Verified by 
The possibility that animals might have imaginations strikes some as anthropomorphic—we are projecting our own qualities into the minds of animals. This was a problem for psychology for a long time and helped fuel the era of behaviorism, a time when animals weren’t supposed to have thoughts. Or, at least we weren’t supposed to talk about them.
Tolman, famous for ascribing animals' cognitive maps, had different ideas. Based on his observations, he proposed that rats considered alternatives before they made a choice. He called this vicarious trial-and-error learning. An animal can experience the consequences of its decision before it makes that decision.
Tolman thought this because he saw the rats in what appeared to be rat-like consideration of a future choice. This is something I often see in other faculty, my children, and the reflection of my computer screen.
To assess this idea of vicarious trial-and-error thinking, neuroscientists began recording from animal’s brains to "see" what they were thinking. Using brain imaging techniques that can record from cells in an area of the brain called the hippocampus, researchers could see that certain cells were active in specific places. We now call these hippocampal place cells.
But researchers also discovered something they didn’t quite expect: Animals imagine alternatives before they take them. The place cells in the animal’s hippocampus show patterns of activity that sweep ahead of the animal’s position in a contiguous, connect-the-dot pattern. These forward sweeps go around corners and travel down the arms of mazes, considering different alternatives in succession.
These forward sweeps, sometimes called sharp-wave ripple complexes, are now understood to enhance learning. If they are interrupted, animals learn less well. They happen while animals sleep in apparent dreams. They also happen when animals are awake, in those moments of rat-like reflection. Like other forms of exploration, they are more frequent when animals have less experience with a decision. In other words, the less that animals know about an environment, the more likely they are to have imaginings about it.
But is this really imagination, or are the animals simply replaying past experiences?
Tolman observed that rats could take novel but more direct paths through a maze when they became available, even though they had never experienced those precise paths before. This was part of his motivation for proposing cognitive maps. His logic being, it would take a map to make these kinds of intelligent inferences about the world. But do animals really think their way down those paths before they take them?
A number of researchers have now shown that rats can imagine novel events during their sharp-wave ripple complexes. In one study, Pfeiffer and Foster (2013) allowed rats to travel from a home location in an open-field experiment in search of a reward. When the rats found the reward (yummy liquid chocolate), they could travel back home and receive another reward (more liquid chocolate!).
This experiment wonderfully combines both a random exploration phase with a predictable go-home phase. Except, the go-home phase, if it’s to be a direct path, is likely to be a path the animal has never taken. The rats take a wandering path to find the unpredictable chocolate and then (if they can) a direct path home (form a random location) to get the home chocolate.
What Pfeiffer and Foster observed using neural recordings from hippocampal place cells is that during the home phase, when the animals could take a direct route, the animals would imagine the trajectory towards the home before taking it. The rats didn’t simply imagine highly experienced locations, but preferentially imagined routes homeward when a reward was located there. Moreover, they imaged this route regardless of what direction they were pointing in. They also imagined multiple paths to the home location.
These results, and many others like them, suggest that animals (at least rats) can imagine routes they’ve never taken and can do so in a goal-directed fashion. These imaginings appear to reflect planning behavior used to guide future action because animals most often took one of the paths they were imagining.
The fact that animals can imagine things, dream, and perhaps even tell themselves a kind of story based on their own desires, is a fascinating idea. It’s not likely to surprise anyone who has watched their dog bark and chase objects while sleeping in the middle of the living room floor. Except, your dog could be dreaming about something it’s never actually done, like fighting dragons (or at least tractors).
If they can imagine, one can’t help but wonder what kind of cognitive system is required for such a feat. What other implications do imaginings create? Does an animal that can deliberate about its future enjoy some modicum of free will?
Might an animal that imagines also know that it is imagining? Might it know the difference between the real and the imagined? Might it, therefore, know the difference between its real self and what it could be?
Facebook/LinkedIn image: eva_blanco/Shutterstock
References
Pfeiffer, B. E., & Foster, D. J. (2013). Hippocampal place-cell sequences depict future paths to remembered goals. Nature, 497(7447), 74.
Tolman, E. C. (1948). Cognitive maps in rats and men. Psychological review, 55(4), 189.
