Syllabification occurs within the domain of phonological words. The syllable structure of given strings can be modified through the phonological or morphological processes in the name of resyllabification or ambisyllabification. Moreover, syllabification occurs repeatedly beyond the boundary of domain of phonological words. Syllable is dominated by a phonological word, because a lower prosodic unit should be licensed by the upper one. But when a certain syllable is dominated by a phonological phrase, it is the case of the allocation of syllable structure in a broader domain beyond the phonological words. Phonological words should be not necessarily identical to the morpho-syntactic constituents, so certain prefixes and members of compounds in Korean can be phonological words which provide domains for syllabification. There are two kinds of resyllabification, which apply in the domain of phonological words and larger domains such as phonological phrases or intonational phrases. Ambisyllabification can also occurs in the domain of phonological words and larger domains. This system, a rule-based process, which assumes the underlying representation and a certain phonological rules allows continuous syllabification, so one cannot determines which of the resyllabification and ambisyllabification is a better choice for explanation of the given strings in a certain stage. But in the output-based approach, the optimality theory, any candidates showing any kinds of syllabification which least violates the constraints which are ranked one another are selected as the most optimal form in the prosodic structures. So the motivation of syllabification, whether it is resyllabification or ambisyllabification, is unnecessary in the optimality theory.