Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 1
RELAÇÕES ENTRE A ESCRITA E A FALA NA PRIMEIRA LÍNGUA E SUA
INFLUÊNCIA NO USO DE INGLÊS COMO LÍNGUA ADICIONAL: UMA
REFLEXÃO SOBRE O ENSINO DE PRONÚNCIA
RELACIONES ENTRE ESCRIBIR Y HABLAR EN LA PRIMERA LENGUA Y SU
INFLUENCIA EN EL USO DEL INGLÉS COMO LENGUA ADICIONAL: UNA
REFLEXIÓN SOBRE LA ENSEÑANZA DE LA PRONUNCIACIÓN
THE RELATIONSHIP BETWEEN WRITTEN AND SPOKEN FIRST LANGUAGE AND
ITS INFLUENCE ON THE USE OF ENGLISH AS AN ADITIONAL LANGUAGE: A
REFLECTION ABOUT THE TEACHING OF PRONUNCIATION
Rodrigo Garcia ROSA1
e-mail: rodrigo_21usp@yahoo.com.br
Como referenciar este artigo:
ROSA, R. G. Relações entre a escrita e a fala na primeira língua
e sua influência no uso de Inglês como Língua Adicional: Uma
reflexão sobre o ensino de pronúncia. Rev. EntreLinguas,
Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529.
DOI: https://doi.org/10.29051/el.v9i00.17208
| Submetido em: 19/09/2022
| Revisões requeridas em: 12/01/2023
| Aprovado em: 07/03/2023
| Publicado em: 27/04/2023
Editora:
Profa. Dra. Rosangela Sanches da Silveira Gileno
Editor Adjunto Executivo:
Prof. Dr. José Anderson Santos Cruz
1
Faculdade Cultura Inglesa (FCI), São Paulo SP Brasil. Professor dos cursos de licenciatura em língua inglesa
e bacharelado em tradução da língua inglesa. Coordenador de pós-graduação na mesma instituição.
Relações entre a escrita e a fala na primeira língua e sua influência no uso de Inglês como Língua Adicional: Uma reflexão sobre o ensino
de pronúncia
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 2
RESUMO: Este artigo propõe uma discussão acerca das relações grafofonêmicas e
fonografêmicas em português brasileiro, segundo as propostas de Capovilla e Casado (2014a,
2014b), para a avaliação da produção ortográfica (na escrita) e falada (na leitura em voz alta)
de aprendizes brasileiros de inglês como língua adicional. Mais especificamente, é apresentada,
por meio de uma reflexão que coteja a teoria supracitada, uma proposta de entendimento sobre
as paralexias e paragrafias possivelmente encontradas na produção linguística de aprendizes de
inglês. Com base nas relações entre a escrita e a fala (relações grafofonêmicas) e entre a fala e
a escrita (relações fonografêmicas) do português brasileiro, discutem-se casos em que relações
consagradas nesta língua podem afetar a produção da fala e da escrita de aprendizes de inglês.
PALAVRAS-CHAVE: Escrita ortográfica. Leitura. Inglês como LA
RESUMEN: Este artículo propone una discusión sobre las relaciones grafofonémicas y
fonografémicas en portugués brasileño, según las propuestas de Capovilla y Casado (2014a,
2014b), para la evaluación de la producción ortográfica (escrita) y hablada (en lectura en voz
alta) de aprendices brasileños de inglés como un idioma adicional. Más específicamente, se
presenta, a través de una reflexión que compara la teoría mencionada, una propuesta para la
comprensión de las paralexias y parálisis que posiblemente se encuentren en la producción
lingüística de los aprendices de inglés. Basado en las relaciones entre la escritura y el habla
(relaciones grafofonímicas) y entre el habla y la escritura (relaciones fonografémicas) del
portugués brasileño, se discuten casos en que las relaciones consagradas en esta lengua
pueden afectar la producción del habla y la escritura de los aprendices de inglés.
PALABRAS CLAVE: Escritura ortográfica. Lectura. Inglés como LA.
ABSTRACT: This article discusses the graphophonemic and phonographemic relations in
Brazilian Portuguese, according to Capovilla and Casado (2014a, 2014b), to evaluate the
written and spoken production of Brazilian learners of English as an additional language. More
specifically, it is presented, through a reflection that compares the theory above, a way of
understanding the paralexia and paragraphs possibly found in the linguistic production of
learners of English. Based on the relationships between writing and speech (graphophonemic
relationships) and between speech and writing (phonographemic relationships) in Brazilian
Portuguese, we discuss cases in which the relationships established in the first language can
affect the production of speech and writing by learners of English
KEYWORDS: Spelling. Reading. English as an additional language.
Rodrigo Garcia ROSA
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 3
Introdução
O ensino de línguas modernas é reconhecidamente uma área multifacetada com práticas
plurais, que se valem de estudos de áreas irmãs como a educação, a linguística, a psicologia e
a sociologia. Isso se deve, em parte, à busca constante dos que se dedicam a esse ofício por
modelos de ensino mais eficientes e que se mostrem mais adequados à realidade social, cultural
e psicológica dos aprendizes dessas línguas. Dentre esses modelos, destacam-se os estudos de
aprendizagem de línguas adicionais
2
, que têm por objetivo explicitar os mecanismos
(conscientes ou não) que levam aprendizes a formarem regras, formularem hipóteses e a criarem
padrões da língua a qual estão sendo expostos (GASS; BEHNEY; PLONSKY, 2013).
No entanto, é importante salientar que, se por um lado, os estudos de aprendizagem de
uma língua adicional (doravante LA) servem como ótimos subsídios teóricos. Esses estudos
não se justificam apenas à luz da elaboração de modelos pedagógicos, como apontam Gass,
Behney e Plonsky (2013), uma vez que pesquisas dessa natureza integram estudos mais amplos
sobre a linguagem, a cognição e o comportamento linguístico de falantes, sobretudo nos estudos
de orientação cognitiva (ELLIS; CADIERNO, 2009).
Em termos linguísticos, a aprendizagem de uma LA não se limita à investigação de um
único nível linguístico. Ao contrário, os estudos nessa área de investigação concentram-se,
igualmente, na aprendizagem dos sistemas fonológico, morfológico, lexical, sintático,
semântico, sociolinguístico e pragmático pelos falantes.
Neste artigo, propõe-se uma discussão com base nos conceitos de cifrabilidade e
decifrabilidade
3
do português brasileiro, segundo as propostas de Capovilla e Casado (2014a,
2014b), para a avaliação da produção ortográfica (na escrita da LA) e falada (na leitura em voz
alta na LA) de aprendizes de inglês como LA. Para tal, discutem-se as relações entre a escrita
ortográfica e a leitura em voz alta na língua portuguesa sob os dois ângulos; ou seja, da escrita
ortográfica para a leitura em voz alta (relação grafofonêmica) e da compreensão auditiva da
fala para a escrita ortográfica (relação fonografêmica). Após esta apresentação, discute-se o
2
O termo língua adicional, em substituição à expressão tradicional língua estrangeira, é aqui adotado com o
intuito inclusivo de neutralizar os traços semânticos de estranheza, exterioridade e distanciamento que a palavra
estrangeira pode evocar. Partilha-se, portanto, da ideia de que língua adicional pode ser encarada como um
hiperônimo mais neutro e abrangente do que termos como língua estrangeira ou segunda língua (LÔPO RAMOS,
2021).
3
Segundo Capovilla e Casado (2014a, 2014b), os conceitos de cifrabilidade e decifrabilidade, referem-se
respectivamente, aos processos de transformação da fala em grafemas, isto é, o que ocorre na escrita (cifrabilidade)
bem como ao processo de transformação da escrita em fala, isto é, o que se faz na leitura em voz alta
(decifrabilidade).
Relações entre a escrita e a fala na primeira língua e sua influência no uso de Inglês como Língua Adicional: Uma reflexão sobre o ensino
de pronúncia
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 4
nível de transferência das relações grafofonêmicas e fonografêmicas da língua portuguesa para
a LA e reflete-se sobre como tais relações podem gerar paralexias (leitura incorreta de palavras)
e paragrafias (escrita ortográfica incorreta) em inglês como LA. Dessa forma, o artigo,
sobretudo, propõe uma análise de uma característica linguística da primeira língua e os efeitos
que tais características podem ter na produção de inglês como LA.
O artigo apresenta uma breve análise da literatura produzida por Fernando Capovilla e
seus colaboradores quando da elaboração do modelo de cálculo de cifrabilidade e
decifrabilidade do português brasileiro, através de um levantamento dos principais passos
metodológicos que levaram os pesquisadores à conclusão do projeto. Segue-se a isso uma
discussão crítica das possibilidades de utilização dos principais resultados desse projeto na
avaliação da produção escrita e oral de aprendizes de inglês como LA. Por fim, apresentam-se
as considerações finais aventadas à luz da reflexão feita ao longo deste texto.
A necessidade da quebra do código do português brasileiro
Os processos de leitura e escrita aplicados a alfabetizandos de diferentes faixas etárias,
geralmente, usam técnicas de avaliação sob ditado como recurso para verificar o estado das
rotas fonológicas e lexicais dos indivíduos, a depender da fase de desenvolvimento em que se
encontram (ELLIS; YOUNG, 1988 apud CAPOVILLA; CASADO, 2014a, 2014b). No entanto,
os ditados, que em si servem bem aos propósitos a que se prestam, carecem de um material de
insumo mais apropriado que as atuais listas de palavras.
Segundo Capovilla (2015), as listas de palavras utilizadas não constituem paradigmas
sistemáticos, pois as palavras, tipificadas simplesmente como regulares e irregulares, por
exemplo, apresentam características e relações intercambiáveis, sobretudo no âmbito das
correspondências entre sons e grafemas. Segundo Callou e Leite (2003)
no primeiro estágio do processo de alfabetização, a criança (ou adulto) vai
aprender que as letras servem de símbolo para os segmentos fônicos da língua.
A dificuldade inicial é que a hipótese de biunivocidade letra-fone é limitada e
varia de dialeto para dialeto. Um exame de erros de ortografia na escola
deveria constituir um método valioso de investigação para o linguista, pois
esses erros refletem geralmente uma falta de correspondência entre o sistema
de fonemas e o sistema de grafemas (CALLOU; LEITE, 2003, p. 46-47).
À luz das características apresentadas acima, Capovilla e Casado (2014a, 2014b)
propõem a substituição das listas de palavras assistemáticas comuns por tabelas geradas a partir
Rodrigo Garcia ROSA
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 5
de um paradigma matricial em que uma variável nominal de tipo, como regular ou irregular,
cede o seu lugar para uma variável intervalar de grau médio de decifrabilidade grafofonêmica,
ou seja, na correspondência entre grafemas da língua escrita e fonemas da fala, e outra variável
de grau médio de cifrabilidade fonografêmica, ou seja, na correspondência entre os fonemas da
língua falada e os grafemas da escrita. Capovilla e Casado (2014a, 2014b) desenvolveram um
programa, que calcula o grau de cifrabilidade fonografêmica e decifrabilidade grafofonêmica
do português brasileiro, a fim de contribuir para a problemática levantada por Callou e Leite
(2003).
A seguir, apresentaremos de forma sucinta as principais etapas percorridas no desenho
do modelo, no que diz respeito tanto ao conceito de decifrabilidade, quanto ao conceito de
cifrabilidade.
A decifrabilidade grafofonêmica da palavra escrita
A decifrabilidade grafofonêmica da palavra escrita tem como objetivo explicar o
processamento de leitura em voz alta e prever paralexias, ou seja, erros de pronúncia durante a
leitura em voz alta, que surgem das relações estabelecidas entre grafemas (unidades de escrita)
e fonemas (unidades fonológicas da língua). Para chegar a um modelo com tal capacidade
explicativa, Capovilla propõe os seguintes passos:
(i) Verificação das possibilidades de realização fonético-fonológicas de diferentes
grafemas por meio de leitura em voz alta. Por exemplo, o grafema <x> pode
corresponder, em português, aos fonemas /ʃ/ como em ‘axé’, /s/ como em ‘extensão’,
/z/ como em ‘exato’, mas também /gz/ como em ‘oxácido’ (CAPOVILLA, 2015);
(ii) Verificação das diferentes possibilidades de realização ortográfica de diferentes
fonemas por meio de atividades de escrita sob ditado auditivo;
(iii) Definição da frequência absoluta e relativa de cada relação grafofonêmica;
(iv) Definição do índice ponderado de decifrabilidade da relação entre cada grafema e
suas possíveis realizações fonético-fonológicas.
O objetivo que se tem na definição da frequência absoluta e relativa de cada relação
grafofonêmica (item iii) é o de se estabelecer dois tipos de relações, isto é:
Relações entre a escrita e a fala na primeira língua e sua influência no uso de Inglês como Língua Adicional: Uma reflexão sobre o ensino
de pronúncia
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 6
a) Relações onde o fonema mais frequente, portanto, canônico da relação, será um
candidato provável a usurpar o grafema de uma palavra menos frequente em uma
atividade de leitura em voz alta (exemplo: oxácido ser pronunciado com a fricativa
pós-alveolar surda /ʃ/ para o grafema <x> ([o'ʃasido]) ao invés de [o'gzasido] por
conta da relação canônica existente entre o grafema <x> e a fricativa /ʃ/);
b) Relações menos frequentes e menos canônicas nos quais o fonema da relação se liga
‘excepcionalmente’ com o grafema em questão. Nesses casos, uma força
estatística que, em atividades de leitura em voz alta, pode agir para dar espaço para
um fonema mais frequente tomar o lugar do excepcional. Trata-se da relação
descrita em (a) ao contrário.
Estabelecida a decifrabilidade grafofonêmica, resta ao modelo fazer o processo inverso,
isto é, definir a cifrabilidade fonografêmica.
A cifrabilidade fonografêmica da palavra falada
A cifrabilidade fonografêmica da palavra falada tem como objetivo explicar o
processamento de escrita sob ditado e prever paragrafias, isto é, erros na escrita em situações
de ditados, que surgem das relações estabelecidas na ordem fonema-grafema. Para chegar a um
modelo com tal capacidade explicativa, Capovilla (2015) propõe os seguintes passos:
(i) Verificação das possibilidades de realização ortográfica de diferentes fonemas por
meio de escrita sob ditado. Por exemplo, o fonema /s/ pode ser grafado com <c>,
como em ‘cena’, <s> como em ‘cais’, <z> como em ‘paz’, <ç> como em ‘caça’,
<sc> como em ‘descer’ e <ss> como em ‘passar’;
(ii) Verificação das diferentes possibilidades de realização fonético-fonológica de
diferentes grafemas através de atividades de leitura em voz alta;
(iii) Definição da frequência absoluta e relativa de cada relação fonografêmica;
(iv) Definição do índice ponderado de cifrabilidade da relação entre cada fonema e suas
possíveis realizações ortográficas.
O objetivo que se tem na definição da frequência absoluta e relativa de cada relação
fonografêmica (item iii) é o de se estabelecer dois tipos de relações, isto é:
Rodrigo Garcia ROSA
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 7
a) Relações onde o grafema mais frequente, portanto, canônico da relação, será um
candidato provável a usurpar o fonema de uma palavra menos frequente em uma
atividade de escrita sob ditado (exemplo: [ɐ'bojs]) ser grafado com <s> para o
fonema /s/ (abois) ao invés de aboiz por conta da relação canônica existente entre o
fonema /s/ e o grafema <s>;
b) Relações menos frequentes e menos canônicas onde o grafema da relação se liga
‘excepcionalmente’ com o fonema em questão. Nesses casos, uma força
estatística que, em atividades de escrita sob ditado, pode agir para dar espaço para
um grafema mais frequente tomar o lugar do excepcional. Trata-se da relação
descrita em (a) ao contrário.
A seção seguinte apresenta uma análise sobre como o modelo de (de)cifrabilidade do
português pode, potencialmente, auxiliar em questões relativas a paralexias e paragrafias em
aprendizes brasileiros de inglês como LA.
A (de)cifrabilidade grafofonêmica e fonografêmica e o ensino de inglês como LA
Como foi mencionado na introdução, uma parte dos estudos sobre aprendizagem de
LA se concentra nos mecanismos que levam aprendizes de línguas a formar hipóteses e
estabelecer padrões linguísticos mentais em vários níveis de análise linguística. Dentre esses
mecanismos, é bastante difundido na literatura especializada (ELLIS; ROBINSON, 2008;
ELLIS; CADIERNO, 2009; ELLIS, 2013; GASS; BEHNEY; PLONSKY, 2013) o papel
fundamental de certas construções da primeira ngua em como se formarão algumas
construções análogas em LA. Em outras palavras, o aprendiz, sobretudo o iniciante, fará
transferências diretas e/ou relacionadas de sua primeira língua para a língua em aprendizagem
em todos os níveis de análise, ou seja, no nível fonético-fonológico, morfológico, sintático,
semântico e pragmático. Especificamente sobre o nível fonético-fonológico, temos que
depois de uma certa idade, as referências sonoras de uma pessoa são
totalmente conectadas com sua língua nativa. É quase como se a língua nativa
funcionasse como um filtro para quaisquer sons estrangeiros. Caso o som seja
idêntico, não problemas. No entanto, se o som for diferente, esse será
Relações entre a escrita e a fala na primeira língua e sua influência no uso de Inglês como Língua Adicional: Uma reflexão sobre o ensino
de pronúncia
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 8
reinterpretado de acordo com o sistema sonoro da língua nativa (GODOY;
GONTOW; MARCELINO, 2006, p. 18, tradução nossa).
4
Apesar de o papel da transferência ser amplamente discutido, inclusive no âmbito da
fonética e fonologia (cf. ECKMAN, 2012 para um panorama histórico dos estudos em aquisição
de fonologia de LAs), o se pode dizer o mesmo sobre o efeito exercido pelas relações entre
os sistemas de escrita e fala da primeira língua sobre o sistema de escrita ou mesmo sobre a
pronúncia de uma LA. Em particular, pouco se discute como as relações de correspondência
entre os sistemas de escrita e sistemas fonológicos das primeiras línguas dos aprendizes podem
influenciar a aprendizagem de escrita e pronúncia de uma LA. A reflexão que se propõe aqui
se insere nessa lacuna e, para tal, utiliza-se do confronto entre as línguas portuguesa e inglesa.
Português e inglês: Duas línguas de tipologias grafofonêmica e fonografêmica distintas
As diferenças entre os níveis de opacidade grafofonêmica e fonografêmica entre línguas
como o inglês e o português são evidentes até mesmo para os aprendizes mais experientes de
inglês, como LA. Se o português, por um lado, apresenta uma relação mais direta entre os
grafemas e seus respectivos fonemas, o mesmo não pode ser dito de línguas como o inglês
5
, em
que sequências gráficas nem sempre têm uma relação de um para um com os fonemas (CELCE-
MURCIA; BRINTON; GOODWIN, 2010).
A ortografia da língua inglesa é notoriamente imprevisível. Por exemplo, a
vogal /i:/ pode ser representada na escrita de variadas maneiras. Todas as letras
sublinhadas nas palavras que se seguem representam /i:/ me, see, sea, believe,
receive, pizza, people, key, quay, quiche, Portuguese, foetus. A maioria dos
outros fonemas podem ser grafados de modos variados, em especial as vogais
(CARLEY; MEES; COLLINS, 2018, p. 03, tradução nossa).
6
4
Original: After a certain age, a person’s sound references are totally connected to the native language. It is almost
as if the native language acted as a filter for any foreign sound. If the sound is identical, there is no problem. If the
sound is different, it is reinterpreted according to the native sound system.
5
Capovilla (c.p.) chama atenção para que, comparativamente, os alfabetizandos de português brasileiro estão em
uma posição de vantagem em relação aqueles cujas primeiras línguas é inglês ou francês, pois essas duas línguas
são reconhecidamente mais opacas no que se refere às suas relações grafofonêmicas. No entanto, se comparada a
línguas como o alemão, o italiano ou o espanhol, a língua portuguesa torna-se, em termos relativos, mais opaca
nas suas relações grafofonêmicas, pois nas línguas mencionadas, a relação entre grafemas e fonemas é muito mais
direta e objetiva, havendo pouca variação.
6
Original: English orthography (i.e. spelling) is notoriously unreliable. For instance, the vowel /i:/ can be spelt
in numerous ways. All the letters underlined in the following words represent /i:/ me, see, sea, believe, receive,
pizza, people, key, quay, quiche, Portuguese, foetus. Most other phonemes can be spelt in many different ways,
especially vowels.
Rodrigo Garcia ROSA
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 9
Como exemplo adicional no âmbito do sistema consonantal, peguemos a africada, pós-
alveolar, sonora [dʒ] que representa o som dos seguintes grafemas grifados modular, gentle,
jitters adjust, edge. Além disso, como apontado por Carley, Mees e Collins (2018), o mesmo é
válido para vogais, que, além dos processos fonológicos de assimilação de traços de nasalização
ou modificação por retroflexos e laterais (vowel coloring, como em bird e label), exibem uma
relação grafofonêmica muito mais opaca do que a existente em português brasileiro.
Se os alfabetizandos de inglês como primeira língua têm de lidar com um sistema de
escrita pouco transparente, no que diz respeito às correspondências entre escrita e fala, quando
pensamos em aprendizes brasileiros de inglês como LA, à opacidade existente na língua inglesa
somam-se as incongruências estatísticas grafofonêmicas e fonografêmicas inerentes ao
português. Em outras palavras, há uma clara transferência de padrões grafofonêmicos e
fonografêmicos da primeira língua para a LA. A esse respeito, diz Thaïs Cristófaro-Silva:
Argumento que a construção do sistema sonoro da língua estrangeira é
baseada, primordialmente, no sistema sonoro da língua materna com
interferência direta deste. No caso do(a) falante brasileiro que aprende inglês
como língua estrangeira, ele(a) deve ter um referencial sólido do sistema
sonoro do português (CRISTÓFARO-SILVA, 2015, p. 10).
Dessa forma, podemos concluir que a aprendizagem dos sistemas sonoro e gráfico da
LA tem como base, ainda que somente no início do processo de aprendizagem, o conhecimento
acerca dos mesmos aspectos linguísticos da primeira língua. Rafat (2016) demonstra, com
dados experimentais de falantes de inglês sem conhecimento prévio de espanhol, a importância
da (in)congruência entre os sistemas ortográficos e sonoros da primeira língua na produção da
LA. Os dados mostraram que a falta de ligação entre fonemas e grafemas na primeira língua
resultou em um número maior de paralexias na produção de LA.
Sendo assim, se a primeira língua apresentar incongruências de correspondência entre
os sistemas ortográficos e sonoros com diferentes forças de frequência para cada uma das
relações, pode-se esperar que os aprendizes transfiram tais relações, sobretudo aquelas mais
consagradas em sua língua, quando do uso da LA.
Considere por que línguas estrangeiras são difíceis de compreender quando as
ouvimos. Mesmo se conseguimos lê-las, os sons de suas letras e combinações
de letras podem ser diferentes dos sons correspondentes às mesmas letras e às
combinações de letras em nossa ngua nativa [...] a razão é que fatores
cognitivos e contextuais influenciam nossa percepção do sinal recebido. Por
exemplo, o efeito de restauração fonêmica envolve integrar o que conhecemos
com aquilo que ouvimos quando percebemos a fala (STERNBERG, 2016, p.
312-313).
Relações entre a escrita e a fala na primeira língua e sua influência no uso de Inglês como Língua Adicional: Uma reflexão sobre o ensino
de pronúncia
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 10
Podemos concluir, portanto, que o conhecimento das relações grafofonêmicas e
fonografêmicas na primeira língua pode influenciar a aprendizagem de uma LA, sobretudo em
contextos de ensino onde a aprendizagem se dá por meio da intermediação de textos escritos na
LA, como aponta Steinberg (2006).
Ler a nossa própria língua resulta ler segundo a nossa pronúncia individual.
Ler uma língua que não a nossa pode trazer uma série de problemas, pelo
menos no estágio inicial do aprendizado. O aprendizado em geral se por
meio de um texto. Um texto vem escrito num sistema […] cujos símbolos
representam fonemas, sílabas ou morfemas, respectivamente […] Quando
duas línguas, isto é, a materna e a que se deseja aprender, empregam o mesmo
sistema, então nos deparamos com problemas de interferência. No caso do
português e do inglês, as duas línguas empregam o sistema alfabético latino.
Esse sistema tenta representar os fonemas. Os símbolos empregados são
praticamente os mesmos nas duas línguas, só que nem sempre representam os
mesmos fonemas. Poderíamos dizer que os códigos são diferentes. A mesma
letra não representa o mesmo fonema e este não é representado pela mesma
letra em cada uma das línguas (STEINBERG, 2006, p. 66).
A ideia de transferência de relações, como foi dito, pode ser um obstáculo no processo
de aprendizado do sistema da LA. Contudo, o cenário pode ser agravado se pensarmos mais a
fundo nas relações consagradas na primeira língua que podem ser projetadas pelo aprendiz no
sistema fonológico da LA. Essa complicação adicional pode ser compreendida mesmo com um
exemplo em primeira língua. É o que ocorre na relação grafofonêmica em português brasileiro,
exemplificada na palavra oxácido, discutida em Capovilla (2015) ou em outras menos
infrequentes como êxodo, exímio e sintaxe. A ideia central é que, por ser uma palavra de
baixíssima frequência
7
e por ser grafada com uma relação não-canônica entre o grafema <x> e
a sequência fonológica /gz/, a palavra torna-se suscetível a uma relação grafofonêmica mais
frequente, isto é, com <x> sendo pronunciado /ʃ/, gerando assim, a paralexia [o'ʃasido] ou, de
modo similar, em palavras como execrável, eximir e exumar que, diante de suas evidentes
baixas frequências ou desconhecimento por parte dos alfabetizandos, poderiam abrir espaço
para a relação <x> e /ʃ/, nesse caso intrusiva e responsável pela paralexia. Contudo, é crucial
notar que a interferência e consequente transferência da primeira língua para a LA depende de
diversos fatores, além das puramente grafofonêmicas e grafofonêmicas, como o contexto
fonológico, ser ou não cognato, nível de proficiência, conhecimento específico da palavra, bem
como diferenças individuais do aprendiz (ESCALANTE, 2016).
7
Ocorrência alguma foi encontrada no Corpus do Português www.corpusdoportugues.org
Rodrigo Garcia ROSA
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 11
Se analisarmos um paralelo simples, podemos dizer que, para estudantes iniciantes de
inglês, como LA, as palavras que são usadas com frequência em inglês podem parecer tão
infrequentes quanto oxácido é em sua própria língua. Além disso, por terem pouco
conhecimento na língua que estão aprendendo, parece-nos plausível levantarmos a hipótese
segundo a qual relações grafofonêmicas estatisticamente mais proeminentes na primeira língua
podem exercer uma pressão negativa sobre o sistema da LA, gerando assim paralexias
(HAYES-HARB; CHENG, 2016). Em outras palavras, podem ocorrer paralexias na pronúncia
de uma palavra estrangeira por conta da utilização de estratégias de canonização grafofonêmica
segundo o paradigma da primeira língua. Um exemplo dessa situação pode ser dado com o
grafema <ch> que, em português, está mais frequentemente associado à fricativa /ʃ/ (exemplos:
chá, chuva, chapéu, chop, chave e achar) e em inglês à africada /tʃ/ (exemplos: achieve, change,
chow, each e chives). Dada a disparidade de associações da relação grafofonêmica em ambas
as línguas, não nos parece implausível esperar que palavras como achieve, change, chow, each
e chives possam ser pronunciadas com /ʃ/ por aprendizes iniciantes, por essa ser uma associação
consagrada em português brasileiro.
É fato que a relação entre <ch> e /tʃ/ em inglês não é uma relação de um para um,
podendo o grafema realizar-se como a velar /k/ (exemplos: charisma, chemistry e stomach) ou
como a pós-alveolar /ʃ/ (exemplos: chic, machine, Chicago e Michigan); no entanto, até mesmo
a associação entre <ch> e /k/ não encontra reforço no paradigma do português brasileiro.
Já nos casos em que os aprendizes erroneamente recorrem às relações fonografêmicas,
segundo o paradigma do português brasileiro, essa estratégia pode levá-los ao cometimento de
paragrafias na escrita de uma palavra estrangeira por conta da canonização fonografêmica na
cifragem de um fonema estrangeiro. Esse caso pode ser exemplificado com a labiodental surda
/f/ que, em inglês, se associa frequentemente aos grafemas <f>, <ff> e <ph>, mas também com
o dígrafo <gh>. A baixa frequência da associação entre /f/ e <ph> ou <gh> em português, por
exemplo, pode levar aprendizes a produzirem paragrafias como as primeiras palavras na
sequência de pares a seguir: *fase/phase, *frase/phrase, *fisician/physician, *graf/graph,
*touf/tough e *enouf/enough na produção escrita em inglês.
Outro caso de relações fonografêmicas intrusivas de primeira língua que podem exercer
uma pressão negativa refere-se a fonemas existentes na LA e que inexistem no sistema
fonológica da primeira ngua. Esse é o caso do linguodental /θ/ do inglês (exemplos: thanks,
thin, bath e oath) que não figura no sistema fonológico do português brasileiro e que é
frequentemente produzido por falantes iniciantes com fonemas substitutos que partilham com
Relações entre a escrita e a fala na primeira língua e sua influência no uso de Inglês como Língua Adicional: Uma reflexão sobre o ensino
de pronúncia
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 12
/θ/ algum traço fonológico, tais como o ponto de articulação, caso da oclusiva dental /t/, o modo
de articulação, caso da fricativa labiodental /f/ ou ainda a fricativa alveolar /s/ (GODOY;
GONTOW; MARCELINO, 2006).
Em suma, podemos também analisar, sob a mesma perspetiva, casos em que, mesmo
que as relações grafofonêmicas e fonografêmicas sejam similares entre a primeira língua e a
LA, elas podem causar paralexias e paragrafias na produção da LA por conta da aplicação de
regras fonêmicas inerentes à primeira língua que não se mostram iguais na LA. Essa é a regra
conhecida durante a alfabetização do português onde o grafema <s> intervocálico sofre um
processo de vozeamento, sendo pronunciado com a fricativa sonora /z/, como em caso, básico,
asa e sica. Tal relação em português reforça as paralexias em LA (exemplo: case */keɪz/,
house */haʊz/, basic */ˈbeɪzɪk/ e vase */veɪz/) que, por sua vez, recebe reforço em dados da
própria LA como em asset, essay e bossy em que o dígrafo <ss> é pronunciado como /s/ e que
pode gerar a falsa conclusão de que as regras na LA se assemelham com as do português.
Considerações finais
Neste breve artigo reflexivo, apresentamos a proposta de Capovilla e Casado (2014a,
2014b) para a (de)cifrabilidade grafofonêmica e fonografêmica do português brasileiro e
analisamos o efeito que essas características linguísticas da primeira ngua podem ter na
produção de uma LA, levando-se em consideração os casos de transferência. A proposta foi
avaliada a partir dos principais passos metodológicos de Capovilla e Casado (2014a, 2014b)
para que possamos compreender melhor o processo que tornou o modelo possível. Por fim,
propusemos uma reflexão sobre a utilização de um modelo para informar questões referentes à
aprendizagem de uma LA, sobretudo do inglês, por falantes nativos de português brasileiro. Os
exemplos listados e discutidos com mais detalhes são capazes de demonstrar que as relações
grafofonêmicas e fonografêmicas do sistema do português brasileiro podem exercer uma
pressão negativa sobre a aprendizagem do sistema fonológico do inglês, assim como podem
levar alunos iniciantes a cometerem paragrafias e paralexias na LA.
Contudo, é importante salientar que a discussão sobre os exemplos apresentados no
contraste entre o português brasileiro e o inglês não termina com esta reflexão. Ao contrário,
tais exemplos requerem uma intervenção experimental que os coloque sob uma análise empírica
para confirmar ou refutar as visões aqui apresentadas que, por ora, mostram-se ainda como
hipóteses a serem mais profundamente exploradas em trabalhos futuros.
Rodrigo Garcia ROSA
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 13
REFERÊNCIAS
CALLOU, D.; LEITE, Y. Iniciação à fonética e à fonologia. Rio de Janeiro: Zahar, 2003.
CAPOVILLA, F. C.; CASADO, K. Quebrando o código do português brasileiro: Como
calcular a cifrabilidade de qualquer palavra falada e a decifrabilidade de qualquer palavra
escrita. São Paulo: Memnon, 2014a.
CAPOVILLA, F. C.; CASADO, K. Voz brasileira na Nova Ortografia: As Vozes das
Letras: CD-ROM e Manual. São Paulo: Memnon, 2014b.
CAPOVILLA, F. C. Nossa Língua Portuguesa: Como avaliar o grau de cifrabilidade de
qualquer palavra falada e o grau de decifrabilidade de qualquer palavra escrita nas normas
lusitana e brasileira. In: PEREIRA, R. (org.). Abordagem Multidisciplinar da
Aprendizagem. 1. ed. Viseu, Portugal: Qual Consoante, 2015. v. 1, p. 405-504
CARLEY, P.; MEES, I. M.; COLLINS, B. English Phonetics and pronunciation practice.
Nova York: Routledge, 2018.
CELCE-MURCIA, M.; BRINTON, D. M.; GOODWIN, J. M. Teaching Pronunciation: A
course book and reference guide. Nova York: Cambridge University Press, 2010.
CRISTÓFARO-SILVA, T. Pronúncia do inglês para falantes do português brasileiro. São
Paulo: Contexto, 2015.
ECKMAN, F. R. Second Language Phonology. In: GASS, S. M.; MACKEY, A. (org.). The
Routledge handbook of second language acquisition. Nova York: Routledge, 2012.
ELLIS, N. C.; ROBINSON, P. An Introduction to Cognitive Linguistics, Second Language
Acquisition, and Language Instruction. In: ROBINSON, P.; ELLIS, N. C. (org.). A
handbook of cognitive linguistics and SLA. London: Routledge, 2008.
ELLIS, N. C.; CADIERNO, T. (org.). Constructing a second language. In: Annual Review of
Cognitive Linguistics. Amsterdam: John Benjamins Publishing Company 2009. p. 111-290.
ELLIS, N. C. Second language acquisition. In: TROUSDALE, G.; HOFFMANN, T. (org.).
Oxford Handbook of Construction Grammar. Oxford: Oxford University Press, 2013.
ESCALANTE, C. Intervocalic Voicing and Regressive Voicing Assimilation in L2 Spanish
/s/. Divergencias. Revista de estudios lingsticos y literarios, Arizona, v. 14, n. 1, 2016.
Disponível em: https://divergencias.arizona.edu/2022/10/26/volumen-14-numero-1-
primavera-2016/.Acesso em: 16 jun. 2021.
GASS, S. M.; BEHNEY, J.; PLONSKY, L. Second language acquisition: An Introductory
Course. 4 ed. Nova York: Routledge, 2013.
GODOY, S. M. B.; GONTOW, C.; MARCELINO, M. English pronunciation for
Brazilians. São Paulo: Disal, 2006.
Relações entre a escrita e a fala na primeira língua e sua influência no uso de Inglês como Língua Adicional: Uma reflexão sobre o ensino
de pronúncia
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 14
HAYES-HARB, R.; CHENG, H. W. The influence of the Pinyin and Zhuyin writing systems
on the acquisition of Mandarin word forms by native English speakers. Frontiers in
Psychology, Switzerland, v. 7, 2016. DOI:10.3389/fpsyg.2016.00785. Disponível em:
https://www.frontiersin.org/articles/10.3389/fpsyg.2016.00785/full. Acesso em: 16 jun. 2021.
LÔPO RAMOS, A. A. Língua adicional: um conceito “guarda-chuva”. Revista Brasileira De
Linguística Antropológica, Brasília, v. 13, n. 1, p. 233-267, 2021. DOI:
10.26512/rbla.v13i01.37207. Disponível em:
https://periodicos.unb.br/index.php/ling/article/view/37207. Acesso em: 16 jun. 2021.
RAFT, Y. Orthography-induced transfer in the production of English-speaking learners of
Spanish. Language Learning Journal, London, v. 44, p. 197-213, 2016.
DOI:10.1080/09571736.2013.784346. Disponível em:
https://www.tandfonline.com/doi/abs/10.1080/09571736.2013.784346?journalCode=rllj20.
Acesso em: 16 jun. 2021.
STEINBERG, M. Inglês Norte-Americano: pronúncia e morfologia. São Paulo: Nova
Alexandria, 2006.
STERNBERG, R. Psicologia cognitiva. São Paulo: Cengage learning, 2016.
CRediT Author Statement
Reconhecimentos: Agradeço aos pareceristas deste periódico pelos comentários.
Financiamento: Não há.
Conflitos de interesse: Não há.
Aprovação ética: O trabalho não passou por nenhum comitê de ética.
Disponibilidade de dados e material: Os dados não estão disponíveis para acesso.
Processamento e editoração: Editora Ibero-Americana de Educação.
Revisão, formatação, normalização e tradução.
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 1
THE RELATIONSHIP BETWEEN WRITTEN AND SPOKEN FIRST LANGUAGE
AND ITS INFLUENCE ON THE USE OF ENGLISH AS AN ADITIONAL
LANGUAGE: A REFLECTION ABOUT THE TEACHING OF PRONUNCIATION
RELAÇÕES ENTRE A ESCRITA E A FALA NA PRIMEIRA LÍNGUA E SUA
INFLUÊNCIA NO USO DE INGLÊS COMO LÍNGUA ADICIONAL: UMA REFLEXÃO
SOBRE O ENSINO DE PRONÚNCIA
RELACIONES ENTRE ESCRIBIR Y HABLAR EN LA PRIMERA LENGUA Y SU
INFLUENCIA EN EL USO DEL INGLÉS COMO LENGUA ADICIONAL: UNA
REFLEXIÓN SOBRE LA ENSEÑANZA DE LA PRONUNCIACIÓN
Rodrigo Garcia ROSA1
e-mail: rodrigo_21usp@yahoo.com.br
How to reference this paper:
ROSA, R. G. The relationship between written and spoken first
language and its influence on the use of English as an Aditional
Language: A reflection about the teaching of pronunciation.
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-
ISSN: 2447-3529. DOI:
https://doi.org/10.29051/el.v9i00.17208
| Submitted: 19/09/2022
| Revisions required: 12/01/2023
| Approved: 07/03/2023
| Published: 27/04/2023
Editor:
Profa. Dra. Rosangela Sanches da Silveira Gileno
Executive Deputy Editor:
Prof. Dr. José Anderson Santos Cruz
1
Cultura Inglesa College (FCI), São Paulo SP Brazil. Professor of English language undergraduate courses
and English language translation undergraduate courses. Coordinator of graduate studies at the same institution.
The relationship between written and spoken first language and its influence on the use of English as an Aditional Language: A reflection
about the teaching of pronunciation
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 2
ABSTRACT: This article discusses the graphophonemic and phonographemic relations in
Brazilian Portuguese, according to Capovilla and Casado (2014a, 2014b), to evaluate the
written and spoken production of Brazilian learners of English as an additional language. More
specifically, it is presented, through a reflection that compares the theory above, a way of
understanding the paralexia and paragraphs possibly found in the linguistic production of
learners of English. Based on the relationships between writing and speech (graphophonemic
relationships) and between speech and writing (phonographemic relationships) in Brazilian
Portuguese, we discuss cases in which the relationships established in the first language can
affect the production of speech and writing by learners of English
KEYWORDS: Spelling. Reading. English as an additional language.
RESUMO: Este artigo propõe uma discussão acerca das relações grafofonêmicas e
fonografêmicas em português brasileiro, segundo as propostas de Capovilla e Casado (2014a,
2014b), para a avaliação da produção ortográfica (na escrita) e falada (na leitura em voz alta)
de aprendizes brasileiros de inglês como língua adicional. Mais especificamente, é
apresentada, por meio de uma reflexão que coteja a teoria supracitada, uma proposta de
entendimento sobre as paralexias e paragrafias possivelmente encontradas na produção
linguística de aprendizes de inglês. Com base nas relações entre a escrita e a fala (relações
grafofonêmicas) e entre a fala e a escrita (relações fonografêmicas) do português brasileiro,
discutem-se casos em que relações consagradas nesta língua podem afetar a produção da fala
e da escrita de aprendizes de inglês.
PALAVRAS-CHAVE: Escrita ortográfica. Leitura. Inglês como LA
RESUMEN: Este artículo propone una discusión sobre las relaciones grafofonémicas y
fonografémicas en portugués brasileño, según las propuestas de Capovilla y Casado (2014a,
2014b), para la evaluación de la producción ortográfica (escrita) y hablada (en lectura en voz
alta) de aprendices brasileños de inglés como un idioma adicional. Más específicamente, se
presenta, a través de una reflexión que compara la teoría mencionada, una propuesta para la
comprensión de las paralexias y parálisis que posiblemente se encuentren en la producción
lingüística de los aprendices de inglés. Basado en las relaciones entre la escritura y el habla
(relaciones grafofonímicas) y entre el habla y la escritura (relaciones fonografémicas) del
portugués brasileño, se discuten casos en que las relaciones consagradas en esta lengua
pueden afectar la producción del habla y la escritura de los aprendices de inglés.
PALABRAS CLAVE: Escritura ortográfica. Lectura. Inglés como LA.
Rodrigo Garcia ROSA
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 3
Introduction
Modern language teaching is a multifaceted area with plural practices that draw on
studies from sister areas such as education, linguistics, psychology, and sociology. This is due,
in part, to the constant search for more efficient teaching models that are better suited to the
social, cultural, and psychological realities of modern language learners. Among these models
are the studies of additional language learning
2
, which aim to explain the mechanisms
(conscious or not) that lead learners to form rules, formulate hypotheses, and create patterns of
the language to they are being exposed (GASS; BEHNEY; PLONSKY, 2013).
However, it is essential to note that, on the one hand, studies of learning an additional
language (henceforth LA) serve as significant theoretical subsidies. Moreover, these studies are
not only justified in light of the development of pedagogical models, as Gass, Behney, and
Plonsky (2013) point out, since research of this nature integrates broader studies on language,
cognition, and linguistic behavior of speakers, especially in cognitively oriented studies
(ELLIS; CADIERNO, 2009).
In linguistic terms, learning LA is not limited to the investigation of a single linguistic
level. Instead, studies in this area of research focus equally on speakers' learning of the
phonological, morphological, lexical, syntactic, semantic, sociolinguistic, and pragmatic
systems.
In this article, we propose a discussion based on the concepts of cipherability and
decipherability
3
of Brazilian Portuguese, according to Capovilla and Casado's (2014a, 2014b)
proposals, for the evaluation of orthographic production (in LA writing) and spoken production
(in LA reading aloud) of learners of English as LA. To this end, the relations between
orthographic writing and reading aloud in LA are discussed from two angles, i.e., from
orthographic writing to reading aloud (graphophonemics relation) and from auditory speech
comprehension to orthographic writing (phonographemic relation). After this presentation, we
discuss the level of transfer of graphophone mic and phonographemic connections from
Portuguese to LA and reflect on how such relations can generate paralexies (incorrect word
2
The term additional language, replacing the traditional term foreign language, is adopted here to neutralize the
semantic traits of strangeness, exteriority, and detachment that the word foreign can evoke. We share that
additional language can be considered a more neutral and comprehensive hyperonym than foreign or second-
language terms. (LÔPO RAMOS, 2021, our translation).
3
According to Capovilla and Casado (2014a, 2014b), the concepts of cipherability and decipherability refer
respectively to the processes of transformation of speech into graphemes, that is, what occurs in writing
(cipherability) as well as to the process of change of writing into discourse, that is, what is done in reading aloud
(decipherability).
The relationship between written and spoken first language and its influence on the use of English as an Aditional Language: A reflection
about the teaching of pronunciation
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 4
reading) and paragraphs (incorrect spelling) in English as LA. Thus, the article mainly proposes
an analysis of a linguistic feature of the first language and the effects that such elements may
have on the production of English as LA.
The paper presents a brief analysis of the literature produced by Fernando Capovilla and
his collaborators when elaborating on the model for calculating the cipherability and
decipherability of Brazilian Portuguese through a survey of the main methodological steps that
led the researchers to the conclusion of the project. This is followed by a critical discussion of
the possibilities of using the main results of this project in evaluating the written and oral
production of learners of English as a foreign language. Finally, we present our final
considerations in light of the reflections made throughout this text.
The necessity of breaking the Brazilian Portuguese code
Reading and writing processes applied to literacy learners of different age groups
generally use assessment techniques under dictation as a resource to check the state of
individuals' phonological and lexical routes, depending on their developmental stage (ELLIS;
YOUNG, 1988 apud CAPOVILLA; CASADO, 2014a, 2014b). However, dictations serve their
purpose well but lack a more appropriate input material than current word lists.
According to Capovilla (2015), the word lists used do not constitute systematic
paradigms because words typified simply as regular and irregular, for example, have
interchangeable characteristics and relationships, especially in the realm of correspondences
between sounds and graphemes. According to Callou and Leite (2003)
In the first stage of the literacy process, the child (or adult) will learn that
letters serve as symbols for the phonic segments of the language. The initial
difficulty is that the assumption of letter-phoneme biunivocity is limited and
varies from dialect to dialect. Therefore, examining spelling errors in school
should be a valuable method of investigation for the linguist because these
errors generally reflect a mismatch between the phoneme and grapheme
systems (CALLOU; LEITE, 2003, p. 46-47, our translation).
In light of the characteristics presented above, Capovilla and Casado (2014a, 2014b)
propose replacing the shared asystematic word lists with tables generated from a matrix
paradigm in which a nominal variable of type, such as regular or irregular, gives way to an
interval variable of medium degree of graphophonemics decipherability, that is, in the
correspondence between graphemes of the written language and phonemes of speech, and
Rodrigo Garcia ROSA
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 5
another variable of medium degree of phonographemic decipherability, that is, in the
correspondence between phonemes of the spoken language and graphemes of writing.
Capovilla and Casado (2014a, 2014b) developed a program that calculates the degree of
phonographemic cipherability and graphophonemics decipherability of Brazilian Portuguese to
contribute to the problem raised by Callou and Leite (2003).
Below, we will briefly present the main steps in the model's design concerning both the
concept of decipherability and the concept of cipherability.
The graphophonemic decipherability of the written word
Graphophonemic decipherability of the written word aims to explain read-aloud
processing and predict paralexies, that is, pronunciation errors during read-aloud, which arise
from the relationships established between graphemes (units of writing) and phonemes
(phonological units of language). To arrive at a model with such explanatory capacity,
Capovilla proposes the following steps:
(i) Checking the possibilities of phonetic-phonological realization of different
graphemes by reading aloud. For example, the grapheme <x> can correspond, in
Portuguese, to the phonemes /ʃ/ as in ‘axe’, /s/ as in ‘extensão’
4
, /z/ as in exact’, but
also /gz/ as in ‘oxácido’
5
(CAPOVILLA, 2015);
(ii) Verification of the different possibilities of orthographic realization of different
phonemes through writing activities under auditory dictation;
(iii) Definition of the absolute and relative frequency of each graphophonemic relation;
(iv) Definition of the weighted decipherability index of the relation between each
grapheme and its possible phonetic-phonological realizations
The goal one has in defining the absolute and relative frequency of each
graphophonemic relation (item iii) is to establish two types of relations, that is:
a) Relationships where the most frequent, and therefore canonical, phoneme of the
relationship will be a likely candidate to usurp the grapheme of a less frequent word
4
Meaning: extension.
5
Meaning: oxyacid.
The relationship between written and spoken first language and its influence on the use of English as an Aditional Language: A reflection
about the teaching of pronunciation
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 6
in a read-aloud activity (example: oxyacid be pronounced with the post-alveolar
deaf fricative /ʃ/ for the grapheme <x> ([oc'ʃasid]) instead of [oc'gzasid] because of
the canonical relationship between the grapheme <x> and the fricative /ʃ/);
b) Less frequent and less canonical relations in which the phoneme of the relation links
'exceptionally' with the grapheme in question. In these cases, there is a statistical
force that, in read-aloud activities, can make room for a more frequent phoneme to
replace the exceptional one. This is the relationship described in (a) in reverse.
Once graphophonemic decipherability has been established, it remains for the model to
do the reverse process, that is, to define phonographemic decipherability.
The phonographemic cipherability of the spoken word
The phonographemic cipherability of the spoken word aims to explain the processing of
writing under dictation and to predict paragraphs, that is, errors in writing in dictation situations,
which arise from the relations established in the phoneme-grapheme order. To arrive at a model
with such explanatory capacity, Capovilla (2015) proposes the following steps:
(i) Verify the possibilities of orthographic realization of different phonemes by writing
under dictation. For example, the phoneme /s/ can be spelled with <c>, with in cell
phone’, <s> with in cars’, <z> with in alcatraz, <ç> with in ‘caça’
6
, <sc> with in
sceneand <ss> with in kisses’;
(ii) Verification of the different possibilities of phonetic-phonological realization of
different graphemes through reading-aloud activities;
(iii) Definition of the absolute and relative frequency of each phonographemic relation;
(iv) Definition of the weighted cipherability index of the relation between each phoneme
and its possible orthographic realizations.
The goal in defining the absolute and relative frequency of each phonographemic
relation (item iii) is to establish two types of relations:
6
Meaning: hunt.
Rodrigo Garcia ROSA
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 7
a) Relationships where the most frequent, and therefore canonical, grapheme of the
relationship will be a likely candidate to usurp the phoneme of a less frequent word
in a dictated writing activity (example: [ɐ'bojs] be spelled with <s> for the phoneme
/s/ (abois) instead of aboiz because of the canonical relationship existing between
the phoneme /s/ and the grapheme <s>;
b) Less frequent and less canonical relations where the grapheme of the relation links
'exceptionally' with the phoneme in question. In these cases, there is a statistical
force that, in dictated writing activities, can act to make room for a more frequent
grapheme to take the place of the exceptional one. This is the relationship described
in (a) in reverse
The next section presents an analysis of how the Portuguese (de)cipherability model can
potentially help with issues related to paralexies and paragraphs in Brazilian learners of English
as LA
Graphophonemic and phonographemic (de)cipherability and the teaching of English as
LA
As mentioned in the introduction, part of the studies on learning English as a foreign
language focus on the mechanisms that lead language learners to form hypotheses and establish
mental linguistic patterns at various levels of linguistic analysis. Among these mechanisms, it
is already widespread in the specialized literature (ELLIS; ROBINSON, 2008; ELLIS;
CADIERNO, 2009; ELLIS, 2013; GASS; BEHNEY; PLONSKY, 2013) the fundamental role
of certain first language constructions in how some analogous constructions will be formed in
LA. In other words, the learner, especially the beginner, will make direct and related transfers
from their first language to the language being learned at all levels of analysis, that is, at the
phonetic-phonological, morphological, syntactic, semantic, and pragmatic levels. Specifically
on the phonetic-phonological level, we have that
After a certain age, a person's sound references are connected to their native
language. It is almost as if the native language is a filter for foreign sounds. If
the sound is identical, there are no problems. However, if the sound is
different, it will be reinterpreted according to the sound system of the native
The relationship between written and spoken first language and its influence on the use of English as an Aditional Language: A reflection
about the teaching of pronunciation
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 8
language (GODOY; GONTOW; MARCELINO, 2006, p. 18, our
translation).
7
Although the role of transfer is widely discussed, including in the realm of phonetics
and phonology (cf. ECKMAN, 2012 for a historical overview of studies in LA phonology
acquisition), the same cannot be said about the effect exerted by the relations between the
writing and speech systems of the first language on the writing system or even on the
pronunciation of an LA. In particular, there is little discussion of how the correspondence
relations between the writing and phonological systems of learners' first languages can
influence the learning of writing and pronunciation of an LA. The reflection proposed here is
inserted in this gap, and, for such, it is based on the confrontation between the Portuguese and
English languages.
Portuguese and English: Two languages with different graphophonemic and
phonographemic typologies
The differences between the graphophonemic and phonographemic opacity levels
between languages like English and Portuguese are evident even for the most experienced
learners of English like LA. While Portuguese has a more direct relationship between
graphemes and their respective phonemes, the same cannot be said of languages like English
8
,
where graphical sequences do not always have a one-to-one relationship with phonemes
(CELCE-MURCIA; BRINTON; GOODWIN, 2010).
The spelling of the English language is notoriously unpredictable. For
example, the vowel /i:/ can be represented in writing in various ways. All the
underlined letters in the following words represent /i:/ me, see, sea, believe,
receive, pizza, people, key, quay, quiche, Portuguese, foetus. Most other
phonemes can be spelled in different ways, especially vowels (CARLEY;
MEES; COLLINS, 2018, p. 03, our translation).
9
7
Original: After a certain age, a person’s sound references are connected to the native language. It is almost as if
the native language acted as a filter for any foreign sound. If the sound is identical, there is no problem. If the
sound is different, it is reinterpreted according to the native sound system.
8
Capovilla (c.p.) draws attention to the fact that, comparatively, Brazilian Portuguese literate learners are in an
advantageous position compared to those whose first languages are English or French because these two languages
are known to be more opaque in their graphophonemic relations. However, when compared to languages such as
German, Italian, or Spanish, the Portuguese language becomes, in relative terms, more opaque in its
graphophonemic relations because in the languages above, the relationship between graphemes and phonemes is
much more direct and objective, and there is little variation.
9
Original: English orthography (i.e., spelling) is notoriously unreliable. For instance, the vowel /i:/ can be spelled
in numerous ways. All the letters underlined in the following words represent /i:/ me, see, sea, believe, receive,
pizza, people, key, quay, quiche, Portuguese, foetus. Most other phonemes can be spelled in many different ways,
especially vowels.
Rodrigo Garcia ROSA
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 9
As a further example within the consonantal system, let us take the Africanized, post-
alveolar sound [dʒ] that represents the sound of the following graphemes grapheme modular,
gentle, jitters adjust, edge. Moreover, as pointed out by Carley, Mees, and Collins (2018), the
same is true for vowels, which, in addition to the phonological processes of assimilation of
nasalization features or modification by retroflexes and laterals (vowel coloring, as in bird and
label), exhibit a graphophonemic relationship that is much more opaque than that existing in
Brazilian Portuguese.
If literate learners of English as a first language have to deal with a writing system that
is not very transparent in terms of the correspondence between writing and speech, when we
think of Brazilian learners of English as a second language, the opacity of the English language
is compounded by the graphophonemic and phonographemic statistical incongruities inherent
in Portuguese. In other words, graphophonemic and phonographemic patterns are clearly
transferred from the first language to the LA. In this regard, says Thaïs Cristófaro-Silva:
The construction of the sound system of the foreign language is based
primarily on the sound system of the mother tongue with direct interference
from it. In the case of the Brazilian speaker who is learning English as a
foreign language, they should have a solid reference to the Portuguese sound
system (CRISTÓFARO-SILVA, 2015, p. 10).
Thus, learning the sound and graphic systems of HL is based, even if only at the
beginning of the learning process, on knowledge about the same linguistic aspects of the first
language. Rafat (2016) demonstrates, with experimental data from English speakers without
prior knowledge of Spanish, the importance of (in)congruity between the orthographic and
sound systems of the first language in LA production. The data showed that the lack of
connection between phonemes and graphemes in the first language resulted in more paralexies
in LA production.
Thus, if the first language presents mismatches of correspondence between orthographic
and sound systems with different frequency strengths for each of the relations, it can be
expected that learners will transfer such relations, especially those more established in their
language when using the TL.
Consider why foreign languages are difficult to understand when we hear them. Even
if we can read them, the sounds of their letters and letter combinations may be
different from the sounds corresponding to the same letters and letter combinations in
our native language [...] the reason is that cognitive and contextual factors influence
The relationship between written and spoken first language and its influence on the use of English as an Aditional Language: A reflection
about the teaching of pronunciation
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 10
our perception of the received signal. For example, the phonemic restoration effect
involves integrating what we know with what we hear when we perceive speech
(STERNBERG, 2016, p. 312-313, our translation)
We can conclude that knowledge of graphophonemic and phonographemic relations in
the first language can influence the learning of an LA, especially in teaching contexts where
learning takes place through the intermediation of written texts in the LA, as Steinberg (2006)
points out.
Reading our language results in reading according to our pronunciation.
Reading a language other than your own can cause many problems, at least in
the early stages of learning. Learning takes place through a text. A text is
written in a system [...] whose symbols represent phonemes, syllables, or
morphemes respectively [...] when two languages, i.e., the mother tongue and
the language to be learned, use the same system, then we run into interference
problems. In the case of Portuguese and English, both languages employ the
Latin alphabetic system. This system attempts to represent phonemes. The
symbols used are almost the same in both languages but don't always represent
the same phonemes. The codes are different. The same letter does not
represent the same phoneme, which is not defined by the same letter in each
language (STEINBERG, 2006, p. 66).
As already mentioned, the idea of relation transfer can be an obstacle in learning the LA
system. However, the scenario can be aggravated if we think deeper about the relations
enshrined in the first language that can be projected by the learner into the LA phonological
system. This additional complication can be understood even with a first-language example.
This happens in the graphophonemic relationship in Brazilian Portuguese, exemplified in the
word oxyacid, discussed in Capovilla (2015), or other less infrequent ones such as exodus,
eximious, and syntax. The central idea is that, for being a word of very low frequency
10
and
because it is spelled with a non-canonical relationship between the grapheme <x> and the
phonological sequence /gz/, the word becomes susceptible to a more frequent graphophonemic
relationship, that is, with <x> being pronounced /ʃ/, thus generating paralexies [oc'ʃasid] or,
similarly, in words like execrable, exempt and exhume which, because of their evident low
frequencies or lack of knowledge by the learners, could open space for the relation <x> and /ʃ/,
in this case, intrusive and responsible for the paralexies. However, it is crucial to note that the
interference and consequent transfer from the first language to the TL depends on several
factors other than purely graphophonemic ones, such as phonological context, being cognate or
10
No such occurrence was found in the Corpus do Português www.corpusdoportugues.org
Rodrigo Garcia ROSA
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 11
not, proficiency level, word-specific knowledge, as well as individual learner differences
(ESCALANTE, 2016).
If we look at a simple parallel, we can say that for beginning English learners like LA,
words used frequently in English may seem as infrequent as oxyacid is in their language.
Moreover, because they have little knowledge of the language they are learning, we can
hypothesize that graphophonemic relations that are statistically more prominent in the first
language may exert negative pressure on the LA system, thus generating paralexies (HAYES-
HARB; CHENG, 2016). In other words, paralexies can occur in pronouncing a foreign word
because of the use of graphophonemic canonization strategies according to the first language
paradigm. An example of this situation can be given with the grapheme <ch> that, in
Portuguese, is most often associated with the fricative / ʃ / (examples: chá
11
, chuva
12
, chapéu
13
,
chop, chave
14
and achar
15
) and in English to africative /tʃ/ (examples: achieve, change, chow,
each e chives). Given the disparity of graphophonemic relation associations in both languages,
it does not seem implausible to expect that words like achieve, change, chow, each and chives
can be pronounced with /ʃ/ by novice learners since this is an established association in
Brazilian Portuguese
It is a fact that the relationship between <ch> and /tʃ/ in English is not a one-to-one
relationship, the grapheme can be realized as the velar /k/ (examples: charisma, chemistry and
stomach) or as a post-alveolar /ʃ/ (examples: chic, machine, Chicago and Michigan); However,
even the association between <ch> and /k/ does not find reinforcement in the Brazilian
Portuguese paradigm.
In cases where learners erroneously resort to phonographemic relations, according to
the Brazilian Portuguese paradigm, this strategy may lead them to commit paragraphs in writing
a foreign word because of phonographemic canonization in the scrambling of a foreign
phoneme. This case can be exemplified with the deaf labiodental /f/, which, in English, is
frequently associated with the graphemes <f>, <ff>, and <ph>, but also with the digit <gh>.
The low frequency of association between /f/ and <ph> or <gh> in Portuguese, for example,
may lead learners to produce paragraphs as the first words in the following sequence of pairs:
11
Meaning: tea.
12
Meaning: rain.
13
Meaning: hat.
14
Meaning: keys.
15
Meaning: found.
The relationship between written and spoken first language and its influence on the use of English as an Aditional Language: A reflection
about the teaching of pronunciation
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 12
*fase/phase, *frase/phrase, *fisician/physician, *graf/graph, *touf/tough and *enouf/enough in
written production in English.
Another case of intrusive first-language phonogram relations that can exert negative
pressure refers to phonemes that exist in the TL but do not exist in the phonological system of
the first language. This is the case of the English linguodental /θ/ (examples: thanks, thin, bath
and oath) which does not figure in the Brazilian Portuguese phonological system and is often
produced by beginners with substitute phonemes that share with /θ/ some phonological trait,
such as the point of articulation, like the dental occlusive /t/, the mode of articulation, like the
labial fricative /f/ or the alveolar fricative /s/ (GODOY; GONTOW; MARCELINO, 2006).
In short, we can also analyze, from the same perspective, cases in which, even if
graphophonemic and phonographemic relations are similar between the first language and the
TL, they may cause paralexies and paragraphs in the TL production due to the application of
phonemic rules inherent to the first language that are not the same in the TL. Such is the rule
known during Portuguese alphabetization, where the intervocalic grapheme <s> undergoes a
voicing process and is pronounced with the voiced fricative /z/, as in case, basic, loose e music.
Such a relationship in Portuguese reinforces the paralexies in LA (exemplo: case */keɪz/, house
*/haʊz/, basic */ˈbeɪzɪk/ e vase */veɪz/) which, in turn, receives reinforcement in data from LA
itself, as in asset, essay and bossy in which the digit <ss> is pronounced as /s/, which may lead
to the false conclusion that the rules in LA are similar to those in Portuguese.
Final considerations
In this brief reflective paper, we present Capovilla and Casado's (2014a, 2014b)
proposal for graphophonemic and phonographemic ((de)cipherability of Brazilian Portuguese
and analyze the effect that these linguistic features of the first language can have on the
production of a TL, taking transfer cases into account. The proposal was evaluated from the
main methodological steps of Capovilla and Casado (2014a, 2014b) to understand better the
process that made the model possible. Finally, we proposed a reflection on using a model to
inform issues concerning the learning of an LA, especially English, by native speakers of
Brazilian Portuguese. The examples listed and discussed in more detail can demonstrate that
the graphophonemic and phonographemic relations of the Brazilian Portuguese system can
exert negative pressure on learning the English phonological system and lead beginning learners
to commit paragraphs and paralexies in LA.
Rodrigo Garcia ROSA
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 13
However, it is essential to note that the discussion about the examples presented in
contrast between Brazilian Portuguese and English does not end with this reflection. On the
contrary, such models require an experimental intervention that puts them under an empirical
analysis to confirm or refute the views presented here, which, for now, are still hypotheses to
be further explored in future works.
REFERENCES
CALLOU, D.; LEITE, Y. Iniciação à fonética e à fonologia. Rio de Janeiro: Zahar, 2003.
CAPOVILLA, F. C.; CASADO, K. Quebrando o código do português brasileiro: Como
calcular a cifrabilidade de qualquer palavra falada e a decifrabilidade de qualquer palavra
escrita. São Paulo: Memnon, 2014a.
CAPOVILLA, F. C.; CASADO, K. Voz brasileira na Nova Ortografia: As Vozes das
Letras: CD-ROM e Manual. São Paulo: Memnon, 2014b.
CAPOVILLA, F. C. Nossa Língua Portuguesa: Como avaliar o grau de cifrabilidade de
qualquer palavra falada e o grau de decifrabilidade de qualquer palavra escrita nas normas
lusitana e brasileira. In: PEREIRA, R. (org.). Abordagem Multidisciplinar da
Aprendizagem. 1. ed. Viseu, Portugal: Qual Consoante, 2015. v. 1, p. 405-504
CARLEY, P.; MEES, I. M.; COLLINS, B. English Phonetics and pronunciation practice.
Nova York: Routledge, 2018.
CELCE-MURCIA, M.; BRINTON, D. M.; GOODWIN, J. M. Teaching Pronunciation: A
course book and reference guide. Nova York: Cambridge University Press, 2010.
CRISTÓFARO-SILVA, T. Pronúncia do inglês para falantes do português brasileiro. São
Paulo: Contexto, 2015.
ECKMAN, F. R. Second Language Phonology. In: GASS, S. M.; MACKEY, A. (org.). The
Routledge handbook of second language acquisition. Nova York: Routledge, 2012.
ELLIS, N. C.; ROBINSON, P. An Introduction to Cognitive Linguistics, Second Language
Acquisition, and Language Instruction. In: ROBINSON, P.; ELLIS, N. C. (org.). A
handbook of cognitive linguistics and SLA. London: Routledge, 2008.
ELLIS, N. C.; CADIERNO, T. (org.). Constructing a second language. In: Annual Review of
Cognitive Linguistics. Amsterdam: John Benjamins Publishing Company 2009. p. 111-290.
ELLIS, N. C. Second language acquisition. In: TROUSDALE, G.; HOFFMANN, T. (org.).
Oxford Handbook of Construction Grammar. Oxford: Oxford University Press, 2013.
ESCALANTE, C. Intervocalic Voicing and Regressive Voicing Assimilation in L2 Spanish
The relationship between written and spoken first language and its influence on the use of English as an Aditional Language: A reflection
about the teaching of pronunciation
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 14
/s/. Divergencias. Revista de estudios lingsticos y literarios, Arizona, v. 14, n. 1, 2016.
Disponível em: https://divergencias.arizona.edu/2022/10/26/volumen-14-numero-1-
primavera-2016/.Acesso em: 16 jun. 2021.
GASS, S. M.; BEHNEY, J.; PLONSKY, L. Second language acquisition: An Introductory
Course. 4 ed. Nova York: Routledge, 2013.
GODOY, S. M. B.; GONTOW, C.; MARCELINO, M. English pronunciation for
Brazilians. São Paulo: Disal, 2006.
HAYES-HARB, R.; CHENG, H. W. The influence of the Pinyin and Zhuyin writing systems
on the acquisition of Mandarin word forms by native English speakers. Frontiers in
Psychology, Switzerland, v. 7, 2016. DOI:10.3389/fpsyg.2016.00785. Available:
https://www.frontiersin.org/articles/10.3389/fpsyg.2016.00785/full. Access: 16 June 2021.
LÔPO RAMOS, A. A. Língua adicional: um conceito “guarda-chuva”. Revista Brasileira De
Linguística Antropológica, Brasília, v. 13, n. 1, p. 233-267, 2021. DOI:
10.26512/rbla.v13i01.37207. Available:
https://periodicos.unb.br/index.php/ling/article/view/37207. Access: 16 June 2021.
RAFT, Y. Orthography-induced transfer in the production of English-speaking learners of
Spanish. Language Learning Journal, London, v. 44, p. 197-213, 2016.
DOI:10.1080/09571736.2013.784346. Available:
https://www.tandfonline.com/doi/abs/10.1080/09571736.2013.784346?journalCode=rllj20.
Access: 16 June 2021.
STEINBERG, M. Inglês Norte-Americano: pronúncia e morfologia. São Paulo: Nova
Alexandria, 2006.
STERNBERG, R. Psicologia cognitiva. São Paulo: Cengage learning, 2016.
Rodrigo Garcia ROSA
Rev. EntreLinguas, Araraquara, v. 9, n. 00, e023011, 2023. e-ISSN: 2447-3529
DOI: https://doi.org/10.29051/el.v9i00.17208 15
CRediT Author Statement
Acknowledgements: I thank the referees of this journal for their comments.
Funding: None.
Conflicts of interest: None.
Ethical approval: The work was not submitted to any ethics committee.
Data and material availability: Data is not available.
Processing and editing: Editora Ibero-Americana de Educação.
Proofreading, formatting, normalization and translation