Disentangling Encoding and Retrieval

Doctor Josh Dubnau explains that memories may be present (encoded) but not accessible (retrievable). Scientists have devised a number of experiments for teasing apart encoding from retrieval.

There is an interesting distinction to be drawn between the presence of a memory and the ability to access that memory. To make an analogy, if I ask you to remember your phone number, you can immediately rattle off your phone number; that memory is present in your brain and you have no trouble retrieving it. If I ask you to remember the phone number of a friend that you haven’t called in 25 years, you might remember that and you might not. Perhaps you don’t remember that, and then the next day, driving your car, you think "I remember that person’s phone number" and it comes back to you. That shows that that memory was present in your brain, but you had a retrieval difficulty and you didn’t access it. This is a very difficult thing for experimentalists to disentangle when studying memory in animals, because if you do an experimental manipulation, and it causes the animal to be unable to behave in a way that tells you that it remembers something, you don’t know if you’ve actually erased the memory or prevented its retrieval. The only way we can disentangle these issues is to do manipulations that are reversible. So if you do an anatomical lesion experiment, where you surgically remove a chunk of brain from a rat, and then the animal can’t remember something you’ve taught it in the past, you don’t know whether you’ve blocked retrieval or erased the memory. But if you had a way to put that chunk of tissue back in exactly the way it was, and then the memory comes back, you would say either that the memory was in the chunk that you took in and out which it could be, or that the disruption you did affected the animal's ability to retrieve the memory. In fruit flies, because we have a long rich history of genetics, we have the ability to genetically do that experiment. We can target a particular gene switch to a small number of neurons, and then by raising or lowering the temperature of the animal we can silence those neurons at the higher temperature or allow them to function normally at the lower temperature. When we do that we can ask first whether that population of neurons is involved in some way in a particular memory, and we can disentangle whether those neurons are involved in retrieving the memory or storing it because if we take those neurons offline and the animal can’t access the memory, but we restore them again a moment later and then the animal can remember, we infer that those neurons are involved in the retrieval of the memory.