Fresh episode of Bad Takes is out, about the theory that Joe Biden could and should smoothly step aside in 2024.

I don’t like to read about politics in my downtime and instead tend get interested in random non-political topics.

For a while I was very into popular works on linguistics, but because I have politics lurking in my brain even when I don’t really want to, I often found myself noting parallels between the history of language and modern political issues — in particular, the interesting difference between how languages evolve when (like most languages) they are spoken primarily in insular communities and how they evolve when they are spoken by large, diverse groups.

I’m in Jamaica right now, and the formal official language here is, of course, English. The country’s major newspapers are written in English, government officials give their formal speeches in English, and English is the language of instruction in formal education. But in informal settings, Jamaicans — especially lower-class, rural, and older Jamaicans — actually don’t speak English, they speak an English-derived creole called Patois.

Patois is sometimes mistakenly characterized as ungrammatical English or slang or just a thick accent. But while it’s certainly true that bilingual Patois/English speakers may incorporate Patois into their English as slang or make grammatical errors because English isn’t their first language, Patois is a different (though related) language that isn’t fully comprehensible by monolingual English speakers. In the clip below, the newscaster and the port chief operating officer are both speaking Standard English with a Jamaican accent, but the food vendor the newscaster interviews in the middle of the segment is incomprehensible to me — but the intended audience of a Jamaican TV news program is expected to be able to understand both languages.

What’s interesting, though, is that the linguistic relationship between English and Patois isn’t the same as the relationship between, for example, Spanish and Portuguese or Swedish and Danish. In those cases, the languages are like siblings with a close linguistic parent. Patois, though, is a creole language, a language that arises from the collision of multiple languages, a situation that was especially common in the era of the transatlantic slave trade. And creole languages are interesting because they teach us something general about language contact situations and the difficulties that arise from the reality that adult humans are bad at learning new languages.

New languages of the New World

All around the world and throughout history, adults who don’t speak mutually intelligible languages and need to communicate tend to develop a pidgin. A pidgin normally has a very limited vocabulary because it meets the specialized needs of the contact situation, often trade and some basic elements of war and peace. It’s not a language that people use at home or for religious or literary or scientific purposes or in formal settings. And because the range of ideas speakers need to express is not especially broad, pidgins don’t generally have complex grammar. The goal is essentially a shared roster of nouns and verbs that people on both sides of the language divide can agree on, aided by a lot of pointing at things.

But when slavers brought Africans to plantations in the New World, the language situation was more complicated than a typical pidgin.

Enslaved children were brought up in complex linguistic circumstances in which the adults around them were communicating with each other largely in pidgin — the slavers did not speak the languages of the enslaved, and the enslaved people themselves came from diverse communities. And of course the enslaved children weren’t sent to school or formally educated in any particular language. But children, unlike adults, are really good at learning languages. And so those who grew up in communities where a lot of pidgin was spoken and possessed native-level fluency in the pidgin naturally started to flesh the pidgin out, coining or borrowing words for a wider range of things and adding grammatical features to generate a full roster of tenses and otherwise broaden the range of expression.

And the results were creole languages. A creole normally has a vocabulary that is largely based on one language — the lexifier — which in the New World is typically the language of the local colonial power. So Haitian Creole has a French-based vocabulary, Chavacano has a Spanish-based vocabulary, and Patois has an English-based vocabulary. But even though the vocabulary is the most superficially recognizable aspect of a language, these creole tongues end up being more different from the lexifier languages than it appears on the surface because they have different grammar.

In English, we create a present progressive tense (“I am coming”) instead of a simple present (“I come.”)

In Jamaican they use “mi” for “I” and “kom” for “come,” both of which are recognizably English-based. But the present progressive isn’t “mi koming,” it’s “mi de kom.” Patois speakers use either “de” or “a” (“Mi a go a maakit”) to make a present progressive. Negation works on a different principle:

• Mi no nuo we fi se1, I don’t know what to say

• Mi no nuo we fi du, I don’t know what to do

• Mi no nuo we im de, I don’t where he is

• Mi no nuo we im niem, I don’t know what his name is

• Mi no nuo we im a go, I don’t where he is going

And the different present progressive intersects with the different negation to give us:

• Mi naa du notn, I am not doing anything

• Di pikni-dem naa du dem lesn, The children are not doing their lessons

In English, we mark past tense by conjugating the verb: “I cooked the callaloo” rather than “I cook the callaloo.” We also have a lot of verbs that are irregular and don’t use this “ed” convention (eat/ate or sing/sang). In Jamaican, you don’t do those conjugations: the present is “mi nyam di callaloo,” and the past is “mi ben nyam di callaloo.” Most of the Jamaican tenses have this quality, relying on helper words rather than conjugations. Patois speakers use “wehn” as a helper to indicate continuous action in the past, and “a-go” or “de-go” indicates action in the future.

Every language is different, but this seems to be a common feature of creoles. Adults find it difficult to learn conjugations, so in the pidgin phase, they’re eliminated. Then as subsequent generations build up a complete language, they use helper phrases to do the work of conjugations.

English as a long-lived creole

I think the language situation in Jamaica and other Caribbean islands is fascinating and I could go on, but I’d probably only make mistakes. And part of why I find learning about creoles so interesting is that English itself shows some signs of being a kind of long-ago creole.

Old English, for example, had a system of grammatical gender that has since faded away. Because this seems to have started in the regions of England that were conquered by Vikings at around the time they were doing the conquering, it’s widely believed that the loss of gender is a simplification due to language contact. Old Norse and Old English are pretty similar, so Norse-speaking invaders could more or less communicate with their English-speaking wives. But the inflection systems were different, adults struggle to learn new grammar, and the easiest thing to do was just ignore it. Kids raised in mixed households or mixed communities learned to speak without grammatical gender, and this was part of a more general loss of inflections in the transition to Middle English.

Linguists broadly characterize languages as either analytic or synthetic, which initially confused the hell out of me because I majored in philosophy and was familiar with a very important and totally different analytic-synthetic distinction. But the basic idea is that a highly synthetic language has a complicated system of inflections to denote verb tenses, noun cases, and other things you may know about if you’ve ever studied Latin or Russian. A highly analytic language instead relies on the order of the words and lots of little helper words to do this work. Creoles are examples of very analytic languages.

English is also a pretty analytic language. You can see both the helper words and the conjugation in the English past tense, but as you’ll probably recall from French2 or Spanish class, the English conjugation system is quite simple in comparison.

English’s analytic nature seems to stem from the language’s turbulent history. England never suffered the exact situation of the slave plantation economy. But our language reflects the collision of different Anglo-Saxon dialects, contact with the conquered Celtic-speaking peoples, conquest by Norse speakers, a second conquest by speakers of Norman French, and then a lot of conquering around the world.

All this language contact has bequeathed to English a very large vocabulary and a very weird spelling system. For those of us who were otherwise good at school but struggled to spell, it’s natural to think of English as an unusually difficult language to learn. It’s kind of ridiculous that before the advent of ubiquitous spellchecking, I consistently made more spelling errors in my native language than in my otherwise-not-good French.

But historically, most human communication is oral, and now thanks to the internet we have a lot of writing that largely replicates the features of oral communication. English is actually grammatically fairly simple, with a lot of analyticity. English also shows a clear trajectory toward becoming simpler over time, which you can see in the histories of irregular verbs — the past tense of help used to be “holp,” but we now say “helped.” Quantitative analysis shows that rarely used English verbs regularize faster than common ones. There’s something a bit unintuitive about this, but it has to do with how people learn languages.

The push and pull of linguistic innovation

The thing about frequently used irregularities is that precisely because they are so frequent, even a language learner who’s really struggling is beaten over the head with them. The fact that the past tense of “go” is “went” and not “goed” is annoying, but because people talk about "going" all the time, everyone eventually figures it out.3 But for a rarely used verb, you just don't hear the irregular form very much. If you ever find yourself needing to say it, you’ll just assume it’s regular and say it wrong. If you do this while you’re a little kid and you have educated parents who correct you, you’ll learn the irregular form. If you’re bookish and read a lot, you’ll get lots of exposure to rare words and you’ll learn the irregular form. But if you’re an English-language learner or growing up in an immigrant household or just don’t read very much, you’ll probably say it wrong.