Syllable structure divides segments into onset, nucleus, and coda; phonotactic constraints determine which consonant clusters are licit word-initially or word-finally—constraints that vary widely across languages. Optimality Theory ranks universal constraints (ONSET, NOCODA, consonant hierarchy) differently to generate cross-linguistic syllable patterns.
Catalog syllable structures and consonant clusters across diverse languages; rank OT constraints to predict which structures appear in each language.
Syllables are not mere phonetic groupings but phonologically active units that anchor features and stress; syllabification itself is linguistically relevant, not pre-determined by phonetics.
A syllable is the basic rhythmic unit of spoken language, and its internal structure is not arbitrary. Every syllable has a nucleus — almost always a vowel — which is the sonority peak of the syllable. Surrounding the nucleus are the onset (consonants before the nucleus) and the coda (consonants after the nucleus). The nucleus and coda together form the rime, which is the part of the syllable relevant to rhyme and stress. The structure can be represented as a tree: syllable → (onset) + rime → nucleus + (coda). Brackets indicate optional elements: onsets and codas can be absent, which is why "a" and "at" are both well-formed syllables.
Phonotactics refers to the constraints on which segments can appear in which positions. Not all consonant sequences are equally permitted: English allows "str-" as an onset but not "tl-"; Japanese permits almost no coda consonants; Arabic tolerates complex codas that English speakers find difficult to pronounce. These patterns are not random — they follow from the sonority sequencing principle: within a syllable, sonority (a scale running from obstruents < nasals < liquids < glides < vowels) should increase toward the nucleus and decrease away from it. An English onset like "str-" follows the hierarchy (fricative → stop → liquid → vowel rises in sonority), while "tl-" doesn't (stop → liquid, but then violating the sequencing in specific ways). Languages vary in how strictly they enforce sonority sequencing and in where they draw the cutoff for permissible clusters.
From your introduction to Optimality Theory, you already have the tools to analyze these patterns formally. Syllable structure emerges from constraint interaction: ONSET (syllables should have onsets) and NOCODA (syllables should not have codas) are markedness constraints that push toward the universally preferred CV syllable shape. These conflict with MAX-IO (don't delete segments) and DEP-IO (don't insert segments) — faithfulness constraints that resist modifying the input. A language that ranks NOCODA above MAX-IO will delete or devoice coda consonants; a language that ranks MAX-IO above NOCODA will preserve codas at the cost of marked structure. Japanese's near-universal CV syllable pattern reflects very high ranking of ONSET and NOCODA; English's tolerance of complex clusters reflects MAX-IO dominating more broadly.
The deeper point is that syllabification is phonologically active, not just a description of how sounds happen to group. Syllable boundaries determine where many phonological processes apply: aspiration in English (initial stop aspiration applies to syllable-initial stops, not coda stops: "pit" vs. "spit"); flapping (the /t/ in "butter" is syllable-initial in the second syllable, triggering a tap); and stress assignment (stress systems refer to syllable weight — whether the rime contains a long vowel or a coda consonant — not just segment count). Understanding syllable structure is therefore a prerequisite for analyzing almost every prosodic phenomenon in phonology: stress, tone, reduplication, vowel harmony, and the directionality of many segmental rules are all stated at the syllable level or depend on syllabification to be correctly stated.