe like family it can take years to attract IDs.
You don’t need to be certain to put a species name on an observation. If you were out on a hike with a friend and you saw that, would you say, “I think it’s a such-and-such”? If so, label it. If you don’t feel confident, add “tentative ID” or a series or question marks or some even a more elaborate statement like, “I think it’s such-and-such but I don’t really know the vegetation of your area.” Think of this as the first part of a conversation.
Then, watch your notations! If you are doubtful about somebody’s confirming ID, you can always withdraw yours and push the observation back to Needs ID.
For decades the best botanist in southwest Oregon was a self-effacing woman with absolutely no education beyond high school. She knew all the local plants. She was easy to ignore, but if you were smart you’d soon learn that was a mistake. The most formal education my late husband’s farming relatives had was through 9th grade but they taught me a LOT about plant identification (and other things). I myself have done a lot of identifying for some excellent and prolific observers and I would find myself saying, “I don’t know how to ID that fungus”, then stop to think, I’ve seen dozens of examples of that species posted and labeled by this person; I know exactly what it is. Then I’d provide a supporting ID. Formal education is great! (I say that as a person with way more formal education than makes sense.) But what is important is learning, not now you got the knowledge.
Sometimes if I think my ID on my own observation may be “overly bold,” I will at the same time vote Yes on “Can the Community Taxon still be confirmed or improved?” That way if someone comes along and blindly “Agrees” with my ID, it will stay in the “Needs ID” pool for others to weigh in on first, instead of going immediately to Research Grade. I can remove my vote later when I am satisfied with the confidence level.
I understand what you are saying. But when I see an observation that has been stuck at genus for six years without follow-up, I can’t help but wonder how many like you saw it and then passed by because they weren’t confident. How does that help anyone? Recently, I went ahead and started adding sections or subsections to such observations. I have only recently learned about them, and am certainly not at expert level; but in at least a few cases, my going ahead and being bold resulted in someone else adding an ID the very next day – remember, this observation had sat untouched for six years. It turned out I was wrong about the subsection; but the disagreeing ID moved it to subgenus, which is still better than genus.
Point being, a “bold” initial ID, even if it turns out to be wrong, can still be an improvement. Especially if it motivates someone who might otherwise have passed on by.
… maybe it depends a lot in the organisms one is IDing, but in my area of expertise there are hundrets, if not thousands of observations with bold initial IDs no spider experts ever touched, because nobody wants to deal with all the unsatisfying backlash of pushing them back to genus. Those bold IDs just lead to a situation with many wrongly IDed observations, which now give a very wrong impression of the whole situation and which did not even get one confirming ID at genus level, even if that would have been easy to do for any spider person. I am also talking years here! Very unsatisfying for both sides I think.
In the identify tab, it is possible to sort observations by date updated. This includes things like people adding a different ID.
So it may well be that it was not “being bold” (the implication being that previous users had lacked courage) that brought the observation to the attention of an IDer with expertise, but simply the new activity on the observation.
It’s quite possible that the observation had sat untouched simply because nobody had managed to get around to looking at it. I suspect lot of users prioritize newer observations (among other reasons: because looking at old “needs ID” observations can be a rather thankless task) and sometimes stuff just gets missed, particularly if the observation was initially entered without an ID or with a completely wrong ID.
I sympathize with this, and I realize I am probably not the typical user in this respect, but if I make an unjustified species ID on any of my spider observations, please know that you are welcome to push them back to whatever level is suitable. I generally try to ID my spiders in accordance with my limited knowledge, but sometimes I’ll get some wild idea or I’ll fail to check for similar species and I appreciate being corrected.
I do think it’s fine to go a bit beyond one’s level of confidence when ID’ing one’s own observations (whether this is based on book knowledge or experiental knowledge). Making mistakes sometimes is part of the learning process.
At the same time, I really would prefer that users do put at least a little thought into the ID they are adding, because it is generally less work and less stressful to provide a more specific ID for something than to correct a wrong ID.
Most of the wildly wrong and overly specific IDs I see do not seem to be the result of an informed guess, but rather uninformed guesses – following the CV suggestion without looking to see whether it makes sense, or without checking to see whether there might be other related species that look similar. Users who enter honest wrong IDs (e.g., because they have a little bit of knowledge but don’t really understand what the relevant characteristics are) tend to make different mistakes than the ones resulting from indiscriminate use of the CV.
honestly, I’ve hardly ever run into pushback when identifying older observations. Either the uploader is long gone, or they’ve stuck around long enough to be a regular. For better and for worse, you’re most likely to interact with the observer when you stick to obs less than two weeks old.
I do not complain about honest mistakes… that is fine, I have done them a lot myself, thinking I have a strong case for a certain ID and being corrected later…
I talk for example about a pair of species in Europe that from the current body of knowledge cannot be distinguished by habitus… only few observations are rightfully IDed at species level, because the genital features were presented as well. Nonetheless, there are over 2000 observations IDed to species, half of them even RG! That it is only 2000 observations is due to the fact that I and other had started correcting for a while and at the moment a fearless ID-warrior is still going on with it. I had to stop at some point, because the backlash was too much for me and I value my time too much to spend it arguing like this (it is a extreme case for sure, but for those species a lot of people seem to feel emotionally attached … newer users as well as people using iNat for a long time and 1000s of observations). And I suspect other identifiers thinking along similar lines is the reason, why so many of those observations are never touched, because one would need to correct them actually… so better not do anything with them and leave them be. It is a pitty for the observers that just made an honest mistake and are willing to learn… I for sure encountered several of those as well.
There are a lot of other taxa like this in Europe alone that might go back up to 9 years.
So in certain taxa, if you are bold, be prepared that you might not even get a higher level agreement, even if this would be easy to do.
There’s taxa where I go through and correct occasionally if I get bored but… I just don’t have the energy to argue with people, usually. For taxa that don’t have a ton of observations, its not too difficult to correct- there’s a few lesser-IDed taxa that I’m very confident with that I just set to notify me when new observations come through - but when there are thousands of IDs of one taxa its really mentally taxing to try to make a dent in the inertia.
Especially since in most of the cases, I wouldn’t be suggesting another species level ID, I’d just be bumping it up to genus/subgenus/section and that can frustrate folks.
(EDIT: though to be fair, most of my IDs anymore are from looking at sequences, I don’t have as much time to do other IDS)