Lecture 1: Introduction
The mind: The collection of all internal experiences, thoughts an abilities that an organism may have.
The set of experiences, thoughts and, abilities is infinite (breathing/blinking/loving)…
What are the mechanisms driving the mind? 2 ways (we need both):
- Cognitive route: define all brain functions and discover the relationships among those
functions (vision, feature detectors, mechanism to focus vision, oculomotor control…)
- Neuropsychological route: discover the neural circuitry driving brain functions (older
science)
How do we investigate the mind? We do not know what is going on inside, the black box problem.
(how the input leads to the output, the processing is hidden for us). Thee answer is to design and do
experiments we measure stimulus-response relationships to infer the mind’s operations. Really
controlling the input and inferring the type of response.
Cognitive science: try to divide the brain up into different functions. View the brain as an information
processor. Use experimental methods for measuring what’s going on. Donder’s subtraction method
Cognitive psychology assumes that mental processes exist, tries to construct models of internal
processing, examines observable behavior and uses o….. Cognitive psychology is inspired by the first
computers. We conceptualize the brain as an information processing system. How is information
carried and represented in the brain?
Looking inside the brain:
- fMRI: functional magnetic resonance imaging. Active areas of the brain demand oxygen.
Blood (and thus oxygen) travels to those area. Hemoglobin changes the magnetic properties
of an area. Using a very strong magnetic field, we can trace those areas. But the brain
presumable does many things at once. How would you know which brain area is responsible
for what? fMRI tells us where something happens, but not WHAT happens precisely and
what it does.
All information in the brain is carried in the form of electrical signals. Neurons are our “electrical
wires”. In each neuron, you have a cell body. Axon is the arm/leg of the neuron. Axon ends in the
synapse. Around the cell there are dendrites (gathering signals from surrounding neurons), the
receivers. Dendrites causes activity to build up. Activity builds up until action potential. Chemicals
are being released at the end of the synapse. Action potential is an all or nothing process (full
electrical activation or none). How active a neuron is, is dependent on the frequency (how many
action potentials).
The collection of pulses across neurons represents our world. A representation is a code for some
property of the external world (e.g. the color red). The number of neurons needed to code
something (a familiar person) is 1 neuron, however many neurons activated to activate this neuron.
Specificity vs distributed processing: 1 neuron for 1 thing vs a network of neurons to code
something. Specificity: disadvantage: can handle only number of objects of number of neurons and if
a neuron dies, you lose recognition. Distributed processing: each object is represented by a unique
combination of neurons. Advantage: higher capacity and if a neuron dies, the rest of the distributed
pattern still gives you a pretty good idea what the object is. Most of our cognitive functions rely on
distributed processing rather than single cells.
, Lecture 2: Perception
Perception: the process of interpreting sensations. For example: “something stinks”
Sensation: the registration of a physical stimulus by receptive neurons (e.g., activation of olfactory
bulb / activation of visual cortex). A physical, factual thing, not susceptible to interpretation.
There is a thin line between sensation and perception.
Goal of perception: interpreting, recognizing, understanding (‘what is it?’) and interacting with the
world (‘How to respond?’). The core challenge of perception is resolving ambiguity. The inverse
projection: from sensory processing alone, we cannot say anything conclusive about the world. To
resolve ambiguity, there are several theories about it:
- likelihood Principle (Helmholtz): perception corresponds to the most likely physical event
that could’ve caused the sensations
bottom-up vs top-down processing (crucial concept):
- bottom-up: sensory organs provide activation of ‘low’ cortical regions, cascades to ‘higher’
regions. Regions activate more regions
- top-down: ‘higher’ regions influence activation of ‘lower’ regions.
Sensation = bottom-up and perception is more a mixture. Bottom-up processing: everything that is
seemingly automatic, without the involvement of (conscious decision) and top-down processing is
everything else.
Bottom-up processing might not be just sensation. Grouping of local features into global structures
seems to proceed automatically. Gestalt principles: a set of assumptions about things that happen in
an automatic, bottom-up fashion, a mere product of the system’s architecture.
- Principles of similarity (things that do look similar, seem to be similar (color….)
- Principles of proximity (things close together seem to be similar)
- Principle of symmetry (expect things from the outside world to be similar)
- Principle of closure ()
- Principle of continuity (see something extending behind the first one)
- Principle of common fate (things that move in the same direction seem to be similar)
Questioned whether the gestalt principles are all really the result of bottom-up processes: probably
not. Our life experiences bolster the expectation that: similar-looking things belong together; objects
are most often symmetrical…
Top-down processing: this is how we typically conceptualize processing in the brain. Various levels of
processing in the brain and there is interaction among these levels. Word-superiority effect: a letter
is recognized faster if it is in a word than if it is in a non-sensical string. This is due to feedback from
the “word-level” to the “letter level”. Back in the days, the interaction among the different levels
was thought as to be hierarchically structured.
One way to frame the interaction between bottom-up and top-down processing: Bayesian
inference. The likelihood of interpreting something as X, depends on: (1) prior experience, which
shapes the probability of X occurring and (2) the extent to which new data aligns with out prior
experience. With new data, the collection of prior experiences is updated.