Humans don’t use the same type of memory to recall the capitals of the world as they do to flash back to a visit to a friend’s house the week prior. The latter, which allows people to revisit past personal experiences in their minds, is known as “episodic memory,” and was once thought to be exclusive to humans. Although there is still debate about the extent of this type of memory in nonhuman animals, research has suggested that it is shared by groups as diverse as rodents, cephalopods, and birds—and findings published July 25 in Current Biology add bottlenosed dolphins (Tursiops truncatus) and potentially other brainy cetaceans to the list.
Over the past two decades, scientists have endeavored to develop tests to assess episodic memory in various animals. Jays, rats, dogs, and cuttlefish are among those that have successfully passed such tests. Curiously, “whilst dolphins are often thought of by scientists, as well as [by] the general public, to be amongst the most intelligent animals,” there is scarce research “investigating their memory and mental time travel abilities,” University of Cambridge cognitive scientist James Davies tells The Scientist over email. “This was surprising to us,” he adds, so he and his colleagues attempted to fill this gap by exploring whether the aquatic mammals are capable of retrieving details of past personal events.
To achieve this, the team chose an approach that involves testing if the animals can inadvertently store apparently irrelevant details of an event. The event is then followed by an unexpected task in which one of those trivial details is required to solve a problem.
The team used “where” and “who” questions, each on a different trial, as the seemingly irrelevant items needed by the dolphins to perform a memory task that involved retrieving a hidden ball behind a person at one of a dozen spots around the edge of a pool. Each dolphin was first trained to retrieve the object from the water, and then to get it by approaching a person holding it visibly in front of them while ignoring an empty-handed person standing at a different spot, each of which was marked. During this training, the locations were randomized and the person holding the ball differed each time, so that those details were irrelevant to learning the fetching behavior. Then, for the trials, the dolphins were asked to fetch the ball as they had learned to do, but after 10 minutes, something changed—this time, the ball was unexpectedly no longer visible, as it was now behind one of the two people’s backs. In the “where” trials, the ball was hidden in the same spot as in the first event, but both people had been swapped out, while in the “who” trials, the locations of the people changed but the ball remained with the person who’d had it previously.
Eight dolphins went through each of the two tests, separated by at least 48 hours. All dolphins got it right in choosing the correct spot on the “where” trials, and seven achieved success on the “who” trials (the eighth dolphin simply made no choice on that task). The researchers designed the test in a way it avoided the possibility that other kinds of memory were at play, such as their short-term working memory or their familiarity with a location or a person.
“Amongst the different ways that people have tried to develop animal models of episodic memory, in my view, the strongest type of evidence comes from the strategy that they use in this paper,” referred to commonly as “incidental encoding followed by an unexpected question,” says Indiana University Bloomington experimental psychologist Jonathon Crystal, who was not involved in this study. Thus, he adds, the “study is compelling and provides strong evidence” that dolphins possess episodic-like memory.
Kelly Jaakkola, a cognitive psychologist and marine mammal scientist at the Dolphin Research Center in Florida who did not participate in the study, says it “is brilliant in a few different ways.” For instance, the authors “figured out a way to ask dolphins about different aspects of a previous event” without training them to do so, which is important since dolphins could solve the task as a result of training rather than episodic memory. She also praised the team for controlling for short-term memory and familiarity.
Jaakkola mentions a small caveat, though: The experiments were not accurately blinded. Based on previous animal cognition studies, it is well known that people interacting with animals during tests may “have unconscious ways of communicating” the right answers, she says. In this study, it is possible that people may have provided “unintentional cues” leading the dolphins to choose the spot where the ball was hidden. While she says she believes this is not the reason why dolphins succeeded in these trials, it needs to be addressed to “close that loop.”
Davies says the team was “aware of this issue, and whilst this was difficult to completely irradicate,” they minimized the likelihood of the people present during the trials giving these involuntary cues. Their “body position, hand gestures and eye gaze” were fixed, oriented towards the point between the two possible choices. Based on the design and outcome of the trials, he says, “it is unlikely that [the dolphins] used unintentional [or] unconscious human cues to solve the task.”
Overall, Jaakkola says that based on their cognitive skills, dolphins are “a good candidate” for having episodic-like memory, and “this study goes really far in showing that.” The more we look for such capabilities in nonhuman animals, she says, the more species we’ll likely find them in. An exciting question, she adds, is therefore “where do we draw that line? Which animals do have it, which animals don’t, and what sort of cognitive or neurological or social characteristics” do those animals share? That’s going to be “the fun part of the game,” she concludes.