Questions about Welsh

[Jonlang]

The vowel sound between /g/ and /r/ doesn't really seem to matter but moving from /g/ to /r/ without a vowel seems pretty impossible.

It's perfectly possible. Syllabic rhotics are a thing--just ask the Czechs.

I don't doubt it, but not with a Welsh accent. Welshies stretch out vowels pretty much everywhere and you'll hear one with syllabic consonants too. I mean, I have a Welsh accent.

[Yng]

Going off Wiki, it's basically something like ['ɬɔɨ̯.gɨr~'ɬɔɨ̯.gr̩].

The former seems about right. I don't think it's an [E] in the second syllable, it's either 1 or @ or I (probably depending on accent).

Obviously it would be possible to pronounce it as one syllable, and I think it may well have been pronounced as one syllable once. But we epenthesised all these clusters, I think (maybe some dialects more than others) or reduced them.

Thanks, I've got some unlearning to do. For example, I'd learnt llyfr 'book' as [ɬɨvr] (one syllable), whereas it's something like [ɬəvr̩] (two syllables). Another problem with the notion that Welsh spelling specifies the pronunciation.

Yeah, it's [ɬəvr̩] or [ɬəvər̩] or even [ɬəvɨr~ɬəvɪr] I think depending on the speaker - I lean towards the latter pronunciation. Sometimes you will even encounter e.g. llyfr llythr as llyfyr llythyr in eye dialect.

To be fair, if you know syllable restrictions then you know that these clusters cannot appear and you epenthesise them automatically. But there are also some we delete (colofn /kolo:n/, ffenestr /fEnEst/ etc).

[linguoboy]

I don't doubt it, but not with a Welsh accent. Welshies stretch out vowels pretty much everywhere and you'll hear one with syllabic consonants too. I mean, I have a Welsh accent.

Making generalisations about a language based on the substratal effects of an earlier related variety on another language is a hazardous game. As Yng says, the treatment of these final syllabic rhotics has changed over time in Welsh. So, for that matter, has the Welsh-accented English you speak. (Whoever was the last generation of fluent Welsh-speakers in your family, you don't speak English exactly like them.)

[Pole, the]

The vowel sound between /g/ and /r/ doesn't really seem to matter but moving from /g/ to /r/ without a vowel seems pretty impossible.

It's perfectly possible. Syllabic rhotics are a thing--just ask the Czechs.

And stop+liquid codas without syllabic rhotics are also a thing — just ask the Poles.

(For instance, the Polish word for an “ogre” is „ogr”.)

[linguoboy]

The vowel sound between /g/ and /r/ doesn't really seem to matter but moving from /g/ to /r/ without a vowel seems pretty impossible.

It's perfectly possible. Syllabic rhotics are a thing--just ask the Czechs.

And stop+liquid codas without syllabic rhotics are also a thing

Did someone claim otherwise?

[Richard W]

The vowel sound between /g/ and /r/ doesn't really seem to matter but moving from /g/ to /r/ without a vowel seems pretty impossible.

It's perfectly possible. Syllabic rhotics are a thing--just ask the Czechs.

And stop+liquid codas without syllabic rhotics are also a thing

Did someone claim otherwise?

That's how I read Dyolf's statement. Especially if syllabic resonants are counted as vowels, as in Sanskrit.

[Richard W]

To be fair, if you know syllable restrictions then you know that these clusters cannot appear and you epenthesise them automatically. But there are also some we delete (colofn /kolo:n/, ffenestr /fEnEst/ etc).

That's a big 'if'. There's also the issue of rule ordering, plus the fact that synchronic rule ordering (where real) isn't necessarily the same as diachronic rule ordering.

[Richard W]

Thanks, I've got some unlearning to do. For example, I'd learnt llyfr 'book' as [ɬɨvr] (one syllable), whereas it's something like [ɬəvr̩] (two syllables). Another problem with the notion that Welsh spelling specifies the pronunciation.

Yeah, it's [ɬəvr̩] or [ɬəvər̩] or even [ɬəvɨr~ɬəvɪr] I think depending on the speaker - I lean towards the latter pronunciation. Sometimes you will even encounter e.g. llyfr llythr as llyfyr llythyr in eye dialect.

I think there's an issue of synchronic as well as diachronic variation. (That's no big surprise.) I've now found [ɬɨvr] in several places, and I found this quote in the one hundred year-old A Welsh Grammar, Historical and Comparative of John Morris Jones:

But in N. W. dialects the parasitic vowel did not arise in groups containing f; thus in the greater part of N. W. ofn, ‘fear’, cefn ‘back’, llyfr ‘book’, barf ‘beard’ are purely monosyllabic to this day.

I can also find something close on Youtube! [ɬɨvr̩] (2 syllables to my ear!) occurs at least 5 times in the first 43s of 4 Llyfr: Eiconau Cymraeg Ynghyd, and you don't need to know Welsh to pick the word out several times.

[din]

I don't speak/understand any Welsh, but it sounds like [ˈɬɪvr̩] to me.

And does Welsh generally have a high density of /r/s, or is it just me (or that particular sample)?

[linguoboy]

The Welsh dialect survey (Cardiff, 2000) is actually a book I own and llyfr is one of the words it covers, so I had a gander. According to this, the distribution is:

Monosyllabic /ɬᵻvr/ in the North-West and North-East.
Bisyllabic /ɬᵻvr̩/ in the Midlands.
Bisyllabic /ɬəvɪr̩/ (mostly in Ceredigion) or /ɬɪvɪr̩/ in the South.

Syllabic /r/ is more widespread in gafr, but not at all attested for llwfr (though, to be fair, responses are absent for most of the Midlands).

[Jonlang]

The Welsh dialect survey (Cardiff, 2000) is actually a book I own and llyfr is one of the words it covers, so I had a gander. According to this, the distribution is:

Monosyllabic /ɬᵻvr/ in the North-West and North-East.
Bisyllabic /ɬᵻvr̩/ in the Midlands.
Bisyllabic /ɬəvɪr̩/ (mostly in Ceredigion) or /ɬɪvɪr̩/ in the South.

Syllabic /r/ is more widespread in gafr, but not at all attested for llwfr (though, to be fair, responses are absent for most of the Midlands).

Which regions does "Midlands" cover? It's not a term used to refer to anywhere in Wales, but the Midlands is an area of central England which is obviously not the area the book is talking about. The term "Mid Wales" is used to refer to Powys and Ceredigion and sometimes southern Gwynedd.

[Richard W]

Bisyllabic /ɬᵻvr̩/ in the Midlands.
Bisyllabic /ɬəvɪr̩/ (mostly in Ceredigion) or /ɬɪvɪr̩/ in the South.

Doing some more browsing, I found that some dialects (West Pembrokeshire, for example) have largely eliminated the alternation of obscure <y> with clear <y> (and with <w>) (Iosad p152). How much of the anomalous occurrence of clear 'y' in bisyllabic llyfr is accounted for by that, rather than rule ordering?

[linguoboy]

The Welsh dialect survey (Cardiff, 2000) is actually a book I own and llyfr is one of the words it covers, so I had a gander. According to this, the distribution is:

Monosyllabic /ɬᵻvr/ in the North-West and North-East.
Bisyllabic /ɬᵻvr̩/ in the Midlands.
Bisyllabic /ɬəvɪr̩/ (mostly in Ceredigion) or /ɬɪvɪr̩/ in the South.

Syllabic /r/ is more widespread in gafr, but not at all attested for llwfr (though, to be fair, responses are absent for most of the Midlands).

Which regions does "Midlands" cover? It's not a term used to refer to anywhere in Wales, but the Midlands is an area of central England which is obviously not the area the book is talking about. The term "Mid Wales" is used to refer to Powys and Ceredigion and sometimes southern Gwynedd.

[Yng]

The Welsh dialect survey (Cardiff, 2000) is actually a book I own and llyfr is one of the words it covers, so I had a gander. According to this, the distribution is:

Monosyllabic /ɬᵻvr/ in the North-West and North-East.
Bisyllabic /ɬᵻvr̩/ in the Midlands.
Bisyllabic /ɬəvɪr̩/ (mostly in Ceredigion) or /ɬɪvɪr̩/ in the South.

Sounds reasonable. As I said, I'm not very good at phonetics.

Doing some more browsing, I found that some dialects (West Pembrokeshire, for example) have largely eliminated the alternation of obscure <y> with clear <y> (and with <w>) (Iosad p152). How much of the anomalous occurrence of clear 'y' in bisyllabic llyfr is accounted for by that, rather than rule ordering?

Only a bit - that feature is only present in the South as far as I'm aware.

I was thinking about this and I don't think that the epenthetic vowel affects stress either. I had an example but now I can't think of any examples where we don't just simplify the cluster.

[Nortaneous]

What did the Proto-Brythonic copula look like? How did it develop into the Welsh copula? I can't find anything on this.

[Jonlang]

Can anyone explain the difference between the prefixes 'cyd-' and 'cyf-'? They appear to mean the same thing - the coming together of two things, but I think they're used differently somehow.

[linguoboy]

Can anyone explain the difference between the prefixes 'cyd-' and 'cyf-'? They appear to mean the same thing - the coming together of two things, but I think they're used differently somehow.

My initial impression is that cyf- is longer-standing and has a broader range of meaning, but I'd have to look into it more.

[Znex]

How does Welsh intonation work across different sentences? Is it much similar to English intonation or not? Where are the peaks and valleys in pitch?

eg.
Plain statement: Mae'r ci'n rhedeg o'i cwmpas y gath.
Plain statement with emphasis: Y ci sy'n rhedeg o'i cwmpas y gath./Y gath mae'r ci'n rhedeg o gwmpas.
Polar question: Ydy'r ci'n rhedeg o'i cwmpas y gath?
Wh question: Beth sy'n rhedeg o'i cwmpas y gath?/Beth mae'r ci'n rhedeg o gwmpas?/Beth sy'n rhedeg o gwmpas?
Extra wh question: Pam mae'r ci'n rhedeg o'i cwmpas y gath?/Lle mae'r ci'n rhedeg?/Pwy ci sy'n rhedeg o'i cwmpas y gath?

[Yng]

I have no idea how to transcribe the intonation, but you have a few errors.

Plain statement: Mae'r ci'n rhedeg o gwmpas y gath.

Plain statement with emphasis: Y ci sy'n rhedeg o gwmpas y gath./Y gath mae'r ci'n rhedeg oi chwmpas.

This third one sounds a bit weird to me, but we stress y ci and y gath, as in 'it's the dog' in English.

Polar question: Ydy'r ci'n rhedeg o gwmpas y gath?

no particular emphasis

Wh question: Beth sy'n rhedeg o gwmpas y gath?/Beth mae'r ci'n rhedeg o'i gwmpas?/Beth sy'n rhedeg o gwmpas?
Extra wh question: Pam mae'r ci'n rhedeg o gwmpas y gath?/Lle mae'r ci'n rhedeg?/Pwy ci sy'n rhedeg o'i cwmpas y gath?

What's this last one supposed to be? Ci pwy sy'n rhedeg o gwmpas y gath?

the preposition is o gwmpas (from 'compass'). o'i gwmpas/chwmpas means 'around it'.

[Znex]

What's this last one supposed to be? Ci pwy sy'n rhedeg o gwmpas y gath?

Yeah, I think that's what I meant. Also I saw the o'i gwmpas used somewhere for some reason and followed that.

I have no idea how to transcribe the intonation

IPA uses ↗ and ↘ to indicate intonation (a climb in pitch and a fall in pitch respectively), so say for the English sentence:
The dog runs around the cat.
For my idiolect, I'd put arrows in these places:
The ↘↗dog runs around the ↘cat.

Whereas where there's emphasis on dog, as in It's the dog that..., I'd have:
The ↘dog runs around the cat. (where there's no great change in pitch after dog)
Or cat:
The dog runs around the ↘cat. (where there's no great change in pitch before cat)

[Jonlang]

I need some help identifying the historical causes of mutation of adjs after feminine nouns. I asked this on a Facebook group, directly appealing to a Welsh linguist but he's unable to answer, but he did point me in the direction of a £110 book which I can't really afford, so hopefully someone here will have some idea. Below I've pasted my question to him, simply removing the greeting to omit names and added some formatting:

Welsh mutations are, basically, just a natural set of shifts along the lenis-fortis scale. They began as phonological changes and they're predictable, being a mixture of lenition, fortition and eclipsis. Somewhere along the line these changes moved from being purely phonological to morphological - feminine modifiers/quantifiers (I think) used to end in vowels and masculines did not, hence why only feminine nouns mutate after words like un (*oinā (f) *oinos (m) both older forms of the modern un. Either Common Brythonic or Proto-Celtic, I can't remember.) Because *oinā ended in a vowel it caused lenition in the following feminine noun, but the masculine *oinos ends in a consonant, blocking lenition of the masculine noun.

My question is this:
Why do feminine nouns cause lenition to adjectives? Was this once a phonological change (like all fem nouns ending in vowels)? Or did this happen later as a morphological change (like speakers just being used to leniting feminine words by this point)?

This is to help me with working out (and justifying) mutations in my conlang family, but it doesn't have gender, so I need a way to justify adjs mutating after nouns.

EDIT: I've been informed that the feminine nouns tended to end in -a, which caused the lenition. When the final -a was lost due to apocape in Brythonic people carried on leniting feminine words by analogy, hence why all susceptible adjs mutate after feminine nouns. Happy days! So all I need to do for my conlang is have most nouns end in vowels and make sure that the lenitions take place before apocape, so that when the vowels disappear all nouns will cause lenition to adjs by analogy, I think...

[Yng]

Yes. IIRC there was originally something along the lines of -o -a on most masc and feminine nouns but the former was lost before the latter, leaving only feminine nouns (at least in general) with vocalic terminations.

[hwhatting]

Yes. IIRC there was originally something along the lines of -o -a on most masc and feminine nouns but the former was lost before the latter, leaving only feminine nouns (at least in general) with vocalic terminations.

If we're basing this on attested Britannic names, male nouns ended in -Vs (mostly -os) and female nouns mostly in -a, so the reason for the absence of the mutation after male nouns is their final -s.

[Jonlang]

Yeah, just like my example of *oinā (f) and *oinos (m) which both later became un! Un gath (< cath) but un ci (not *un gi).

[Jonlang]

Can anyone help me in finding out how Welsh developed its collective/singulative system? Someone has suggested to me that it could be a Brythonic innovation as it doesn't occur in Goidelic languages (their words, not mine). Were words like coed and plant originally singulars at some point and somehow there was a shift in meaning? Possibly because people more often spoke of many trees and many children rather than just one? Or am I barking up the wrong tree here?