Why Are Sound Changes Systematic?

[@darynvoss7883]

You've done my head in with your hour/our distinction. The only similar case I can think of in my Australian dialect is the word "gone", which by rights should rhyme with con or don, but has a significantly longer vowel that is also somewhat further forward.

[@TovarichBramble]

Glad I’m not the only one! The ant / aunt is also another one

[@enricobianchi4499]

Is it different from the vowel in "corn"?

[@darynvoss7883]

@@enricobianchi4499 completely

[@Old_Catholic]

Do you think, like me, that Kath and Kym invented a new Aussie dialect that didn't exist prior?

[@darynvoss7883]

@Old_Catholic oh it existed...

[@seyeruoynepotsuj]

I'd love to see a linguist progress a medieval conlang (such as one of Tolkien's languages) to see what it might be spoken like in that particular IP's 'modern' setting!

[@unvergebeneid]

I mean Tolkien himself already progressed his languages over the millennia of his world. Of course that doesn't extend to a modern world with Elvish cars and Dwarven computers but if you're just interested in conlangs with an invented history to them, Tolkien is actually a good start.

[@MadhanBhavani]

Something similar I was wondering about: Let's say a kid in the US deep south grew up around parents that distinguish w and wh. Of course, all wh question words are going to be used regularly at home. But what about words like wheezing, whale etc that the kid probably will hear from their schoolteachers first (let's say they don't have the w, wh distinction)? In that case, might they pronounce only common wh words as wh, but words like wheezing with a w, thereby creating an incomplete merger?

[@pattap2826]

That’s likely

[@spooderman9122]

This sounds like something that Jackson Crawford might have in his speech given that he distinguishes between wh and h

[@WGGplant]

I had something of the opposite happen. My mother is from Chicago and I grew up in Alabama. I have a pretty thick southern accent mostly picked up from friends and teachers at school (with wh w distinction in most words). If something like that could come out of my situation, I see no reason why different influences couldnt easily cause a different outcome. There are probably tons of kids who fit your scenario, especially the more urban you get, where teachers are less likely to have much of a regional accent, and families are more likely to have come from other places. In my case though, I've picked up dialectal words my mother has never used, intonation, and (un)grammatical structures from friends. Things a teacher would purposefully avoid teaching, even in the south. But my mother doesn't use them, and i dont have a dad.

[@Glassandcandy]

Children generally don’t pick up language from their parents exclusively, they pick up language from their community/peer-group more than anything. A child in a household that differentiates w and wh likely will not do so if their peer groups external to the household do not. This why first generation anglophones do not pick up their parents accents, even if both of their parents speak with a heavy accent from their mother tongue.

[@gav7428]

@@spooderman9122 He does, he's commented on it before, saying he picked it up from his grandparents

[@qeithwreid7745]

Engaging and humble as always

[@dukereg]

I'd love to hear an explanation of why these processes don't reach a fixed point where they don't change further. At first glance even a local minimum of cost would imply stability for a gradual process like this. Yet every example of an "inefficient" sound or combination developed from previous sounds, again and again across time. Why aren't languages just cycling through increasingly small groups of more efficient sounds? What prevents language settling into small cycles or terminal nodes on the graph of sound change?

[@SylviaBounds]

this isn't really relevant to the central topic but what you said about people coming to expect the new pronunciation as normal made me realize that here in ohio US, many people raise the cat vowel in all contexts, but still keep the can vowel seperate by raising that one even further, and some i have even heard having a more open cap [kʰæp̚] vowel, a slightly raised cat [kʰɛæ̯t̚] before all alveolar spirants, and a sometimes monophthong can [kʰẽn] or [keə̃n] where the nasality is really just incidental, as people understand which phoneme youre using just be the /e/ quality. some speakers have the same raising before velar consonants, but i live far enough south that this has a heavy stigma as a noticeably "midwest" feature, most I think would be more likely to have the vowel breaking if any raising at all or stubbornly sticking to the open æ. I'd be interested to hear your thoughts on the role stigma and social pressures play on the development of people's idiolects

[@BurstVessels]

I don't know enough about linguistics / IPA to to properly describe it, but I have a funny story for you: I knew a young woman from Ohio, from an Italian-American family. She pronounced most of her "cats" with a very typical midwestern accent, but when she said the title of the TV program "The Sopranos" (which she professed to never have watched), she pronounced the "a" with an extremely uncharecteristic "open" sound (to match a proper Italian pronunciation 😂). Very funny example to me of how language can provide a window into psychology, culture, and history.

[@bob___]

Good commentary. We have an ahr/ower pairing for "our" in North America, as well (at least in the part of North America where I'm from, near Washington, D.C.), with the emphatic version often being pronounced as two syllables. I think that the distinction between emphatic "our" and "hour" is felt even when it isn't really heard. "Hour" is less likely to be pronounced as having two syllables. Also, there are local variations on the pronunciation of the "ou" in "hour" and "house," where some speakers in Baltimore may pronounce "house" almost like the Dutch "huis," but I don't think that necessarily applies in the same way to emphatic "our."

[@dudejoe8705]

This is quite interesting to me! I'm from Tennessee (though I'm younger and have more of a general American accent closer to that of Illinois than Tennessee, but certain regional pronunciation differences still come out ofc), and for me "Hour" is much *more* likely to be pronounced as having two syllables while "our" is more likely to be pronounced with one syllable "ahr". Regional accent differences are fascinating! Ever since I consciously learned about weak vs strong forms which most native speakers just pick up intuitively without ever thinking about it, I've been quite fascinated by which situations we might use weak vs strong, and *why* we might use it (in my accent, it's useful to have "hour" to be *more* likely to be pronounced as two syllables than "our", because strong "our" is pronounced the same!). I do think I get what you mean with "feeling" the distinction though even when phonetically ambiguous, I think that has to do with our brains natural ability to pick up on grammatical context, and context in general. Quite amazing if you ask me!

[@Eronoc13]

Young person from Western Montana (understudied region linguistically, we're at a bunch of crossroads for accent and dialectical features but the first academic speech survey here was only in the 2010s), I'd say my "our" is almost exclusively unstressed, for me something like [ˈäɹʷ]. It's exactly homophonous with "are". I'm _aware_ of a stressed form [ˈawɚ] (two syllables, entirely homophonous with "hour"), but I wouldn't ever produce it unconsciously. I find it somewhat interesting that for me, the actual quality of the first vowel has become different in "our" and "hour". If I had to guess, it's due to the R in monosyllabic "our" affecting the preceding vowel in some way. I'd even say that the /a/ in my "hour" is a bit raised, like it wants to become an /æ/, which distinguishes the two words even more. I guess I know what I'm going to be paying attention to in my peers' speech this week.

[@ajs41]

A dinner lady at my school (who I think was from the Stoke-on-Trent area originally, which is about 30 miles north of where the school was) used to do a similar type of thing with the word "sure" as with "our". When she was emphasising the word it would have two syllables.

[@IllidanS4]

How does "sure" get two syllables‽

[@amandachapman4708]

​@@IllidanS4 Like this: shoo-ehr. That's how I say it, as opposed to shaw as it seems Standard English has it

[@watchmakerful]

[12:30] Even more, this "glide" in the [æ] sound is DIFFERENT before [ŋ] and before [n]/[m], so "ban" vs. "bank" have two distinct vowels. A similar phenomenon occurs in Hebrew, where almost any stressed vowel (except [a]) before a word-final [ħ] acquires an unstressed [a] glide.

[@enricobianchi4499]

Nice observation! I think it's due to front vowel raising before velars interacting with front vowel raising before nasals. Why the mid-height offglide to /æ/ is central rather than front might be because of the tendency in English for historically short vowels to have a central offglide when emphasized.

[@frankierose528]

Your hour/our distinction is interesting because there is an a / a: distinction for some people (I think in Southern England) - bag, bad and sad can be pronounced with a longer vowel than lad or pad. I don't know if this would be considered phonemic though.

[@rdklkje13]

Thank you for (indirectly) illuminating the ongoing Scandinavian debate about Danish. Of course we don't walk around with potatoes in our throats as the Swedes and Norwegians would have it. We're just being incredibly efficient with our vocal muscles! And becoming ever more efficient at an astonishing rate. These days I need subtitles to understand the kids on tv (but they seem to understand each other). Heck, a part of me wouldn't be at all surprised if Danish becomes a vowel-and-stød language with no consonants left in just a few more generations 😅

[@overlordnat]

It’s interesting why sound changes for individual words occur in certain accents rather than for all words where that change could apply. Consider how ‘tooth’ has the vowel of ‘foot’ not ‘loose’ in Birmingham and South Wales, the same is true for ‘room’ and ‘groom’ in RP and Cockney and for ‘roof’ and ‘hoof’ in some American accents but only in Northern Ireland would ‘oo’ be consistently closer to FOOT than LOOSE for all words (and the opposite in Scotland of course)

[@mikedavidson1970]

Our friends , starting in an hour . will be arriving hourly at our house. Here in Ohio older speakers can have stressed Our.

[@TSGC16]

Ohio💀

[@greenockscatman]

Sounds like the brain is a lot like a neural network (like a large language model) when it comes to language. Not super surprising, I guess, but it's definitely interesting to see a lot of the same concepts that apply to training AI models come into play when Simon's talking about language learning.

[@jlewwis1995]

I mean yeah, that's where people got the idea for neural network based AI from to begin with, why waste time and effort trying to come up with something original when you can try to use what you know already works

[@theblitz6794]

Lmao "the brain is a lot like a neural network" OH YOU DONT SAY 😂😂😂

[@bobbyg1068]

There's an xkcd cartoon about how any time you do a task manually you can technically say "I trained a neural net to do it" for CS credit

[@vampyricon7026]

Perhaps I'll be a bit critical of this video: I'm not sure why you wouldn't consider [awə aːwə] two separate phonemes (well, [a aː], anyway). At least to me, I've always thought that as long as there's a sound system with minimal pairs, you could construct a phonology out of it. It's just that typically people speak of phonologies of languages rather than idiolects since that's of more general interest. On a similar note, there are several Chinese languages in which [m̩] only appears in one syllable meaning "not" (e.g. in early-20th-century Cantonese, [m̩²¹], with [¹] being the lowest tone height and [⁵] being the highest), and yet it still means something different and contrasts with other syllables. I don't really see the difference between this and the [awə aːwə] contrast, if it's a marginal phoneme. I'm also unsure how much the discussion of calorie use contributed to the video. I think mentioning the sharing of neural circuitry and the assimilation of phones to nearby ones via "convenience" or "laziness" is sufficient, and while minimising calorie use is true, it's too general to be a good explanation. (As the physicist Sean Carroll has joked, everything can be explained with "Because of the laws of physics and the initial conditions of the universe", but that's hardly enlightening.) Still, thank you for the video. It was a good one :)

[@patdmulroy99]

Interesting point about “chaff.” I have a similar theory about the Spanish “joder” which is their F-word. The regular sound changes would predict “hoder” and the H would be silent. However the current spelling reflects an older pronunciation that was common to most words now beginning with H. The normal path was F > H (pronounced as english) > H (silent). E.g. Filius > Fijo > Hijo > (h)ijo Given that the current J sound in Spanish is pronounced like the older H would have been, it seems Joder is frozen in time.

[@SirBoggins]

Last time I was this early, we were still speaking "Anglosaxon!" 😂

[@thecozyconstellation]

🤣 love it

[@varana]

Next video: How the social environment affects all of this - social groups, peer pressure, status perception, regional identity, movement patterns, media, and writing. :D

[@HighWealder]

I'm very aware of how my pronounceation and speech patterns change talking to different people. A friend is from Essex and my whole lexicon and accent somehow shifts when speaking to him, but I don't do it on purpose.

[@haydenismondo]

My son says “NewTube” instead of KZbin, and it’s so tempting not to correct him because I think it’s so cute. He also says “phong” instead of “phone.” Like saying the word phone but with a nasal g sound at the end! It’s so interesting how children form their own sounds and then have to break them later on to learn the “correct” way!

[@AnthonyRosequist]

"[the set of phonemes] can't be too small; take that to the extreme and you'd just be pronouncing all words identically, in which case nobody would be able to decode what you were trying to say" And thus French was born.

[@AdDewaard-hu3xk]

The fact that I don't pronounce the words on the screen the way that you do says a lot.

[@GregoryKC]

In my opinion, it’s the children who are mostly involved in the process of phonetic change, because I don’t know if adults, once used to a certain set of phonemes, ever change this set in their lifetime. Children, however, are new listeners of the language and tend to hear it with a kind of “noise”, resulting in a slight degree of deviation in the way they reproduce the phonemes of a given language. What do you think?

[@qeithwreid7745]

In my accent I can’t make the “why Y” pun because I aspirate the WHy in a big whistle way as a Doric speaker. They sound different to me. I get it as a pun in other languages.

[@seinmstudio]

Simon, please do a breakdown of Tolkien's accent! I beg thee.

[@twig5543]

With the 'chaff' example, surely city dwellers would have had cause to talk about birds, in which case you'd think 'chaffinch' would be pronounced chahff-inch. Do you have any theories about why that didn't change? Or did it but it didn't stick long-term?

[@twig5543]

I love these videos, by the way, both the linguistic and philosophical subjects. An instant watch as soon as one drops.

[@Roland-pw5xj]

There is the idiom, "Sorting the wheat from the chaff."

[@twig5543]

​@@Roland-pw5xj That's where i know the word 'chaff' from. However, I was commenting on the fact that the same syllable appears in 'chaffinch' but did not change even though city folk would likely have used it. Regardless of 'chaff' being used in an idiom, it's a little odd that chaffinch didn't follow the same pattern as the other vowel sounds. Perhaps it's because of the rhythm of the word is different or some other reason, but it's interesting nonetheless.

[@amandachapman4708]

Consider these words: daffodil, faff, gaff, raffish, yaffle.

[@jeff__w]

2:16 “If the individual people aren't paying attention to the sound change why is it so orderly and neat?” I enjoyed the video but I found the assumption that unconscious sound change (repeated a few times at the beginning) would somehow _not_ be systematic (or “orderly and neat”) just, well, _weird._ There’s no reason that various language sounds would not change over time-they sort of “fall towards” a new, “easier” or more efficient-and all speakers in a given language community would be subject to the same pressures giving rise to those changes, whether they knew about them or not. Almost all of _how_ we speak is pretty unconscious, anyway-linguists and phoneticians uncover the patterns and extract the “rules”-and there would be no reason to think that these types of sound changes are any different.

[@kitoglaw8785]

Fascinating video, as always. Just wanted to ask what is meant by suedes of historical linguistics 0:33 . I'm not a native so I might've probably misheard it or something, the subtitles recognize the word as previously stated so if someone be able help, I'll be most grateful

[@TheTedder]

I believe he said swathes.

[@kitoglaw8785]

@@TheTedder thanks a lot, appreciate it

[@Jeremy-hu7yq]

If the brain wants to be as efficient as possible insofar as it uses the smallest possible set of different phonemes/unique sounds, then does that somehow imply that some languages are more efficient than others depending on the number of sounds they 'use'? For instance say, Spanish as opposed to click languages of south Africa (I'm sure this is much oversimplified though). Also what about pitch accent or tones in certain languages like Vietnamese or Mandarin Chinese? I assume these should each be considered different phonemes and not allophones of the same sounds. Sorry if this is overly naïve or nonsensical.

[@nightmaretraveller]

Hi. I know it's a rather random question to ask, but do you think you have the lad-bad split?

[@DrPumpkinz]

How do linguists figure out how people spoke in times before audio recording?

[@paradoxmo]

In the previous phonology video he went over this- search for “Are historical accent reconstructions just nonsense?”

[@LordJazzly]

Coming from what linguistic I did years ago in uni: It's a little bit of complicated analysis of rhyme words and spelling mistakes, and a _lot_ of 'this is all relatively new information we have learnt _specifically because_ we have recordings of historic speech now, and the amount people's speech patterns have changed even since the beginning of the 20th century is amazing'. There were _theories_ about sound change before, but they were controversial and subject to a lot of debate before longditudinal studies of recorded audio were possible.

[@AllotmentFox]

Have you got a new lens? Or are you colour-grading that what looks like a 70s lens? When you said there, pair, fair the first time I said that’s how I say it and then you did the second form and I said, yes, that’s really the way I say it. I now don’t know how I speak

[@LordJazzly]

You might speak both ways; people can and do maintain different speech patterns in different contexts. This is most evident when it comes to bilingualism - when someone switches into an entirely different langauge, it's obvious, because their comprehensibility to different people will change completely - but it can happen with speech registers and dialects as well. If you've ever had a friend who came from a regional area with a distinctive dialect, or had a family that had a distinctive accent, and you noticed that person spoke 'unaccented' English with you but spoke 'with an accent' when talking to their family (quote marks around 'accent' here, because accents are complicated) - that's a similar thing. People can internalise more than one way to say the same thing; it's weird, but cool.

[@AdDewaard-hu3xk]

Tell me how to pronounce 'aunt'. I have no clue.

[@aureltoniniimperatorecomun4029]

An important question is: why the phonemes can't evolve in precedent forms? A vowel could simply mutate in o- u and - o again, but this is impossible, instead is possible something as o- u-ü - ju ecc. It seams the speakers have a " memory" of the precedent forms of the phonemes. This is maybe false for revitalized languages as hebrew or cornish

[@insectoid_creature]

What makes you say phonemes never revert to older forms?

[@WGGplant]

it is 100% possible for a vowel to shift back to a previous form. It's happened in English with the diphthongization of certain vowels.

[@EzraSisk]

The body of literature on sign change is itself I’m guessing

[@nightmaretraveller]

2nd