Abstract:
Conventional reading models, such as the dual-route-model or the connectionist model assume that recoding is the same in all orthographies. Empirical results, however, show that recoding is influenced by characteristics of orthography. In particular, the consistency of an orthography seems to play a central role in recoding. There are only few theories or models that take these results into account: the orthographic-depth-hypothesis (Katz & Frost, 1992), the psycholinguistic grain-size-theory (Ziegler & Goswami, 2005) and the two-cycles-model (Berent & Perfetti, 1995) which takes into account the inconsistency of English orthography. The empirical evidence behind these theories and models is different, but the results suggest unanimously that native English speakers use different (additional) recoding strategies to German native speakers. According to the grain-size-theory (Ziegler & Goswami, 2005) German readers recode new words using a grapheme-phoneme-mapping, whereas English readers also have to take into account larger units, such as onset and rime. Using onsets and rimes, they can form analogies to new words. The two-cycles-model (Berent & Perfetti, 1995) assumes that English natives initially only process the consonants of a word and integrate the vowels only in a second cycle that is slower and more controlled than the first.
Such results are not taken into consideration in didactics of English as a second language very often. In general, prior knowledge based on German learners’ experience of learning their own language is excluded from the English language instruction. Moreover, relevant recoding strategies are not topics in English instruction, nor used in English language training, even though preliminary applications and didactical concepts exist.
In light of this, the question arises, which strategies German learners of English use to recode English words. It is not clear whether they continue using the strategy they used for German – serial recoding – or if they learn new strategies, such as recoding using analogies or using the consonants for orientation.
These questions are explored in two studies. The first study focuses on recoding according to the grain-size-theory, the second study tests the two-cycles-model with German learners of English. For this purpose, students of grades 5, 7 and 9 of a Gymnasium (Grammar school) and students of a university were asked to read English and German pseudowords and words. The recoding strategies were identified based on reading speed and accuracy. According to the grain-size-theory, switching costs when reading pseudowords with and without orthographic neighbours in mixed lists (compared to blocked lists), should emerge when readers switch between recoding strategies. The word-length-effect, which occurs with serial recoding, can also be used as an indicator for the recoding strategy being employed. Results show that more experienced German learners of English take into account larger units when reading English pseudowords than less experienced learners (the word-length-effect is reduced with more experience) whereas small units were used to recode German pseudowords (word-length-effect is not affected by experience level). Switching costs are not reduced with higher experience levels as expected, but students in grades 7 and 9 grade read pseudowords more easily both with, and without, orthographic neighbours, when they are presented in mixed lists rather than in blocked lists. These results suggest, that students with little experience in English learned to recode using analogies, but initially had difficulties in choosing the relevant strategy and suppressing the other one. This ability emerges with more experience in English.
The two-cycles-model was tested using a priming-paradigma. If consonants are processed first and faster than vowels, rapidly presented information about the consonants in target words should facilitate reading. Primes with the same vowels as the target should only facilitate reading when presented for a relatively long period. According to the results of this study, German learners of English do not process consonants and vowels in different cycles. The results suggest that in a first step the graphemes are translated into German phonemes – in general only the consonants of these phonemes can be used for the English pronunciation.
Consequently, the didactics of English as a second language (for German learners), should to take into account the prior knowledge about reading in German, and use it as basis for teaching recoding strategies in English. Teachers of English as a second language should be made aware of these different recoding strategies in German and English.