Friday, 4 February 2011


Steve Doerr writes asking me to comment on
the name ElBaradei, prominent in the news just now. The location of the stress and the pronunciation of the -ei are the bones of contention for me. (I'm not an Arabist though.)

In LPD I give əl ˈbær ə daɪ, with alternatives el-, -deɪ. As far as the main pronunciation is concerned, in this I follow the recommendation of the Oxford BBC Guide to Pronunciation, which draws on the expertise of the BBC Pronunciation Unit and (presumably) the advice of Arabic speakers working for the BBC World Service (now so shamefully to be drastically cut back). The OBGP notes further that “this is his preferred anglicization”.

For some reason I neglected in LPD to supply the Arabic pronunciation, which to the best of my knowledge is [ʔelbæˈɾædʕi].

In Arabic the name is normally written البرادعي ʾlbrādʿī (see for example here), with short vowels as usual not shown. The final consonant is an ʿayn (pharyngealized glottal stop).

Classical Arabic has only three phonemic vowel qualities, /i a u/ (short and long), and I think the same is true of Egyptian Colloquial Arabic. The consequence is that there is considerable allophonic and other variability in the actual qualities of the vowels. So /a/ is typically æ in non-pharyngeal contexts, but in pharyngeal contexts ɑ, while in the definite article it can even range over e and ə. So I do not think there is any call to be dogmatic about whether we make the final sound in English or .

Word stress in Arabic can fall on any of the last three syllables of the word, and is predictable from the syllable structure. The rules are somewhat different in Egyptian from the rules for other varieties. If I am right in thinking that in this name it falls on the penultimate, that would be because of the consonant cluster following the penultimate vowel. But when we anglicize it we can’t cope with ʕ, which we omit, so that there is no cluster and the stress reverts to antepenultimate. (Or am I fantasizing? Is the vowel long, as written, and therefore stressed?) What Steve Doerr presumably doesn’t like is the final stress heard from some newsreaders.

Anyhow, you can see from these rather vague and tentative remarks that I am not an Arabist either. I would welcome comments from those who know more than I do about Egyptian Arabic phonetics.


  1. I'd prefer [ʔelbæˈɾɑdʕi] for the (Modern Standard) Arabic pronunciation, because /ɾ/ causes backing in /æ/. (In Arabist terms, Rayy is a "dark" letter.)

    The broad transcription would usually be given as /ælbæˈɾædʕi/.

  2. I'm very far from being an expert, but I did work in Egypt, and i did take a few elementary courses in Egyptian Colloquial Arabic. If I remember correctly, the definite article was always transliterated as el in teaching materials and usually in business names, street signs etc written in Roman script. In the days when Teach Yourself books were heavily into phonetic transcription, TF Mitchell represented it it as ʕil.

  3. No that wasn't a mistyping. Mitchell used ʕ for a glottal sop and reversed ȝ for ein.

  4. Mitchell states that in Egypt the qualities he transcribes as a and ɑ may vary in the same word from speaker to speaker — depending not only on dialect and the degree of inclination to follow Classical rules, but also on the sex of the speaker.

    Following r he gives the example of garráaɧ (surgeon) in typical women's speech versus garrɑ́ɑɧ in typical men's speech.

  5. Mitchell's statement of the word-stress ('prominent syllable') rule is:

    ULTIMATE if the syllable is long
    ANTEPENULTIMATE if all the last three syllables are short
    PENULTIMATE by default

  6. Regarding those word stress rules, Teach Yourself Arabic gives a different set of rules.

    That book says the last "heavy" syllable in a word carries the stress, where a heavy syllable is a closed syllable with a diphthong or long vowel. Possessive and object suffixes should be removed before the syllables are counted, and prefixes such as al- and wa- can't carry stress. Also, the adjective-forming suffix -iyy (or -iiy) is treated as light unless followed by the feminine suffix -ah.

    I think these rules will give different results, but these are for MSA rather than Egyptian Arabic.

  7. I'm not sure where you get "In Arabic the name is normally written البرادع lbrādʿ" - in Arabic you're missing the final yaa', the [i] sound, and in the transcription you're missing both the initial eh- sound that comes from the Arabic alif and the final [i] as well.

  8. Lane: You're right, it should be البرادعي. (Now corrected)

  9. It's complicated. After the /d/, there might be a short /a/ which might as well be elided. You're right about the missing yaa in spelling and /i/ in sound. Concerning the article, Egyptian Arabic tends less to reduce the whole thing to an /l/ or the lengthening of the following consonant, but I'm not sure the phenomenon's all that alien.

  10. Thanks, JW, for granting my wish.

    As Lane mentioned, I think you missed a vital letter at the end of the name.

    Ignoring the definite article, my assumption was that the name/word baradei was four syllables and that treating the ei as a diphthong of any sort was wrong. That would depend on where you insert the short vowels, so my assumption was that the e goes before the ayin rather than after: ba-rā-de-ʿī. I've tried to find the word in Hans Wehr's Arabic dictionary. There's an entry for بردعة which just says = برذعة (barḏaʿa 'saddle, pack-saddle (for donkeys and camels)'), which has a derivative براذعى (barāḏiʿī 'maker of donkey saddles, saddler').

    I also looked at the Oxford BBC dictionary, but I wasn't sure whether 'preferred anglicization' referred to the spelling or the pronunciation.

  11. Pete

    That rule doesn't cover words where not one of the last three syllables is 'heavy'.

    Mitchell's rule for Egyptian in such words is that the stress in antepenultimate — with a couple of exceptions:

    1. If the last syllable includes a pronominal suffix ending in -h

    2. If the word is a plural form with the first two of the three syllables containing the unusual sequences of vowels u-u and i-i

    Under these conditions, the stress may shift to the penultimate: e.e ramítu(h) she threw it, libísa underpants. The shift is not universally made for the plural forms — especially outside Cairo.

    I don't think there's any clash over adjectival endings, just a difference in transliteration/transcription. It seems that where Teach Yourself Arabic has -iyy or -iiy, Mitchell's Teach Yourself Colloquial Arabic has -i. Or perhaps there's a real difference in sound....

  12. Lipman

    As I understand it, Egyptian doubles the consonant sound only where Classical Arabic does so. And it's not always possible to elide the vowel. That's why Mitchell gives the citation form ʕil

  13. In practice, MSA stress is underspecified, and the effective rule is "Stress falls where your colloquial makes it fall."

  14. Sheikh Daoud (Arabic instructor)4 February 2011 16:01

    Since the Arabic news is almost always broadcast in Standard Arabic and almost all written Arabic is Standard Arabic, it seems to me the pronunciation we ought to follow is the standard rather than the Egyptian. The syllable "ba" cannot be stressed here as it is a short vowel followed by a long one. The "ra" is stressed, and the "i" follows the "عayn".

  15. David Crosbie - yes, I think the rules in Teach Yourself Arabic should be bottomed out with an additional rule saying that if none of the last three syllables are heavy then the antepenultimate carries the stress.

    Also, I should have said that a "heavy syllable" is either a closed syllable with a diphthong or long vowel, or a short vowel followed by two consonants. My bad.

    No, no, of course there's no clash over whether the adjective-forming suffix should be given as -iyy or -iiy. It's just a difference of transliteration as you say, and a difference that you'll notice doesn't affect the stress-placement rules (because both iyy and iiy would be heavy). And it's actually closer to -iy or -ii or even -i, all of which are light, which is why it's excluded from the stress-placement rules.

    But, when followed by the feminine suffix -ah, the full weight of the adjectival syllable reasserts itself, giving -iyyah or -iiyah and shifting the stress to the penultimate. Hence مغربي (maghribiy(y)) /'mæʁribiː/, but مغربية (maghribiyyah) /mæʁri'bijːæ(h)/.

  16. Is this Arabic: The name occurs about 13-15 seconds into the clip.

  17. John Wells said:

    > What Steve Doerr presumably doesn’t like is the
    > final stress heard from some newsreaders.

    Actually, no: it was the stress on the 'Ba-' syllable that I didn't like (as opposed to '-ra-', which Sheikh Daoud confirms as the stressed syllable).

  18. Marijana Krleža4 February 2011 19:26

    Wikipedia gives it as [mæˈħæmːæd mosˈtˤɑfɑ (ʔe)lbæˈɾædʕi].

  19. Steve Doerr

    That clip (and some others) seem to end the name with a long — and therefore stressed — i:.

  20. Marijana Krleža said...

    > Wikipedia gives it as [mæˈħæmːæd mosˈtˤɑfɑ (ʔe)lbæˈɾædʕi].
    Which raises a couple of questions:
    - why ?
    - why would the tonic accent in مصطفى (Mustafa or Mostafa) fall on the middle syllable?


  21. In the YouTube news clips I hear a long and stressed final vowel. But this may well be a refection on my English-speaker's ear. I have difficulty detecting a stressed short vowel in Czech when it's followed by a long vowel. Am I mishearing Arabic in the same way?

  22. OK so if it's البرادعي then that's /ælbæˈɾɑːd(æ)ʕiː/ in Arabic, which would be Anglicised as /ælbəˈrɑːdi/. If, as some of the later comments seem to be saying, it's البردعي then that's /ælbæˈɾɑdʕiː/, giving the same or maybe /ælbəˈrædi/.

    Either way, the Anglicisations you normally here, /el'bærədei/ and /el'bærədai/, are both out of line with the Arabic pronunciation.

  23. Pete


    A shift of stress to the final syllable makes either of those quite a plausible approximation to some pronunciations.

    el approximates to what Egyptians say

    • final stress sounds like (at least) some TV news-speaker pronunciations

    • some diphthong of the ei/ai variety is the nearest most English speakers can get to ʕiː

    I see where the Roman spelling comes from.

  24. I'm certainly no Arabist either, but I used to be very into the comparative study of Arabic vernaculars, and I agree with Pete about the "dark" rāʾ causing backing, certainly in some of the vernaculars, but not distinctively so in the way the emphatics with non-emphatic correlates do, though the situation in spoken Egyptian seems to be more complicated. And I think it does so in Steve's clip (which does seem to be some sort of MSA, though of course it may be a bit like 'near-RP'), but it sounds as if the result is further forward than Pete's ɑ, and the ra is certainly not following Sheikh Daoud's instructions.

    >I also looked at the Oxford BBC dictionary, but I wasn't sure whether 'preferred anglicization' referred to the spelling or the pronunciation. <

    It's hard to believe it referred to the pronunciation: it would be odd to stress the ba when the vowel is only there for epenthetic reasons and would be missing in the modern vernacular but for the el- then causing a banned sequence of three consonants. On the other hand if Lipman's /a/ (or even [a] etc) is missing in -daʿī the resulting sequence of two consonants shortens the long a of the preceding rā.

  25. David, I do see why you hear a long and stressed final vowel in that clip. I think it's the sequence aʿī represented in the Anglicization by –ei or –ai. What I hear the name as is ʔɛlbɐrɐˈdæʕejj. Of course I too have no doubt failed to avoid hearing it in accordance with my interpretations, but for this "shift of stress to the final syllable" as you call it to have any basis in reality, it must be because EITHER the adjectival ending is both long, as with the 'iyy' or 'iiy' you've been talking about, and moreover subject to the stress placement rules, even if I am imagining the /a/ suggested by Lipman bouncing back, OR if I'm not imagining it, that /a/ is what is stressed, in its capacity of being a non-final light syllable that directly follows a (classically, but no longer, it seems) heavy syllable, by rule (2a, 2b) in the Wikipedia article on Egyptian Arabic. This non-imaginary [æ] would give us Steve's four syllables for 'baradei', and even though the difficulty we have in disentangling it from the ʕ might still make the whole thing sound like a diphthong to our infidel ears, he is absolutely right to conclude that we cannot even think oftreating it as a diphthong, not even of any sort. MSA is of course a pretty artificial media language, but that doesn't appear to have stopped the long a of the above-mentioned classically heavy syllable rā getting reduced to ɐ here, just as in the vernacular.

    As for this 'iyy', as I remember the writings of TF Mitchell, a lot of his phonology and phonetics seemed to be mixed up with morphology, but of course that is the way Arabic orthography was set up in the first place. So this unorthographic doubling of y may just be his way of tidying up the relationship between the masculine ending and the feminine ending in –iyya. However monosyllables do have noticeable gemination of final consonants after short vowels, alternating with non-geminated consonants after long vowels, and it's not beyond the bounds of possibility that you get this alternation with final stress if that is what it is in this clip. Obviously j is a consonant for those purposes. Thus etymological (and if you like phonological) /aj/ /aw/ /ij/ /uw/ may have the reflexes [e(ː)] [o(ː)] [i(ː)] [u(ː)] etc medially but various vowels with [jː] and [wː] after them finally.

    It occurred to me that an awareness of this final gemination might prompt some Arabic speakers to use the shadda (ّ) on the consonant, thus explicitly doubling, lit. "strengthening" it. I got 109 Google hits for + البرادعيّ(ElBaradei with the shadda doubling the final yāʾ) as opposed to 3,630,000 for the standard البرادعي without it (including البرادعى with alif maqsura, which seems to be widely used in Egypt for final yāʾ). It appears that Google can be made to search + البرادعىfor the spelling with alif maqsura as distinct from the spelling with final yāʾ, giving about 1,290,000 hits, but that never carries the shadda, and it seems from another google that nobody has tried to give it one (except of course the insane mallamb). The shadda featured as far as I could see on sites which would not be the first choice for anyone seeking a guide to literacy, but is it not of considerable significance for your iyy or iiy and a shift of stress?

    Can some competent Arabist please comment on this as evidence for a native-speaker sense of this allophony?

    Lipman, even if you're only commenting from the perspective of Hebrew, it must be very trying for you to see us all drowning in a puddle like this.

  26. I'm actually more interested in how should one 'properly' pronounce Taḥrīr? Because I'm a bit irritated by all the /tʌˈrɪə/ or /təˈrɪə/. Could it be something like /tʌhˈriːə/ or even /tʌhˈriːr/, which is a bit too rhotic for British English, I guess.

  27. Funny how we seem to hear the same clip different ways! I hear it with the -ra- syllable stressed, and what I interpret as secondary stress on the final syllable (after the ayin), but given extra prominence by an intonational effect that I attribute to the fact that the newsreader is taking a slight pause after delivering a proper name (possibly unfamiliar to some viewers, though that's unlikely in this case!) before resuming ordinary language. I hear a definite -da- syllable (not sure about the vowel) before the ayin, but I don't hear it bearing the primary stress as mallamb seems to indicate. But maybe I'm hearing what I want to hear! (It's nice to feel vindicated.)

    As for Taḥrīr, it quite often seems to come out as tʌˈhɪə (which seems to be taking non-rhoticism a bit far)!

  28. Mallamb

    It sounds as though Mitchell had different policies according to the purpose of the work. In Teach Yourself Colloquial Arabic he explicitly says that he writes -i if that is the sound spoken — even when spelling and grammar create a different expectation.

    The book is an interesting hybrid. The first part is a scholarly account of the phonetics and grammar of Egyptian Arabic. The rest of the book is, pedagogically speaking, rather rubbish — a no doubt accurate but unserviceable set of themed vocabulary lists and phrases. No exercises or progression of any sort. How it got into the Teach Yourself series is a mystery.

    A friend once told me of a talk given by Mitchell on his linguistic analysis of the camel market in Tripoli. Afterwards, my friend's Arabic-speaking colleagues expressed bemusement. They felt as if they'd attended an academic lecture conducted by a provincial camel trader.

    Mitchell consistently represents the masculine singular adjectival ending as -i. By the rules he gives, the alternative -ii would (in Colloquial Egyptian) be stressed.

    Pete's Teach Yourself Arabic covers a different variety and apparently allows for the transliteration -iyy or -iiy. This necessitates a rule specifically redefining the final syllable as 'light' and therefore unstressed. If accurate, this observation would seem to rule out your gemination analysis — for that variety of Arabic, that is.

    What only a specialist can answer is:

    1. Is el Baradei pronounced with a final long vowel (or diphthong or geminate jj) in any variety of Arabic?

    2. Is it possible in any variety of Arabic for a syllable with a short vowel vowel carrying word stress to precede a syllable in the same word with an long vowel?

    3. If so, can the nature of the stress be such as to be made obscure to foreign ears by the prominence of the long vowel?

    A couple of points on the vernacular that may interest you.

    — In over two years, I consistently heard the word qism in the sense 'Academic Department' pronounced with Classical qaf. With one exception. The College gatekeeper once told me something using the Colloquial pronunciation with a glottal stop.

    — In a talk in English, a sociolinguist made the tongue-in-cheek claim to have discovered a missing link. He played a recording of a pious but very under-educated imam discussing some point of theology. Those that understood him smiled at one another, and shook their heads; but, for mine own part, it was Arabic to me.

  29. Steve

    What you hear as a syllable, I hear as the end of the d followed by the start of the ain.

    The fact that i don't hear any stress on the ra syllable doesn't necessarily mean that I deny its existence. I don't hear short vowels stressed in the environment of the Czech feminising ending -ovaa, although I know that Czechs hear them.

    Well, I can detect the stress with conscious effort and the knowledge that it's definitely there to be detected. It would be nice to have similar certainty with Arabic.

  30. For what it's worth, the English Wikipedia article indeed has an "a" after the d, and the Hebrew article has a schwa.

  31. Then there's Megrahi (مقرحي‎)...

  32. I'm afraid we can't assume that the English Wikipedia article has an "a" after the d for any other reason than that the "a" is in a transcription intended to reflect the orthography. The same goes for the ā in rā, and they're not even consistent about that, as they don't mark the final ā of Muṣṭafā, which is also spelt with a length-marking alif, although not the same one as in Barādaʿī, but the alif maqsura I mentioned in my last post, used in its canonic value for ā this time, instead of for ī, which is supposed to be distinguished from it by the two dots (but characteristically not in Egypt, as I also said there).

    Neither of these alifs are glottal stops, for even the possibility of which Hamza is needed in to give them a bit more oomph. They are simply length markers, so in Classical Arabic the syllables aren't "heavy" as in Pete's rules for MSA from Teach Yourself Arabic, in that they are not closed in the ways he mentions.

    So Steve,
    That would be the answer to your question "why would the tonic accent in مصطفى (Mustafa or Mostafa) fall on the middle syllable?" Even though the final a has lost its length, the classical quantity still determines where the stress falls. The stress sometimes doesn’t seem as predictable as some of these guides claim unless you know what length the vowels are supposed to be, does it? I have already pointed out that the rā in Barādaʿī which reflects the orthography loses its length too, pace Sheikh Daoud, and even though you have succeeded in hearing a residual accent on it I don't think you would say you can hear any distinctive length on it.

    Your other question "why mæ?" in what they claim is aphonetic transcription for Egyptian Arabic [mæˈħæmːæd mosˈtˤɑfɑ (ʔe)lbæˈɾædʕi] had better be thought of as rhetorical! In the media language clip there is a wonderful authentic mʊ, but Egyptian Arabic would not by a long chalk be the only vernacular in which there is no vowel at all between the m and the ħ, or at most a schwa which I suppose might be coloured by the æ realization in the next syllable, given the non-emphatic context of the whole. By the same reckoning, I would have thought the [ʔelbæˈɾɑdʕi]Pete gives for MSA more likely to be [ʔelbɑˈɾɑdʕi, given the context of an r behaving as an emphatic.

    But listen to the second occurrence after the pause: that sounds even more unstressed on the ra.

  33. Thanks, Mallamb. I probably need to read Teach Yourself Arabic to understand! I think I've tracked down the name مصطفى in Wehr's dictionary, where it appears under صفا and is transliterated as muṣṭafan ('chosen, selected[...] epithet of Mohammed'). Is the mu- considered a proclitic and therefore outside the accentuation 'zone' perhaps?

  34. Well it's the same mu- participial prefix as in the name Muḥammad itself actually, Steve, but I don't think it accounts for the stress. I think it's probably a moraic thing, as we have seen with other languages discussed here. This is the sort of case in which I don't object to the term "underlying": there's an extra underlying mora in the final -a even when it's short.

    To continue on my previous theme after another of the successive attempts by BT to install BT Infinity (cursed be its name!) which have been leaving me incommunicado for vast tracts of time:

    I appreciate your chopping the thickets of my prose into bullet points, but I'm afraid I don't seem to be able to do point-by-point answers.

    Like you I could be having difficulty hearing stress on the ra, in my case because what I do hear in both the occurrences of the name I have mentioned is a rise on that syllable followed by what sounds like a more prominent stress, and my ear is attuned to Japanese, where in that situation the rise on the earlier syllable would be a non-functional realizational phenomenon. We had a similar discussion about the Russian pitch accent some time ago, with Lipman having difficulty getting someone to perceive it on the correct syllable.

    But it can't be just we infidels who have this impression of the long and stressed final vowel you hear, or the long and stressed sequence of something or other that I hear. It can't be totally alien to native-speaker intuitions, or why would the people who write البرادعيّ (ʔlbrdʕjː) feel inspired to double the final yod consonant in the spelling? Does this not imply that for them that syllable is indeed as heavy as in the transcriptions iyy and iiy Pete referes to, as opposed to being actually closer to -iy or -ii or even -i, all of which are light. And would they have this intuition that it is heavy if it were in fact excluded from the stress-placement rules? I admit it's counterintuitive for me. I would have thought /'mæʁribi/was fine for مغربي (maghribiy(y)), which is going further than Pete's /'mæʁribiː/, but perhaps that's because there's an uncontentiously light penult, whereas with a penult in ʔɛlbɐrɐdæʕi(ː) which is like Schrödinger's cat, anything might happen.

    So I hope I wasn’t misrepresenting TF Mitchell. I do have some recollection of him saying of vernaculars that he writes -i if that is the sound spoken, for example for Libyan Arabic.

    But even if we go along with the final short -i, and the æ I seem to hear in the preceding syllable is stressed, it might be no more surprising in al-Bara'daʿi for al-Barādaʿī than with the final short -a in Muṣˈṭafa for Muṣṭafā, for the reasons I mentioned before. And that dead-and-alive æ might actually be the e in ElBaradei's own alleged 'preferred anglicization' in the Oxford BBC dictionary, which as I said before I feel sure must have referred to the spelling rather than the pronunciation. As for the long and stressed final vowel you hear and the exaggerated ej: I hear (ej being unremarkable for final ij in Egyptian Arabic), that could be an artificial resuscitation of the long i, i.e. Classical ij, which is thought appropriate for the heightened pronunciation of MSA media language.

    Pete said "OK so if it's البرادعي then that's /ælbæˈɾɑːd(æ)ʕiː/ in Arabic, which would be Anglicised as /ælbəˈrɑːdi/. If, as some of the later comments seem to be saying, it's البردعي then that's /ælbæˈɾɑdʕiː/, giving the same or maybe /ælbəˈrædi/." But surely none of the commentators up to that point were saying it could ever be البردعي in the orthography! It's just that in Egyptian Arabic at least there's this loss of vowel length.