de-rhotacism of r

Platytude

Senior Member
English - AU
Hi,

In some languages, there was a rhotacism of *s to some form of r. Has the opposite also occurred (r > alveolar or dental fricative) somewhere? Does Czech ř count as a (to-be) example of this?

P.S. The term is "de-rhotacism" in speech pathology; is it in historical linguistics too?
 
  • Swatters

    Senior Member
    French - Belgium, some Wallo-Picard
    There's a handful of words with /r/ > /z/ in French, but the change was reversed before it spread to every word (this was a century or so prior to uvularisation) : cathedram > chaire > chaise (chair, compare to... the English loanword chair) and beryl + -culos > bericles/béricles > besicles/bésicles (spectacles, compare to the Walloon cognate /bɛrik/)

    An equivalent term to rhotacism with an endpoint of /s ~ z/ would be "sigmatism", but that's mostly used by speech pathologists to refer to a specific lisp
     

    Cenzontle

    Senior Member
    English, U.S.
    D. Lincoln Canfield's Spanish Pronunciation in the Americas (p. 13) gives a list of eight specific areas in Central and South America where the /r/ phoneme (double r in spelling — an apical trill in most dialects) is realized as an alveolar "grooved fricative". Generally called "assibilation", rather than "de-rhotacism".
     

    LeBro

    Senior Member
    Turkish
    In some languages, there was a rhotacism of *s to some form of r. Has the opposite also occurred (r > alveolar or dental fricative) somewhere?

    I wonder if the (alleged) */-r/ > */-z/ change (zetacism) (occured) in (Proto-)Turkic fits the question :confused:
     

    clamor

    Senior Member
    French - France
    Hi,

    In some languages, there was a rhotacism of *s to some form of r. Has the opposite also occurred (r > alveolar or dental fricative) somewhere? Does Czech ř count as a (to-be) example of this?

    P.S. The term is "de-rhotacism" in speech pathology; is it in historical linguistics too?

    In some Armenian dialects, the Classical /ɾ/ has become [ʒ] in a various set of contexts. Note that in the Standard modern varieties it is pronounced as [ɹ~ɾ] with contextual devoicing and allophonic palatalization, so the change to a sibilant fricative is not surprising.
     

    Sobakus

    Senior Member
    In most Sardinian varieties, /s/ and /r/ are completely neutralised word finally: /r/ is pronounced [s] and /s/ pronounced [r] depending on the following sound. So tres baccas [trɛrˈbakːaz̺a] “three cows”, battor panes [batːɔs̺ˈpanɛz̺ɛ] “four loaves of bread”.
     

    Dib

    Senior Member
    Bengali (India)
    In most Sardinian varieties, /s/ and /r/ are completely neutralised word finally: /r/ is pronounced [s] and /s/ pronounced [r] depending on the following sound. So tres baccas [trɛrˈbakːaz̺a] “three cows”, battor panes [batːɔs̺ˈpanɛz̺ɛ] “four loaves of bread”.

    Sanksrit had a strikingly similar development. Word-final /s/ and /r/ lost contrast in most phonetic environments (only regular exception: [a_# W] W = vowel or voiced consonant), and their actual realisation varied widely based on the environment: including both [s] and [r] - as well as other sounds.
     

    Sobakus

    Senior Member
    Off-topic but why the support vowel after the s? Is it an internal development or Italian influence?
    I don't think it's Italian influence – rather, all Romance varieties south of la Spezia-Rimini line that preserve final obstruents (such as -s/-t) exhibit a support vowel after it. This is therefore a repair mechanism in the face of a prohibition on obstruents ending a syllable. The other repair strategy was their assimilation with further morphosyntactic reanalysis, as in most of the Italian area, where it's become limited to stressed syllables and to certain monosyllables (raddoppiamento fonosintattico). This makes the prohibition look very old, and the repair strategies inherited, or at least somehow independently pre-determined as soon as the prohibition arises.

    Since Sardinian exhibits the most original stage in the development of the raddoppiamento (which is integrated into its system of lenition and effectively blocks it), it's reasonable to think that the support vowel situation is also original, and hence in the parent variety of Late Latin assimilation and vowel insertion were alternative or complementary (just as in Sardinian) repair mechanisms in a system of sandhi similar to that of Sanskrit. There (still hopefully) are varieties around Cosenza and southern Lucania that preserve final obstruents, though they seem to use schwa or /i/ as a support vowel instead of copying it as Sardinian does. The now-intervocalic final /t/ is voiced in some of them, just like in almost all of Sardinian.

    I'm not sure how original this is, but many modern (chiefly southern it seems) Indian pronunciations of Sanskrit insert a phrase-final copy vowel even after the coda /s/ undergoes sandhi to become [ h ] (known as visarga, spelled : and transliterated ḥ), which then voices to [ɦ]. This is in effect a combination of these two repair strategies, but may be due to a newly-arisen ban on coda [ h ] which is itself the repair outcome of a ban on coda [ s ].
     
    Last edited:

    Uncreative Name

    Member
    English - United States
    In Icelandic, R is often pronounced something like [z] or [ʒ]. (At least, that's what it sounds like to my ears.) I don't know if this was a de-rhotacization sound change or not, though.
     

    Sobakus

    Senior Member
    In Icelandic, R is often pronounced something like [z] or [ʒ]. (At least, that's what it sounds like to my ears.) I don't know if this was a de-rhotacization sound change or not, though.
    That's probably what the Wikipedia article on Old Norse transcribes as [ɹ̠˔]. This sound developed out of the retracted [z̠] and had its own letter in the runic alphabet, , today romanised as ʀ. In modern Icelandic it's merged with /r/, and the difference is now allophonic. The fricative-like sound normally occurs in the endings /ʏr/, /ɪr/, and is de-voiced unless a vowel follows; it doesn't occur at the start of the word, intervocalically inside the word, or when doubled – listen to the difference, also contrast. So this isn't a new sound change (de-rhotacism), rather it's selective preservation of the original post-rhotacism stage.
     
    Last edited:
    In Tsakonian Greek, the initial /r/ phoneme of Standard Modern Greek is often realised as a /sx/ and before a front vowel a /sç/ eg:
    Classical Greek «ῥίζα» [ˈr̥id͡za] (fem.) --> root > Standard MoGr «ρίζα» [ˈɾiz̠a] (fem.) = Tsakonian «σχίντα» [ˈs̠çiŋd̠a] (fem.) via an intervening «*ῥίντα» [ˈr̥iŋt̠a].

    Αpologies for my late edit, but I just wanted to add that also in Tsakonian the cluster /tr/ (eg: MoGr «τρία»[ˈt̠ria] --> three (cardinal nr)) is realised as the derhotacised /tʃ/ (eg: Tsak.«τσ̌ία» [ˈt̠͡ʃia] --> three (cardinal nr)), while at the same time, the opposite phenomenon is observed with the replacement of final /s/ in MoGr nominals by /r/ (rhotacism or extreme rhotacism), eg: MoGr «τρεις ημέρες» [ˌt̠ri.iˈme̞ɾe̞s̠] = Tsakonian «τσ̌ιρ αμέρε» [ˌt̠͡ʃiɾ.aˈme̞ɾe̞] --> three days.
     
    Last edited:

    sumelic

    Senior Member
    English - California
    In Tsakonian Greek, the initial /r/ phoneme of Standard Modern Greek is often realised as a /sx/ and before a front vowel a /sç/ eg:
    Classical Greek «ῥίζα» [ˈr̥id͡za] (fem.) --> root > Standard MoGr «ρίζα» [ˈɾiz̠a] (fem.) = Tsakonian «σχίντα» [ˈs̠çiŋd̠a] (fem.) via an intervening «*ῥίντα» [ˈr̥iŋt̠a].
    Does the spelling σχ in Tsakonian really represent a cluster /sx/~/sç/, or is it a unitary fricative [ʃ]?
     
    Does the spelling σχ in Tsakonian really represent a cluster /sx/~/sç/, or is it a unitary fricative [ʃ]?
    I think you're probably right, I'm familiar with Michael Deffner's Tsakonian script, which uses specific subscript & upperscript markings to show palatalisation (Zakonische Grammatik, 1881, p177), I'm unfamiliar with the later scripts devised by Hubert Octave Pernot (1934) and Athanasios Kostakis (1951)
     

    berndf

    Moderator
    German (Germany)
    I think the same alternation of r/s occurs in some cases within the English language (cf. I was/you were) and the German language (cf. ich war vs. ich bin gewesen, and das Wesen (=the being)).
    See #16. The /z/ rhotisation described tgere affected all North and West Germanic languages. German ich war is a modern German regularisation to align ich war/wir waren. In early modern High German is was still ich was/wir waren.
     

    Red Arrow

    Senior Member
    Nederlands (België)
    In Icelandic, R is often pronounced something like [z] or [ʒ]. (At least, that's what it sounds like to my ears.) I don't know if this was a de-rhotacization sound change or not, though.
    This also happens in Central Swedish: r can be pronounced [ʐ].

    (Dutch has some s > r: was/waren, vriezen/bevroren, verkozen/uitverkorene, but no r > s as far as I know)
     
    Top