4 ms·

I forgot where I read the idea but some neuroscientists proposed that consciousness in the brain is more a measure of interconnectivity of activated brain regio

I forgot where I read the idea but some neuroscientists proposed that consciousness in the brain is more a measure of interconnectivity of activated brain regions. Basically, whatever brain region communicates with each other allows for mixed matching, whatever that brain region does (synesthetic people are very concrete examples of this). What do you think? Could this be true? I find it interesting to consider but have a hard time understanding how the brainstem fits into this. -- psych. & computer science student with a strong interest in cognitive science & neuroscience

Balgair 11y ago
Continue your passion! Neuro is so cool and we need comp-sci people, at the least to process the data and help out with connecting VGA cables (I shudder at bio people's hard-coding). Interconnectivity in the brain is very stereotyped. Certain areas typically only talk to certain others. The 'connectome' in the brain is being mapped these days and there is a lot to learn, but I think we can say that the connectome is stereotypical. That's good, we can make a road map then! Consciousness is a tougher nut to crack, though. Many questions arise from disease studies like this one: https://www.newscientist.com/article/dn12301-man-with-tiny-brain-shocks-doctors/ https://www.newscientist.com/article/dn12301-man-with-tiny-b... . Obviously from that patient, he lives a normal life (within the IQ standard deviations) and no-one else saw deficits in him worth concern. Other patients report only balance issues and delayed speech with having half of all the body's neurons not develop: https://www.newscientist.com/article/mg22329861-900-woman-of-24-found-to-have-no-cerebellum-in-her-brain/ https://www.newscientist.com/article/mg22329861-900-woman-of.... Still other people have extreme deficits in short term memory from having comparatively small portions removed : https://en.wikipedia.org/wiki/Henry_Molaison https://en.wikipedia.org/wiki/Henry_Molaison . What I am saying is that the brain is a confounding mystery and we need the help of people in comp-sci. If you have the time, try to volunteer at your local university, we need the help.
- stuxnet79 11y ago
  > Continue your passion! Neuro is so cool and we need comp-sci people, at the least to process the data and help out with connecting VGA cables (I shudder at bio people's hard-coding). What do you mean by the connectome is stereotypical? I was reading up on connectomes and was pleasantly surprised to find out that model organisms like C Elegans have been fully mapped but I haven't heard of any large insights being derived from this. In general what are the computational stumbling blocks in Neurosci at the moment? (I have a genomics background) I'd like to do a Phd in Cog Sci or Neurosci (computational) but I've been so stress free ever since moving to industry (from academia) 10 months ago I can't really see myself going back for the long haul.
  - Joof 11y ago
    About the stress of academia... I did some work with a cognitive science / computer science researcher while in undergrad and know plenty of PhDs in various fields. Academia is seeping with stress. Grad students have a very high rate of mental illness. I greatly admire the people who do that work, especially if they are working towards tenure. The work is incredibly fun, but you have to be a certain kind of person to handle the tenure route.
    - stuxnet79 11y ago
      So did you stop at the Bsc level or did you go ahead and do a Masters? Tenure is basically not worth it to me. If I went back to grad school I'd go with the intention of moving right back to industry afterwards. While I don't miss the constant stress I do miss the challenge, progression, joy of discovery and relative independence of Grad School.
  - Balgair 11y ago
    Stereotypical in the sense that most connectomes are the same amoungst the same sub-species. You mention C. elegans as an animal and it's connections are mostly the same no matter the animal. That is a huge finding btw, going in, there is nothing to suggest that they should be all pretty much the same. The neuro stumbling blocks are computational power and knowledge of the modulation of the circuit. We just can't model the circuits in sufficient detail right now and we don't know how to 'cheat' it such that we can still see most of the behavior but can do so in some fraction of the computation time. We need better and more efficient models. Your genomics work would be very useful in the numerical analysis of neural nets. The mammalian brain is also very mysterious still, C elegans is transparent, mice are not. We are having a heck of a time mapping it. The work is slow going, but still moving. Better modeling should accelerate the process and help us narrow where to look and the best bang for our buck in terms of experiments. I would not go back to grad school if I were you then. Grad school is a ton of stress for little reward (in terms of $$$). In terms of desire, it is very individual and mostly depends on the advisor. If you want to go back and have a super-duper advisor already waiting for you, by all means. But the stress load is extreme. 90+ hr weeks for 5-7 years and ~30k/year.
omginternets 11y ago
It's generally accepted that information sharing between distant cortical regions is a hallmark, if not the mechanism of consciousness. From there, there are many takes on that idea. Really, though, it helps to understand what led to the idea if you want to understand the scope of the theory. The first models of consciousness actually weren't; they were instead models of attention and executive control, but researchers began to tease out both the function and hallmarks of conscious processing in these theories. In the late 50's, Broadbent published on the famous "cocktail party effect", which basically states that bottom-up (extrinsic) attention can cause interrupts. In other words, you can be engrossed in a conversation, but still immediately notice your name being called out from across the room. He proposed a simple model that involves two steps: 1. a set of parallel sensory buffers that store information temporarily (later became sensory memory, e.g. iconic memory) 2. a unique, limited-capacity "perceptual system" The big idea is that certain operations can take place in parallel, while others are strictly serial in nature. Things like speaking are serial, but extrinsic attention is parallel and can quite literally interrupt a conversation. From there the model was refined in several ways: Baddeley and Hitch (1975) propose that between each processing step is a "loop" (the most famous of which is the phonological loop) that maintains information by rehearsing. The "central executive" then coordinates, selects, and triggers the maintenance and activity of the peripheral systems that implement the buffers. We now have a basic theory of multitasking by context-switch. Norman and Shallice (1986) further clarify the distinction between automatic, non-conscious processes and voluntary, consciousness-dependent processes. This distinction is what sets the stage for modern, neurocomputational theories and models of consciousness. To review, each of these models has the following in common: 1. A central (albeit ill-defined) component that is associated (albeit incidentally) with conscious perception and deliberate behavior. 2. The notion of an "observer" entity that issues command & control operations. (Note: the philosophical criticisms on this particular point are largely responsible for the ensuing paradigm shift, see Daniel Dennett) 3. Some independent, parallel functions. I like to think of them as goroutines. The modern take is that there is indeed a distinction between automatic-and-parallel processes and their voluntary-and-serial counterparts. The shift is in understanding the nature of this serial controller. We no longer see it as a volitional homunculus, but essentially as a router. The idea was first formally proposed by Bernard Baars (1989) and is called the "global neuronal workspace" theory. The idea is that there are hundreds (if not thousands) of independent local cortical processors that perform specialized tasks, in parallel. Consciousness, then, is the act of allowing these processors to communicate with each other through long-distance cortico-cortical connections (in particular, a fronto-parietal network of pyramidal neuron axons... but I digress), and this routing/broadcast structure is called the Global Neuronal Workspace (GNW). From there, there are many schools of thought about how local information is granted access to the machinery of consciousness. My Ph.D explores the theory that attention may be the gatekeeper of mechanism, the implication being that attention must intervene in order for consciousness to occur. Here's our latest publication, which shows that retroactively orienting attention towards an unseen stimulus can render it conscious after the disappearance of the physical source: http://goo.gl/dYdRcl http://goo.gl/dYdRcl The implication is that we have a "pre-conscious" (see: http://goo.gl/xQXjGe http://goo.gl/xQXjGe for an explanation of the term pre-conscious) stimulus encoded into the visual system, that is not consciously perceived, and that will never be consciously perceived unless acted upon by an outside force. We show that attention can serve as this outside force, thus arguing in favor of the GNW by virtue of the fact that no other theory posits such a level of access control on the machinery of consciousness. >I find it interesting to consider but have a hard time understanding how the brainstem fits into this. I suspect much of the confusion is related to terminology: you have to distinguish between consciousness and consciousness ;) It's well-understood that the brainstem (or rather, not the brainstem per se but the thalamus) regulates consciousness in the sense of sleep/wake/arousal. It also plays a critical role in cortical oscillation, and oscillation implies communication, which is necessary (but not sufficient) for consciousness. >-- psych. & computer science student with a strong interest in cognitive science & neuroscience Keep it up! Cogsci is increasingly computational, so we need hackers. We also need better software engineering... badly!

Balgair 11y ago

Continue your passion! Neuro is so cool and we need comp-sci people, at the least to process the data and help out with connecting VGA cables (I shudder at bio people's hard-coding). Interconnectivity in the brain is very stereotyped. Certain areas typically only talk to certain others. The 'connectome' in the brain is being mapped these days and there is a lot to learn, but I think we can say that the connectome is stereotypical. That's good, we can make a road map then! Consciousness is a tougher nut to crack, though. Many questions arise from disease studies like this one: https://www.newscientist.com/article/dn12301-man-with-tiny-brain-shocks-doctors/ https://www.newscientist.com/article/dn12301-man-with-tiny-b... . Obviously from that patient, he lives a normal life (within the IQ standard deviations) and no-one else saw deficits in him worth concern. Other patients report only balance issues and delayed speech with having half of all the body's neurons not develop: https://www.newscientist.com/article/mg22329861-900-woman-of-24-found-to-have-no-cerebellum-in-her-brain/ https://www.newscientist.com/article/mg22329861-900-woman-of.... Still other people have extreme deficits in short term memory from having comparatively small portions removed : https://en.wikipedia.org/wiki/Henry_Molaison https://en.wikipedia.org/wiki/Henry_Molaison . What I am saying is that the brain is a confounding mystery and we need the help of people in comp-sci. If you have the time, try to volunteer at your local university, we need the help.

stuxnet79 11y ago
> Continue your passion! Neuro is so cool and we need comp-sci people, at the least to process the data and help out with connecting VGA cables (I shudder at bio people's hard-coding). What do you mean by the connectome is stereotypical? I was reading up on connectomes and was pleasantly surprised to find out that model organisms like C Elegans have been fully mapped but I haven't heard of any large insights being derived from this. In general what are the computational stumbling blocks in Neurosci at the moment? (I have a genomics background) I'd like to do a Phd in Cog Sci or Neurosci (computational) but I've been so stress free ever since moving to industry (from academia) 10 months ago I can't really see myself going back for the long haul.
- Joof 11y ago
  About the stress of academia... I did some work with a cognitive science / computer science researcher while in undergrad and know plenty of PhDs in various fields. Academia is seeping with stress. Grad students have a very high rate of mental illness. I greatly admire the people who do that work, especially if they are working towards tenure. The work is incredibly fun, but you have to be a certain kind of person to handle the tenure route.
  - stuxnet79 11y ago
    So did you stop at the Bsc level or did you go ahead and do a Masters? Tenure is basically not worth it to me. If I went back to grad school I'd go with the intention of moving right back to industry afterwards. While I don't miss the constant stress I do miss the challenge, progression, joy of discovery and relative independence of Grad School.
- Balgair 11y ago
  Stereotypical in the sense that most connectomes are the same amoungst the same sub-species. You mention C. elegans as an animal and it's connections are mostly the same no matter the animal. That is a huge finding btw, going in, there is nothing to suggest that they should be all pretty much the same. The neuro stumbling blocks are computational power and knowledge of the modulation of the circuit. We just can't model the circuits in sufficient detail right now and we don't know how to 'cheat' it such that we can still see most of the behavior but can do so in some fraction of the computation time. We need better and more efficient models. Your genomics work would be very useful in the numerical analysis of neural nets. The mammalian brain is also very mysterious still, C elegans is transparent, mice are not. We are having a heck of a time mapping it. The work is slow going, but still moving. Better modeling should accelerate the process and help us narrow where to look and the best bang for our buck in terms of experiments. I would not go back to grad school if I were you then. Grad school is a ton of stress for little reward (in terms of $$$). In terms of desire, it is very individual and mostly depends on the advisor. If you want to go back and have a super-duper advisor already waiting for you, by all means. But the stress load is extreme. 90+ hr weeks for 5-7 years and ~30k/year.

omginternets 11y ago

It's generally accepted that information sharing between distant cortical regions is a hallmark, if not the mechanism of consciousness. From there, there are many takes on that idea. Really, though, it helps to understand what led to the idea if you want to understand the scope of the theory. The first models of consciousness actually weren't; they were instead models of attention and executive control, but researchers began to tease out both the function and hallmarks of conscious processing in these theories. In the late 50's, Broadbent published on the famous "cocktail party effect", which basically states that bottom-up (extrinsic) attention can cause interrupts. In other words, you can be engrossed in a conversation, but still immediately notice your name being called out from across the room. He proposed a simple model that involves two steps: 1. a set of parallel sensory buffers that store information temporarily (later became sensory memory, e.g. iconic memory) 2. a unique, limited-capacity "perceptual system" The big idea is that certain operations can take place in parallel, while others are strictly serial in nature. Things like speaking are serial, but extrinsic attention is parallel and can quite literally interrupt a conversation. From there the model was refined in several ways: Baddeley and Hitch (1975) propose that between each processing step is a "loop" (the most famous of which is the phonological loop) that maintains information by rehearsing. The "central executive" then coordinates, selects, and triggers the maintenance and activity of the peripheral systems that implement the buffers. We now have a basic theory of multitasking by context-switch. Norman and Shallice (1986) further clarify the distinction between automatic, non-conscious processes and voluntary, consciousness-dependent processes. This distinction is what sets the stage for modern, neurocomputational theories and models of consciousness. To review, each of these models has the following in common: 1. A central (albeit ill-defined) component that is associated (albeit incidentally) with conscious perception and deliberate behavior. 2. The notion of an "observer" entity that issues command & control operations. (Note: the philosophical criticisms on this particular point are largely responsible for the ensuing paradigm shift, see Daniel Dennett) 3. Some independent, parallel functions. I like to think of them as goroutines. The modern take is that there is indeed a distinction between automatic-and-parallel processes and their voluntary-and-serial counterparts. The shift is in understanding the nature of this serial controller. We no longer see it as a volitional homunculus, but essentially as a router. The idea was first formally proposed by Bernard Baars (1989) and is called the "global neuronal workspace" theory. The idea is that there are hundreds (if not thousands) of independent local cortical processors that perform specialized tasks, in parallel. Consciousness, then, is the act of allowing these processors to communicate with each other through long-distance cortico-cortical connections (in particular, a fronto-parietal network of pyramidal neuron axons... but I digress), and this routing/broadcast structure is called the Global Neuronal Workspace (GNW). From there, there are many schools of thought about how local information is granted access to the machinery of consciousness. My Ph.D explores the theory that attention may be the gatekeeper of mechanism, the implication being that attention must intervene in order for consciousness to occur. Here's our latest publication, which shows that retroactively orienting attention towards an unseen stimulus can render it conscious after the disappearance of the physical source: http://goo.gl/dYdRcl http://goo.gl/dYdRcl The implication is that we have a "pre-conscious" (see: http://goo.gl/xQXjGe http://goo.gl/xQXjGe for an explanation of the term pre-conscious) stimulus encoded into the visual system, that is not consciously perceived, and that will never be consciously perceived unless acted upon by an outside force. We show that attention can serve as this outside force, thus arguing in favor of the GNW by virtue of the fact that no other theory posits such a level of access control on the machinery of consciousness. >I find it interesting to consider but have a hard time understanding how the brainstem fits into this. I suspect much of the confusion is related to terminology: you have to distinguish between consciousness and consciousness ;) It's well-understood that the brainstem (or rather, not the brainstem per se but the thalamus) regulates consciousness in the sense of sleep/wake/arousal. It also plays a critical role in cortical oscillation, and oscillation implies communication, which is necessary (but not sufficient) for consciousness. >-- psych. & computer science student with a strong interest in cognitive science & neuroscience Keep it up! Cogsci is increasingly computational, so we need hackers. We also need better software engineering... badly!