Summary. This paper expands a new
hypothesis on the causal mechanisms underlying
irrelevant behaviour. It begins with a critical
summary of earlier theories which attempted to
explain displacement activities, but failed to
predict the consistency with which certain types
of behaviour are shown in stressful situations
by s variety of species. Behavioural and
physiological studies suggest that these
behaviour patterns are closely associated with
the incipient activation of sleep. The
functional significance of this link and some of
the causal processes which may be responsible
for it are discussed. Paradoxically, however,
displacement activities occur when animals are
in a state of high arousal. The concept of
arousal is reconsidered in the light of
information theory and assumed to be closely
correlated with the information processing rate
in the nervous system. The relationships between
neural and autonomic arousal are considered in
this context. It is argued that over-arousal may
occur when information handling exceeds the
limited channel capacity of the system, with a
consequent loss of efficiency. It is pointed out
that there are mechanisms capable of controlling
the information influx into the brain, and it is
hypothesized that they are tied up in a feedback
mechanism which regulates arousal and which
involves the activation of a de-arousal system,
correspdnding to the neurological sleep
mechanism. Displacement activities are viewed as
cànsequences of this regulatory
activation of the sleep system. This hypothesis
is then cornpared with existing theories of
displacement and its relationship with them is
discussed.
This evidence indicates that in a number of
species the behaviour patterns shown as
displacement activities, particularly grooming,
are also normally associated with inactivity,
drowsiness and sleep; indeed, the latter is
often itself a displacement response. Whether
this can be generalized to other species and to
other displacement behaviours remains uncertain,
because relevant information is usually lacking
as, for example, in the case of the fanning of
the stickleback (Sevenster; 1961). The idea that
individual behaviour responses are facilitated
by specific levels of arousal or degrees of
wakefulness is not new, and there is empirical
evidence to support it (Bindra, 1959a).
The origin of this association is fairly
obvious. Much of the animal's behaviour can be
assumed to have high priority, in the sense that
it cannot be postponed without impairing the
animal's or its progeny's surviva!: feeding,
fighting, courtship etc. These activities
typically require high degree of responsiveness
to environmental stimuli, with precise timing
and orientation, and often a marked level of
motor activity. In brief, animals performing
this type of behaviour can be said to be in a
state of enhanced wakefulness or arousal.
Other activities, such as grooming and
sleeping, are not so dependent on precise timing
for functional effectiveness; they have low
priority and, can be postponed until the high
priority activities have been carried out,
without markedly impairing survival. Such
behaviour is often characterized by an absence
of precise patterning dependent on external
stimulation; many appear to be centrally
"preprogrammed" motor coordinations that only
require triggering and involve restricted motor
activity. Hence they can be carried out in
states of diminished wakefulness. It is
therefore reasonable to assume that an
organizational pattern which causally linked
these activities has been selected for in
evolution.
There is uncertainty regarding the causal
processes underlying this association. Although
there is some contradiction in the relevant
studies, it may be that the secondary,
long-latency component of the cutaneously evoked
potentials at central levels, which seem to be
related to the occurrence of specific, more
complex types of behavioural responding, here
grooming, is of a larger amplitude during
drowsiness and the early stage of sleep than
during wakefulness or deep sleep. Whether other
mechanisms, such as activation of the grooming
motor coordination centres by the sleep system
play a role, must remain an open issue for the
time being.
A complementary mechanism may also
contribute to the link between sleep and
grooming. Roitbak (1960) and Pompeiano (1965)
have found that in the cat repetitive
stimulation of cutaneous afferent, is an
extremely effective procedure for
generating-sleep. Such type of stimulation does,
of course, arise during grooming and it is
conceivable that if grooming is triggered
perhaps by external stimulation, it could
through this mechanism tend to be followed by
sleep. The implication that the performance of
certain types of activities is conducive to the
induction of sleep or at least relaxation is not
a new idea; in man, for exampIe the performance
of a repetitive behaviour pattern (Oswald,
1962), lying down (postural facilitation, Lind,
1959) or closing the eyes (cut-off, Chance,
1962) is claimed to have such an effect. Here
then, the activation of grooming and perhaps
some other behaviour patterns would be the
primary response, and sleep or drowsiness the
secondary effect following it.
The nature of the connection of some of the
other behaviour patterns with sleep is less
obvious, particularly the staring down and
pecking of birds. Possibly it follows an
efferent suppression of the peripheral visual
fields by the sleep system. This then enhances
behaviour controlled by foveal vision (so called
tunnel vision), which in turn diverts the gaze
away from stimulating surroundings.
Arousal Reinterpreted
The De-arousal Hypothesis
At this point it is possible to return to
displacement activities. As we have seen, the
situations in which displacement activities
occur: conflict, frustration, thwarting and
novelty, are associated with signs of arousal in
the conventional sense; in other words, the
animals show signs of enhanced responsiveness,
marked activity and electroencephalographic as
well as autonomie signs of increased arousal.
Paradoxically, however, we have also seen that
some of the behaviour patterns shown in
displacement are associated, by various
criteria, with the incipient activation of
sleep. In the context of the foregoing this
contrast is highly suggestive of a regulatory
process, where the activation of sleep system
counteracts an increase in arousal.
The displacement situations can be assumed
to provide the animal with more information and
to demand from it more information proeing than
other situations. In conflict, it is exposed to
two contradictory assemblies of information,
between which it has to choose. In the
frustration and thwarting situation, an
established sequence of behaviour leads to
failure, and presumably the animal engages in
finding alternative solutions, a process which
must require intense information sampling and
processing. In the case of exposure to novel
stimuli, the increased information influx is
obvious. More precise data on this point are
needed, but as yet are difficult to come by,
because estimates of information processing
rates presume a reasonably good knowledge of the
processing modes used by the nervous system.
Accepting that the displacement situations
involve relatively high rates of information
hanriling, it seems likely that they could often
entail an overloading of the chaannel capacity
with a consequent deleterious effect on the
organisms efficiency. To counteract overloading
in such situations I suggest that reduction in
the processing rate is achieved by a concomitant
activation of an arousal inhibiting mechanism,
the sleep system, which has the property of
reducing the overall information influx into the
animal. Displacement activities are the
epiphenomenal consequence of this feedback
activation of sleep.
It is not possible to specify with any
certainty at this stage the detailed neural
organization of this regulating mechanism. I
suggest that the level of arousal (i. e. the
information processing rate) is monitored ither
via a polysensory system, like the reticular
formation, or through detection of an electrical
concomitant of a high level of processing, such
as potential changes, or perhaps by measuring
the level of some metabolic consequential
correlate of arousal. When the measure exceeds a
certain level in relation to the channel
capacity, it leads to the activation of the
sleep system or a related deactivating
mechanism.
There is empirical evidence that strong
activation of the reticular form tion leads to a
rebound activation of the sleep system which
then inhibits the reticular formation (Dell,
Hugelin and Bonvallet, 1961 Parmeggiani, 1968).
Whether this is a purely neural process or,
whetl possible sleep hormones (Monnier, Koller
and Hoesli, 1965; Pappenhieinaer Miller and
Goodrich, 1967) play a role, cannot be
decided.
The activation of the sleep system, either
directly or through th inhibition of the
reticular activating system, affects the
afference sensory information through the
centrifugal control mechanisms of se sory
pathways (Koella, 1966). I contend that this
leads to a reduction, of the attention span,
that is to an overall reduction of the
information inflow and hence of the information
processing rate (i. e. arousal), through
efferent inhibition of the primary, i. e. visual
and auditory, sensory systems. Consequently
there is a relative facilitation of the
secondary, i.e. cutaneous and perhaps olfactory,
systems and behaviour mediatsd by them, as
indicated earlier. Other responses may be
activated more directly by the dearousal system
at motor coordination levels, and some of the
activities so induced may in turn be conducive
to a reduction of arousal either by cutting down
sensory input or by generating repetitive
stimulation, facilitating the induction of sleep
and thereby reinforcing the original process
(Mason, 1967).
It is not clear why displacement activities
do not always include the complete gamut of
sleep-related behaviours (Ewer, 1967 b). Two
factors may be responsible: one is the magnitude
and time course of sleep activation involved,
which clearly will affect the range of
behaviours showN. A high dose of barbiturates
gives the animal only the chance of showing one
or two grooming movements before falling asleep,
while a low dose elicits a highly complex and
variable sequence of behaviour. Another
possibffity arises from the fact that the sleep
system cannot be viewed as a unitary center but
rather as complex network with more or less
differentiated subdivisions, as demonstrated,
for example, by the existence of separate though
related mechanisms for slow and fast sleep
(Jouvet, 1967). Which of these subdivisions are
involved in displacement, and how their
activation affects the behaviour of the waking
animal, must remain open.
Any of this behaviour may of course be
further modified by incidental adequate external
and internal stimulation. For example,
displacement grooming might be facifitated by
suitable cutaneous stimulation or a pre-existing
partial activation of the sleep system, while
displacement pecking would increase with hunger
and the presence of grains. A consequence of
this is that a clear-cut distinction between
normal and dislacement occurrences of given
behaviour patterns is not possible, in eement
with most observational data; this may have to
do with the oulty of formulating a precise
definition of displacement.
Although it is proposed that displacement
activities are the reflection a de-arousal
process, they do occur in a context of high
arousal, one may expect that some of the
behaviours shown in these situations are more
direct consequences of the heightened arousal.
Wood- Gush Guiton (1967), for example, found
that chickens would often react th frantic
escape attempts to a thwarting procedure which
regularly yielded displacement grooming and
sleep. Similar escape behaviour reliably be
obtained upon electrical stimulation of the
reticular activating system in gulls (Delius, in
prep.), apparently as a direct effect of high
arousal. Pigeons show similar escape behaviour
in novelty situations, and EEG recordings
support the view that this reflects a state of
extreme arousal (own observations). Similarly,
if the animal is informationally challenged when
it is operating below its maximum channel
capacity, may be assumed that the system's first
response will be to increase its attention span.
Some of the alerting or orienting patterns seen
in stress situations may therefore reflect this
process (Berlyne, 1967). As we have argued,
there are reasons for assuming an overall
correlation between central and autonomic
arousal and so we may expect that displacement
situation will also give rise to autonomic
responses. Schmidt 1950) provides good examples
of this in dogs exposed to frustrative
ttiations, with panting, salivation, defaecation
and urination occurring long with displacement
grooming and sleep. Similar behaviour appears to
be recorded in the open field test for
emotionality in rats and mice van Toiler, pers.
com.).
Finally one would expect the proposed
homeostatic process to fail occasions, resulting
in over-arousal to the extent that the organism
comes behaviourally ineffective. This is
reflected in the so-called freez state many
animals fail into when facing extreme stressful
situations -d which may often be clearly
unadaptive, as when stoats simply grab zing
rabbits (own obervations) or lions kill
apparently paralyzed - debeests or zebras (H.
Kruuk, pers. corn.). Even so, many, species seem
have managed to evolve an adaptive response out
of this freezing beviour: concealment freezing,
death feigning, and so on (Hmnde, 1960).
Comparing the de-arousal hypothesis with
other theories of displacelent, its main virtue
is that it explains the specificity and
consistency of the behaviour shown in
displacement situations and moreover, it
provides a logical basis for their relatedness
to sleep. It is consistent th the fact that
displacement occurs in contexts of high arousal
and that the level of arousal modulates
irrelevant be haviour (Fentress, 1968). It is
concordant with the disinhibition hypothesis
(van Lersel and Bol, 1953), in admitting that
peripheral stimulation can play a role in
determining the displacement behaviour, but has
no, difficulty in explaining its occurrence in
frustration and thwarting situations. It also
accepts that shifts of attention (McFarland,
1966a) are] involved in bringing about
displacement activities. The functions of dis:
placement activities in the context of the
de-arousal hypothesis are consistent with those
suggested by Tinbergen (1952) and Chance (1962)i
avoidance of disruption of neural processes
through a reduction arousal.
When assessing the value of the
various-theories, consideration must be given to
the possibility that the displacement phenomena
are hetero-geneous. It may be, for example, that
the disinhibition hypothesis is satisfactory
explanation of the displacement fanning of
sticklebacks, but that the de-arousal hypothesis
gives a more adequate account of the causal
processes underlying dis placement grooming in
gulls. The fact that grooming is the dominan
displacement pattern in a wide range of species
from insects to man, how: ever, suggests that at
least this type of displacement should be ex
plainable in terms of a single universal
hypothesis.
To close, an alternative line of
explanation, perhaps related to th de-arousal
hypothesis, should be mentioned. As indicated
earlier, the sit t, in which displacement
activities occur can be characterized stressful.
Adrenocorticotrophic hormone injected into the
brain ventricles of cats elicits yawning and
stretching behaviour (Gessa, Pisano, Vargi
Crabai and Ferrari, 1967). These responses are
sometimes seen in conjunction with the more
common displacement activities of the cat. A
secretion is known to follow the induction of
psychological stress an takes place in the
adenohypophysis, but probably also at some
unidentified ventricular secretory organ
(Frankel, Graber and Nalbandov, 1967 Among the
target organs of ACTH is the hippocampus (aus
der Mühle and Ockenfels, 1968), which is
involved in deactivation as mention earlier and
seems in turn to control ACTH secretion (Mason,
Nauth Brady, Robinson and Sacher, 1961/62).
Possibly, displacement acitivities are just one
component of the general adaptation syndrome
(Selye,1959 the organism's response to
stress.
