Borrowing from natural languages

One of the hardest parts about creating a language has to be the vocabulary. At least, that’s always the hardest for me. Maybe you’re different, but I doubt it’s easy for anybody, unless you’re doing one of those “engineered” languages where an algorithm does all the work for you.

Anyway, since creating words is so difficult, and since we do have to have them to, well, make a language, it’s only natural that we look for shortcuts. One of those is the random word generator, as you know, and I’ve spoken on that subject before. Today, however, we’ll look at a different method: borrowing. Specifically, I’m talking about borrowing from an existing language, a real language.

Can it work?

Borrowing from natural languages is fairly straightforward, but it’s easy to go wrong. Obviously, if you just take a bunch of English words wholesale, then you’re not making a separate language. You’ll end up with something closer to a pidgin instead: English words stuffed into foreign grammar. And that’s probably not what you want.

So we need a better strategy, but which one you want to use depends on your goal. Which words you want to borrow will go a long way towards defining the “feel” of your conlang. If you’re taking a bunch of old Anglo-Saxon roots, that’s going to create something that looks much different from a language that only borrows modern technical terms like “internet” or “photovoltaic”.

Also, remember that languages don’t always borrow a whole linguistic paradigm. They’ll tend to take only a root (which might not be the actual root) and derive native terms from there. So even if you borrow “computer”, that’s no guarantee that you’ll be borrowing “computers”, “computing”, and “computation”, too. If you do, it’ll look less natural, because that doesn’t often happen in the real world. And you do want this to look realistic, don’t you?

Details

Clearly, the absolute best way of borrowing from natural languages would be to let your conlang stay in contact with the “source” language (e.g., English) for generations, allowing the loans to build up organically. But we don’t have that kind of time. How can we simulate that evolutionary process in a hurry?

Well, there are a lot of ways. For the modest goal of creating a natural-looking conlang backed by a plausible culture, following the guidelines I’ve mentioned in my “Let’s Make a Language” series will help. Rather than send you to read all of those, though, I’ll boil them down to their essence right here.

First, think about how existing languages borrow words. It’s not at random. It’s usually to fill a need, such as an imported food or a new invention. It could be political or religious in nature, as well, as the large number of Latin and Greek borrowings related to Christianity will attest. But it’s not often for things we already have words for. You don’t see common, basic vocabulary items like “sea” or “dog” being borrowed, because there was never any need. Yes, some specific subsets might come from loans (e.g., “maritime”, “canine”), but these are the exception, not the rule.

Second, languages are only going to borrow from those they have contact with. English today is everywhere, but that wasn’t always so. Japanese got most of its loans from Chinese to start, while Quechua (in South America) took mostly from Spanish. Borrowings, especially in pre-modern times, are going to come first from neighbors, second from conquering or conquered peoples, and last from a “lingua franca”. That does require you to locate your conlang in the real world, but it allows for greater verisimilitude, which is why you’re reading this post in the first place.

Finally, the culture of the conlang itself will determine what it borrows. Initially, it will move to fill gaps in its lexicon, and what those gaps are can create a different feel for the language. To create one contrived example, imagine a small culture undergoing a push for equal rights for women. It’s been mostly male-dominated up to now, and the vocabulary reflects that. But it has contact with French, which has gendered occupational titles. So it might borrow a few feminine forms here and there, or maybe even the -eur/-euse distinction as a whole. If the movement goes far enough, the existing (native) masculine words may be reinterpreted as gender-neutral forms, giving rise to a new dichotomy. Then, as more modern occupations become available (to men and women alike), the language would borrow terms for them, then modifying them to fit the new standard.

The same principle works pretty much everywhere. A perceived need is filled by taking from a nearby or well-known language that has already filled them. It works in all fields, under any circumstances. You can even see it at work today, among smaller natural languages. Look around, and you’ll see how many have borrowed, in some form, “telephone”, “television”, “automobile”, and a whole host of others. Of course, they wouldn’t need those words if they didn’t have those concepts, but that’s neither here nor there.

In other wor(l)ds

The same principle works in non-modern settings. You’d have to do a lot more work to come up with plausible borrowings from, say, Sumerian or Etruscan, but we know they provided loanwords to their neighbors. Remember, though, that older times imply less connectivity, less globalism. (Not always, as the Roman Empire proves, but it’s a good rule of thumb.) That also means more dialects, which can provide a bit more variety in your loans.

You can even generalize this to other worlds, though this one’s a lot more difficult. At some point, you’re making a whole “conworld”, rather than just a conlang, and that’s a different article for a different time. Still, the basic principle of “borrowing to fill in the gaps” works anywhere.

For a conlang intended to be spoken by a hypothesized real-world people, take from those languages that are supposed to be their neighbors. A culture hidden in an inaccessible corner of the Amazon isn’t going to start getting European loans until 1492, at the earliest. More likely, it’ll take some time for influence to diffuse that far, possibly even centuries. Likewise, central Africa isn’t going to get much Chinese influence until almost right now.

In a way, this whole process is reminiscent of the creation of an auxiliary language. But it still retains the artistic style, the creative flair of an a priori conlang. It’s almost like an intermediate form, you might say. A happy medium.

Let’s make a language, part 27c: Religion (Ardari)

Although I’ve been intentionally vague on the whereabouts of Ardari, it’s definitely less attached to European culture than Isian. To that end, it has few true borrowings for religious terms, instead relying on reinterpreted roots from the native belief system. Angels and devils, for instance, are firar and ghemar. A priest is an ekòna—but modern reformation has led to the creation of a feminine variant for what was once a masculine-only term: ekòni.

The Ardari priesthood might not have been a bastion of equality, but the faith itself was. Most supernatural beings in the old polytheistic system came in male and female forms, so there are pairs like tsora and tsori, or fira and firi, and these stand alongside the neuter terms used as the default.

That’s not to say that Ardari doesn’t import religious terms from other languages. It does, but it uses native words for most of the basic concepts. The sole loan in the list below is tyorymat “religion”, a conceptual term that only came in once Ardari speakers of eras past needed to talk about religion as distinct from faith. Other borrowings are made instead to describe concepts specific to one religion, such as santös “saint” (from Latin sanctus), èklecha “church” (from Latin ecclesia), or mazhid “mosque” (from Arabic masjid).

Word List

  • angel: fir (or gendered fira/firi)
  • devil: ghem
  • fairy: lyun (or gendered lyuna/lyuni)
  • faith: mitraz
  • ghost: qoj
  • god: tsor (or gendered tsora/tsori)
  • heaven: èlyas
  • hell: uldall
  • holy: mirs
  • magic: bräz
  • priest: ekòna (also modern feminine form ekòni)
  • religion: tyorymat (distant borrowing from theo-)
  • ritual: plan
  • sacred: grès
  • soul: jull
  • to bless: konye-
  • to curse: dakya-
  • to pray: nyes-

Let’s make a language, part 27b: Religion (Isian)

Isian, as we have seen, has borrowed more than a few terms from European languages. That shows up again in the matter of religion. Its speakers are mostly Christian, thanks to an earlier period of conversion and reformation. Before that, however, they had a polytheistic faith similar to many of their neighbors.

Remnants of this still show through in terms like alam “god”, which stands alongside the Latinate loan Domo “Lord”. The latter refers specifically to the God of Christianity, while the former, native, word can be used for any deity. It’s also more amenable to derivation, such as alanchi “demigod” or alamel “godly”. Domo on the other hand, is essentially fixed in form.

Other borrowings include engel “angel” and sacrel “sacred”, though the second is more of a calque. The word helin, meaning “ghost” or “spirit”, may also be related to the Germanic root underlying English “holy”. And it’s clear that priests have always been considered “holy men”, as the Isian word for them is a direct compound: chisam.

Word List

  • angel: engel (borrowed, possibly from Germanic)
  • devil: nukh
  • fairy: su
  • faith: sahe
  • ghost: helin
  • god: alam (Christian God usually trans. as Domo)
  • heaven: timiro
  • hell: hasilo
  • holy: chi
  • magic: ampen
  • priest: chisam (lit. “holy-man”)
  • religion: caltir
  • ritual: ronden
  • sacred: sacrel (borrowed from Latin/Romance)
  • soul: mit
  • to bless: leya
  • to curse: murgo
  • to pray: barda

Let’s make a language, part 27a: Religion (Intro)

As with the last part, we’re going to delve into a topic that may be a bit controversial. This time around, it’s the other half of church and state: religion.

For some languages, the whole subject is unnecessary. Quite a few, even among fictional conlangs, won’t need too many words for religious concepts. Auxiliary languages can likely get by with borrowing the needed terminology. And a far-future sci-fi setting might consider religion to belong to an earlier era.

On the other hand, even if the hypothetical speakers of your conlang don’t need to talk about their religion, that doesn’t mean they won’t want to talk about any religion. So it helps to have a bit of vocabulary specifically tied to the subject.

Gotta have faith

Religion and spirituality, in some form, have been around since the earliest days of humanity. Even if it’s nothing more than simple ceremonial burial, you can find evidence of the practice from the Stone Age, and some of our oldest human creations are religious in nature. It stands to reason, then, that a few basic ideas are going to be universal. The specifics might be wholly different even between two neighboring cultures, but they’ll both likely have some common ground in the fundamentals.

According to those who study the field (I don’t), religion of any kind probably started when someone first asked, “Why?” Why is the world like it is? Why do the seasons change? Why do people die? Maybe they begin as simple answers to those questions and more, or a shared set of stories, myths, and legends that only increase in popularity as they are told and retold over the generations.

This bare summary already gives us fertile ground for linguistic roots. The concepts most common to all religions are very likely going to be represented by native terms: faith, prayer, blessing and cursing, gods (or a monotheistic God, such as the case may be), an afterlife. Depending on the culture, you can also add in those placed in charge of religious matters, whether priests, shamans, or something else entirely. The ceremonies, rituals, and rites will also be in this field; they’ll likely be too specific to translate directly, but the words describing them won’t be.

As the folklore surrounding a religion grows, it necessarily gains a bit of verbal cruft. Even in Western Christianity, you’ve got quite a lot of vocabulary, from saints to bishops to crusades. (Note that many terms associated with Christianity, like “crusade”, tend to be related to “cross” or its analogues in Latin, Greek, and the Romance languages. That’s certainly not a requirement, but more of a historical quirk.)

Not only does a growing religion gain more words, but it also spreads across the lexical space, as it envelops closely related fields. Western faiths might all be monotheistic, but they each have a collection of supernatural beings, including (to use Christian-specific terms) angels, devils, demons, and ghosts.

This is where the twin forms of borrowing come into play. First, a highly organized religion will be able to spread its message far and wide, sending its specific terms to new places on the lips of its priests. So many English religious words come from Latin and Greek for this very reason. Similarly, Arabic loans related to Islam pop up everywhere from the western coast of Africa to the farthest reaches of Indonesia.

The second bit of borrowing comes when a new religion overtakes an old one. Here, it’s not so much that new words are borrowed, but the old ones may be reinterpreted, then spread in their new connotations. An example might be English ghost, which seems to have spent the last thousand years or so cycling between referring to a malevolent supernatural entity, the haunting spirit of the deceased, or even a kind of supernatural essence (as in the word spirit, itself a Latin loan). Fairies got their own bit of folk reinterpretation, while possibly-wise daemons became always-evil demons.

No matter what your conlang’s speakers believe, they’ll have a number of words specifically for their religion. The native terms will be made for that. If, along the way, the people were converted to some other faith, then they’ll likely take it in one of those two ways. Either they’ll import the words they need (spirit, bible, angel) from the “official” language of their church, or they’ll take some of their own and remake them to fit the new worldview: ghost, holy. Coinages tend to come about for new ways of thinking about the religion, and even then they aren’t made from whole cloth.

Discourse particles in conlangs

Speech is a funny thing. It really is, if you think about it. Compared to written language, it’s a lot less fixed in form. When we talk, we don’t speak perfect English (or whatever your language is). The words we say sometimes bear little resemblance to those we write.

One of the ways the two forms of language differ is that spoken language tends to include a lot of “filler” words. Writing doesn’t need them (unless you’re recording dialogue, obviously), but they come into our speech naturally, because we can’t talk as fast as we can think. It’s common to stop speaking while we come up with the next words we want to say, but silence is uncomfortable and ambiguous, so dropping a meaningless syllable or two here and there helps keep the conversation moving.

Filling the gaps

Pretty much any language meant to be spoken is going to have filler syllables or words. They won’t always be the same, but they’ll be there. For English, we’ve got a sizable collection, ranging from neutral syllables (“er”, “uh”, “um”) to entire phrases lacking in actual content (“you know”, “you see”, “I mean”).

Grammatically, these words and phrases add nothing to a sentence. They don’t really have a part of speech, even if their constituent parts normally would. Nor do they have semantic content, though some can indicate by their presence a very subtle difference in mood. Beyond that, they serve no purpose but to fill a gap. In short, they’re just…there.

Curiously, filler syllables, like English “er” and “uh”, do tend to have a few things in common across linguistic boundaries. It’s by no means universal, but there are patterns to be found.

For one, these syllables tend to have “neutral” vowels: unstressed, short, lax, central. Middle of the road, if you will. Even those closer to the corners of the vowel space, like “ah” (/a/), won’t be emphasized. In five-vowel systems, /e/ seems to be the most common filler phoneme. Seven-vowel systems often use /ɛ/ or /ɔ/. If there’s a schwa phoneme (or even allophone), you’ll likely find it here.

The consonants used in filler syllables also show a slight pattern. Whatever they are, they don’t tend to have a lot of “force”. Like vowels, they’ll tend to be the lighter, weaker type of phone. Something like /h/ is common in languages that have it. (Probably the weakest filler of all is /hʌ/, “huh”. Amazingly enough, recent research suggests that it might be universal among human languages.) Other “weak” consonants, like /ɹ/, /l/, /j/, etc., can also show up, but you’re probably not going to find too many voiced stops.

For both vowels and consonants, it’s also possible that a filler syllable uses an allophone that isn’t normally found in the language. Languages with smaller vowel systems that lack reduction might still have fillers with a schwa. A consonant inventory lacking /h/ might still have “huh”. The sounds would be analyzed by speakers as variants of something else, of course, but filler doesn’t even have to follow the normal phonotactics: English “yeah” and “nah”, which have filler uses, both show a final /æ/, a combination of phone and position that occurs nowhere else in the lexicon.

Discourse

Moving on from syllabic filler, we come to the broader category of discourse particles. These include the fillers we saw above, but also a set of language-specific words that have meaning in other contexts, but lose it when used in this manner.

For English, we’ve got a good collection: “like”, “well”, “so”, and many, many more. All of these are actual words, and we use them often in other places, but they become meaningless filler with ease. Overused, they, among others, tend to indicate something amiss with the speaker; “like” and “you know” are stereotypical markers of low intelligence or immaturity, for instance.

Other languages have their own sets of discourse particles. Like filler syllables, these words and phrases do show some patterns. They tend to be made from utterances that already have little semantic content, or that don’t add appreciably to a sentence. “Like”, in translation, is very common, though this may be English influence showing up.

Construction

Discourse particles and other forms of filler are perfectly natural in spoken language. In writing, of course, they should be avoided, with the exception of recording actual speech, but they are not, in themselves, harmful. And if they are natural, then a conlang made to look natural might want to consider having them.

How best to make them? For filler syllables, it’s not too hard. “Huh” is possibly universal, and it can be used as an interrogative with nothing more than the proper intonation. Others are slightly harder, but look for the most neutral sounds your conlang has, and start with those. Central or mid-open vowels, approximants, nasals, nothing that requires much mouth effort. You might even think of a filler syllable as a kind of glide between words—like glide sounds, there’s simply not a lot of friction involved.

Actual words and phrases are a bit tougher. Again, look for those that already don’t do too much. Your conlang’s word for “like” might be a good start, as are simple demonstratives like “this” and “that”, colloquial forms of “yes” and “no”, and generic confirmation phrases such as “right”, “you know”, or “I said”. Here, the type of language you’re making will have some influence: isolating languages might be more likely to have longer filler phrases, especially if they have restrictive phonologies—in this case, all the “good” words are already in use.

However you do it, remember that the point of filler words is to be, well, filler. They don’t mean anything, not really. But that’s not to say they’re truly meaningless. They serve their purpose, and that purpose is to give us time to think about the “real” words we want to use.

Let’s make a language, part 26c: Government (Ardari)

As is ever the case, Ardari is more likely to construct terms of its own. This is certainly true in the realm of government, where a number of words are derived from the root verb tysan- “to rule”. Indeed, in the short list below we already see tysanönda “authority, right to rule”; tysanègh “government, center of rule”; and tysanyn “rule, regulation, law”. Note also that these are native terms, not borrowings, though Ardari does have a few of those, including zhudis “court” (probably from “judicial” or something similar) and polisa “police force”.

These two are examples of modernization at work. The area where Ardari is spoken hasn’t entirely transformed into a modern Western democracy. There are plenty of elements of monarchy, aristocracy, and bureaucracy still around. At present, it might be best described as a parliamentary monarchy, closer to the UK than the US. It’s certainly not decentralized, however: the word for a province, dalrit, being a transparent derivation from dal “nation” is proof enough of that.

Most other terms are native, and they often have other connotations besides those shown here. For instance, makhèla, here glossed as “army”, can also connote any gathering of forces. (Phrases can be used to disambiguate: dalin makhèla “national army”; idyaze makhèla “attack force”; illin makhèla “rebellion”.) The same goes for byara “navy”, with creations such as dable byara, literally “land navy” but actually referring to an amphibious assault.

Word List

  • army: makhèla
  • authority: tysanönda
  • border: aroned
  • capital: präzdoza (lit. “great city”)
  • court: zhudis (borrowing, cf. “judicial”)
  • crime: karha
  • free: arin
  • government: tysanègh (lit. “place of rule”)
  • judge: tölera
  • law: gla
  • official: (tysanèghin) fèlokön (shortened nafèlokön “worker”)
  • nation: dal
  • navy: byara
  • peace: sèsym
  • police: polisa
  • province/state: dalrit
  • right (a right): èkhros
  • rule: tysanyn
  • tax: èzas
  • to control: träm-
  • to elect: soslin-
  • to permit: ejoten-
  • to prohibit: èkoten-
  • to punish: laqas-
  • to rule: tysan-
  • to vote: jamull-
  • war: jova

Let’s make a language, part 26b: Government (Isian)

Isian, when it comes to its government, fits into the usual “small” Western mold. It follows the typical European-style parliamentary system, with a number of parties vying for power. But it still has vestiges of a monarchy, too, a time when the land was ruled by a lakh or king. Today, even that is long gone, but remnants survive in words such as lactor “province”, literally a “king’s land”. The fact that nashil had to be borrowed should say something about Isian speakers’ centralization…or lack thereof.

The rule of law is also respected in its modern form. There are courts, judges, trials, etc., and they are (mostly) fair and just. On the national level, there is a standardized police force (borrowed term polisi), while localities generally have a holtedos—roughly speaking, a city guard or neighborhood watch. Officers of these are generally called holtem, which can also be more literally translated as “guardian”, while national policemen are instead polisimi.

In these tumultuous times, it’s also important to note that Isian speakers are protected by a defensive army (ancha) and navy (busa), though the latter is quite small and admittedly anemic. Historically, it has always been thus; Isian is a language of land forces, not sailors. But peace (histil) has reigned for a long time, and the last war involving speakers of the language is simply known as Cabrigo: The Great War.

Word List

  • army: ancha(s)
  • authority: awtorit (modified borrowing)
  • border: obres
  • capital: lireblon
  • court: caje
  • crime: cofan
  • free: mir
  • government: orisanas (from orisi “to rule”)
  • judge: teldem (from telde “to judge”)
  • law: rokh
  • official: rokhesam (from rokh + sam, lit. “law-man”)
  • nation: nashil (modified borrowing)
  • navy: busa(s)
  • peace: histil
  • police: holtedos (also holtem or borrowed polisi)
  • province/state: lactor (from lakh + tor)
  • right (a right): mas
  • rule (regulation): liyo
  • tax: ferma(s)
  • to control: camida
  • to elect: jiro
  • to permit: likha
  • to prohibit: nasco
  • to punish: agri
  • to vote: banki
  • war: cabri

Let’s make a language, part 26a: Government (Intro)

Ah, government. One of those things you can’t really live with, and you can’t really live without. But let’s not get political today. There’s enough of that going on without us adding our two cents. Instead, let’s talk about the ways government can be filtered through the lens of language.

Rules of rule

First off, remember that there’s no real “hierarchy” of government systems. None of them are truly natural, and all have their problems. (Democracy, so the quote goes, is the worst form of government…except for all the others.) Rome had a republic 2000 years before the US, the UK and Japan both have hereditary monarchs who wield little actual power, and although the West may have backed away from theocracy centuries ago, it’s still alive and well in the Middle East.

Thus, there’s not really a point where we can say that this form of government can’t be native to a language’s culture. Maybe you won’t find technocratic socialism in a Middle Ages society, and it’s doubtful that a far-future Earth will have brutal autocrats in charge. But the concepts don’t really change much. No matter how a land is governed, some things are near universal.

Most English terms regarding this subject are derived from Classical sources, usually Latin and Greek: republic, democracy, judicial. That’s okay; political science really started to get going around the same time that the classical revival was on, so it’s no surprise that the thinkers of that era chose the old tongues. And in many cases, the concepts themselves come from those same sources.

But there’s no reason it can’t be different elsewhere. Maybe a language’s culture developed a form of representative government on its own. Growing out of some kind of tribal council, for example. Then, republic would very likely be translated by a native root. So would elect and a lot more. As an alternative, they may use a “calque”, or loan translation, translating the meaning of the borrowed word instead of taking its form directly.

Law and order

Government, however, encompasses more than just the rule of a nation-state. It can also include a lot of basic societal structure. How are cities managed? Who protects those who cannot protect themselves? What about handling war, crime, or just simple grievances?

In many cases, a language will have native roots for the most fundamental concepts. Every civilization has something like cities; that’s practically the definition of civilization. War is, alas, within our very nature. Codes of law date back millennia, whether Hammurabi or the Hebrews. At its heart, you might even call all government merely different varieties of conflict resolution, and one of those philosophical conlangs could definitely find something there.

However you do it, the “oldest” things are likely to be native, but it’s not always certain. That’s especially true if you trace a culture’s history. There might be periods of occupation or enslavement, where they were wholly subservient to another people; in this case, many words would be borrowed from the “overlord” tongue, and not always with positive connotations. On the other hand, an imperialist culture would probably have more native terms for its institutions, rarely borrowing at all from its victims. (After all, they’ve already got all the words they need to rule.)

High and low

Now, all this pertains to government as it’s seen by its citizens. Not everybody is a political scientist, so most will tend to use the words they know. On the other hand, those who work with or study government will develop a much broader vocabulary. Legal codes tend to use verbose language, as they need to be definitive. Political treatises very often need to draw very fine lines, so they’ll need complex words or phrases to subdivide the concept space. And there’s nothing stopping the government itself from creating new words to describe its actions. (We’ve got a few of those in English, such as gerrymandering.)

In all, this causes a kind of diglossia. There’s the “high” language of the learned, describing nuances that most people neither want nor need to know, using words and phrases like autonomous, technocratic, subcommittee, or continuing resolution. Then you have the “low” vernacular, which talks plainly of votes and armies and judges and the like. It’s like this in many fields, particularly the scientific, but it somehow seems “thicker” when government gets involved.

Not every language will be subject to this process. It’s more likely to occur with agglutinating languages, because they’re practically made for it. Isolating types, on the other hand, will have more complex phrases instead of words, and those phrases might retain a bit of their transparency. That’s not to say the nuances won’t be there, but they’re expressed in different ways.

Moving on

No new part next month, as I’m working on a few other writing projects that need to take precedence. But I’ll be back in September with Part 27, covering the other half of the phrase “church and state”. Yes, we’ll be looking at religion, and I won’t deny that I’m glad I’m taking a month off before I deal with that.

Let’s make a language – 2017 translation special

It’s been awhile since the last time we saw our two “Let’s make a language” conlangs in action. Since I’m a bit preoccupied right now, to the point where I’m not really “feeling” the series as much, I thought I’d change that. So, here we’ll see another short bit of text translated into our two favorite fictional languages.

A year and a half ago, I used the Babel Text. This time around, it’s another linguistic classic, Aesop’s fable of the North Wind and the Sun. I chose it because it’s very popular for comparative language research, and—oh, never mind. I picked it because it’s short.

The North Wind and the Sun

The version of the fable I’m using is lifted directly from Wikipedia. I’m not entirely sure which translation it is, so yours may be slightly different. Nonetheless, it’s not that difficult a text, except for some fairly complex grammar bits.

The North Wind and the Sun were disputing which was the stronger, when a traveler came along wrapped in a warm cloak.

They agreed that the one who first succeeded in making the traveler take his cloak off should be considered stronger than the other.

Then the North Wind blew as hard as he could, but the more he blew the more closely did the traveler fold his cloak around him;

and at last the North Wind gave up the attempt. Then the Sun shined out warmly, and immediately the traveler took off his cloak.

And so the North Wind was obliged to confess that the Sun was the stronger of the two.

Isian

First up is Isian, and here’s what our little story looks like in it:

E Ul Naf wa e Sida nactas ed e folosin mid, ha ta usangam cosas cu disine ta him capat ijedo.

Is awcos ed ke costan hi ades los e usangam cu asine ey capat cal par pasa folosin dir a.

Ad e Ul Naf dosemay hi furus, nu i cu furus, e usangam otasis ey capat im oto;

ar tarkinas ni, e Ul Naf madeshis cu gati. Ad e Sida him hi shalis, ar necamay e usangam asines ey capat.

Teti e Ul Naf tole kinadid e Sida tes e folosin a o es naw.

Grammar

Mostly, the hardest part about translating this story is the sometimes complex sentence structure. Fortunately, Isian is fairly close to English in word order and the like (except for the postpositions thing), so it’s a lot more straightforward than it might seem at first glance.

That said, there are a few pitfalls. Lo, for instance, is a new verb meaning “to cause”. As you might expect from that gloss, it’s a causative verb. Its object is a subordinate clause, which makes things a bit hairy in line 2.

Most of the rest is just finding out which word to use. The indefinite pronoun ed is used to mean “which one” on line 1, and ijedo has a secondary meaning of “when” or “at the time of”.

Vocabulary

Not too much new vocabulary in this one, as you can see. Then again, there’s not a lot of text to translate, and it’s a bit repetitive.

  • to dispute/argue: nacta
  • about/concerning: mid
  • strong: folos
  • to travel: usanga
  • cloak or coat: capat
  • to cause, make happen: lo
  • to take off, remove: asine
  • other: dir
  • but: nu
  • to release, let go: madeshi
  • to try, attempt: gati
  • to shine: shali
  • immediately, at this moment: necamay
  • to admit, confess: kinadi

Ardari

Now we’ll move on to Ardari, and you can’t believe how much this one made my head spin.

Kyama Fawatövi Chitö qa èlldad am dortö mantö krazènedyt, tym jechinkön fynine pärine ilya ky vi tonad.

Ys bèt torydyid: Mantö qa jechinköntö ani pärine sudram ky twèralyët am dor zèt lembejëdall.

Drä, Kyama Fawatövi am dor èll ky furadökhan, adam furad jechinköntö ani pärine anön òs sòvadjyn,

Kyama Fawatövi èftanyntö zhajopad. Drä, Chitö fynin blajadjyn, jechinköntö ani pärine nyasab sudramad.

Ghinyas Kyama Fawatövi qa Chitö am dortö weghetö èllda is ky kómaryd.

Grammar

As with Isian, most of the grammar should be obvious. Here’s what’s not:

  • -khan is a conjunction clitic. When added to a verb, it’s like saying, “, but…”

  • len- is a derivational affix for verbs, roughly connoting “in a specific way”. Here, it’s used to create lembejë- “to think of” from bejë- “to see”. (Usually, this will make a ditransitive verb, but Ardari is fairly free about dropping verbal arguments.)

  • The verb kóma- “to cause” works essentially the same way as Isian lo, taking a subordinate clause as an object. Again, this is much different from the English method of “that” followed by a clause.

Vocabulary

Note that, in this list of new vocabulary, I haven’t included regularly derived terms like jechinkön “traveler”.

  • to dispute, argue over: krazèn-
  • strong: dor
  • when: tym
  • to travel: jechin-
  • coat, jacket: pärin
  • to consider, see as, think of: lembejë-
  • should (adverb): zèt
  • to take off: sudram-
  • but (conj.): -khan
  • more: adam
  • near, close: myll
  • to give up, surrender: zhajop-
  • to attempt, try: èfta-
  • to shine: blaja-
  • immediately: nyasab
  • to make, cause: kóma-

Conclusion

And there we go. It was a bit later this year, and it wasn’t nearly as much as last time, but we’ve successfully added another extended text to our repertoire for Isian and Ardari. Hopefully, you can do the same for your conlangs, whatever they are.

Next month, we’ll pick back up with the usual topic-based posts. Specifically, we’ll be looking at government, because nobody’s doing that these days, right?

Will there be more translation exercises? Probably. Not this year, most likely, but stay tuned. Who knows? 2018 might have something even better in store.

Random generation for conlangs

Making a language is hard work. As anyone who knows programmers knows, hard work is not something we like. Not hard work that we have to do ourselves, that is. If we can find a way to get a computer to do it for us, well, that’s entirely different.

On the surface, it seems pretty simple. Language creation is a lot of time-consuming, repetitive work. We have to make hundreds or thousands of words. We have to work out grammatical rules. The list goes on, and it’s almost a sure thing that, at some point, the computer-savvy conlang creator is going to ask, “Can I automate this?” And thus is random generation born.

The pros

It’s not that bad an idea, if you do it right. Computers are a lot faster than the human mind when it comes to such things. They can generate a million random words in the time it takes us to think of one. So shifting some of the burden to the uncaring machines seems almost natural. After all, they come up with random numbers all the time, don’t they? So what’s the difference.

For words, it’s not even that big a deal. If you’re working with an artistic conlang, or something otherwise not bound by the sequences of sounds used by natural languages, there’s almost no reason not to randomly generate at least some roots. Not whole words, mind you, because they won’t necessarily fit the grammatical and morphological patterns of your language. No, it’s better to generate the basic roots, then inflect them however your conlang does that—if it does at all.

Word generators are easy enough to find, and they aren’t all that hard to make. (I’ve done it no fewer than four times.) Most of the good ones—not mine—give you a lot of nice options. They let you do frequency distributions, so some sounds are more common than others. A few even provide variable substitution, syllabic constraints, allophony, and you get the idea. I won’t say that random word generation is a solved problem, but it’s definitely not new. If the circumstances of your conlang allow it, and you think you can get good results, go for it.

At other levels, random generation is a bit more iffy. Gleb is a part of an (abandoned?) attempt at randomly creating every part of a language; the link goes to a phonology generator, which…works. It’s a start, though, not an end result. I used a few of its outputs as the initial seeds for the conlangs of my “Otherworld” setting, but all of those required a lot more polish before they were anything approaching usable.

It’s also possible to generate random grammar rules and the like, but the field is so vast, and the different parts of a language so interwoven, that I’m not sure you could ever write a program that could give you something plausible. If you could, you’d probably be halfway to strong AI already. I’d like to see it, though.

The cons

On the flip side, random conlang generation has an awful lot of downsides that make the process unsuitable in a lot of instances.

First and foremost, any conlang intended to be an auxiliary language almost certainly can’t use randomness at all. It simply doesn’t fit the criteria. These languages are supposed to be either familiar to a broad population, simple enough for anyone to learn, or engineered based on linguistic principles. None of that really meshes well with random words, much less any other part of a conlang.

Second, unless you’re willing to go through a lot of trouble, the output of a generator isn’t always that great. Yes, the numbers it uses internally will likely be entirely random, but mapping those numbers to letters, phonemes, words, etc., in a way that doesn’t look, well, computer-generated, is an exceedingly difficult task. Take any of those word generators I mentioned above. Sure, you can get a good list out of them. More likely, however, you’re going to be looking at thousands of nonsense letter sequences that do nothing but waste your time.

Next, as I said above, most of the grammatical portion of a conlang isn’t really amenable to randomizing. Language structure is so full of causation and correlation, of universals and implications, that it’s just not that random. It’s more like a chain of logic, but with a few forks in the road that give the creator a bit of leeway. You don’t have to, say, have a past tense just because you have a future one, but it’s more common if you do, and random generators would have to account for that. In the end, there are so many variables, so many special cases, that the programmer effectively has to make a language just for the generator.

I could go on, but I’ll leave with one final argument against randomness, one I don’t entirely support. Nonetheless, it is popular, so I’d be remiss if I didn’t at least mention it. That reason is art. If you’re of the belief that conlangs are art, you might balk at the very notion of using computers to create them from random numbers. That would be no different from trying to use an RNG to make artistic textures, or images, or worlds, and we see how often those go awry. (Minecraft and the like use noise functions, which are entirely different. In many cases, they’re actually predictable; the same world seed will give you the same world.)

If you consider a created language to be an artistic work, then part of its allure is in the way it is crafted. We, the makers, choose words based not on algorithms, but aesthetics. It’s a more…philosophical argument, in my opinion, but I can see the reasoning.

Conclusion

In summary, my thoughts on the subject are as follows. Use random word generators if your conlang supports them. Don’t use them for anything where roots have to be derived from some other principle than your own mind. A phonology generator can be a good starting point, but not a finished product. Generators claiming to create grammar and the like for you probably aren’t going to give you something sensible, much less usable.

That’s not to say you can’t have fun with random generation. It is fun, and I can’t count the number of names I’ve used that ultimately derive from a list created by a computer program. (Some have even been published! Well, they will be in the coming weeks.) It’s not yet a substitution for the hard work of our minds, however, and it may never be. By the time a computer can create a language from nothing more than a random sequence of numbers, we may have bigger problems than conlangs becoming a commodity.