Consider the distribution of [r] and [l] in Korean in the following words:
rupi
"ruby"
mul
"water"
kiri
"road"
pal
"big"
saram
"person"
sul
"Seoul"
ratio
"radio"
ilkop
"seven"
ipalsa
"barber"
Are [r] and [l] separate phonemes or allophones of the same phoneme? If you conclude that they are allophones of the one phoneme, state the rule that explains their distribution. They are allophones of the same phoneme phoneme (because there are no minimal pairs). In this case, their distribution is determined by location. The sound [r] occurs only at the beginnings of syllables, and the sound [l] occurs only at the ends of syllables. So, the rule would be either /r/
[l] when syllable-final
or /l/ [r] when syllable-initial. Either response would be acceptable, given the data you have.
Here are some Japanese words in phonetic transcription. Based on t hese data, are [t], [tš], and [ts] in complementary distribution? Which are allophones and which are phonemes? (NOTE: Treat [tš] and [ts] as if th ey were a single sound - which they are as palatal and alveolar affricates, affri cates, respectively.)
tatami
"mat"
tomodatši
"friend"
utši
"house"
tegami
"letter"
totemo
"very"
otoko
"male"
tšitši
"father"
tsukue
"desk"
tetsudau
"help"
šita
"under"
ato
"later"
matsu
"wait" "wait"
natsu
"summer"
tsutsumu
"wrap"
tšizu
"map"
kata
"person"
tatemono
"building"
te
"hand"
The sounds are in complementary distribution, and they are all allophones of the same phoneme (because there there are no minimal minimal pairs). The sound [t] only only occurs before [a], [e], and [o]. The sound [ts] occurs occurs only before [u]. The sound [tš]
occurs only before [i]. So, the best rule to describe the distribution of these sounds would be /t/
[ts] before [u]
and /t/ [tš] before [i]. If you wanted to simplify the rule even more, you could say /t/ an affricate when before a high vowel.
Consider these phonetic forms of Hebrew words. Are [b] and [v] allophones of one phoneme? Can you formulate a rule to explain their distribution? Does the same rule, or lack of a rule, apply to [p] and [f]? If not, why not? (NOTE: [?] represents a glottal stop, and [x] represents a pharyngeal fricative.)
bika
"lamented"
litef
"stroked"
mugbal
"limited"
sefer
"book"
šavar
"broke" (masc.)
sataf
"washed"
šavra
"broke" (fem.)
para
"cow"
?ikev
"delayed"
mitpaxat
"handkerchief"
bara
"created"
ha?alpim
"the Alps"
The sounds [b] and [v] are allophones (because there are no minimal pairs). As in the first example above, the key here is location in the syllable. The sound [b] only occurs at the beginning of a syllable. But unlike the earlier example, [v] does not necessarily always occur syllable-final. In a word like [šavar], the [v] isn't exactly syllable-initial or syllable-final. So, in this case, [b] is always syllable-initial, and [v] is never syllable-initial, but [v] occurs everywhere else. The rule would then have to be /v/
[b] when syllable-initial.
It couldn't be put the other way around (since [v] isn't only syllable-final). As for [p] and [f], the same rule applies: /f/
[p] when syllable-initial.
Phonemes, Allophones and Phones Introduction In the last session on auditory phonetics and speech perception we talked about categorical perception and the fact that what your ear picks up and what your brain makes of it may be two different things - your ear may perceive signals from a continuum and your brain sorts these signals into discrete categories without you noticing. In this session we will explore this issue in more detail. Pl ease consider the table below: Audio Input
What Your Brain Makes Of It
[aha]
/aha/
what comes out of your head
what is in your head
concrete
abstract
phones
phon emes
phonetics
phonology
Reading For background reading try Crystal (1987), pp. 160. You can also have a look at Fromkin/Rodman (1998) and their section on phonology. Their general introduction is easy to read but mind that they don't always use IPA symbols for their transcriptions and that they mainly use examples in American English. In the exam at the end of term you will have to use IPA symbols and stick to either British English/Received Pronunciation or American English/GenAm. In Clark/Yallop (1995) you will find information on
phonemes in section 4.2, pp. 91-94 allophones in section 4.3, pp. 94-99 free variation in section 4.11, pp. 117-119
Terms Term
Definition
Example
Phoneme
a fundamental unit in phonological structure For English: speech sounds that contrast or distinguish words /b/, /p/, ... abstract mental representations of the phonological units of language the smallest meaning-distinguishing unit in speech
Minimal pair
beat /bi:t/ vs. feet two words which have the same number and order of sound segments except for one segment /fi:t/ and which have a differ in meaning makes a
difference in meaning
Minimal set
a number of words which have the same number beat /bi:t/ vs. feet and order of segments except for one sound /fi:t/ vs. seat /si:t/ segment that occurs in the same place in the vs. meet /mi:t/ etc. string and makes a difference in meaning
Allophone in a wider sense
any realisation of a phoneme synonym to phone
Allophone in a narrow sense
predictable phonetic variant of a phoneme the choice of the allophone in a narrow sense is the dark and light rule-governed /l/ in English phonetically similar to each other, i.e. must share most of the phonetic features
Phone
any single phonetic segment irrespective of what its phonological status may be
Free variation
when two or more phones or phonemes can occur economics with in the same position in a word without changing initial schwa or /e/ the meaning or /i:/ butter with medial /t/ or a glottal stop instead
Complementary distribution
allophones which never occur in the same context, i.e. in the one context one allophone occurs and in another the other one occurs (cf. Fromkin/Rodman's example of Superman and Clark Kent)
the slightly different /k/ and /ŋ/ in the word Kingkong
dark and light /l/ in English oral and nasal vowels in English
Phonemic and phonetic transcripts When you transcribe speech you can have both a phonemic and a phonetic transcript. Phonemic transcripts only take into account those segments which are necessary to distinguish the meaning of the word or utterance you describe. That is, phonemic transcripts work with phonemes. Another term for this kind of transcript is broad transcription. Phonemic transcripts are always given in slashes: /.../. Phonetic transcription includes phonetic information, too, and gives phonetic details which are not included in the phonemic transcript. Phonetic transcriptions uses for example allophones and, depending on how much detail you want, any other phonetic detail of the sound segment. Another term for the phonetic transcript in narrow transcript . Phonetic transcripts are always given in square brackets: [...]. In order to do a phonemic transcript of speech you need the phonemic inventory of the language. For a phonetic transcript, however, you do not need a list of the phonemes, in fact, you don't even have to understand the language you are meant to describe - you "just" note down what you hear (which is hard enough ...).
For precise definitions of phonemic and phonetic transcription see Trask ( 1996).
Phonological Analysis In order to set up the phonemic inventory of a language, i.e. a list of all the meaningdistinguishing vowels and consonants, you need to do a phonological analysis of the language. In the structuralist tradition, a phonological analysis consists of two steps: segmentation and classification. As a prerequisite, you ideally want all the words of the language transcribed in as much detail as possible (which is impossible ...). In step 1 you segment your sequences of sounds into minimal units. You just use your ears to establish these phones. You will give all the phones in square brackets []. Step 2 can be summarised in the command "Look for minimal pairs!". This means that you try to find all the words that only differ in one speech sound or phone and have a different meaning. With the help of these minimal pairs you can then you try and classify the phones that you found in step 1 as
seperate phonemes, i.e. one sound is different and the meaning of the word is different
If you cannot find minimal pairs, you will have to look for the distribution of your phones. Phones can be allophones of phonemes and can come as
free variations, i.e. the sounds are different but the meaning stays the same allophones in complementary distribution, o i.e. you can find two different sounds but no minimal pair in which the y both occur and in one context you will always have the one phone and in another you o will always have the other and the phones are similar with respect to their articulation o
In a later step you can also find the rules which determine which allphone is used in which context. For more information on phonological rules you can try Fromkin/Rodman (1998) pp. 300. In class, we will get back to phonological rules in two weeks time.
Exercises 1. Decide whether the following pairs of words are minimal pairs or not and give reasons! o Oma : Opa Rand : Rat o o Rad : Rat bitten : bieten o
Rosen (pronounced with an alveolar trill) : Rosen (pronounced with an uvular trill) Buch : Bücher o o dir : Tier Rasen : rasen o Sache : Sachen o o Milch : mild blau : Bau o o Weg : Steg chunk : hunk o 2. Transcribe the word Kingkong phonemically and phonetically, giving as much detail as you can for the phonetic transcript. 3. In English, are [h] and [ŋ] phonemes or allophones of one phoneme? Find arguments for and against each classification by using o minimal pairs o contexts in which they occur (complementary distribution?) phonetic similarity o o
Make a decision and give your reasons! Answer: In English, you will not find a minimal pair for the phonemes /h/ and /ŋ/. Also, you will find that /h/ never occurs in word-final position and /ŋ/ never in word-initial position. Still, they are phonetically different in terms of their articulation: /h/ is a voiceless glottal fricative and /ŋ/ is a voiced velar nasal. Thus, they do not fulfill the phonetic similarity criterion which is required for allophones and therefore they have to be considered phonemes. 4. When you click on this link , a pdf document with some exercises on complementary distribution of allophones in various languages will open. We will go through the example with Greek together in class. After that please work together in groups of at least three people. Choose one example, discuss it and answer the questions. Click here to get to the answer sheet.
Homework 1. Bring lists giving the phonemic inventories of English (British English/RP or American English/GenAm) and German. In addition, find lists of the respective allophones which occur in complementary distribution. You will need these lists to do phonetic and phonemic transcripts.