Cognitive Philosophy Cognitive Philosophy
Is There a Difference Between Memory and Imagination?

If you remember something wrong, is the label “memory” still accurate? Does the label of memory necessitate a 1:1 correspondence with the past? If not 1:1, how much correspondence with the past is necessary for us to still be comfortable using the label of memory? More importantly, if we can talk about a memory being in error, or even completely fabricated (i.e. – “false memories”), then at what point can we then say there is a meaningful difference between memory and imagination?

I’ve written about memory a handful of times on this blog. In one post I criticized the notion of implicit memory. In there I argued that what we call implicit memory is simply a way to describe the behavior of an organism based off its neurophysiological structure in the present, by connecting that behavior to changes in neurophysiology that occurred during past interactions. But that this notion is not what we commonly refer to when we talk about “memory”. So in this post, when I talk about memory I’m specifically referring to “the act of remembering,” memory as conscious access to, and awareness of, past experiences.

What I’d like to do first is describe a body of research that calls into question our standard conception of memory as some sort of accurate recording and subsequent access to past experiences, then I’ll move into a more explicit conceptual connection between these two seemingly distinct things that we refer to as memory and imagination. First, before we even get into the act of recalling an experience, let me point out that a memory doesn’t really correspond to a past event, but rather a past experience. What do I mean by this? We know that the act of experiencing does not always correspond to the total reality of the events going on around us. For instance, just as a few examples, we know that your prior knowledge, your current emotional state, and where your attention is directed at any given moment will all greatly influence the nature of your experience, and thus, even the initial memory you form may be a grossly distorted representation of what was actually occurring around you. This here is already a slightly problematic point to consider, since we have to acknowledge that even from the very beginning, our memories are susceptible to all sorts of distorting effects based on cognitive processes unrelated to the truth of the circumstances in the world outside of ourselves. So already, memory is not about events, but about experiences.

The notion that memories are not accurate representations of the past is not new. These days the cognitive science literature has moved away from talking about a “memory” being stored in the brain, but rather, a “memory trace.” This memory trace is a partial aspect of your memory, which is actually a constructive process of remembering that happens in the moment, built up from the memory trace and…what? This is precisely the problem. If we accept that this stored fragment (the trace) and the memory (the subjective experience of recalling a past event) are not the same thing, then we are committed to the notion that the stored fragment is just a contributor to memory, something that plays a part in contributing to some sort of new emergent entity which we’ve labeled a memory. Compare this to the imaginative process. Imagination also needs to be constructed in the moment from fragments of other thing stored in the brain. If I imagine myself winning the lotto or winning an Oscar or sky diving, somewhere in my brain needs to be stored representations of what a lotto is, what the Oscars are, what an Oscar statue looks like, what planes look like, what parachutes look like, what skydivers look like, etc…Imagination is certainly a creative process, but it is only made possible by constructing an imaginative experience from stored fragments of memory that it can be built up from, in something that seems to me very much akin to memory process discussed above. One way to look at it this is simply that imagination is not possible without memory, but, of course, I’m making a stronger claim than that. I’m claiming that the process of remembering and imagining are not so easily distinguished.

Moving on though, we also know that memories are not stable constructs. Our brains are always undergoing synaptic change. As our neurons fire, neuronal connections are created, strengthened, pruned, etc…Memories are thought to the be the result of a set of neuronal connections made in the hippocampus, the memory trace I discussed above. Recalling an event is thus, at least in part, reactivating this set of neurons. The problem is that neurons can’t fire without some sort of change occurring in the strengths of their connections. The phenomenon that describes the change in our memories due to this reactivation (and the process that subsequently occurs) is known as reconsolidation. What this means is that all sorts of things at the moment of recollection can influence aspects of the memory you are recalling. Similar to the actual memory formation process, the influence of your prior knowledge, your current emotional state, other things you are thinking about, and where your attention is directed can all play a role in changing the memory you are recalling. And this process is going to happen every time you remember something. In a way, the more you remember something, the less likely that memory is going to remain accurate; you are constructing a new entity every time you remember.

Now we’ve reached an even bigger problem though. If memory is constructed from traces, and yet, because of the fact of synaptic change, even these very traces are constantly undergoing change, then it seems that some time down the line the only thing that connects your current recollective experience to the actual original experience, is a causal chain connecting the current memory trace to the original memory trace. But there are all sorts of causal connections between two objects or events in the universe, and on it’s own this fact doesn’t seem to solve all scientific problems. There’s a causal connection between your neurophysiology at this moment and your neurophysiology 10 years in the past. That might give us justification for saying you’re the same person, but in itself doesn’t seem to justify a distinction between the act of remembering and the act of imagining based on that neurophysiology.

Many researchers have noted that memories are simply imaginative reconstructions of past events; that the experience of remembering is shaped as much by a rememberer’s expectations and general knowledge regarding what should have happened, and what could have happened, as what actually did happen. Cases have been documented where knowledge about what we expect to happen can become incorporated into a new memory, even when that expected event didn’t occur. We also know that often times we create false memories by activating certain concepts or categories related to the intended objects of memory. For instance, if I give you a list of words to memorize that include words like candy, sugar, honey, chocolate, etc…many people will later confidently remember reading the word sweet, though it never appeared in the list.

I’ll run through a few more ways memory goes wrong. In recent decades there has been a string of patients, who either under hypnosis or questioning by their therapists have retrieved long repressed memories of sexual abuse, witnessing satanic rituals, or alien abduction. Many of these memories came under fire when it was realized that certain psychiatrists tended to have an inordinate amount of patients with the same type of repressed memory. Psychiatrist A’s patients had all been abducted by aliens, B’s had all been molested, etc…We now know that inserting false memories is incredibly easy. Elizabeth Loftus has been successful in inserting various false memories into subjects. Whether through vivid imagining or simply through repetition, we tend to be unable to distinguish between real memories and imaginings. It’s also been shown that hypnosis can create an environment in which subjects are willing to call just about any mental experience a “memory.”

So far I’ve talked about all sorts of empirical connections between memory and imagination; I’d like to spend a bit of time focusing on some conceptual ones as well. I mentioned above how memory is necessary for imagination, and how imagination can be said to draw from the memory area. But remembering draws from the memory area too, in what way are these processes different? Well, at first glance, memory (the act of remembering) is supposed to be about things that happened in the past, and imagination about things that have never occurred. But isn’t this an epistemological issue? i.e. – there’s a matter of fact about the situation, and we may be able to determine it, but is there actually any difference from the standpoint of the processes functioning in the brain?

Let’s step back for a moment and think about the function of both memory and imagination from an evolutionary perspective. It seems to me that the function is the same in both cases. No thought occurring in the present can have any effect on the past; necessarily, it can only have an effect on the future. And the effect of these processes is to aid in action selection. Memory and learning have a survival advantage inherent in them; otherwise the capabilities would have never evolved. Thinking about stuff in the past allows us to learn and apply the knowledge from past experiences to better act in the future. Imagining counterfactual situations allows us to think about possible events and outcomes and incorporate that knowledge into future action. What we call memory and what we call imagination are intimately tied in allowing organisms to better act in their environments. Sure, these days we can lie around and reminisce about past events, or day dream of what ifs and what might have beens, without necessarily being relevant to survival, but a) this doesn’t speak against the evolutionary origins and purposes of these processes and b) as I’ve written before (in regards to memory), those processes still play that same functional role, just not as obviously.

I am not quite taking the extreme position I seem to be. I don’t want to imply that we can never be in mental states that accurately, or at least mostly accurately, correspond to events that we experienced in the past. What I am arguing though is that these mental states that we call remembering are not cases of you accessing some sort of stored experience out of your memory vault. Every time you ‘remember’ a past experience you are not accessing some sort of stored ‘thing’ in your brain. You are constructing an entirely new experience in something akin to the imaginative process, and while what you construct this experience out of will have some sort of causal connection to synaptic changes made at the time of the original experience, and while there are ways to ensure that this imaginative construction is more justifiably in correspondence with the original event, that original event is gone forever; all that exists is your imagining in the moment. Let’s not be too tied to calling this experience a memory, without at least being cognizant of in how many ways this experience fails to fit that role.

arrow13 Responses

  1. 46 mos, 1 wk ago

    Why can’t this attention-driven imaginative reconstruction of past experiences be called memory? You may be right that the particular instance may not be “a memory” or that there is no such thing as “a memory” in the way we commonly understand it, but it does not follow from the token that the underlying process is not that of Memory. Just as certain tokens do not share statistical properties with the types they are associated with, it may very well be that Memory is not made of memories but is indeed a principle of correspondence as a statistical process…

    TL:DR just because there are no memories doesn’t mean there is no Memory.

  2. Greg
    46 mos, 1 wk ago

    Ramon, I don’t necessarily think I’m committed to disagreeing with your point. I ask a lot of leading questions in that post, but I don’t really come out and say “there is no such thing as memory”. I’m trying to get people to question their commitment to certain naive assumptions about memory.

    In my link on facebook I used some more bombastic language, but even in the UTEP group, i softened it from saying “memory doesn’t exist” to “memory doesn’t exist in the way you think it does“. In either case though, that was a way to get people interested in the post, which actually takes a softer stand. I even end the post with a qualification: “Let’s not be too tied to calling this experience a memory, without at least being cognizant of in how many ways this experience fails to fit that role.”

    edit: I want people to be aware of the fallible/constructive nature of this thing we call memory. But I’m not entirely pessimistic about the nature of memory. I even point out that there are ways that we can better ensure that this reconstruction is in greater correspondence with the original event (though I don’t in this post discuss specific ways to do so).

  3. 45 mos ago

    I have two questions:

    First, in the making of The Lord of the Rings movie trilogy, (some sort of) digital video camera was used to record images of actors in costume. At a later time, someone with a computer added magic and dragons. According to your above argument, should the images of the actors appearing in the finished movie not be considered “recordings”, because maginative scenes of monsters and magic were laid over these images using a parallel digital process?

    Second, you posit that, like memory, imagination must be based on experience. Upon what experience do you suppose the implanted false memories of alien abduction might have been based? :)

  4. Greg
    45 mos ago

    I don’t see the first question as a relevant analogy at all. One of my points in regards to memory, which I’ve discussed in other memory posts and which is widely accepted in cognitive science now (contra classical conceptions of memory as recording), is that initial memories are never recordings.

    As for your second question, the answer would be a mix. But likely movies and tv shows with alien abductions, pictures of aliens in books, etc… You’re reading too literal a definition of “experience” into what I’m arguing. The point is that to imagine something, anything, you have to have experience (interpreted broadly) with the concepts used to engage in the imaginatory process. If I’ve never seen pictures of aliens or seen a movie with aliens, if I’ve never heard the word alien or been made aware of the fact that there are other planets and solar systems and galaxies on which life might exist, and never thought about space flight or anything like that, it’d be very difficult, impossible I argue, to imagine being abducted by aliens. Now, maybe there is some way you want to argue that even this person I’ve just described can somehow imagine being abducted by aliens. But if they can concoct that story, they still have to reason from somewhere. Maybe they think, “okay, people come in all different colors. And there are all sorts of animals on this planet and they all look very different. And even though we’re the only ones who talk and have technology, maybe it could be the case that other creatures could develop with intellect like ours. Now…imagine if there were other planets where these creatures lived, and they made things like planes, except these planes could travel between planets, etc…” That’s possible, but even that story depends on all sorts of experiences to serve as a basis for all the imagination. boxes that fly. creatures that look different. etc…

    You might be interested in this post of mine:

  5. 45 mos ago

    “is that initial memories are never recordings.”

    Precisely. If the neural capture of information cannot be considered a recording, because a parallel but creative process may overwrite the information, then why should the DIGITAL capture of information be considered a recording, when a parallel but creative process can overwrite that information? The only difference is the medium of storage; both system even obtain information from the same amplitudes of electromagnetic energy.

    “impossible I argue, to imagine being abducted by aliens.”

    But if people were truly incapable of creating new ideas–if imagination were purely a process of re-creation, rather than creation, as you suggest, then that FIRST person would never have been able to conceptualize alien life. So abduction stories would ever have been put forth (unless you’re arguing that at least the first one actually took place). But we don’t need to discuss the possibilities of xenomorphs to demonstrate imagination as a creative power: lateral-thinking exercises will do nicely. Most famously, Zen koans. Riddles the correct to which cannot be derived logically from the question asked, thereby requiring that imagination be used to bring forth something entirely new.

  6. Greg
    45 mos ago

    “why should the DIGITAL capture of information be considered a recording, when a parallel but creative process can overwrite that information? The only difference is the medium of storage”

    No, that’s part of what I’m getting it. The act of forming a memory isn’t a recording…memory is NOT a process of encoding events in neurons. Cognition is NOT a matter of information processing. It is because cognition is not these things that thinking of memory as a recording is problematic. To the degree that pictures capture all the electromagnetic energy that humans are capable of interacting with, and to the degree that our medium of projection faithfully reconstructs all those features, I think it makes sense to call them accurate recordings. Even if that’s not literally true, I’m currently not aware of instances where that terminology would cause interactive problems between agents. On the other hand, while I’m OK with calling the process I talk about above “memory,” I’m only okay with doing so if we as rational agents have a better understanding of what we’re talking about when we talk about memory. Because having the naive conception of memory as accurate information encoding, storage and retrieval IS problematic and leads to interactive errors between agents.

    “But if people were truly incapable of creating new ideas–if imagination were purely a process of re-creation, rather than creation, as you suggest, then that FIRST person would never have been able to conceptualize alien life.”

    Not true, I told a story of how that could occur. It involved a lot of inferences and generalizations by concept and category associations. Again, I think you viewing that word “experience” too literally. All I’m saying is that for a person to imagine a spaceship, being familiar with the concept of spaceship helps, but it’s not necessary, as long as the agent has had enough experiences/language development/conceptual development that they have the tools to engage in the creative process from something. Did you take a look at that post about sensory experience? That post, coupled with the two posts of mine about the chinese room and language development will give you a better idea of where I stand on these aspects of cognition.

    Here’s the first one that is mostly setup, and the follow up post fleshes out my own thoughts on the matter.

  7. Katie
    36 mos, 4 wks ago

    I am being a pain (I’m aware) but I would like to use some of this information on one of my essays in school, and I need this to be a credible source. Is there any way I could get a list of your sources or anything like that? Thank you.

  8. Greg
    36 mos, 3 wks ago

    Hi Katie, various aspects of the post were informed by various sources. It’s a bit much to list everything for you. The book I have linked below, Searching for Memory, was the source of much of the psychology behind memory and the constructive nature of memory. I link to Elizabeth Loftus in the post and you can find her work on false memory insertion by way of that. If you’re interested in anything else, please email me through the site’s contact page. Thanks, and good luck!

  9. 36 mos, 2 wks ago

    Hi Greg,

    I’m doing Masters work on the Mind-Body Problem. I just came on your blog by a friend’s link. Your article is very good, it’s an interesting point of view on memory! I put this blog in my favorites ;)

  10. Greg
    36 mos, 2 wks ago

    Thanks for the kind words Davy. Glad you enjoyed it!

    Mind-Body problem eh? You’ve chosen a nice and easy topic for your masters work!

  11. Rishi
    28 mos, 2 wks ago

    A fantastic read. Thank you for sharing. Learned much from it.

  12. 20 mos, 4 wks ago

    enjoyed it a lot.

    searched something in this line of thinking. when this idea occured to me, heavily reading locke, bergson etc, after a long period of thinking, my best assumption is that we differentiate memory by its repetetivness (to which i think you refer) and empathy/identification.

    that is to say, we empathize with our memories, we take them as part of our identity. if we do so with that which is regularly considered as imagination (especially when it is clearly harmful), we are suffering from madness, phobia or a certain disturbance.

    i am looking to read about such things… guess i will try that book :-)

  13. 16 mos, 4 wks ago

    Interesting thought!
    While reading the verb “to revisit” came to my mind, which links “to visit” (somehow related to imagining) to its literal meaning, to remember, by adding reptitivity to it (“re”).
    Also interesting is that convincing liars usually explain their skills by fusing imagination and memories. This is another cool term: “magical realism”, instead of asking what actually happened you think about what could have perfectly happened? I have no idea about the relation between region-specific brain activity and it’s function, but it would be interesting to see whether for well trained Pinocchios the activity boundaries between imagination and memory smear out :)

Leave A Comment