Are drawings evidence?

General
81

The discussion in this thread about skill level in drawing reminds me of why I’m practicing my drawing. The basic idea of the book I’m using is that you don’t learn to draw to be an artist, you learn drawing to communicate. Similar to learning writing or math, the goal is not to be an author or mathematician.

I’m using Drawing Textbook by Bruce McIntyre. He’s not a fan of all drawing lessons being relegated to art classes. He sees drawing as a basic skill to function in society, like writing (by hand or typed), reading, or math. Sure many artists can draw, but drawing is useful for so much more. Just a couple hours of practice and I can see progress. It’s also kind of fun.

[Edited with book info.]

9 Likes
82

I think drawing is half control over the tool and half seeing what you want to use the tool for. Most people can control a pen well enough to write letters; the next step is noticing which details you want to include. How many joints on a limb and which direction do they bend? How many spots and where on the body? How long is this part relative to that part?
For our normal everyday thinking we conceive of eagles or ladybugs with abstract names/symbols, and not until we try to put them on paper do we realize that those conceptions aren’t fully-fleshed-out representations of those organisms.

I also wanted to point out that there are same factors that can visibly distinguish a “drawing” from a “field sketch”. A classic field sketch often includes multiple drawings of the organism from different angles (they don’t need to be good drawings), including within its environmental context, and includes notes capturing behaviour and colouration details that couldn’t be illustrated.
There are a fair number of observations in the field sketch project that are just isolated portraits of an organism, and to be honest for many organisms it’s rare to get that kind of view of those species (e.g. a clear view of the entire body of a small bird for an extended period). Especially when no notes or context is provided, I think it’s understandable for identifiers to be suspicious about those kinds of observations. But the response should be gentle advice and request for more information, rather than anything like identifying as Human or marking no evidence of organism.

7 Likes
83

I think it is important to distinguish between intentional fakes and unintentional inaccuracies. I agree that very few people are going to knowingly draw something they didn’t see just to fool us. But a lot of people might, for example, look a a guide book, decide what they think it was, and then in doing the drawing be influenced by the field marks mentioned in the guide book. Human memory is intensely influenceable.
So if we were designing a new system, I would say that drawings from memory are not evidence. However, that is not the situation we are in. As you and several others have said, drawings have long been considered evidence, and changing the data standards on an established system is almost always a bad idea. So drawings, from memory or not, are evidence.

3 Likes
84

Maybe yours do, but my field sketches rarely showed more than one view and never showed the environmental context.

2 Likes
85

I guess it makes sense that the style for birds and plants would be different.

1 Like
86

I loved the discussion here. So much food for thought as a naturalist with a heavy field journal practice.

I started with iNat in June 2018 to document my observations of wildlife for later study. I was interested in the capabilities of the computer vision and contributing to that, but I also valued the fact that my boots on the ground have the potential to contribute to broader scientific study (that it was research grade), and actually help protect the wild populations and habitats we all love and depend on.

But is the iNat system value of Research Grade used for advancing the computer vision, or is it really a useful (or limiting) filter for scientific research?

Then, enter illustration and field sketching to that can of annelids…

I started field journaling because my photography simply couldn’t capture all the observations I made. Some were at night or twilight, some too far or too deep in water or foliage, some too large and complex while some are too small and obscure for my entry level phone camera.

Pen, paper, a hand lens and binoculars turn out to be excellent tools to record all those other observations of bats, birds, fish, fungi, insects and plants, in an arguably less-invasive and more accessible way than the expensive gear required to advance in the study natural history via photography alone.

The consensus here is that illustrations or sketches are evidence of an organism, when interpreted in good faith. But I think the contention and the debate in this topic is that
the question of Drawings as Evidence begs the further question: Evidence of what?

Can, and should, illustrations be used as evidence for advancing iNaturalist’s computer vision?

And

Can, and should, illustrations be used as evidence of presence of organism in a specific time and place, as scientists are concerned with the observations collected by iNaturalist?

Are two different questions.

And that’s where my question above comes in, How do “research grade” and “presence of organism,” in regard to how we tag and file illustrations and field sketches, come in to play?

Ideally (from my perspective), illustrations should be reviewed as honorably as other evidence types, to the scientific value level they are able to contribute given community review, and stored separately than photos, like sound recordings, in the computer vision (until a reliable body of all ‘illustration and sketch’ observations are built up for a reliable addition to the CV camera feature).

Until iNaturalist has that solid stock of field drawings to contribute to the computer vision, I suggest we approach this subject for a best-case-scenario for scientists (where field journaling entries are filtered out of the computer vision model, but not filtered off the map as casual grade).

So, as I continue to upload my field observations from the past few years, (many of which are journal entries alone) I’d like your advice on how to tag and process my uploads.

What will be the best way to have my drawn observations available to scientists looking for quality information on taxa presence and behavior, while also not confusing the growing computer vision?

Thank you all. Looking forward to more conversations here! Please check out my profile for examples of my field journal observations, and let’s discuss their merit as examples in light of this conversation! https://www.inaturalist.org/people/staysustainable

2 Likes
87

The premise behind this thread just seems like more gatekeeping to me. I posted this on another platform a few years ago, and with some adjustment of wording, it can be seen to deal with the same issue:

You know I’m not much of a luddite, but come on – back in the day, scientists could generate graphs and plots with a ruler and a pen. Now that everything is expected to be computer generated, I get so bogged down in the nuts and bolts of coding the plot, I can’t actually get to the data analysis part. What do my coding skills have to do with my data analysis skills?

2 Likes
88

Interesting discussion. I’m still somewhat skeptical of drawings as evidence for iNat purposes but I appreciate the various arguments for accepting them.

The days of birders doing sketches of rare bird sightings for evaluation by committees are likely largely gone. But it was acceptable in the pre-digital days when bird photography was less common. Nowadays almost any rare bird sighting calls for photo evidence. Today there are a lot of birders with telephoto cameras, and they need not be very experienced birders to document a rarity. Note that in the sketch days, the experience of the observer was likely a factor in accepting a sketch record or not.

As someone who used to evaluate the validity of rare mammal sightings in my state based on descriptions and rarely sketches (no photos) of what was seen, I’m skeptical of human memory and the ability to judge size and note important characteristics. When the observer says they checked the internet to see what a jaguar looks like before they decided that’s what they saw I’m even more skeptical.

4 Likes
89

Here’s a drawing I submitted of a red-headed woodpecker that decided to do a loop-de-loop in front of the car as we were pulling into a campground:

large.jpeg1024×1024 62.6 KB

It shows only the details that I saw and could clearly remember: The solid, bright red head, the black wings, and the white stripe across the back.

My drawings either get accused of having not enough detail, or having too much detail, which apparently means I “cheated” and looked at a guide book. It’s just unecessary gatekeeping and endless shifting of goalposts.

Here’s another drawing I made of a different woodpecker. I drew the illustration right there at the park on my phone to add a time and place marker, since I could only get photos on my digital camera, not my phone.

I don’t have a car, so any time I see things I know how to identify in a car (where the driver Does Not Care and will not stop even for the coolest thing ever on an empty road in the middle of the day), I have two choices: Draw it (white ibises), or take a motion-blurred photo (groundsel tree) from inside the moving car.

Both options have gotten me accused of “destroying the sanctity of iNat” and “polluting the data with useless observations”. It’s an extremely immature and gatekeeping response. I would love if I had a car and had the ability to pull over wherever I wanted to get detailed photos of things, but I don’t. Not even for something as exciting as my first sighting of a fox squirrel, on what’s supposed to be a fun camping trip.

I post drawings for observations I care about that I couldn’t get a photo for. Sometimes it’s because I’m taking a bag of trash out at 3AM and didn’t think to bring my phone wiht me for the 20ft walk to the curb. Sometimes it’s just because the car is moving way too fast for my camera to turn on before the plant is out of sight behind us.

Sometimes, it’s something you’d expect to see on an anti-littering commercial.

Or the first time I ever saw a tufted titmouse after moving halfway across the country to an entirely new ecosystem.

If I am uploading a drawing, it’s because it’s an observation I care about. I don’t draw every single thing I see but couldn’t get a picture of, despite what some people have accused me of.

My drawings might seem “purposefully stylized” but that’s just a side effect of having only a free drawing app on my phone handy most of the time, and trying to prioritize making sure the identifable details / the details I saw are clear, even if it doesn’t end up as the most realistic drawing ever.. I’ve tried making the drawings more detailed, but then people just assume that I’m “cheating” and not actually drawing from memory.

If I had the ability to carry around a sketchbook and pencil every time I go out, to draw a detailed traditional sketch, I would. But that’s just not possible. And again, would end up with the goalposts shifting once again, to arguing that if the drawing is too detailed, then it must be faked.

Arguing that drawings are likely to be faked for some reason is nonsensical, because it’s just as easy, or easier, depending on what’s being done, to fake a photograph in some way. I once found an observation that someone had posted of a prop facehugger from Alien (1979), (no, I unfortunately don’t have a link to it, I don’t think I favorited it, lol. If you find it, let me know), but that doesn’t mean all photos are faked. If someone wanted to fake an observation, it’d be way faster and more sucessful to just steal a random image from google, or use an image generator to fake one, than it would be to take the time to make a drawing and deal with the almost inevitable backlash.

TLDR: Yes, drawings are evidence, and people need to chill out when it comes to them, and how high or low a quality they think they are. As was said above, the CV is here to help US, we are not here to feed it. And there will never be more drawings than there are photos or sound recordings. Taking a photo is a lot easier than making a drawing.

I’ve had to start upvoting all the DQA buttons for my observations of drawings because someone will inevitably have not read the observation guidelines and will incorrectly vote it down. And even doing that is enough to get some people mad at you to the point of blocking.

iNaturalist is meant to help people to connect with nature, and connect with others. Insisting that every single observation must meet the most rigorous standards of scientific photography defeats the purpose.

My new-to-me phone takes photos of inordinate size. Each one is 1GB all by itself and there’s nothing I can do to stop this. This makes transfering them to my laptop for easier sorting all but impossible, when I have a backlog of 3k photos and videos from before, during, and after a camping trip to Florida. Receiving nothing but negativity when I upload a few drawings does not help me feel enthused at all of the gargantuan task of sorting through these 3,000 photos, even now that I’ve finally gotten started.

I just wish people could just look the other way if they don’t want to identify a drawing instead of leaving negative comments. It does not help with seasonal depression nor the task of sorting through a huge backlog of photos. If I could take a photo of every awesome animal I see, I would, believe me. But just because I can’t isn’t going to stop me from making an observation if it’s something I care about, even if it seems minor or unimportant to others.

5 Likes
92

At least one of your offensive commenters has since been suspended.

I can see why you use All the DQAs.
But - recent evidence is intended against fossils.
And single subject against obs with multiple photos, each of a different sp.

93

A few notes, which I hope you will take kindly for they are sent that way:

  • Try to remember that just as all people do not know you/about your seasonal depression, you do not know all people/their level of understanding of what is/is not accepted on iNaturalist. It is tempting to think everyone who marks a drawing as Human is doing so against you, but in reality, we are all here learning, not just about nature but about the intricacies of this platform. Some may just not yet have learned that drawings are acceptable on iNaturalist because in reality, they are such a small percentage of Observations.
  • It is tempting to add funny things like “Not today Satan” as you did in the late night fox Observation but it may make a drawing harder to take seriously for some, there being a wonderful breadth of neurodiversity here.
  • I am not a biologist (ask anyone here!) so I do not know if you already do this, but since your drawings are of a more simple type, I would be hesitant to upload at species unless I was sure I saw every needed definitive marker of species and reflected that in my drawing. (My own Observations are photos and multi-photos when I see something I think is exciting, but because I am not a biologist, I often do not have all the features so things by necessity are left at genus or higher by the Identifers. (That’s OK! I still saw it.)
  • Since your Observations are meant for you, as a record of the awesome animals and plants you see, let them be that. This is to say maybe consider letting go if they become Casual through DQA. I say this not because it is right for people to downvote them, rather because it is exhausting to try to manage other people and if your goal is to make a personal record of awesome species, it may also be an unnecessary battle to fight, especially in the face of depression, seasonal or otherwise.
  • Finally, re: your new phone. Adjust the photo settings. You are probably saving photos at the factory default, which is the “best” quality and yes, enormous. You can adjust it to where you need not fiddle at all. (Only took me like a month to find it when I got a new phone recently.)

Be well. Happy iNatting!

9 Likes
94

I was recently messaging with a different user who’s gotten a lot of negative feedback on their sketch observations recently. Here’s what I suggested:

These extra details (including environmental context in the drawing and additional context in the notes) aren’t necessary but they increase trust from other people evaluating them. They’re what in psychology would be called “costly signals”; they’re relatively easy to produce honestly and more difficult to produce deceptively.

Like anyone can draw a generic profile of an animal from a field guide. It takes a bunch of extra creative effort to come up with pretend environmental and behavioural information. If you already have the additional context in front of you, it’s not that much extra work to include it in your observation.

@nonbinary-naturalist many of your sketch observations do include those details, but some of the more controversial ones don’t and that’s likely why. You don’t have to, but not doing so comes with the tradeoff that people will intuitively be more suspicious. Including an explanation of why you couldn’t get a photo will likely help as well (partly because that’s a pretty relatable experience). There’s often frustration on both sides of the observer/identifier dynamic - identifiers are constantly seeing low effort observations from students and new app users stealing photos, uploading photos from books, etc. Yeah they should check if the observer is an iNat regular before saying anything, but they’re not always going to remember to.

3 Likes
95

(I think you meant to reply to @nonbinary-naturalist , not me, correct?)

1 Like
96

someone has to be the first one to explain to these people what the rules are, though. Which means it is my job when it happens on my observations, or an observation I see.

1 Like
97

yeah, unfortunately, as I said above, that just brings more people accusing me of “cheating” and drawing it long afterward. The goalposts keep shifting. If I draw the whole scene, it’s cheating. If I draw it too detailed, it’s cheating. If I draw it too simply, it’s too stylized. No matter what I do, someone is going to get mad about it. Which is why the site itself needs to make it clearer that illustrations are allowed, and what the appropriate way to interact with them is. (IE: Not automatically and incorrectly downvoting in the DQA)

1 Like
98

Yes. Correct. But that someone does not have to be you. Self-care is important, so if you (or anyone) are struggling with depression or even just having a tired day, it is fine to use the flagging system and let a Moderator handle it, that is all.

Pick your battles but self-care first. :)

6 Likes
99

That illustrations are allowed is written out clearly, and there is a flagging system in place for when individuals knowingly or unknowingly (I find it less exhausting to assume good intent but sometimes yes, it is obviously knowingly) flout the written guidelines.

100

Ok, so I recently had this situation (exact details are either obscured or changed so I can illustrate the situation without calling anyone out).

Someone posed a drawing of a passerine with a red crown and IDed it to Passeriformes. The observation was in a southern-ish state and the observed date was sometime this winter.

I added a comment, “compare with Ruby-crowned Kinglet”. The observer said something along the lines of “maybe, but I remember it being more brown overall.” Someone else then added an ID of Redpoll, and the observer confirmed.

While a Redpoll is a possible winter species in the observed state, it is very unlikely outside an irruption year (which this year is not). No confirmed sightings have been recorded in the state on iNat or eBird this winter (and relatively few in the U.S., in general).

I bumped the observation back to Passeriformes and selected “no” for “can the Community Taxon still be confirmed or improved?” with a note explaining the above. The observer responded with a comment along the lines of “please don’t make the observation casual just because you dislike the fact that it is a drawing.”

I want to emphasize, I did not select “no” for “evidence of organism”. But if an observation had a blurry photo and all I could tell was it was a small passerine with a red crown, I would similarly bump back the ID to the lowest confirmable taxon. So, while I do think drawings can count as evidence, that doesn’t mean they aren’t subject to at least the same level of scrutiny as photo observations, and in my opinion, they should be subject to more scrutiny. Also, just because someone doesn’t think a particular drawing is evidence, doesn’t mean they think no drawings are evidence.

11 Likes
101

I’ve had to post the relevant text from the guidelines as a note or comment to avoid having my drawings downvoted as “no evidence.” And these are drawings that I made from a specimen in hand.

6 Likes
102

I’ve been watching this discussion fairly avidly. I wouldn’t post any of my old nature drawings—while they’re usually dated, time stamps and field notes are missing, and I’d be kind of self-conscious even if I hadn’t seen what the various illustrators have been going through. (By the way, @nonbinary-naturalist, I love your Turkey Vulture!) I do have a recent one (relatively speaking; from this past September) that I’ve been waffling about posting: a chance sighting of an Abert’s Squirrel being chased across our yard by an Eastern Fox Squirrel.

Abert’s Squirrels are not really supposed to be in this part of Denver; they’re a Foothills-and-higher species, and we’re much closer to the Eastern Plains, with Ponderosa Pines kind of thin on the ground. That being said, there is one RG observation from our part of town. I know the species on sight, and that’s why I was kind of startled. I didn’t even have a cell phone available to grab a shot.

This seems, to me, to be the kind of observation that would be incredibly valuable to anyone studying range expansion, or even looking for vagrant species. How often do they stray? Is there a particular part of Denver Metro outside of their ususal range that they find hospitable? Are they hitchhiking with tourists? Nobody would even think to ask the questions if the data wasn’t there to prompt them. Is it the best data? Well, no; but a photograph wouldn’t have been either; trying to pull a focus with the speed at which the squirrels were moving would have been difficult at best.

I haven’t made up my mind yea or nay, although I’m leaning more towards ‘maybe I should post it’. And in the meantime, I’ll keep reading the development of the conversation. :slightly_smiling_face:

6 Likes