I’m working at the desk in my office in the early afternoon. Behind me, my dog, Lily, begins to stir, knowing that we typically take a play break around 1pm. I ask her to find her frog toy. She runs up three flights of stairs and returns quickly victorious. We play some tug and then do a training session where she’s learning to discriminate between different directions. Later, we go out into the backyard. A gust of wind blows the cover of the barbecue pit. Lily startles, hackles raised, and races back inside. Lily, I think, you know that’s a barbecue pit, which has been there for the two years you’ve lived with us, just as you know what time it is and where your frog is and which way is left or right. But does she?

Dog playing with puzzle toy — How do we know if a dog knows something? We don’t have access to a dog’s inner thoughts or feelings, so we have to puzzle out how knowledge is at play in something we can observe, namely the dog’s behavior.

Wait, though! If we’re trying to observe how knowledge shows up in behavior, we should probably define what knowledge is in the first.

So, that’s a bit complicated. My background in philosophy reminds me that thinkers have been debating the nature of knowledge, let alone its definition, for centuries. Yet, many people have a pretty good idea of what it means to know something. For our purposes in this post, we can think of knowledge in terms of cognition, insight, and understanding. Following Sara Shettleworth (2001), we can think of cognition as including “perception, learning, memory and decision making, in short all ways in which animals take in information about the world through the senses, process, retain, and decide think on it.” In short, cognition involves perceiving something through the senses, such as temperature, intensity of sound, and odor profile. It also involves processing that sensory information, remembering it, and acting on it.

Here, we can use nose work as a good example. When we first introduce our pups to a primary reinforcer, such as a treat, in boxes, they begin to form an association between food and boxes. So, when they come into an area that has boxes strewn about, they begin to search for food. As we begin introducing different environments, substrates, and other contexts, they start to problem solve – they (ideally) begin to apply things learned in other contexts to the one they’re in now to achieve the goal of sourcing the food. When we introduce odor, we make it salient to the dog by first pairing it with food. Now, the odor of birch is associated with hot dogs and predicts that hot dogs might become available. Now, the dog recognizes that a trailer hitch, a toolbox, a picnic bench, or a podium might all be the source of odor and therefore also predictor of hot dogs or a favorite tug toy. The dog learns to pay attention to thresholds and move out wide to come back into the scent cone. To be sure, we don’t know that a dog recognizes a hitch as a hitch or a toolbox as a toolbox or even a hot dog as a hot dog or the scent cone as a scent cone.

This is a tricky part. We can directly observe what seem to be cognitive processes, such as the dog recognizing the relationship between cause and effect. Why else might my dogs park under my toddler niece’s highchair in anticipation of dropped food? What we--the human we--can’t tell is the degree of the dogs’ consciousness. As Shettleworth (2001) reminds us,

This does not mean it is impossible to study patterns of behavior in other animals analogous to behavior accompanied by distinct states of awareness in humans, but such work eventually meets an impenetrable barrier: the animals cannot report verbally on their experiences.

They can’t tell us what and how they think or think about thinking, or, at least, we haven’t yet figured it out if they are already communicating that. So, when talking about knowledge, we have to be careful about attributing more than we have evidence for, but we should also be careful of denying what the animal may actually be showing us.

We have all seen our dogs problem solve, especially in nose work. What we haven’t seen, though, is what the dog thinks about that problem solving. We might see our dogs’ loose, wiggly bodies when we pull out our odor kit, or we might see some tight mouths and flattened ears when motorcycles rev past the search area. We can recognize that dogs seem capable of some degree of self-awareness. For example, when dogs were tasked to retrieve an object attached to a mat the dogs were standing on, many dogs recognized their body as an obstacle and moved out of the way to retrieve the item (Lenkei et al. 2021). So, we can point to evidence that dogs do have some kind of self-awareness or consciousness (a whole nother complicated topic!) at play in problem solving. We still don’t know, though, the extent of that awareness or the degree to which it is analogous to human experience (de Waal and Ferrari 2010).

Still, that problem solving capacity and flexibility in different environments suggest that dogs are not simple input-output machines. They can be creative and innovative. They seem to demonstrate some kind of understanding by being able to respond to novel environments. Many of us have probably already figured that out for ourselves, so what does this matter?

I think it matters because we owe it to our dogs to recognize what they might really be capable of. In nose work, the human end of the leash gets to stand back and watch dogs masterfully solve problems and lend a bit of support when they get stuck. We have witnessed our dogs gain confidence from the first moment they’re willing to put their head in a box to clearing an auditorium effortlessly.

Yet, we’ve also probably all been in a search where the dog seems confused, where we feel compelled to cue them multiple times to search or get to work, frustrated that they know how to do this and aren’t doing it. And maybe we’re also a bit embarrassed! Here’s where our dogs can actually teach us a lot. Dogs typically approach the world with a lot of curiosity, and they’re more able to be curious when they are supported (Merola et al. 2011). So, how can we be curious about what they’re showing us? Have they had enough experience searching in this kind of environment? Did they have an upsetting parking lot experience? Have they noticed the crows in the trees more than the presence of clove? Do they have a hot spot under their harness? Do they just need a minute to collect themselves? Are we going into the search enabling them to choose how the search will go? How can we better work together as a team (Topál et al. 1997)?

Our dogs are already capable of a lot, and we probably already knew that. But what else can we discover?

References

de Waal FBM, & Ferrari PF. Towards a bottom-up perspective on animal and human cognition. Trends Cogn Sci. 2010; 14: 201–207.

Lenkei, R., Faragó, T., Zsilák, B. et al. Dogs (Canis familiaris) recognize their own body as a physical obstacle. Sci Rep 11, 2761 (2021). https://doi.org/10.1038/s41598-021-82309-x

Merola, Isabella & Prato-Previde, Emanuela & Marshall-Pescini, Sarah. (2011). Social referencing in dog-owner dyads?. Animal cognition. 15. 175-85. http://doi.org/10.1007/s10071-011-0443-0

Shettleworth, S. J. (2001). Animal cognition and animal behaviour. Animal Behaviour, 61(2), 277–286. https://doi.org/10.1006/anbe.2000.1606

Topál, J., Miklósi, Á., & Csányi, V. (1997). Dog–human relationship affects problem-solving behavior in the dog. Anthrozoös, 10(4), 214–224. https://doi.org/10.2752/089279397787000987

What and How do We Know What Dogs Know?