Elements of episodic-like memory in non-human animals: theory, methodology, and evolution

Abstract

Episodic memory is the declarative memory system that involves the conscious recollection of personal events and when absent, results in profound losses to the typical human conscious experience. The debate surrounding its uniqueness to humans has seen a lot of controversy, but as evidence for conscious episodic recall in humans is centred around language-based reports, major issues arise when investigating its presence in non-human animals. Therefore, the term ‘episodic-like memory’ is used to reflect the behavioural characteristics of human episodic memory in the absence of evidence for consciously experienced recall. Various behavioural paradigms have been developed to test episodic-like memory, namely the what-where-when memory, incidental encoding and unexpected question, and source memory paradigms. Whilst some authors argue that non-episodic solutions could be used in solving some of these tests, each methodology targets different aspects of human episodic memory though an animals’ behaviour (e.g., the what-where-when paradigm tests the contents of an episodic memory, whilst the incidental encoding paradigm tests for the process of automatic encoding of incidental information in episodic memories). Therefore, in the first experiment of this thesis, I extend the work conducted with corvids on the what-where-when paradigm by developing a novel methodology based on the incidental encoding paradigm to test episodic-like memory in Eurasian jays. The jays successfully encoded and recalled incidental visual information associated with experimenter made ‘caches’ (i.e., food underneath cups), thus providing converging lines of evidence across multiple paradigms that caching corvids possess an episodic-like memory ability. As such, episodic-like memory is often thought to be an adaptation to the cognitive demands of food caching behaviour, for which jays are highly specialised. Accordingly, I explore this hypothesis by testing a non-caching avian group: the parrots. I test various parrot species with the same methodology used to test the jays, so that their performance can be directly compared. The parrots failed to encode or recall incidental visual ‘cache’ information, possibly supporting the caching hypothesis. However, this methodology is based on spatial memory, for which cachers may have experienced enhanced selection for, and episodic-like memory may exist within other domains, such as social information. Therefore, I test another non-caching species with similar sociality to parrots: dolphins, on a non-caching-based incidental encoding methodology that includes testing the recall of incidental social information. Dolphins successfully encoded and recalled incidental social information, indicating that other informational domains can be encoded within episodic-like memories. That said, dolphins were also successful in recalling incidental spatial information, suggesting that episodic-like memory may be a generalised ability in some species, as it is in humans. Furthermore, as studies with captive, hand-raised subjects cannot accurately recreate the conditions under which the cognitive ability evolved, and merely show what a species can do rather than what they actually do in nature, conducting studies in the wild is imperative. Therefore, in my final experimental chapter, I combine the what-where-when memory and incidental encoding paradigms to comprehensively test the memory abilities of wild, free-living, non-caching parids. I used automated computerised feeders with custom-built programmes that allowed for experimental manipulation on an individual-level basis in the field. The birds showed evidence of episodic-like memory in both tasks. Consequently, I present multiple lines of converging evidence for episodic-like memory in a wild population of generalist foragers, suggesting that episodic-like memory, and the selection pressures that favour it, may be more widespread than previously assumed. Similarly, as episodic memory encompasses a variety of cognitive structures and processes, research on episodic-like memory in non-humans should follow this multifaceted approach and assess evidence across various behavioural paradigms. Subsequently, I examine whether various key features of human episodic memory are conceptually represented in episodic-like memory across phylogenetically and neurologically diverse taxa, identifying similarities, differences, and gaps in the literature, concluding that the evidence is mixed. Overall, I conclude that future research should utilise an episodic-like memory ‘test battery’, with a combination of multiple paradigms, and conduct tests with wild subjects in situ.