It's "sodder", not "soLder", damnit! (for us USAians)

[tooki]

And whose pronunciation do you use?  In American English, "bitter" and "bidder" are pronounced the same way.  And let's not even start the "Mary, merry, marry" thing again 

No, TimFox is absolutely correct: they are not the same. Very similar, but not the same. The tt in "bitter" is an unvoiced flap, while the dd in "bidder" is a proper voiced dental stop.

But can you say that with certainty? Doesn't it depend on the speaker?
I'd say you're correct when the speaker is using the "proper" pronunciation for "bitter" as opposed to "bidder", but to my ear,  I'd say most speakers here (U.S.) are not so precise and pronounce them almost identically. Pretty sure I do so myself.

No, what I described is how I think a typical speaker says it in everyday speech, not in a deliberate “proper” pronunciation, which would be a voiceless dental stop, not the voiced one I described.

[tooki]

No, TimFox is absolutely correct: they are not the same. Very similar, but not the same. The tt in "bitter" is an unvoiced flap, while the dd in "bidder" is a proper voiced dental stop.

Are we describing or proscribing pronunciation?  I can assure you that in casual speech, this native speaker of American English pronounces "bitter" and "bidder" identically. 

I do understand and can hear the difference between flaps, stops, etc.: when I was studying Linguistics I had phonology examinations where I was graded on my ability to identify, differentiate, and produce all kinds of sounds.  Flaps vs. stops was an easy one (compared to, say, all the different "clicks" in Zulu).

Sure, there are some (many?) people who do pronounce these differently and think they are making the same sound - but I'm not one of those people   I can post a recording if you like.

I am describing.

[paulca]

While we are there...

inflammable.

On first meeting this and other "in" prefixes, even as a native speaker, its likely you will take the inverse meaning initially.  Until someone points out the "in" means MORE not less and it's not "un" either.

"Infamous"

Yet... when we come to "intolerable" does not make the person "highly tolerable".

[IanB]

Well, inflammable comes from inflame.

Therefore intolerable comes from intole? Except there is no such word. So inflammable and intolerable cannot be compared.

[golden_labels]

Also let’s forget not^(corr:1) that the pronunciation given in dictionaries is an approximation. No phonetic notation accurately expresses minuscule details of actual speech.

Trying too hard to follow such approximations, against how people really speak, is an error. Often a case of hypercorrection.

^(corr:1) Corrected ~9 hours later.

[TimFox]

While we are there...

inflammable.

On first meeting this and other "in" prefixes, even as a native speaker, its likely you will take the inverse meaning initially.  Until someone points out the "in" means MORE not less and it's not "un" either.

"Infamous"

Yet... when we come to "intolerable" does not make the person "highly tolerable".

The original word in English is “inflammable”, from the verb “to inflame”, equivalent to French and Spanish usage.
Mistaken opinion that “in-“ means “not” in this context caused American insurance companies to insist on “flammable” instead.  The correct word for “not burnable” is “nonflammable”.
The medical term is inflammation, never “flammation”.

[Aldo22]

Why do you actually have to specify a language when using IPA readers?
I thought that wasn't supposed to matter with IPA?

https://ipa-reader.com/
https://www.leskoff.com/s01763-0
etc.

Try ɬanˈɡoɬen

[IanB]

Why do you actually have to specify a language when using IPA readers?
I thought that wasn't supposed to matter with IPA?

Because the IPA is an approximation.

Two words with the same IPA representation can sound different when spoken. There can be subtle differences in vowel length, tone, and accent even though the phonemes are nominally the same.

[Aldo22]

Why do you actually have to specify a language when using IPA readers?
I thought that wasn't supposed to matter with IPA?

Because the IPA is an approximation.

Two words with the same IPA representation can sound different when spoken. There can be subtle differences in vowel length, tone, and accent even though the phonemes are nominally the same.

But doesn't that make it a little, um... pointless?

If I select the Italian voice, I won't have a clue how to pronounce “Llangollen” in Welsh.

I thought the IPA was meant to get around that problem.

[Analog Kid]

Why do you actually have to specify a language when using IPA readers?
I thought that wasn't supposed to matter with IPA?

Because the IPA is an approximation.

Two words with the same IPA representation can sound different when spoken. There can be subtle differences in vowel length, tone, and accent even though the phonemes are nominally the same.

But doesn't that make it a little, um... pointless?

If I select the Italian voice, I won't have a clue how to pronounce “Llangollen” in Welsh.

I thought the IPA was meant to get around that problem.

As did I. Otherwise, isn't it kind of ... useless as a guide to pronunciation?
I thought IPA was supposed to be a universal description of human pronunciation, or at least as close an approximation to that as possible.

[IanB]

The IPA is meant to be used academically as a guide to the sounds found in human languages, and to describe them in terms of the mouth shape used to make them.

As a guide, it can help you understand approximately how to make the sounds, but it can never tell you how to speak words in a native accent.

Saying it is useless because it cannot do everything an audio recording can do is a bit of a strange statement, is it not?

"It cannot do everything, therefore it is useless and not worth having at all." 

[Analog Kid]

As a guide, it can help you understand approximately how to make the sounds, but it can never tell you how to speak words in a native accent.

Well, why not?
What is lacking in the descriptive scope of IPA phonemes that prevent that from being possible?
Is it levels of subtlety that go far deeper than possible with describing individual sounds that humans can make with their mouths?
How precise would one need to be in order to make it possible to do what you said (speak words in a native accent)?

[golden_labels]

But doesn't that make it a little, um... pointless?

If I select the Italian voice, I won't have a clue how to pronounce “Llangollen” in Welsh.

I thought the IPA was meant to get around that problem.

Because if a typical English speaker wishes to pronounce Polish sz, ś, ż/rz, ź, cz, dż, drz, or dź,⁽¹⁾ it is very hard to explain them using English phonology, further complicated by different dialects of English, and it’s impossible to write it down in a compact manner. Equally, try to express Czech ř in English.

Phonetic alphabets, like IPA, offer a way to know what the sound even is. It’s an approximation only, but otherwise you’d have no idea at all. If you don’t believe, guess these words: “dzisiaj”, “szumieć”, “drzwi”, “ważyć”, “Wałęsa”, “chorąży”, “chrząszcz”, “vařečka”. Now check their actual pronunciation at Wiktionary.

⁽¹⁾ With ć, ś, ź, and dź being different from ci, si, zi, and dzi.

[helius]

A problem with any phonetic alphabet is that adjacent phonemes can be blended to varying degrees. In "The Language Instinct", Steven Pinker gives some examples: The initialism "NPR" when spoken rapidly by most native speakers sounds more like "MPR" because the plosive P blends into the preceding N to make a M sound. But when speaking more slowly, the N sound stays more intact. It's hard to represent this varying blend in terms of a fixed phonetic transcription.

[Aldo22]

Saying it is useless because it cannot do everything an audio recording can do is a bit of a strange statement, is it not?

"It cannot do everything, therefore it is useless and not worth having at all." 

Well, if I have a number notation system that tells me that five is seven when a Scotsman interprets it and two when an Italian reads it, then I'd say it's useless.

[tooki]

As a guide, it can help you understand approximately how to make the sounds, but it can never tell you how to speak words in a native accent.

Well, why not?
What is lacking in the descriptive scope of IPA phonemes that prevent that from being possible?
Is it levels of subtlety that go far deeper than possible with describing individual sounds that humans can make with their mouths?
How precise would one need to be in order to make it possible to do what you said (speak words in a native accent)?

You can encode that with IPA, using additional diacritics, but it’s only done when contrasting the specific expressions in different accents. Normally, over-detailed IPA is avoided a) because it’s then not general enough to apply broadly, b) will vary significantly both between speakers but also on context, intonation, and even some random variation (even within one single speaker), and c) it’s very difficult and time consuming to do.

[Zero999]

I've changed my opinion on one thing, thanks to this thread which has sent me down the rabbit hole of watching historical phonology videos, I've realised we have a fairly good idea of what people's accents long before the era of sound recording.


I originally wanted to find out whether it was sodder first and the Brits started pronouncing the L or the other way round. Not only have I not found the answer, there doesn't appear to be strong evidence either way, but its raised more questions.

Why have certain sound changes only propagated over a certain geographical distance, then run out of steam? Take the strut-foot split for example. In early modern English, the U vowel in the words foot and strut sounded the same, so the two words rhymed, yet in the mid-17 century, Us split into two different sounds. Interestingly this did not affect Northern England. There's an isogloss running through the Midlands. which divides where the words wood and blood rhyme and don't rhyme. I'm just south of this isogloss, put doesn't rhyme with strut for me, yet I recently asked someone from the next town to the north and she said they do. Contrast this with non-rhoticity, the lack of the R sounds, which started a little later yet spread further north, but not so much in the south and west.

[golden_labels]

Trying to get a native English speaker to hear or make the distinction between "proszę" (please) and "prosię" (little pig) is quite challenging indeed    English "sh" falls somewhere in between these two sounds.

This points to another important use of phonetic alphabets. They tell the reader there is a difference they themselves would otherwise never notice.

To add to the confusion, verb “prosić” is both the infinitive of “proszę” and verbification of “prosię”. The latter means “to farrow,” nowadays usually used in perfective aspect, reflexive “oprosić się.” Imperative of “prosić” (in the begging sense) is “proś,” which is different from “prosi” (3rd person singular of “prosić” in both senses). I challenge English speakers to notice the distinction without IPA.

It’s usually easy to spot Polish people speaking English, because English and Polish ‘r’ use generally different sounds. ɹ in English, r in Polish. Most Polish speakers pronounce “solder” with a trill at the end. One can hear the difference in word “folder” found in both languages: English vs Polish.

I've changed my opinion on one thing, thanks to this thread which has sent me down the rabbit hole of watching historical phonology videos, I've realised we have a fairly good idea of what people's accents long before the era of sound recording. (…)

A great video.

We also have evidence from where and when people were not consciously describing a language. That’s because words migrate between languages. A word may become fossilized, with their original features, and preserved until modern times. That word may have undergone changes together with the entire language carrying it, but we know what these changes were. That word may be written down, approximating its actual sound (rather than the original spelling) at the time it was used. We do have wordplays, which make sense only if we accept a particular pronunciation of the words involved. We have poetry. Finally, people make mistakes in spelling and with lower literacy levels they become common and repetitive. All this in quantities sufficient to matter in statistical sense.

[paulca]

It is quite common to have a map which says something, which appears "native language".  When you translate it you find it just means "Hill".  Or "Lake" or "Woodland".

So the map place name translates to "The hill hill".

The locals when this "mapper" arrived and began asking what places where called may not have been a good native speaker, so when he pointed at the hill and tried to ask, "What do you call that?", the native person looked confused and said, "A hill?"  And so that what the map calls it.

Examples, made up:  Mount Sleive.  Carn Hill.  Riverburns.

[Analog Kid]

Kind of like "Mount Fujiyama", right?

[Uup]

Kind of like "Mount Fujiyama", right?

Yes. However, in Japan Mt Fuji is referred to as Fujisan. Fujiyama is more of a western inference.

[TimFox]

The two suffixes -yama and -san use the same kanji:  you have to know which one to use (san is correct for Fuji).

[golden_labels]

And Nile, which simply means “river.” Both its current name and how Ancient Egyptians would call it.

Sometimes it goes the other direction. Australia being a good example: from terra australis incognita.

[Uup]

The two suffixes -yama and -san use the same kanji:  you have to know which one to use (san is correct for Fuji).

Just to add a little context; Fujiyama, as a loanword into English, is correct and the de-facto standard outside of Japan.

Also, whilst not sounding as bad or depraved to my ears as hearing "sodder" I sometimes have trouble comprehending the English accent, used in parts of the UK, where certain consonants are not pronounced unless they are at the beginning of a word. Using the example a few posts above, the word bitter, with 't' sound silent, would be pronounced 'bi-er'.

[IanB]

Also, whilst not sounding as bad or depraved to my ears as hearing "sodder" I sometimes have trouble comprehending the English accent, used in parts of the UK, where certain consonants are not pronounced unless they are at the beginning of a word. Using the example a few posts above, the word bitter, with 't' sound silent, would be pronounced 'bi-er'.

You mean like "The water in Majorca don't taste like what it ought to?"

https://www.eevblog.com/forum/chat/its-sodder-not-solder-damnit!/msg6381111/#msg6381111