Introduction / Chapter 1
Learning: biological processes that facilitate adaptation to one’s environment
What causes behaviour? Something that makes you respond (always a stimulus,
whether you know it or not), and something that has value (motivational value -
individual differences and dependent on context, must motivate you to behave).
We have motivational value because we have previous encounters with the
stimuli and have memory for it - memories are a product of learning!
This occurs because we have a nervous system that responds to stimuli by
looking back into past experience with the stimuli, taking into account how
valuable the stimuli is at that moment, and behaving accordingly.
Rene Descartes introduced the concept that your brain is involved in making you
behave. In a sense he was a dualist meaning he believed that the mind and the
physical world influence behaviour. Involuntary behaviour occurs by taking stimuli
in through the senses, transmitting this to the brain, which then decides how the
muscles should respond. Involuntary actions were the only ones believed to be
shared by animals as well. Voluntary actions were unique to humans and
occurred by taking in stimuli through the environment, sending the information to
the brain, which was then linked to the mind through the pineal gland. The mind
then sent responses through the pineal gland to the brain and then to the
muscles. This view of learning was important because it emphasized that stimuli
must exist for behaviour and that the brain is also involved.
Stimuli are detectable changes in the environment that can either be neutral (no
motivational value), appetitive (making you approach stimulus) or aversive
(making you back away from the stimulus).
Responses are a quantifiable reaction to the stimulus and can be either learned
(memory) or unlearned (innate, genetic predisposition). Within these two
categories, responses can be somatic (observable responses) or autonomic (e.g.
changes in heart rate, temperature etc.).
Rene Descartes also believed that we were born with all of the knowledge we
have (nativism) while John Locke believed that we are born as blank slates and
that we acquire knowledge through our experiences (empiricism). Turns out both
are right depending on the behaviour.
Hobbes believed in the distinction between involuntary and voluntary responses
(like Descartes) but believed that the mind operated in a lawful and predictable
way (unlike Descartes). He suggested hedonism; where people behave in a way
to seek pleasure and avoid pain.
Association (proposed by the empiricists) is the fact that simple sensations are
combined into more complex ideas and these associations are the building blocks of learning. There are primary rules to associations (contiguity -
repeatedly occurring together will be associated together, similarity and contrast)
and there are secondary rules (there are a number of factors that influence
formation of associations - frequency, recency, intensity etc.).
Ebbinghaus used nonsense syllables to show that strength of associations’
increases with training and that order effects associations. Bell showed that there
are separate nerves for sensory and motor information. Swammerdam showed
that it takes more than mechanical irritation of a nerve to make a muscle
Watson was the father of behaviouralism, which stated that things that cannot be
observed should not become data for science. He said that if you know the laws
and you know the stimuli, you should be able to predict the behaviour of the
individual. Essentially, you are the victim of your experiences. Watson conducted
the famous “Little Albert” experiment in which a mouse (CS) was paired with a
loud sound (US) which then conditioned the baby to be afraid of the mouse and
expanded to anything that had fur.
Darwin proposed that there was nothing special about humans. Since we have
evolved from animals it is hard to pin point where the “mind” developed and thus
we can either say that animals must have minds too or that no one has minds.
Romanes suggested that intelligence can be determined by the ability to learn
(whether you make new adjustments or you modify your old behaviours).
Sechenov proposed that stimuli do not always elicit a reflexive response directly.
Nervism was introduced in the mid 1800s and stated that all physiological
functions are governed by the nervous system and that there are biological
mechanisms behind our behaviour. Pavlov used this approach when studying
how reflexes could be shaped by experience. He found that new reflexes could
be formed through association (classical conditioning). Classical conditioning is
an S-S* association or an association between a stimulus or signal and a
biologically significant stimulus such as food, water, drugs, sex or social
interaction which produce a biological response whether you like it or not.
Classical conditioning fails to tell us why people engage in particular behaviour.
In conditioning trials we see responding increase (learning curve, pairing CS with
US) and then in extinction trials we see responding progressively decrease (CS
and no US). What is interesting is that responding can come back and it is the
basis of relapse. Why does responding come back? Drug associated cues (e.g.
being in the same context), the drug itself (e.g. the potato chip effect or “the one
that doesn’t eat the chocolate has the problem”) and stress (chemical, physical
etc.). Dollard and Millar believed that research on nonhumans could provide
information that better helps us understand human behaviour (less expensive,
simpler, more easily controlled).
B.F. Skinner was a radical behaviouralist and suggested operant or instrumental
conditioning (S-R learning). He said that a signal elicits a response that is then
reinforced (positively or negatively) through the presence or absence of the
biologically significant stimulus. This presence or absence will then affect future
behaviour upon presence of original signal.
Performance is the action of an organism at a particular time and is often used as
an indicator or evidence of learning. There are other sources of behavioural
change such as fatigue (physical exertion that results in gradual reduction in
intensity), changes in stimulus conditions (lights go down, everyone goes silent in
movie theatre), maturation (cannot reach cookie jar until tall enough), alterations
in motivational or physiological state of the organism. Learning can be analyzed
at the level of the whole organism (behavioural), neural circuits and
neurotransmitters (neural system or network) and neurons / synapses
(molecular, cellular, genetic). There are 3 alternatives to research with animals
(refine research so there is less suffering, replace animals with other means or
reduce the number of animals used in statistical techniques.
Elicited behaviour is like a reflex. Reflexes consist of 2 or 3 neurons making up
the reflexive arc. Reflexes make up the majority of behaviours in newborns
(withdrawal reflex, grasping reflex, occlusion reflex). Goal-directed behaviours
are a set of behaviours that are directed towards something the organism is
trying to achieve. They can either be instincts (genetically programed, resistant to
change, require little to no learning, has stereotypical nature to them) or learned
(require learning, flexible and adaptive).
The psychological theory of instincts (William James) says that instincts are
motivators of behaviour, they are the “fuel” to make someone behave and this
fuel comes from our genetic background. James’ theory says that there are an
infinite number of instincts which thus makes it difficult to verify and almost
untestable but what we can take away from it is that there needs to be energy
that makes you behave.
The ethology theory of instincts (Lornez, Tinbergen) says that instincts are
behaviour and they differ in their degree of sensitivity to change in the
environment. The behaviours can either be appetitive or consummatory.
Appetitive behaviours are learned, flexible and open to modification while
consummatory are genetically determined, require little to no learning and are
insensitive to changes in the environment. Consummatory responses are also
known as fixed action patterns. An example of a consummatory behaviour would
be chewing and swallowing. This is a behaviour that occurs in the same way almost every time. An example of appetitive behaviour knows there is food under
a cup and pushing over the cup to get to the food. This theory states that all
behaviours need a stimulus and energy.
Key or sign stimuli are stimuli that produce a fixed action pattern or
consummatory response. Social releasers are a type of sign stimuli in which
behaviour from one individual evokes a response in another member of the
species, for example, yawning.
Homeostasis is the tendency to maintain balance or internal equilibrium. Our
bodies have a variety of set points and if there is a disturbance to any of them,
will fight in the opposite direction (opponent response) to get back to equilibrium.
This is the basis of “buy happiness, get sadness for free.” The primary process
(effect of the drug) creates a disturbance in the body which in turn, makes the
body mount an opponent process that generates the opposite emotional reaction.
There is an important balance. Imaging there is a tank that is filled with fluid that
is held in by a valve. This valve is also attached to a scale in which if weight was
added, the valve would slide open. The opening of the valve would allow the fluid
in the tank to fill up 1 to 6 drippers depending on how much fluid is allowed
through. There are then two ways to open the valve, one is if there is so much
fluid in the tank, the mere pressure opens the valve and the other is if there is
enough weight on the scale to push the valve open. For example say we take
hunger. Over the course of the day you get hungrier and hungrier, thus adding
more and more fluid to the tank. You will reach a point in which you are SO
hungry that you eat which means the pressure has opened the valve. Now, say
you are full, there is very little pressure on the valve due to fluid in the tank, but
you are presented with your favourite food, then the weight on the scale
increases and you are likely to eat it. The weight on the scale is thus the “perfect
Drive is a motivational construct associated with the maintenance of homeostatic
balance. When homeostatic balance is disrupted it creates a need. This need
creates a drive, energy to want to reduce the need. This energy is converted into
behaviour to reduce the need, which in turn, reduces the drive and restores
homeostasis. We can study drive by changing the amount of deprivation in rats.
We have the rats in a chamber, with the goal in another chamber, separated by a
bridge that gives them foot shocks - thus creating a conflict situation. We are
looking to see how many crossings the rats make for exploration, water, food and
sex as a measure of the strength of drive. We see that after one crossing, the
number of crossings for exploration goes down (seen it once, I know what is over
there). It is hard to deprive a rat of sex but there is a maximum amount of
crossings and then it stabilizes. For both food and water, as the amount of
deprivation goes up, so does the number of crossings, however it is a peak
function which probably reflects the decrease in energy after not having ate or
drank. Hull’s drive theory uses the equation sEr = D x sHr to predict when an animal will
do a behaviour, where sEr is the strength of the behaviour, D is the strength of
the drive and sHr is the strength of the habit (something that has been learned
and becomes stronger through how often it is followed by a satisfying event -
reduction of drive is satisfying event). In our experiment we manipulated the
number of reinforced repetitions (strength of habit) and measured the number of
reactions to extinction. From here we had two groups, one that was not deprived
for long and another that was. We see that both groups produced same curve but
the increase in drive pushed the curve upwards. If you have a strong habit but no
drive, you will not produce behaviour and likewise if you have a strong drive but
no habit (have no idea what to do), you will not produce behaviour. Therefore
drive is necessary to learn something and when you encounter that drive in the
future you are likely to take part in the same behaviours to reduce the need. Hull
said that drive is a general pool of energy and therefore there is only one type of
drive however when drive is activated, stimulus-drive (sd) is activated in parallel
which is drive dependent which then directs previous habits and performance of
specific responses. Hull eventually realized that the characteristics of the goal
object also influence the motivation of the organism and added the incentive
value (k) to his equation. Incentive value is learned, relative to something else
and relative to the motivational state. Rats will run faster when goal goes from 1
pellet to 16 pellets compared to 16 to 16 or 32 to 16.
There is an optimal level of motivation and it is dependent on the task. The
Yerkes Dodson law states that there is an inverse relationship between task
difficulty and optimum level of motivation - the easier the task, the higher the
level of optimal motivation. Drive reduction is also not necessary for learning like
Hull suggests. Latent learning is learning that occurs but does not show itself
until the circumstances have it. We see this through rats that are put in a maze
with no goal at the end. Then we take these rats and new rats that have never
seen the maze before and see who gets to the goal first. The rats that had been
in the maze before get to the goal faster thus proving that they were learning
before but the learning was not expressed because there was no goal.
Habituation is the decrease is responding due to repeated exposures, e.g.
clothes, smell etc. If you want people to keep eating, each bite should be a
different taste so they don’t get habituated. The opposite is the case for if you
want to lose weight - it is better if all bites taste the same. Habituation takes
longer if we are not paying attention and according to the other race effect, we
are better at discriminating a face if the individual has small differences from a
different race than us. Habituation is both stimulus specific and response
specific. Sensitization is when you have an incremental increase in responding
due to repeated exposures (one experience does something to your nervous
system such that when you experience a smaller version of that you have the
same response as the original, major experience. Sensitization is not highly
stimulus specific. Modification of calcium intake in neurons directly impacts the amount of
neurotransmitter released. The magnitude of neurotransmitter release influences
the magnitude of the response.
Take slug and implant 2 electrodes, one on sensory neuron and the other on
motor neuron. Then you do habituation training and see that the amount of
sensory information (input) is the same but the amount of neurotransmitter
release to the motor neuron has decreased. In the sensitization experiment we
shock the tail of the slug. Since sensitization is not stimulus specific, all
responses of slug are amplified. Now an interneuron alters the activity of the
sensory neuron causing it to release more neurotransmitter (serotonin) to the
motor neuron and activating it to a greater extent. One shock causes an effect
that lasts a few minutes. Five shocks cause a growth of new synapses and a
longer lasting effect and this is the basis of PTSD.
Chapter 3 and 4
US is unconditioned stimulus (food, water, sex, drugs, social interaction, had very
little experience with to determine it’s value) and UR is unconditioned response
(salivation, increase in attention, response that just happens, can be a reflex or
more complex than that). CS is conditioned stimulus that will then be used as a
predictor of the US and the CR, conditioned response, is the response that
accompanies the CS.
We measure CR’s by presenting the CS alone in a test trial (without the bias of
the US) and from this we can measure the latency (how soon), the magnitude
(how much) and the probability of responding (how often).
Conditioned excitation is when the CS is a predictor of the presence of the US
(CS+) and conditioned inhibition is when the CS is a predictor of the absence of
the US (CS-). There are two main ways to get conditioned inhibition. The first is
differential inhibition procedure, in which CS1 is paired with US and CS2 is
paired with no US and therefore is inhibitory. Secondly you can do a conditioned
inhibition procedure in which CS1 is paired with the US and then CS1 & CS2 are
paired with no US in which case, CS2 is a conditioned inhibitor (learn that CS1 is
useless when CS2 is present). Conditioned inhibition can be tested by a
summation test, when the conditioned inhibitor is paired with a new CS to see if it
elicits a response. If a response occurs, the conditioned inhibitor was actually not
a conditioned inhibitor but if a response does not occur, it was. This can also be
tested using the retardation of acquisition test in which we try and make the
conditioned inhibitor into a conditioned excitor. If it was a conditioned inhibitor, it
should take more trials to turn it into a conditioned excitor than if it was not a
There are three criticisms to the S-S* learning model. One is that we are not
making a response to the CS, instead the animal is salivating, for example, because of mere exposure to the food so many times. This concept is called
pseudo conditioning and we can test this by using a control group that is only
given food and see if the control group starts to show the conditioned response
to the CS - which they do not. The second criticism is that there is sensitization to
the CS and thus we get salivation because we have increased the sensitivity to
the CS. We test this by having a control group that is only presented the CS and
then tested to see if they salivate to it and the answer is they do not. The last
criticism is that instead of creating a CS-US association, that we are creating a
CS-Response association (by passing the US completely). We test for this in 2
different ways. The first was is through US devaluation or post-conditioning
devaluation. In these cases we look to see if the US association is important in
making the response. Rescorla paired the CS and the US together and then in
one group just exposed the US and this group did not show a response when
presented with the CS. A similar experiment was conducted but the US was
instead paired with sickness in which the group also did not respond when just
the CS was presented. The second line of evidence is CS-CS associations. One
of these associations is called second order conditioning in which a CS is paired
with a US to the point that the CS elicits the response. Then the CS is paired with
a second CS with no US. We then see that the second CS also elicits the same
response as the first CS thus transferring it’s conditioning power to another
stimuli. An ex