I gotta be honest. I'm kinda stoked more than one person requested my conlang to be featured in one of your videos. I should also point out that FluidLang wasn't created to be any sort of interlang or auxiliary language, so it can't really be compared to languages like Esperanto - it was just a fun little project that I wanted to develop, and I don't see what's so bad with that. Still, I'm flattered.
What do you mean when you say that putting more than 3 vowels in a language is a bad idea? I know of no language with such a small amount of vowels.You seemed to point to arabic, but as far as i know, arabic has 6 vowels, and arabic is an extreme example, and if you rank languages by the ammount of vowels in them, then adjust for number of speakers, the average language has roughly 8 vowels.
Arabic has just 3 vowels, I think, but they all have length variations, so you end up with 6 distinct phonemes (like a word _lan_ may both be read differently and mean a different thing to _la:n_ (totally making words up here, don't know a single syllable in Arabic)). And the points made were all towards FluidLang supposedly being an interlang (which it apparently isn't) and therefore meant to be an _easy_ way to communicate for most of the world. In such a language, being easy to be spoken for majority of people is absolute key to success. Combine this with the fact that it's far easier to ditch your 3 excessive vowels than to adopt 3 completely new ones, and you'll see why 3-vowel system is _very_ beneficial for an interlang. That goes for interlang only, though. in Artlang, there are virtually no boundaries to creativity (though obviously the more practical the more realistic the better, d'oh).
Irish has no words for yes and no the verb in the question is repeated and for negation a negator is added. and there are many other languages like that.
I was just about to make that same comment, but using Scottish Gaelic as the example. If anything, it's better than a yes-no system: when giving their response, the listener is also confirming that they correctly heard the verb and tense of the question.
My mother tongue (Czech) has a yes and no system, but you can totally get away with just repeating the verb in a positive or negative form, and you won't even sound weird. In fact, it's a safer way because there's less room for confusion in case of negative questions (in Cs you can do stuff like _Don't you know that? -> _*_Yes(, I do)._*_ vs. _*_Yes(, I don't)._* which doesn't work well in En... but you may make stuff like _Don't you know Peter is sick? -> _*_No(, I don't)._*_ vs. _*_No(, he isn't)._* clearer). But I think most, if not all, languages can do that, to varying degrees of fluidity of speech.
That's actually a thing in Hiberno-English too. Irish people will commonly say e.g. "I do" or "I don't" in response to a yes or no question, rather than "yes" or "no." That's one of the ways Irish has influenced Hiberno-English.
3:54 "there most always be words for yes and no!" i speak irish and we don't have a yes or a no. we use verbs instead and native speakers often shorten them. sometimes u can use right or wrong or true or false instead too.
you don't even need a word for yes and no if you don't have a word for right and wrong. you could go all mandarin and just repeat the previous verb from the sentence
Man, I would totally play a drinking game watching Conlang critic. - Vötgil mentioned - It's bad because it's not minimalistic - At this point I always do this thing, but this time I won't - A feature that Jan actually finds good - Sexism found in a language - Rating system gets changed - How do you even pronounce it? (Surprisingly rare one, to be honest) - Why, why would you write [something] as [some weird combination of symbols]
But that doesn't work like that in practice. These sounds are allophones and despite some dialects pronounce 'a' an some 'e' and some both in different circumstances, they treat them as one. Like, have you known that you pronounce different "e' sounds as you speak English? Can you recognise them at all or pronounce them on purpose, even at the 'wrong' places? I've seen English speakers trying to learn an Uralic language (e.g. like my native) and they had issues even to hear the difference between these new vowels, not just with pronouncing them. And what about Ubykh, a Caucasian language what only has two vowels, 'e' and 'a'? :D By allophones it goes up a bit, linguists can't argee but some do the counting even up to six. But all of these are coordinated by surrounding consonants, native speakers don't think they are different sounds. You can count on the French and English skills of native Arabic speakers, or speakers of other similar languages, but that approach already questions the reason behind your auxlang, doesn't it?
That Hangul orthography is terrible too. It uses /pʰ/ for /b~β/ but uses /k͈/ and /t͈/ for /g~ɣ/ and /d~ɾ/, and it doesn't bother marking the distinctive vowel length. Not to mention the fact that it uses an obsolete letter that is unsupported by most fonts and can't even be typed on a normal korean keyboard while there are perfectly valid alternatives that aren't obsolete, that could be used instead.
The reason that the creator uses diacritics to express long vowels is because doubling vowels could get confusing to most English speakers. Two Os can make a u sound. Two Es make a i.
I have been studying Latin for several months now and haven't seen any words for yes and no. It's because there aren't any. XD You just repeat the statement or emphasized word to say yes and repeat the statement and negate it to say no.
You don’t even need words for yes and no if you have right and wrong! Finnish, instead of “yes”, affirms the question asked, for example, “do you walk?” Is answered “I walk” (instead of “yes”), and it works just fine.
I notice that this channel is quite against the use of prefixes and suffixes in ials. I think that there isnt a problem with having many prefixes and suffixes in an ial. As long as it's regular, I don't think it will be difficult to get used to some prefixes and suffixes. (And this is coming from a native speaker of English and Chinese).
jan really overemphasizes learnability over anything else in auxlangs, specifically his idea of what makes a language learnable. Apparently speakers of one language having to learn one or two unique sounds or a new grammatical concept for the sake of clarity completely invalidates an auxlang despite auxlangs being universally easier to learn than any natural lingua franca
Jazz Not to be that guy, but “jan” is just Toki Pona for “human.” Misali is his actual name. I do agree on your points about grammar and phonology, though.
@@moonlightning8269 if you think of it like assigning scores, a language can have 9/10 for learnability, coz -1 for prefixes (or something else) and so still be one of/the best.
Doesn't even make that much sense that prefixes are always bad. Hell, a very easy way to make the contrary of the world is to take a work and add a prefix or suffix that means "not", way easier than to learn a new word - very Newspeaky, I know, but real languages do that very often, like Chinese. Or English.
Jan misali:"if you aren't using the whole Latin alphabet you shouldn't be using diacritics" Gaelic: Speaking of Gaelic it doesn't have words for yes and no
'If' isn't quite the same as 'since,' 'since' is more related to 'because' and 'if' is more like 'when;' in some particular contexts they seem similar, but overall they're not really the same. Compare: "I'll go to the party if Jeff will bring his harmonica," "I'll go to the party when Jeff will bring his harmonica," "I'll go to the party since Jeff will bring his harmonica." And "I won't go to the party since it's raining," "I won't go to the party because it's raining," "I won't go to the party if it's raining." Notice how in each trio of examples, the first two are not exactly identical but they're much closer to each other than they are to the third. It is however very true that 'since' and 'but' are extremely different words, almost opposites in fact, since 'since' exists to establish a relationship between two things (either in respect to time, or causation), but 'but' establishes the absence of a relationship (while also negating the object in some situations). Examples: non-negating 'but' = "I don't like Jeff but I'll go to the party anyway" (aka "I don't like Jeff, and irrespective of that fact I'll go to the party), negating 'but' = "everyone but Jeff can come to the party" (aka "everyone minus Jeff can come to the party").
Correction: Arabic has 3 separate vowels BUT has multiple other vowel sounds! Certain letters, like ط ظ ص ض make "a" closer to "o", as well as in the name of God. And Ive heard that an "eh" sound exists when an alif "ا" is written with a hamza below it (dont know how to write on keyboard)
Alright, someone make a language that features a phonemic burp. It is a language that one can only perhaps maybe utter a word or two in every once in awhile.
@@arcticflower7223 No, but that is. Unless you can explain how I've done something awful by writing that (two years ago, at that). "Awful" seems to imply "immoral". I would agree in hindsight that it's a bit disgusting, but in no way is it immoral. I simply had this thought because I thought it would be interesting to see a language that includes a sound which is difficult to produce by one's own volition, and seven people seemed to agree. But you think I'm "awful", or have a tendency to do "awful" things. Alright. Thanks for letting me know, ArcticFlower. Appreciate it.
mfw virtually all varieties of arabic besides MSA (which isn't often used outside of teaching or formal contexts afaik) have a pretty standard 5-ish vowel system i know this video is old but i'm pretty sure even wikipedia says this
1:35 So because a common language has only 3 vowels, no more than 3 vowels is acceptable? Languages with few vowels tend to have more consonants. (With a few exceptions such as Polynesian languages.) So what's the greatest number of acceptable consonants? Must a language be minimal as regards both vowels and consonants to be acceptable?
For international auxiliary languages specifically, where you want to maximize learnability to make it accessible to everyone, it's good to have fewer phonemes, and to pick the ones that are most common across the world's languages. It's easier to adapt to a small, familiar phonology, than to a big one that includes sounds you don't know. But there isn't really a magical ideal maximum number when it comes to consonants. It's just that vowel systems are easier to pin down and categorize. Also, jan Misali pretty directly says in the video that five vowels is an acceptable number too, so more than three is acceptable.
Hey, don't hate on case! (I say unbiasedly, not acknowledging that my own conlang has case) I can see why it's not good for an InterLang, but playing with word order is a lot of fun and really not that bad if you're willing to take the time to learn the language.
You're right, but from my experience learning Latin (and high school Spanish) I've found that noun declensions and verb conjugations aren't really a major hurtle, at least when it comes to simple, everyday speech. Once you get into the habit of conjugation it's really not that difficult, and if you want to pick up an InterLang then you're willing to make that adjustment. The core issue with creating an Interlang is that natural language is so divers that no grammar is going to feel intuitive and easy for everyone.
True that not any one one grammar will feel intuitive and simple for everyone, but it is also true that some forms of grammar are easier to learn. For example, Lojban. It might not feel natural right away. In fact, it probably won't. Nevertheless, I think it took me about a week to get the hang of the entire thing basically. Not saying I have a problem with Latin and it's declension and conjugations, but they don't exist out of necessity for expression, but out of habit and tendency. :)
I think case is fine for an IAL as long as it is simple and regular. For example Hungarian is very complex but very regular, so just take that regularity and reduce the number of cases.
There isn't anything wrong with case. However one should consider which case distinctions are truly important rather than blindly copying say Latin. Almost every conlang which has case has exactly the same cases as Latin for no real reason. Except: Quenya. Quenya distinguishes genitive, dative, allative, locative and instrumental: but does NOT distinguish accusative (except in it's most archaic form). That should be a sign to all conlangers that the accusative is really quite superfluous. It can be dealt with through word order alone. Quenya even allows the accusative to be expressed solely through pronominal endings on verbs: a feature which I'm surprised FluidLang didn't copy given the intentions behind making it. When Aragorn finds the seedling of the White Tree he simply says "utúvienyes!" To break that down: we have the past tense verb stem "utúvie", the first person long pronominal ending "nye" and the third person short pronominal ending "-s": the sentence means "I have found it!" The pronominal ending given first is assumed to be the subject (as that's the pronominal ending you would expect to find on the end of a verb) and then the second ending is obviously the object of the verb. In fact, I wouldn't give FluidLang as much credit as Jan does. Quenya seems to be better, grammarwise, at fulfilling the stated goals of FluidLang. It just seems like they went in totally the wrong direction with this conlang. If what they wanted was precision and density, they should have been looking at Greek and Quenya when it came to grammar.
I making an international language that has 4 vowels, it uses the 12 most spoken languages to compile a vocabulary easy to learn Its vowels are A /ə ~ ɑ ~ ɒ ~ ɐ ~ ɔ/ E /ɛ/ I /i/ O /o/ Is this good?
Chaosdude7111 From what I've read, the vowels vary by dialect, much like English, but it is noted that long a is on occasion pronounced æ: rather than a: as it is defined. Still, this minute difference does not affect the fact that Arabic has three vowels, each of which can be either short or long.
Some of the criticisms seem to show a distinct bias towards English and away from Algonquian languages... so, I take offense! :) It kindof underscores the problem with interlangs though.
rad. if you end up looking at it this is the document we worked on docs.google.com/spreadsheets/d/1q9OdOaKMYlg6zeTeiax63Q7qbFiGQ5OjRTVKjJNfg0s/edit?usp=sharing
Hindi has 5 basic vowels with 2 dipthongs. To transliterate English words, 2 additional letters are used to indicate those vowels. (Actually almost 10 new diacritics were introduced but most people are familiar with only 2) What I'm getting at is that even Hindi speakers who can't speak a word of English can still easily pronounce those new vowels even though their language has a 5 vowel system. Would people really find it THAT difficult to adapt to new vowels?
if they were all phonemic, yes. Accents can really play havoc with vowel sounds in speech. Although English has many vowel sounds: no single word requires more than three (as far as I'm aware). Most words are either monosyllabic or repeat vowels. Technically, you only need three vowel sounds, plus diphthongs, in order to speak English intelligibly: although there are five vowels total in the language. This is because there are triplets of words with the same consonants like "red", "rod" and "rid". it doesn't particularly matter for intelligibility whether you pronounce the vowels correctly: as long as you clearly distinguish them from one another.
So this is a pretty arbitrary episode to leave this comment on but I'd like to say that watching this and the w and c videos has been really interesting as someone who is only fluent in English but can read Hebrew fluently, although not understand it. One thing I notice a lot is the order of letters in alphabets - one thing I noticed was that abgd equivalents are together and first in both Hebrew and Greek (the two alphabets I know at least somewhat aside from latin) and lmnop also kinda works in Hebrew (lamed mem nun tzadik ayin pey fey)
All arabic languages seem to have at least 5 vowels. Algerian, have 5; Saidi, have 7; Egyptian, Baghdad, Najidi, Hejazi have 8; Urduni, Damascus, have 10; Baghdad Jewish, have 11; Lebanese, have 12; Khaliji have 14; Tunisian, have 16. Virtually all dialects of arabic have at least 5 vowels, and, apart from some outliers, have the standard 5 vowel inventory when length is ignored.
I don't think verb inflections are that much inferior to pronouns and adverbs - for an interlang, a table of verb inflections feels roughly the same difficulty as a table for pronouns/adverbs, assuming they convey the exact same meaning. If you take to the extreme, in that literal sense you could take every pronoun/adverb and append it to a verb to make an inflection serving that exact purpose, or vice versa. I'm not hating on either, I just don't think that, from a practical point of view, there should be such a strong preference for one or the other. I'm definitely not a pro, though, and I'm happy to be wrong, so I'm open to corrections and/or examples that would disprove my claim.
That's not the point. The point is for E V E R Y B O D Y to speak an auxlang, not "the most people". So of course if something like 30 or even fewer, like 15% of the speakers of the world have only 3 vowels in their native language, that cannot be overseen and it doesn't matter how many vowels the others languages have. The auxlang needs to set standard on the one that has the fewest vowels in order to be accessible to the most people.
***** I mean im busy trying to develop a system of languages but am struggling because its a weaksauce hobby I did minimal research in. And if someone trained or more interested in linguistics wanted to come along and help that would be great.
EVEN VOTGIL knew that and IT'S VOTGIL
Deutsch Amerikaner it’s votgil
No, it's Vötgil (with ö).
@@obviativ123 ís the ö prounounced
@@theuniversalstegosaurus7911 I don't know
@@theuniversalstegosaurus7911 Yes, it is pronounced.
I gotta be honest. I'm kinda stoked more than one person requested my conlang to be featured in one of your videos.
I should also point out that FluidLang wasn't created to be any sort of interlang or auxiliary language, so it can't really be compared to languages like Esperanto - it was just a fun little project that I wanted to develop, and I don't see what's so bad with that. Still, I'm flattered.
my bad. FluidLang is good as an artlang, but I was told that it's an interlang, so I treated it accordingly.
I'll clear it up in the next episode.
What do you mean when you say that putting more than 3 vowels in a language is a bad idea? I know of no language with such a small amount of vowels.You seemed to point to arabic, but as far as i know, arabic has 6 vowels, and arabic is an extreme example, and if you rank languages by the ammount of vowels in them, then adjust for number of speakers, the average language has roughly 8 vowels.
Arabic has just 3 vowels, I think, but they all have length variations, so you end up with 6 distinct phonemes (like a word _lan_ may both be read differently and mean a different thing to _la:n_ (totally making words up here, don't know a single syllable in Arabic)).
And the points made were all towards FluidLang supposedly being an interlang (which it apparently isn't) and therefore meant to be an _easy_ way to communicate for most of the world. In such a language, being easy to be spoken for majority of people is absolute key to success. Combine this with the fact that it's far easier to ditch your 3 excessive vowels than to adopt 3 completely new ones, and you'll see why 3-vowel system is _very_ beneficial for an interlang.
That goes for interlang only, though. in Artlang, there are virtually no boundaries to creativity (though obviously the more practical the more realistic the better, d'oh).
Conlang Critic Would you review artlangs as well? I do suppose that would be a highly subjective type of reviewing, but still...
Irish has no words for yes and no the verb in the question is repeated and for negation a negator is added. and there are many other languages like that.
I was just about to make that same comment, but using Scottish Gaelic as the example. If anything, it's better than a yes-no system: when giving their response, the listener is also confirming that they correctly heard the verb and tense of the question.
scotsmanRS
That is an excelent point and would help me a lot if such a thing was common in English I might start using it.
My mother tongue (Czech) has a yes and no system, but you can totally get away with just repeating the verb in a positive or negative form, and you won't even sound weird. In fact, it's a safer way because there's less room for confusion in case of negative questions (in Cs you can do stuff like _Don't you know that? -> _*_Yes(, I do)._*_ vs. _*_Yes(, I don't)._* which doesn't work well in En... but you may make stuff like _Don't you know Peter is sick? -> _*_No(, I don't)._*_ vs. _*_No(, he isn't)._* clearer). But I think most, if not all, languages can do that, to varying degrees of fluidity of speech.
That's actually a thing in Hiberno-English too. Irish people will commonly say e.g. "I do" or "I don't" in response to a yes or no question, rather than "yes" or "no." That's one of the ways Irish has influenced Hiberno-English.
Gaelic is cooked in the head though
3:54 "there most always be words for yes and no!" i speak irish and we don't have a yes or a no. we use verbs instead and native speakers often shorten them. sometimes u can use right or wrong or true or false instead too.
fucking latin doesn't have a yes and no
you don't even need a word for yes and no if you don't have a word for right and wrong. you could go all mandarin and just repeat the previous verb from the sentence
Mandarin has right and wrong for yes and no though. But we do usually repeat the verb
You can often do that in English too.
toki pona moment
@@franimer2107 ni li tenpo pi toki pona!
@@official-obama jan pi ale ala li sona e toki pona :(
2:46 drink whenever Jan gets mad at the IAL and makes comparison with Vötgil
Man, I would totally play a drinking game watching Conlang critic.
- Vötgil mentioned
- It's bad because it's not minimalistic
- At this point I always do this thing, but this time I won't
- A feature that Jan actually finds good
- Sexism found in a language
- Rating system gets changed
- How do you even pronounce it? (Surprisingly rare one, to be honest)
- Why, why would you write [something] as [some weird combination of symbols]
whenever misali*
@@jaydentthis tumblr states that jan is an okay way to shorted his name, but that it isnt capitalized, as per toki pona
@@ashstargloww here it is capitalized but i see wat u mean
Speakers of 3 vowel languages like Arabic don't have trouble with five vowel languages because often, they allophonically have 5 or even 6 vowels.
Yes, if I had it on me, I could cite from Comrie's "The World's Major Languages" that most colloquial dialects of Arabic have phonemic mid vowels.
But that doesn't work like that in practice. These sounds are allophones and despite some dialects pronounce 'a' an some 'e' and some both in different circumstances, they treat them as one. Like, have you known that you pronounce different "e' sounds as you speak English? Can you recognise them at all or pronounce them on purpose, even at the 'wrong' places? I've seen English speakers trying to learn an Uralic language (e.g. like my native) and they had issues even to hear the difference between these new vowels, not just with pronouncing them.
And what about Ubykh, a Caucasian language what only has two vowels, 'e' and 'a'? :D By allophones it goes up a bit, linguists can't argee but some do the counting even up to six. But all of these are coordinated by surrounding consonants, native speakers don't think they are different sounds.
You can count on the French and English skills of native Arabic speakers, or speakers of other similar languages, but that approach already questions the reason behind your auxlang, doesn't it?
Allophonically having 5 vowels is not the same as phonemically having 5 vowels, so it can still pose a problem
Not to mention that pretty much all modern Arabic dialects/languages have five vowels. MSA doesn‘t, but nobody speaks that as a first language.
6 vowels is too much? _Laughs in Finnish_
Finnish?!? Cute, *laughs in Danish*
Danish? Cute *laughs in IPA*
@@slamwall9057 IPA? Laughs in English
ipa cyootte laffse inne ǎ̼̺̪᷆̋̏̂᷅ɚ̺̃
excuse me whaaaaat?
That Hangul orthography is terrible too. It uses /pʰ/ for /b~β/ but uses /k͈/ and /t͈/ for /g~ɣ/ and /d~ɾ/, and it doesn't bother marking the distinctive vowel length. Not to mention the fact that it uses an obsolete letter that is unsupported by most fonts and can't even be typed on a normal korean keyboard while there are perfectly valid alternatives that aren't obsolete, that could be used instead.
yeah, that is terrible. if I was better at Korean, I'd've definitely gotten into that in the episode itself.
Yeah.. Sometimes ㄱ (g) has the same pronounciation as ㅋ (k')
Like in 가다
But in the word 내가 its like a ㄲ (g'). Its frustrating.
I don't think there are even Unicode Hangul characters with ㅸ as a component.
That's because the creaters do what most English speakers do: assume romanization is accurate.
@@danielholowaty2648 probably because of shifts in the language since king Sejong published Hangul
FluidLang: "You always need words for yes and no"
Chinese: 错
"你总是需要'是'和'不是'的词。“
”错。”
That would be a hella rude conversation haha
This was so great. Glad to have discovered this. I have some neat ideas and wonder what you'd think about it, but I'll need time first
* Listens to bit about vowels
* Looks at Danish
* Tries not to break down crying.
As a dane, hearing that only 3 vowels are needee kind of fucks my mind.
weird way to say chokingonapotatoian
The reason that the creator uses diacritics to express long vowels is because doubling vowels could get confusing to most English speakers. Two Os can make a u sound. Two Es make a i.
It's not hard to adjust to though.
You're totally right
@@purplemosasaurus5987 By that Logic Conlangs Shouldn't use j for /j/ because English Speakers might pronounce it as in Jar
@@JohndoeBlocks Whatever, I don't want to start a comment-section-argument again, I'm tired
These videos make me appreciate common languages a lot more.
You should totally try to create you own language, it’d be fun to create and to watch the updates
I have been studying Latin for several months now and haven't seen any words for yes and no. It's because there aren't any. XD You just repeat the statement or emphasized word to say yes and repeat the statement and negate it to say no.
It is the same in Irish and Scottish Gaelic.
"Non"
Zenytram Searom That’s the negative adverb “not”. It wasn’t regularly used as an interjection in Classical Latin.
Ita est / non est i'd say
Oh I see, like if someone said "is the stove preheated" then "preheated" means yes, and "not preheated" means no.
"the show that gets facts wrong about your favourite conlang"
2:54 I can't put into words quite how terrible that Hangeul transliteration looks...
"*sigh* the FluidLang subreddit" - jan Misali
this quote lives in my head rent-free
"there must always be words for yes and no!"
mandarin, with like a billion speakers: 不需要
You don’t even need words for yes and no if you have right and wrong! Finnish, instead of “yes”, affirms the question asked, for example, “do you walk?” Is answered “I walk” (instead of “yes”), and it works just fine.
1:42 *Deviantart flashbacks*
Misali: "If you're not using the entire Latin alphabet, you shouldn't have diacritical marks."
Vietnamese:
vietnamese is a tonal language
@@milkymilsy And?
(Besides, not all of the diacritics indicate tone)
You know things are a bit goofy when there's more vowels then consonants X3
I notice that this channel is quite against the use of prefixes and suffixes in ials. I think that there isnt a problem with having many prefixes and suffixes in an ial. As long as it's regular, I don't think it will be difficult to get used to some prefixes and suffixes. (And this is coming from a native speaker of English and Chinese).
jan really overemphasizes learnability over anything else in auxlangs, specifically his idea of what makes a language learnable. Apparently speakers of one language having to learn one or two unique sounds or a new grammatical concept for the sake of clarity completely invalidates an auxlang despite auxlangs being universally easier to learn than any natural lingua franca
Jazz Not to be that guy, but “jan” is just Toki Pona for “human.” Misali is his actual name. I do agree on your points about grammar and phonology, though.
@@moonlightning8269 if you think of it like assigning scores, a language can have 9/10 for learnability, coz -1 for prefixes (or something else) and so still be one of/the best.
Doesn't even make that much sense that prefixes are always bad. Hell, a very easy way to make the contrary of the world is to take a work and add a prefix or suffix that means "not", way easier than to learn a new word - very Newspeaky, I know, but real languages do that very often, like Chinese. Or English.
Jan misali:"if you aren't using the whole Latin alphabet you shouldn't be using diacritics"
Gaelic:
Speaking of Gaelic it doesn't have words for yes and no
man that was such a bad take. I completely disagree with that now
@@HBMmaster whoa, responding to a comment on a 3 year old video
@@HBMmaster enjoy crying
@@HBMmaster my curiosity is piqued
which version of gaelic? irish? scots? (and does welsh count as gaelic?)
'If' isn't quite the same as 'since,' 'since' is more related to 'because' and 'if' is more like 'when;' in some particular contexts they seem similar, but overall they're not really the same. Compare: "I'll go to the party if Jeff will bring his harmonica," "I'll go to the party when Jeff will bring his harmonica," "I'll go to the party since Jeff will bring his harmonica." And "I won't go to the party since it's raining," "I won't go to the party because it's raining," "I won't go to the party if it's raining." Notice how in each trio of examples, the first two are not exactly identical but they're much closer to each other than they are to the third. It is however very true that 'since' and 'but' are extremely different words, almost opposites in fact, since 'since' exists to establish a relationship between two things (either in respect to time, or causation), but 'but' establishes the absence of a relationship (while also negating the object in some situations). Examples: non-negating 'but' = "I don't like Jeff but I'll go to the party anyway" (aka "I don't like Jeff, and irrespective of that fact I'll go to the party), negating 'but' = "everyone but Jeff can come to the party" (aka "everyone minus Jeff can come to the party").
Can you do Solresol
just updated the Big List.
I can't wait to hear you sing it
Correction:
Arabic has 3 separate vowels BUT has multiple other vowel sounds!
Certain letters, like ط ظ ص ض make "a" closer to "o", as well as in the name of God. And Ive heard that an "eh" sound exists when an alif "ا" is written with a hamza below it (dont know how to write on keyboard)
Alright, someone make a language that features a phonemic burp. It is a language that one can only perhaps maybe utter a word or two in every once in awhile.
That's an awful thing to write.
@@arcticflower7223 No, but that is. Unless you can explain how I've done something awful by writing that (two years ago, at that). "Awful" seems to imply "immoral". I would agree in hindsight that it's a bit disgusting, but in no way is it immoral. I simply had this thought because I thought it would be interesting to see a language that includes a sound which is difficult to produce by one's own volition, and seven people seemed to agree. But you think I'm "awful", or have a tendency to do "awful" things. Alright. Thanks for letting me know, ArcticFlower. Appreciate it.
Unless you’re Rick Sanchez APPARENTL
i really want that lojban 42 shirt ):
2:32 Which letter would that be?
mfw virtually all varieties of arabic besides MSA (which isn't often used outside of teaching or formal contexts afaik) have a pretty standard 5-ish vowel system
i know this video is old but i'm pretty sure even wikipedia says this
1:40 you don't need the pipe symbol; "internationali[sz]ed" is correct.
For next episode:
(1)Toki Pona, (2)Esperanto, (3)Xynder
What is Xynder?
1:35 So because a common language has only 3 vowels, no more than 3 vowels is acceptable? Languages with few vowels tend to have more consonants. (With a few exceptions such as Polynesian languages.) So what's the greatest number of acceptable consonants? Must a language be minimal as regards both vowels and consonants to be acceptable?
For international auxiliary languages specifically, where you want to maximize learnability to make it accessible to everyone, it's good to have fewer phonemes, and to pick the ones that are most common across the world's languages. It's easier to adapt to a small, familiar phonology, than to a big one that includes sounds you don't know.
But there isn't really a magical ideal maximum number when it comes to consonants. It's just that vowel systems are easier to pin down and categorize.
Also, jan Misali pretty directly says in the video that five vowels is an acceptable number too, so more than three is acceptable.
This doesn't make it compatible
Agma Schwa made a video on what the least phonemes needed for a naturalistic language… I think there was somewhere between 6 and 8
@@m90ewhat about the most phoneme? i want to have alot because i hate allophone
you should do another Jack Eisenmann conlang, maybe pegakibo? just a thought.
just updated the Big List.
Reviewing Pegakibo doesn't really make much sense since it's supposed to be a personal language.
2:37
wait, what? english only has 5 vowel letters.
Yeah it has 5 LETTERS, but like a million vowel SOUNDS.
@@fiveseventy yeah, i know. i speak english.
I looked at the languages for subtitles after you started talking about it, and there was Esperanto.....
Please do Pmitxki!! There are only 2 videos, but it sounds really interesting
just updated the Big List.
@@HBMmaster How about Kotava ?
2:56 grammatical gender?
2:30 ?
right? did he really expect them to use for /ɑ/ lol
Love your videos I watched them all like 5 times. Have you ever tought about doing a video on Novial?
3:30
"the same number as Toki Pona also seven more"
I spat out my drink
No two words ever mean exactly the same thing or they wouldn't co-exist.
disagree. Sometimes two words mean exactly the same thing, but in different dialects of the same language. A famous example in Dutch is friet vs patat
What beef do you have with reddit? Or is it only the fluidlang sub?
what if I don't want to use q for /œ/
Fluidlang's vowels are "oo ee oo aa aa, ting tang walla walla Bing bang"
LOL.
How would a 20 vowel language adapt
my first thought on seeing no labiodental consonants was "oh hey, yinglets could speak this"
I know you won't get to it for many many months, but could you check out Vallenan?
just updated the Big List.
HBMmaster do Siwa please. The guy who made it also made a hella long grammar.
just updated the Big List.
HBMmaster Have you put Solresol (or the allegedly improved version Sarus) on that list yet?
just updated the Big List.
Hey, don't hate on case! (I say unbiasedly, not acknowledging that my own conlang has case)
I can see why it's not good for an InterLang, but playing with word order is a lot of fun and really not that bad if you're willing to take the time to learn the language.
He critiques Conlangs in the context they are meant to. As such, an Interlang is criticized by its ability to *be* an interlang.
You're right, but from my experience learning Latin (and high school Spanish) I've found that noun declensions and verb conjugations aren't really a major hurtle, at least when it comes to simple, everyday speech. Once you get into the habit of conjugation it's really not that difficult, and if you want to pick up an InterLang then you're willing to make that adjustment.
The core issue with creating an Interlang is that natural language is so divers that no grammar is going to feel intuitive and easy for everyone.
True that not any one one grammar will feel intuitive and simple for everyone, but it is also true that some forms of grammar are easier to learn. For example, Lojban. It might not feel natural right away. In fact, it probably won't. Nevertheless, I think it took me about a week to get the hang of the entire thing basically. Not saying I have a problem with Latin and it's declension and conjugations, but they don't exist out of necessity for expression, but out of habit and tendency. :)
I think case is fine for an IAL as long as it is simple and regular. For example Hungarian is very complex but very regular, so just take that regularity and reduce the number of cases.
There isn't anything wrong with case. However one should consider which case distinctions are truly important rather than blindly copying say Latin. Almost every conlang which has case has exactly the same cases as Latin for no real reason. Except: Quenya. Quenya distinguishes genitive, dative, allative, locative and instrumental: but does NOT distinguish accusative (except in it's most archaic form). That should be a sign to all conlangers that the accusative is really quite superfluous. It can be dealt with through word order alone. Quenya even allows the accusative to be expressed solely through pronominal endings on verbs: a feature which I'm surprised FluidLang didn't copy given the intentions behind making it. When Aragorn finds the seedling of the White Tree he simply says "utúvienyes!" To break that down: we have the past tense verb stem "utúvie", the first person long pronominal ending "nye" and the third person short pronominal ending "-s": the sentence means "I have found it!" The pronominal ending given first is assumed to be the subject (as that's the pronominal ending you would expect to find on the end of a verb) and then the second ending is obviously the object of the verb. In fact, I wouldn't give FluidLang as much credit as Jan does. Quenya seems to be better, grammarwise, at fulfilling the stated goals of FluidLang. It just seems like they went in totally the wrong direction with this conlang. If what they wanted was precision and density, they should have been looking at Greek and Quenya when it came to grammar.
I making an international language that has 4 vowels, it uses the 12 most spoken languages to compile a vocabulary easy to learn
Its vowels are
A /ə ~ ɑ ~ ɒ ~ ɐ ~ ɔ/
E /ɛ/
I /i/
O /o/
Is this good?
I hardly distinguish /ɔ/ and /o/, and im supposed to do that, just saying (im italian)
Hey, Hungarian does the diacritics for vowel length thing.
I love your videos
Can you do Ido and Novial?
just updated the Big List.
Wrong IPA
i love making weird grunts and have it be academic
You should review Talossan.
just updated the Big List.
Doesn't Arabic have at least four vowels? There's a distinction between "a" and "æ", right?
Chaosdude7111 From what I've read, the vowels vary by dialect, much like English, but it is noted that long a is on occasion pronounced æ: rather than a: as it is defined. Still, this minute difference does not affect the fact that Arabic has three vowels, each of which can be either short or long.
In my dialect: no
TECHNICALY
Arabic has really as many vowels as English, if not more
Some of the criticisms seem to show a distinct bias towards English and away from Algonquian languages... so, I take offense! :) It kindof underscores the problem with interlangs though.
At least they didn't put in the Welsh ydw, ydy, ydyn, oedd, do, wyt...
I don't even know what I would call that system
Can you PLEASEEE talk about septopod?
You should look at sabldiri, the language that developed during reddit's robin April fools joke.
just updated the Big List.
rad. if you end up looking at it this is the document we worked on docs.google.com/spreadsheets/d/1q9OdOaKMYlg6zeTeiax63Q7qbFiGQ5OjRTVKjJNfg0s/edit?usp=sharing
ṑ looks like it's wearing a Robin Hood hat.
The wikipidia made a mistake about Russian phonology, the consenent ʑ: or a soft zhj doesnt exist in Russian
FluidLang: *is supposed to be an interlang*
Also Fluidlang: *here is a letter with 2 diacritcs stacked on top of each other: ṑ*
Translations ? Well I've found my summer job.
How do I do the translation?
there's a link in the description.
Oh wait Esperanto translation is done scratch that.
god, i'm a beginner conlang maker and hearing about grammar hurts my brain. I DON'T UNDERSTAND NOUN TYPES OR WHATEVER :-:
fortunately, you don't need complicated grammar to make a compelling conlang. see: Toki Pona.
thank shempie
So is Toki Pona your favorite conlang?
you'll get the answer in the next episode! or in the aUI episode, where I just said the answer unprompted
HBMmaster you will have already gotten had previously you had been watching
I guess I was actually just asking for reasons, but you'll cover that.
Hindi has 5 basic vowels with 2 dipthongs.
To transliterate English words, 2 additional letters are used to indicate those vowels. (Actually almost 10 new diacritics were introduced but most people are familiar with only 2)
What I'm getting at is that even Hindi speakers who can't speak a word of English can still easily pronounce those new vowels even though their language has a 5 vowel system.
Would people really find it THAT difficult to adapt to new vowels?
if they were all phonemic, yes. Accents can really play havoc with vowel sounds in speech. Although English has many vowel sounds: no single word requires more than three (as far as I'm aware). Most words are either monosyllabic or repeat vowels. Technically, you only need three vowel sounds, plus diphthongs, in order to speak English intelligibly: although there are five vowels total in the language. This is because there are triplets of words with the same consonants like "red", "rod" and "rid". it doesn't particularly matter for intelligibility whether you pronounce the vowels correctly: as long as you clearly distinguish them from one another.
high valyrian!
So this is a pretty arbitrary episode to leave this comment on but I'd like to say that watching this and the w and c videos has been really interesting as someone who is only fluent in English but can read Hebrew fluently, although not understand it. One thing I notice a lot is the order of letters in alphabets - one thing I noticed was that abgd equivalents are together and first in both Hebrew and Greek (the two alphabets I know at least somewhat aside from latin) and lmnop also kinda works in Hebrew (lamed mem nun tzadik ayin pey fey)
you should redo this video since you reviewed it as an interlang, not an artlang
if I was going to redo old episodes, this wouldn't be the first one
don't forget: this is the show that gets facts wrong about your favorite conlang!
0:13 of course they used that font
4:00 ya dingus
please i need the lojban vore shirt
idk y I had subtitles in netherlandish so I thinked that you made subtitles in fluidlang 😮
All arabic languages seem to have at least 5 vowels. Algerian, have 5; Saidi, have 7; Egyptian, Baghdad, Najidi, Hejazi have 8; Urduni, Damascus, have 10; Baghdad Jewish, have 11; Lebanese, have 12; Khaliji have 14; Tunisian, have 16. Virtually all dialects of arabic have at least 5 vowels, and, apart from some outliers, have the standard 5 vowel inventory when length is ignored.
I love how your voice shows about as much disdain for reddit as the average 4chan anon.
I'm requesting Venedic and it's corected ortography called poilschi
just updated the Big List.
appreciate it
do a video on dama diwan
um why is the second half of the episode repeated twice?
not enough vowels
If and because is actually pretty close, you just flip the clauses. (and some subtle changes in semantics, of course.)
why exactly is the said like that?
…where did the ~fluidlang subreddit~ go
Maybe the same word for completely different meanings recognizes the irregularities of real languages
I've studied Arabic. It has more than 3 vowels.
I don't think verb inflections are that much inferior to pronouns and adverbs - for an interlang, a table of verb inflections feels roughly the same difficulty as a table for pronouns/adverbs, assuming they convey the exact same meaning. If you take to the extreme, in that literal sense you could take every pronoun/adverb and append it to a verb to make an inflection serving that exact purpose, or vice versa.
I'm not hating on either, I just don't think that, from a practical point of view, there should be such a strong preference for one or the other.
I'm definitely not a pro, though, and I'm happy to be wrong, so I'm open to corrections and/or examples that would disprove my claim.
the audio...
Review Talossan!
just updated the Big List.
2:54
Korean:what?
Chinese is an eight-vowel language. And look how many people speak it,
That's not the point. The point is for E V E R Y B O D Y to speak an auxlang, not "the most people". So of course if something like 30 or even fewer, like 15% of the speakers of the world have only 3 vowels in their native language, that cannot be overseen and it doesn't matter how many vowels the others languages have. The auxlang needs to set standard on the one that has the fewest vowels in order to be accessible to the most people.
Doesn't make it easier for non-Chinese to learn it.
What is four two?
trinkises
wait, wrong conglang
Hey, I can translate it to spanish and catalan
nice
man, remember when community subs were a thing? yeah,
1:03 /h/
0:21
1:26 Minecraft villager
anyone interested in helping develop a system on languages?
***** I mean im busy trying to develop a system of languages but am struggling because its a weaksauce hobby I did minimal research in. And if someone trained or more interested in linguistics wanted to come along and help that would be great.
***** a family. Thanks