Memory, p.30
Memory, page 30
Yet another observer began very confidently: ‘It is the interior of a house. There are three figures. One is tall, the second less tall, the third less tall still. The figures are leaning against a pillar or wall. Probably I am looking at a copy of some old master. It might well be “The Woman Taken in Adultery”.’ He soon withdrew this, and said that there were two figures only. He thought that the picture was one of Charles the First and Henrietta, and to this view he adhered throughout the remaining observations, though with no great certainty.
From KONRAD LORENZ, King Solomon’s Ring (1949), translated by Marjorie Kerr Wilson (1952)
In a territory unknown to it, the water-shrew will never run fast except under pressure of extreme fear, and then it will run blindly along, bumping into objects and usually getting caught in a blind alley. But, unless the little animal is severely frightened, it moves, in strange surroundings, only step by step, whiskering right and left all the time and following a path that is anything but straight. Its course is determined by a hundred fortuitous factors when it walks that way for the first time. But, after a few repetitions, it is evident that the shrew recognises the locality in which it finds itself and that it repeats, with the utmost exactitude, the movements which it performed the previous time. At the same time, it is noticeable that the animal moves along much faster whenever it is repeating what it has already learned. When placed on a path which it has already traversed a few times, the shrew starts on its way slowly, carefully whiskering. Suddenly it finds known bearings, and now rushes forward a short distance, repeating exactly every step and turn which it executed on the last occasion. Then, when it comes to a spot where it ceases to know the way by heart, it is reduced to whiskering again and to feeling its way step by step. Soon, another burst of speed follows and the same thing is repeated, bursts of speed alternating with very slow progress. In the beginning of this process of learning their way, the shrews move along at an extremely slow average rate and the little bursts of speed are few and far between. But gradually the little laps of the course which have been ‘learned by heart’ and which can be covered quickly begin to increase in length as well as in number until they fuse and the whole course can be completed in a fast, unbroken rush.
Often, when such a path-habit is almost completely formed, there still remains one particularly difficult place where the shrew always loses its bearings and has to resort to its senses of smell and touch, sniffing and whiskering vigorously to find out where the next reach of its path ‘joins on’. Once the shrew is well settled in its path-habits it is as strictly bound to them as a railway engine to its tracks and as unable to deviate from them by even a few centimetres. If it diverges from its path by so much as an inch, it is forced to stop abruptly, and laboriously regain its bearings. The same behaviour can be caused experimentally by changing some small detail in the customary path of the animal. Any major alteration in the habitual path threw the shrews into complete confusion. One of their paths ran along the wall adjoining the wooden table opposite to that on which the nest-box was situated. This table was weighted with two stones lying close to the panes of the tank, and the shrews, running along the wall, were accustomed to jump on and off the stones which lay right in their path. If I moved the stones out of the runway, placing both together in the middle of the table, the shrews would jump right up into the air in the place where the stone should have been; they came down with a jarring bump, were obviously disconcerted and started whiskering cautiously right and left, just as they behaved in an unknown environment. And then they did a most interesting thing: they went back the way they had come, carefully feeling their way until they had again got their bearings. Then, facing round again, they tried a second time with a rush and jumped and crashed down exactly as they had done a few seconds before. Only then did they seem to realise that the first fall had not been their own fault but was due to a change in the wonted pathway, and now they proceeded to explore the alteration, cautiously sniffing and be-whiskering the place where the stone ought to have been. This method of going back to the start, and trying again always reminded me of a small boy who, in reciting a poem, gets stuck and begins again at an earlier verse.
From COLIN BLAKEMORE, Mechanics of the Mind (1977)
Penfield hoped to discover in each patient an area in the brain which, upon stimulation, would arouse the same curious mental aura that ushered in the epileptic attacks. That would be the spot to destroy, Penfield argued; and this approach was remarkably successful. But it also gave him the chance to discover the functions of the other parts of the cerebral cortex. Stimulation of the motor cortex caused jerky twitches of the muscles which the patient could not control; excitation of the touch area produced strange sensations felt by the patient in his skin; stimulation of the visual cortex made the patient see flashes of light or swirling coloured forms in his visual field. But when Penfield moved his stimulating electrode to the temporal lobe and the hippocampus itself, the experiences of the patient were not mere fragments of movement or sensation. They were whole episodes of existence, plucked from the patient’s previous life. The person would suddenly be transported into the past and would feel himself eavesdropping on a familiar scene.
One of Penfield’s patients was a young woman. As the stimulating electrode touched a spot on her temporal lobe she cried out, ‘I think I heard a mother calling her little boy somewhere. It seemed to be something that happened years ago … in the neighbourhood where I live.’ A moment later the same spot was stimulated. ‘Yes,’ she said, ‘I hear the same familiar sounds; it seems to be a woman calling; the same lady.’ Then the electrode was moved a little and she said, ‘I hear voices. It is late at night, around the carnival somewhere – some sort of travelling circus. I just saw lots of big wagons that they use to haul animals in.’
There can be little doubt that Wilder Penfield’s electrodes were arousing activity in the hippocampus, within the temporal lobe, jerking out distant and intimate memories from the patient’s stream of consciousness.
Memory, its physical structure, is an unsolved challenge. It is, perhaps, the central question, rather like the problem of the structure of DNA, deoxyribonucleic acid, for molecular biology and genetics. But theories about the nature of memory have still not really progressed beyond the stage of description through analogy. Analogy has often been a valuable step in the discussion of biological problems, but it is, of its nature, constrained by the technological development of the time or the level of scientific knowledge in other fields. The restrictive nature of argument by analogy is aptly illustrated by historical models of memory. They are almost all based on the devices used by man himself to store information.
According to Aristotle, sensory impressions entered the head with such force that they left physical inscriptions in the brain, like a scribe engraving on a wax tablet. This idea, that the mind is a tabula rasa on which experiences are literally written, was espoused by the Empiricist school of philosophy. ‘Let us then suppose the mind to be, as we say, white paper, void of all characters, without any ideas;’ wrote John Locke in 1690, ‘how comes it to be furnished? Whence comes it by that vast store, which the busy and boundless fancy of man has painted on it with an almost endless variety? … To this I answer in one word, from experience.’
Even current models of memory dwell on analogies with existing, artificial methods of storing information. Mental memory has been compared with the electromagnetic polarisation of ferrite rings in the core memory of a computer, and with the ‘distributed’ image of a hologram – a device that stores a record of a three-dimensional scene by photographing the interference pattern produced by illuminating it with laser light.
Each analogy has a certain attraction because it mirrors some particular special feature of memory. The permanence of lines scratched in a wax tablet or written on paper mimics the durability of real memory. The speed of access to the memory of a computer core is reminiscent of the remarkably rapid way in which neuronal memory can be consulted.
Because the information in a holographic place is ‘distributed’, a somewhat degraded reconstruction of the entire stored image can still be retrieved even when part of the plate is destroyed. Now the psychologist Karl Lashley, working in the first half of this century, described a similar kind of resistance to local injury in the store of information in the rat’s brain. The actual representation of remembered events is almost certainly in the cerebral cortex, but Lashley found that small areas of damage in the rat’s cortex simply blurred the animal’s ability to perform tasks that it had previously learned. The degree of degradation of memory was roughly proportional to the area of damaged cortex. He concluded that the cerebral hemispheres have a kind of ‘mass action’ in the remembering process. In the same way, the steady attrition of about 50,000 nerve cells that unavoidably die each day in the cerebral cortex of man does not rob us, piecemeal, of individual elements of memory; it gradually removes the edge from remembered events and stunts the power to capture new ones. Lashley expressed his frustration in failing to track down the physical substance of individual stored remembrances in a famous scientific paper in 1950. ‘I sometimes feel,’ he wrote, ‘in reviewing the evidence on the localization of the memory trace, that the necessary conclusion is that learning just is not possible.’
From J.Z. YOUNG, Programs of the Brain (1978)
All animal memories can be considered as stored information for the performance of skills. We only know that an animal has learned by studying its actions. I will not say that this disposes of Gregory’s second category, of memory for events.fn1 We humans certainly do have the facility to recall single events that occurred both recently and long ago. Capacity to do this may be one of our unique features and some workers would recognise two categories here, semantic memory, for meanings of words and concepts, and episodic memory for personal experiences. In order to find out what the memory system of the brain is like it is probably wiser to begin by studying the memory for skills rather than events. This, of course, does not mean that the two sorts of memory are unrelated, nor that we must restrict ourselves to animals.
A physical event outside the body provides ‘information’ only if the nervous system identifies it as a symbol, a sign that something is likely to have a particular relevance for life, which may be good or bad. The smell of food is not eatable, nor do nerve impulses in the olfactory nerve provide calories, but an adult animal or man can learn to go towards those smells that are likely to yield food but to avoid others. What we now want to know is what changes occur in the brain during such adult learning. How do particular patterns of nerve impulses come to have symbolic meaning?
An animal or man will only learn if he tries to do things. A creature that sits still and does nothing cannot learn anything. Put the other way round, this reminds us once again that whatever the memory is, it is certainly a part of an action system. It is not a passive structure nor a picture, nor a model in any literal representational sense. It is particularly important to stress this because, following Bartlett and Craik many people, including myself, have developed the concept that by learning we build a model in the brain. I am using now the concept of programs or hypotheses or plans partly because people are so apt to consider a model as a static thing. We have defined the programs of the brain as plans for action and we are now asking how these are learned. But of course whether we speak of models or programs of plans, they must have a physical basis. Even if human memory records are set up ‘in our minds’, I think all scientists at least would agree that when we learn there must also be a change in our brains.
Out of all the immense amount of work that has been devoted to memory, there have emerged three types of theory about the nature of the change that establishes a memory record. There might be (1) a change of standing pattern of activity, or (2) a change of some specific chemical molecules, such as the instructional molecules of RNA or (3) a change in the pathways between neurons within the nervous system. This last is probably the main basis for stable memory records, but all three methods are worth examining.
The brain is continually active and one sign of this is its electrical activities, whether action potentials or slower potential changes such as can be seen in electroencephalograms. Several workers have shown that as an animal learns there are changes in the electrical activity of the brain. The late J. Olds and his colleagues at the California Institute of Technology developed a technique by which after implanting electrodes in the brain of a rat, they could follow the electrical activity of various regions before, during, and after learning a simple task such as pressing a lever to obtain a pellet of food. The whole experiment can be done in two or three days. Changes in electrical activity were found in many parts of the brain. Very interestingly, the earliest signs were in certain regions near the hind end of the brain, including the medial forebrain bundle, which we shall later identify as a possible reward pathway. Changes in the thalamus come later in the learning process, and changes in the cerebral cortex later still.
Unfortunately, we do not yet know how these electrical changes operate to produce the new behaviour of the animal. An early suggestion was that learning sets up cycles of activity in the brain and that these serve to maintain a particular pattern of activity. The importance of self-re-exciting chains of nervous action was first realised, I think, by Alexander Forbes of Harvard in 1922, dealing with the spinal cord, and then by Lorenté de No in 1933 for the control of eye movements. I learned of it from Lorenté during a visit to St Louis in 1936. I then noticed that in the brains of cuttlefishes there are very beautiful self-re-exciting circuits. We found that cuttlefishes show an interesting simple sort of memory. When their prey, say a prawn, disappears out of sight, they will follow it. But they cannot do this if the circuit has been interrupted by a cut. So it may be that the nerve impulses going round the circuit serve as it were to keep the representation of the prawn acting upon the cuttlefish, even when the image is no longer on the retina. This has never been proved to be the explanation of this case, but the circuit arrangement is very striking and there is more evidence for such a mechanism in octopuses.
Many workers have since used the concept of self-re-exciting chains as parts of their theories of memory: D.O. Hebb is often cited as the originator of the concept. It is unlikely that this is the main basis for long-term memory, but such circuits may play an important part in setting up the record. Shocks to the brain received immediately after an event prevent the establishment of a long-term memory record of it. This is a fact that is well attested by those who suffer accidents, and is confirmed by many experiments with electrical shocks given to men or animals. A shock received say one hour later has no such effect on the memory. Footballers questioned immediately after they had been concussed were able to give information about what happened, but half an hour later could remember nothing. There had been no transfer to the long-term store. So the conception has grown up that the record is formed in two stages (perhaps more). In the first or short-term memory, the record is maintained by an ongoing activity, perhaps of the cyclic form I have discussed. Thereafter, a process known as consolidation takes place and the record is firmly established. This long-term memory record surely must involve some stable physical change, for records of single events can remain for up to 100 years in man. Cyclic activity could not last for a fraction of this time without becoming distorted. Moreover, procedures that must upset activity, or retard or even stop it, such as shock, anaesthesia, or cooling nearly to freezing point, do not disrupt the memory in an animal.
From KONRAD LORENZ, The Foundations of Ethology, translated by R. W. Kickert (1982)
Habit formation, in the sense of becoming accustomed to a stimulus configuration to the point of making it indispensable, plays a very important part in the formation of social bonds in many different animals, as well as in man. For those animals in which the process is limited to a strictly definable period during ontogeny, as is the first bond formation in humans, it bears a certain resemblance to the process of imprinting. One example of the similarity is this: a newly hatched, inexperienced greylag gosling first reacts by ‘greeting,’ and a little later by following, any moving object which gives, in response to its ‘lost piping,’ a series of rhythmically repeated sounds within a certain range of pitch. After having ‘greeted’ in this manner a human being two or three times, the gosling refuses to react in the same way to any other object, its real mother included. The irreversibility of this ‘object fixation’ – as Freud would call it – is characteristic of imprinting.
One of the great mysteries of imprinting is to be found in the fact that it fixates behavior onto the species and not onto the individuality of the object. Once the gosling’s following response has been imprinted on humans, the gosling cannot be made to follow a goose, but the human individual fostering it can be exchanged for another without any diminishing of the following response. Also, a gosling hatched by its real mother occasionally changes over to another goose family and remains with that family.
Imprinting of the following response is succeeded by a process of habit or custom formation. During a timespan of roughly twenty hours, the gosling does not reliably recognize its parents as individuals. Interestingly enough, the parents’ voices are recognized slightly sooner than are their physical forms. The process of getting to know the parents is demonstrably independent of reward or punishment; if a gosling happens to wander away from its family soon after leaving the nest, it will try to join any strange family with goslings of approximately the same age. If the latter are more than two days old, their parents will take exception to the little stranger and bite it, in a mild and inhibited manner, but still strongly enough to make it utter distress calls and to run away. If one reunites the wayfarer with its own family, it shows that it has definitely not profited by its disagreeable experience; on the contrary, goslings which have once joined a strange family tend to repeat this error. It is as if even a short following in the wake of the ‘wrong’ parents tends to blur the image of the true ones.
