根據 Kiparsky (1982) 所提的「表意明確論」，凡和字義有關的訊息會在該字的表面結構存留下來。因此對於某些西班牙文方言中，字尾/s/消失的情況應該不會發生。即使發生了，也應該有某種替代的結構來取代這消失的/s/。本文延用Hammond (1978)和Figueroa (2000)的研究方法，研究目的乃使用實驗的方式，來探究西班牙安大路西亞（Andalusia）方言中是否存在某種替代結構，像是把母音延長或是改變母音來取代消失的/s/字尾。六位安大路西亞方言母語人士參與此項研究，並以他們的語言樣本制作了一份包含四大類組的語言辨別測驗。之後，再請二十五位也是安大路西亞的母語人士完成該測驗看他們是否能分辨因/s/消失而形成的同音異義字。最後再將這些字用光譜儀分析。研究結果發現安大路西亞方言並無任何替代結構來取代消失的/s/字尾，但確有足夠的證明顯示，音節中若有/s/消失了，則該音節的母音發音長度較原字平均長約24.4%，例如，79%的參與者能夠成功地辨識buque [ˈβu k ] ‘船’vs. busque [ˈβuø.ke] ‘尋找’。這份研究結果與Hammond (1978)對邁阿密古巴人的西語方言研究和Figueroa (2000)對波多黎各的方言研究一致。

According to the Distinctiveness Conditions proposed by Kiparsky (1982), semantically relevant information tends to be retained in the surface structure. As a result, one would anticipate either the retention of /s/ or some compensatory mechanism to convey the information lost by its deletion in those dialects of Spanish that systematically weaken to the point of deletion said segment in syllable- and word-final positions. Following the methodologies of Hammond (1978) and Figueroa (2000), the objective of this study is to experimentally explore in Andalusian Spanish the possible existence of a phonological quantitative and/or qualitative compensatory mechanism, realized as either vowel lengthening or an open-closed vowel alternation, in an attempt to disambiguate those tokens rendered homophonous by the deletion of final /s/. Speech samples were elicited from six native speakers of Andalusian Spanish. These speech samples were then organized into a four-part perception test and administered to twenty-five participants who were also native speakers of this same dialect area. Finally, spectrographic analysis was performed on all tokens used in the perception test. This study found no systematic evidence of any compensatory mechanism at work in the vowels preceding /s/ → [ø] in word-final position. However, this investigation did find sufficient evidence for the phonemicization of vowel duration in syllable-final position within the word. An average increase of 24.4% in the length of the vowel preceding the [ø] allophone of /s/ provided the participants with adequate acoustic cues to correctly distinguish pairs such as buque [ˈβu.ke] ‘ship’ (n.) and busque [ˈβuø.ke] ‘look for’ (3SG.PRES.SUBJ) at a rate of 79.0%. The aforementioned findings concur with those of both Hammond (1978) for Miami-Cuban Spanish and Figueroa (2000) for Puerto Rican Spanish.