There is one thing that I rather repeatedly heard from them – and which I never listen to:

“Novices always take on impossible dogs”

Understand Schnauzers:

“Schnauzers, the sort of dog who is slow to learn”,
”You are going to work much harder to obtain results with this breed”

I agree that as a general tendency, these dogs and particularly the Standard variety for what I have seen, has strong susceptibilities for reinforcement of behaviours which are not easily manipulated in the training ring: all the sequences for chasing rodents, not a toy, a warm, furry mammal.

I find that they are also serious dogs, who strongly commit themselves in what they want to do and renouncing that to take the activity you suggest has a high cost. In other words, they quit the situation, with a damaging effects on relationships. In other words, you make partnership with them –and they are then gold, or you make nothing.

It also came to my mind more than once that when compared to what I saw, for instance in Malinois Shepherds, Schnauzers were quite difficult to automate, that is to engage in habitual processes, and were resistant to addiction.

Searching the recent research literature in neurosciences, I first came across this definition:

“Habits arise from goal-directed learning as a particular behaviour becomes automated with repetition and its performance becomes less dependent on the outcome” (1).

In other words, a behaviour becomes habitual when it becomes elicited by a cue, independently of the outcome. This is not only a theoretical view. Activity of neurons in a part of the mammalian brain termed dorso-lateral striatum (DLS) has been shown to be required for habit formation, as well as to their maintenance. Of course, purely habitual processes, completely disconnected from consequences, never exist in real life. The fact that an effect exists does not mean that it prevails, or corresponds to a law. Other systems always balance these effects. Indeed, activities of a neighbouring neuron population, the dorso-medial striatum (DMS), is known to be essential to the dependence of behaviours on outcomes (goal-directed behaviours): rats in which the activity of these neurons is inhibited become completely insensitive to changes in outcome value in a two-choice, two-outcome task.

But how do the neurons, supporting habit formation, interact with those required to goal-directed behaviours, strongly dependent on outcomes? How do the DLS and DMS parallel learning systems, respectively promoting habitual and goal directed behaviours, influence each other?  We may think that the activity of the latter prevails at early stages of learning and that with repetition, the brain transitions to a state where the former becomes predominant. But things are often more complicated than expected when it comes to how the brain works and it is not the case.

In a discrimination learning context activities, it has been shown that both activities co-exist during early and late learning AND that habit promoting systems make early learning slower. Animals learn faster, more easily, when habitual systems are (artificially) inhibited [2]. In other words, the capacity to engage in habitual processes, which is an advantage in stable environments or for the efficiency of common behaviours (no need to assess outcomes), has a cost. It makes early stages of learning slower.

How have breeds been selected? What do we term a super-learner? It depends on which sort of activity we want to favour. I have always found that Schnauzers preferred non-repetitive tasks, leaving them a broad space of expression, in environments and tasks requiring a strong behavioural flexibility: searching activities, man-trailing, search-and-rescue.

They have all my empathy for this trait… perhaps a differential susceptibility to engage neuronal systems, promoting goal directed processes, versus habit promoting ones, may lie behind these characteristics? Terrace’s pigeon was perhaps not in the best condition to learn, with a subtle variation to an established behaviour, that had perhaps become habitual, in the scientific meaning of the term.

We should be more specific when we refer to learning abilities.

Thank you; Kay, for never telling me that Schnauzers are slow learners, stubborn, impossible dogs…

[1] O-Hare et al. 2016. Pathway-specific striatal substrates for habitual behaviour. Neuron 89: 472-479.

[2] Bergstrom et al. 2018. Dorsolateral striatum engagement interferes with early discrimination learning. Cell Reports 23: 22-‘-2272.