Old and New Tongues: Constructed Languages and The Wheel of Time

Ninte calichniye no domashita, Agelmar Dai Shan,” Moiraine replied formally, but with a note in her voice that said they were old friends. “Your welcome warms me, Lord Agelmar.”

Kodome calichniye ga ni Aes Sedai hei. Here is always a welcome for Aes Sedai.” He turned to Loial. “You are far from the stedding, Ogier, but you honor Fal Dara. Always glory to the Builders. Kiserai ti Wansho hei.

With Tor.com’s new “Reading the Wheel of Time” series working its way through the Eye of the World, it seems like now would be a good time for a refresher on how the Old Tongue works in Randland. If you haven’t read The Wheel of Time, there might be spoilers below. Go read the books now, maybe! I’ll still be here in a year. (For clarity’s sake: There’s a weak spoiler for book nine, a strong spoiler from seven, and definite spoilers for the first three books).

A refresher won’t take long, as the Old Tongue is surprisingly simple. Here’s how Robert Jordan describes it: “The actual words are based on many words. I have used Turkish, Arabic, Russian, Chinese, Japanese, and for a hint of the familiar, I used a little Gaelic, too. Because fantasy languages always have Gaelic in them. That’s just the way it goes. But I deliberately made the grammar and structure complicated.” Want more clarity on the grammar? Ok. “[The] grammar and syntax are a blending of English, German, and Chinese with some influence from a set of African languages read about long ago, and all but the oddities of structure long since forgotten.”

It is around this point that it should be clear I was pulling your leg: The Old Tongue is not simple. It’s a mishmash of a dozen different languages, thrown together into a Trolloc’s cauldron and stirred with a lot of artistic license. For the majority of fantasy readers, this is OK. Throw in some more apostrophes. Add a few z’s, maybe some capital letters in the middle of the word (HIja’, I’m looking at you, Klingons). We’ll muddle through.

But there’s a small percentage of us who want more: Those of us who try and learn Quenyan, those who translate Facebook into our favorite conlang, and those of us who are curious why Ninte calichniye no domashita, Agelmar Dai Shan translates as Japanese for “I am glad I came up with you” on Google Translate.

What can we make of the Old Tongue? Not much. We can say it sounds pretty, and we can learn by rote some of the grammar rules and intricacies that Robert Jordan invented for it. Without a large lexicon, we won’t be able to speak it on its own. There is a large dictionary in The Wheel of Time Companion which you could use to bootstrap your efforts, but there’s still only around a thousand words, and a lot of them technical. (Mashadar isn’t very useful when you’re trying to get your brother to pass the butter at dinner.) Mostly, we can use our knowledge to make the world seem more real, which was, I believe, the point of creating it in the first place.

On the other hand, if we start pedantically splitting hairs and looking at how the Old Tongue is used in the books, we can also use it to make the world seem less real.

There’s a couple of ways of doing this. The first is to focus exclusively on single words or phrases, and exclaim that they make no sense according to what we know about language. For instance, as others have pointed outTia mi aven Moridin isainde vadin (“The grave is no bar to my call”) doesn’t follow the same word order as the other phrases in the Old Tongue that Jordan used earlier. This is probably because Jordan simply hadn’t gotten around to thinking about Old Tongue syntax when he wrote the first book. Jordan loved languages—he purportedly had dozens of dictionaries and grammars in his study. But he was first and foremost a writer of fiction, and sometimes other matters take precedence over figuring out ergative structure in subordinate clauses.

Of course, we could also point out that this scene didn’t happen this way, either. And where is the light source coming from?

Another example, within a single word, is from Towers of Midnight, where Faile mentions one of her ancestors: “Nikiol Dianatkhah was a drunkard, despite being known as one of our greatest kings.” This is decidedly weird. I couldn’t find any other character or name in all of the books with a <kh> phoneme in their name, and that’s not for lack of trying. They appear in other fantasy languages—who could forget the keen edge of a Dothraki arakh, for instance—but we never see it in the Old Tongue. But here it is in a name. This suggests that the name was either a result of subtle language change in Saldaea, or it was misspelled by the publisher (sorry Tor), or Sanderson couldn’t read Jordan’s handwriting, or Sanderson made it up (which would also be OK, I think, as he did a fantastic job with the series), or it was a one-off and no one could spell Niki’s name throughout his life—or it was simply an example of poor language planning on the part of the author. I’m much more inclined to think it is that last one.

After all, no one is assuming that this series is attempting to precisely reflect reality—we all know it is a work of fiction, written by an author who may be focusing on different aspects as demanded by the story. If no one batted an eye at Tolkien for not having any clear currency in his world, then who are we to judge Robert Jordan, a veteran and a pipe collector, who allows us to see hills in terms of cavalry attacks and who teaches us that pipes with amber bits exist? After all, Jordan is on record as saying that he translated the language for us.

Wait, what?

Yes. All of the time we thought that Rand was speaking English, he was actually speaking some sort of New Tongue. Robert Jordan wasn’t just hearing the muse when he wrote, he was actively translating what she said to him.

Again, who am I to judge? A persnickety, entitled, and small-minded linguist,  that’s who. Because frankly, I don’t think that that excuse makes any sense. Let’s go into detail on the reasons why, focusing on regional differences in Randland, how phonology and orthography normally work, and why Robert Jordan’s translation excuse doesn’t cut it.

First, let’s briefly talk about how names are useful when trying to understand languages. Onomastics is the study of the origin and use of proper names. By looking at how people and places are named, you can get a pretty good idea of what the language looked like when spoken by those people or in that area. Normally, this is pretty clear: Paris, Lyon, Marseilles all sound French, which makes sense, because they are French cities. Boston, New Hampshire, and Manchester all sound English, largely because New England was settled by the English; similarly, Connecticut, Nantucket, and Massachusetts are harder for English speakers to pronounce because they aren’t English words at all, they’re Wompanoag.

Take a closer look at those three names: Connecticut, Nantucket, and Massachusetts. The words look and sound different than the other examples. Even from these three examples, you can tell that there’s too many t’s and k’s and n’s for English. This is because the phonology (or sounds) of Wompanoag is different. If we had more text, we could learn more about the language. For now, what we have is a theory that the construction of words (called morphology) is different. Here’s an example of written Wompanoag: Nooshum keskqut quttianatamanack hoowesaouk. Given our few examples, that is right in line with what we’d expect! (For those of you who are curious, that’s the punch line to my joke, “How do you say the Lord’s Prayer in Massachusetts?” which no one, to date, has ever found funny).

So, let’s get take this back to Randland and the Old Tongue. The <kh> example I mentioned above is a good one for understanding what I mean here—the phoneme (bit of sound) ought to tell us about the language that it came from. But it doesn’t, in this case. You’ll most likely not find another word that uses this combination.

Unfortunately, this isn’t exactly an isolated example. My least favourite word in the entire series is cuendillar. Did you know that it is pronounced with a Spanish /j/ sound? Why? Illian isn’t pronounced that way—just cuendillar. And there aren’t any Spaniards in Randland, last I checked. Or take the Niamh Passes mentioned briefly in Fires of Heaven—did you know that is probably pronounced “nee-v”? It’s a Gaelic word for a princess (as I wrote about, here). Jordan had to throw in some Gaelic, but you can’t just throw in a word from a language without bringing along a ton of phonological, orthographical, and (in some cases) syntactic baggage. For instance, domashita sounds exactly like a Japanese verb form, –mashita, which is the past tense, like in 分かりました wakarimashita, “I understood” (it also means “homeowners” in Bulgarian according to Google Translate, but whatever).

The problem is much more systematic, and it’s why I wanted us to focus on names: if everyone all speaks the same language, why are their names so different?

Moiraine Damodred would never be mistaken for an Andoran. Cairhienin all have names like Talmanes Delovinde, Barmanes Nolaisen, or Colavaere Saighan. Andorans from the Two Rivers all have names like Jac al’Seen, Jaim Dawry, or Ren Chandin. With names like Brandelwyn al’Vere, I wouldn’t be surprised to find a Bill Ferny, either. Shienarans have names like Easar Togita, Blaeric Negina, Joao, Qi, or Ragan. There are some people with names that could fit in multiple cultures—where is Takima Deraighdin from, for instance?—but on the whole, it’s pretty easy to tell where someone in The Wheel of Time comes from by looking at their name.

You can also tell from their accent, of course. Seanchan people talk slowly, Cairhienin lilt, and Two Rivers people talk noticeably differently than other Andorans. Of course, we don’t hear the accents when reading—we have to have the phrase “and he said this in Tairen accent” included in the text. Sanderson at one point mentions a “rural Illianer accent,” which I wish I could have heard (and which makes you wonder why the accent was less marked in the city than outside of it). Occasionally we have farmers sounding like they came from Appalachia, which says more about the author’s biases than anything else.

We know that the Old Tongue had accents, too. Birgitte tells Mat that “One sentence you’re an Eharoni High Prince and the next a First Lord of Manetheren, accent and idiom perfect.” We also know that there are differences in syntax. Murandians use frozen constructions like “Sure and that’s none of your ways, is it?”, Taraboners love topicalization, and Illianers do be hard to take seriously.

However, none of these differences should explain the naming problem. We’d expect everyone’s names to get closer to modern English, not to get more Cairhienin or Tairen. We do have an inkling that this is happening: The etymology of Far Madding is interesting—we’re told directly that it changed from Aren Deshar to Aren Mador to Far Madding. We know that Cairhien’s proper name is Al’cair’rahienallen, Hill of the Golden Dawn. But these are isolated examples, and on the whole, I would argue that, overall, the evolution toward modern English isn’t happening in Randland names.

Robert Jordan did like talking about the Old Tongue during interviews. He consistently says that everyone spoke it in the Age of Legends, and, after the Breaking, there was never enough time for populations to diverge enough to speak their own language. Every thousand years, a disaster would happen that brought everyone together again, and stopped any regionalisation: the Trolloc Wars, and Artur Hawkwing. (And, as Lan muses in New Spring: “And now, close enough to a thousand years after Hawkwing’s empire died, the Aiel came, burning and killing. It had to be a Pattern.” Which raises a whole ton of eschatological questions, linguistics aside.) He also says that conquests did this for the Seanchan, too. In Winter’s Heart, we read that “History fascinated Egeanin, and she had even read translations from the myriad of languages that had existed before the Consolidation began.”

But languages don’t die just because someone conquers your country. That’s often when you start hiring translators. Yes, one language may take over for economic reasons, over centuries—or, in cases where the majority of the population dies off, the native languages may also become extinct (like Wompanoag, which currently has around five native speakers). But there should be remnant or substrate languages all over the place—particularly in the hinterlands, like the small villages south of Shienar where Hardan used to be. Or in the Two Rivers. Or with the Aiel.

Jordan gave another excuse for the lack of language change—printing presses. They’ve have been around since the Breaking, and that froze a lot of the language differences. But I don’t think this is a good enough explanation, and it doesn’t match how languages work, to my knowledge. Literacy is fairly low in Randland. There are only a few books, and there don’t seem to be many schools for farmers. You can’t freeze a language using books effectively, especially if not everyone is reading. You’d also have to have the same presses and books operating in Arad Doman as in Mayene, and, besides The Travels of Jain Farstrider, we don’t have enough information to know if this is true or not. Without public libraries, I suspect it isn’t. Jordan points to Shakespeare a lot—but how many of us can fully grasp every line of Coriolanus without a dictionary handy? Or Beowulf? Sir Gawain and the Green Knight was written in a variant of English, but if it wasn’t for scholars like Tolkien—a translator—it might not be in wide circulation now.

The Old Tongue is consistently described as a more complex language than what is currently spoken in Randland. That’s actually a bold claim, in Linguistics; only in the past few decades have scientists been able to say definitively that language complexity differs and changes across languages. What I would argue is that the world is not homogenous just because of intervals of high-contact every millennium. According to the linguist Peter Trudgill, languages can become more complex in a few circumstances:

The major complexity-producing social factors include: small size, dense social networks, large amounts of shared information, high stability, and low contact. … [change requires] considerable periods of time in order to develop undisturbed and go to completion. My sociolinguistic-typological point of view is that in large, high-contact, unstable communities with loose social networks, such lengthy periods are less likely to be available.

This makes sense to me, and it’s the argument that Jordan makes—but I don’t see that there is enough contact not to warrant different languages, and I would argue that 1000 years is far too small a timescale. Look at the Germanic branch of languages, alone. A thousand years ago, Anglo-Saxon would have been pretty similar to Old Norse, Dutch, and High German. But I speak a Germanic language, and after spending two years in Germany, I still wouldn’t be able to phrase “Blood and bloody ashes, there are Trollocs running over our fields” in German without some significant amount of effort.

So, there should be more languages, but there aren’t. There are accent and slight grammar differences, but they don’t explain why names are different. Jordan’s arguments for monolingualism are weak—conquests aren’t enough to force monolingualism on everyone, and literacy doesn’t stop the presses of language change, either. Finally, modern English in Randland doesn’t look anything like the Old Tongue because it’s been “translated” for us. We’re sadly left with a world that makes less sense than it did when we started admiring the realism of the Old Tongue.

But—was the Old Tongue translated, too? That’s what Tolkien did. Almost everyone in his books spoke Westron, where we read English. The Rohirrim had their own language, which Tolkien translated into Old English for us to show how it was related to Westron. Robert Jordan never says that he translated the Old Tongue, but there are some subtle hints. For instance, zemai, t’mat, and oosgai look a lot like maize, tomato, and whiskey. The nadra-bush mentioned by Sanderson in Towers of Midnight looks an awful lot like naddre, which is the Old English word for adder. These words are too familiar to their modern descendants to be anything other than progenitors. But could Carai al Caldazar ever turn into “For the Red Eagle!”? I doubt it.

It’s at this point that even the most avid linguaphiles among us are forced to face the truth: Robert Jordan was certainly a keen amateur linguist, but he wasn’t an academic—he was a writer first, and a linguist second. The handwaving excuse of “I translated it” wasn’t meant to be serious, it was meant to stop the reader from digging in and finding inconsistencies. Because, ultimately, they don’t matter. What matters is that we see a world filled with history and wonder, with words of power and words of lore, echoing through the ages. Instead of raising a hand like an arrogant high-schooler, we’re supposed to take up arms at Mat’s battlecries (I know I’ve pumped my fists in glory a few times in his chapters).

It sure is fun to try and find the cracks though, isn’t it?

Richard Littauer is a fantasy author, poet, linguist, and noun phrase. He tweetsreviews books, cooks over philosophy, and runs Word Hoard Press for original work in dead languages, among other things.

32 Comments

Subscribe to this thread

Post a Comment

All comments must meet the community standards outlined in Tor.com's Moderation Policy or be subject to moderation. Thank you for keeping the discussion, and our community, civil and respectful.

Hate the CAPTCHA? Tor.com members can edit comments, skip the preview, and never have to prove they're not robots. Join now!