The puzzle of the memory trace

At the beginning of the Twentieth Century an animal that acted within a puzzle-box was considered a black box itself. Afterwards a wide range of techniques has been developed that help us to look inside the " black box." An intriguing question that scientists and psychologists alike have been striving to answer for centuries is how and where memory processes occur in the brain. In his Lecture to the Royal Society of 1894, Santiago Ramó n y Cajal proposed a theory of memory storage: memory is stored in the growth of new connections of neurons. This prescient idea was in good part neglected for half a century as students of learning fought over newer competing ideas (Kandel, 2001).

German philosopher Hermann Ebbinghaus is fondly remembered as the founder of the scientific study of memory. Ebbinghaus published a book called “On Memory: An Investigation in Experimental Psychology” in 1885. Basically, his research was concentrated on the memorisation of nonsense syllables. The fact that it is easier to remember short sequence of nonsense syllables than a long one seems to be completely trivial. But from modern physiologists’ point of view, this clearly demonstrates that human memory differs from the memory of computer or, say, tape recorder, because machines store all obtained information, if they have room for it, till that moment when they are switched off. This experimental approach for studying mechanisms of memory was applied to studying animal intelligence and preceded early behaviouristic and neurophysiological researches.

Ideas of early behaviourism moved one of the founders of modern neurophysiology, Karl Lashley, to the search for the “engram”, material tracks of learned units. At the age of 16, Lashley studied general zoology and comparative anatomy with Albert M. Reese at the University of West Virginia. Reese appointed him departmental assistant. One of the new assistant's first tasks was to sort out various materials in the basement. Among them he found a Golgi series on the frog brain and proposed Reese that he draws all of the connections between the cells in order to know how frog’s brain works. Only later did Lashley realise that functional variables such as spatial and temporal summation, excitatory and inhibitory states, and micro-movements of elements influencing synaptic contact need not be represented microscopically. The lesson is that neurones are not inert and static, like soldered wires. They are live metabolising cells with synaptic contacts that vary.

By formal training Lashley was a geneticist, and his first 30 publications, until 1917, were dedicated to genetic and behavioural problems. Two of them were co-authored with John B. Watson, the founder of behaviourism, with whom Lashley studied physiology at Johns Hopkins University. The problem, Lashley suggested in the 1920s, was the omission of the brain from the Watsonian Stimulus-Reaction formula. In 1929, Lashley wrote his famous monograph, " Brain mechanisms and intelligence." In his well-known article “In Search of the Engram” published in 1950, Lashley summarised his 33 years of research and theory of memory into two principles:

1. The Equipotentiality Principle: all cortical areas can substitute each other as far as learning is concerned.

2. The Mass Action Principle: the reduction in learning is proportional to the amount of tissue destroyed, and the more complex the learning task, the more disruptive lesions are. Thus, every brain region partakes (to some extent) in all brain processes

In other words, Lashley believed that learning was a distributed process that could not be isolated within any particular area of the brain. Furthermore, it was not the location of the lesion that was important (within reason), but the amount of tissue destroyed that determined the degree of behavioural dissociation. In sum, his theory stated that within functional regions (such as the visual association cortex), all parts of the region were equally effective in carrying out the function normally served by the entire cortical region.

In 1949 Lashley’s student, the Canadian physiologist Donald Hebb presented the most successful theoretical view of the general nature of the engram. His " Theory of Cell Assemblies" suggested another model of brain functioning based on a simple but powerful intuition: that strengthening and weakening of connections depend on how often they are used. If a connection is never used, it is likely to decay, just like any muscle that is not exercised. Hebb's theory supports the view that changes that occur during learning develop among interconnections of neurons throughout wide areas of the brain. Particular kinds of learning have been proven to involve the development of particular circuits of neurons. The engram does not appear to be localised, but its existence cannot be questioned. Hebb discarded the term “neural lattice” that was introduced by Lashley in favour of the term “cell assembly”. Hebb succeeded in presenting a theory of behaviour based as much as possible on the physiology of the nervous system as well as on establishing physiological psychology as part of behaviour theory. A good reason for his integrative approach was that he worked both on emotions and intelligence in chimpanzees and surging on human brain. He said that five years of working with chimpanzees gave him much more ideas concerning nature of human than the rest of his scientific career.

A relative latecomer to academia, Hebb had worked as a schoolteacher, farmer, labourer and novelist before starting to study psychology. When he began as a part-time graduate student at McGill University in Montreal, one of his supervisors, Russian physiologist Boris Babkin, who worked with Pavlov at one time, gave him an advice to work experimentally with animals and that turned Hebb to the work of Kohler and Lashley. Working with Lashley, he received his PhD from Harvard. He then took on a fellowship with famous brain surgeon Wilder Penfield at the Montreal Neurological Institute, where his research revealed that large lesions in the brain often have little effect on a person’s perception, thinking, or behaviour. Penfield was trying to treat patients with intractable epilepsy. While patients were fully conscious, though locally anaesthetised, he opened their skulls and tried to pinpoint the source of their epilepsy. Penfield and Jasper (1954) probed the conscious brain (which has no pain receptors) with electrodes and asked the patients to describe their sensations as various parts of the brain were stimulated. Some patients “heard” conversations that had taken place years before, some heard music, and some seemed to find themselves with old friends, long deceased. They seemed almost to relive the experiences, rather than simply remember them. The powers of recall under such circumstances were phenomenal.

Penfield’s experimental surgery led him to a dramatic discovery. Stimulation anywhere on the cerebral cortex could bring responses of one kind or another, but he found that only by stimulating the temporal lobes (the lower parts of the brain on each side) could he elicit meaningful, integrated responses such as memory, including sound, movement, and colour. These memories were much more distinct than usual memory, and were often about things unremembered under ordinary circumstances. If Penfield stimulated the same area again, the exact same memory popped up - a certain song, the view from a childhood window - each time. It seemed he had found a physical basis for memory (Penfield and Rasmussen, 1950).

In 1942, Hebb worked with Lashley again, this time at the Yerkes ‘s Laboratory of Primate Biology. He then returned to McGill as a professor of psychology, and became the department chairperson in 1948. The following year, he published his most famous book, which made McGill a centre for neuropsychology. The basics of his theory can be summarised by defining three of his terms (see Herenhahn et al., 2001). First, there is the Hebb synapse. Repeated firing of a neuron causes growth or metabolic changes at the synapse that increase the efficiency of that synapse in the future. This is often called consolidation theory, and is the most accepted explanation for neural learning today. Second, there is the Hebb cell assembly. There are groups of neurons so interconnected that, once activity begins, it persists well after the original stimulus is gone. Today, people call these neural nets. And third, there is the phase sequence. Thinking is what happens when complex sequences of these cell assemblies are activated. Hebb first distinguished between immediate memory (short-term memory, or STM) and long-term memory (LTM). We will return to this matter in Part III.

Further development of knowledge about memory processes in the brain has been closely connected with development of methods. Before the middle of the 20-th century, the techniques for investigating the functions of the brain were relatively primitive. Most research on the relationship of the brain with behaviour was conducted by removing sections of brain (ablation) or damaging them (lesion). The distinction between memory trace and sensory circuits permits several interpretations of localisation studies. In such studies, animals are typically trained on some response, and a brain structure, hypothesised to be involved in the learning of that response, is destroyed. Post-lesion behaviour is compared to pre-lesion behaviour. If the lesion abolishes the learned behaviour, it is tempting to conclude that the destroyed structure must have contained the memory trace of the behaviour.

In order to determine whether learning-induced changes actually develop in a given brain structure, more information than can be garnered from lesion studies must be obtained. After 1940 the widespread use of the microelectrode and its appurtenant technology (electronic amplifiers, oscilloscopes, etc.) had a revolutionary impact on neurophysiological theories of behaviour. More advanced methods have been developed then, including electrical stimulation of and recording from selected brain areas; chemical stimulations; and much more precise limited lesion techniques (such as stereotaxic surgery). These researches have greatly increased our knowledge of brain functioning and its relationship to learning. The nerve impulse was observed in almost all parts of the peripheral and central nervous systems and came to be regarded as the universal currency, whether it was induced by electrical stimulation or sensory stimulation or occurred " spontaneously".

Indeed, the progress in neuroscience was based on oscillations from moving behaviour theory away from the physiology of the organism for behavioural explanation, and again into the physiology of the organism itself.

Searching for the engram continues even today. But what we perceive to be the engram has changed. No longer do researchers assume that a single anatomical structure in the brain changes when learning takes place. One of central questions in modern neuroscience concerns the nature of the changes in a nerve cell underlying learning and memory. Researchers have joined together a series of different techniques to try to reveal how the brain can memorise and learn. Observations and cognitive tests with people who lost part of their brain due to accidents or strokes provide information on the functions of different regions of the brain. Studies on the effects of chemo-therapies reveal information about neurotransmitters. Electroencephalograms and PET scans reveal activity patterns of different parts of the brain while people carry out cognitive tasks. During brain surgery with awake patients, the electrode stimulation of different parts of the brain as well as measures of specific neuronal activities help map the impacts of different neural stimuli, and of different cognitive experiences. The introduction of genetic engineering techniques to manipulate genes that code for particular proteins has brought a new level of accuracy to the study of neuronal basis for memory processes.