In the study, Prof. Fried observed the neural activity in the brains of 13 epilepsy patients, as the patients watched clips from TV shows like Seinfeld and The Simpsons. A short while after, the test subjects were asked to describe what they remembered from the video clips. During recall, the exact same neurons that had fired while viewing a clip fired once again while the subject was recalling it. Soon, the researchers were able to predict what clip the subjects would recall just by looking at the neurons that lit up seconds before the recall experience was vocalized. Link
This is an exciting development in short-term memory research on the hippocampus. It is also validation for David Hume's theory of mind. Hume thought that the way we differentiate perceiving, remembering, imagining, and reasoning is by their phenomenal properties; how vividly we experience them. Specifically, he supposed that remembering was closest to perception--and generally a more vivid and lively experience--than imagining. It would be good for his theory if the neurons used during perception were more activated when remembering than when imagining. Because even if the way we differentiate between mental processes is more complicated than Hume's picture, there still might be some reliable physical differences that help us monitor the source of our thoughts.
If Hume was part of Prof. Fried's experimental team, then he'd probably like to contrast the neural activity of imagining and remembering the Simpsons. Of course, there are very difficult methodological problems with such an experiment as anyone imagining the Simpsons is probably accessing memories in order to furnish the imagining. Also, how could we test the imagining, remembering and perceiving of the same stimuli within subjects without conflating the results? Perhaps subjects could be asked first if they had seen a particular episode (and removed from the study if they had), then asked to imagine what will happen, then shown the video and subsequently asked to remember what they saw. If at any point subjects remembered that they had seen the episode before, then they would leave the study. Whilst this experimental design is not ideal; it might turn out that imagining produces a quantifiable difference in neuron firing in the hippocampus than remembering.
There has been a lot of philosophical criticism of Hume's theory on the basis that our imaginings can be more vivid than some of our memories, so vivacity can't be the way we differentiate between them: e.g., imagining lying on a beach on a tropical island on a perfect day might affect us more than remembering climbing a tree at our childhood home. I have no doubt we use many different systems to evaluate our mental experiences: We reason about ideas and contrast them with other experiences we've had, we infer the likelihood of something being true based on our theory of memory formation and retention: e.g. we know that older memories fade and horror movies can create vivid nightmares. Nevertheless, it seems likely that one of the systems we use is connected with 'how much like perception' a mental experience is. When our mentalizing seems real, we judge it more likely to be a memory than something we just imagined. New research on neuron firing patterns seems to back up this practise.