Interesting features in your conlang

[steewi]

Inspired by the 4th tense in Lojban thread.

If you have your own conlang, what features have you encoded (or want to encode) in it that you don't find in many languages?

For those without conlangs, what features or distinctions would you like to see used in English?

For example, I like the idea of having morphemes that identify the origin of the information presented - for example, end the sentence with "yan" if you saw it happen, "kun" if you heard it happen, "man" if someone told you they saw it happen and "pun" if you heard a rumour that it happened.

A conlang I always wanted to make would encode cardinal directions in four dimensions (not all mathematical ones, though). You'd have North, East, South and West, but also Up and Down would be treated the same way. The fourth dimension is a perceived moral dimension, so in the same way you'd say someone went south, you could also say they "went bad-wise" or "went good-wise". This distinction would play havoc when they tried to do multidimensional mathematics, but that's beside the point, its speakers are pre-industrial. The distinction is not particularly important, because we essentially say it already in English - you can say "Harry went bad," and so on - but the way they think about the deixis is different. That part is definitely a work in progress.

For more ideas, see Zompist's Language Construction Kit, or share your own.

[Dingbats]

One thing that's pretty nifty in my conlang is that it's active-stative, but taken even further. So the two primary cases are agentive and patientive. In a normal active lang, the direct objects would always be marked with the patientive, and the subjects could be either agentive or patientive depending on control or volition. In my lang, the objects too can be marked with either case. So you get:

Dog-AG sees cat-PAT
"The dog sees the cat" (voluntarily, that is "looks at")

Dog-PT sees cat-PT
"The dog sees the cat" (unvoluntarily, translated as "sees")

Dog-AG sees cat-AG
"The dog sees the cat and the cat sees the dog" (reciprocal, but with most of the focus on the dog)

Dog-PT sees cat-AG
"The dog sees the cat" (unvoluntarily, and the cat is the cause of the action taking place, so more or less a causative)

That's one of the major things. There are lots of little neat details everywhere too.

[steewi]

An interesting mixture of case and focus, marked by word order *and* case marking. That's a pretty cool feature. It would drive me batty for a while until I got used to it, but that doesn't make it not worthwhile.

[ZLVT]

a) "we" would have enough forms to ensure that everyone knows what combination of people is being reffered to

b) Tenses would describe when the time in question took place and how the action in question related to it e.g.
up to and including [that moment]
up to but not including [that moment]
up to including and after [that moment]
starting from and after [that moment]
around [that moment] but not at [that moment]

[ave_matthew]

Dingbats, that rocks.

Mine just has a chitload of cases, and I keep adding more
I do like my verbs though.
Take to love (hmeantio):

Spoiler:

Tenses: present(hmenatiomo) past(hmenatiom) futur(hmenatimo)
each tense can be modified with +'la or +'lo to show it hapened just before or after respectivly
eg. hmentiomo'lo => I'm just about to love
Aspects :
Imperfect(3)

hmeantiémo: that which is ending, but which has been happening since before I remember

hmeantiamo: that which is ending and that I remember it's begining

hmeantiìmo: that which just recently began, and which is ending

habitual(2)

hmenatiõ: that which I do habitualy but know that one day I will stop doing

hmenatiõo: that which I do habitualy and am not certain if I shall ever stop

eternal(1)

hmeantiõa: that which started with or without my knowing, and which will never end

there is also a conditional declenation and a desiritive one as well.
There are a whole bunch of combining adfixes.

and the other thing I like is that there are two "forms" of the genetive case (they probably have names I don't know):
+é
shows that you OWN the thing, and that it will obey you and that you control it.(my camera).
+so
shows that well it is related to you in some way the shows reciprocal possesion, neither of the pair really own the other. (my mother).
the main difference is that +é is for inanimate things and +so for animate things, but not strictly.
eg. joré(my) oma(animate antecedant pronoun) -> my slave. jorso kufa -> my dog(the one that let do whatever it wants)
although dogs are animate, work animals take +é

For example, I like the idea of having morphemes that identify the origin of the information presented - for example, end the sentence with "yan" if you saw it happen, "kun" if you heard it happen, "man" if someone told you they saw it happen and "pun" if you heard a rumour that it happened.

I have a adfix that shows that it's your opinion but that in your opinion there is no other option or reality.

Spoiler:

a sample of jorain

fishprayert.jpg (24.83 KiB) Viewed 6870 times

[ave_matthew]

I have a neat idea, why don't we also post sound clips with accompanying transliteration. **assuming that this is possible**
can the site do that?

[Kizyr]

The most distinguishing feature in the conlang I'm playing around with is that it has no consonants. It's entirely comprised of vowel sounds. Now, I have a more liberal idea of vowels (i.e., as long as there's no friction between two parts of the mouth, so there's a sound similar to an American English "R"), and there are 3 tones. But, between tones and diphthongs, and regular vowels, I can get between 100 and 200 distinct sounds.

Because of how it's constructed, the language itself is based on a syllabary. I'm modeling the writing style like Hangul, in that each syllable is its own character, but the character itself is comprised of its phonetic components (so diphthongs will be a combination of the base vowels). But I'm working on how to make it so that the writing style is easier to write and distinguish for reading. Only problem is that I have an aesthetic preference for geometric shapes.

Vocabulary and grammar are far less interesting. I'm mostly toying with this idea to see if I can make a language without consonants.

For example, I like the idea of having morphemes that identify the origin of the information presented - for example, end the sentence with "yan" if you saw it happen, "kun" if you heard it happen, "man" if someone told you they saw it happen and "pun" if you heard a rumour that it happened.

Interestingly enough, Japanese already has similar distinctions. I'm not sure if you're intentionally inheriting the idea from there, or if it's just coincidence.
E.G.,
起きる = happens
起きそう = looks like it happen(ed)
起きるそう = heard that it happens (hearsay based)
起きるよう = heard that it happens (evidence based)

And of course, you can explicitly say things like "I heard that..." or "It's said that...". While you can do that in English anyway, it's much more common in Japanese speech by comparison. KF

[steewi]

Oh, I know it occurs in languages. It can be surprisingly difficult to come up with something completely original for a conlang when it comes to marking interesting bits. There are whole language families in the Amazon, so I've heard [citation needed] where it's obligatory to mark the origin of the information you're stating.

Another one imported from Australian languages (and some Papuan languages) that I like is similar to one mentioned above, where the tense marking shows the distance in the past/future, thus a verb can be marked as:
recent past (in the last day)
past (in the last week or so)
distant past (ages ago)

and similar for future:
near future (I'm going to do it soon)
future (I'm going to do it)
distant future (I have plans to do it, but not for some time)

That's not the exact way that Australian languages do it, but that's the way I'd set it up if I were incorporating the idea in a language.

Edit:

Oh, and the writing system I use has incorporated punctuation that signifies a sort of irrealis (sarcasm, light-hearted humour, etc. Anything that's not meant to be taken literally). I came up with it before I got into the habit of using emoticons, and it looks better than having a smiley face in the middle of a bunch of text.The writing system is boustrophedon, but otherwise boringly alphabetic.

[BrainMagMo]

Even better is the Hodiernal tense, which indicates that something happens today.
+Hodiernal present and Hodiernal past, for what it seems to mean.

I wanted both a dynamic v. static AND a telic v. atelic distinction.
However, I can't POSSIBLY think of what a telic static aspect would mean, or how it could come about. Perhaps I just don't understand dynamicy enough.

[Cryopyre]

This is AWESOME.

I must construct my own language.

[Dingbats]

This is AWESOME.

I must construct my own language.

It's very tedious. There's so much of it that's just stuff that has to be considered but isn't that interesting, but if you focus only the juicy stuff the lang will never be usable.

[Cryopyre]

what's so tedious, one can make a simple language quite easily.

I already have one forming in my mind.

[Qoppa]

what's so tedious, one can make a simple language quite easily.

I already have one forming in my mind.

A language, not a cipher.

Read Zompist's Language Construction Kit, internalize it all, and then read it again. You need a phoneme inventory to start with, and then some allophony to get a phonology. Then you need to consider the morphology (for nouns, adjectives, verbs, adverbs, etc) and syntax of your language. You need to figure out word order and how phrases are formed, how tense, aspect, plurality, gender, case, evidentiality, etc. are used in your language (if at all). Then you need to build up the vocabulary, which generally means creating some sort of derivational morphology to go along with your already extensive inflectional morphology. ... and so on. In short, you'll need to know/learn quite a bit about linguistics.

I'm not trying to discourage you, but you should keep in mind that unless you want to make a boring English cipher, there's a lot to consider. Creating something interesting that isn't just a cheap rip-off of English or another language is a lot of work, but also very rewarding. I post on the Zompist bboards, and we see a lot of really crappy n00b conlangs which are underlyingly just English with a couple of things changed. Creating something that is unique in its own right is quite a challenge.

[steewi]

what's so tedious, one can make a simple language quite easily.

I already have one forming in my mind.

A language, not a cipher.

Read Zompist's Language Construction Kit, internalize it all, and then read it again. You need a phoneme inventory to start with, and then some allophony to get a phonology. Then you need to consider the morphology (for nouns, adjectives, verbs, adverbs, etc) and syntax of your language. You need to figure out word order and how phrases are formed, how tense, aspect, plurality, gender, case, evidentiality, etc. are used in your language (if at all). Then you need to build up the vocabulary, which generally means creating some sort of derivational morphology to go along with your already extensive inflectional morphology. ... and so on. In short, you'll need to know/learn quite a bit about linguistics.

I'm not trying to discourage you, but you should keep in mind that unless you want to make a boring English cipher, there's a lot to consider. Creating something interesting that isn't just a cheap rip-off of English or another language is a lot of work, but also very rewarding. I post on the Zompist bboards, and we see a lot of really crappy n00b conlangs which are underlyingly just English with a couple of things changed. Creating something that is unique in its own right is quite a challenge.

Of course, sometimes a rip-off language can be fun as well. It depends on your motivation and goals. But I do understand what you mean. Definitely.

[Cryopyre]

Hmmm...

Well I'm not a moron, but picking up a book for a weekend's project would probably not be worth my time.

Now for the "cipher's" gimmick, one word clauses.

[ZLVT]

Of course, sometimes a rip-off language can be fun as well. It depends on your motivation and goals. But I do understand what you mean. Definitely.

Aye, I once began to make a simplified completely regular English with the intent that it would be close enough for natives to understadn but simple and regular enough for foreigners to learn. It's a lingua franca, why not help it along? But I got bored and had things to do. One day though...

[ave_matthew]

just in case anyone would be interested, there is a collaborative conlang project going on right now via the conlang mailing list.
it can be found at the following website, the goal is to make the language, using only the language.
http://fiziwig.com/tak/

[stolid]

How many of you actually created a language or instead just forked and modified one? I am thinking about just forking one because creating one from scratch seems a little ridiculous.

[steewi]

How many of you actually created a language or instead just forked and modified one? I am thinking about just forking one because creating one from scratch seems a little ridiculous.

Pretty much all I've come up with is modified/forked natlangs with interesting or experimental features. My reasoning is that when I come up with my own word-roots and structures, they tend to be very unnatural seeming. That's good if you're making them for aliens, but for a human culture, it's useful to have a natlang base to work off, even if you take it in interesting new directions. Disclaimer: Personal opinion. YMMV.

[fennecfanatic]

My conlang, Narkhokul, does not distinguish consonants on voicing. The inspiration for this came from Toki Pona.

The consonants are: t/d, p/b, s/z, k/c/g, th/dh, ph/f/v, sh/zh, kh/ch/x/gh/q, r, l, m, n
(I chose the representations th/ph/sh/kh to mirror t/p/s/k.)
Vowels are i, e, a, o, u

Words do have preferred pronunciations, though. The word /draphen/, for example, has the canonical pronunciation with voiced /d/ and unvoiced /ph/ but other pronunciations (/trafen/, /traven/, or /draven/) are simply less common, somewhat like the different acceptable pronunciations of American English "either", "neither", and "vase". (I can't speak to whether these are acceptable in other variants of English, as I'm just not that familiar with them.)

[MHD]

I'm highly inspired by the partially defined language of The Old Gods (knwo as R'lyehian) that H. P. Lovecraft wrote about. For my language I use an automatic word generator following certain rules about how words are composed, currently I lack a good way of representing the phonetical letters.
It uses postfix and prefix for anything from grammatical cases to verb tenses and makes no effort to distinguish word classes and so on. I plan to make the language of the gods a real one and we shall all revel in the "not-present" and "possibly" verb tenses, muhahah!

[Link]

I haven't actually written anything down, but if I were to spend time on that, my conlang would feature the following:

• Each letter is associated with one phoneme and vice-versa. That means if you read a word, you can immediately pronounce it correctly, and if you hear a word, you can write it correctly.
• Stress in syllables is not particularly important, and may vary depending on context, convention and pronounceability.
• In writing, "stretched" sounds (like "aaaaaaaah!" when somebody is falling or Gollum's famous "preciousssssss") are denoted with a dot; the length of the "stretching" depends on the size of the dot relative to the character size.
• Capitals are completely absent; names will have a prefix or suffix (haven't decided yet), and the start of a sentence is indicated simply by the end of the previous one (if any).

I imagine the alphabet would be something like this (in the format of "grapheme [ IPA ]"):
a [ a ] á [ aː ] b [ b ] c [ tʃ ] d [ d ] ð [ ð ] è [ ɛ ] e [ ə ] f [ f ] g [ ɡ ] h [ h ] ì [ ɪ ] i [ j ] í [ iː ] j [ dʒ ] k [ k ] l [ l ] m [ m ] n [ n ] ò [ ɒ ] o [ ɔ ] ó [ oː ] p [ p ] r [ r ] s [ s ] ç [ ʃ ] t [ t ] þ [ θ ] u [ u ] v [ v ] w [ w ] ẃ [ ɹ ] x [ χ ] y [ y ] z [ z ] ź [ ʒ ]
So 36 letters compared to the Latin 26. Most consonants are the same as those in English, with a few that have been moved around, and a few extra ones for phonemes that don't usually occur in English (notably, [ y ] and [ χ ]).

[Chrishy]

The one pronoun that English requires:

(s)he

It's honestly just neuter, much like "it", but retaining the "person" aspect of "he" and "she". Simple, yet useful as hell.

edit: I definitely know I'm the 90384092384092383rd person to say that. It's far from original, but definitely helpful. It also will have the aforementioned "we" specifications (once I figure a way for it to not require 15 words for one English pronoun equivalent). I'm working a ton with pronouns atm. haha

[Makri]

Dog-AG sees cat-PAT
"The dog sees the cat" (voluntarily, that is "looks at")

Dog-PT sees cat-PT
"The dog sees the cat" (unvoluntarily, translated as "sees")

Dog-AG sees cat-AG
"The dog sees the cat and the cat sees the dog" (reciprocal, but with most of the focus on the dog)

Dog-PT sees cat-AG
"The dog sees the cat" (unvoluntarily, and the cat is the cause of the action taking place, so more or less a causative)

I would love to see what a child who acquired this language would make of this system. In particular, I have some doubts that the last two can be acquired in the way intended.

[RabbitWho]

I soon found it was frustrating as there were far more words than I'd originally realized.

[vaguelyhumanoid]

My conlang, KoˀZézak, distinguishes past tense and distant or mythic past tense:

Šivkovetikaim-He/she lived
Šivkovetiku-He/she lived long ago

as well as distinguishing dual, plural, and mass plural :
Šivkoviðu-The two of them live
Šivkoveþka-They live
Šivkovžai-Many of them/all of them/the whole group of them live


It's also agglutinative as hell, assuming hell is agglutinative.

[Makri]

There has been argued to be a distinction between distributive and collective plural in some languages. I don't know how much there is to that; but if that does exist, then your idea of plural vs. mass plural would amout to mirroring that in the verbal domain. But maybe you would expect paradigm gaps with certain verbs... Some might not have collective readings. By the way, it has also been argued that collective plurals have a different semantics from noun phrases like "the group of ...". The latter could not plausibly receive special verbal agreement, except in a noun class system.

PS: šiv is really a nasty stem for "live". If it has to be so close anyway, then why not make it živ so people won't be confused?

[vaguelyhumanoid]

What do you mean by "close"?
If you mean too close to English, the vowel isn't even the same, it's pronounced [ʃiβ], though it could be better sounding, definitely. In fact, I didn't even think of the English word until after thinking of the KoˀZézak word in the first place, but it's nice to have feedback.
And yes, "mass" also is used for "collective", in fact nouns have a mass/collective plural mirroring that of verbs.
Dual nouns, however, are used only in literary forms, similarly to the past anterior tense in the Romance languages.

[Makri]

živ is the Slavic root for "live".

[vaguelyhumanoid]

Have any suggestions for alternatives?

[Dingbats]

Dog-AG sees cat-PAT
"The dog sees the cat" (voluntarily, that is "looks at")

Dog-PT sees cat-PT
"The dog sees the cat" (unvoluntarily, translated as "sees")

Dog-AG sees cat-AG
"The dog sees the cat and the cat sees the dog" (reciprocal, but with most of the focus on the dog)

Dog-PT sees cat-AG
"The dog sees the cat" (unvoluntarily, and the cat is the cause of the action taking place, so more or less a causative)

I would love to see what a child who acquired this language would make of this system. In particular, I have some doubts that the last two can be acquired in the way intended.

I've since abandoned that project, but I'm interested in what you said. What would be the problems with learning that system?

[stolid]

I'm still in the design phase, but my language is going to be typed in Cyrillic. I don't know any Russian and only halfway know their alphabet, but I'm working on it. Using another script just seems cool (not to mention if you want to keep others from guessing the otherwise most obvious cognates when written). Do any of you stray from the roman/latin alphabet?

My main goal in my language is compactness and taking some of my favorite features from the languages I know. I'm also kinda wanting a way to obfuscate notes in plain sight (hence the alphabet and combo of language stuff). Kinda off topic, but have any of you considered writing other languages in alphabets they weren't meant to be in (i.e. Spanish in cyrillic or something in the korean alphabet)?

[Makri]

I've since abandoned that project, but I'm interested in what you said. What would be the problems with learning that system?

I think the reciprocal wouldn't work because you would have to have each of both phrases occupy each of two syntactic positions, and furthermore, you would have to exclude that the dog also sees himself. That's just not how reciprocals work in natural language. What languages do is take a plural subject and a reciprocal pronoun as an object or use reciprocal marking on the verb.

And the last is simply a causative, which is also never expressed that way in natural language. That's because the causer position is syntactically higher than the causee position, so the cat would end up in the actual subject position. In addition, causatives of verbs like see are (almost?) never morphologically unmarked.

The problem with your idea is that it ignores how asymmetrically the arguments of a verb are treated in natural language.

[Alces]

How many of you actually created a language or instead just forked and modified one? I am thinking about just forking one because creating one from scratch seems a little ridiculous.

I've created most of my conlangs from scratch. It's not much more work than deriving one from another language; I mean when forking you have to come up with sound changes, grammar changes, etc., while creating from scratch you can freely add in any grammatical structure or word you want.

[breadlord]

I messed around for a bit with the idea for a conlang were all words are stems that are be conjugated to make nouns, adjectives, adverbs and verbs, so that any word can be a noun, adjective, verb or adverb. For example ni (Good) can be conjugated to make ni-bee (Good - Noun), ni-bah (Good - Adjective), ni-boo (Goodly) and ni-bee (To make something good).

[Felstaff]

We are the knights who say Good?

[Roĝer]

I messed around for a bit with the idea for a conlang were all words are stems that are be conjugated to make nouns, adjectives, adverbs and verbs, so that any word can be a noun, adjective, verb or adverb. For example ni (Good) can be conjugated to make ni-bee (Good - Noun), ni-bah (Good - Adjective), ni-boo (Goodly) and ni-bee (To make something good).

You should have a look at Esperanto, it does exactly this, and it makes the language very interesting, as well as easy to learn.

[breadlord]

I'll have a look at Esperanto then, I'm knew to conlangs so I don't really know much about existing ones

[tesseraktik]

I've been considering creating a language with a minimalistic vocabulary (through a combination of which "more complex" concepts can be expressed) and 5-10 "empty" words.
At the beginning of a long text or dialogue (such as a speech, an essary, a lecture, an IM conversation, et cetera), you'd be allowed to give meaning to one, some, none or all of these empty syllables in a way that you hope your readers or fellow conversationalists will understand (for instance, if you assume that everybody who reads your essay knows decent English, you may use English to assign meanings to these words).

Now, while I doubt such a language would catch on, the idea is that if it were used enough, over time it could serve as a test of sorts: Which words do we need to express ourselves in a comfortable and convenient manner, and which ones can be constructed from other concepts?

The inspiration behind this is that I've found it very interesting that Klingon has no words for "Hello!" or "Good-bye!", as this forces you to re-think the way in which you conduct your conversations. I'm particularly fascinated with words that many of us don't feel we truly understand, and sometimes wonder what would happen if there were no word for love:
Would people start using bullkier sentences to express love?
If a language without a standard word for love were left to its own devices for a long time, would one be created? Would there be only one, that would carry almost exactly the same definition as the English word "love", or would there be multiple ones for expressing various conceps that are all covered by our word "love"?

So, in creating the language mentioned above, I probably wouldn't include a word for love. Then, after the language had existed for a while, if anybody had actually used it, I'd see if every essay, lecture, conversation or the like that talked about love defined one of these "empty words" as "love", or if they found ways to say what they wanted to say using only the standard vocabulary.

Does that make any sense what-so-ever?

[Iulus Cofield]

I'm not sure of the value of a language that needs a second language to function. Why not just make your language have additional "empty" pronouns and the vocabulary to define those pronouns in itself?