the binding problem

[gib65]

I'm putting this here in General discussion because I don't know where else it would go.

 

It's a question about the "binding problem", the problem in psychology and neuroscience that questions how it is that the brain "binds" all the input it gets into one unified perception of reality. No one has discovered a central node - or "CPU" if you will - in the brain where all incoming signals converge, and so it's puzzling how the brain knows that all the disparate information it receives is coming from the same thing: reality.

 

Here's my question: why couldn't the binding problem be resolved by assuming that where ever these separate signals get processed in the brain - even if they are sporadic - they can intercommunicate with each other?

 

Here's some diagrams to illustrate what I mean:

 

 

This figure shows the conventional assumption that all signals need to converge in a central point in order to be integrated.

 

 

But this diagram shows what I'm saying - all signals get processed by their own specialized centers, but then enter a general network of centers (where cognitive processing occurs) that intercommunicate, and therefore can inform one another of each all the information coming in from those channels.

[scicop]

Um..i really don't get this post. Its been well established that there is intercommunication between the sensory systems as well as those that control other processes including thought and autonomic systems.

 

In indeed there are special regions that process specific inputs (such as cortical regions) as well as regions that integrate (thalmus, hippocampus, amydala, nucleus accumbens etc). All these question are addressed within the realm of neuroscience, and people such as Joe LeDuex, Rodolfo Llinas, John Rinzel, Bruce McEwen, have made and are still making sense of these circuits. The reality is there are lot of intergrative steps, that is not necessarily a strict feed forward way of intergrating information, there is also back-tracking and error correction.

 

There is no one glorious point of intergration. Even within nuclei there is further structural and functional separation as well as diverse neuronal (and glia) populations with differering connectivities! Just as an example, a simple one, there is the Ventral Tegmental Area (VTA)(midbrain, and mediator of reward associated behaviors), it has a nigral (more lateral and close to the SN) and brachial (close to midline) of dopamine containing neurons. One population projects mainly to the pre-frontal cortex where as the other to the shell and core of the Nu.Accumbens. These neurons also differ in their inputs. Furthermore there are GABAergi neurons in the VTA as well! Some project outside and some make collaterals within the structure. Not to mention the connectivity of the other nuclie that recieve VTA DA contacts and the feedback from VTA targets back to the VTA!

 

The point is that there is alot of complexity with neuronal circuitry, neuronal populations, and nuclei connectivity, and I just highlighted one nuclei. To think there is one spot for your binding issue is incorrect and flawed by your diagrams.

[gib65]

Well, okay! That's why I was confused. I knew the brain had many interconnections, and so the latter diagram was more accurate, but I thought the binding problem had to do with the expectation that all signals converged somewhere in the brain.

 

If the latter diagram is more accurate, why is the binding problem still considered a real problem? Or is it?

[Rocket Man]

the brain has many neurons, some of which are devoted to very specific things, a study done on epileptics to remove afflicting neurons shows that in two patients, the surgeons found specific neurons which only lit up when the subjects were shown images of a particular persons face. the binding problem is only a problem when you try to relate a brain to a computer. the brain is stroage and processing all in one. a stimulus will fire specific neurons to work out what it is, those neurons will go about looking at the relational netwoks to find the most appropriate course of action. theres some info on a thread called "reversed visions"

[scicop]

I dont think "binding problem" is really a problem, given that I've never heard of it. At least from my days as a grad student and then post-doc at well respected institutions with the neurosciences, I understand that the overall aim of cognitive neuroscience research is not to implicate one certain area of the brain for a certain function (as it does have its implications).

 

Rather the aim is to understand how the connectivity/neuronal/glia populations contribute to global functions. Understanding how specific areas may have "most" control over a certain behavior can have (and HAVE had) pharmacological implications, should a region/target specific delivery/pharmacologic agent be developed. i.e. Drugs for psychosis target receptors (D2, 5HT) that are expressed in certain regions/neuronal subtypes in prefrontal cortex and stratial regions, where as drugs for alcohol dependence (mu opioid antagonist) have implicated function in the VTA and/or Nu/Accumbens.

 

As you can see the field of neuroscience is big (the biggest conference of the year in the US) that involvles alot of different diciplines. Even cognitive neurosci dudes are going molecular and genetic, as well as visa-versa.

 

So, simple 2D diagrams are a thing of the past (ancient history), its more 4D today.

[gib65]

You brought up the pre-frontal cortex a couple times, saying that a lot of neural groups project to this area. What roll does the pre-frontal cortex play in this regard? Does it integrate a lot of variegated information?

[scicop]

PFC, yeah, it mediates in a few behaviors such as motivation (behaviors that lead to uncontrolled consumption of drugs..i.e. addiction), to some extent agression, and so on. However so does the BLA, VTA, NuAc ext. Although it can partake in an integrative role, it can not do so without other nuclei. Can't just look at what inputs it has, you have to look at its out-puts as well as its feed-back, as well as self regulatory systems (i.e. DSI, DSE) It may have lets say dopaminerigc input from the VTA, but its also has glutamatergic output back to the VTA. all regulated by DSI/DSE mechanisms (to name one) within their one nuclie (at synaptic points) Not to mention connectivity to other nuclei, I think thalmus is in there some where but its been a while since I've done the neuroanatomy thing...neuroanatomy is something I don't really give a crap about, I'm more molecular/cellular. Again, can't make a claim there is just one integrative spot. Its basically the brain is one big giant "network"or integrative node, if we must it in computer terms. Gee..i guess each one of us is a node..hmmm....

 

As I said, can't represent brain function with simple 2D node maps. There is alot of regulation, both inter-nuclei, inter-cellular (local and non-local) as well as INTRA-cellular controls. So to your binding problem..I guess you'd have to incorporate the ENTIRE brain as you point of "binding". If such a term exists, I have not heard of the concept. Maybe theoretical mumbo jumbo people use that term.

 

Just an short edit: can't look at the PFC as just the PFC,within the PFC thereis a medial PFC, and a lateral PFC, defined by the efferent output targets and afferent inputs origins..purely neuroantomical/connectivity definition. Same goes for NuAcc..there is the Shell and Core, with the Caudate Nucleaus there is the patch and matrix....so again..alot of subdivisions based on connectivity alone. Lets stay away from neuronal populations...dopaminergic here, GABAergic there, medium spiny there, mossy fibers here....it gets crazy! Don't even make me mention Glia!!!!! geez..everyone is barking up that tree now..including big pharma.

[gib65]

Thanks Scicop,

 

BTW, here's a link on the "binding problem":

http://en.wikipedia.org/wiki/Binding_problem

[scicop]

Thanks for the link, yeah..basically what i'm hinting at, its hard to implicate one specific part for "the binding problem" rather it is a combination of all the regions for one big region..the entire brain! . Basically it boils down to understanding the parts, summing them together to get a picture (again the entire brain) And that's what we scientist do (well former for me..at the moment). A 4D picture..but a picture nontheless.

 

First time i've heard of this binding theory, I don't think alot of neurosci people focus on it, I think its more of a philosophical issue at least at the "binding" level. In due time I think a picture will emerge, as their are some smaller ones existing already (i.e. hippocampal/thalamic/cortical circuitries) and so one. I think a problem is that its hard to get funding for these type of studies. Most studies have a pathology oriented reasons for their pursuit rather than just basic understanding, but can't condemn it, we've learned alot!