116
language, and in this paper, the records are analysed with particular reference to indications of automaticity.
117
children with little exposure to literacy practices, mediocre instruction and irregular reading practice may not achieve automaticity at all, and therefore fail to become truly proficient.
Automaticity is rarely achieved if learning is begun after adolescence, and adults learning to read a new script report that even though they see all the letters, they need to consciously work out what words they represent, and never reach the same ease of perception of words in scripts they learned before adolescence (Abadzi, 1996). Where adults do achieve automaticity in word recognition in a new script, it appears to be only after decoding the words
approximately 3000 times (Abadzi, 2012).
A
UTOMATICITY IN DIFFERENT LANGUAGES AND ORTHOGRAPHIESLiterature searches yield an overwhelming amount of information relating to reading in the English language, and much less relating to reading in other languages (Pretorius and Mampuru 2007). This tends to facilitate comparisons with English more than with other languages. However, in the context of this paper, comparisons with English might be completely appropriate since isiZulu and English are the most used languages in KwaZulu- Natal (van der Merwe & van der Merwe, 2006) and readers of isiZulu are most likely to read in these two languages. It is doubtful that findings from studies of reading in English can be assumed to apply across languages (Share, 2008), and as shown below, it is possible that readers of English and isiZulu must develop contrasting skills to cope adequately with the demands of each orthography.
The orthography of English may be particularly conducive to the development of
automaticity. In comparison with other languages, English is characterised by an extremely irregular relationship between its phonemes and the letters used to represent them (Borgwaldt et al., 2004), and it is a disjunctive language, whose words are short even in comparison with other European languages: the average word length in newspaper text is 4.6 letters in English, 4.7 in Danish, 4.9 in Swedish, 5.6 in German and 7 in Finnish (Björnsson, 1983).
118
VISUAL HETEROGENEITY
English has a very wide range of permissible syllables and letter combinations, making for a similarly wide range of visual word forms, and hence high visual dissimilarity amongst its words: information from Project Gutenberg, which digitalises English literature, shows that the only three letter string likely to occur more than 10 times in 400 words of English text is
“the”, and that no four letter string is likely to occur more than 3 times (Project Gutenberg, 2013). This indicates a great deal of variation among English syllables and consequently high heterogeneity among its word forms. This variability of form implies high distinguishability among words, and this and its high frequency of single syllable words (Pretorius and
Mokhwesana, 2009), could facilitate the development and exercise of automaticity.
Studies of learning to read in different languages indicate that the degree of syllabic complexity affects the acquisition of reading skills and the development of automaticity.
Unsurprisingly, automatic recognition is most easily established where simple visual patterns represent particular word sounds (Abadzi, 2011). In languages whose orthography features clusters of consonants, and graphemes consisting of more than single letters, it is more difficult to learn letter to sound decoding skills than in languages with less complex phonology (Seymour et al., 2003).
It is obviously easier to distinguish between letters and groups of letters that are visually disparate than those that are similar, and more practice is needed before readers can readily distinguish among similar patterns (Abadzi, 2011). It follows that automaticity is likely to develop more easily in languages whose orthography features strong visual dissimilarity amongst word forms, and with more difficulty in languages where printed words are not obviously dissimilar. This has particular relevance here because, as explained below,
complex agglutinated isiZulu words and some combinations of their components often appear visually similar although they may have no semantic relation.
119
ORTHOGRAPHIC CONSISTENCY AND TRANSPARENCY
It appears to be easier to learn to read in languages with consistent orthographies, where readers can rely on direct letter – sound correspondence. Inconsistent orthographies compel readers to interpret large grain size units of text (Georgiou et al., 2008), including automatic recognition of whole words. Among European languages, the ability to read words
accurately takes longest to develop in English, which has the most opaque orthography amongst them (Seymour, et al., 2003, Georgiou, Parrila et al., 2008). Children learning to read in languages with consistent, transparent orthographies (such as Italian, German, Greek, Spanish and Finnish) can read words accurately at the end of grade one (Ziegler et al., 2010).
In comparison, children reading the orthographically inconsistent and less transparent languages of Portuguese, French and Danish show less accuracy in reading words. Children learning to read in English, with the least transparent and least consistent orthography in Europe, achieve only 34% accuracy at the end of grade one (Ziegler et al., 2010), and lag behind German children in accurately reading words even at age 12, suggesting that it is “the hardest European orthography to acquire” (Hutzler et al., 2004). Comparable information on learning to read in African languages is not yet available.
Children who have learnt to read in languages that are more orthographically consistent than English are better at reading pseudo words than children who have learnt to read in English (Seymour et al., 2003). Their success with reading these pronounceable non-words suggests that children learning to read in the more orthographically consistent languages develop reading strategies directly linked to grapheme–phoneme decoding, based on the use of small grain size orthographic cues (Ziegler & Goswami, 2005). This appears to suggest that they come to rely primarily not on the instant visual recognition of whole words, but rather on the rapid reconstruction of the speech sounds on the basis of letter-sound correlation. In contrast, children learning to read in English must develop reading skills associated with this letter- sound correlation as well as the use of a mental lexicon with larger grain size orthographic cues (Seymour et al., 2003). The combination of these differing strategies possibly enables readers to cope with the complex graphemes and wide range of apparently contradictory rules they face as they learn to read words as visually similar but phonologically distinct as
“tough”, “though”, “thought” and “through”.
120
There are wider variations in individual progress in learning to read in languages with inconsistent orthographies, with more cases of dyslexia amongst their readers (Seymour et al., 2003). This supports the suggestion that learning to read in opaque orthographies
demands the development of a wider range or greater use of cognitive skills than learning to read in consistent, transparent orthographies. In addition to using large grain size cues as suggested above, readers of opaque orthographies may need to draw more on knowledge of syntax and vocabulary than readers of transparent orthographies (Share, 2008), and they may have to develop greater proficiency in using contextual cues.
However, could it be possible that all these learning demands might ultimately pay off for English readers in the form of significantly increased automaticity, with the associated benefits of extremely rapid and efficient reading in comparison with readers of languages which do not require the development of these strategies?