From Hydraulic Drives to Fluent Skills: Why Some Behaviour Problems Don't Go Away

For a long time, dog trainers have leaned on Lorenz's hydraulic model to explain instinctive behaviour. The idea was simple. Dogs build up internal pressure for certain behaviours, and if that pressure isn't released, it leaks out elsewhere as problem behaviour.

Historically, that helped trainers make sense of predation, drive, and vacuum behaviours. It gave us a way to explain why suppressing behaviour often made things worse.

But the model doesn't actually fit what we see in practice.

It assumes behaviour needs releasing. It assumes that expression reduces future motivation. It treats behaviour like a substance that fills up and empties out. None of that maps well onto how learning, repetition, and biology interact in real dogs.

A more useful way to think about this: skills and fluency

Dogs are not born neutral. They arrive with behaviours that are already fluent at a biological level. These are behaviours the dog doesn't need to learn, rehearse, or be taught to enjoy. They are already good at them.

A Border Collie is fluent at controlling movement. A guarding breed is fluent at scanning and alerting. A bully type is fluent at grabbing and holding.

Fluency matters because fluent skills behave differently to trained behaviours.

You do not practise a fluent skill to empty it. You practise it to sharpen it. Repetition increases speed, confidence, and automaticity. This is why allowing unlimited rehearsal of intrinsically reinforcing behaviour often makes the problem stronger, not weaker.

At the same time, trying to suppress a fluent skill entirely doesn't remove it. It simply means the dog never learns when, where, or how to use it appropriately. Under stress, arousal, or uncertainty, the dog defaults to its strongest, most fluent behaviour.

That is what people often call vacuum behaviour. In reality, it is unregulated fluency.

The dog is not leaking pressure. The dog is falling back on what it already knows how to do best.

How this changes the way we approach common problems

Take jumping up.

Jumping is often intrinsically reinforcing. It works. It gets contact. It creates engagement. Trying to eliminate it completely usually fails, especially in social, people-oriented dogs.

Instead of asking "How do I stop jumping?" the better question is "Where is jumping allowed, and what replaces it everywhere else?"

This is where mutually exclusive behaviours come in, what most trainers know as Differential Reinforcement of Incompatible Behaviour, or DRI.

A sit is incompatible with jumping. A dog cannot jump and sit at the same time. By teaching and reinforcing sit in greeting contexts, you are not suppressing jumping through force. You are giving the dog a clear alternative that physically cannot coexist with the problem behaviour.

And jumping doesn't have to disappear from the dog's life entirely. It can still exist in appropriate contexts. Play, sports, roughhousing, consent-based interaction. The skill is not removed. It is regulated.

The same principle applies across behaviour work

Herding dogs do not stop controlling movement. They learn when control is allowed and when it is not. Guarding breeds do not stop alerting. They learn discrimination and off-switches. Dogs with strong grab-bite tendencies do not lose that skill. They learn rules, thresholds, and context.

The goal is not release and it is not suppression. The goal is regulated fluency.

That means deliberately practising the behaviour in specific contexts, at controlled intensity, with clear start and stop cues, paired with incompatible behaviours, management, and a solid learning history everywhere else.

When this is done well, you see a clear trajectory. The behaviour appears in fewer inappropriate contexts. It becomes easier to interrupt. The dog recovers faster under pressure. Flexibility increases rather than decreases.

That is how you know you are working with the dog's biology instead of fighting it.

Moving beyond the hydraulic model

Lorenz's model helped us get part of the way there. But a fluency-based framework gives trainers something better. It respects instinct. It respects learning. And it gives you practical levers you can actually pull in real cases.

Dogs are born with skills. Your job is not to pretend they are empty. Your job is to teach them when those skills belong.

This is exactly the thinking behind the EPIC protocol in the Behaviour Bible. EPIC (Enhance, Prevent, Inhibit, Control) gives trainers a structured way to work with intrinsically reinforcing behaviours without suppressing them or allowing uncontrolled rehearsal. If you want to go deeper on predation, drive, and behaviour change that actually fits real dogs check the Behaviour Bible.