Minds and Computers : An Introduction to the Philosophy of Artificial Intelligence

conditioned reflex arc.
Given these contemporaneous historical factors – the burgeoning
of psychology, the flourishing of positivism and the development of
Pavlov’s theory of reflex arcs – psychology was ripe for a paradigm
shift. This was brought about by the American psychologist John
Watson (1878–1958).
20
  

---

Watson was strongly in the grip of positivism and consequently
rejected the legitimacy of introspection as a psychological tool,
arguing that in order to be a positive science, psychology should
concern itself only with observable behaviour. In doing so, he recon-
ceptualised psychology – which had been seen as the science of
human consciousness – as the science of human behaviour.
Watson was interested in discovering the innate reflex arcs which
governed human behaviour and investigating the circumstances
under which reflex arcs could be conditioned in humans. To this end
he engaged in experiments on babies and young children which would
certainly never be approved by today’s ethics committees.
To demonstrate the presence of innate reflex arcs in young babies,
Watson showed that certain environmental conditions, such as
sudden loud noise or a sudden loss of support (i.e. dropping them),
would evoke fear behaviour in babies. Some might think these results
less than remarkable.
In investigating conditioned reflex arcs, Watson and his colleague
Rosalie Rayner experimented on Albert B., an eleven-month-old
infant. Albert would be presented with a white rat. When he reached
out in curiosity to touch the animal, the experimenters would make a
loud noise by striking a steel bar close to his head, thereby evoking
fear behaviour in Albert. After only seven such presentations of the
rat in conjunction with the loud noise – five of which were a week later
than the first two – it was found that Albert would exhibit fear behav-
iour on presentation of the rat alone. This conditioned response was
subsequently found to be enduring and, further, transferable to
similar stimuli. Albert would exhibit the same fear response on pre-
sentation of a white rabbit and, to a lesser extent, on presentation of
a dog, a fur coat or a Santa Claus mask.
Following Watson, psychological behaviourism became dominant
and held sway in psychology until roughly the late 1960s. During this
period it became customary to carry out experimental work on rats
where possible, beginning a tradition in psychology sometimes
described – endearingly or pejoratively – as occupied with ‘rats and
stats’.
By far the most influential psychological behaviourist after Watson
was another American, Burrhus Frederic (B. F.) Skinner (1904–90).
Skinner was interested in determining the most e
ffective means of
conditioning reflex arcs. He invented a device, now called a ‘Skinner
box’, in which rats could be placed. This box contained a lever which
the rats were able to depress which could be set to deliver a food pellet
when actuated. By varying the conditions under which the actuation

21

---

of the lever would yield a food pellet, Skinner was able to modify the
behaviour of the rats accordingly.
Unlike classical Pavlovian conditioning, this behavioural modifi-
cation was contingent not just on the stimulus preceding the behav-
iour (a ringing bell, a rat together with a loud noise) but also on the
environmental stimulus following the behaviour. Skinner called this
operant conditioning, and developed an associated theory of how best
to e
ffect operant conditioning by controlling the positive and negative
reinforcement of certain behaviours.
Skinner argued that such operant conditioning could be widely
employed as a social engineering technique. He suggested that crim-
inal behaviour was better treated with behavioural modification tech-
niques rather than punished through incarceration, and he published
a widely read novel, Walden Two, which outlined his utopian vision
of a planned society governed by operant conditioning. As one might
expect, these ideas were met with a mixture of commendation and
condemnation.
By the end of the 1960s, behaviourism in psychology had waned in
popularity, in favour of the newly emerging cognitive psychology.
There are good reasons for the loss of faith in the behaviourist con-
ception of psychology.
For one thing, it became clear that positivism is, simply put, a false
doctrine. Modern science is frequently in the business of theorising
about unobservable entities. Such treatment of unobservables does
not make theoretical physics, for instance, ‘pseudoscientific’.
More importantly, it became increasingly clear that many essential
aspects of mentality are simply not directly connected to observable
behaviour. The mechanisms which underpin memory, the capacity to
draw inferences and the ability to comprehend language do not seem
to be necessarily correlated with any particular kind of behaviour. As
such, these cognitive functions are not amenable to investigation in
the behaviourist tradition.
While it is the case that each of the three traditions of psychology
we have examined are now consigned to history, it is also the case
that modern psychology preserves certain techniques from each of
them. Where, however, each of these traditions were distinctly
methodologically homogenous, modern psychology is markedly
methodologically heterogeneous. There is still a place in psychology
for psychophysical research, self-reporting is still a widely used tech-
nique and there is still a role for the observation of behaviour. There
is no longer any place though for the methodological monism which
characterised these early periods of psychology.
22
  

---

3.5 PHILOSOPHICAL BEHAVIOURISM
Psychological behaviourism, as we have seen, is a methodological
view – a doctrine concerning the way in which one should go about
doing psychology. Philosophical behaviourism, in contrast, is an