Phonetic rendering of names - Explanation
What is IPA?
IPA stands for "International Phonetic Alphabet" (or "Association"), and it was created in an attempt to devise an alphabet that could cover the rich variety of sounds humans use in their speech.
Due to the evolution of spoken language there are few languages whose written form unambiguously shows how to pronounce a word. English is particularly bad in this, for example "tear" is pronounced differently in "tear (teardrop)" and "to tear apart". Alphabets are rarely strictly phonetic, but IPA was designed to be just this: one letter for each sound.
It was first established in 1896 by an international organisation of phoneticians [external link] and is still being revised. The alphabet, with over 100 standardised letters, is mainly based on the latin alphabet, but with many additions, made by adding greek letters, adding diacritics, turning letters upside down, left to right and many other small modifications (hooks, tails, etc).
Probably everybody has come across IPA in dictionaries, where it is used to show the pronunciation, and it is used by all phoneticians and linguists. It is an accepted standard in widespread use with freely usable computer fonts available.
Using IPA for names
So why does MG use a phonetic alphabet when it is supposed to be an easily recognizable visual glyph-language that is independant of pronunciation? While it would be simple to create easily recognizable glyphs of major cities that have famous landmarks, or for famous people that are constantly in the news, this is not possible for places or people that are not famous. And unlike Chinese, where people might have names such as "Universe" or "Cloud" and the capital is called "Northern Capital", names for example in English rarely have such an explicit meaning, or they have lost this meaning over the centuries. Most names in most languages are purely phonetic. An etymologist might know the origin of the name "London", but in everyday life the word and its sound stands for nothing but the city.
Like all glyphs in MG each name will be linked to an explanation that translates the name into other languages. So for example the glyph showing the IPA letters "dɔʏtʃlant" would be linked to an explanation page containing "Deutschland" in German, "Germany" in English", "德国" in simplified Chinese, "Allemagne" in French, etc.
The IPA name always represents the standard pronunciation of the name in the local language, following the unbiased and non-centric approach of the whole language.
It is normally possible to make a good guess at the meaning even if one doesn't know the exact pronunciation of each character. For example "r", "ɹ"ˌ "ɾ"ˌ "ɽ"ˌ "ɼ"ˌ "ʀ" and "ʁ" will all sound more or less similar to the English "r". The same goes for "a", "ɑ"ˌ "æ" or "n", "ɴ"ˌ "ŋ"ˌ "ɲ". For recognizing a name, this is often enough.
Famous names of cities often have slightly different pronunciations in different languages (such as Köln/Cologne and München/Munich in German/English), but because they are well-known it is likely that the local pronuncation is also well-known. And the less known a name is, the more likely it is that the "translated" name is identical to the one in the local language.
Nevertheless it is important to learn some basic IPA for more efficient reading, and for the cases when a text is not linked, such as in printed form.
In the case of writing a text in MG, the input will be made in the local language, and the IPA form is fetched from the database.
Tonal languages
There also exists a method for marking tones in IPA (tonebars), for use in tonal languages such as Chinese but there is no good support in unicode for contour tones (variations of tone levels). For MG we hence devised our own system of marking the tones above the word, making better use of the available glyph space.
As usual, five tone levels are distinguished, and the system is intuitively understandable, though of course nevertheless difficult to pronounce for the untrained speaker.
-
An example for tonal presentation in MG: the capital of China, Beijing, with the two toneboxes identifying the Mandarin tones 3 "dipping" and 1 "high"
-
A nice introduction to tones (with sound samples) can be found here: [external link]
IPA tutorials
There are several good tutorials on IPA freely available on the internet. A particularly useful chart with sound samples can be found at: http://www.ling.hf.ntnu.no/ipa/full/
Problems and Solutions
-
Names can change: sometimes the names of places change (e.g. "St. Petersburg" to "Leningrad" and back).
The database with the names will be updated reflecting the changes, for example adding a new synonym for the new name of the city, pointing to the old name (or vice versa).
-
Ethnic groups: some cities are made up of different ethnic groups, such as the city of Jerusalem, each calling the city by the name in their own language. Or for example Brussel/Bruxelles with its Dutch- and French-speaking population.
In this case the forms in the two (or more) languages will be available and the users/writers can decide which one to use. And the two forms will be synonyms, referring to the same concept, rendered in two different equivalent forms.
-
Dialect: people of one country usually speak with different dialects, sometimes very similar but sometimes very different.
We believe the best solution is to use the accepted "high language", the one which is used by news presenters and which is normally regarded as somewhat neutral. It is normally also the variant that is taught to foreign students learning a language. One may rightfully argue that the city belongs to the people that live in it, so it should have the name that they call it by. But this makes the system even more complicated, and there may be the case that a city has several dialects. The most unambiguous and straightforward solution is to use the standardised "news-reader official language".
-
Biased: IPA is largely based on latin and greek alphabet, so it could be perceived as biased, "eurocentric".
One alternative could have been creating a new unbiased system from scratch, but since IPA is an established international system which all serious dictionaries rely on, and since it is a system that everyone has probably encountered, we adopted it.
Furthermore, in addition to rendering names phonetically, the original form of the names is kept, in their original script. People can learn to associate even an unfamiliar script with the country (e.g. Korean script) or they can read the phonetic transliteration.
MediaGlyphs.org
This page uses Unicode (UTF8)
Last modified:
Mon Aug 18 06:53:36 PDT 2008
First appearance:
Wed Dec 4 18:15:56 GMT 2002