"Learned Helplessness" - a new spin on an old theory
Updated: Dec 23, 2021
- by Diana Kastner and Katharina Pieper
I have been trying to find just the right words to bring this message to you as gently as possible, but in the end, there is only one way to state a claim as bold as this one, and that is to say it as it is:
"Learned Helplessness" is not learned
Yup, you read this right. After more than 50 years of believing that repeated or prolonged aversive stimulus teaches our dogs that no matter what they do, the outcome is the same, we have come to find out that this is not how it works. But who in this world would ever go against solid research such as that of Seligman and Maier, which we have adopted as the gospel truth since the 1960s?
First, please allow me to bring back the story of L. David Mech, who, in the 1960s did research of a different kind and later published his first book "The Wolf: Ecology and Behavior of an Endangered Species". Though L. David Mech wasn't the first wolf expert to conclude that wolves asserted dominance and subsequently followed a well-established alpha hierarchy (Rudolf Schenkel released his then-groundbreaking paper "Expressions Studies on Wolves" in 1947), he was certainly the main player in spreading the dominance (or alpha) theory - especially in the dog world. From the mid 80s on, L. David Mech spent summer after summer living amongst wild, free-roaming wolves on Canada's Ellesmere Island. And that's where the unthinkable happened: He learned that his very own original theories were flawed; that rather than wolves living in unrelated packs and fighting for dominance and alpha status, wild, free-roaming wolves live in harmony within family packs where they care for one another, hunt together, play together, rest together and teach each other. Based on these new findings, L. David Mech has debunked his original work. Yet, the myth persists! In cases when we accept something as being factual, it is difficult for us to let go of it, even if the original source of said fact courageously admits that the old fact is flawed.
Today, we are witnessing a similar situation, in that the authors of an original theory relentlessly searched for a more up-to-date truth, and this new truth so happens to debunk the old. And just as a vast majority of the public clings to the myths of the alpha theory, it might be challenging to convince the dog training world of this new truth behind learned helplessness; a truth that was daringly uncovered by no other than Maier and Seligman themselves.
Psychology and Neuroscience
When Seligman and Maier originally experimented with naive beagles and street-smart mongrels, they did so strictly from a psychological perspective. While all dogs underwent the same general process, the dogs were split into three groups. To begin with, all dogs were placed into hammocks, where group 1 received a series of foot shocks but was trained to push a plate that would allow them to turn off the shock (ESC group). Group 2 received the same series of foot shocks but were not given any control over the outcome, thus making the aversive stimulus inescapable (INESC group). The dogs in group 3 were lucky enough to have not received any foot shock to begin with (Zero group).
A set time after the hammock experience, the dogs were moved to a device known as the shuttlebox. The box consisted of two compartments, divided by a knee-high barrier that could easily be jumped by any of the participating dogs. The floor on one side of the shuttlebox was set to deliver foot shocks once again. It so happens that the majority of the INESC dogs went passive and accepted the new shock as their inescapable fate. However, the dogs from the ESC group and the Zero group performed equally well in jumping the barrier and to safety.
Over the years, Seligman and Maier conducted a number of experiments that led them to coin the resulting passivity observed in the INESC dogs as "learned helplessness".
Learned: because they theorized that the INESC dogs learned they had no control over the outcome of an aversive event, and they learned that they would have no control over the outcome in future aversive events.
Helplessness: because nothing one does matters.
Seligman and Maier theorized that this was a cognitive process.
At some point, Seligman and Maier parted ways. While Seligman went on to focus on human psychology, especially on the correlation between learned helplessness and depression, Maier found himself drawn to study the brain. Finally, this newly found interest in neuroscience prompted him to search for the neurocircuit of learned helplessness. And this is where it gets really interesting. This is where science shifted gears and did a complete 180 on an old theory that was based strictly on observable behaviors. And the outcome is nothing short of amazing: The original theory of learned helplessness has it backwards. Rather than learning that the animal has no control, the animal learns control; rather than the animal expecting lack of control in future aversive situations, the animal learns to expect control; and rather than passivity in the absence of control being cognitive, this passivity is a deeply subcortical default survival mechanism.
The procedures used in finding this circuit was quite complex and invasive. By now, Maier and Seligman have moved from experimenting on dogs to rodents (Maier) and humans (Seligman). With the help of various tracers, agonists and antagonists, Maier and his team of researchers were able to arouse or inhibit different brain activities in the rat brain and thus map out the circuit that facilitates not only the default mechanism of learned helplessness, but also the learned mechanism of deactivating it. To read the author's manuscript detailing the research, please follow this link.
The Learned Helplessness Circuit
The groundbreaking news about the mechanism of learned helplessness (from here on referred to as "default passivity") is that it is set in motion by default any time an animal is exposed to a threat or otherwise aversive stimulus. The brain regions involved in this circuit are the dorsal raphe nucleus (DRN), the amygdalae, the striatum and the periaqueductal gray (PAG). Within this circuit, the DRN acts as a hub. In a nutshell, a threat activates multiple brain regions, which in turn activate the DRN. The DRN then releases higher amounts of serotonin, which signal to the PAG and striatum to inhibit fight and flight, as well as to the amygdalae for heightened fear and anxiety.
However, the subsequent state of passivity (PASSIVITY/ANXIETY) is not an immediate occurrence. While initial serotonin levels spike quickly at the onset of an aversive event, reaching the threshold that finally induces passivity is a gradual process, which explains why only prolonged aversive stimuli result in what we have come to know as learned helplessness (default passivity). And just as it takes a while to reach said state of default passivity, once passive (due to sensitized serotonin receptors in the PAG and striatum), it will take a number of days following an inescapable stress event for these sensitized serotonin receptors to desensitize to once again allow the animal to become more active.
The following cluster of behaviors observed during the state of default passivity are behaviors that can be linked to either passivity or anxiety (hence PASSIVITY/ANXIETY):
Reduced aggression (PASSIVITY)
Exaggerated attention to external cues (ANXIETY)
Potentiated opioid reward (PASSIVITY)
Potentiated fear conditioning (ANXIETY)
Slowed fear extinction (ANXIETY)
Reduced food and water intake (PASSIVITY)
Neophobia (fear of novel foods) (ANXIETY)
Reduced preference for sweet tastes (PASSIVITY)
Reduced juvenile social exploration (ANXIETY)
Reduced social dominance (PASSIVITY)
Deactivating the LH Circuit
With exposure to escapable shock (or other aversive stimuli), and thanks to neuroplasticity, the ventromedial prefrontal cortex (VPC) learns to:
DETECT: detect control,
ACT: first build a new neural pathway between the prelimbic area (PL) of the VPC and the DRN by creating proteins (plasticity proteins to be exact) and then deactivating the DRN via this new pathway when control is DETECTed
EXPECT: expect control in future aversive events
Please note that recognizing the mere presence of control (rather than the act of controlling) will deactivate the DRN.
Once the ACT pathway is created, immunization (see below) will allow it to be activated (and the DRN deactivated) even when the aversive situation is inescapable. However, since immunization is not permanent, it is important to keep "refreshing" this pathway, that is, keeping it active by exercising control.
The flow of encountering a Threat
The beauty of this mechanism is that, initially, it allows time for defensive behaviors and passivity is only activated when control is not detected. Learning control is vital, life experience is vital, allowing dogs to cope with life and with stressors in THEIR ways is vital.
Questions to consider
Why would default passivity (LH) be important for survival?
As we know, threats engage defensive reflexes; however, these reflexes are quite energy intensive, and if unsuccessful, it might be more beneficial to inhibit them and conserve energy for use in situations when they are more successful again. As Maier stated in his article, "evolutionarily, 5-HT is very old and has been involved in controlling and shifting the balance and flow of energy from the beginning".
If control leads to expect a favorable outcome, why wouldn't ESC subjects outperform those that have never learned to escape?
According to the article, serotonin builds gradually. Thus, by the time serotonin reaches the threshold that causes passivity, the dog already escaped to safety. Therefore, the non-shocked subjects learn control before serotonin reaches levels that would cause passivity. "This first, very rapid DETECTion of control would then inhibit the DRN in the future".
How did the street dogs compare to the beagles in the original experiments?
To get to the bottom of this question, we need to first identify the condition in which these dogs lived prior to the experiments. Even though the histories of the street dogs was unknown, it was probably fair to assume that these dogs have lived adventurous lives filled with misfit, challenges and obstacles. The beagles on the other hand were cage-raised and largely kept in social/emotional isolation from early puppyhood until the time of testing.
Following a single exposure to prolonged inescapable shock, the street dogs outperformed the beagles during the shuttlebox experiment. While the beagles did not attempt to escape even after only one prolonged inescapable shock, the majority of the street dogs had no problem jumping the barrier. However, two exposures to prolonged inescapable shock were enough to activate the DRN and render the dogs passive. (source)
It was during these early experiments when Seligman and Maier considered past life experiences - and thus experiences of control - to be vital in preventing learned helplessness. They likened this to a process of immunization, a process for which they found a neurobiological explanation in their later work (see above).
What is the connection between dopamine and default passivity (LH)?
While dopamine is not directly involved in the learned helplessness circuit, dopamine levels decrease when an animal is in a state of default passivity (LH) (source). This shouldn't come as a surprise, considering that the function of dopamine is to modulate motivation, anticipation and expectancy.
How does the endogenous opioid system play into default passivity (LH)?
A separate study has found that stimulating the endogenous opioid system by opioid agonists such as morphine "reversed the escape deficit that was caused by shock pretreatment" (please click here for more information). What this means is that endogenous opioids (opioids that are released within the organism), including endorphins, can reverse the default passivity induced by prolonged aversive events.
While this is definitely interesting, this too shouldn't come as a surprise, given the positive effects endogenous opioids have during stressful events. And if you have been following Affective Dog Behavior, you are familiar with the many ways in which we can help our dogs release endogenous opioids. For more information, please contact one of our ADB coaches.
How can we "teach" control?
The re-evaluation of the original studies has highlighted the importance of an individual having the impression of being able to control a situation - even if this control is not performed at all. We can help our dogs strengthen the ACT pathway by, on the one hand, allowing dogs as much control as possible in everyday life and, on the other hand, allowing and supporting dogs to deal with stressful events in THEIR ways.
IMPORTANT REMINDER: This does not mean that we can forcefully expose our dogs (or other animals) to stressful events, prolonged or otherwise. For one thing, that would be abusive and cruel, secondly, this in itself would mean that we are taking control away from our dogs, which is exactly what can lead to default passivity (LH).
In addition to the uncovered neurobiology, Seligman's research with humans has shown that focusing on future events is more therapeutic than focusing on the past. Naturally, asking dogs to focus on future events may be a bit challenging, but we can certainly help our dogs live more in the moment; provided we find a way to let go of what might have gone wrong in their past.
UPDATE 12-22-21:
Even though it has only been a few days since this article's initial release, it has already brought about some good and objective discussions. There were three items of concern in particular that are worth taking up again and clarifying:
Concern 1: Despite the solid research presented, some readers are convinced that Learned Helplessness IS learned and easily observed in dogs that give up and "shut down" as they comply with handler demands after repeated use of aversive stimulation and the dog learns that giving up and complying are the only ways to stop the punishment.
While this scenario is not a case of "learned helplessness" but one of operant learning. and the very nature of this statement in which the dog's response (giving up) changes the outcome (punishment stops) is the exact opposite of the definition of "helplessness", where "nothing one does matters". In other words, in order for a dog to go into a state of "learned helplessness", the aversive stimulus would have to be inescapable and thus continue even after the dog complies or gives up.
Having said that, the use of aversive tools and techniques can and, in most cases, will contribute to a multitude of behavioral and possibly physical and mental health issues that do not fall into the category of "learned helplessness". Therefore, we are strongly advising against the use of aversive tools and techniques.
Concern 2: Ongoing research of LH has advanced beyond what neuroscience has to offer and said research has been well-documented for decades. In human psychology, many forms of depression and trauma have been linked to the phenomena of learned helplessness, in which one learns to detect and expect a lack of control.
There are two aspects to this we might need to keep in mind: (a) Many of the studies that have been performed with human subjects have been strongly influenced by the early theories that one learns to detect and then cognitively expect a lack of control. This in itself may lead to different conclusions than had the researchers known that it was the presence of control which is detected and expected. (b) The differences between human and other animal subjects are stark. While it is possible to experiment with cage-raised "naive" non-human subjects that have next to no social interactions and no life experiences, all human subjects come pre-packaged with a much more complex cortical brain programmed through many years of social interactions, emotional ups and downs, environmental stimuli and subsequent life experiences and problem solving. Unlike non-human subjects, when a human subject participates in psychological experiments, it is voluntarily and with an awareness that these experiments come with ethical boundaries and time limitations.
Concern 3: Research done in lab environments does not give us the necessary answers to help us understand real life situations.
This has always been one of our main concerns when we look at behavioral research, especially since the behaviors exhibited in a lab environment do not take into consideration the multitude of stimuli an animal would be exposed to in the real world. However, the research to locate a specific neurocircuit is invasive and includes the injection of tracers, agonists and antagonists to map out parts of the brain during a number of behaviors in a variety of situations.
Food for thought
Hopefully, we all can learn something when we are faced with stories of righting a wrong, of busting a myth or debunking old science. Science is a lot like the ocean; it never stops moving. It comes in waves and persistently washes ashore new discoveries. As canine professionals who pride themselves as being science-based, we have an obligation to stay on top of the latest research, or else we lose credibility. Worse yet, we end up doing our dogs a great disservice.
Knowledge is for those who have the wisdom to seek it!
by Diana Kastner