Representation and Memory

Written ‘a while back’, by Barney Low

When I saw a firetruck for the first time in my wild and misspent toddlerhood, there is a good chance that it was driving along without the siren blaring and lights flashing, nor indeed with any flaming building with billowing smoke in sight. Now however, when I think of the word firetruck, my mind immediately leaps to such images, perhaps with some offshoots into subjectively related concepts such as firemen, the battery of my fire alarm, or the f***ers who set off the fire alarms at unholy hours during my stay in my university’s halls of residence.

With the specific example of the firetruck, prior to this encounter I had probably developed a fairly extensive neuronal representation of what a firetruck might be associated with, having been a huge fan of ‘Fireman Sam’ as a child. If you had spoken the word ‘firetruck’ to me before the first time I saw one in real life, the pre-existing neuronal network linked to the auditory stimuli of the word would probably have had a few synaptic connections with the word ‘Sam’ and the visual image of his stop motion figurine, others might connect with sections of the brain associated with excitement, or the vague notion of sitting down to watch television.

As the real life vehicle drove past me for the first time, the basic visual stimuli to enter my eyes would be of a large, red, rectangular version of a car, driving along a road. It seems likely that I would have asked my Mum or Dad to confirm my suspicions; that I could store these novel stimuli under the neurone labelled ‘firetruck’. For the first time, a raw sensory exposure to this object of interest to me would be networked in, with links made to the colour, size, noise, and vehicular nature of the firetruck. At this point I think it is necessary to try and explain how the networking of the brain allows us to recall, at will, the image, sound, taste, emotional significance or functionality of an idea or object, and how this ability is intrinsically linked to perhaps the most significant difference between other animals and man – language.

For the sake of simplicity, I will try to explain the recall of long term memory with the example of a visual stimuli, but arguably the same principles apply to all kinds of recall. The human brain contains a lot of neurones, with some of the more conservative estimates guessing at around 86 billion, and with each neurone synapsing with an average of 7000 other neurones. When presented with an object, for instance an old teddy bear, the raw data  of light-dependent depolarization in the retina undergoes a great deal of processing before it enters the conscious field. There will be individual neurones that fire for almost every conceivable visual detail of an object; some that fire off when presented with a humanoid shape, others that fire off when presented with the particular colour of the bear, or indeed a lump of colour of a particular size or shade. Others still might fire in response to the specific pattern of light and shade that reveals the furry texture of the bear’s fur, or indeed the recognisable shape of most eyes we are exposed to. This pattern of firing would be preserved by one of the hallmark phrases of neuroscience – “neurones that fire together, wire together”, where a group of representory neurones that happen to be interconnected with most if not all of the stimuli, would strengthen their synaptic links with all of the lower echelon neurones, like the roots of a single bud.

The first time we saw the bear, we may have only had the word ‘teddy bear’ under which to label our furry friend, but having settled on a name such as ‘Louis’ or (I wasn’t a particularly imaginative child) ‘Panda’, that word would then have links back to all the various neurones that code for the different traits of that particular teddy. With language, we need only think the name of our childhood toy, and, without the original stimulus of the teddy in front of us, the neurone representing it would simultaneously fire off all the different sensory stimuli that had been activated alongside it in the past, as well as the stored representations of the bear’s texture, and emotional and sentimental significance.

The beauty of this system is that it allows each lower echelon neurone to fire off for many different objects, ideas or thoughts. A neurone that codes for circular shapes might fire off in response to anything from an apple to a clock, and would even fire off (though slightly less exuberently) when presented with the raw sensory data for a pear, far more than it would in response to a sock or a firetruck. It is by matching profiles of the firing (and indeed the specific frequency of firing) of many hundreds or thousands of neurones to a select few, higher echelon neurones, linked with others coding for verbal labelling and other associations, that we are able to so easily resubject ourselves to past experiences, thoughts and sensations at will, in ways that other animals have shown little evidence of being capable of. It also explains why the representation of individual concepts or ideas seems so diffuse when inspected using imaging techniques such as fMRIs, with the many hundreds of traits that make them up being spread broadly around the various sections of the brain.