welcome to Conlang Critic, the show that gets
facts wrong about YOUR favorite conlang! I’m jan Misali, and in this episode, we’ll be
looking at the verbless artlang, Kēlen. when I first heard the concept of Kēlen,
I was immediately sceptical. no verbs? that can’t be right. how would that even work?
surely, what Kēlen does is relabel its verbs in a way that disguises their function, or
maybe there’s just a large set of things that you can’t say in Kēlen. right? Kēlen is a fictional language created by
Sylvia Sotomayor, who has been working on it since 1980. at the time of writing, the
most recent update to it was made in June 2019. in universe, it’s a non-human language,
spoken by elves called the Kēleñi. there is definitely some external lore associated
with Kēlen, but the language largely exists as a standalone work. Kēlen’s consonants are: /m n ɲ ŋ/
/p t s c k/ /ɸ θ ʃ ç x/
/l ʎ/ /r/ okay, I really like this inventory. it’s
not especially unique or anything, but it’s well balanced in a way where everything definitely
fits together nicely. big fan of the palatal series and that labial fricative. now you might be wondering about the placement
of /s/ there in the stop row, which is definitely not a stop consonant. Sylvia explains that
in universe it was historically an affricate, [ts], that ended up shifting into a fricative
for most speakers. it makes sense to still consider it a stop due to the way it behaves
phonotactically. stops and fricatives are phonemically voiceless,
but they are sometimes voiced, specifically when one comes between two phonemically voiced
sounds. the difference between how they behave is that some speakers pronounce fricatives
as voiced in all contexts, which doesn’t apply to /s/. this is, I think, enough to justify the apparent
mislabeling. that and the fact that it makes the chart narrower. most of the resonants here also have geminated
forms listed, specifically all the nasals and all the liquids except the palatal lateral. I don’t think it’s strictly necessary
to analyze these as separate phonemes, for reasons I will explain in a bit. Kēlen’s vowels are: /i u/
/iː uː/ /e o/
/eː oː/ /a/
/aː/ it’s just the five vowel system plus length,
pretty basic. there also exists a marginally phonemic close central vowel, only appearing
in some dialects, which lets me rearrange this whole dang chart. there’s also a few diphthongs, which are: /je/
/jeː/ /aj aw/
/aːj aːw/ Sylvia doesn’t explicitly define Kēlen’s
syllable structure on her website any further than “consonant-vowel-consonant”. however,
I was able to get a feel for what stuff is allowed just by looking at the canon wordlist,
which leads to the main reason I don’t think it makes sense to call the double resonants
separate phonemes: there are plenty of examples in the wordlist of doubled versions of other
consonants. so I think the more sensible analysis is to
say that the doubled consonants are just a consonant appearing in the coda of one syllable,
then appearing again in the onset of the next syllable, especially since there are no examples
of any double consonant appearing anywhere other than between two vowels. though I guess to get a full understanding
of the phonotactics, I’d have to check every single possible pair of consonants to see
if they’re attested in the wordlist, and that sounds like a lot of work, I don’t
really want to do that. I mean, I’d have to check like, what, eight nif pairs of phonemes?
yeah, I’m not doing that. ... so here’s a table showing which pairs of
consonants exist in words in the canon dictionary. I guess I should say that just because there
aren’t any attested words that have a specific cluster doesn’t necessarily mean that that
cluster isn’t allowed. it’s instead more useful to look at the big picture. even though only seven consonants appear word
finally, all but one appear syllable finally, the exception being the palatal lateral. granted,
that’s only if you consider the geminate consonants to be one consonant twice in a
row rather than phonemes in their own right. let’s remove all the rows for consonants
where the only time they appear in the coda is when they’re geminated. looking at it like this, you can see that
some types of consonant clusters completely don’t exist: there are no sequences of a
stop followed by a fricative, or two different fricatives in a row. the only stop-stop sequences
that exist are ones where the first stop is /s/, which is definitely making me kinda regret
going along with the analysis that /s/ is a stop. oh well. I think Kēlen’s phonology is pretty good,
even though it obviously wasn’t the main focus. it definitely serves the language well,
and has a nice aesthetic. before I get into Kēlen’s orthographies,
I’m gonna start with the Romanization. really the only thing I don’t like here
is the use of <w> for /ɸ/. it does work for the common voiced allophone [β], but even
then <v> would be better for it than <w>. everything else makes sense. I do, however, have a super nitpicky complaint
about the way it’s presented. on her website, Sylvia, for some reason, uses slashes to mark
the Romanization? and she does the same thing when giving examples from other languages.
this caused seconds of confusion where I assumed based on convention that this was supposed
to be phonemic transcription. I assume this is a holdover from some web 1.0 restriction,
but like, come on, slashes are for phonemic transcription, and orthography should use
angle brackets. anyway, Kēlen has two orthographies. the
main one is an alphabet that looks like this. the line across the top of everything makes
it look a whole lot like Devanagari, but functionally it’s just an alphabet. its design is somewhat
featural. for example, fricatives are written with letters that look like their corresponding
stops but with an extra line added to the right, and there’s a few place-of-articulation
diacritics used consistently for nasals and liquids. you might have noticed that on this chart
that there’s apparently a letter for some sort of palatal trill. that is, of course,
the letter for /r/ with the same palatal diacritic used for writing the two palatal resonants.
since there isn’t an actual palatal trill phoneme, it’s used for the cluster /rç/. the place-of-articulation diacritics are also
used for writing diphthongs, which is pretty neat. this alphabet is solid, but it’s easily
the less interesting of Kēlen’s two writing systems. Kēlen’s other orthography is its Ceremonial
Interlace Alphabet, which as the name implies is an alphabet used for ceremonial purposes
with an interlace aesthetic. before going over how it works, I’d just
like to say how beautiful this is! I really like interlace as a decorative style, and
using it for a writing system is such a cool idea! everything ends up looking like ribbons
knotted together in these really cool patterns. now, since this is meant to be used more decoratively
than for actual writing, it can be kinda hard to tell what’s actually going on when looking
at stuff written in the Ceremonial Interlace Alphabet, which is completely by design. I
mean, this doesn’t even look like writing! this design actually says taxōnīkōn, which
means “conlanger”. since the Ceremonial Interlace Alphabet is
completely decorative and not meant to be used for writing lengthy texts, long vowels
are written the same as short vowels. now, if you look closely at this design, you’ll
see that most of it is one long continuous band, and the way that band is knotted is
what spells out the letters for T-A-X-O-N-I-K-O-N, and then just for good measure it loops back
on itself. by the way, when I was making this video’s
thumbnail, I made the bands thinner to make it easier to appreciate the braided structure.
this is supposed to say “Kēlen”. I guess I don’t have that much to say about
it, but dang I think this might be my new favorite writing system, it just looks so
good! from October 2009 to December 2011, Sylvia
Sotomayor did a thing on her blog called “Kēlen Word of the Day”, where every day she highlighted
a specific Kēlen word. this way of presenting vocabulary one word at a time in a context
that’s separate from the actual documentation gives a little bit more structure than just
reading through the dictionary, partially sorted by category and partially walking through
translations of different texts. there were a lot of highlights that I took
screenshots of as I read through it, and I don’t really have that much to say about
any of them individually other than “hey this is neat” so right now I’m just padding
out the script so that they’re not going by too fast for you to see them. as a whole, Kēlen vocabulary definitely works
to show who the Kēleñi are and what their culture is like, which is of course by necessity,
as there isn’t a full work of fiction to go with it. even though the Kēleñi aren’t
human, they aren’t really alien. their culture definitely seems like it would fit in just
fine on Earth. the only real indication in the vocabulary
that this isn’t supposed to be a human language are the words that refer to things that don’t
exist, but even then they’re usually directly compared to something familiar. not that I’m
complaining, of course. the worldbuilding on display in Kēlen’s vocabulary is very
impressive. anyway, pronouns. the four-way grammatical number distinction
made with the pronouns doesn’t actually appear anywhere else; only personal pronouns
have separate dual and paucal forms. also, the reduced forms there are used as nonspecific
abbreviations of the other pronouns regardless of number, and they can also be used as relative
pronouns. oh also, Kēlen’s numbering system uses
octal, base eight. fun! we now arrive at the main event, the grammar.
the whole idea of Kēlen is that it doesn’t have verbs. how does that even work? well,
one thing at a time. nouns in Kēlen are given prefixes that mark
animacy. this is not inherent to a given noun, and in fact weather something counts as animate
or inanimate can vary from speaker to speaker. ranging from only referring to one’s own
kin as animate to referring to literally everything as animate. the “possessed” prefixes are used for
when something is a part of somebody, usually referring to body parts, but it’s sometimes
used for other things. as briefly mentioned before, there’s also
grammatical number. the plural is marked with the -i suffix, or -ien for animate nouns. okay, that’s enough about that. now let’s
talk about how Kēlen works without verbs. so, what Kēlen has instead of verbs is a
set of four words called “relationals”, which are really just verbs. so, exactly as
I suspected, Kēlen just relabels its verbs as something else. that’s not really fair though. Kēlen’s
relationals work like verbs functionally, but they don’t carry any semantic meaning.
and like, the fact that there’s only four of them still shouldn’t just be written
off. there definitely is truth to the claim that Kēlen is a verbless language. the first relational is la, the “relational
of existence”. it roughly corresponds to the verb “to be”. it’s used for saying
that something exists or that two things are the same. you know, general copula stuff. there’s
also a past tense te and a negative wa, which count as being the same relational I guess. the second relational is ñi, the “relational
of change”. it’s like la, but it’s used for change of state instead of static state. “causing something to become something else”
is a really general concept that encompasses a lot of things that you’d normally use
separate verbs for. the way it works actually reminds me a lot of how transitive verbs work
in Toki Pona, my favorite conlang. rather than having a designated verb for something
like “melt”, you instead use an adjective and put some particles around it to talk about
a change of state; instead of saying that heat melts ice, you say that it causes ice
to become melted. it’s a really useful concept that you can get a lot of mileage out of. the relational ñi can also be inflected for
agent, replacing a pronoun that normally would be necessary. the third relational is se, the “transactional
relational”. it roughly corresponds to the verb “to give”. it’s used for talking
about giving and taking things. it’s also used for talking about talking, in the sense
that speaking is “giving words”, and it’s also used for introducing a topic. just like how ñi can be inflected for agent,
se can be inflected for “source” and “beneficiary”. the resulting table is definitely a lot to
take in, but you can still tell that it’s not just completely arbitrarily assigning
things. a word like sexelte fits in with the other inflections of se with the same source
and with the same beneficiary. the final relational is pa, which doesn’t
have a fun descriptive name. it roughly corresponds to the verb “to have”. it’s actually
an extension of something that can be used with la. as you can probably guess, it’s
used for saying that something has something, both for actual possession and for when something
has an abstract property. there’s really not that much to this one,
but it is still just as versatile as the others. looking at this system as a whole, with these
four relationals, I can’t help but think that this could’ve worked with fewer than
four. I really don’t think they’re all strictly necessary. the relational of change
could’ve been merged into the relational of existence with creative use of word order. I mean, earlier I compared the relational
of change to how transitive verbs in Toki Pona work, and in Toki Pona, the stuff handled
by Kēlen’s relationals of existence and change are both covered by the same grammatical
particle. I also think the transactional relational
and the one without a nickname could be merged into one thing by way of the relational of
change. I mean, in a sense, giving someone something is the same as causing them to have
it, right? that said, I don’t think having the bare
minimum number of these relationals was what Sylvia had in mind. I mean, Kēlen isn’t
really an engineered language, it it. it definitely has elements of an engelang, but at heart
it’s a fictional language and an art language. in a sense, Kēlen actually has a lot in common
with Drsk. they’re both fictional languages that have elements that fit more with the
design philosophy of an engineered language. it’s just that Kēlen does a much, much
better job at it. the following text is an excerpt from Aesop’s
The North Wind and the Sun, translated into Kēlen by Sylvia Sotomayor. se mūrāna masīrien; se malō;
iēlte teteñ anþēŋŋeni ien la ma pa antāken anānexa makēñ;
il aþ ñi marāona nīkamma jalūra jacālle rājanō;
teteñ anñēīki ien la ma pa antāken anānexa ma ñamma jalūra rū marāona pēxa cēja;
il ñamma anūri nā ā mūrāna masīrien il ñamma jalūra rā sāen āñ nō ā marāona; all in all, I like Kēlen a lot. from the
aversions of linguistic universals to the piecemeal worldbuilding to this divination
thing to my favorite writing system for any conlang, I could feel the passion and dedication
that Sylvia Sotomayor put into it. I’m not sure if it’s my favorite fictional language,
but it’s definitely in my top three. Kēlen is a wonderful conlang that I’m glad I had
the chance to take an in depth look at. thanks for watching, and I’ll see you next
time, where I’ll be reviewing Lingwa de Planeta. o tawa waso!
o pali e ale! sina ken, tawa mi!
mi ken kute e kalama sina
tan insa pilin mi hey everyone. in case you didn’t notice,
I changed the name of this channel to “jan Misali”. if you’re wondering why, I made
a short video announcing the change. Conlang Critic isn’t going anywhere. that said,
I do actually have one big video planned to go between this video and the upcoming Lingwa
de Planeta episode. I’m unreasonably excited about this one. oh, I should do a big announcement!
yeah, that’ll be cool.
