Arbitrary Language: The Signifier/Signified Relationship
An occasionally overlooked root tenant of Cognitive Linguistics is the relationship of signifier and signified outlined by Ferdinand de Saussure.
Roughly speaking, Saussure defines these terms as follows. The "signifier" means the word that represents an entity. Every word you're reading now is a signifier. And every signifier is related to the signified, or the "sound-image," of a concept. The letters d-o-g are together the signifier for "dog" and the sound-image is your concept of a dog as it exist when you silently say the word in your head.
It gets more interesting when Saussure argues that the relationship between the signifier and the signified is arbitrary. To take the earlier example, there is no real reason the the letters (and the phonemes they represent) d-o-g should really mean that fuzzy creature whose tail wags so hard its whole body shakes when you get home. The letters d-o-g refer to a dog in English by nothing more than convention.
In Norwegian dog is h-u-n-d. If Arabic is represented using the Latin alphabet, dog is a-l-k-a-l-b. Likewise, Chinese for dog is g-o-u. There is no reason beyond habitual use by a community for any signifier to be linked with any signified.
Carried to it's conclusion, Saussure's argument holds that humans have arbitrarily divvied up the world through linguistic convention. We decided what was dog, or chair, or house. We assigned signifiers to signifieds to enable communication, and not because any thing is inherently the thing we say it is. To put it another way, our furry friends exist but they only exist as "dogs" because a community of speakers decided to call them that.
Prototypes in Cognitive Linguistics
Cognitive Linguistics takes Saussure's realization of the arbitrary relationship between signifier and signified and builds on it with the idea of prototypes.
Prototype theory is deceptive simple. At it's heart it simply asks, how do you know what any given entity is? How, to return to our example, did you learn what a "dog" is? The answer is that you were repeatedly exposed to many instances of a dog.
But here's the wild card: no one else in the world was exposed to the same "dogs" at the same time. So your idea of dog is slightly different. You can rest this by asking anyone to describe what they see in their head when you say "dog." Some will say a Golden Retriever, while for others it will be the Corgi they grew up with. Both are types of dog, but each individual has a different prototype for the concept based on their own personal experience.
Having different prototypes for something as supposedly clear as "dog" means that there is no chance for perfect communication. I cannot say any word and expect anyone else to hold the exact same concept in their mind as a result. Of course, communication between speakers of a language is still possible since prototypes for any given concept are substantially similar. Still, they are never identical.
Nietzsche, Nihilism, and Cognitive Linguistics
Reading Nietzsche recently I was stuck by the similarity between the abstract relationship between signifier and signified, prototype theory, and Nietzsche's seemingly pervasive argument that there is no "truth"--that "knowing" something comes not from realizing a truth about the universe "but schematizing, so as to impose on to chaos as much regularity and form as suffices for our practical needs."
It seems that nihilism and current linguistics arrive on their own at a similar realization about the world and human perception: that we are the creators rather than discoverers of meaning.
Though I have a background in Cognitive Linguistics, I'm relatively new to Nietzsche. Any thoughts or comments on the argument and conclusions presented here would be welcome.
Select Sources:
Women, Fire, and Dangerous Things, George Lakoff
Linguistic Categorization, John Taylor
Kritische Studienausgabe, Frederick Nietzsche
Nietzsche, Nihilism, and the Philosophy of the Future, edited by Jeffery Metzger