4 ms·

Is Geometry a Language That Only Humans Know?

8 points by gone35 5y ago

ggm 5y ago
I'm not convinced by this test. I am pretty sure higher apes can grok shape sorting tests with complex 3D geometry to win food prizes, They have both the dexterity and motivation. the shape sorter bucket square peg in a round hole problem? They get it. The test as I understand it posed 2D representative choices, and the abstraction may not work well. There are stories that some initial west/tribe interactions fell down on 2D drawings of 3D things, they were "read" incorrectly. Performing the uplift "thats a perspective line drawing" is arguably culturally informed at times.(this may be urban myth. I am still looking for the references to this) Many spatial geometry problems in predator-prey involve speed estimation and interception to target. From the start point, for the two vectors of predator and prey, thats a triangle, under continuous re-calculation: their present positions, and the intercept point. I don't see how the animal can be held NOT to understand this. Parellax in vision, Auditory response to binaurial signals (echo location) all consider triangles. Look at the positions animals take to each other in times of stress, distribution to minimise risk. The choices of arrangement will form the apex of a minimum surface be it square or triangle or polygon, depending on how many of each type of risk present. Maybe I'm over-thinking it, but I think this is a classic solipsist "know" problem: the authors don't "know" what the ontology of knowledge really is, with respect to animals other than humans.
- raxxorrax 5y ago
  Given the computing power needed for 3-dimensional perception alone, it is astonishing how bad we are at basic geometry in the first place. But if the complexity of geometric and non-geometric forms can be increased and there is a difference in extra effort for animals and humans, it should be testable.
dr_dshiv 5y ago
Makes me think about stroboscopic geometry or the geometric patterns you see when you rub your eyes. Where are those shapes coming from? I think some neural wetware can be instinctually tuned to resonate with particular shapes—like snakes or faces. But I would guess that large collections tend to resonate with certain geometries naturally. Like, to the perfect symmetry of circles or to crisp contrasts or straight lines or repeating angles. In other words, I think some geometry is inherent to neurons once there aren’t other instinctual matters to deal with. This should be empirically testable.
- ksaj 5y ago
  I think it was Anton LaVey, but I'm not sure, who thought that the parallelogram was a most evil simple shape, and pointed to how it makes people feel uneasy. I figured if this is true, it might be because it's the shape you see if something square is about to fall over on you.

8 points by gone35 5y ago

ggm 5y ago

I'm not convinced by this test. I am pretty sure higher apes can grok shape sorting tests with complex 3D geometry to win food prizes, They have both the dexterity and motivation. the shape sorter bucket square peg in a round hole problem? They get it. The test as I understand it posed 2D representative choices, and the abstraction may not work well. There are stories that some initial west/tribe interactions fell down on 2D drawings of 3D things, they were "read" incorrectly. Performing the uplift "thats a perspective line drawing" is arguably culturally informed at times.(this may be urban myth. I am still looking for the references to this) Many spatial geometry problems in predator-prey involve speed estimation and interception to target. From the start point, for the two vectors of predator and prey, thats a triangle, under continuous re-calculation: their present positions, and the intercept point. I don't see how the animal can be held NOT to understand this. Parellax in vision, Auditory response to binaurial signals (echo location) all consider triangles. Look at the positions animals take to each other in times of stress, distribution to minimise risk. The choices of arrangement will form the apex of a minimum surface be it square or triangle or polygon, depending on how many of each type of risk present. Maybe I'm over-thinking it, but I think this is a classic solipsist "know" problem: the authors don't "know" what the ontology of knowledge really is, with respect to animals other than humans.

raxxorrax 5y ago
Given the computing power needed for 3-dimensional perception alone, it is astonishing how bad we are at basic geometry in the first place. But if the complexity of geometric and non-geometric forms can be increased and there is a difference in extra effort for animals and humans, it should be testable.

dr_dshiv 5y ago

Makes me think about stroboscopic geometry or the geometric patterns you see when you rub your eyes. Where are those shapes coming from? I think some neural wetware can be instinctually tuned to resonate with particular shapes—like snakes or faces. But I would guess that large collections tend to resonate with certain geometries naturally. Like, to the perfect symmetry of circles or to crisp contrasts or straight lines or repeating angles. In other words, I think some geometry is inherent to neurons once there aren’t other instinctual matters to deal with. This should be empirically testable.

ksaj 5y ago
I think it was Anton LaVey, but I'm not sure, who thought that the parallelogram was a most evil simple shape, and pointed to how it makes people feel uneasy. I figured if this is true, it might be because it's the shape you see if something square is about to fall over on you.