Transcription
- Kinds of transcription: narrow vs. broad
- Suggestions/tricks for making better transcriptions
- Common mistakes
- Vowels in unstressed syllables
Kinds of transcriptions
There is no such thing as the transcription of a word. Strictly
speaking, you can only transcribe how, for example, Kevin Russell uttered
that word cat at 12:58:03 pm on January 15, 1997. You can
transcribe that utterance as exactly as possible, within the limits of
your hearing and the conventions provided by the IPA.
If you want to go beyond that, to try to describe how Kevin Russell
pronounces the word in general, or further still to how English speakers
pronounce it in general, then you have to start making abstractions -- you
have to decide which details to include and which details to ignore.
It's common to distinguish between two kinds of transcription, based on
how many details the transcribers decide to ignore:
- Narrow transcription: captures as many aspects of a specific
pronunciation as possible and ignores as few details as possible. Using
the diacritics provided in the IPA, it is possible to make very subtle
distinctions between sounds.
- Broad transciption (or phonemic transcription): ignores as many
details as possible, capturing only enough aspects of a pronunciation to
show how that word differs from other words in the language.
For example, consider the difference between the vowels in [liv] and
[lv].
- For English, a narrow transcription would note the difference between
the [i] and [], and
possibly more subtle details as well. A broad transcription would also
need to note the difference, because the two words mean different things.
- For Canadian French, a narrow transcription would note the
difference. Both [i] and [] occur in the language, but they never contrast, that
is, they cause a difference in meaning, so a broad transcription could
ignore the difference and write both as [liv]. [Footnote]
With the symbols we have so far, we are able to make broad transcriptions
of Canadian English.
Footnote: Well, if there's no
difference in meaning between the two, then why not write it [lv]?
This would in fact be perfectly legitimate. But one of the unspoken
principles of broad transcription is that, when you're given a choice
between two symbols and when all other considerations are equal (sometimes
even when they aren't), you'll pick the one that's easier to type.
Suggestions/tricks for improving transcriptions
Here are a few suggestions which some people
have found helpful for getting more accurate broad transcriptions
of English.
-
Pretend you're someone who can't spell (e.g., an advertising
executive). How would you misspell this word if you wanted to
deliberately misspell it? For example, you might misspell
knight as night or nite -- which gives you
some clues about which sounds are really there are which aren't.
-
Compare the word to other words whose transcriptions you're
more certain of. If two words are homonyms, their transcriptions
should be identical. If two words rhyme, their transcriptions
should end the same way.
-
Decide how many sounds the word has and what the sounds
sound like before you worry about which symbols to use for the sounds.
-
We read left to right -- there's no law you have to write
that way. Don't feel you have to get the symbol for one sound
perfect before you move to the next. If you know the first
consonant and the last consonant but aren't sure of the vowel in
the middle, get the consonants down on paper and worry about the
vowel later.
-
When you're unsure of an individual sound, consider other
words where that sound occurs. For example, if you're not sure
what symbol to use for a vowel, what word would you get if you
put that vowel between h_d or b_t?
-
Read your transcription back out loud. Make sure it says
what you think it says.
-
When you read your transcription back out loud, pretend
you're a very stupid computer who can't do anything more than
play little sound clips one after the other. If the only thing
that sounds strange about your reading is the pauses ([d --
--
g]), then
your transcription is probably right. If it sounds like the
computer is trying to say a different word or a nonsense word ([d
-- o -- g]), you'd better try again.
-
Pay attention to what you're doing with your body. Often
phoneticians who are trying to transcribe an unfamiliar sound
will imitate the sound as closely as they can and then choose the
symbol more on the basis of what they're doing with their vocal
tract during the imitation than on what it sounds like. If it
doesn't feel like an [n], if your tongue body is
touching your soft palate rather than your tongue tip touching
just behind your teeth, then it's not an [n].
-
Practise. Practise. Practise.
- Do the exercises in the textbook.
- Get more textbooks. Do their exercises too.
- Do the exercises on the web page.
- Transcribe words in your head while waiting for the bus.
- Write your grocery lists in IPA.
- Write your diary in IPA.
- Read stories in IPA
whenever you have insomnia.
- ...
Some common mistakes
Square brackets
Always use square brackets around your transcriptions in
order to distinguish them from ordinary text.
Silent letters
There's no such thing as an unpronounced symbol in a
phonetic transcription.
[c]
The IPA symbol [c] represents the sound you make with your
tongue body hitting your hard palate. English doesn't use this
sound. If you're ever tempted to use [c] in transcribing the
speech of a normal English speaker, you're almost certainly
wrong.
ng, nk
The sound usually spelled ng has the symbol []. You
should not add a [g] unless there is actually a [g] pronounced --
does it sound more like singer or like finger?
Clusters spelled nk (and often nc) also usually
have this sound: [k].
j
Remember: y is [j] and j is [d].
The A's
Be careful not to confuse "typewritten" [a] and "script"
[].
These sounds are made differently. In some languages, switching
one for the other can change the meaning of a word.
Capital letters
Don't use capital letters where English spelling conventions
do. The waiter Bill and the bill he brings you are
pronounced identically and must have the same transcription. In
IPA, smaller versions of the capital letters are often used for
completely different sounds. Calling someone [BIl] is generally
considered rude.
Unstressed vowels
If an unstressed syllable has a neutral vowel, don't
transcribe it with a full vowel because it's spelled that way.
See the next section.
Next: Vowels in stressless syllables
Previous: Stress, diphthongs
Up: Table of contents