This post is the text of a video essay (posted on YouTube here):
When we’re on the inside of a circular causal loop, it’s usually easier to see how our own actions are influenced by the behavior of others than it is to see how our actions influence the behavior of others.
We know two people in a relationship, for example, and we can see how one side’s behavior affects their partner – negatively or positively – and vice versa. If Mike would just stop storming off to his man cave, Jennifer would be way less critical. And if she would stop complaining so much and express more gratitude, he would be a lot more likely to help out and listen to her needs. Why can’t they both see it?! Each seems focused mostly on one side of the circular pattern. What they each see is an accurate yet incomplete half circle.
This propensity that humans have to see only half of a circular dynamic is what I call circular causality blindness. And I think it’s a pretty strong human tendency that could viewed as a byproduct of more foundational cognitive biases, also known as “brain bugs”, or neuropsychological glitches in the ways we perceive reality – glitches that often evolved for some adaptive reason, but that have downsides as well, including warping our reality to some extent.
Three cognitive biases known as the fundamental attribution error , the actor-observer bias, and self-servicing bias, which could be found in a typical psychology 101 textbook, explain a lot of human perception and relational dynamics, and I think the three of them in conjunction can illuminate why humans tend to suffer from circular causality blindness.
The fundamental attribution error is our tendency to overemphasize internal personality traits and under-attribute situational factors when we make sense of the behavior of others. The classic example for the fundamental attribution error is how we explain others’ bad driving in terms of their character as we think, “what an impatient jerk”, whereas our own driving mistakes are clearly the result of context (“I’m late for a meeting”, “I’m tired”, “the sun was my eyes”, etc).
Compounding this reality distortion, we also tend to do the reverse when we make sense of our own behavior – overestimate our situational factors and underestimate our dispositional traits. This tendency is associated with the actor-observer bias: we feel like “the actor” in life looking at our own actions, so we focus on “the stage” – for example, “I failed the test because the questions were bad”. But when looking at others, we feel like “the audience”-observer looking at “the play”, and then we focus on the internal traits of the actors (“they failed the test because they were lazy and didn’t study”).
Then there is the wonderful, self-serving bias, our tendency to attribute our successes to internal factors like skill, talent, and effort and our failures to external factors like bad luck and unfair circumstances. If we get a job promotion we’re likely to believe it is due our hard work and talent, but if we don’t get the promotion, we’re likely to explain it in terms of an unfair process or being dragged down by colleagues, or other circumstances outside our control.
You might notice that the self-serving bias, which causes us to focus on our inner traits, and the actor-observer bias, which causes us to focus on our context, contradict each other when we’re looking at our successes. The way it works is that the self-serving bias overrides the actor-observer bias in situations where we’re having success, whereas in situations of failure, the two biases reinforce each other. If we ace a test, the self-serving bias dominates and will tempt us to conclude it’s because we’re intelligent and fast learners, not because the test was easy (which is what the actor-observer bias would have told us). But if we fail the test, it’s because the test was unfair, which is consistent with both self-serving bias and actor-observer bias.
All three of these “brain bugs”, which do distort reality, can have some adaptive value, which is why they evolved in the first place. You might also notice that they are all interrelated and have a lot of overlap.
One commonality between them is that they help us to simplify complex experiences, which can save energy and cognitive effort. They can also all help to preserve a positive self-image, especially in comparison to others, which can sometimes help us confront a challenging world with more false-yet-adaptive confidence.
But any reality distortions can also have downsides, which is perhaps the best reason to understand and be aware of them, so we can consciously correct for them when it’s helpful to do so, such as in interpersonal and intimate relationships, or when the complexity of the situation matters and shouldn’t be filtered out.
AI tells me that over 200 cognitive biases have been identified and classified. I don’t think that means that these 200+ tendencies to warp reality are distinct, and it’s likely that many of them essentially point at the same thing from a slightly different angle. Sometimes when a taxonomy like the one called “cognitive biases” gets very detailed and complex (often due to isolated rather than collaborative work by its contributors), it’s useful to consolidate and connect the ever-growing list of categories and concepts. I think one way to consolidate and more deeply comprehend the three biases I’ve mentioned here is to notice how they combine to create circular causality blindness.
Let’s apply them to the example of a romantic intimate relationship: our own behavior is our partner’s external situation, and their behavior is our external situation. According to the fundamental attribution error, we’ll underestimate the impact of our behavior (their context) on them. Meanwhile, the actor-observer bias causes us to overestimate the impact of their behavior (our context) on the choices we make. And the self-serving bias often reinforces this perception, especially in situations like conflict in which two partners let each down and then explain their “failure” in terms of the of context of the other person’s prior choices, rather than looking at their own psychological defenses and underlying wounds that keep the negative pattern going and that they otherwise could work through with conscious acknowledgment and dedication.
When both interconnected fellow human beings do this, the result is two parties holding two different and half-complete conceptualizations of a circular dynamic. Both sides can basically see one half of a circle, the half that the other doesn’t see — a half picture that says, “they affect me a lot but I don’t really affect them”.
With only that half image, we typically make just a small or moderate effort to change, and if we don’t see an immediate or strong effect in the other, we conclude that they are incorrigible, possibly due to habits and conditioning set in stone during their childhoods, impervious to our influence. This is how we overemphasize their internal personality traits. And when the other person does that one thing that drives us crazy and we react to it, we tend to feel reasonably justified by the external situation that they’re creating for us, and we underemphasize the role of our own personality traits in our response – traits that stem from our genes and our cumulative life conditioning. We don’t generally see the to extent to which our inborn dispositional makeup and the experiential templates from our past shape the narratives that we form to explain our partner’s actions and the relationship problems.
We don’t see very well how we’re in a perpetual game of relationship tennis, setting them up for the next swing. Imagine hitting a ball into a thick fog that totally hides the other player and the way that our shots influence theirs, giving us the perception that balls are just coming at us with random speeds and directions, independent from how we hit them. That’s a bit like how circular causality feels to us when we’re inside of, and only one half of, the reciprocal system. And sometimes we can even feel like the ball being hit, rather than a player in a back and forth exchange.