University Studies 15A: Consciousness I

University Studies 15A:Consciousness I The Neurobiology of Consciousness

First topic for discussion: the final paper https://eee.uci.edu/11f/87670/FIP_WritingAssignment_2.pdf

Please note the requirements for the paper: “Your thesis statement should make an arguable and specific claim. After thinking about the specific features on which you wish to focus and as you prepare your argument, try writing several statements defending the importance of those features.” And“Remember that a good claim will need to be arguable, meaning that you must provide logical evidence to prove your point and a peer must be able to argue against your claim.” This perhaps puts a particular burden on people who wish to argue a dualist position that neurobiology does not and cannot account for that which is at the basis of consciousness: you will need a specific, arguable claim.

Neurobiology of Consciousness What is “consciousness?” In philosophical terms: 1. Access Consciousness: what one can report 2. Phenomenal Consciousness: the subjective quality of experience In neurobiological terms: 1. Attention: brain resources fixated on an object or event 2. (Phenomenal) Consciousness: the brain parses an object or event for meaning and assesses the need to act in relation to that object or event. The two neurobiological systems are different. However, as we shall see, they blend into one another when we turn at last to a current, widely shared neuroscientific model for consciousness as an integrated system.

Neuroscientists often think in terms of the state of attentional resources. When one is fully awake, one can devote the most resources:

We already have seen that the thalamus plays a central role in the arousal of the attentional system. The thalamus also plays a key role in gating attention. It is connected to almost every important cortical region:

When we talked about the processes of decision, we discussed saliency as a key factor in the neuronal calculus of decision. There seems to be a “quick-and-dirty” bottom-up saliency system for visual attention that works through the pulvinar in the thalamus:

In contrast, the “voluntary” control of visual attention comes from the prefrontal lobe (which then interacts with the pulvinar).

Neurobiologists tend to conflate the attentionalsystem with consciousness. Baars & Gage in “Putting it all together” focus on how sensory data gain the attentional resources of the system. The attentional system focuses on particular objects of concern, but the brain is still at work noting other sensory inputs that fall outside explicit attention. Neuroscientists have explored high-order processing outside attentional focus in great detail. One type of experiment involves masking visual images: showing them very briefly and then showing a replacement image. The subjects have no “conscious” awareness of having seen the first image, but, for example, they show semantic priming responses. That is, shown an angry face, a subject will recognize an “angry” word faster than a “happy” word. That is, the semantic system of the brain has parsed the “unseen” face and there is an expectation or at least an activation of “angry” meanings.

“Conscious” processes outside attention and the Default Mode Network I have discussed the Default Mode Network before: it is the set of cortical regions that are active when no attentional tasks are active: Cingulate cortex Posterior parietal VMPFC DLPFC MTL complex

The Default Mode Network (DMN) thus involves the executive functions of the DLPFC, the emotional integrations of the VMPFC, the wide range of assessment functions of the anterior cingulate cortex and the integration of all this in the posterior parietal cortex. It usually is described as a system for specifically introspective assessment. A crucial aspect of the DMN is that it is inhibited when the focal attentional system becomes active: those integrative resources are recruited by “attention” instead. Thus, even though there are priming effects for unattended visual data, the visual data that receives attention produces far more widespread interpretive, integrative activation in the frontal lobe and the posterior parietal cortex.

With or without attention, this spread of activation and communication between regions is made possible by the vast system of bidirectional connectivity between the regions:

When the visual system transmits the neural activations for an image but that image is not registered as demanding attention, it is processed by all the usual neural networks for assessing the image: The “face” is “reassembled as a visual object in the inferotemporal cortex. The “face” is processed for emotional valences (“happy,” “sad”) with appropriate activations in the amygdala, VMPFC, orbitofrontal cortex, the insula, the MLT episodic memory system and the rest of the semantic memory system, and thus contributes to the sets of neural networks active at the moment (otherwise there would be no priming effect). This ensemble of activations spreading throughout the cortex is, I suggest, the neuroscientific mechanism behind phenomenal consciousness. Things of which we are phenomenally conscious do not necessarily impinge upon explicit “consciousness.”

Explicit Consciousness When neurobiologists construct models for consciousness, they primarily focus on the systems that process objects and events that have focal attention. To keep the terminology clear, I shall call this sort of consciousness focused on data that receives attention explicit consciousness. There are good reasons for looking at explicit consciousness (the combination of attention and “conscious” integration) as the core of conscious experience. First, and most compelling, is the fact that when the attentional system is turned off by turning off the thalamus, consciousness of any sort disappears. Prof. Gage discussed anaesthesia on Tuesday: it primarily works through the thalamus (and perhaps the PPC). Not only is attention eliminated, but the sort of cortical activity associated with the Default Mode Network disappears. Patients luckily do no know “what it is like” to be operated on.

Secondly, biologists see the function of explicit consciousness as the evolutionary “point” of consciousness. Biologists stress the value of being able to integrate vast amounts of information (both current and historical) about current circumstances in shaping responses In this view, controlling focus on a particular object or event is extremely important. Hence Baars and Gage note the “winner-take-all” nature of explicit consciousness.

Models for Explicit Consciousness There are two basic models that are in fact complementary. There are also other models that are minor variations on the basic approach. 1. The Global Workspace Model:

The idea is that there is a core of interconnected high-level integrative “work spaces.” This is the network at the center of the diagram. These work space networks draw on other networks that essentially function as automatic local (unimodal) data organizing units.

Remember the importance of long-range connectivity: The Global Workspace was developed as an artificial neural network model to see if an artificial neural network with the right sort of connectivity could produce the sort of behavior observed in the cortex. The answer was “Yes:” a core network using affective data for evaluation could selectively draw on peripheral nodes in a “winner-take-all” process of integrating data.

Bernard Baars, in developing the model, added a particular interpretive view: He suggested that we can think of the front (executive) part of the cortex as “watching” the sensory back half of the cortex. That is, he reinterpreted and preserved the traditional images of the “theater” and “homunculus.” However, these two concepts are not inherent in the model, which, quite the contrary, uses the sort of calculus of “emotional” value to control attention we have discussed (i.e., no homunculus). And the model also requires that cortical processing be widely distributed rather than localized (no theater, but maybe streaming video)

2. Consciousness as Thalamocortical Loop The thalamocortical loop model requires the sort of integrative areas described in the global workspace model But, the model requires the synchronization of neural firing between the integrated neural networks. The thalamus provides the mechanism for synchronizing the complex looping of firing.

The data for this model looks very good. 1. It integrates the role of the thalamus in controlling attention. 2. It explains why consciousness in general needs a functioning thalamus. 3. The thalamus has the right sort of connections to serve this role:

4. The model explains the need for all the reentrant (top-down) connections between higher level cortical regions and lower level sensory cortices: consciousness requires the synchronization of firing patterns. 5. The model explains why “masked” images do not become explicitly conscious: explicit consciousness requires a synchronization between higher and lower networks that take at least 100ms to establish.Masked images did not persist for that long, and they were replaced by other images that could establish the necessary synchronization.

Baars and Gage ask whether this sort of explicit consciousness requires a self. I hope by now it should be clear that the “self,” encoded in the visceral and “emotional” values of memory and on-going perceptionas well as in the very structure of episodic and semantic memory, is integral to the entire system. One can remove on-going “self” data from the processes of consciousness by disabling the posterior parietal cortex (and producing out-of-body experience, etc.), but the brain, by its structure, development, and on-going dynamics has a “self.” If we understand “self” in this deeply embedded, deeply embodied way, then we surely can say that the self underlies conscious experience.

University Studies 15A: Consciousness I