THINKING, IMAGERY AND MEMORY 



167; 



NATURE OF MEMORY 



Memory is a process whereby organized time-space 

 events are carried forward in time. This has been 

 referred to as the 'time-binding' feature of the brain. 

 So ubiquitous are the phenomena of learning through- 

 out the animal kingdom that it has been difficult to 

 conceptualize the nature of the recording process. In 

 spite of a voluminous literature of experimental 

 studies on the nature of learning, one psychologist has 

 written recently: ''It is a blot upon our scientific 

 ingenuity that after so many years of search we know 

 as little as we do about the physiological accompani- 

 ments of learning" (22). There is fairly general agree- 

 ment that organized experience is somehow, some- 

 where represented in the nervous system as substantive 

 memory 'traces.' There is virtually no agreement as 

 to what order of anatomical or electrochemical event 

 comprises the memory trace (9, 18, 28). Speculations 

 concerning their possible nature arc reviewed by 

 Galambos & Morgan in Chapter I. XI of this Hand- 

 book. 



Katz & Halstead (28) have proposed the nucleo- 

 proteins, RNA, as a possible storage mechanism for 

 the engram. Recent investigations of F. Morrell (per- 

 sonal communication) would appear to lend some 

 support to this notion. He has obtained selective 

 staining of nerve cells, stained for RNA, in an induced 

 epileptic focus transferred homologous!)- between 

 brain hemispheres at a time rate compatible with 

 'learning.' 



Experimental investigations of memory max be 

 divided into those which attempt to describe necessary 

 conditions for the development of a memorv trace 

 (including Pavlovian conditioning, operant con- 

 ditioning, human learning), and those which attempt 

 to trap the memory trace in time and space (ablation 

 studies, drug effects, metabolic effects, transfer effects) 

 It is impossible to review all of this work here but the 

 following publications will prove helpful (22, 23, 32, 

 37, 57). There is no known way compatible with life 

 by which an individual memory trace can, with 

 certainty, be erased from the neural pool once it has 

 been laid down. In studies based upon well-estab- 

 lished artificial habits ablation may disturb the per- 

 formance of organized acts, as Lashley (37) and others 

 have shown; but in general, retraining usually reveals 

 some savings or sparing of the original trace. Lashley 

 has concluded from his programatic study of the 

 problem that the motor cortex does not participate 

 directly in the transmission of the trace pattern. 

 Furthermore, there is no evidence that the transmis- 



sion of impulses over well-defined isolated patterns 

 from one part of the cortex to another is essential for 

 performance of complicated habits. Citing his work 

 on the rat and monkey, along with that of Sperry also 

 on the monkey (53), Lashley concluded that the as- 

 sociative connections or memory traces of the con- 

 ditioned reflex do not extend across the cortex as 

 well-defined arcs or paths. "The evidence thus indi- 

 cates that for sensory-motor habits of the conditioned 

 reflex type, no part of the cerebral cortex is essential 

 except the primary sensory area. There is no trans- 

 cortical conduction from the sensory areas to the motor 

 cortex and the major sub-cortical nuclear masM'v, 

 thalamus, striatum, colliculi and cerebellum do not 

 play a part in the recognition of sensory stimuli or in 

 the habit patterning of motor reactions" (37). 



It is known that specific memory traces may be un- 

 altered through wide variations in body temperature 

 in warm-blooded animals, through the diurnal cycle 

 of sleep and wakefulness, through associated electrical 

 stimulation of various parts of the brain, through 

 electrically induced convulsions, through biochemi- 

 cally induced convulsions, and through wide vari- 

 .11 inns in the systemic biochemical milieu. Clinical 

 observations based upon humans, on the other hand, 

 have repeatedly revealed selective lushes of memory 

 I the agnosias on the input side and the apraxias on 

 the output side) associated with focal trauma, space 

 occupying lesions or brain disease, including senility. 

 It has been possible to evoke a verbal report of 

 'images' or 'memories' in approximately 30 per cent 

 of the human brains stimulated electrically by Pen- 

 lield and his associates to elucidate underlying epi- 

 lepsy. However, the possibility exists that such evoked 

 memories are 'leakage 1 effects associated with some 

 epilepsies and not with others since the) have never 

 been observed during stimulation of a normal 

 hemisphere. 



An encouraging approach to the trace problem has 

 been introduced recently by Myers (46, 47). Ad- 

 dressing himself to the problem of interocular or inter- 

 hemisphere transfer of pattern discrimination in cats, 

 Myers has made an interesting discovery. The 

 afferent connections from each eye were restricted to 

 the ipsilateral brain hemisphere in nine cats by surgi- 

 cal section of the crossed fibers in the optic chiasma. 

 The cats were then trained to perform visual pattern 

 discriminations with a mask covering one eye. When 

 the mask was later shifted to the opposite eye, it was 

 found that the discriminations could be performed 

 correctly with the untrained eye (and hemisphere!. 

 In six additional cats, prepared surgically in the above 



