How Does Epistemology Address Memory as a Source of Knowledge?

The field of Epistemology often begins by examining visual perception, a primary doorway we enjoy to the world. But memory—our capacity to store and recall information that we were acquainted with in the past—is often overlooked as a source of knowledge. This is a mistake. For memory connects us to the past and helps us make sense of the present. In this article we’ll unpack some of the epistemology of memory and why it’s an important source of knowledge.

Think about a memory card in your phone or camera. We can first distinguish its capacity from its function. The card’s capacity is just as described—it’s a ‘container’ for information. Its function, similarly, is to store information in this container. But the card’s capacity and function are not enough for it to be a memory card. You must also be able to retrieve the information stored within its capacity. 

Human memory is very similar. We can think of it as a capacity (or ‘faculty’) to store (‘remember’) information that we can later access (‘recall’). The distinction between remembering and recalling might sound subtle, but it’s crucial. Otherwise we would be unable to explain a temporary inability to recall, say, a word until someone prompts us: that we do eventually recall the word confirms that we remembered it—it was stored in our memory—but were unable to access or retrieve it. 

The point to take away is that we can think of memory in three ways: as a capacity or faculty, a storehouse of memories, and as something we can access to recall these memories.  

Remembering is Factive

You have a dinner date with a friend but show up at the wrong restaurant. Upon realizing your mistake, you call your friend and say you believed you were meeting elsewhere. Would it be fair to say you remembered you were meeting elsewhere? It seems not: you surely believed you were meeting elsewhere, and you recalled you were meeting elsewhere, but you did not remember this. The reason is that memory is factive: like perceiving something (rather than just having a sensory experience, like a hallucination), remembering requires that what you remember is true. 

But remembering is more than just knowledge of past facts. You have to have yourself been acquainted with these facts. For example, most people know the United States fought a civil war. But no one remembers the civil war because no one living today experienced it. Instead, people remember others’ testimony that the US fought a civil war. 

So to remember something it is not enough that it is a past fact you can recall from your memory. It must also be a past fact with which you yourself were acquainted. Otherwise you do not remember the event itself, but only someone’s reporting about it. 

We have all had misleading memories. But veridical memories—ones that accurately represent to us facts from the past—are distinguished by the fact that they are causally linked to the fact that makes them true. One reason I only seem to remember having locked the door is that I did not actually lock it, and therefore this fact cannot have caused my memory that I did. Likewise, the reason I remember seeing you last week is because I had an experience of seeing you, and this experience is causally connected to the memory of seeing you that I have stored and recalled. 

This causal requirement for remembering is almost certainly necessary for memories to be sources of justification for our beliefs about the past, that is, things that make our beliefs about the past reasonable. But a complete epistemology of memory must consider much more. For example, memory is similar to perception because it relies on perception to furnish the storehouse of facts waiting to be recalled. But, unlike perception, memory is very likely not imagistic: we can remember locking the front door without being able to recall a visual representation of this event, but we cannot perceive locking the door without an attendant visual representation.

This brings us to a crucial point. Though memory is not an originary source of knowledge, depending as it does on other sources like perception, it is an epistemically direct source of knowledge. This means that veridical memories give us knowledge of the past without our having to make inferences about the sources through which we first acquired that knowledge. More simply, to know on the basis of a memory we do not also have to know the memory was formed through a reliable perceptual process. Instead, it is plausibly enough that the memory is causally connected to its source, and that our recollection of the memory from our memorial storehouse is reliable.