@NicoleCRust @bwyble

That rule was derived from anatomy too.

@NicoleCRust @bwyble @DrYohanJohn

shameless plug, read my CONB on this topic with Adam K, for my more edited, fleshed out opinion!
https://www.sciencedirect.com/science/article/pii/S0959438822001246

(just insert "microcircuit motif" whereever you read cell-type, Adam K. and I disagree about which phrase is the better one :p)

I get the argument about confusing implementation and computation, but I agree with with @NicoleCRust that the idea of canonical cortical computations has been super influential, especially / at least in vision (which is all of computational neuro anyway, amiright?)

I think the idea of simple, repeated computation is kind of necessary / permissive for a certain types of "grand unified theory" that are very intuitively appealing exactly because they squash together computation and implementation into one little thing that is intuitively understandable in words. The fact that these "theories" are conceptually "small" down to implementation mean that people understand them, they catch on, they drive research.

Now whether this is positive or not, is an open debate. But I think there's no way we can say that ideas like predictive processing, backprop, divisive normalization, and maybe even convolution havent been wildly influential in the field.

To state this in a slightly more aggressive way than I feel: I do think the idea of a canonical microcircuit is very useful, because it's studiable! If we just start off by assuming everything is brain soup it's very easy to just give up and assume we'll never understand implementation. I rather start off with the assumption that there exist architectural motifs that matter, and try to take that as a hypothesis from which to start, than just admit defeat.

So maybe the answer is, it depends on the level of explanation you are lookign for. If you dont care baout multilevel understanding then "how brain regions connect in networks" black box type of understanding is enough. I personally think it's only step 1, and then understanding more details of implementation is step 2. Furthermore, I've come to think that the implementation detaiuls will probably constrain and help us understand the higher level.

@achristensen56

I agree completely! In fact this aligns very nicely with what I said about normative arguments. It would definitely be nice if we could think of each cortical area/column as processing its inputs in a generic manner. But it's an empirical question whether the idealization is true.

And it may well be that a second term in the infinite series (which I am considering in part because of my anatomist colleagues) may add functional flexibility.

@NicoleCRust @bwyble

@achristensen56 @NicoleCRust @bwyble

In other words, there is a whole continuum between "columns are interchangeable" and "each cortical column is a unique snowflake". Systematic architectonic variation has been observed: the challenge now is for a few computational neuroscientists to imagine a few functional/computational stories for why it's there.

This is why I used the term "cortical spectrum": it's not giving up, any more than the EM spectrum requires giving up.

@NicoleCRust @DrYohanJohn @achristensen56 @bwyble

This discussion makes me think that if the cortical columns are elementary circuits for computations that are compounded to form the cortical surface, does this imply that these computations necessarily involve topographical maps?

I find it hard to imagine that the same computations could be performed if the microcircuits were randomly placed on the cortical surface. The total length of the wiring would not allow it, while minimizing the length of the wiring would produce topographic maps (retinotopy, orientation maps, somatosensory maps, ...).

@laurentperrinet @NicoleCRust @DrYohanJohn @achristensen56 I think there are several reasons for using topographic representations. One of them is reducing intracolumnar wiring costs as you say.

But also, topography is a way to innately encode information about the environment. In the physical world, sampled light from nearby locations is more likely to be from the same object/surface/location, and so it makes sense to cluster the processing of light according to its location on the retina. This makes it easier for attentional modulation to select the cortical areas processing a given object/boundary. The same argument can be made for the auditory, somatosensory, motor domains, and this can be extended to higher order cognitive aspects as well, like language, etc.

@bwyble @laurentperrinet @NicoleCRust @achristensen56

Yup. Topography is a great way to group signals in a way that preserves spatial relations (including in abstract spaces).

What sorts of topographic map have been found in higher cognitive areas? Even place cells show no topography, given remapping etc.