A crosslinguistic study is underway concerning children's protracted phonological development (i.e. speech sound disorders). The current article reports pilot Spanish data for this study from two 4-year-old boys with protracted phonological development. The purposes of the pilot study were to: (1) develop and evaluate a word list for elicitation that could be used across Spanish dialects and that sufficiently sampled Spanish word lengths, stress patterns, word shapes and phonemes; and (2) to derive hypotheses for the larger study, based on patterns found in these children's speech, and a review of the literature. The two speakers showed some developmental patterns reported for other languages (e.g. constraints on production of liquids and word-initial consonants in unstressed syllables) but also patterns that may reflect Spanish phonological inventories, allophony and frequencies. These data helped consolidate the Spanish word list for elicitation and led to questions for the ongoing study concerning word structure, multisyllabic words, liquids, fricatives and vowel sequences.

The current study provides preliminary reference data for word structure development in a Spanish variety with restricted coda use, both by age and types of word structures. Between ages 3 and 5 years, global measures (whole word match, word shape match) distinguished children with typical versus protracted phonological development. By age 4, children with typical development showed near-mastery of word structures, whereas 4- and 5-year-olds with PPD continued to show syllable deletion and cluster reduction, especially in multisyllabic words. The results underline the relevance of multisyllabic words and words with clusters in Spanish phonological assessment and the utility of word structure data for identification of protracted phonological development.

Models of language learning and processing differ in their level of emphasis on the storage of individual meaningful units versus combinations of meaningful units. While there is evidence for the storage of larger stretches of speech, a separate issue is how much such stored forms contribute to processing, as compared to morphologically simpler forms. We examine the acquisition of one aspect of the phonology of Valley Zapotec: complementarity of segmental length based on subsegmental features: vowels before fortis consonants are short (VCː), and vowels before lenis consonants are long (VːC). This complementarity is found for fortis consonants in morphologically simple forms with final stress (simple nouns, verbs with full subject noun), but not in morphologically complex forms with a final unstressed syllable (diminutive nouns, verbs with pronominal subject clitic). During one period of development, Zapotec-learning children overgeneralize the complementarity from morphologically simple to morphologically complex forms (with u-shaped learning likely). The child’s processing of complex forms in language production is based more on simple forms than on the complex forms themselves. We identify five possible explanations of these results. Insofar as combinations of morphemes are stored at this young age, they are relatively ineffective at influencing processing during language production.

It is well known that F0 at vowel onset can be influenced by a preceding consonant. That influence varies significantly across languages and consonant types, and may function as a perceptual signal to consonant manner. It has further been suggested that tone languages may behave differently from non-tone languages in this respect, with a shorter duration of consonantal perturbation [Hombert, Studies in African Linguistics, 1977]. Previous studies include a limited range of consonant types, and too few tone languages to test Hombert’s proposal. This study presents the results of an acoustical investigation of the effects of implosives and prenasalized stops on the F0 of a following vowel in Shona, a tone language. It is found that implosives have a similar raising effect on F0 at vowel onset than that of voiceless (aspirated) stops, contrary to expectations based on previous studies [Wright and Shryock, Journal of the Phonetic Association, 1993]. It is also found that prenasalized consonants behave as nasals, having no effect on the F0 of the following vowel, again contrary to expectation [cf. Trithart, Studies in Bantu Tonology, 1976; and Hombert, Studies in Bantu Tonology, 1976]. Finally, duration results do not support Hombert’s position regarding tone languages.