Lecture 8: Global Network Theory

Question 1

Describe the general idea of global neuronal workspace theory in detail

Answer 1

1. On the first graph we see that the early feedforward processing is independent of the awareness. The amount of activation is more or less the same for all making levels, around 100ms.
2. Then, always on the feedforward sweep, when this information travels up in the hierarchy, towards the temporal regions of the ventral stream (what pathway). The strength of activation decays gradually depending on the strength of the input, so the consciousness level. The stronger the input (the weaker the masking), the higher the activation.
3. Afterwards, in some of these situations, there is crossing of the consciousness threshold (yellow, purple and light blue). These activate the global ignition by Global Neuronal Workspace neurons in parietal and frontal cortex. This is an all-or-none phenomenon, the information either crosses the threshold or not. So the ones who passed the threshold show almost the same strength of activation with each other, same goes for the ones who couldn’t pass.
4. Once the information activated these GNW, we also get a global feedback. In these feedback activations we see the same all-or-none profile. So the above-thresholds sent feedback and others didn’t. We call it feedback when the information is sent from higher level regions to lower level regions. E.g. from prefrontal cortex to the visual cortex.

Question 2

According to GNWS theory, what happens if there is a weak or interrupted bottom-up stimulus strength and attention is absent?

Answer 2

Subliminal (unattended):

• very little activation
• Activation is already weak in early extrastriate areas
• little or no priming
• No reportability

Question 3

According to GNWS theory, what happens if there is a weak or interrupted bottom-up stimulus strength and attention is present?

Answer 3

Subliminal (attended):

• Strong feedforward activation
• Activation decreases with depth
• Depth of processing depends on attention and task set
• Activation can reach semantic level
• Short lived priming
• No durable fronto-parietal activity
• No reportability

Question 4

According to GNWS theory, what happens if there is a sufficiently strong bottom-up stimulus strength and attention is absent?

Answer 4

Preconscious:

• Intense activation, yet confined to sensorimeotor processors
• Occipito-temporal loops and local synchrony
• Priming at multiple levels
• No reportability while attention is occupied elsewhere

Question 5

Give an example of when you may be preconscious of a stimuli

Answer 5

Attentional blink, inattentional blindness

Question 6

What happens when attention is present and the stimulus strength is sufficiently strong?

Answer 6

• Orientation of top-down attention
• Amplification of sensorimotor activitry
• Intense activation spreading to parietofrontal network
• Long distance loops and global synchrony
• Durable activation, maintained at will
• Conscious reportability

Question 7

At which stages is conscious access possible?

Answer 7

Only at the conscious stage

Question 8

What does this model also claim through these stages?

Answer 8

That attention plays a very important part in consciousness

Question 9

Describe the procedure of a study which demonstrates how attention can modulate unconscious processing

Answer 9

In the masked no-go condition, subjects go through 3 trials randomly. They either have a conscious no-go meaning that they see a grey dot and inhibit their response, they have a masked/ unseen grey dot, or there is no dot at all. So subjects make an association between inhibition (no-go) and the grey dot, we analyse their reaction to the unconscious grey dot.

In the control condition, subjects again go through 3 trials but this time the unconscious no-go and the conscious no-go has different stimuli so there is no association between the grey dot and the no-go reaction.

Question 10

What were the EEG results of this study?

Answer 10

The EEG results show that only the attended unconscious stimuli traveled far, although the early processing was the same for all. Therefore, attention can shape the processing of unconscious information, without making them more conscious.

Question 11

How does this study show the difference between subliminal and preconscious information?

Answer 11

When we attend to a stimulus but mask it very strongly, this can lead to deeper feedforward processing, but still don’t lead to any conscious access.

Question 12

What does SOA represent?

Answer 12

SOA represents the delay we give between the screens, so regards the mask

Question 13

Perception is not driven really by the physical stimulation, so how does subjects have different percepts?

Under identical SOAs and other identical conditions, how is that they sometimes see it and sometimes don’t?

How do some stimuli cross the threshold of consciousness and others do not?

Answer 13

This is because of the internal fluctuations in brain activity that dictates whether we will see a stimulus or not. We are much more likely to see a stimulus when we are in high state, compared to when we are in a down state. Therefore this internal fluctuations decide whether the threshold is reached or not.

Question 14

What area of the brain plays an important role in vigilance?

Answer 14

Thalamus plays an important role in the awakening of the entire network. It relates to changes in state, switch from unconscious to conscious brain. Regulates the arousal levels of the entire cortex.

Question 15

What changes are observed when our vigilance is low? (2)

Answer 15

Spontaneous activity is drastically reduced, and the ignition proper disappears, even strong sensory input quickly decays without triggering global ignition.

The network behaves as a sleepy, anaesthetized brain. It responds to stimuli but only in its peripheral sensory areas and activation typically fails to climb all the way up to workspace areas and ignite a full-blown cell assembly.

Question 16

Describe in detail what is meant by the workspace in GNWT in regards to activation

Answer 16

In the local modules, we can have activation even when they are isolated from the global workspace. This activation is not consciously accessed.

In regards to the diagram in the notes:
(green: conscious, high strength and attention) When we have a very strong stimulus and the information is attended, then that information may be processed by (periphery sensory regions) the local modules all the way up to the prefrontal cortex and ignite the system and we can get a broadcasting of information and the other sensory regions are also informed by the stimulus that is presented.  T1 travels all the way up to the global workspaces and the other local module also gets informed by that and turns into green. 

(orange; Preconscious: high strength, no attention) when we have a preconscious target that is for example the T2 in an AB task that isn’t perceived. Because there is already a stimulus accessed by the global workspace, that blocks the entrance of this informations entrance. So it cannot reach the global workspace and its information cannot be accessed consciously. Even though it is strong, there is no attention to it so it cannot be globally broadcasted.

(red; Subliminal: weak strength) the input is there but the information dies while traveling to the global workspace so doesn’t reach there. It doesn’t ignite the system.

Question 17

How can we assess what regions of the brain are part of the Global Workspace?

Answer 17

Results of white matter tracking in the brain. It shows us certain hubs in the brain that are relatively far to each other, yet are strongly connected. Strong functional and structural long-distance connectivity between different “default mode” areas in prefrontal, parietal, temporal and posterior cingulate cortex (part of global workspace).

Question 18

Name five ‘hubs’ included in this global workspace

Answer 18

Inferior Parietal Cortex (IPC)
Posterior Cingulate Cortex (PCC) and precuneus
Medial Prefrontal Cortex (mPFC)
Anterior Cingulate Cortex (ACC)
Medial Temporal Lobe (MTL)

Question 19

Brain connections are visualised in a diagram on the slides. Describe this

Answer 19

Red dots labeled as brain areas with connections between the red dots being displayed as blue lines. In the image, the larger the red dot is, the larger its regions connections with other parts of the brain.

Question 20

What is meant by rich club phenomenon?

Answer 20

Rich club phenomenon is when hubs of a network tend to be more densely connected among themselves than aides of a lower degree. These rich clubs are also mostly part of the Global Workspace.

Question 21

How do these rich clubs relate to the global workplace?

Answer 21

So the idea is that this global workspace is a densely connected network of fronto-parietal and temporal cortices. These regions are strongly connected with each other but also highly connected with other regions

Question 22

Do sensory regions or association cortices have stronger and longer distance connections?

Answer 22

It is a phenomenon of the brain that association cortex has stronger and long-distance profiles than the sensory regions.

Question 23

Name two other ways in which association cortices are special

Answer 23

Regarding variability association cortices has the largest variability between people. It also matures late and possesses functional properties that are different from sensory areas.

Question 24

What network overlaps a lot with GNWT?

Answer 24

There is strong overlap between GNWS and default mode network. Default mode network is when there is no task-related activity in the brain and it is in its ‘default mode’. The subject isn’t doing anything.

Question 25

So, in a sentence, what does the GNWT posit is the neural correlate of consciousness?

Answer 25

Brain wide information sharing (and implies connectivity and re-entrant loops), global availability of information (GWS) is what we subjectively experience as a conscious state (“identity relationship.”)

Question 26

What two different processes determine unconscious and conscious processing?

Answer 26

There are important differences between local processing modules (unconscious and isolated activity) and global workspace processing (conscious and integrated activity)

Question 27

Give a very broad description of the regions involved in the GNWS

Answer 27

Frontal cortex, parietal cortex, cingulate cortex, precuneus, temporal cortex

Question 28

Describe research carried on humans out by racist daddy Lamme which lends credence to this theory

Answer 28

Receptive fields become bigger along the cortical hierarchy and the processing complexity of the information also increases. Feedforward goes from early to higher level regions. Feedback reports from higher level to lower level regions.

Lamme measured the spiking activity of a neuron in V1, to 2 different stimuli (Both almost static- diagonal lines, Left has the orientation of the lines in the centre reversed creating the percept of a square). Although these stimuli are different, the receptive field of this V1 neuron was actually looking at the exact same thing. Therefore the neuron gets the exact same information from both stimuli.

However, on the spiking activity we see that there is a difference. The neuron reacts much stronger to the stimulus on the left. This difference is possible because of the feedback that that neuron receives.

Question 29

Describe research carried on primates out by racist daddy Lamme which lends credence to this theory (2)

Answer 29

Lamme then did research on higher-level lesioned brains. Then it was seen that the early response in the normal situation was identical to the one with the lesioned brain (the first peak on the graph). So the feedforward sweep is unaffected, it is sent to the higher-level regions. But as these regions are lesioned, they can’t send a feedback information. This was also done on anesthetized monkeys and in was again seen that the contextual modulation was absent. So no feedback processing.

They also did it after training the monkeys to respond to the peripheral shape. They were rewarded it if they reported seeing the shape, and they would fixate in the middle if they didn’t see the peripheral shape. So now we can separate between the cases the monkey has seen the stimulus vs didn’t see the stimulus. In the seen trials, there is again a contextual modulation but it wasn’t present for not seen trials. So feedback in higher order areas isn’t triggered and monkey doesn’t see the peripheral stimulus.

Question 30

How has TMS also been utilised to provide evidence for these processes?

Answer 30

Researchers presented flashes of light and tried to interfere with the feedback sweep. TMS at visual cortex (back of the brain) at 100 ms after stimulus presentation “blocks” feedback and reduces conscious experience.

Shows causal evidence for the necessity of feedback processing for conscious experience