Introduction Flashcards
history
behaviourism (1950s, 60s) - Skinner, Watson
cognitive science (1970s, 80s)
cognitive neuropsychology
cognitive neuroscience
behaviourism
Psychology should only concern itself with observable behaviour, which can be explained without recourse to internal mental events
Behaviourists focused on inputs and outputs
Inputs = external stimuli
Outputs = behaviour
Tried to establish rules governing the translation of inputs into outputs (i.e. learning)
cognitive science
Internal mental events are essential for explaining behaviour and can be characterised as the computational operations of a computer program
- Around time when computers popular
Chomsky: Language development cannot be explained in purely behavioural learning terms – too complex for this
Development of computers provided new technological metaphor
Mind as the software running on the hardware of the brain
The specific form of the hardware is irrelevant
Cognitive scientists studied behaviour (reaction times etc.) but inferred the operation of ‘mental modules’
- Often just inference
cognitive neuropsych
Behaviour can be explained by internal mental events and these events can be localised to discrete brain regions
Driven by observations of loss of specific functions after defined brain lesions, e.g. hippocampus lesions leading to memory problems
Double dissociation logic
- 2 patients
- Patient 1: Lesion to area A: Function X impaired but function Y spared
- Patient 2: Lesion to area B: Function Y impaired but function X spared
cognitive neuroscience
Driven by the development of new tools for measuring and manipulating brain function
At least in its early stages, adopted the logic of cognitive neuropsychology and combined it with the extra precision afforded by these new tools
functional Magnetic Resonance Imaging (fMRI)
Electroencephalography (EEG)
Transcranial Magnetic Stimulation (TMS)
Improved lesion mapping methods (sMRI)
Magnetoencephalography (MEG)
the search for the mind in the brain
Can we reduce all mental phenomena to physical (brain) processes?
Perhaps easy to accept that motor function or language might be reducible to neuronal function
But what about intelligence?
Do you think it is plausible that one day we will be able to describe the richness of flexible, intelligent human behaviour in terms of the firing of neurons?
the answer from cog neuroscience
Cognitive neuroscience doesn’t require you to accept that complex mental phenomena can be reduced to the firing of individual neurons
Cognitive neuroscience suggests there is an intermediate level of description at the level of neuronal systems
Systems do not necessarily have to be mapped to discrete brain regions (although they may be)
tools of cognitive neuroscience
computerised cognitive testing
neuropsychology
fMRI
EEG
MEG
TMS
psychophysiology
monkey single unit (neuron) recordings
computerised cog testing
Measuring reaction times (RTs) and/or accuracy in different conditions
Relies on subtraction logic: If you have 2 conditions that differ by a single process, the difference in RT/acc between the two conditions should reflect the operation of that process
- E.g. Condition 1 (active condition) – decide which of two letters is an X
- Condition 2 (control condition) – press when you see any letters
- Condition 1 – Condition 2 = selective attention
neuropsych - tools
Inferring the function of brain regions from the pattern of deficit when damaged
Originally double dissociation logic
More recent studies use large samples and lesion mapping techniques
Allows definitive causal statements about brain-behaviour relationships
Problem is the brain regions are so large that such statements sometimes quite imprecise
stroke
thrombus in carotid artery breaks off and travels to the cerebral artery in the brain
thrombus lodges in the cerebral artery causing a stroke
fMRI
Visualising functional brain activity during task or at rest
Actually measures Blood Oxygen Level Dependent signal (BOLD) – a proxy for neural activity
- Differences in magnetic properties of oxygenated and deoxygenated blood
fMRI adv
High spatial resolution (mm3)
fMRI disadv
Low temporal resolution (seconds)
univariate fMRI - single DV
More accurately “mass univariate”
Brain divided into cubes or “voxels”
‘Activation’ in each voxel is a dependent variable
Each voxel analysed independently of others
End up with a brain map showing which voxels are ‘activated’
multivariate fMRI
Brain again divided into voxels
However, this time voxels are not treated independently
Here, we examine patterns of activation across groups of voxels
Hence, sometimes referred to as multivoxel pattern analysis (MVPA)
Computer algorithm (pattern classifier) trained to learn the patterns of neural activation associated with different conditions
Then given a new data set and asked to predict which condition the subject is currently experiencing based on their neural activation patterns
Above chance classification suggests the brain region in question encodes information about the conditions
EEG
Recording electrical (neuronal) signals from the scalp
EEG adv
Very fine temporal precision (measurement every ms)
EEG disadv
Poor spatial resolution – declines as go deeper into the brain
2 ways EEG used
Event-Related Potentials (ERPs)
Examination of oscillations in different frequency bands (during task or at rest)
ERPs
Subject performs some task involving repeated trials of 1 or more conditions
EEG response to trials of each condition are averaged together to form an average waveform (an ERP)
E.g. P300 – ‘oddball’ signature - surprise
Examination of oscillations in different frequency bands (during task or at rest)
EEG recordings occur in rhythmic, repetitive activity patterns
These are described in terms of their frequencies (Hz)
E.g. Gamma, delta, theta, alpha, beta
Different roles for different frequency bands in cognitive processing – e.g. theta in working memory
E.g. synchronization across different brain regions – how they communicate
MEG
Records magnetic activity from scalp
MEG adv
High temporal precision (ms)
