In the animal kingdom, the astonishing power of the number instinct
thereader.mitpress.mit.edu
thereader.mitpress.mit.edu
I watched a video of a duck who, reasons unknown, decided to navigate a waterfall with her six ducklings. Two made it. Does she know that she has fewer babies now? I have seen similar videos where a heron will snap or two or three ducklings while the mother is unable to do anything about it. It occurred to me that zero ducklings might make it per year, and that this would be a normal thing to happen to a mother duck. A continuity of memory and the ability to enumerate loss could result in learned helplessness; would nature end up selecting for creatures that live in the blind optimism of now, with neither past to compare against nor future to dread on either side?
This pretty much involved two of us paddling around the lake in a canoe searching for goose nests, one person would distract the geese, while the other snuck up to the nest and shook the egg until you felt them break inside. The geese would return to the nest, and continue hatching the eggs and be fairly unaware and unperturbed when they didn't hatch.
It had to be done that way though. If the eggs were just removed, the geese become stressed and upset and end up just laying another nest.
If you have cats, and one of them dies, the other may spend months looking for their companion in the usual places, howling away, but for progeny, litters of kittens average about four or so (by nipple math!). If a kitten dies or is somehow lost, the mother will likely miss them on an individual basis, but will the change in the number of kittens in a kitten pile also cause a feeling of loss?
What about with the aforementioned ducks? You might have twelve ducklings and a reduced capacity in the mother to recognize them individually; she might instead look at the ducks collectively. I saw a bit of a nesting duck leading her babies out for their first foray into the water. She would quack and every so often another duckling would pitch themselves out of the nest-hole and onto the water. Eventually the quacking didn't get her any more ducklings and she proceeded onward. She had an astounding nineteen ducklings.
However, that number is going to decrease over time. Will she have a creeping experience of fewer and fewer ducklings? Or is it something like "many, three, two, one"? And what kind of internal state comes about? Certainly we see that loss in the larger primates, and even in our own pets, but when it gets down to birds, many of them do not have a lot of brain matter to work with -- what's their state? Is there enough memory and experience to overcome instinct? "All of my babies died last year and the year before ..."
Overall, and this is just the vaguest of grim suspicions, what if, at each increase in memory and care for offspring, evolution must select against those organisms experiencing a kind of despair, a learned helplessness, at the futility of the ever-larger spheres of their existence? It's not an issue if you're just a handful of neurons wiggling your way toward the next source of food, but it could become large enough to struggle against the survival instinct, and I wonder if numeracy is one of those critical struggles.
Male pseudoscorpions smell the number of competitor males that have copulated with a female and adjust by progressively decreasing sperm allocation as the number of different male olfactory cues increases from zero to three.
How in the world was this determined? It's presented so neatly, and I certainly believe it, but it's fascinating to think of the procedures that must've been taken to verify such a claim. You'd need a scorpion sperm counter, perhaps, and I'm not sure Amazon sells one.
Human sperm tests are done using a similar procedure depending on what you are looking for.
and a spreadsheet: https://web.archive.org/web/20070224203309/http://www.fertil...
There's a high-quality video here: https://www.youtube.com/watch?v=T9_XkaXCXqc&ab_channel=Advan...
I'm just surprised it's theoretically possible to do it yourself. Tests are upwards of $100.
$200 is still a hefty price though.
Here's a promotional video from 2008 from a maker of such devices [1]. Here's a 2015 video from them showing a newer model [2].
I'm not sure how they actually work. Given the age it probably isn't machine learning.
The same company also sells a home test kit, under the name Yo, which might use machine learning [3]. It uses your smart phone camera. There is also a PC/Mac version that looks like it includes its own camera. $69.95 for either kit, $45.95 for refills for the consumables for 2 tests, $89.95 for 4 tests.
[1] https://www.youtube.com/watch?v=krakvOT-7xI
http://calteches.library.caltech.edu/51/2/CargoCult.htm
Maybe the don't "adjust", but it is a fear reaction because of the imagined proximity of other males, which kind of distracts from the act ...
How they measured it? Perhaps a fake female with a reservoir, which would introduce even more variables.
"Discriminating Males and Unpredictable Females: Males Bias Sperm Allocation in Favor of Virgin Females", doi:10.1111/j.1439-0310.2011.01928.x
It's fantastic.
Animals can surely assess and "feel" quantities via the senses and via their own automatic body reactions (hormones etc...). As well as have many pre-programmed quantities in their genome.
But a "number" is purely an abstract concept. I don't think we can ever be sure an animal is actually "counting" things. Instead from what we know about how fear, excitement, etc is regulated via hormones it seems fairly probable that they are simply seeing, hearing, touching and this triggers a feeling that guides them to a certain behavior (through hormone secretion).
Unfortunately he's been to the vet rather too often in his life. It's in the neighbourhood, so if we're out and he thinks I'm taking him there, he starts to think about ways to avoid it, and pulls me along alternative routes. Every walk's a negotiation between me and his determination to avoid the vet.
Anyway, he got attacked on Monday and I had to take him to get stitched up. This made him extra wary of having to go there again, and it resulted in him trying to drag me along a path he has never walked down in his life, because it's a) through a kids' playground and b) a dead end. But he THOUGHT it would take him away from danger.
So yeah, you can say he was responding to hormones if you like, I'm sure that's true. But he was acting on a hunch, or a hope, not on knowledge.
This is a pretty wonderful place; if you ever need PT for your dog and you live on the SF Peninsula, I highly recommend them. So does Brownie.
After the first few visits she had the route memorized. When I turned off 101 and onto the city streets, she got more excited at each turn as we approached their parking lot.
She's made an amazing recovery. Here's how she was doing in March:
https://www.youtube.com/watch?v=wSGAf3oo6Jc
And even better since then: now she runs and chases squirrels in the yard. She's still a little wobbly, but I can see why she was so excited about her visits to Scout's House.
Most treat oriented dogs will happily take an extra treat, and keep looking for them if there's a potentially indeterminate number.
My experience is that for small treats, 1/2 sized treats are just as good as ones that you haven't split. It's not like you're feeding the dog, you're treating it to reward/train. This is often with high value treats like bits of lunch meat that's ripped up to give them. (This is like a couple of grams for a 20kg dog, a snap and a gulp and it's gone.)
If the dog is used to getting 2 treats after an activity, that's what they'll wait for. If they're used to one treat after coming in, that's what they'll wait for.
If you, for some horrific reason, don't have the usual number of treats at the right time, you will be hounded for them. If you split them so that there are the right number but a smaller overall quantity, They're fine with it.
My experience is: * Dogs now how many treats they get for a specific context, including that there might be a long string of them. * Within reason, it's the number of actions rather than the volume. * Be very careful what actions you're actually rewarding.
Dogs are pretty smart. I'd be surprised if the smarter ones aren't on the verge of tool use. (collie pups are scary smart. greyhounds on the other hand.... not so much)
(source, dog fosterer, ~15 in the last couple years. )
Would genuinely love for someone to point out a reputable scientific paper that goes into this and actually draws some conclusive evidence.
Many human cognitive abilities are unconscious, like, in most cases, face recognition, so while I am not typically doing deliberate conscious mental math when distinguishing between my mother's face and my drinking buddy's face, it doesn't make sense to call it a "conditioned reflex" either, it seems to me. My brain is doing some sort of thought that I am not consciously aware of. That doesn't mean thought isn't happening.
I think they might have been thinking of a conditioned response, but then again you can't classically condition an animal to do something it's not already capable of. It's only the associations that are changed
The fact that they're showing different behavior in different circumstances shows some basic idea of quantity, and what is the expected quantity, and at least the difference between 1, 2, n-1, and indefinite quantity.
My dog would take it up the stairs, stand it on end, and cartwheel it down the stairs - dispensing all the treats at once for easy consumption. Kind of blew my mind that she came up with it. Her brother never figured it out.
Why wouldn't they? In evolutionary terms, we followed the same path to get here.
Then why can't I fly?
We know that language had a role in our ability to count. Most primitive societies with limited language only counted 1 or 2 and everything else was "many". It's why many languages have a "half" and a systematic one-thirds, one-fourth, one-fifth, etc. If we had "numbers" and the ability to count, we'd say "one-twos" rather than half. This is true for most cultures/languages.
Also, if you think animals grasp the concept of number, do you think they have the ability to understand negative numbers?
The ontological question if both representations of the concept of a number are truly equal is not really important if we can show equivalent outcomes.
My point is that these 2 activities are clearly different, and that non-humans clearly engage in the last one but I'm not so sure I have ever seen clear evidence they engage in the first one.
* https://www.wellbeingintlstudiesrepository.org/cgi/viewconte...
or bees (OK, only to 5, but still):
* https://www.sciencedirect.com/science/article/pii/S258900421...
In these cases I would say -- even if they don't have the ability to tell us that they can reason with exact numbers -- we can observe experimentally that they do. And I don't see any reason to assume that the abstraction that is represented in the animal mind is fundamentally different from our notion of it.
EDIT: That could also be an indication for mathematical Platonism, when I think about it. But alas, as long as we don't observe a species that develops mathematics from the basic forms of countability that would be a stretch too far... Or to put it differently -- if there is a species that could reasons about numbers beyond the basics of counting, given enough time they would also be able to axiomatize mathematics under an equivalent of ZFC Set theory.
I see a lot of meandering about methodologies in the chimp paper and if you see the Concusions paragraph you can see how weak it is:
>"CONCLUSIONS: Studies of numerical competence in the chimpanzee continue to provide new insights into the range and capacity for quantitatively based information processing in this species. In general, the rebirth of studies of animal counting currently suggests that this area remains a rich and fruitful source of contributions to our understanding of animal cognition and behavior. And for a truly comparative perspective, it will be important for researchers to challenge their creativity, by continuing to devise new methods for tapping capacities toward counting in a variety of species, including nonhuman primates, rats, birds, and additional new species for whom no data currently exists"
The bee paper is more interesting but it also doesn't reach the conclusion that bees can count. At least it offers a context for the framework used and the conclusions reached.
What they did is simulate a neural network that has the bees visual data as input and prove that a simple "inexpensive" computation can in theory scan quantity of perceived objects.
Well isn't that just the same as proving that a neural network algorithm can simulate an computer accumulator register?
> "Within this framework we have shown that counting and numerical ordering are computationally inexpensive, provided the animal employs an active, sequential scanning of pattern elements."
The numerical instincts in this article are one example. Other useful abilities that would make them more general purpose, but missing from most AI systems: fast adaption, basic understanding of complex high-dimensional environments, integrating streams of information, self-maintenance, etc.
I feel like the author uses several examples that are based on relative size, and also examples that appear to require actual counting (ie the cowbirds at the end). It kind of muddies the issue, but overall interesting ideas
poor example: https://www.youtube.com/watch?v=tDQw21ntR64
(Humans are also herd animals, which has repercussions for safety measures in public spaces.)