On the Interpretation of the Nonexistent: Nonoccurring Syllable Types in Spanish Phonology

Abstract

This paper has two basic aims: (i) to show that phonotactical patterns that do not exist are relevant for establishing the phonology of languages; and (ii) that nön-existent phonotactical structures provide an argument against purely formal analyses of phonological structure.The position taken here is that phonotactical analyses should be based on possible syllable types, not on the existing ones. My evidence comes from the Castilian dialect öf Spanish, whieh is my own. Being fairly conservative, Castilian Spanish prövides an adequate source of data for generalizations on syllable structure since syllable-and word-final consonants are relatively well preserved. I will present my examples in conventional spelling unless it is necessary to specify a phonetic detail.Recent approaches to phonotactical patterns, such äs HARRIS (1983), are based on surveys of syllable types both existent and non-existent. The existing types are the basis for setting up the distributional rules governing possible syllable shapes. The question we inight ask is whether non-existent phonological structures can be equated with impossible ones.Let us look at non-occumng syllable patterns. We may distinguish between two kinds of syllable types:(i) Ungrammatical syllable types. (ii) Syllable types that simply do not appear.

Spanish Taps and Trills: Phonological Structure of an Isolated Opposition

“…Harris (1983: 68) observes that native Speakers reject forms such äs *chinchorro but a subsequent note (p. 144) regarding informants' explanations of their rejection suggests that lack of familiarity with such words may be an important factor (cf. Pensado 1984). 18 I have informally surveyed a number of native Spanish Speakers from different countries and social strata, and while all find words like *chinchorro to be odd, none rejects such a combination with the vehemence provoked by segment-level combinatorial violations.…”

The intricate connection between diphthongs and stress in Spanish

“…In short, it was claimed that the reason for banning antepenultimate stress in such words is that penultimate syllables are heavy and attract stress. Contrary to these analyses, I follow Alvord (2004), Bárkány (2002), Ohannesian (2004), Pensado (1985) and Roca (1997) and do not interpret the banning of antepenultimate stress before a heavy penultimate as clear evidence of their bimoraicity. These latter studies have shown that, even if the above-mentioned structures are scarce, they are not completely absent from Spanish vocabulary, as illustrated in 10 It is true that the words in (10) are not very common and that antepenultimate stress in forms of the type 'σ.CGV.σ, 'σ.CVG.σ and 'σ.CVC.σ is not generally found in Spanish vocabulary.…”

“…However, instead of assuming that penultimate CVC, CGV and CVG are heavy and that is the reason for banning such structures, there is an alternative interpretation of the facts. Since antepenultimate stress was forbidden in Latin in words with heavy penultimates, the near-absence of 'σ.CGV.σ, 'σ.CVG.σ, 'σ.CVC.σ in Spanish vocabulary can be seen as an historical accident (Pensado 1985). That is, antepenultimate stress in words containing a CVC, CVG or CGV in the penultimate syllable is rare but not ungrammatical.…”

The Acquisition of Subset and Superset Phonotactic Knowledge in a Second Language

“…tlapería 'paint/hardware store. '23 Harris (1983) claims that these are accidental gaps, offering an example Jruschev 'Khrushchev', which is disputed byPensado (1985).Downloaded by [University Library Utrecht] at 01:48 18 December 2017…”