Well, with Christmas over and nothing to do...
Ygggdrasil said:
To try to summarive my current views on the subject:
1) The human brain has evolved mechanisms for spatial reasoning and pattern recognition, but is not born with the physical intuitions dicussed in the "Physics for Infants" article posted by Pythagorean (
#40).
2) The physical intuitions come about by the interaction of the spatial reasoning and pattern recognition mechanisms with a world that obeys the laws of classical mechanics (i.e. via learning). This seems to be supported by some of the evidence posted by atyy suggesting that these intuitions develop over time (
#69).
One reason I favor this model is by analogy to the way our visual processing circuitry develops. While it might make sense for the way our eyes are wired to the brain to be pre-determined by genetics, the wiring actually occurs in response to simulation of the eyes by the environment (as shown by the classic
monocular deprivation experiments done by Hubel and Wiesel). Of course, certain genetic factors influence the process (for example, certain neurotrophic factors define a specific critical period during which the wiring can occur), but this is a nice example that clearly demonstrates how neural circuitry develops in response an individual's experiences.
Of course, this model essentially specifies the inevitable development of physical intuition because no infant will experience a world not governed by classical mechanics. So, in this sense, you can say physical intuitions are pre-determined. Of course, it would be interesting (but highly unethical) to test whether raising an infant in some virtual reality that presents different physical laws could alter the learned physical intuitions of the child or whether the infant still develops innate physical intuitions consistent with the real world. Perhaps such experiments might be possible using virutal reality systems for studying mice (for example
http://www.nature.com/nature/journal/v461/n7266/full/nature08499.html).
I generally agree with you, but I still think the innate side is being (significantly) undersold. First of all, let me just say that I really appreciate your approach to discussion and I value your perspective; I know neuroscience is a huge diverse topic and different people know it at different scales and aspects (from molecular to cells to systems to behavior) and our introductory textbooks are always changing and always wrong somewhere.
I will start with monocular deprivation, since I find it relevant to the context. It helps demonstrate the
false dichotomy of learned vs. innate. Then I will talk about the vestibular system and some of the innate wiring between visual and vestibular systems that lend to physical intuition.
Monocular Deprivation
While some of the wiring is a result of learning, that is really more at the level of engrams than over all wiring structure, and this is a functionally relevant distinction. In particular, it appears that invariant aspects of our environment (such as would be physics) are more innate, while variant aspects (shapes, colors, lighting) would be more learned. The literature:
“ The basic structure of cortical maps is therefore innate, but experience is essential for specific features of these maps, as well as for maintaining the responsiveness and selectivity of cortical neurons.”
The Role of Visual Experience in the Development of Columns in Cat Visual Cortex
http://www.sciencemag.org/content/279/5350/566.abstract?sid=45909110-de27-4bbf-bf1d-ce471a76064e
“We argue that these spontaneous patterns may be better understood as part of an “innate learning” strategy, which learns similarly on activity both before and during visual experience. With an abstraction of spontaneous activity models, we show how the visual system may be able to bootstrap an efficient code for its natural environment prior to external visual experience, and we continue the same refinement strategy upon natural experience.”
Innate Visual Learning through Spontaneous Activity Patterns
http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1000137
“A comparison of the layout of the two maps formed under these conditions showed them to be virtually identical. Considering that the two eyes never had common visual experience, this indicates that correlated visual input is not required for the alignment of orientation preference maps.”
Development of identical orientation maps for two eyes without common visual experience
http://www.nature.com/nature/journal/v379/n6562/abs/379251a0.html
“This suggests that the initial development and layout of orientation preference maps are determined by intrinsic processes that are independent of visual experience.”
Development of orientation preference maps in area 18 of kitten visual cortex.
http://www.ncbi.nlm.nih.gov/pubmed/9283830
“Previous experiments indicate that the shape of maps of preferred orientation in the primary visual cortex does not depend on visual experience. We propose a network model that demonstrates that the orientation and direction selectivity of individual units and the structure of the corresponding angle maps could emerge from local recurrent connections.”
Intracortical origin of visual maps
http://www.nature.com/neuro/journal/v4/n4/full/nn0401_431.html
Further evidence shows rescue and prevention of monocular deprivation, implying that the negative effects of monocular deprivation are a result of expected inputs. The system is expecting inputs that have spatial structures consistent with our physical world and when you block light input, the system begins making correlations on intrinsic noise. The evidence of prevention and rescue:
http://europepmc.org/abstract/MED/9987024
http://www.nature.com/nature/journal/v378/n6553/abs/378189a0.html
http://www.pnas.org/content/103/22/8517.short
Additionally, horses can walk the same day they are born without running into walls, so while being born earlier in our developmental period has in increased effect on outcome from learning for us and cats, it shouldn’t diminish the innate hierarchy of the circuits and how they connect to sensory organs before learning takes place (or while learning takes place in the case of early birthers like humans).
What’s neat about the visual system… is it has an intrinsic
orientation for down, thought to be based on the vestibular system’s graviception.
“it has been suggested that the cortical vestibular network is involved in the perception of our spatial orientation relative to the gravitational vertical (17, 24, 25,27).”
Representation of Visual Gravitational Motion in the Human Vestibular Cortex
http://www.sciencemag.org/content/308/5720/416.abstract?sid=a51c9b39-95e3-48d4-9765-adcb7f2622df.
Vestibular System
Consider how learned and innate are coupled through overall wiring architecture vs. local synaptic connections when considering how vestibular, visual, motor systems, and hippocampus all talk to each other to form fine connections in the first place.
The vestibular system and it’s wiring with respect to the visual system are a conserved trait (across many, if not most, vertebrates). The vestibular system is essentially an acceleration detector. It can detect:
1) Angular acceleration (through the semicircular canals)
2) Linear acceleration (through the utricle, for horizontal movement and the saccule for vertical movement)
3) Gravity (also otolithic organ)
So we’ve basically evolved a little Newtonian experimental lab inside of brains that can confer to us the laws of motion. Gravisensors have particularly interesting implications in our innate sense of gravity and is believed to be coupled to our visual system (as described above). Another well-known innate coupling between visual and vestibular systems is the vestibular-occular reflex which compensates head movements with eye movements.
Interestingly:
"According to Einstein's equivalence principle, linear accelerations experienced during translational motion are physically indistinguishable from changes in orientation relative to gravity experienced during tilting movements. Nevertheless, despite these ambiguous sensory cues provided by the primary otolith afferents, perceptual and motor responses discriminate between gravity and translational acceleration. "
http://www.ncbi.nlm.nih.gov/pubmed/11710454
a note
Many of the fundamental assumptions of classical physics are "wrong" (or only true "in the limit" if you like). Not surprisingly, these limits are at the scales of our senses whereas the conflicting modern concepts are outside of them. Properties of the classical realm are much closer to our intuition than reality. These include continuity, locality, solidity, etc. Our intuition for the physical world underlies classical physics. We impose our perception and intuition for the world on the discipline of physics. And this is exactly the reason why classical physics is "wrong": because our sensory systems measure things on a limited scale of reality: the classical scale. And our brain has already been wired in an appropriate hierarchy to interpret those signals successfully by the time we start learning the finer details of our environment.
