I think it's a good way, or at least a perfectly legitimate one. My love affair with natlang linguistics began with conlanging and each one helps me understand the other better
Funnily, when living a few years in China, as a native English speaker, I found if I tried to pronounce tones no-one could understand me, but if I ignored the tones I could make myself understood well enough. I am assuming that my attempts to use tones was muddling my pronunciation up beyond the point of comprehensibility, while when I stopped trying so hard, the listener could pick up enough of my words to use context to infer which tones were needed for what I was saying to make sense! My opportunities to actually speak Chinese were somewhat limited too, since I was there as an English Language teacher, so my job was to speak English all day, and even away from work, anyone who spoke even a smattering of English, on realizing I wasn't actually a local, wanted to practice it. (I'm Asian, but 2nd/3rd gen Australian, even my mother doesn't speak any other language than English. Related funny story: one of my teaching colleagues was a bright-red-haired Scottish lad who had studied Chinese at university to an impressively high fluency, and if we were out together, people would assume I was his translator and try to speak to me. Which he then had to translate for my benefit!)
I can definitely see that. One of the things I've noticed is that if the use of tones doesn't yet come naturally to the speaker, it can slow down their speech to the point it can sometimes be com pared some what to spea king Eng glish one syl la ble at a time, which makes things more difficult to understand because it's less clear from the rhythm where one word ends and the next one starts, and the listener can more easily forget the earlier parts of the sentence due to the combination of the time delay and the extra cognitive load. (Also, new speakers tend to exaggerate tones more than usual, which unintuitively also tends to make it harder to understand.) In contrast, it's easier for native speakers to understand "tone-less" Chinese more easily because tone-less Chinese is already pretty common in everyday life: Songs can't usually be sung with tones intact, so understanding song lyrics by ear necessarily requires the listener to be able to identify words from context without tone being present.
@@vikiai4241 There are other contexts as well. If you're shouting or whispering, tonal distinction can easily get lost, but the language can still be understood. A good analogy is whispering. If you whisper properly, you don't produce voiced consonants, meaning, theoretically, "Pat the bat in the bag on the back" would come out as "Pat the pat in the pack on the pack." Realistically, since combinations like that are somewhat unlikely, context normally allows it to be clear what is being intended, just as you can understand a ventriloquist even though they cannot produce many phonemes in their acts.
Chinese tones tend to merge and flow into each other in speech, unlike say Vietnamese or Cantonese, which doesn't do that. So you could be pronouncing all the tones as they would be pronounced in isolation, thereby pronouncing it unnaturally. This makes it really hard to understand. A good example of not entirely correct but get the flow speech would be John Cena. An example of someone that tries to pronounce everything is Mark Zuckerberg. He speaks really slowly, so it helps but it still requires a lot of attention to figure out what he is saying.
I have to mention Punjabi. It often doesn't get included when talking about tonal languages but it is kinda the only Indo-European language that is fully tonal. The three tones in Punjabi are: level, low-rising and high-falling. The tones occur due to the loss of the voiced aspirated stop series. Word initial voiced aspirated stops became voiceless unaspirated stops and cause a high-falling tone in the proceeding vowel. Word medial and word final voiced aspirate stops became voiced unaspirated stops and cause a low-rising tone in the preceeding vowel. In a lot of dialects (including my own) h in the coda position is lost and causes a low-rising tone. Other than that this was an amazing video as always. Note: I'm writing this comment based off of my memory so some of the information might be incorrect.
Punjabi native speaker and Punjabi teacher here! You are exactly right! The letter ਘ (gh) for example, in the beginning of a word is pronounced like a kà (high-falling) and at the end like a ga (level tone). ਘਰ - ghar /kàr/ ਸਿੰਘ - singh /sing(a)/ I was really hoping this video would mentioned Punjabi because of its uniqueness in being the only alive IE language with tones, so I'm happy to see another (presumably) Punjabi speaker in the comments
@@theguy5898 Not the only IE languages with phonemic tonality, Swedish, Norwegian, Lithuanian, Latvian, Serbo-Croatian, Slovenian and Limburgish have or at least in most dialects have tonal qualities more or less. :)
@@jasminekaram880 all of those have only pitch accent, not tone. There's a difference. Btw Serbo-Croatian speaker here. Edit: it's Slovene, not Slovenian.
@@wtc5198 I know quite a few linguists consider “pitch-accent” to just be a restricted tonal system, especially the Norwegian-type (which includes Serbo-Croatian) where it is commonly analysed as a marked tone attaching to the stress.
I natively speak Hokkien (a Sinitic language), and what I've learned that makes it unique is that nearly every syllable has two possible tones. Let's call them R and S for the sake of labelling. Which of the two possible tones gets used depends on the syllable's function and position in a phrase/sentence. For example (in the Philippine variety, where I come from): 四 "four" - R /si˥/ or S /si˥˩/ 九 "nine" - R /kaw˧˥/ or S /kaw˥/ 十 "ten" - R /tsap˧/ or S /tsap˧˥/ This "two-toneness" also happens at the word level, not just at the syllable level: 十四 "fourteen" - RR /si˥ tsap˧/ or RS /si˥ tsap˧˥/ 四十 "forty" - RR /tsap˧ si˥/ or RS /tsap˧ si˥˩/ 十九 "nineteen" - RR /tsap˧ kaw˧˥/ or RS /tsap˧ kaw˥/ 九十 "ninety" - RR /kaw˧˥ tsap˧/ or RS /kaw˧˥ tsap˧˥/ 四十四 "forty four" - RRR /si˥ tsap˧ si˥/ or RRS /si˥ tsap˧ si˥˩/ 四十九 "forty nine" - RRR /si˥ tsap˧ kaw˧˥/ or RRS /si˥ tsap˧ kaw˥/ 九十四 "ninety four" - RRR /kaw˧˥ tsap˧ si˥/ or RRS /kaw˧˥ tsap˧ si˥˩/ 九十九 "ninety nine" - RRR /kaw˧˥ tsap˧ kaw˧˥/ or RRS /kaw˧˥ tsap˧ kaw˥/ This phenomenon is not to be confused with tone sandhi (as many people do), which refers to tones changing depending on the surrounding phonetic environment. In contrast, the "two-toneness" is a purely grammatical thing; the R and S forms have nothing to do with each other. It's still a mystery to me how such a tone system could have developed naturally, and how native speakers subconsciously understand all this. I still haven't heard of any other language with this sort of "two-toneness".
I think tonogenesis is currently the best theory on how tones evolve and it's definitely best for conlangers to just stick to it but we also shouldn't forget that this current model can't explain all tonal developments. For example Middle Chinese had 4 tones, level, rising, falling and "checked" (which may or may not have had a unique contour tone but was definitely checked by a syllable final stop). When in Late Middle Chinese voicing was starting to disappear in obstruents, this system split largely the way you'd expect but curiously the 3rd tone (called 上 sháng or shǎng "rising") following a voiced obstruent merged into the falling tone (Standard Chinese 4th tone: HL). Otherwise, the 3rd tone just became the modern day 3rd tone (HLH). What's even more mysterious is the modern distribution of the Middle Chinese checked tone after voiceless obstruents because there's actually no pattern to be found. This may or may not have been due to dialectal mixing though. Lastly there are also other weird developments. For example in the Beijing dialect of Mandarin Chinese the MC 1st tone following a voiceless obstruent became a high level tone, while in Jinan this exact same tone became a falling-rising tone. This development is exactly opposite to the development of the MC rising tone after a voiceless obstruent: Beijing falling-rising, Jinan high level. And it's not like those dialects are that far apart from each other - just about 400km.
the irregular outcome of the checked tone after voiceless obstruents is only really found in Mandarin. Other Sinitic languages (and I believe, even more peripheral varieties of Mandarin) tend to have pretty regular correspondences, so given Mandarin's unique sociolinguistic position compared to the rest of Sinitic, the dialect mixing explanation seems pretty solid
Tonogenesis is also not the end all be all of tonal evolution. Tonogenesis can cover evolution of nontonal language to tonal languages, but tonal natlangs have tone sandhi systems which we have no good theory on their development. For example, compare Hokkien and Mandarin tone sandhi, these came from the same original language and couldn't be more different with their tone sandhi. Further (except for Cantonese), every real tonal language has a tone sandhi system and I would argue that a tone sandhi system is what REALLY sells a tonal language, more than the existence of tones itself Edit: Yes, even Vietnamese, see my comments below. It is difficult to notice tone sandhi if you speak a tonal language natively for various reasons.
@@sciencyazn1509 there are plenty of tonal languages that don't have any form of tone sandhi, the entire Kra-Dai family lacks it entirely as does the nearby Vietnamese
@@multinet9037 from wikipedia's tone Sandi page: "Tone sandhi occurs to some extent in nearly all tonal languages, manifesting itself in different ways."
400 kms are (better: were) a huge distance in China, unless you were on the Grand Canal... In Italy in the same distance you pass from the Gallo-Italic Lombard dialects to properly Italian Tuscan.
Your timing with this video is impeccable. Only just yesterday was I looking for resources on how tones evolve in a language and today I see that you've released this video. Many thanks. It was quite informative and I might try my hand at some kind of tonal language now. ♥
Well, considering 13 thousand people already have watched it, the likelihood that someone just started working with tones is very likely. So I wouldn't say it was a timing, if it was inevitable.
One possibility with tones that I think is really interesting is the possibility of tonal morphemes, where a tone on a vowel in a verb can actually carry some meaning. For example, a verb might change tense/aspect/mood or a noun might change case via a shift in tone. In Tiv, for example, past tense is marked by a floating low tone while habitual is marked by a floating high tone. And in Angas, floating high tones and floating low tones attach to vowels in nouns in specific cases, that themselves already bear tone, forming contour tones if the basic lexical tone and the tonal morpheme being attached differ.
this video seems very good at discussing the tonogenesis in things like the Chinese languages, Vietnamese, and Thai, but doesn't match so well to my understanding of how the pitch accent of Continental Scandinavia developed (where it's largely to do with the presence or absence, and location of any now-lost syllables in the Early Old Norse form of the word). Unfortunately I don't know the details of which properties lead to which accent, but this seems to be the consensus in the literature. It would be interesting to see a discussion of this sort of tonogenesis in the future (and it absolutely should be considered tonogenesis, given the existence of tone sandhi, it's pretty clear that pitch accents and largely independent syllabic tones are on a sliding scale rather than entirely distinct entities)!
Iirc, Scandinavian tones developed from peak delay. Basically, stress in Proto-Germanic was believed to have a falling contour. Proto-Germanic was also polysyllabic like IE Languages. However in the transition to North Germanic, much of the unstressed syllables were lost creating many monosyllabic words. In the remaining polysyllabic words, the tone shifted one syllable back, creating Scandinavian Tone 2. In some monosyllabic words ending in a resonant, an epithetic vowel was added but did not affect tonal placement, creating tone 1
Javanese (the most spoken regional language in Indonesia) might be at the earliest state of tonogenesis. Proto-Malayo-Polynesian voiced and voiceless stops are pronounced more or less the same in Javanese, the differences are manifested in the following vowels. Just as shown in 3:32, vowels following the original PMP voiced stops are produced with lowered larynx and have a somewhat breathy phonation, called slack voice, resulting in lower pitch. The opposite happens for vowels following PMP voiceless stops, usually described as stiff voice. Some researchers liken this to the development of tones/registers in a lot of Mainland Southeast Asian languages. Maybe in the future pitch will be the only thing to distinguish them, making Javanese a fully tonal language, or maybe some other qualities will take over instead. Who knows. If you're an Indonesian, you know exactly what I'm talking about. This is what gives Javanese their distinctive 'medhok' accent.
My friend, look up Iau (it’s a natlang in Papua New Guinea). One of the few languages I keep physical copies of articles of so that I can prove I didn’t dream it up.
Chichewa would like a chat, lol (Seriously though look up Chichewa grammar on Wikipedia if you want you brain to explode, it's more in-depth than the articles on most European languages)
I think Old Chinese had a bit of this? I recall a number of characters simply being endoactive/exoactive versions of each other according to the etymology section on wiktionary.
When the 'what to do with tones' video comes out, I hope Blackfoot's system gets some coverage, as it applies contour in a very different way than we expect to see in most languages (rejecting any firm notion of tone levels) and as such is excellent fodder for inspiration. Very basically, the Blackfoot tone system begins each word with a high tone and by default lets it fall over the course of the word, (which can be pretty long because Blackfoot is polysynthetic). However, certain subsequent syllables can have a "high tone", where the tone is raised, not to the level of the start of the word, merely somewhere above the previous syllable, before continuing to fall toward the end of the word.
This is legitimately super helpful! I always shy away from using tones because it’s hard to get the hang of it all. Also I love how you gave multiple conlang starter examples! I will definitely implement tones more often in my languages
This is perfectly timed. I was just thinking of revisiting my Sino-Tibetan conlang and worrying about how to make my tones less arbitrary has been the main thing putting me off from doing it. Great work as always!
I think it's important to mention though that most tonal languages fall under a "pitch accent" and true complex tonal languages like the Chinese languages are comparatively rare.
I've thought of a tonal language where it starts doesn't matter, just whether it's rising, falling, or flat. Thus, you can get MH or LH for rising, ML or HL for falling, or M, H, or L for flat. It's relative to the start of the sound rather than relative to the mid-tone.
That's kind of how it works, as I understand it. In theory, you have these nice distinct tone levels that are the yardstick for all tones, but when you actually use tones in real speech, they slightly alter eachother. This is why tones aren't like musical notes. If I play Eb on a flute in a variety of different musical contexts, it will always sound the same no matter what other notes are near it. If I speak a rising tone that goes from level 2 to level 5 in a variety of tonal contexts, it will have allophonic differences depending on the surrounding tones. For instance if we had a rising toneme LH with an allophonic variation MH following a high tone, then the phonemes /ba(LH) ba(LH)/ would actually be pronounced [ba(LH) ba(MH)], but the distinction would never differentiate words because it's contextual. Hopefully that's not too confusingly explained.
@@Salsmachev while my friend was learning Mandarin, what I was told about tones is it's relative to the person's mid-tone during the conversation. What I was thinking of is that it doesn't matter what the person's mid-tone is, or even if it is a mid-tone, just that the sound is rising, falling, or flat. So it could be L and a MH in the same word, but the L is the same as though it were an M, but because it doesn't change, it's just a flat.
@@HuffleRuff Yes, but there's a difference between how we describe tonemes versus their actual realisation in phonetic environments. Let me give a non-tone example. In English we have both aspirated and unaspirated p, like in pot and spot respectively. They're a single phoneme though, because there's a perfectly predictable pattern to which one you use. You'd never aspirate the p in spot, for example. So aspirated and unaspirated p are allophones of a single phoneme. You can (and do) get multiple allotones of a toneme the same way you get multiple allophones of a phoneme. So a toneme that's pronounced LH by default might have a predictable MH pronunciation in certain environments as well. For the purposes of a phonemic transcription (or a language textbook), we'd just call it a rising tone and not care about whether it's MH or LH, but in a phonetic transcription (or a really good language textbook), we'd differentiate and if you used the wrong allotone, native speakers would likely notice. So, for instance, if you used MH after a high tone, it might sound weird the same way that aspirated p in an sp cluster sounds weird. If you're talking about random variation without any underlying phonetic/semantic/grammatical pattern, that's not something that's likely to occur in a natural language.
@@Salsmachev It doesn't detract from your example but musicians playing Eb in a slightly different way depending on context IS a thing, especially on (unfretted) strings. On things with fixed buttons/frets it's instead about tuning to adapt to the key you're in. It's because the modern equal temperament tuning is a compromise and actually "pure" intervals will not neatly line up with a 12 tone scale, so playing by ear on an instrument that allows it you will naturally adapt to what key or even chord you're playing at the moment
It's worth noting that while tonogenesis is a good place to start, a real tonal language has extra features, the main one being tonal sandhi. As a speaker of many tonal Sinitic languages, a tonal language without tonal sandhi comes off as VERY strange. I know of only one Natlang without tonal sandhi, and that's Cantonese. Tonogenesis mostly gets the idea of tonal evolution correct from nontonal languages, but we have little theory as to how tonal languages evolve into different tonal languages, mostly because we are almost completely theoryless on the evolution of tonal sandhi systems. For anyone starting with tonal languages, you should study some real tonal sandhi systems to really sell the language. My recommendations are standard Mandarin, Hokkien, Suzhounese, and maybe a few Bantu languages with their downstepping systems. Since we have literally no theory on how tone sandhi systems develop, you can only really check out real life examples and try to take inspiration from these Edit: Yes, even Vietnamese has tone sandhi, see my comments below. Tone sandhi is unnoticed by native speakers for the most part, but it is almost always there. Cantonese debatably may have tone sandhi, but this remains controversial depending on the definition of tone sandhi
It might be worth noting that even in Cantonese, 肥肥地 and 瘦瘦地 are pronounced fei4-fei2-dei2 and sau3-sau2-dei2 instead of fei4-fei4-dei2 and sau3-sau3-dei2. On another note - if it's not too personal of question, do you work with languages professionally? And if yes, what do you do with them? I'm also interested in linguistics and Sinitic languages specifically, although I can't yet boast with really speaking more than one of them. I'd love to have some "perspective" on what one can do with those interests.
@@bioniclegoblin6495 @Bioniclegoblin this is actually not considered tonal sandhi (arguably?). Granted I'm not familiar enough with this to say, but if you pair up the characters and become a different word AND the tones are slightly different that's not considered sandhi for various reasons. I don't know if this reduplication changes the meaning, so you'd have to be the judge. I do not work in linguistics, translating, interpretation, or language coaching. Linguistics and language learning are a hobby of mine, particularly Sinitic languages for their tonal sandhi and I am working towards learning a language of each major tonal sandhi type in China and then a few other languages in Asia. I do not professionally work with languages
@@sciencyazn1509 Thanks for your answer anyway! As for the Cantonese, while this case of tone change doesn't count as tone sandhi, I find it an interesting example of the "extra features" of a tonal language. Good luck with your language leaning!
@@d.b.2215 my comments are being deleted because I'm sending a link or something. But Vietnamese has tone sandhi. I cite: Stress and tone Sandhi in Vietnamese Reduplications by Anh-Thu T. Nguyen and John C. L. Ingram (it's online, but I can't post it because of the link ban) If you speak a tonal language natively, you are almost never going to notice tone sandhi. I didn't realize I spoke a language with tone sandhi until I started studying linguistics and Hokkien. Indeed, in Vietnamese the tone sandhi is not as prominent, but even in languages with crazy tone sandhi like Hokkien, natives do not notice any tone sandhi.
This is so much better explained than anything else I could find out there - thank you! Now I just have to go and redo all my tonal language's tones (not so thank you)...
Stupid joke I thought of after seeing the Mandarin example of tone near the beginning: I accidentally called my mother a horse so she said "Don't you use that tone with me!"
I have been looking for a video like this for so long! All of the other resources on tonogenesis that I could both find and understand were not very helpful for me. This clears up a lot for me on a practical level.
Khmer is considered a non-tonal language but the Phnom Penh dialect is developing a low dipping tone, specifically where secondary R in initial consonant clusters like "srey, trey, pram," etc. is dropping out. So srey > sey, pram > pam, with low dipping tone, etc.
The main reason I don't implement tone in a conlang is it would make it impossible for me to speak it. Cases I can deal with because I studied Latin, and morphosyntactic alignment is so intuitive to me that I accidentally made my conlang an ergative-absolutive one before I even knew what these words meant. The grammar stuff in general I can handle, it's producing sounds that don't exist exist in one of the languages I speak that I'm incapable of.
Similar for me. I could produce tones with little effort, but using them using them lexically is so counter-intuitive to me that I'd have to always read the words with tones marked to not mix anything up... and guess their meanings from those differentiated just by tone. As for the sounds, the only problem I have with consonants is mixing up those more exotic from my POV. The vowels though... I can never remember which "exotic" vowel was which, because bunches of them sound largely the same to me :D Heck, I'm not sure about the English vowels, but that's not my native language - I can't even tell which IPA vowel is the "y" in my native Polish :D So yeah, no fancy vowels nor tones for my conlangs, all the fun is in those unusual consonants.
ARTIFEXIAN DID YOU STEAL MY LINGUISTICS PAPER TOPIC Seriously though, this would have been SO helpful a couple weeks ago when I was trying to write a paper about, what do you know, tonogenesis! Thank you so much!
@@eugeneng7064 Eh... I'd describe it more as tonal languages apply tone at the syllable level, but pitch accent languages applying it at the word level. E.g. polytonic Greek only distinguishing rising/falling on the stressed syllable
@@justineberlein5916 so how does this work with tonal languages like Mandarin which have toneless syllables? Or syllables where the tone disappears? Or word level (not syllable) tone?
one idea I had was, for an avian race, the language would be, as I call it "Notal" instead of tonal, where instead of using relative "tones" you instead use absolute "notes" in one idea, the entire language is made exclusively out of notes, with no consonants or vowels at all
In Cantonese and some other southern Chinese languages you can have words / syllables which are only a nasal such as /m/ and can still carry a range of tones. I'm not sure what they derived from and how they got this way through tonogenesis.
Most syllabic consonants in Southern Sinitic had an associated vowel in their earlier stages that got deleted. However, the tone still remains after vowel deletion and attached itself to the new consonant
These videos are really high concept and rich with ideas, it's interesting to see how far evolved it's been since watching your first conlang videos. As a small but seperate suggestion, I personally am not the biggest fan of the background music, I find it somewhat distracting, so maybe if you'd like to it'd be very cool if it'd be something you'd consider an alternative to.
@@rauhamanilainen6271 Well description says Udo Grunewald. The few videos I found on YT does have some microtonal stuff and funky rythms in there so it fits, but I couldn't find the specific track, so it's not 100% that it's him (very likely though, it's a pretty unique name I'm guessing)
I don't know what it is about tone but I simply cannot comprehend any of this. I feel like what I need most is a relation back to real world languages and how they sound.
Yay!! I've been waiting for this for a while! Thanks, Edgar! :D Now I have a question. Can a reasonably synthetic language develop tones? Or is it just limited to analytic ones?
1:27 lol at Cantonese we arguably have 2 rising tones (plus common sandhi) vs 0-1.5 falling (tone 4 is analyzed as either low falling or just super low flat; tone 1 can become a high-mid falling tone in connected speech depending on the following tone) Doesn't Thai have 2 rising tones? Their "high" is mid-high rising, then a separate low-high rising. Only falling tone high-low falling. Or include the low tone as low-falling
Could you make a video like this on... basic sound changes and generating new consonats in a conlang? This was a very simple way of explaining this, and I mean that as a compliment. Most videos on sound change get complex real fast.
This actually makes sense. Tones and prosody in general are a scary part of phonology across linguistics. This is because it is almost paraverbal, and speakers create tones without even noticing, I think. At least I now understand why Cantonese has six tones!
"No natural language specifies exact pitches". Time to make a conlang like that! But first, I ought to finish my first conlang and try making a regular tonal language. It would use solfege. Either the entire community has perfect pitch or all conversations start with "do".
Random thought: Any precedent for a tonal language having different rules for diphthongs vs vowels? Or, say, linguistic evolution converting tones into diphthongs or vice versa? My gut says it's plausible and could be a fun direction to evolve a conlang.
Entirely plausible, and I know there are examples, though I can't recall specific ones at the moment. The pitch (and pitch contour) of sounds, as indicated in this video, is an important component of phonetics in all languages - it just often isn't *phonemic.* Vowels are indeed distinguished partially by pitch differences (e.g. back vowels are lower than front vowels), so it's entirely plausible these could be reanalyzed as purely tonal differences.
I learned some historical Chinese phonology and got surprised by this video. We learned about tones but in a retroconstructive ways. Just curious about does these theory are related to Chinese and if other languages are also considerd?
Yes, in Scandinavia, tones develop from peak delay. Basically it is believed that Proto-Germanic’s stress had a falling contour which was inherited into North Germanic. Proto-Germanic had many polysyllabic words that became monosyllabic in North Germanic. In the remains polysyllabic words, the falling contour shifted one syllable back, creating tone 2. Later, certain monosyllabic words gained an epithetic vowel but did not shift the tone back, creating tone 1. Pitch-accent is also a terrible label because it conflates two very distinct types of tone systems and force-fits them into a stress-like analysis.
Quick question, is Spanish a Tonal Language? It’s because there are some words that are distinguished by tone like si sí, hablo habló. However there is only one tone, making it seem more like a stress indicator than tone.
All my conlangs have ended up with tones. It just feels lıke somethıng ıs mıssıng, lıke ın thıs sentence, wıthout them. Probably tonal native language influence hehe. Oh don't forget another way to develop tonemes: Compounding! If there's an allophonic tone difference between monosyllabic and multisyllabic words, it can become phonemic by compounding. That's how Norwegian and Swedish got their pitch systems. In dialects with word-final vowel elision the system has been developed further, making minimal pairs in monosyllabic words.
Found out my native language, Afrikaans, is undergoing tonogenesis. Turns out I distinguish certain words purely by tone (leather - léér, Learn - lèèr), and tonally inflect the first syllable of words based on the voicedness of the preceding consonant (branch - ták, roof dàk) Note: the tone markings are non standard orthography. My parents do not make these distinctions
In Portuguese depending on the tone that you speak the sentence is interrogative or affirmative. Ex: Você está bem.(affirmative) ; Você está bem?(interrogative). Um olá do Nordeste(Brasil)! e Obrigado pelo video!
I wouldn't consider the Swedish pitch accent system anywhere near as complex as a tonal system. There's no consonant deletion involved, there are just variations in prosody between certain words with identical spellings and different meanings. It's extremely basic though, to the point of almost being unnecessary as context overrides its importance. One who speaks Swedish as a second language could be considered fluent long before they've mastered it, because they're only likely to do so as they begin to lose their foreign accent.
pitch accent systems and systems with largely independent tones on each syllable are very much on a sliding scale. Whilst languages like Mandarin is usually conceived of as having independent tones on each syllable, some syllables are atonal, and tone sandhi effects (which I believe will be discussed next video) lead to the tone in one syllable being at least partially controlled by the tone in another more prominent syllable. In Shanghainese (a variety of Wu Chinese, not Mandarin), this tone sandhi is so complete that the tone in the first syllable of a word almost always fully determines the tone in the following syllables (with a few cases where the earlier tones are determined by a later one instead), which sounds an awful lot like a pitch accent system obviously in Swedish the pitch accent is still carrying far less functional load (there are nowhere near as many minimal pairs as in Mandarin for instance), but the fact that something like Shanghainese can develop from the same place as Mandarin's tones (with various Sinitic varieties occupying a wide range of space between the two) and end up somewhere with a "tonal" system very close to a Baltic style "pitch accent" system (which are themselves, only a little more complicated than Swedish's, allowing a few different tones on the accented syllable) shows that these shouldn't really be considered distinct entities, but opposite extremes on a sliding scale
Theres more tonality in swedish than just pitch accent, a bunch of words can radically change meaning if you get the pitch wrong. Do it wrong, and you will end up saying weird stuff that might be hard to figure out even with decent context.
@@hamstsorkxxor That's just not true though. Let's take the number one cited example in Swedish, "anden" (the duck or the spirit). The only time you wouldn't know what was being said without correct use of pitch accent is when the word is being said alone, used in a sentence the context would always be crystal clear. Even if you said the word on its own, people would probably immediately know you're trying to touch upon the subject of pitch accent. And as mentioned above, not even all dialects of Swedish have pitch accent, but they all have these words.
Cool, I still have a conlang in my head with just a few consonants and a lot of vowels. As I could give the tones in the vowel seperate symbols, I could up the number of vowels that way. Or just (when I look at english), every different pronounciation of a vowel gets a different symbol.
Tones has meaning in Norwegian too... Bønder ber bønner over bønner... Without the use of tones, and stress changes, there's no way to differentiate the three nouns from each other
So I'm curious. Going by this origin theory, the majority of tonal languages would have simple syllable structures. Are there any natural tonal languages with CVC, CCVC, or CCVCC? 🤔
The Chatino languages have complex onsets, as well as tones. Nivkh also breaks this correlation somewhat. In the Karuk language, an interesting effect is found where the language allows both tone and complex syllables, but that placement of tone in a word largely avoids complex syllables.
Can you retain voicing distinction on some consonants while others disappear into tonality? Can your language reaquire voicing distinction after losing it to produce tonality?
I enjoy these videos simply as a linguist who isn't interested in making a conlang but who wants to understand natlang concepts better.
I think it is a good use of this sort if thing.
I think it's a good way, or at least a perfectly legitimate one. My love affair with natlang linguistics began with conlanging and each one helps me understand the other better
Funnily, when living a few years in China, as a native English speaker, I found if I tried to pronounce tones no-one could understand me, but if I ignored the tones I could make myself understood well enough. I am assuming that my attempts to use tones was muddling my pronunciation up beyond the point of comprehensibility, while when I stopped trying so hard, the listener could pick up enough of my words to use context to infer which tones were needed for what I was saying to make sense!
My opportunities to actually speak Chinese were somewhat limited too, since I was there as an English Language teacher, so my job was to speak English all day, and even away from work, anyone who spoke even a smattering of English, on realizing I wasn't actually a local, wanted to practice it. (I'm Asian, but 2nd/3rd gen Australian, even my mother doesn't speak any other language than English. Related funny story: one of my teaching colleagues was a bright-red-haired Scottish lad who had studied Chinese at university to an impressively high fluency, and if we were out together, people would assume I was his translator and try to speak to me. Which he then had to translate for my benefit!)
I can definitely see that. One of the things I've noticed is that if the use of tones doesn't yet come naturally to the speaker, it can slow down their speech to the point it can sometimes be com pared some what to spea king Eng glish one syl la ble at a time, which makes things more difficult to understand because it's less clear from the rhythm where one word ends and the next one starts, and the listener can more easily forget the earlier parts of the sentence due to the combination of the time delay and the extra cognitive load. (Also, new speakers tend to exaggerate tones more than usual, which unintuitively also tends to make it harder to understand.)
In contrast, it's easier for native speakers to understand "tone-less" Chinese more easily because tone-less Chinese is already pretty common in everyday life: Songs can't usually be sung with tones intact, so understanding song lyrics by ear necessarily requires the listener to be able to identify words from context without tone being present.
@@william_sun Interesting point regarding song lyrics. I hadn't considered that but it does make a lot of sense!
@@vikiai4241 There are other contexts as well. If you're shouting or whispering, tonal distinction can easily get lost, but the language can still be understood. A good analogy is whispering. If you whisper properly, you don't produce voiced consonants, meaning, theoretically, "Pat the bat in the bag on the back" would come out as "Pat the pat in the pack on the pack." Realistically, since combinations like that are somewhat unlikely, context normally allows it to be clear what is being intended, just as you can understand a ventriloquist even though they cannot produce many phonemes in their acts.
Chinese tones tend to merge and flow into each other in speech, unlike say Vietnamese or Cantonese, which doesn't do that. So you could be pronouncing all the tones as they would be pronounced in isolation, thereby pronouncing it unnaturally. This makes it really hard to understand.
A good example of not entirely correct but get the flow speech would be John Cena. An example of someone that tries to pronounce everything is Mark Zuckerberg. He speaks really slowly, so it helps but it still requires a lot of attention to figure out what he is saying.
@@keith6706 In English, aspiration helps. Whisper "pa, pa, pa", then whisper "ba, ba, ba". The first one's aspirated, and the second one isn't.
Artifexian: "tones appear by deleting consonants"
French: **Let me introduce myself**
oh gods a tonal French lol
bòjour, mò ami àglái
@@lukesmith8896 I tried to pronounce it, and yup, it sounds very weird 😂
French had already inherited stresses from Latin, but then managed to yeet those out later.
No lexical tone or stress, just prosodic tone/stress.
@@samuelmarger9031 yup
I have to mention Punjabi. It often doesn't get included when talking about tonal languages but it is kinda the only Indo-European language that is fully tonal.
The three tones in Punjabi are: level, low-rising and high-falling. The tones occur due to the loss of the voiced aspirated stop series. Word initial voiced aspirated stops became voiceless unaspirated stops and cause a high-falling tone in the proceeding vowel. Word medial and word final voiced aspirate stops became voiced unaspirated stops and cause a low-rising tone in the preceeding vowel. In a lot of dialects (including my own) h in the coda position is lost and causes a low-rising tone. Other than that this was an amazing video as always.
Note: I'm writing this comment based off of my memory so some of the information might be incorrect.
Punjabi native speaker and Punjabi teacher here! You are exactly right!
The letter ਘ (gh) for example, in the beginning of a word is pronounced like a kà (high-falling) and at the end like a ga (level tone).
ਘਰ - ghar /kàr/
ਸਿੰਘ - singh /sing(a)/
I was really hoping this video would mentioned Punjabi because of its uniqueness in being the only alive IE language with tones, so I'm happy to see another (presumably) Punjabi speaker in the comments
@@theguy5898 Not the only IE languages with phonemic tonality, Swedish, Norwegian, Lithuanian, Latvian, Serbo-Croatian, Slovenian and Limburgish have or at least in most dialects have tonal qualities more or less. :)
@@jasminekaram880 all of those have only pitch accent, not tone. There's a difference. Btw Serbo-Croatian speaker here.
Edit: it's Slovene, not Slovenian.
@@wtc5198 I know quite a few linguists consider “pitch-accent” to just be a restricted tonal system, especially the Norwegian-type (which includes Serbo-Croatian) where it is commonly analysed as a marked tone attaching to the stress.
Not true
I natively speak Hokkien (a Sinitic language), and what I've learned that makes it unique is that nearly every syllable has two possible tones. Let's call them R and S for the sake of labelling. Which of the two possible tones gets used depends on the syllable's function and position in a phrase/sentence. For example (in the Philippine variety, where I come from):
四 "four" - R /si˥/ or S /si˥˩/
九 "nine" - R /kaw˧˥/ or S /kaw˥/
十 "ten" - R /tsap˧/ or S /tsap˧˥/
This "two-toneness" also happens at the word level, not just at the syllable level:
十四 "fourteen" - RR /si˥ tsap˧/ or RS /si˥ tsap˧˥/
四十 "forty" - RR /tsap˧ si˥/ or RS /tsap˧ si˥˩/
十九 "nineteen" - RR /tsap˧ kaw˧˥/ or RS /tsap˧ kaw˥/
九十 "ninety" - RR /kaw˧˥ tsap˧/ or RS /kaw˧˥ tsap˧˥/
四十四 "forty four" - RRR /si˥ tsap˧ si˥/ or RRS /si˥ tsap˧ si˥˩/
四十九 "forty nine" - RRR /si˥ tsap˧ kaw˧˥/ or RRS /si˥ tsap˧ kaw˥/
九十四 "ninety four" - RRR /kaw˧˥ tsap˧ si˥/ or RRS /kaw˧˥ tsap˧ si˥˩/
九十九 "ninety nine" - RRR /kaw˧˥ tsap˧ kaw˧˥/ or RRS /kaw˧˥ tsap˧ kaw˥/
This phenomenon is not to be confused with tone sandhi (as many people do), which refers to tones changing depending on the surrounding phonetic environment. In contrast, the "two-toneness" is a purely grammatical thing; the R and S forms have nothing to do with each other. It's still a mystery to me how such a tone system could have developed naturally, and how native speakers subconsciously understand all this. I still haven't heard of any other language with this sort of "two-toneness".
I think tonogenesis is currently the best theory on how tones evolve and it's definitely best for conlangers to just stick to it but we also shouldn't forget that this current model can't explain all tonal developments.
For example Middle Chinese had 4 tones, level, rising, falling and "checked" (which may or may not have had a unique contour tone but was definitely checked by a syllable final stop). When in Late Middle Chinese voicing was starting to disappear in obstruents, this system split largely the way you'd expect but curiously the 3rd tone (called 上 sháng or shǎng "rising") following a voiced obstruent merged into the falling tone (Standard Chinese 4th tone: HL). Otherwise, the 3rd tone just became the modern day 3rd tone (HLH). What's even more mysterious is the modern distribution of the Middle Chinese checked tone after voiceless obstruents because there's actually no pattern to be found. This may or may not have been due to dialectal mixing though.
Lastly there are also other weird developments. For example in the Beijing dialect of Mandarin Chinese the MC 1st tone following a voiceless obstruent became a high level tone, while in Jinan this exact same tone became a falling-rising tone. This development is exactly opposite to the development of the MC rising tone after a voiceless obstruent: Beijing falling-rising, Jinan high level. And it's not like those dialects are that far apart from each other - just about 400km.
the irregular outcome of the checked tone after voiceless obstruents is only really found in Mandarin. Other Sinitic languages (and I believe, even more peripheral varieties of Mandarin) tend to have pretty regular correspondences, so given Mandarin's unique sociolinguistic position compared to the rest of Sinitic, the dialect mixing explanation seems pretty solid
Tonogenesis is also not the end all be all of tonal evolution. Tonogenesis can cover evolution of nontonal language to tonal languages, but tonal natlangs have tone sandhi systems which we have no good theory on their development. For example, compare Hokkien and Mandarin tone sandhi, these came from the same original language and couldn't be more different with their tone sandhi.
Further (except for Cantonese), every real tonal language has a tone sandhi system and I would argue that a tone sandhi system is what REALLY sells a tonal language, more than the existence of tones itself
Edit: Yes, even Vietnamese, see my comments below. It is difficult to notice tone sandhi if you speak a tonal language natively for various reasons.
@@sciencyazn1509 there are plenty of tonal languages that don't have any form of tone sandhi, the entire Kra-Dai family lacks it entirely as does the nearby Vietnamese
@@multinet9037 from wikipedia's tone Sandi page: "Tone sandhi occurs to some extent in nearly all tonal languages, manifesting itself in different ways."
400 kms are (better: were) a huge distance in China, unless you were on the Grand Canal... In Italy in the same distance you pass from the Gallo-Italic Lombard dialects to properly Italian Tuscan.
Your timing with this video is impeccable. Only just yesterday was I looking for resources on how tones evolve in a language and today I see that you've released this video. Many thanks. It was quite informative and I might try my hand at some kind of tonal language now. ♥
Same! I was making a tonal language yesterday!
Well, considering 13 thousand people already have watched it, the likelihood that someone just started working with tones is very likely. So I wouldn't say it was a timing, if it was inevitable.
That's a tone of information you packed in this video!
I'll show myself out.
One possibility with tones that I think is really interesting is the possibility of tonal morphemes, where a tone on a vowel in a verb can actually carry some meaning. For example, a verb might change tense/aspect/mood or a noun might change case via a shift in tone. In Tiv, for example, past tense is marked by a floating low tone while habitual is marked by a floating high tone. And in Angas, floating high tones and floating low tones attach to vowels in nouns in specific cases, that themselves already bear tone, forming contour tones if the basic lexical tone and the tonal morpheme being attached differ.
this video seems very good at discussing the tonogenesis in things like the Chinese languages, Vietnamese, and Thai, but doesn't match so well to my understanding of how the pitch accent of Continental Scandinavia developed (where it's largely to do with the presence or absence, and location of any now-lost syllables in the Early Old Norse form of the word). Unfortunately I don't know the details of which properties lead to which accent, but this seems to be the consensus in the literature. It would be interesting to see a discussion of this sort of tonogenesis in the future (and it absolutely should be considered tonogenesis, given the existence of tone sandhi, it's pretty clear that pitch accents and largely independent syllabic tones are on a sliding scale rather than entirely distinct entities)!
Iirc, Scandinavian tones developed from peak delay. Basically, stress in Proto-Germanic was believed to have a falling contour. Proto-Germanic was also polysyllabic like IE Languages. However in the transition to North Germanic, much of the unstressed syllables were lost creating many monosyllabic words. In the remaining polysyllabic words, the tone shifted one syllable back, creating Scandinavian Tone 2. In some monosyllabic words ending in a resonant, an epithetic vowel was added but did not affect tonal placement, creating tone 1
@@henrywong2725 wasn't PG stress initial? how could it have moved back?
@@wtc5198 by “back” I meant “right” :p
@@henrywong2725 ah ok that makes sense, thank you
This reminds me of my favorite real time strategy game...
Tonal War: Three Kingdoms
Tonal War: Three Contours.
When you play the Game of Tones, you tone or you die.
Javanese (the most spoken regional language in Indonesia) might be at the earliest state of tonogenesis. Proto-Malayo-Polynesian voiced and voiceless stops are pronounced more or less the same in Javanese, the differences are manifested in the following vowels. Just as shown in 3:32, vowels following the original PMP voiced stops are produced with lowered larynx and have a somewhat breathy phonation, called slack voice, resulting in lower pitch. The opposite happens for vowels following PMP voiceless stops, usually described as stiff voice. Some researchers liken this to the development of tones/registers in a lot of Mainland Southeast Asian languages. Maybe in the future pitch will be the only thing to distinguish them, making Javanese a fully tonal language, or maybe some other qualities will take over instead. Who knows.
If you're an Indonesian, you know exactly what I'm talking about. This is what gives Javanese their distinctive 'medhok' accent.
wah baru tau saya pdhl sy orng jawa wkwk
Me: *listening to this in the background while I wittle away at an unrelated task.*
"Ah, yes. Tones and toenails, and Nasal tones. Interesting."
I’m watching this while making letters for my language
So far I only got the vowels but I’m getting the plosives now
I think it would be cool to make a conlang where tones are used for declension and conjugation.
That’s called grammatical tone and it would be interesting to see in a future video. This video only discussed lexical tone.
My friend, look up Iau (it’s a natlang in Papua New Guinea). One of the few languages I keep physical copies of articles of so that I can prove I didn’t dream it up.
Iau or Turu now has a detailed Wikipedia entry if you care to take a look.
Chichewa would like a chat, lol
(Seriously though look up Chichewa grammar on Wikipedia if you want you brain to explode, it's more in-depth than the articles on most European languages)
I think Old Chinese had a bit of this? I recall a number of characters simply being endoactive/exoactive versions of each other according to the etymology section on wiktionary.
When the 'what to do with tones' video comes out, I hope Blackfoot's system gets some coverage, as it applies contour in a very different way than we expect to see in most languages (rejecting any firm notion of tone levels) and as such is excellent fodder for inspiration.
Very basically, the Blackfoot tone system begins each word with a high tone and by default lets it fall over the course of the word, (which can be pretty long because Blackfoot is polysynthetic). However, certain subsequent syllables can have a "high tone", where the tone is raised, not to the level of the start of the word, merely somewhere above the previous syllable, before continuing to fall toward the end of the word.
This is legitimately super helpful! I always shy away from using tones because it’s hard to get the hang of it all. Also I love how you gave multiple conlang starter examples! I will definitely implement tones more often in my languages
Omg I legit questioned how do language tones evolve like 2 hours ago! thank you all seeing warlock, much appreciated.
The language at 10:14 is like Cantonese. The only difference is that in Cantonese they have a low falling tone instead of a high falling tone.
This is perfectly timed. I was just thinking of revisiting my Sino-Tibetan conlang and worrying about how to make my tones less arbitrary has been the main thing putting me off from doing it. Great work as always!
I am currently working on a tonal conlang, so this video is much appreciated. Looking forwards to the next one
I think it's important to mention though that most tonal languages fall under a "pitch accent" and true complex tonal languages like the Chinese languages are comparatively rare.
Really like how concise and yet very informative and easy to follow the video is
Are you gonna go into pitch accent in part 2?
I hope he is, I speak a pitch accent language and have always wanted to create a conlang with pitch accent.
I've thought of a tonal language where it starts doesn't matter, just whether it's rising, falling, or flat. Thus, you can get MH or LH for rising, ML or HL for falling, or M, H, or L for flat. It's relative to the start of the sound rather than relative to the mid-tone.
That's kind of how it works, as I understand it. In theory, you have these nice distinct tone levels that are the yardstick for all tones, but when you actually use tones in real speech, they slightly alter eachother. This is why tones aren't like musical notes. If I play Eb on a flute in a variety of different musical contexts, it will always sound the same no matter what other notes are near it. If I speak a rising tone that goes from level 2 to level 5 in a variety of tonal contexts, it will have allophonic differences depending on the surrounding tones. For instance if we had a rising toneme LH with an allophonic variation MH following a high tone, then the phonemes /ba(LH) ba(LH)/ would actually be pronounced [ba(LH) ba(MH)], but the distinction would never differentiate words because it's contextual.
Hopefully that's not too confusingly explained.
@@Salsmachev while my friend was learning Mandarin, what I was told about tones is it's relative to the person's mid-tone during the conversation. What I was thinking of is that it doesn't matter what the person's mid-tone is, or even if it is a mid-tone, just that the sound is rising, falling, or flat. So it could be L and a MH in the same word, but the L is the same as though it were an M, but because it doesn't change, it's just a flat.
@@HuffleRuff Yes, but there's a difference between how we describe tonemes versus their actual realisation in phonetic environments. Let me give a non-tone example. In English we have both aspirated and unaspirated p, like in pot and spot respectively. They're a single phoneme though, because there's a perfectly predictable pattern to which one you use. You'd never aspirate the p in spot, for example. So aspirated and unaspirated p are allophones of a single phoneme. You can (and do) get multiple allotones of a toneme the same way you get multiple allophones of a phoneme. So a toneme that's pronounced LH by default might have a predictable MH pronunciation in certain environments as well. For the purposes of a phonemic transcription (or a language textbook), we'd just call it a rising tone and not care about whether it's MH or LH, but in a phonetic transcription (or a really good language textbook), we'd differentiate and if you used the wrong allotone, native speakers would likely notice. So, for instance, if you used MH after a high tone, it might sound weird the same way that aspirated p in an sp cluster sounds weird.
If you're talking about random variation without any underlying phonetic/semantic/grammatical pattern, that's not something that's likely to occur in a natural language.
@@Salsmachev not talking random variation.
@@Salsmachev It doesn't detract from your example but musicians playing Eb in a slightly different way depending on context IS a thing, especially on (unfretted) strings. On things with fixed buttons/frets it's instead about tuning to adapt to the key you're in.
It's because the modern equal temperament tuning is a compromise and actually "pure" intervals will not neatly line up with a 12 tone scale, so playing by ear on an instrument that allows it you will naturally adapt to what key or even chord you're playing at the moment
I've been needing this video for so long. I could never wrap my head around the concept of tones until now.
Thanks for explaining it.
It's worth noting that while tonogenesis is a good place to start, a real tonal language has extra features, the main one being tonal sandhi. As a speaker of many tonal Sinitic languages, a tonal language without tonal sandhi comes off as VERY strange. I know of only one Natlang without tonal sandhi, and that's Cantonese. Tonogenesis mostly gets the idea of tonal evolution correct from nontonal languages, but we have little theory as to how tonal languages evolve into different tonal languages, mostly because we are almost completely theoryless on the evolution of tonal sandhi systems.
For anyone starting with tonal languages, you should study some real tonal sandhi systems to really sell the language. My recommendations are standard Mandarin, Hokkien, Suzhounese, and maybe a few Bantu languages with their downstepping systems. Since we have literally no theory on how tone sandhi systems develop, you can only really check out real life examples and try to take inspiration from these
Edit: Yes, even Vietnamese has tone sandhi, see my comments below. Tone sandhi is unnoticed by native speakers for the most part, but it is almost always there. Cantonese debatably may have tone sandhi, but this remains controversial depending on the definition of tone sandhi
It might be worth noting that even in Cantonese, 肥肥地 and 瘦瘦地 are pronounced fei4-fei2-dei2 and sau3-sau2-dei2 instead of fei4-fei4-dei2 and sau3-sau3-dei2.
On another note - if it's not too personal of question, do you work with languages professionally? And if yes, what do you do with them? I'm also interested in linguistics and Sinitic languages specifically, although I can't yet boast with really speaking more than one of them. I'd love to have some "perspective" on what one can do with those interests.
@@bioniclegoblin6495
@Bioniclegoblin this is actually not considered tonal sandhi (arguably?). Granted I'm not familiar enough with this to say, but if you pair up the characters and become a different word AND the tones are slightly different that's not considered sandhi for various reasons. I don't know if this reduplication changes the meaning, so you'd have to be the judge.
I do not work in linguistics, translating, interpretation, or language coaching. Linguistics and language learning are a hobby of mine, particularly Sinitic languages for their tonal sandhi and I am working towards learning a language of each major tonal sandhi type in China and then a few other languages in Asia. I do not professionally work with languages
@@sciencyazn1509 Thanks for your answer anyway!
As for the Cantonese, while this case of tone change doesn't count as tone sandhi, I find it an interesting example of the "extra features" of a tonal language.
Good luck with your language leaning!
No tonal sandhi in Vietnamese either. We have a maximum of six tones, and by and large things stay where they are.
@@d.b.2215 my comments are being deleted because I'm sending a link or something. But Vietnamese has tone sandhi. I cite: Stress and tone Sandhi in Vietnamese Reduplications by Anh-Thu T. Nguyen and John C. L. Ingram (it's online, but I can't post it because of the link ban)
If you speak a tonal language natively, you are almost never going to notice tone sandhi. I didn't realize I spoke a language with tone sandhi until I started studying linguistics and Hokkien. Indeed, in Vietnamese the tone sandhi is not as prominent, but even in languages with crazy tone sandhi like Hokkien, natives do not notice any tone sandhi.
10:26 that is soooo close to Cantonese tone distribution, only that we have a low falling tone instead of a high falling.
Great video, was just wondering about this topic!
This is so much better explained than anything else I could find out there - thank you! Now I just have to go and redo all my tonal language's tones (not so thank you)...
Cantonese: *Defies all your rules*
Taiwanese Hokkien: "Don't you just want to go ape shitt"
Stupid joke I thought of after seeing the Mandarin example of tone near the beginning: I accidentally called my mother a horse so she said "Don't you use that tone with me!"
I have been looking for a video like this for so long! All of the other resources on tonogenesis that I could both find and understand were not very helpful for me. This clears up a lot for me on a practical level.
Khmer is considered a non-tonal language but the Phnom Penh dialect is developing a low dipping tone, specifically where secondary R in initial
consonant clusters like "srey, trey, pram," etc. is dropping out. So srey > sey, pram > pam, with low dipping tone, etc.
The main reason I don't implement tone in a conlang is it would make it impossible for me to speak it.
Cases I can deal with because I studied Latin, and morphosyntactic alignment is so intuitive to me that I accidentally made my conlang an ergative-absolutive one before I even knew what these words meant. The grammar stuff in general I can handle, it's producing sounds that don't exist exist in one of the languages I speak that I'm incapable of.
Similar for me. I could produce tones with little effort, but using them using them lexically is so counter-intuitive to me that I'd have to always read the words with tones marked to not mix anything up... and guess their meanings from those differentiated just by tone.
As for the sounds, the only problem I have with consonants is mixing up those more exotic from my POV. The vowels though... I can never remember which "exotic" vowel was which, because bunches of them sound largely the same to me :D Heck, I'm not sure about the English vowels, but that's not my native language - I can't even tell which IPA vowel is the "y" in my native Polish :D So yeah, no fancy vowels nor tones for my conlangs, all the fun is in those unusual consonants.
Learning to pronounce a new sound is considerably easier than tones I think.
Wow! That was both very efficient and inspiring. Thank you!
ARTIFEXIAN DID YOU STEAL MY LINGUISTICS PAPER TOPIC
Seriously though, this would have been SO helpful a couple weeks ago when I was trying to write a paper about, what do you know, tonogenesis! Thank you so much!
This is so helpful for a conlanger. It's a how-to guide. Thank you.
You should make a video about Pitch-Accent languages.
Pitch-accent is basically basic tone covered in the video. Some languages have additional rules while others don't
@@eugeneng7064 Eh... I'd describe it more as tonal languages apply tone at the syllable level, but pitch accent languages applying it at the word level. E.g. polytonic Greek only distinguishing rising/falling on the stressed syllable
@@justineberlein5916 so how does this work with tonal languages like Mandarin which have toneless syllables? Or syllables where the tone disappears? Or word level (not syllable) tone?
@@eugeneng7064 pitch accent languages only ever have tone on the stressed syllable
@@wtc5198 Japanese has no stress but is considered to have pitch accent. Mandarin has similar things to unstressed syllables but is not pitch accent
I was in the process of making a tonal language, this really helps! Thank you
one idea I had was, for an avian race, the language would be, as I call it "Notal" instead of tonal, where instead of using relative "tones" you instead use absolute "notes"
in one idea, the entire language is made exclusively out of notes, with no consonants or vowels at all
There is already a language like this, called Solresol.
Question: can tone loss evolve back into consonants and vowel or does it just work the other way around?
I don't know. The latter is much more likely, but I think it would be interesting to see consonants influenced by tones, followed by tone loss.
IIRC the unaspirated stops in Navajo are allphonically voiced before low tone-vowels but voiceless before high tone-vowels.
It's more common for tones to just disappear without a trace, causing a bunch of mergers.
Often when time is lost the artifacts are phonation (creaky voice, breathy voice etc) on the vowels.
In Cantonese and some other southern Chinese languages you can have words / syllables which are only a nasal such as /m/ and can still carry a range of tones. I'm not sure what they derived from and how they got this way through tonogenesis.
Most syllabic consonants in Southern Sinitic had an associated vowel in their earlier stages that got deleted. However, the tone still remains after vowel deletion and attached itself to the new consonant
That is so cool.
I can't help thinking about easy ways of noting tone in writing. I don't have conclusions, bot some thoughts.
These videos are really high concept and rich with ideas, it's interesting to see how far evolved it's been since watching your first conlang videos. As a small but seperate suggestion, I personally am not the biggest fan of the background music, I find it somewhat distracting, so maybe if you'd like to it'd be very cool if it'd be something you'd consider an alternative to.
New video, nice!
I like the use of microtonal music in this vid about tones
Interesting observation. Would you perhaps have a link to the song (or at least the song name)?
@@rauhamanilainen6271 Well description says Udo Grunewald. The few videos I found on YT does have some microtonal stuff and funky rythms in there so it fits, but I couldn't find the specific track, so it's not 100% that it's him (very likely though, it's a pretty unique name I'm guessing)
5:58 that feels like a very random thing to say
dude thank you so much, I was looking for a guide for this stuff
I don't know what it is about tone but I simply cannot comprehend any of this. I feel like what I need most is a relation back to real world languages and how they sound.
Love the Irish accent joke in the title there.
Yay! Finally tones!
Very informative. I love your videos:)
Yay!! I've been waiting for this for a while! Thanks, Edgar! :D Now I have a question. Can a reasonably synthetic language develop tones? Or is it just limited to analytic ones?
Navajo has (register) tone. Most pitch accent languages are synthetic. I'm not sure about real contour tone languages
@@wtc5198 thanks :)
Will you talk about Swedish-like tone systems or only this stuff, just wondering?
1:27 lol at Cantonese
we arguably have 2 rising tones (plus common sandhi) vs 0-1.5 falling (tone 4 is analyzed as either low falling or just super low flat; tone 1 can become a high-mid falling tone in connected speech depending on the following tone)
Doesn't Thai have 2 rising tones? Their "high" is mid-high rising, then a separate low-high rising. Only falling tone high-low falling. Or include the low tone as low-falling
Complex tonal languages are actually extremely rare. The most of em gave pitch accents which are accents and less required to speak said language.
First time I've caught the preview
Another excellent video
In the game of tones, you win or you die.
I did make a Lang family with split tones this year and it really warped my brain lmao
Where is Part 2?
Please, make it!
Could you make a video like this on... basic sound changes and generating new consonats in a conlang? This was a very simple way of explaining this, and I mean that as a compliment. Most videos on sound change get complex real fast.
"Gravy, now let's mash..."
Thank you, now I'm hungry.
This actually makes sense. Tones and prosody in general are a scary part of phonology across linguistics. This is because it is almost paraverbal, and speakers create tones without even noticing, I think. At least I now understand why Cantonese has six tones!
"No natural language specifies exact pitches". Time to make a conlang like that! But first, I ought to finish my first conlang and try making a regular tonal language.
It would use solfege. Either the entire community has perfect pitch or all conversations start with "do".
I saw a private video in the your language playlist yesterday and knew something was comming
You should make a video on Animal and Colour names.
I don't know why I thought this was going to be about color tones and vexillology lol
Put all the examples of "ma" into a single sentence, and you have, "Did you scold the stoned mother on the horse?"
They are either rising "ma?" or falling [villager sound]
So I thought this was going to be about colour tones, and either about where dyes come from, or colour names in language
Great video.
Random thought:
Any precedent for a tonal language having different rules for diphthongs vs vowels? Or, say, linguistic evolution converting tones into diphthongs or vice versa? My gut says it's plausible and could be a fun direction to evolve a conlang.
Entirely plausible, and I know there are examples, though I can't recall specific ones at the moment. The pitch (and pitch contour) of sounds, as indicated in this video, is an important component of phonetics in all languages - it just often isn't *phonemic.* Vowels are indeed distinguished partially by pitch differences (e.g. back vowels are lower than front vowels), so it's entirely plausible these could be reanalyzed as purely tonal differences.
Burmese, I think, some language in that area I'm sure but I think it's Burmese, maybe Thai
I learned some historical Chinese phonology and got surprised by this video. We learned about tones but in a retroconstructive ways. Just curious about does these theory are related to Chinese and if other languages are also considerd?
Where is the second video about tones? I cant seem to find it...
ua-cam.com/video/iaHYIXBvL94/v-deo.htmlfeature=shared
Here
I have only ever done the simplest of those ways, I did not even know everything else.
Could you do a video on pitch-accent and how it develops? For example, I reckon that Swedish didn't get its pitch accents this way [tonogenesis].
Yes, in Scandinavia, tones develop from peak delay.
Basically it is believed that Proto-Germanic’s stress had a falling contour which was inherited into North Germanic.
Proto-Germanic had many polysyllabic words that became monosyllabic in North Germanic.
In the remains polysyllabic words, the falling contour shifted one syllable back, creating tone 2.
Later, certain monosyllabic words gained an epithetic vowel but did not shift the tone back, creating tone 1.
Pitch-accent is also a terrible label because it conflates two very distinct types of tone systems and force-fits them into a stress-like analysis.
Quick question, is Spanish a Tonal Language? It’s because there are some words that are distinguished by tone like si sí, hablo habló. However there is only one tone, making it seem more like a stress indicator than tone.
No. It is stress, not tone.
All my conlangs have ended up with tones. It just feels lıke somethıng ıs mıssıng, lıke ın thıs sentence, wıthout them. Probably tonal native language influence hehe.
Oh don't forget another way to develop tonemes: Compounding! If there's an allophonic tone difference between monosyllabic and multisyllabic words, it can become phonemic by compounding. That's how Norwegian and Swedish got their pitch systems. In dialects with word-final vowel elision the system has been developed further, making minimal pairs in monosyllabic words.
Found out my native language, Afrikaans, is undergoing tonogenesis. Turns out I distinguish certain words purely by tone (leather - léér, Learn - lèèr), and tonally inflect the first syllable of words based on the voicedness of the preceding consonant (branch - ták, roof dàk)
Note: the tone markings are non standard orthography. My parents do not make these distinctions
Nice. I was just wondering if I wanted tones in my language.
In Portuguese depending on the tone that you speak the sentence is interrogative or affirmative.
Ex: Você está bem.(affirmative) ; Você está bem?(interrogative).
Um olá do Nordeste(Brasil)! e Obrigado pelo video!
That is the same in just about every non-tonal language.
Tones in Sinitic and some Southeast Asian languages (Viet and Thai) developed as from more complex consonant clusters.
tonal genesis evangelion
I wouldn't consider the Swedish pitch accent system anywhere near as complex as a tonal system. There's no consonant deletion involved, there are just variations in prosody between certain words with identical spellings and different meanings. It's extremely basic though, to the point of almost being unnecessary as context overrides its importance. One who speaks Swedish as a second language could be considered fluent long before they've mastered it, because they're only likely to do so as they begin to lose their foreign accent.
Not all variations of Swedish even have this semi-tonality, whereas all forms of Chinese are tonal.
pitch accent systems and systems with largely independent tones on each syllable are very much on a sliding scale. Whilst languages like Mandarin is usually conceived of as having independent tones on each syllable, some syllables are atonal, and tone sandhi effects (which I believe will be discussed next video) lead to the tone in one syllable being at least partially controlled by the tone in another more prominent syllable. In Shanghainese (a variety of Wu Chinese, not Mandarin), this tone sandhi is so complete that the tone in the first syllable of a word almost always fully determines the tone in the following syllables (with a few cases where the earlier tones are determined by a later one instead), which sounds an awful lot like a pitch accent system
obviously in Swedish the pitch accent is still carrying far less functional load (there are nowhere near as many minimal pairs as in Mandarin for instance), but the fact that something like Shanghainese can develop from the same place as Mandarin's tones (with various Sinitic varieties occupying a wide range of space between the two) and end up somewhere with a "tonal" system very close to a Baltic style "pitch accent" system (which are themselves, only a little more complicated than Swedish's, allowing a few different tones on the accented syllable) shows that these shouldn't really be considered distinct entities, but opposite extremes on a sliding scale
Theres more tonality in swedish than just pitch accent, a bunch of words can radically change meaning if you get the pitch wrong. Do it wrong, and you will end up saying weird stuff that might be hard to figure out even with decent context.
@@hamstsorkxxor That's just not true though. Let's take the number one cited example in Swedish, "anden" (the duck or the spirit). The only time you wouldn't know what was being said without correct use of pitch accent is when the word is being said alone, used in a sentence the context would always be crystal clear. Even if you said the word on its own, people would probably immediately know you're trying to touch upon the subject of pitch accent. And as mentioned above, not even all dialects of Swedish have pitch accent, but they all have these words.
@@hamstsorkxxor how isn't that pitch accent?
this is the name of a futurama episode
Does tonogenesis from voicing distinction also work in codas?
Cool, I still have a conlang in my head with just a few consonants and a lot of vowels.
As I could give the tones in the vowel seperate symbols, I could up the number of vowels that way.
Or just (when I look at english), every different pronounciation of a vowel gets a different symbol.
Tones has meaning in Norwegian too... Bønder ber bønner over bønner... Without the use of tones, and stress changes, there's no way to differentiate the three nouns from each other
Artefexian: Gives an extremely complex tonal language with 6 tones.
Cantonese: _Laughs in 9._
Technically Cantonese has 6 tones. The additional three come from tones 1 3 6 placed into a syllable with a stop coda
Thank Rakl'tena for Artifexian. This is arriving just in time for a group of conlangs in my story.
Might you please consider making a Video on Diminutives and Augmentatives at some point in the future?
i'm putting basic tone distinction in my conlang Xotan [ɕotæn]. Thanks!
So I'm curious. Going by this origin theory, the majority of tonal languages would have simple syllable structures. Are there any natural tonal languages with CVC, CCVC, or CCVCC? 🤔
The Chatino languages have complex onsets, as well as tones.
Nivkh also breaks this correlation somewhat.
In the Karuk language, an interesting effect is found where the language allows both tone and complex syllables, but that placement of tone in a word largely avoids complex syllables.
@@the_linguist_ll Neat.
Yes, but I think that that's usually if those languages had tone to start with and therefore didn't have to go through consonant loss
Swedish
just as im working on a tonal language family :))
What about the other way? How do tones affect other sound changes?
“All languages make use of pitch or tone”
Signed languages: “am I a fucking joke to you?!”
Lol
Can you retain voicing distinction on some consonants while others disappear into tonality? Can your language reaquire voicing distinction after losing it to produce tonality?
1:24 He sounds like A minecraft villager. LOL
Interestingly, my main conlang that I am working on is non-tonal, but evolved from a tonal language.