Like Being a Dog

Back in 1974 Thomas Nagel published the now-famous paper What is it like to be a bat? It was an examination of the mind-body problem. Part of Nagel’s argument includes the notion that we can never really know what it’s like to be a bat. As W.G. Sebald said, “Men and animals regard each other across a gulf of mutual incomprehension.”

But in What It’s Like to Be a Dog: And Other Adventures in Animal Neuroscience (2017) neuroscientist Gregory Berns disagrees. In his opinion Nagel got it wrong. The Sebald Gap closes from both ends. Firstly because animal minds aren’t really that different from ours. Secondly because we can extrapolate our experiences to those of dogs, dolphins, or bats.

I think he has a point, but I also think he’s misreading Nagel a little.

Maybe not misreading so much as focusing on a lesser point — the inscrutability of bats or other animals. But (despite the name of the paper) bats, and what it’s like to be one, actually occupy only a portion of Nagel’s paper — about two of the eight-plus pages.

Granted the seven paragraphs that do discuss them mention them a lot: 25 times. And Nagel does therein write:

It will not help to try to imagine that one has webbing on one’s arms, which enables one to fly around at dusk and dawn catching insects in one’s mouth; that one has very poor vision, and perceives the surrounding world by a system of reflected high-frequency sound signals; and that one spends the day hanging upside down by one’s feet in an attic. In so far as I can imagine this (which is not very far), it tells me only what it would be like for me to behave as a bat behaves. But that is not the question. I want to know what it is like for a bat to be a bat. Yet if I try to imagine this, I am restricted to the resources of my own mind, and those resources are inadequate to the task. I cannot perform it either by imagining additions to my present experience, or by imagining segments gradually subtracted from it, or by imagining some combination of additions, subtractions, and modifications.

Yet this is exactly what Gregory Berns argues we can do. We can extrapolate from our own experience, plus neuroscience shows us that animal brains function very much like human brains, so we can believe that their experience has something in common with ours.

Certainly anyone who is close to dogs believes that. (Dogs are a unique case. They certainly aren’t the smartest animal, but they are the most domesticated and communicative.)


While I think Berns has a point, I end up siding with Nagel and Sebald. In one of the paper’s footnotes (#8) Nagel argues:

The imagination is remarkably flexible. My point, however, is not that we cannot know what it is like to be a bat. I am not raising that epistemological problem. My point is rather that even to form a conception of what it is like to be a bat […] one must take up the bat’s point of view. If one can take it up roughly, or partially, then one’s conception will also be rough or partial.

I think ultimately even Berns would have to admit that what he has in mind is “rough or partial.”

So despite his bringing it up a number of times, Berns doesn’t really mean what Nagel means. More importantly, that we can’t know what it’s like to be a bat isn’t Nagel’s main point.

At the end of the paper Nagel sums up his main point, writing, “it seems unlikely that any physical theory of mind can be contemplated until more thought has been given to the general problem of subjective and objective. Otherwise we cannot even pose the mind-body problem without sidestepping it.”

So it isn’t so much that we can never understand what it’s like to be a bat but that, in the first place, there is something it is like to be a bat and, further, that this universal experience of being a bat is shared among bats.

By extension, there something it is like to be a human, and it is (as far as we can tell from millennium of description) shared universally, albeit with variation, among humans. Nagel’s view boils down to that we need to include and understand subjectivity — a unique property of more advanced brains.

I think he’s exactly right about that. Subjective experience — what most think of as consciousness — is (so far) a unique property of advanced brains, and human consciousness is (so far) a unique property of human brains. We don’t yet understand how it emerges from brain function, yet it is a central fact and facet of our existence.

Nagel is saying we need to take it seriously (and I’m sure Berns agrees with him on that point 100%).

§ §

Dr. Gregory Berns, MD, PhD, is a neuroscientist and teacher at the Computation and Cognitive Neuroscience Lab in the Emory University School of Medicine in Atlanta, Georgia.

His work respective to this book involves fMRI imaging of animal brains. He founded and leads the Dog Project where they train dogs to crouch down and be still enough in an MRI machine that brain imaging was possible while they performed experiments.

Several chapters of the book discuss the imaging of the brains of other animals. These other animal brains, which all came from deceased animals, include coyotes, sea lions, dolphins, and two Tasmanian Tigers (which, as you may know, are believed extinct).

Berns is a strong animal advocate, and one great aspect of his work is that the dogs are given the freedom to choose whether to participate or not. They never use anesthetics (which would make the experiments impossible anyway), and they never bind the dogs in any way. They do spend a great deal of time training the dogs, first with a tube, then with a tube and noises, and so on until the dog is able to withstand the noise of the actual machine.

They even make a point of having a stair so the dog actively walks up into the MRI machine under its own power and choice. And, after spending much time and effort in the training, during which many dogs washed out, some dogs couldn’t take the real thing and refused to participate. Their choices were respected, which I think is so cool.


Animal advocacy is an especially strong theme in the latter chapters when Berns discusses scanning the brains of “Tasmanian Tigers” (Thylacines) he was able to obtain.

The case of the Thylacine is particularly poignant because it’s a unique and interesting animal, and it seems clear human action removed it from the world. Supposed sightings abound, and many hope to find a living Thylacine, but most experts think that hope is founded in shame.

Berns goes so far as to paint a fictional picture of the last days of the last Thylacine, which was in a zoo where, per Berns, it was basically neglected to death. It’s an almost startling bit of storytelling in an otherwise science-based book, but I didn’t mind it, and I can see why Berns went there. It is kind of a heart-breaker.

Speaking of which, Berns also tells a tale about his Medical School experience with bodies. He wasn’t bothered, and barely remembers, the human cadaver his lab group spent a semester dissecting, but forever vividly burned into his heart is the single day in the anesthetics lab where he killed a dog.

Back then medical schools used animals as teaching aids (this practice has gone the way of the buggy whip), and this lab involved pound dogs that were pre-anesthetized, bound down, and their chests opened to expose their hearts. The students administered various drugs and were able to view the results on the heart. The final stage was an injection of potassium chloride into the heart to stop it.

The thing was, potassium can take up to ten minutes to work, so the teacher gave the students the option of using a scalpel to cut the pulmonary artery and letting the dog almost instantly bleed out. Berns took that option — a memory he can never forget.

I think any of us who have taken a beloved dog to the vet on that final visit can relate. Fortunately we live in a world that is, albeit slowly and with backsliding, growing more mature regarding its views on the ethical and humane treatment of animals.

§ §

Berns begins by listing several key principles of brains:

The first is: Animals have brains because they need to do things. He explains:

What an animal needs to do, though, depends both on its physical form and on the environment in which it is situated. Although brains do process information, information processing is necessary only to the extent that it facilitates action. Moreover, animals have control over the information they process, which is called active perception. Buried in this tight relationship between brain and body is the animal’s mind.

The next principle is: Animals have brains to tailor their actions to their environment.

In other words, animals do not exist in isolation. They are embedded in the world around them, and part of the function of a brain is to link the external world to the animal’s decision-making systems, and ultimately, its body.

The third principle is: Animals have brains so they can learn.

Berns points out that a few neurons suffice for stimulus-response association, a form of learning, but more sophisticated brains allow an animal to learn much more about complex outcomes in their environment. He writes:

The answer is that brains have evolved to do more than take in information and act on it. A sophisticated brain constantly runs simulations of possible actions and outcomes, like you do when you are playing chess. As the behavioral repertoire of vertebrates increased, the complexity of their brains had to increase accordingly.

Which brings us to the fourth and final principle: Brains simulate possible actions and future outcomes so as to make the best possible decision for the situation at hand.

From there he segues into a discussion about brain size compared to body size and the encephalization quotient (EQ) that provides a better measure than brain size alone.


One chapter, Buridan’s Ass, gets into some of the more challenging experiments Berns designed for dogs. For one thing, he wanted a way to attempt the marshmallow test or something like it. The experiments he came up with are fascinating.

He also wanted to test whether dogs have preferences between praise from their human or food. (The chapter name comes from these tests.) Not surprisingly, it turns out dogs vary in their preferences. Some are all about the treats, others prefer their owners.

The test involves a simple T-shaped maze with the human sitting at one end and a bowl with food at the other. The dog is let into the room and can see both; which will it go to first?

(It occurs to me that, first with cable and now with streaming, we all live in the world of Buridan’s Ass. How often have we stared, brain-locked, at the list of things we could watch and found it almost impossible to decide?)

A very good dog ready for scanning!

In the chapter Talk to the Animals, Berns cites Rico and Chaser, two border collies famous for their large vocabulary, and sets off to determine whether dogs do understand labels — do they have a semantic system and, if so, how does it function? He writes:

We humans take it for granted that a name refers to the whole object. But there is no reason to expect other animals to think like us when it comes to language. Dogs could be feature-bound where we humans take a gestalt view. The evidence was scant, but a few studies did support my idea that dogs mapped words to objects in a fundamentally different way from humans.

The experiments were challenging, but successful enough for Berns to conclude that, firstly dogs react to novel words, which makes sense — reacting to novelty is kind of what brains are specialized to do — and it indicates that dogs differentiate the sounds well enough. More interestingly:

In addition to novelty, the dogs seemed to process the words in terms of actions associated with the objects. In our experiment, both objects could be nudged or picked up with the mouth. So even though we taught the dogs two words, it is possible that we didn’t see any difference in semantic representation because the associated action was always the same. An action-based semantic system would make sense for an animal. For an animal lacking language capability, there would be no need to symbolically represent the names of things. But knowing whether a thing should be picked up, chewed, or eaten would be very important.

Later in the chapter, in irony of his own assertion about closing the Sebald Gap, Berns writes:

I have written a lot about the similarities between the brains of humans and other animals, but when it comes to language, we must acknowledge fundamental differences.

I can only agree. 😀

§ §

Which brings me back to Nagel and Sebald. As I noted, Berns does admit to the gap and does so throughout the book. But he also feels we can bridge it, if perhaps as Nagel put it, in a “rough and partial” way.

I think there is truth to that, but I also think we never can really know what it’s like to be another human being. We can try to “walk a mile in their shoes” but ultimately the only viewpoint we truly know is our own.

Millennia of literature and art tells us that the viewpoints of other humans are similar, even very similar, but, as someone I once knew put it very eloquently, “No matter how close you are to someone, no matter how tightly you hold them, in that last moment before you fall asleep, you are alone.”

We are all forever alone in our own heads. The great thing about literature, art, and music, is that it can bring in some company to visit.

Stay something it is like, my friends! Go forth and spread beauty and light.

About Wyrd Smythe

The canonical fool on the hill watching the sunset and the rotation of the planet and thinking what he imagines are large thoughts. View all posts by Wyrd Smythe

21 responses to “Like Being a Dog

  • Wyrd Smythe

    Third book review in a row! As I mentioned in the previous post, I did a lot of reading in the previous months!

    I didn’t, in the post, mention that Berns reminds me a lot of Dr. Clive Wynne, who also specializes in the study of dog minds. See my post Wynne: Dog is Love.

  • Wyrd Smythe

    As a bit of synchronicity, this new video about Tasmanian Tigers just popped up in my YouTube feeds:

  • SelfAwarePatterns

    Sounds like an interesting book. Berns seems to have the standard neuroscience view. The parts about movement are, I think, particularly important. We often forget why brains evolved. Remembering that mental life is part of the causal chain for movement helps keep things in perspective.

    The part about working with dogs in scanners sounds particularly interesting. It seems like something that would require an enormous amount of patience. Although I saw something a few years ago about a study that scanned fly brains, but never looked up how in the heck that is done. There have also been studies scanning baby brains, something that also seems like it would require deep reservoirs of patience.

    • Wyrd Smythe

      Definitely a very interesting book!

      Could the babies have been asleep, or were those cases of needing them awake to do stuff? It did take a lot of patience to train the dogs — the owners were involved and did the actual experiments during scanning — and, as mentioned, dogs washed out all along the line up to, and including, the actual experiments. Some dogs were clearly not trainable to begin with — just too active.

      Of course some didn’t like the noise. As you know, MRI is very noisy, and to get the data they wanted they were running the system at the limit, so it was extra noisy. Even with training the dogs wore headgear to protect their ears. (You see it in the picture with Berns and — I think — his dog, Callie.)

      • SelfAwarePatterns

        The babies were awake. The goal was to see how closely their brain patterns matched adults on various types of stimuli. Obviously you can’t give infants tasks to do while they’re in there, but you can show them pictures or stimulate various parts of their body to see how the brain reacts. I think they had to have a selection process, since many infants were too fussy or freaked out with the machine.

        Stanilas Dehaene, in his book on consciousness, described studies to see how closely their brain patterns match those of adults for consciousness oriented stimuli, although those were done with EEG and related technologies. The results were that they showed similar cortical patterns, but significantly slower. That makes sense since their cortex isn’t mature yet, with a lot of myelination and synaptic pruning yet to occur.

      • Wyrd Smythe

        Come to think of it, some babies (and dogs for that matter) do seem fine with a very noisy environment. It’s a bit like how those who sleep next to train lines can find it hard to sleep if the trains stop running.

        Berns mentions that room temperature superconductors would revolutionize MRI technology and make possible very small scanners people (or animals) could wear while doing tasks. That would really put the “functional” in fMRI. And I hadn’t heard of this, but Berns mentions a coil blanket that can be used for scanning difficult body parts (he gives the example of shoulders).

        (Back when I learned that myelination doesn’t finish until about age 20 it cracked me up because it kind of justifies the age limit of 21 for drinking. It’s always wise to remember alcohol is the one hydrocarbon-based solvent we can consume and not die pretty much immediately. Drink enough alcohol, and it is quickly fatal, but one has to really make an effort. There are stories of guys who drank 21 shots in quick order to celebrate their 21st birthday and then dropped dead. Anyway, there’s actually good reason to keep it away from your brain until it’s finished baking. The interesting thing to me is that the age limit predates our knowledge of myelination. It’s observationally based, which suggests the myelination really matters.)

      • SelfAwarePatterns

        That sounds interesting on the portal MRI scanners. I had to have my shoulder scanned years ago, and they put it in an awkward harness, which was uncomfortable, and I had to keep completely still for long stretches at a time. When dealing with an already painful body part, a blanket sounds a lot better.

        But the biggest issue with fMRI is, while its spatial precision is excellent, it’s temporal precision isn’t. EEG is the opposite. I always wonder why studies only use one or the other. I know they can’t both be used at the same time, but it seems like when studying the way the brain lights up, you’d want data from both methods, even if with different subjects. But there are probably equipment access logistics I’m oblivious to.

        There is something kind of interesting about the fact that, just as our brain is reaching optimum functionality, we start degrading it with various drugs, particularly drugs that degrade the performance of the prefrontal cortex, which is the last thing to mature.

        Maybe happiness is having a degraded PFC.

      • Wyrd Smythe

        It might depend on the nature of the study? I don’t know but it my limited experience it’s true that I’ve seen studies with EEG examples and studies with various MRI or PET scans, but I don’t recall ever seeing them linked together.

        Berns describes using DTI (Diffusion Tensor Imaging) of the MRI data to reconstruct the white matter pathways of the brains. He was interested in what areas linked with what in comparison with which areas lit up in the fMRI data. He seems interested in mapping the brain, so maybe EEG just isn’t useful.

        Humanity has enjoyed mind-altering substances ever since it discovered them! As with all powerful tools, it’s a sword with two edges.

  • diotimasladder

    Berns’ horrific medical experience makes me wonder what the hell whoever came up with that idea was thinking. Yeah, let’s have students euthanize dogs, see how that goes over. Then they have to make a decision as to whether to let the dog suffer or kill it instantly in a gruesome, traumatic way. Geez.

    On dogs mapping words in a fundamentally different way, I’d want to hear more. (Different, ok. But fundamentally? I dunno…)

    “We humans take it for granted that a name refers to the whole object. But there is no reason to expect other animals to think like us when it comes to language.”

    I imagine that “taking for granted” that words correspond to specific objects takes a great deal of background language already in place; who’s to say we don’t originally pick up language the same way a dog does? Plus, don’t dogs know that their owner’s names (Mommy and Daddy) correspond to their owners? Isn’t that names referring to whole objects, or did he mean something else?

    • Wyrd Smythe

      It’s certainly not the least of our yee olde practices we question now! I got the impression there may have been a deliberate sense of weeding out those that can’t take it, but the stated reasoning (per Berns) was that students would be able to view the action of drugs on the heart. Technology provided better options, for one thing, and Berns makes the point that it was never really necessary to see a heart speed up or slow down. (Which is why I wonder why there might not have been some “ordeal by fire” thing going on.) It is true that vets euthanize animals all the time, and medical doctors deal with death all the time, but (channeling Berns again) there is time enough for that later in the student’s career. (There might be something to be said for weeding students out as soon as possible if they don’t have the stomach, but our relationship with dogs is so deep that it seems a poor choice.)

      Regarding dog vocabulary, Rico and Chaser aside (both border collies, interesting to note), most experiments trying to teach dogs specific words for objects haven’t done well. Such experiments are invariably noun-based. Berns thinks dogs might be verb-based (which, yeah, probably isn’t fundamental 😉). The words “mommy” or “daddy” might, to a dog, refer to his interaction with those people. Even their names might be linked more to the attention that’s linked to it than any sense it identifies them.

      It’s a matter of what the sounds are linked to. You might be right about children, I have no clue. Maybe noun-based is something we acquire through training. Maybe verb-based is the more basic way for simpler minds. Sam, for instance, seemed to know the difference between “toy” and “ball” but she might have linked those words to what we did with those objects. Bentley, likewise, doesn’t seem to have much noun vocabulary, but the terms she does seem to know all do seem linked strongly with actions (“walk”, “park”, “ride”, “dinner”).

      • diotimasladder

        Hell, I’m not sure I generally think of nouns in a totally non-verby way, certainly not with words like ‘ice cream.’ 🙂

      • Wyrd Smythe

        Ha! 😀 😀

        (Although,… even with all sorts of verb-y associations added to it, deep down, my bet is that you know “ice cream” is a noun referring to a frozen milk confection. If anything, that’s a key distinguishing feature of human intelligence to me — we understand the notion of abstract classes of things!)

      • diotimasladder

        I’m not sure dogs don’t know some abstract classes of things, although ice cream as a ‘frozen milk confection’ might be a bit of a stretch.

      • Wyrd Smythe

        Could be. I was convinced Sam understood the difference between “toy” and “ball” — the former being anything of hers that wasn’t the latter. For that matter, they also seem to have a sense of the class “mine” (although who knows if it’s a verb-y thing for them or a noun-y thing).

      • diotimasladder

        Interesting. I think that’s what Geordie gets too—balls are one thing, toys are another. And I’m sure he gets possession. If I give out treats to him and his girlfriend, he won’t touch hers unless he’s sure she’s not planning on eating it (lucky for him, she’s a picky chihuahua). Of course, all of this happens pretty rapidly by our standards. I like to think of it as the equivalent of asking, “Are you gonna eat that?” while at the same time reaching across the table to take someone’s food off their plate. I wonder, though, whether he would just steal it from her if no people were around to watch. He could do it easily and she probably wouldn’t care. It could make an interesting experiment to find out whether Geordie really is the good boy I think he is. 🙂

        As for possession being verby or nouny, I’m guessing it’s probably pretty verby!

      • Wyrd Smythe

        I’ve been impressed by how quickly dogs recognize something as belonging to them. Not just treats, those are obvious, but toys, balls, and blankets — stuff that, on some level, would seem to have no immediate value to them. They seem to have a higher level understanding then that. I suppose some of it comes from the human offering that whatever. They’d have some conditioning along that line.

        Geordie has a little girlfriend! You gotta do videos! 😀

        Yeah, possession does sound pretty verb-y. MINE!!

  • Michael

    Another really interesting post that I much enjoyed reading, Wyrd. For me you hit on a fascinating question right at the end, which I might reframe as follows: can any subjective mind truly know what it is like to be another? Meaning… maybe the attempts by a human mind to imagine what it is like to be a bat are “rough or partial” but perhaps that is a special case of a general state of affairs? Is the notion that we cannot truly understand what it is like to be another living being perhaps the fundamental fact of the matter, and the difference in species just a compounding factor in the level of difficulty?

    As I read this I couldn’t help wondering about those members of various indigenous cultures who have purported to embody of directly experience the qualities of consciousness of various animals. I actually think there is something to this, and that while we may forever be uncertain exactly what it is like to be a bear, it is also true that those who invest a great deal of time and study–and conscious intent–may in fact encounter what bear consciousness is like. They will likely discover that in reality every bear is unique, and those unique attributes may be sensible to those who possess sufficient relationship with the animals to encounter those properties.

    There’s a subtle distinction there: I may never know what exactly what it’s like to be Donald Trump, for instance, but at the same time I know the qualities of his consciousness as compared to those of my mother are distinct in many ways. There are differences in the qualities of different beings that we simply do experience and sense. It is this sensing and encountering of distinction that I think is possible, and I think that if prejudicial barriers were lifted, we would find that our capabilities to embody and understand other beings are really quite profound.

    And I do agree much of this may hinge on Nagel’s final point: it would help if we better understood what subjectivity actually is…


    • Wyrd Smythe

      “Is the notion that we cannot truly understand what it is like to be another living being perhaps the fundamental fact of the matter, and the difference in species just a compounding factor in the level of difficulty?”

      Absolutely! I think there is the “Sebald Gap” between our ability to understand other humans versus all animals. (Even though Berns constantly argues against the Gap, he also ends up constantly admitting to it.)

      I had notes for getting into this, but the post hit my word count limit. I may yet return to the idea; there seems enough there for a post.

      “I actually think there is something to this, and that while we may forever be uncertain exactly what it is like to be a bear, it is also true that those who invest a great deal of time and study–and conscious intent–may in fact encounter what bear consciousness is like.”

      Indeed. When a culture has a history of life in an environment, it becomes expert on it. Often such cultures are without what we would define as science, but that means they’re also not distracted by how science expands us into the whole universe. Instead they focus all their intellectual power (and their brains are just as good as ours) on what’s around them. It’s kind of like how little kids can be so laser focused on the environment they notice stuff adults never do.

      Just knowing dogs all my life,… no two are alike anymore than two humans are!

      “I think that if prejudicial barriers were lifted, we would find that our capabilities to embody and understand other beings are really quite profound.”

      I’m with you in general, certainly with animals, but with humans,… well, I am something of a raging misanthrope, but (in my mind) for good reason. As I wrote in today’s post (about humans):

      It isn’t me, it’s you. You fuckers piss me off in more ways than I can count. At the same time, humans fascinate me. Other minds are so interesting, even when they’re clouded with bullshit. (Corrupted with evil I can’t deal with. My general response to evil is: “Martha, get my gun!”)

      It isn’t always a matter of understanding someone so much as needing to defend against their choices sometimes.

  • Tons of TV | Logos con carne

    […] this year I read a lot (see: this, this, this, or this). Lately I’m watching more TV, trying to whittle away at various watch lists. (For a […]

  • Knowing Other Minds | Logos con carne

    […] My post last month about Dr. Gregory Berns and his studies of animal minds ran long because I also discussed Thomas Nagel and his infamous paper. Dr Berns referenced an aspect of that paper many times. It seemed like a bone of contention, and I wanted to explore it, so I needed to include details about Nagel’s paper. […]

And what do you think?

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: