Do Animals Dream?

What nonhuman animals experience while sleeping can be called dreaming, philosopher David Peña-Guzmán writes in a new book. This impacts how we should think about their interior lives.
A red panda sleeps on a branch.
Holger Ströder for Getty. 

Many people who live with animals have had the experience of watching their pets exhibit odd behaviors while sleeping, like twitching their feet, barking or whining, or moving their eyes back and forth. It’s common for pet owners to say things like, “Look, they’re dreaming,” and wonder if their terrier, when he grunts in a satisfied-seeming manner, is dreaming about chasing squirrels and foxes through an endless and perfect grassy field.


But the actual scientific literature on whether nonhuman animals dream is sparse. Despite thousands of studies on animals and sleep, it’s been left to the side whether nonhuman animals experience the same kind of dreamstates that humans do—where we lose our teeth, miss a high school exam despite being an adult, or see surreal combinations of people and places. 

Since animals can’t describe what went on in their minds after they wake up, scientists have been wary of speculating about the nature of their unconscious, said the philosopher David Peña-Guzmán, an associate professor of humanities and liberal studies at San Francisco State University.

In a new book, When Animals Dream: The Hidden World of Animal Consciousness, Peña-Guzmán pushes back against this instinct. He argues that there is enough evidence on nonhuman animal sleep to claim that yes, animals do dream—and this claim raises many interesting philosophical and moral questions that are worth exploring, starting with wondering why there's been a reluctance to say that animals can dream like we can. 

It’s only very recently that the topic is being broached at all. The first modern science paper on animal dreaming was published in 2020, in the Journal of Comparative Neurology. With the title, “Do All Mammals Dream?,” it was the first contemporary use of the terms “dream” and “dreaming” to refer directly to animals other than humans. 


Motherboard talked with Peña-Guzmán about what convinced him that animals can dream, animal nightmares and sleep talking, and what animal dreaming implies about the obligations we have to animals. 

This interview has been edited for clarity and length. 

Motherboard: It was striking to read that the first scientific article explicitly about dreaming in animals was published so recently—in 2020.
David Peña-Guzmán: A second one just came out a couple of months ago after my book was already in print. It's only very, very recently that people are putting these things together, at least in the 20th and 21st century.

But there is a longer history of people wondering whether animals dream or not, what are some of those earlier examples?
The most famous character here is [Charles] Darwin, who writes about the dreams of animals in The Descent of Man, which came out in 1871. By the time that book came out, of course, Darwin had already published On The Origin of Species and had created a major upheaval in Europe, and really all over the world, with the theory of evolution. 

In On The Origin of Species, Darwin was very careful not to mention human beings. He just laid out the logic of the theory of evolution and talked about the diversification of biological forms and organic life. But of course, everybody knew what the implications of that book were. When the book proved to be controversial anyways, Darwin decided to write a second book which was The Descent of Man where he doubled down on his claim that humans are also implicated in this theory.


It’s in that book that you get what people called the birth of comparative psychology, looking at animal minds and human minds and seeing the parallels that you can find that indicate a common origin. He talks about animals' perceptual systems and our perceptual systems. He talks of animals having abstract concepts. Animals having memory. Animals having social relations and plastic social hierarchies. One of the things that he talks about, though not in great length I have to say, is dreaming.

For Darwin, there is reason to believe that the kinds of subjective experiences that we have when we fall asleep, that we call dreams, also occur in a large number of other animals. He mostly focuses on mammals. He had this insight that the capacity to generate a subjective experience in the middle of the sleep cycle is not an exclusively human capacity.

I talk about Darwin, and also a few other scientists and philosophers from the early 19th century who pick up on this insight. For example, George Romanes, who was one of Darwin's proteges, who also wrote a lot about comparative psychology and comparative anatomy, who talked about the dreams of animals. José Miguel Guardia, who was a Spanish philosopher scientist, takes this and starts putting it in conversation with philosophical questions about the nature of dreaming about the imagination, and about what this tells us about the sensibility of other animals, about how they experience the world.


There are these important figures in the 19th and even in the 18th century who recognize that there are enough behavioral parallels in our sleep cycle and the sleep cycle of other animals to say, yes, they're having their own kind of virtual realities when they fall asleep. What they dream about, we don't know. But we have good reason to think that they dream.

This kind of excitement that existed in the 19th and 18th centuries suddenly dies in the 20th century. I call the 20th century the silent century, because this is the period when certain norms of hypothesis formation in science and, in particular, in psychology change. Suddenly it's not okay to be speculative about what might happen in the minds of other animals. So there is a sort of movement away from those questions.

Even during this “silent period”, were scientists still studying what happens in nonhuman animals while they were sleeping, just not using the word dreaming?
They definitely don't use the words dream or dreaming or any variation of that. What they do is they focus primarily on identifying physical events that can be tracked and measured in the laboratory. That includes two things. It includes initially sleep behaviors like what the animals are doing with their limbs, with their bodies, or with their eyes. But later, when neuroscience really takes off, it starts including neural states. What is happening in the brains of animals that we can study, say, with EEG technology?


You have a very large body of evidence in the 20th century about animal sleep. In fact, one of my difficulties in writing this book was moving through what are, I would say, thousands of scientific publications about animal sleep.

And yet even though there is this explosion of empirical research across the board, what I found most fascinating was the absolute absence of any references to dreams. I started to think that that absence was somewhat symptomatic of a resistance on the part of scientists to talk about the subjective component of sleep.

Of course, they had no problem talking about what neural pathways get activated, what physiological functions increase or decrease, what happens to respiration rate, to blood pressure, all of that. But the moment you ask the question of experience, are the animals feeling, experiencing, or living through something? Scientists would very quickly avoid that question and they would circumvent it. Sometimes they would just straight up deny it. They would say, no, animals don't dream. Or most of the time they would take an agnostic position like, we cannot talk about this because we just cannot know.

Part of what the book tries to do is to recognize the very real limitation that we all face when we study the minds of other animals, because after all, we cannot put ourselves in their shoes. But also avoid what I think to be an excessive form of agnosticism that refuses to take a reasonable leap when there is sufficient empirical and philosophical evidence for that.


One argument for agnosticism might be that since a nonhuman animal can’t describe to us what they’re feeling, or their subjective experience, then that’s why we must remain unsure. But isn’t it the case that the study of human dreams doesn’t only rely solely on subjective descriptions? There are other measurements at play.
There are two ways to approach this question. One is to say that there has been this double standard applied to human and nonhuman animals when it comes to dreaming in the case of human beings. As you point out, there have been a good number of studies that make all kinds of claims without hesitations about dreaming purely on the basis of behavioral and neuroscientific research. You look for certain indicators that tend to be reliable markers of dream experience. You find them, and then you talk about the dreaming that is likely occurring at that particular time.

When it comes to animals, people exhibit this undue conservatism where even when those same indicators are there at the same stage of the sleep cycle, in the same parts of the brain, following the same pattern as in the human case, there is an unwillingness to make the same kind of leap.

It's a double standard that applies not just to dreams, but really across the field of animal cognition. How do you really know that animals are remembering? How do you really know that animals are imagining? How do you really know that animals are feeling pain as opposed to merely undergoing physical damage to the body?


The second way to address this question is to talk about some of the assumptions behind that claim, which is that you need language in order to even begin to consider the possibility of dreaming. We know that that's not the case in humans. And for instance, if you must wait for the possession of human style language, then you can never talk about the dreams of children. And we know that children dream.

There's more to dreaming than just remembering the dream and talking about it. By now, a lot of dream researchers will recognize that, in fact, dream reports can be quite unreliable, that we might not really remember our dreams exactly as they were. 

In the book you cover a number of compelling studies on animal sleep that, to you, support the claim that animals are actually dreaming, and we should call it dreaming. Is there any particular work that really convinced you?
Yes, in the section when I talked about animal nightmares, and research that was conducted by researchers at Peking University, who effectively induced nightmares in rats. They did this by subjecting two different groups of rats to different forms of trauma.

One group of rats was given electrical shocks to the feet, so physical harm. The other group of rats was given psychological trauma, which took the form of letting them watch the physical trauma that was given out to the first group.

What's really fascinating is that both groups ended up experiencing nightmares. One group on account of having undergone a kind of physical torture, and the other one because of the purely psychological burden of having seen that happen to conspecifics.


The rats, much like humans who experience highly traumatic events and episodes in their lives, would then suffer chronic nightmares that would draw them out of sleep. They would fall asleep, then at the moment where they would go into a R.E.M sleep, they would start replaying scenes of the trauma and panic and wake up with all the indicators of very high levels of stress. 

This is something I saw in various studies: the physiological markers of dreaming. the fact that at a specific moment in the sleep cycle, a lot of animals experience a reversal of the typical direction in which physiological changes go. Typically, when you fall asleep, your heart rate and respiration go down. But then during this critical window, which is when I think dreams happen, things turned upside down and suddenly heart rate increases, respiration rate increases, blood pressure increases. This is also the same period when animals start making, in the case of mammals in particular, facial grimaces, and signs of distress that indicate that they are bothered. There is an arousal that is happening in the middle of the night that I don't think can be explained without appealing to an emotionally charged experience.

As somebody who is interested in the role that the body and embodiment plays in the constitution of our mental life, that was a really important indicator for how to talk about dreams, is to talk about that arousal, the emotional, felt, embodied nature of all dreaming.


Speaking of embodiment, one of my favorite anecdotes in the book is the example of chimpanzees trained in American Sign Language (ASL) potentially “sleep talking” by signing while sleeping, making the words for “COFFEE” and “GOOD.” 
The case of the chimpanzees in particular helps us move away from the notion that even if animals dream, they must just dream innate scenarios of evolutionary significance. In this case, these chimpanzees are obviously dreaming something with linguistic content. But of course ASL is not something that chimpanzees do in the wild.

That's something that they learn in the artifice that is the laboratory. It means that animals, at the very least, mammals, but I think this really cuts across the board, are capable of dreaming of things that are particular to their individual life history. They dream of experiences they've had, of predicaments they face, of the sorts of circumstances that they face on a day to day, depending on the kind of life that they live.

It really brings to the foreground the fact that dreaming varies very radically from individual to individual, because you take another identical chimpanzee who was not raised in either in in in captivity, who was not taught sign language, and I don't think that their dreams were they would be similar in that way.

All of this is interesting on its own, but as a philosopher, animal dreams have a moral dimension for you. It’s not only that we should study this more, but animal dreaming has implications about animals and our relationship to them—if they do dream. What’s the connection between dreaming and morality?
Part of it has to do with a very peculiar Catch-22 situation that we find ourselves in. The Catch-22 is that the more scientific research that we do on animals, the more we investigate the way they experience the world, the more we learn about what they are capable of doing. Often, to our surprise. 


The more that we learn about them, the more that our image of who they are is enriched, and the more we need to question what it is that we are doing to them in the first place, that got us to the recognition of the kinds of beings that they are. That's a kind of paradox about which I've written in the past, that sometimes science can have this very curious effect of, through its very own success, ultimately undermining its own justification.

If we realize that animals have all these capacities, then at some point the moral question enters the scene of whether we are entitled to do the things that we do to animals that show these capacities, whether it's a capacity for pain, a capacity for sociality, a capacity for building social relations, or as in this case, a capacity for dreaming.

There is a very close connection between dreaming and what I call moral status, which is an animal's entitlement to moral respect. The connection here is that dreaming is a kind of consciousness. It is a subjective awareness that includes what I call affective consciousness. It's necessarily emotional and affective and felt and embodied. So any animal who dreams is a self who feels things, because by definition, that's what a dream is. It's when you imagine in the context of sleep, that certain things are happening to yourself, and you feel those things happening.

The Revival of Stoicism

We can draw the conclusion from the fact that other animals dream that they have a certain kind of consciousness which, following the philosopher Ned Block, I call phenomenal consciousness. This is the kind of consciousness that is associated with having a phenomenal state, like feeling things, seeing things, sensing things. I follow a certain school of thought in philosophy that believes that this kind of consciousness is what entitles us to moral status: that the reason why we matter from a moral perspective is precisely because we have phenomenal states, because we are sentient beings who sense and suffer.

Most people who write about the connection between consciousness and morality typically focus on things that happen when we're awake, like, do we feel pain? Do we understand our own pain? Do we have other phenomenal states that might give us that moral status? But one of the arguments that provides that connection between morality and dreaming is precisely that dreaming is one of those phenomenal states that also gives us that moral status.

It doesn't really matter what animals dream about or how they dream. It's just the brute fact that they dream that, in my view, has these pretty fundamental moral implications. It means that those animals have moral interests that we are obliged to take into consideration. And those can include their interest in not being harmed, their interest in continuing to live, their interest in cultivating and pursuing their own life projects unimpeded by human action, and so on.

Do you think we’ll never know what animals dream about? And are you okay with retaining agnosticism towards that, even if you stand by the claim that animals do in fact dream, and all the moral consequences that come along with that?
When it comes to the moral question, I think the question of dream content is secondary. It's just the capacity for dreaming itself that automatically grants animals access to what I call the moral community—the community of beings that owe each other respect and that have ethical responsibilities towards one another, or at least who are entitled to moral care from moral agents.

Why We Sigh

That's not to say that we can never know what other animals dream about, because there are some limited cases where we can make pretty good inferences to the best explanation. I talk a lot about research on rodents and mental replay that actually gives a pretty close sense of exactly what animals dream about in particular cases. With birds, we know, at least based on studies of brain activity, that we can pinpoint exactly what musical notes birds are singing at a particular moment in their sleep cycle, just by comparing what's happening to their brain during sleep to what happens to their brain while awake. Sometimes you actually have this really unexpected access to the dreams of other animals. But the moral argument does not hinge on that.

Follow Shayla Love on Twitter.