How Language Creates Suffering
or, what the fuck are relational frames?
This is a long one, so hold on to your butts.
Also, as a heads up, I will be doing some serious paraphrasing from multiple sources here. I am not really including references (other than a few links), but will provide them if requested.
As many of you know, I've been doing a lot of training in Acceptance & Commitment Therapy (ACT, said like the word). I've done two trainings — ACT for Beginners and Trauma-Focused ACT — with Russ Harris, author of The Happiness Trap and a renowned ACT trainer, and I'm doing ACT Immersion with Steve Hayes, one of the founders of ACT, which focuses on ACT as a "process-based therapy." This means taking a transdiagnostic approach and targeting underlying psychological processes rather than constellations of symptoms. This fits in neatly with the dimensional approach to diagnosis and treatment that I talked about over on Twitter (rather than the categorical approach of the DSM and medical model). There's a good, fairly accessible article that explains the process-based approach here, and a research paper that gets into the nitty-gritty here.
It's not hard to see that I've swallowed the ACT Kool-Aid. It's quickly become my main modality, but one of the best things about ACT, in my opinion, is that many other approaches can be used within ACT as long as they are done in an ACT-congruent manner. For instance, Progressive Muscle Relaxation is often used in the treatment of trauma. With its focus on relaxation and making tension essentially "go away," it's not consistent with the process of experiential acceptance that we would be working to develop. However, if you make it more of a noticing and mindfulness exercise, focusing on the different feelings of tensed muscles vs. non-tensed muscles, with the goal of increasing bodily awareness and acceptance of different body states, rather than cultivating relaxation (the idea that we can make stress or difficult emotions go away), it becomes ACT-congruent. We can even point out that sometimes these exercises will result in feelings of relaxation, but we can't reliably count on them to do so, and the likelihood of doing so is inversely related to how stressful the situation is. The surer way to manage difficult feelings, then, is to learn to acknowledge, allow, and accommodate them by opening up space around them. In other words, can we let the difficult feelings/thoughts/sensations be there, if it means being able to focus on doing what is important? This is essentially the question at the very heart of ACT; simplistically, can we accept what we do not have the power to change and commit to acting to change the things we can?
ACT is also compatible with polyvagal theory, resourcing exercises from EMDR, skills training from DBT, Internal Family Systems, somatic therapies, and so on, as long as our focus is not the reduction of symptoms or distress per se, but developing the six core processes of psychological flexibility. In use as a trauma treatment, this makes ACT unique in that it addresses both "bottom-up" processes in the autonomic nervous system, brain stem, and limbic system, and also "top-down" cognitive processes in the neocortex. (ACT is sometimes erroneously viewed as a "top-down" approach to trauma because it is, at its heart, a cognitive therapy (third wave baybeee!); however, the processes of contacting the present moment and experiential acceptance are very compatible with techniques that target the body, nervous system, and the "emotions and survival" parts of the brain — and that work often, but not always, comes first.
As much as I've been into ACT, and as much as I feel I understand it well, I still struggle with the philosophical foundation of ACT, a psychological theory of language called Relational Frame Theory (RFT). For whatever reason, RFT is a much more difficult concept to wrap my head around. Many ACT practitioners, including Russ Harris, will say that a thorough understanding of RFT is not necessary to practice ACT, however it does help strengthen one's understanding and ability to target the core pathological processes of psychological inflexibility. And I'm nothing if not a completionist. (Otherwise, why would I take three ACT trainings in about as many months?)
RFT has been elusive to me. It is one of those things that I feel like I understand on an intellectual level, but am unable to explain to others — which often leads me to think that I don't actually understand it at all! I suppose this essay is an attempt to help me learn to do that, so I appreciate your willingness to let me go on about it. And I hope that as you read this, you find something useful in it.
The somewhat short (but still geeky) description of RFT is that it proposes that the building blocks of language and cognition are "relational frames." Non-human animals acquire cognition and object recognition through association, which is why your dog knows that when you take out the leash, it's time to go outside. It's a straight sort of one-to-one, unidirectional pairing of an association that was learned through operant conditioning (the pairing of a stimulus with reinforcement or punishment, strengthening or weakening certain behavior respectively). We'll come back to associative learning and operant conditioning in a moment.
Humans, on the other hand, make bidirectional links between objects based not only on the strength of an association (if mailman = bad gets reinforced in your dog, that association is strong and will result in behavior change) but also the type of relationship between the objects. For example, an apple can be associated with an orange, but there are numerous relationships attached to that association: both are fruits, both are edible, they’re different colors, they’re similar in shape, they both grow on trees, one is usually juicier than the other, they taste different, have different textures, and so on. Those are all relationships between objects, and those relationships all fit within certain categories of relationships known as "relational frames." Some of the most common relation frames are similarity/sameness ("papa" is the same as "dad," "soccer ball" is similar to "volleyball," "hardcore" is like "punk rock"), opposition ("graceful" is the opposite of "clumsy"), distinction ("it is not warm outside" -- okay, so is it hot or cold or something else? all we know is it is not warm), comparison ("papa" is bigger than "baby," "A+" is better than "A," and so on), temporal/causal (before and after, if ... then, because of, etc.) and diectic which are frames that specify a relationship to the speaker ("I," "you," "mine," "yours," "here," "there," "now," "then," "left," "right").
Still with me? I hope so.
All of these relational frames are then put in networks that grow increasingly complex and which allow for verbal abstraction. What I mean by that is the way humans can be talking about something without directly talking about it, or how one word in a certain context can imply all kinds of other words and ideas. For example, if I say "I'm sad" after my favorite baseball team loses, it is very different from saying "I'm sad" after my grandmother dies, even though it's the exact same sentence. We can make this distinction because of all the various networks we have that contain the word/idea of "sad" in mutual relationships with all the other contextual cues of the situation (verbal observation of my affect, knowledge of the events, etc.). Words or ideas become networked with arbitrary contextual clues, which could explain certain trauma responses, for example, or why you just don't like tomatoes despite never having any obviously aversive experience with them. These networks of relational frames then influence our behavior — which starts to give us a clue as to how language causes suffering. So hold on tight.
This is all in contrast to B.F. Skinner's view that children learn language through the (relatively) straightforward process of operant conditioning and "shaping"--children's use of language is reinforced and thus strengthens the behavior, in essentially the same way you teach your dog to “sit” or “play dead.” (There is a somewhat famous debate — in certain circles, at least — between Skinner and the linguist Noam Chomsky, who instead proposed an innate "language acquisition device" because operant conditioning and shaping cannot fully explain children's development of language; in other words, language is not simply learned behavior.) RFT allows that operant conditioning plays a role, and could be considered a post-Skinnerian theory of language and verbal behavior, but the explanation it provides is more complex and more complete than Skinner’s. Part of this complexity is the important role of "arbitrarily applicable relational responding."
I'll be honest, that's pretty much the exact place I get lost. I think I did an okay job of defining "arbitrarily applicable relational responding" above (relational frames being networked with arbitrary contextual clues), but I'm just ... not sure! It's pretty dense stuff, honestly. I realize part of it is probably because my understanding of Skinner's work is fairly superficial, limited to what I learned in college psychology classes, and all of this is built on top of Skinner's theories about verbal behavior. And I’m not a behavior scientist! There’s a reason I didn’t continue on to a Ph.D. program!
To get a sense of how dense RFT is, check out this introduction by Steven Hayes. If this is the dumbed-down version, I don't know what I'm supposed to do!
Well, it turns out there is a more dumbed-down version, thankfully. There's actually a little ditty that Steven Hayes wrote to help: Learn it in one / derive it in two / put it in networks / that change what you do. I think it works better as a mnemonic to help you remember it once you've already learned it than to learn it in the first place, though.
The ditty, along with Steve explaining RFT is a much more accessible way than the link above, can be found here.
Now, to the question at hand -- how does language cause suffering?
These networks of relational frames form the basis of human language and cognition. And we can arbitrarily relate pretty much anything to anything else in any possible way. As far as we know, no other animal is capable of these relational frames, even our closest relatives in the primates, even the smartest pets (no matter how smart you think your dog is, Doug). We're unique in this ability, and it's probably why, for good or bad, we're the dominant species on the planet. It's also why we experience suffering in a uniquely profound manner.
According to Hayes and his fellow researchers, these networks of learned relationships are what make up the human "mind,” meaning they have perhaps an outsized influence on everything we do and think. In a sense, they don’t just inform our cognitions, they form the basis of our cognitions.
For a useful example, take this exercise from Hayes’ self-help book, Get Out of Your Mind and Into Your Life:
Our specific sets of relational frames give us access to many helpful skills. We are expert verbal problem solvers. It’s why we’re on top. But these relational frames also open us up to distress and suffering. Memories of past events can cause present pain. The knowledge that airplanes sometimes crash can cause a fear of flying. We can predict bad events and ruminate about things we did wrong in the past. We can compare ourselves to others, or against an ideal, which leads to yearning for what we don't have (and perhaps never will).
These relational frames also make up the core of the type of problem-solving we tend to do very well: get rid of the problem. If the trash is full, take it out to the bin and chuck it. The opposite of “full” is empty. The opposite of “here” is “there.” The opposite of “present” is “gone.” If we don't like that our shoe has a tear in it, we get it fixed, or buy a new shoe. We always find a way to get rid of the problem. Those relational networks give us the cognitive skills to be able to do that.
But what happens when we turn those same problem-solving skills inward, toward emotional states, thoughts, physical sensations? How do we get rid of those?
Of course, this all developed within an evolutionary context. Just as we can see how relational frames can cause us problems, we can see how, from an evolutionary standpoint, thinking in this manner was adaptive and evolutionarily advantageous. It helped us excel at verbal problem-solving, and many of the thought patterns we struggle with today — what in traditional cognitive behavioral therapy they call "irrational thoughts" or "cognitive distortions" — actually exist to keep us safe.
In the context it evolved in, the human brain really only had one major function: to help us survive. All of our most basic needs or innate drives are about survival. Food and water? Keep us alive. Shelter? Keeps us safe. Reproduction? Ensures the survival of the species. Our want for material things? We need resources to successfully make it through the winter. Our instinct toward self-defense and defending our families? Our need to bond? Our yearning to make meaning of the world? Our seemingly hardwired predilection toward sorting people into ingroups and outgroups? All of it. All of it is for survival.
So let's look at the so-called "cognitive distortions." We worry, catastrophize, predict the worst, avoid things that scare us. Well, if there might be a tiger in that cave, we better steer clear of that cave. And we better be prepared for the worst, because if we aren't, we could die. And if there's something that looks like another person out there, on the horizon? It could be someone from the outgroup coming to drive a spear through our guts. Watch out! Steer clear!
We extrapolate from past events to predict the future. Of course we do, because three moons ago, there was a bear in that patch of woods, and it could come back. And I don’t want to fuck with a bear, do you?
We ruminate on painful memories. And again, of course we do. If we survive an encounter with a dangerous animal or another clan, we'd better go over the events, again and again, to know what we did right and what we could do better in the future. The insight gained will save us in the future.
We compare ourselves with others because if we don't fit in, we'll be alone, and we won't last long on our own. We pay close attention to discomfort and anxiety and come up with reasons not to face that source of that discomfort; this is tied to neuroception, the way the brain and nervous system are always scanning for danger, beneath our conscious awareness and control, to keep us safe. Even greed, hopelessness, treating painful internal events as being as real as physical threats — all of it is to help us survive.
The neuroceptive processes — along with all of the ability to create relational frames, and all of the processes of thinking that are derived from those networks of relationships — evolved when survival was difficult, and anything that gave us that little ping of anxiety needed to be paid close attention to. And we couldn't dwell on any one thing for long — we were basically moving from survival event to survival event. And in this context, get rid of the problem is clearly the best way to survive.
For most of us, that's no longer true. Most of us live comfortable lives, relatively speaking. Most of us have shelter, friends and family, access to food and water, don't have to worry about sabretooth tiger attacks, and so on. And so we tend to dwell on these thought processes for longer and in greater depth than even our recent ancestors. But it turns out that these "cognitive distortions" aren't "distortions" at all -- they're how the mind works in order to ensure safety, protection, and survival. What we think of as pathological processes are actually adaptive, becoming maladaptive not because we're using them wrong, or because they no longer do what they were “designed” to do, but because the context they're being used within has changed.
And that new context really hasn't been the context for very long if you think about it. Our closest ancestors showed up somewhere between five and seven million years ago, and the first primitive homo sapiens about 300,000 years ago. The "modern" homo sapiens followed 200,000 years ago, and the human brain developed into the size and shape of its modern form about 100,000 years ago. It's not clear exactly when the ability to organize things into relational frames developed, but it is clear that the context of the human brain has been one of survival for hundreds of thousands of years, if not millions of years, longer than it's been one of domestication and industrial agriculture, relative comfort and safety, boredom, social media and remote work, and so on.
Language, through relational frames, gave us the tools to survive and solve problems. Those same tools now cause us suffering. They drive us to get rid of problematic thoughts, emotions, and sensations. Not because we're dumb, or broken, or mentally ill, but because context and function are no longer aligned. And because these processes are essentially hard-wired in, we start at a disadvantage.
It isn't hopeless, though, and we don't have to rely on the next great evolution of the human mind to save us.
The human brain has plasticity and can be rewired. New neural pathways can be developed (new relational frames), but just like wearing in a new path through an overgrown field, it takes time. We keep getting pulled back into the old rut of problem solving and psychological inflexibility, so we have to actively move back on to the new path until the new rut will consistently hold us and guide us in the direction of psychological flexibility and valued living. And in time, the old rut will fade away.
In this way, the relative ease with which we make relational frames causes us suffering, but might also hold the key to our salvation.