Orthography and rhyme judgement: the interplay between phonology and orthography on rhyme judgement

Orthography and Rhyme Judgement 1 Reference: Castro, S. L., Caló, S., & Gomes, I. (2003). The interplay between phonology and orthography on rhyme judgement. Psychologica, 34, 143-154. Running head: ORTHOGRAPHY AND RHYME JUDGEMENT The Interplay Between Phonology and Orthography on Rhyme Judgement São Luís Castro and Susana Caló Universidade do Porto, Porto, Portugal Inês Gomes Universidade Fernando Pessoa, Porto, Portugal Address for correspondence: São Luís Castro FPCE - Universidade do Porto Rua do Campo Alegre, 1021 P 4169 - 004 Porto, Portugal Tel +351 22 607 9756, Fax +351 22 607 9725 E-mail: [email protected] Orthography and Rhyme Judgement 2 English: The interplay between phonology and orthography on rhyme judgement Abstract A Word Rhyme Judgements task for neuropsychological assessment in European Portuguese is presented. The task was modeled from Kay, Lesser & Coltheart (1992). It requires a yes/no decision on whether word pairs rhyme or not. Half of the pairs rhyme; half of these have the same final spelling (CEDO/DEDO), half have different final spellings (POÇO/FOSSO). Control pairs, half of which share the final spelling, do not rhyme (MEDO/CREDO; MOÇO/NOSSO). Thirty-two college students performed this task in the spoken and written modalities (order counterbalanced). Clear effects of modality, in favor of the spoken presentation, and of orthography, in favor of same spelling rimes, even in the spoken presentation, were observed. The findings are consistent with the assumption that orthography interacts with phonology even when the task does not require explicit attention to orthography. The usefulness of the materials and task for the assessment of language is upraised. Key words: Rhyme Judgement, Orthography, European Portuguese, Language, Neuropsychological Assessment Orthography and Rhyme Judgement 3 Português: Ouve, rima ou não? Ortografia e fonologia no julgamento de rima Resumo Apresentamos aqui uma tarefa de Julgamento de Rimas em Palavras para avaliação neuropsicológica em português. A tarefa foi elaborada a partir da sua homónima de Kay, Lesser & Coltheart (1992), e consiste em decidir se um par de palavras rima ou não. Metade dos pares de palavras rimam; metade das rimas escrevem-se da mesma maneira (CEDO/DEDO), outra metade de maneira diferente (POÇO/FOSSO). Pares de controlo, metade dos quais têm a mesma sequência ortográfica no final, não rimam (MEDO/CREDO; MOÇO/NOSSO). Trinta e dois estudantes universitários realizaram esta tarefa na modalidade falada e na modalidade escrita (ordem contrabalanceada). Observaram-se efeitos claros de modalidade, em favor da apresentação falada, e de ortografia, em favor das rimas escritas da mesma maneira. Estes resultados são consistentes com a noção de que a ortografia interage com a fonologia mesmo quando a tarefa não requer atenção explícita dirigida à ortografia. A utilidade destes materiais e desta tarefa para a avaliação da linguagem é enfatizada. Palavras chave: Julgamento de Rima, Ortografia, Português Europeu, Linguagem, Avaliação Neuropsicológica Orthography and Rhyme Judgement 4 Français: Phonologie et orthographie déterminent le jugement de rime Résumé Nous présentons ici une tâche de Jugement de Rime des Mots pour l'évaluation neuropsychologique en portugais. La tâche a été préparée d'après Kay, Lesser & Coltheart (1992), et consiste à décider si une paire de mots rime ou pas. La moitié des paires rime; la moitié de ces rimes a la même séquence orthographique (CEDO/DEDO), l'autre moitié a une séquence orthographique différente (POÇO/FOSSO). Les paires de contrôle, la moitié desquels a la même séquence orthographique finale, ne riment pas (MEDO/CREDO; MOÇO/NOSSO). Trente-deux étudiants universitaires ont réalisé cette tâche dans la modalité auditive et dans la modalité écrite. Des effets clairs de modalité, en faveur de la présentation parlée, et de l'orthographie, en faveur des rimes de séquence orthographique identique, ont été observés. Ces résultats sont consistants avec la notion théorique d'après laquelle l'orthographie interagit avec la phonologie même quand la tâche ne demande pas l'attention explicite dirigée vers l'orthographie. L'utilité de cette tâche pour l'évaluation neuropsychologique du langage est soulignée. Mots clés: Jugement de Rime, Orthographie, Portugais Européen, Langage, Évaluation Neuropsychologique Orthography and Rhyme Judgement 5 Introduction The neuropsychological assessment of language poses special problems because language is highly abstract and, at the same time, highly specific. In Portuguese, for example, there are two words for language, linguagem, the general mental faculty, and língua, the concrete form it takes when it is spoken or written. Because languages, línguas, differ markedly in the forms they take (phonemes, words, and sentences), assessment has to address the fundamental aspects that are known to be relevant across languages, but it must also be tailored to the specific phonological, orthographic, and morphosyntactic aspects, among others, of the language system being assessed. Some years ago, Kay, Lesser and Coltheart (1992) developed an interesting tool for the assessment of language. The Psycholinguistic Assessments of Language Processing in Aphasia, or PALPA, consist of 60 tests designed to address specific aspects of language, from the auditory processing of nonword minimal pairs to the comprehension of written sentences. Language is viewed as a complex system composed by partly autonomous components or modules, and each task is designed to provide information about the integrity of these modules. Assessment is based on a hypothesis testing model in which the neuropsychologist probes the deficits and their causes by using individually tailored groups of tests, according to the observed pattern of impaired and preserved functioning of the various language components. Because it is modular and finely tuned, the PALPA battery is a useful resource for clinical assessments as well as for research on language. Originally developed in English, it was then adapted for the Spanish language (Valle & Cuetos, 1995). Some of its tests have also been adapted for Portuguese, and were used to monitor changes in, for example, letter knowledge and sentence comprehension in early through middle childhood, as well as differences between highly schooled and less schooled adults (Gomes & Castro, Orthography and Rhyme Judgement 6 1996, 1998, 1999, in press). In this paper, we present materials and results for a rhyme judgement task in Portuguese (European variant), that was developed on the basis on the Word Rhyme Judgements task from the PALPA battery (Test 15 from Kay et al., ibd., pp. 1-6). The Word Rhyme Judgements task is designed to test phonological short-term storage, input processing abilities and segmentation skills (ibd., p. 1). It uses pairs of rhyming words that are spelled either differently, like BEAR/CHAIR, or in the same way, BONE/CONE (phonological vs. orthographic rhyme, respectively). Each rhyming pair has a corresponding control that does not rhyme and that is spelled either differently, e.g., FLAIR/YEAR, or the same, e.g., TONE/GONE (phonological vs. orthographic controls, respectively). The word pairs may be presented in written (the Written Version of the task) or auditorily (the Auditory Version). An effect of orthography is quite likely in the written version, since the visual similarity or dissimilarity of the words in each pair may help, or interfere with, the rhyme judgement. In the spoken modality, a naive expectation would be that the incongruity in spelling would not play a role, since the task requires attention to the sounds of the words and these are heard instead of being seen as letter sequences. Kay et al. (ibd.) do not give results for this task (for most of the other PALPA tests, means and standard deviations for small groups of healthy participants are presented). However, other studies in English have shown that orthographic knowledge influences auditory word recognition. For example, it is easier to detect that two words rhyme when they have the same orthographic endings, even if the task is performed in the spoken modality (Seidenberg & Tannenhaus, 1979). In auditory lexical decision tasks, words whose endings that can be written in more than one way produce slower responses, and more errors, than words whose endings can be written in only one way Orthography and Rhyme Judgement 7 (Ziegler & Ferrand, 1998). Thus, one would expect to find an effect of orthography on the auditory version of the rhyme judgment task even if the phonological short-term storage systems are intact. If, however, these systems are damaged, judgements on rhyme would probably rely more on the visual representation of the words, and the effect of orthography would be stronger. Thus, the size of the orthography effect can be an index of the integrity of the phonological short-term storage systems and/or of the robustness of phonological representations. In the Spanish version of the PALPA battery, the rhyme jugdment task (Juicios de Rima de Palabras: Version Auditiva y Escrita, Valle & Cuetos, 1995, p. 177) comprises only three conditions because all of the non-rhyming pairs are spelled differently (e.g., LOSA/LISA). Furthermore, of the 20 rhyming word pairs, only 5 differ in spelling (e.g., UVA/TUBA; an example of the remaining pairs is CHAL/REAL). The distinction between orthographic and phonological rhymes is lost because Spanish is a shallow orthography: if two words rhyme, they are also spelled in the same way. Though both are Romance languages, Portuguese and Spanish have important differences on phonology and orthography. The Portuguese vowel system has more vowel types, for example, and the Portuguese orthography is classified as intermediate in the opaque vs. transparent dimension (Castro & Gomes, 2000, p.133; Pinheiro, 1994; Seymour, Aro, & Erskine, 2003). Print to sound matching is typically predictable on the basis of contextual rules and fixed correspondences, but matching in the reverse direction - from sound to print - is less straightforward (for a fuller account, see Gomes, 2001; Castro & Gomes, in preparation). In Portuguese, thus, the link between phonology and orthography is less consistent than in Spanish, though more consistent than in English. Does this difference have an impact on rhyme judgement, and on the effect of orthography? The following experiment was carried out with the purpose of examining whether there is an effect of Orthography and Rhyme Judgement 8 orthography on auditory rhyme judgement in Portuguese, an orthography where the mapping between phonology and orthography is only partly consistent. An additional goal was to provide descriptive statistics on a Portuguese version of the Rhyme Judgments Task from the PALPA battery. Method Participants Thirty two right-handed college students (18 female and 14 male) from 20 to 30 years of age (M = 23.6, SD = 2.6) volunteered to take part in the experiment. All were native speakers of European Portuguese with normal or corrected-to-normal vision, and normal hearing according to self-report. The participants were observed during July 2003. Materials The Word Rhyme Judgements task from the PALPA battery (Kay et al., 1992) uses 60 pairs of monosyllabic words, half of which rhyme (the other half does not). Half of the rhyming and of the non-rhyming pairs share the spelling of the rime (say, -OSE in the rhyming POSE/PROSE pair and in the non-rhyming ROSE/LOSE pair). The Portuguese materials were based on the same principle. We selected word quadruplets that share the final orthographic segment such that two of them rhyme and two do not, and such that four groups of word pairs could be built: same spelling rhyming and non-rhyming word pairs (e.g., respectively, CEDO/DEDO and MEDO/CREDO, the latter /-edu/ vs. /-Edu/), and different spelling rhyming and non-rhyming word pairs (e.g., POÇO/FOSSO, and MOÇO/NOSSO, the latter /-osu/ vs. /-Osu/). Bdisyllabic C(C)VCV words were chosen because monosyllables are relatively rare in Portuguese. The number of word pairs was 40 (10 by condition) because Portuguese is more regular than English and thus presents fewer candidate words. The words were selected from the Porlex database (Gomes, Orthography and Rhyme Judgement 9 2001; Gomes & Castro, 2003), and were roughly equated in length, frequency and neighborhood density across the four groups (see Table 1). The complete list of the word pairs is reported in the Appendix. ________________________ Table 1 about here ________________________ These materials were prepared for a computer controlled presentation, which allowed us to measure reaction times in milliseconds as well as accuracy. For the spoken version of the task, the stimuli were digitized, edited and saved in individual Wave files. Words were spoken by a female native speaker of Portuguese (SC) and recorded from a soundproof booth using ProTools LE, version 6.0 (1991 - 2002) at a sampling rate of 48 KHz and 16-bit resolution. The perceptual quality of the recordings was checked by the authors and by a third listener blind to the objectives of the experiment (LC). If rated positively, words were selected for further edition. This involved creating individual files for each word starting with 20 ms of silence before the onset and after the offset of the word. Two lists containing the 40 word pairs in pseudorandom order were prepared, one for the written presentation, and another for the auditory one. The constraints were that no more than three consecutive pairs belonged to the same condition, and that a rhyming pair and its corresponding control were in different halves of the lists. Procedure The presentation of the stimuli and recording of the responses were controlled by Superlab Pro, version 2.0 (2001), run on a Toshiba Satellite 1800-400 portable computer attached to a Cedrus RB610 response box. Participants were tested individually in a quiet, dimly lit, room. They were explained that they would be Orthography and Rhyme Judgement 16 References Castro, S. L., & Gomes, I. (2000). Dificuldades de aprendizagem da língua materna [Language dificulties in children]. Lisboa: Universidade Aberta. Castro, S. L., & Gomes, I. (in preparation). Language-specific aspects of learning how to read and spell: A cognitive analysis of the European Portuguese orthography. Damian, M. F., & Bowers, J. S. (2003). Effects of orthography on speech production in a form-preparation paradigm. Journal of Memory and Language, 49, 119132. Gomes, I. (2001). Ler e escrever em Português Europeu [Reading and writing in European Portuguese]. Unpublished doctoral dissertation, University of Porto, Faculdade de Psicologia e de Ciências da Educação, Portugal. Gomes, I., & Castro, S. L. (1996). Avaliação do processamento da linguagem em adultos pouco escolarizados [The assessment of language processing in lowschooled adults]. In L. S. Almeida, S. Araújo, M. M. Gonçalves, C. Machado, & M. R. Simões (Eds.). Avaliação psicológica: Formas e contextos, vol. 4 (pp. 233-240). Braga: APPORT. Gomes, I., & Castro, S. L. (1998). Language processing and schooling in European Portuguese. In P. Reistma, & L. Verhoeven (Eds.), Problems and interventions in literacy development, (pp. 113-127). Dordrecht: Kluwer Academic Publishers. Gomes, I., & Castro, S. L. (1999). Language processing in Portuguese children aged 4 to 10 years. In M. G. Pinto, J. Veloso, & B. Maia (Eds.), Proceedings of the 5th International Congress of the International Society of Applied Psycholinguistics. Porto: Faculdade de Letras da Universidade do Porto. Orthography and Rhyme Judgement 17 Gomes, I., & Castro, S. L. (2003a). Porlex, a lexical database in European Portuguese. Psychologica, 32, 91-108. Gomes, I., & Castro, S. L. (in press). Emparelhamento palavra/figura: Uma prova para a avaliação da compreensão escrita por crianças e adultos com diferentes graus de educação [Word/picture matching: A task for the assessment of written comprehension by children and adults]. Actas do Congresso em Neurociências Cognitivas. Kay, J., Lesser, R., & Coltheart, M. (1992). Psycholinguistic assessments of language processing in aphasia: PALPA. Hove: Lawrence Erlbaum Associates. Nascimento, M. F. B., Casteleiro, J. M., Marques, M. L. G., Barreto, F., & Amaro, R. (n.d.). Léxico multifuncional computorizado do Português Contemporâneo [Multifunctional computacional lexicon of contemporary Portuguese]. Available: http://www.clul.ul.pt/sectores/projecto/lmcpc.html/ [2002, Dec. 13]. Pinheiro, A. M. V. (1994). Leitura e escrita: Uma abordagem cognitiva. Campinas: Editorial Psy II. Pro Tools LE (version 6.0) (1991-2002) Computer Software. Palo Alto, USA: Digidesign. Seidenberg, M. S., & Tanenhaus, M. K. (1979). Orthographic effects on rhyme monitoring. Journal of Experimental Psychology: Human Learning and Memory, 5, 546-554. Seymour, P.H.K., Aro, M., & Erskine, J.M. (2003). Foundation literacy acquisition in European orthographies. British Journal of Psychology, 94, 143-174. Stone, G. O., Vanhoy, M., & Orden, G. C. V. (1997). Perception is a two-way street: Feedforward and feedback phonology in visual word recognition. Journal of Memory and Language, 36, 337-359. Orthography and Rhyme Judgement 18 Superlab. General purpose psychology testing software (version 2.0) (2001) [Computer Software]. San Pedro, USA: Cedrus Corporation. Valle. F., & Cuetos, F. (1995). Evaluación del procesamiento linguístico en la afasia (EPLA): from Psycholinguistic Assessments of Language Processing in Aphasia (Kay, J., Lesser. R., & Coltheart, M., 1992). Hove: Lawrence Erlbaum Associates. Ziegler, J.C., & Ferrand, L. (1998). Orthography shapes the perception of speech: The consistency effect in auditory word recognition. Psychonomic Bulletin & Review, 5, 683-689. Ziegler, J. C., Tan, H. L., Perry, C., & Montant, M. (2000). Phonology matters: The phonological frequency effect in written Chinese. Psychological Science, 11, 234238. Orthography and Rhyme Judgement 19 Appendix List of Word Pairs, WP, by Condition # Same spelling rhyming WP # Same spelling non-rhyming WP 1 cedo - dedo 1 medo - credo 2 cego - prego 2 lego - grego 3 selo - gelo 3 grelo - belo 4 fera - bera 4 hera - cera 5 festa - sesta 5 testa - cesta 6 foca - toca 6 doca - boca 7 bloco - foco 7 floco - coco 8 moda - soda 8 roda - boda 9 tolo - golo 9 bolo - colo 10 mota - cota 10 lota - gota Different spelling rhyming WP Different spelling non-rhyming WP 1 fato - cacto 1 mato - pacto 2 neto - tecto 2 recto - preto 3 mesa - proeza 3 presa - reza 4 nicho - lixo 4 bicho - fixo 5 prior - teor 5 motor - pior 6 rocha - broxa 6 tocha - coxa 7 poço - fosso 7 moço - nosso 8 coze - dose 8 gnose - doze 9 bruxo - bucho 9 caucho - fluxo 10 supor - propor 10 furor - bolor Note. Practice trials: rhyming fada - nada, globo - lobo, lilás - cartaz, face - classe; non-rhyming comboio - saloio, mulher - lazer. Orthography and Rhyme Judgement 20 Author Note São Luís Castro and Susana Caló, Faculdade de Psicologia e de Ciências da Educação da Universidade do Porto, Porto, Portugal. Inês Gomes, Universidade Fernando Pessoa, Porto, Portugal. This research was supported by a FCT grant to the Center for Psychology at the University of Porto (Language Group). We thank Alexandre Castro Caldas for his interest in our contribution to this special issue. Correspondence concerning this article should be addressed to São Luís Castro, Faculdade de Psicologia e de Ciências da Educação, Universidade do Porto, rua do Campo Alegre, 1021, P 4169 - 004 Porto, Portugal. E-mail: [email protected] Orthography and Rhyme Judgement 21 Table 1. Characteristics of the words averaged for each condition (standard deviation in parentheses) Rhyming Pairs Non-Rhyming Pairs Variables Same Spelling Different Spelling Same Spelling Different Spelling Number of letters 4.20 (0.41) 4.70 (0.66) 4.30 (0.46) 4.70 (0.57) Number of phonemes 4.20 (0.41) 4.40 (0.68) 4.25 (0.54) 4.45 (0.69) Word frequency 379.74 (509.18) 282.89 (495.36) 477.05 (322.50) 464.17 (1038.88) Number of orthographic neighbours 11.25 (4.66) 4.30 (3.70) 9.85 (4.92) 5.0 (3.67) Number of phonological neighbours 8.95 (4.88) 6.90 (7.5) 6.65 (3.70) 4.9 (5.28) Position of uniqueness point (in number of letters) 4.15 (0.37) 4.45 (10.69) 4.25 (0.44) 4.50 (0.61) Position of uniqueness point (in number of phonemes) 3.85 (0.59) 3.90 (0.72) 4.0 (0.56) 4.05 (0.76) Auditory length (in ms) 800 (107) 800 (72) 716 (177) 787 (88) Note. Auditory length in milliseconds measured using ProTools LE (version 6.0). Word frequency collected from the Corlex database (Nascimento, Casteleiro, Marques, Barreto, & Amaro, n.d.). The remaining measures collected from the Porlex database (Gomes, 2001; Gomes & Castro, 2003). Orthography and Rhyme Judgement 22 Figure Captions Figure 1. Mean percentage of correct responses (a) and of response latencies (b) as a function of modality and orthography for rhyming word pairs. Figure 2. Mean percentage of correct responses (a) and of response latencies (b) as a function of modality and orthography for non-rhyming word pairs. Orthography and Rhyme Judgement 23 Figure 1. Correct Responses Yes Rhyme 6 7 8 9 10 Spoken Written Modality n correct Same Ortho Different Ortho Response Latency Yes Rhyme 500 600 700 800 900 1000 1100 1200 Spoken Written Modality RT in ms Same Ortho Different Ortho Orthography and Rhyme Judgement 24 Figure 2. Correct Responses No Rhyme 6 7 8 9 10 Spoken Written Modality n correct Same Ortho Different Ortho Response Latency No Rhyme 500 600 700 800 900 1000 1100 1200 Spoken Written Modality RT in ms Same Ortho Different Ortho