If Counter Conditioning Does Not Remove Old Learning, What Can It Do?
Extinction and counterconditioning don’t remove old learning — they add new learning that only works reliably when the brain is reminded to use it.
One of the most common misunderstandings in behaviour training is the idea that once a fear or unwanted behaviour has been “trained out,” it is gone for good.
However, Bouton (2002), Lindsay (2005) Pavlov (1927), Konorski (1948),Pearce and Hall (1980) and Wagner (1981) show this is not how the brain works. Instead, the research show’s that extinction and counter-conditioning do not erase old learning. Rather, they create new learning that exists alongside the original association. Whether the new or old behaviour appears depends heavily on context and memory retrieval, not willpower or stubbornness
What Extinction and Counter-Conditioning Actually Do
When we use extinction (exposure without consequence) or counter-conditioning (pairing a trigger with something positive), the brain does not delete the original association[TH1] .
Instead, it learns a second rule:
Old learning: “This predicts danger / stress / discomfort.”
New learning: “This is safe / neutral / positive.”
Both remain stored.
The trigger therefore becomes ambiguous. The brain decides which meaning to use based on the situation it is in at that moment.
This is why a dog can appear calm and “fixed” in one setting, yet react strongly in another.
Why Behaviour Often “Comes Back”
Bouton et al identified several predictable ways that previously improved behaviour can return, not because training failed, but because the brain retrieved the old learning instead of the new one.
1. Change of environment
A dog that is calm during training sessions may react again:
in a different park
on holiday
with a different handler
The calm behaviour was tied to the training context, not fully generalised.
2. Time passing
Even without new bad experiences, behaviour can resurface after weeks or months. Time itself acts like a new context, making it harder for the brain to retrieve the newer learning.
3. Stress or a single negative event
A stressful incident can rapidly bring back an older response, because stress strongly favours well-established learning patterns.
“Rats that have received tone-shock pairings followed by tone–food pairings show reinstatement of fear if they are exposed to the shock again (Brooks et al 1994).”
4. Rapid relearning
If the trigger becomes scary again, the return of the behaviour is often much faster than the first time it developed. This happens because the original learning was never removed.
What This Means When We Work With You?
We Remember what Brenda Aloff says;
“Behaviour is a lifelong commitment to helping your dog get It right.”
We appreciate that this research explains why statements like “he knows better” or “she’s being stubborn” are inaccurate.
Your dog is not choosing to ignore training. Your dog’s brain is instead retrieving a different memory.
Effective behaviour change therefore depends on helping the brain reliably access the new learning, especially under stress and in new environments.
How To Make Behaviour Change More Consistent.
We Train Across Many Contexts
i.e Different locations, distractions, times of day, handlers, and emotional states. Aside from ensuring you always have someone when you need, this is one of the arguments for us supplying two or three different team members for the same dog. This is especially important if owners are unable to do a lot of the work themselves as, multiple handlers gives us a better chance of the work generalising to you.
We also look to train in locations that you will go to.
“Mystkowski et al’s results suggest that about 30% of the distress that had been reduced during treatment was recovered when the spider was tested in the different context. (Only about 7.5% returned in the same context.) Thus, therapy for an anxiety disorder might indeed cause new learning that is at least partly specific to the context in which it is learned” (Bouton, 2002)
We Expect Fluctuations
Setbacks are normal and expected. They do not mean the training failed but we need management tools to ensure safety depending on the issue at hand- eg longlines, muzzles, harness.
Use Reminders
I.e we teach alternative Cues such as come here ‘watch me’ as well as just pairing the fear evoking trigger with something positive e.g food.
How we Use Food.
Food rewards are paired with a verbal marker (e.g. “good”), creating a conditioned reinforcer through classical conditioning (Ivan Pavlov). Once established, the marker allows precise timing of reinforcement and can be delivered immediately following the desired behaviour, even when food cannot be given instantly. Food reinforcement is then placed on a variable schedule, whereby the marker is consistently provided but food is delivered intermittently. Behaviour maintained on variable reinforcement schedules is more resistant to extinction than behaviour reinforced on continuous schedules, resulting in greater behavioural consistency.
While the marker itself can activate reward-related neural pathways once conditioned, its reinforcing value remains dependent on continued, intermittent pairing with primary reinforcement. Removing food reinforcement entirely leads to a gradual weakening of the marker’s effectiveness.
Space Training Over Time
Repeated learning across weeks and months is more stable than intensive short bursts. We do not do short intensive residential training, rather we educate you to have the tools to put the work in yourself over a long period and/ or ask you to find a budget you can commit to over a longer period where we are working in your area.
“…memory research suggests that material learned over spaced trials is better retained over longer intervals (Glenberg, 1976)”
Where we do find residential training helpful is if a much more controlled environment is needed to get over an initial hurdle- i.e if in London your dog is so reactive we can’t get the work started safely or the logistics of what your dog needs is just too costly in London. Moreover continued repeated repetition and exposure to the trigger in an alternative positive way certainly improves the chance of that more desired behaviour being accessed in that context. Thus it can be very valuable if that is a means to access consistent training, but Pavlov et al (1927) show this work must generalised with multiple locations and handlers and be continued to be positively reinforced in order to see these changes being accessed more consistently.
Prioritise prevention
Most of all we encourage everyone getting a dog to book a pre owner consultation to help choose the right breed for you, from the right litter and the right pup from that litter and then get started straight away or at your earliest chance to prevent issues.
Summary
Extinction and counter-conditioning are some of the best tools we have, however they are not erasers.
They build new learning that must be:
retrieved
reinforced
and practised consistently through life.
Generalised and or practiced in specific contexts- eg around your home.
Understanding this allows trainers and owners to stop blaming the dog, stop blaming themselves, and instead design training plans that work with the brain rather than against it and accept the principle of helping your dog get it right as opposed to take my dog and fix it.
Reference
Bouton, M. E. (2002). Context, ambiguity, and unlearning: Sources of relapse after behavioral extinction. Biological Psychiatry, 52, 976–986.
Aloff, Brenda (2001) Aggression in Dogs. Wentachee. Dogwise.
Brooks DC, Bouton ME (1994): A retrieval cue for extinction attenuates response recovery (renewal) caused by a return to the conditioning context. J Exper Psychol Anim Behav Processes 20:366–379
Glenberg AM (1976): Monotonic and nonmonotonic lag effects in paired-associate and recognition memory paradigms. J Verbal Learn Verbal Behav 15:1–16.
Lindsay, S. R. (2000 & 2005) Handbook of Applied Dog Behavior and Training, Ames, IA: Iowa State University Press.
Pavlov, I. P. (1927) Conditioned Reflexes: An Investigation of the Physiological Activity of the Cerebral Cortex. Oxford: Oxford University Press.