A suggestion from a reader, emailed to me:
I'm not press, but this seemed the only access to the Wikipedes. (Mind you, I am a former newspaper columnist.) As a lexicographer, I admire Wikipedia, and offer one suggestion. Ask your contributors to indicate pronunciation where needed. For instance, I had to look up an actress with the first name of "Cote." Is it [COAT]? [ko-TAY]? [KO-tay]?
Quite a few of our articles on people include the pronunciation, but it's far from universal. Would anyone be interested in a drive to fill out pronunciations? (What groups of Wikipedians are fans of this stuff?) Adding a field in the infoboxes might help too.
- d.
On Jan 18, 2008 6:02 AM, David Gerard dgerard@gmail.com wrote:
A suggestion from a reader, emailed to me:
I'm not press, but this seemed the only access to the Wikipedes. (Mind you, I am a former newspaper columnist.) As a lexicographer, I admire Wikipedia, and offer one suggestion. Ask your contributors to indicate pronunciation where needed. For instance, I had to look up an actress with the first name of "Cote." Is it [COAT]? [ko-TAY]? [KO-tay]?
Quite a few of our articles on people include the pronunciation, but it's far from universal. Would anyone be interested in a drive to fill out pronunciations? (What groups of Wikipedians are fans of this stuff?) Adding a field in the infoboxes might help too.
On a related note, it would be helpful if we, you know, used a pronunciation guide that makes sense to most English-speaking people. The IPA thing is nonsense, and I have never been able to glean more than bits of redundant pronunciation information from it.
Johnleemk
On a related note, it would be helpful if we, you know, used a pronunciation guide that makes sense to most English-speaking people. The IPA thing is nonsense, and I have never been able to glean more than bits of redundant pronunciation information from it.
Agreed. IPA is great for people that know it, so it's worth including, but a simple phonetic spelling should also be provided, or even just a "rhymes with X".
On 18/01/2008, Thomas Dalton thomas.dalton@gmail.com wrote:
On a related note, it would be helpful if we, you know, used a pronunciation guide that makes sense to most English-speaking people. The IPA thing is nonsense, and I have never been able to glean more than bits of redundant pronunciation information from it.
Agreed. IPA is great for people that know it, so it's worth including, but a simple phonetic spelling should also be provided, or even just a "rhymes with X".
But how do you pronounce word X? English pronunciation varies very, very widely (even in tiny England the range of pronunciation of simple words like "rub" and "new" is quite large) At least IPA has consistency, and if I really need to know what those funny squiggles mean I can click on the link in the template to see exactly that.
On 18/01/2008, Tony Sidaway tonysidaway@gmail.com wrote:
On 18/01/2008, Thomas Dalton thomas.dalton@gmail.com wrote:
On a related note, it would be helpful if we, you know, used a pronunciation guide that makes sense to most English-speaking people. The IPA thing is nonsense, and I have never been able to glean more than bits of redundant pronunciation information from it.
Agreed. IPA is great for people that know it, so it's worth including, but a simple phonetic spelling should also be provided, or even just a "rhymes with X".
But how do you pronounce word X? English pronunciation varies very, very widely (even in tiny England the range of pronunciation of simple words like "rub" and "new" is quite large) At least IPA has consistency, and if I really need to know what those funny squiggles mean I can click on the link in the template to see exactly that.
IPA is great for giving precise pronunciations, but simpler methods are great for people that just want a general idea so they can read the article (unpronounceable words can make it hard to read something, even silently - it disrupts the flow). Oh, any less of the "tiny", if you don't mind...
On 18/01/2008, Thomas Dalton thomas.dalton@gmail.com wrote:
IPA is great for giving precise pronunciations, but simpler methods are great for people that just want a general idea so they can read the article (unpronounceable words can make it hard to read something, even silently - it disrupts the flow). Oh, any less of the "tiny", if you don't mind...
Some articles have ogg files with the pronunciation, which I have found rather helpful in some cases.
On Fri, 2008-01-18 at 15:47 +0000, James Farrar wrote:
On 18/01/2008, Thomas Dalton thomas.dalton@gmail.com wrote:
IPA is great for giving precise pronunciations, but simpler methods are great for people that just want a general idea so they can read the article (unpronounceable words can make it hard to read something, even silently - it disrupts the flow). Oh, any less of the "tiny", if you don't mind...
Some articles have ogg files with the pronunciation, which I have found rather helpful in some cases.
Yeah, that seems to be one of best solutions: IPA + audio. I find the rhymes with x solution rather amateurish and encyclopedic to be fair. As soon as I get a decent microphone, which I'll actually order straight after posting this, I'll start recording names for foreign people from German, Austria and Switzerland and possibly do some work on Chinese names (or get a friend to sit down and do it).
Ian [[User:Poeloq]]
Thomas Dalton wrote:
IPA is great for giving precise pronunciations, but simpler methods are great for people that just want a general idea so they can read the article (unpronounceable words can make it hard to read something, even silently - it disrupts the flow).
Mispronouncing a word when you are reading silently doesn't matter. Nobody can hear you. If you insist on having the pronunciation for every word that you read you must be a very slow reader. The Chinese are just as literate despite the fact that Chinese characters have only limited connection to pronunciation. With English names the Chinese characters would work just as well.
Ec
On 18/01/2008, Ray Saintonge saintonge@telus.net wrote:
Thomas Dalton wrote:
IPA is great for giving precise pronunciations, but simpler methods are great for people that just want a general idea so they can read the article (unpronounceable words can make it hard to read something, even silently - it disrupts the flow).
Mispronouncing a word when you are reading silently doesn't matter. Nobody can hear you. If you insist on having the pronunciation for every word that you read you must be a very slow reader.
My usual experience is that people stumble the first time, decide on a pronunciation, and file it away. As you'd expect, this is very common among children who read voraciously and somewhat in advance of their peers - they encounter a word they never hear spoken, so guess the correct pronunciation.
This then causes gales of hilarity in later life when they happen to use the word in conversation for the first time, and discover everyone else has a different pronunciation of it...
(Examplar: is it "web two" or "web two point oh?" It's a term pretty much everyone will encounter in text long before in speech, and I noticed at a conference last year that there was a fairly even divide - and pretty random distribution - between which way people parsed it...)
Andrew Gray wrote:
(Examplar: is it "web two" or "web two point oh?" It's a term pretty much everyone will encounter in text long before in speech, and I noticed at a conference last year that there was a fairly even divide
- and pretty random distribution - between which way people parsed
it...)
I guess it makes me some kind of eccentric to be saying "web two point zero." :-)
Ec
On Jan 18, 2008 8:12 PM, Ray Saintonge saintonge@telus.net wrote:
Andrew Gray wrote:
(Examplar: is it "web two" or "web two point oh?" It's a term pretty much everyone will encounter in text long before in speech, and I noticed at a conference last year that there was a fairly even divide
- and pretty random distribution - between which way people parsed
it...)
I guess it makes me some kind of eccentric to be saying "web two point zero." :-)
Ec
Aww, crap! Now _I'm_ gonna think about it that way!
On 18/01/2008, Ray Saintonge saintonge@telus.net wrote:
Thomas Dalton wrote:
IPA is great for giving precise pronunciations, but simpler methods are great for people that just want a general idea so they can read the article (unpronounceable words can make it hard to read something, even silently - it disrupts the flow).
Mispronouncing a word when you are reading silently doesn't matter. Nobody can hear you. If you insist on having the pronunciation for every word that you read you must be a very slow reader. The Chinese are just as literate despite the fact that Chinese characters have only limited connection to pronunciation. With English names the Chinese characters would work just as well.
It doesn't matter, but it is nice to be close rather than just guessing a random possible pronunciation. There are two ways you can read any given word, it can be a word you're familiar with (it's in your "sight vocabulary") and you just take it in at a glance, or it can be a word you aren't familiar with and you need to "sound it out" (of course, once you're an experienced reader, sounding out is done much quicker than a young child would do it). The kind of words you're likely to not know the pronunciations of are words you aren't familiar with, in which case being able to sound it out is useful. You can just skip over it and just think "oh, that word again", but it breaks your flow and slows reading down. I don't speak Chinese, but I imagine it's similar to English is that respect - most characters will be in your sight vocabulary and it doesn't make any difference how they're written, you just know it as an image. Characters you are less familiar with, you would presumably have to slow down to read. The fact that you can't just "sound out" the word might make it harder to learn new words when reading, I don't really know, perhaps you just get used to it the same as we get used to the way English works.
On 1/18/08, Tony Sidaway tonysidaway@gmail.com wrote:
On 18/01/2008, Thomas Dalton thomas.dalton@gmail.com wrote:
On 1/18/08, John Lee johnleemk@gmail.com wrote:
On a related note, it would be helpful if we, you know, used a pronunciation guide that makes sense to most English-speaking people. The IPA thing is nonsense, and I have never been able to glean more than bits of redundant pronunciation information from it.
Agreed. IPA is great for people that know it, so it's worth including, but a simple phonetic spelling should also be provided, or even just a "rhymes with X".
But how do you pronounce word X? English pronunciation varies very, very widely (even in tiny England the range of pronunciation of simple words like "rub" and "new" is quite large) At least IPA has consistency, and if I really need to know what those funny squiggles mean I can click on the link in the template to see exactly that.
I agree. As long as there exists a system with some degree of widespread acceptance we should build upon it rather than against it.[1]
I noticed one user's tool which apparently adds hovering "help text" to help interpret IPA symbols, e.g. <span title="fOOd">u</span>, which might be worth experimenting with.[2]
On the other hand if the constructed IPA is more consistent than *any* natural language, wouldn't it be relatively easy for a machine to read it, aloud?
—C.W.
[1] This may be why Citizendium is a colossal waste of bandwidth and Veropedia is slightly interesting. :p [2] http://en.wikipedia.org/wiki/User:Nohat/IPA.js
On 18/01/2008, Charlotte Webb charlottethewebb@gmail.com wrote:
On the other hand if the constructed IPA is more consistent than *any* natural language, wouldn't it be relatively easy for a machine to read it, aloud?
Why, why, *why* has this not been suggested before? Marvellous!
(also, an image-generation program to create heraldic crests from descriptions. that'd be much the same sort of concept)
On Jan 18, 2008 6:25 AM, John Lee johnleemk@gmail.com wrote:
On a related note, it would be helpful if we, you know, used a pronunciation guide that makes sense to most English-speaking people. The IPA thing is nonsense, and I have never been able to glean more than bits of redundant pronunciation information from it.
IPA is, technically, automatically convertible to correct speech by software. Phonetic spellings won't be.
None of the existing free software speech syntheizers that I am aware of support IPA as a phonetic input format currently. I spend a couple of hours trying to write a IPA->festival phonetic converter, but my own lack of understand of IP (and phonetic representations) got in the way.
This really should be done, not only would it improve the ease and coverage of getting a pronouncation but it would also reduce the risk of IPA vandalism.
(If you search the archives you should be able to find me talking about this a couple times in the past, ... I can't be bothered right now.)
On 18/01/2008, Gregory Maxwell gmaxwell@gmail.com wrote:
IPA is, technically, automatically convertible to correct speech by software. Phonetic spellings won't be. None of the existing free software speech syntheizers that I am aware of support IPA as a phonetic input format currently. I spend a couple of hours trying to write a IPA->festival phonetic converter, but my own lack of understand of IP (and phonetic representations) got in the way. This really should be done, not only would it improve the ease and coverage of getting a pronouncation but it would also reduce the risk of IPA vandalism.
This is a *brilliant* idea. Please, spread it around!
(I'm just picturing it: "MediaWiki. Dependencies: Festival.")
- d.
On 18/01/2008, David Gerard dgerard@gmail.com wrote:
A suggestion from a reader, emailed to me:
As a lexicographer, I admire Wikipedia, and offer one suggestion. Ask your contributors to indicate pronunciation where needed. For instance, I had to look up an actress with the first name of "Cote." Is it [COAT]? [ko-TAY]? [KO-tay]?
Quite a few of our articles on people include the pronunciation, but it's far from universal. Would anyone be interested in a drive to fill out pronunciations? (What groups of Wikipedians are fans of this stuff?) Adding a field in the infoboxes might help too.
The problem is that Pronunciation Is Hard [TM].
We get this request a lot on OTRS, and the universal response is "we try to, but coverage is patchy". There have been drives in the past, but they've foundered on:
a) not too many people are motivated to do it; b) in many cases, the people who are motivated to do it aren't the ones who also know how it's pronounced; c) we insist on using IPA rather than "pronounced KO-tay".
c) is kind of the sticking point. Very few people are confident and competent with IPA; to most of us, it is at best something that we encounter in dictionaries and try not to think too hard about. This means that our editors aren't likely to add it (because they don't feel comfortable using it, and certainly don't want to spend fifteen minutes with a lookup table and some incomprehensible runes to get a valid word), and that our readers aren't likely to use it even if it's there.
But we don't have much option! Anything else is simply ineffective for a project with such a wide and diffuse base of contributors and users - saying "rhymes with X" or some kind of phonetic spelling is only reliable if we assume that everyone pronounces X, or interprets those syllables, the same way.
[This is also the reason why giving dictionary-style definitions for normal nouns - "tomato" - would rapidly fall over; too much local variation. Names and other proper nouns have a "right" pronunciation regardless of where you are, though, so we can get away with it]
And we just have too many contributors, too wide a readerbase, to be able to say that with any degree of confidence.
On Fri, Jan 18, 2008 at 05:22:13PM +0000, Andrew Gray wrote:
Names and other proper nouns have a "right" pronunciation regardless of where you are, though, so we can get away with it
If that were true, the situation would be much easier. It isn't uncommon for pronunciation of a name to vary significantly from one dialect to another.
Carl
On 18/01/2008, Carl Beckhorn cbeckhorn@fastmail.fm wrote:
On Fri, Jan 18, 2008 at 05:22:13PM +0000, Andrew Gray wrote:
Names and other proper nouns have a "right" pronunciation regardless of where you are, though, so we can get away with it
If that were true, the situation would be much easier. It isn't uncommon for pronunciation of a name to vary significantly from one dialect to another.
Yes, but we can argue that the one they themselves use is the "right" one :-)
You and I might differ on how "Beckhorn" is pronounced, but any third party would say that your interpretation is the one that ought to go after your name.
On Fri, Jan 18, 2008 at 05:47:44PM +0000, Andrew Gray wrote:
You and I might differ on how "Beckhorn" is pronounced, but any third party would say that your interpretation is the one that ought to go after your name.
I would argue that the third party should use the pronunciation of my name that is most common and well-regarded among members of his intended audience. This is especially relevant when the name is originally in a language different than the one being used to speak. Using the original, foreign pronunciation could easily be perceived as an affectation.
Carl
On 1/18/08, Andrew Gray shimgray@gmail.com wrote:
On 18/01/2008, Carl Beckhorn cbeckhorn@fastmail.fm wrote:
On Fri, Jan 18, 2008 at 05:22:13PM +0000, Andrew Gray wrote:
Names and other proper nouns have a "right" pronunciation regardless of where you are, though, so we can get away with it
If that were true, the situation would be much easier. It isn't uncommon for pronunciation of a name to vary significantly from one dialect to another.
Yes, but we can argue that the one they themselves use is the "right" one :-)
"Birmingham"
Andrew Gray wrote:
[This is also the reason why giving dictionary-style definitions for normal nouns - "tomato" - would rapidly fall over; too much local variation. Names and other proper nouns have a "right" pronunciation regardless of where you are, though, so we can get away with it]
Even with local names and proper nouns that's not the case. In North America French names are particularly vulnerable to mangling when they are anglicized. French speakers can only be bemused by the way that Americans pronounce the name of the capital of Iowa.
Ec
On 1/19/08, Andrew Gray shimgray@gmail.com wrote:
c) we insist on using IPA rather than "pronounced KO-tay".
c) is kind of the sticking point. Very few people are confident and competent with IPA; to most of us, it is at best something that we encounter in dictionaries and try not to think too hard about. This means that our editors aren't likely to add it (because they don't feel comfortable using it, and certainly don't want to spend fifteen minutes with a lookup table and some incomprehensible runes to get a valid word), and that our readers aren't likely to use it even if it's there.
But we don't have much option! Anything else is simply ineffective for a project with such a wide and diffuse base of contributors and users
- saying "rhymes with X" or some kind of phonetic spelling is only
reliable if we assume that everyone pronounces X, or interprets those syllables, the same way.
[I have a linguistics degree with a whole subject worth of phonetics, so I am Qualified To Speak On The Subject]
IPA *is* hard. It's a real bitch to write in wikitext, it's not easy to get right, and people don't even want it. I also don't think it's appropriate for expressing the pronunciation of English names/words/places, because they should really be given in a phonemic pronunciation (eg, the pronounciation of "Prahran" should only communicate to the reader that it's puh-RAN, without describing the Australian accent as well). IPA is probably the best tool for foreign words though.
So, for non-IPA contexts, rhyming is good. We Australians are well used to seeing American rhyming schemes given. Sometimes it's odd - I saw a pronunciation for "boffin" given as "BAW-fuhn", but we get the gist. People play up the differences in regional accent far too much - all we're trying to communicate is the rough structure of the name, what syllables are stressed, whether "Reibl" is an "ee" or an "eye" sound.
[This is also the reason why giving dictionary-style definitions for normal nouns - "tomato" - would rapidly fall over; too much local variation. Names and other proper nouns have a "right" pronunciation
Giving a pronunciation for a common English word is a bad idea. Giving a pronunciation for a rarer word, like "gilliflower" is helpful.
In summary: - Only provide pronuncations where it's actually helpful or there is an ambiguity - Use IPA for foreign words, some rhyming scheme for English words
Steve