How Does Language Evolve?
Natural selection is often at play, but new research suggests that language evolution can be a lot more random than we think.
Be honest: does it bother you that the words "haterade" and "fitspo" were recently canonized by Oxford Dictionaries?
Be honest again: are you such a traditionalist that you still use the Old English "Gōdne mergen" when you see people in the morning?
Older generations love to bemoan the youth and their slang, but then again, they’re not exactly speaking Chaucer’s English, either. Change is normal, natural and healthy, and even if the prescriptivists and descriptivists love to argue over what constitutes legitimate evolution, the fact remains: language refuses to stay in place.
There’s also a good deal of debate in the linguistics community over how languages evolve, however.
Do You Catch My Drift?
In some cases, languages change because the needs of the speakers change, particularly as new technology and ways of life require new ways to talk about them. In other cases, language changes simply because no two people are alike, and we are all, linguistically, products of the environment and people we interact with. Languages change when speakers come into contact with new populations (think colonialism and trade), and they change when social groups adopt their own distinctive norms.
Though it’s generally accepted that languages evolve according to natural selection (a sort of Darwinism for words), University of Pennsylvania biology professor Joshua Plotkin recently applied the principles of evolutionary biology to the study of linguistics, and he made a good case for why language, like genes, can change at random. Just like gene mutation, languages transform as they’re passed down from one generation or geographical region to the next: a process known as linguistic drift. The process of "replicating" language over time is imperfect, and it’s shaped by input from parents, siblings, peers and the larger community.
Plotkin and his colleagues performed statistical tests to weed out cases where natural selection is at play (versus drift). When these tests were applied to online repositories (like the Corpus of Historical American English), they found cases where natural selection favored regularized verb tenses ("smelled" instead of "smelt," for example). They also found cases where natural selection pointed to irregularity: "dove" instead of "dived." They posited that "dove" replaced "dive" around the same time that cars were becoming popular, which means we were using words like "drive" and "drove," given that we might have a tendency to choose words that sound alike. This is not unlike the observation noted by Grammar Girl in a recent podcast episode, which is that sound changes eventually affect every word that carries the sound: "pater" to "father," "pire" to "fire," "ped/pod" to "foot."
The Random Factor
Plotkin also analyzed three major changes in the English language: the "ed" ending in past-tense verbs, the adoption of the word "do" as an auxiliary verb ("you say not" became "you do not say"), the evolution of negation (from the Old English "ic ne secge" to the Middle English "I ne seye not" to the Early Middle English "I say not"). He determined that selection was likely why negative sentence structures changed, but the other two were likely the results of random linguistic drift, which, examined up close, is actually a series of fits and starts that looks like gradual change from a distance.
Of course, Plotkin wasn’t the first researcher to tackle the mechanisms of linguistic drift. Researchers William Hamilton, Jure Leskovec and Dan Jurafsky published a study that examined some of the major tendencies of semantic change. They pointed out that nouns are more likely to change due to irregular cultural shifts (like new technology) while verbs more often change due to regular semantic drift, largely because they are more semantically mutable than nouns.
Should this emphasis on randomness pick up steam in the field of linguistics, we’re likely to see a lot more appreciation for the mysteriousness of organic evolution — and perhaps fewer attempts to explain away processes that can sometimes defy logic.
Think about it this way: we can identify the grammar of doge speak, but we probably can’t totally explain how we actually got to this point. Much wow!