Perception and the Unreliability of Episodic Memory

Source: George Milton/Pexels

Recent research suggests that repeated "replays" of episodic memories—i.e., memories of personal episodes from our past—can help improve our ability to visually distinguish between scenes, faces, and objects in our surroundings (e.g., Görler et al, 2020; Brogaard & Sørensen, in press).

A plausible explanation of how this happens is that repeated replays of episodic information can help fine-tune the memory templates you use to determine the identity of what you are looking at in your surroundings (Brogaard & Sørensen, in press).

The finding that episodic memory can improve our perceptual capacities may seem rather puzzling at first. Specifically, it appears to be at odds with data that suggest that episodic memory is highly prone to error (e.g., Schacter, 1999). If, indeed, episodic memory is highly unreliable, how can it help improve our perception?

The Unreliability of Episodic Memory

A large body of literature has shown that episodic memories are highly prone to errors, distortions, and other imperfections (see e.g., Schacter, 1999).

One such memory imperfection is trauma-induced forgetting. Traumatic experiences tend to lead to a narrowing of our attention (Easterbrook, 1959; Christianson & Loftus, 1991). During a car crash, for example, you are likely to focus more intensely on certain details while ignoring others. Later you may be unable to recall many details about the car crash.

Another kind of memory imperfection is suggestibility. When we are given false, but probable, information that is compatible with existing information stored in episodic memory, we often incorporate the false information into our stored memories (Loftus, & Hoffman, 1989; Schacter, 1999). So, when you think back to something you experienced in the past, you may recall the false information as if it were true.

Source: Alexandr/Podvalny

Say you remember that your parents took you to Disney World on your 13th birthday. If you are later told (incorrectly) that you got lost in the amusement park that day, you may subsequently "remember" that you got lost there, even though it never happened.

Still other research suggests that repeated recalls of an episode from your past can alter the information stored in episodic memory (Loftus & Palmer, 1974). Over time, the information you retrieve about the episode thus becomes increasingly less accurate.

One reason that information stored in episodic memory often changes with repeated recalls is that memories are stored in bits and pieces rather than in a single "file." So, when you think back to an episode from your past, the brain must first put the fragments back together.

This constructive process can easily introduce memory errors. While piecing together memory fragments, the brain often incorporates erroneous information, which can result in vivid but inaccurate recollections of what actually happened.

Solving the Puzzle of Memory and Perception

But if recollections of episodes in our past are highly prone to errors and distortions, how can repeated recall help improve our ability to visually distinguish between items in our surroundings?

Or to put it differently: If epistemic memory is highly imperfect, why don't repeated replays of our episodic memories worsen rather than improve our perceptual capacities?

Source: Matheus Bertelli/Pexels

One way to explain how repeated replays of episodes from our past can improve our perceptual capacities is to look closer at the nature of memory replays. Replays of episodic memories occur not just when we deliberately recollect episodes from our past but also during dream sleep (e.g., Diekelmann & Born, 2010; Inostroza & Born, 2013).

Unlike deliberate recollections of episodes from our past, replays of episodic information during dream sleep haven't been shown to be particularly prone to errors or distortions.

This suggests a testable hypothesis: Involuntary replays of episodic information during dream sleep may play a greater role in fine-tuning our perceptual capacities than our deliberate recollections of episodes from our past.

Memory Essential Reads

Why Cramming Doesn't Work

Shaping the "Episodes" of Our Memory

Hyper-Focus May Improve Memory

Another way to explain how episodic memory can improve our perceptual capacities despite its general imperfection is to look closer at the role attention plays in laying down memories.

Source: Pixabay/Pexels

Traumatic experiences can, as noted, cause us to focus more narrowly and intensely on some details while completely ignoring others, leading to trauma-induced forgetting. But the converse is also true: our recollections of the details we hyper-focused on at the time tend to be more robust and accurate than episodic memory in general (Brogaard & Sørensen, in press).

This leads to a second testable hypothesis: Replays of episodic details we hyper-focused on during a traumatic event may play a greater role in fine-tuning our perceptual capacities than replays of episodic details we didn't pay much attention to.

Expertise May Improve Memory

Still another way to explain how episodic memory can improve our perceptual capacities despite its general unreliability is to look closer at which kinds of remembered episodes are most suitable for improving perception. In line with this idea, Fang and colleagues (2018) found that recollections of episodes from our past tend to be more accurate and detailed when involving categories in areas of great familiarity or expertise.

Source: Tina Nord/Pexels

Say you are a bird expert and see a type of bird you know a lot about and later think back to that episode. Your recollection of that episode is then likely to be more detailed and accurate than if you had recalled an episode involving a less familiar category.

This suggests a third testable hypothesis: Recollections of episodes involving familiar categories may help to further improve our ability to visually distinguish between items within those categories. Recalls of episodic information involving less familiar categories, by contrast, may not play this sort of role.

These three hypotheses are all compatible with the finding that episodic memory generally is highly unreliable (cf. Brogaard & Sørensen, in press).