7 replies to this topic

[gib65]

Where does thinking occur? Isn't the frontal lobe the seat of cognition? I've heard stories, however, about subjects who were asked to think of particular things while their brains were being imaged, and the parts of their brains that showed more activity were those involved in the sensual/perceptive parts. For example, if one subject was asked to think of a banana, areas in the visual cortex would become active. Or if a subject was ask to think of a song, areas in their auditory cortex would become active. But then what roll does the frontal lobe play? Could it be that the sensory/perceptive areas are active when visualizing tangible objects whereas the front lobe is active when thinking about abstract ideas?

[gib65]

Well, does anybody know what that phenomenon is called, at least? That is, the phenomenon of the sensory cortex becoming active just by visualizing objects?

[ashennell]

Hi,
sorry I've been busy.

The results you refer to in your first post are correct. The motor and perceptual parts of the brain are indeed active during visualisation, thinking, imagining or what ever else you want to call it.

As for abstract thought - the situation is not so clear. I have seen some evidence, from language research, that would suggest that abstract thought is given a spatial representation or some way interpreted in terms that are similar to the constraint placed on our perceptual systems. e.g. graphs are sueful for visualising functions. I'm seem ot recollect that many of the great scientists relied on a good visualisation of the problem in order to solve it.

I'm sure that the PFC is critical in abstract thought but I think that by interpreting abstract ideas in this way would allow other cortical areas to become involved.

[gib65]

e.g. graphs are sueful for visualising functions.

Yes, I've noticed this too. Something interesting that I noticed about this - and I don't know if there's anything to this (maybe you could enlighten me) - is that the most useful graphs are those based on nodes and connecting edges or arrows. For example, the structure of a business is often depicts by departments represented by nodes, and the way these departments interact with each other represented by arrows (a flow chart, essentially). This is uncannily like the networks that neurons form with each other where the nuclei represent the nodes and the axons represent the adges. Could it be that graphs of this sort work so well because they literally provide the brain with a complete and precise "blue print" for how to build a neural network - one that, when built, will make possible the understanding of the concepts represented by the graph?

I'm sure that the PFC is critical in abstract thought but I think that by interpreting abstract ideas in this way would allow other cortical areas to become involved.

Do you think that the PFC "queries" the other, more sensory/perceptual areas of the brain for information? I have this idea of querying that seems to explain this. That is, if the PFC wants to visualize something, it needs information on how to build such a visualization. For example, if it wants to visualize a house, it needs to know what a house looks lke vis-a-vis lines, shapes, colors, textures, etc. Therefore, it "queries" the more sensory or perceptual parts of the brain for information on how to do this. Do you think this sounds plausible?

[ashennell]

the most useful graphs are those based on nodes and connecting edges or arrows. For example, the structure of a business is often depicts by departments represented by nodes, and the way these departments interact with each other represented by arrows (a flow chart, essentially).

I have an opinion about this rather than a definate answer.

I think that the cortex is optimised for dealing with sequential information / temporal structure. It seems our motor system is organised into a repertoire of automatic behaviours - learned sequences of actions. I believe perceptual chunking and cognitive chunking are a result of exploiting predictable sequences of events - dependancies between events. Converting data into a narrative can improve memory consolidation. Anyway, the point, data presented in a format that can be easily converted into sequential patterns is probably easier to comprehend.

Do you think that the PFC "queries" the other, more sensory/perceptual areas of the brain for information? I have this idea of querying that seems to explain this. That is, if the PFC wants to visualize something, it needs information on how to build such a visualization. For example, if it wants to visualize a house, it needs to know what a house looks lke vis-a-vis lines, shapes, colors, textures, etc. Therefore, it "queries" the more sensory or perceptual parts of the brain for information on how to do this. Do you think this sounds plausible?

Honestly - no, not at all. If we visualise a house the visualisation occurs in the visual cortex. The information does not need to be transfered to the PFC for us to perceive it. This seems like a hang=up from the old cartesian theatre
view (although this often occurs without realising it). Once we have one part of the cortex to represent something there is no point shifting it elsewhere to represent it again.

If we were actually looking at a real house, our visual areas would contain house-related info, what would the PFC be doing? Well, in the visualisation of a house we are essentially trying to partially recreate this situation.

Maybe this is not clear.

If you are visualising a house there is probably a good reason, maybe the PFC is processing info about the reason.

[gib65]

So, tell me something. Have brain scans ever turned up observations showing any difference at all between visualizing something and actually seeing it? I mean, is there any difference in the brain activity when someone looks at an apple (say) compared to visualizing that same apple, or is the brain activity under both condition EXACTLY the same?

[ashennell]

I can't say for sure without a quick review.
I'll have a quick look at the data over the weekend and get back to you.

However, there is one problem I can foresee, Scientists are kind of limited in what they can show people in a MRI. Also, I think that a picture of an apple will provide different activation to a real apple. Data from electrode studies is not going to be too useful in answering this - I guess.

There must be some difference between viewing a real scene and imagining it but for the vision to 'mean' the same (or similar) as the real input it needs to recreate some of the same patterns of activity in the same parts of the brain.

[gib65]

There must be some difference between viewing a real scene and imagining it but for the vision to 'mean' the same (or similar) as the real input it needs to recreate some of the same patterns of activity in the same parts of the brain.

Agreed.

If there was absolutely no difference between visualizing and apple and seeing one - in terms of brain activity - this would open a pandora's box of philosophical dilemmas. I suppose if you believed in free-will, this would be the Holy Grail of experimental evidence, but if you were a hardnosed determinist/materialist this would not sit well at all. To think that you could get blatantly different behavior from perfectly identical brain states is unnerving to a determinist/materialist. I mean, you can easily ask someone to raise their left arm (for example) if they were merely visualizing an apple, but raise their right arm if they were actually seeing it - obviously different behavior. But then you bring the MRI into the scenario and observe that, in both cases, there is no difference in brain activity whatsoever, and you become want for a physicalistic explanation for what's really causing their arm - either the right one or the left - to be lifted.