Recording Speech. Notation of Sounds and Phonemes
Recording Speech. Notation of Sounds and Phonemes
Recording Speech. Notation of Sounds and Phonemes
Vowels
1
For practical purposes students are required to know and use phonemic (or broad) transcription.
This marks only the relevant features of sounds (i.e. it is used for rendering only the phonemes of a
language) and is especially suitable for learning the correct pronunciation. This is the transcription
most pronunciation charts (e.g., in dictionaries) are based on. It does not normally make use of
additional symbols (diacritics), except for the ones mentioned in this section. By contrast, phonetic
(or narrow) transcription, marks all features of soundsrelevant, irrelevant and those resulting from
the acoustic environmentand includes a wide range of diacritical signs. This course will occasionally
rely on phonetic transcription (e.g., in the context of studying allophones).
The chart provided here has been adapted after the Oxford Students Dictionary.
1
Consonants
d did [dId]
z zoo [zu:] N sing [sIN]
k kit [kIt]
S she [SI:]
Z vision ["vIZ@n] l leg [leg]
Semivowels
j yes [jes]
w wet [wet]
Standard British English (SBE), the symbol [a] is used when rendering diphthongs
and triphthongs only. The reason for this consists in the fact that the vowel rep-
resented by this symbol is phonetically distinguished from the two aforementioned
monophthongs and it also embodies qualities different from those signalled by the
cardinal vowel /a/ (historically, the nucleus of these compound vowel sounds has
undergone a process of vowel harmony/vowel attraction). Thus, words like cut,
farm, time or town should be transcribed, respectively, [k2t], [fA:m], [taIm] and
[taUn] (such transcriptions as [kat], [fam], [t2Im] and [t2Un] are incorrect.
a) [O] vs.[o] The sounds rendered by these symbols are phonetically different.
Present day SBE uses the former. The latter is a peculiarity of some regional
and age accents, as well as of General American (GenAm). For transcriptions
of SBE you should use the first symbol, e.g., hot [hOt].
b) The sounds rendered by these symbols are phonetically different.
[O:] vs. [o:]
For some time the tendency in present day SBE has been to shift towards the
latter. However, there isnt yet common agreement between specialists as to
which prevails. For now, you can reliably use the former, e.g., sport [spO:t].
c) The two symbols are equivalent and render the same vowel phonemes.
[U] vs. [u]
2
for the former. Use whichever you want, but use them consistently (i.e. do
not alternate them in one and the same transcription), e.g., put [pUt] or [put].
d) [E@] vs. [e@] The sounds these symbols render are phonetically different (be-
cause the nuclei of these diphthongs are distinct). Nevertheless, some pro-
nunciation charts fail to consider this, most of them preferring the latter for
its simplicity. In addition to this, there is a tendency to realize the respective
diphthong as an allophone whose nucleus coincides loosely with the monoph-
thong (vowel No. 3 [e]). This is not, however a generalised phenomenon. The
ambiguity is enhanced by the fact that the nucleus of the allophone [E@] does
not fully coincide with the qualities suggested by the cardinal vowel. Until
agreement is reached, you can reliably use the symbol [E@] for transcriptions
of SBE.
e) [@] vs. [@:] These two symbols should never be confused, as they represent
3. Diphthongs and triphthongs are long composite vowel sounds, but in English their
component vowels are short. As a result, it is useless and inappropriate to mark
any of their constituents as long. Thus, [E@:] or [a:U] are incorrect transcriptions
(the correct ones being [E@], respectively, [aU]).
4. Special care should be taken in using the symbol [ j]. This stands for a semivowel
which is restricted in distribution to initial and prevocalic positions in a syllable,
e.g., you [ ju:] or tune [tju:n]. Therefore it cannot be used for the glide [i] in
diphthongs and triphthongs. Such renderings are incorrect: [a j] [ej] [a j@]. The
correct symbols in these cases are [aI] [eI] [aU]. Note that in English, unlike in
Romanian, diphthongs and triphthongs do not contain semivowels! For the same
reason, the semivowel [w] cannot be used as the nucleus of the diphthong [U@] or
as a glide in [aU] and [aI@]. Therefore, such transcriptions as town [tawn] or flower
["flaw@] are incorrect.
5. Some non-native speakers (esp. the German and Hungarian) should be attentive
to the pronunciation and transcription of words containing the sounds [v] and [w].
They are different phonemes, the former being a labio-dental voiced fricative (a
consonant), while the latter is a bilabio-velar semivowel (a gliding voiced sound).
6. Foreign speakers should make sure they do not confuse the following voiceless
voiced consonantal pairs: [ S ] [Z], [tS ] [dZ], [T] [D]. Refer to their descriptions in
the table of English consonants and consult a dictionary if you have doubts about
using them in words.
7. Do not confuse [N] and [n]. They are both nasals, but the former is velar, the latter
is alveolar. Consider the peculiar contexts for using the velar consonanti.e. only
3
in the vecinity of another velar sound.
2. The slashes / and double slashes // may be used within a transcribed sequence
delimited by brackets in order to signal interruptions (shorter or longer, single
slash or double slash, respectively). However, to simplify the transcription and
avoid confusions, you may omit them.
3. Other diacritics that may be allowed or required are the parentheses (round brack-
ets) ( and ) and the stress (accent) marks " and . The parentheses signal sounds
that are optional in pronunciation, such as a schwa [@] that may be omitted, e.g.,
suddenly ["s2d(@)nli], or an additional yod [j], as in tune [t( j)u:n] or new [n( j)u:]
(the yod is usually not pronounced in such positions in GenAm). The two stress
marks differ in that the former signals a primary (strong) stress, whereas the lat-
ter is used for a secondary (weak) stress, which may be the case with some longer
words, usually derivatives, e.g., metaphorical [met@"fOrik(@)l].
4. It should be kept in view that all symbols used in transcription are lower case. No
sound or word (be it a proper name) is rendered by upper case/capital symbols.
5. Unlike with transcriptions of some languages (e.g., French), the asterisk * is not
used in transcriptions of texts spoken in English. An asterisk would signal a
proper name transcribed in a sequence, but, since transcription renders speech,
not orthography, it is unjustified. Consequently, do not use it.
6. Do not use punctuation marks required by orthography. The full stop, colon,
semicolon, quotation marks etc. are inappropriate when transcribing speech, as
transcription is not a means of signalling orthographical conventions (see No. 5
above).