I am very grateful to JJM who has answered my plea for information about what the French do with feminine nouns referring to masculine creatures (e.g. sentinelle, recrue) (see his note to “(s)he vs they”), and says that the French have less difficulty separating gender from sex than English speakers do.

Now comes the worrying part: Language Log reported last week that the French no longer agree on the genders of nouns: “Fifty-six native French speakers, asked to assign the gender of 93 masculine words, uniformly agreed on only 17 of them. Asked to assign the gender of 50 feminine words, they uniformly agreed only 1 of them”.

Does this mean that we foreigners can now ignore the problem, as we are likely to be “wrong” only as often as the native speakers?

The first dictionary definitions of essential are ‘vital’, ‘absolutely necessary’. However there is one place where we see the word every day when this is not its meaning, even if the users of the word would like us to think it is. Manufacturers of soaps, shower gels, shampoos and the like are keen to tell us when they have added “essential oils” to their products. There are alternative names for these oils: volatile oils or ethereal oils, but of course these names would not serve the ends of the manufacturers – they do not give the impression that without the oils we will die!

All that essential oils do is add a pleasant smell to the product: “essential oil – an oil present in and having the characteristic odour of a plant etc., from which it can be obtained by distillation” (Oxford English Reference Dictionary). Essential in this sense is the adjective formed from essence in its subsidiary meaning: “an extract obtained by distillation etc., especially a volatile oil; a perfume or scent, especially made from a plant or animal substance”.

Essential? I don’t think so!

Jack Windsor Lewis has brought up the subject of the letter ‘j’ and its interpretation by English speakers when it comes in non-English words. It is a problem: in the Germanic languages, plus Polish, Czech and Italian it is regularly pronounced [j] (i.e. like the English consonantal ‘y’), but in Spanish it is a voiceless velar fricative [x], while in French and Portuguese it is a voiced palato-alveolar fricative [ʒ]. Only in English, among the languages best known in English-speaking countries, does ‘j’ (and sometimes ‘g’) represent a palato-alveolar affricate [dʒ]. So, what are English speakers to do when confronted with ‘j’ in a word or name of foreign origin? Jack mentions adagio, Beijing, Gigli, raj, and Taj (Mahal). To these, I can add an even stranger one: Abidjan, the capital of Côte d’Ivoire, in which the letter combination ‘dj’ makes it quite clear that the affricate is intended. Nevertheless, BBC reporters – even those stationed in West Africa – frequently pronounce this with a fricative rather than an affricate.

In one name, where the orthography has initial G, the ‘mistake’ is to use a velar plosive instead of the affricate: Genghis Khan. The -gh- in the middle gives away the fact that we have borrowed this spelling from Italian (perhaps as far back as Marco Polo), and that therefore while the medial consonant is a velar plosive, the initial one is intended to be an affricate. Confirmation of this comes from the German spelling given in the Duden Aussprachewörterbuch: Tschingis, and the names of various present-day Central Asians: Chingiz, although oddly, French seems to have opted for the opposite to the English mistake, and gone for two fricatives: Gengis, pronounced [ʒɛ̃ʒis] – at least according to the Larousse Dictionnaire de la prononciation. We are obviously afraid of the affricate – perhaps we know that it is rare in European languages, and assume that therefore it can’t ever be the right sound in a foreign word. But to my ears the result is not that we replace the affricate with the straight forward palato-alveolar fricative [ʒ] that occurs in pleasure [‘pleʒə], which is usually lip-rounded and laminal, but with a less lip-rounded, and often apical articulation.

Has anyone else noticed this?

My post on the use of third person pronouns and the problems of sexism in language has generated quite a few comments. In fact, although we can bemoan the lack of a neutral third person singular pronoun in English, at least we can get round it by using ‘they’ and making the sentence plural. The Romance languages have it even harder: the third person plurals are also gender-specific. How do French, Spanish, Italian, etc. feminists cope with that?

In that paragraph, I used the word “gender-specific”. I meant precisely that: that the Romance language pronouns show the grammatical gender of the nouns they refer to. They do not indicate the sex of the animate beings. Or do they? What does French do when confronted with the necessity for using a pronoun for the second reference to army recruits? The French for recruit is “la recrue”, despite the fact that until fairly recently, they will all have been male. Similarly “la sentinelle”. I have read quite extensively in French, but I don’t recall ever coming across a solution to this conundrum. Any evidence will be very welcome …

John Wells’ blog has been talking about the mispronunciation of ‘chorizo’ in a TV advert, in which the ‘z’ is pronounced /ts/, as if this were an Italian (or maybe German) word rather than either /θ/ or /s/, depending on the variety of Spanish the speaker uses.

There are at least two other cases of orthographic ‘z’ being misinterpreted by English speakers. One is standard, the other rather more doubtfully acceptable. Both are Greek prefixes: schizo-, almost invariably pronounced /’skıtsou/ and piezo-, frequently heard as /’pi:tsou/, although “correctly” it ought to be /paı’i:zou/. All the current pronouncing dictionaries give many versions of the second of these, but the Oxford BBC Guide, as usual more prescriptive, allows only the pseudo-Greek version, not mentioning any of the hybrid German/Greek/English interpretations.

In today’s blog, John Wells quotes a long piece by Eric Armstrong, a voice, speech and dialect coach. I have no complaint with any of the content of what he says, but towards the end, the pronoun he invariably uses for “actor” is “she”, “her”: “But silliness has great value to an actor! It frees her up, lets her connect to those new sounds in a joyous, unfettered manner, stripping away all those value judgments of sounds (and ultimately symbols) as being “mere math”. And once she can hear those sounds in her own mouth, to feel the physical action required, the visualize the action of her articulators, she is ready to begin to learn how to write them down systematically with IPA.”

Am I the only person who finds this pandering to the extremist feminist lobby offensive? The English language has no neutral pronoun for the third person singular, either as subject or object, but to use “she” and “her” in this way brings me up short every time. I am given the impression that it is only Eric’s female students who value silliness. Earlier in his piece, Eric has avoided the problem by keeping to the plural form. Why could he not continue? This is an increasing tendency among American writers. I do not believe it does anything to enhance the feminist cause, and simply annoys large numbers of otherwise peaceable citizens.

To quote John Wells from yesterday: “End of rant”.

The re-spelling system used in the recent BBC Oxford Guide to Pronunciation introduces some dubious practices. For instance, the short vowel (John Wells’ FOOT vowel) is represented there by UU. This is counter intuitive, especially as the GOOSE vowel is re-spelled as OO. This means that the closer vowel is represented by what is usually thought of as a more open vowel symbol. Consequently, Rushdie is re-spelt ‘ruush-di’ in the BOGP. I have used the more traditional BBC re-spelling system for non-adapted typewriters, which has ‘oo’ for the FOOT vowel, and ‘oo:’ for GOOSE. Not perfect, but at least it doesn’t contradict IPA practice. Hence ‘ROOSH-di’.

Petr Roesel asks about the history of the Pronunciation Unit, and various other questions.

The Unit had a predecessor body called the Advisory Committee on Spoken English, that was set up by the BBC’s first Director General, John Reith, in 1926 – when the C still stood for Company rather than Corporation. This committee originally had six members, including the then Poet Laureate, Robert Bridges, as Chairman, and George Bernard Shaw as Deputy Chairman. The Secretary was Arthur Lloyd James, at that time a lecturer in phonetics at the School of Oriental Studies, London University, but later to become the Professor of Phonetics there. Daniel Jones, the Professor of Phonetics at University college, London, and author of the English Pronouncing Dictionary, was also a member. The initial job of the committee was to make recommendations to announcers on the pronunciation of words that presented them with some difficulty – either because they were unfamiliar, or because they had two or more current pronunciations. The linguists on the committee were well aware that they were not deciding on correctness, but Bridges and Shaw, and the other non-professionals on the committee, believed that they were helping to maintain high standards in English usage. When Bridges died in 1930 (aged 86), Shaw (who was merely 74 years old) took his place as Chairman, and other luminaries were asked to join. The Committee eventually reached a figure of thirty members, and as anyone who has worked on committees knows, the greater the number of members, the smaller the amount of work completed. However, apart from the pronunciation of vocabulary words, the Committee set about discovering the pronunciations of place names and personal names, first in the British Isles, and later in the wider world. For British names in particular, this is something about which some degree on unanimity can be achieved – establish who “owns” a name (the bearer if a person, or the inhabitants if a place), and follow them. Inevitably there are a few disagreements, such as the quality of the vowel in the place Bath: long (as in the south) or short (as in the north)? In some cases, the inhabitants of a place disagree among themselves: Shrewsbury: -ew as in ‘sew’, or as in ‘shrew’? There are even families some of whose members pronounce the family name in one way, others in a different one. Foreign names are more problematical, as I have pointed out in other posts, since the differences between English phonology and that of the language from which the word or name is borrowed have to be accommodated in some way.

The Committee was suspended at the outbreak of the Second World War, and replaced (originally ‘for the duration’, but so far the War has apparently not ended!) by the Pronunciation Unit, staffed by two Scottish maiden ladies: G.M.(‘Elizabeth’) Miller, and Elspeth Anderson (‘Andy’), and a clerk. Despite an increasing workload – more radio and TV networks, more daily hours of broadcasting – this remained the entire staff until 1957, when a third linguist was appointed. I never met either Elizabeth or Andy, but their influence was still felt when I joined the Unit in 1979, when I succeeded yet another Scot, Mrs Hazel Wright, as head of the Unit, with the title Pronunciation Adviser. One of Elizabeth’s last successes was the publication by OUP of the BBC Pronouncing Dictionary of British Names (1971), and I was the editor of the second edition which came out in 1983 (paperback 1990). This is now out of print, and neither the BBC nor OUP seems interested in a third edition, which I feel is a shame, as there is no equivalent available.

I eventually persuaded my management in 1984 that we needed more staff, and a fourth linguist was appointed, but five years later, in one of the BBC’s periodic hair shirt periods of retrenchment, that post was withdrawn when one of our number resigned. From the start of my career in the Unit, I was keen to improve the technology employed, and suggested as early as 1983 that the index, by then numbering the hundreds of thousands, could be computerized, and a synthesized voice added so that our service would be available 24 hours a day. No money was forthcoming for this until the early 90s, but eventually by 2002, the index was available to every employee on the BBC desktop, complete with the audio component. I also got the official name of the Unit changed to Pronunciation Research Unit, to reflect more accurately the nature of the work we were doing.

In 2001 another round of cuts required volunteers for redundancy, and since my management had been hounding me for the last four years, I decided that enough was enough, and put my name forward. It was accepted with alacrity, and the post of Pronunciation Adviser was closed. The Unit is now managed, so far as I understand, by a non-linguist, although one of the three remaining linguists (the clerk’s post disappeared with computerization) has the title Co-ordinator. Her name is Catherine Sangster, and her CV can be found on the BBC website.

I hope this answers Petr’s questions?

Petr Roesel asks if the pronunciation of Salman Rushdie that I gave as the “correct” one is an accurate reflexion of the pronunciation in Urdu.

All words borrowed from one language into another, whether names or not, have to be adapted in order to fit the phonology of the borrowing language. Sometimes, when the borrowing language is a well-established literary language and has its own writing system, it prefers to adopt a pronunciation that reflects the spelling as interpreted by the borrowing language. For instance, the word football has been borrowed by French with the English spelling, but with the second syllable pronounced ‘bal’ (as in ballet), whereas Spanish has borrowed a close approximation to the English pronunciation, changing the spelling to fit: fútbol.

In the special case of names, the BBC always aimed to make a recommendation that was as close to the native pronunciation as was possible while using mainly English sounds. I am not a speaker of Urdu, but I think I am safe in saying that in Salman Rushdie’s name, the two vowels of the first name are both open, that the second is long while the first is short, and that stress falls on the second syllable. In the family name, the first vowel is back and close rather than open, the second front and close, and the stress is on the first syllable. I would expect a French or German speaker to use a uvular R at the beginning of the family name, a dental L, N and D, and a clear L in the first name, where English speakers would have an approximant R, alveolar L, N, and D, and a dark L. While the French speaker would have no trouble with the initial voiceless S, a German might substitute a voiced sound, although Austrians would find the voiceless sound natural to their dialect of German.

As a reminder, the pronunciation recommended to the BBC by friends of the author is salMAAN ROOSHdi (-al as in “pal”, -oo as in “book”, and the stressed syllables in capitals). It is what you might call “Urdu with an English accent”.

Starting in the early 1990s, I was increasingly asked at the BBC how we should be pronouncing the names of the years following 1999. Until then, there was only one way of naming the years: by grouping the numbers in twos, i.e. 1899 was eighteen ninetynine, or, in the case of the last year of a century, by saying, for example, nineteen hundred. The choice for the future appeared to be between two thousand and twenty hundred, and then two thousand and one and twenty oh one, etc.
For 2001, I always thought that two thousand and one was inevitable, as that was the title of the Stanley Kubrick film from the 1960s, and for over thirty years it had been pronounced like that, but the rest of the decade would have to wait until we got there. The oddity was 2000, which was (and is) rarely called two thousand, and never twenty hundred. The usual pronunciation is the year two thousand, and very often the words “the year” precede “2000” in writing as well.
Charlotte Green was, I think, unique among BBC newsreaders and radio announcers in calling the subsequent years twenty oh …, and this habit was questioned on Radio 4’s Feedback (BBC domestic radio’s ‘listeners’ letters’ programme). But early in 2006, she succumbed to the adverse criticism, and started to say two thousand and … like all her colleagues.
Interestingly, for years following 2009, we hear much more the twenty oh … style, and I suspect that at New Year 2010, we shall start to hear this regularly.
It is very disappointing that Charlotte received hate mail for this pronunciation. There was nothing wrong in English terms with the style she had adopted, and the British public should be more tolerant. However, I know from bitter experience that there is nothing that makes British people more angry and abusive than a perceived mispronunciation.
In the case of year pronunciations, the French have always had two styles, and as a student of French at secondary school, I was made aware of this very early on: 1968 is either mille neuf cent soixante-huit, or dix-neuf cent soixante-huit. Why should English be restricted to a single way? There’s more than one way of skinning a cat!