Training Anger Management

How does state-dependent learning affect skill-building?

Imagine this.

You take a well-regarded communications course and learn the skills of curiosity, empathy, and self-expression. You come home and use the skills with your partner, which they love. Until…things go wrong.

You’ve been practicing, and feel like you’re really getting a hang of these tools. But, a few days later, your partner forgets about a work party of yours and double-books themselves. You’re forced to go alone. Almost everyone else has a date, and you feel awkward.

When you get home, you yell at your partner. “How could you embarrass me like that??”

They wave their hands, trying to calm you down. “It wasn’t intentional. I forgot I had another commitment.”

You’re on a roll now, though. “This always happens. You put your social commitments over mine. This relationship feels so one-sided sometimes. I do such nice things to care for you, and then when I need you, where are you?”

Your partner says, “Calm down, honey. Use some of those great communication tools you learned at the workshop last week.”

This only incenses you further. Who are they to tell you to calm down? To use communication tools? THEY need tools, not you!

You’re seeing red. You know you’re being a bit of an asshole, and are probably going to regret this later. But you can’t seem to stop yourself. You can’t find your curiosity or your empathy. Your self-expression has no bounds.

You trained so hard - why can’t you access those well-curated, well-paid-for tools now?

Fighting Better

I and my husband have been teaching conflict trainings for the past 4 years. We’ve had about 400 students go through our courses, and hundreds more attend the Fight Labs we’ve held online for even longer. (They’re free - join us for one!) For those classes, one of the biggest challenges we’ve had to overcome is: how do we get students to apply skills they learn in a regulated state to situations where they are very upset?

If you’ve ever been angry, you know that you lose access to a lot of your curiosity and empathy when you’re triggered. You may care about the other person(s) you’re fighting with, but in the moment, you can’t move your attention outside your own sphere. (Check out this previous Substack article for the neurobiology of this.)

For this reason, Authentic Relating and Circling aren’t great conflict tools. We had to turn to the literature on negotiation, mediation, and de-escalation to find better ones. But even after we found the tools - which we teach in our Art of Difficult Conversations course every year - we had an even trickier issue crop up.

Our students found the tools mentally fascinating. But we were pretty sure that when it came to actual fights, they weren’t being assholes any less than before.

(I use that phrase deliberately. In fights, we may think that the other person is the problem. But usually, unless we’re in a genuinely abusive relationship, it’s at least 50% us. Either we’re practicing bad communication when triggered, or we’re being “good” in a way that sets our partner(s) off.)

Why couldn’t our students apply tools they learned in class to their actual conflict scenarios?

To find out, I had to delve into the research on a psychological phenomenon known as “State-Dependent Learning”, or SDL.

This article is about what I found - and how we might be able to use it to help ourselves, and our students, learn better.

The Church of SDL

The idea of State-Dependent Learning has been around since the 1700s. In 1784, French aristocrat Marquis de Puysegur conducted experiments on hypnosis. He found that subjects couldn't recall events from their hypnotic state until hypnotized again. Similar hypotheses and experiments were made about dreams, fugue states, and stimulus response up until the 1960s, when Donald Overton started running formal experiments on how drugs influence memory and learning. As one book defines it:

State-dependent learning is observed when subjects experience training under one of two internal states (a state induced by the presence or absence of a drug in Overton’s case) and [are] tested under the opposite internal state. The common finding is that when subjects are trained in a nondrug state and tested in a nondrug state, behavior consistent with training is observed. Conversely, if subjects are trained while in a drug-induced state (e.g., amphetamine) and tested while not in a drug state, a decrement in performance is typically observed…

…If the internal state of the subject during testing is the same as during training, regardless of whether it is a state induced by administration of a drug, performance consistent with training is observed.

In other words, if you’re put into a specific state to learn a skill, and then tested in the same state, you’ll do better on that skill than if your state is different between learning and testing.

State-dependent learning is similar to state-dependent memory, i.e. why you don’t remember what happened when you got drunk…presumably until you get drunk again. (I will now need to do a test in the next few weeks: get very drunk, do something dumb, then get drunk again to see if I can remember the details of what I did. For science!)

This effect has been most tested as it relates to drugs. For example, some researchers had men smoke either marijuana or a placebo, and then learn a task such as putting 7 objects in order. The next day, the researchers divided each group in two, and once again had half the men get high.

The results were conclusive: men made 3 times the number of errors when they HAD been exposed to marijuana during learning, and were NOT exposed during re-testing. This result held even though smoking pot had a minor negative effect on general recall, as the table below shows. “D” stands for Drug, “N” for No drug.

The Drug-Drug group made far fewer errors than the groups that smoked pot in one setting but not the other

In a short foray into “WTF Science”, these guys got off easy in just smoking pot. For some reason, the method of testing many of these hypotheses was to train participants in stressful environments, and then deliver shocks to their feet and/or immerse them in very cold water to see if researchers could replicate the stressful environment when testing learning recall. “Similar recovery effects have been found when retention deficits were induced with hypothermia (Vardaris et al., 1973)”. Ah, the halcyon days of Milgram and Zimbardo, before we had pesky things like ethical requirements and institutional review 🙄

And in other research I couldn’t help but cite: the effect of state-dependent learning is so widespread that it even shows up within the tiny worm c.elegans. When scientists tried exposing the worm to an odor (benzaldehyde) and ethanol at the same time, the ethanol had no effect on the worm’s reaction to the odor. But when the scientists pre-exposed the worm to ethanol and the odor at the same time, the worm would ONLY show an effect to the odor when ethanol was also present. So - get c.elegans drunk around a pretty odor, and it will fall for it again…but only if it’s drunk.

SDL has been so generally tested and proven, in so many animals (including humans), that it now seems likely that ALL memories and reactions are in some way state-dependent. A man who beats his wife when drunk can be a totally different person when sober. And if you teach someone skills while they’re regulated, they won’t remember them when they’re upset.

I see two ways of using this research in practice.

The first is in transferrable learning. The second, in habit and ritual.

A teaser of what we’ll talk about, if you are a paid ($5/month) subscriber:

• How to fake anger response in the body

• What early Circling and encounter groups did wrong

• How to upset people JUST enough to train them

• How master martial artists take tests

• Why students during the pandemic got worse grades

• Using your environment to cue state shifts

Let’s get to it!