Have you, like us and our colleagues at Eicorn, come to the conclusion that the single most important thing to get right in life is the stuff that goes on inside your head? The quality of your thinking is more important than anything else to establish a great and happy life, good relationships, good work, great investments, and so on. Let’s call it “success” to keep it simple.
The idea that your thought-process is key to success compete with other human attributes such as passion, diligence, grit, luck, and more. Most such things do indeed play a part in success, but your thinking (and your decision-making in particular) is what determines your direction: Towards or away from success. You, dear valued reader, have probably realized this simple but very powerful truth, and are now exploring the intellectual upgrade packages called Mental Models.
Mental Models is a concept quickly increasing in popularity, and several contemporary self-improvement and management books are based on the topic of Mental Models. After reading several such books, and as well as many of the sources that all the new books seem to commonly reference, it struck us that every author and influencer on this subject appears to have overlooked (more or less) the fact that the Mental Model is itself a Mental Model.
As such, it should be thoroughly explored and discussed, modelled and re-modelled, pedagogically packaged and applied as much as possible. Because, the Mental Model of the Mental Model, let’s call it MM2, is potentially a Super Mental Model on steroids, given that it conceptually should describe the inner workings of all other Mental Models. What if a good deliberate tweak to the MM2 can upgrade most or all of your mental models at once? Seems like a nice hack, so let’s try it!
Mental Models are usually described something like “intellectual representations of the world, allowing people to understand reality by using simplified thought-models of complex actualities.” Some people have the opinion that all human reasoning, so basically most of your conscious thinking, is based on Mental Models. Others lean more towards the idea that Mental Models are only constructed, deliberate thinking tools, and that all other “simplified thought-models of complex actualities” should be called something else. We sometimes see people state something like “there are about a thousand Mental Models”, which would suggest that there is also a limited number of complex actualities, or “case-at-hand” as Ray Dalio prefer to define specific issues. Unless there is something else than Mental Models that help us make sense of the world.
Of course there are also various academic efforts to define Mental Models, some with more merit than others. But in general we find them (the ones we have tried to digest) quite difficult to convert into actionable insights. This excursion to define the MM2 takes a practical point of view to explore the singular Mental Model, in contrast to Charlie Munger’s model of a latticework describing the effective application of plural Mental Models. We are not trying to be philosophical for the sake of philosophy, but to create a foundation for learning so that we all can upgrade our brains a little bit. Oh, what the heck… Let’s upgrade them a lot while we are at it!
Our approach to creating this definition of the MM2 has been to respectfully disregard all previous attempts to describe the Mental Model as a model and start from scratch. But, the existing attempts are of course still influential, part of our learning and how we have evolved our thinking around the topic. In fact, we do not feel like we have to claim that much at all is new about this definition, it is mostly a synthesis of other peoples’ ideas, facts, insights and additional intellectual contributions. Still, we think this approach is very actionable and does add significant value to the growing interest in Mental Models.
The fabric of Mental Models
We are not supporting the opinion that all human reasoning, or conscious thinking, is based on Mental Models. If that was the case, we would just dilute the Mental Model idea to a holistic thinking concept that spans more or less every intellectual activity, and that would not be very useful. Therefore, we are proposing three conditions that must be true for us to have entered a Mental Model domain:
- What we think about must be a system
- It must include possible state-changes
- It must include uncertainty
Since this clearly do not apply to everything that passes through our minds, there must be other intellectual concepts than Mental Models that we use to think. One such concept is General Knowledge. Many things can be general knowledge and Mental Models at the same time. Let’s use an apple as the example. If you think about the apple as an object, our three conditions above are not met. You knowing that the apple is an apple is just general knowledge.
If you think about the apple as a function instead, you instantly enter the world of Mental Models. If you pick the function “food” the apple all of a sudden is part of a system that convert calories and nutrients to energy and tissue and other good stuff in the body. It must change its state: You can’t swallow it whole, and it must be digested before it can turn into energy or tissue, etc. And you can’t know for sure that it will not cause an allergic reaction, or be rotten inside, and so on. Switch function of the apple to “projectile” and you have a different Mental Model in your head.
Note that function is not the criteria, but the three conditions are. Also note that thinking about objects does not always put us in the general knowledge domain. If the object is a complex system, let’s say an engine, we mostly end up in a Mental Model directly. Is the engine running or not? Is it hot if I touch it? How much fuel does it have? etc. A typical giveaway for if the brain has entered Mental Model mode is if it has a battery of questions about the thought or situation being processed.
Similarly, the general knowledge domain is not restricted to only objects, sounds and cognition in general. Consider simple calculations, those you just can’t get wrong, like 1+1. Sure, addition is a system, 1+1 is clearly a state-change in the making, but there is no uncertainty. You don’t have to apply a model in your head to get this calculation right, its solution is general knowledge. However, the solution to for example 14+27 may not be, or is even likely not to be, general knowledge. You have to think about the answer, and there is a real possibility that you get the answer wrong. Voilà! You have entered the realm of a Mental Model to get this one right.
Why is it important to distinguish this difference between types of thinking? Because it illustrates that the idea that all conscious thinking is based on Mental Models is too broad, and that the idea of only about a thousand deliberate Mental Models is too narrow. It also illustrates that Mental Models are in play subconsciously as much as consciously, whenever those three conditions are met. For example, you rarely use the food Mental Model described above consciously. But if it didn’t run subconsciously you would likely be dead within days.
How Mental Models run inside our brains
A second useful distinction of the MM2 are the autonomous versus deliberate application modes. Most Mental Models are autonomous, habitual. They do their thing consciously as well as subconsciously with full autonomy, i.e. you can just tag along for the ride. Take the food Mental Model as an example, it just works (take general perspective) regardless of how conscious or not you are about applying it what you eat.
The real power of Mental Models as tools for intellectual change and better thinking lies in our ability to take over the control of our Mental Models, make changes to them and use them deliberately. If we stick with food as an example, becoming a vegan is a deliberate change and use of the food Mental Model. However, once a vegan menu is the norm the Mental Model will go back to autonomous mode and work automatically in the background to fulfil your goal. Autonomous or deliberate application modes are not good or bad, they are just different in nature. Some Mental Models are always deliberate; they are simply too rare, too hard, or too unnatural to internalise deeply enough to become autonomous models.
Why do we have Mental Models?
Ideally the WHY-question is answered before anything else, but in this case we thought that the MM2 basics are important to understand in order to fully grasp what Mental Models are for. The best way to answer the question “why do we have Mental Models?” is to make use of some good old First Principle Reasoning, a great Mental Model that you should learn if you don’t already know it.
So, a First Principle of our ability to think is to have a brain. Why did organisms develop brains? Because it gave them an advantage in the competition for resources. Why did they get an advantage? Because the brain is able to make predictions, with the implications that animals with brains could foresee state-changes within food and habitat systems, and alter the probabilities of success and failure for certain tasks. Such as finding food or cover.
Mental Models are, from a practical point of view, essentially extraordinary tools for making predictions about how reality works. The decisions we make, consciously as well as subconsciously, are bets on those predictions. As an example, the 80/20 rule is a Mental Model predicting that most of the effect within a system will come from only a small fraction of that system’s state-changes. Second Order Thinking is a Mental Model predicting that you will uncover big uncertainties within any system if you can work out several orders of state-changes. And so on.
So, we use these predictions to make decisions, big and small. But we also seem to use Mental Models to fill our brains with meaning about the world we live in. You have probably heard the saying that “our heads are meaning-making machines”. This meaning-making is the result of Mental Models doing their thing, often to deal with the chemical marinades served by the brain, for the brain.
You may for example get an angry stare by a manager at work. The amygdala in your brain will immediately try to determine Fight or Flight (a very powerful primal Mental Model) and within a split second your nervous system will be flooded with a mix of cortisol and adrenalin. Regardless of the fight or flight outcome, all kinds of mental models will likely autonomously start to make predictions about the nature of the situation, in order to serve you meaning. Does it mean that the manager does not like you, or just that he or she had a bad day? Or, if you have a people-reading Mental Model that is badly calibrated on the uncertainty condition, could it be that you read the manager’s face incorrectly?
The point we want to make is that we mostly encounter the Mental Model concept in intellectual contexts dealing with difficult decision-making, typically within a business context. But our brains use Mental Models just as much to generate predictions that can explain how we feel, our social status, and what the world means at large. This is an important insight because it explains a lot of human behavior related to beliefs and values. For example, to believe in an afterlife is a Mental Model that predicts that there is more to come later, after the life/death state-change, therefore you need not fear death so much. Monogamy is a Mental Model that, amongst several different things, predicts the strong likelihood of being held in high regard by at least one person over an extended period of time, i.e. the social foundation for a happy life.
The Five Cognitive Dimensions of Mental Models
So far, we have explored the fabric of Mental Models, their two different execution modes, and a few plausible reasons for why we have the extraordinary ability to form Mental Models. All these aspects of the MM2 can be used to examine any individual Mental Model with the hope to understand it better and make better use (including no use) of it.
But there is one more key aspect to consider: How Mental Models fit into what we chose to call the cognitive dimensions of human thought and behaviour. The five cognitive dimensions described here are loosely inspired by the first six levels of Leary’s and Wilson’s proto- (or pseudo- if you will) scientific, but highly intuitive, eight-circuit model of consciousness, best described in Wilson’s book Prometheus Rising.
Most importantly, the model argues that one can’t rise above any circuit that has a current activation energy above a certain threshold. It is conceptually the same idea as “managers get promoted to the level of their incompetence”, i.e. when their incompetence is above a certain threshold level there is no grounds for further promotions. Due to our most common cognitive biases the five cognitive dimensions of Mental Models typically work the same way. If you for example have a Mental Model about the origins of mankind with a high cognitive conviction level (i.e. “activation energy”) in the faith-dimension, e.g. a strong belief in God, you are very unlikely (or even incapable) to consider a different Mental Model in a higher dimension explaining the origins of mankind with logic or evidence.
The Five Cognitive Dimensions of Mental Models are:
Our most primal Mental Models, like Fight or Flight, are hard-coded into our brains and DNA. They are not even based on our own experiences but on our ancestors’ experiences, natural selection and other evolutionary forces. For example, research in neuroscience indicates that we can’t make deliberate changes to the triggers in the Fight or Flight model, we can only (at best) make deliberate changes to how we react to those triggers. The oldest and most basic cognitive dimension is best described as instinct-based. Unfortunately, we can’t easily get rid of the crap that is hard-coded into this dimension, we can only manage it.
As humans became a bit more sophisticated and evolved through the cognitive revolution, or however we came to be what we are, the meaning-making machines between our ears started to make things up. Pretty soon, faith was invented and with it came religion and a myriad of pretty crazy Mental Models based on beliefs. Don’t get us wrong, we recognise that the stories of origination that exist around the world are full of philosophy, wisdom, beauty and real life advise that are great and timeless contemplations. But, as prediction-models to get a clear understanding of the world they kind of suck.
So, the second cognitive dimension of Mental Models is belief-based. Luckily, we have total control over this dimension and can deliberately change those Mental Models if we wish to do so. Interestingly though, it seems like we are losing some of that good old meaning-making capability of Mental Models as we upgrade them to higher cognitive dimensions. Some of the greatest thinkers of all time, among them Albert Einstein, seem to have been able to hold on to meaning-making belief models, while selectively upgrading world-predictive models to higher cognitive dimensions.
Ego is probably as old as faith, if not older. It is the, mostly unfortunate, human condition that fuels pride, stubbornness (at least the bad kind) and tribalism (definitely the bad kind). Preference-based Mental Models range from wishful thinking to ideology, i.e. from naivety to genocide. Preference is clearly the egoistic cousin of faith, which is evident in for example secular communities that used to be religious. People have ditched their faith, but prefer to hold on to many, if not most, of the Mental Models they used to have as believers, including the less intelligent ones.
Preference-based Mental Models is also the cognitive dimension of most idealists, who interact with the world as they would prefer it to be rather than how it is. We have also included Mental Models that exist due to social compliance in this dimension, which are often much like idealistic ideas. The Semmelweis reflex explains how Mental Models from any of the other cognitive dimensions can end up in the preference bucket due to ego, social norms, moral constructs, etc.
Some 2500 years ago Greek philosophers, among them the stoics, started to talk and write about Logos: The way to think and reason about non-evident matters, a.k.a logic. To form theories, which really are Mental Models based on inductive arguments, in turn based on observations of the world (“theory” is derived from the Greek word “theorein”, meaning “to look at”), is the foundation of science.
Many, if not most, popular Mental Models of our time fall into this category simply because we use them as inductive arguments. E.g. “the apple is a tasty snack, therefor the pear must also be a tasty snack”. While faulty inductive logic can be very dangerous, e.g. “I’m not allergic to apples, therefor not allergic to pears”; well-crafted theory can be very powerful as the foundation of Mental Models. Powerful, as in good at predicting reality.
Evidence-based Mental Models, the superior cognitive dimension for understanding the world, generally comes from logic-based models that are tested in the real world and found to accurately predict reality. The evidence-based models range from the very simple ones that we get from experience, e.g. “the stove is hot after cooking on it”; to the very complex ones, e.g. Einstein’s theory (nowadays more of a law than a theory) of General Relativity, derived from logic.
When we consider the general Mental Models that are currently getting more and more popular to deliberately use as tools to sharpen our intellects, some but not all fall into this dimension of superiority. For example, a growing body of evidence suggest that multitasking is generally a bad idea. We can broadly apply that evidence-based insight to fight belief-, preference- and even logic-based Mental Models that upholds the practice of multitasking.
With a different very popular Mental Model, the 80/20 rule, the application of the model is much more delicate and we will quickly fall into induction-land if we are not careful. While it is a proven optimization principle in many disciplines, we can’t take for granted that it is a good predictor of reality or results in other areas. For example, in programming, 80% of errors typically come from 20% of the code, so identifying the 20% and focusing on that code is an evidence-based model proven to work. From here it is easy to make hasty generalizations and for example start reading 20% of patients’ journals to cure them by 80%, or why not to focus only on the 20% of lovemaking that accounts for 80% of the pleasure? Moving evidence-based Mental Models away from the evidence and into the inductive logic dimension can quickly turn into a slippery slope ride down to the lower levels of predictive capabilities.
What to make of this discursion
For now, this is how we see the Mental Model of the Mental Model, the MM2. We consider it to be a logic-based model supported by some evidence, though it must be critically contested that such evidence is taken out of its original context. I.e. proceed with caution and please propose contributions or suggest revisions. Based on our understanding of Mental Models, however, we are already using these insights in the following manner and would argue that you should too:
- Apply the Five Cognitive Dimensions of Mental Models model to all of the Mental Models you use deliberately. It will help you see where there is relative strength and weakness in the different models you use, and where you should focus your efforts on lowering your conviction levels so that you can aspire towards Mental Models in higher cognitive dimensions. Most probably this will result in a Pereto distribution, i.e. a minority of models will likely be evidence-based, but have the most impact, and the bulk of models will end up in the lower level.
- Identify your autonomous Mental Models that clearly fall into the first three categories of the cognitive dimensions; instinct, belief and preference; and see if you can start using deliberate versions that are upgraded with logic and evidence in the same way, i.e. by first lowering your conviction levels for the current model in use. If possible, engage in deliberate practice to internalise the upgraded versions of these Mental Models. In time you can let them go back on autopilot, with the upgrades installed.
- Identify future situations where you are certain that you may fall for using Mental Models of the lower levels of predictive precision. Create cognitive triggers for those situations that force you to deliberately chose how to think instead of autonomously reacting. E.g. name folders on your computer to keep you in check, like: “Recruiting – by evidence not preference” instead of just “Recruiting”. Or keep a journal. Or record your own guided meditations. Or set recurring alarms on your smartphone with messages to remind you of the change you are trying to make. Get creative.
- One Mental Model that we use autonomously all the time is the “Situation” model, i.e. the sorting algorithm in our brains trying to inform us about what is happening (make a prediction) and why (constructing meaning). In the Freudian school, ID (“instinctual desires”) has the first say, which conceptually corresponds to the amygdala response in neuro-scientific terms. But after that initial hard-coded response we can make a lot of deliberate changes to our thinking patterns.Upgrading our “Situation” Mental Model to higher cognitive dimensions gives us superpowers. The best examples of this in the business world is probably Ray Dalio’s Case-at-hand approach and Charlie Munger’s Always invert Both guys have managed to rig their minds with very low conviction levels for Mental Models that work outside of their respective Circle of Competence. In summary: Mastering the Situation Mental Model makes you better at making accurate predictions about the world.If all these things are new concepts to you we advise you to read the books “Principles” by Ray Dalio and “Poor Charlies Almanack” by Charlie Munger and Peter D. Kaufman. Virtually all Mental Model enthusiasts have read (and love) those two books, for good reasons.
- A second Mental Model that runs too much on autopilot for most people is the “Decision-making” model. Of course, decisions look very different depending on the situation, and how you make decisions is to a large degree contextually influenced. But the act of making decisions is its own Mental Model: The system to change, i.e. by a decision, is you, arguably a very complex system. The decision is per definition a state-change, or the prevention of a state-change. And if there is no uncertainty, then you are just wasting your time deliberating on a decision that is already made.
Our next article: The Mental Model of Charlie’s Latticework
This article focused on understanding the inner workings of the Mental Model as a singular thought process. We have already started to work on what we think is the obvious next chapter: How a plural of Mental Models interact with each other, autonomously as well as deliberately, and how it is possible to apply the MM2 model to Charlie Munger’s idea of a latticework of theory to become great thinkers. Make sure not to miss it by following Eicorn on LinkedIn.
Thank you for reading! Please let us know your thoughts, if this was helpful to you, or maybe just a waste of time. After all, we want some more evidence before we can accept or reject the MM2.