If the first view were tenable, if order prevailed throughout the whole material universe, if it were manifest wherever
the laws of physics operate, I should have no reason to doubt
the materialist when he asserts that all order is wholly the
result of the operation of those laws. But I have shown in
the last lecture that the materialist's faith in the ability of the
laws of physics to create order is unsound and refuted by
facts. We have to conclude that there is no order in the
rough untouched world of lifeless things.
I propose to show in this lecture that the second view, the
doctrine of universal randomness, is equally untenable. So
we are left with the third view, according to which some
things are ordered and some not. My immediate task will be
to consider the nature of order and to show why a philosophy
cannot be sound in which the reality of order is disregarded.
But before I do this I must define what order is.
Of several objective and scientifically valid criteria by
which one can distinguish between random and ordered
systems or events I shall have time to mention only three.
They are based respectively on considerations of probability,
on the second law of thermo-dynamics, and on the ways in
which predictions are made about random and ordered
systems.
In this example one may judge from results whether a
principle of randomness or a principle of order has applied
to the sequence in which the cards were dealt. When a deal
shows a freak distribution players often draw the conclusion
that the pack was not well shuffled. If a poker player has the
best cards in his hand whenever he is the dealer one may
justifiably suspect that a selective principle has replaced a
principle of randomness. Thus card players are quite familiar
with the first criterion for distinguishing between order and
randomness. It is the frequency with which certain specific
configurations occur, in the above example the configurations presented by the hands dealt.
But in the other world, the world that is touched by life,
events do not justify any such assumption. Selective principles
do operate. A motor-car may serve as our first example. Its
assembly cannot be attributed to a principle of randomness.
The components are not arranged in the factory like cards
in a well-shuffled pack. Their movement from the store to
the assembly shop cannot be described as just flying about.
They do not merely shake down into the finished vehicle.
If the assembly of a motor-car depended on a principle of
randomness it would be a very rare event, rarer than any
freak hand at bridge. Even if we did not know that the
assembly was the work of workmen who select, control,
guide the course of events, the simple fact that motor-cars
are not at all rare would lead us to the conclusion that their
assembly did not come about by chance.
This argument is not new, but it is sometimes woefully
misunderstood. Earlier in the present century it was rather
fashionable to misapply it. The favoured illustration at that
time was not a motor-car but the works of Shakespeare. It
was customary to say that, given time enough, a monkey at a
typewriter, tapping the keys at random might produce the
complete works of that poet. The argument was used in
support of the theory that all order results from the operation
of physical laws that themselves are non-selective and represent only a principle of randomness. More recently
analogous theories about the way the human brain works
have gained some favour. The notion is that the brain incorporates a principle of randomness, a principle that, like
the monkey in the other illustration, is incapable of selecting,
guiding, controlling anything, and that this random element
is required to produce the ordered performance of the brain.
The theory is a little surprising to an engineer; for no
designer of man-made machinery would expect to improve
his design by incorporating a principle of randomness into it.
However, I doubt whether that effort to present order as
created by a random environment will be any more long-
lived than previous similar theories have been. Some very
simple facts will, I think, prove too strong for that very
simple faith.
The facts are, let me repeat, the frequency with which
specific configurations occur. The monkey of chance would
deal any bridge hand one might care to specify once in a .
while. But it would deal every conceivable other hand as well.
It might perhaps type out Shakespeare's works once in a
while during the whole of eternity, but it would produce
every other combination of the letters of the alphabet as well.
Now we do have all conceivable bridge hands and identical
hands are very rarely repeated; so we conclude that they
result from randomness. We do not have every conceivable
combination of letters in the books that appear and publishers
do print off thousands of identical volumes in one edition.
Here again the repetition of identical sequences of letters
and the absence of all the alternative sequences would suffice
to prove that the works of Shakespeare have not, in fact,
been produced by the monkey of chance.
Those who introduced this monkey into the argument did
not use it to explain the works of man but the occurrence of
living organisms. But that does not make it any more logical.
To say that once in the remote past atoms of the elements
that constitute protoplasm would, if in random motion, reach
every conceivable configuration including, once in a long
while, that of living substance would explain perhaps the
occurrence of one morsel of protoplasm. But that is not what
has to be explained. There are more oak leaves in the world
than motor-cars. They are all very much alike. If one were
to calculate the probability that atoms in random motion
would form such a configuration one would reach a figure
vastly less than the probability of any particular hand at
bridge. It is because of the frequency with which oak leaves
occur that they cannot be attributed to a principle of randomness. The rule for applying this criterion can be expressed
quite simply. Estimate the frequency with which an observed
configuration would occur on the assumption that it was the
product of randomness. Then estimate or count the frequency with which the configuration does occur. If the latter
is much greater than the former, one has to conclude that a
selective principle operates. By this criterion all organic
structures are ordered and all structures in the rough untouched world of lifeless things are random.
Water in a mountain lake, for instance, contains potential
energy by virtue of its position. It loses this as it flows down
the valley, and what appears in its place is kinetic energy
represented by turbulence in the water. Similarly steam
under pressure in a boiler contains some potential energy. If
the steam is blown off to atmosphere it loses this and what
appears instead is kinetic energy in the form of turbulence
in the air. In these examples the conversion from one form of
energy to another is immediate. And it is almost complete; If
the water is not allowed to spend its energy in creating swirls
and eddies, but is guided through a water turbine, the conversion is delayed and incomplete. Some potential energy can
be retained, though in another form, perhaps as electricity.
The same thing can be made to happen if the steam is
guided from the boiler to a steam engine. In both examples
the reason why a rather large fraction of the initial potential
energy becomes available is that the working fluid is not
allowed to move entirely at random but is guided, controlled,
subjected to order.
It will be noticed that order is not linked with potential
energy as such but with the process by which it is converted
to kinetic energy. If those who have said so carelessly that
potential energy is the same thing as order had spoken instead of its differential coefficient in respect to time they
might perhaps have arrived at a more correct statement.
Predictions in Physics and in Biology
The last of the criteria for which there is time is provided
by the methods that prove successful in making predictions.
Let us consider first what methods are employed in physics
and then what are employed in biology.
First be it noted that most of the observations made by
physicists are not on random but on ordered systems.
Scientists would not get far if their only laboratory were the
rough untouched world of lifeless things. In a laboratory in
which useful work can be done there must be order. And
this is the reason why precise predictions are made there.
One may predict, for instance that in a given circuit the
ratio of current to voltage will confirm Ohm's law; but only
if the materials have been properly selected and the circuit
carefully constructed to give that result. In a laboratory all
random events that might affect readings are excluded;
circuits are shielded from stray magnetic fields; chemical substances are so prepared as to be free from chance impurities;
temperatures are controlled; draughts are excluded. The
predictability of a result in a laboratory experiment is a
measure of the degree of order maintained. The precision
with which experiments are repeatable does not prove that
it is in the nature of matter to behave in an orderly manner
but only that it is in the nature of scientists to do so.
This is why we cannot use methods of making predictions
as a means of distinguishing between order and randomness
if we consider only predictions about systems subjected to
the order of a laboratory technique; we must consider random
systems. So let us turn our attention to some examples.
Neutrons in a lump of uranium may serve. They move
in all directions, undergo many random collisions, and yet
what happens in that lump can be predicted with great
accuracy. The reason is in what is called the law of averages.
There are so many neutrons in a sizeable lump that their average behaviour is consistent. A freak distribution is possible,
just as a freak distribution of cards at bridge is possible. But
it is so improbable that one need not reckon with it.
It is the same with the individual molecules of a substance
in a test tube. The substance has been submitted to a diathesis; it has been carefully prepared, measured into the test
tube in a specified quantity, mixed with measured quantities
of other pure substances, maintained at a specified temperature. But the individual molecules are uncontrolled; they are
in random, erratic motion. Yet a chemist who knows what
the substances are can predict what will happen in the test
tube with great accuracy. This is only because he relies on
the law of averages. His prediction will only be confirmed if
the behaviour of the molecules is random. A Clerk Maxwell
demon who got inside the test tube and introduced some
sort of order into the behaviour of the molecules would falsify
the prediction; unless, of course, the chemist knew what the
demon was up to.
The point of these illustrations is that one can predict the
result of the average behaviour of a number of particles in
random motion if one has selected the particles and knows
what they are. But can one predict the paths that will be
taken by individual particles? Can one predict the configuration that they will assume at any future moment of time?
One can, of course, if the movement is not random and if
one knows the kind of order that is applied. Thus a skilful
billiard player can predict with reasonable accuracy what
configuration will be assumed by a selection of billard balls
after he has propelled them. But can one do the same when
things are merely flying about? The answer is neither a
simple 'no' nor a simple 'yes'.
An eclipse can be predicted with great precision and I
fear that some will take this as proof that the movement of
the stars is ordered and not random. But that is not the true
reason. The system in question includes virtually only three
bodies: the sun, the earth, and the moon. For the effect of
other heavenly bodies is very slight. These three bodies move
in only one kind of field of force, a gravitational one. They
vary greatly in size, so that the sun's field has a predominating effect and the effect of the moon on the earth's orbit is no
more than what calls for a small correction to the orbit that
would be followed if there were no moon. The three bodies
are widely separated, so that the gravitational field near each
is almost symmetrical around a centre; it is, in other words,
an almost uniformly divergent field. The three bodies are
too far apart ever to collide.
All this means that the system composed of sun, earth,
and moon is a peculiarly simple one; it is close to what might
be called an ideal or a formal system; in this exceptional
instance a system is to be found in the rough untouched
world of lifeless things of the kind that scientists must
usually construct laboriously for themselves. The solar
system has, moreover, been studied for a long while and the
fields in which the bodies move have been charted with
great accuracy. Those are the reasons why an eclipse can
be predicted with precision. The reason is emphatically not
that the solar system conforms to any principle of order.
Indeed it can be shown that if there is no principle to prevent
the three bodies from following a random movement they
must follow the observed orbits.
Let now some different system be imagined; one in which
the number of heavenly bodies is increased ever so slightly,
be it only from three to four. Let their sizes be approximately
equal. Let them be so close that collisions can occur. Then
their relative positions at any moment of time are still determinate; but no human mathematician could calculate those
positions. When many particles are jostling and colliding
in many kinds of field of force and when they are subjected
to indiscriminate influences from neighbouring bodies the
configuration that they adopt at any given moment may still
be virtually determinate, at least if the particles are large, but
in practice it is very, very far from being predictable.
This is why a meteorologist, for instance, can at best foretell
the weather a few days ahead and then only with considerable
uncertainty. He may be fairly sure that the sea will be rough
at a certain place and a certain time. But he could not say
when a particular wave will break at a particular spot along
the beach. He may know that there will be clouds in the sky.
But he cannot predict their exact sizes, shapes, or number.
The physical sciences are rightly said to be precise sciences.
But one must understand correctly what that statement
means. Its justification is only very rarely provided by any
observation outside a laboratory. In the physical sciences
predictions are very precise about things that have been
made by man, such as the experimental gear that he sets up
in his laboratories; they are usually very imprecise about
things that are only found in rough nature, such as clouds and
ocean waves.
Let us now turn our attention to another illustration of
prediction. Take a physicist to a place in the country and tell
him that where he stands the ground contains an object with
a specific structure; it is not man-made; it has been found
there. Its millions of small component parts can move and
change their relative positions. It is surrounded by earth,
water, and air. Nothing will be done to shield it from its
environment. Then ask the physicist to examine and measure
the object carefully and to predict what configuration its
components will adopt after a stated passage of time. The
answer that he must give as a physicist is obvious. He will
say that it is difficult enough to predict the future configuration of only three bodies when they are moving under the
influence of the forces exerted by each on the other two and
that when there are millions of them the task is quite beyond
him. He will remind you that you cannot even tell him what
the unshielded influences are going to be that may bring
sundry forces to bear on the component particles of the
object and he will insist that these will have a decisive effect.
Now let us imagine that you have brought another companion with you. He is a botanist and recognizes the object
as an acorn. He can tell you that an oak-tree is quite likely
to stand in that place in ten years' time To make this prediction he does not need to examine the relative positions of
the atoms of which the acorn is composed; the method by
which an astronomer predicts eclipses is not a part of his
technique. He has not measured out the ingredients of the
acorn as a chemist measures the substances put into a test
tube when he is making a repeatable experiment. The
botanist may hardly know what all the substances in the
acorn are and he does not use such knowledge as he may have
about them when he predicts that it will grow into an oak-tree. The botanist does not say, as the physicist must, that he
can make no prediction without knowledge of the forces that
will be brought by the environment to the delicate and complicated system that he is studying. He knows that the effect
of these forces is not usually decisive. The basis of predictions made in botany is totally different from that for predictions in the physical sciences.
The predictions are also more detailed and precise. While
a meteorologist can predict vaguely that there will be some
clouds and a geologist that a landslide will produce a heap of
debris at the foot of a cliff, the botanist can predict exactly
what the shape and size of the oak leaves will be. He knows
how the branches will grow, what the bark will be like, how
the leaves will unfold from their buds in the spring, when
flowers and new acorns will appear. He can even predict
what minute detail of structure will be observed if any morsel
of leaf, bark, or stem is looked at under a microscope. He
knows what chemical substances will be found in each kind
of structure, how many chromosomes each cell will contain,
how chains of amino-acids will be orientated.
Here we have on the one hand the extremely imprecise
predictions made in meteorology and geology; the most apt
generalizations that can be made about clouds and heaps of
debris is that they represent random configurations. On the
other hand, we have the extremely precise predictions that can
be made in the biological sciences. Between these extremes,
but with a wide gap on each side, are the predictions that can
be made in a laboratory or an engineering shop. They are,
as I have said already, about things made and not about
things found. When a substance has been carefully prepared,
a certain degree of purity ensured, and the temperature and
other experimental conditions are controlled, a chemist can,
for instance, predict that a certain kind of crystal will precipitate out of a solution. When an engineer is shown the
blueprints and specification for a motor-car he can predict
what will come off the assembly line. To do so he does not
have to calculate the forces that will propel the cover on to
the cylinder and each nut on to its bolt. He knows that the
fitter will bring the right kind of force into action. The
knowledge that enables the engineer to predict that a motorcar will come off the assembly line is basically unlike the
knowledge that enables an astronomer to predict an eclipse
and basically like the knowledge that enables a botanist to
predict an oak-tree. It is not knowledge of forces, movements, masses. It is not detailed knowledge of existing configurations. A motor-car will result whether the bolts were
stored in bin A or in bin B or came from a neighbouring
town. The acorn will germinate whether the earth was a little
dryer or a little wetter, whether the wind that brought its
carbon as carbon dioxide in the air blew from the west or
from the south. The configurations presented by motor-cars
and oak-trees are not critical to external circumstances as
those presented by bodies in random movement are. This is
the reason why it is possible to predict those configurations
correctly without precise knowledge of the external circumstances. The knowledge that enables a person to predict a
motor-car or an oak-tree is knowledge of the laws, the rules,
the specification, whatever one likes to call the principle
that must be met so that the observed order may occur.
The third criterion by which to distinguish between random and ordered systems and events can now be stated. If
an event can only be predicted with the help of knowledge of
the positions and motions of the particles concerned and of
the forces brought to bear on those particles the event is a
random one. If it can be predicted without that knowledge
the event is an ordered one. This criterion is as objective as
the other two. It does not depend on any subjective judgement. By this criterion all living substance reveals order; and
this order is more complete, more detailed by far, than the
order ever found in the world of man-made machines or even
in the best conducted laboratory.
Let us turn to the second of the theories often found
attractive. The number of people who like to think that the
whole of reality conforms to one single basic principle may
not be great; but again I can understand such a point of view.
If it were true the universe would be agreeably simple,
comprehensible. We might hope, once we had discovered
that single, unifying principle, that we should know everything, a thought that flatters human vanity. Progress in
physics, moreover, has raised men's hope that it may be so.
In the rough untouched world of lifeless things there is
probably one such basic principle. I explained in the second
lecture that since the time of Newton a great simplification
and unification of physical science has been achieved. We
know now that it is not correct to think that there is one law
for pendulums, another for the falling of apples, another for
planets, yet another for rain clouds. All those statements that
are called laws in physics are special cases of more basic
principles. I also suggested that they can probably all be
unified as expressions of the most basic principle of all, a
principle that might be formulated as the law that there are
no laws.
It is among things that are touched by life, among plants
and animals and among the many things that man, the tool-using animal,
produces that this great, simple, unifying,
negating principle does not hold. In this other world things
do conform to laws. And the laws are not of the unifying
type, be they man's or Nature's. The rule of the road differs
between countries; each town council makes its own by-laws;
each school of painting follows a different aesthetic principle;
each species of trees grows in a different way. The notion of
uniformity has its attraction. The popular appeal of totalitarianism in politics proves it. I can understand why it is so
often supported by faith. But it cannot be supported by any
facts that I have been able to discover.
The third of the cherished beliefs, namely that space is
the container of all active reality is certainly held by that most
fallible instrument, the judgement of common sense. Another way of formulating it is to say that what is must be
somewhere. The notion that I had to put before you in the
first lecture of a non-material influence without location is
probably by far the biggest obstacle to acceptance of any
dualistic philosophy. That diathetes, be they called God, life,
mind, or the soul, can be nowhere and yet do things is a most
puzzling concept. It does provide the theologian with a
difficult problem. But it is not one of doctrine, it is one of
presentation. It arises, as I said in the first of these lectures,
because religion has to be conveyed to ordinary persons,
who have but little philosophical training. The theologian
may be able to refute but he is not allowed to ignore the
judgement of common sense. Though this judgement is such
a slender reed, religion must often lean on it.
This is why a preacher may himself not think it necessary
to regard space as the container of all active reality and yet
find it difficult to answer the question where God is. He
often seeks an easy way out of the difficulty by declaring that
God is everywhere. But if one of his parishioners, old Mrs.
Smith say, asks him where her own soul is will he tell her
that it is everywhere? Does he mean literally that God is
everywhere, that like a tenuous gas He is spread evenly over
the whole of space? Of course not. I am sure that the trained
theologian does not interpret the word 'everywhere' literally.
He is well aware of the distinction between being everywhere and acting at every place. When he says that God is
everywhere he means that God is capable of exerting His
influence everywhere. And when he says that old Mrs.
Smith's soul is in her body he means that it can control every
part of her body. It is not relevant in religion whether this
control is exercised from a particular place, or from every
place, or from no place.
In science it is easier today to abandon the conviction that
space is the container of all active reality than it would have
been at the beginning of this century. For many of the views
about space that are held by common sense have had to be
rejected. Among them is the notion that it is something
absolute, something conceptually distinguishable from
matter. To a modern physicist space itself has physical properties. He no longer thinks it quite precise to say that matter
is in space; he prefers to regard it as coincident with space. To
a physicist space is not a container at all. He thinks of space
as something more limited, in concept as well as in extent;
and that allows him to think of realities that do not form a
part of space. But to accept all that is not to deny that
the notion of an active influence without location calls for a
very substantial intellectual effort. As I have said before
one cannot achieve a coherent philosophy without working
for it.
Lastly there is the theory that a non-material influence
necessarily acts consciously. This view is very ingrained.
The growth of a tree is a completely unconscious process,
and from this it is often concluded that it is not controlled
by any non-material influence, but is the result of the unaided
action of matter on matter. A similar conclusion is often
reached about the instinctive behaviour of man and the
lower animals. When a bird builds a nest or a man performs
some involuntary movement the act is said, even by those
who support some form of dualism, to be caused in the same
way as the orbit of a planet, the falling of a stone, a storm at
sea, the passage of water in a mountain stream. Dualists,
oddly enough, often agree with monists that there is no basic
distinction between events in the rough untouched world of
lifeless things and the great majority of events to be observed
in the organic world. Only on the rare occasions when an
action is performed consciously, they say, does it reach a
level where it can only be explained as controlled by a non-material influence.
Such a view may pass the test of religion; for religion is
only concerned with actions for which the individual can be
held responsible. But I have shown in these lectures that
it cannot pass the test of science. A sense of proportion might
have led us to the same conclusion. Mind, at least conscious
mind, is a very late arrival in the history of evolution. It
controls but a minute fraction of the organic world. Even
those of man's activities that are fully deliberate are by no
means fully conscious. When a person raises an arm, though
it be a quite voluntary action, he has no conscious knowledge
whatever of the behaviour of the millions of component parts
of that wonderful mechanism, the human brain; he does not
know what signals reach his muscles through what nerves;
he does not consciously select, co-ordinate, control the
thousands of muscle fibres that perform their complicated
drill so that his arm may come to be raised. Even when we
are conscious of what we do, we remain profoundly unconscious of how we do it. Besides, when we raise an arm the
action is far more often unconscious than conscious. Even in
man consciousness is but a fitful affair. Our basic condition
is unconscious.
It is irksome to have to admit this; for we like to think
that we know much more about our actions than we can ever
hope to know. It is also unflattering to human vanity to have
to regard man's most distinctive possession, his conscious
mind as important only to himself and as no more important
in the general scheme of things than any feature of some
lower form of life that has proved its survival value.
The notion that any process that is completely vegetative,
like the growth of an oak-tree, can be controlled by an influence that is wholly devoid of consciousness is also incompatible with the judgement of common sense. To that
superficial arbiter it seems impossible that any influence could
achieve, without being conscious of its activities, what no
man could do when fully awake and alert. So common sense
prefers to assume that the growth of an oak-tree is not controlled at all.
Such are among the reasons why a philosophy in which
consciousness is presented as of cosmic significance and
every unconscious process is attributed to the unaided action
of matter on matter is more comforting than the more
austere, but I venture to claim also more coherent, philosophy that I have put before you. In conclusion let me summarize this in as few words as possible.
There are cogent, indeed irrefutable, reasons for the
assumption that some events are random and others ordered.
Matter is, by its nature, incapable of creating order. So
only the random events can be attributed to the unaided
action of matter on matter. When order is observed, its
cause must be some non-material influence. A convenient
collective name for all the non-material influences that may
have to be postulated is diathetes. As space is inseparable
from matter, diathetes cannot have their existence in space.
A number of purely objective facts provides evidence of the
reality of one particular class of diathete, often called life.
The evidence shows that this exercises so detailed a control
over living substance that the arrangement of individual
atoms is subjected to a specified order. Except in the rarest
instances the diathete life operates completely unconsciously.
The exceptions are a small fraction of the activities of man
and, perhaps, of some of the higher animals.
As a palliative to the emotional and intellectual unattractiveness of some features of this philosophy let me, in
conclusion, add one consoling thought. If consciousness
has but little cosmic significance, it has at least done for
Man something more and finer than to ensure his survival
in the struggle for existence. It has provided for him a
window, though it be but a little one, through which he can
look on another world, on a world that differs from the one
that is presented by his organs of sense perception; and
through that tiny square Man gains what has never been
seen before in the whole long history of evolution, a fleeting
glimpse of self.
Top of Page
