In checking the bibliography of a book I'm copyediting, I hit an article titled "La typologie des catalogues dhes: un rseau gnalogique thmatis." I was stopped in my tracks by the bizarre (to me) word hes; I could make no sense of it and could barely pronounce it (/ee/ sounds very strange all by itself, though of course it's common in words like cr). Was it a typo? But Google and Wikipedia came to the rescue (what did we do before them?)—hes redirects to Catalogue des femmes, which links to the English Catalogue of Women, where the alternate name of this ancient poem is explained thus:
In antiquity the poem known as the Megalai Ehoiai (Greek: Ἠοῖαι or Ἤοιαι; Latin Eoeae, Ehoeae, Eoiae, etc.), from the formula ἢ οἵη (ē hoiē), "or such a woman (as)", which introduces new sections within the poem, is also possibly the same, unless there are two poems in the same style - we know both only from quotations.The French article is more certain of their identity: "Il est galement connu sous le nom dhes ou es..." The form es looks even stranger. At any rate, a unique name with a unique etymology. (And how to pronounce it in English—ee-HEE-ee? Yikes!) Posted by languagehat at October 20, 2009 10:39 AM
The word is totally new to me, but there is no doubt about its French pronunciation /ee/ as you say, whether or not there is an h in the middle.
Posted by: marie-lucie at October 20, 2009 11:10 AMJust think if the Champs-lyses had been valleys instead. Then they'd be the Valles-lysees. French, the logical language.
Posted by: John Cowan at October 20, 2009 11:33 AMehes is placenta in my language, ekh is mother
so eeHeeee does not sound anything yikes to me
Just think if the Champs-lyses had been valleys instead. Then they'd be the Valles-lysees.
Nice try, but no, you don't add an extra e if there is already one at the end of the word.
Posted by: marie-lucie at October 20, 2009 02:40 PMWhat's all this learned pussyfooting about a "catalogue"? Surely the title means "Big mommas".
Posted by: Grumbly Stu at October 20, 2009 03:14 PMI thought you pronounced the first 's' in Champs-lyses ([ʃɑ̃zeliˈze]). Wouldn't the same apply to Valles-lyses?
Posted by: Bathrobe at October 20, 2009 07:04 PMBathrobe, of course! (but what is the tilde doing on your z?). JC wanted to add an extra e before the plural suffix -s.
Posted by: marie-lucie at October 20, 2009 07:33 PMread: if ekh is "mother, what is "father"?
Posted by: marie-lucie at October 20, 2009 07:35 PMbut what is the tilde doing on your z?
I see it over the preceding vowel; must be an issue with your browser.
Posted by: language hat at October 20, 2009 07:47 PMYou know, it is never remotely germane, but I actually have a fondness for my periodic exposure to a new Mongolian word that sounds or means something vaguely similar to the topic at hand.
Posted by: Z. D. Smith at October 20, 2009 07:48 PMBathrobe, apologies; LH, thank you: I should have realized that the tilde belonged on the "a" not on the "z" where it landed as it travelled to my screen.
Posted by: marie-lucie at October 20, 2009 07:55 PMYou know, I see it over the z as well, on both Firefox and Safari.
Posted by: Z. D. Smith at October 20, 2009 08:08 PMIt's on the ɑ here, both Firefox and Safari as well. So it may be a font issue but that's far from certain.
Posted by: Kellen Parker at October 20, 2009 08:47 PMfather is etseg, ekh etseg or etseg ekh, the order doesn't matter, is formal for parents
informally father is aav, mother is eej
sorry, i just have this reaction to the words yikes and eew, start contradicting, should restrain myself from reacting
so why ehees is yikes i did not get
Thank you, read.
Posted by: marie-lucie at October 20, 2009 09:53 PMmy pleasure, m-l
Posted by: read at October 20, 2009 10:10 PMI, too, would like to know how to pronounce this word in English.
But I would even MORE like to know how to pronounce it in Modern Greek!!!!! According to my negligible knowledge of MG, all the diphthongs in this word are pronounced /i:/, as is the eta; and aren't word-initial rough breathings (i.e. the h sound) omitted? So in MG, if I am right, it would be ee-EE-ee !!!
The only halfway-feasible English pronunciation I could imagine would be to pronounce the 'h', and do the standard /e/ -> /eɪ/ conversion, yielding 'ey-HEYS',
Posted by: Z. D. Smith at October 21, 2009 12:47 AM@dveej: Sadly, αι is now /e/, so it would be /i.ie/ . We'd update the inflection now, because the modern 1st declension plural is the old 3rd declension plural; so Ηοίες, /i.ies/.
I'm sure we do this kind of nominalisation in English all the time, but I can't think of an instance of a pluralised function word like this. The closest I can think of is Shakespeare's pun "As's of Great Weight".
Posted by: Nick Nicholas at October 21, 2009 01:24 AMm-l and Z.D., I get it over the z too. I and m-l both have macs, I know, so it may be a mac vs windows thing.
Posted by: A. J. P. 'Contemporary' Crown at October 21, 2009 02:41 AMread, I'm always cool with you posting mongolian info, but what kind do you use? would it be feasible to give examples in cyrillic (which most here should be able to deal with and if they can't they should)?
Failing that, is there a particular system of romanization that you use?
Posted by: michael farris at October 21, 2009 03:13 AMThe word is totally new to me, in Malaysia we pronounce "ehe" like a little laugh.
m-l and Z.D., I get it over the z too. I and m-l both have macs, I know, so it may be a mac vs windows thing.
Well, actually, I have a Mac, too, and it's over the 'a'. Not sure why this is happening.
At any rate, should it be pronounced [valez elize] or [vale elize]?
Posted by: Bathrobe at October 21, 2009 05:56 AMread, speaking of Mongolian words, there is a brand of sports shoe in China called 'Erke'. The moment I saw it I thought "Monglian!" I asked the staff where the company was based, and I was told "Fujian". So it wasn't some ethnically enlightened Chinese adopting a Mongolian name --- it was just another capitalistic chump adopting whatever sounded best. :)
Posted by: Bathrobe at October 21, 2009 05:59 AMso why ehees is yikes i did not get
Surely it's clear from what I said ("And how to pronounce it in English ee-HEE-ee? Yikes!") that my "Yikes!" referred to the suggested pronunciation, which if you read it out loud will probably sound pretty funny even though you're not a native speaker. (Note that "ee" represents the vowel in English "see," not whatever it would represent in Mongolian.)
Posted by: language hat at October 21, 2009 08:17 AM@B&MF :)
yes, Erke sounds Mongolian, if it's erkh - means rights, like in human rights or if it's an adjective means spoilt (like in 'spoiled kid')
i don't use any particular system of romanization, just substitute Cyrillic letters with the Latin ones, o/u all get the same though like in the old script, should be no problem
when people write, for example ou or oy to write , it looks not right to me and even annoying, keep things simple i say
that's easier for us, of course, but for the foreigners learning the language maybe it's better to know if there is o or
i'll try to use translit which i've learned from here
sorry, for the off-topicness
I've had the feeling I don't know much about character sets, fonts and Unicode. I do know that I hated the subject with a passion - a lot of my IT colleagues hate it too. Nevertheless, on the occasion of this problem with Bathrobe's IPA string appearing differently under different OS in different browsers, I decided to look at the subject again. To my surprise, I found an incredibly simple explanation for what's happening.
It's best to think about this business in a way very different from the one implicitly used by the various commenters here - and different from the one I had previously used, which made the subject appear difficult to me. I'm afraid Hat may need to avert his eyes here, because the appropriate approach to understanding is RADICAL CONSTRUCTIVIST. In other words, what we see in our browsers is what we have (implicitly or explicitly) constructed our browsers to show us.
"Constructed" does not mean "constructed out of nothing", but "constructed out of the materials to hand, such as our perceptions and our browsers". When you bake a cake, you have a certain constructive freedom using flour, water, eggs and sugar. You don't create the cake ex nihilo. You can make many different kinds of cake with those ingredients, but you can't make eggplant souffl out of them. Note I am not claiming that flour, water, eggs and water are the basic ingredients out of which the universe is made, i.e. they too are subject to reconsideration.
In brief, you have to select an appropriate font in your browser to see the "proper" representation of Unicode numbers.
Posted by: Grumbly Stu at October 21, 2009 08:51 AMPart 2:
Here are the technical details about the rendering problem. At my German blogsite is a snapshot of Bathrobe's Unicode string in different fonts:
1. What you see in Bathrobe's comment is the rendering, by your browser, of a string of Unicode symbols.
2. Unicode defines a list of "Unicode numbers", to each of which corresponds a different visual mark. Unicode is intended to provide a common basis for the display of certain visible marks in browsers, text editors etc.. (referred to in the following simply as "browser"), This means that the definition of the visual marks has to be independent of any browser that may currently exist.
To make absolutely clear what each visual mark is, the marks must be drawn on a non-software-interpretable medium, such as paper, that has nothing to do with browsers or the internet. A photograph of the marked paper can be published in the Internet without creating problems for the users, since browsers, text editors etc. cannot interfere with photographs.
3. Browsers that "support Unicode" are supposed to work in the following way. When a Unicode number arrives, the browser checks the "character set" it is currently using. A character set is a precursor of Unicode, and there are thousands of them, because they were "standardized" in different ways by different IT people in different countries, working independently of each other.
Like Unicode, a character set is a list of numbers, to each of which corresponds a different visual mark. The problem is that not all software uses the "character set" that you think it should, in the sense that it shows you the visual marks you expected. That depends sometimes on the operating system (OS) as well.
4. A "font" is a basically kind of icing on a character set. Browsers using different fonts will show you, say, different visual marks when you type an "e" on an American keyboard - but they are all supposed to "look like an 'e'", just cursive, sans-serif etc. Notice the conventions involved here "'e' on an American keyboard", and "what does it mean to look like an 'e'".
Posted by: Grumbly Stu at October 21, 2009 08:52 AMPart 3:
What's been confusing people here, I think, is this way of regarding the matter:
* There is a THING out there, the particular phonetic representation which Bathrobe wants to appear in his (?) comment
Isn't it strange that everyone seems certain they know what particular THING Bathrobe means, although all report that different character sequences appear in their browsers? How could they know what this one THING is, since everyone sees different things? But what are these "different things" that people are seeing? The answer is, a lot of assumptions are being made, i.e. "realities" are being imagined, i.e. constructed. The systematics of Unicode, fonts and character sets are conventions being used in these constructions, and each convention is itself a construction.
There's nothing wrong with assumptions and constructions, of course. The problems start when you believe that your momentary assumptions and constructions are "really real realities", that take you once and for all outsie the self-referentiality that is there when you look closer.
Strangely, none of the fonts whose renderings people have reported here are working as they should. The Unicode number that is being improperly displayed is the second one. According to the University College London IPA transcription in Unicode, Bathrobe's Unicode number corresponds to the visual mark that I would describe as "an 'e' with a tilde over it, not an 'a'".
Posted by: Grumbly Stu at October 21, 2009 08:54 AMand I was reacting only to the word yikes, I don't know why but I have antipathy to the word, one says yikes about something, the same feeling the word invokes? in me
I know it's strange, can't help myself
but i'm glad there is no any strange meaning containing words in English sounding close to ehees, I remember I couldn't use my nickname in Japan, b/c it sounded shrimp in Japanese
are
Posted by: read at October 21, 2009 09:13 AMYour nickname is エビ ?
Posted by: A. J. P. 'tempura' Crow at October 21, 2009 09:16 AMaga, i mean yes, it sounds like abd, but my folks tend to further shorten it to ab only, close to エビ
b/c my name is very long, 11 letters, nobody could pronounce it, so had to shorten it, but the last part of the name which is commonly used by my friends, a very common short name, i could never use there too, b/c it contains the syllable chim which is also in Japanese kinda like a strange sounding word, i did not know that and suggested to call me like that first and my roommate at that time immediately laughed and explained me why
"Unicode number" makes me think of the Unicode Numeric Forms block (PDF).
In official terminology the numbers assigned to each character are their "character codes" and the 〹 or ꯍ entities are "numeric character references".
Posted by: caffeind at October 21, 2009 11:27 AMThanks, Stu, for that extended treatment.
In fact, Bathrobe has a terrible confession to make. Bathrobe could have dazzled us all by inputting the Unicode characters himself using the IPA palette that he downloaded from Moses Hall's site, which makes it a cinch to find and input Unicode characters (the equivalent IPA Character Map for Windows users is here). But Bathrobe didn't do that, did he? Bathrobe was too lazy to think up the IPA and input it himself. Instead he copied and pasted it from Wikipedia (blush).
But Bathrobe is still wondering: presumably the IPA at Wikipedia shows up correctly on most people's browsers. Why on earth would copying and pasting cause it to screw up?
Over to you, Stu!
Posted by: Bathrobe at October 21, 2009 11:31 AMActually, looking at the page Grumbly Stu directed us to, the character e with a tilde over it (i.e. ẽ) is not actually an ẽ at all. It occurs in a list of "Non-spacing diacritics and suprasegmentals", and the non-spacing diacritic referred to, hex no. 0303, is . I think the 'e' is there just to provide an example of in use.
Posted by: Baθroʊb at October 21, 2009 11:42 AMActually, ɑ̃ is encoded as & # 5 9 3 ; & # 7 7 1 ; (I've spaced it out to prevent it from showing as ɑ̃). This encoding appears to be correct, namely:
& # 5 9 3 ; = ɑ
& # 7 7 1 ; = ̃
So why is it showing up differently in other browsers/fonts?
Posted by: Bathrobe at October 21, 2009 11:56 AMBathrobe, that's because a fatal flaw in some font rendering system, like Verdana that is commonly used: old Vernada interprets the combining character as something that goes before the intended character other than the sanctioned standard of positing after the intended character as one would type. Apparently, some but not all of this error has been redressed in Windows, but some browsers (especially Firefox with Pango support) and some OS used some hacks to place that correctly.
Even though I am using a Mac, I personally cannot replicate the rendition error, so I think it's more to do with the font really.
Here is a test: b́ɑ̃z
Posted by: 28481k at October 21, 2009 01:07 PMActually, I successfully resurrect the error, though because I don't have my own hosting, I would not post the results here: basically, it's the new smart fonts that seemed to be suffered from placement error, that it's placed on the next character DejaVu fonts, Charis SIL, all these smart fonts seemed to be unsmart.
Posted by: 28481k at October 21, 2009 01:45 PMWell, I see I didn't get the details quite right. From my Anglophone perspective, the basic principle of Unicode should be, but perhaps isn't quite:
A numbered list of visual marks
Then there are fonts and their rendering systems, and also, apparently and unfortunately, "Unicode combinations for superposition" like Bathrobe identified at the UCL page I linked (I hadn't looked closely). Is that kind of combination something specific to the IPA system, or is it part of Unicode? I.e. is the IPA system a font rendering system based on Unicode, or are there things like "backspace and try to put a tilde over the previous character" in Unicode??
28481k diagnoses a font rendition problem here, which seems plausible.
In any case, I hope the constructivist message came across ... There are more things that we rely on when "perceiving things" than we are usually aware of. (Of course I know nobody wants to talk about this, because it's crazy stuff ...)
Posted by: Grumbly Stu at October 21, 2009 01:49 PMBathrobe: should it be pronounced [valez elize] or [vale elize]?
By "of course", I meant that the s at the end of the first word (the noun) should "of course" be pronounced [z], in your hypothetical Valles-lyses (thus [valez elize]) as well as in the existing Champs-lyses. These would be fixed, frozen expressions naming specific places (les Champs-lyses is a direct adaptation from Latin, which took it from Greek). But you would not normally pronounce [z] (except in very formal speech or reading classical poetry aloud) if you just added an adjective to an ordinary noun, such as in des champs entiers 'whole fields' or les valles alpines 'the Alpine valleys'.
Posted by: marie-lucie at October 21, 2009 02:21 PMMarie-Lucie: I would make the z elision (is that the right word - sliding the end of the first word in the start of the second ?) in both those last examples, so, roughly, des chamszentier and les vallayszalpine.
Does that mean I speak very formally - it would surprise me - or am I just pronouncing them incorrectly - more likely, my French was learned on the hoof, so to speak.
Posted by: Paul at October 21, 2009 05:02 PM"But you would not normally pronounce [z] (except in very formal speech or reading classical poetry aloud) if you just added an adjective to an ordinary noun, such as in des champs entiers 'whole fields' or les valles alpines 'the Alpine valleys'."
I'm surprised to hear you say that, m-l, because both of those liaisons sound perfectly normal to me, not formal in any way. And wouldn't the not pronouncing of the [z] introduce a kind of strange glottal stop? And isn't it precisely to avoid that kind of thing that you hear all these fausses liaisons like "on ne sait-z-o" or "Il a rpondu-z- ceux qui lui demandaient ..."?
Presumably, then, les valles lves would produce the effect that John intended?
Posted by: Bathrobe at October 22, 2009 08:33 AMNo, the effect I intended was three e's in a row, which apparently doesn't work. Ah well, he shrugged americanly.
Posted by: John Cowan at October 23, 2009 03:19 PMPaul: I would make the z elision (is that the right word - sliding the end of the first word in the start of the second ?)
The right word is liaison (literally "linking"); lision means "eliding" (omitting a sound).
in both those last examples, so, roughly, des chamszentier and les vallayszalpine. ... Does that mean I speak very formally - it would surprise me - or am I just pronouncing them incorrectly - more likely, my French was learned on the hoof, so to speak.
To me, you would sound formal. It is not a matter of "correct" or not, as both pronunciations are correct, but one is more formal, the other one more colloquial.
bruessel: both of those liaisons sound perfectly normal to me, not formal in any way.
bruessel, does your name means that you are Belgian? there may be a different standard there.
And wouldn't the not pronouncing of the [z] introduce a kind of strange glottal stop?
No, there is no glottal stop in a standard pronunciation (except perhaps to emphasize the beginning of a word), only a succession of vowels. But perhaps in Belgium or thereabouts French speakers use a glottal stop because of the influence of the neighbouring Germanic languages (Flemish/Dutch, German, Luxemburgish) which do have a glottal stop before a vowel in word-initial position.
And isn't it precisely to avoid that kind of thing that you hear all these fausses liaisons like "on ne sait-z-o" or "Il a rpondu-z- ceux qui lui demandaient ..."?
I don't think the point is to avoid two vowels in succession, since there are many words which contain such a sequence, but this is an instance of hypercorrection, an attempt to sound more educated or formal by introducing a consonant which is only heard in formal speech under conditions of liaison (since the more formal the speech event, the more liaisons are expected).
Unfortunately, which consonant to use depends on knowing how the word is written, hence people who are not very familiar with the written form of the language will introduce the [z] (which is the most common liaison consonant) in places where it does not belong. For instance, there is a [z] liaison Je ne sais o "I don't know where" or "God knows where" (a frozen expression said on its own - not followed by more words) but not in On ne sait | o where there is no liaison, so introducing a [z] there is not justified except as an illiterate analogy with the Je ne sais o.
Bathrobe: Presumably, then, les valles lves would produce the effect that John intended?
Yes. I would say les valleZ leves in reading a text aloud, for example to a class, but les valles | leves in normal speech (actually, I would prefer les hautes valles since lev with the meaning "high" is a little more formal than haut).
Posted by: marie-lucie at October 23, 2009 04:21 PM(My reply to Bathrobe is not technically the right reply here as John's point was not liaison, as he explains just before my post).
Posted by: marie-lucie at October 23, 2009 04:24 PMm-l: Why doesn't On ne sait o exhibit liaison in formal contexts as On ne sait t-o? I know it doesn't, I just don't understand why not, historically speaking.
Most but not all non-rhotic varieties of English have dropped /r/ restored in liaison (not AAVE, though); unsurprisingly, there is also unhistoric intrusive /r/ in pseudo-liaison situations, as in JFK's famous "Cuber" for "Cuba" (before vowels only, of course).
Posted by: John Cowan at October 25, 2009 08:09 PMJC: m-l: Why doesn't On ne sait o exhibit liaison in formal contexts as On ne sait t-o? I know it doesn't, I just don't understand why not, historically speaking.
Je ne sais o is a fixed expression, but On ne sait o isn't. Besides, with a liaison On ne sait o would sound exactly like On ne sait tout "one doesn't know everything" (which would be another archaism). I guess that the liaison in the common phrase Je ne sais o is a leftover from the time when there was liaison everywhere (and ne was more important than pas in negation), but On ne sait o is not a common phrase in Modern French.
Normally, with a sentence complement to the verb savoir there would be no liaison between the verb and the complement, as in both Je sais | o il est "I know where he is" and Je ne sais pas | o il est "I don't know where he is".
Posted by: marie-lucie at February 17, 2010 10:03 PM