"I remember it well"

Hermione Gingold and Maurice Chevalier in Gigi (1958)
sing "I Remember It Well"

When Maurice Chevalier and Hermione Gingold reminisce about their younger days in the musical Gigi, it is clear that despite Chevalier's insistence that he remembers it well, their memories differ considerably in detail and precision.  Such variation in how precisely we can remember previous events is not only an affliction of older adults: we are all used to the vagaries of our memories, with some events remembered with crystal clarity and others recalled only indistinctly.

Studies of human memory often focus on how we remember some experiences but forget others, distinguishing only between ‘successful’ and ‘unsuccessful’ memory. However, as Chevalier and Gingold demonstrate, our memory for successfully remembered events can vary widely in quality, differing in the kinds of detail we can remember and how precise our memory for those details is. In addition to these ‘objective’ measures of how well we remember, our memory for an event can also subjectively feel more or less vivid to us based on our conscious experience of reliving the episode, regardless of how accurate our memory actually is. To date, there has been limited understanding of how such substantial differences in memory accuracy and experience occur.

Example study display of 3 objects
presented on a background picture

In our new paper published in eLife, Franka Richter and Rose Cooper developed a novel behavioural task that enables a measure of the likelihood that a memory will be remembered to be separated from measures of both the quality of that memory and the vividness with which it is experienced. In the task, participants studied visual displays each consisting of several different everyday objects presented on a background picture of a landscape or building. The objects varied in three characteristics: their colour (on a continuous spectrum), orientation (how they were rotated around 360 degrees), and their location on the background picture. 

We tested participants’ memory for these characteristics by asking them to recreate the colour, orientation, and location of the objects using dials to change the appearance of the object in a continuous manner (see video below). By testing how precisely participants were able to recreate these three characteristics of the objects from memory, we were able to tease apart the likelihood of remembering from the quality of memory. Before we asked participants to recreate the objects, we also presented them only with the background picture and asked them to remember the objects that were presented on this background and rate how vividly they perceived their memory of the objects to be. By asking them to rate the vividness of their memory on a slider from “not vivid” to “very vivid” we obtained a record of how rich the memory felt to participants.

In the current study, we discovered that specific brain mechanisms underlie these three distinct aspects of memory using a brain imaging technique known as functional magnetic resonance imaging (fMRI). The results of our study demonstrated that the hippocampus, a brain region widely associated with memory retrieval, responded to successfully remembering (in contrast to forgetting) an event, regardless of its quality. That is, the hippocampus indicated whether an object characteristic was remembered no matter how vague this memory was.  In contrast, a region towards the back of the brain known as the angular gyrus tracked the precision of the memory that was recalled, such that activity in this region increased with how close participants were to the correct object location, orientation degree, and colour shade.  Lastly, a third region, the precuneus, which has been associated with imagination and experiencing memories from a first person perspective, showed activity that was specific to the vividness with which participants experienced their memories.

Distinct brain regions support different aspects of remembering

Teasing apart these properties of memory in the brain provides a novel approach that could be used to examine and possibly detect subtle early memory differences associated with neurological disorders that previous methods might not have been able to capture. For example, while it has been long known that damage to one of the areas mentioned above, the hippocampus, can cause severe memory problems, studies of patients with damage to more posterior (towards the back) areas of the brain have suggested that these regions are involved in more subtle aspects of memory, such as quality or detail. For instance, people who suffer strokes or brain tumours located towards the back of the brain are typically not amnesic, but their memories can lack detail and richness, and they may report low confidence in their memories, perhaps indicating a reduction in the precision or vividness of memory representations.

Similarly, people with posterior cortical atrophy (a possible variant of Alzheimer’s disease associated with progressive degeneration of posterior brain areas) can exhibit difficulties with visual processing earlier in the progression of the disease than memory deficits can be detected, which could affect the quality of patients’ memory representations even if the ability to successfully remember the gist of an event remains. Using a variation of the task described above might help to characterise and detect these potential deficits earlier than traditional tests, providing earlier access to treatment for such individuals. Moreover, combining the fMRI data with newly developed brain stimulation techniques might facilitate therapeutic interventions to target specific problems with memory retrieval in these populations.

Reference:
Richter, F.R.*, Cooper, R.A.*, Bays, P., & Simons, J.S. (2016). Distinct neural mechanisms underlie the success, precision, and vividness of episodic memory. eLife, 5, e18260. (* joint first-authors)