The Myth of Canine Hierarchy - What Science Tells Us About Dog Intelligence

Read the article on Substack 

Yes, they found breed differences. But not where you'd expect them, and not in the ways our cultural narratives would predict.

Border Collies, those supposed canine Einsteins, excelled at impulse control—which makes sense if you think about it. Herding requires incredible restraint, the ability to resist the urge to simply chase and instead channel that predatory instinct into something useful. But guess what? They weren't universally superior across all cognitive tasks.

German Shepherds and Malinois, breeds we associate with intense focus and trainability, actually scored lower on impulse control tests. Why? Because their jobs often require immediate, decisive action. When you're detecting explosives or apprehending suspects, hesitation isn't a virtue—it's a liability.

The most striking finding, though, was what they didn't find: no significant breed differences in short-term memory or logical reasoning. None. Zero. That brilliant Border Collie might be no better at remembering where you hid the treats than your supposedly "average" mixed breed.

This is Heliox: Where Evidence Meets Empathy

Independent, moderated, timely, deep, gentle, clinical, global, and community conversations about things that matter.  Breathe Easy, we go deep and lightly surface the big ideas.

Thanks for listening today!

Four recurring narratives underlie every episode: boundary dissolution, adaptive complexity, embodied knowledge, and quantum-like uncertainty. These aren’t just philosophical musings but frameworks for understanding our modern world. 

We hope you continue exploring our other podcasts, responding to the content, and checking out our related articles on the Heliox Podcast on Substack. 

Transcript

0:25

This is Heliox, where evidence meets empathy. Independent, moderated, timely, deep, gentle, clinical, global, and community conversations about things that matter. Breathe easy. We go deep and lightly surface the big ideas. You know, have you ever looked at your dog and wondered if their breed really tells you something fundamental about how they think or act? Oh, definitely. Like, you know, do Labradors actually love people more than, say, a German Shepherd? Or are Border Collies always the undisputed sort of intellectual champions of the canine world? Right, the stereotypes. Well, today we're taking a deep dive into the fascinating science of dog cognition and get ready, because some of those long-held assumptions, they might be about to get a serious shakeup. That's right. And this isn't just, you know, a casual chat about dog behavior. We're really zeroing in on a truly groundbreaking study. One that offers a much clearer, I think, more nuanced picture of what makes different dog braids tick. And what makes it so different? Well, what really sets this research apart is its meticulous focus on individual breeds, rather than these broad, sometimes, frankly, misleading categories. Ah, I see. And as you'll find out, that really detailed approach makes, well, it makes all the difference. So where are these insights coming from? What's the source? It's a big study by Sara Janilla and her colleagues. It was published in Scientific Reports, a pretty well-regarded journal. Okay. The title is Breed Differences in Social Cognition, Inhibitory Control, and Spatial Problem-Solving Ability in the Domestic Dog. You get the idea. It's thorough. Sounds it. So our mission here today is to kind of pull out the most important nuggets from all that research. Exactly. We want to help you understand how different dog breeds genuinely stack up in areas like, say, problem solving or their social smarts. Even something as simple seemingly as how they greet a new person. Get ready for some real aha moments, I think. Okay, let's start at the beginning then. It feels like we've always had these ideas about breeds. Shepherds are smart. Hounds follow scents. But why did we need another study? What were the gaps? That's a really good question, because actually, if you look back, studies on dog cognition and behavior were often quite contradictory. Oh, really? Why was that? Well, there were a couple of big reasons. First, a lot of the older studies tended to group breeds by their original job, you know, like herding dogs or retrievers or maybe by how closely related they were genetically. Okay, that sounds kind of logical on the surface. What's the problem with doing it that way? Well, it sounds intuitive, right? But grouping them like that often just smoothed over or masked really important differences within those groups. Think about it. The pressures that lead to breeding change over time, you know? And just because dogs look similar or have similar ancestors doesn't mean they behave the same way now. Right. What a breed was first bred for might not be its main thing anymore. Precisely. So that broad brush approach could easily hide significant variations. It just wasn't detailed enough. I get it. So lumping them together obscures the diversity. Exactly. And the second big hurdle for a lot of past research was honestly sample size. Often they just had small numbers of dogs or only looked at a few breeds. Which makes it hard to say anything definite about breeds in general. Totally. You just can't draw broad, reliable conclusions. And this is precisely where Winchila and her team really aimed to make a difference, to bring some clarity. And the scale of this one really stands out. Over 1,000 dogs. 1,002, you said. That's huge. It's massive for this kind of research. Really impressive. And crucially, they didn't just lump them. They looked at 13 specific breeds plus a mixed breed group. Okay. And for each of those specific breeds, they made sure they had at least 40 dogs. 40 per breed. Wow. Yeah. So that rigorous individual breed focus allowed for these really detailed sort of apples to apples comparisons that, well, previous studies just couldn't manage. And it wasn't just a quick test, right? You mentioned a whole battery of tests. That's right. They used a standardized set called the SmartDotty test battery. Ten different tests in total. Ten. Yeah. Covering both social cognition, how they interact with humans, and non-social cognitive stuff, like problem solving, plus just general behaviors, too. Right. So it's giving a much fuller picture than just, say, obedience training success. Definitely. A much more holistic view of how dogs think and behave. So let's get into the nuts and bolts. How did they actually test these thousand plus dogs? What did it look like for the dogs involved? Okay, so the participants were mostly pets, you know, privately owned dogs. There are a few police dogs in the mix, too. And they were all between one and eight years old. The testing itself happened indoors in Finland. Owners were there, which is good for the dog's comfort. Yeah, keeps them relaxed, hopefully. Right. But the owners were told to stay quiet and basically neutral during the actual tasks. No helping. Ah, I bet that was hard for some owners. And motivation. I guess treats were involved. Oh, absolutely. Food rewards were key. Or sometimes a favorite toy if the dog wasn't super food motivated. Makes sense. And they actually advised owners not to feed their dogs right before the testing, just to make sure they were keen for those rewards. All right. Now, you said they also looked at general behaviors, not just the cognitive stuff. Yeah, they did these sort of behavioral snapshots. For instance, a greeting test. How did the dog react when an unfamiliar person came into the room? Okay, like were they scared or friendly or maybe too friendly? Exactly. They rated it on a scale. Fearful, indifferent, friendly, or overexcited. They also measured activity level using one of those FitBark monitors. Oh, yeah, like a Fitbit for dogs. Kind of, yeah. It gave activity points just a general measure of energy. And they looked at exploration, too. How willing was the dog to check out this new indoor space? Did it stick close to the owner, or was it running around sniffing everything? Okay, that gives a nice baseline. Now, let's dig into the brain teasers, the cognitive challenges, the cylinder test for inhibitory control. That sounds interesting. How does that work? It's a classic test, actually. Picture this. There's food visible inside a clear cylinder, like a plastic tube open at both ends. The dog sees the food. Its first instinct is probably just to try and push straight through the clear barrier. Yeah, we bash its nose on it. Exactly. But to actually get the food, the dog has to stop itself, inhibit that direct lunge, and realize it needs to go around to one of the open ends. Ah, so it's about resisting the immediate obvious impulse for a smarter indirect solution. Precisely. It measures what they call motor inhibitory response. Can they suppress that inefficient action for a more effective one? Got it. Impulse control. I can picture some breeds being better at that than others already. What about social smarts? The human gestures test. Oh, this is crucial for understanding how dogs work with us. Basically, the dog had to choose between two bowls. Only one had food. And the human helped. Well, yes. The human tester used different signals to show which bowl had the food. Things like pointing dynamically or just a quick point, pointing with their foot. Even just looking at the right bowl. Yeah, even just using their gaze. It tests how well they pick up on our human ways of communicating, which is a huge part of the dog-human bond. Absolutely. Then there's the V-detour. Spatial problem solving, you said? Sounds a bit like the cylinder. Similar concept, yeah. And it also taps into that inhibitory control. This time there's a transparent V-shaped fence between the dog and the food. Okay. So again, they can see the reward. But to get to it, they actually have to move away from it initially to go around the point of the bee. Ah, that's tricky. Moving away from what you want requires some thought. It does. It's counterintuitive, tests their spatial reasoning, and again, that ability to not just charge straight ahead. And the unsolvable task. I'm picturing some frustrated dogs here. Yeah. What was the goal of that one? Ah, yeah, maybe some were. It's a great test for looking at persistence, but also how much they rely on humans when stuck. How did it work? So first, the dogs solved an easy version of getting food out of a box. Simple. Then they were presented with the box again. They could see the food inside. But this time, it was impossible to open. It was sealed. Oh, great. A little bit. But the researchers watched. How long did the dog keep trying on its own? Did it try different strategies? Or did it quickly turn to the owner or the tester, like, hey, help me out here? Or did it just give up and wander off? It reveals a lot about their problem-solving style and their tendency to seek human help. fascinating and what about this memory versus gesture test that sounds like it really pits their own knowledge against what we tell them this one is really cool i think so the dog clearly sees the tester put food into one of two bowls no tricks okay so they know where it is they should know but then the human points emphatically towards the empty bowl uh a conflict exactly Does the dog trust its own eyes, its own memory of where the food went? Or does it follow the human's cue even though it's wrong? Wow. What does that tell us? It's a deep dive into their social cognition. How much weight do they give to human signals versus their own direct experience? It's kind of a test of trust in a way. That's really thought-provoking. And they also tested basic logical reasoning and memory too. Yeah, quickly on those. Logical reasoning was simple inference. They saw one bowl was empty, so the treat must be under the other one. And memory was a spatial short-term memory test. Hide a treat under one of three bowls, wait for increasing amounts of time up to two and a half minutes, and see if the dog remembers where it is. Right. Standard memory challenge. So, yeah, you can see it was a really comprehensive set of tests aiming to capture lots of different facets of how dogs think. Okay, we've got the setup. We've got the tests. Now for the big reveal. What did they actually find? Did the science confirm our stereotypes or throw them out the window? Well, it's a bit of both, really. The study found definite significant breed differences in quite a few areas. Like what? Like understanding those human gestures we talked about, following misleading gestures, that memory versus gesture test, spatial problem solving in the VD tour, inhibitory control in the cylinder test. Okay, so lots of differences there. Yeah. Also in persistence and how much they look to humans in that unsolvable task, how they greeted strangers, their general activity level, and how much they explored the new place. So yes, clear differences across breeds in those specific things. And did those differences seem to line up with what the breeds are, you know, known for? Their jobs? Often, yes. There were some really strong connections. Take inhibitory control, that cylinder test. The herding breeds, like border collies and Australian shepherds, they scored highest. They showed really strong impulse control. Which makes sense for herding, right? You can't just chase the sheep wildly. Exactly. You need to inhibit that predatory chase instinct and channel it. But then breeds like the Melanoi and German Shepherds, often used in roles needing really high responsiveness, quick action, they actually scored lower on that specific test. Interesting. So lower impulse control might actually be beneficial for their jobs, like protection or police work. Potentially, yeah. It suggests selection for different traits depending on the task. And what about looking for help from humans in that unsolvable task? Which breeds did more of that? Well, breeds like the Kelpie, Golden Retriever, Aussie Shepherd, and Border Collie spent more time looking to the human, seeking interaction. Again, breeds often kept as pets or working closely with people. Right. It fits roles needing that close human partnership pets, herding dogs that work tightly with a handler, retrievers working cooperatively. But then there's the flip side, the independent problem solvers. Exactly. And guess who topped that chart? Let me guess. German Shepherds in Malinois. You got it. They were the most likely to be completely independent in the unsolvable task, hardly spent any time asking for help. Which again, fits their roles, like detection work, where they need to work on their own initiative. Precisely. It suggests that independence is a trait that's maybe been selected for, or at least is valued in those kinds of working roles. Okay, so we do see clear breed differences, and they often connect quite logically to function. But you said it was a bit of both. Where didn't they find big differences? This is the surprising part, right? This is the really big headline, I think. The study found no significant breed differences in short-term memory or logical reasoning. Wow. No difference across those 13 breeds. None that were statistically significant in these specific tests. So that common idea that, you know, breed X is just generally smarter or has a better memory than breed Y across the board, this research really challenges that. So my Labrador might be just as good at remembering where I hid his toy as a border collie, even if the collie aces the impulse control test. Based on the study's findings for short-term spatial memory and logical inference, yes. It suggests intelligence isn't just one single scale. It's more like a toolbox with different tools, and breeds might excel at using different ones. That is a huge aha moment. It really reframes how we should think about dog intelligence. It's much more specific. Absolutely. And what's fascinating is how this points towards artificial selection. Meaning how humans bred them. Exactly. The findings strongly suggest that these traits where they did find differences, social cognition, problem-solving style, inhibitory control, These have likely been under what scientists call diversifying artificial selection. Diversifying meaning. Yeah. We've pushed breeze in different directions. Precisely. Humans have maybe intentionally, maybe unintentionally, bred dogs for specific cognitive skills based on what we wanted them to do. We've literally shaped their minds for different jobs over generations. We've sculpted their brains. Yeah. But you also hinted it's not a perfect match, right? It's not as simple as herding dogs are like X, retrievers are like Y. That's such a vital point. While lots of the results do align nicely with breed function, it's definitely not a neat, tidy, one-to-one thing. This study really hammered that home. How so? What were the inconsistencies? Well, for example, even within the herding group, there was loads of variation. The Finnish Laffin, also a herder, it scored lowest in understanding human gestures. Lowest, but other herders were high. Right. Kelpies and Malinois, also technically herding breeds, were among the highest goals on those same gesture tests. Huge difference within one functional group. Wow. Okay. That completely blows apart the idea of just grouping them by job. It really does. And another one, the Australian Shepherd was super persistent in the unsolvable task, like almost never gave up. Right. But the Kelpie, also a herder, was actually one of the breeds most likely to abandon the task. Polar opposites, behaviorally, in that context. Yeah. But both herders. Exactly. And even look at retrievers. The golden retriever was quite different from the Labrador retriever in how they handled the unsolvable task and the gesture tests. Even though they're both retrievers, bred for similar things, seemingly. Right. So if you pull all this together, what's the big takeaway message here? It sounds like looking at broad categories like herding or retriever just doesn't cut it if you want to understand dog cognition properly. That's absolutely the core message. You risk missing incredibly important distinctions, real variations, if you only look at those broad groups. To get a genuine understanding, you have to investigate individual breeds. Grouping them can really obscure the fascinating diversity that's actually there. Makes perfect sense. Yeah. Now, we should also touch on limitations. No study is perfect. What should we keep in mind about this one? Good point. Well, first, the dogs were mainly from Finland, and many were involved in dog sports. So they might not be perfectly representative of all pet dogs worldwide. Okay, the population might be a bit skewed towards enthusiasts. Possibly. Also, while they found these breed differences, the study couldn't completely control for every single dog's individual training history or life experiences, you know. Right. Nature versus nurture is always tricky. But does other research give us clues about how much is genetics versus training for these traits? That's a key question. And yes, other studies do suggest that genetics plays a pretty substantial role. things like understanding human gestures, inhibitory control. Research indicates genetic relatedness accounts for a good chunk of the variation we see. And even when studies have tried to control for training history, significant breed differences still pop up in many of these cognitive tasks. So while training matters, there seems to be a strong underlying genetic component to these breed tendencies. Got it. And just to be clear, the fact they didn't find differences in memory or logic here doesn't mean those differences don't exist in any dogs. Right, exactly. That's crucial. This study looked at 13 specific breeds. It didn't include, say, ancient breeds or hounds or terriers. Maybe if you tested those groups, you would find differences in memory or logical reasoning. So more research need there, as always. As always, absolutely. Okay, so let's bring this home. What does all this mean for you, the listener, maybe cuddling with your own dog right now? Well, we've taken this deep dive right into what makes different dog breeds cognitively unique or sometimes surprisingly similar. Yeah, from impulse control to asking for help, even just saying hello to a stranger. We've seen that, yes, significant measurable differences do exist between individual breeds. And often these differences kind of echo their history, their original purpose. But, and this is the big but, we've also seen where those breed stereotypes might just fall apart. especially in areas like general memory or that specific kind of logical reasoning they tested. Right. And this isn't just, you know, trivia for dog lovers. It's really about understanding the specific cognitive blueprint of your dog, your companion. How does understanding that help? Well, it can genuinely help you appreciate your dog's natural tendencies more. Maybe understand why they struggle with certain things or excel at others. It helps predict their behavior better. And it can even help you make more informed choices if you're thinking about getting a specific breed, matching their cognitive style to your lifestyle or what you want in a dog. It fosters a better relationship, really. A more informed, maybe more harmonious relationship by understanding their innate settings, so to speak. Well put. Okay, so final thought to leave everyone with. If we humans have spent literally millennia shaping dog minds through selective breeding for all these specific tasks, what does this ongoing evolution, this partnership in intelligence, tell us about our own unique bond with dogs? And thinking ahead, you know, as our understanding deepens, what cognitive traits might we, maybe even unconsciously, be selecting for in our dogs next? What does the future of the canine mind look like shaped by us? Thanks for listening today. Four recurring narratives underlie every episode. Boundary dissolution, adaptive complexity, embodied knowledge, and quantum-like uncertainty. These aren't just philosophical musings, but frameworks for understanding our modern world. We hope you continue exploring our other podcasts, responding to the content, and checking out our related articles at heliocspodcast.substack.com.