Tag: conlang

For the first time in this series, not only will we be able to treat Isian and Ardari in the same post, but we’ll actually look at them at the same time. We can do this thanks to the similarity in the way they treat numerals. Sure, there are differences, and we’ll see those as we go, but the highlights don’t change that much from the “simple” Isian to the “complicated” Ardari.

The numerals

First off, both conlangs use a decimal system, like most languages in common use today. Both are based around the number ten, but in slightly different ways. Ardari is a more “pure” decimal language, although it has a little bit of vigesimal contamination; Isian, on the other hand, likes to work with hundreds for larger numbers. Although that may sound odd, think about how we do it in English: a million is a thousand thousands, a billion a thousand millions, and so on.

Before we get to the meaty grammar bits, here’s a table of numeral words in both conlangs. It shows all numerals up to twenty, all the multiples of ten up to a hundred, and a few selections to illustrate the numbers in between.

Number	Isian	Ardari
1	yan	jan
2	naw	wegh
3	choy	dwas
4	khas	fèll
5	gen	nibys
6	hod	sald
7	sowad	chiz
8	nicul	ghòt
9	pir	ang
10	pol	kyän
11	poloyan	vänja
12	polonaw	braj
13	polochoy	kyävidas
14	polokhas	kyävèll
15	pologen	kyuni
16	polohod	kyävisald
17	polosowad	kyävichiz
18	polonicul	kyävijòt
19	polopir	kyäveng
20	nopolic	darand
21	nopoloyan	darandvi jan
22	nopolonaw	darandvi wegh
30	choypolic	dwaskyän
33	choypolochoy	dwaskyänvi dwas
40	khaspolic	wedarand
50	gempolic	byskyän
60	hobolic	dwasrand
70	subolic	chiskyän
80	nilpolic	fèldarand
90	pirpolic	änkyän
100	cambor	grus

In both languages, the default form of a numeral is as an adjective. For Ardari, this requires adjective inflection for the first four, including changing for the gender of their head nouns. On the Isian side, every number but yan “one” will have a plural head noun, but there is otherwise nothing to worry about.

We can use numerals directly as nouns in Ardari, just like any adjective, but we can’t in Isian, since it doesn’t allow adjectives without head nouns. Instead, we can use the “dummy” noun at: naw at “two things”. (For “one”, we’d use the singular yan a.)

Creating higher numbers in Ardari is, surprisingly, fairly straightforward. As you can see in the table above, numbers like 21 are constructed using the linking conjunction -vi, which appears on everything but the last noun or adjective in the phrase. Thus, darandvi jan is literally “twenty and one”. This pattern extends throughout the system: 123 is grusvi darandvi dwas.

In Isian, things get a little hairier. Up to 109, you take the “tens” numeral, strip off the final -ic, add on a linking -o-, and add the “ones” numeral: nopolic “twenty” plus yan “one” equals nopoloyan “twenty-one”. Past that, you have to make a phrase like polopir cambor at wa nilpolochoy “1,983”, but this takes you all the way to 9,999.

For positively huge numbers, you need more numerals. Isian has two native higher powers: jagor “ten thousand” and ilicor “million”, which can be used just like cambor “hundred”. As an example, the large number 1,048,576 would be represented in Isian by the mouthful ilicor at wa khas jagor at wa nilpologen cambor at wa subolohod. Yes, our way looks more compact, but imagine writing it out.

Ardari instead has separate words for each power of ten up to a million: ulyad “thousand”, minyir “ten thousand”, ovòd “hundred thousand”, and akrèz “million”; these can be “stacked” into a -vi phrase with the others. Our same example in the paragraph above, 1,048,576, then becomes akrèzvi fèll minyirvi ghòt ulyadvi nibys grusvi chiskyänvi sald. (As a shorter alternative, one can simply recite the digits in order, putting yvi before the last: jan zu fèll ghòt nibys chiz yvi sald.)

That last example shows the Ardari word for zero, zu. Isian has one, too: anca. However, it has an added wrinkle in that it doesn’t work the same way as the other numerals. To say “zero” as a noun, instead of using anca at “zero things”, you say anocal, the Isian word for “nothing”.

Our number is up

That’s all there is to it for counting numerals in our conlangs. They’re fairly simple, mostly because I stuck to a decimal number system. If you want to use something more “exotic”, like base-12, well, have fun with that. I’ve tried, and it’s a lot harder than it looks. Still, the “dozenal” people don’t seem to mind. Also, there’s a lot of grammar stuff I could have added, and we haven’t covered ordinal numbers, but those can come later. We can count in our languages now, and that’s good enough for the time being.

After learning how to speak, counting is one of the first things children tend to figure out, for obvious reasons. And language is set up to facilitate learning how to count, simply because it’s such an important part of our existence as human beings. The familiar “one, two, three” of English has its counterparts around the world, though each language has its own way of using them.

These numerals will be our focus today. (Note that we can’t really call them numbers in a linguistic context, because we’re already using the term “number” for the singular/plural distinction.) Specifically, we’ll look at how different languages count with their numerals; in math terms, these will be the cardinal numbers. In a later post, we can add in the ordinal numbers (like “first” and “third”), fractions, quantities, measurements, and all that other good stuff. For now, let’s talk about counting.

Oh, and since numerals lie at a kind of intersection of linguistics and mathematics, it’ll help if you’re familiar with a few concepts from math. While we won’t be going into things like positional number systems—I’ll save that for a post about writing systems, far into the future—the concept of powers will be important. More information shouldn’t be that hard to find on the Internet, so I’ll leave that in your capable hands.

Count the ways

How a language counts is highly dependent on its culture. Remember that counting and numeral words predate by far the invention of writing. Now think about how you can count if you can’t write. One of the best ways is by using parts of your body. After all, it’s always with you, unlike a collection of stones or some other preliterate method. Thus, bodily terms often pop up in the context of numerals.

In fact, that’s one of the simplest methods of creating numerals: just start numbering parts of your body. A few languages from Pacific islands still use this today, and it’s entirely possible that it’s how all ancestor languages did it. Words for the fingers of one hand usually cover 1-4, with the thumb standing for 5. After that, it depends on the language. Six could be represented by the word for the palm or wrist, and larger numbers by points further up the arm. In this way, you can continue down the opposite arm, to its hand, and then on to the rest of the body.

Once you need to work with larger numbers, however, you’ll want a better way of creating them. The “pointing” method is inefficient—you need to remember each point on the body in order—and there are only so many body parts. This is fine for a hunter-gatherer society, and many of those have a very small selection of numerals (anywhere from one to five), using a word for “many” for anything higher. But we “advanced” peoples do need to refer to greater quantities. The solution, then, is to use a smaller set of numerals and construct larger ones from that. That’s how we do it in English: “twenty-five” is nothing more than “twenty” plus “five”.

For our language, the key number is 10. Every number up to this one has its own numeral, while larger ones are mostly derived. The only exceptions are words like “hundred” and “thousand” which, incidentally enough, represent higher powers of 10. Thus, we can say that English uses base-10 counting—or decimal, if you prefer fancier words.

At the base

Every language with a system of numeral words is going to have a numerical base for that system. Which number is used as the base really has a lot to do with the history of the language and how its people traditionally counted. Not every number is appropriate as a base; Douglas Adams once said that nobody makes jokes in base-13, and I can state with confidence that nobody counts in it, either. Why? Because 13 is awkward. It’s a prime number with essentially no connection to any part of the body. Since counting probably originated with body parts, there’s no reason for a culture to ever develop base-13 counting. Other numbers, though, are quite suitable.

• Decimal (base-10) counting is, far and away, the most common in the world. Look at your hands, and you’ll see why. (Unless, of course, you don’t have ten fingers.) Counting in decimal is just the finger counting most of us grew up with, and decimal systems tend to have new words for higher powers of 10. In English, we’ve got “hundred” and “thousand”, and these are pretty common in other decimal languages. For “ten thousand”, we don’t have a specific native word, but Japanese (man) and Ancient Greek (myrioi) do; the latter is where we get the word “myriad”.

• Vigesimal (base-20) is not quite as widespread as decimal, but it has plenty of supporters. A few European languages use something like base-20 up to a certain point—one hundred, in fact—where they switch to full decimal. But a “true” vigesimal system, using powers of 20 instead of 10 (and thus having separate words for 400, 8,000, etc.), can be found in Nahautl (Aztec) and Maya, as well as Dzongkha, in Bhutan. Like decimal, vigesimal most likely derives from counting, but here it would be the fingers and the toes.

• Quinary (base-5) turns up here and there, particularly in the Pacific and Australia. Again, it comes from counting, but this time with only one hand. It’s far more common for 5 to be a “sub-base” in a greater decimal system; in other words, 10 can be “two fives”, but 20 is more likely to be “two tens”. The alternative, where the core terms are for 5, 25, 125, and so on, doesn’t seem to occur, but there’s no reason why it can’t.

• Duodecimal (base-12) doesn’t appear to have an obvious body correlation, but it actually does. Using the thumb of one hand, count the finger bones on that hand. Each finger has three of them, and you’ve got four non-thumb fingers: 3 × 4 = 12. There are a few languages out there that use duodecimal numerals (including Tolkien’s Quenya), but base-12 is more common in arithmetic contexts, where its multiple factors sometimes make it easier to use than decimal. Even in English, though, we have the “dozen” (12) and “gross” (144).

• Other numbers are almost never used as the “primary” base in a language, but a few can be found as “auxiliary” bases. Base-60 (sexagesimal), like our minutes and seconds, is entirely possible, but it will likely be accompanied by decimal or duodecimal sub-bases. Some languages of Papua New Guinea and thereabouts use a quaternary (base-4) system or, far more rarely, a senary or base-6 system. Octal (base-8) can work with finger counting if you use the spaces between your fingers, and a couple of cultures do this. And, of course, it’s easy to imagine an AI using octal, hexadecimal (base-16), or plain binary (base-2).

Word problems

In general, numerals up to the primary base are all going to be different, as in English “one” through “ten”. A few powers of the base will also have their own words, but this will be dependent on how often the speakers of a language need those higher numbers. “Hundred” and “thousand” suffice for many older cultures, but the Mayans could count up to the alau, 20⁶ or 64 million, China has native terms up to 10¹⁴ (a hundred trillion), and the Vedas have lots of terms for absurdly large numerals.

No matter what the “end” of the scale, most of the numbers in between will be somehow derived. Again, the more often numbers are used, the more likely they’ll acquire specific terms, but special forms are common for multiples of the base up to its square (100 in decimal, 400 in vigesimal, and so on), like our “twenty” or “eighty”. Intermediate numbers will tend to be made from these building blocks: multiples and powers of the base. How they’re combined is up to the language, but the English phrasing, for once, is a pretty good guide.

Some languages work with a secondary base, and these may affect the way numeral words work. Twelve and twenty can almost be considered sub-bases for English with words like “dozen” and the peculiar method of constructing numbers in the teens. Twenty is a stronger force in other European languages, though. French is an example here, with 80 being quatre-vingts, literally “four twenties”. In contrast, a full vigesimal system can function just fine with the numeral for twelve derived as “ten and two”, using 10 as a sub-base, although I’m not aware of an example. Any factor can also work as a sub-base, especially in base-20, where 4 and 5 both work, or base-60, where you can use 6 and 10.

Irregularity is everywhere in natural languages, and that includes numerals. There always seem to be a few outliers that don’t fit the pattern. English has “eleven” and “twelve”, of course; it gets them from Germanic, as do many of its cousins. Spanish, among others, has veinte for 20, whereas other multiples of ten are constructed fairly regularly from their “ones” (treinte, etc.). Other examples abound.

Fitting in

How numeral words fit into a language is also a major variable. Sometimes, they’re a separate part of speech. Or they can be adjectives. Or nouns. Or some combination of all three. If they’re adjectives or nouns, then they may or may not participate in the usual grammar. Latin, for instance, requires small numerals (up to four) to be inflected, but everything larger is largely fixed in form. English lets numerals act as adjectives or nouns, as needed, and some dialects allow nouns following adjectival numerals to ignore grammatical number (“two foot of rope”, “eight head of cattle”). It’s really a mess most everywhere.

For a conlang, it’s going to come down to the necessities. Auxlangs, as always, need to be simple, logical, and reasonable, so it’s best not to get too crazy, and this extends to all aspects of numerals. You’re not going to get many followers if you make them start counting by dozens! (Confession time. I did this for a non-auxlang over ten years ago, and I still forget it uses duodecimal sometimes! Imagine how that would be for a language intended to be spoken.)

Fictional languages get a little bit of a pass. Here, it’s okay to go wild, as long as you know what you’re doing. Non-decimal bases are everywhere in conlangs, even in “professional” ones like Tolkien’s. With non-humans, you get that much more rope to hang yourself with. Four-fingered aliens (or cartoon characters) would be more likely to reckon in an octal system than a decimal one. Depending on how their digits are made, you could also make a case for base-6 or base-9, by analogy with Earthly octal and duodecimal finger counting. Advanced races will be more likely to have a sophisticated system of higher powers, like our billion, trillion, etc. And so on.

More than any other part of this series, numerals are a part of a culture. If you’re making a conlang without a culture—as in an auxlang—then think of who the speakers will be, and copy them. Otherwise, you might need to consider some of the aspects of your fictional speakers. How would they count? How would they think of numbers? Then you can start making your own.

How do we form questions in Isian and Ardari? The answer, you will see, is quite simple.

Isian

Isian, fittingly, doesn’t have a lot of question “morphology”. Yes-no questions are made in the simplest possible manner, by nothing more than rising intonation. This means, however, that the meaning of, say, so il til can be one of two things. With falling or level pitch, it’s a statement “you are there”. Go up in pitch as you near the end, and it becomes so il til?, the interrogative “are you there?”

The answer to such questions will usually be a simple sha “yes” or num “no”. If you need more, Isian allows you to add it by repeating the verb: sha, en “yes, I am”. (Note that I’m using English punctuation as a convenience, but also because there would be a slight pause between answer word and verb.)

If you prefer the vernacular, you’ve got shasha, which is more like “yeah”; noy is the negative counterpart, and its best translation might be “nope”. A wishy-washy reply would be momay “maybe”, while genuine ignorance can also be expressed by ekh “I don’t know”.

Negation in Isian is accomplished with the adverb an, as you’ll recall, and this extends neatly into the realm of the question. We can just as easily ask so an il til? “aren’t you there?” We don’t have to worry about double negatives, though; proper responses would be sha, en or num, an en.

Isian even gives you a couple of tags. These are highly discouraged in formal speech or writing, but common among friends and family. The one that concerns us most is ey, which works like English “isn’t it” and friends: so il til, ey? thus means something more like “you’re there, aren’t you?”

For the more general wh-questions, we have a family of fronted interrogatives:

• con “who” (only used for people)
• cal “what” (never used for people)
• cazal “where”
• carec “when”
• canyo “why”
• cadro “how”

These go at the front of a sentence, which is otherwise unchanged, except for a bit of rising intonation at the end. An example of each might be:

• con so il? “who are you?”
• cal to e? “what was that?”
• cazal so wasal? “where are you going?”
• carec is cosa? “when did they come?”
• canyo so kil to “why do you say that?”
• cadro so il “how are you?”

The more formal a situation, the more answer is required. Common speech can get away with single-word answers, but writing might need whole sentences. The rules are broadly similar to those in English, but Isian is overall more relaxed.

Ardari

Ardari’s interrogatives are built around the particle qö, which begins all questions. For yes-no questions, it’s all you need, other than the requisite intonation: qö sy pren èllè? “are you there?”

Valid responses will start with è “yes” or kyu “no”, usually repeating the verb in more formal speech and writing. Thus, there is a distinction between è “yes” and è èllo “yes, I am” in Ardari.

The same particles, when placed at the end of a sentence, can also function as tags expressing an expected reply. In these cases, the question particle isn’t needed, only the intonation: sy pren èllè, kyu? “you’re there, aren’t you?”

For wh-questions, the basic premise remains the same. The particle qö goes at the beginning of the sentence, but the question word stays where it is. As for the question words themselves, Ardari has eight of them, shown here with examples:

• qom “what”: qö qom pralman èlla? “what was that?”
• qomban “who”: qö sy qombane èllè? “who are you?” (lit. “you are whom?”)
• qomren “where”: qö sy qomren chinès? “where are you going?”
• qomlajch “when”: qö ajo qomlajch toned? “when did they come?”
• qoman “which”: qö sy qomane lyebè? “which do you like?”
• qabre “how”: qö ysar zalman qabre troded? “how did they know that?”
• qömjas “how many”: qö a qömjasòn byzrell perada? “how many books does he have?”
• quld “why”: qö ti quld ajnadyt? “why was she crying?”

Of these eight, qom, qomban, and qomren inflect like neuter nouns, while qoman and qönjas act like neuter adjectives. The rest function as adverbs. In all cases, if they would be the first word in a sentence, Ardari allows you to omit the initial qö, as it’s subsumed into the question word itself. (They’re all derived from it, in case you hadn’t noticed.)

Answering these questions requires only the bare minimum. Ardari is very lenient on how you reply, and even in formal situations you can get away with a response of only a word or two. For instance, qö sy qomren chinès? above can be answered with just mynin tyèk “my house”. Even inflections can be largely ignored in this form, though you’ll need them for an extended answer: my mynin tyèke lim chinos.

One added wrinkle involves single-word answers of pronouns. In this case, Ardari uses the vocative, which otherwise doesn’t appear often. Thus, qö sy qombane èllè? can be answered with myne “Me!” Simple my, on the other hand, would be ungrammatical.

Next up

I know the question you might be asking right now. “What’s in the next part?” The truth is, I don’t know yet. I’m thinking about taking a bit of a diversion into more general conlanging issues. We’ll get back to the step-by-step guide to making languages a little bit down the road. Whatever I decide, I’ll see you next week.

How are you? What’s up? What am I talking about?

Up to this point, our look at language has focused primarily on the declarative, statements and utterances of fact or conjecture. That’s great, because those make up the largest part of a language, but now it’s time to move on. Why? Because we need to ask questions.

Asking the question

How do we ask a question? In English, you already know the answer, and it’s pretty complicated. Worse, it’s complicated in different ways depending on what kind of question you’re asking. So let’s take a step back.

Questions (interrogatives, if you prefer the more technical term) are, at their core, requests for information. We don’t know, so we have to ask. We’ve already met a couple of cases where we didn’t know something, like the subjunctive mood, and “interrogative” can indeed count as its own mood. But questions are a little different, because they are directed at the listener with the intention of receiving an answer.

If you think about it, you’ll find that questions fall into a few different categories. One is the yes-no or polar question; as its name suggests, this kind expects one of two answers: an affirmative (“yes”) or a negative (“no”). Examples of polar questions in English might be “Are you going with us?” or “Did you see that?” For English, yes-no questions are marked by “inversion”, where the verb (or an appropriate auxiliary, like do) is moved to the front of the sentence, and that’s fairly common in its relatives and neighbors, such as German and French. It was even more common in the past, as anyone reading Shakespeare or the King James Version of the Bible would know.

Another kind of question is usually known as the wh-question, after its most distinctive feature in English. These are the ones that request a specific bit of information like identity, location, or reason, asking things like “Who are you?” or “Where are we going?” In our language, they employ one of a handful of question words (“who”, “what”, “where”, “when”, “why”, and “how”), that most often appear the beginning of a sentence. This type of question also has inversion, but only after the question word has moved into place.

Alternative or choice questions make up a third type. “Do you want grape or orange?” is an example showing how this one works. Options are presented, with the expectation that the answer will be among them. This one allows, even begs for, an answer in the form of a simple stating of the preferred choice. This sort of elliptical response (a sentence consisting solely of “Grape,” for example) is very common, especially in speech, no matter what the formal grammar of a language might say.

Tag and negative questions, the last two of the major types, are similar to each other in that they both presuppose an answer, but they go about it in different ways. Negative questions use a negated form of a verb, as in “Aren’t you coming?” Tag questions, on the other hand, are formed as indicative statements “tagged” by an additional interrogative bit at the end: “You’re coming, aren’t you?” Strictly speaking, these are both polar questions, in that they invite a yes/no response, but the prototypical yes-no question (“Are you coming?” in this example) has a more neutral tone. Negatives are asked from a position of expecting a negative reply, while tag questions work more for confirmation or even confrontation.

Keep asking

English, again, is pretty complicated when it comes to questions. Polar and wh-questions use inversion, while wh-questions add an interrogative word into the mix. Tag questions basically have their own set of interrogative words (“you know”, “isn’t it”, and so on) that go at the end of a sentence, turning a statement into a question. All in all, there’s a lot to worry about, and other languages have their own systems.

There is one universal, however, and that is intonation. Nearly all known human languages, mo matter how they form polar questions, have a specific way of marking them. The intonation, or pitch level, of yes-no question sentences always rises from beginning to end. In English, it’s even possible to have this as the sole indication of a spoken interrogative, as in the statement “you’re coming” versus the question “you’re coming?” Some other languages, such as Spanish, only allow this method, as opposed to the inversion usual in English. (Question marks serve essentially the same purpose in these cases, but for the written form of the language.)

Looking around the world, you’ve got a few other options, though. You can add an interrogative mood marker to the verb, as in Turkish and others; this is probably going to be more common in languages where verbs already have a lot of marking. Another option is an interrogative particle, which can go just about anywhere. Polish has czy at the beginning of a question, which Esperanto lifted directly as ĉu. Japanese has the sentence-ending ka (phrases ending in “…desu ka?” are known to every lover of anime), fitting its hardcore head-finality. Latin puts nē in a kind of “second” position, after the questioned part; it also has the similar num for negative questions and nonne for positives.

Chinese, among others, takes a different approach, sometimes referred to as A-not-A. Here, the polarity is redefined in the form of an alternative question: a rough translation might go something like, “Is he there or not?” (“He is/is not there?” comes closer to the original, at the expense of being horrible clunky.) Another option, more likely to be found in colloquial speech rather than formal grammar, is through liberal use of tag questions or something like them.

Tag questions themselves are likely to be marked only by the tag and its intonation, as above. Wh-questions, on the other hand, have potential for more variation in their formation. Many languages use question words like those in English, and they are commonly moved to the front of a sentence, functioning as their own question particles. That’s not the case everywhere, however; although it has a specific connotation in English, we can still ask, “You want what?” (Unlike polar questions, intonation isn’t a guide here. English continues to use rising pitch for wh-questions, but Russian, for instance, doesn’t.)

The answer

Asking a question is one thing. Answering it is quite another. And answers to questions have their own grammar and syntax beyond what a normal statement would require.

Very many languages, maybe even all of them, allow a speaker to omit quite a bit when responding to a question. “Yes” and “no” can be sentences all by themselves in English, as can “si” in Spanish or “non” in French. Not every language, though, has equivalents; some instead repeat part of the question in a positive or negative form. Still others have two versions of “yes” and “no”, with one pair used for answering positive sentences, the other for negatives. (Even those that don’t can vary in the meaning of “no” when it answers a negative question. Does that create a double negative? It does in Japanese, but not English.)

Beyond polar questions, how much of a reply you need often depends on what you’re being asked. In general, a lot of languages allow you to express only the most specific part of a phrase under question: “Where are you going?” can be answered by “Home.” A fuller answer would be “I’m going home,” but the short form is perfectly acceptable in speech, and not only in English.

Further questions

So that’s it for questions in general. Next, we’ll look at the very specific question of, er, questions in our conlangs.

Now that we have the theory out of the way, adverbs—whether words, phrases, or clauses—aren’t going to be too bad, for either Isian or Ardari. We already got a glimpse of them in both languages, back in the Babel Text, but now it’s time to see them for real.

Isian

As always, we start with Isian. As you may recall, Isian adjectives normally can’t appear without a head noun. Well, now they can, and that’s how we make most adverbs.

In Isian, we use postpositions, and the postposition hi is our go-to for adverbs. It’s the equivalent of English -ly, Spanish -mente, and so on, making adverbs out of adjectives. Examples might include ichi hi “beautifully” (from ichi) or bil hi “well” (from bil “good”, with no stem change like in English). Couldn’t be simpler.

We can fit these into sentences by placing them just about anywhere. Just before or just after the verb phrase are the most common, though. An example might be sha seri ichi hi “she sings beautifully”, which could also be written sha ichi hi seri.

Little hi can also work for phrases, with almost the same meaning. Take the sentence mi doyan hi cheren im “I see him as my brother”. Granted, it’s a little on the metaphorical side, but it illustrates the point. (You can write this one as cheren im mi doyan hi if you like, but that way emphasizes the object “him” rather than the adverbial phrase.)

For full clauses, we need a little bit more grammar. First, we have the general conjunction ha, which introduces adverbial clauses. In certain informal situations, we don’t have to put it in, but it’s mandatory otherwise. Second, since Isian uses postpositions, it also has “clause-final” conjunctions. Thus, the words that would translate as “before”, “after”, and so on appear at the end of the phrase, not the beginning, as in English.

These two rules cover most of what we need to know, and we can already make quite a few clauses. Here’s a couple of examples:

• is hamas ha is inamas pane, “they ate before they went to sleep”
• mit las an wasanda ha is likhas mida todo, “we couldn’t go because they wouldn’t let us”
• em cosata ha cheren es abradi terta, “I came to see the mountains”

For all subordinate clauses like this, Isian’s default is independent. Dependent clauses are only allowed in a few cases, namely those of purpose or cause. (Desire or wanting, using the verb doche, allows dependents, too, but that’s not really an adverbial.)

To construct a dependent clause, all you need to do is use the infinitive form of the verb, which is the bare verbal stem (or 3rd-person singular present, which has the same form) preceded by cu. Thus, we might have cu chere “to see” or cu lenira “to read”. From there, the clause mostly follows English rules, except that the conjunction goes at the end, if it’s there at all.

Of our examples above, only the third can be rewritten as a dependent: em cosata cu chere es abradi. The first indicates time, which isn’t allowed to be “deranked” in this fashion, while the second has different subjects in the main and subordinate clauses. (Like English, we only get to use the infinitive version when the subjects would be the same.)

Ardari

Ardari, curiously enough, starts out easier than Isian: adjectives can be used as adverbs directly. They don’t inflect like this; they’re just…there. An example is ti ojet ajanga “she sings sweetly”.

Strictly speaking, Ardari doesn’t have simple adverb phrases, so we’ll skip ahead to the clauses. For this conlang, there’s a distinction in those. “Purpose” clauses (along with “wants” and perception, though these aren’t adverbial in nature) are always dependent, but everything else is normally independent.

These two groups are distinct in their position, as well. Dependents always precede the head verb, while independents are allowed to follow it, one of the few flaws in Ardari’s head-final nature. But independent clauses can be moved around freely, even fronted, like in English.

If that weren’t bad enough, adverbial clauses of time can appear in either form. In speech, it’s considered better to use the dependent form unless you absolutely need them at the beginning of the sentence. Writing prefers independents, mostly at the end of the sentence.

Okay, but how do we do it? For the independent clauses, there’s almost nothing to do. Put the adverbial clause after the main one, then put the appropriate conjunction at the end: my syne zejman anyerodyill salmotya byu, “I’ll give you these because I love you”. Since Ardari is otherwise head-final, the simple fact that something follows the verb is a sign that we have an adverbial clause.

Dependents are a little harder, but not much. As with Isian, we need an infinitive verb. For Ardari, it’s the verb stem followed by ky: dyem ky “to buy”, ivit ky “to see”. This goes at the end of the clause, followed by the normal conjunction: my fèse dyem ky chinod, “I went to buy food”.

Of course, there’s a slight problem of ambiguity that could crop up here. Because these clauses appear before the verb, with nothing to mark them off as special, we don’t really know when they start. In practice, though, it’s not that bad. Context helps. (Plus, it’s natural. No language is fully regular and unambiguous.)

Now, knowing all of this, we can get back to adverbial phrases. Ardari handles them like they were a special kind of dependent clause, using the infinitive form of the copula, èll ky: zall èll ky “like this”. (Perhaps in the future, this might evolve into an adverb-making suffix -èlky. Who knows?)

That’s it

Once again, it’s harder to describe something than to put it into action. That was the case with relative clauses a few weeks ago, and it’s the case today. But now we have adverbs, which fills in just about the last box in our list of parts of speech. Almost any kind of statement is possible now.

Next time, we’ll look at questions. Not the kind you certainly have, but the kind speakers of a language will be asking. We’ll see how they’re made and how we can make them.

As we move into Act II of our language-making show, let’s pick up one of the loose threads from last week’s Babel Text: adverbs. When I say “adverb”, though, I’m not just talking about words like English “hardly” and “badly”, but any word or phrase that changes and refines the meaning of a verb. That includes certain phrases that we can call adverbial or subordinate clauses. We’ll see what those are in just a minute, but we should first back up and think about the very idea of an adverb.

The forgotten one

Adverbs, broadly speaking, are to verbs what adjectives are to nouns. They modify the meaning, allowing us to express finer distinctions. Verbs, remember, represent actions, so adverbs are what we use to tell how an action happens. Examples like “she sings happily” or “the clouds are hanging menacingly” show the most familiar of these adverbs.

Of course, there’s always more to the story. Not all adverbs really modify verbs. Some in English, for instance, modify whole sentences. Grammar pedants don’t like it, but that’s what has happened with words such as “hopefully”. And English also has words like “manly” that look like adverbs but fill the role of an adjective.

And then there are languages that don’t actually have a separate collection of adverbs at all. Many of these have no problem allowing an adjective to modify either a noun or a verb; in the latter case, it functions like an adverb, even though there’s no indication that it is one.

(If that weren’t enough, there is another way of defining adverbs: as grammatical words that don’t fit into any other category. That’s a negative definition that isn’t exactly helpful to those making their own languages, but it’s useful to know. Some languages do see adverbs this way, as a closed class of words separate from the other parts of speech, with the more common “adverbs” being derived regularly from adjectives.)

Really making

The largest group of adverbs (or what would be called adverbs) in most languages includes those derived from adjectives and meaning something like “in an X way” or simply “like X”. In English, we can make most of these with the -ly suffix: “real” becomes “really”, etc. Plenty of other languages have their own counterparts that are very similar in use, including Spanish -mente and Japanese -ku, to name only two.

Another common option is, well, nothing at all. Adjectives in many languages can be used directly as adverbs. In these cases, they might not be inflected as usual for case or number (since they won’t be modifying a noun), and they likely won’t appear in their customary position. But those would be the only ways you could tell the difference.

Every other word derivation is possible, too. You can have suffixes, prefixes, extra words before or after, and just about anything else you can think of. For widely-used adverbs, irregularities might arise, especially if the adjective itself already has them. English good is one example, forming the adverb well, to the consternation of schoolchildren everywhere. (The regular goodly also exists, but it has a much different connotation.)

Finally, adverbs aren’t necessarily always derived from adjectives. Words like “soon”, for example, are only adverbial. (It’s not a coincidence that most of the ones you can think of have something to do with time.) It’s perfectly possible to make an adverb from a noun or even a verb, as well. But these probably aren’t going to be made into a simple word; we need a phrase.

As an adverb

When one word just won’t do, adverbial phrases come to the rescue. What are they? Well, it’s right there in the name. An adverbial phrase is nothing more than a phrase that acts like an adverb. (Coincidentally enough, that last sentence perfectly illustrates my point: “like an adverb” is, in fact, an adverbial phrase!)

In English, many adverbial phrases are essentially prepositional phrases used as adverbs. They’re more likely to use “temporal” prepositions like before or when, since those don’t make as much sense for nouns, but anything is possible.

Grammatically speaking, the same is true. If a language allows them, adverbial phrases tend to take the same form as prepositional (or postpositional or whatever) phrases. It’s easy to see why, as the adverbial function only generalizes the idea of prepositions.

Because the clause

The adverbial clause, on the other hand, is a totally different animal. Here, we’re not talking about a little noun phrase, but a whole clause. It could be an entire sentence (an independent clause) or only a fragment unable to stand alone (a subordinate clause). Either way, it also works as an adverb, so it’s a good idea to look at it here.

The key difference between adverbial phrases and clauses is that a clause has a predicate. It’s usually a verb, but some languages only require something with a verbal meaning. (A language with a zero-copula construction, for example, could conceivably have a subordinate clause with only a subject and an object.)

Some of these verbs will be inflected like any other verb in the language. Take, for instance, this English sentence: “It started raining while I was walking home.” The marked part is the adverbial clause, and you can see that, except for the conjunction while, the clause could stand on its own as a sentence.

Now, on the other hand, let’s instead say, “I saw the rain while walking home.” This time, we still have a predicate (walking home), but it can’t stand alone. The special form of the verb, walking instead of walk, is our cue for this.

In English grammar, we call the first example an independent clause, while the second is a dependent one. Some linguists instead refer to them as balanced and deranked clauses, respectively. Either way, the difference between them is clear: one can be “broken out” into its own sentence, while the other can’t.

Counting the ways

Adverbial clauses come in a few different categories. Each has a different meaning and a different set of conjunctions that connect it to the rest of the sentence.

Here are the primary types of clauses, each with a brief definition and an example sentence. We’ll use them later. In the examples, the conjunction that introduces the adverbial clause is emphasized, and the clause itself is everything that follows.

• Purpose: the purpose of an action; “I went home so that I could take a shower”

• Time: when something happens, relative to some other time or event; “the boys played in the sand when they went to the beach”

• Reason: the reason why something happens; “I can’t come because I am sick”

• Place: the position or location of an action or event; “they like it where they live”

• Manner: the way something is done; “this book wasn’t written how I would have liked it”

• Condition: a possibility or consequence, an “if-then” situation; “bad things will happen if you go out in the storm”

There are a couple of others, but they work about the same. Clauses indicating results are similar to those of reason, and concessions are pretty close to conditions. Comparisons are worthy of their own topic, which will come a bit later.

Any of these clauses, though, can be used as adverbs. In English, as you can see above, they often follow the verb, like an object; this isn’t absolutely necessary, and any one of them can be rearranged to put the adverbial clause at the front.

Note, too, that they’re all independent. Taking that away isn’t quite as easy, and it doesn’t always work. It does in some cases, though, as long as the subject of both clauses is the same. We could say, for example, “I went home to take a shower“, creating a dependent clause. Mostly, English prefers “balanced” clauses, to use WALS terminology, permitting “deranked” as an occasional option. (In terms of style, dependent clauses sound slightly more formal or less “personal”, at least to me.)

Constructing the clause

While the general definition of an adverbial clause isn’t that dependent on a specific language, how they’re formed is. For English, as you can see, you first need a conjunction. Then, you have the clause itself. For dependent (or deranked, if you prefer) clauses, the verb appears as either an infinitive or a gerund, depending on what you’re trying to say; either way, it’s not the usual inflected form that you’d use in a “proper” sentence. Independent (balanced, hence the name) clauses have fully inflected verbs, although that isn’t saying much in English.

But how do you do it in a conlang? Well, that truly depends. They’re probably going to look a lot like prepositional phrases, however you do those. Verb-final languages will likely end an adverbial clause with the conjunction, and the clauses themselves will tend to be farther forward in the sentence. SVO or VSO languages would go the other way, more like English.

But this kind of phrase isn’t a core part of a sentence, so there’s nothing to stop it from “floating”. Adverbial clauses can show up anywhere. English allows them at the front, in the back, and even in the middle. Of course, you can be strict, too, if you like. You aren’t going to see many adverbs at the end of a Japanese sentence, after all.

Next up

Next week, we’ll look at how Isian and Ardari tame these monstrous clauses. Then, it’s time to answer something you’ve probably asked once or twice: how do we ask a question?

We’re taking a bit of a week off this time. Don’t worry, there’s more to come later on, but today is the first day of the new year, a time to take a break, a time to reflect on the 366 days to come. (2016 is a leap year, remember.) We’ve gone through ten parts of this series already, and we’ve come a long way. But there’s still a longer way to go, although the pace for this year won’t be quite as hectic.

So, to celebrate this new year of conlanging, here’s what we’re going to do. Today, you’ll get to see the first significant text in each of our two languages, Isian and Ardari. That text is the Babel Text, the first nine verses of Genesis 11. Sure, it’s a religious writing, but that’s okay, because we’re not interested in it for its theology, but for its linguistics.

Babel

The Babel Text is one of those “classic” tests of a conlang. It really has it all, grammatically speaking: tenses, moods, aspects, and all those different kinds of clauses. (Some of them we haven’t seen yet, but we can deal.) Plus, the story itself is about language, the Biblical account of the making of the world’s languages. Essentially, it’s a fable, one originally meant to be told orally. We’ll write it here, though, since that’s easier.

If you want to play along at home, you can use whichever version of the text you like. I’ve gone with this one, derived from the NRSV:

1. Now the whole earth had one language and the same words.

2. And as they migrated from the east, they came upon a plain in the land of Shinar and settled there.

3. And they said to one another, “Come, let us make bricks, and burn them thoroughly.” And they had brick for stone, and bitumen for mortar.

4. Then they said, “Come, let us build ourselves a city, and a tower with its top in the heavens, and let us make a name for ourselves; otherwise we shall be scattered abroad upon the face of the whole earth.”

5. The LORD came down to see the city and the tower, which mortals had built.

6. And the LORD said, “Look, they are one people, and they have all one language; and this is only the beginning of what they will do; nothing that they propose to do will now be impossible for them.

7. Come, let us go down, and confuse their language there, so that they will not understand one another’s speech.”

8. So the LORD scattered them abroad from there over the face of all the earth, and they left off building the city.

9. Therefore it was called Babel, because there the LORD confused the language of all the earth; and from there the LORD scattered them abroad over the face of all the earth.

So that’s what we’re working with. (Yours, of course, may be different.) Now let’s see how it looks in our made-up languages.

Isian

First off, here’s the full text translated into Isian:

1. Nec hi, e sota sata fanas yan sangoy wa es ilir beti.

2. Ad ha is mishas si way keres, is cosas ta abe e Shinar tor i, ar is dalegas.

3. Ad is kis lan es, “Cosa, ad tinte gados, ar becre sota hi sim.” Ar is fanas gadocat sencat afich, wa arcay empan afich.

4. Toc hi, is kis, “Cosa, ad oste lan ir ta eblon, wa ta farin ke ey poy e timirot i, ar tinte ni lan ir. Loydaro mit hade par nos basagima e fayan e sata o sos.”

5. E Domo esto cosas ha i chere e eblon wa e farin ke ostec im nakhit at terta.

6. Ad e Domo kis, “Chere, is e yan cudisa, ar sota sim o fana yan sangoy, ar ne yahi e cu nawe ed ke te im is o. Anocal ke is wachi cu te nec nos e arisosan im ir.

7. Cosa, esto wasa, ar golbata si sangoy til ni, ha is an nos noyta le gonas terta.”

8. Teti, e Domo hade basagis sim til fo e fayan e sata o sos, ar is tarcas cu oste e eblon.

9. Teyno, i par pasa Babel, ha e Domo golbatas til e sangoy e sota sata o todo; ar e Domo basagis til fo e fayan e sata o sos.

Grammar

Most of the grammar is what we’ve already seen. Some of it, though, is new, particularly the adverbs. We’ll get into the details in later parts but here’s a preview that tells you all you need to know to get through the text:

• Isian adjectives (and some nouns) can be changed into adverbs by putting hi after them. This works much like the English suffix -ly: ichi “beautiful” becomes ichi hi “beautifully”.

• Adverbs usually go before their verbs, but they can also be moved to the front of a sentence.

• A whole prepositional phrase can be used like an adverb by putting ha before it.

• Conjunctions are close to their English counterparts. They’re simple little words that link sentences; in the Babel Text, for example, we have ad and ar, both roughly meaning “and”, but with slightly different connotations. Again, we’ll give them a closer look later on.

We get our first tastes of word derivation here, too:

• The suffix -cat, when attached to a noun, changes the meaning to that of a material: gado “a brick”; gadocat “brick”.

• One way to make an adjective into its antonym is by the prefix/suffix pair a-an. We see this in arisosan “impossible”, which is actually derived from risos “possible”.

• Finally, the suffix -nas can convert a verb into a kind of “abstract” noun: go “to speak”; gonas “speech”.

We’ll see a lot more of derivation in later parts of the series.

Vocabulary

Isian	English
abe	a plain
ad	and then…
ar	and (conjunction)
arcay	tar (or bitumen)
arisosan	impossible
basagi	to scatter
bet	word
cudisa	people, group
domo	lord
empan	mortar
esto	down, downward
farin	tower
gado(s)	a brick
gadocat	brick (material)
golbata	to confound
gonas	speech
hade	away
ilir	same
keres	while
loydaro	otherwise
misha	to move
nakhit	mortal
nec	now
noyta	to understand
pasa	to call by a name
sangoy	language
sencat	stone (material)
si	east
tarca	to stop doing
teti	so, because of this
terta	so as to…
teyno	therefore
til	there
timiro	heaven
toc	then
tor	a land
wa	and (noun phrase)
wachi	to desire, wish
yahi	only
yan	one

Ardari

Now, let’s switch to Ardari. Again, here’s the full text:

1. Lokhi omaritö jane kolrachevi sun lagrelltös nyas perodjyn.

2. Ysar sälltö tov tapsined ky vi, Chinare me dablan wi mokiti lim tonedjyn, ysar jeren wizèledjyn.

3. Lataj ry isedjyn, “Tonje, tyolton grätje, ajon warhan sechaje.” Gwanan bòte tyoltanvi, pyuryse bòte pamöre peredjyn.

4. Drä isedjyn, “Tonje, präzdanvi, qa me khaj èlyasòndös wi kombran lataj da mollje. All grätje, sinran omarini sòletö ori oprös utuweryll.”

5. Tsoratö qa sèlokynar molledadyt präzdantövi kombrantö ivit ky tèghdaradjyn.

6. Tsoratö isad, “Ivitje. Ysar jane banöladan èllejyn, ysar me laz jane kolrache perejyn, zalman qa aghell me sòto ky èlla. Duqom qa agh ky märyke ysar da urburdosdill.

7. Tonje, tèghdarje, ysoj kolrache jeren kamrulje, lataj me simënda rejvetell kyus.”

8. Èlladjyn Tsoratö ysar jeren tov omarini sòletö ori utuweradid kyus, ysar präzdantö moll ky uq.

9. Yse Babèle filtyda, Tsoratö omarini kolrache jeren kamruladjyn, jeren tov Tsoratö ysar omarini sòletö ori utuwerad byu.

Grammar

Ardari looks much different, doesn’t it? Much more complicated, too. As with Isian, we haven’t really gone over all the grammar bits you need, so here’s a primer:

• Ardari lets you use most adjectives directly as adverbs, with no changes needed.

• Nouns, noun phrases, and some adjectives instead require you to follow them with èll ky. (This is the infinitive form of èll- “to be”.) An example would be kone èll ky “like a man, manly”.

• “Subordinate” clauses are complicated enough that the full story will have to wait. Some of them let you use a bare verb stem followed by ky, like above, and you use them as a postpositional phrase before a sentence’s head verb. Others appear mostly as normal, but they follow the verb. (This is the only way Ardari lets you put something after the main verb of a sentence.)

Ardari’s words also tend to have more subtle shades of meaning, and these don’t always line up with their English translations:

• nyas means “now”, but only as an adverb
• drä, meaning “then”, connotes a time long in the past
• jeren “there” is used for things very far away; closer things instead use pren
• oprös normally works as an adjective meaning “other”; as an adverb, its meaning becomes “otherwise”
• kyus denotes an effect or implication
• èllad literally means “it was”, but it’s also used to introduce a subordinate clause
• filt- “to know as” is a ditransitive verb, like “to give”

We’ll see conjunctions later on, but we have two here. They’re suffixes, not bare words, so you might not have even noticed. They both mean “and”, but -vi is used for noun phrases, while -jyn is for verbs. To use them, you suffix them to each head word (noun for -vi, verb for -jyn) except the last one.

And then we have a few regular derivations we can point out:

• -ölad (alternate form -ëlad) creates “mass” nouns for substances, collections, and things like that.

• ur- negates adjectives; urburdos “impossible” is the antonym of burdos “possible”.

• -önda (alternate form -ënda) creates abstract nouns from verbs: sim- “to speak”; simënda “speech”.

Vocabulary

Ardari	English
banölad	a people
bòte	instead of
dabla	land
drä	then
èlyas	heaven
filt-	to know as, call by
ghinyas	therefore
gwana	stone
jan	one
jeren	there (far)
kamrul-	to confuse, garble
kolrach	language
kombra	tower
kyus	so that, because of
lagri	word
lòkh	whole
märyk-	to propose, plan
moki	a plain
nyas	now
oprös	other, otherwise
pamör	mortar
pyurys	pitch, tar
rejvet-	to understand
säll	east
sèlokyn	mortal
simënda	the act of speech
sun	same
tapsin-	to migrate
tèghdar-	to descend
tsor(a/i)	a god
tyolta	brick
uq-	to stop doing
urburdos	impossible
utuwe-	to scatter
warhan	thorough
wizèl-	to settle
zhi	thus, in this way

Conclusion

So there you have it: the first full text in both Isian and Ardari. I hope you’re playing along at home, and you’re close to making your own translation of the Babel Text (or whatever you prefer).

Starting in the next part, we’ll be filling in the blanks that I had to leave in here. That should keep us occupied for a while. And then we’ll need some more words.

By this time next year, Isian and Ardari should be radically different. It’s my hope that 2017 will open with something far more…intense. By then, our conlangs will be well on their way to general usability. They won’t be complete, mind you, because when can you say a language is complete?