Questions: Allomorphy and Morphophonological Processes

Question 1

The English words 'cats,' 'dogs,' and 'buses' end with the sounds /s/, /z/, and /əz/ respectively. From the perspective of morphological analysis, these three sounds are best described as...

• A. Three different plural morphemes, each assigned to different noun classes
• B. Three allomorphs of a single plural morpheme, conditioned by the phonological context of the stem
• C. Free variation — English speakers choose arbitrarily among /s/, /z/, and /əz/
• D. Two distinct morphemes: a regular plural and a sibilant-harmony plural

Answer: (B)

All three are realizations of the same underlying plural morpheme. The distribution is fully predictable: /əz/ follows sibilants (to avoid two adjacent sibilants), /s/ follows other voiceless consonants (voicing assimilation), and /z/ follows voiced sounds and vowels (the default voiced form). Because the allomorph choice is determined by phonological context, these are phonologically conditioned allomorphs of a single morpheme — not separate morphemes or free variants.

Question 2

A student notices that the past tense of 'go' is 'went,' sharing no phonological material with the base form. This is an example of which morphological phenomenon?

• A. Phonologically conditioned allomorphy, because the verb class determines the ending
• B. Regular morphological alternation triggered by a following vowel
• C. Suppletive allomorphy, where the allomorphs share no phonological relationship and must be stored in the lexicon
• D. Zero morphology, because no suffix is added to create the past tense

Answer: (C)

Suppletion is the term for paradigm alternations where forms share no phonological resemblance — the relationship is purely lexical and must be memorized. 'Went' shares nothing phonological with 'go'; compare this with phonologically conditioned allomorphy (like plural /s/ vs. /z/), where the rule is derivable from the phonological context. High-frequency forms in natural languages frequently develop suppletive paradigms because they get memorized as whole-form entries rather than computed rule-by-rule.

Question 3

Allomorphy is phonological disorder — different forms of the same morpheme vary unpredictably depending on dialect or speaker preference.

• T. True
• F. False

Answer: False

Allomorphy is systematic, not disordered. Phonologically conditioned allomorphy is fully predictable from the phonological environment and applies consistently across all speakers of a language. Even morphologically or lexically conditioned allomorphy (like English strong verbs) follows class-level patterns — the sing/sang/sung pattern applies to a whole set of verbs (ring/rang/rung, swim/swam/swum). Only suppletive forms are idiosyncratic, but even these are uniformly shared across speakers.

Question 4

The allomorph selected for a phonologically conditioned morpheme is determined by properties of the adjacent sounds, not by arbitrary lexical indexing.

• T. True
• F. False

Answer: True

Phonological conditioning means the allomorph distribution follows a sound-based rule: /s/ vs. /z/ in English plurals is determined by the voicing of the stem's final consonant, and /əz/ is triggered by sibilant consonants. These rules operate on phonological features (voiced/voiceless, sibilant/non-sibilant), not on memorized word-by-word lists. This is what distinguishes phonological conditioning from lexical conditioning, where the choice must be remembered per-lexeme (as with irregular plurals like 'ox/oxen').

Question 5

Why must suppletive allomorphs be stored individually in the lexicon rather than derived by phonological rule, and what does this reveal about the organization of linguistic knowledge?

Model answer: Suppletive allomorphs (e.g., go/went, good/better) share no phonological material, so no rule could derive one from the other. A phonological rule maps one sound sequence to another predictably; suppletion has no predictable input-output mapping. Linguists conclude these forms are stored as separate lexical entries linked by paradigmatic relationships, not computed at the time of use. This reveals that linguistic knowledge includes both rule-governed productivity (applied online) and stored exceptions (indexed to specific items).

The concentration of suppletion in high-frequency items is a cross-linguistic universal: frequent forms get memorized directly because retrieval is more efficient than rule application for items accessed thousands of times. The distinction between 'rule' and 'list' is central to generative morphological theory and reflects the dual architecture of language processing.