Written by Ash B. · 7 min read

If you’ve ever watched your dog lose their mind over a treat you tucked under a couch cushion, only to see them stare blankly when you moved that same treat somewhere they didn’t see, you’ve already run an informal cognition experiment. What you’re testing, whether you meant to or not, is object permanence, the understanding that an object still exists even after it disappears from view. This isn’t a cute party trick question. It’s a core piece of how a brain models the world, and dogs turn out to be genuinely interesting subjects for studying it.

What Object Permanence Actually Means

The concept comes from Piaget’s work on human infant development, where babies move through predictable stages of understanding that things don’t cease to exist just because they’re out of sight. Early on, infants act like “out of sight, out of mind” is a literal law of physics. Eventually they develop a full mental representation of objects and their locations, even through complex, unseen movement.

Researchers borrowed this framework to test animals, dogs included, because it’s a clean way to measure something otherwise invisible: whether an animal is working off pure sensory input in the moment, or actually holding a representation of the world in its head when the sensory input goes away.

Dogs land somewhere past the earliest infant stages, and that’s not a small thing. A lot of animals never get there at all.

The Easy Test: Visible Displacement

In a visible displacement task, the dog watches an object get hidden, say, a treat placed under one of several cups, in full view. No trickery, no delay behind a screen. The dog just has to remember where the thing went and go get it.

Dogs are reliably good at this. It’s not a coin flip performance, it’s consistent, confident retrieval. This tells us dogs form short term spatial memory of an object’s last known location and can act on that memory purposefully, rather than just reacting to whatever’s directly in front of them.

This is the same skill behind a dog watching you walk into another room with the leash and immediately parking themselves by the door. They’re not confused that the leash left their visual field. They know it still exists, and they know roughly where it’s headed.

Where It Gets Interesting: Invisible Displacement

Invisible displacement is the harder version. Here, an object gets hidden inside something (a cup or container), that container is moved behind a screen out of the dog’s sight, the object is secretly deposited somewhere, and then the now-empty container reappears. To solve it, the dog has to infer that the object was left behind somewhere it never actually saw the object go.

This is where performance gets messier. Some dogs can solve single-step invisible displacement, which already puts them ahead of a lot of other species. But stack multiple invisible moves in sequence, and accuracy drops. The mental math of “it was in there, that went behind there, therefore it must be there” is a heavier cognitive lift than tracking something you watched the whole time.

One classic error shows up here too: dogs will often search the location where they last found the object before, even after clearly watching it move somewhere new. This is called the A-not-B error, and it’s not just forgetfulness. It looks a lot like a fight between memory of the new location and a strong, kind of stubborn habit of “that’s where the good thing was last time, so that’s where I’m digging.”

Egocentric Memory vs. True Representation

A big question in this research isn’t just whether dogs can solve these puzzles, it’s how. There’s a difference between a dog holding a genuine mental map of “the object is at that specific location in space” versus a dog just remembering “I was moving in this direction when I lost track of it, so I’ll keep moving that way.”

The second strategy is called egocentric tracking, meaning it’s based on the dog’s own body and movement rather than an independent map of the environment. It’s a perfectly functional strategy a lot of the time, which is exactly why it’s hard to tell from the outside whether a dog “understands” object permanence in the deeper sense or is just very good at dead reckoning based on its own motion.

This matters because it changes what we’re actually impressed by. A dog nailing an invisible displacement task might be doing sophisticated mental representation, or might be running a simpler but still pretty effective spatial trick. Both are legitimate cognitive tools. They’re just not the same tool.

The Domestication Wildcard: Reading Us Instead of the Object

Here’s the part that makes dogs a weird and useful case study compared to, say, primates working on the same tasks: dogs are exceptionally responsive to human social cues, like pointing or gaze direction, and will use those cues to solve a search problem instead of relying purely on their own tracking of the object.

This isn’t cheating, cognitively speaking. It’s arguably a different kind of intelligence, one that’s been selected for hard across domestication. A dog that watches your hand and your eyes and updates its search accordingly is solving the same practical problem (finding the missing thing) through a completely different channel than a dog doing pure spatial memory work.

It also means that when dogs “fail” a pure object permanence test, it doesn’t necessarily mean the underlying capacity isn’t there. Sometimes it means the dog is waiting for you to just tell them where it is, because historically, that’s worked out great for them.

What This Actually Means for Separation Behavior and Boarding

This is where the science stops being trivia and starts being relevant to actual dogs in actual homes. Object permanence for a physical treat is one thing. Whether a dog holds any equivalent representation about a person who has left the house, or left them at a boarding facility, is a related but separate question, and one this research doesn’t fully answer.

What we can say is that search behavior, pacing, and distress at a person’s departure aren’t proof the dog thinks the person has ceased to exist. It’s more accurately described as an attachment and arousal response, not a cognitive failure to grasp permanence. A dog can understand an object or person still exists somewhere and still be distressed by the separation itself. Those are two different systems, one cognitive, one emotional.

This is a distinction worth sitting with if you’ve ever felt guilty leaving a dog who whined at the door. The whining isn’t evidence of confusion about whether you still exist. It’s a normal response to a relationship they value being temporarily out of reach.

Practical Ways to Use This

None of this is just academic. It has real applications for enrichment and training.

• Hide and seek games with visible displacement (dog watches you hide it) build confidence and satisfy search drive with a high success rate.
• Invisible displacement puzzles, like snuffle mats or puzzle feeders where the dog can’t see the food go all the way in, are genuinely harder mental work, not just physical work. That’s a feature, not a bug, for dogs who need more cognitive load in their day.
• If your dog keeps searching an old hiding spot instead of the new one, that’s not stubbornness or a “dumb dog” moment. That’s a documented pattern (the A-not-B tendency) and it usually resolves with a few more repetitions.
• Dogs that seem to rely heavily on watching your hands or face during a search game are using social cues as a legitimate strategy, not failing to think for themselves.

Understanding this also just makes you a better observer of your own dog. Watching whether they solve a problem through spatial memory, through scent, or through staring at your face waiting for a hint tells you something real about how that particular dog’s brain likes to work.