This is the first part of a two-post series. The Part 2 is at The Thoughtful Animal blog on our network.

"Let’s see if you bastards can do ninety." And with that, the DeLorean accelerated forwards and traveled back in time thirty years.

Unfortunately, in life outside Back to the Future, we do not yet have the option of time travel. The closest thing we have at the moment is mental time travel. As humans, we have the ability to let our minds wander back in time to an experience of our choice, whether it be thirty years, or thirty minutes.

The ability to remember a particular event or episode (so-called ‘episodic memory’) is not the only type of memory we have. We also have memory of facts of the world, for example, ‘the sky is blue’, or ‘Jack Bauer is a hard man’, without having a specific episode in mind – we just seem to know it.

Episodic memory is considered to be a key aspect of our consciousness. Indeed we find it difficult to imagine being ourselves without our personal memories. This topic has not only captivated philosophers and science fiction writers; it is also a large field in science. One area has been to try and learn whether other animals apart from humans have memories of past events in their lives. Apart from it being an interesting question in itself, knowing this could help us better appreciate human episodic memory, and aid our understanding of diseases which affect it (like Alzheimer’s).

But how do we go about finding out if other animals remember events from their past? In humans, we have a simple way of doing this: we ask. Unfortunately with non-human animals it is not so easy. First, we have to think about what specifically we mean by an ‘episodic memory’. Endel Tulving, the godfather of episodic memory, said that it was a form of consciousness that allows people to think about the subjective time in which they live and ‘mentally travel’ in that time. Unfortunately this definition is somewhat futile when coming to test it in other animals. More recently two scientists, Nicola Clayton and Anthony Dickinson, proposed a new definition of episodic memory in order to look for it in non-human animals. This definition was that the behaviour would have to fit three criteria: the animal must have the ability to recall ‘what, where and when’; to integrate this information; and then deploy this information flexibly.

As this is quite a lot of information to take in in one go, let us break down that last sentence. Firstly, when we recall a memory from our past, say, our first date, the key components of that memory are ‘what’, ‘where’ and ‘when’. The ‘when’ part of the memory would be, say, fifteen years ago. The ‘where’ part of the memory might be an ice cream parlour (yes I am lapsing into an American Hollywood movie cliché). And perhaps the ‘what’ part of the memory is that the date was really awkward and it finished with you concluding that the opposite sex were not all they were cracked up to be. However, we don’t recall these three separate pieces of information as separate entities: they are all integrated as a whole.

This brings us to the second of Clayton and Dickinson’s criteria that an episodic memory has to fulfill: the memory must be an integrated whole (not just where and when it was, or what and where it was). The final criterion an episodic memory must adhere to is that it must be able to be deployed flexibly in new situations. For example, we can recall an event from the past, and use that information to apply it to a new situation given our current knowledge about the world. This is something we do countless times a day, perhaps without even realizing it.

Now that we have broken down episodic memory into its composite parts it is certainly more attainable than a vague sense of our memories in a wispy floaty form. However, it is still difficult to see how it might be demonstrated in non-human animals.

Clayton and Dickinson tackled this challenge through an innovative set of experiments with western scrub jays. A key thing to know about these birds is a particular natural behaviour they have: they store food (by burying it) to retrieve at a later date. To begin with, the researchers gave birds the opportunity to learn about different types of food. They were given peanuts and wax worms to bury in sand-filled ice cube trays, and then dig up at a later date. During this training phase the birds learned that the worms (which they greatly preferred) decayed much faster than the peanuts. They therefore also learned that if a long period of time had passed then there was not much point in digging them up, as they had gone rotten.

The birds were then tested in a scenario where they were given new ice cube trays where they were forced (by restricting access to one side of the tray) to store one type food on one side of the tray, and then the other type of food on the other side of the tray five days later. Four hours after they buried the second food type they were allowed to dig up both of the food items.

The order in which they were allowed to store the two food items differed. This meant that when worms were stored first they would have decayed by the time the birds recovered the food. However, when the nuts were stored first both foods would still be edible. The birds behaved as if they remembered what they had done: when they buried the worms first, then the nuts five days later, they recovered the nuts and ignored the rotten worms. However, when they buried the nuts first, and their favourite wormy snack was still fresh, they dug that up instead.

So it seems that these birds remembered ‘what’ food was buried (the peanuts or the wax worms), ‘where’ they buried it (the location in the ice tray), and ‘when’ they did so (four hours or five days previously). They also integrated these three components into a single memory, as not doing so would have resulted in them making mistakes when digging up the stored food.

However, there was a possibility that the birds were not remembering, but simply using a visual cue or smell to choose which part of the tray to dig for food. To make sure this was not the case, another test was carried out. This time the birds were allowed to store their food in the ice cube tray as before, but when the tray was presented to them the second time, the food was removed and it was filled with clean sand. The birds dug in the places that they would have found the food if it had not been removed by the researchers, and did so in the same pattern as before: digging for nuts when the worms were old, and for worms when they were fresh.

Even though this was fairly convincing evidence for an episodic type of memory in scrub jays, there was still another possible way that the birds could have been solving the task. Perhaps they were just not as good at remembering worms as peanuts. This would mean that when the birds stored the worms first, it was not that they were purposefully avoiding them when they had decayed: they had just forgotten about them.

To make sure this was not the case, a second group of scrub jays were trained on worms which, after they were buried by the birds, were always dug up by the scientists (without the birds seeing) and replaced with fresh worms. This meant that these birds never learned that worms decayed over time. When these birds were tested, they always dug up the worms more often than the nuts, even when more time had elapsed. This shows that the first group of birds were indeed avoiding the worms because they had learned that they had decayed, and not just because scrub jays forget worms more easily.

So, scrub jays did seem to integrate the ‘what-where-when’ information into an ‘episodic-like’ memory. It is called ‘episodic-like’ because we can never truly prove that an animal is remembering the past in a similar manner to how we remember it, even if they show all the same behavioural signs. Some might argue that since this is the case, we may as well just call it ‘episodic’ and be done with it, but that can be for the philosophers to worry about.

You may remember that there was a third criterion for episodic memory: the ability to use memories flexibly in new situations. This means integrating an episodic memory with a knowledge-about-the-world type memory like ‘the sky is blue’. For example, you might remember exactly which pub you left your lucky hat in last night, but then you later learn from a friend that you looked like a bit of an idiot in that hat, and everyone was laughing at you most of the time anyway. You might have been going to retrieve your hat, but now that you have integrated this new knowledge about it (and more importantly, how you look wearing it), you decide on balance it is probably not worth retrieving.

So, what would be an example of this in the bird world? Say a scrub jay stores some worms, but then after having buried them learns separately that worms degrade more quickly than it originally thought. This second part of information is not stored in the original episodic memory, but is more general knowledge about the world. Can a scrub jay then use its memory and this new knowledge to change its behaviour? It was found that the scrub jays could: when they buried food but then learned from another store they had that those worms had degraded faster than expected, they then changed their preference in the original tray to the nuts.

These experiments are exemplary because they not only were the first conclusive evidence for episodic-like memory in non-human animals, they also are a good demonstration of the scientific method in action. Initially the scientists found that birds seemed able to remember food-storing episodes in their lives. However, they then addressed all the possible caveats they could think of in their original experiment, or the ways that they could disprove their finding. The follow-up experiments held up their original result and made it even more convincing.

Photo credits:

Time: Carlos Henrique Reinesch

Western Scrub Jay: Cory Chen


Clayton, N., Yu, K. & Dickinson, A. (2001) Scrub jays (Aphelocoma coerulescens) form integrated memories of the multiple features of caching episodes. Journal of experimental psychology. Animal behavior processes 27(1): 17.

Clayton, N.S., Bussey, T.J. & Dickinson, A. (2003) Can animals recall the past and plan for the future? Nature Reviews Neuroscience 4(8): 685-691.

Clayton, N.S. & Dickinson, A. (1998) Episodic-like memory during cache recovery by scrub jays. Nature 395(6699): 272-274.

Clayton, N.S. & Dickinson, A. (1999) Scrub jays Aphelocoma coerulescens remember the relative time of caching as well as the location and content of their caches. Journal of Comparative Psychology 113(4): 403.

Tulving, E. (1985) Memory and consciousness. Canadian Psychology/Psychologie Canadienne 26(1): 1.

Now, proceed to Part 2 of this post, over at The Thoughtful Animal blog.