<BR><BR>Hey everybody, I’m new here! Some of you I know already, others I look forward to getting to know.<BR><BR>Anyways, one issue that I have been interested in, and which I do not think has received as much attention as it deserves, is that of giving a theoretical account of what a brain state is. Crucial to making progress on this issue is distinguishing between two different kinds of states that will figure in our science of the brain. It is only after we settle this theoretical issue that we can make progress on other philosophical issues like the viability of the identity theory or the multiple realizability of mental states. Here is how I formulate the distinction taken from my paper <A href=)
What is a Brain State? (Phil Psych 19(6)).
I want to introduce a distinction between brain states and states of the brain (cf. Chalmers’ specific and background NCC’s (Chalmers 2000)). Particular brain states occur against, and only against, background states of the brain. By ‘states of the brain’ I intend to include such states as being awake or being asleep; although there may turn out to be more than these two once we fully explicate what one is. However many there are, intuitively a state of the brain is the overall state that the brain is in and this strikes me to be a very different kind of thing from such local states like perceptual or motor representations. In the framework I am advocating trying to explain brain states in terms of states of the brain (as Putnam wanted to do) is to conflate these two kinds of states. Any theoretical account of the brain and its states should explain both kinds as it seems natural to expect each kind of state to play a distinct explanatory role. It may, of course, turn out that the distinction ultimately collapses but absent such an argument I propose to see how far I can get with it. I now want to argue that certain chemical states are good candidates for states of the brain, whereas certain electrical states are good candidates for brain states.
I go onto argue in the paper that brain states are sychronized neural activity in a particular frequency and that states of the brain turn out to be the gating and modulating of neural activity by the brain stem and so turn out to be global chemical (neuromodulators) states. These two states of the brain correspond to the various states of creature consciousness (in the case of states of the brain) and the representational/intentional content of mental states (i.e. the transitive conscious states). If this is right, then there is very little reason to think that brain states are in fact multiply realized here on Earth as synchronization in a frequency seems to be a strategy that all evolved brains use. There are other implications if this account of what a brain state is turns out to be right (we are going to have to wait for multiple-cell recording technology to become more advanced); like that fMRI’s do not give us pictures of brain states and also that Rosenthal’s kind of higher-order theory of consciousness, where the higher-order thought and the first-order target states are seperate states, will ultimately turn out to be Kriegl’s same-order account that treats the two states as distinguishable parts of a complex whole.
12 Comments
Comments are closed.
The link is messed up.
Thanks for the tip…I think I fixed it now…
Dear Richard,
sorry for the late comment, I hope it can still be of some use.
— BTW we never introduced ourself but we met at Pete Mandik’s talk at CUNY —
I want to add something about the idea that brain states are synchronized neuronal oscillations in the gamma frequency range; and that synchrony corresponds to the representational/intentional content of mental states.
(1) It seems to me that synchrony of oscillations might indicate relations among neuronal groups but that it does not represent the content of neuronal representations.
The object that we perceive is decomposed in primary features — feature maps –in early visual areas. Whenever a specific stimulus is signaled the corresponding feature maps activate. Thus, the activity of a feature map codes for or represents characteristic of the perceived stimulus.
The role of oscillation is, instead, to signal which are the features that belong to the perceived object. Suppose you have a green horizontal and a red horizontal lines: feature maps for green, red, horizontal, and vertical are activated. But how does the visual system know which feature belong to what?
Synchronization of oscillatory activity might be a way to tag feature maps that code for features of the same object. Thus, neurons coding for features belonging to the same object synchronize their activities (probably at around 40 Hz) and desynchronize their activities when they code for features belonging to different objects. In this sense, feature maps represent the content and synchrony bound related features together.
Binding by synchrony is a fascinating hypothesis but it also has some problems: For example, it cannot account for binding of novel features, since it is quite unlikely that neurons coding for these new combinations are linked together. Since connections among neurons are what make synchrony possible, without them you cannot explain how novel conjunctions are bound.The tendency, nowadays, is to link binding by synchrony and attention for explaining how binding of features and novel features is accomplished (Crick and Koch, 1990, Sem.Neurosc. ; Fries et al, 2001, Science).
(2) About a possible relation between synchrony of neuronal oscillatory activity and transitive consciousness:
Synchrony of oscillation is part of a complex computational process that results in visual awareness. Here, I think there are two problems: (1) visual awereness might not have the same intentionality of ‘being conscious of’, since it might be nonconceptual; and (2) Synchrony is linked to awareness only in the sense that amplifies signals that need to be further processed.
All of that turned to be longer than I thought!
Best,
Michela
Hi Michela,
Better late than never!!!
Sorry we didn’t get a chance to meet…do you regularly come to the CUNY CogSci meetings or was it only for Pete’s?
I think you raise some good points, so thanks for the comment.
RE 1: In a way what you say here is not a threat to what I say. I claim that when one sees a red horizontal bar the brain represents that via a specific pattern of synchronous firing of certain neurons. Though I didn’t emphasize it in the post (but I do in the paper), this depends on neurons having ‘preferences’ or ‘being tuned’ to certain physical properties. You say that it is the activity of the neurons in feature maps that represent or code for specific features and the synchrony is merely a way to group the features into a unified whole and you are in part correct. Synchrony is proposed to solve the ‘binding problem’ as you say, but Singer also proposes that it is a general strategy used by the brain and can code for features and motor acts (as well as ‘relatedness’)…so the hypothesis is that there is synchronous activity that codes for each feature and when various neuronal assemblies coding for the separate features come to fire in synchrony then we have a representation of an object and not merely features of an object. So when you say “the activity of a feature map codes for or represents characteristic of the perceived stimulus,” I agree except that I think that that activity will turn out to be synchrony in a frequency…
As for the novel stimuli objection to synchrony…it is not an impressive objection. For, it is commonplace that experience changes the connectivity of the brain. The connections in the brain are not static but ware constantly changing…besides which since most of the things we see are built up out of the same basic features and experiments show that the same neuron can participate in many different neuronal assemblies, synchrony is a great way to explain how novel stimuli …however if something is SO novel that it can’t be explained in this way then, sure, attention is fine. IF only someone knew what attention was!
RE 2: transitive consciousness is when you are conscious of something. You become conscious of something by (a) sensing it or (b) thinking about it. If synchrony in a frequency turns out to be a way to explain how we sense things that it is and explanation of transitive consciousness. It explains how I can be conscious of a red horizontal bar. How? Because I have neurons that are sensitive to those properties (yeah, yeah, I know the story really starts with ‘I have certain kinds if light sensitive opsins’ but skip it for now as that is part of the story about how I come to have neurons that are ‘tuned’ to certain physical properties) and those neurons fire in a certain way resulting in a mental representation of the red horizontal bar. Whether said representation is conceptual or not doesn’t seem to matter to the point that I am making.
As for your second worry about synchrony merely amplifying a signal that needs further processing…Even if this were the only way that synchrony were related to awareness it wouldn’t be all that bad. What else is it to be conscious of something in the world than to have that thing affect your behavior and thinking…so whether I consciously see some stimulus or not it primes me to say one thing or another, or to feel one way or another, or whatever…so too whether I consciously see the other car on the road I am able to slow down enough to avoid hitting it and so on…amplifying the signal seems like a good way to recruit resources and to guarantee better/more rapid processing…but at any rate I don’t agree that this is the only relation to awareness that synchrony has. Aside from the story I told above about seeing a horizontal bar, you should look at the Buzsaki model of memory formation and storage that I talk about in the paper…the whole theory is formulated in terms of various frequencies…Gamma is not the only frequency!! The brain seems to utilize different frequencies for different purposes…mostly due (it seems) to the particular propoerties of the kind of neuron the frequency is generated by…anyway I hope that helps.
Hi Richard:
Just a brief comment: I can see the point for supporting synchrony approaches. They offer an elegant model to explain brain states and processes. My main concern is about the lack of empirical evidence to support some of their claims (Lamme & Spekreijse, 1998; Shadlen and Movshon, 1999; Thiele and Stoner, 2003; Palanca and De Angelis, 2005)
<>
We are probably referring to two different ways to consider synchrony. I am only referring to the binding by synchrony hypothesis (Von der Marlsburg 1981; Singer & Gray, 1995). In this case, there is a sharp distinction between features representations and synchrony: Features are represented by cells’ firing rates and synchrony relates features together. Of course, a new representation emerges once the features are link together. So you can ask what is the content that synchrony adds to the percept representation. But the content of the single feature maps is represented solely by neurons’ firing rates.
<> >
Attention can be defined operationally as a mechanism that selects brain states to be further processed. Evidence for the existence of attention can be found in every journal every month…
<
My reference for supporting the hypothesis that synchrony augments the chances of a signal to be further processed is Fries et al. 2001. I read the paper and I found it convincing. However, I agree that the relation between synchrony and awareness can be explained in some other ways.
<>
I hope to be able to come regularly
Cheers,
Michela
Hi Michela,
Thanks for the reply.
I share your worries about the empirical stuff…the matter is not settled and more work needs to be done (but to be fair, there is alos a lot of support for the theory. Some that even contradicts some othe claims in your references see Miguel Castelo-Branco et al 2000 in Nature; Biederlack J, et al 2006 in neuron; WOMELSDORF, T 2007 in Science; Larimer & Strowbridge 2007 in Nature)
I don’t think we are disagreeing about the synchrony/single neuron stuff…my only point was that it may turn out that the features are represented by cell assemblies instead of single neurons. I mean is it really supposed to be the case that whenI see red only ONE neuron starts firing? But I agree that this is the way that the theory is standardly cast.
The attention literature is a mess. Everybody means something different by it…though I agree that there are papers on ‘it’ regularly.
I haven’t been able to come to the meetings since Pete’s talk because I am teaching summer school but perhaps I’ll see you in the Fall, I would be happy to talk about this stuff if you wanted.
Binding of separate features must occur in space as well as time. So synchrony/concurrence of evoked features cannot alone explain perceptual binding.
For a detailed account of brain mechanisms that *can* support phenomenal binding see:
https://people.umass.edu/trehub/
Hi Arnold, if I may call you that,
Thanks for the comment!
I will try and take a look at some of your stuff, and maybe it will help, but I’m not sure that I see what the problem is supposed to be…is it that an experience of (say) a blue triangle must not only bind the blueness and the triangulaity but also that it must be located at some place in space? If so I don’t see that there is immediately a problem. For, if there are ‘sensory qualities’ then locating an object in space will just be having some other neurons in the synchronous cell assembly. I know that there is a debate over whether there are such spatial qualities (e.g. Austen Clark denies that there are) but there are those who argue that there are (e.g. David Rosenthal)…or were you suggesting some other problem?
Richard,
The problem is this:
When you look at a blue triangle, your phenomenal experience is that of a blue triangle some distance from you, at a particular 3D location within your egocentric visual space. Synchronous neuronal discharge barely begins to explain the
feature-binding required in this perceptual experience. The 2D retinal images project to the visual cortex where the triangular shape and its blue quality are each represented by neuronal activity in spatially separated visual regions. How do we explain the precise phenomenal spatial registration of the color blue within the contour of the triangle in 3D egocentric space when shape and color are idependently processed in 2D retinocentric brain maps? This the binding problem that the retinoid model solves.
Thanks for the response.
I must confess that I still don’t see a problem in principle. As I said if it is plauible that we have an internal representation of space then the color’s being at a certain location in egocentric space is explained (ex hypothesi) by the neurons that represent the location of teh color firing in synchrony witht he neurons that represent the pther features of the object. The fact that the image on the retina is 2D doesn’t seem to pose a problem because (as far as I know) the visual system moves from 2D to 2 1/2D to 3D by well known computations (i.e depth cues derrived from binoccular information etc)
You wrote “The fact that the image on the retina is 2D doesn’t seem to pose a problem because (as far as I know) the visual system moves from 2D to 2 1/2D to 3D by well known computations …” Even if there were consensus about the computations by which our visual system might move from 2D TO 3D representation, I thought your interest was in explaining how the brain works in order to achieve perceptual/phenomenal binding (brain states?). In this case, it seems to me, you should be looking for an explanation of binding in terms of brain mechanisms and systems rather than in a possible mathematical description. Also, the problem is not simply one of going from a binocular 2D representation to a 3D representation. It is the crucial and much more difficult problem of making the neuronal transformation from patterns in 2D retinocentric space to properly bound 3D representations in egocentric space. Temporal binding of separate visual features by synchronous brain activity is the easy part. Neuronal conversion of visual features in 2D retinocentric space, over a succession of saccades, into a coherent phenomenal experience in 3D egocentric space is the much more difficult problem.
Synchrony is a brain mechanism, or so I argue.
Also, if, as I have suggested, there are spatial qualities (not only of locations but also for shapes) then these are just more features that need to be bound, and synchrony is the perfect candidate for that. So, I still don’t see what the objection is supposed to be. Granted that the retinoid model may be a possible explanation of how this is done, so may synchrony.