Why rational arguments rarely change minds

Mark Pottenger


Built into our Brains

"Give me a child until he is seven and I will show you the man." is attributed both to Aristotle and to St. Ignatius Loyola, founder of the Jesuits. Whoever said it, it encapsulates a very important fact of human nature. Most humans will never fundamentally change beliefs they learned during their early lives. Our existing beliefs are strongly defended by biases built into the way our brains and minds work, including Confirmation Bias, Motivated Reasoning, and many more.

A very useful model of thinking is described in Thinking, Fast and Slow: System 1 is fast and parallel-processing, and evolutionarily older, and System 2 is slow and single-processing, and evolutionarily newer. I described a little of this model in an earlier musing. A site I found gives a good follow-up on some details.

The System 1 label mostly overlaps with the older labels subconscious, unconscious, pre-conscious, non-conscious, non-self-aware, reflexive, non-reflective, etc., but doesn't carry the baggage and associations many of those labels carry from other mental models.

These systems do not map to completely separate parts of the brain, though some areas more strongly support one system or the other. They also do not run independently—both system are active and interacting all the time.

I think that triggers, a term I have seen used increasingly over several years, are a System 1 phenomenon. System 1 quickly responds to a trigger with whatever trained reaction is has stored, and System 2 might not catch it.

A key insight with using this model is that System 2 can easily get fatigued, and when this happens, System 1 output will get used (treated as reality) without any System 2 fact-checking / editing / cleanup / monitoring. Since System 1 is full of biases and heuristic short-cuts that return fast but sometimes unreliable answers, a physically or mentally or emotionally tired person is very likely to give the appearance of living based on unthinking trained reactions rather than conscious thought. The beliefs System 1 retrieves start from a core accepted in our earliest years of life. A huge weakness of System 1 is that it substitutes familiarity or easy retrieval for truth, so recent or frequent lies are treated by System 1 as true. System 2 is also subject to biases and errors, but as a focus of conscious attention there is a better chance of catching them than with an unchecked System 1 response.

People are terrible at telling fact from fiction, partly because fictions are usually presented in ways that better fit our mental (System 1) preferences for stories than the ways facts are usually presented. System 1 is more likely to accept a good story than dry facts that take more energy-expensive System 2 processing to make sense. If a person is too tired for System 2 to be active, the story is treated as reality.

Several studies have suggested that conservatives are driven by fear and negativity more than liberals are. If this holds up, they probably have less energy available for System 2, therefore spending more time operating with just System 1 unthinking trained reactions.

I view most human characteristics as varying along measurement axes, and assume the same is true for the natural activity level of System 2 for each person, though I have not found anything in my reading so far that mentions any experimental confirmation for that assumption. I know I am a better proofreader than many people I know, which anecdotally confirms the assumption, since proofreading is a System 2 activity. If this assumption is valid, it means that even before environmental stresses pile on, some people can use System 2 more easily and for more time than other people can. One site I viewed while searching for data said System 1 is 98% of our thinking and System 2 is 2%, but it didn't say if those are mean values or how much they might vary, or even if they are completely arbitrary (made up) percentages.

I have seen many mentions of a current lack of and need for Critical Thinking in modern life, but almost all such mentions ignore a key problem. Critical Thinking needs a healthy and active System 2. If a person does not have the energy for System 2 thinking, it doesn't matter how much training they get in Critical Thinking skills—they won't (can't) use them. Presenting rational arguments to such people is a waste of time.

Science fiction over many decades has explored a lot of ideas, some of which relate to current problems in our society. In 1951, The Marching Morons by C. M. Kornbluth showed a possible future in which the average IQ (defined in current usage as having a mean of 100) of most of the population was 45 and the burden of keeping society functioning fell on a very small number of smarter people. In 1957, Crisis In 2140 [aka Null-ABC] by H. Beam Piper and John J. McGuire showed a possible future in which only 10% of the population was literate and bearing the burden of keeping society functioning. In 1973, The Stone That Never Came Down by John Brunner depicted a falling-apart society saved by changing human nature to make it harder for people to ignore facts. Modern U. S. society now includes a substantial minority raised on misinformation and disinformation promulgated by politicians and businesses that can't stand exposure to reality from the light of science.

A study published in 2018 found that in social media (the study analyzed Twitter data), lies spread farther and faster than truths. Section 230 of the United States Communications Decency Act (if I got the right citation) largely protects Internet sites from liability for information provided by third-party users. This aspect of human nature combined with this U. S. law creates what economists call a perverse incentive. Since lies spread farther and faster, sites that make their money based on traffic will make more money by allowing lies than by enforcing honesty. Since monitoring content to enforce honesty also costs more, sites have monetary incentives to promote (or at least not discourage) lies. I don't think the current proliferation of mass delusions will stop until the U. S. Federal government changes the incentives by establishing host site and broadcaster liability for spreading lies, even those posted by third parties. Enforcing honesty does not eliminate free speech—people will still be able to say whatever they want one-on-one. Honesty in broadcasts and on the Internet will just remove the amplifiers currently spreading lies rapidly as dangerous mental infections.

Even if honesty enforcement started today, it would still not cure many of the problems besetting the modern U. S. A. (and other countries) because there is still that large minority of the population already raised on and deeply attached to brainwashing / disinformation / lies / misinformation / propaganda instead of to scientific facts. Most of those people, through mental inertia, will continue to base the rest of their lives on the lies that they have accepted. They will accept / believe / trust and copy / echo / forward / like / retweet new lies that fit their current frameworks of lies. Reaching a more rational society will require enough decades to bring up new generations with deeply learned pro-science basic knowledge and biases instead of deeply learned lies.

With Intent

All the mental limitations discussed above apply to all human communications, even when the people involved are not intentionally distorting anything. When we add considerations of intent, the picture gets even darker. Among the entities that intentionally distort communications (create new lies) are advertisers, authoritarian/totalitarian regimes, Bircher billionaires, businesses that do things that are bad for people or society or planet Earth, demagogues, ideologues, and many politicians. Most intentional distorters would not change their messages even if the broadcaster and site host shields went away—they would simple bear any extra costs to continue putting out their distorted message because they want to control people through those messages. I suspect that a small subset of lie generators actually believe their own lies, because the new lies tie in to accepted networks of already accepted lies and these people never learn better because they live in System 1. Reining in the damage all these sources do to our world will take much more drastic action, which will be very difficult within the current framework of U. S. protection of free speech. I suspect it will end up taking some sort of restriction on paid speech, and cleaning up years of court rulings that treat money as speech and claim that corporations have the speech rights of people.

Among the many human traits with wide ranges of personal expression are conscience and empathy. I strongly suspect that intentional liars tend toward a weak conscience and low empathy.

Toxic Tales

Because people respond more to stories than to facts, human culture has been influenced for millennia by toxic tales that convince unthinking majorities of the population to allow themselves to be exploited by minorities. These toxic tales include the divine right of kings (that monarchs rule because God gave them the right), that the rich deserve riches (even when they come from rigged laws), that all poor people are lazy, that (ethnic / national / cultural / racial / sexual / class / etc.) group X (minority or majority) has rights to rule over group Y (minority or majority) or has more value, that institutionalized discrimination doesn't exist, that removing discrimination is a new form of discrimination, that corporations have human rights, etc. Removing these toxic tales from a culture is a major undertaking, especially when a large subset of the population accepts the tales as truth, and a lot of power and privilege is enabled and reinforced by these tales.

Bad News

Keeping informed about the world involves a deep dichotomy. Most reporting of "news" only reports actions or events that diverge from normal daily patterns or represent possible dangers. This leads to a heavy bias toward reporting on negativity. A person who wants to keep aware enough of current events to avoid surprises has to work out a personal censoring or filtering system to avoid getting overwhelmed by negativity. I still haven't found a completely satisfactory balance—I still read a daily newspaper, but I do a lot more skimming of the news than I used to.

There are also belief biases built into all sources of news or information, so we all need to be as aware as possible of any biases in the sources of news we choose to use.

Some people simply avoid the mundane world. Historically, there were hermits, anchorites, monasteries, and nunneries. All of these required and still require acceptance by and/or support from the people in the mundane world around them. The world is a lot more intrusive than it used to be, but a contemplative life away from the mundane world is still possible.


Because there is vastly more to know about the world than any one person can learn from personal experiences, we all rely directly or indirectly on experts.

From the Oxford American Dictionary:

"expert n. a person who has a comprehensive and authoritative knowledge of or skill in a particular area."

"authoritative adj. able to be trusted as being accurate or true; reliable."

There are no universal experts, since no human has "comprehensive and authoritative knowledge" about all fields of knowledge. A politician or actor or businessman or broadcast host is not an expert who can be trusted about a scientific topic unless they have also studied that science enough to have "comprehensive and authoritative knowledge" about it. Even scientists who have studied other sciences are not experts about all the sciences they have not studied. If anyone is ever caught committing fraud or lying about a field, they fail the authoritative part of the definition, so they have zero credibility about that field for the rest of their life, and must abandon any claim to be an expert about that field.

Bertrand Russell suggested three rules about experts:

"1. When the experts are agreed, the opposite opinion cannot be held to be certain.

2. When they are not agreed, no opinion can be regarded as certain by a non-expert.

3. When they all hold that no sufficient grounds for a positive opinion exist, the ordinary man would do well to suspend his judgment." (Russell, Skeptical Essays,1921)

From Profiles of the Future by Arthur C. Clarke:

"When a distinguished but elderly scientist states that something is possible, he is almost certainly right. When he states that something is impossible, he is very probably wrong."

Many supposed experts are really only "white coats"—people who assume an unjustified mantle of authority. "White coats" is a reference to many years of advertisements in which someone wearing a white coat was presented as a presumed authority.

I will emphasize an essential point: a person can only be an expert in one field (or a small number of fields)!

When speaking of "the" future (actually, possible futures): if the language is not probabilistic and qualified/hedged, it is not expert. Certainty about the future is possible for demagogues, but not for experts.


I said a bit about scientific knowledge in my Trust musing. I grew up reading references to hard physical sciences and soft psychological sciences (such as anthropology, economics, psychology, and sociology), but those are not very good labels. The solidity and verifiability of conclusions fit those labels, but they are actually reversed if you are discussing ability to get results: physical sciences are easier to research than psychological sciences. All sciences build models, and as computer technology has improved over the last several decades a lot of that model-building has moved to computers, but all models still need measurements of quantifiable things, which are much easier to get for physical sciences. The things psychological sciences try to measure are often interior mental or emotional states, which can only be inferred or determined based on subjective reports by study subjects, since no technology yet developed actually reads minds and emotions (though some technologies are getting awfully close). This currently unavoidable fuzziness of the sciences that cover economics, politics, and human behavior is one reason merchants of doubt have been able to make far too many people distrust all sciences.

In most sciences, the natures of the quantifiable things to be measured are actually parts of the model. Models give us language to describe things and frameworks for how to think about them. For example, many economic models use totally rational and self-interested individuals as the basic decision makers, and leave the actual complexities and irrationalities of real people for behavioral economists to deal with. Almost every model that attempts to predict possible futures that are affected by human behavior produces results that have huge ranges because of the difficulty of predicting human behavior. E.g., early attempts to model what to expect in the COVID-19 pandemic completely missed the politicizing and self-harming irrationality that ensued. Climate models give multiple predictions based on different levels of change in human-caused greenhouse gas emissions, with actual behavior often turning out worse than the most pessimistic models. As another example, the System 1 and System 2 separation of mental functioning is a psychological model. Once that model was formulated, many old observations could be fitted into it and new experiments could use it to make predictions. Observations without a model are rarely science. A hypothesis is a first step in a first draft of a model. If enough confirming observations and successful retrodictions and predictions accumulate, a hypothesis can grow up to be considered a theory.

Elaborate models can be programmed on powerful computers, but then they must be tested and validated. This is much easier for physical sciences than for psychological sciences, but even there it is easy to create an incomplete model. E.g., I have programs that let me calculate orbital motion for almost any asteroid with published elements, but I know my programs fail for asteroids that make very close approaches to planets. My programs are incomplete models, treating all bodies as point sources of gravity and missing refinements needed to handle the effects of the shapes of the planets.

The General Circulation Models used to model climate are incredibly complicated, incorporating a very large number of factors. They are tested by simulating climate from a past date up to the present, then run forward to predict possible futures based on various assumptions. The testing by retrodiction has been essential in determining some of the factors that must be included. Even with all the knowledge of the world already built into these models, new scientific discoveries still have to be added to them. Good scientists always stay aware that models and simulations, no matter how good, are still not reality.

Most good science takes time! Equipment can take years to build. Experiments, especially in biological sciences, can take years to run. A lot of confusion about COVID-19 arose from trying to formulate policies based on incomplete data. Researches are still learning more about COVID-19 every day, and it will take years to get clarity about long COVID.

Levels of Awareness

I have mentioned the terminology of Levels of Awareness in earlier musings (here and here).

All of the discussion above is strictly within First Level Awareness, which is reactive (the world acts on people and people react to the world, often in a knee-jerk reflex manner [System 1 in the discussion above]) and prone to seeing people as victims.

Second Level Awareness is active, with people able to initiate thoughts and act on the world. Almost all self-help and self-actualization systems are trying to help people achieve Second Level Awareness, and good ones realize that this awareness can never be imposed from outside the person. Second Level Awareness requires System 2 conscious thought. If a person is too busy or distracted or tired or stressed to ever activate their System 2, they will not spend any time in Second Level Awareness. Lessons and examples and techniques can help people achieve epiphanies, but they can't guarantee them. Each person is a unique mix of many characteristics (personal values on every scaled psychological axis ever invented to describe people) and life experiences, and the possible triggers for epiphanies are just as unique.

In general, someone locked into System 1/First Level Awareness will never even realize that System 2/Second Level Awareness is possible. Only someone who has experienced and noticed moments of System 2/Second Level Awareness (Awareness of Awareness) will realize that more of it might improve their life. Anyone working to increase the time they spend in System 2 and Second Level Awareness will accumulate a toolkit of techniques. Because of our unique psychologies and experiences, every toolkit will be different. Just as one person may prefer a screwdriver with changeable tips and another person may prefer a set of separate screwdrivers, our mental/emotional/spiritual toolkits will all be personalized. We can ask people we know and trust for ideas, then test them to see if they work for us. We can research models and techniques, preferably ones that have been scientifically studied, but keep cautions in mind. Many studies get published even if they have problems: small sample sizes, bad design, bad use of statistical techniques, no replication, etc. Even when the studies are valid and the results look good, the studied techniques may not be useful for us because of our unique psychologies.

In building our personal toolkits, we can pick and choose individual models and tools from many possible sources. Every existing system of therapy or self-improvement or mental development is based on some model of human mental and/or physical nature, and one or more techniques to manipulate and improve a person's mind and/or body. The degrees to which these systems have been validated by good scientific studies vary widely. Someone looking for personal tools can research many systems to look for pieces of models that make sense for them, and for techniques that they can actually practice.

Some possibilities to investigate include:

Acceptance and commitment therapy (a subset of CBT).

Analytical psychology (Jungian analysis). Includes anima and animus, archetypes, the collective unconscious, complexes, extraversion and introversion, individuation, the Self, the shadow, synchronicity, and psychological types.

Astrology gives a psychological model of 12 personality components, with every person having a unique personal mixture of those components. The components of what Zipporah Dobyns called the twelve-letter alphabet are Aries (I do my thing.), Taurus (I enjoy the sense world.), Gemini (I see, conceptualize, and talk.), Cancer (I save, protect, nourish, and assimilate.), Leo (I rejoice in expansion.), Virgo (I work competently.), Libra (I enjoy balance.), Scorpio (I penetrate, control, absorb or eliminate according to my desire.), Sagittarius (I trust, value, and direct my life according to my understanding.), Capricorn (I carry out the Law.), Aquarius (I seek new knowledge and brotherhood for all mankind.), and Pisces (I dream of love and beauty and am absorbed in the whole.).

Cognitive behavioral therapy (which can be an umbrella term including many named systems of therapy).

Gestalt therapy.

Individual psychology, founded by Alfred Adler.

Linguistic relativity. Language influences thought, as is obvious in all my references to models.

Maslow's hierarchy of needs.


Neuro-linguistic programming.

Psychoanalysis. Freud's id, ego, and super-ego model, work with the unconscious mind, and many other ideas that have spread to mainstream culture.


Relaxation techniques.

Tai chi.

Transcendental Meditation.

Twelve-step programs like AA.


Examples of individual exercises include:

Try affirmations, though they are only useful for some people.

Establish and use a personal mantra, such as love, joy, happiness.

Practice experiencing a feeling (such as happiness) that you want to increase in your life. If it is too hard to directly get yourself into that feeling, find what does let you achieve the feeling as an effect. This could be doing a specific sort of visualization, reading something in a particular genre, listening to a certain kind of music, watching something, thinking of a particular place, etc.

See Lizards and Models for a follow-up.

Copyright © 2022 Mark Pottenger

Musings Home