The 2011 Smiley Award Winner: Okuna

I am very pleased to present the 2011 Smiley Award to Okuna (formerly known as Tokana), an immense and detailed personal language created by Matt Pearson. Okuna has long been recognized as one of the best naturalistic conlangs ever created, and I'm delighted to present it with the Smiley Award. Congratulations, Matt!

Smiley Award 2011

http://pearson.conlang.org/


A Legend

By the time I first joined the online conlanging community back in 2000, everyone already knew about Tokana and revered it. Tokana, in fact, was background knowledge for everyone on the Conlang-L at the time, and so no longer needed to be discussed, only referred to—something I found rather intimidating. Add to that that the grammar itself was somewhat hard to get a hold of (Matt maintained his own copy, and it eventually did get on the web [and it's still there, in fact], but it took a while), and one may be able to sympathize with newcomers who, at the time, had the impression that Tokana was the best conlang ever created, but that no one was allowed to see it. By 2002, Tokana had already achieved legendary status.

And then Matt went and made it better.


Sticking With It

As I discussed before, there are two types of conlangers: those who work on a single language throughout their lives, and those who work on dozens. And while Matt did work briefly on another language (the conlang used in the sci-fi series Dark Skies), for the most part, he's been a one conlang guy. Even so, most monogamous conlangers experience a kind of linguistic epiphany at some point during the life of their language, and they end up radically changing it in some way. Sally Caves did it; Sylvia Sotomayor did it; and so did Matt Pearson. The unique thing about Tokana's transformation into Okuna, though, is that this transformation happened after the language had already attained quite a bit of notoriety amongst conlangers.

Having worked quite a bit on a number of languages, I can say from experience it's the toughest thing in the world to come to a point where one is forced to say, "This isn't working. This needs to change". Many of us give it up (I know I sure did). Some do a total teardown and keep only the name, and maybe a couple words here and there. Matt did something a little different. Matt kept a lot of the original vocabulary of Tokana, as well as a lot of the morphological exponence, but then kind of started over and reimagined the case system.


It Ain't Broke: Let's Fix It

[Note: For a detailed explanation of what I'm describing here, take a look at Matt's LCC1 presentation, or, for a more up-to-date version, Matt's article in Fiat Lingua.]

Tokana used to have a case system that should strike conlangers as quite familiar. It was an ergative-absolutive language, where the ergative marked agents, the absolutive subjects and themes, and the dative marked recipients and goals. Thus, you might get a sentence that looked like this:

  1. Na mikal tsitspè kopo.
    /3SG.ANI.ERG boy break.CPL=3SG.INA.ABS pot/
    "The boy broke the pot."

[Note: In the above example, case is encoded by a preposed article. The article that marks "pot" as absolutive is an enclitic on the verb.]

That's the ordinary way of things with an ergative-absolutive language (after all, "break" is about as transitive a predicate as there is). Then Matt decided to make things less ordinary. Based on a standard ditransitive clause (shown below [but not immediately below]), Matt decided to shake things up and change the way cases were assigned in Tokana.

Sticking with his three core cases (with the absolutive becoming the nominative), Matt revised their definitions as follows:

  1. Ergative: Marks a noun phrase as denoting an event participant (typically animate and volitional) which is the source of an action, or initiates a change of state in some other participant.
  2. Dative: Marks a noun phrase as denoting a participant which specifies the goal or endpoint of a telic event.
  3. Nominative: The "elsewhere" case: Typically marks a noun phrase as denoting a participant which occupies (or comes to occupy) a location, which is transmitted from an agent or source to a goal, or which mediates in some way between an agent or source and a goal.

These definitions were derived from a standard ditransitive sentence (whose case frame didn't change from Tokana to Okuna, bearing in mind that what was the absolutive became the nominative). See below:

  1. Sakialma mikail kytò uktiyi.
    /Sakial.ERG boy.DAT present.NOM give.PRF/
    "Sakial gave the present to the boy."

The idea is that the standard ditransitive clause features an agentive source of action which affects a theme up to a certain point (a goal), and that this schema can be applied to situations other than the standard ditransitive clause. Matt's experiment was to take this schema and what was already present in Okuna and see how far he could run with it—and it turns out he could run with it quite a ways.

The neat thing about Okuna's new system is that in addition to getting the meat and potatoes (direct objects, indirect objects, experiencers, etc.), it also allows Matt to encode subtler shades of meaning without resorting to prolix expressions (e.g. prepositional or adverbial phrases). For example, Okuna also has an allative case. One might expect that it'd be used in translating something like, "The children went to the house". Instead, the dative is used, as the "house" denotes the natural endpoint of "going", in this particular case (or rather, the dative is used to indicate that the "house" is the endpoint of the action). This frees up the allative case for use elsewhere, allowing for this nice distinction:

  1. Lhatè kotoi etyit.
    /children.NOM house.DAT go.PRF.PL/
    "The children went to the house."

  2. Lhatè kotoua etyit.
    /children.NOM house.ALL go.PRF.PL/
    "The children went towards the house."

Thus, if you have this over-arching pattern that your core cases cover, the other cases can be used to mark exceptions to the rule, producing the kind of natural economy we see in natural languages. It also produces some interesting quirks. Take our first sentence up there, for instance. The way it actually comes out in Okuna is shown below:

  1. Mikalma kopoi tsitspyi.
    /boy.ERG pot.DAT break.PRF/
    "The boy broke the pot."

That is, the boy enacted a change of state with the pot, taking it from a state of being whole, to a state of being broken. Since "broken" is the natural endpoint of that change of state, the pot is marked with the dative, rather than the nominative. That frees the nominative up for use elsewhere. So while you can express an instrument with the instrumental case, shown below:

  1. Mikalma kopoi konomme tsitspyi.
    /boy.ERG pot.DAT hammer.INS break.PRF/
    "The boy broke the pot with a/the hammer."

You can also express the instrument with the nominative:

  1. Mikalma kopoi konome tsitspyi.
    /boy.ERG pot.DAT hammer.NOM break.PRF/
    "The boy broke the pot with a/the hammer."

One of my favorite examples features the verb "enter". In a standard "enter" clause, it acts just like an intransitive verb of motion, with the dative denoting the endpoint, but it can also be used ditransitively, in which case it looks just like "give":

  1. Sakiale kotoi lhyuyi.
    /Sakial.NOM house.DAT enter.PRF/
    "Sakial went into the house."

  2. Sakialma keule kotoi lhyuyi.
    /Sakial.ERG chair.NOM house.DAT enter.PRF/
    "Sakial brought the chair into the house."

This one's a nice example of essentially what Matt did throughout the language: He took the pattern from one verb frame and applied it to a different one. Basically, he took what he had, and, without inventing anything new, created a new language.


Some Favorites

There's a bunch of cool stuff scattered throughout the Okuna Grammar and Okuna Lexicon documents. Before getting to that, though, take a look at those documents! How dare you learn to use LaTeX and then put it to such good use, Matt?! What are the rest of us who are too lazy to learn or use LaTeX supposed to do?! Dang, man!

One of my favorite bits of Okuna grammar is the converbial construction. In Okuna, you can use converbs to cover almost everything. Let's take a look at some examples.

First, some background. A converb is a non-finite verb form that combines with another verb to create a kind of complex predicate. That's the basic definition; how it works is language-specific. What's cool about Okuna is that converbs do an awful lot of stuff.

A language like English tends to encode manner either directly on the verb, or as an adverb, e.g.:

  1. "She spoke quickly."

  2. "She trudged into the room." (I.e. "She entered the room and did so trudging.")

In Okuna, such constructions are handled with a converb, as shown below:

  1. Na lotsane lihke itifahpauat.
    /3.ERG branch.NOM cut.CNV PRG.remove.ICPL.IMP.NPL.PL/
    "They are trying to cut the branches off."

  2. Na kiote etsampyiot.
    /3.ERG quick.CNV speak.ACT.PRF.REC.PL/
    "They spoke to one another quickly."

Lot of extra stuff going on in (12), but the point is it's not "They cut the branches off" but "They removed the branches by means of cutting" (or perhaps, "They cuttingly removed the branches"). Anyway, in addition to things like this, converbs take care of other things handled in various ways in English. Take, for example, how the complex predicate "pelt" is teased apart in the following example:

  1. Sakialma sekeit naka solhe ikahtama.
    /Sakial.ERG rat.DAT stone throw.CNV PRG.hit.IMP.DPL/
    "Sakial is pelting the rats with stones."

Literally, that's Sakial is throwing stones, and also hitting the rats with them. In fact, naka, "stone" in (14) is actually incorporated into the verb (which is why it has no case), forming a kind of super predicate with two verbs and a noun (something like "to rock-hit throwingly"). Pretty cool!

And there's more! The same construction accomplishes a lot of what English uses auxiliaries for. Here are a couple examples using "finish" and "keep on" or "continue":

  1. Ma ahotsine nalhe uslyi.
    /1SG.ERG corn.NOM plant.CNV finish.PER/
    "I finished planting the corn."

  2. Sakialma halmai tale sehtyi.
    /Sakial.ERG book.DAT read.CNV go-on.PER/
    "Sakial kept reading his book."

It even works in cases like this one, which I don't even get:

  1. Ohui tat kas iase mu ietuosat ne?
    /fruit those.NOM already eat.CNV enough PRG.REL.ripe.IMP.PL Q/
    "Is that fruit ripe enough to eat yet?"

I guess "ripe enough" is one predicate, and then the converb works to turn "eat" into, like, an adverbial or something...? Anyway, it works, and it's pretty cool!

Of course, my favorite part of any language is the lexicon. The lexicon makes up 99% of a language as it is, so it stands to reason that some of the most exciting stuff would exist exclusively in the lexicon. One of the spots where a conlanger can have some fun is taking regular derivational processes and producing less than regular results (like "emergency" from "emerge"). Here are some fun ones I found that are products of Okuna's derivational morphology.

Okuna has a series of directional nouns formed by adding -ot or -ut to the root. Some of them are pretty straightforward (e.g. sihkasta "to go up stream" > sihkasout "upstream"), but this is also where the cardinal directions come from:

Base Directional Form
Okuna English Okuna English
ahopiau "noon" ahopiaut "south"
koset "evening" kosetot "west, southwest"
kotsim "morning" kotsimot "east, southeast"
ise "snow" iseut "northeast"
suku "wind" sukuot "northwest"
kotsinkis "late morning" kotsinkisot "southeast"
kosehkam "early morning" kosehkamot "southwest"

In case you're wondering, the word for "north" isn't missing, it's just not obviously derivational in nature (it's heut). It looks like it might be derived in the way the others are, but if it is, it would've derived from the copula (a possible literal meaning might be "the place where existence began"...?).

Another neat set comes from the verbalizing suffix -(i)t. Here are some examples:

Base Verbal Form
Okuna English Okuna English
loi "eye, opening" loita "watch, look at"
minap "bone marrow" minahta "be vital, powerful"
tasi "liquid, fluid" tasta "melt, thaw"
uos "shape, form" uosta "create"
his "star" hista "lead, guide"
kun "four" kunta "separate (into groups or piles)"
elia "lithe" elita "relax"

Personally, whenever I think of power, I think of bone marrow. Love that coinage.

I don't want to throw up too many, but some of my favorites come from the -on derivations. This suffix derives abstract verbs from other verbs and some nouns. Here are some examples:

Base Verbal Form
Okuna English Okuna English
eka "be empty" ekona "be hungry"
halhka "be dry" halhkona "be thirsty"
ksohe "darkness" ksohona "be mysterious"
tlota "be heavy, dense" tlotona "be dangerous, unsafe"
niokta "return" niokona "remember"
suha "leave, exit" suhona "forget"
uake "urine" uakona "need to pee"

Heh, heh! That last one is surely something we all know; not sure how many conlangs have a word for it (let alone a derived one).

I could go on (there's plenty to say about Okuna), but I can only go so far without, essentially, recapitulating the grammar in full—and given that the grammar is over 300 pages long, it'd probably serve you better to simply read the grammar. So, without taking any more time...


How Okuna Has Made Me Smile

I can't help but wonder if this example sentence was borne of real life circumstances:

  1. Sù alhme, otieuni seuki tifoksa hialò.
    /rain through.INS garden.DAT weed remove.must.IMP today/
    "In spite of the rain, (we) need to weed the garden today."

Notice how the "we" is unstated. More like the garden needs to get weeded and ain't nobody here but us chickens, if you dig.

Also, just for fun, here's a list of Okuna words that look mighty Finnish, if you ignore the etymologies (and the vowel harmony rules of Finnish):

  • kihuaitanen "carving"
  • namuohtyipa "feasible"
  • elipaimanen "medicine"
  • kylhouanen "container"
  • kysieihpanen "essay"

But what really made me smile was the following, which comes from Matt's Fiat Lingua article:

Interestingly, my attempts to make my system as coherent and consistent as possible have led me to posit various patterns and syncretisms which don’t exist in any natural languages that I know of, but which should exist in some language if my theories about event structure are on the right track.

Ultimately, this is the ideal for a naturalistic conlang. In trying to evaluate the level of naturalism in a conlang, one is often forced to compare it to other natural languages, in which case the entire exercise becomes something akin to a game of BINGO: the conlang grammar calls out a feature, and the evaluator checks it against the list of extant features from natural languages. If the feature is on the list, the conlang is good; if it isn't, it needs work. This method of evaluation is unsatisfactory for many different reasons, not the least of which is that creating a conlang based on the "does x language do it?" method often results in rather unnatural conlangs. After all, while it might be true that some language will do everything found in a given conlang, that doesn't make that conlang naturalistic. So what does?

It seems like the most satisfactory method of naturalistic conlang construction is to do precisely what Matt did with his case system—that is, evolve it. Imagine a set of speakers with a language at time x that need to use what they have to tackle y. The answer is rarely "coin a new word" or "create a new affix". Speakers of natural languages often use what they have at their disposal first before resorting to neologism. And provided the choices are principled (as Matt's choices in evolving Tokana into Okuna were), the result will be naturalistic regardless of whether the particular set of features one ultimately comes up with exist in an extant natural language.

As conlanging itself is still a young art, I don't think we've seen a masterpiece yet. Okuna approaches that threshold, though. We're lucky to have not only an outstanding and comprehensive work like Okuna to serve as inspiration, but also to have a wonderfully thorough grammar to serve as a model for conlang description. Personally, I feel like I've always admired Tokana/Okuna from afar, like a satellite. It was nice to finally jump in and admire firsthand the work and wit that went into it. Congratulations, Matt! Thanks for allowing us to bear witness to the continuing evolution of your incredible language.

Back to the Smiley Award Main Page

This page was last modified on Friday, August 3, 2018.
This website was last modified on .
This page can be viewed normally, as a milk or dark chocolate bar, in sleek black and white, or in many other ways!
All languages, fonts, pictures, and other materials copyright © 2003- David J. Peterson.

free counters