In this article
Two dogs tumble across the grass. There are teeth, flying paws and a soundtrack that sounds rather dramatic. Then they spring apart and dive back in. From the path, it can be hard to decide whether you are watching a lovely game or the beginning of trouble.
Dogs Trust advises reading the whole body and the situation. Play can include growling, barking and gentle mouthing, alongside loose, exaggerated movements. Sound alone cannot settle the question1.
Watch the conversation between both dogs
Dogs Trust lists bouncy bodies, changing roles, matching movements and breaks. Interrupt repeated pinning, intense one-sided chasing or difficulty responding to an owner1.
Imagine watching a brief clip that begins with one dog already on its back. You have missed how the pair arrived there and what happens next. Hold the same caution in the park: the few seconds before and after an exciting moment belong in your assessment too.1
What a play bow can tell us
Front end down, bottom up: the bow is one of the easiest pieces of dog conversation to recognise. In a study at an Austrian research centre, researchers examined 136 bows during play between 10 dog puppies. In 135 cases, both puppies were within each other's visual field2.
The puppies were about two to five months old, hand-reared in peer groups. The researchers coded behaviour immediately before and after bows already occurring within a game. They were studying a particular sample and situation, rather than every kind of invitation a pet dog might offer2.
Bows tended to follow pauses and precede more active play. The findings supported restarting play; they did not support the prediction that the bow clarified potentially ambiguous offensive behaviour. A bow is therefore useful information, with limits to what this experiment establishes2.
An earlier study of adult pet dogs also examined what the two partners did around a bow. Bows most often followed brief pauses. Afterwards, the dogs commonly moved together or performed complementary actions, such as one running away and the other chasing3.
The authors likewise interpreted this pattern as reinitiating play. Their published abstract reports a comparison of behaviours around a signal. It does not supply a test in which every subsequent chase was checked against the owner's judgement that both dogs remained comfortable3.
That distinction matters when you apply a research finding. “A bow happened” answers a question about the sequence you saw. “Both dogs are still comfortable” asks you to consider the continuing interaction. A study of one signal's timing cannot finish that second assessment for you3.
Dogs sometimes copy a friend's play face
Another study filmed 49 dogs in a Palermo dog park. Researchers looked for rapid copying of play bows and relaxed open-mouth expressions, with a matching response occurring within a second. The central response comparison used 21 dogs with enough observations for the required comparisons4.
Those dogs matched a bow with a bow, and an open-mouth expression with the same expression, more often than the comparison movements predicted. Copying was more frequent between friends than less familiar partners. Sessions containing rapid mimicry also lasted longer4.
It is an appealing glimpse of how finely a game can be coordinated. However, observing longer play alongside copying cannot show which causes which. The authors could not disentangle that relationship. Their proposed connection with emotional sharing remains an interpretation, rather than a direct reading of either dog's feelings4.
There is no need to time your dog's face with a stopwatch. The research helps explain why play can look like a conversation, while leaving the everyday question of whether this particular encounter should continue open4.
Taking turns is useful, without a scorecard
In a separate analysis of 203 dog-park play sessions, researchers examined how uneven play was and how long it lasted. More asymmetrical sessions were shorter. Familiar and unfamiliar partners did not differ in their overall level of asymmetry5.
This challenges an overly tidy picture in which every friendly game must divide each action equally. The study's abstract describes changing relationships between play patterns and duration. It provides no universal percentage of chasing, winning or being underneath that owners must count5.
The authors suggested shorter sessions might help dogs manage uneven play and limit escalation. That is an explanation of the observed pattern. It should not be retold as proof that uneven play is harmless, or that dogs will always stop before anyone gets hurt5.
A review of dog play also cautions against treating play as a general certificate of good welfare. The authors describe different functions for play and stress the importance of the animals' histories and the setting. Its meaning depends on more than the presence of playful behaviour6.
So an enthusiastic wrestler and a willing partner are one situation; the same wrestler paired with a dog who keeps trying to leave is another. The review's warning about context is a reason to keep that distinction visible, even when an activity has the familiar label “play”6.
You do not need to settle the evolutionary purpose of a game before making room for an individual dog. The useful lesson from the review is that a broad category cannot stand in for an assessment of this dog, with this partner, here6.
Choose a suitable meeting, then allow a break
Battersea recommends asking the other owner before a greeting, respecting a refusal and matching play styles. Before off-lead meetings, a reliable recall matters. A dog who enjoys one companion need not be introduced to every dog you pass7.
Battersea describes balanced games with short breaks and recommends rewarding your dog for coming away. A growl deserves attention rather than punishment7. For repeated anxiety or aggression, arrange a vet check before help from a qualified behaviourist8.
A calmer interruption
You can decide that a meeting has been enough without naming a culprit. “We're going to walk on” is a complete explanation to another owner. There is no prize for extending the game until the dogs, or the people watching them, have reached their limit.
When the game needs to end
PDSA lists discomfort signals including turning away, trying to move off and crouching. Snarling, snapping or lunging call for space. If another dog is behaving aggressively, move away calmly and promptly; where possible, use a fence, gate or car as a visual barrier8.
A dog who has been bitten needs veterinary attention straight away, even when the surface injury seems minor: deeper damage can be hidden. Arrange a vet check for a sudden change towards aggression too, so possible underlying health problems can be assessed8.
A useful park-side decision does not need a perfect diagnosis of the encounter. You can notice that one dog needs more room, call time while it is still manageable, and take your own dog somewhere quieter. Enjoying play includes giving it an ending.7
Sources and further reading8 sources
The small numbers in the article link to the source behind each fact. Sources open on their own websites.
- How to understand your dog's body language
- Investigating the Function of Play Bows in Dog and Wolf Puppies (Canis lupus familiaris, Canis lupus occidentalis)
- Investigating the function of play bows in adult pet dogs (Canis lupus familiaris)
- Rapid mimicry and emotional contagion in domestic dogs
- Unveiling the "secret" of play in dogs (Canis lupus familiaris): Asymmetry and signals
- Why do dogs play? Function and welfare implications of play in the domestic dog
- How to introduce your dog to other dogs on walks
- Dealing with a dog attack