Liber, Montréal, 1995
The Quest for Pleasure
Table of contents
1 My Damascus Road
2 From Physiology to Experimental Psychology
3 'Do-it-yourself' can be Advantageous
4 Pleasure is Adapted: the Laws of Temperature Regulation Apply to
Taste and Odor Pleasure
5 Alliesthesia: Pleasurable Means Useful
6 Ponderostat and Generalization
7 The Behavioural Final Common Path
8 Psychological Verifications
9 From Experimental Psychology to Micro-economics
10 Pleasure Optimizes Behavior
11 Happiness and Joy
12 The Step into Morals
This little book describes a research on pleasure, taking place over the last thirty years.
The pleasure under study was not a specific pleasure, such as sexual pleasure or the
pleasure of playing a game, but was rather pleasure in general as a motivation of our
behavior. When the time to find a title came, I first thought of Ergo in order to underline
the process taking place in mind when, examining the premises of a problem, the
researcher is driven to an irresistible conclusion. Then I thought of The Obsession of
Pleasure, to indicate the omnipresent fixation on the problem under study, until the
liberating 'Eureka'! But the researcher's "unsatiable curiosity", like that of the Elephant
child, is neither contemplative nor sad. In the experimental sciences the passage à l'acte is
the poursuit of successive answers to never ending series of questions. This dynamic
and joyous process is evoqued by the present and final title, The Quest for Pleasure.
Acording to Descartes, one can read the great book of Nature following two ways. The
first way consists in analyzing more and more finelly the mechanisms that operate the
outside world and our own body. This path is the most congested in our modern time
where technology produces more and more performant instruments which permit to
better and better probe the infinite small and the infinite great. The second way is
opposite and goes from simple to complex. In that case research starts with "the
simplest objects which are the easiest to recognize; then research continues with more
complex objects of study and step by step, climbs to aim at the understanding of the
most complex ones (Descartes R. Discours de la méthode, 1644). The following pages will
deliberately take the second venue. Contrary to what many think, science is not
Let us recall some commonplace knowledge. Nature is complex and new properties
emerge from complexity. With the step from matter to life appear in the living beings
the properties that cannot be predicted and explained from the properties of matter.
Living beings tend to maximize their energy content and they reproduce. This does not
exist in the world of matter. All the laws of matter continue to apply to living systems
but one cannot deduce the properties of living beings and the laws of life from the
properties of matter and the laws of physics and chemistry. Similarly, new properties
emerge with the step from living organisms to thinking organisms. By definition
thinking beings think. The mental properties of the brain and the laws of psychology
cannot be predicted and explained from the properties of living beings and the laws of
life. Yet, all the laws of life continue to apply to the brain, siege of thinking. A third
step is the passage to collectivities. New properties emerge in societies and the laws of
sociology cannot be predicted from the laws of psychology.
Yet, although the various sciences differ from one another by their fields of
investigations, their levels of complexity, and their technologies they possess a common
method. Whatever the science the scientific method is the cement which unites the
products of our curiosities when we explore nature. This common cement consists in
sharing evidence and describing the laws of nature.
It should not be impossible, therefore, to reunite all the sciences into a coherent sum,
from matter to whole mankind society, from quark to Teilhard de Chardin's noosphère.
It must be possible to follow a logical continuum and cross from one science to another.
Biologists have long accepted to cross from matter to life. Psychologists have started a
similar thought process from life to thinking1. The process remains scientific as long as
observable facts can be identified and studied.
The present book does not pretend to propose a general unifying theory of the sciences.
It will simply describe the itinerary of one scientist whose research led him by chance
through several of the recognized sciences. Following this itinerary will lead us from
the simple to the complex. Our starting point will be my thesis research for the degree
of Doctor in Medicine. Step by step we shall turn from physiology to physiological
psychology, then to psychology, then to micro-economics. Such an itinerary was not
anticipated, neither on the starting block nor during the race, but one teaching received
by the runner was that the Ariane thread of scientific research always leads to surprises.
Actually, one of the greatest surprises was that, what appeared to me as an obvious
phenomenon, the feeling and the role of pleasure, in the sixties was treated neither in
physiology nor in psychology texbooks and that a huge aspect of our daily actual
experience remained terra incognita for science. Freud had admited intuitively, yet in a
rather abstruse way 2, that pleasure is important. Further, Freud's methodology was
non-scientific, since, by definition, psychoanalysis defies objectivity. Scientific
psychology responded to psychoanalysis with behaviorism. According to the
behaviorist trend of scientific psychology, science must study only 'public data', i.e.
observable and measurable facts. Under their influence psychology became the science
of behavior. As a result, a whole generation threw the baby with the bath water and
scotomized all 'private data', i.e. all mental facts. Pleasure, the motor of our behavior,
was judged as an epiphenomenon, deserving no study, by the Behaviorists. Yet,
pleasure, private datum, is an objective facts for those who experience it. There is no
reason why the objects taking place within the mind, mental facts, should not be studied
scientifically. All it takes is to proceed objectivelly and not to satisfy oneself with
1(Cosmides L., Tooby J., BarkowJ. H., "Introduction: evolutionary psychology and conceptual
integration", J. H. Barkow, L. Cosmides, J. Tooby eds, The Adapted Mind, New York, Oxford University
Press, 1992, 3-15.)
2 "what is pleasure for the inconscious is displeasure for the conscious"; see Safouan M. L'échec du principe
de plaisir, Paris, Éditions du Seuil, 1979.
intuitions like the psychoanalysts. What makes science is the method, not the field of
study. At this point it is necessary to recall the difference between awareness and
knowledge. Everybody before Archimede was aware that one can float on water, and
before Newton that apples fall from appletrees. Everybody is aware the a sensation can
be pleasant. Such an awareness is not synonymous with knowledge of the laws of
physics and, more modestly, of those of sensory psychology.
Another surprise was that the laws of this psychology of pleasure, once recognized were
not accepted by my contemporaries. "Nothing is more difficult to study than the
obvious"3. The psychologists L. Cosmides and J. Tooby, founders of a new discipline of
modern psychology, Evolutionary Psychology, describe with humor the three steps
followed by many scientists when they encounter new ideas 4:
Step 1 "It is not true", then
step 2 "All right, it may be true but it is not important," then
step 3 "It's true and important, but we have always known it."
Therefore this book is not a texbook, extensively treating all aspects of pleasure, but
simply describes the solitary itinerary of a research against the main stream. In order to
share my conviction, I tried to describe the inner logic leading step by step research.
Rather than a book of science, this is a book on science, a peculiar science. This is why
the bibliographical references are reduced to a minimum, just sufficient to provide the
chronological markers and to reassure the reader on the validity of the arguments
presented5. Interested readers will find there, in addition to the original data, all the
discussions which took place at that time and the complete literature lists. The sequence
of the first eleven chapters describes as many series of experiments which will perhaps
be indigest to the lay reader. This sequence is nevertheless necessary to provide the
minimal scientific basis for the final chapter with reasonning and hypotheses which
would be, without them, only intuitive. Finally, I hope that the reader will share my
pleasure in this quest.
3 Tenney S. M. "For all is but a woven web of guesses", News in physiological science , 1993, 8: 51-53.
4 Cosmides L., Tooby J. "From evolution to adaptations to behavior. Toward an integrated evolutionary
psychology", Wong R. Biological Perspectives on Motivated Activities, Norwood New Jersey, Ablex
Publishing Co., p. 13.
5Bibliography is the sum of all the works published before and related the article a scientist is being
writing. It is of fundamental importance that an author does not miss predecessors when weaving the
fabric of science.
MY DAMASCUS ROAD
"There must exist a sensitivity to local cold in the hypothalamus, and here is why. I
shall summarize our steps, and explain why we should expect to find there a cold-
Professor Joseph Chatonnet is speaking. We are in 1959 in his physiology laboratory in
the Faculty of medicine of Lyon. I am here to complete my medical studies. I wanted to
become an anthropologist, but for that I would have had to go to Paris; being from
Grenoble, I was not really inclined to go to Paris. Hence, I chose instead to go to Lyon, a
neighboring town, proud of its Physiology department, and I began working on my
thesis. Research is not just having ideas or asking questions. These ideas and questions
must fit in the accepted scientific construction. The first contact with research brings in
the traditional teaching, in which the Socratic relations between thesis supervisor and
his disciple could be described more as companion-like, than any other teaching.
Between supervisor and graduate student the dialog must be continuous. A thesis starts
with the questions the supervisor asks, not the student. This explains why the value of a
school resides in the value of its teachers, not in that of its students, contrary to what
Joseph Chatonnet has accepted me as his student and, now, he is explaining the
research program that he offers me as a thesis subject. While he speaks, an experiments
is going on nearby. A dog is put in a cold room and its oxygen consumption is being
recorded. The dog looks through the window straight at us, its tail wagging from left to
right. My supervisor is a specialist in temperature regulation, a function ill known of
the public, but extremely important as it is what permits the constant balance of body
How could I have guessed that the following five minutes were going to change
my life in such a dramatic way that my whole career orientation would be so brutally
shattered by the slow, well versed explanations of my teacher who would pull me out of
anthropology, where I always thought I wanted to go and in which I had oriented my
studies, and shove me hard into experimental studies in the field of physiology.
Now, Joseph Chatonnet presents the question he is asking to himself, the
question that we shall try to answer in my thesis:
"We know that shivering is the main mean the body uses to combat the cooling down
of its temperature. Shivering burns energy without producing mechanical work, hence
making the energy consumed by the shivering to be released as heat. We are
interested in the signals that trigger shivering. In 1893, from indirect observations, the
French physiologist Charles Richet proposed that there may be two types of shivering:
central shivering and superficial shivering, triggered respectively by a drop in the deep
body temperature, and a drop of skin temperature. A drop of skin temperature would
cause what Richet called the "superficial shivering", whereas the drop of deep body
temperature would cause the "central shivering". This "two-shivering" notion was later
abandonned by other physiologists for the apparently following good reasons. While
the sensitivity of the skin to temperature is obvious, a sensitivity to temperature deep
in the inner core of the body does not arise a sensation. In addition, the inner body is
always hot and the nervous structures, thus protected against cooling, would have no
way of being stimulated by an outer drop in temperature. The fight against cold is
generated exclusively by the drop of skin temperature. This is the present conception
in all physiology textbooks."
After this exposé of the leading theory, Joseph Chatonnet then goes on with his
own contradictory theory:
"I think Charles Richet was on to something and that there exists a sensitivity to cold
in both the skin, the surface of the body and the internal parts of our body. With my
collaborator Maurice Tranche, we have used the preparation of a spinalcordless dog
as described by Doyen Hermann for the study of the regulation of arterial pressure.
This preparation has served us well to understand the circulatory function but also to
study temperature regulation. A spinalcordless animal is completely paralyzed since
its muscles do not receive any orders from the central nervous system. Its skin is
also completely anesthetized since all the messages to the brain travel through the
spinal cord that has been destroyed in these animals. Only the head and the neck
stay innnervated, meaning sensible and mobile. Incidentally, they suffer of no pain,
since all their pain affenrences have been supressed with the spinal cord. Now, if we
place around and on top of the head of such an animal a heating craddle similar to
those used in hospital to warm the patients after surgery, we thus neutralize all
possible cold inputs to the brain which cannot receive any cold stimulus from any part
of the body. Then, if we cool the deep layers of the animal and thus lower its internal
temperature, e.g. by letting it drink cold water, or by placing a pack of ice on its belly
and rear legs, we see shivering take place in the neck and head of the animal. Yet,
no sensory message can originate from the head, nor from the rest of the skin since
the animal is spineless. Hence, this shivering cannot be a superficial shivering,
triggered by a cooling of the skin. It must be a central shivering, triggered by the
cooling of the internal parts of the body."
This research for my thesis is my first real contact with the thought process in
biology. Our Logic teacher in terminal classes of high school, always insisted on the
necessity for science to find an evidence that can be shared before any conclusion may
be drawn. Later, in the school of medicine, physiology was taught by true successors of
Claude Bernard and nothing was to be taken for granted unless proven by
experimentation. However, there was a huge difference between the theoretical
teachings received ex cathedra, and the dazzling of this first confrontation with the
unknown, and of the first "therefore" which is the result of previous experimental results
and a lot of thinking.. The "therefore" places us in another viewpoint, from which we
shall be able to look at the situation with a new eye. In the case of temperature
regulation, following Joseph Chatonnet, we now take the deep body sensitivity to cold
as highly likely, not as a supposition. My superviser now presents the plans for an
experiment to prove his hypothesis, and decribes its expected results:
"As a first experiment for your thesis, I would like you to find where in the body this
internal cold sensitivity resides. I can already tell you it is most likely to lie in the
brain, somewhere near the hypothalamus. Why? Because the American
neurophysiologist Magoun, has surgically implanted warming electrodes in the
hypothalamus of a dog and the local rise in temperature, created by the passing of
high frequency current, was followed by panting, a reaction of fight against heat, in
that case the change in hypothalamic temperature. Hence, those nervous structures
have something to do with the fight against temperature change. Barbour, another
American physiologist, has implanted a probe in the brain by which he flowed cold
water, thus lowering the temperature of the hypothalamus. He obtained a first sign of
reaction against cold: the skin blood vessels closed. As you know, cutaneous
vasoconstriction, that limits the loss of heat towards the environment takes place
when a dog is threatened by cold. But the importance of this result has been
contested because other teams could not replicate the results. This is exactly why
the scientific community denies the possibility of deep body sensitivity to cold,
because nobody can obtain shivering by cooling the brain. Interestingly, though, the
heating of the brain area brings the classic fight against heat. Animals usually adopt
an extended posture, their skin is vasodilated, and they start to pant. This breathing
pattern is the Dog's most powerful mean to fight a temperature rise, because
evaporation absorbs much heat from the mouth, the nose, and the upper airways.
Since this experimental result is obvious, all physiologists accept that the heat loss
response is triggered mainly by the deep body temperature whereas the cold defense
reaction responds mainly to skin temperature. How am I supposed to explain the
contradiction between the incapacity of others to obtain convincing results when they
look for a deep body cold sensitivity, and my own conviction that there exists a
hypothalamic sensitivity to cold? I believe that the answer to that question is in the
fact that the animals used in our predecessors' experiments were placed under
general anesthesia for evident ethical reasons. I have the conviction that anesthesia
has profoundly masked the brain sensitivity to cold. You may tell me that Magoun
and Barbour's experiments with hypothalamic warming were also done under general
anesthesia and that they gave convincing results, but I will answer that hyperthermia
is far more dangerous for a dog, and a human as well, than hypothermia. We may
die from a 5°C rise, but not from a 5°C drop in body temperature. Hence, we may
think that the function fighting against hyperthermia is stronger and can better resist
general anesthesia than the cold defense. Maybe our predecessors did not get
results in their experiments against hypothermia because their animals were under
An now Joseph Chatonnet explains how he plans to prove his hypothesis:
"I would like for you to create a drop in the local temperature of the hypothalamus in
unanesthetized dogs and to record the animals' reactions. Of course the surgery will
be done under general anesthesia, but the experiment, which is painless, will be done
after they awake. I expect that you will obtain vasoconstriction in the skin and
shivering in the muscles."
All the following will prove Joseph Chatonnet right, as will be seen now.
First, I had to make a probe that I could implant into the brain, a thermode, that
would produce the local drop of hypothalamus temperature. The first model was just a
concentric tube with a central injector of cold water. The second model was much tinier,
a concentric needle with central injection of propane under pressure. The use of
propane under pressure, was the idea of Michel Tornier, the engineer who helped me
build the thermode. Thus propane expanded near the tip of the needle, mildly cooling it
and the surrounding tissues. The cooling needle was chronically implanted under
general anesthesia. After healing and full recovery, the dog did not seem to mind the
thermode. He manifested a real interest in its being a subject in the experiments which
broke the monotony of its days, an gave signs of joy as we patted it.
Fig. 1.1 An example of recording the shivering of an unaneasthetised dog during short
periods of very minute cooling (black bars) of its hypothalamus. It can be seen that
cooling triggered bursts of shivering. The cooling was so minute that the thermocouple,
located at some distance, hardly recorded the temperature drops.
Even tough we had placed a heating craddle above the dog's body and head, the
dropping of the hypothalamic temperature brought out all the signs of fight against
cold6. Within seconds of the beginning of hypothalamic cooling, the dog was already
shivering and its fur was horripilated its back and neck. These signs disappeared as
soon as we stopped the cooling7. The animal being held warm by the heating craddle,
there was nothing to disprove the fact that the dog possessed an internal body
sensitivity to cold, since there could have been no source of cold other than its
hypothalamus. In addition, we felt reasonnably comfortable with a hypothalamic
localisation as the same locus had been shown previously to be home of the sensitivity
6 J. Chatonnet, M. Tanche, & M. Cabanac. Influence d'un abaissement localisé de la température de la
région diencéphalique sur l'activité musculaire thermorégulatrice chez le chien évéillé. J. Physiol. Paris,
1960, 52: 48-49.
7 Professor Chatonnet, although he is a cool, and calm, and collected Lyonnais, could not hide his
excitement with these results: "Look, look! Cabanac, and take good notes!".
Claude Bernard's classical experimental method, but the logic of which is anterior
to Claude Bernard, has two volets: stimulation and ablation. In the past, any study of a
nerve consisted first in transecting it and observing the results of the section, and then
stimulating successively the peripheral and the central truncs of the cut nerve. To study
an organ suspected to be an endocrine gland the first step was the ablation of the gland,
then the grafting of the gland or the injection of gland extracts. Ablation, stimulation,
both volets complete each other. Each a contrario proves the other one. Ablation, when
followed with an effect, proves that we have suppressed an active agent.
The suppression of the cold sensitivity in the hypothalamus was, therefore the
inevitable second volet of the experiment. Through our interventions while under
anesthesia, we achieved the electrolytic coagulation of the base of the hypothalamus.
This intervention consisted in flowing an electrical current with two electrodes
implanted in a zone to destroy. Electric current destroyed the nervous tissue, and the
sensitivity to cold. After the recovery from anesthesia, the animals were acting
normally, but shivered only with drop of skin-level temperature. They had lost their
central shivering, but were still apparently normal otherwisem 8. This proved that the
hypothalamus was therefore implicated in the process of temperature regulation and
that a sensivity to cold there lied there, in the hypothalamus.
The preceding story, that of the hypothesis of the hypothalamic sensivity to cold
proposed by Joseph Chatonnet, is a perfect example of the scientific path leading to a
discovery. Every step is an example of the process leading from a state A to a state B. In
the A state, the scientist makes an initial observation or obtains results that are
contradictory to the official thesis. He is then in a state of discomfort because he himself
believes in this ruling official theory. Yet, he must accept that his own results render
unlikely the accepted ruling theory. This is a privileged situation, however, because it
brings in a new way of looking at the ruling accepted theory; now, you have found a
crack in the wall. The scientist is now in the B state: he tries to find a replacing theory, in
which the new facts and the old knowledge must be blended. In the preceding pages,
we have seen Joseph Chatonnet in a state of discomfort with the theory that warm
sensitivity is internal and cold sensitivity is superficial. To escape this discomfort and
move to the B state he proposed that there exists a cold sensitivity deep located deep in
the core of the body, probably in the hypothalamus.
In experimental sciences, the next step is the bringing forward of a new
experiment to provbe the new hypothesis. Actually, the experimental results showed
Joseph Chatonnet right. The experiments raisedus one degree higher on the ladder
leading to knowledge of temperature regulation. We now understand better how our
own body works. This new understanding will place us in a priviledged situation to
8 M. Tanche, J. Chatonnet, & M. Cabanac. Exclusion elective de la sensibilité "centrale" hypothalamique
au froid. Proc. of the Congress of the International Union of Physiological Science, II, 1962, p. 484.
look at what occured during an experiment in a completely different field, the
physiology of respiration.
FROM PHYSIOLOGY TO EXPERIMENTAL PSYCHOLOGY
Pierre Dejours, a professor at the Faculté de Médecine de Paris, lays on a cot in a
swimsuit. A mask is covering his nose and mouth, bringing the expired air into a
spirometer to measure the volume of air he is expiring, and into gas analyzers to
calculate his intake in oxygen and production of carbon dioxide. Over him, the window
of his laboratory is wide open on Saints-Pères street and let the freezing cold air from
the of Paris invade the laboratory, in this cold December of 1962. Two powerful fans
flood him with this arctic cold air. He is riddled with shivering from his top to his feet.
Today, he serves as a guinea pig for an experiment that he himself devised to answer a
question that upsets him: What is the influence of thermal shivering on ventilation? He
has showed in the past that the hyperventilation that accompanies muscular work is
both the result of humoral and nervous signals. This means that the adaptation of the
ventilatory flow rate depends on signals coming from the blood and sensitive nerves.
Now, he wants to know what happens whith shivering, a muscular activity that does
not produce real movement. Does the ventilatory flow respond to humoral signals or to
nervous signals? The answer is what he expected: the muscular contraction of shivering
'hooks' the ventilatory flow rate 9 in a way similar to that he already described with
normal muscular activity. But today, we are still only collecting the data that will only
be analyzed later.
After Algeria, where I served as a physician and almost decided to continue my
life in medicine, I am ending my military service as a researcher in Val de Grace, the
historical military hospital, where we work on experimental hemorragic shock. Two
days a week my military superiors allow me to work outside and I chose to work with
Pierre Dejours to complete my post-doc formation. Since I know temperature regulation
a little, Pierre Dejours has proposed me this research. Today I am the experimentator, I
collect the data with Albertine Lacaisse, the third member of the team. Tomorrow I shall
be the subject, and I am a little anxious when I see the obvious discomfort endured
today by Pierre Dejours.
The following day, when my turns comes and when they open the window, I
experience what a true cold discomfort means. Not very often are we exposed to a cold
able to bring thermal shivering in only minutes. All our lives we are shielded by our
clothes and we live in a mini tropical climate created by them and protected also by our
houses. If ever we are able to shiver, it is most often only after hours of rest in a cold
9 M. Cabanac, A. Lacaisse, P. Pasquis & P. Dejours. Caractères et mécanismes des réactions ventilatoires
au frisson thermique chez l'homme. C.R. Seances Soc. Biol. 1964, 158: 80-84.
environment and we quickly increase the insulating power of our clothes to shield us of
from the cold. Today, as a subject in this experiment, I am only dressed in a swimsuit
and a cold wind blows on my poor whole body skin. To escape the freezing wind I
would wish to be able to huddle up, and that both my back and my front laid on the
mattress at the same time. The cold wind that seemed benign at the beginning of the
cold exposure quickly becomes extremely unpleasant. I can stand it, but the unpleasant
cold feeling invades my whole body. Soon I start to be shaken with intense shivers, and
I am longing for this session to reach its end.
Suddenly, everything stops. I am seized with a sudden wave of delicious, sweet
heat. At the same time all shivering stops and my muscles become completely relaxed.
Yet, the fans are still on and I can feel that my skin is still very cold. This strange
phenomenon has already been described by the Swiss-English team of Barcroft and
Verzàr, as I would find out later, under the terms of "basking in the cold"10. The warm
sensation lasts a few seconds, then the discomfort and the shivering return. A few
minutes later, the warm sensation happens again and alternates with cold discomfort
periodically until the end of the session.
Here we see how valuable it is for a researcher to remain as close as possible to
the raw data collected. Interestingly enough, this basking in the cold, that had been
described in passing by its discoverers as a curiosity of Nature, was never really
scientifically studied. To me, it take a peculiar significance since my thesis was on deep
body cold sensitivity. To know that there is such a cold sensitivity buried in the depth
of our body gives me a new way to look at this basking in the cold, and I have an inkling
that the hypothalamic sensitivity might intervene in the perception of thermal cold
discomfort. An experiment with purely physiological aims, such as the measurement of
ventilatory responses to shivering turns out to allow me to do an experiment on myself
with unexpected perceptions, that would have slipped me if I had not been a subject
myself. Could it be that the perception of cold discomfort would be attributible to
internal receptors, in the hypothalamus, that would be sensitive to the drop in body
temperature and the transient breach of warm comfort would be the result of the intense
heat production the depth of my body? No physiologist can doubt that the proof that
there exists a hypothalamic cold sensitivity found with the the Dog is also trus with the
Human species. Therefore, there exists in my hypothalamus, i.e. in the core of my body,
neurons the cooling of which trigger a fight against cold. Could my intense shivering
have produced a poweful heat wave that would have stopped my cold defence for a few
seconds? If so, this would mean that the cold sensitivity in the hypothalamus
commands not only shivering and skin vasoconstriction, but also the perception of cold
discomfort in a cold environment.
10 J. Barcroft & E. Verzár. The effect of Exposure to Cold on the Pulse Rate and Respiration in Man. J.
Physiol, London, 1931, 71: 373-380.
On second thought, such a hypothesis seems to make sense. Indeed, one can see
that simple behaviors such as postural adjustements, but also more sophisticated
behaviors such as clothing and house building, heating and climatization serve
temperature regulation. At each instant, we perform behaviors that improve
temperature regulation, without even being conscious of the fact. For example, we live
permanently in a micro climate that we carry with us under our clothes. If behavior is to
serve temperature regulation, then there must be some conscious perception that
determines its adaptation, one way or another, to body needs. It is then not so srprising
that the perception of thermal comfort obeys to signals neighboring those which control
Now, we must find out if this hypothesis fits reality. As soon as I can, I rush to
the library to see if I can find what science known on thermal comfort, a field that was
taught neither in the medical curriculum nor in the science curriculum of my studies.
What is known is quite limited. For 1963 science, thermal comfort depends only on skin
temperature. Thirty three degrees is the skin normal and comfortable temperature; as
soon as the skin temperature is warmed above this temperature the subject reports
warm discomfort 11. As soon as skin is cooled below 33°C, the subject report cold
discomfort. No mention of deep core temperature. Yet, basking in the cold exists, I
Discharged from militairy service, I accept with joy Professor Chatonnet's offer of
to move to Lyon and to join his team. He allows me to run my own research and study
this basking in the cold, a very liberal attitude for the time.
To study the basking in the cold phenomenon, I must repeat Pierre Dejour's
experiment all over, but instead of measuring the respiratory variables, I must
specifically explore the subjects' thermal comfort during exposition to heat and cold. We
can do it by periodically asking the subjects on their state of thermal comfort or
discomfort, and by simultaneously recording their body temperatures, from the skin
and the inner body. Thus organized, the experiments show unexpected results 12. As
soon as a subject is exposed to cold, his rectal temperature goes up, contrary to what we
may expect. At the same time, the subject starts complaining about the cold sensation
and the discomfort of shivering. When the comfort and the shivering become
intermittent, when the muscular relaxation and euphoric comfort alternate with the
discomfort and the shivering, then the rectal temperature is higher than in the beginning
of the session. Of course, rectal temperature is not that of the hypothalamus. It is well
known among temperature physiologists that the time course of the temperature in
these two loci may be very different. Anyway, this first exploration of thermal
11J. D. Hardy, Control of heat loss and heat production in physiological temperature regulation. Harvey
lecture, 1953-1954, 49: 242-270.
12 J. Chatonnet, M. Cabanac & S. So. Évolution de la sensation thermique au cours de l'exposition au froid
chez l'homme. J. Physiol. Paris, 1964, 158: 305-307.
cdiscomfort upon cold exposure is not at all encouraging. On the other hand, the skin
temperature recording shows that as soon as the subject is exposed to cold, his skin
temperature drops brutally, which is not surprising, but also that during the alternance
of comfort and discomfort skin temperature remains very low. The episodic comfort
cannot find its source in a warming of the skin. At most, skin temperature has a
tendency to oscillate at low amplitude in phases with the short small episodes of
comfort. The oscillations of the temperature of the skin are due to modest waves of
periodic vasodilatation 13, which take place in bare skin areas. This phenomenon has
been described as "hunting" by the British physiologists who discovered it 14.
Although we cannot attribute the perception of comfort to core temperature, on
the basis of this experiment, it remains that the state of comfort described in
psychophysiology textbooks as being the result of a skin temperature leveling at 33°C is
not valid to cold exposure. Indeed, the onset of cold discomfort in the skin temperature
of the cold-exposed subjects drops might be understood as the sole sign the brain takes
in consideration to produce thermal comfort. Yet, our experiment deos not confirm the
classical view that confort depends only on skin temperature. The alternating phase of
comfort and discomfort takes place some minutes after the onset of cold exposure, when
skin temperature has dropped and continues to drop. The short bursts of basking in the
cold take plce during the whole session with skin temperatures widely different at the
beginning and at the end of the session. The absolute value of skin temperature cannot
explain the basking in the cold. The official theory of comfort, basing the experience of
comfort on a lukewarm skin temperature is no valid therefore and must be revised.
Another point must be checked nevertheless; it cannot be discarded that the small
oscillations of skin temperature during the basking episodes are the source of this
comfortable basking. It is well known to temperature physiologists that the
temperasture sensors in the skinwork not only as thermometers, sensitive to
temperature, but also as sensors of changes. i.e. they are quite sensitive to the direction
of a change in temperature. Thus they might amplifythe temperature signal they send
to the brain. It is possible that the small temperature changes in the skin, caused by the
periodical vasodilatations of the hunting phenomeneon, are amplified as signals by the
phasic sensitivity of the skin temperature sensors. If this is the case the next problem
will be to explain where do the hunting phenomenon takes its origin. Whatever the
anwer to that problem, it would remain that the hunting takes place a very different
skin temperatures and, thus, that the perception of thermal comfort cannot be explained
13 release of the tonus in the skin blood vessels, resulting in a warming of the skin. The opposite process
14 T. Lewis, J. Haynal, W. Kerr, E. Stern & E. M. Landis. Observations upon the reactions of the vessels of
the human skin to cold. Heart, 1930, 15: 177-208.
R. H. Fox. Cold-induced vasodilatation in various areas of the body surface of man. J. Physiol.
London, 1962, 162: 289-297.
only on the skin signal. This first experiment is not sufficient to answer our initial
It is necessary to remove the ambiguity introduced by the complex way of
working of the skin temperature sensors. The aim is to control skin temperature. A way
to reach that goal consists of water immersion. Water is 25 times more conductive of
heat than air. Water is thus capable of extracting almost all the heat contained in skin
and clamps skin temperature at the bath temperature, provided the bath is well stirred
in order to prevent the formation of insulating layers of water at the surface of the skin.
Thus we can fix the problem of changing skin temperature met in cold-exposed subjects
and render constant skin temperature. In a well-stirred bath, if changes in the
perception of thermal comfort take place over time they should not be attributed to skin
temperature because skin temperature will have remained stable during the bath. On
the other hand, we may expect slow drifts in the subject's deep temperature for the same
reason that keeps skin temperature steady: water is highly heat-conductive. The
experiment will hence consist in asking the subject to report whether he feels
comfortable and in recording the changes in rectal temperature in order to know the
sujects core temperature. The verbal report, whether the subject feels comfort or
discomfort, will be correlated to the rectal temperature, representative of the inner body
temperature, and to the bath temperature, representative of the skin temperature. Such
a method should permit to warm up as well as to cool down a subject with only minor
changes in bath temperature. The method will permit to study thermal perception of
warm as well as cold discomfort.
Subject M. J .P. 14 mar 1964
r ec t al
0 30 60
Fig. II.1 Warm discomfort (open triangles) in a subject immersed in a warm
thermostated bath. Discomfort, known from the subject's verbal response, takes place
only at about min 32 when deep core temperature (rectal temperature) begins to rise
above 37°C, although mean skin temperature was warm since the beginning of the bath.
From, Chatonnet & Cabanac, 1965 15.
Immersion into a 38°C-bath is perceived by the subject as being comfortable. This
comfort remains for about 10 min, then gradually stoops down to a state of indifference
towards the bath temperature. In turn, indiffernce merges into a feeling of discomfort.
After a half hour, the subject is clearly in a state of dicomfort; he is too hot (Fig. II.1). If
we compare this evolution to that of body temperatures we notice that the skin
temperature has stayed the same all along because the bath temperature is narrowly
regulated. On the other hand, core temperature has progressively risen above its initial
value. The cause of this rise is to be found in the subject's own heat production which
cannot leave the body because the bath is not cooler than the body. Thus the heat
produced within the body accumulates in the body. In addition, because the bath is
slightly warmer than the body, some heat also enters the body this way, although the
gradient is weak. The overall result is a progressive rise of core temperature parallel to
the slow onset of warm discomfort.
To explain the time course of thermal discomfort, we may discard fluctuations in
skin temperature, because bath temperature was maintained constant during the whole
session. On the other hand, the pattern of rectal temperature, could explain first the
absence of discomfort at the beggining of the bathwhen rectal temperature was low,
then the subject's warm discomfort which took place when rectal temperature rose.
Immersion into a 33°C-bath (Fig. II.2) is perceived as nearly indifferent, at least as
not-uncomfortable, especially if the inner body of the subject is a little higher than
averag (we shall return to this in the next chapter). However, after fifteen minutes, the
subject feels a thermal discomfort, he is cold and starts to shiver a little. How to explain
both cold discomfort and shivering? In the same way as we could explain warm
discomfort in a stable 38°C-bath. Skin temperature is clamped by the bath and remains
constant and close to 33°C, but rectal temperature dropped noticeably because the high
heat conductivity of water cools the body even when the difference between water and
body temperatures is only 4.5°C as is the case here. The lower core temperature is the
source of cold discomfort and shivering. In order to verify that skin temperature enters
for nothing in this process, let us raise the bath temperature up to 38°C. This will warm
up the skin temperature only, while the rectal temperature will keep dropping for the
next 20 min because thermal exchanges within the are slow and there is strong inertia
before the core care warm up. In this situation, the subject is aware that the water
15 J. Chatonnet & M. Cabanac, The perception of thermal comfort, Int. J. Biometeorol. 1965, 9: 183-193.
temperature has gone up, but he still feels cold and continues to feel uncomfortable as
long as his rectal temperature stays low.
Fig. II.2 Cold discomfort (solid triangles) in a subject immersed in a lukewarm
thermostated bath. Discomfort, known from the subject's verbal response, takes place at
about min 8 when deep core temperature (rectal temperature) begins to drop above
37.7°C, athough mean skin temperature remains steady. The discomfort persists at the
end of the bath even when the bath, and the skin with it, is warmed up to 38°C. From,
Chatonnet &Cabanac, 1965 15.
Where are we now? According to the generally accepted theory, the perception
of thermal comfort depends only on skin temperature. Comfort would be the conscious
experience from a 33°C-skin. Any displacement from that temperature would cause
discomfort. Yet, contrarily to that theory, in the water-immersed subjects’ skin
temperature was not sufficient to explain thermal comfort. We may hence conclude that
thermal comfort depends on thermal sensors inside the body, most probably in the
One of the advantages of doing research in a school of medicine is that contact
with hospitals is easy and facilitates access to pathological situations neighboring our
research subjects. Thus, two studies would come corroborate our conclusion about the
perception of thermal comfort. The first one of them was the simple observation and the
interrogation of patients undergoing gastric freezing as a treatment for their stomach
ulcers. This method, in vogue during the sixties, consisted in placing a balloon in the
patient's stomach and running a below freezing-temperature liquid in it. At the same
time, the skin of the patients was well heated with a heating blanket. Temperature
recordings confirmed that skin temperature remained stable while core temperature
cooled down during the gastric freezing. Even though the patients were well heated in
their exterior, they still shivered and complained of thermal discomfort, during the
whole therapeutic session 16. The second pathological situation was a rare disease called
"erythrodermia". The patients' skin is red, as indicated by the the name of the disease,
because the blood vessels are dilated. It follows that skin temperature is very high, that
the patient loses a lot of heat and shivers. The patients also complained of permanently
feeling cold, although having a high skin temperature which they acknowledged of
being warm 17. These two pathological situations show that thermal comfort depends on
signals aroused in the depth of the body, indicative of core temperature.
16 J. Chatonnet, H. Thiers, M. Cabanac & J. Pasquier. Sur l'origine de l'impression de confort thermique.
Lyon med. 1966, 50: 1387-1392.
17 H. Thiers, J. Chatonnet, M. Cabanac & P. Michel. Le frisson et l'inconfort thermique des
érythrodermies. Leur valeur séméiologique. Leur interprétation physiopathologique. Lyon med. 1966, 50:
These results as a whole have been obtained by introspective exploration of the
subjects and the patients. Such a method, as well as the results and the conclusions
meant a qualitative leap in a physiologist's methodology as well as in his sphere of
activity. In the sixties, the study of behavior did not really have its place in physiology
laboratories. Meanwhile, the library shows me that scientific psychology is strictly
behaviorist, in its defence against the excesses of psychoanalysis, and forbids itself to
resort to introspection or use of "mentalistic" concepts. It is thus in a quasi-desert that I
enter. This adventurous leap into psychology is hopefully not new, as illustrious
physiologists of the past have done it, like the French Claude Bernard who studied thirst
perception, the American Walter B. Cannon hunger perception, and the Russian Ivan
Pavlov who, while he was interested in saliva secretions, discovered a new way of
linking a sensory signal to a physiological response, the conditioned reflexes. Actually,
this thought process is a logical course for any physiologist as soon as he realizes that
behavior is for all animals and humans a way to satisfy the needs of the body. Animal
life would not be possible without the control of behavior to serve these needs.
However, a beginner physiologist needs these reassuring examples of the great
predecessors to leave, with some quaking, his sphere of competence and dare embark
'DO-IT-YOURSELF' CAN BE ADVANTAGEOUS
I am just out of an experiment where I was the subject. The next subject will be
here in a few minutes. Although I am still dripping with sweat from the hot bath, I do
not have enough time to cool myself because I must clean the bathtub and the
equipment and prepare everything for the next session. At the beginning of my carreer
my research means are limited I cannot pay the salary of a graduate student or a
technician and must do everything by myself. After I sterilized the various probes I
must still clean the bathtub for the next subject. I spread some Chlorox and vigorously
brush the whole tub. This exercise still augments my hyperthermia18. After cleaning, I
rince. Suddenly, I realize that I am rincing with very cold tap water and further, that I
let complacently that cold water run on my brush-active hand. This sensation is
pleasant. Extraordinary! As the cold is biting the water temperature must be around
15°C. Yet, according to Psychology textbooks, only temperatures around 33°C are
pleasant. How come this biting cold water feel pleasant? Could it be because my body
temperature is high, indeed it must be high, I am still sweating? Quick, quick, I must
The next subject was programmed to serve in a study on the influence of
hyperthermia on thermal comfort. Instead of a simple questionnaire on his perception
of comfort /discomfort, as scheduled, I must explore his sensations when I warm and
cool his skin. What should be done is to draw his attention towards his skin thermal
sensation and ask him whether he feel pleasure or displeasure. Fig. III.1 presents the
experiment. It is important that we control all the subject's body temperatures. To
achieve that , the subject will remain in water; thus his skin temperature will be that of
water and his deep temperature will drift slowly up if the bath is warm, and down if the
bath is cold. At the same time he will receive local temperature stimuli on one hand
only where the subject will answer for pleasure and displeasure. A small tank filled
with water is sufficient to immerse the whole hand. It will serve as a temperature
stimulator. Water temperature will be recorded from a thermocouple attached to the
subject's hand. In order to be confident that the skin temperature is the same as, or close
to, water temperature in the small tank, it is sufficient to let the hand immersed for half
a minute, provided the subject moves his hand and stirs the water. At the end of the 30-
sec period, the subject declares whether the temperature sensation at his hand is
pleasant, unpleasant, painful, or indifferent. To gain more precision, the subject
18 Physiologists and physicians call hyperthermia the state of a subject who is gorged with heat and
whose body temperature is higher than normal. Hypothermia is the reverse condition when body
temperature is too low. N.B. Hyperthermia is different from fever. During fever the temperature
'wanted' by the body is higher than 37°C, while during hyperthermia the body 'wants' to stay at 37°C.
describes his pleasure or displeasure with a number on the following scale: +2 very
pleasant, +1 pleasant, 0 indifferent, -1 pleasant, and -2 very unpleasant.19
Fig. III.1 The method used to investigate the pleasure or the displeasure aroused in the
left hand. The subject is immersed in a well-stirred thermostated bath, in order to
control his mean skin temperature. The skin temperature of the stimulated hand is
continuously recorded as well as deep core temperature (both rectal and œsophageal).
From, Cabanac, 1969 20.
In many following experiments the subjects reported identical sensations of
pleasure and displeasure. Beneath 14-15°C and above 45-46°C, stimuli were decribed as
very unpleasant. No wonder, the thresholds for temperature-aroused pain are 15°C for
cold and 45°C for warm. Painful stimuli were of course reported as unpleasant by the
Between these temperatures, the subjects had no problem using the minus 2-plus
2 rating scale. But the pattern of the responses were widely different with different
body core temperatures. When the subjects were hyperthermic, in a warm bath,
regulated at 38°C, they described as pleasant allthe cold stimuli in which they dipped
19 Years later, psychophysicists taught methat this scale was not the best I could have instructed the
subject to use because in that scale, 2 is not the double of 1. Numbers are only guide marks on the scale
to help the subject make up his mind. Using a maximal rating at 2 tends to bias the results when
sensation is intense. Rather than such a 'category scale', it would have been better to use a 'magnitude
estimation scale'. But, for the time being, it does not matter much because, although imperfect, the
method provided convincing clear results.
20 M. Cabanac, Plaisir ou déplaisir de la sensation thermique et homéothermie, Physiol. Behav. 1969, 4:
their hands. Even stimuli the temperature of which was lower than the cold pain
threshold, was occasionnally described as pleasant biting cold. On the other hand,
warm stimuli were described as unpleasant. The sensation from the subject's whole skin
immersed in the bath payed little or no role in the perception of pleasure or displeasure
in the hand. This appeared when the bath tempature was abruptly modified with
pouring several buckets of ice into the bath; thus, the bath temperature dropped by
some 15°C but if that manœuvre was rapid the subject's body core temperature
remained high for several minutes. During these minutes, a warm hand remained
unpleasant and a cold hand remained pleasant as long as the subject remained
hyperthermic. Then, under the influence of the cool bath, his deep core temperature
slowly fell and little by little his percetion of pleasure reversed. Cold stimuli were then
perceived as unpleasant and warm stimuli as pleasant.
Fig. III.2 Ratings (in ordinates) given by the subject of Fig. III.1 to describe his pleasure
or displeasure when his hand is stimulated by the temperatures shown in abscissæ.
Each stimulus lasts 30 s. The subject respond verbaly by giving a number. Round dots
indicate the responses obtained while the bath was warm, and triangles when the bath
was cold. Solid symbols: the subject was hyperthermic; empty symbols: the subject was
hypothermic. It can be seen that a given stimulus can be experienced as pleasant or
unpleasant depending on the subject's internal state, not his mean skin temperature.
From, Cabanac, 1969 20.
The case of a hypothermic subject in a cool 28°C bath is exactly symmetrical. The
subject describes as pleasant all the warm stimulations of his hand. Even a painfully hot
stimulus can be described, occasionally, as pleasant. On the other hand, all cold stimuli
are reported as unpleasant. The influence of the skin temperature of the rest of the body
immersed in the cool bath, remains minor or nil. Indeed, if enough hot water is poured
into the bath tub and raises the bath temperature by ca.10°C, a warm hand remains
pleasant and a cold hand unpleasant, as long as the subject remains hypothermic.
The subject's deep body temperature is the source of the pleasure or the
displeasure aroused by a peripheral stimulus (Fig. III.2). The technical hitch (incident)
of the ice-cold water on the hand of the tub cleaner revealed a new fact that appeared
neither in psychology texbooks, nor a fortiori in physiology texbooks: that a given
identical stimulus could arouse two different sensations, one pleasant, the other
unpleasant. If we look closely at the situation and try to analyse what is the cause of this
bizarre dual sensation, we find that everything reamined constant, except the subjects'
deep body temperature. There must be the explanation of this phenomenon. How is it
that this observation, although quite simple and even obvious, escaped those who
studied thermal comfort earlier? Simply because wanting to be rigorous they studied
resting subjects in steady state. A human subject at rest in a lukewarm air environment
evolve spontaneously towards a 37°C deep core temperature and a 33°C mean skin
temperature, and there feels comfotable. But this steady state is somewhat artificial
because our activity tends to augment our body heat production and our behaviors take
us un cooler or warmer environments. Thus our core temperature tends to fluctuate
permanently above or below the regulated level of 37°C21. It is quite understandable,
therefore, that nature have produced animals and humans that regulate not their skin
temperature but their deep temperature which reflects the overall balance of heat gained
and heat loss. In turn it is understandable that nature have poriviledged core
temperature as the main source of thermal comfort, because thermal comfort determines
The fact that deep temperature determines the pleasure of skin sensation entails
an important consequence: sensory pleasure is adapted to the body's need. If a stimulus
is pleasant when it tends to correct an internal deficit, the seeking of pleasure should be
useful for maintaining the stability of the internal milieu of the body. As long as the
deep layers of the body will be too cool or too warm, corrective skin stimuli will feel
pleasant and the seeking of pleasure will be useful. To a hypothermic subject a warm
skin will be pleasant and a cool skin unpleasant. Sensory pleasure seems to be the
motor of behaviors adapted to the defense of the body's physiological integrity since
pleasant stimuli correct the internal deviation of deep temperature from the level
'wanted' by nature, and unpleasant stimuli would exagerate the deviation. The seeking
21 to make is simpler ,37°C is given here as the level wanted by our body. Actually, we know that the
temperature 'wanted' by our biological 'thermostat' oscillates roughly from 36.6 and 37.3 with a
minimum near 4:00 h in the morning and a maximun at about 16:00 h in the afternoon. The laws of
thermal pleasure remain true during this oscillation. The internal signal responsible of the pleasure of
the peripheral sensation is the difference between the set-point and the actual deep temperature, as will
be seen in the following pages.
of thermal pleasure, the avoidance of thermal displeasure are useful because they lead
to seek lead to favorable environments. This is how behavior is adapted to the needs of
Later the Tunisian psychiatrist Essedik Jeddi showed that the usefulness of thermal
pleasure is semantically prolonged22. The connotation of the words 'warm' and 'cold' in
a given language, depends on the climatic environment of the ethnic group using that
language. In the languages of Europe, where the climate is lukewarm, cool, or cold,
'warm' possesses positive and 'cold' negative connotations; a warm wellcome, a cold
reception. The reverse situation can be found in languages of Africa where the climate
is warm or hot. There, warm is hostile, and cold beneficial: 'bouroud', in North-African
Arabic and 'sedlemlol' in Senegalese Bambara-Ouolof both mean cool and are
synonymous with pleasant. In the same languages, 'sokhara' and 'tang' both mean
warm and annoyance.
Taking note that pleasure seems to be the index of usefulness, opens infinite
perspectives that are so exciting, even dazzling, that it is necessary, before
exptrapolating further, to check its validity in the very field where the phenomenon is
noticed. In other words, it is necessary, now, to explore whether the pleasure and the
displeasure evoked by skin thermal stimuli, obeys the laws of temperature regulation
which have been studied in physiology. Quick, quick, as soon as next week, I shall try
to answer that question, through the experimental method of course.
22 Jeddi E. "Confort du contact et thermorégulation comportementale", Physiology and behavior, 1970, 5:
PLEASURE IS ADAPTED : THE LAWS OF TEMPERATURE
REGULATION APPLY TO THERMAL PLEASURE
Two experiments will permit to verify the validity of the hypothesis according to
which the research of pleasure and the avoidance of displeasure, in thermal the case of
temperature sensation, induce thermoregulatory behaviors: on the one hand, a study of
thermal pleasure and displeasure in feverish subjects, and, on the other hand, a
behavioural selection by the subjects themslves of the skin temperatures that are most
Fever and Pleasure
We know that the deep body temperature is ca. 37°C in the human species. This
temperature is subject to a regulation, meaning that temperature is constantly
maintained stable by a complex apparatus. This apparatus includes themperature
sensors, which operate as thermometers, and defense responses. The temperature
sensors are located in the skin and in the hypothalamus. The defense responses are skin
vasomotricity, that makes of the skin a variable radiator, further, more powerful
responses, sweating and shivering. A characteristic feature of any regulation is the
existence of what is usually called the 'set-point', a sometimes virtual signal, that serves
as a landmark for the level that must be regulated.
When the regulated variable, here the inner body temperature, deviates from the set-
point, the regulation uses its defense responses to return it back to normal, close to the
set-point. Specialists in control and regulation call 'error signal' the difference between
the actual value of the regulated variable from that of the set-point. Regulation consists
of the elimination of the error signal. The responses are used for that purpose and
return the regulated variable to its initial normal value close to the set-point. Thus,
hyperthermia is followed with sweating, the evaporation of which serves to cool down
the skin and then all the body. Symmetrically, hypothermia is followed with muscle
shivering, the heat production of which warms the body up. The true value of the set-
point is difficult to know because it is a signal, not an organ. It is the triggering of a
defense response that indicates that there is an error signal, i.e. that the regulated
variable is not equal to set-point. The presence of a defense response is thus the indirect
method to know the set-point. The set-point is the value where the defense responses
disappear. As, in humans, the set-point of temperature regulation is 37°C, when there is
a positive error signal, for example if the core body temperature is at 38°C (38-37=+1),
the subject is said to be hyperthermic and sweats. When there is a negative error signal,
for example if the core body temperature is at 36°C (36-37=-1), the subject is said to be
hypothermic and shivers.
As long as behaviour is used in thermoregulation purposes, we must be able to
think similarly of thermoregulatory behaviour as of shivering and sweating, and in turn
of pleasure which triggers behaviour. If sensory pleasure aroused by thermal
stimulation of the skin is adapted to the defense of body temperature, it must follow the
determinism of sweating and shivering. We have seen in preceding chapter that that
seemed to be the case in healthy subjects.
The studying of feverish subjects is an excellent and simple way of verifying the
existence and modus operandi of the set-point in the case of the biological thermostat. Let
us compare what takes place during hyperthermia and during fever. In hyperthermia,
core temperature is higher than 37°C and the subject sweats to cool his body down
through evaporation. During fever, body temperature is also higher than 37°C, but the
body reacts differently. Instead of sweating, the feverish subject shivers in the
beginning of the fever attack until his temperature has risen. Then, shivering stops
because body temperature and set-point are equal, there is no more error signal and core
temperature stays more or less stable. When fever ends, abundant sweating takes place,
until core temperature returns to 37°C. Everything is as if the biological thermostat
defended a higher body temperature during fever; that can be summarised as an
elevation of the set-point. If this is what actually takes
Fig. IV.1 Ratings (in ordinates) given a the subject during two different baths to describe
his pleasure or displeasure when his hand is stimulated by the temperatures shown in
abscissæ. During both sessions the experimental conditions are strictly identical (bath
33°C, rectal temperature 38.2-38.5°C), but the open symbols were obtained while the
subject was feverish from influenza, and the solid symbols in a control session while the
subject was in good health. Compare with Fig. III.2. From, Cabanac, 1969 20.
place, and if pleasure allows to adjust behaviour to thermoregulatory purpose, then we
must endeavour to explore the pleasure of thermal sensation in feverish subjects. We
may expect that pleasure should defend the feverish value of the thermoregulatory set-
The experiment will be simple. It will be a simple replica of the experiment
described in the preceding chapter with subject who were in good health. With an
alternance of cold and hot baths, I record the verbal reports of pleasure and displeasure
by subjects, in response to a broad gamut of temperature stimuli, between the
thresholds for cold and warm pain. Each subject will serve twice in the experiment.
Once, he will be feverish feverish, the next time he will be in good health. In this cold
1968 spring, influenzas are numerous among the Lyon faculty of Medicine staff and it is
easy to recruit feverish subjects. The first feverish volunteer is my colleague René
Lacour, from the Department of Physiology.
Any scientist who sets up an experiment always pretends to himself that he acts
in perfect objectivity; he also presents his work and results to colleagues in knowing
societies with a poker face. It is important to hide his inside passion which would be
very suspect to others. However, he secretly hopes to be correct and is full of joy when
he is. In this present case, my results confirm magnificently my working hypothesis
(Fig. IV.1). A feverish subject uses all the range of verbal responses to describe great
pleasure or displeasure when I heat up or cool down his hand. These responses are
modulated as a function of his deep core body temperature in exactly the same pattern
as during health. A feverish subject is normal, therefore, and the patterns of responses
with and without fever are identical, with one important exception: the subject's internal
body temperatures are different in the two experimental sessions.
When body core temperature of the subject in good health is between 38,2°C and
38,5°C, he is hyperthermic and feels pleasure from cold stimulation and displeasure
from warm stimulation. Whereas, when the subject is feverish, although his body core
temperature is the same as above, between 38,2°C and 38,5°C, his pleasre/displeasure is
inverted. He feels pleasure from warm stimulation and displeasure from cold
stimulation. His response during fever his hence reset from the one in good health. All
is identical from one session to the other, except for the thermoregulatory set-point.
During fever, the higher set-point makes him respond in a hypothermic way as long as
his actual temperature as not equaled that of the set-point. In a subject in good health,
the error signal is suppressed near 37°C, whereas it is suppressed beyond 38.5°C in the
For more than a century, the physiologists have known with Liebermeister that
fever is a resetting of the biological thermostat to a higher temperature 23. Now, they
know also that fever if beneficial since Matthew Kluger showed them that lizards that
are unable to raise their temperature behaviourally die when injected with microbes, but
that others, which can raise their temperature by basking under an infrared heating
lamp, survive this injection 24. Our team has contributed to this knowledge by showing
that other ectothermic vertebrates frogs, and that animals as ancients as scorpions, and
as simple as leaches, use their behaviour to become feverish when injected with
pyretogens. In addition, crickets die like Matt Kluger's lizards when prevented to warm
themselves during infection 25. Fever is, thus, the manifestation of a defense mechanism
in response to infection. It is not good, therefore, to systematically 'treat' fever when it is
the only symptom of sickness.
It is now possible to bring an answer to the question of the influence of the
biological thermostat set-point on thermal sensation. Pleasure is sensitive to this signal
which is essential to thermoregulation: the error signal, the difference between actual
core temperature and set-point temperature. Sensory pleasure, therefore, obeys to a
fundamental law of temperature regulation. From this point of view, pleasure is the
same as sweating and shivering. The pleasure that is aroused by temperature stimuli
appears well adapted to the defence of body temperature.
In the above experiments, the subjects described their pleasure/displeasure
experienced when presented with temperature stimuli. Subject's sensation was
recorded in response to a stimulation imposed by the experimenter. There is every
indication that pleasure is an indicator of usefulness. What we need to know now is
whether, in real life, sensory pleasure is used as a signal for the needed
thermoregulatory behaviour. In other words, does the seeking of thermal pleasure
produce a behaviour that may be considered by any temperature physiologist as truly
thermoregulatory. To answer this question will be the aim of the next experiment. We
just saw that the thermoragulatory set-point was essential in determining pleasure. It
remains to be verified whether the other signals of temperature regulation 26 are taken in
account to trigger thermoregulatory behaviours, in a manner similar to what is known
in the command of shivering and sweating. Also, instead of just exploring the response
to imposed stimuli, it is the behaviour of the subject himself that will be recorded.
23 C. Liebermeister, Handbuch der Pathologie und Therapie des Fiebers, Leipzig, Vogelwelt, 1875.
24 M.J. Kluger, Fever, its Biology, Evolution, and Function, Princeton N.J., Princeton University Press,
25 M. Cabanac, Phylogeny of fever, in J. Bligh & K. Voigt, Thermoreception and Temperature Regulation,
Berlin, Springer-Verlag, 1990: 285-296.
26 We saw previously that the three main signals of temperature regulation are, skin temperature, deep
core temperature, and set-point temperature.
Figure IV.2 presents the experimental setup. The subject is immersed up to his
chin in a thermostated well-stirred bath, in order tocontrol the major part of his skin
temperature. The bath achieves a slow drift of his deep body temperature, due to the
high thermal conductivity of water. The subject's only instruction is to keeep a
maximum of pleasure at his left hand, by using the command of electrovannes with his
right hand, from within the bath. For an hour or two, the bath temperature is
periodically modified in order to explore a vast range of the subject's core and skin
temperatures. During that time, the subject behaves and seeks pleasure by using the
electrovannes. His left-hand preferred temperature is permanently recorded by
attaching a thermocouple to one of his fingers.
Fig. IV.2 The subject wears an œsophageal thermocouple the tip of which is located
behind his heart. This will permit the recording of deep core temperature (Tes) changes
with minimal lag. The subject's left hand, out of the bath, is thrust in a large rubber
glove. The tip of each of the glove fingers receives a small tube conducting continuously
water pushed by the pump (P). The outlet sends back water to where it was pumped
out, one of the two thermostated tanks. From inside the bath, the subject can, with his
right hand, operate the electrovalves that command the water circuitry to and from the
glove. Thus, he can feed the glove with eithe hot or cold water. From, Cabanac,
Massonnet, & Belaiche, 1972 27.
An example of the results is given in Fig. IV.3. We can see that the subject used
well the possibility that was given to him to play with his left-hand temperature to find
pleasure. It is immediately obvious that the time course of the temperature selected by
the subject tends to go in mirror opposition to that of his deep core body temperature.
Fig. IV.3 An example of recordings from the experimental set-up of Fig. IV.2. Tes,
œsophageal temperature; Tb, bath temperature; Tg, glove temperature, self-selected by
the subject himself as the most pleasant. The solid dots, superimposed near the glove
temperature recording show the theoretical preferred temperature, as predicted from
equation 4-1. From, Cabanac, Massonnet, & Belaiche, 1972 27.
In the seventies, it was a fashion for physiologists was to present physiological
results with mathematical models 28. The results of the experiment allow us this
approach. Systematical analysis of the recordings, point by point every minute, in
27 M. Cabanac, B. Massonnet, & R. Belaiche, Preferred temperatures as a function of internal and mean
skin temperatures. J. appl. Physiol., 1972 , 33: 699-703.
28 To be fair, I must admit that this is not only a fashion; to be able to present a mechanism in
mathematical terms, it is a prerequisite to measure quantitatively each parametre, a discipline which
implies a knowledge much deeper than that implied by the mere description of the time course of a
several subjects and several sessions each, leads to a mathematical model 29 summarising
the thermal preferences in the group of subjects:
Tpref=-0.3Tbath(Tes-36.3)+44 equation 4-1
where Tpref is the preferred temperature at the left hand, Tbath et Tes are bath and
deep core temperatures; all these temperatures are in degree Celsius. A quick check will
verify the validity of this model: by assuming that the subject is hypothermic, with an
esophageal temperature of 36.3°C. This cancels the second arm of the equation, in turn
the first factor is cancelled and the prefered temperature predicted by the model turns to
be 44°C, which is hot, but lower than the hot pain threshold.
The preferred temperature as predicted by the model (equation 4-1) is shown in
Fig. IV.3 by the black dots, superimposed to the recording of the subject's behavioural
choice of preferred temperature. We can see that the actual and theoretical preferred
temperatures coincide acceptably, as the dots are never very different from the
recording. We must be aware that what is at stake is the pleasure experienced by the
subject, a psychological variable, not a physical or physiological variable. The model
catches the eye on a piece of information which had remained unnoticeded in the
previous experiment. The subject's mean skin temperature, close to the bath
temperature, enter into play to modulate temperature preference. Its influence is
minimal when compared to that of core body temperature, in which a minute half-
degree change can reverse a subject's thermal preference, but it is still to be recognised 30.
We may draw two lessons from the results of this experiment. First, if we
compare the mathematical model of equation 4-1 to those that describe the autonomic
responses to hyper- and hypothermia, sweating and shivering, we can recognise close
resemblance. Behavioral and autonomic models look much alike. For example, the
equations experimentally obtained for heat production and for sweating of human
subjects are as follows:
Qm= 42(36.5-Tes)(32.2-Ts)+8(32.2Ts) equation 4-2 31
29 A mathematical model of a physiological function is a description of that function.The Physiologists
use two different types of models: the firstis a theoretical descriptionof the function under study; such a
model incorporates hypotheses and its predictions can be compared to the results of experiments. The
second is simply descriptive. Equation 4-1 belongs to this second type.
30 Everything is as if mean skin temperature modified the set-point for temperature regulation: a warm
skin lowers the set-point, a cold skin, raises it. This may explain the rapid rise of core temperature on
sudden cold exposure.
31 E.R. Nadel, S.M. Horvath, C.A. Dawson, A. Tucker, Sensitivity to central and peripheral thermal
stimulation in Man. J. appl. Physiol. 1970, 29: 603-609.
Qevap= 132Tes+20Ts-5500 equation 4-3 32
with, Qm (in Watts) as the rise in metabolic heat production above the basal value at
rest, Qevap (in Watts per square metre of skin) as the rise in evaporative heat loss, Tes
as the deep core body temperature, and Ts as the mean skin temperature (all
temperatures in degree Celsius).
The similarity between equation 4-1, describing the behaviour of thermal
pleasure seeking, and equations 4-2 and 4-3, describing the autonomic responses, is
striking, despite minor differences of parametres. The similarity manifests itself in the
fact that the same variables, enter into play, including mean skin temperature. This
similarity confirms the rooting of sensory pleasure into physiology and its biological
role, at least in the case of temperature sensation.
The second lesson, beyond the simple shape of the equation or the value
ofparametres, lies in the fact that the seeking of pleasure can be predicted from
physiological variables like mean skin temperature and core body temperature.
Pleasure, therefore, is not an abstract superstructure, a mental piece of information out
of the reach of science, but is, on the contrary, an explorable variable, emerging from
biological complexity. Not only can we study it qualitatively, but we can also measure
quantitatively its determinism. To seek pleasure will produce a thermoregulatory
behavior. To seek pleasure will be the motor of a behavior adapted to a physiological
finality. This aspect is one of the difficulties encountered in these studies: everyone
knows these mechanisms out off personal experience, and, from introspection, thinks he
knows all about sensory pleasure. Such a knowledge, though, is not science.
Te finding that the pleasure of temperature sensation is physiological in its origin
and purpose brings a question to mind: sensory pleasure in evidently not limited to the
pleasure of thermal sensation. Could it be that law put in evidence in the case of
thermal sensation be equally valid in other orders of sensation? Could it be that the
adaptation of sensory pleasure to the needs of the organism be a general phenomenon?
Let us try to answer this question. To mention sensory pleasure immediately
brings to mind sexual pleasure. However, any experimental study in that area would
raise ethical concerns. But other sensations also are source of acute pleasure, the best
example would be the sensations associated to food intake. Let us turn toward them.
The first thing to do is to check whether psychologists have already studied the
pleasures of gustatory and olfactory sensations. The literature shows that indeed some
32 J.A.J. Stolwijk, B. Saltin, A.P. Gagge, Physiological factors associated with sweating during exercise.
Aerospace Med. 1968, 39: 1101-1105.
authors have turned their attention to it 33, but as prisoners of their times, they limited
their studies to animal experiments. Alas, the will to be rigorous led to throw the baby
with the bath water. By the end of the sixties, in our university library, the only
accessible publication that would describe a study of gustatory pleasure showed that
sweet solutions are pleasurable at all concentrations, that acid and sour solutions are all
displeasurable at any concentration, and that salt is pleasurable at low concentrations
but turns unpleasant when concentration rises. On the whole, it appears that the
knowledge of gustatory sensation is in the same situation as that of thermal sensation a
few years back: the official theory is that sensation depends only on the stimulus; there
is no mention of the physiological internal state of the tasting person. My next step will
consist, therefore, in exploring the gustatory pleasure while modifying the subject's
internal state., i.e. to repêat with taste all the experiments on temperature sensation.
33 P.T. Young, The role of affective processes in learning and motivation. Psychol. Rev. 1959, 66: 104-123;
C. Pfaffman, The pleasures of sensation, Psychol. Rev. 1960, 67: 253-268.
ALLIESTHESIA: PLEASURE INDICATES USEFULNESS
This morning, the subject who comes to the laboratory has not eaten since
yesterday evening. The subject is my boss, Professor Chatonnet, whom we met on
Chapter I; he volunteered to serve as a subjecty in this experiment. I give him a tofee
and he savours it with obvious pleasure. When the tasting is completed, he gives a
rating to describe his pleasure, on the same scale and the same landmarks as the
temperature expreriments of the previous pages. Plus two: very pleasant. Five minutes
later, we do it again, a second tofee; plus two again: very pleasant. Five minutes later
we do it again, then again five minutes later, and again, and again. Since the subject was
fasted, his successive ratings remain plus two, very pleasant, for a long time. But now
comes a plus-one rating, pleasant only. Soon, a zero follows; the toffee is indifferent, no
more pleasant but not unpleasant. Let us continue. After some more zeros, suddenly
the subject crosses the Rubicon river and decides, to his own surprise, to rate negatively
the sensation aroused by the toffee. After that decision which made him hesitate, the
subject feels more confident and, after some additionnal toffees does not hesitate to give
a rating of minus two, very unpleasant, to describe the sensation aroused by the sweet.
What happened? Was the stimulus different, to modify so drastically the
sensation? Impossible, these are industrially produced sweets; they are made by the ton
and their chemistry is fairly constant. What changed is the subject due to continuous
intake of identical tofees. Something inside his body has been modified and it is that
modification which has resulted in a pleasant sensation turning into an unpleasant one.
We must try to identify this internal signal. However there is one preliminary necessary
checking. We must verify that it is the repeated intake of toffee and not just the repeated
oral stimulation, the taste, that turned a pleasurable sensation into a displeasurable one.
For this purpose, it is necessary to normalize the experimental conditions, make them so
simple that they are absolutely reproducible, and only modify one parameter at a time.
Sucrose in water, always at the same temperature, in the same volume, but with variable
concentration shouldachieve that goal. After each stimulus, the subject will rince his
mouth with water after rating the sensation, and expectorate the rincing, in order to
prevent a possible drift of natural rincing by saliva secretion that might occur within a
session. Also, the subject will always be fasted since the last evening supper.
In careful observance of these conditions, the first thing to do is to reproduce
with sweet water the results of the toffee experiment. When the results are available,
they are identical to those with toffee: the intake of sweet water (50 ml with 10 g of
sucrose in water), first very pleasurable in fasted subjects becomes, when repeted, less
pleasant, then indifferent, then unpleasant, and eventually very unpleasant. However,
if the subject expectorates the sweet samples instead of drinking them, the pleasurable
sensation stays pleasant throughout all the session. This sets the hypothesis that
displeasure might be caused by repetition. The ingestion itself is responsible of the
transformation of pleasure into displeasure (Fig. V.1). Besides, the same phenomenon
takes place when a subject drinks all the sweet water in a single load, when the same
charge is injected directy into the stomac gasrtically directly into the stomach with an
esophageal tubing. This last experiment rules out deglutition as a factor of this
transformation of sensation. The change from pleasantness into unpleasantness is
caused by some internal modification of the subject.
Fig. V.1 Time course of the hedonic ratings reported by a fasted subject tasting sweet
water samples repeatedly. Pleasure (positive ratings) and displeasure (negative ratings)
are experienced in response to the same gustatory stimulus, a sample of sweet water
presented repeatedly every third minute. Solid dots: the subject expectorated the
samples after tasting; open symbols: the subject swallowed the samples and thus
accumulated a heavy sucrose load in his stomach. In the latter case, it can be seen that
the same sweet taste that aroused pleasure in the fasted subject aroused displeasure in
the satiated subject. From, Cabanac, 1971 34.
The research to identify the internal signal responsible for this transformation
from pleasure to displeasure took years and still goes on presently. Its description, will
be found elsewhere 35, and does not have its place in this little book. At this stage (For the
time geing), let us just draw the lesson learnt then when these experiments took place:
afasted subject sense a stimulus with nutritive value as agreable; after repeated intake
the sensation becomes disagreable. We find here the pattern obtained with thermal
sensation: a stimulus that facilitates homeostasis is pleasant, but becomes unpleasant
when it is useless or noxious. To seek pleasure leads the subject to a useful behavior,
34 M. Cabanac, Physiological role of pleasure, Science, 1971, 173: 1103-1107.
35 M. Fantino, Nutriments et alliesthésie alimentaire, Cahiers Nutr. Diétét. 1995, 30: 14-18.
intake of food when the subject is fasted. To avoid displeasure will also lead the subject
to a useful decision, toend a meal when satiated.
The same research, by Roland Duclaux 36 with olfactive stimuli, shows that only
the pleasure aroused by food stimuli is affected by the intake of a meal. The odors of
orange, of liquors, or of cheese, which are pleasant when fasted, become repugnant after
the subject is satiated 37. However, the non-food odors are not affected by the state of
satiety; smells of bleach, ink, or aftershave remain pleasant or indefferent and are not
changed by a meal.
The decrease of pleasure and the rise of displeasure of gustatory and olfactory
sensations participate in the subject's satiation, putting a quantitative term to food
intake. This phenomenon appears especially well suited to control food intake as only
sensations triggered by foods are modified. The usefulness of pleasure and displeasure
in olfaction and taste seems hence adapted in priority to determine the behavior of food
Fig. V.2 Figure 2. Sensation seen as a multidimensional event in response to a stimulus.
Quality, describes the nature of the stimulus; intensitity, describes the intensity of the
stimulus; hedonicity (pleasure or displeasure), describes the physiological usefulness
36 R. Duclaux, J. Feisthauer, M. Cabanac, Effet du repas sur l'agrément d'odeurs alimentaires et non-
alimentaires chez l'homme. Physiol. Behav. 1973, 10: 1029-1034.
37 'Satiation' is the dynamic process leading to 'satiety'.
(survival value) of the stimulus. Duration is added to the three other fundamental axes
of sensation. From, Cabanac, 1995 38.
Thus, we may draw a general lesson from this new observation: an identical
pattern predicts the onset of pleasure and displeasure in sensory modalities as different
as thermal, gustative or olfactive perceptions. A same given stimulus, thermal, gustatory,
or olfactory can be perceived as pleasant or unpleasant depending on the stimulated subject's
internal state. Because this seems to be a general phenomenon, it would be appropriate
and convenient to find a word for it. The advantage of a word, is that it permits to
avoid repeating a whole sentence each time we call the concept. With physiologist
Stylianos Nicolaïdis, we decide, after a long discussion, to coin the word alliesthesia.
The etymology of the word is Greek because both of us are loitered nostalgics of
our classical background; alliosis-esthesia means, literally, changed sensation. The word
indicates that the hedonic dimension of sensation canchange in respomnse to biological
modifications or good or bad experiences in the subject's past. The word is not yet
universally adopted, but can be found in some dictionaries and encyclopedias 39.
Negative alliesthesia is the evolution of pleasure towards less pleasure or towards
displeasure while the stimulus remains constant. Positive alliesthesia is the evolution
towards less displeasure or more pleasure, also in response to a constant stimulus.
We can now place sensory pleasure and displeasure, the hedonic dimension, in a
schematic descriptive model of sensation (Fig. V.2). Actually, sensation holds four
dimensions: duration, quality, intensity and hedonicity, but only three can be
adequately represented in a plane, a fourth one can be only suggested. Duration is
implicit. Quality is the mental analog for the nature of the stimulus that generates the
sensation. Intensity is the analog of the intensity of the stimulus. Hedonicity, pleasure
or displeasure, indicates the usefulness of the stimulus for the body. During alliesthesia,
sensation moves on the hedonicity axis.
It is reasonable to think that the central mechanism responsible for pleasure is not
specific of each sensation, thermal, gustatory, olfactory, but, on the contrary, that it is in
common to all sensations. The clinical examination of two patients who suffer from a
rare syndrome, congenital indifference to pain, give a good indication of this and will
confirm this intuition 40. These patients can feel the pains. Blindfolded, they can tell
which stimulus is given to them, "prick", "pinch", "burn", etc., but they never manifest
any sign of emotion, or nor do they respond with a retreat movement. During hot or
38 M. Cabanac, On the origin of consciousness, a postulate and its corollary, Neurosci. biobehav. Rev.,
1995, 20: 33-40.
39 e.g. Encyclopedia od Neurosciences refxxxxxxxxxx.
40 M. Cabanac, P. Ramel, R. Duclaux, M. Joli, Indifference à la douleur et au confort thermique. Presse
méd. 1969, 77: 2053-2054.
cold baths, they manifest no alliesthesia in response to thermal stimuli. In addition, they
seem to be devoid of any hedonicityity towards food and negative alliesthesia does not
take place after ingesting a glucose load. An interrogatory reveals their almost complete
inadaptation of their behaviour towards cold environment as well as good food. This
leads them to a behavior strictly limited to rationality. To clothe, they cannot rely on
their senses, every morning they must find out the outside temperature. Spontaneously,
they would eat for survival, not for the pleasure of enjoying the haute cuisine available in
Lyon. Their behavior tends to confirm a contrario the role of sensory pleasure in
adapting behaviour to physiological goals.
If alliesthesia manifests itself for sensations as different as thermal sensation,
taste, and olfaction, it is likely that all sensory pleasure exists to signal to the subject who
experiences it the presence of a stimulus to seek and consume, and displeasure a
stimulus to avoid for the well being of physiology. Of course, it is very tempting to
verify this hypothesis. Such a verification can be done by searching whether we find the
same laws binding thermal sensations with thermoregulatory behaviour on one hand,
and the sensations aroused by foods with ingestive behavior on the other hand. Now, in
the case of thermal sensation, the usefulness of a stimulus is judged by the central
nervous system in regard with the internal state of the body. The internal body
temperature is compared to the thermostat set-point, 37°C during good health and 38.2-
38.5°C for the feverish subject of Fig. IV.1. The error signal is therefore responsible for
the activation of the autonomic regulatory responses and of alliesthesia which will in its
turn arouse a thermoregulatory behaviour. If the same laws apply to alliesthesia in the
various sensory modalities, we may expect that an underlying regulatory mechanism
similar to that of temperature regulation operates with food intake. Some non-identified
and unknown error signal may work to determine the to-olfagustatory alliesthesia
coming forth with a meal in response to food stimuli. The English physiologist Hervey
pinpointed how body weight is constant during our adult life 41. Could it be that the
alliesthesia for food sensations be linked to our body weight?
41 G. R. Hervey, Regulation of energy balance, Nature, 1969, 223: 629-631.
PONDEROSTAT AND GENERALISATION
Today, the three of us start an experiment; there is Herbert Spector, an American
PhD who came for a two-year post-doctoral stay to work with us in Lyon, Roland
Duclaux who is finishing his medical study curriculum, and me. We test test gustatory
and olfacctory alliesthesia on ourselves and correlate the results with our own body
weights. There will be three series of measurements. The first one with be today. Then
we shall lose weight untill the second series of measurements. Finally, we recover our
original body weight for the last test. It is a simple experiment but requiring long
persistent effort. The purpose of the fist measure is just to taste a range of five sweet
stimuli and to describe the pleasure experienced on the scale the numerical landmarks
of which are now familiar to us: from -2, very bad to +2, very good 42. Afterwards, we do
the same operations with odorant stimuli. The subject smells a battery of eleven scent
bottles containing orange syrup more or less concentrated. The subject keeps his eyes
closed to better concentrate on the odour, inhales the uncorked flaccon, then gives a
rating of his pleasure or displeasure.
The results show that fasted subjects like the sweet sensation. The sweeter, the
better (Fig. VI.1, top). This is true also with the orange odour.
Once we have collected all this information, the subject drinks 50 g of glucose
dissolved in water, for a total volume of 200 ml. One hour later, the subject receives
again the whole battery of five sweet samples and eleven of orange odours, and gives a
rating of pleasure/displeasure for the sensation aroused by each stimulus. The results
show that the subjects do not like the sugary flavor, nor the orange odour anymore.
These sensations are now unpleasant. The ingestion of a gastric load of concentrated
glucose has produced in them a clearcut negative alliesthesia (Fig. 6-1 on top). This
result is not new, it is the simple replica of the experiments described in the previous
chapter; the protocol was slightly different, but the principle remains the same. Nothing
is new so far, but we are now coming to the heart of the experiment.
42 Of course, we shall try to operate as objectively as possible. Each subject is fasted since the last
evening. Successive sesssions will take place in identical conditions and at the same time of day. The
samples tasted will be at the same temperature and same volume. Each stimulus is presented 15 sec,
then the subject expectorates the solution. The various samples at different concentration of sucrose are
presented in random order to avoid any sequence effect; they are presented at regular intervals. After
the subject rates his pleasure or displeasure, the subject rinces his mouth with water a constant
temperature. Once a series of concentrations has been tasted, the whole range is tasted again to verify
Fig. VI.1 Verbal responses decribing the hedonic experience of two subjects receiving
sweet stimuli (left) and orange odours (right). In all six boxes the subjects were tested
before and after receiving a concentrated glucose load in their stomach, 50 g in 200 ml.
The upper two boxes were control sessions on ad-libitum fed subjects. It can be seen
that the glucose loads rendered unpleasant the alimentary stimuli, sweet taste and
orange odour. In the middle boxes the subjects had lost weight (5.2 kg, left; 3.0 kg, right)
from drastic hypocaloric diets during several weeks. In lean subjects the gastric load
has no more satieting effect. Several weeks later, the ad-lib fed subjects returned
spontaneously to their original weights; the gastric loads are efficacious again (lower
boxes). From, Cabanac, Duclaux & Spector, 1971 43.
All three subjects will have to lose weight, now. We decide arbitrarily at ten
percent of their body weight what the two heaviest participants have to lose, and five
percentfor the leanest who has not too much excess fat. Now, starts an interesting
adventure which provides another example of the advantage for an experimenter to be
his own subject. Until that day, I never really cared about my weight. It is stable, thus
confirming Hervey's observations. Without my noticing, meals succeed meals assuring
the satisfaction of my nutritional needs, quantitative and qualitative. It is, indirectly, the
evidence that behaviour is indeed at the service of physiology, without necessarily
reaching a high level of awareness. But now, everything is different. I must diminish
my present food intake from about 7000 kJ per day, to about 2000 kJ per day. Just to
give some landmarks of human needs of energy, a hard labor worker needs about 25 000
kJ per day; a sedentary worker, 7000 to 8000 kJ per day; and a child about 4000 kJ per
How can we know the energy of our food? It is necessary to estimate the content
of each dish and weight everything before eating. I must bring a scale and weigh all my
food at home and everywhere. This is the only acceptable way to estimate the energy
content of the meagre livelihood laying in my plate. I must also buy a personal scale to
weigh myself everyday, and check that I am losing enough weigh to reach the aim
defined for the experiment. On the first days, the results are spectacular. My weigh
collapses drastically and I expect to soon reach my goal. After a few days, however, I
am soon disillusioned. I live what the nutritionists know well when they treat obese
patients: my weigh decrease slows down and soon becomes almost nil from one day to
the other. Yet I do not eat anything, or at least I feel like I am eating nothing. I am
permanently hungry. During the night, I dream of gorgeing on mountains of food, a
feeling of guilt wakes me up, and to my relief I understand that I have not succumbed,
but I am still hungry. When I weigh myself, I take my watch off to gain, or to loose, a
few additional grammes. Although it is a scientific experiment, I catch myself to be
tempted to cheat by placing my feet on the edge of the scale because I noticed that the
reading was lower in that position. At last, fifty one days after starting the hypocaloric
diet, on the blessed day of 11 December 1970, the weight read on the screen of the scale
is what it needs to be; my goal is reached. The other two subjects have both lost ten
percent of their initial weight. We may enter the second step of the experiment, explore
the olfacto-gustatory alliesthesia, and, hopefully, end this draconian diet.
The rigorous reproduction of the control protocol (before body weight loss) gives
some interesting results (Fig. VI.1 in the middle). When we are fasted in the morning,
the ratings after sweet tastes and orange odours superpose to those obtained before the
loss of weight. The method is reliable, therefore, when the responses are stable more
than seven weeks after the initial measurements. However, the ratings obtained after
the intake of the fifty grammes of glucose are very different from those obtained seven
weeks earlier. The ratings after ingestion are identical to those before ingestion. The
drinking of fifty grammes of glucose does not change sensation. Alliesthesia is
vanished. We may end the starvation diet that we were kept for weeks.
To celebrate the end of our hypocaloric have lunch with a true banquet at
L'Auberge Savoyarde, an excellent Lyonaise restaurant neighbouring the Faculty of
Medicine. There, each of us eats effortless about 19 000 k/j in one single meal. From
this day, we are following an ad libitum intake, and Christmas and New Year period
does not really induce moderation on food consumption. Nevertheless, it will takes us
more than three months to return to our initial weight. This return is accomplished
effortless, without thinking. In the middle of March, that is it. It is now time to do our
third series of tests to verify whether our sensations are back to normal, like our
weights. Actually, The same protocol, sensations aroused by sweet samples, then by
orange odours when fasted, then after drinking fifty grammes of glucose in two
hundred millilitres, faithfully reproduces the pattern of sensations recorded five months
earlier. Alliesthesia for the sensations aroused by alimentary stimuli is back.
What is the teaching to be drawn from this experiment? The ingestion of glucose
is followed by a negative alliesthesia in the subjects whose body weight is normal , but
alliesthesia disappears in the subjects that have thinned. These results correspond to a
defense of body weight, similar to the defense of core body temperature. Everything
seems to be as if body weight was regulated; as as if a biological regulation of body
weight existed, similar to the biological thermostat which regulates temperature. To
describe this regulatory apparatus, we coin the neologism "ponderostat" 43, from pondero
(mass, weight), and -stat (stabiliser). Of course, it is a Greek-Latin barbarism, but a quite
convenient one. The subject who does not pay attention to his body weight eats ad
libitum and alliesthesia takes charge and sets a limit to the amount of energy ingested.
Spontaneously, his weight stabilizes at a value neighboring the set-point. When the
energy diet shows a deficit and the subject loses weight, an error signal from the body
weight occurs and slows down the onset of negative alliesthesia. This, naturally, leads
the subject to extend his food intake, without even being aware of it. The duration of
each meal is simply a little longer, satiety is delayed.
These results obtained on normal subjects allow us to formulate a hypothesis on
the nature of obesity. During the period when we drastically reduced our food intake,
we experienced the "Obese's agony", even though our initial body weights were far from
being that of obese subjects. We may wonder, therefore, if obesity is not due to a
resetting of the ponderostat set-point at a higher value. The obese person who fights his
obesity reduces his food intake and his body weight. Thus, he creates a distressing error
signal and each meal turns into a fight against this signal. On the other hand, the obese
person who pays no attention to his weight does not suffer from hunger. This
hypothesis will be elegantly verified by the Parisian physician Bernard Guygrand, who
will show that when obesity settles itself up, for still unknown reasons, the alliesthesia
for sweet stimuli disappears during the dynamic period when the patient gains weight.
At this stage, the patients have sensations and a behavior similar those of subjects who
are losing weight, as in the previous experiment. When obesity is constituted, as long of
course as the patient is not fighting against his disease, alliesthesia reappears. The
patient recovers his feeling of satiety at the end of a meal and a behaviour similar to
43 M. Cabanac, R. Duclaux, H. N. Spector, Sensory feedbacks in regulation of body weight: is there a
ponderostat? Nature, 1971, 229: 125-127.
those of normal subjects. The stability of our body weight can thus be explained by an
active mechanism of regulation where the ponderstat works in a way similar to that of
the thermostat. According to this analogy, obesity is similar to fever; in these two states,
the regulation operates with higher set-points than during good health.
It is likely that some mental anorexiae correspond to the reverse picture where,
for an unknown reason, the ponderal set-pont is strongly lowered, bringing the patient
to a discomfortless cachexia. However, a dramatic anecdote comes to mind to illustrate
the perreniality of the ponderostat. When the starving prisonners
Fig. VI.2 One subject is fed for weeks only with a very monotonous liquid food (left two
boxes). As a result he looses weight without suffering. It can be seen that the weight
loss, several weeks later, did not change the sensory response to a glucose gastric load:
the stimuli become unpleasant after the load. When the same subject, lost the same
weight from drastic restriction on normal food, he remained hungry all the time and the
gastric load could not render the sweet sensation unpleasant (lower right box). From,
Cabanac & Rabe 44.
of the concentration camps were liberated in 1944, they had to be locked up again to
control their food intake, so big was their hunger. Contrarily to the anorexic patients
with similar body weights, their ponderstat was normal and pushed them to eat
endlessly, which brought the first ones to death, before their liberators realised that it
was vital to limit their access to food.
Edward Rabe, a young American scientist on post-doctoral research stay in Lyon,
and I will try to verify whether the analogy between ponderostat and the thermostat
goes even further. We shall explore whether the peripheral sensory input -olfaction and
taste in the case of food intake- plays the same role as skin temperature in the
thermostat. In equation 4-1, Tbath, mean skin temperature, intervened to adjust the
thermostatset-point. We shall try to see whether the same phenomenon takes place with
the ponderostat. The experiment will be conducted on naive subjects, but we serve also
as subjects to experience it ourselves, too 44. During several weeks, we only eat Renutril
(TM), a synthetic food containing all the nutritive elements needed in a balanced diet,
but not very attractive, being without the rich flavours of regular foods. The goal is
precisely to diminish the sensory message associated to normal food intake. If the
ponderstat works like the thermostat, foods rich in flavours should elevate the set-point,
and foods with no flavor should bring it down. The obese's agony that we experienced
while we were doing the previous experiment with restricted diet, was due to the fact
that we were eating normal food at the family table, with a lot of flavours. But now,
with Ted Rabe, we shall ingest exclusively water and this synthetic food, for weeks; as a
result our food intake goes down and our body weight with it. Contrary to the forced
weight loss experiment, with normal foods, this time we lose weight, without effort,
without suffering. On the first few days, we drink five cans of the synthetic liquid food.
But, as time goes by, our intake goes down to only three cans per day. We are not
hungry and the process is effortless. In a few weeks, our weight loss is similar to that of
the previous experiment. The absence of hunger, the easy weight loss are explained
when we run the experiment to explore alliesthesia in response to sweet stimuli.
Alliesthesia is still present. The presence of alliesthesia, in spite of the large weight loss,
indicates that the body-weight set-point has dropped (Fig 6-2). The impoverishment of
the sensory input has, therefore, lowered the ponderstat set-point. The situation is
similar to what we know with the thermostat. A warming of the skin lowers the set-
point of the biological thermostat; a cooling raises it. Here, an impoverishment of the
food flavours lowers the ponderstat set-point; a sensory enrichment raises it.
44 M. Cabanac, E. F. Rabe Influence of a monotonous food on body weight regulation in humans.
Physiol. Behav. 1976, 17: 675-678.
Before we can reach a conclusion for this chapter, let us make two preliminary
remarks and list some a priori contradictions to the usefulness of pleasure. The first
remark is that experiments done on the experimenters themselves are always
questionable, for other scientists, and perhaps still more so for the experimentators
themselves. Biases, whether conscious or not, can intrude and blurr or even falsify the
results, or the interpretations of the results. It was extremely important, therefore, that
we have been able to reproduce these results with naive subjects who wew completely
foreign to the research and who ignored the working hypothesis of the above studies.
The second remark is that the concept of ponderostat does not imply that the
body weight itself is be regulated. On the contrary, all the available physiological
evidence indicates that all the physiological variables that are subject to regulation are
tensive variables. A tensive variable, like temperature, pressure, or concentration, is
characterised by the fact that if we sum up two identical systems, their added state stays
constant; if we mix two containers of water at 40°C, the temperature of the mixture
remains 40°C. The antonym of tensive is extensive, which characterises variables that
add, like heat, mass, and volume. Hence, in our body tensive variables like blood
pressure, glucose or calcium concentration in the milieu intérieur, or core temperature,
will be subject to regulation, but not extensive variables like the mass of blood, the mass
of glucose or calcium, or the amount of heat contained in our body. It is clear that the
mass of fat in the body is not a tensive variable. It is another intensive variable,
therefore, a correlate of body mass, that will be subject to the regulation that ends in
body weight stability. Nevertheless, I find legitimate to keep the notion of ponderostat,
because his weight is what is of interest to the patient and his physician when when
they are confronted with a weight pathology. Body weight measurement, in addition, is
easy and directly accessible. Finally, the real variable actually regulated is still unknown
Generalisation: Sensory Pleasure Shows Usefulness
This chapter illustrates the breadth, the extent, of the laws applying to sensory
pleasure. Indeed, we recognised with gustatory and olfactory sensations and food
intake, similar phenomena as those secured with thermal sensation and
thermoregulatory behaviour. In both cases, the seeking of pleasure and the avoidance
of displeasure leads to behaviours useful to recover and maintain homeostasis. In both
45 Several hypotheses link the amount of fat stored in the body (body mass) to tensive variables in the
body. Their description would be out of the scope of this book. let us simply recall that G. R. Hervey
(Regulation of energy balance, Nature, 1969, 223: 629-631) was the first to propose that the concentration
of a circulating steroid hormone could index indirectly the amount of fat contained in the body because
steroid hormones are liposoluble. It was also shown the the fat cells, the lipocytes, produce a hormone
the concentration of which indicates to the centers the amount of fat contained in the body
cases, alliesthesia adapts the hedonic dimension of sensation to the short term needs of
the body. In both cases, the internal signal is an error signal, meaning a difference
between the regulation set-point and the actual value of the given variable being
regulated. In both cases, pleasure is servo controlled to a long term regulation.
Similarities remain even if the variable actually regulated has not yet been identified in
the case of the ponderostat. A this stage, it is tempting to conclude that, in the sensory
domain, pleasant equals useful. We shall return to the optimising role of sensory
pleasure and on the tie beteen pleasure and usefulness, but we can already find in what
has preceded and in Addison disease good exemplification of this principle.
First, the spontaneous time course of body weight and of food intake of subjects
under a restrictive diet shows that indeed all goes as if behavior was in conformity with
the rule of pleasure meaning usefulness.
The behavior of subjects who had themselves control of the skin temperature of
their left hand brings an even more objective and stronger argument. We should
remember that these subjects were instructed to try to obtain the greatest pleasure out of
their hand. Maximization of pleasure makes them chose temperatures useful to their
homeostasis. In fact, the water circulating in the glove they use, does not just simply
stimulate the sensory receptors of the hand skin. It also extracts, or provides, heat from,
or to, the hand, depending on whether the water is cold or hot. The hand of a
hyperthermic subject can lose up to 70 w by simply lingering in cold, agitated water.
This is equivalent to all the metabolic heat production of a subject at rest. To seek this
little pleasure during hyperthermia, is thus sufficient to balance the whole heat
production of the rest of the body.
People suffering from Addison syndrome develop an appetite for salt that long
appeared to be abnormal. We know now that, due to a deficit in desexycorticosterone, a
suprarenal steroid hormone, their kidneys cannot retain their sodium which is lost in
urine, and hence they develop a deficiency in salt. When presented with salt solutions,
the more salty, the more they like it. In this case, to seek their pleasure saves their lives.
With Marc Fantino, we had the opportunity to verify the truth of this, at the Antiquaille
Hospital in Lyon, where professor Mornex had invited us to meet one of his patients.
The session took place with a battery of vials placed on a hospital chariot by the patient's
bed. The patient expectorated the salty samples in his room sink.
With admirable naïvety, in 1971 I sent the manuscript of a theoretical article
assigning sensory pleasure as a starter for behavior useful for physiology, to the very
prestigious journal: Science. The magazine being so multidisciplinary, I felt it would be
an appropriate support for an article half-way between physiology and psychology.
Odly enough, the article was accepted, even though it was only based on introspective
experimentation 46. After that, the Psychologists could not ignore sensory pleasure any
more, but they mostly focused on the olfacto-gustatory sensations, forgetting the studies
on thermal pleasure, my starting point, that allowed a broader viewpoint. This resulted
in a series of works, in the seventies, by various laboratories to verify my allegations. As
is often the case in such circumstances, the experimental conditions were not the same
as mine and the results were hence sometimes different; that brought a cloud of doubt
over my whole theory, which inconvenienced the behaviorist establishment. I had to
wait more than twenty years for all smoke to clear from confirmation by other
laboratories and the theory be accepted. It must be recognize, also, that we canfind a
priori contradictions to the idea that pleasure means usefulness of a stimulus and that
maximisation of sensory pleasure optimizes the the body functions.
Even a superficial examination permits us to pinpoint some apparent
contradictions to the law stating that pleasant equals useful. They deal mostly with our
eating habits that makes us dig up our grave with our teeth, with the use of
psychotropic drugs like caffeine, nicotine, alcohol, cocaine, heroin, all supposed to bring
pleasure but nevertheless toxic, and with the deliberate selection of unpleasant
behaviours, such as getting up early to go to work. Before trying to answer these
objections, it should be to pointed out that if our intuitive knowledge of the human
mind, based upon introspection, has not progressed since Cro-Magnon, our scientific
knowledge of it is only rudimentary. What seems to be, at first glance an objection, may
simply reflect our lack of knowledge of extended finalities. With evidences as
constraining as those developed in the preceding pages, it is allowed to take these
objections with a grain of salt (with wit? with humour?). nevertheless, tet us try to face
Eating habits. People eat what they like. Thus, they may exclude from their diet
some useful, or even necessary, nutriments and thus develop nutritional deficiencies.
We could think that these eating habits are an exception to the laws of useful pleasure.
However, alimentary preferences are acquired and then stabilised in young animals as a
function of their nutritional needs. Also, an adult animal who comes across a new
stimulus tha might become food, is cautious before consuming it. It is only when the
post-ingestive effects of a minute ingested sample are beneficial that the animal takes a
habit in them. The stabilisation of food preference is beneficial thus when preventing
intoxication by unknown foods with new flavors. This natural mechanism may be
fooled simply by culinary arte facts, which can uncouple the sensory signal from its
46 Years later, one of the anonumous reviewers who had recommended that Science should publish my
article, met in a congress, confessed that the reading of my manuscript had been for him "a bowl of fresh
air". At the time I did not understand why. Later on, I learned that behaviorism then dominated
experimental psychology, up to the point of choking it.
biological meaning. We can therefore, that In human kind's the natural environment
the statement that sensory pleasure means usefulness is valid.
Another objection, psychotropic drugs . The case with drugs resulting in addiction is
neighboring, but a little more complex. Is drug consumption an exception to the law of
useful pleasure, since we find no usefulness for them? It is obvious that the relation
between pleasure and useful behavior escapes a superficial examination in the case of
the behaviours leading to the consumption of psychotropic drugs. We may oppose four
answers to this objection:
1) From a Darwinian point of view, it is not necessary that a mechanism be
hundred percent favourable to give a selective advantage to its owners. All that is
needed is that one trait be slightly better in helping the chances to survive of an
organism and for the trait be transmitted to future generations. In this light, we may
compare pleasure with curiosity. Nobody can deny the great selective advantage that
curiosity is to obtain information on food sources, good environments, presence of
predators, and reproductive partners. Yet, curiosity can end badly if the prey lurks
around too close to the predator. It could be similar with pleasure. It would suffice
have that pleasure be statistically advantageous to justify it's existence. But I think the
connection is tighter.
2) The existence of heart diseases does not invalidate the vital function of this
organ. A pathology of the machanism of pleasure would not invalidate the usefulness
of pleasure. Mental pathology, masochistic, sadistic, violent behaviors, as well as drug
use may all be considered as pathological distortions of the fundamentally useful
mechanisms that have permitted the animals, and then the humans, to survive through
the millions of years of evolution.
3) We must distinguish the first try of a drug from a settled toxicomania. The first
contact is not out of pleasure, but out of curiosity. Even though curiosity is noxious in
this case, nobody can deny the survival value of curiosity for the individual and the
species. It is the chronic use of psychotropic drugs that can result in an addiction about
which we may question whether pleasure means usefulness. Once toxicomania is
installed, the trap is closed, the subject is prisoner of his drug, coffee, tobacco, or hard
drugs, but it is too late. Drugs might activate nervous paths and centres, that are basis
to pleasure. What we have here, therefore, is an arte fact similar to the use of food
additives like saccharine, that carries a signal devoid of nutritional contents.
4) Finally, we may, with prudence, defend the idea that there is some usefulness
in toxicomania. The post-ingestive effects of occasional alcohol, may be considered as
beneficial if this drug -rich in metabolisable energy, in addition- temporarily lifts social
inhibitions or psychological discomforts. The smoker who enters into a room where
everybody turn around to look at him, finds useful to bring out a cigarette, to pout a
little with it, and casually to light it up. The teenagers couple, a little scared on their first
date, find very useful the little rituals of tobacco smoking that help to avoid the
intimidating glance of the partner, and screen their paralysing emotion.
From drugs bringing heavy addiction, like nicotine 47, the withdrawal syndrome is
so uncomfortable that the drug may be understood as useful, on the short term, by the
user because it ends the major discomfort of the syndrome, rightly considered
pathological in the view of the user.
On a more serious tone, we may discern in some toxicomanias, a last resort before
committing suicide for some hopeless people. If the drug contributes to help alleviate a
desire of death in the intoxicated user, its noxiousness is more social than individual. In
this perspective, farthing from death may be considered useful. It is obvious that this
short-term usefulness is paid by a long-term addiction that the user cannot estimate.
Choice of disagreeable and painful conducts. The mere saying that pleasure could be
the ultimate motor to all behavior brings naturally to mind one last apparent
contradiction: the case when we decide to produce a painful behaviour. Examples of
behaviours with unpleasant consequences, but nevertheless chosen, are the innumerable
results the necessities of life. Everybody knows the weariness of working, the
frustration of altruism, the fatigue oftraining for a competition, or the torture of a hard
diet to loose weight. Yet, our day-to-day life is filled with these choices of unpleasant
conducts, that seem to indicate that in 'true life' we meet with more displeasurable-
usefulness than pleasurable-usefulness. Could it be that we do not seek pleasure? Or
that pleasure is not useful? The answer to these questions will follow in later chapters;
but we first must make a little excursion into Ethology and meet the "Behavioural final
47 The pharmacological effects of nicotine are lessviolont than those of other drugs like cocain or herain,
but the very strong dependence towards nicotine makes of it a hard drug.
THE BEHAVIOURAL FINAL COMMON PATH
Ethology is the science of behavior stemmed from zoology. Two great postulates
are at the basis of Ethology. According to the first postulate, animal behavior is optimal
by definition since it was produced by natural selection over millions of years. For the
Ethologists, optimality is estimated from the reproductive efficiency of an individual
and a species. Such a definition of Optimality is therefore defined on the long-term,
privileged view point for the Zoologists, but does not exclude the short term, province
of the Physiologists. Behavioral motivations are numerous as survival from predators,
reproduction, alimentary needs, most of temperature regulation, need to sleep are all
satisfied behaviorally. The British ethologists, McFarland and Sibly 48 used the
expression "behavioural final common path" to describe and underline that many
different motivations converge towards a single means of satisfaction. Such a term is
especially eloquent for a physiologist. Indeed, it points towards the beautiful homology
between behavior with the basic organization of the nervous system responsible for the
motor response. The great English physiologist was first to name "final common path",
the alpha moto-neurone whose cellular body is located in the ventral horn of the spinal
chord and on which converge multiple excitatory and inhibitory paths. In the same way
as the alpha-moto-neurone integrates these influences and generates only one command
to the peripheral muscle, either inhibition or excitation, in the same way behavior
integrates simultaneously contradictory motivations that cannot be satisfied
simultaneously (e.g. the need to sleep with the need to eat) and must manage a large
mass of additional signals before acting. This illuminating image leads to the second
According to the second postulate, at each instant the animals satisfy behaviorally
their most urgent motivation. When they theorized on the "behavioural final common
path", McFarland and Sibly stated that the central nervous system responsible for
behavior, must rank on a priority scale all motivations simultaneously present according
to their urgency. To achieve such a ranking the brain must dispose of a common
currency. Thus, a trade-off will be able to take place among motivations, one being
downgraded for the benefice of another one that will rank first. Animals occasionally
renounce to feed because of the presence of a predator, or of a reproductive partner. In
that case immediate survival or reproductive necessity were ranked higher than the
necessity to feed. The common currency makes possible the trade-off between
motivations and the ranking according to the urgency of the need to access to behavior.
48 D. J. McFarland, R. M. Sibly. The behavioural final common path. Philos. Trans. roy. Soc. London ,1975
The reading of McFarland and Sibly's "behavioural final common path", then that
of other published works of McFarland's gave me [N.B. I am not sure of whether present or
past tense is best in English. In French I used the present tense in order to be both more factual
and more vivid. M.C.] a strong intellectual excitement because these texts succeed in
unifying biology to economical science through their border line disciplines, ethology
and micro-economics; the latter is the branch of economical science dealing with the
behavior of individual persons or agents and the principles of their decision making.
Yet the situation was still more stimulating because it led me to have an inkling of the
implication of a third order of science, psychology. Sensory pleasure might be the
common currency postulated by McFarland and Sibly. Let us examine further this latter
McFarland and Sibly's theoretical analysis is satisfying for the ethologists. Such
an analysis shows its excellency when, and by, permitting a mathematical description of
behavior. However, if the theory describes from outside, phenomenologically, how
decisions are made, it does not inform on the mental cognition of the decision making
process. Yet, the process of decision making is a mental one. McFarland and Sibly
explain the how of behavioral decisions, but there remains a problem to understand the
why of behavior. Now, we just saw in the preceding chapters that sensory pleasure
appeared to be linked to the physiological needs of the body, that the pursuit of sensory
pleasure leads to behaviors adapted to satisfying the body's physiological needs.
Sensory pleasure might answer the question left open by the Ethologists in the
theorization on the behavioral final common path and the common currency. Might it
be that sensory pleasure would be the common currency that allows the trade-off
among motivations, and their ranking by order of priority? From a theoretical point of
view it is permitted to accept the hypothesis.
The reasoning goes as follows:
Sensory pleasure permits the optimization of behavior by triggering
useful behaviors. Yet, if the first postulate of ethology, according to
which behavior is always optimal, is true sensory pleasure should
optimize any function that ranks first on the ladder of the motivations
ranked by order of priority. Pleasure thus should optimize various
motivations. If this it so, then the next fundamental question is: might
pleasure be the common currency postulated as necessary on
According to such a hypothesis, it may be accepted that the trade-off between
motivations takes place simply by comparing the magnitude of the various pleasures
aroused if satisfying the various competing motivations. The motivation whose
satisfaction will fulfil the strongest pleasure will be satisfied first. The second rank will
be occupied by the motivation whose satisfaction will provide the highest remaining
The theory permits to render count also of the selection of behaviors that procure
displeasure. The affective dimension of any perception being the common currency
would allow to associate a behavior producing pleasure to another behavior producing
displeasure. The decision would take place according to the result of the simple
algebraic sum of pleasures and displeasures aroused by various behaviors, as in the
theoretical example of figure VII.1.
If a behavior 1 arouses a pleasure that increases from p to P, the subject will tend
to achieve such a behavior which procures pleasure.
If a behavior 2 arouses a displeasure that increases from d to D, the subject will
tend to avoid such a behavior which procures displeasure.
However, if behavior 1 and 2 are associated, yoked, the procurement is both P
and D, the subject will tend to achieve these yoked behavior if in the affective result the
pleasure of P+D is larger than the pleasure of p+d (or the displeasure smaller). In the
opposite case, of course the subject will abstain.
Behavior 1 p ------> P yes
Behavior 2 d ------> D no
+ p + d -----> P + D yes
with pleasure P > p, and displeasure D > d
and with total pleasure P + D > p + d
Figure VII.1: Theoretical presentation of two behaviors 1 and 2 and on how a behavior
that arouses displeasure may be chosen by a subject. A behavior (behavior 2) that
produces displeasure can be chosen by a subject if another behavior (behavior 1) that
produces pleasure is simultaneously chosen. The necessary and sufficient condition for
the behavior 2 to occur (action) is that the algebraic sum of affective experience
(pleasure) of the yoked behaviors is positive (P+D>p+d). Capital letters P and D
indicate larger affective experience than respective small letters p and d. From,
Cabanac, 1992 49.
49 M. Cabanac. Pleasure: the common currency. J. theoret. Biol. 1992 155:, 173-200.
This association of a positive element providing pleasure with a negative element
lacking pleasure evokes irresistibly other but similar processes taking place in living
beings. It is common knowledge that the living beings are capable to synthetize energy-
rich molecules in order to store this energy for future use. Such a synthesis is
spontaneously improbable because by costing energy to the environment it apparently
violates the second principle of thermodynamics. Living cells nevertheless synthetize
routinely energy-rich molecules 50. The transfer of energy and its storage in energy-rich
molecules is possible only because a larger amount of energy is liberated simultaneously
from the destruction of other energy-rich bonds. It is the yoking of an exothermic
reaction with an endothermic reaction that allows the storage of energy to take place (N.B.
It should be recalled that the overall balance of the yoked reactions nevertheless
augments total entropy of the system in its environment, as nature does not violate the
second principle of thermodynamics).
One can recognize some analogy between cellular thermodynamics and
behavioral decision making processes. In a manner similar to energy transfer, the
yoking of two behaviors, one arousing pleasure and the other displeasure permit
behaviors that arouse displeasure and therefore that are spontaneously improbable to
occur because the net balance is positive in favor of pleasure.
It remains to verify whether that theoretical construct resists experimental
verification, both in physiology and in psychology. The experiments devised to probe
the theory will therefore leave the study of pleasure in only one sensory modality, to
explore the field of multidimensional pleasure. In these experiments several sensations
will be controlled simultaneously and the resulting algebraic sum of pleasures will be
correlated with the actual behavior chosen by the subject. These experiments were
thought out after a slow ruminating on the theoretical implications of the behavioural
final common path and the common currency, over several years. They were started
only after my moving to Québec in 1980.
50 See for example A. L. Lehninger, Bioenergetics. The molecular basis of biological energy transformations.
1965 New York, W. A. Benjamin Inc.
The notion of sensory pleasure being the common currency between motivations
has been posted in the preceding chapter from a limited number of experiments
regarding exclusively olfacto-gustatory and thermal sensations. This was limited field,
therefore. In order to test the hypothesis, other experiments should be done in other
sensory fields. Further, it would be desirable to stimulate simultaneously incompatible
motivations, thus forcing the subject to rank them for access to the behavioural final
common path. Only results that have to do with pleasure and displeasure of very
different sensations will allow to generalise the previous conclusions. For this purpose,
we shall now explore the various sensations aroused by muscular exercise. Three
situations of conflicting motivations will be studied in the following pages: fatigue vs.
against thermal discomfort, fatigue in the chest vs. fatigue in the lower limbs, and sweet
taste vs. sauer taste. In each experiment, pleasure of one sensation will only be obtained
a the cost of displeasure in another sensation.
Fatigue vs. Thermal Discomfort
In a climatic chamber, a subject wearing only sports shoes and bathing trunks
walks on a motor-powered treadmill. The speed is fixed at 3 km/h and doesn't change,
but the ambient temperature of the room and the slope of the treadmill can be changed
by the experimentor. There will be three series of one-hour sessionss. The first serie of
sessions, will consist of twenty five one-hour walks. During each of these twenty five
sessions the subject will walk at imposed ambient temperature and treadmill slope.
These two parametres will systematically combine 0, 6, 12, 18, and 24% slopes, with 5,
10, 15, 20, and 25°C ambient temperatures. All twenty five sessions will thus differ from
one another, either from the slope or from the ambient temperature. The result is a two-
dimension matrix or graph of the twenty five different conditions with physical exercise
on the Y axis going from light to exhausting, and ambient temperature on the X axis
going from frankly cold to rather warm.
Near the end of each session, we ask the subject rate not his sensations but the
pleasure or displeasure of them. The subject, thus, gives one rating for the pleasure or
the displeasure of his physical exercise (Y axis) and a second rating for the pleasure or
the displeasure of his temperature (X axis). On the graph describing the twenty five
sessions, we draw the dots that the subjects gives us about the his hedonic feelings in
both dimensions, fatigue and thermal comfort. What is to be checked is whether
behaviour is brought about by the algebraic sum of pleasures and displeasures (x+y).
The sum of both ratings is put on the diagram for each of the
Fig. VIII.1 Behaviour tends to place a subject in pleasurable area of a motivational
space. This figure superimposes results obtained in one subject who walked at a
constant 3 km.h-1 on a treadmill over several sessions.
a) The iso-clines were obtained in the first series of sessions as follows: a
combination of treadmill slopes and ambient temperature (Ta) was imposed, and the
subject rated separately his discomfort aroused by fatigue and by cold. The lines show
the algebraic sum of ratings triggered by the various combinations of environmental
temperature with intensity of exercise; positive ratings indicate a pleasurable sum,
negative ratings indicates an unpleasant sum. The lines are therefore iso-hedonic lines.
b) The dots were obtained in the second series of sessions. They show the actual
locations in the map sought by the subject in ten sessions. Either ambient temperature
was imposed on the subject (5, l0, l5, 20, or 25°C) and he could select slope (H) i.e.
intensity of exercise, or the slope of the treadmill was imposed (0, 6, 12, 18, or 24 percent
slope) and the subject could choose environmental temperatures (J). For example, when
Ta was imposed at 10°C, the subject requested to raise the treadmill up to 23.5 percent
slope and the H is located in the white area (pleasure) beyond the + 0.5 pleasure incline.
When the treadmill was imposed at 0 percent slope, the subject requested Ta to be
raised to 26°C and the J is located in the white area (pleasure). From, Cabanac &
LeBlanc, 1983 51.
twenty five sessions. The overall result is a 'geographycal map' 52, giving us a map
showing the subject's overall pleasure and displeasure on a two dimension-type sensory
space (Fig. VIII.1).
Throughout the second series of sessions, we shall compare the subject's
spontaneous behavior with his xy map of bi-dimensional pleasure-displeasure. One of
the dimensions will be imposed by the experimentator and the subject will adjust
himself the second dimension. First, the treadmill slope is imposed to the subject; thus
the amount of heat produced in his muscles is imposed. The subject is left with free
access to the thermostat and the ambient temperature, for one hour walk at constant
speed; thus, he will be able to lose more or less heat into the environment. On five
sessions, we impose the slopes and he raises or lowers ambient temperature
temperature ad libitum .
Eventually, in the last series of walking, we place the subject in the reverse
situation: the experimenter imposes the ambient temperature and the heat loss or gain
from the environment, and the subject choses the treadmill slope and the rate of heat
produces in his muscles, this for five sessions, all at different ambient temperatures.
Now, we can indicate, as dots, the subject's behavioral choices for these last ten
sessions (two series of five sessions) on the map of bi-dimensional pleasure/displeasure
(Fig. VIII.1, b). Triangles show the slopes chosen by the subject when ambient
temperatures were imposed; and circles show the ambient temperatures selected when
slopes, and thus the intensity of heat produced within the body were imposed.
Two main conclusion may be drawn from the above results. First, both clusters
of dots are similar, if not superimposed. Triangles and circles belong to the same
population. The subject fights his hyperthermia by lowering his environmental
temperature and hypothermia by raising the intensity of his muscular work. He
actuates indifferently either one of the variables at hand, x or y, to produce the same
result. Second, all dots but one fall in the white area of the map, i.e. in the zone where
the algebraic sum of pleasure plus displeasure in temperature and exercise is positive.
What is the actual location of the three dots left out of the white, pleasure, area on the
map? It can be seen that even the last dot is placed in the area of minimal displeasure;
the subject did not have the freedom to modify the 8% slope that was imposed on him
by the experimenter. Even though the algebraic sum is slightly negative in these
51 M. Cabanac, J. Leblanc, Physiological conflict in humans: fatigue vs cold discomfort. Am. J.
Physiol.1983, 244: R621-R628.
52 Once the ratings are reported on the 'map,,
particular point, it is likely that the subject minimized the negative algebraic sum. We
can see on the map that the dot is located in the prolongation of two white peninsulas of
the positive zone. In the whole totality of this experiment, behavior is well determined,
therefore, by a tendency to maximize the algebrical sum of pleasure plus displeasure,
whenever that is possible.
Chest vs. Lower Limbs
While perfecting the protocol of the previous experiment, I am the subject of pilot
studies imposing on myself to walk on the treadmill at various slopes and speeds. I
notice that fatigue of walking on the treadmill takes place mainly in two areas of my
body: the chest and the lower limbs. When the speed of the treadmill accelerates, an
unpleasant feeling, then real painappears in the muscles of the legs and thighs. When
the slope of the treadmill rises, these discomforts do not show immediately, but in their
place unpleasant sensations of short breath and cardiac thumping start to arise in the
chest area. With a little closed-eyes concentration, it becomes easy to focuse on my body
sensation and distinguish the feelings from both areas. In turn, it is then easy to rate
independently the pleasures, or here rather the displeasures aroused in these different
parts of my body. Now, with two different sources of hedonic feeling, it becomes
possible to do an experiment to verify iwhetherf these pleasures or displeasures add
algebraically, as was the cas in the preceding experiment. Exactly the same
experimental scheme can be used. The first step consist in establishing the xy map of
pleasure and displeasure, lower limbs vs. chest, in a series of sessions with imposed
slopes or imposed speeds. Then, in a second series of sessionss, the speed of the
treadmill is imposed and the subject can change the slope. Finally, in the last series of
sessions, the slope is imposed and the subject can adjust the speed. In the sessions when
the subject can adjust the intensity of his exercise by playing with either slope or speed,
he is given the task of climbing an elevation of 300 m. Then the session ends.
The results show that the chosen slopes and speeds all fall one single population
of dots (Fig. VIII.2), whether the subject can manipulate the speed of the treadmill (y
axis), or its slope (x axis). In the three boxes of the figure, the dots showing a subject's
behavioural choices regarding slope and speed, are superimposed with the maps of
hedonic ratings; only the iso-hedonic lines for low intensities of displeasure are drawn
on the picture. In the top box, the behavioural results are compared with of the
threshold (-1) and clearcut (-2) displeasure in the lower limbs. The behaviour shows
some correlation with the lines of displeasure in the lower limbs, but this is obviously
not a very high correlation. In the middle box of the figure, the behavioural results are
compared with the lines of displeasure in the chest area. Again, here there is some
correlation, but here also this correation is not very high. However, when, in the lower
box, the dots which show the subject's behaviour are compared to the isohedonic lines
showing the algebraic sum of displeasure ratings in the lower limbs plus in the chest
(simple sum of the ratings shown in the two above boxes) the correlation turns out
excellent and statistically highly significant. We must remember that the cloud of eight
dots incorporates the sessions when the subject could manipulate slope and those when
he could manipulate speed. Thus, subjects play with the variable they have access to,
either x or y, to minimize the integrated sum of their hedonic perceptions.
Fig. VIII.2 Subjects sum al