Cognitive Science
Atkinson & Shiffron 1986, maaan. I am back in Highschool.
A chunk is a set of items that can be remembered as a whole.
According to Miller a chunk is 7+-2 chunks. This was based on remembering chunks of unrelated numbers. It seems STM can hold more now a days.
Baddeley’s working memory model.
Apparently they have drawn localisation to different parts of the brain from his model.
Squire’s taxonomy of LTM
Memory can be declarative or non-declarative.
Declarative memory can be facts or events.
Non- declarative memory focuses on procedural memory, priming and perceptual learning, simple classical conditioning etc.
How can we know what is caused by what we know what is caused by our intuition.
Cognitive penetrability.
It is within our capacity to assign a different referent to a pronoun. So we can change the meaning of a sentence based on the context we have. They are caused by a difference in our background knowledge. This means that a sentence or thing is cognitively penetrable.
When a pattern is cognitively penetrable then it allows us to conclude that it does not arise from a property of the cognitive architecture.
Our expectations governs a lot of what we perceive. Cognitive penetrability in part is about what kind of expectations we can change.
What is a mental representation.
A critical part in the psychological explanation of many phenomena, but also a source of great confusion.
- A represented world
This is the domain that representations are about. This is about the world outside of the cognitive system. Everything outside of that can be a mental representation.
- A representing world
When we have a certain mental representation of eg. temperature, our focus is only on that thing which we represent. This is about the place of specific abstractions. In all known representational systems we lose information about the represented world in the representing world.
It is related to the represented world by a set of rules. Rules about mapping from one to the other.
Generally speaking there are two types of representations. There can be analog representations, like using logical rules. But they can also be numerical for example.
Mental representations are structures within the brain (Dan Williams 2018) that function like familiar external representations like map, but in a different form.
Computer simulations.
General problem solver, one of the first simulation models.
It could tacke problems such as logic tasks and simple games. By breaking problems down into goals and subgoals it could finish these.
It follows means-ends-analysis.
Crypt-arithmetic problem was one that it could sove, much like a person would.
DONALD + GERALD = ROBERT D=5
ACT-R model (Anderson, 2004)
Basically a model that puts together a lot of memory models.
Knowledge compilation can be about proceduralisation, you prepare inte production knowledge – you get used to a general rule related to solving problems to do it more efficiently, or it can be about composition, which is about cutting out unnecessary pieces of information in order to focus on what is really important.
Bergen är i födslovånda, en löjlig råtta kommer att födas.
Connectionist networks, computational methods to show information.
Distributed representations are quite efficient. They have 2__n different representations possible from a simple square in terms of how to solve a problem.
Cognitive scientists are focused on functions of the mind. The questions are about our mental software. It is strange to work on a computer however when only working with the software. You need to administer the hardware to get a software reaction.
Only focusing on the software is called functionalism.
This led to the trilevel hypothesis:
Intelligent systems are roganised at 3 distinct levels: physical, symbolic, and semantic.
Understanding how the brain works is probably related to this kind of schema.
Explaining the physical chips on the calculator is physical, explaining the information processing steps would be symbolic, whilst the semantic explanation may be the most abstract which basically prove the behaviour of the calculator through application.
The brain controls all reaction of the body and thinking. 20% of our blood goes to 2% of our body. The brain requires an enormous amount of oxygen to keep functioning. The brain is divided into 2 hemispheres connected by the corpus collosum. The outer layer is the cerebral cortex which has 4 main areas: frontal lobe, parietal lobe, occipital lobe and the temporal lobe.
Some stroke patients don’t stop seeing certain parts of their visual field but they do not give the thing attention.
Deeper areas have more primal functions related to them, and have therefore probably developed earlier in evolution.