Assessing Darwin

In the 25 years I have been a behaviour analyst, I can recall three instances in which my repertoire required a significant change. These three events, although very different, all challenged my existing knowledge and required me to take a step back – back to principles, the literature, and to learn something new. The process each time has been both exciting and terrifying.

The first was when, at 26 years of age, I found myself having to lead the clinical team and design the intervention in a major early intervention research programme.

The second was when the COVID pandemic hit and Italy (my home country) went into one of the strictest lockdowns in the world (no outside time other than to go to the supermarket, pharmacy or medical appointments) and the families I was supporting found themselves at home having to deal with, in some cases, severe challenging behaviour and no in-home help.

And the third was raising Darwin: where the personal and the professional lines blurred into one. While not once I thought of giving him up, there were many days where I thought of giving up and resigning myself to a life of martyrdom and social isolation. It was only through the help of a number of professionals and the reliance on solid behaviour-analytic principles that I was able to do a little bit every day, knowing that any movement forward, however miniscule, was something we could build on the next day.

In the next few blogs, I will outline the process behind Darwin’s several intervention plans. This post focuses on the assessment of Darwin's behaviour to strangers and prevention strategies.

I apologise for the length of the post: the exercise of gathering and organising my thinking into the various sections is largely for my benefit, but I hope to illustrate some of the similarities between the application of behavioural approaches to challenging behaviour, regardless of the learner population.

As a behaviour analyst, I approach each new problem by going through the same two-part multi-step process.

Part 1: assessment

a) A behaviour analysis: an analysis of the controlling variables, a functional assessment. This is the equivalent of asking the learner “Why are you doing this?” Behaviour is rational – which is to say, it works, it has a function. In a science of behaviour, we look for orderly relations; a functional assessment enables you to discover the relationship between the motivations, the specific evoking stimuli, the behaviour (a product of both genetic influences and the environment) and the maintaining reinforcers. So, it’s a little bit more complex than a simple ABC – something happens first, the learner does something, something happens as a result. Darwin’s functional assessment was less precise than the one I would do for my human learners, but I felt I had sufficient descriptive data (ABC) to demonstrate a functional relationship between the evoking conditions, the topographies observed and the consistent consequences. When possible I did run a quasi-functional analysis (mainly single-function with a matched control) where the precursor reflexive response (eyebrow changes, ears back, tense body) in a given context would result in the stranger moving away or us moving away (test condition) with the matched control (the stranger at a distance).

b) A risk assessment: this is ongoing and is also included in the functional assessment. My risk assessment was again not dissimilar to what I would normally use in my clinical practice, and it largely comes down to describing: Should the challenging behaviour occur, what is:

· the risk of injury to the learner, others and the environment

· the frequency of the behaviour,

· the severity,

· the extent to which it restricts access to the social/learning environment,

· the likelihood of success of the current behaviour (whether it works to access the maintaining consequence)

These criteria were applied to:

Staring, grumbling (a very low growl), growling, lunging, barking; but of course, if these behaviours were not successful in contacting the maintaining consequence (the evoking stimulus was not removed), extinction would likely do its job and select the next behaviour: nipping or biting. I did not want to get there. There were also several other reflexive responses (ears back, stiffening of the body, eyebrow changes, cheek puffing) which, preceding the operant response, were generally an indicator that something in the environment was happening, but I wasn’t always able to identify the specific variables. My analysis of these responses was often post-hoc, they became indicative of the presence of certain conditions, and in time I learned to turn them on and off by changing some aspect of the environment to test my reasoning and be able to link them to certain stimuli. For example, when an intact male was in the vicinity, Darwin would stop, move his head forward and sniff hard. When a person moved in a place out of context (e.g., someone with a cowboy hat walking in an otherwise empty woods) then he would turn and stiffen, before attempting to take off.

The antecedent conditions I had previously taken data on as evoking the challenging behaviour were:

· An off-leash intact dog approaching (this was sometimes inevitable as in the UK most dogs are offleash)

· An on-leash dog staring, lunging or barking at him

· Being directly approached by a stranger who attempted to interact (looked at him, held out their hand, talked to him directly)

· Strangers walking toward him out of the blue (runners out of a bush, people turning a corner and facing us)

· Strangers coming near where we he was sitting or lying down (in a park)

· Strangers carrying (large camping gear, tripods, large sticks) or wearing unusual things(a helmet, a Santa's hat, large and loud headphones) and walking toward us

· Strangers coming to the house – outside (in the garden, at the door) and inside (even when behind a gate)

c) An assessment of basic needs: diet, sleep, physical activity, leisure/enrichment, socialisation. All of Darwin’s basic needs were being met, as confirmed by the professionals I worked with. Darwin was taken out on a long line, free to sniff and run (when I could drop the line in empty fields or off-lead when I hired enclosed fields) for up to 2 hours a day and never less than 1.5 hours. He had a raw meaty bone every night for supper which he took up to 30 minutes to eat. He also used his nose to find dry food that I regularly scattered in the garden. He slept through the night in his crate, slept for 3 hours in the morning after our long walk, and took several naps throughout the day. I arranged to meet other neutered or female dogs a few times a week so that he could continue to maintain his social skills after leaving daycare. We found a lovely small daycare with a few dogs (no more than 5) who accepted him twice a week. When this second daycare closed, the trainer we had worked with when Darwin was a puppy accepted him for two group walks a week with females and neutered males.

d) A full medical assessment: Pain was something that was mentioned by several professionals as a contributor to problem behavior. Personally, I did not see any evidence that pain was either a causal or a mediating factor. The behaviour could be easily turned on and off under very specific conditions. In the interest of being thorough, Darwin underwent a full blood and allergy assessment, was checked by a canine physiotherapist for any “mechanical” issues. He was in good health, a lean and muscular big dog.

e) Assessment of current adaptive skills: what could Darwin already do, and which behaviours were under stimulus control? There were not many; he could sit on request, he could go to middle and that’s about it… I had really not been able to shape or teach much, as I had not yet been able to find an effective reinforcer. Any time I attempted something he would switch off, even when it was in the context of a game. I just never managed to get past very simple ups and downs and food chasing and even those never evoked much enthusiasm. I was really tired of hearing trainers saying “you gotta find the one thing he loves and use that…” I began to sympathise with all the teachers and therapists who have come to me for help saying “he’s really hard to motivate, to engage”. Oh! How dismissive and patronising I must have sounded when I would reply: “you just gotta be creative and find the one thing he’s willing to do anything for…”

f) Contact relevant professionals: I recently heard a colleague say “you do not need to be your dog’s BCBA.” Personally, I disagree. It’s like saying to parents “you don’t need to be your child’s therapist; just be a parent.” The truth is that parents are a child’s first, permanent and most important educators, and they have to be given the knowledge and the tools. In fact, research clearly shows that parental involvement in intervention is a fundamental variable in successful outcomes. I believe my colleague meant “you do not need to be the person designing and planning the intervention for your dog; the skills you have as a BCBA do not necessarily translate to the canine.” With this, I partially agreed, but I don’t believe it has to do with being a behaviour analyst. It has to do with being an ethical clinician. When your learner’s behaviour falls outside your experience and realm of competence, refer or seek supervision. That I did. Most of the professionals I contacted used an emotional model of understanding behaviour. They employed behavioural procedures, but their interpretation was that behaviour was caused by emotions. As a radical behaviourist, I had a lot of problems with this interpretation as being circular, where the cause is inferred from the effect, and the only evidence for such cause is the behaviour to be explained. Emotions must be entered in the contingency analysis, but they do not cause behaviour nor do they maintain it. To change private events, if such change is actually observable, you change behaviour. In fact, no intervention exists that can alter "emotions" without environmental manipulation, regardless of what the implementer may believe is responsible for the change. But more on that in another post. After a lot of searching, I found specialists who didn’t just recommend permanent management, but who were willing to supervise my work as a behaviour analyst dealing with a novel population.

g) Developing a contingency analysis: Darwin’s challenging behaviours in relation to people were maintained by negative reinforcement and largely escape. In other words, once the problem stimulus presented itself and avoidance was not possible, it became important to remove it, so he would engage in active movement toward it to ensure that it went away. These triggers are defined as Conditioned Motivating Operations of the Reflexive type (CMO-R) – changes in the environment that, for the learner, reflect a worsening set of conditions and establish the removal of themselves as the reinforcer. An additional variable was my handling of the situation (tension on the lead), due to my anticipation of challenging behaviour, became essentially the Sd for engaging in the behaviour itself and discriminative of escape. My mechanics really needed to change, I knew only too well that I was also part of the problem.

Part 2: Intervention

There were many components to Darwin’s intervention programme and we are by no means done. I suspect that we never will be, realistically, and that I may always need to be vigilant. That’s the problem with challenging behaviour; it can often resurge – but one can minimise the probability by maintaining intervention somewhat constant and programming for generalisation. Because we had several challenges (people and intact dogs), each was tackled as a separate problem with a different intervention.

When I work with clients with challenging behaviour, I generally follow the same four-step approach: (1) removing the EO completely to bring behaviour to zero occurrences; (2) teaching in small increments the replacement behaviour prior to placing it back under EO conditions in dedicated sessions only while maintaining EO removal in day to day life, (3) emergency procedures, and (4) programmed generalisation. Through the many courses I have taken on dealing with dogs' aggression toward people, I have found that this sequence is shared by many expert trainers, and while the interpretation is not always behavioural or conceptually systematic, the interventions are always operant.

I found that one of the main differences between the various approaches was in the use of consequences. Here there seems to be a great divide in the dog training world. There are those committed to a positive reinforcement only approach, those who support a functional or constructional approach (that uses the actual natural reinforcer – in this case negative reinforcement to build functionally equivalent replacement behaviours), and those who use a positive punisher to immediately interrupt the problematic contingency and positive reinforcement to build the alternative behaviour. There is not much support for either punishment only or escape extinction. In summary, most of the procedures call for a gradual manipulation of the EO (the trigger), some emphasise prevention (management) more than intervention, and all seem to focus on teaching an alternative behaviour, with differences in how to get there.

As a behaviour analyst who, at times, has had to deal with severe and life-threatening challenging behaviour in less than optimal environments, with untrained parents and staff, I try to maintain an open mind. I learned early on to rely on all behavioural principles, to explore all the evidence-based behavioural approaches available and applicable within the boundaries of my profession’s code of ethics, my scientific conscience, and my experience as a clinician.

a) Removing the EOs completely to avoid strengthening the problematic contingency. I learned that in dog training this is called management, removing the triggers to avoid rehearsal of the target behaviour, which in the case of strangers looked like this:

· Avoid all public places with people in close proximity (cafes, shops, markets, car parks).

· Avoid pavement walks where people could come in and out of shops or street corners.

· Go to outdoor places where clear vision is ensured for at least 200 yards on a long line at all times and avoid all footpaths.

· Put Darwin in a separate room when strangers come to the house with something to engage him in (Kong, meaty bone, chewy).

This brought the behaviour down to zero – no EO, hence no motivation to behave… But, prior to moving to teaching, I knew I had to address the low engagement, low drive to do anything, the fact that nothing seemed to excite him, the effective reinforcer issue… Every trainer we have met labelled him as an anxious, nervous, sensitive, soft or unmotivated dog. But Darwin could also be very bold, never shied away from exploring enclosed spaces (boxes, tunnels), was never phased by loud noises, accepted all sorts of physical manipulations from me, was fine being left on his own for up to 4 hours at the time, and would engage in a lot of energetic approach movements (helicopter tail, undulating his body, soft eyes, smiley mouth) when people he knew visited the house or when he met one of his dog friends.

If he was a human, I would have said that at times he seemed clinically depressed (and I won’t bore you with an operational definition). I suspected an underlying chemical imbalance (genetic? lack of something during the critical development period? poor early socialisation? diet? Who knows?..). I asked the vet for a referral to a vet behaviourist, one of the few in the country. My aim was to come out of it with a prescription for an SSRI, but I did not tell this to the vet, as I wanted independent verification. It was a useful consult, although initially, I did not think so. Darwin gave his best impression of an anxious and depressed dog. He did not bark or lunge at the doctor, the nurse and the two students; he just gently grumbled when they said hello or if they moved a bit too close, and then proceeded to sit himself on the sofa, from which he did not move and nor did the humans from their place. He spent the whole 2.5 hour consultation roughly in the same position, crouched, behind my back, occasionally yawning, lip licking, seemingly napping, but always with one eye open to ensure all remained in exactly where they were meant to be. And of course, he did not accept any treats.

The first two hours were spent going through universal protocols for stranger danger (management with gates and crates when visitors came, long line when out, never off leash, minimise stress, increase sleep time, teaching settling on a mat, bring disengagement under a cue such as “watch me”, treat and retreat) and intact males (avoid all dogs, drop the leash if intact dog approaches and chemical castration implant once fully grown to observe effects of lack of testosterone). Apart from increasing sleep time, which was truly a revelation, I remember sitting there and thinking that I had just spent a fortune on being told something I had already spent a fortune on being told and was already doing. About 30 minutes before our time was up, the doctor very tentatively asked how we felt about medication and if it was something we might at all consider. I asked him if he thought that somehow my behaviour (my handling, lack of exposure during lockdown) was the cause of Darwin’s behavioural issues. He said something I will never forget: that while some of Darwin’s behaviours had a definite trigger and learning history, that I would have to be an incredible professional to operantly condition all those reflexive responses and his general low motivation. I was a good behaviour analyst, but not that good! We agreed that a long-releasing SSRI was needed and settled for fluoxetine at the lowest dose and to touch base after 3 months.

I knew medication was not going to turn Darwin into the friendly food-motivated Labrador I had dreamed of, but I hoped for two collateral outcomes that would at least enable me to teach more effectively: First, an increase in food taking; and second, a longer latency in responding to certain triggers, with enough seconds for me to give a cue for an alternative behaviour once the topography had been established. In other words, I was hoping for a biochemical alteration as an overarching setting event that would affect sensitivity to certain stimuli. Ideally, this would enable me to have an initial stepping stone to change the functions of these stimuli, from CMO-Rs to Sds for a different behaviour to produce and establish through differential negative and positive reinforcement, a different learning history.

At the start of my career, as a young and somewhat presumptuous behaviour analyst, I used to be very opposed to medication to address challenging behaviour– until I began working with young adults and children with autism with phobias and severe anxiety. In some of these cases, medication opened a window large enough for the intervention to begin taking place and for operant procedures to do their work. I was hoping for the same effect for Darwin. In addition, as soon as he was fully grown (18 months), he received the Superlorin castration implant (available in the UK, but not in the US). Between the medication and the reduction in testosterone, a new world opened up for Darwin – one made of interesting food flavours to learn new behaviours for, of balls and sticks to retrieve, of birds and squirrels to chase, and of ponds to swim in. Finally, I had enough potential reinforcers to compete with the environment and enough movement to teach and shape into alternative behaviours. Darwin’s reinforcement landscape (and mine!) having been reduced to stamp size, gradually increased to postcard and to A4 size. We are aiming for poster size next!