The object-word problem

Consider a cat—the actual feline—and consider the word “cat”—the spelling C-A-T. How do you look at the object, the actual cat, and infer how the word “cat” is spelled? Similarly, how do you look at the word “cat” and derive the associated feline object?

Obviously you can’t, and it doesn’t make sense to try. We could use any word to refer to actual cats, like “shmerk” or “bligblorp”. And we could use the word “cat” to refer to any object, like actual dogs or actual potatoes. There is no logical association between cats and the word “cat”, just a historical process that happened to produce that outcome.

But suppose you think that there should be some logical connection between objects and the words that are used to refer to them, or that admitting that there isn’t any such connection is problematic for some reason. Like maybe you think that if you admit that we could use any word to refer to any object, then people will start using arbitrary words to refer to whatever they want, all communication will become impossible, and society will collapse. Then you might develop the object-word problem, the observation that there is no way to use the properties of an object to infer which word refers to it, nor the properties of a word to infer which object it refers to.

The object-spelling problem is the observation that it is apparently impossible to derive the word that refers to an object from the facts of that object and vice versa. It is analogous to the is-ought problem, wherein it is apparently impossible to derive an ought from an is and vice versa. The is-ought problem exists because oughts are preference orderings of ises (see also), so without specifying a preference order over a set of ises, there is no way to determine their ought. Furthermore, there are many ways to put a preference order on a set of ises, and there is no way to figure out which preference order you ought to use without putting a preference order over the set of preference orders, etc.

But while there may be no purely abstract method for picking one preference order over another, there is an empirical method: the actual processes of competition and cooperation that happen in the real world that determine what the ises are and what ought is ordered over them. The facts of the matter and the evaluative claims over those facts are both built by the economic process of collective intelligence. The distinction between is and ought is useful, but not a natural kind, despite what our first-person experiences seem to tell us.

Object-word relationships were determined by a historical process in which objects, words, and object-word relationships competed and cooperated with each other to produce the current set of outcomes. The word “sandwich” migrated its relationship over time from referring to a sandy market town to referring to a food item by way of a person’s title, which is vaguely like how an organism’s niche will change over the course of its evolutionary history. Objects can also compete for words, like two companies in a legal battle over a trademarked term. The equilibrium at any point in time also shapes future developments, leading to slang and new words for new concepts, and old concepts can be changed by applying new words to them, or even to new words to other concepts that the concept in question is connected to.

The processes of competition and cooperation among the parts of a collective, however we categorize their collective nature, will produce the evaluative processes that associate words with objects and oughts with ises without reference to some overarching evaluation rule because the ordering of the parts does the evaluating. It’s a historical empirical process, not an atemporal philosophical argument.