Again and again people trying to do anything complicated for a long period of time are beset by the same issue: What do our words refer to, since they clearly don’t refer to truth conditions of the person who originally wrote them? The answer is that “refer” is a misnomer, and it’s an especially confusing one because it almost works.
This is perhaps not hard to see at the frayed: When I say something is “beautiful” I can hardly be talking about a single aesthetic standard that everyone abides by. What am I referring to? One could suggest I am referring to the feeling of experiencing beauty. This is a very worrying answer to me. How do I know what it means to experience beauty, especially in a communicable enough way that I can assume your “experience of beauty” (as defined by what you feel when you say something is beautiful) is similar enough to my own that we can meaningfully connect?
In theory, this is impossible. In reality, people tend to call many of the same things beautiful, and only disagree about certain strands of beauty. This is perceived as taste, though it could be equally well interpreted as a result of the imprecision of our coordinating mechanism. If we believed people were rational fixed agents we could leave well enough alone here. But we don’t believe that. We believe that people are a kind of mammal, more specifically a primate, more specifically they are brought up in a culture, and for the vast majority of them they are brought up by their biological parents under specific, community-accepted child-rearing practices. Analyzing the implications of this is out of scope, but saying the following is not: Much of what is perceived as beauty comes from a process of acculturation. This makes it very unclear whether the feelings people feel about beauty are all the same and “taste” differs or whether the experience of beauty between people is fundamentally different, but the rituals around beauty (“Looking sharp today!”) construct our notion of the concept.
Words come to mean the things that they can get done. They definitely refer to things, but those references are never exact. They are hammered into place by the need for exactitude, and the need is often very, very low.
If your parents are uncomfortable with your current dating partner, do they feel comfortable complimenting how beautiful or handsome they are? If so, they will use this as a tool to relieve perceived awkwardness. The common view would be that this “dilutes” or “corrupts” the meaning. This is a misunderstanding of how meaning functions.
The meaning of words is not a reference to a thing. Words are a way of establishing a referent, by slicing away or building-up possible referents in the given context. Words, however, can only be defined by other words or by experiences in which words did or did not get you what you wanted. When you can read a wedding invitation and you show up at the right time and the right place, the words worked. When my China-born friend asks for a napkin but wanted a tissue (the same word in most common Mandarin dialects), the words did not work, and one updates how they can be used.
It is not all transactional. When I read a novel and some of it is tough going but some of it really speaks to me, I can come back to the tough parts and leverage my success with the other parts and my understanding of narrative structure to try to investigate what the tough parts mean. There are no guarantees in this system, and the possibility of practically-private language is easy to confirm. Whether such a language “couldn’t” be understood is a misunderstanding of what language is, and Wittgenstein was right.
Words expand to be useful, not because they do anything, but because that’s how we use them.
One example of this is written laws, e.g. the Living Constitution in the U.S. which expands and contracts with what certain people feel the need to see in it in order to adjust rules to the reality we inhabit, a reality so alien that its writers and ratifiers could hardly have planned for it. Of course, the constitution was—in theory—supposed to be torn-down and rewritten every once in a while, but it turns out politics is a difficult problem, highly prone to emergent effects. Instead of rewriting, we use reinterpretation, which is a very powerful tool.
An easier to explicate example is that of “Computer”. Computer was originally an occupation—it was a person who did calculations. It’s not hard to see how this extended to mechanical implements, but on the way the occupation sense shriveled-up and died. In 2020 it would be rude to call someone a computer. In fact, people often talk about “Human Computers”—people with extraordinary mathematical gifts for concrete calculations, so the human part needs to be added as a kind of contrast to what a computer is “supposed” to be in the mind of the listener.
This is not very surprising stuff. But as usual, what’s easy to see in practice is hard to explain in theory. When a literal banana grows arms and starts talking about words failing we should be skeptical, but curious. Bananas have a unique perspective, and might have deeper insight into the underlying mechanisms and assumptions with which we operate. The above banana has presented the argument that words are failing because they are understood through a discourse but manipulated and trusted as if they are a definite reference, e.g.
I could hardly disagree that these words are used in deceptive ways, even by people with no means to deceive. Everyone has met someone who cites studies they haven’t read and don’t understand. For many people that someone is me.
The words have failed us, but it is the frame we put them in that caused the inevitable betrayal. I blame Lady Macbeth for planting the seed of betrayal in Macbeth’s mind, even if Macbeth is a killer.
The achievements of physics have driven other fields insane—they are having trouble thinking out loud without being tempted to claim that their words have absolute referents. This impulse is not new (e.g. Freud), but the facade of capital-S Science makes it possible to believe in a short-term pathway to objectivity. Words haven’t failed, our intuition about what they’re referring to is being coopted by a new school of what things can mean at global scale. Bananas have known this for a while, coining the term global knowledge game. In a globalized world, to describe things across culture we need shared standards, we need to reach for objectivity. This is necessary and right—but it is easy to fake and still achieve wealth and fame. That is the perverse incentivize of self-policing knowledge institutions.
The solution is not trying to fix words to the things they refer to now or upon their invention. Words are what they’re used to be and as their function shifts their meaning shifts along with this. The finer point is that words often shift to occupy a space that would have been difficult or impossible to describe merely with words or examples. The idea of internet browsing is hard to describe. It is not merely the act of navigation. It is not merely the sum, average, or median of all internet activity. It is a notion, that takes its cues from people browsing spaces and our attempts to describe such behavior in animals. We need words to drift in order to explain ourselves—but we need more skepticism about what it means to explain one’s self to someone with very little shared context. Science, as it is today, is disincentivized from ironing this out, as it will lower the status of high status individuals. But the internet is in a good position—if it can pump life into communities whose goal it is to be widely understood, instead of standing at the top of niche hierarchies. Citizen Science is good, but Citizen Discourse is better.
Getting there is a problem of incentives. That’s a problem for another post.