A phonologist wants to write a single rule that applies to all nasal consonants (/m/, /n/, /ŋ/) but no other sounds. What property of feature theory makes this possible?
AEach phoneme is treated as an unanalyzable atom, so the rule can simply list the three phonemes
BThe sounds /m/, /n/, /ŋ/ form a natural class defined by sharing the feature [+nasal], allowing one rule to reference the whole group
CPhonological rules always apply to exactly three sounds at a time
DFeature geometry predicts that only three nasals can exist in any language
Natural classes are the key motivation for feature theory. Instead of listing /m/, /n/, /ŋ/ individually (which would be stipulative and miss the generalization), the feature [+nasal] picks out exactly these sounds as a class. A single rule referencing [+nasal] then applies to all of them — and will automatically extend to any additional nasal phonemes a language might have. This is a genuine empirical prediction, not just shorthand.
Question 2 True / False
A binary feature like [+voice] means that sounds are either mostly voiced or substantially voiceless, with no gradient realization possible.
TTrue
FFalse
Answer: False
This confuses the phonological level with the phonetic level. At the phonological level, [+/-voice] is a categorical, discrete distinction used to contrast phonemes (e.g., /p/ vs /b/ in English). At the phonetic level, voicing is a continuous acoustic and articulatory property — the same speaker may produce /b/ with varying degrees of voicing depending on position in a word, speaking rate, and surrounding sounds. Features capture categorical phonological contrasts; phonetics captures gradient realization.
Question 3 Short Answer
Why do phonologists analyze phonemes into sub-segmental features rather than treating each phoneme as an unanalyzable primitive?
Think about your answer, then reveal below.
Model answer: Features allow phonologists to capture generalizations about which sounds pattern together in phonological rules. If phonemes were atoms, rules that apply to natural classes (all nasals, all voiceless stops) would require listing each member individually and would miss the fact that membership is predicted by shared properties. Features make those shared properties explicit and allow a single rule to cover the whole class, making the theory both more concise and more explanatory.
The deeper point is predictive power: feature theory predicts that only certain groupings of sounds should behave uniformly in phonological rules — those defined by a shared feature value. If we observe that a rule applies to /p/, /t/, /k/ (all voiceless stops) but not /b/, /d/, /g/, feature theory explains this with [+consonantal, -voice, -continuant]. Treating phonemes as atoms would make it coincidental rather than principled.