Chapter 6
1. Minimal pairs. Sample answers:
Initial
Medial
Final
a. /k/—/g/
cold/gold
bicker/bigger
tuck/tug
b. /m/—/n/
mice/nice
simmer/sinner
sum/sun
c. /l/—/r/
lake/rake
cold/cord
feel/fear
d. /b/—/v/
ban/van
saber/saver
dub/dove (near minimal pair)
e. /b/—/m/
ban/man
clabber/clamor
rub/rum
f. /p/—/f/
pail/fail
supper/suffer
leap/leaf
g. /s/—/ʃ/
sell/shell
masses/mashes
lease/leash
h. /ʧ/—/ʤ/
chin/gin
etches/edges
rich/ridge
i. /s/—/z/
sip/zip
fussy/fuzzy
mace/maze
2. Rules relating spelling to pronunciation. In the patient’s system of spelling–to-pronunciation, the
following are true:
In reading:
a corresponds to /a/ or /æ/
3. Complementary Distribution and “A Case of Identity.” Sherlock Homes solved the mystery of Miss
Mary Sutherland’s missing fiancé by noting that Mr. Hosmer Angel (the missing fiancé) and Mr. James
Windibank were in complementary distribution, i.e. they never appeared in the same place at the same
4. Korean.
Part One
<LL>a. [r] and [l] are allophones of one phoneme.
b. No, they do not occur in any minimal pairs.
Part Two
a. [s] and [ʃ] are allophones of the same phoneme. They are in complementary distribution: [ʃ] appears
before [i] and [s] appears before all other vowels or word finally.
b. This distribution can be written as the following phonemic rule:
8. English [aɪ] ~ [ʌɪ].
a. The final sounds in column A are [-voice] and those in column B are [+voice].
b. The words in column C end in vowels.
9. English palatalization. Palatalization occurs whenever /t/, /d/, /s/, and /z/ (alveolar obstruents) are fol-
lowed by the palatal glide /j/.
10. Japanese [t] ~ [ʧ]~ [ʦ].
a. Yes, these sounds, [t], [ʧ], and [ʦ], are in complementary distribution as they never occur in the same
environments.
d.
tatami
/tatami/
ʦukue
/tukue/
ʦuʦumu
/tutumu/
tomodaʧi
/tomodati/
teʦudau
/tetudau/
ʧizu
/tizu/
uʧi
/uti/
ʃita
/ʃita/
kata
/kata/
tegami
/tegami/
ato
/ato/
koto
/koto/
totemo
/totemo/
maʦu
/matu/
tatemono
/tatemono/
otoko
/otoko/
deguʧi
/deguti/
te
/te/
ʧiʧi
/titi/
naʦu
/natu/
ʦuri
/turi/
11. Paku.
i. Yes, stress is predictable. It falls on the penultimate (next to last) syllable.
12. English stress.
a. The following are essentially phonemic transcriptions, except for [ə], the symbol for all unstressed
vowels.
A
B
C
/əstanɪʃ/
/kəlæps/
/əmez/
/ɛgzət/
/ɛgzɪst/
/ɪmpruv/
/ɪmæʤən/
/rəzɛnt/
/sərpraɪz/
/kænsəl/
/rəvolt/
/kəmbaɪn/
/əlɪsət/
/ədapt/
/bəliv/
/præktəs/
/ɪnsɪst/
/əton/
b. The final syllable of the verb is stressed if it ends with a consonant cluster; otherwise, the stress falls
13. English phonotactics.
Word
Possible
Not
Possible
Reason
a.
[pʰril]
x
b.
[skriʧ]
screech
c.
[kʰno]
x
English phonotactic rules prohibit an initial cluster
composed of two stops.
d.
[maɪ]
my
e.
[gnostɪk]
x
English phonotactic rules prohibit an initial cluster
composed of two stops.
f.
[jũnəkhɔrn]
unicorn
g.
[fruit]
x
A glide always occurs between front and back high
vowels in English.
h.
[blaft]
x
i.
[ŋar]
x
English phonotactic rules do not permit the velar nasal
to occur word-initially.
j.
[æpəpʰlɛksi]
apoplexy
14. Hebrew.
a. [b] and [v] are allophones of one phoneme and are in complementary distribution.
[b] occurs word-initially and after consonants while [v] occurs only after vowels.
15. Maninka.
a. (1) –li
(2) –ni
b. Yes, the phonetic variants are predictable. The form is -ni if the last consonant of the stem is a nasal
and -li otherwise. Notice that the last consonant of the stem does not have to be the last segment of the
stem for the nasalized variant to appear.
16. Luganda.
a. No, nasal vowels are not phonemic in Luganda. Yes, they are predictable since they only occur before
nasal consonants.
similar environments.
17. Japanese morphophonemics. (Cf. exercise 10.)
a.
“call”
/yob/
“write”
/kak/
“eat”
/tabe/
“see”
/mi/
“leave”
/de/
“go out”
/dekake/
“die”
/sin/
“close”
/sime/
“swindle”
/kata/
“wear”
/ki/
“read”
/yom/
“lend”
/kas/
“wait”
/mat/
“press”
/os/
“apply”
/ate/
“drop”
/otos/
“have”
/mot/
“win”
/kat/
“steal a lover”
/neto/
18. Ojibwa.
a. The morpheme “I” has the allomorphs [nit] and [ni].
The morpheme “you” has the allomorphs [kiʃ] and [ki].
b. Yes, the allomorphs for “I” are in complementary distribution. Their distribution is predictable; [nit]
19. Myanmar nasals.
There is a four-way contrast between the phones [m], [m
̥], [n], and [n
̥], clearly exemplified by the fol-
lowing minimal set:
[ma] “health”
20. Wakanti.
a. The phonemic form of the negative morpheme is /n/.
b. The allomorphs are [n], [m], [ŋ].
c. The nasal consonant [n] assimilates in place of articulation to the following consonant. For example,
[aɪæ̃mb + ɪc] “iambic”
/anweli/
“I didn’t climb up”
/inoa/
“You didn’t fall”
/anie/
“I don’t hunt”
/unlamaba/
“We don’t put on top”
21. French
a. The two forms for “small” are [pəti] and [pətit]. The two forms for “our” are [no] and [noz].
b. [pəti] occurs before words that begin with: [t], [l], [n]
[no] occurs before words that begin with: [t], [l], [n]
22. English /b/ deletion.
a. The two allomorphs of each root morpheme are:
[bm] ~ [bmb]
[kʰrʌ̃m] ~ [kʰrʌ̃mb]
[aɪæ̃m] ~ [aɪæ̃mb]
23. Hebrew metathesis.
24. Japanese vowel devoicing.
a. [u] and [i] may occur voiceless.
b. Yes, the voiceless vowels are in complementary distribution. The voiceless vowels occur between
26. German. Both German and English obey the constraint: Obstruent sequences may not differ with respect
27. Speech errors.
1. The final consonants of gone and seed are swapped. Normally, the vowel in gone would be nasalized
2. The ti in stick and the mu in mud were swapped. The [t] in stick is not aspirated, because it occurs af-
3. This example is a little more complicated. It looks like first the entire CC onset [sp] in speech
switched with the [pr] in production: resulting in preach and speduction. Then, perhaps because the re-