Last time, we talked about writers on the aphantasia spectrum–a reduced detail of mental imagery (whether visual or audio or any other sensory modality, although I will use the visual one for examples). Interestingly, with few exceptions, the same level of detail (from almost-photographic to vague to near-nonexistent) applies not just to voluntary or involuntary imagining but also to memories and dreams. What is going on here? Could there be some commonality between these forms of mental imagery?
Indeed, there is–at least, this is what current neuroscience tells us. In fact, most of our mental imagery is ultimately based on memories. We can’t imagine an apple, for example, if we have never seen one. In that hypothetical case, if someone describes an apple to us as, say, something spherical and red (or green), then we can imagine a red (or green) sphere–if we had seen a sphere previously–but it won’t be anything like an apple, until we actually are introduced to one. However, if we have seen thousands of apples, then imagining one (to a varying degree of detail, of course) shouldn’t be a problem. Except that our imagined apple may not be exactly the same as any of the applies we have seen before. It may be slightly different.
This has to do with malleability of memories. Our brain isn’t like a computer or a camera that can preserve a photo with extremely fidelity, down to a pixel. But whereas digital storage has this fidelity–and, as a consequence, a complete reproducibility (the same image can be copied many times, without changing)–as an advantage, the malleability of our organic memories has an advantage, too. A complementary one, in a sense.
You see, the brain has evolved in animals and not in plants because we’re capable of complex and voluntary movement. Its main purpose is to predict the outcome of movement–and even though the brain’s functions have increased in complexity, ultimately everything, even the most abstract thoughts, are based on some internalized motion in one way or another. Although this subject is much deeper, perhaps to be addressed in a separate blog post, the important thing to remember right now is that the brain is a prediction engine–and memories help it predict the future by comparing the current to the past.
If some visuals or sounds or smells or anything else reminds us of some danger we have come across before, we may find ourselves on high alert–regardless of whether we consciously have noticed the trigger or not. Conversely, if something in the present reminds us of a pleasure in the past, then our emotional background may be colored accordingly and our behavior may change–again, regardless of whether we’re actually conscious of that fact or not. But these are fairly crude examples. Memories can help us with assessing the present in many far more subtle ways.
But since our memories will rarely match the present faithfully, then some degree of variation in the memories would help the brain detect a close but imperfect match. Our brain is a massively parallel system, with multiple neuronal ensembles running simulations in parallel but with slightly different variations, competing for a better match–with the winner rewarded by increasing the synaptic connections in that neuronal ensemble, which would increase its weight of contribution in the future. In a way, we run the simulations with many slightly different versions of the same memories to see which one better helps us in the present.
For that, a source of variation is necessary. It is present in the brain in the form of a pink noise that manifests in the neuronal activation patterns–and, via them, observable, for example, in the brainwaves. I already talked about how pink noise is different from other kinds of noises earlier. Suffice it to say that it’s the most complex kind of noise, between the chaos of white noise and the relatively order of brown noise. By being a highly complex randomizer, it turns the brain into the ultimate butterfly effect machine, where billions of metaphoric needles are standing on end every milliseconds, deciding on which way to fall–with even tiny variations leading to highly divergent results in the end.
This is the, admittedly much simplified, essence of the Neural Darwinism paradigm, introduced by Gerald Edelman. Previously, he got his Nobel Prize for discovering that an important part of our immune system (now called the adaptive immune system) works by the evolutionary principle. Before him, people used to think that the immune system just has some kind of “blueprints” for different kinds of pathogens. But this is very inefficient, won’t work for something that doesn’t match any blueprint–and isn’t, in fact, how the immune system works. Instead, a certain kind of immune cells undergo mutation at a far accelerated rate (unlike most of the cells in our bodies, those don’t have a fixed genome). When a new pathogen is detected, they go into an overdrive, in effect accelerating their “evolution,” until a match to the pathogen is found. That way, our immune system can adapt to any pathogen past and future, even if it comes from outside the solar system.
Gerald Edelman also proposed that the brain works by the evolutionary principle. My description above of the brain as a massively parallel system where the best prediction by neuronal assemblies is constantly being selected and reinforced, with pink noise as a source of randomness for variation (that is to say, mutation) is a simplified illustration of his Neural Darwinism paradigm. This is also the reason our memories are so malleable–and for why this is an improvement over the digital fidelity. Our mental imagery, too, is malleable for the same reason, as it ultimately comes from altered memories, perhaps recombined in novel ways.
What’s more, malleability of memories even helps us process direct sensory input. For example, when the brain constructs our field of vision (what we see at the current moment), it often fills in the relatively unimportant parts (the ones we aren’t focused on) not with the actual data received but with something similar from memory, a little changed to match–because it’s computationally less expensive. A fairly good chunk of our present is, in fact, stitched out of our memories this way–as is beautifully described in Gerald Edelman’s Remembered Present.
The same applies to dreams. But we will take that subject on in the next blog post.