Portuguese Morphophonology: A Generative-Markedness Approach

§I.0.0 A non-linguist seeing Portuguese Morphophonology: A Generative-Markedness Approach on a list of publications, or actually finding it in a bookstore, might wonder: Who would read such a book, much less buy it? Indeed, many phonologists in either circumstance might react to such a title by expressing bewilderment that a book with that orientation would or could be published in the twenty-first century. Admittedly, strictly generative treatments of phonological phenomena have gone out of style in the last 20–25 years, and linguists have adopted approaches such as autosegmental, metrical, lexical, and prosodic phonology as well as Optimality Theory to address phonological phenomena that have challenged linguists using analytic orientations such as generativism and its predecessors. However, as insightful as some of these post-generative approaches may have been in certain areas, they have rarely surpassed the analytic insight and generalizations possible in generative phonology’s best instantiations. In addition, many latter-day practitioners have become much more interested in their theories than in, especially, particular languages and their grammars (cf. Gross, 1979).

§I.0.2 Still, linguists familiar with Portuguese may ask themselves whether their field needs another book-length treatise on Portuguese (morpho)phonology given that three (one in Portuguese, one in French, and one in English) have been published between 1975 and 2000. But some of those linguists might hope that a new treatise will finally provide a clear-headed, coherent, principled, and reasonably thorough study of Portuguese morphophonology.

§I.0.3 This book’s precursors: Mira Mateus’s Aspectos da fonologia portuguesa (1975 [1982]), d’Andrade Pardal’s Aspects de la phonologie (générative) du portuguais (1977), and d’Andrade’s and Mateus’s combined effort, The Phonology of Portuguese (2000), are, despite their important insights, less than clear-headed, coherent, principled, and thorough. I have reviewed all three (Brakel, 1977b, 1982a, 2008), and this introduction begins by outlining specific objections to their work.

§I.1.2.0 Mira Mateus originally posited an underlying unspecified /N/ to account for the nasalization of vowels both in words such as canto [′kẵⁿ.tŭ], ‘corner’ and mão [′mẵ^ũ] ‘hand’. She used /N/ as a nasalizing diacritic uniting both the phonetic process of tautosyllabic nasalization canto /kaNt+o/ ➔ [′kẵⁿ.tŭ] and of morphophonemic nasalization mão /maN+o/ ➔ [′mẵ^ũ]. In her revision, she opted for /n/ as the nasal consonant responsible for these vocalic nasalizations, even though the first kind is assimilatory and the second morphophonemic.

§I.1.2.1 Her revised version of Aspectos proposes three degrees of nasalization: the strongest occurring when a morphophonemic nasal consonant is deleted: /man+o/ ➔ [′mẵ^ũ], the next strongest is when the consonant is tautosyllabic: [′kẵn.tŭ], and the weakest occurs when a heterosyllabic nasal consonant produces nasal resonance on a preceding vowel: ano [′ẵ.nu] ‘year’. She once more fails to acknowledge that the strongest case of nasalization is the result of a morphophonemic process, whereas the other cases entail low level assimilatory phenomena.

§I.2 D’Andrade Pardal’s Aspects de la phonologie (générative) du portugais was published in 1977, and, similar to Mira Mateus’s first edition of Aspectos, its most up-to-date bibliographical entry is 1973. To his credit, d’Andrade Pardal tried not to postulate abstract or unspecified segments for underlying forms, but unfortunately he identified the nominative plural morpheme as /s/ and had to acknowledge (192) that it is never realized as such in the type of Portuguese he describes. Because this morpheme surfaces as [z]‌ when intervocalic, d’Andrade Pardal was forced postulate underlying geminate /ss/ for word-internal /s/ to emerge as [s], e.g. the past subjunctive morpheme in falasse ‘might speak’ would be +sse#. However, there is no convincing argument to be made for underlying geminate segments within the domain of a single Portuguese morpheme, orthography notwithstanding. Beyond this, d’Andrade Pardal relied heavily on diacritics to make his underlying forms susceptible or immune to specific rules in their derivations. For the preterite he proposed an underlying /+vi+/ morpheme, which disappears in the course of the derivation of preterite forms, save, perhaps, the /i/ in the 1^st person singular: falei [fă.′le^ĭ] ‘I spoke’, comi [kũ.′mi] ‘I ate’, and saí [să.′i] ‘I left’. He presents derivations for words ending in the abstract noun forming suffix, variously written as ção and são (e.g. criação [kri.ă.′sẵ^ũ] ‘creation’ and adesão [ă.δə.′zẵ^ũ] ‘adhesion’), and derives all of them from underlying forms containing the past participle morpheme +t+. Pardal’s book falls short because it provides scant articulatory or phonotactic motivation for his rules and it eschews derivations for some more problematic nominal forms. As a treatise based on late 1960s and early 1970s work, and not withstanding its valuable insights, it, too, was out of date before it was published.

§I.3.2.2 Mateus’ and d’Andrade’s exposition is so repetitive that many interesting and complicated data (e.g. clitics) have perforce been omitted. Nothing has been gained in this volume over the insights the authors propound in their individual efforts.

§I.3.3 Beyond Mira Mateus’s and d’Andrade Pardal’s uncritical acceptance of early generative approaches to phonology and the incoherence in their mutual autosegmental endeavor, there is another matter that has rendered their work unsatisfactory. They approach the Portuguese sound system’s patterns from a phonological point of view, but one cannot address Portuguese phonology adequately without secure footing in that language’s morphology and its morphophonemics. The language’s truly impressive consonantism and vocalism are conditioned by its morphology, and nowhere in their work is there anything resembling a theory of morphology. Thus, there is ample room, indeed a need, for a new book that approaches the Portuguese sound system with greater clarity, completeness, and, most of all, with a coherent approach to morphology.

§1.4.0 However, before proceeding to the book itself, this introduction must address work on the Portuguese sound system that has recently emerged.

§1.4.2 Whereas all these chapters help inform the work in the present volume and are cited and discussed in chapters 3, 4, 6, 7, 8, 9, 10, 12, 13, and 16, their authors have not worked in concert toward advancing a grammar of the phenomena examined. Indeed, these authors have eschewed the analysis of Portuguese words based on underlying patterns on which sets of rules operate to generate the actual forms that occur in the language. In their attempt to describe and at times explain extant patterns in the language through the analysis of surface manifestations of Portuguese words, several authors postulate phantom segments in attempts to make their analyses more ‘coherent.’ Or they maintain that, in the interest of making their postulated patterns consistent, specific morphemes and phonemes that are present in the data do not count as counter-examples to their analyses—they dismiss these morphemes and phonemes as extra-metrical. In dealing with various patterns, they often allude to rules that purportedly account for these patterns, but they do not formalize said rules, which makes testing the assertions these alleged rules purport to support impossible. Like Portuguese Phonology, the chapters addressing morphological matters do so without any explicit theory of morphology. The scant formalism one finds fails to unite disparate but related phenomena—at most it appears to indicate that the authors adhere to one or another post-generative approach to phonology. In short, the chapters pertinent to this study are piecemeal studies of phenomena which, while they do not point toward a coherent and principled account of Portuguese morphophonology, in the above-mentioned chapters they do provide evidence that supports the advancement (presented here) of a consistent and over-arching account of the morphophonology operant in the Portuguese language.

§I.5.3 Since its aim is to address morphophonology, the grammar will not be based on any particular version of contemporary Portuguese. The phonetic transcriptions provided will portray, unless otherwise stipulated, a non-specific, conservative, standard European rather than Brazilian pronunciation. But the items (phonemes, morphemes, and boundary affinities) adopted or postulated obtain in all standard varieties of Portuguese. With a set of dialect-specific rules applied to the morphophonemic transcriptions this grammar enables, one could easily advance a set of rules to derive pronunciations typical of contemporary Lisbon, Oporto, Rio, Recife, Porto Alegre—indeed of any variety of spoken Portuguese.