Carlos Silva

Among Portuguese mainstream phonologists (e.g. Mateus & Andrade 2000), Regressive Vowel
Harmony (RVH) has been assumed as the only type of vowel assimilation in European Portuguese
(EP). Indeed, progressive harmony is conceived of as, if not impossible, at least highly exceptional,
almost idiosyncratically exclusive of specific lexical entries (ibid.). However, in some dialects of EP
Progressive Vowel Harmony (PVH) proves very frequent and productive. This is namely the case of
some dialects spoken in the Madeira island (Segura da Cruz & Saramago 1999), where
(1) an /I/ in a stressed syllable diphthongizes: {[í]>[ɨj] = |@,@|Stress ℘ |I, I|Glide}.
(2) if the precedent syllable vowel has a |U| element as its head, PVH is triggered. The
consequences of the harmony process can be checked in realizations such as a) [ˈɨj] = {|@,@|Stress
℘|I, I|Glide} Ex.: [puˈlɨjsjɐ], b) [ˈuj] = {|U,U|Stress ℘|I, I|Glide} Ex.: [puˈlujsjɐ] c) {|U,@|Stress
℘|I, I|Glide} Ex.:[puˈlʉjsjɐ]. The diphthongization can be explained as a mechanism of assigning
syllabic weight, making the nucleus appear as if it was fulfilled by a long vowel, as represented
by Backley (2011: 48). The PVH is triggered by the |U| element only (which then behaves, in
this respect, asymmetrically to |I|).

So far, the Historia Regum Britanniae of Geoffrey by Monmouth has not been taken into consideration by scholars of Arthurian Studies. However, this historiographical text provides a set of remarkable information and meanings which relaveal not only the cultural environment of the the first half of the 12th century, but also the political and legitimacy goals of an author writing under the patronage of the Anglo-Normand dynasty. Among the narrative motifs present in this work, the figure of king Cassibelaunus is worthy of a careful analysis, since it can be interpreted as a potential antecedent to Robert of Boron’s version of the Excalibur episode, in which the famous challenge of the sword in the stone successfully completed by Arthur makes him king.

RESUMO. O arquipélago da Madeira, (re)descoberto em 1419, foi, durante os séculos XV e XVI, povoado sobretudo por portugueses cuja origem geográfica é ainda muito discutível. O certo é que, por ser uma região de grande variedade, os seus múltiplos dialetos, especialmente os mais isolados, tornam-se interessantes para os estudos linguísticos, por um lado pela conservação de marcas bem vincadas de fases arcaicas do português, por outro pelos diversos paralelos fonéticos com várias regiões do continente, desde Trás-os-Montes até Sagres, e, por fim, pelas inovações próprias que desenvolveu durante quase 600 anos de autonomia linguística. ABSTRACT. Madeira's archipelago was (re)discovered in 1419. During the fifteenth and sixteenth centuries, it was colonized mainly by portuguese people whose geographical origin is still doubtful. The truth is that the dialectal variety in the region makes its many dialects, especially the most isolated, fascinating for linguitics' researchers. On the one hand because they often preserve marks from archaic phases of portuguese language; on the other hand, because there are several phonetic parallels between these dialects and those who are spoken in different regions of the mainland from Trás-os-Montes to Sagres, and finally, because many innovations were developed during almost 600 years of linguistic autonomy.

Among Portuguese mainstream phonologists (e.g. Mateus & Andrade 2000), Regressive Vowel Harmony (RVH)
has been assumed as the only type of vowel assimilation in European Portuguese (EP). Indeed, progressive
harmony is conceived of as, if not impossible, at least highly exceptional, almost idiosyncratically exclusive of
specific lexical entries (ibid.). However, in some dialects of EP Progressive Vowel Harmony (PVH) proves very
frequent and productive.
This is namely the case of some dialects spoken in the Madeira island (Segura da Cruz & Saramago 1999),
where
1) an /I/ in a stressed syllable diphthongizes: {[í]→[ɨj] = |@,@|Stress  |I, I|Glide}.
2) if the precedent syllable vowel has a |U| element as its head, PVH is triggered. The consequences of the
harmony process can be checked in realizations such as:
[ˈɨj] = {|@,@|Stress |I, I|Glide} Ex.: [puˈlɨjsjɐ]
[ˈuj] = {|U,U|Stress |I, I|Glide} Ex.: [puˈlujsjɐ] trigger form: /puˈlisjɐ/ ‘police’
[ˈʉj] = {|U,@|Stress |I, I|Glide} Ex.:[puˈlʉjsjɐ]
The diphthongization can be explained as a mechanism of assigning syllabic weight, making the nucleus appear
as if it was fulfilled by a long vowel, as represented by Backley (2011: 48). However, in Portuguese, there is no
phonological contrast between long and short vowels, so this element copying cannot be accepted, leaving behind a
vowel empty of elements.
The PVH is triggered by the |U| element only (which then behaves, in this respect, asymmetrically to |I|). So,
why do the phonetic realizations – [uj], [ʉj] and [ɨj] – differ? We can explain it as a special case of PVH, triggered by
|U| or both PHV and RVH, triggered by |I|, as the two processes active at the same time.
To sum up, Element Phonology, combined with Autosegmental Phonology, offers a strong explanation for
harmony processes. Furthermore, this case of VH itself shows the asymmetry between the elements. |U|- labiality- and
|I|-palatality- and how they neutralize one another, whenever the result of the VH is [ɨj] = |@,@| |I, I|.