Understanding is One Hundred Percent Misunderstanding
What does it mean for something to have “gotten across”?
It means that the same thing existing in my head has now stood up in the other person’s head. But is there any way to verify that? The fact that the other person nodded correctly is merely the fact that the circumstances for nodding coincided.
Whether there was agreement cannot be verified
In 1960, Willard Quine presented a scenario in his book Word and Object.
A linguist is walking with a speaker of a language they know nothing about. A rabbit runs past them. The speaker says, “Gavagai.”
The linguist wants to write in their notebook: Gavagai = rabbit. But that doesn’t settle it.
What the word refers to might be a single animal called a rabbit. It might be a temporal cross-section of a rabbit at this exact moment. It might be a collection of rabbit parts that haven’t been separated yet. Or it might mean that “rabbit-ness” is manifesting there.
Regardless of which it is, the situation in which the speaker says “Gavagai” remains exactly the same. Because pointing at a rabbit is simultaneously pointing at a cross-section of a rabbit, and pointing at a part of a rabbit. You cannot distinguish between them even by pointing.
What Quine demonstrated is that there are multiple translations that can explain the other person’s behavior in its entirety, and there is no fact anywhere to determine which one is correct. Even if you spend more time with them, even if you ask more questions, the same thing happens.
There is simply no way to verify whether agreement was reached.
Communication was built without verifying agreement
Twelve years prior, in a completely different place, someone turned communication into mathematics.
In 1948, Claude Shannon published a paper titled “A Mathematical Theory of Communication.” Most of today’s communication stems from this paper.
Near the beginning, it says:
These semantic aspects of communication are irrelevant to the engineering problem.
He goes on to write that what is important is that the actual message is one selected from a set of possible messages.
He decided not to deal with what was meant. He dealt only with which one was selected.
From there came compression, error correction, and the calculation of channel capacity. Telephone networks and the internet are built upon this. The fact that these characters are reaching you now is a result of it.
For nearly eighty years, global communication has functioned. During that time, whether meaning agreed has never once been inspected. There is no mechanism for inspection built into it.
Communication was created and operated without ever verifying the agreement of meaning.
Changing what we measure: from agreement to reproducibility
These two perspectives may seem to face opposite directions, but they point to the same single spot.
The philosophical side said there is no means to verify agreement. The engineering side showed that everything could be built without verifying agreement. From neither side does a scenario emerge where meaning is transferred as-is.
What is actually happening is this: I send something. The other person receives it and reconstructs it using the materials they have within themselves. The resulting construction is made of the other person’s materials. Therefore, it is different from mine.
If we call that a misunderstanding, then understanding is one hundred percent misunderstanding. I don’t mean this in a negative sense. Simply that nothing exists but misunderstanding.
This is not an argument for giving up on communicating. It is an argument for replacing what we measure.
That is exactly what Shannon did. He declared that meaning would not be handled, but that didn’t mean he measured nothing. He rigorously measured whether the same thing appears on the other side when the same thing is sent. How much it breaks down due to noise, how much redundancy must be added to restore it, how much can pass through this channel. Error correction is a mechanism that guarantees whether something arrived without ever once looking at the meaning.
Agreement cannot be measured. Reproducibility can be measured.
Therefore, I stop asking if it “got across.” Instead, I look at whether the same thing happens again. When the same thing is placed, whether the other person changes or time passes, does an image of the same shape stand up? If it does, then even if it is a misunderstanding, it is reproducing. If it doesn’t, then the way I am assembling things on my end is different every time.
Agreement cannot be designed. Reproducibility can be designed.
I believe that refining how we communicate is that very task. It is not the work of improving the inside of the other person’s head, but the work of ensuring that the same input produces the same output.