display | more...

"This is just a word I coined myself. Just as the etymology is a branch of linguistics that focuses on the history of words, their forms, and their changes of meaning over time, the meta-etymology is the "etymology of the 'beyond'", that is, the whole process of reasoning about a word that goes beyond the conlang itself, and that metaphorically travels throughout the natural languages. Put it simpler, the meta-etymology is the thoughts that led me to make a word the way it is now." - reddit user Askadia on the constructed languages subreddit, November 2017

Linguists spend a great deal of time tracking the history of words: how their spellings and pronunciations change over time, how they transmit between languages as loanwords and cognates, how their meanings evolve to become more specific or more general, more pleasant or more negative in connotation. The Oxford English Dictionary and the Online Etymology Dictionary both represent significant time and effort committed to tracking the earliest known uses of a given word, and to keeping track of a given word's "family tree" among multiple languages.

Linguists spend just as much time tracking the history of meanings. "Shirt" and "skirt" in English once were a single merged word in Old English; so were "eye" and "egg." In Irish Gaelic, however, these words do not share the same merged origins: "súl" is not connected in a "family tree" to "ubh," and "léine" is not related to "sciorta," even though we can see a clear relatedness between "sciorta" and "skirt." In some way, English once featured a single larger concept of a garment, which later split apart into "shirt" and "skirt," and over time we have widened the separation between these two concepts, so that they are fully separate classes of garment. A "shirt" may be synonymous with a "blouse," but a "blouse" is definitely never a "skirt." In ancient Ireland, there were two likely possibilities for how their words developed: either there was never a single larger garment-concept that divided, at all, and "léine" never coincided with "sciorta," or else "léine" was always the primary garment-concept, and "sciorta" was adapted into the language later, through contact with people referring to their garments as "skirts."

The field of semantics consults the ways meanings change, as a surface-level phenomenon that can be documented through human usage: people write down the words they use, and we can track at what point in time a different word came into use for a given concept.

What we effectively cannot track, however, is the deep-level psychological metamorphosis of meaning, that represents the underlying concepts that words represent in our communication. Languages give us a shorthand for externally depicting the categories within our minds, into which we place the ideas we encounter. "Dog" is a very complicated and specific concept that would be difficult to explain to someone who had never encountered any mammal at all other than human beings... yet I defy any well-socialized human to fail to classify a dog as a dog, to a degree that satisfactorily matches the definition used by other humans in the vicinity. Dogs come in many shapes, sizes, colours, and habitats. Some of them look much more similar to a sheep, at a distance, or a wolf, or a small fluffy sphere. There are animals which can make the same noises as dogs, yet certainly are not dogs. Generally speaking, humans class dogs as dogs, with nearly perfect consistency to one another. "Dog" is a simple utterance of sound which encapsulates the entire class of concepts being described, allowing any other human to share comprehension with great precision.

Other concepts are far less consistent in their classification, from one person to another. Whether you are a chef or a botanist may determine whether you class "tomato" as vegetable or fruit. When you say "fidelity," you may be referring to the precision and resolution of an audio clip, or you may instead mean commitment within a relationship... and in both these cases, the definitions supplied only abstract the matter further, because what does "commitment" even mean? Does it refer to monogamous exclusivity, or to the intention to persist in the relationship until either party's life has ended, or both of these, or neither? Abstract concepts frequently rely on still more abstract concepts to clarify the speaker's intentions, and as a result, instead of a single simple utterance mapping to a single specific but complex map of networked concepts... the language used becomes a long row of single bare lightbulbs defining the perimeter of a vast, dim warehouse at night. They map around the edges of what the speaker personally, precisely means,. Every speaker maps the meaning differently, and indirectness and redundancy are the only option available to allow separate minds to meet in the middle. Jacques Derrida addressed this gulf of concept-map-comparison, in his descriptions of deconstruction and différance.

Meta-etymology is how conlangers, the inventors of recreational, artistic, constructed languages, demonstrate the mapping of concepts that result in the selection of elements to create a word in their conlang. A conlang is not a nebulous entity with an undocumented origin and imprecise, sprawling primary data. It is a deliberate, procedural, methodical creation with finite vocabulary and unambiguous origins. A conlanger has a unique advantage on a linguist who is trying to map a natural language, or natlang: the solitary conlanger knows the entire concept map of every single word he creates, because he is the only person who uses his conlang at all, and the only person with any hand in building it. He enjoys a unique precision and directness that we native speakers of natural languages can never fully enjoy, except with the most concrete and universal concepts, like "dog." As such, meta-etymology is a subfield of linguistics which is virtually exclusive to conlangs, since only conlangs can supply definitive, invariant data.


Iron Noder 2017, 25/30