What does your dog understand about your facial expressions?

You have had a difficult day. You sit down, your dog approaches, and you think, “He knows I am upset.” It is a familiar experience, and the companionship can mean a great deal. But what has your dog actually noticed?

Perhaps your expression changed. Perhaps you moved more slowly, spoke differently, or sat somewhere unusual. Several explanations can fit the same interaction. Research into dogs’ responses to human faces helps us examine one part of that question, while leaving room for what we still do not know.

A study by Hernández-Pérez and colleagues, published in iScience in 2026, used brain imaging to investigate responses to photographs of human facial expressions. The dogs were awake and trained to remain still voluntarily during scanning. This allowed researchers to compare brain activity while different expressions were presented.

One experiment involved eight dogs viewing happy and neutral faces. The responses differed. A second experiment involving 12 dogs compared happiness, anger, fear, and sadness. Analyses distinguished happiness from negative expressions, and also fear from anger and sadness. Anger and sadness were not clearly distinguished from each other in that analysis.

These findings support the ability to discriminate between some expressions under the study conditions. They do not demonstrate that dogs understand what caused the photographed person’s emotion, or that the dogs themselves experienced the same feeling.

For me, that distinction is central to interpreting the research. Detecting a difference, identifying its meaning, understanding its cause, and deciding how to respond are separate questions. A brain scan cannot answer all of them simply because the activity changes when the picture changes.

Think about a person looking at a photograph of someone frowning. They may see that the expression differs from a smile. They might describe it as anger, concentration, discomfort, or confusion. Without the surrounding situation, they may be unsure. We should allow similar uncertainty when interpreting what a dog understands from a face.

The small samples also matter. Dogs able to participate in this type of experiment have undergone substantial training. Their performance offers useful evidence, but does not tell us how every dog, at every age and in every situation, will respond. It would be a mistake to turn a group finding into an expectation that an individual dog must meet.

Still photographs create another boundary. At home, a facial expression arrives with movement, voice, distance, and a history of interactions. Those features can change together. If your dog steps away when you frown and lean forward, the observation alone cannot tell us which feature mattered most.

Owners sometimes ask whether they need to watch their faces constantly around their dogs. I would not take that message from this study. Ordinary life includes tiredness, frustration, sadness, and joy. This experiment did not test whether hiding those expressions improves a dog’s welfare.

Instead, I would use the research as an invitation to describe interactions more carefully. “She knew I was angry” contains an interpretation. “When I stood up and spoke, she moved behind the sofa” gives us a sequence we can examine. Both descriptions may come from an attentive owner, but the second leaves more possibilities open.

Imagine finding a torn cushion and looking towards your dog. The dog lowers its body and looks away. It is easy to combine those events into a story about guilt. Yet this facial-expression study did not investigate guilt, understanding household rules, or remembering a particular action. It cannot settle what that response means.

The same care applies to more comforting interpretations. A dog resting beside someone who is crying may provide welcome companionship. We can appreciate that moment while remaining uncertain about the dog’s understanding of the tears. Scientific caution does not make the relationship less meaningful.

When I discuss these situations with owners, I want to know what happened before the response and what happened next. Was the dog already nearby? Did the person call them? Did the dog approach freely or remain at a distance? Did contact continue comfortably, or did the dog withdraw?

These questions are useful because they anchor the conversation in events. They also make it easier to notice variation. A dog might approach on one occasion and move away on another. Rather than calling that inconsistency or lack of affection, we can ask what differed between the two situations.

A simple written description can help. Record the setting, what you did, and what the dog did. Keep the account short enough to maintain. There is no need to stage an emotional scene or deliberately make the dog uncomfortable to obtain an example.

If an ordinary interaction is recorded, a short video may help clarify the sequence. It should show enough context to understand the movement of both dog and person. A close-up of one facial expression can be compelling, but may leave out the information needed to interpret the exchange.

There is also a practical communication question. When asking a dog to do something, consider whether your words, movements, and expectations are clear. If the response surprises you, review the whole interaction before assuming the dog understood your expression and chose to disregard it.

That is a practical interpretation, rather than a training programme established by this experiment. The study compared brain responses to pictures. It did not test a particular way of teaching dogs, diagnose behaviour problems, or show that changing facial expressions alone resolves difficulties at home.

If your dog regularly appears uncomfortable during interactions, describe those situations to your veterinarian. Bring specific observations instead of feeling that you must already know the explanation. An account of what happens is useful even when you cannot say why it happens.

What interests me about this research is the more precise question it allows us to ask. We have evidence that dogs can distinguish certain facial expressions. We can now investigate how that information is used alongside everything else a dog encounters.

The next time your dog looks at your face, there is room for both affection and curiosity. Notice the response, consider the surrounding situation, and resist filling every gap with certainty. Understanding our dogs starts with giving their behaviour careful attention.

Reference

Hernández-Pérez et al. (2026). Dog brain representations of human facial expressions: Encoding happiness and differentiating specific negative expressions. iScience. DOI: 10.1016/j.isci.2026.116900