Skip to first unread message
AfriBats
May 25, 2016, 5:48:12 AM5/25/16
to iNaturalist
Hi everyone!
I stumbled over this "taxon" http://www.inaturalist.org/taxa/490704-Psychodidae-st1, which was created to accommodate an observation of an unidentified moth fly (www.inaturalist.org/observations/1456250).
I flagged this taxon and reached out to the user who created it as I think this approach, if adopted widely, would lead to taxonomic confusion. As outlined in the curator guidelines, iNat is not a place to argue about taxonomy (although it still happens for good reasons and quite frequently), and iNat has decided to stick to established taxonomies where this is possible. I think there are indeed good reasons not to create parallel universes within iNat, especially as these are unlikely to link up with any external database.
In this particular case, the moth fly could be 1) an undescribed species or 2) described but simply not yet identified to species or even genus level. Stephen Thorpe, who created the dummy taxon, wrote "This site is about species, not about names. Why should the lack of an "official" name (binomen) prevent us from mapping and tracking an interesting species? Please drop the issue, as it is not constructive to persue it further."
My suggestion was to use Tags, Observation Fields, or the Description to create some sort of user-defined ID (as would be done with OTUs in taxonomy before a species is formally described) rather than creating a completely undefined taxon within iNat's taxonomy that is of little if any use to iNat's community (as it lacks a description, characters etc that other users could use).
I'm curious to hear what others think, and hope we can find a consensus. Cheers, Jakob
I stumbled over this "taxon" http://www.inaturalist.org/taxa/490704-Psychodidae-st1, which was created to accommodate an observation of an unidentified moth fly (www.inaturalist.org/observations/1456250).
I flagged this taxon and reached out to the user who created it as I think this approach, if adopted widely, would lead to taxonomic confusion. As outlined in the curator guidelines, iNat is not a place to argue about taxonomy (although it still happens for good reasons and quite frequently), and iNat has decided to stick to established taxonomies where this is possible. I think there are indeed good reasons not to create parallel universes within iNat, especially as these are unlikely to link up with any external database.
In this particular case, the moth fly could be 1) an undescribed species or 2) described but simply not yet identified to species or even genus level. Stephen Thorpe, who created the dummy taxon, wrote "This site is about species, not about names. Why should the lack of an "official" name (binomen) prevent us from mapping and tracking an interesting species? Please drop the issue, as it is not constructive to persue it further."
My suggestion was to use Tags, Observation Fields, or the Description to create some sort of user-defined ID (as would be done with OTUs in taxonomy before a species is formally described) rather than creating a completely undefined taxon within iNat's taxonomy that is of little if any use to iNat's community (as it lacks a description, characters etc that other users could use).
I'm curious to hear what others think, and hope we can find a consensus. Cheers, Jakob
Stephen Thorpe
May 25, 2016, 6:21:11 PM5/25/16
to iNaturalist
Hello,
I urge strongly against jumping to conclusions here without a full rational consideration of reasons both for and against the careful use of tag names. As I said, iNaturalist is about species, not about names. Species are species whether they are named or unnamed, so they should be treated in the same way (i.e. mapped etc.) regardless. I'm not sure why Jakob appears to have "a bee in his bonnett" about this issue, but his reasons against tage names are "obscure" at best. He says "I think this approach, if adopted widely, would lead to taxonomic confusion". Really? How exactly? What kind of confusion? He says "iNat is not a place to argue about taxonomy". I agree! Using tag names involves nothing remotely related to "arguing about taxonomy"! I just don't see any connection here. He says "iNat has decided to stick to established taxonomies where this is possible". Again I agree! But it is not possible to stick to an established taxonomy when there is no name for a species (i.e. there is no established taxonomy for that species). I am only too happy to find an equally good alternative way of mapping and tracking unnamed species, if there is one, but I'm pretty sure that there isn't. Jakob's suggestion of using Tags, Observation Fields, or the Description to create some sort of user-defined ID isn't very satisfactory. For example, it doesn't facilitate accurate user stats on numbers of species uploaded, it doesn't automatically create seasonality information (i.e. no. of observations in each month), and it doesn't allow the IDs to be easily updated if a species does finally get a name (which is easily done by merging tag name into the new name). Overall, I suggest that the careful use of tag names (and only curators can create them, so we need not fear all users creating them willy nilly) has more benefits than costs, and that it improves the scientific credibility of the site.
Cheers, Stephen
I urge strongly against jumping to conclusions here without a full rational consideration of reasons both for and against the careful use of tag names. As I said, iNaturalist is about species, not about names. Species are species whether they are named or unnamed, so they should be treated in the same way (i.e. mapped etc.) regardless. I'm not sure why Jakob appears to have "a bee in his bonnett" about this issue, but his reasons against tage names are "obscure" at best. He says "I think this approach, if adopted widely, would lead to taxonomic confusion". Really? How exactly? What kind of confusion? He says "iNat is not a place to argue about taxonomy". I agree! Using tag names involves nothing remotely related to "arguing about taxonomy"! I just don't see any connection here. He says "iNat has decided to stick to established taxonomies where this is possible". Again I agree! But it is not possible to stick to an established taxonomy when there is no name for a species (i.e. there is no established taxonomy for that species). I am only too happy to find an equally good alternative way of mapping and tracking unnamed species, if there is one, but I'm pretty sure that there isn't. Jakob's suggestion of using Tags, Observation Fields, or the Description to create some sort of user-defined ID isn't very satisfactory. For example, it doesn't facilitate accurate user stats on numbers of species uploaded, it doesn't automatically create seasonality information (i.e. no. of observations in each month), and it doesn't allow the IDs to be easily updated if a species does finally get a name (which is easily done by merging tag name into the new name). Overall, I suggest that the careful use of tag names (and only curators can create them, so we need not fear all users creating them willy nilly) has more benefits than costs, and that it improves the scientific credibility of the site.
Cheers, Stephen
Stephen Thorpe
May 26, 2016, 5:01:02 PM5/26/16
to iNaturalist
Thinking about it, if you are worried about the use of tag names, then there is, I suggest, a simple solution which I'm sure the developers could implement rather easily, if you requested it. We just need 3 classes of names (instead of just 2), i.e. scientific names, common names and tag names. Tag names would function exactly like scientific names, except:
(1) Research Grade would be disabled for observations IDed with tag names;
(2) they would be clearly marked as tag names, e.g. by having "[Tag name]" appended to them; and
(3) they would not be displayed in italics.
Cheers, Stephen
(1) Research Grade would be disabled for observations IDed with tag names;
(2) they would be clearly marked as tag names, e.g. by having "[Tag name]" appended to them; and
(3) they would not be displayed in italics.
Cheers, Stephen
James Bailey
May 27, 2016, 7:07:31 PM5/27/16
to iNaturalist
I've mentioned before about the possibility of "non research-grade" taxon but I don't recall the result. I think there was a concern about having 100s of "non-published" names. iNaturalist isn't necessarily taxonomy central but I think the goal is to at least have sources backing the scientific names we use. That of course cannot happen with a new or undescribed species until the relevant publication exists.
Besides not letting research grade happen, the other possibility is that perhaps research grade submits them at only genus level instead of not at all.
James
Stephen Thorpe
May 28, 2016, 6:55:00 PM5/28/16
to iNaturalist
Hi James,
I think there is some confusion about the issues here. I agree that "non-published names" should not be used on iNaturalist, but by that I mean names that look like published names (e.g. Jamesius baileyi), but I see no problem with something like Psychodidae sp. ST1, which is self-evidently just a dummy or tag name. There is no relevant difference between recognising a species and giving it a tag name, vs. recognising a species and finding a published name for it (i.e. identifying it). Since iNaturalist is primarily about species, not names, the important thing is recognising species. Why should we miss out on tracking/mapping a species just because there is no published name for it?
Cheers,
Stephen
I think there is some confusion about the issues here. I agree that "non-published names" should not be used on iNaturalist, but by that I mean names that look like published names (e.g. Jamesius baileyi), but I see no problem with something like Psychodidae sp. ST1, which is self-evidently just a dummy or tag name. There is no relevant difference between recognising a species and giving it a tag name, vs. recognising a species and finding a published name for it (i.e. identifying it). Since iNaturalist is primarily about species, not names, the important thing is recognising species. Why should we miss out on tracking/mapping a species just because there is no published name for it?
Cheers,
Stephen
AfriBats
May 29, 2016, 8:33:51 AM5/29/16
to iNaturalist
It seems I haven't been able to explain my concerns, so I'm trying again:
iNaturalist is about observing the living world, and putting - established - names on these observations as far as this is possible based on the presented evidence (photos and/or sounds, in some cases maybe with a link to a specimen in a natural history collection, or a DNA sequence on genbank, BOLD...).
In many cases, especially in 'obscure' and neglected groups (mostly small stuff), observations are not IDed down to species but frequently stuck at various higher taxonomic levels, depending on the expertise of the iNat community.
In the specific case, it could be that 1) the species is indeed described and identifiable from the photos, but there's currently no one on iNat with the necessary expertise to provide the available name, that 2) the the species is described, but belongs to a group that cannot be IDed from photos alone, or that 3) it represents an undescribed taxon.
If 1), then ID at an appropriate, higher taxonomic level [family, subfamily, genus] with the prospect of having an expert identifying it further. If 2), also ID at higher taxonomic level, and flag it under Data Quality Assessment that community ID cannot be improved. If 3), ID at higher taxonomic level, and use Tags, Observation Fields, or Description to highlight that the observation belongs to a new taxon. Once described, the then available name can be added to iNat's taxonomy and applied to the observation.
Adding dummy taxa to iNat's taxonomy helps no one but the user who created these as there's no associated description that could be used by others. It also adds subjectively defined OTUs to an otherwise established taxonomy, where described names have been gone through peer review, taxonomic revision etc. Last but not least, such names won't align with any taxonomies when data are shared with external DBs.
If it's all about gathering spatio-temporal data of what a user considers an identifiable species (but without knowing its name), then using 1 of the 3 options (Tags, Obs Field, Description) should do the job to narrow down a search within the next higher, established taxon.
Jakob
iNaturalist is about observing the living world, and putting - established - names on these observations as far as this is possible based on the presented evidence (photos and/or sounds, in some cases maybe with a link to a specimen in a natural history collection, or a DNA sequence on genbank, BOLD...).
In many cases, especially in 'obscure' and neglected groups (mostly small stuff), observations are not IDed down to species but frequently stuck at various higher taxonomic levels, depending on the expertise of the iNat community.
In the specific case, it could be that 1) the species is indeed described and identifiable from the photos, but there's currently no one on iNat with the necessary expertise to provide the available name, that 2) the the species is described, but belongs to a group that cannot be IDed from photos alone, or that 3) it represents an undescribed taxon.
If 1), then ID at an appropriate, higher taxonomic level [family, subfamily, genus] with the prospect of having an expert identifying it further. If 2), also ID at higher taxonomic level, and flag it under Data Quality Assessment that community ID cannot be improved. If 3), ID at higher taxonomic level, and use Tags, Observation Fields, or Description to highlight that the observation belongs to a new taxon. Once described, the then available name can be added to iNat's taxonomy and applied to the observation.
Adding dummy taxa to iNat's taxonomy helps no one but the user who created these as there's no associated description that could be used by others. It also adds subjectively defined OTUs to an otherwise established taxonomy, where described names have been gone through peer review, taxonomic revision etc. Last but not least, such names won't align with any taxonomies when data are shared with external DBs.
If it's all about gathering spatio-temporal data of what a user considers an identifiable species (but without knowing its name), then using 1 of the 3 options (Tags, Obs Field, Description) should do the job to narrow down a search within the next higher, established taxon.
Jakob
Stephen Thorpe
May 29, 2016, 5:19:48 PM5/29/16
to iNaturalist
Yes, I understood your "concerns" the first time, but they are at best minor concerns which are outweighed by other reasons. It does sound a bit like you are simply trying to dictate that everybody else follows your own preferences. I'm not going to explain yet again why tage names are a good idea overall, and not, as you seem to think a bad idea overall. One thing though, you say "Once described, the then available name can be added to iNat's taxonomy and applied to the observation". Yes, but this would have to be done one observation at a time, whereas a tag name can be simply merged once into the new name, saving much time and effort. You also say "It also adds subjectively defined OTUs to an otherwise established
taxonomy, where described names have been gone through peer review,
taxonomic revision etc." Well, that's a trifle naive! At any rate, even if described names are perfectly objective and reliable, the user still has to make the identification, and this step will only be as reliable as the user, so we are no better off.
Stephen
Stephen
Stephen Thorpe
May 29, 2016, 5:23:38 PM5/29/16
to iNaturalist
PS: Oh, and before anyone tries to claim that I am simply trying to dictate that everybody else follows my own preferences, I am not trying to dictate that anyone else uses tag names. They, including Jakob, are free to use them or not as they see fit. However, Jakob is trying to stop me from using tag names. Why would he want to do that??
Stephen
Stephen
James Bailey
May 30, 2016, 12:15:56 PM5/30/16
to iNaturalist
I think one of the actual issues here is that we can't submit thing like "Genus st. 1" to datasets like GBIF. One of iNaturalist's biggest challenges is ensuring that submitted data follows the most even taxonomic ground possible. That's why we are always fighting with what taxon names to use. That is somewhat besides the point, but I think the only real problem here, if it is arguably a problem at all, is that tag names are either personal to one person or only applicable to a select group of people. The species has to have a formal description and a published name for it to readily fit into the database.
Else there is nothing stopping people just creating random new species tags for species that may or may not already be described. I certainly don't think that is what you are doing Stephen, but our "system" isn't one to give people the benefit of the doubt.
In any case carry on I suppose, at this point there is no harm in the practice.
James
Stephen Thorpe
May 30, 2016, 5:43:48 PM5/30/16
to iNaturalist
Hi James,
Thanks for your comments. I don't see it as a problem that we can't submit thing like "Genus st. 1" to datasets like GBIF. For a start, we should surely put the interests of iNat before those of GBIF, so if tag names are useful to us, then that's the main thing. That said, GBIF isn't significantly disadvantaged. Take the example of PSYCHODIDAE sp. ST1. Because it is a tag name, an observation isn't going to get to "research grade", so it won't be submitted to GBIF. The alternative is that the observation is IDed to family (PSYCHODIDAE) and raised to "research grade" and submitted to GBIF. But, psychodids are common everywhere, so a family-level ID of one from Auckland isn't actually of any use to anybody. So, GBIF have missed out on an observation (is that our problem?), but only an observation that isn't actually of any use to anybody! Is this a big deal? In general, IDs above species level aren't really very useful. Tag names give us the opportunity to turn them into more useful observations, by identifying them to species, just with a tag name. In reality GBIF is full of all sorts of problematic data anyway, including non-existent species names, so this issue isn't adding anything new to their problems.
You say "tag names are either personal to one person or only applicable to a select group of people", but I just don't see that at all! It also doesn't matter if tag names are created for species that are already described. A tag name is just a way to treat a set of observations as a single species, for the purposes of tracking/mapping/counting species. If a published name for a species presents itself, then we just merge the tag name into it. Nobody would use a tag name in preference to a published name for a given species! Again, I just don't see any significant problems here, but I do see benefits.
Cheers,
Stephen
Thanks for your comments. I don't see it as a problem that we can't submit thing like "Genus st. 1" to datasets like GBIF. For a start, we should surely put the interests of iNat before those of GBIF, so if tag names are useful to us, then that's the main thing. That said, GBIF isn't significantly disadvantaged. Take the example of PSYCHODIDAE sp. ST1. Because it is a tag name, an observation isn't going to get to "research grade", so it won't be submitted to GBIF. The alternative is that the observation is IDed to family (PSYCHODIDAE) and raised to "research grade" and submitted to GBIF. But, psychodids are common everywhere, so a family-level ID of one from Auckland isn't actually of any use to anybody. So, GBIF have missed out on an observation (is that our problem?), but only an observation that isn't actually of any use to anybody! Is this a big deal? In general, IDs above species level aren't really very useful. Tag names give us the opportunity to turn them into more useful observations, by identifying them to species, just with a tag name. In reality GBIF is full of all sorts of problematic data anyway, including non-existent species names, so this issue isn't adding anything new to their problems.
You say "tag names are either personal to one person or only applicable to a select group of people", but I just don't see that at all! It also doesn't matter if tag names are created for species that are already described. A tag name is just a way to treat a set of observations as a single species, for the purposes of tracking/mapping/counting species. If a published name for a species presents itself, then we just merge the tag name into it. Nobody would use a tag name in preference to a published name for a given species! Again, I just don't see any significant problems here, but I do see benefits.
Cheers,
Stephen
Ken-ichi
May 31, 2016, 3:28:00 PM5/31/16
to inaturalist
Hi folks,
I think you all raise some valid points, but I would really like to
hear from some other people. Is there anyone out there other than
Stephen who would use provisional names? I've heard privately from one
systematist on iNat who would, but I think deciding on a solution
would benefit from some other perspectives.
I think the main problem here is that the Curator's Guide just isn't
clear on the subject of provisional names, so mea culpa, but hopefully
we can come up with a reasonable policy here. It seems like so far
everyone agrees observations associated with provisional names should
not leave iNat, e.g. should not achieve Research Grade status and be
shared with GBIF, etc. Great, we can totally do that. Next question:
does this actually benefit anyone within iNat? Clearly it benefits
Stephen, and perhaps other specialists who might actually use this
kind of categorization in efforts to describe new species. Does that
benefit exceed the cost of confusion over some names that aren't tied
to any kind of externally verifiable description? Is there really that
much cost there if the only people using or making these provisional
names are a handful of iNat site curators?
I think a technical solution similar to what Stephen described (allow
provisional taxa but don't allow observations of provisional taxa to
become Research Grade) would probably satisfy those who want
provisional taxa without freaking out the rest of us too much, but
again, it would be good to hear from people other than Stephen and
Jakob.
FWIW, Mushroom Observer supports provisional names (or at least
unpublished names), and I *think* BugGuide does.
> --
> You received this message because you are subscribed to the Google Groups
> "iNaturalist" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to inaturalist...@googlegroups.com.
> To post to this group, send email to inatu...@googlegroups.com.
> Visit this group at https://groups.google.com/group/inaturalist.
> For more options, visit https://groups.google.com/d/optout.
I think you all raise some valid points, but I would really like to
hear from some other people. Is there anyone out there other than
Stephen who would use provisional names? I've heard privately from one
systematist on iNat who would, but I think deciding on a solution
would benefit from some other perspectives.
I think the main problem here is that the Curator's Guide just isn't
clear on the subject of provisional names, so mea culpa, but hopefully
we can come up with a reasonable policy here. It seems like so far
everyone agrees observations associated with provisional names should
not leave iNat, e.g. should not achieve Research Grade status and be
shared with GBIF, etc. Great, we can totally do that. Next question:
does this actually benefit anyone within iNat? Clearly it benefits
Stephen, and perhaps other specialists who might actually use this
kind of categorization in efforts to describe new species. Does that
benefit exceed the cost of confusion over some names that aren't tied
to any kind of externally verifiable description? Is there really that
much cost there if the only people using or making these provisional
names are a handful of iNat site curators?
I think a technical solution similar to what Stephen described (allow
provisional taxa but don't allow observations of provisional taxa to
become Research Grade) would probably satisfy those who want
provisional taxa without freaking out the rest of us too much, but
again, it would be good to hear from people other than Stephen and
Jakob.
FWIW, Mushroom Observer supports provisional names (or at least
unpublished names), and I *think* BugGuide does.
> You received this message because you are subscribed to the Google Groups
> "iNaturalist" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to inaturalist...@googlegroups.com.
> To post to this group, send email to inatu...@googlegroups.com.
> Visit this group at https://groups.google.com/group/inaturalist.
> For more options, visit https://groups.google.com/d/optout.
Charlie Hohn
May 31, 2016, 3:46:02 PM5/31/16
to inatu...@googlegroups.com
I don't feel strongly about dummy names, I will say that if we figure out a way to do it that makes sense I'd be in favor of creating groupings between species and genus for groups of species that often can't be differentiated in the field. That being said I think I proposed that in the past and was outvoted, so that is fine too. I realize it's a slippery slope and what seems impossible to differentiate for one person isn't for another... so maybe it is for the best.
Maybe in the long term we could establish a subset of curators for each taxonomic group, and only let them create dummy taxa (maybe by consensus) so that they don't proliferate. Though it does not sound like that is likely to happen at this point.
--
============================
Charlie Hohn
Montpelier, Vermont
Charlie Hohn
Montpelier, Vermont
BJ Stacey
May 31, 2016, 3:46:48 PM5/31/16
to inatu...@googlegroups.com
I would definitely use them. I have at least two species that are know to BugGuide as distinct species but are as of yet no formally described to science. Using such "temporary" names might also be of a benefit to researcher looking for data points for these species.
Sam Kieschnick
May 31, 2016, 4:01:13 PM5/31/16
to iNaturalist
I would not use dummy names, and I would think it would cause more confusion for the casual user. Take my two cents for what they're worth.
In the herbarium, researchers would study at a taxon, and they'd create these additional "sensu lato" and "sensu stricto" folders for the same species... So, we'd have three separate taxa represented: species, species (sensu lato), and species (sensu stricto). It was a major pain in the butt when the research left and NEVER came back to clear up that mess... As we had new specimens in the herbarium, rather than filing them in the new folders (s.s. or s.l.), we just filed them in the species folders.
Personally, when I cannot find the species on bugguide or a plant list, I would rather keep it at the genus level than narrowing it down to a new taxon. I'd leave in a comment, but for organization's sake, I would 'file it' in the higher taxon of genus.
~sam
In the herbarium, researchers would study at a taxon, and they'd create these additional "sensu lato" and "sensu stricto" folders for the same species... So, we'd have three separate taxa represented: species, species (sensu lato), and species (sensu stricto). It was a major pain in the butt when the research left and NEVER came back to clear up that mess... As we had new specimens in the herbarium, rather than filing them in the new folders (s.s. or s.l.), we just filed them in the species folders.
Personally, when I cannot find the species on bugguide or a plant list, I would rather keep it at the genus level than narrowing it down to a new taxon. I'd leave in a comment, but for organization's sake, I would 'file it' in the higher taxon of genus.
~sam
Charlie Hohn
May 31, 2016, 4:09:23 PM5/31/16
to inatu...@googlegroups.com
What about a way of tracking things as "To genus only but pending species description" or "undescribed species within this genus" or something? It would be different than casual grade beacuse the holdup isn't inability to ID, it is that the species itself needs to be described! T hen it would be easy to find those later.
Stephen Thorpe
May 31, 2016, 5:53:11 PM5/31/16
to iNaturalist
Hi Ken-ichi,
Just a couple of very important clarifications in response to your comments:
(1) you say "Does that benefit exceed the cost of confusion over some names that aren't tied to any kind of externally verifiable description?"
Actually, this is conflating two very distinct issues. A large number of species (particularly ones described long ago, and including many very common and well known species) cannot be identified from their original descriptions, and there may not even be a modern redescription. So, having a published name and having an "externally verifiable description" are two very different things;
(2) given (1), I reiterate that iNat is primarily about species, not about names, so species should be treated the same way regardless of whether they have a published name. It is arbitrarily to exclude species for which we don't have a published name. These species are in fact likely to be more interesting and important in some ways than named species, so we should facilitate data retrieval about these nameless species by treating them as a unit and not losing the data within that of a larger taxon (e.g. the large family Psychodidae in the present example).
Cheers,
Stephen
Just a couple of very important clarifications in response to your comments:
(1) you say "Does that benefit exceed the cost of confusion over some names that aren't tied to any kind of externally verifiable description?"
Actually, this is conflating two very distinct issues. A large number of species (particularly ones described long ago, and including many very common and well known species) cannot be identified from their original descriptions, and there may not even be a modern redescription. So, having a published name and having an "externally verifiable description" are two very different things;
(2) given (1), I reiterate that iNat is primarily about species, not about names, so species should be treated the same way regardless of whether they have a published name. It is arbitrarily to exclude species for which we don't have a published name. These species are in fact likely to be more interesting and important in some ways than named species, so we should facilitate data retrieval about these nameless species by treating them as a unit and not losing the data within that of a larger taxon (e.g. the large family Psychodidae in the present example).
Cheers,
Stephen
Scott Loarie
May 31, 2016, 6:12:20 PM5/31/16
to inatu...@googlegroups.com
I agree with Sam, filing things under the parent taxon works for me and adding some sort of additional complexity would incur a confusion cost that would outweigh the benefits
--
You received this message because you are subscribed to the Google Groups "iNaturalist" group.
To unsubscribe from this group and stop receiving emails from it, send an email to inaturalist...@googlegroups.com.
To post to this group, send email to inatu...@googlegroups.com.
Visit this group at https://groups.google.com/group/inaturalist.
For more options, visit https://groups.google.com/d/optout.
--------------------------------------------------
Scott R. Loarie, Ph.D.
Co-director, iNaturalist.org
California Academy of Sciences
55 Music Concourse Dr
San Francisco, CA 94118
--------------------------------------------------
Scott R. Loarie, Ph.D.
Co-director, iNaturalist.org
California Academy of Sciences
55 Music Concourse Dr
San Francisco, CA 94118
--------------------------------------------------
Stephen Thorpe
May 31, 2016, 6:32:44 PM5/31/16
to iNaturalist
I strongly disagree with Scott. Did he read my previous comments? What "works for him" doesn't necessarily work for all. As I said earlier, nobody is suggesting that it be made mandatory to use tag names. Scott doesn't have to use them if they don't interest him. But for me, working mainly with little known taxa of terrestrial inverts, it is crucial to be able to group observations together as species. Otherwise, potentially significant data gets totally lost and scattered under a large parent taxon. I still fail to see any substance in the claim that "adding some sort of additional complexity would incur a confusion cost that would outweigh the benefits". It is easy to say, if one is so inclined, but it is way too vague a statement to carry any meaning. We should be doing what we can to improve the scientific value of iNat. We should not be mere "nay sayers".
Stephen
Stephen
Sam Kieschnick
May 31, 2016, 6:47:17 PM5/31/16
to iNaturalist
Stephen has been an AMAZING curator (looking at his stats -- 3903 taxa
curated; 510 taxon changes
added; 1 flags
resolved), so I think if the taxon created has a diligent curator that monitors that taxon (even if it's a yet to be published species/subspecies), then I think it's ok to have these sorts of 'dummy taxa.' I do think it's important to have a detailed discussion about these created taxa though -- giving some of the reasoning behind why it's not a 'closely related' species. And I would foresee this being a comparatively rare instance with the number of species at the sp. level on iNat.
I just don't think that I would personally use them, and the many folks that I correspond with on iNat most likely wouldn't use them... Would there be dramatic confusion if taxa were created that don't have a specific epithet yet? Eh, probably not. But I would indeed hope that a creator of those taxa would be responsible for them and monitor them with the updated publications (as Stephen fits this perfectly)... Again, my experience in the herbarium dealt with researchers that were just present for a few weeks, played with the specimens, created some newly name folders, annotated a few specimens, and then bailed... It only created an unresolved mess.
Isn't taxonomy that way though? Infinitely unresolveable?!? :)
I just don't think that I would personally use them, and the many folks that I correspond with on iNat most likely wouldn't use them... Would there be dramatic confusion if taxa were created that don't have a specific epithet yet? Eh, probably not. But I would indeed hope that a creator of those taxa would be responsible for them and monitor them with the updated publications (as Stephen fits this perfectly)... Again, my experience in the herbarium dealt with researchers that were just present for a few weeks, played with the specimens, created some newly name folders, annotated a few specimens, and then bailed... It only created an unresolved mess.
Isn't taxonomy that way though? Infinitely unresolveable?!? :)
Stephen Thorpe
May 31, 2016, 6:59:55 PM5/31/16
to iNaturalist
Hi Sam,
Thanks for your comments. Yes, I am only too keen to ditch tag names for published names as they come to hand, and I know about most new publications relating to the N.Z. insect fauna before they are even published, so I'm a "coiled spring" for trying to resolve as much "mess" as possible as quickly as possible. My tag named species are listed in my continually updated browsable lists of N.Z. species on iNat comments pages. Again, a species is a species and all should be treated the same regardless of whether mor not they have a published name, and a published name is just that, it does not necessarily mean that the species is in any sense "objectively identifiable", certainly not necessarily identifiable from the accompanying description and/or illustrations. Species are often only identifiable by direct comparison to the type. So, using tag names adds nothing new to problems of identification.
Cheers,
Stephen
Thanks for your comments. Yes, I am only too keen to ditch tag names for published names as they come to hand, and I know about most new publications relating to the N.Z. insect fauna before they are even published, so I'm a "coiled spring" for trying to resolve as much "mess" as possible as quickly as possible. My tag named species are listed in my continually updated browsable lists of N.Z. species on iNat comments pages. Again, a species is a species and all should be treated the same regardless of whether mor not they have a published name, and a published name is just that, it does not necessarily mean that the species is in any sense "objectively identifiable", certainly not necessarily identifiable from the accompanying description and/or illustrations. Species are often only identifiable by direct comparison to the type. So, using tag names adds nothing new to problems of identification.
Cheers,
Stephen
Donald Hobern
Jun 1, 2016, 3:26:36 AM6/1/16
to iNaturalist
I agree that use of tag names would be really useful in many cases (probably in inverse proportion to how well-described the local fauna/flora is). There are dozens of species I know from my Australian garden which are clearly diagnosable but have no published name. Some of these have informally published working names (e.g. the "Moths of Victoria" series of booklets is doing a good job of giving diagnostic criteria for clear and obvious separate species), Most of these sit within large genera. An identification to genus is easy but represents much wasted information. I can say (with as much confidence as I can for many named species) that tag name X should be applied to a whole set of observations. Leaving the record just identified to genus reduces the probability that these observations will be discoverable and usable in the future.
I would think that a good solution would be for all tag names used in iNaturalist to be associated where possible with a resource that provides information on the undescribed taxon. Printed resources like "Moths of Victoria" or web resources like Popple's "The cicadas of central eastern Australia" (http://sci-s03.bacs.uq.edu.au/ins-info/index.htm) would be suitable. Where this is not possible, a short text (possibly with images) could be provided when the tag name is added.
From a (GBIF) infrastructure standpoint, I certainly hope that future versions of a comprehensive "catalogue of life" include mechanisms for tag names, BOLD BIN numbers, etc. to be captured and treated as recognised names related to species concepts, and for that reason I'd always prefer to see us collect data now in ways that maximise the possibility of making such links in the future. Observations which are associated with tag names but which otherwise meet research grade criteria seem to me to be part of the data space GBIF should ultimately be able to handle.
Donald
--
Jon Sullivan
Jun 1, 2016, 6:51:26 AM6/1/16
to iNaturalist
Hi everyone,
Interesting conversation. I would use tag names for IDs, and I do see a use for them. There are a great many invertebrate species without names, and some of them are common where people live and well recognised (as Donald just mentioned). It would be quite helpful if entomologists on iNat, like Stephen, could provide provisional names that we could use to explore the distribution, phenology, and natural history of these species in advance of their formal descriptions.
I've used fields to do this up until now, e.g., NZ Lycaena common copper complex tag name, NZ Lycaena boulder copper complex tag name, NZ Lycaena glade copper complex tag name, aff. (tag name). That's not been an elegant solution. As Donald said, keeping the IDs at genus-only isn't satisfying, especially for large genera (like Lycaena). You also can't easily make a map of those undescribed species tagged with fields. It's also not clear how all these observations we're tagging with fields could be quickly batch updated to the new names once those taxa are described. Plus, in the case of Lycaena in New Zealand, it has been encouraging people to instead use outdated names, like Lycaena salustius which was once applied to all our "common coppers" but is now thought to have been wrongly used and applies to a different taxon completely, split off from the rest of Lycaena rauparaha. That makes its own mess.
In the case of our NZ Lycaena tag names, there is a publication behind them (Brian and Hamish Patrick's 2012 Butterflies of the South Pacific) and they're working on the genus revision for NZ at the moment.
If we did allow provisional names on iNat, I like the idea of requiring citations or descriptions (and DNA barcodes). Nothing that constitutes a formal species description, but rather an informal guide for others on how to use the name.
I also agree that there's potential for confusion here, and potential for a disorganised proliferation of competing and contradicting tag names. I don't see it as the iNat's community's role here to decide what's a species and what's not, and it would be good if we weren't the place where lots of new tag names were introduced. I'd be more comfortable if provisional names added to iNat at least were expected to have some form of external corroboration. I'm not sure if that would meet the needs of people like Stephen though who are both the people recognising new undescribed species and adding observations of them to iNat.
I'm not sure how best to handle the risks here. I agree that the tag names need to be at least labelled clearly as separate entities from described species and not granted research grade status. Would we let any curator add these tag names? It might be better to open this to only those with a demonstrated taxonomic expertise, and in that case for only their area of expertise. They could then promote others in a way that curators can do for regular users now. And perhaps at least one curator should review and OK another's tag names? I'm not advocating for this, just floating ideas. Regardless, we'd need some detailed curator guidelines on what's appropriate here.
With my plant hat on, this is also the slippery slope that could lead to named cultivars of garden plants and domestic breeds of animals. We've just been using fields for these and provisional tag names that we expect to become described species are conceptually different. Still, we'd want to think through a solution that wasn't then used to load up an explosion of cultivar names. (Others may disagree.)
Cheers,
Jon
Interesting conversation. I would use tag names for IDs, and I do see a use for them. There are a great many invertebrate species without names, and some of them are common where people live and well recognised (as Donald just mentioned). It would be quite helpful if entomologists on iNat, like Stephen, could provide provisional names that we could use to explore the distribution, phenology, and natural history of these species in advance of their formal descriptions.
I've used fields to do this up until now, e.g., NZ Lycaena common copper complex tag name, NZ Lycaena boulder copper complex tag name, NZ Lycaena glade copper complex tag name, aff. (tag name). That's not been an elegant solution. As Donald said, keeping the IDs at genus-only isn't satisfying, especially for large genera (like Lycaena). You also can't easily make a map of those undescribed species tagged with fields. It's also not clear how all these observations we're tagging with fields could be quickly batch updated to the new names once those taxa are described. Plus, in the case of Lycaena in New Zealand, it has been encouraging people to instead use outdated names, like Lycaena salustius which was once applied to all our "common coppers" but is now thought to have been wrongly used and applies to a different taxon completely, split off from the rest of Lycaena rauparaha. That makes its own mess.
In the case of our NZ Lycaena tag names, there is a publication behind them (Brian and Hamish Patrick's 2012 Butterflies of the South Pacific) and they're working on the genus revision for NZ at the moment.
If we did allow provisional names on iNat, I like the idea of requiring citations or descriptions (and DNA barcodes). Nothing that constitutes a formal species description, but rather an informal guide for others on how to use the name.
I also agree that there's potential for confusion here, and potential for a disorganised proliferation of competing and contradicting tag names. I don't see it as the iNat's community's role here to decide what's a species and what's not, and it would be good if we weren't the place where lots of new tag names were introduced. I'd be more comfortable if provisional names added to iNat at least were expected to have some form of external corroboration. I'm not sure if that would meet the needs of people like Stephen though who are both the people recognising new undescribed species and adding observations of them to iNat.
I'm not sure how best to handle the risks here. I agree that the tag names need to be at least labelled clearly as separate entities from described species and not granted research grade status. Would we let any curator add these tag names? It might be better to open this to only those with a demonstrated taxonomic expertise, and in that case for only their area of expertise. They could then promote others in a way that curators can do for regular users now. And perhaps at least one curator should review and OK another's tag names? I'm not advocating for this, just floating ideas. Regardless, we'd need some detailed curator guidelines on what's appropriate here.
With my plant hat on, this is also the slippery slope that could lead to named cultivars of garden plants and domestic breeds of animals. We've just been using fields for these and provisional tag names that we expect to become described species are conceptually different. Still, we'd want to think through a solution that wasn't then used to load up an explosion of cultivar names. (Others may disagree.)
Cheers,
Jon
Charlie Hohn
Jun 1, 2016, 10:19:04 AM6/1/16
to inatu...@googlegroups.com
Conversely, as a plant guy, what is the downside of having cultivars on iNat? Granted they are not natural, but sometimes they are important ecologically. For instance there has been some discussion of the idea that cultivars of native plants in New England or California, USA, are providing much less pollinator value than the normal native varieties. Some ability to track cultivars as part of pollination data, etc could be really important. For instance the ability to tag them in the 'interaction:' fields. Or at least some way to formally differentiate a cultivar from the 'regular' species. Tags are good, I use them a lot, but they may not be enough for this.
As for proliferation, who cares? Globally, data storage is pretty much increasing faster than humans are able to create data and creating a separate cultivar doesn't seem like it would add much to iNat server costs. Same for tag species. What is the real harm of them? As long as they can only be created by some curators, and as long as people don't create duplicates of described taxa, I don't really see the problem.
As for proliferation, who cares? Globally, data storage is pretty much increasing faster than humans are able to create data and creating a separate cultivar doesn't seem like it would add much to iNat server costs. Same for tag species. What is the real harm of them? As long as they can only be created by some curators, and as long as people don't create duplicates of described taxa, I don't really see the problem.
At its heart this discussion touches upon a larger issue of how to expand inat's usability for everyone while still keeping it as a powerful scientific tool. I have a lot of opinions on this and won't get into them all here, but I think it is absolutely VITAL to keep and expand iNat as a powerful scientific tool, even if it gets slightly in the way of usability to the most casual user. If we lose validity as a scientific tool it would be a massive loss not just to scientists, but to those who use the site hoping their data is useful for science (basically almost all users). And I am not saying we are going in that direction at all but IMHO it's very important to be aware of that issue and proceed carefully as iNat expands at an exponential rate (which is great!). If tag names are scientifically invalid, let's not use them. if they might annoy some users in a mild way, I don't see that as a valid reason not to use them. 99% or more of our users will never even see a tag name.
--
You received this message because you are subscribed to the Google Groups "iNaturalist" group.
To unsubscribe from this group and stop receiving emails from it, send an email to inaturalist...@googlegroups.com.
To post to this group, send email to inatu...@googlegroups.com.
Visit this group at https://groups.google.com/group/inaturalist.
For more options, visit https://groups.google.com/d/optout.
AfriBats
Jun 1, 2016, 10:21:59 AM6/1/16
to iNaturalist
Hi everyone
Great to have this discussion unfolding with a diversity of views and aspects to consider! I'll throw in a few thoughts, some already mentioned above, which hopefully help clarifying in which specific cases tag/dummy names might be needed.
My initial concern stems from seeing a dummy taxon on iNat to which I can't relate because it lacks a (published) description. Any taxonomy is obviously a human construct of arranging and naming the diversity of life, with all known issues and shortcomings. However, the Linnaean taxonomy as used by iNat is meant for communicating the diversity of life. The foundation of this taxonomy is a published description together with type specimens deposited somewhere (yes, there are many cases where the description is vague, or the taxon unidentifiable, or the specimens are lost). Introducing tag names has the potential to create confusion because an undefined name (= lacking a basic description / characters / data) doesn't serve the very purpose of a name, which is communication.
Stephen insists that iNat is all about species (and seems to imply that his tag names would potentially allow identification of every observation to species level). With that statement I disagree as we shouldn't ignore the constraints set by iNat's observation data - most are photo(s) plus the when and where information. Clearly, many small organisms documented that way, especially in species-rich regions, cannot be identified beyond a level coarser than species.
I think we should be very explicit in which cases tag names are needed. Above I outlined 3 possibilities why the moth fly isn't IDed to species level:
1) the species is indeed described and identifiable from photos, but there's currently no one on iNat with the necessary expertise to provide the available name
2) the species is described, but belongs to a group that cannot be IDed from photos alone
Great to have this discussion unfolding with a diversity of views and aspects to consider! I'll throw in a few thoughts, some already mentioned above, which hopefully help clarifying in which specific cases tag/dummy names might be needed.
My initial concern stems from seeing a dummy taxon on iNat to which I can't relate because it lacks a (published) description. Any taxonomy is obviously a human construct of arranging and naming the diversity of life, with all known issues and shortcomings. However, the Linnaean taxonomy as used by iNat is meant for communicating the diversity of life. The foundation of this taxonomy is a published description together with type specimens deposited somewhere (yes, there are many cases where the description is vague, or the taxon unidentifiable, or the specimens are lost). Introducing tag names has the potential to create confusion because an undefined name (= lacking a basic description / characters / data) doesn't serve the very purpose of a name, which is communication.
Stephen insists that iNat is all about species (and seems to imply that his tag names would potentially allow identification of every observation to species level). With that statement I disagree as we shouldn't ignore the constraints set by iNat's observation data - most are photo(s) plus the when and where information. Clearly, many small organisms documented that way, especially in species-rich regions, cannot be identified beyond a level coarser than species.
I think we should be very explicit in which cases tag names are needed. Above I outlined 3 possibilities why the moth fly isn't IDed to species level:
1) the species is indeed described and identifiable from photos, but there's currently no one on iNat with the necessary expertise to provide the available name
2) the species is described, but belongs to a group that cannot be IDed from photos alone
3) it represents an undescribed taxon
I guess we all agree that dummy/tag names should be only included in iNat's taxonomy to accommodate the 3rd case. For such cases, I like Jon Sullivan's suggestion to somehow formalize how and where tag names are created. As we're creating and using names to communicate, I think the crucial point is to have some sort of reference point. If there's a printed monograph or web resource (Donald's examples above) that includes an informal description (and maybe already a 'dummy name') of a new taxon, then fine - a tag name could be added to iNat's taxonomy with a reference to that source. If it's the iNat user who discovers (or knows of) a new species without any sources to refer to, then I think we should come up with a solution to have an informal description of a tag name that can be used by others than the person creating that name.
Cheers, Jakob
Cheers, Jakob
James Bailey
Jun 1, 2016, 4:08:17 PM6/1/16
to iNaturalist
Depending on what field you work with tag names are either useful or useless. If you work with entomology, other invertebrates, or to some extent several areas of marine life, then tag names are very useful indeed.
As for the comment above by jakob I agree that at this time we should focus on tag names used in the methodology of scenario number 3. Whether or not we should then include "sensu stricto/lato" or "aggregate" tags** is a different matter.
**an aggregate is for instance "Oligia strigilis agg". This means that a moth is 1 of 3 inseparable species closely related to Oligia strigilis, instead of any 1 species out 100s of Oligia species as simply genus level would indicate.
James
Stephen Thorpe
Jun 1, 2016, 5:35:53 PM6/1/16
to iNaturalist
Let me try once again to be clear:
There are two steps to identifying a specimen (or an observation):
(1) recognise the species as distinct from other species. If you can't tell it apart from other species, then you can't identify it!
(2) find a name for the species.
Sometimes, we can only get as far as step (1). In these cases, we use a tag name, until such time, if ever, that we find a name for the species.
The above protocol is perfectly in line with how iNat functions at present, and just extends it a little by introduction of tag names. Any comments above relating to "confusion", "mess", etc., are red herrings.
There is no more or less need for a diagnosis to accompany a tag name on iNat as there is for a diagnosis to accompany a published name, and I'm sure that very few identifiers on iNat refer to published diagnoses of what they are identifying. For many common species, the original description is diagnostically useless anyway (e.g. most names from Linnaeus 1758, etc.)
iNat users will differ in their abilities to do both step (1) above and step (2), but they already differ in their abilities to identify observations. Tag names introduce no new problems. They simply function to group observations together into species, observations which would otherwise be lost under a higher taxon (possibly as high as family).
Jakob said "Stephen insists that iNat is all about species (and seems to imply that his tag names would potentially allow identification of every observation to species level)."
I intended no such implication! All I em saying is that for cases of distinctive species for which we have no name, tag names prevent us from having to ignore that species as a recognisable entity. They allow us to track, map, count and talk about that species using the core iNat framework and infrastructure.
There really is nothing in this issue to warrant objections. It is a win-win. Not sure why some people aren't seei
Stephen
There are two steps to identifying a specimen (or an observation):
(1) recognise the species as distinct from other species. If you can't tell it apart from other species, then you can't identify it!
(2) find a name for the species.
Sometimes, we can only get as far as step (1). In these cases, we use a tag name, until such time, if ever, that we find a name for the species.
The above protocol is perfectly in line with how iNat functions at present, and just extends it a little by introduction of tag names. Any comments above relating to "confusion", "mess", etc., are red herrings.
There is no more or less need for a diagnosis to accompany a tag name on iNat as there is for a diagnosis to accompany a published name, and I'm sure that very few identifiers on iNat refer to published diagnoses of what they are identifying. For many common species, the original description is diagnostically useless anyway (e.g. most names from Linnaeus 1758, etc.)
iNat users will differ in their abilities to do both step (1) above and step (2), but they already differ in their abilities to identify observations. Tag names introduce no new problems. They simply function to group observations together into species, observations which would otherwise be lost under a higher taxon (possibly as high as family).
Jakob said "Stephen insists that iNat is all about species (and seems to imply that his tag names would potentially allow identification of every observation to species level)."
I intended no such implication! All I em saying is that for cases of distinctive species for which we have no name, tag names prevent us from having to ignore that species as a recognisable entity. They allow us to track, map, count and talk about that species using the core iNat framework and infrastructure.
There really is nothing in this issue to warrant objections. It is a win-win. Not sure why some people aren't seei
Stephen
Jon Sullivan
Jun 1, 2016, 6:36:52 PM6/1/16
to iNaturalist
Good point, Charile,
Perhaps I spoke too soon. Yes, there are a lot of readily recognised cultivars and breeds. In some situations I can see that it is useful to ID observations to this level (e.g., when they're in interactions with wild species, or have escaped cultivation). If a provisional names tool could also be used to ID observations to cultivar, I don't suppose that would be a bad thing.
I've been getting by with fields for these but that approach has the problem of typos and variations in spelling and formatting for the same names. It would be a headache to try and analyze the iNat data at the cultivar level for this reason. For example, looking now at the Cultivar field, I see that some people are typing in the cultivar name (like I've been doing) and others are just typing "yes". I once set up a field, Phormium cultivar, with a drop down list of the recognised traditional cultivars of Phormium (NZ flax) in New Zealand. I doubt we'd want to maintain lots of fields like this. A cultivar/breed category within a provisional name tool that tacked names onto the iNat names tree would be more elegant. You could even have those hidden by default on the taxon pages.
Cheers,
Jon
Perhaps I spoke too soon. Yes, there are a lot of readily recognised cultivars and breeds. In some situations I can see that it is useful to ID observations to this level (e.g., when they're in interactions with wild species, or have escaped cultivation). If a provisional names tool could also be used to ID observations to cultivar, I don't suppose that would be a bad thing.
I've been getting by with fields for these but that approach has the problem of typos and variations in spelling and formatting for the same names. It would be a headache to try and analyze the iNat data at the cultivar level for this reason. For example, looking now at the Cultivar field, I see that some people are typing in the cultivar name (like I've been doing) and others are just typing "yes". I once set up a field, Phormium cultivar, with a drop down list of the recognised traditional cultivars of Phormium (NZ flax) in New Zealand. I doubt we'd want to maintain lots of fields like this. A cultivar/breed category within a provisional name tool that tacked names onto the iNat names tree would be more elegant. You could even have those hidden by default on the taxon pages.
Cheers,
Jon
Jon Sullivan
Jun 1, 2016, 7:00:09 PM6/1/16
to iNaturalist
Here's a couple of recent examples of observations where a tag name feature would be helpful:
www.inaturalist.org/observations/3354386
This is an observation from yesterday. It's the recognisable and common New Zealand mushroom Coprinellus aff. disseminatus. I looks like the Coprinellus disseminatus found in the Northern Hemisphere but Jerry Cooper tells us that the NZ form has genetic and morphological differences and is undescribed. The observation currently has the ID Coprinellus. Other NZ observations have been mistakenly IDed to Coprinellus disseminatus (to research grade). It would be very helpful to have a Coprinellus aff. disseminatus provisional name under Coprinellus on the names tree that we could use for IDs.
http://www.inaturalist.org/observations/2586248
This is an observation from January of the New Zealand orchid currently known as Spiranthes aff. novae-zelandiae ("Motutangi"), in the process of being described. It's also now IDed to just Spiranthes but in the comments there's agreement that it's this undescribed species.
In both cases, we're trying to capture this with the aff. (tag name) field, but that's not ideal.
Cheers,
Jon
www.inaturalist.org/observations/3354386
This is an observation from yesterday. It's the recognisable and common New Zealand mushroom Coprinellus aff. disseminatus. I looks like the Coprinellus disseminatus found in the Northern Hemisphere but Jerry Cooper tells us that the NZ form has genetic and morphological differences and is undescribed. The observation currently has the ID Coprinellus. Other NZ observations have been mistakenly IDed to Coprinellus disseminatus (to research grade). It would be very helpful to have a Coprinellus aff. disseminatus provisional name under Coprinellus on the names tree that we could use for IDs.
http://www.inaturalist.org/observations/2586248
This is an observation from January of the New Zealand orchid currently known as Spiranthes aff. novae-zelandiae ("Motutangi"), in the process of being described. It's also now IDed to just Spiranthes but in the comments there's agreement that it's this undescribed species.
In both cases, we're trying to capture this with the aff. (tag name) field, but that's not ideal.
Cheers,
Jon
Stephen Thorpe
Jun 1, 2016, 7:25:22 PM6/1/16
to iNaturalist
I would say that the two examples given by Jon well illustrate the confusion and mess that can result from NOT using tag names!
Stephen
Stephen
AfriBats
Jun 3, 2016, 6:56:40 AM6/3/16
to iNaturalist
There are two steps to identifying a specimen (or an observation):
(1) recognise the species as distinct from other species. If you can't tell it apart from other species, then you can't identify it!
(2) find a name for the species.
Sometimes, we can only get as far as step (1). In these cases, we use a tag name, until such time, if ever, that we find a name for the species.
I tried to clarify which potential situations might underlie cases of only reaching step 1. I'm fine with tag names if this applies to organisms that are likely to belong to an undescribed taxon. However, if tag names would encompass everything that a user considers distinct from something else, and without any underlying reference or informal description, I would be opposed to using tag names in such cases.
There is no more or less need for a diagnosis to accompany a tag name on iNat as there is for a diagnosis to accompany a published name, and I'm sure that very few identifiers on iNat refer to published diagnoses of what they are identifying. For many common species, the original description is diagnostically useless anyway (e.g. most names from Linnaeus 1758, etc.)
I'm not only referring to the original description, but the whole body of references and resources associated with a published name: taxonomic revisions, monographs, field guides, websites. When I say "this belongs to Turdus merula", Linnaeus' description might have been vague or even useless, but we all understand what that name means.
Along these lines, I'm proposing that tag names should be restricted to cases where there is good evidence that they represent undescribed taxa. Such names should be accompanied by a link to an external source providing some sort of informal description (Donald's and Jon's examples above). I also like Jon's suggestions how this could be handled in cases where there's no informal description available.
AfriBats
Jun 3, 2016, 7:06:31 AM6/3/16
to iNaturalist
Hi Jon
I definitely see that in such cases, it would be good to bundle observations under an identifiable and provisional name. However, the solution with a dedicated observation field already allows to filter and even map them accordingly, e.g.
http://www.inaturalist.org/observations?taxon_id=56315&field:aff. scientific name (tag name)=aff. disseminatus
But I guess it's cumbersome, and prone to errors for those not familiar with the taxonomic issues of the respective groups.
Cheers, Jakob
Here's a couple of recent examples of observations where a tag name feature would be helpful:
www.inaturalist.org/observations/3354386
I definitely see that in such cases, it would be good to bundle observations under an identifiable and provisional name. However, the solution with a dedicated observation field already allows to filter and even map them accordingly, e.g.
http://www.inaturalist.org/observations?taxon_id=56315&field:aff. scientific name (tag name)=aff. disseminatus
But I guess it's cumbersome, and prone to errors for those not familiar with the taxonomic issues of the respective groups.
Cheers, Jakob
Colin Meurk
Jun 3, 2016, 4:17:50 PM6/3/16
to iNaturalist
i'd be in favour of tag names providing 1) there was some reference somewhere to what it means - that is what are the characters that distinguish the species from its close/similar relatives - not a whole formal description, but the key differences; 2) there was some way of ensuring there wasn't proliferation of unverifiable 'names' - maybe requiring a second opinion from another curator. this doesn't mean that the other person necessarily can judge the merits of the taxon, but rather that they are a safeguard against some robot or malicious user generating vast numbers of spurious names. in most cases it will just be ticking the box. and 3) it should only be available within a curators specialist field. on a related topic, i support finding some way of including 'aff', agg, sensu lato, etc into names (as per Jon) so that one is not forced to endorse something that may be too narrowly or incorrectly defined - so far.
Stephen Thorpe
Jun 3, 2016, 9:21:02 PM6/3/16
to iNaturalist
Jakob said: "I'm fine with tag names if this applies to organisms that are likely to belong to an undescribed taxon"
My reply: Well, many of my tag names are widely agreed by entomologists to represent undescribed taxa (though they could be mistaken). Others are just next to impossible to find a name for. Effectively, these two scenarios amount to much the same thing. I am not putting tag names on species for which a published name could easily be found!
Stephen
My reply: Well, many of my tag names are widely agreed by entomologists to represent undescribed taxa (though they could be mistaken). Others are just next to impossible to find a name for. Effectively, these two scenarios amount to much the same thing. I am not putting tag names on species for which a published name could easily be found!
Stephen
Stephen Thorpe
Jun 4, 2016, 12:49:36 AM6/4/16
to iNaturalist
Colin,
I think it somewhat unreasonable to require specification of distinguishing characters for tag named species, given that many of us identify species more by familiarity with the way that they "look" than anything else. Should I suggest that you specify distinguishing characters (or a published reference to such characters) for every species that you identify. For example, we can all easily recognise <em>Gingko biloba</em>, but it would be quite difficult to put diagnostic characters into words in a useful fashion that would distinguish it from <em>Phyllocladus</em> etc.
Cheers,
Stephen
I think it somewhat unreasonable to require specification of distinguishing characters for tag named species, given that many of us identify species more by familiarity with the way that they "look" than anything else. Should I suggest that you specify distinguishing characters (or a published reference to such characters) for every species that you identify. For example, we can all easily recognise <em>Gingko biloba</em>, but it would be quite difficult to put diagnostic characters into words in a useful fashion that would distinguish it from <em>Phyllocladus</em> etc.
Cheers,
Stephen
Stephen Thorpe
Jun 4, 2016, 10:41:12 PM6/4/16
to iNaturalist
Just to clarify my last comment (rereading it, it doesn't quite sound as I intended it): I do not strongly object to providing some diagnostic information for tag named species that I recognise. I can do that. I'm just questioning whether it is really necessary, since (i) I always try for diagnostic photos; and (ii) the user is no better off with many described species, for which there is little or no available literature to facilitate ID. I am happy to provide what I can, of course, bearing in mind that words have to be interpreted by the reader, which may not be easy!
Cheers, Stephen
Cheers, Stephen
Reply all
Reply to author
Forward
0 new messages