LINGUIST List 11.2186

Tue Oct 10 2000

Sum: Phonological Change Driven by Imitation

Editor for this issue: Lydia Grebenyova <>


  • Bill Palmer, Phonological Change Driven by Imitation

    Message 1: Phonological Change Driven by Imitation

    Date: Tue, 10 Oct 2000 11:20:33 EST
    From: Bill Palmer <>
    Subject: Phonological Change Driven by Imitation

    For Query: Linguist 11.1890

    On 8 September 2000 I posted the following query:

    "Phonological change driven by imitation: Does anyone know of any instances of phonological (or for that matter lexical or grammatical) change resulting from an idiosyncratic feature of the idiolect of a prominent personality? Several times over the years I have heard a couple of anecdotes about phonological change driven by the imitation of a monarch with an unusual pronunciation. One anecdote has the Spanish interdental fricative arising from imitation of a Spanish monarch with a lisp. The other has the northern European shift to a uvular rhotic arising from the imitation at court of a French monarch who had this as an idiosyncratic feature of their idiolect, subsequently spreading to Paris, then France, then neighbouring countries to the east, explaining the areal and not genetic nature of this change. Is there any truth in either of these anecdotes? If so can anyone help me with references?

    Many thanks to all those who replied:

    Glynis Baguley Laurie Bauer Adrian Clynes Timothy Jowan Curnow Kirk Hazen Keith Johnson kejU.Arizona.EDU Mark J. Jones Andrew McCrum Elena P. Milkova Bruce Moren Simon Musgrave no name given James Ritchie Eric Russell-Webb Marie-Lucie Tarpent Stijn Verleyen

    (Apologies to Andrew McCrum and Bruce Moren for losing their email addresses.)


    Both the lisping Spaniard anecdote and the story about the uvular 'r' are clearly widely known, but almost universally discounted. The reactions of those who responded to my posting ranged from skepticism to outright dismissal. However, a number of other examples of the phenomenon were suggested, and I will return to those later in this summary.

    Andrew McCrum, suspects this phenomenon does occur but has found little evidence. He referred me to: Ohala, John (1981) 'The listener as a source of sound change'. Papers from the Regional Meetings of the Chicago Linguistics Society: 17.

    Kirk Hazen knows of no evidence of a prominent person actually leading a language change, but added: "From the point of view of a quantitative variationist, some one person was the first to pronounce what later became a change, but most often that first pronunciation goes completely unnoticed". He referred me to "Erik Thomas's dissertation (1995 maybe?)".


    Mark J. Jones referred me to an article by Mona Lindau in Fromkin (ed.) (1985) Phonetic Linguistics, in honor of Peter Ladefoged, on what links the members of the rhotic class, which may have details on this and other changes from apical to uvular /r/.

    James Ritchie discounts the possibility that the uvular rhotic in German was a prestige borrowing from French or elsewhere, saying that "many of these ideas were generated through positions about language which were based more on emotional charges than on science. Some of these ideas are found in Trautmann (1879) and Vischer (1889). Evidence for a uvular r in German predates a uvular in French by at least 100 years (see Runge 1973:233). It could just as well have been a borrowing into French from German much the way Germanic h certainly was borrowed in the Middle Ages (see Stephen Cutts 1994). There is also much acoustic evidence for a uvular r in early Germanic (see Ritchie 2000)."

    Eric Russell-Webb questions the direction of the posited cross-linguistic influence: "Firstly, there is no direct evidence which attributes the "shift" from apical (i.e. /r/) to dorsal (i.e. /R/) articulations to any one factor. In both French and German...there is tremendous historical confusion. On the one hand, you have germanisten, such as Jesperson, who attribute the change in German (c. 18th century, for Hochdeutsch at least) to French influence. On the other, there are records of French philologues who attribute the same change in French (earlier for the standard, used in erudite circles... "le francais de l'acad�mie") to Germanic influence. Other anecdotal and dialectal evidence suggests that, within the aire of each language, there were dialects using the dorsal /r/ even when the apical form was solidly standard: in the Tyrol, Graubunden and parts of southern Bayern in the German-speaking parts of Europe; among certain "petit bourgeois" social classes in France. What is clear in the historical evidence is that there was a shift in each of the respected standards and that this shift favoured dorsal /r/. No comprehensive research has attributed, with any degree of precision or credibility, the genesis or catalyst of such shift to any one factor, be it biomechanical/ phonetic, social, or purely phonological. Within each language, however, the massive adaptation of dorsal /r/ by all or most concerned populations is easily attributed to social factors (e.g., the prestige of erudite sociolectes, the advent of universal education in the 19th century, the development of mass communications), once the change took place in those segments of the language-population susceptible to exert influence over others."

    Marie-Lucie Tarpent discussed the origin of the French shift at some illuminating length:

    "The other day I was reading the book by Brian Joseph and H.H. Hock 'Language history, language change and language relationship', and was surprised to see this attributed to a mispronunciation by Louis XIV! As a French person raised in France I studied quite a bit of French history, and although not a specialist I am fairly familiar with the history of the French language, and this is the first time I had heard of this anecdote, which seems totally spurious."

    "It seems quite well-documented that the uvular pronunciation developed first as a stigmatized feature among the Parisian lower class, the apical r being general at that time. This was still the case at the end of the 19th century where this pronunciation was called "parler gras" 'to speak fat' and the R, "r grasseye'" (the pronunciation [gRa] itself, with low back [a], examplifying the feature in question); (although by the time the R became more general in urban areas many people had no idea what the verb "grasseyer" meant any more and used it to mean a stigmatized pronunciation that they themselves did not use). Anyway, during the Revolution when it was important for the former upper classes to keep a very low profile, and often to go into hiding, many people started to use the R in an effort to try to blend with the common people. A (short-lived) reaction to this set in after the revolution when fashionable people affected the almost r-less pronunciation of Josephine de Beauharnais, later Napoleon's wife, who was from Martinique where the planters were apparently influenced by the slaves' pronunciation (this could occur if the children were raised by slave nurses, cf the African influence in Southern US speech) (the present r of Martinique, etc is actually a weak voiced velar, not uvular fricative, as occurs also in the speech of many Africans). Later the pronunciation of a rather weak uvular fricative became general in Paris and also spread to other urban areas (where it might have been adopted first by local revolutionaries), but the reinstated royal court and the old aristocracy, especially the ones who had emigrated to escape the revolution, clung to the apical r, as did most of the country people. It is only in the 20th century that R has become standard, but many older people especially in rural areas still use r. Still, one feature which is stigmatized and which is typical of uneducated people in some areas of Paris (who speak with what French people call "accent parisien", a derogatory term similar to "cockney" or "Brooklyn accent" in English-- note that all these varieties are associated with both a class and a local area of a very large city), is a strongly fricative uvular R, sometimes even pharyngealized."

    "It should be obvious from all this that attributing the long-stigmatized shift r > R to Louis XIV, of all monarchs the most preoccupied with his image, is preposterous."

    "I think that the shift may have started with the double rr, for instance in the effort to differentiate the words gue're (grave accent) 'hardly' and guerre 'war', which are homophonous in present-day French. My main reason for thinking this is not very scholarly but is based on the pronunciation of my grandfather, who was a native speaker of Occitan (where only r was used, as in Catalan or Spanish). When speaking French, he used the apical r in most contexts, but a very strongly fricativized or vibrated, long uvular R, which was obviously an effort for him to pronounce, for words written with double rr. I think the same thing may have happened with the old Parisians (ie r:rr > r:R) and R become generalized to all the instances where other speakers had r."

    "As for the "areal spread", if that is indeed what it is, one would have to know at what time it occurred in Germany: the pronunciation was limited to a stigmatized, local French group at the beginning of strong French influence (Louis XIV); the Huguenots who emigrated to Protestant countries during the reigns of Louis XIV and Louis XV were mostly educated, urban people who would not have used this either, and neither would the French e'migre' nobles during the revolution, since it was r that was associated with nobility and royalty. Nor is the change limited to Germany since R is apparently also typical of the uneducated speech of Durham in England (where R might result originally from the sequence rh ?)."

    "It is possible that a royal or prestigious origin of this and other non-phonemic but socially meaningful shifts of pronunciation was invented later in order to give respectability to a previously stigmatized feature."


    The idea of a monarch-led intedental in Spanish received equally little support. Several respondents noted that if the Spanish interdental was the result of a lisp it would have to have been a selective lisp. Stijn Verleyen believes that P. Lloyd talks about the anecdote in the first volume of his book "From Latin to Spanish" (?1991), and referred me to an article by Andr� Martinet, 'The unvoicing of old Spanish sibilants' (Romance Philology, 1951), which "explains the shift from apico-dental s to interdental c on structural bases". Mark J. Jones suggests that the change was part of a reorganisation of the Medieval Castilian fricative system, and referred me to John Green's article on Spanish in Routledge's Romance Languages survey for more details). He also referred me to a paper by Kirk A Widdison in Rivista di Linguistica (1995) (special edition on sound change) that deals partly with this.

    In Keith Johnson's version of the anecdote the King was one of the Habsburgs and had a strong lisp ("if you've seen portraits of any of the Habsburgs, you'd see that this is a strong possibility"). The people wanted to talk like the king so they adopted the voiceless interdental fricative as the sound represented by the letter 'z' and by the letter 'c' before front vowels. However, Johnson points out: "If the king really did have some physical defect which prevented him from pronouncing the [s], it would have been generalized to every instance of /s/, not just those reflected by orthographic {z} and {c} when the latter appears before /i e/, but not in instances of /s/. And, of course people didn't know how to read back then, so orthography isn't much of an issue. Now, that's not to say that people couldn't imitate him incorrectly, and get his lisp for only a subset of the contexts in which he himself used it, but this is not likely, since cases of /s/+V[+ant] contrast with theta+V[+ant]. Also, if I'm not mistaken, the deaffrication of [ts] and [dz], the former of which was the predecessor of the Castillian theta, was already well underway, if not complete, at the time of the Habsburgs... There are records of Arabic and Hebrew writing showing orthographic symbols for an interdental fricative in the Middle Ages, which might indicate that the process preceded the Habsburgs by a LONG time."

    Tim Curnow dismisses the idea of a lisp, "or else there was a very, very clever, very learned Spanish king. The Spanish interdental fricative turns up everywhere that Latin historically had a 'c' [k] before a front vowel, whereas wherever Latin had an 's' [s] before a front vowel, Spanish has an /s/ [s]. So he managed only to lisp in all those words that were derived from Latin words which historically had a [k] not an [s]... It's generally thought that the story is something like this: At some point Spanish had two different s sounds, an apical (from Latin s) and a laminal (from Latin k). In an attempt to keep these two more clearly separate, the laminal moved further and further forward, ending up interdental (that makes it sound like it was deliberate, but you get the idea). But then I suppose it's faintly possible that there was a king once who had a 'lisp' that meant that he said the laminal (but not the apical) as interdental - but I don't think there's any support for that view (or any way of testing it as a hypothesis)."


    With lisping Austrian Spaniards and a working class uvular Sun-King consigned to the dustbin of linguistic urban myths, we can turn to other examples, where monarchs remain a popular posited source of change. Although remaining sceptical, Glynis Baguley provides this addition to the list:

    "In British English, the pronunciation of the <ei> bit of either/neither can be /i:/ or /aI/; I don't know what the geographical or social distribution is but I think it's probably quite complex. The /aI/ pronunciation (which I use) has been attributed to George I, who was German and never really learnt English thoroughly. I think I've also heard him credited with the /t/ in the pronunciation of the river Thames."

    Bruce Moren points to "the heavy borrow from French into Swedish as a result of the coronation of Jean Baptiste Bernadotte (a Frenchman) as the Swedish King in the 19th century. He was a French General who was "adopted" by the Swedish King (who had no heirs). The Bernadottes are still the ruling family. I have also heard it said that the royal family still has 'a slight French accent when speaking Swedish'."

    >From influence from a different first language to influence from a different first dialect and a more specific example, Elena P. Milkova (St.Petersburg, Russia) drew attention to an instance of this phenomenon in Soviet Russia:

    "Certain phonetic (not pure phonological though) changes driven by imitation took place in Soviet Russia: Brezhnev as well as Gorbachov after him both shared a striking Southern dialect feature: they pronounced fricative [gamma] instead of regular [g]. It soon became a matter of fashion to pronounce fricative 'g' among other party leaders, especially among not so important leaders in provincial cities. They copied such feature of pronunciation even if they did not have it previously. Now this fashion is forgotten but in theory it could become more widespread and could even became a norm. Notice that such "innovations" happen mostly in tyrannies."

    Simon Musgrave and Adrian Clynes both mentioned a similar and even more specific example in Indonesian. It is worth quoting Adrian Clynes at some length on "the following facts I observed while living in Indonesia, about how President Suharto (apparently) led a language change, if only among some of his bureaucrat lackeys."

    "Indonesian/Malay has an applicative verbal suffix {kan}, normally pronounced [kan]. A characteristic of some older (mostly aged 70+) Javanese is to pronounce {kan} as [kn] ( = schwa). This is no doubt because of the influence of a largely functionally equivalent cognate affix in Javanese [their first language], which is pronounced [akn]."

    "This was a well-known, often commented upon, feature of the speech of the former dictator Suharto, who reigned from 1966. Living in Bali in the late seventies to mid 80's, it was very noticeable that some of the local middle-to-high government officials also used this pronunciation when making speeches, even though they were younger, and ethnically Balinese, not Javanese. It seemed clear that by using the "Suharto pronunciation" (my term) they were signalling that they too frequented the Corridors of Power, that they too held Power."

    "The Suharto pronunciation was not only used by some Balinese officials; it became quite widespread among (some) bureaucrats throughout Indonesia. One brave, very well known TV commentator, for years had presented a program on 'how to speak correct Indonesian'. He lost the job in the mid-eighties, apparently because he dared to criticise the affectation on his show."

    "The conditions in the 70's and 80's were ripe for such a dictator-led linguistic change, particularly in Indonesia, where differences in the power/status of speaker and hearer have traditionally been signalled linguistically. A far-sighted linguist would have kept tabs on subsequent changes in pronunciation with the coming of 'reformasi' and the overthrow [was it in 1997?] of the hated Suharto and family. I wasn't able to, but I'll bet that those 'lisping' Balinese officials have long since jetissoned their now-no-longer-prestigious [kn]s, and returned to the safety of [kan]."

    "P.S. It is just possible that the change actually began to occur earlier, and was led by the previous president Sukarno, who may well have had the same pronunciation as Suharto, being also Javanese and older than Suharto. I never thought of asking anyone at the time, but I have never heard any comments to that effect. It would be nice to get the views of older Indonesians on this."


    And from the sublime to the ridiculous or vice versa, Andrew McCrum observed: "The only similar instance I know of involves a stress shift amongst some British English speakers where the word harass shifted stress from the first to the second syllable. This, I believe, was introduced by the actor Michael Crawford whose catch phrase it became in the British television series 'Some mothers do 'ave 'em' about twenty years ago."

    Likewise, Laurie Bauer identify several instances of cross-dialect lexical influence arising from specific television or film characters: "the use of 'dude' as a term of address outside the US in the 1980s can probably be attributed to the Teenage Mutant Ninja Turtles. Similarly, the use of the tournure illustrated in 'That's brilliant -- Not!' (discussed at some length on LINGUIST some time ago) was attributed to Wayne's World."