From the moment that achievement problems have been call or classified as learning disabilities instead of mental retardation, up to now, there is consistent research conducted in the field of learning disabilities. One of the problems that have received most attention, is reading difficulties. Findings are supportive of the theory that reading disabilities are due to a deficit in phonological abilities. No causal relationship has yet been established, but the assessment of these abilities has been proved to be the best predictor of the reading level at every age. However, in the last decade (due not only to the expansion of research in this field, but to the "globalization" phenomena and "bilingualism"), researchers had tried to generalize the findings observed in the English speaking population to another languages, and are still researching how this definition of reading difficulties applies to different languages that also have different orthographies.
            There are several reasons to question whether the same patterns of relationships would be obtained for children learning to speak and read other languages. First, the phonological structures of languages differ, and these linguistic differences might be expected to influence the development of certain types of skills.
             Differences in the consistency or transparency of written language is another factor that could lead to cross linguistic differences in patterns or reading acquisition. Although all alphabetic languages reflect the spoken language at the level of phonology, some to this more consistently (transparently) than others (Bruck, M. , Genesse, F, Caravolas, M. , 1997).

            Therefore, is not only the regularity of the language but also the unit of language employed (phoneme, morpheme, etc.), what will affect the development of reading (Wimmer & Goswami, 1994) and the strategies involved in that process.

            In terms of regularity, we can think in a continuum where languages are classified from the more transparent / shallow orthography to the more opaque / deep orthography (see graph).

            A transparent orthography is that with a consistent mapping from spelling to sound. Grapheme-phonemes correspondences (GFC) are easier to detect. The best method for teaching is phonics in that children need to learn the correspondences between grapheme-phoneme to be able to decode any word or even non-words. Automatization of this process is expected to occur after one year of exposition to phonics instruction, where the focus is transferred to timing and reading comprehension. In this orthographies, the initial logographic stage is omitted. However, once the GFC is automatized, orthographic strategies such as "sight vocabulary" are needed for acquiring fluency.

            An opaque orthography is that where the underlying rules are less consistent, more complex in terms of being context-sensitive and operating at different phonological levels. Graphemes or combinations of letters might have a different phonemic association depending on the context. The reader should be able to deal with phonetic as well as contextual strategies. Usually, most of the words are learned by sight. For these orthographies, might be more adaptive to learn spelling patterns by rote memory, and then use various strategies such as analogy (similar endings, patterns) to try to read new words. Children might go through a logographic stage (visual access to the lexicon), and then go through the alphabetic stage (Wimmer & Goswami, 1994).

            In terms of unit of language employed, languages go from syllables (morphemes) to phonemes. Thus, from symbolic to alphabetic systems, all languages have a writing system which has to be learned to read. Alphabetic systems are quite simple in that once acquired, the reader is able to read fluently any text presented. Phonemes are isolated sounds, that once combined are easy to read. Symbolic systems, require more memory and orthographic skills, and are limited to the prior knowledge of the reader and his ability to deduce meaning from the context. Morphemes are the least units of meaning, therefore, memorization it is required for reading (sound and meaning) in order to be able to combine them and deduce meaning.

             Landerl (1997) suggested that in transparent /regular languages, even dyslexic children were found to exhibit high levels of phonological awareness and to read accurately as long as their reading time was unlimited. However:

Even in a language with a consistent orthography, accurate word decoding is at the beginning an effortful process of print-to-sound translation (phonological recoding), and high accuracy levels are obtained only at the expense of a low reading speed. In this phase, as for american children, differences in phonological awareness will be relevant. Thus, when speed and not accuracy becomes the dominant factor in reading, as is characteristic for learning to read in more consistent orthographies such as Dutch and German, phonological abilities ceases to have additional influences.

In languages with an inconsistent orthography, complete word decoding accuracy seems to be reached after a much longer period and, as a result, phonological abilities  remains important for subsequent reading acquisition.

         Furthermore, instruction method should have an impact in the development of pre-reading skills. It appears that in languages with transparent orthographies children do not receive specific phonic instruction and letter knowledge, until they reach grade 1. This is just the opposite as occurs in English instruction, where children are taught literacy skills since kindergarten. These might be due to the fact that transparent languages are straightforward to learn, and children acquire the basic skills without that much effort. Whereas for English, they need to master a great deal of rules due to the irregularity of the language.
 

Glossary

GFC: grapheme-phoneme correspondence