#a name from another language just has the phonology to make it sound plausibly like it could be from this other language

[clannfearrunt]

I’ve created a very stupid problem for myself where the names I’ve given to Rill’s parents are unpleasant 2 say in Japanese (for me) so I need to give them different names in Japanese BUT they can’t be Japanese names because they’re not remotely from Squidcountry so I need to keep their names Not Japanese to signify this, but

[apolesen]

A phonological analysis of Cardassian

@illogicalbroccoli and I decided to sit down and analyse the Cardassian sound-system. This can serve as a guide for coining names and words. We have also included a description of what a Cardassian sounds like. 

Limitations

This analysis is based on the names of all named Cardassians in alpha-canon (i.e. the TV series) (with exceptions of names that are clearly names of the production team that are added in writing just as place-holders), taken from the category ‘Cardassians’ on Memory Alpha. We have also taken into account the names of the sectors of the Union Capitals and the fan-coined names for body parts: chufa, chula, chuva (the spoons), prUt, ajan, vit (genitals). (We have also gone through the Cardassian names in Memory Beta and have considered how they would fit in, but not used them as basis for our analysis.) This is a fairly good sample, but with any sample, some phonemes and combinations of phonemes might not be represented. 

We are treating the way these words are written as Federation transcription. That means that sometimes, we might find inconsistencies in it. We have tried to not invoke transcription errors too often, but at times we have simply to keep the system plausible. 

The final thing is that Cardassian vocal organs are probably different from human ones. Do Cardassians have an uvula, do their teeth look the same, what is the range of movement of their tongue? However, in this post, I am assuming that their vocal organs are close enough to ours that we can use the same terminology. 

Terminology

Stop - a consonant where the air-flow is completely cut off. These are /p, t, k, b, d, g/. If you can say it drawn out, it’s not a stop.

Fricative - another type of consonants where the air-flow is not completely cut off. If it’s a consonant you can say drawn out, it’s a fricative, such as /f/.  

Sibilant - sibilants are strictly a kind of fricative, but are worth mentioning in their own right. They are s-sounds. If it sounds like something a snake might say, it’s a sibilant. 

Affricate - a sound that starts as a stop and ends as a fricative, for instance English chap. 

Cluster - any sounds that commonly occur together. Here I specifically mean consonant clusters, such as /pl/, /gr/, /rm/ etc. 

Phoneme - a distinct sound in a specific language. Phonemes are identified by minimal pairs, words where the only thing that is different are the sounds we are investigating. For instance, English rent and lent are different words, and shows us that /r/ and /l/ are phonemes. 

Allophone - an allophone is how a phoneme presents. A phoneme might manifest differently depending on its position in a word. As with phonemes, there are specific to languages. What is allophonic in one language can be phonemic in another. 

Slashes are used for indicate phonemes, i.e. sounds. Angled brackets indicate graphemes, i.e. written signs. 

I have used some IPA (International Phonetic Alphabet) signs in this post, but have endeavoured to give examples wherever I can. A good resource is the website internationalphoneticalphabet dot org, where you will find a link to “IPA chart with sounds”. You can simply click the sign in question and it reads it to you. 

The Cardassian phonetic system

We are proposing the following system: 

p - t- k - q - v

b - d- g - ɢ - f - θ - ʃ - j - h

ʧ - ʦ

r - l - m- n- ŋ - s - z

a - e - i - ɔ - o - u - ei

In the following sections, I will go through them and explain our reasoning and give examples. 

Stops

Cardassian has a balanced stop system, much like that of English: 

/p/ - Paldar

/t/ - Tain

/k/ - Kelas

/b/ - Belor

/d/ - Damar

/g/ - Garak

However, the voiceless velar /k/ causes us some problems. This sound is sometimes written with <k> (e.g. kanar), but sometimes it is written with <c>, e.g. Broca, Preloc, Crell, Lemec. However, <c> is also used for a sibilant in the names Ocett and Macet. We could put this down to an error in transcription, but that is a dull out. Let us take another piece of data into account - the <gh> in Ghemor. At first glance, this might be an aspirated stop (a sound that occurs allophonically in English and sounds a bit more ‘breathy’), but if so, we would expect more aspirated stops. We could say that the examples of <c> as a velar are aspirated, but it does not ring true to me. Therefore, I suggest that <c> and <gh> represent uvular stops: 

/q/ - Crell

/ɢ/ - Ghemor

The uvula is the thing that hangs down at the back of your mouth. In order to make this sound, you pronounce an ordinary velar (/k/ or /g/) but also bring up the back of your tongue and touch your uvula. These sounds do not occur in English, but does in Arabic and some Caucasian languages. 

It is worth noting that there is one Cardassian name with a <q> in beta-canon, Andrul Taqut. This might be another example of the uvular. 

Fricatives, affricates, liquids and nasals

The Cardassian fricative system is not as well-balanced as its stop system:

/v/ - Revok

/f/ - chufa

/θ/ - Thrax

/ʃ/ - Shoggoth

/j/ - Ziyal

It has both /v/ and /f/, but the latter only appears word-internally (e.g. chufa - there is a beta canon example of Fhret, but this is an odd cluster and I am going to put it aside). There is /θ/, transliterated <th> (like in English thirst), but no voiceless version (like English there). There is also /ʃ/ (English sheer) and /j/, (English yet).

As for affricates, there are two: 

/ʧ/ - chuva

/ʦ/ - Marritza

The sound /ʧ/ is that of English chap and /ʦ/ is Italian pazzo (a /t/ and /s/ sound very close together).  

The nasals, liquids and sibilants are much the same as in English: 

/r/ - Garak

/l/ - gul

/m/ - Mila

/n/ - Nor

/ŋ/ - Lang

/s/ - Silaran

/z/ - Ziyal

What about the times <c> is used for a sibilant? It is noticeable that it only occurs word-internally and between vowels. We have three names with <s> word-internally and between vowels: Rusot, Moset and Russol (assuming the gemination does not show it is something else). Both the first and last cases are flanked by back vowels, so I suggest that they can be disregarded. We are then left with Ocett, Macet and Moset. I will suggest that <c> is used (by a very sloppy transcriber) to present an allophone of /s/ which occurs when the phoneme /s/ follows after a vowel and precedes a /e/. Usually, allophones are not written in different ways, so we would be correct writing these names as Osett and Maset. 

Furthermore, there is a /h/, which often occurs within words. 

Vowels

With the vowels, we must note that there is a problem as English vowels are very messed up. Do we assume that the orthography or the pronunciation is closer to the Cardassian truth? Most of the time, we think the pronunciation is closer, but with some provisions for English lenition. We decided upon this system: 

/a/ - Garak

/e/ - Enabran

/i/ - Ziyal

/ɔ/ - Terok Nor (always transcribed with <o>)

/o/ - Ulani (sometimes transcribed with <o>, sometimes with capital <U>)*

/u/ - Dukat

*) We are assuming an inconsistency of transcription here, simply to allow for the orthography in prUt which does not occur in alpha canon.

There is also one diphthong (or at least one that we can find evidence of): 

/ei/ - Tain **

**) Even if this is written with an <a>, the first part of this diphthong is clearly an /e/. 

Notably, Cardassian does not have a schwa, /ǝ/, which is a sound in the very middle of the mouth like in chocolate. This occurs when an unstressed vowel is weakened to the extent that we don’t bother to put our tongue up or down but simply keep it in a fairly neutral position. Because 

Vowel length is marked, and appears to orthographically be shown by double vowels, as in Boheeka, Aamin. At least once, these are separated by an accent, Darhe’el. The accent also occurs in the name Pa’Dar, where it rather seems to show a shortening of the preceding vowel. I am going to chalk this up to a misunderstanding in the transcription. 

A note on the uses of <h> in beta canon

I have discussed <gh> in Ghemor. However, it is worth mentioning that there are some odd uses of <h> in beta-canon names: Khevet, Efheny, Rhukal, Rhemet, Mhevet. 

Most of these can, in my opinion, be dismissed. Khevet could be an uvular that has been written that way by influence of the transcription of Ghemor. Rhukal and Rhemet are probably examples of the phoneme /r/. It might be an allophonic variation not shown in names such as Rusot and Rugal (perhaps Bajoran doesn’t have this variation and that’s why Rugal’s name is written without it?). As for Efheny, I do not think that the <h> is there to show a change in the /f/, but that there is supposed to be a hiatus between them, Ef-heny. When it comes to Mhevet, I think we need to look at the character in question. Arati Mhevet is from North Torr, which is a poor and close-knit community, and precisely the kind that might have a distinct dialect. I propose that the <mh> represents a labial that is found in the North Torr dialect and not in ‘standard’ Cardassian. 

Permitted clusters

Languages are governed by some internal rules of sounds that can and can’t go together and only in certain positions. For instance, initial /ks/ is generally not permitted in English, which is why words such as xylophone and xenophobic are often pronounced starting with something like /z/. Similarly, English does not allow initial /pt/ and /ps/, which is why words starting with these, e.g. pterodactyl and psychologist, often are pronounced with the <p> silent. 

In Cardassian, we can see some general rules for permitted clusters. 

Clusters consisting of a stop plus /r/ are generally permitted. Liquid plus stop is rare, as are fricative-initial clusters. 

However, this obscures a lot. 

Stop-initial clusters: 

pr, tr, kr, br, dr

kl, gl

gn

-ks#

When it comes to stop plus /r/, all are permitted with the exception of /gr/. In the case of stop plus /l/, only velars, /k/ and /g/, are allowed. The only stop-nasal cluster allowed in /gn/. 

The hash indicates the end of a word; the combination /ks/ is only allowed at the end of words (there is a Cardassian name Thrax). 

Liquid-initial clusters are: 

rb, rd, rt, rl, rn, rm, rv 

ld, lt 

Liquid+stop clusters are generally rare, and /r/-initial clusters are far more common than /l/-initial ones.

Nasal-initial clusters are: 

nd, nt, np, mp

As for clusters starting in fricatives or sibilants, there are only two: 

θr, sk

A Cardassian Accent

What a person sounds like when speaking a language which is not their first language is down to what phonology they are used to. For instance, a native speaker of Hindi may pronounce sounds that are not retroflexed in English as retroflexed, as that is a common sound in Hindi. The allowed clusters in a language also often carries over. 

As we have established what sounds and clusters occur in Cardassian, we can extrapolate what a Cardassian-speaker might sound like if they were speaking English. The amount of accent one has depends on a great number of things - age of second language acquisition, previous bilingualism, general aptitude for language, amount of progress. In my opinion, someone like Garak would not have much of an accent, but people with less experience of other languages and less exposure to native English-speakers will have a more noticeable accent.

As there is no voiceless fricative /ð/, like in English there, so a Cardassian-speaker might say “dere”. The /f/ phoneme is never initial, so while a Cardassian-speaker might be able to pronounce English sofa well, they might struggle with foal and instead say “voal” (which might lead to interesting misunderstandings). They would have trouble with English diphthongs, of which there are a lot. They might also make distinctions that an English speaker would not, such as using an uvular velar in certain positions. 

As the consonant clusters that are allowed in Cardassian are fairly few, a Cardassian-speaker will struggle with some of the clusters in English. Words like star or morsel might cause them trouble. They might add a vowel in between - “satar” - or drop one of the segments “mosel”. 

We have not taken things like accenting patterns and word melody into account, as there is not really any data that allows us to reconstruct it, but that too would change how a Cardassian accent would sound. 

[wrath-of-conlang]

Fantasy Phonoaesthetics 4- Human

Okay, this one requires some explanation.

See, normally if human language is even addressed, it’s presented as common, or the default, and is therefore usually represented as the language the work is written in, so, usually English. If there’s a human conlang at all it’s usually designed to be the least “exotic” sounding of all the languages, and will often be based on English or Latin or another widely spoken/taught Indo-European language.

So I wanted it to sound super fucken’ weird.

I basically just heaped weird phonemes onto it. I thought about what the least common phonemes were and then threw my favorites of those onto it. Dental fricatives, sure, rare vowels, why not. Clicks? Hell yeah.

That’s the really experimental element of this language to me. I’d never studied a language with clicks before (and still to this day haven’t really, besides learning a very basic amount about Xhosa and reading some of a grammar of N|uuki). So honestly, I don’t know how plausible this phonology is, but on the other hand, that’s kind of the point. These are all sounds that exist in human languages so why shouldn’t the default humans have them, right?

(You might be asking, what about ejectives? Well, I can’t pronounce ejectives very well and I wanted to be able to at least say my example sentences out loud. So. Maybe one day I’ll figure out what ejectives sound like and be able to use them but for now that will just have to be a dream.)

Phonological inventory

p b                         k g                           i                        u

f v       θ ð s z ʃ ʒ     x                                ɪ

                     t͡ʃ  d͡ʒ                                   e         ə          o

m            n             ŋ                                æ                 ʌ  ɔ  

w                                                                  a

              ɹ

ʘ      ǀ ! ǂ ǁ             ʔ 

Wow, is this gonna suck to make an orthography for or what?

Orthography

Again, I’m trying to use only ascii characters for ease of typing. (I handwrote the notes for this originally, so it wasn’t an issue until right this second.) Luckily think I had the foresight not to include diphthongs or long vowels, so we can use digraphs for some of the odd vowels.

/æ/ > <ae>      /ɪ/ > <ii>     /e/ > <ee> and /ə/ > <e>     /ʌ/ > <uu>    /ɔ/ > <au>

Kind of ugly but functional. The tricky part is, of course, gonna be wrangling those consonants. Unfortunately, I used some consonant sequences so we’re gonna have to be a little more careful with digraphs here.

But, we can take cues from Xhosa and represent some of those clicks as letters that aren’t used in the normal orthography. Let’s see what we’ve got to work with.

taken - <p b k g f v s z m n w>

It’s pretty obvious to use <r> for the r sound and since there’s no /t d/ we can use <t d> for <θ ð>. Also, if we use <h> for /x/ then <x> is freed up for another sound.

So now we have to represent 11 sounds / ʃ ʒ t͡ʃ d͡ʒ ŋ ʘ ǀ ! ǂ ǁ ʔ / with six letters, < q y j l x c >. So we’ll probably need SOME digraphs. Of course, if we want we can also dip into punctuation, numbers, and even capital letters if we feel like going the Klingon route (which I am not opposed to, but can get complicated with some kinds of formatting).

A lowercase L looks the same as a <|> and there’s no /l/ in the language, so let’s use that. /ʒ/ can be <j>, /c/ can be /ʃ/, and since there’s no gemination in the language /t͡ʃ d͡ʒ/ can be <cc jj>. <ng> might as well be /ŋ/. <’> is usally used for a glottal stop, but I like <q> for that position, so <’> can be reused as <ʘ>. <x> can be /ǂ/, and y can be /ǁ/, and ! can stay as it is.

Got that? If not, don’t worry, here it is again.

Clicks:

ʘ labial click ‘         ǀ dental click l              ! alveolar click !

ǂ palatal click x        ǁ lateral click y           ʔ glottal stop q

Other special consonants:

θ - t          ð - d          ʃ - c      ʒ - j        t͡ʃ  - cc       d͡ʒ - jj

It’s convoluted but I tried.

Grammar

Nouns and pronouns

This language has a case system, because I guess I decided to start hating myself somewhere past the halfway point of this project. I just lifted the case system straight out of Latin, too, but I think I should get a pass considering the extreme stress the phonology is causing me.

Pronouns decline according to the following table.

                      singular                      plural

            1           2           3            1           2               3

nom    ta          veem       ii           tacc      veemacc   iicc

acc    tae         veemi       i          tacc     veemicc      icc

gen   tad         veemad     iid        tade    veemade    ide

dat   tang       veeng        iing      tange   veenge       iinge

abl   ya          veyim         yi         yacc      veeyimiicc  yicc

This orthography is a monster, I’ve created a monster.

Nouns have three declensions. The first declension is for words ending in -a.

                        singular                      plural

nom                     -a                             -acc

acc                      -ae                           -aecc

gen                     -ad                           -ade

dat                      -ang                        -ange

abl                     -ya                           -yacc

The second declension is for nouns ending in -ii.

                        singular                    plural

nom                 -ii                               -iicc

acc                  -i                               -iicc

gen                  -iid                           -iide

dat                  -iing                         -iinge

abl                  -yi                             -yicc

The third declension is for nouns ending in everything else. The schwa in parentheses is used as needed to break up consonant clusters.

                     singular                          plural

nom               -(e)m                             -(e)macc

acc                -(e)mi                             -(e)micc

gen               -(e)mad                           -(e)made

dat                -(e)ng                              -(e)nge

abl                -yim                                 -yimiicc

Nouns can be negated with the prefix bo!-/bo-

Verbs

I modeled verbs after Japanese again. They have four forms.

Base: -------          Negative: -ba

Past:  -ux              Past negative: -uxbe

Multiple negation is generally required.

There’s also a system of verb prefixes that can be added along with the suffix.

yiyi-: to be glad of the verb

fute-: to be unhappy of the verb

In context, these can be understood as imperative forms, and can be translated as something like, “I would be happy if you did this” or “I would be unhappy if you did this”.

Adjectives

Adjectives take the same ending as the noun or pronoun they modify.

Vocab

be- jjung  love- fa!qga  eat- reelong  see- zog  help- wa’u

permit/allow- !uw  think- zaqa

home- zeeba  friend- hoa  water- ciltii  tree- ngaufii  name- xura  dog- ccixa

human- tuja  orc- gii!a  elf- gabra  dwarf- sonuli  goblin- yida

good- yi-  bad- fut-  tall- ‘ek-  short- ving-  hot- daeq  cold- !uang- 

blue- nguung-  desperate- yomii!-

goodbye/peace- fua

or- yoyo  again- huvam

Example sentences

Jjung xura tad ___.                        Veemi ta fa!qga.

be name.NOM 1SPRON.GEN        2SPRON.ACC 1SPRON.NOM love

My name is ____.                          I love you.

Ta zeebang jjung.                        Ta hoae zogux.

1SPRON.NOM HOME.DAT be      1SPRON.NOM friend.ACC see.PAST

I’m at home.                                 I saw my friend.

Ta bongauficc reelongba.                            Ta tuja jjung.

1SPRON.NOM NEG.tree.ACC eat.NEG        1SPRON.NOM human.NOM be

I don’t eat trees.                                          I am a human.

Ta ‘eki ngaufi zog.                                      Ccixa nguunga jjung.

1SPRON.NOM tall.ACC tree.ACC see        dog.NOM blue.NOM be

I see the tall tree.                                      The dog is blue.

Yiyizog.                                                    Futezog.

POS-REQ.see                                         NEG-REQ.see

Look!                                                       Don’t look.

Ta futewa’uba, ta yomii!a jjung.

1SPRON.NOM NEG-REQ.help.NEG, 1SPRON.NOM desperate.NOM be

Help me, I am desperate. (The negation on the verb has a connotation of “if you don’t help me, bad things will happen”.)

Jjiim...fua, tad...tad ccehaila. Huvam veeyim yoyo xuraya veemad ta !uwba zagaba tae.

Jim...peace, 1SPRON.GEN...1SPRON.GEN t’hy’la.NOM. again 2SPRON.ABL or name.ABL 2SPRON.GEN 1SPRON.NOM think.NEG allow.NEG 1SPRON.ACC

Jim...Goodbye, my...my t’hy’la. I won’t permit myself to think of you or even your name again.

If you were paying very close attention, you’ll notice that in loanwords with L, the L becomes a dental click, not an R. I don’t know why I did that, but I honestly kind of like it.

Last one is goblins, and if you thought this orthography was a mess, wait til you see me trying to figure out how to represent a tonal system with ascii characters.