4. Neural Correlates

Question 1

why are neuro-correlates interesting

Answer 1

This is interesting as it can explain why priming happens

Question 2

ERP study on semantic/ associate priming

Answer 2

Schweinberger, 1996:

• ERP study on semantic/associate priming
• Participants shown primes or shown neutral stimuli and then target
• Related (highly associated) target stimuli elicit more positive ERPs between 400-500 ms than unrelated targets

 N400 priming effect

• N400 effect is similar for within- (name-name, face- face) and between-domain priming (name-face, face-name)
• Unrelated condition is relatively more negative than the related 400ms-600ms

Question 3

What does the ERP study on semantic/ associate priming indicate?

Answer 3

This is good because it is not the same stimulus but not relation but it is clear we can access semantic information more easily if it is pre-emtive- this is in line with studies. Also is later in time then other things we have looked at (N170, N250) this is later so follows the idea that it is a step by step process.
Get this effect irrelevant of whether you present face and name, name and face ect. Can always get this difference irrelevant of the domain
This is also nice because replicates behavioural studies

Question 4

What would IAC predict for neural correlates of priming effects

Answer 4

According to IAC should get a priming effect for same category and would assume that would get a greater priming effect with associated categories. Basically according to IAC if we have something at all we should get something.

• This would suggest a different organization of identity-related semantics!

Question 5

neural correlates Evidence for IAC and what did they want to find?

Answer 5

Wiese and Schweinberger, 2011

• IAC explains semantic/associative priming via shared SIU.
• Alternatively, co-occurrence (association) may explain priming effect.
• This would suggest a different organization of identity-related semantics!

Question 6

what did Wiese and Schweinberger, 2011 do?

Answer 6

Priming experiment using associated (e.g., Angelina Jolie with Brad Pitt), same category (e.g., Hugh Grant with Brad Pitt), and different category (e.g., John Lennon with Brad Pitt) prime/target pairs
P

Question 7

what did Wiese and Schweinberger, 2011 find?

Answer 7

• N400 effect for both associative and categorical priming supports IAC- serves as a neurocorrelate for priming

Question 8

explanation for findings by Wiese and Schweinberger, 2011

Answer 8

Explanation for this is probably because the prime and the target share semantic information- not because they have a link in that you see them together all the time but is the similar semantic info. They are not linked by a link together but linked by higher level semantic info

Question 9

overall what does Wiese and Schweinberger, 2011 study show us?

Answer 9

Tells us about how semantic info is represented in the mind- not that everything that you know about someone is stored as part of them but instead that it is abstract and that everyone who shares this semantic info is linked to this abstract- so tells us about knowledge storage in the brain.

Question 10

but what is the issue with the Wiese and Schweinberger, 2011 study?

Answer 10

difficult to disentangle all if this information
when you use celebrities as stimuli you can not disentangle visual co-occurance and semantic info- is hard to find two celebrities who co-occur but have no similar semantic information.

Question 11

what would be ideal to disentangle?

Answer 11

Ideally what you would want if you want to disentangle the idea of visual co-occurrence and semantic info is that you would have one condition where you would have only co-occurrence and nothing and one condition where only shared semantic info and nothing else

Question 12

why is it so hard to completely disentangle

Answer 12

It is very hard to completely disentangle co-occurrence and semantic overlap with celebrity stimuli (e.g., highly associated pairs also share semantics)
Last experiment had pairs that shared only semantic information and can argue that this is a step in the right direction but it is very difficult to find pairs of celebs that co-occur and have nothing in common it doesn’t happen.

So from an experimental point of view this is difficult to do….at least with celebrities

Question 13

how can the problem of disentangling be overcome?

Answer 13

So instead can try and do it in a learning experiment

Question 14

learning experiment study of neuro-correlates, what did they do?

Answer 14

Wiese & Schweinberger, 2015:

• Participants learn pairs of pre-experimentally unfamiliar faces (5 training sessions spread across 2 weeks) in different conditions: 1. Co-occurrence/shared semantics, 2. Co-occurrence/no shared semantics, 3. No co-occurrence/shared semantics, 4. No co-occurrence/no shared semantics
• Sometimes the semantic information was identical for both of the pairs- these people co-occur and share a lot of semantic information
• Can also have two people who co-occur but share no semantic information.
• Can also have two people who never co-occur but have the same semantic information (they are never seen together)
• And can have condition where individuals never co-occur and share no information
• So in this paradigm can disentangle this information
• Post learning all of these conditions do a EEG experiment with priming
• Prime at 100ms and target at 1000 ms

Question 15

learning experiment study of neuro-correlates, what did they find?

Answer 15

Wiese and Schweinberger, 2015:

• Clear priming effect for co-occurrence/shared semantics relative to the unrelated condition.
• N400 priming effect for co-occurrence/shared semantics relative to the unrelated condition.

Question 16

what do the results from Wiese and Schweinberger, 2015 suggest

Answer 16

Nothing going on for only semantic information and only co-occurrence information. So having only co-occurance or only semantic does nothing so at least when we learn new faces we need both semantic and co-occurrence information.
 Both visual co-occurrence and semantic feature overlap structure semantic person memory.

Question 17

what can be said are weaknesses in the Weise and Schweinberger, 2015 study?

Answer 17

But can argue that what we have here is relatively weak new representations and that with more robust representations effects of priming would be bigger for co-occurance and semantics alone. More robust information it seems to be the case that semantics are more important (celeb stimuli) but we can’t be certain about this because is hard to disentangle

Question 18

what overall can we learn from disentangling studies

Answer 18

At learning new faces you need both co-occurrence and semantic and this means that how relations are organised in head at least when learning new people it is semantic info and co-occurrence that are both driving it.
Tells us something about how our semantic network is organised

Question 19

talk about areas in brain for individual info

Answer 19

• Neuropsychological studies suggest an important role of the anterior temporal lobe for semantic person memory.
• Right anterior temporal regions may be particularly important for person-related semantics.
• In the Haxby et al. (2000) model, information about the identities of familiar people forms part of the extended system.
• Biographical information is stored in anterior temporal regions.

Question 20

evidence for which areas in brain responsible

Answer 20

• Thompson et al. (2004):
• Patient J.P. showed severely impaired person-specific semantics, with relative preservation of knowledge about animals and objects.
• Patient M.A. showed the opposite pattern.
• J.P. had predominantly right temporal atrophy, M.A. had predominantly left temporal lesions.