Morpho-phonology formalizes interactions where morphological structure conditions phonological rules. Affixes may trigger sound changes in roots (English plurals: [ɪz] after sibilants). Phonological rules apply cyclically—first to root+affix, then to progressively larger words. This interface explains allomorphy, apparent exceptions to phonological rules, and preservation of underlying contrasts in morphologically sensitive positions.
You already know that allomorphy is the phenomenon where a single morpheme surfaces with different phonological forms depending on context — the English plural morpheme appears as [s], [z], or [ɪz] depending on the final segment of the root. You also know that phonological rules map underlying representations to surface forms. The morphology-phonology interface asks: when both systems are active at once, how do they interact? The central insight is that morphological structure is not invisible to the phonology — the boundaries between morphemes, and the order in which morphemes combine, can determine which phonological rules apply and in what order.
Consider what happens when you add the suffix *-ity* to *sane*: you get *sanity*, not *sanety*. The underlying /eɪ/ in *sane* shifts to [æ] before the suffix. But add *-ness* to *sane* and nothing changes: *saneness* preserves the [eɪ]. The suffix *-ity* is a class I affix that bleeds into the root and triggers phonological changes; *-ness* is a class II affix that attaches "outside" the phonologically active domain. This asymmetry cannot be explained by phonology alone — the identity of the morpheme determines the phonological behavior. This is the interface problem in miniature.
Cyclic application is the theoretical response. The idea is that phonological rules do not apply once to the completed word form; they apply in successive cycles, one per morphological layer. In the first cycle, the rule applies to the root alone. In the second cycle, it applies to the root-plus-innermost-affix unit. In the third cycle, to the full derived word. Rules that were already applied in an earlier cycle may be "frozen" — their effects persist even if the derived environment would now trigger a different rule. This captures why some phonological alternations are visible in derived forms while others are opaque to further derivation.
The practical payoff is that the interface explains why phonological rules appear to have "exceptions" that are really morphologically conditioned. When a rule applies only when a class I suffix is present, when stress shifts only at certain morphological boundaries, or when a vowel undergoes vowel shift in some derivatives but not others, the cause is not phonological irregularity — it is the structure of morphological constituency. Understanding the interface means you no longer treat these as arbitrary exceptions: they are systematic effects of the layered interaction between the grammar's two combinatorial systems.