Knowledge is Domain Specific : an elusive, inscrutable and profound truth

The statement that knowledge is domain specific has the ring of something blandly matter-of-fact and dull. It is, in fact, of earth-shattering importance and it explains a great deal concerning the limitations of human cognition.

For instance, professors of statistics and probability are generally unable to apply their knowledge to ordinary, everyday affairs and they come up with the wrong answers. They know how their expertise applies to certain domains; perhaps reverse mortgages, or the likelihood of chimpanzee number three getting to mate, but do not know how it applies to used car purchases, dating apps, or paint prices.

If even the professors are unable to apply what they have learned, it seems rather pointless to insist that psychology undergraduates learn the concepts and methods. However, this is simply the situation of all of us in every field. Those psychology students should potentially be able to apply stats to certain kinds of experiments, though possibly completely unable to apply the same concepts to anything outside that narrow topic.

The domain specificity of knowledge explains the need for explanations. Explanations take something a person is already familiar with and relates it to something unfamiliar. It is not possible to explain the unfamiliar by appeal to something also unfamiliar.

If it were possible for someone to extrapolate automatically from something he knows to all circumstances where the concept might usefully apply, explanations would be redundant and human beings would get a lot closer to omniscience.[1]

Ironically, even the insight that knowledge is domain specific is domain specific and it is necessary painfully to learn, typically one domain at a time, how it applies. It is fascinating to recognize how the concept applies yet again.

Creativity is sometimes described in this way; as often taking something known and using it in a novel way, which would mean applying it in a new context.

There is a tendency to think of examples as expendable nothings that serve only to communicate a point or principle. However, showing how principles apply is often where all the real work is done.

Students sometimes consider an article or book repetitious because they focus on the principle and do not realize the work being done by showing how the principle applies to one domain after another.

The phrase “knowledge is domain specific” can be found in Nassim Nicholas Taleb’s book Antifragile. Antifragile is a neologism coined by Taleb because he knows of no word in any language that refers to something that gains from stress, uncertainty, variability and the random. There are things that are robust; that hold up well against stress. But then there are things like organisms and the economy that benefit from stress. Exercise, whether it be walking, running, weightlifting, yoga, is stress. So long as the stress is not too extreme, it is extremely beneficial. In fact, the less exercise someone gets, the more fragile he becomes. The immune system needs stress in order to stay healthy and strong. It is not merely a matter of resisting harm. It needs small “harms” just to keep functioning properly.

One not every good student commented about Taleb – “he keeps repeating himself” – not realizing that to simply learn the definition of “antifragile” or “optionality” is nearly worthless. It is seeing how the concepts apply and to what they apply that is the interesting part.

The notion that knowledge is domain specific is very hard to write about because its wisdom has infinite application and each example of knowledge being domain specific can go on forever – showing over and over again how each concept applies in one context after another. Each example can potentially fill endless books.

A failure to realize the domain specificity of knowledge ruins what would otherwise be a very brilliant book. The book is The Systems Bible by John Gall, the first edition having been called Systemantics. Gall compiles a list of general principles of very wide application and of significant truth. Examples include the tendency of systems to get bigger and bigger at the rate of 6% a year. The bigger and more complex a system gets the more unpredictable its results. If a job can be done without a system, do not start a system. Systems are attempted solutions to problems. Systems create their own new problems. If a new problem is encountered it will usually not be possible to apply a preexisting system to solve the new problem because that will not have been the problem the system was designed to fix. Most large complicated systems do not work. E.g., systems for assessing “learning outcomes” for students. If a system does work by some chance, leave it alone. Often people try to make a system work by pushing on it and trying to get the system to work harder. This usually just makes things worse. It is not possible to judge what someone is actually doing by his title within the system. A “shipbuilder” of giant ocean-going ships does not actually build ships. At most he arranges financing for them. Systems become their own reason for existing. Most effort is put into maintaining the system, typically at least 90%. If a system is 10% efficient it is working wondrously well. An example would be charities, or businesses. Most of the money produced by a business goes back into the business in the form of salaries and other expenses.

Thomas Sowell worked for the Labor department in 1960 as a summer job, studying the minimum wage in Puerto Rico. He found that as the minimum wage was raised, employment went down. Unions blamed lower employment rates on hurricanes that had harmed the sugar plantations. Sowell realized that to test this theory statistics were needed to show how much sugar cane was left standing after the hurricanes. His more senior colleagues were horrified because the minimum wage law was providing one third of the income of the department of labor.[2] His question threatened the self-interest of the government department. Sowell then realized that the Labor department’s business is primarily about maintaining the Labor department, not about solving problems. In the Labor department’s defense, this is true of nearly any system – that is why there should be as few systems and government departments as possible.

The huge problem with Gall’s book is that he lists the principles but makes the enormous mistake of then saying, to paraphrase – “now that you are familiar with the principles you will now see how they apply to very wide aspects of life.”  Despite all his insights, he failed spectacularly by seemingly not being aware that human beings are unable to extrapolate to widely varying domains, and certainly not all of them. Gall treated the extrapolation/application as though it were a matter of course and would take care of itself. It does not.

In the first edition of the book he solicited examples of applications from his readers. He did in fact receive some. Disappointingly, in the revised versions, these examples were ridiculously brief. Sometimes a line or two and there were still very few of them. Of course, a library could be filled with examples of malfunctioning, metastasizing systems. It would have helped, however, to show how the principle applied just to, say, government departments, or university administration. Readers could then hope to get ideas about further applications in other possible domains.

If someone were paying for ideas it might make sense to pay five dollars for the idea, but one hundred dollars for each successful application of the idea, depending upon profundity.

Sometimes an idea is applied too broadly. It becomes an idée fixe. Many years ago Ken Wilber wrote of a mindset he called “leveling.” It is egalitarianism run amok. When confronted with nearly any sort of hierarchy whatsoever, the hierarchy is rejected, including the distinction between good and evil (moral relativism,) between one culture and another (cultural relativism), grammatical English and ungrammatical, income distribution, prizes, awards, Valentine’s Day cards; on and on and on. It can get to the point where practically whatever someone says, if the utterance “leveling” were made, it would amount to the same thing. A similar phenomenon can currently be seen with labels like “sexist,” and “racist.” They become bats to hit people with, with no thinking or arguing required.

Since radical egalitarians rank “leveling” as better and higher than ranking, and view themselves as morally superior to those who explicitly believe in moral superiority and inferiority, their view involves a contradiction and is thus false.

The leveling idea is normative in character. It is a desire to obliterate distinctions and to turn everything into a goopy mess of “acceptance.” Except, many deviations from leveling will not be met with acceptance at all but violent denunciation.

By being aware of the phenomenon of “leveling,” it is possible to connect seemingly disparate phenomena and see that they are all part of one pathology.

Unfortunately, in trying to write about the domain specificity of knowledge, the proof is in the examples and the examples never end. I, like John Gall, must hand the ball off to the reader and remind him to keep the idea in mind – particularly in the context of pedagogy. Do not imagine that very much work has been done if only a principle is communicated – and paradoxically – the notion that knowledge is domain specific is just one more example!

A Musical Analogy

Recently, a friend of mine, Thomas F. Bertonneau, and I found that we shared a common musical fascination. He with the musical form passacaglia [the “g” is silent] where a “ground melody” plays while over the top another melody is layered, somewhat similar to the musical form chaconne, or ciacone, where a chord progression plays underneath variations in the treble.

Blues, rock, and some jazz solos therefore are chaconnes where a musician improvises over a repetitious chord progression played by other members of the band.

Since I was a teenager, I have been attracted to and interested in an inverted version of that where the treble stays the same and the chords or bass line changes. What seemed interesting is the way in which the meaning of the musical phrase in the treble kept altering as the rug was pulled out from under it – as the fragment was layered over context after context; an interesting combination of repetition and novelty by continuously reframing the picture. Each time the musical connections, harmonies and associations varied.

When I recently gave a musician the CD version of a “mixed tape” he commented that several of the songs had this feature. I had not even noticed. I just knew that I liked the songs.

This application of the domain specific nature of knowledge is taking place on a poetic and intuitive level.

Meaning involves connection to context. Place the same item in a different context and the meaning changes. Singing a happy song at a children’s birthday party or in a death camp mean two quite different things.

I repeatedly had the experience while working on my dissertation twenty two years ago of realizing how a familiar idea applied to multiple phenomena. There would be a sense of things clicking into place; like some kind of semi-mystical experience. It even seemed to have a visual component in my imagination. The original concept seemed like a brand new idea again. In a way it was because it now meant something quite novel.

Each time a principle is applied, if done correctly, a new meaning of the principle is revealed; a new connection to context. It is insight.


[1] I imagine someone has had this thought before, about the relationship between knowledge being domain specific and explanations. If so, I would like to know about it.

[2] Comments from titled “Thomas Sowell: I started as Marxist and Dept of Labor changed my view.” How the department’s income and the minimum wage law were related is not explained.

26 thoughts on “Knowledge is Domain Specific : an elusive, inscrutable and profound truth

  1. I think this is fascinating. I had never thought of it that way. But (I think) it gets to the core of why I am so uncomfortable with the idea of a ‘Theory of Everything’ sought by physicists, sociologists, economists, etc. It is overtly trying to apply specialized principles outside of the area where the principles are useful. Sure, to badly paraphrase Mark Twain, there are some areas where it Rhymes. But there will never be a Theory of Everything which adequately explains…well, everything. Seeking to understand first a Theory of Physics, then a Theory of Sociology, then a Theory of Economics, is additive to an individuals knowledge. But one cannot then attempt to connect all three. It would be like taking the first half of Moby Dick, the middle of Dune, and the last part of Les Miserables, taping them together, and calling it a ‘Book of Everything’. Sure, it’s got pieces of greatness, but you won’t learn anything about each piece by reading it cover to cover.

    After writing all that, I REALLY, hope that i’m understanding your article properly.

    • @Scoot – Yes. I’m not promoting a theory of everything and like you I have always disliked the phrase. It’s hubristic. I’m only writing about knowledge as principles or concepts that can be usefully and appropriately applied in different contexts. Misapplication and trying to apply ideas where they do not belong is not helpful.

      The key idea is just that we are unable to immediately see in what other contexts an idea might be illuminating and this puts a real limit of human cognition. Also, that it is not enough to know a concept, theory, or idea – knowing how and when to apply it and to what is where a lot of the work is done.

      • This was shared with me by a friend. The Venerable Fulton Sheen shares your insight: Knowledge is not universal. He mentions it in passing, almost as if it’s a given. It’s a lesson that bears repeating and emphasizing. I’m glad I read your article first, or i wouldn’t have caught it.

        Thank you!

  2. So, I guess that an elementary school teacher who knows that 2 + 2 = 4 can only apply that in her classroom. Whenever presented with the same arythmetic operation outside his/her field of expertise, the teacher will not be able to perform it.

    ” Children who are not exposed to bullying, teasing, hatred and the emotionally uncomfortable, will become less and less able to cope with such things when they do inevitably encounter them. The remedy for being “triggered,” suffering great floods of tears and wailing and emotional distress, is to be exposed to the triggering phenomenon repeatedly and often in a manageable manner.”

    As someone who has been systematically bullied in 95% of the activities of his life between the ages 6-13, you have absolutely no idea what you are saying.

    • Hi, APC – thanks for reading and commenting. What you are doing is called a strawman argument where you make someone’s position more extreme than it is and then attack an argument they are not making. If you read carefully you’ll see that I make some effort to include qualifiers like “often,” “generally,” “frequently,” and the like. And no the elementary school teacher will not know every instance in which 2 + 2 = 4 might be applicable. E.g., looking at some complicated piece of equipment on a space station. It is not a question of not being able to do the addition. It’s about recognizing every single instance in which it might be useful.

      Overprotected children will of necessity be far less emotionally robust than children who have not been. Perhaps the key phrase in your case might be “manageable manner.” The case I have in mind is the so-called snowflake students who think they should not be “triggered.” It might be worthwhile finding out why “you were systematically bullied in 95% of the activities of his life between the ages 6-13.” That doesn’t sound good. If you have a therapist, does he/she say you should now avoid social encounters that might make you upset?

      • Yes. Apparently bullies and bully victims tend to share a connection. The victims of bullying tend to also be the type who bully and come from fatherless homes.

      • To your saying that I fall in a strawman argument, I would retort that the basic for operations of math are universal and not domain specific. I circumscribe myself to the 4 basic operations of math in order to keep things simple and avoid red herrings.

        Regarding your hypothetical consideration of the case of the space station, I would retort that yours is an argumentum ad ignorantiam (very typical of the Critical Thinking Theory), where you say something like “A is not A because maybe B is C”.

        Finally, considering your sayings about bullying, “managed bullying”, I would say that
        – bullying is in many cases perpetrated not by an individual, but by a group
        – the victim is often found to have a weak defense/support (either because the victim comes from a dysfunctional family, which was my case; or by a shattered one)
        – the victim finds himself/herself in a circumstance where the ones to take action are either blind to the abuse (bullies are not stupid, in many cases they abuse in the absence of authorities or where the authorities are blind to the abuse) or indifferent
        – the psychological effects of bullying may be so deep that the victim actually starts to think that he/she is deserving of such abuse.

        This pseudoargumentation that bullies are not that bad or are not fully accountable for they behavior and victims should “grow up” is what in many cases keeps things going.

        Not very long ago, Pope Francis said literally that “buying is the devi’s work”. This is not the saying of a fantasy novel writer, but of a person with at least 12 years of formation if the field of theology/philosophy/ethics/religion, and who is the Vicar of Christ.

        Granted, overprotected children have development issues and there is a need for mothers to let go so that children become men; however, bullying is a completely different ballgame.

        If in grown adults bullying can lead to suicide or psychological trauma, what do you think that it does to a child?

      • APC – I agree with your statements about bullying. Bullying is scapegoating and should not be tolerated. By analogy with exercise, stress in the form of teasing and social difficulties is good, but bullying is the equivalent of being dropped 100 feet. No good comes from it and just harms a person. I’m just removing the paragraph since it was only supposed to be illustrative anyway, rather than trying to fix it.

        Unfortunately, your attempt at sarcasm in the first paragraph was incoherent and just amounts to trolling and infantilism, possibly unintentional, so I’m removing it. (For readers of this the assertion was that the title of the article indicated that I was incapable of creating a valid argument).

        I do believe that you are misunderstanding the phrase “knowledge is domain specific.” This is knowledge as known by a knower – hence the talk of extrapolating.

        2 + 2 = 4. OK. Please list all appropriate applications of this fact, past, present and future. Your inability to do this is what is being highlighted.

        Why is this important? Because it is sometimes considered sufficient to teach general principles and regard examples as irrelevant, or as a mere teaching tool for understanding the principle. If this were true, human beings would be geniuses compared with our actual abilities.

        Even if a truth were to be universally applicable, each knower must engage in the task of identifying each domain to which it applies.

        The knowledge of an individual human being is domain specific.

        It is impossible to anticipate all the contexts in which an item of knowledge or concept might usefully and appropriately be applied.

        For instance, bullying is a form of scapegoating. Understanding scapegoating, the mob ganging up on a victim and bonding in shared hatred, might count as knowledge. Let’s say this is true. SJWs are able to identify the scapegoating of “lower” victims but are blind to the scapegoating the higher, e.g., when feminists scapegoat men, or activists scapegoat whites, or the president. The fact that scapegoating applies here is unknown to them. They only recognize the dynamic in limited domains.

        I’m a philosophy professor and have taught critical thinking – though I don’t think much of it apart from some informal fallacies – but I have never heard of “critical thinking theory” per se and doubt that it exists. Your attempt to categorize my space station example as a fallacy is erroneous. Hopefully it will now be clearer why.

        There is a thought/idea/principle and then there is context. “Domain” means context. Just because a principle might usefully apply in a context does not mean that an individual is aware of that fact. You will not understand this concept if you just keep focusing on the principle.

      • Prof. Cocks:

        … The fact that scapegoating applies here is unknown to them. They only recognize the dynamic in limited domains.

        Quite. And this phenomenon goes beyond various illustrations of the scapegoating mechanism, as you have intimated. I’m thinking specifically in terms of the old leftist trope declaring that “you can’t legislate morality.”

        Now, it might well be that the authors and main purveyors of such abject nonsense know precisely what they are doing for nefarious purposes, but those unthinking among us who pick up the slogan and run with it, as though it were even possible to follow in a moral world, among moral beings, and as if it were true, I tend to believe have no conception of the impossibility of adhering for a moment to its illogical precepts.

      • Thanks, T. Morris:
        Yes. It is sometime hard to distinguish between being obtuse and being intentionally obtuse. The fact that Neo-Marxists elevate the “cause” above the truth certainly confuses matters.

      • I would say that Professor Cocks if anything understates the danger. Absent a thorough understanding of context, one is not only apt to miss applications of a principle but also to misapply it with reckless overconfidence.

        Suppose A has a quarter and B has three pennies. B is supremely confident in his arithmetical skill but doesn’t know the value of coins, and he insists that his change is more valuable than A’s because 3>1, and anyone who says otherwise is a math denialist. In vain does A protest that 3>1 is not this issue in dispute.

        Few would be that silly, but Prof. Cocks is a philosopher and so deals with principles that are very abstract but don’t always seem very abstract, that are difficult to apply but this difficulty is often forgotten. And yet–this is my discomfort with a great deal of metaphysics–if we can’t agree on the rules for applying a principle, do we really understand it at all?

      • “This pseudoargumentation that bullies are not that bad or are not fully accountable for they behavior and victims should “grow up” is what in many cases keeps things going.”

        If someone asks you to point to where this was said or implied, you wouldn’t be able to answer.

        There is no shared responsibility. A bully is responsible for his acts. Saying that the person being bullied is also responsible for his acts does not take anything away from that previous statement. The reason things keep going this way is that people are told they aren’t responsible for anything.

  3. Pingback: Knowledge is Domain Specific : an elusive, inscrutable and profound truth | Reaction Times

  4. So, if I may ask: does the word “domain” mean “subject” or “context”?

    And if I may also ask: does this have anything to do with our finite intellectual capacity? Are you arguing that because our intellectual capacity is finite, then our knowledge is domain specific because it is limited by our intellectual capacity?

  5. Why Oppenheimer couldn’t change a tire. I see this is a little in today’s engineers; they’re seemingly very good at starting at spreadsheets but come up short in the spontaneous creativity design department.

  6. “[1] I imagine someone has had this thought before, about the relationship between knowledge being domain specific and explanations. If so, I would like to know about it.”

    I’m not certain of where this has appeared, but it sounds a lot like something Michael Polanyi would say, do you study him?

    • Hi, nota bene,

      Thanks for reading.

      That sounds likely. I haven’t actually read Michael Polanyi but he has been recommended to me as somebody I might find congenial. I think I started “Personal Knowledge” but got annoyed with the first couple of pages. But I would expect to have a lot in common with him.


Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.