The Khan Academy did a video where they asked random people in the streets what the internet is, a question to which almost all gave some ridiculous answers —confidently. A young girl said it is the WiFi because, according to her, "no WiFi, no internet", and a grown-up said, "It is like plumbing, it’s always moving".

The phenomenon this tries to prove is that we most times possess untested knowledge about things, which may turn out to be incomplete or outrightly wrong. This is called The Illusion Of Explanatory Depth (IOED for short).

Yet, the reality of tacit knowledge proves that it is human to know and not be able to explain in great detail, or in any details for that matter.

The Illusion Of Explanatory Depth

The term IOED was coined by Leonid Rozenblit and Frank Keil in 2002, stating: "Most people feel they understand the world with far greater detail, coherence, and depth when they really don’t".

The two were able to prove this oddity through multiple multiphase studies. In the first phase, Rozenblit and Keil asked participants to rank how well they understood fabrications such as a cell phone, crossbow, or a sowing machine; and in the second phase asked the participants to put down a well-detailed explanation of how each fabrication works. Afterward, the participants were finally asked to re-rank how well they understood each one.

Repeatedly, the outcome of the study showed that ratings of self-knowledge plummeted dramatically from phase one to phase two, after participants were faced with their inability to explain how the fabrications operate.

Beyond fabricated objects, the IOED extends to more abstract things like politics, tax laws, human rights, and many other things. It seems we don’t know as much as we think we do, and we sometimes possess a misunderstanding in the place of an understanding.

Explanations So Far

In part, the reason for this may be because what we think we know about a thing is usually dominated by how we feel about it.

Prof. Steven Sloman, a cognitive psychologist at Brown University explained: “If you’re forced to give an explanation, you have to really understand, and you have to confront the fact that you might not understand. Whereas when you give reasons, then you do what people do around the Thanksgiving dinner table. They talk about their feelings about it, what they like, what they don’t like.”

Also, it may well be as a result of an inability to differentiate between facts and 'know-hows’. The award-winning physicist Richard Feynman observed, "I learned very early to differentiate between knowing the name of something and knowing something".

You can know what the term philosophy means intuitively, know how to use it in sentences, derive meanings when someone else uses it, potentially misuse it, and yet not be able to give a concise definition of the term.


If we are to put simply the arguments for the Illusion of Explanatory Depth, it would be that: we really know only when we know in detail, and being able to articulate what we know —or think we know— is a way of telling if we actually know in detail.

Of course, knowing in detail is your best bet if you want to possess an understanding that is more than what the average person has, especially since some people  — though loud— are ignorant of a topic; but, using the ability to articulate as a way of telling if you know in detail isn’t a good one.

Tacit Knowledge

I find it interesting that the category of knowledge that validates the IOED falls under tacit knowledge, which is a kind of knowledge that is in itself difficult to transfer(reveal) to another person, through writing or verbalization.

This kind of knowledge was first pointed out in the works of Michael Polanyi, Personal Knowledge in 1958 and in a later work, The Tacit Dimension. He asserted that "we can know more than we can tell".

He later went on to state that it is not just that there is knowledge that cannot be effectively communicated by verbal means, but also that all knowledge is rooted in tacit knowledge. With tacit, we often are not aware of the knowledge we possess.

Examples of tacit knowledge are the kind of knowledge required to use a cell phone, shoot a crossbow, or even, use a sewing machine. We can know these, but not know how we know them, or give a detailed description of what we know about them.

So if I had taken part in any of Leonid Rozenblit and Frank Keil's experiments, I would have felt confident that I knew enough and chosen to give a high ranking, and then when my inability to give details comes up, which has been wrongly used as a benchmark for measuring how much I know, I will be forced to agree that I do not know as much as I had thought I knew.

Understanding How This Is Possible:

Further, to better understand how possible it is that we can know more than we can say, we turn to the field of Neuro-Linguistic Programming (NLP) and in particular to the works of Richard Bandler and John Grinder. The duo wrote the very first work on NLP, in their 1975 book titled The Structure Of Magic.

Human beings, according to them, use language in two ways:

  • To represent our experiences. Where we are using language as a representational system, for creating a model of our experiences. We do this through active reasoning, thinking, fantasying, and rehearsing.
  • Secondly, we use language to communicate our model or representation of the world to each other; through talking, discussing, writing, lecturing, and singing.

The latter forms a part of what they call the Surface Structure, which comprises what we voice out when we try to communicate our perception of the world; while the former forms a part they call the Deep Structure, which is the perception one has  — how it is mentally.

For a better understanding, let’s look at these two sentences:

  • The lion killed the man; and
  • The man was killed by a lion.

Here, the words and their arrangements, Surface Structure, differ, while the message conveyed, which is the Deep Structure, is the same.

This minor difference between the both, as a result of an arrangement of the words in a different order, they called Permutation Transformation.

However, Surface Structure can differ from Deep Structure to such a degree that the Surface Structure when communicated will fail to pass the underlying Deep Structure message. In this type of difference, the complete logical semantic representation, the Deep Structure, may fail to appear in the Surface Structure.

For instance, if I want to say that John spoke rudely to Mr. A, but merely say that John talks rudely; though I know what I meant, the Surface Structure, by being different, can mean something else to someone else. When Surface Structure differs from Deep Structures, they may at best fail to pass the Deep Structure message accurately, or at worst assume a different meaning.

When we try to communicate our minds, we unconsciously make word choices that can either express accurately what we intend to or fail woefully at passing our message. It takes great skill and a high level of consciousness to be able to convey effectively one’s thoughts or knowledge.

If I’m lacking in the skill or in the requisite self-awareness to detail what I know effectively, should it then be taken to mean I’m ignorant?

An attempt at measuring what I know by what I’m able to put down, without consideration of the fact that I may be unskilled in the ability to put things down, is disingenuous

New to LessWrong?

New Comment