Memory is the collection of records of perceptions. The production of such records is
the essential aspect of observation. Records can be stored in human memory, i.e., in the
brain, or in machine memory, as in computers, or in object memory, such as notes on
paper. Without memory, there is no science, no life – since life is based on the records
inside the DNA – and especially, no fun, as proven by the sad life of those who lose their
Ref. 710 memory.
Obviously every record is an object. But under which conditions does an object qualify
as a record? A signature can be the record of the agreement on a commercial transaction.
A single small dot of ink is not a record, because it could have appeared by mistake,
Motion Mountain – The Adventure of Physics available free of charge at www.motionmountain.eu Copyright c Christoph Schiller November 1997–January 2009
812 30 the story of the brain
for example by an accidental blot. In contrast, it is improbable that ink should fall on
paper exactly in the shape of a signature. (The simple signatures of physicians are obviously
exceptions.) Simply speaking, a record is any object, which, in order to be copied,
has to be forged.More precisely, a record is an object or a situation that cannot arise nor
disappear by mistake or by chance. Our personal memories, be they images or voices,
have the same property; we can usually trust them, because they are so detailed that they
cannot have arisen by chance or by uncontrolled processes in our brain.
Can we estimate the probability for a record to appear or disappear by chance? Yes, we
can. Every record is made of a characteristic number N of small entities, for example the
number of the possible ink dots on paper, the number of iron crystals in a cassette tape,
the electrons in a bit of computer memory, the silver iodide grains in a photographic negative,
etc.The chance disturbances in any memory are due to internal fluctuations, also
called noise. Noise makes the record unreadable; it can be dirt on a signature, thermal
magnetization changes in iron crystals, electromagnetic noise inside a solid state memory,
etc. Noise is found in all classifiers, since it is inherent in all interactions and thus in
all information processing.
It is a general property that internal fluctuations due to noise decrease when the size,
i.e., the number of components of the record is increased. In fact, the probability pmis
for a misreading or miswriting Challenge 1203 ny of a record changes as
pmis ∼ 1/N , (500)
where N is the number of particles or subsystems used for storing it.This relation appears
because, for large numbers, the so-called normal distribution is a good approximation of
almost any process; the width of the normal distribution, which determines the probability
of record errors, grows less rapidly than its integral when the number of entities is
Challenge 1204 ny increased. (Are you able to confirm this?)
We conclude that any good record must be made from a large number of entities.The
larger the number, the less sensitive thememory is to fluctuations. Now, a system of large
size with small fluctuations is called a (physical) bath. Only baths make memories possible.
In other words, every record contains a bath.We conclude that any observation of a
system is the interaction of that system with a bath. This connection will be used several
times in the following, in particular in quantum theory.When a record is produced by a
machine, the ‘observation’ is usually called a (generalized) measurement. Are you able to
Challenge 1205 s specify the bath in the case of a person looking at a landscape?
From the preceding discussion we can deduce a powerful conclusion: since we have
such a goodmemory at our disposition, we can deduce that we are made of many small
parts. And since records exist, the world must also be made of a large number of small
parts. No microscope of any kind is needed to confirm the existence of molecules or
similar small entities; such a tool is only needed to determine the sizes of these particles.
Their existence can be deduced simply from the observation that we have memory. (Of
course, another argument proving that matter is made of small parts is the ubiquity of
Page 256 noise.)
A second conclusion was popularized in the late 1920s by Leo Szilard.Writing a memory
does not produce entropy; it is possible to store information into a memory without
increasing entropy.However, entropy is produced in every case that thememory is erased.
Motion Mountain – The Adventure of Physics available free of charge at www.motionmountain.eu Copyright c Christoph Schiller November 1997–January 2009
the story of the brain 813
It turns out that the (minimum) entropy created by Challenge 1206 ny erasing one bit is given by
Sper erased bit = k ln 2 , (501)
and the number ln 2 ≈ 0.69 is the natural logarithm of 2. Erasing thus on one hand
reduces the disorder of the data – the local entropy–, but on the other hand increases the
total entropy.As is well known, energy is needed to reduce the entropy of a local system.
In short, any systemthat erasesmemory requires energy. For example, a logical AND gate
effectively erases one bit per operation. Logical thinking thus requires energy. It is also
known that dreaming is connected with the erasing and reorganization of information.
Could that be the reason that, when we are very tired, without any energy left, we do not
Challenge 1207 s dream as much as usual?
Entropy is thus necessarily created when we forget.This is evident when we remind
Ref. 715 ourselves that forgetting is similar to the deterioration of an ancientmanuscript. Entropy
increases when the manuscript is not readable any more, since the process is irreversible
and dissipative.* Another way to see this is to recognize that to clear a memory, e.g. a
magnetic tape, we have to put energy into it, and thus increase its entropy. Conversely,
writing into a memory can often reduce entropy; we remember that signals, the entities
that write memories, carry negative entropy. For example, the writing of magnetic tapes
usually reduces their entropy.
The capacity of the brain
“Computers are boring.They can give only
answers.
(Wrongly) attributed to Pablo PicassoThe human brain is built in such a way that its fluctuations cannot destroy its conte”nts.
The brain is well protected by the skull for exactly this reason. In addition, the brain
literally grows connections, called synapses, between its various neurons, which are the
cells doing the signal processing. The neuron is the basic processing element of the brain,
performing the basic classification. It can only do two things: to fire and not to fire. (It
is possible that the time at which a neuron fires also carries information; this question is
Ref. 718, Ref. 719 not yet settled.)The neuron fires depending on its input, which comes via the synapses
from hundreds of other neurons. A neuron is thus an element that can distinguish the
inputs it receives into two cases: those leading to firing and those that do not. Neurons
are thus classifiers of the simplest type, able only to distinguish between two situations.
Every time we store something in our long term memory, such as a phone number,
new synapses are grown or the connection strength of existing synapses is changed. The
connections between the neurons are much stronger than the fluctuations in the brain.
Ref. 716 * As Wojciech Zurek clearly explains, the entropy created inside the memory is the main reason that even
Maxwell’s demon cannot reduce the entropy of two volumes of gases by opening a door between them
in such a way that fast molecules accumulate on one side and slow molecules accumulate on the other.
(Maxwell had introduced the ‘demon’ in 1871, to clarify the limits posed by nature to the gods.) This is just
another way to rephrase the old result of Leo Szilard, who showed that the measurements by the demon
Ref. 717 create more entropy than they can save. And every measurement apparatus contains a memory.
To play being Maxwell’s demon, click on the www.wolfenet.com/~zeppelin/maxwell.htm website.
Motion Mountain – The Adventure of Physics available free of charge at www.motionmountain.eu Copyright c Christoph Schiller November 1997–January 2009
814 30 the story of the brain
Only strong disturbances, such as a blocked blood vessel or a brain lesion, can destroy
neurons and lead to loss of memory.
As a whole, the brain provides an extremely efficient memory. Despite intense efforts,
engineers have not yet been able to build a memory with the capacity of the brain in
the same volume. Let us estimated this memory capacity. By multiplying the number of
neurons, about 1011,* by the average number of synapses per neuron, about 100, and also
by the estimated number of bits stored in every synapse, about 10, we arrive at a storage
capacity for the brain of about
Mrewritable ≈ 1014 bit ≈ 104 GB . (502)
(One byte, abbreviated B, is the usual name for eight bits of information.)Note that evolution
hasmanaged to put asmany neurons in the brain as there are stars in the galaxy, and
that if we add all the synapse lengths, we get a total length of about 1011 m, which corresponds
to the distance to from the Earth to the Sun. Our brain truly is astronomically
complex.
In practice, the capacity of the brain seems almost without limit, since the brain frees
memory every time it needs some new space, by forgetting older data, e.g. during sleep.
Note that this standard estimate of 1014 bits is not really correct! It assumes that the only
component storing information in the brain is the synapse strength. Therefore it only
measures the erasable storage capacity of the brain. In fact, information is also stored in
the structure of the brain, i.e., in the exact configuration in which every cell is connected
to other cells.Most of this structure is fixed at the age of about two years, but it continues
to develop at a lower level for the rest of human life.Assuming that for each of the N cells
with n connections there are f n connection possibilities, this write once capacity of the
brain can be estimated as roughly N
√
f n f n Challenge 1208 ny log f n bits. For N = 1011, n = 102, f = 6,
this gives
Mwriteonce ≈ 1016 bit ≈ 106 GB . (503)
Ref. 720 Recent measurements confirmed that bilingual persons, especially early bilinguals, have
a higher density of grey mass in the small parietal cortex on the left hemisphere of the
brain. This is a regionmainly concerned with language processing.The brain thusmakes
also use of structural changes for optimized storage and processing.
Incidentally, even though the brains of sperm whales and of elephants can be five to
six times as heavy as those of humans, the number of neurons and connections, and thus
the capacity, is lower than for humans.
Sometimes it is claimed that people use only between 5% or 10% of their brain capacity.
This myth, which goes back to the nineteenth century, would imply that it is possible
to measure the actual data stored in the brain and compare it with its capacity to an
impossible accuracy. Alternatively, the myth implies that the processing capacity can be
measured. It also implies that nature would develop and maintain an organ with 90%
overcapacity, wasting all the energy and material to build, repair and maintain it. The
myth is wrong.
*The number of neurons seems to be constant, and fixed at birth.The growth of interconnections is highest
between age one and three, when it is said to reach up to 107 new connections per second.
Motion Mountain – The Adventure of Physics available free of charge at www.motionmountain.eu Copyright c Christoph Schiller November 1997–January 2009
the story of the brain 815
The large storage capacity of the brain also shows that human memory is filled by the
environment and is not inborn: one human ovule plus one sperm have a mass of about
1mg, which corresponds to about 3 ⋅ 1016 atoms. Obviously, fluctuations make it impossible
to store 1016 bits in it. In fact, nature stores only about 3 ⋅ 109 bits in the genes of
an ovule, using 107 atoms per bit. In contrast, a typical brain has a mass of 1.5 to 2 kg,
containing about 5 to 7 ⋅ 1025 atoms, which makes it as efficient as the ovule. The difference
between the number of bits in human DNA and those in the brain nicely shows that
almost all information stored in the brain is taken from the environment; it cannot be of
genetic origin, even allowing for smart decompression of stored information.
In total, all these tricks used by nature result in the most powerful classifier yet
known.* Are there any limits to the brain’s capacity to memorize and to classify? With
the tools that humans have developed to expand the possibilities of the brain, such as paper,
writing and printing to help memory, and the numerous tools available to simplify
and to abbreviate classifications explored by mathematicians, brain classification is only
limited by the time spent practising it. Without tools, there are Ref. 721 strict limits, of course.
The two-millimetre thick cerebral cortex of humans has a surface of about four sheets of
A4 paper, a chimpanzee’s yields one sheet and a monkey’s is the size of a postcard. It is
estimated that the total intellectually accessible memory is of the order of
Mintellectual ≈ 1GB , (504)
though with a large experimental error.
The brain is also unparalleled in its processing capacity. This is most clearly demonstrated
by the most important consequence deriving from memory and classification:
thought and language. Indeed, the many types of thinking or language we use, such as
comparing, distinguishing, remembering, recognizing, connecting, describing, deducing,
explaining, imagining, etc., all describe different ways to classify memories or perceptions.
In the end, every type of thinking or talking directly or indirectly classifies observations.
But how far are computers fromachieving this! The first attempt, in 1966, was
a programming joke by JosephWeizenbaum: the famous chatterbot programEliza (try it
at www.manifestation.com/neurotoys/eliza.php3) is a parody of a psychoanalyst.Eventoday,
over 40 years later, conversationwith a computer program, such as Friendbot (found
at www.friendbot.co.uk), is still a disappointing experience.We need to understand the
reasons for this slow development.
Curiosities about the brain
The brain plays strange games on the people that carry it.Modern research has shown
that students can be distinguished into five separate groups.
1. Strong students
2. Uninterested students
3. Students that overestimate themselves (often, but not always, boys)
4. Students that underestimate themselves (often, but not always, girls)
5. Weak students
* Also the power consumption of the brain is important: even though it contains only about 2%of the body’s
mass, is uses 25% of the energy taken in by food.
Motion Mountain – The Adventure of Physics available free of charge at www.motionmountain.eu Copyright c Christoph Schiller November 1997–January 2009
816 30 the story of the brain
This has to be kept in mind when teaching classes. To which group do/did you belong?
* *
Teachers should all be brain experts.The brain learns best when it has an aim. Howmany
teachers state the aim of their class at its beginning? Students need different motivations:
potential applications, curiosity, competition, activation of already acquired knowledge,
impressing the opposite sex, or exploring the unknown. And they need it on different
levels of difficulty. Which teacher provides this mix? Students have different ways to create
concepts: using words, sounds, images, emotions, body sensations, etc.Which teacher
addresses them all in his lessons?
…. recently watched a movie , a Bollywood movie (i must not shy while telling its a Bollywood movie in spite of Nepali movie coz every one of my age prefer hollywood or Bollywood movie to nepali movies) named ‘Rock Star’ where the meaning to the rock star was unlike what we think of real rock star..Rockstar myt have been influenced by pains or some sorts traumas but does that mean anybody wid painful hearts get voice to be a rockstar?? Had that happened many of us would have been a celebraty…And like these rock stars need to fuck unconditionally to keep them going…what the hell to the meaning of ‘Rock star’…Showing middle fingers at the public doest show that u hav performed as good as Jim Morrison…
