It's funny the rabbit holes fiction writers get dragged down sometimes.
This latest one occurred because of two things that happened kind of at the same time. First, in my work-in-progress, a fall-of-civilization novel called In the Midst of Lions that in the current national and global situation is seeming to cut a little close to the bone, one of the characters is a linguist who saw what was coming and wrote a conlang -- a constructed (invented) language -- so he could communicate with people he trusted without it being decipherable by enemies.
So of course, to make it authentic, I've had to write the language, following in the footsteps of the Star Trek folks with Klingon and J. R. R. Tolkien with Quenya and Sindarin (two of the languages of the Elves). My MA is in linguistics (yes, I know, I spent my career teaching biology; it's a long story) so I know a good bit about language structure, and I wanted to make the language different enough from the familiar Indo-European languages to seem (1) an authentic language, not just a word-for-word substitution, and (2) something a smart linguist would think up. Unfortunately, my specialty is Indo-European languages, specifically Scandinavian languages. (My wife gives me grief about having studied Old Norse. My response is that if the Vikings ever take over the shipping industry, I'm gonna have the last laugh.)
The second thing, though, was a nice kick in the rear that came from a question on Quora that asked, "What is the hardest language to learn to speak fluently?" By "hardest" most people assumed "for speakers of English," which went right to what I was interested in -- finding out what would seem odd/counterintuitive (and therefore difficult) for English speakers.
Well, this is what led me directly into the research labyrinth, literally for hours.
One respondent answered that the hardest ones would be the Northwest Caucasian languages of Georgia, Azerbaijan, and Armenia -- a group made up of Abaza, Abkhaz, Adyghe, Kabardian, and Ubykh -- the last-mentioned of which became extinct in 1992 when the last native speaker died of old age. These languages form an isolate family, related to each other but of uncertain (but undoubtedly distant) relationship to other languages.
So naturally, I had to find out what's weird about them. Here's what I learned.
Let's start out with the fact that they only have two vowels, but as many as 84 consonants depending on exactly how finely you want to break them up based on the articulation. They use SOV (subject-object-verb) word order, plopping the verb at the end of the sentence, but that's hardly unique; Latin does that, giving rise to the old quip that by the time a Roman got to the verb in his sentence, his listeners had forgotten who all he was talking about.
But in the parlance of the infomercial, "Wait, there's more!" The Northwest Caucasian languages use agglutination -- gluing together various bits and pieces to make a more specific word -- but only for verbs. In these languages, a verb is actually a cluster of parts called morphemes that tell you not only what the core verb is, but the place, time, manner of action, whether it's positive or negative, and even the subject's and object's person.
Then, there's the fact that they're ergative-absolutive languages. When I hit this, I thought, "Okay, I used to know what this meant," and had to look it up. It has to do with how the subject and object of a sentence are used. In English (a nominative-accusative language), the subject has the same form regardless of what kind of verb follows it; likewise, the object always is the same. So the subject of an intransitive verb like "to walk" is the same as the subject for a transitive verb like "to watch." (We'd say, "she walked" and "she watched [someone or something];" in both cases, you use the form "she.") The object form of "he" is always "him," regardless of any other considerations in the sentence.
Not so in the Northwest Caucasian languages, and other ergative-absolutive languages, such as Tibetan, Basque, and Mayan. In these languages, the subject of an intransitive verb and the object of a transitive one have the same form; the subject of a transitive verb is the one with the different form. (If English was an ergative-absolutive language, we might say "He watched her," but then it'd be "her walked.")
So there are lots of things that seem normal, obvious even, which in fact are simply arbitrary rules that we've learned are universal to English, but which are hardly universal to other languages. It always puts me in mind of the Sapir-Whorf hypothesis, which is that the language you speak shapes your cognitive processes. In other words, that speakers of languages differently structured from English literally perceive the world a different way because the form of the languages force different conceptualizations of what they see.
I've gone on long enough about all this, and I haven't even scratched the surface. There are tonal languages like Thai, where the pitch and pitch change of a syllable alter its meaning. There are languages like Finnish and Japanese where vowel length -- literally, how long you say the vowel for -- changes the meaning of the word it's in. There are inflected languages like Greek, where the ending of a word tells you how it's being used in the sentence (e.g., in the phrases "the cat walked," "she pet the cat," "it's the cat's bowl," "give the food to the cat," and "the dog is with the cat," the word "cat" would in each case have a different suffix).
So I have some work to do to make my conlang something that would be believable to a linguist. Or, in the context of the story, something an actual linguist would invent. Of course, being that it's only one small piece of the story, in the end I'll probably use something like a dozen phrases total from the language, so it'll be a lot of work with very little useful result.
But hey, if J. R. R. Tolkien did it, who am I to criticize?