Home

Mental Models from Deep Survival

February 09, 2020

For those following along at home, a delicious link (yes, it is a dirty affliate link like all other book review blogs):

Would I recommend reading this to an alternate universe self that had not read this book yet? Yes! Lots of useful knowledge and highly engaging.

TL;DR Deep Survival looks at individual/group behavior and thought processes in life or death situations. The takeaways on minimizing stress, training good automatic responses, and maintaining accurate mental maps apply far beyond these extreme circumstances.

What I learned in Boating School is…

System 1 vs. System 2

Courtesy of Kahneman and Tversky’s psychology research synthesized in Thinking, Fast and Slow, we can picture the human brain as operating under one of two modes. The first mode, called, System 1, includes our fast, automatic, and uncontrolled reflexes while the second mode, System 2, enables slower, reflective, and controlled behavior. While not perfect, this mental model of brain function fits many observed properties of the brain. For example, skilled tennis players can hit a tennis ball with almost no conscious effort. At first, tennis players need to consciously think how to grip the racket, how to step, how to windup, and how to follow through, but eventually, for the fortunate and persistent, these techniques become automatic appear as one fluid motion.

Automatic System	Reflective System
Uncontrolled	Controlled
Effortless	Effortful
Associative	Deductive
Fast	Slow
Unconscious	Self-aware
Skilled	Rule-following

Table courtesy of Nudge (2008)

These two systems are not independent and at times act in direct opposition. System 2 may set plans that System 1 later automatically overrides. How many times have you intentionally set an alarm, then hit snooze, and then later cursed your “lazy” self as you rush to get somewhere on time?

In this scenario, system 2 sets an alarm to allow enough time in the morning to arrive somewhere on time. Then, system 1 effectively hijacks control over a groggy, sleep-deprived brain. Later, system 2 reflects on system 1’s behavior, often with some guilt. The battle is not over yet. System 2 makes further plans to try and outwit that sleep-hungry self with dozens of alarms and alarm clocks that roll around and hide. Over time, if system 2’s plans triumph over system 1’s hunger, waking up with the alarm (or even without it) will become a conditioned, automatic behavior.

In terms of brain anatomy, the two components associated with these two modes of thinking are: the amygdala and the cortex. The cortex acts as the seat of cognition while the amygdala helps produce emotional reactions.

When someone jump-scares a friend, their amygdala receives the sensory input milliseconds before the visual section of their cortex as a “big shadowy form” (Deep Survival, 66). They don’t know if they were scared by a bear or by an unsuspecting housemate until after the emotional center has a chance to react.

The amygdala, after sensing a possible threat, hijacks cognition by releasing the hormones epinephrine and norepinephrine. In response to these hormones, the body also raises cortisol levels, leading to increases in heart rate and blood sugar. Of course, we colloquially refer to this phenomenon as the “flight or fight” response.

Deep Survival Cover — Imagine the moment of pure terror after walking through your house and encountering one of these muscular critters

Impact of Stress on Cognitive (System 1) Function

In Deep Survival, Laurence Gonzales writes about the interplay and performance of these two systems during life-threatening situations. High-stress causes normal cognitive function to completely shut down. Gonzales writes, “most people are incapable of performing any but the simplest tasks under stress” (38). In other words, stress inhibits System 2 and promotes System 1, shutting down the ability to think analytically. Once the “survival” instinct kicks in, intuition dominates over insight.

When system 1 thinking dominates, we are prone to experiencing inattentional blindness. The now famous basketball and gorilla experiment demonstrates that humans selectively process sensory inputs. When System 1 kicks in, the body narrows in on those sensory inputs instinctively deemed most valuable. Cognition, particularly expensive machinery for the body to sustain, shuts down. This tendency can lead to catastrophic results such as scuba divers instinctively pulling off their masks in response to feeling suffocated.

Gonzales makes a helpful analogy to the interplay between intuition and cognition: “To succeed we have to use the reins of reason on the horse of emotion” (29). Here, the reins represent our cognition and the horse represents our intuition. If the horse gets spooked, much like our intuition, the jockey holding the reins will be along for the ride as seen in the jump-scare reaction.

Mental Mapping:

We construct mental maps of our world in order to simplify the complexity of our surroundings. We know, generally, how trees will sway in the wind or other pedestrians act while walking in the street. We don’t consciously look at each person we pass and make full evaluations on the relative threat they pose. Constantly paying full attention to each of these sensory inputs would be overwhelming. As Laurence Gonzales puts it, “mental models are surprisingly strong; working memory is surprisingly fragile” (75). Mental models provide one basis for how we view and interact with the world.

What happens when our mental mapping is incorrect? We encounter potentially life-threatening situations. For example, in Deep Survival the snowmobile riders who tried “hammerheading” or riding their snowmobiles up the mountain face and then drifting back down. Their mental map of the mountain was of a solid, indestructible incline. This model is, of course, an oversimplification. A more detailed map for snow covered mountains is that certain critical angles form perfect avalanche conditions: too flat and the snow won’t move, too steep and no snow will accumulate.

Mental maps break when the real complexity of a system suddenly breaks through simplified human understanding. The same seems true in programming when underlying complexity breaks through an abstraction. Numbers, for the most part, work as expected. Complexity arises because computers have a fixed amount of memory. Integers are not infinite. Floating point numbers are not infinitely precise. Once the complexity starts to leak through, the programmer must understand enough to properly handle that in their own code.

In a survival situation, “those who do the best have the most realistic understanding of their situation” (122). Gonzales talks about the human tendency to “bend the map” when lost. We want to believe that we know where we are, so we try and bend reality to fit our mental map. Once a map gets bent, regaining an accurate assessment of reality is challenging. A sudden snap realization can lead to panic.

Coming to terms with reality sounds quite zen. As Gonzales points out, death itself is something that a human must accept into their mental model: “Survivors don’t defeat death, they come to terms with it.” There’s a similar zen to be found in programming: how many problems arise from trying to get reality to fit code when we the proper approach would be to get the code to fit reality?

Mental maps go far beyond survival situations. Deep Survival suggests that keeping them up to date and constantly reevaluating, especially when evidence doesn’t fit, are essential. Which mental models have you updated recently?

Here are some that have helped me update mine recently:

Takeaways

An ounce of prevention is worth a pound of cure

— Benjamin Franklin

Because system 2 (cognition) is expensive, our most repeated behaviors become conditioned into our system 1 (intuition) as a way to save on costs and increase the speed of action. For example, while driving hitting the brakes when you see red lights is automatic. At first this behavior is learned. Red lights don’t innately mean stop. We build a mental model mapping red to stop to putting the foot on the brake pedal. Then, behavior is conditioned over time to become habit. Getting to the destination safely and quickly is the reward.

These habits, which accumulate over a lifetime of conditioning, can either serve us or hinder us. Automatic behavior resists change. Once so deeply ingrained, these behaviors must be rooted out through stress, retraining, conditioning, etc.

In reflection, many of these habits accrue incidentally rather than intentionally. For instance, the evolutionary ones, such as the jump-scare reaction, evolved hundreds of thousands of years ago in environments far different from modern civiliation. Other habits, such as always making my bed in the morning, evolved over the course of childhood. I would wake up, make my bed, and then receive praise from my mom (I am thankful for this habit though there are certainly less desirable ones like feeling frustrated when losing in video games or sports).

The habit loop from The Power of Habit illustrating the cycle for conditioned behavior

While system 1 behavior happens automatically and beyond conscious control, outside of these scenarios that require immediate and automatic response, we can reflect back on past system 1 behavior. When was the last time you felt like system 1 behavior really took over? Feeling hangry while waiting in line at McDonald’s and then starting a fight after a sweet, old lady accidentally bumped you in line?

Once stuck in the system 1 thinking mode, humor can help to diffuse stress and retake the reins of control over the situation. When handling an amygdala hijack, Gonzales would also suggest the survival school acronym: SOTP (or STOP, short for stop, observe, think, plan). These steps try to force a reboot of system 2. Think of how many angry emails or text messages could have been avoided with this simple survival technique!

Even better than reducing unproductive time spent in system 1 (e.g. frustrated while debugging something you just know should work) is avoiding the switch in the first place. Some research suggests that both meditation and adequate amounts of sleep may help to reduce emotional reactivity. From Matthew Walker’s Why We Sleep, without adequete sleep,

“We cannot rein in our atavistic impulses – too much emotional gas pedal (amygdala) and not enough regulatory brake (prefrontal cortex). Without the rational control given to us each night by sleep, we’re not on a neurological – and hence emotional – even keel.”

These factors are also important to keep in mind when reflecting on a situation that went poorly or when stuck in a negative habit rut.

Bottom Line

Deep Survival presents a rich analysis of cognition under extreme circumstances. I’d like to take stock of some more of my own mental models as well as habits which form the basis of thought and action. By uprading bad habits and faulty mental models, my System 2 brain tells me I will become a better person (or at least more intentional).