Abstract
Recently, a research program has emerged that aims to show that animals have a memory capacity that is similar to the human episodic memory capacity. Researchers within this program argue that nonhuman animals have episodic-like memory of personally experienced past events. In this paper, I specify and evaluate the goals of this research program and the progress it has made in achieving them. I will examine some of the data that the research program has produced, as well as the operational definitions and assumptions that have gone into producing that data, in order to call into question the ultimate value of the episodic-like memory research program. I argue that there is a gap between the claims that the research program makes and the data it uses to support these claims, and that bridging this gap is essential if we want to claim that human episodic memory has a meaningful analog in animals. I end with some suggestions of how to potentially fix these problems.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
The “like” in “episodic-like memory” is meant to capture a departure in the method of studying episodic memory: in humans, much work has been focused on the phenomenological aspect of remembering one’s own past (what is feels like to remember a past experience), but the episodic-like memory research program claims that episodic memory can be studied behaviorally, without needing to include the phenomenology. This divorce with the typical way of studying/thinking about episodic memory allows the research program to study episodic memory in animals, who cannot uncontroversially report on what their experiences feel like.
It should be noted that by “personally experienced past events” here I simply mean memory of an event that happened in the animal’s personal past—that is, the event happened to the animal, or involved the animal in some way. A memory that involves, in Suddendorf and Busby’s words, mental time travel, or the reliving of past events. The animal need not be a person or have any kind of robust sense of self.
Although my focus here is on comparative research on episodic memory specifically, I think many of the points I raise can be applied to other sub-fields of comparative psychology. As I will end up arguing, many of my criticisms stem from the way that the episodic memory phenomenon is described (in both animals and humans), and how this description matches up with the particular goal of showing that the episodic memory system works, mechanistically, in a similar way across species. Thus, I believe my criticisms probably extend to sub-fields that have similar goals and have similar problems with phenomena definition. Although I think my criticisms can be applied outside of the comparative episodic memory program, I also think it is important to individually examine each sub-field, its goals, and its data.
I say “can and do” utilize here, but I think there is an important distinction to be made between being able to use some capacity and actually using that capacity, which I will discuss later in this paper. However, it is not clear to me whether Clayton and Dickinson (or the other researchers within the episodic-like memory program) are arguing that scrub jays regularly do use episodic-like memory (like in their daily caching behavior in the wild), or simply that they are able to form episodic-like memories under experimental conditions, but perhaps do not do so normally. To preview upcoming arguments, I think what really matters is the “do” in the can/do distinction, and that it doesn’t make much ecological sense to say that scrub jays are using episodic-like memories—that is to say, that they are not fulfilling the “do” condition.
One way of overcoming the worries about overtraining is to design tasks that involve only one training period. Object recognition tasks are one such design. These tasks take advantage of rodents’ natural exploration of novel objects and tendency to explore novel objects more than familiar ones. This task design can be applied to study episodic-like memory in rodents in a way that does not involve overtraining. For example, Dere et al. (2005) designed a task that involved exposing mice to two different sets of novel objects that were each arranged in a particular spatial layout. Both sets of objects were presented once, with a time delay between each presentation. One set of objects was thus more recently presented than the other set. The test trial involved an array that included two of each type of object (four objects total), with all but one in the original spatial position they were presented in. Mice spent more time exploring the older items than the more recent items, and more time exploring the old object that was out of place than the old item that was in place. The authors argue that this pattern of exploration suggests that the mice have episodic-like memories of their initial encounters with the objects because their more frequent exploration of the older, displaced object shows that they remembered what the object was, when it was presented (longer ago in time), and where it was presented—thus, it is a demonstration of episodic-like memory from only two training trials. Although this experimental design overcomes overtraining worries, it comes with its own set of concerns—for instance, the fact that the mice could be making judgments based on familiarity (stronger memory trace for more recently encountered set of items), rather than explicit recall, along with the assumption that novelness can be ranked, and that this ranking corresponds to different amounts of exploration. Most importantly though, overtraining is only one concerning feature of episodic-like memory tasks, and this experiment does not overcome the worry that the animals are not demonstrating a type of memory that works similar to human memory, and the worry that demonstrating www memory does not mean that the animals remembered a personally experienced past episode. I discuss these worries more below.
Although, as Felipe de Brigard has noted to me, a similar problem can arise within human populations as well—different human populations may have different ways of doing a task, even their performance is identical. However, this highlights the importance of having an idea of just how, mechanistically, the task is performed. If we know something about the neurobiological workings underlying task performance, we are in a better position to say that two populations—be it two human populations or a human and animal population—are performing the task in the same way.
The fact that remembering what–when–where information does not entail remembering a past episode has also been noted by Suddendorf and Busby (2003). They argue that if we characterize episodic memory as www memory, then there is no implication that the phenomenon of episodic memory involves remembering a past episode. I return to this idea in the final section of this paper.
Along with permutations of “what–when–where”, including “what–where” and “what–when” (see Allen and Fortin 2013 for brief review).
Other variations of this experiment have attempted to capture both the “remembering” aspect of episodic memory and the “what–when–where” content aspect. For instance, in one experiment King had to temporally order photographs of foods he had recently eaten (Schwartz et al. 2005).
The relationship between the remember/know distinction and the episodic/semantic distinction is not entirely clear cut: although the remember/know paradigm was originally used to tests hypotheses about differences between the episodic and semantic memory systems (Tulving 1985), other researchers who study the remember/know distinction have not committed themselves to the claim that remembering and knowing are done by different memory systems, like Tulving has. Instead, most argue for more subtle distinctions in processing between the two kinds of remembering involved in remembering and knowing. Regardless of whether or not the remember/know data support a claim about memory systems, though, it is still the case that if there is such a thing as a distinct episodic memory system, then the remembering involved in episodic memory is most likely going to involve the remembering processes that are studied in remember/know paradigms, and so data claiming to demonstrate episodic-like memory should show that remembering processes, rather than knowing processes, are being used.
Some authors have recently tried to operationally define episodic-like memory with an eye specifically towards the neuropsychological mechanisms that underlie it. I return to these definitions in Sect. 6.
Of course, this depends on what we mean by “distinct memory system”. Cognitive systems, including memory systems, are defined by a combination of the information-processing task, the rules of operation, and the neural implementation (Schacter and Tulving 1994; Michaelian 2011). Perhaps episodic memory has its own information processing task with a certain implementation, but if this information processing task itself involves several distinct information processing tasks each with their own implementation, then there is a sense in which episodic memory is made up of several different systems. Just what is the right sense of “distinct system” is up for debate, but I believe it has something to do with what the explanatory goals are, an idea I return to in Sect. 6.
Their strongest argument for the similarity of human and rat processing of the sequences comes from their “ordinal transfers” condition. In this condition, one item for one sequence (the X from WXYZ sequence, for example), would be transferred into the same ordinal position in a different learned sequence (X would replace the B in the sequence ABCD). The degree to which an ordinal transfer is judged to be out of sequence indicates the degree to which the subject represents the sequence using item-to-item associations versus using item-in-position associations. If the subject uses only item-to-item associations to remember the sequence, the argument goes, then the X in the sequence AXCD should always be judged to be out of sequence, since X should only ever follow W. If, however, sequences are coded in terms of the positions of their items, then the X in the sequence AXCD should not count as out of sequence. Allen et al.’s results suggest that both humans and rats utilize both item-to-item and item-in-position associations.
The authors do not claim that the rats’ performance in this study is evidence that they have episodic memory, but they do argue that their study suggests that this task can be used on rat models to study the underlying mechanisms of episodic memory. Furthermore, Allen et al. (2014) cite studies of rat memory for sequences as evidence that rats have a capacity for episodic memory.
For an example of a neural comparison that is too broad to really support the comparative psychological claim, take the argument that the hippocampus is an evolutionarily old structure, and since it is involved in human episodic memory, it would make sense, neurobiologically, that other animals with hippocampi have episodic memory as well (Salwiczek et al. 2010). Once the comparison of hippocampi becomes more fine-grained, this argument seems much less plausible. For example, although scrub-jays have a relatively large hippocampus compared to other bird species (Pravosudov and de Kort 2006), avian hippocampi do not receive higher-order, multimodally integrated information from the brain areas responsible for integrating sensory information (the neocortical association areas in mammals, and the dorsal ventricular ridge in birds) (Rattenborg and Gonzalez 2011). Given that reciprocal connections between neocortical association areas and the hippocampus are thought to play a crucial role in episodic memory encoding and retrieval in humans (Simons and Spiers 2003), the lack of such connections in scrubs jays suggests that they cannot have an episodic memory system that works in a similar fashion as a human episodic memory system.
Clayton and Russell (2009) also argue for a broad conception of episodic memory, what they call a “perspectival memory trace”. Their definition is problematic for similar reasons: its breadth allows it to capture a variety of behaviors that are unlikely to all be served by similar neural mechanisms. Thus, it is detrimental to the overall goal of the research program.
And if such careful consideration is, in fact, required, then we probably do not really need the animal model, since we will already be required to know quite a bit about how human episodic memory works (see Steel’s (2007) discussion of the “extrapolator’s circle”).
My point here is inspired by Bechtel and Mundale’s (1999) discussion of grains of explanation: there different degrees of precision we can use to describe both a phenomenon and the mechanisms that explain that phenomenon. None of these grains of explanations is the correct one; instead what matters is that the grain of description of the phenomenon matches the grain of description of the mechanisms that explain it. Given that the comparative program wants to be able to compare the mechanisms across species, the description of the mechanisms needs to be fine-grained enough to support good cross-species comparisons—the kind of comparisons that will allow us to determine if a particular species would make a good model for studying human memory disorders, for example. Thus, the description of the episodic memory phenomenon should be comparatively fine-grained.
By “multiple different mechanisms” I mean “multiple very different mechanisms”—assuming that both humans and nonhuman animals have episodic memory, we can always expect there to be some differences in the way the mechanisms underlying this phenomenon work, since there are, after all, species differences in the way brains are built. The differences simply should be small enough that they do not preclude the claim that the mechanisms work in the same way, and, importantly, the differences should not be so great that they change the way we characterize the episodic memory phenomenon across species.
References
Allen, T. A., & Fortin, N. J. (2013). The evolution of episodic memory. Proceedings of the National Academy of Sciences USA, 110, 10379–10386.
Allen, T. A., Morris, A. M., Mattfeld, A. T., Stark, C. E. L., & Fortin, N. J. (2014). A sequence of events model of episodic memory shows parallels in rats and humans. Hippocampus, 24, 1178–1188.
Balda, R., & Kamil, A. (2006). The ecology and life history of seed caching corvids. In M. F. Brown & R. G. Cook (Eds.), Animal spatial cognition: Comparative, neural, and computational approaches. www.pigeon.psy.tufts.edu/asc/balda/.
Bechtel, W., & Mundale, J. (1999). Multiple realizability revisited: Linking cognitive and neural states. Philosophy of Science, 66, 175–207.
Cheng, S., & Werning, M. (2015). What is episodic memory if it is a natural kind? Synthese,. doi:10.1007/s11229-014-0628-6.
Clayton, N. S., Bussey, T. J., & Dickinson, A. (2003). Can animals recall the past and plan for the future? Nature of Review and Neuroscience, 4, 685–691.
Clayton, N. S., & Dickinson, A. (1998). Episodic-like memory during cache recovery by scrub jays. Nature, 395, 272–274.
Clayton, N. S., Griffiths, D. P., Emery, N. J., & Dickinson, A. (2001). Elements of episodic-like memory in animals. Philosophical Transactions of the Royal Society London B, 356, 1483–1491.
Clayton, N. S., & Russell, J. (2009). Looking for episodic cognition in animals and young children: Prospects for a new minimalism. Neuropsychologia, 47, 2330–2340.
Crystal, J. D. (2010). Episodic-like memory in animals. Behavioral Brain Research, 215, 235–243.
Degrange, A. R., Fitzpatrick, J. W., Layne, J. N., & Woolfeden, G. E. (1989). Acorn harvesting by Florida scrub jays. Ecology, 70, 348–356.
Dere, E., Huston, J. P., & De Souza Silva, M. A. (2005). Episodic-like memory in mice: Simultaneous assessment of object, place and temporal order memory. Brain Research Protocols, 16, 10–19.
Fouquet, C., Tobin, C., & Rondi-Reig, L. (2010). A new approach for modeling episodic memory from rodents to humans: The temporal order memory. Behavioral Brain Research, 215, 172–179.
Glenberg, A. M. (1997). What memory is for. Behavioral Brain Science, 20, 1–55.
Michaelian, K. (2011). Is memory a natural kind? Memory Studies, 4, 170–189.
Morris, R. G. (2001). Episodic-like memory in animals: psychological criteria, neural mechanisms and the value of episodic-like tasks to investigate animal models of neurodegenerative disease. Philosophical Transactions of the Royal Society of London B: Biological Science, 356, 1453–1465.
Povinelli, D., & Vonk, J. (2003). Chimpanzee minds: Suspiciously human? Trends in Cognitive Science, 7, 157–160.
Pravosudov, V. V., & de Kort, S. R. (2006). Is the Western scrub-jay (Aphelocoma california) really an underdog among food-caching corvids when it comes to hippocampal volume and food caching propensity? Brain and Behavioral Evolution, 67, 1–9.
Rattenborg, N. C., & Gonzalez, D. M. (2011). A bird-brain view of episodic memory. Behavioral Brain Research, 222, 236–245.
Russell, J., Alexis, D., & Clayton, N. S. (2010). Episodic future thinking in 3- to 5-year-old children: The ability to think of what will be needed from a different point of view. Cognition, 114, 56–71.
Salwiczek, L., Watanabe, A., & Clayton, N. S. (2010). Ten years of research into avian models of episodic-like memory and its implications for developmental and comparative cognition. Behavioral Brain Research, 215, 221–234.
Schacter, D. L., & Tulving, E. (1994). What are the memory systems of 1994? In D. L. Schacter & E. Tulving (Eds.), Memory Systems (pp. 1–38). Cambridge, MA: MIT Press.
Schwartz, B. L., Hoffman, M. L., & Evans, S. (2005). Episodic-like memory in a gorilla: A review and new findings. Learning and Motivation, 36, 226–244.
Simons, J. S., & Spiers, H. J. (2003). Prefrontal and medial temporal lobe interactions in long-term memory. Nature Reviews Neuroscience, 4, 637–648.
Skov-Rackette, S. I., Miller, N. Y., & Shettleworth, S. J. (2006). What-where-when memory in pigeons. Journal of Experimental Psychology and Animal Behavior, 32, 345–358.
Stanley, J., & Krakauer, J. W. (2013). Motor skill depends on knowledge of facts. Frontiers in Human Neuroscience, 7, 1–11.
Steel, D. (2007). Across the boundaries: Extrapolation in biology and social sciences. Oxford: Oxford University Press.
Suddendorf, T., & Busby, J. (2003). Mental time travel in animals? Trends in Cognitive Science, 9, 391–396.
Suddendorf, T., & Corballis, M. C. (2007). The evolution of foresight: What is mental time travel and is it unique to humans? Behavioral and Brain Sciences, 30, 299–313.
Suddendorf, T., & Corballis, M. C. (2010). Behavioral evidence for mental time travel in nonhuman animals. Behavioral Brain Research, 215, 292–298.
Tulving, E. (1972). Episodic and semantic memory. In E. Tulving & W. Donaldson (Eds.), Organisation of memory (pp. 381–403). New York: Academic Press.
Tulving, E. (1985). Memory and consciousness. Canadian Psychology, 26(1), 1–12.
Tulving, E. (2002). Episodic memory: From mind to brain. Annual Review of Psychology, 53, 1–15.
Tulving, E. (2005). Episodic memory and autonoesis: Uniquely human? In H. S. Terrace & J. Metcalfe (Eds.), The Missing Link in Cognition (pp. 4–56). New York, NY: Oxford University Press.
Woolfeden, G. E., & Fitzpatrick, J. W. (1990). Florida scrub jays: A synopsis after 18 years of study. In P. B. Stacey & W. D. Koenig (Eds.), Cooperative Breeding in Birds (pp. 239–266). Cambridge: Cambridge University Press.
Yonelinas, A. P. (2002). The nature of recollection and familiarity: A review of 30 years of research. Journal of Memory and Language, 46, 441–517.
Zhou, W., & Crystal, J. D. (2009). Evidence for remembering when events occurred in a rodent model of episodic memory. Proceedings of the National Academy of Science USA, 106, 9525–9529.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Malanowski, S. Is episodic memory uniquely human? Evaluating the episodic-like memory research program. Synthese 193, 1433–1455 (2016). https://doi.org/10.1007/s11229-015-0966-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11229-015-0966-z