From Wikipedia, the free encyclopedia - View original article

Jump to: navigation, search
Not to be confused with R-colored vowel.
Sound change and alternation

Rhotacism (/ˈrtəˌsɪzəm/)[1] may refer to an excessive or idiosyncratic use of the letter r, the inability to pronounce (or difficulty in pronouncing) r, or the conversion of another consonant into r.

The term comes from the Greek letter rho, denoting "r".


In medicine, rhotacism is the inability or difficulty in pronouncing the sound r. Thus, many speech pathologists call this problem de-rhotacization, because the sounds lose their rhotic quality rather than becoming rhotic.

The rhotic sounds are usually the last ones a child masters. Some people never learn to produce them; they substitute other sounds, such as the velar approximant, the uvular approximant, and the uvular trill (often called "French R"). In English, the most common occurrence of this type is a pronunciation perceived as closer to [w].

The Looney Tunes character Elmer Fudd is famous for his exaggerated rhotacistic speech, as in, "Be vewy, vewy quiet; I'm hunting wabbits." The cartoon characters Homestar Runner and Tweety talk in much the same way, and the comedian Gilda Radner spoofed Barbara Walters' speech problems in her recurring Saturday Night Live character "Baba Wawa". Another example is the depiction of Pontius Pilate in Monty Python's Life of Brian. In the Only Fools and Horses episode Stage Fright, a singer's rhotacism is a central plot device. In popular culture, examples include Rick from The Young Ones, Barry Kripke (from The Big Bang Theory), Roy Hodgson, Edward Ka-Spel, Jonathan Ross, Mark Owen, Brian Walden, David Zayas, Frank Muir, Sister Wendy Beckett, Lucy Worsley, Nick Heidfeld, the priest in The Princess Bride (portrayed by Peter Cook) and the UK politician Roy Jenkins. Other examples are interviewer Barbara Walters, actresses Kay Francis and Marlene Dietrich and King George VI.

Rhotacism is more common among speakers of languages that have a trilled R,[citation needed] such as Swedish (except in the provinces of Skåne, Halland, Blekinge, Öland and southern Småland), Finnish, Italian, Polish, and Spanish. Many Quebec rural regional accents along the St. Lawrence corridor are characterized by rhotacism.


In linguistics, rhotacism or rhotacization is the conversion of a consonant (usually a voiced alveolar consonant/z/, /d/, /l/, or /n/) to a rhotic consonant in a certain environment. The most common may be of /z/ to /r/.[2]


The southern Tosk dialects, on which modern standard Albanian is based, changed /n/ to /r/ while the northern Gheg dialects did not.[2] Compare:


In Aramaic, proto-Semitic n is often changed to r:


Ancient Basque *l has changed into a tapped R between vowels in Basque.[3] This can be observed in words borrowed from Latin, for example.

Irish and Scots Gaelic[edit]

In Irish and Scottish Gaelic a prevocal /kn/ cluster developed into /kr/ often with nasalization of the following vowel as in cnoc [krɔ̃xk] ('hill').[2]

Germanic languages[edit]

All surviving Germanic languages, members of the North and West Germanic families, underwent a change of /z/ to /r/, implying a more approximant-like rhotic consonant in early Germanic.[4] Some languages have regularized, giving all forms an r. Gothic retains s or z, since it did not undergo rhotacism.

Proto-GermanicGothicOld NorseOld English
Modern English
Old Frisian[5]Dutch(Old High German)
Modern German
*was,1st/3rd sg *wēzum1st plwas, wēsum
var, várum
(wæs, wǣron)
was, were
was, wēren 
was, waren
(was, wārum)
war, waren
*fraleusaną,inf *fraluzanazp.part.fraliusan, fralusans

(forlēosan, forloren)
forlese, forlorn
urliāsa, urlāren 
verliezen, verloren
(farliosan, farloren)
verlieren, verloren


Pronouncing the letter "r" is common in many dialects of American, Canadian, Irish, Welsh and Scottish English and less common in the English of most of England, Australia, and New Zealand. Lenition of intervocalic /t/ and /d/ to [d] or [ɾ] is also common in many modern English dialects (e.g. <got a lot of> (phonemically /gɒtə lɒtə/) becoming [gɒdə lɒdə] or [gɒɾə lɒɾə]). Contrast is maintained with /ɹ/ because it is never realized as a flap in these dialects of English.[2]


In Central German dialects, esp. Rhine-Franconian and Hessian, /d/ is frequently realized as [ɾ] in intervocalic position. This change also occurs in Mecklenburg dialects.


In Korean, the consonant ㄴ, typically /n/, can be realized as /l~ɾ/, as in the surname Roh, and will assimilate into a following or preceding ㄹ /l~ɾ/.

Romance languages[edit]


This reflects a highly regular change in pre-classical Latin. Intervocalic s in the oldest attested Latin documents (assumed to have been pronounced /z/) invariably became r. Intervocalic s in Classical Latin suggests either borrowing (e.g. rōsa) or reduction of an earlier ss (e.g. pausa < paussa, vīsum < *vīssum < *weid-tom). Old s was preserved initially (septum), finally, and in consonant clusters.

The English word honour or honor is derived from Anglo-Norman honour, which in turn was derived from Late Latin honor, earlier honos, which became honor by analogy with the oblique stem of honoris (genitive).

The consonants d or l changed to r before another d or l, so that the same consonant would not appear twice in a row (dissimilation).

This phenomenon was noted by the Romans themselves:

In many words in which the ancients said s, they later said r... foedesum foederum, plusima plurima, meliosem meliorem, asenam arenam

Varr. De lingua Latina, VII, 26, In multis verbis, in quo antiqui dicebant s, postea dicunt r... foedesum foederum, plusima plurima, meliosem meliorem, asenam arenam


In Neapolitan rhotacism is seen in a shift from the sound of "d" to an "r" sound:

(Italian vs Neapolitan)

and, to a lesser extent, from the sound of an "l" to an "r" sound:


In Old Portuguese, rhotacism occurred from the «l» sound to the «r» sound, mainly in consonant plus el clusters, as in the words obrigado, "thank you", originarily from "obliged [in honorably serving my Sir]", praia, "beach", prato, "plate" or "dish", branco, "white", prazer, "pleasure", and praça, "square".

In contemporary Brazilian Portuguese, rhotacism of /l/ in the syllable coda is characteristic of the caipira dialect, while further rhotacism in the nationwide Vernacular include planta, "plant", as [ˈpɾɐ̃tɐ], lava, "lava", as /ˈlarvɐ/ (thus homophonous with larva, worm/maggot), lagarto, "lizard", as [laʁˈɡaʁtu] (in dialects with guttural coda r instead of a tap) and advogado, "lawyer", as [ɐ̞de̞vo̞ʁˈɡadu]. These non-standard patterns are largely marginalized, as rhotacism is regarded as either sign of speech-language pathology or part of the characteristics of illiterates' speech.


Rhotacism in Romanesco consists of a shift from "l" to "r" when it is followed by a consonant, similar to what occurs in certain Andalusian dialects of Spanish. Thus, Latin altus (tall) which in Italian is alto in Romanesco becomes arto. In ancient Romanesco it also happened when "l" was preceded by a consonant, as in the word ingrese (English), but the modern way of speaking has lost this characteristic.

Another change related to r was the shortening of the geminated rr. This is not rhotacism. So the words errore, guerra and marrone (error, war, brown) in Romanesco become erore, guera and marone.


Romanian rhotacism consists of a shift from intervocalic "l" to "r" and "n" to "r".

Thus, Latin caelum (meaning 'heaven' or 'sky') became Romanian cer, Latin fenestra (meaning 'window') becomes Romanian fereastră, and Latin felicitas (meaning 'happiness') Romanian fericire.

Some northern Romanian dialects and Istro-Romanian also further transformed all intervocalic [n] into [ɾ]. This occurred only with words of Latin origin.[7] For example, Latin bonus became Istro-Romanian bur, as compared to standard Daco-Romanian bun.


This phenomenon appears in Andalusian Spanish (particularly in Seville, where "l"s at the end of a syllable preceding another consonant are replaced with "r"s, e.g. saying "Huerva" instead of "Huelva". The reverse is done in Caribbean dialects, even at the beginning of words, e.g. saying "Puelto Lico" instead of "Puerto Rico".


In Sanskrit, words ending in -s other than -as become -r in sandhi with a voiced consonant:

This is not a case of rhotacism proper, since r and s are simply allophones in those positions.

South Slavic languages[edit]

(This section relies on the treatment in Greenberg 1999[8])

In some South Slavic languages, rhotacism occasionally changes a voiced palatal fricative [ʒ] to a dental/alveolar tap or trill [r] when it occurs between vowels. For example:

The beginning of the change is attested in the Freising manuscripts, a written document from the 10th century AD, which shows both the archaism (ise 'which' < *jь-že) and the innovation (tere 'also' < *te-že). It is also found in individual lexical items in Bulgarian dialects, e.g., дорде 'until' (< *do-že-dĕ), and Macedonian, e.g. сеѓере (arch. 'always'). However, the results of the sound change have largely been reversed by lexical replacement in dialects in Serbia and Bosnia beginning in the fourteenth century. Dialects in Croatia and Slovenia have not only preserved more of the lexical items with the change, but have extended grammatical markers in -r- from heterogeneous sources that formally merged with the rhotic forms that arose due to the sound change, e.g., Slovene dialect nocor 'tonight' (< *not'ь-sь-ǫ- + -r-) on the model of večer 'evening' (< *večerъ). The reversal of the change is evident in dialects in Serbia where the -r- formant is systematically removed, e.g., Serbian veče 'evening'.


  1. ^ "American English Dictionary: Definition of rhotacism". Collins. Retrieved December 13, 2013. 
  2. ^ a b c d Catford (2001:178)
  3. ^ Trask, R. Larry (2008), Wheeler, ed., A Historical Dictionary of Basque, University of Essex, p. 29, retrieved January 22, 2011  |first2= missing |last2= in Editors list (help)
  4. ^ Catford (2001:179)
  5. ^ D. Hofmann, A.T. Popkema, Altfriesisches Handwörterbuch (Heidelberg 2008).
  6. ^ robus1; rōbur. Charlton T. Lewis and Charles Short. A Latin Dictionary on Perseus Project.
  7. ^ Nandris (1963:255–258)
  8. ^ Greenberg (1999)