REGISTRO DOI: 10.69849/revistaft/dt10202508221235
Tatiana Bermond dos Santos Rodrigues
ABSTRACT
This paper aims to present the palatalization of /n/, a phonological variation registered in Rodrigues (2017), under the perspective of the syllable model of non-linear phonology. Upon discussing the theoretical frameworks that underlie the model, anchored in Kahn (1976), Clements & Keyser (1983), Silva (2013) among others, this study aimed to investigate the marks left by variants on the syllabic structure of words and to reflect on the possible clues in each structure that justify the variable processes of the variant (nɪ̯). As a result, based on a general analysis, the primary reason for these sound changes is the reorganization of syllables. This is mainly because the beginning sound of the last syllable becomes simpler. The process also reverses a pattern where stress falls on the third-to-last syllable in a word, a change that happens when sounds are completely deleted or turned into palatal sounds. However, this doesn’t include sound changes caused by secondary palatalization.
Keywords: Palatalization. Syllable Model. Non-linear phonology.
INTRODUCTION
The Amazonian sub-dialogue was recorded by Nascentes (1953) as one of the dialects of the northern part of Brazil. One of the most significant studies recording the Amazonian sub dialogue, specifically the Amazonian dialect, originating from the state of Amazonas, was carried out by Cruz (2004) through the development of ALAM – Linguistic Atlas of Amazonas, which included an investigation in nine municipalities representing the nine microregions of the state.
ALAM Phonetic Chart 94 presents the realization of the palatalized variant of /n/ in three locations in Amazonas, the variant [nʲ] recorded in the speech of three informants, one in Barcelos, another in Parintins and another in Tefé, and the variant [ɲ], in the speech of two informants, one in Benjamin Constant and another in Parintins. And in all locations, the vocalization variant of /n/ in /ɪ̯/ was recorded.
Through unsystematic observation, it was detected that these variants are also present in the Manauara speech, however, there is no such record for Manaus in ALAM, as this city was not one of the locations of the research carried out by Cruz (2004). This perception motivated the development of the thesis on the variable (nɪ̯) in the context of paroxytone words ending in –nia and –nio in Manaus, as registered in Rodrigues (2017).
This paper presents a selection of data from the thesis mentioned, focusing on the objective of verifying changes in the syllabic structure of words produced with palatalized, non nasal, vocalized and erased variants, looking for evidence of syllabic motivation.
THEORETICAL FRAMEWORK: THE SYLLABLE MODEL OF NONLINEAR PHONOLOGY
The study of the syllable has evolved from the Prague School, through structuralism, to modern generative theory, and was incorporated into generative phonology.
Harris & Bell-Berti (1984) note that physiological research on the syllable began with the work of Stetson (1951), who suggested a physiological concept for the syllable. In the authors’ interpretation, “the syllable was physiologically defined by an initiating and a terminating burst of activity from the muscles of the chest wall, the internal and external intercostal muscles, resulting in a distinct chest pulse for each syllable. (Harris & Bell-Berti, 1984, p.90)
Following the assumptions of Stetson (1951), also defended by Abercrombie (1967), Silva (2013) defines that the syllable is then interpreted as a movement of muscular force that intensifies, reaching a maximum limit, after which there will be a progressive reduction of this force. The author shares the syllable scheme proposed by Cagliari (1981):
Figure 01: Diagram of muscular effort and syllabic force curve, based on Cagliari (1981, p.101).
Silva (2013) explains that the diagram above represents the three parts that make up a syllable: a nuclear part, which must be filled by a vowel segment, and two peripheral parts, which must be filled by consonant segments. The author also states that if the syllable is composed solely of the vowel segment, it will fill all parts of the syllable structure.
Furthermore, the nucleus of a syllable is not necessarily accented.
Collischonn (2005) emphasizes that the works of Hooper (1976) and Kahn (1976) were the precursors to the concept of the syllable as a phonological unit. This concept would later serve as the basis for the theory proposed by Clements & Keyser (1983), which assigns a phonological status to the syllable. According to the author, Kahn (1976) formulated a theory of syllable structure with a representation of independent layers, as shown in the figure below:
Figure 02: Representation of the syllable structure based on Kahn (1976). (Collischonn, 2005, p.101)
In the previous representation, the symbol σ represents the syllable and is directly linked to the consonant and vowel segments. This representation is classified by Mendonça (2003) as a flat structure model, where there are no subconstituents.
When describing the syllabic pattern of the Portuguese language, Guimarães (1927) postulates that syllables in Portuguese can consist solely of vowels and diphthongs or of both in combination with consonants. In syllables composed of vowels and consonants, there is a clear tendency toward the normalization of the syllabic group, beginning with a consonant and ending with a vowel. Thus, when words are joined together, the final consonant of the first word contacts the initial vowel of the second word. In spoken language, the latter forms a syllabic group of the normal type—consonant plus vowel.
The CV syllabic pattern postulated by Guimarães (1927) was one of the foundations of the theory proposed by Clements & Keyser (1983) in their book “Fonologia CV: uma teoria gerativa da sílaba.”. The authors propose that the relationship between the syllable and its segments should be measured by a CV tier, generating three levels of representation: segmental, CV, and the syllable level, represented by σ.
Silva (2013) interprets that in Clements & Keyser’s (1983) proposal, the distinctive features [consonantal] and [syllabic] are replaced by the categories C (for consonants) and V(for vowels). The author exemplifies Clements & Keyser’s (1983) proposal with the syllabic structure of the word “vida” (life):
Figure 03: Representation of the syllabic structure of the word vida based on the generative theory of the syllable, CV Phonology, proposed by Clements & Keyser (1983). (Silva, 2013, p.203)
Silva (2013) highlights that this model aims to discuss the interaction between phonological processes and syllabic structure, as well as to define a typology for the syllabic inventories of natural languages.
Collischonn (2005) points out another theoretical branch for the description of the internal structure of the syllable, this time proposed by Selkirk (1980). This is classified by Mendonça (2003) as a binary model with rhyme. Collischonn (2005) illustrates the representation proposed by Selkirk (1982):
Figure 04: Representation of the internal structure of the syllable based on Selkirk (1982). (Collischonn, 2005, p. 102)
In the representation above, (A) represents an onset and (R) a rhyme, which, in turn, consists of a nucleus (Nu) and a coda (Co).
Silva (2013) presents the following example of a syllabic pattern in the Portuguese language, along the lines of the binary model with rhyme:
Figure 05: Standard syllabic structure of the Portuguese language. (Silva, 2013, p.206)
The processes of syllabification are discussed in Bisol (2005) and Silva (2013). The former explains the universal conditions for good syllable formation, which are sonority sequence and prosodic licensing. The latter describes the same conditions, but with different names: the principle of sonority and the conditions of syllabic licensing.
Through the sonority scale, it is possible to “correlate the relative sonority of a segment with its position within the syllable.” (Collischonn, 2005, p. 111) The author instructs that the nucleus of the syllable should be occupied by the most sonorous element, while the least sonorous ones will occupy the onset and coda. She further instructs that “when there are sequences of elements within the onset or coda, these present increasing sonority toward the nucleus.”
Silva (2013) proposes the following diagram of sound hierarchy:
Figure 06: Hierarchy of sonority. (Silva, 2013, p. 207)
Furthermore, in the regulation of syllabic structure, there is the principle of prosodic licensing, proposed by Ito (1986). According to this principle, all prosodic units at a given level must belong to hierarchically superior structures (Collischonn, 2005). Because of this principle, every segment must be associated with a syllable.
In an attempt to formalize phonological phenomena through syllabic structure, Kaye & Lowenstamm (1985) proposed that C and V, corresponding, respectively, to consonant and vowel, be replaced by “x,” representing pure positions. This type of representation has been widely used in nonlinear phonology.
RESULTS: APPLICATION OF THE NON-LINEAR PHONOLOGY SYLLABLE MODEL
This section is dedicated to the syllable model of nonlinear phonology, aiming to investigate the marks left by variants on the syllabic structure of words and to reflect on the possible clues in each structure that justify the variable processes of the variant (nɪ̯).
Note that the objective of this section is not to analyze the phonological processes that produce non-alveolar variants, using syllable theory. Rather, it seeks to verify the changes brought about in syllable structure when the new variant is inserted.
The words Virgínia, Betânia, and crânio will be analyzed, representing all the phonological processes that generated the five variants of (nɪ̯) identified in this research. The syllabic structure of the three words will be presented following the model proposed by Kaye & Lowenstamm (1985), that is, the binary model with rhyme, but with the use of “x” instead of the letters C and V to represent pure positions. The underlying form of the word will be presented in phonetic transcription. Each word will be presented in a sequence with the variants recorded for it in this research.
First, a brief discussion is necessary about an element that is present in all words in the Manaus corpus. This is the glide [j], present in the rising diphthong endings –nia and –nio. What would be the position of this element in the syllabic structure? Silva (2013, p.171) recommends that “glides are always associated with a vowel and can never be the nucleus of the syllable (and consequently, a glide cannot receive an accent).” Therefore, the glide should be positioned at the onset of the syllable, adopting Bisol’s (2005) proposal.
Since the glide is not interpreted as a vowel in the syllabic structure, the endings –nia and –nio can be considered to have the CCV structure, which is represented by a branched onset, that is, complex syllables.
The first word is Virgínia, which was pronounced with two variants: alveolar and deletion.
1. Virginia – with the alveolar variant
What is noticeable in the syllabic structure of the word Virgínia, with the production of the alveolar variant, is the sequence CVC–CV–CCV, where the stressed syllable falls on the penultimate syllable of the CV sequence. Therefore, the complex structure of the branched onset in the last syllable -nia is noticeable. Also noticeable is the process of phonetic nasalization, in which the nasal consonant spreads the [+nasal] feature to the preceding vowel, but the nasal consonant itself does not undergo fall.
2. Virginia – with the variant [ø]
Structure representing the nasalization of the preceding vowel and the deletion of the nasal segment:
The structure above represents the spreading of the [nasal] feature from /n/ to the preceding vowel, followed by the deletion of the nasal segment.
Final structure after the deletion of the nasal segment:
In the syllabic structure of the word Virgínia, produced by the deletion variant, we find the sequence CVC – CV – CV. Therefore, we notice the simplification in the last syllable, where the onset ceases to be branched. The deletion cancels the tier of the nasal segment, which formed the onset of the last syllable in the CCV structure. The motivation, then, appears to be the simplification of the syllabic structure.
The next variant to be analyzed is full palatalization, with the word Betânia.
3. Betânia – with the variant [ɲ], total palatalization
In the syllabic structure of the word Betânia, with the production of the full palatalization variant, the sequence CV–CV–CV is observed, where the stressed syllable falls on the penultimate syllable of the CV sequence. The change observed in the syllabic structure, caused by full palatalization, is the simplification of the onset of the last syllable, which, instead of two segments, carries only one. Although the preceding vowel was nasalized before the palatalization process, it is assumed that phonetic nasality is maintained with the presence of the segment /ɲ/ in the following syllable. Thus, this vowel is not interpreted as a nasal vowel, but rather as nasalized.
The disappearance, however, of the glide/semivowel /ɪ̯/ due to the process of total palatalization also seems to be a clue to the attempt to avoid the proparoxytone appearance of this group of words, since, in this way, the word no longer ends in the posttonic rising diphthong, one of the premises for classifying an apparent proparoxytone.
The next variant to be analyzed is secondary palatalization, with the word crânio.
4. crânio – with the variant [nj], secondary palatalization.
Secondary palatalization did not cause any changes to the syllabic structure of the cranium, which in the production of the alveolar variant presents the sequence CCV-CCV. The only change is in the representation of the nasal segment, which now presents the assimilation of the palatal feature from the semivowel /ɪ̯/.
The last variant to be analyzed is vocalization, also in the word Betânia.
5. Betânia – with the variant [ɪ̯], vocalization.
Structure representing the vocalization of /n/ and the composition of complex onset by geminate consonant/semivowel:
In the vocalization of /n/ into /ɪ̯/, consonant clustering or gemination occurs in ɪ̯ ɪ̯. The syllabic structure is represented by the assumptions of Cedeno et al. (1999), who point out that the phonological difference between a simple consonant and a geminated consonant corresponds to the fact that, at the CV (Consonant-Vowel) level, the simple consonant has a single C, while the geminated consonant has two Ces, both linked to a segmental melody.
Final structure with resyllabification:
What operates in this context of resyllabification is the principle of obligatory contour, which prohibits the adjacent sequence of identical units in phonological representations, in addition to being a way to avoid the CCV sequence, with a complex onset in the last syllable. The CV level of the syllabic structuring of the word Betânia with application of the vocalization variant is represented by X, as proposed by Kaye & Lowenstamm (1985). By representing the geminate group ɪ̯ ɪ̯in this way, the diphthongization formed by the vocalization process of /n/ is represented, as well as the lengthening in the duration of /ɪ̯/.
In the final syllabic structure, we can see the progressive nasalization that occurs in the coda semivowel due to contact with the preceding nasalized vowel. In other words, progressive nasalization occurs as a result of resyllabification.
The structural sequence changes from CV – CV – CCV, from the original structure with the production of the alveolar variant, to CV – CVC – CV.
Taking into account that the words in the Manaus corpus are considered apparent proparoxytones, and admitting the hypothesis of an attempt to avoid the appearance of proparoxytones, it is considered that Collischonn (2005) highlights something relevant about paroxytone words: There is a generalization: When the penultimate syllable is heavy, the accent will never fall on the antepenultimate syllable.
Searching for the notion of light syllable and heavy syllable, we take as reference Cagliari (1981) who consider the CV sequence to always be light, and CVC syllables can be considered light if the nucleus elements are counted, or heavy, if the rhyme elements are counted, which is always taken into consideration in the Portuguese language.
Thus, it can be seen that in the original syllabic structure, with the production of the alveolar variant, the accent falls on the CV sequence syllable, which is considered light. This may alleviate the doubt regarding its proparoxytone appearance, although this doubt arises in the written form rather than in the spoken form of the word, as there was no change in pronunciation that would make the word sound less paroxytone or more proparoxytone. The CVC sequence in the stressed syllable, then, would highlight the nature of the accentuation of this group of words as paroxytone, attenuating the proparoxytone appearance.
Since all variants of (nɪ̯) were analyzed in the syllabic context, in the next section the phonological analyses are related to the syllabic analyses in order to delimit their contributions to the research.
FINAL CONSIDERATIONS: Contributions of Syllabic Analysis
Syllabic analyses may reveal factors that favor the production of paroxytone (nɪ̯) variants ending in -nia and -nio in Manaus, as well as factors that favor variation in a broader sense.
Starting with one variant at a time, the following survey is conducted:
Total Palatalization
Syllabically, structural analysis shows that the occurrence of total palatalization simplifies the last syllable, transforming the onset from complex to simple, and also deletes the /ɪ̯/ segment responsible for the posttonic rising diphthong at the end of the word. This ending is a premise for the classification of an apparent proparoxytone.
Secondary palatalization
Syllabic analysis reveals no changes in the structure of the final syllable, maintaining the original structure, including accentuation and classification as an apparent proparoxytone. The only change, therefore, is the assimilation of the palatal feature by the nasal segment.
Vocalization
Syllabically, the simplification of the last syllable and the resyllabification process were recorded, generating a falling diphthong in the penultimate syllable, changing the structure of the stressed syllable to the CVC sequence, that is, a heavy syllable, which favors the application of the paroxytone accentuation rule, avoiding the proparoxytone appearance of this group of words.
Deletion
Syllabic analysis indicates the simplification of the last syllable of the word, with an onset without branching.
In summary, the data collected, compared with phonological analyses recorded in Rodrigues (2017), reveal what is demonstrated in the following table:
A general analysis indicates that syllabic restructuring, mainly with the simplification of the onset of the last syllable and the eventual deconvolution of the proparoxytone appearance of words, in cases of deletion and total palatalization, appears to be the greatest motivator and likely driver of phonological processes, except for secondary palatalization.
REFERENCES
ABERCROMBIE, D. Elements of General Phonetics. Edimburgo: Edinburgh University Press, 1967.
BISOL, L. Introdução a Estudos de Fonologia do Português Brasileiro. 4a ed. Porto Alegre: EDIPUC – RS, 2005.
CAGLIARI, L. C. Elementos de Fonética do Português Brasileiro. Tese de livre docência. Campinas: Unicamp, 1981.
CLEMENTS, G. N.; KEYSER, S. J. CV phonology: a generative theory of the syllable. Linguistic Inquiry Monograph. Cambridge, Mass.: MIT Press, n. 9, 1983.
COLLISCHONN, G. A sílaba em português. In: BISOL, L. Introdução a Estudos de Fonologia do Português Brasileiro. 4a ed. Porto Alegre: EDIPUC – RS, 2005.
CRUZ, M. L. de C. Atlas Linguístico do Amazonas – ALAM. Rio de Janeiro: UFRJ, Faculdade de Letras. Tese de Doutorado em Letras Vernáculas, 2004.
GUIMARÃES, O. Fonética Portuguesa. Coimbra: Coimbra Editora, 1927.
HARRIS, K. S ; BELL-BERTI, F. In: RAPHAEL, L. J.; RAPHAEL, C. B.; VALDOVINOS, M. R., eds. Language and Cognition: Essays in Honor of Arthur J. Bronstein. Springer, 1984.
HOOPER, J. Na introduction to natural generative phonology. New York: Academic press, 1976.
ITO, J. Syllable Theory in Prosodic Phonology. Tese (Doutorado, PhD) – University of Massachusetts, 1986.
KAHN, D. Syllable – based generalizations in English Phonology. Tese (Doutorado, Phd) – Cambridge, Mass: MIT, 1976.
KAYE, J. D.; LOWENSTAMM, J. Compensatory lengthening in Tiberian Hebrew. In: Studies of compensatory lengthening. Dordrecht: Foris, 1985.
MENDONÇA, C. S. I. de. A sílaba em Fonologia. Working Papers em Linguística (Impresso), Florianópolis, v. 7, n.1, p. 21-40, 2003.
NASCENTES, A. O linguajar carioca. Rio de Janeiro: Organização Simões, 1953.
RODRIGUES, T.B.dos S. Estudo da Variável (Nɪ̯ ) em Manaus: Variação Fonológica em Paroxítonas Terminadas em -nia e -nio. Tese de Doutorado. Programa de Pós-graduação em Estudos Linguísticos. Faculdade de Letras. Belo Horizonte: UFMG, 2017.
SELKIRK, E. On prosodic structure and its relation to syntactic structure. Bloomington: IULC, 1980.
SILVA, T. C. Fonética e fonologia do português: roteiro de estudos e guia de exercícios. 10.ed. São Paulo: Contexto, 2013.
STETSON, R. H. Motor Phonetics: A Study of Speech Movements in Articulation (2nd Ed.). Amsterdam: North Holland, 1951.