Do Dogs Lie? What a Food-Box Study Really Shows

Author

Luboslav

Published

October 5, 2026

Reading time

5 min read

Topics

ContextObservation Method

Your dog nudges an empty bowl shortly after breakfast. To one person, that means hunger; to another, a shameless attempt to claim a second serving. The action is visible, but the story about what the dog believes is ours.

A food-choice study found misleading behavior, but interpreting it as human-style lying requires another step. [S1] For everyday episodes, start with what happened and what it achieved before guessing intention.

▶Play video

Three questions hidden inside the word lie

First, could a familiar action have produced a useful result before? If tapping a bowl once brought food, repeating that action can have an ordinary practical explanation. You do not need to decide that the dog is making a false statement.

Second, did the action actually mislead someone? A person might interpret the bowl tap as evidence that breakfast was forgotten. The effect can be misleading even when the dog's understanding of that interpretation remains unknown.

Third, did the dog intend to create a false belief? That is a stronger claim than either repetition or a misleading outcome. Treat these as separate questions rather than allowing the funniest explanation to answer all three.

The experiment offered a reason to misdirect

In a 2017 study, Marianne Heberlein and colleagues trained dogs with two unfamiliar people: one gave them food; the other kept it. Dogs could lead a person to three locations containing preferred food, less-preferred food, or nothing. Afterwards, they could lead their cooperative owner to a location, creating another opportunity to obtain food. [S1]

On the first test day, dogs took the cooperative person to preferred food more often than chance predicted and more often than they took the competitive person there. On the second day, preferred-food choices with the competitive person fell below chance; dogs also led that person to the empty location more often than they led the cooperative person there. The authors interpreted these results as tactical deception. [S1]

A useful result is not a window into a dog's mind

The task shows different choices with different partners. It does not establish what dogs understood about human beliefs or explain every misleading household episode. This is an interpretation of the study's scope. [S1]

You can recognize an effective strategy without treating it as a character flaw. A dog does not need to be either an innocent machine or a miniature con artist for its behavior to deserve careful attention. Those labels make a good punchline, but neither tells you what to change.

Two ordinary scenes, several possible explanations

Consider an illustrative household with two people feeding the dog. One serves breakfast; the other later sees the dog waiting beside the cupboard and adds a second portion. Before calling this a scam, ask whether both people have a clear record of the meal. Waiting by the cupboard tells you where the dog is, not whether it is claiming to be unfed. A shared feeding record resolves the human uncertainty.

In another invented example, a dog carries a toy toward the back door and a person opens it. The dog then abandons the toy and investigates the garden. Perhaps the person inferred a request to play when the dog wanted access outside. Perhaps the toy was simply part of a familiar sequence. The change in activity is worth noticing, but it cannot settle intention on its own.

Change the arrangement instead of prosecuting the dog

For repeated food requests, agree who feeds the dog, what has already been served, and what happens when another request arrives. Keep food out of reach and provide a clear alternative, such as settling nearby while you finish a task. If you reward that alternative, include edible rewards in the planned food allowance rather than adding an untracked second meal.

Review whether everyone is counting the same portions, including snacks. A record prevents both duplicate feeding and the opposite mistake: assuming somebody else fed the dog when nobody did. Blaming the dog can leave the real problem untouched: a person supplied food without checking the routine.

For confusing access requests, make the decision explicit: choose an appropriate outdoor break or offer play indoors. You do not need to stage a trap to catch the dog lying. Consistent responses give you a more useful basis for comparison than scolding, teasing, or repeatedly changing the rules.

Describe the pattern before naming the motive

Replace “he fooled me again” with a precise observation: “He waited at the cupboard after breakfast, and the second person gave him food.” Then ask what happens when the feeding record is clear. Does the behavior recur with both people? Does it occur only near mealtime? What response follows it?

Those questions can guide a practical change without pretending to read the dog's thoughts. When a request becomes clearer, the useful achievement is better communication and fewer accidental outcomes. Winning an argument about whether your dog is morally honest adds little to that result.

For a concise way to record your own observations, use the Dog Behavior Context Checklist.

Sources

[S1] Heberlein, Marianne T. E.; Manser, Marta B.; Turner, Dennis C. (2017). Deceptive-like behaviour in dogs (Canis familiaris). Animal Cognition, 20, 511–520. https://doi.org/10.1007/s10071-017-1078-6