How important is teaching phonemic awareness to children learning to read in Spanish?

Goldenberg, C., Tolar, T.D., Reese, L., Francis, D. J. Bazan, A.R., & Mejia-Arauz, R. (2014). How important is teaching phonemic awareness to children learning to read in Spanish? American Educational Research Journal, 51(3), 604-633.


Here is a new and fascinating wrinkle in the ongoing debate over what should be emphasized in an effective early literacy program: Does the answer depend on the language a child is learning to read? Do the skills to be taught vary based on whether a child is learning to read in her/his home language, or in a second language? Do reading skills learned in one language transfer to reading in a second language? The research study described here explored all of these questions, and caused me to do some deep pondering on the contextual factors that can affect what kind of early reading instruction is best for a given child or group of children. The results reported for Goldberg et al’s study confirm my own belief that there is no “one size fits all” in early literacy instruction, but this research also made me think in new ways about how important spoken language is and how inextricably tied it is to the processes of reading and learning to read.

The authors looked at the nature of early reading instruction provided to three groups of first and second graders: 1) children born in Mexico and attending Mexican public schools, with all instruction in their home language, which was Spanish, 2) children either born in Mexico or with at least one Mexican-born parent who were living in the U.S. and attending U.S. schools, and who were receiving their first reading instruction in Spanish, which was their home language, and were in bilingual programs, and 3) children born either in Mexico or with at least one Mexican-born parent who were living in the U.S. and attending U.S. schools, and who were receiving their first reading instruction in English, which was not their home language but a second language, and who were in full-immersion English programs. Versions of the same measures were used to measure literacy outcomes for all three groups of children. These included measures of phonemic awareness and measures of oral language and reading, which were administered four times over a two-year period (fall and spring of both first grade and second grade). In addition, parent questionnaires were gathered and classroom reading instruction was observed over time.

One of the most interesting pieces of the article was its description of Mexican early reading instruction, which tends to be more comprehension-based than code-based, which differs from typical U.S. early reading instruction, which has been predominantly code-based, with a heavy focus on phonemic awareness, in recent years. Mexican children look more at whole words, syllables, and onset-rime combinations than at individual phonemes. The authors briefly discuss differences between Spanish and English that may account for this. Spanish seems to be a more “transparent” language than English, with more trustworthy phoneme-grapheme correspondences than English, and less ambiguity for beginning readers to cope with. Anyone who has worked with beginning readers or struggling readers is well aware of the problems the orthography of the English language presents. Readers of English need multiple decoding strategies that often require them to use multiple cueing systems simultaneously. Many readers do figure out how to navigate this complex process successfully, but there are those who struggle with it, some for a long time or even a lifetime.

The most striking finding here was that the children in Mexico who were learning to read in their home language, but without an emphasis on phonemic awareness, quickly caught up with and surpassed both of the groups of children who were learning to read in the U.S. on the measures of oral language and general reading ability by the end of second grade. Both groups of children in the U.S., who were receiving phonemic awareness instruction, either in their home language of Spanish or in their second language of English, scored better than the Mexican children on phonemic awareness, but did not make as much progress as the Mexican children on oral language and general reading ability. This occurred even though the Mexican children started out at a lower rate of achievement, and also were in schools that would be considered seriously “disadvantaged” in the U.S., with a shorter school day (Mexican schools operate in two “shifts” per school per day due to overcrowding), large classes (possibly 50 in a class), and classroom libraries that would seem almost nonexistent to U.S. educators.

The rationale given here is that decoding in the Spanish language is much more straightforward than in English, so phonemic awareness is acquired more naturally in Spanish through exposure to texts, and explicit instruction in phonemic awareness is unnecessary. That immediately raises some questions. First, could the phonemic instruction provided to the U.S. children who were learning to read in Spanish be inappropriate or at least less than optimal for those children’s literacy learning? Or is it still worthwhile to teach phonemic awareness in Spanish to children who will later be expected to learn to read in English? Will phonemic awareness skills, while not as useful in Spanish, be helpful later when reading in English? Will there be transfer across languages? Or will being able to “crack the code” and experience reading for meaning in one language be enough to facilitate second language reading, with phonemic awareness not being the key factor, but instead reading success and engagement?

There is definitely evidence in this study that reading skills do transfer across languages, though the findings here must be interpreted cautiously (it is a study in authentic environments, and therefore strict control was not possible). I believe there is also evidence here that it is beneficial for Spanish-speaking English learners to first learn to read in their home language. That seems most powerfully illustrated here by the evidence that the Mexican children, despite what U.S. educators would consider “disadvantages”, still outperformed both groups of U.S. children on general reading measures by the end of second grade. Learning to read in one’s home language first would logically seem to be easier for a young child; the focus could be on learning to decode words and read for meaning, without the added burden of having to translate words from English to Spanish. When learning to read in a second language, the need to translate would put up a barrier between sounding out a word and recognizing it.

The key question is whether there will be later difficulties if a child subsequently must learn to read in English. It is possible that phonemic awareness taught in Spanish could help with the transfer to reading English. More research definitely needs to be done on how early reading instruction in the home language might be structured to ease the transition from reading in Spanish to reading in English. If we could figure out how best to do that, the bilingual model might be accepted more by factions in the U.S. who have opposed it. The goal of any literacy instruction program is to find the most optimal models that fit the needs of the children who will be learning to read. It is clear from the evidence presented in Goldberg et al’s research that English immersion programs, with initial reading instruction only in English, are not the optimal way for Spanish-speaking English learners to learn to read. There is still much work to be done to determine what would be the best way for these learners, but the research reported here will be helpful for those undertaking that work.