Is Implicit Memory Actually Memory?
Implicit memory is normally thought of as a type of memory where past experiences influence current thought and behavior, but without conscious awareness of those previous experiences. How does this work? Is there a “thing” in the brain we can point to and say “this is an implicit memory?” And is “memory” even the best term for it?
Let’s start with organisms that have no explicit memory, and with a simple biological organism without even the ability to learn. It has a certain cellular structure, and that structure is static for all its life, it has no memory. Sensory signals come in, and the organism acts in predetermined ways based on its structure. The output is directly coupled to the input (through a series of complex interactions, yes, but they might as well be mechanical if no change can occur). Now, what is the relationship between the sensory signal coming from the environment, and the organism? I want to point out two ways to look at this. One, the sensory signal implicitly defines the contents of the external environment. The organism doesn’t know this of course, it’s not conscious of the information contained within the sensory signal, but the information is there none the less. A scientist could extract information about the external environment simply by analyzing the sensory signals in the appropriate ways. Two, the sensory signal serves as input that causes the organism to act in different ways (remember, it has a static physiological structure, it can’t choose. So the signal directly causes behavior.).
There are types of bacteria that interact in these simple sorts of ways. They travel up sugar gradients, and if they find themselves not traveling up a sugar gradient, they tumble, stop tumbling, and then start traveling again. Rinse and repeat. Their sensory discrimination consists of ‘sugar’ and ‘not sugar’. And they have predefined ways of behaving based on one or the other sensory input. Thus, sensory signals define ways in which an organism can, and in a simple organism, does, behave in its environment. The important take away from both of the previous points is that we are not making any proclamations about the internal states of the organism. We don’t have to say that the organism “sees” or “experiences” or “represents” the outside world. The sensory stimulus and the structure of the organism determine the behavior of said organism. It is only us humans, analyzing the sensory stimulus, who could construct a coherent picture of the outside world, given the right tools.
Now take an organism that can make synaptic changes based on previous experiences. Its relationship to the environment via its sensory inputs are the same (inputs provide opportunity for interaction), except this time changes in its neuronal structure occur, allowing different behavior based on the same input. In a famous experiment, Eric Kandel studied the roots of memory formation by habituating an Aplysia ( a sea slug) to a certain stimulus (a squirt of water), and studying the physiological changes that occurred as it changed from its standard behavior of retracting its gill when presented with this stimulus, to a decrease and eventual cessation of response activity. The Aplysia “learned” that the stimulus wasn’t harmful and stopped retracting its gill after a while. What happened here to allow this? In the case of the Aplysia the only changes to the organism were in a small network of sensory neurons, motor neurons, and interneurons. This network of neurons functions as a directly coupled system, producing a standard output based on a standard input (think of the knee jerk response). The changes that occurred in these synaptic connections were such that the sensory input of the water squirt now didn’t cause the retraction behavior (stronger signals from the inhibitory interneurons or some such, a biology person would know better than me).
Now here is where we would often say that the Aplysia now has an implicit memory for the experience of being squirted by water, and that’s why its behavior has changed (it’s often a slippery slope from habituation and sensitization to learning and to implicit memory). But I want to argue this conception that the change in neuronal structure encodes a memory, or stores a past experience, is misleading. What happened in the Aplysia is technically what we should call learning, a process that changed the synaptic structure of the nervous system, thus leading to different behavior by the organism based on sensory input. Yes, the organism can better differentiate its environment because of previous experiences or events, but it doesn’t “remember” those experiences (or at least we have no justification for saying so at this point). It just “acts” in response to sensory input based on its current neuronal structure. In the moment of action, the Aplysia is no different from the simple bacteria I described above. Behavior stems from current neurophysiology. We don’t have the warrant to say that any sort of subjective experience accompanies it.
Is there room for the label of implicit memory here at all? What about the very specific change in the neuronal system? Take a snapshot of the neuronal connections before learning (or habituation as neuroscientists would put it), and a snapshot after, and could we call the difference between the two patterns “the implicit memory”? Thus implicit memory would just be a term we use to describe the neurophysiological change in a system (not the act of changing, that’s learning, but the literal physical difference) based off of past events. This seems better than a conception that implies the organism actually remembers anything. But are we sure we want to define implicit memory as the physical change in a system that allows it to behave differently? Taken to the extreme wouldn’t we have to say that the dent in the door of your car is an implicit memory for the accident your car was in. That might be an easy way to speak, and be unlikely to cause problems in every day conversation, but if we then try to talk about human experience and consciousness, this lackadaisical use of language leads to all sorts of problems.
It’s not really the physical neurophysiological change that IS an implicit memory, but rather there are presuppositions contained within any behavior that has been learned, about the events that caused those changes (whether in humans, dogs, or computers). And these presuppositions could “in a sense” be said to correspond to an implicit memory for the event that caused the neurophysiological change to happen i.e. – The question I am asking is, is this a linguistic distinction or a physical distinction we are making? The organism behaves “as if” there was a memory, when all there is neurophysiological structure. So is implicit memory just a term we use to describe behavior that has come about as a result of structural change in a central nervous system? As opposed to behavior that stems from some sort of non accessible memory?
I think this is problematic, as I think this notion of “implicit memory” confuses our conceptions of explicit memory. We talk about implicit and explicit memory, both as forms of “memory”. But when we normally talk about memory we are not just talking about behavior; we are talking about something that can be remembered. A past event that can be thought about. Something that can be reflected on. We don’t just behave in ways that have come about because of past interaction based on synaptic change, we actually “remember” the events that caused that synaptic change. This is not simply a semantic quibble I have with labels, it seems to me that calling the former notion “implicit memory” misleads us when thinking about what memory actually is in regards to our normal conception of memory, and thus how cognitive science can account for it and explain it. What is this *thing* in the brain that is a memory?
How does implicit memory relate to explicit memory? There’s a long way to go from changes in neurophysiology that implicitly encode past experiences, to the ability to recount a story, to reflect on past experiences. To not just behave in ways where the effect of past events are implicitly contained within your actions, but to explicitly remember the details of these past events. But if we continue to view behavioral change as a result of neurophysiological change as a result of interactions with the environment, as an organism acting on implicit memory, it leads us to the conclusion that all we have to do to explain explicit memory is connect implicit memories to our conscious areas, and BAM, we’ve got memory. This is naïve though, and not the first time I’ve written about the problematic nature of this conception of memory in relation to consciousness. As if there’s a “you”, and separately there are “explicit memories” and also separately there are “implicit memories”. And that your conscious circuitry can become aware of the explicit memories because the encodings for them are housed within an area that connects with the “you” area, while the implicit memories are housed in another area without that direct connection. But “you” are not housed in your prefrontal cortex, even if neuroscience determines that the prefrontal cortex is necessary for conscious experience to arise. The idea that there is a you, who is able to access a memory is an illusion. There’s not a you AND a memory, there’s just you in a particular conscious state, which we call remembering, where your conscious thoughts correlate to events that happened in the past (we hope). Your conscious experience, your sense of self, emerges from the same neuronal firing, the same physical processes that “encodes” for a memory. Finding some subset of neuronal connections that when firing seem to allow us to remember a particular event or fact does not actually tell us much of anything useful about how consciousness arises, or what the relationship is between neurons and experience.
Is this something that has much relevance for our everyday lives? No, probably not (but thanks for reading anyway!). But in so far as the difference is important for research and theories of cognition and consciousness, I think that not only do we oversimplify the problem of explaining what memory is (as in, how synaptic change can allow conscious beings to remember specific events) with our standard conceptions, but these conceptions will also direct further research and theories in a direction that may turn out to be a dead end in the long run. That huge problems arise by using language loosely is not an original idea in philosophy (or in general for that matter), but it’s interesting to note just how much our conceptual framework is dependent on our use of language. Memory, like perception, just serves as a useful case study to help us work through more fundamental problems in cognitive science, and by trying to become more clear about these sorts of issues, it will have profound effects on the way we think about consciousness and the nature of the self.