Neuroscience and literacy acquisition

It is a pleasure for us to present the following special section of the International Journal of Behavioral Development on literacy development and variability, with a focus on neuroscience methods. A cursory search of publications on children’s processing as measured using neuroscience techniques suggests that most of this research appears in journals dedicated particularly to neuroscience in some capacity or to other relatively specialized journals (e.g., on autism, dyslexia, or other specific developmental anomalies). We are delighted to offer this special section on literacy skills and neuroscience in a journal that has a long history of broad themes in developmental psychology, including both children and adults. With this special issue, we present some of the most up-to-date thinking in the area of reading from a neuroscience perspective. These articles, by some of the best known and respected researchers in the area, have been written for those who are not necessarily specialists in neuroscience but who have a special interest in the very wide-ranging area of literacy development and impairment.

Literacy research presents a particular challenge for developmental psychologists because learning to read and to write involves some skills that are expressly developmental and others that are clearly focused on explicit teaching and learning. Language and metalinguistic skills that develop with maturation are combined with intensive, effortful teaching with scaffolding by parents, teachers, or other tutors. Along with the tension between development and education, this special section also highlights the importance of culture, which presumably encompasses various components of the literacy learning experience including language, script, home environment, and government and school approaches to teaching reading and writing. With these many competing variables, what are the universals and specifics of literacy development and variability? The papers in our special section suggest some answers.

The special section contains five empirical studies. Three of the studies used the event-related potential (ERP) technique with EEG data in order to investigate changes in the N170 (N1) component in the course of literacy acquisition or short-time reading training in children or adults. Two studies used a correlational approach in order to associate reading-related behavioral measures with neural measures stemming from developmental MRI and EEG recordings. The languages and writing systems in these studies ranged from Chinese and Japanese to English, Spanish, German, and Finnish. In the following, we present a short description of each study emphasizing some of their contributions to the field of neuroscience and literacy acquisition.

Zhao, Zhao, Weng, and Li (2017) conducted a literacy training study in Chinese preschoolers (4–6 years old) in order to test whether different training methods would modulate how the brain becomes initially tuned for print. They compared a visual identification training with a free writing method and measured changes in the N170 component of the ERP. Results showed that training effects in the visual identification training were found over the left and right hemisphere (depending on the contrasts tested), while those in the writing training were generally most pronounced in the right hemisphere. The results suggest that different types of reading instruction lead to distinct changes at the neural level that are related to visual expertise, and that such effects can even be found after a relatively short training period.

Sanchez-Vincitore, Avery, and Froud (2017) reported ERP results from a rare group of people who had learned to read only as adults and compared them to a group of adults who had learned to read in childhood. While the N170 in the childhood learners was left-lateralized for words and also larger for words than symbols over the left hemisphere, the N170 in the adult learners did not distinguish between words and symbols and was bilateral for words rather than left-lateralized. The results suggest that reading if acquired late in adulthood is not supported by the same neural mechanisms compared to reading learned in childhood, presumably due to a lack of automatization as the authors speculate.

Brem et al. (2017) investigated in a single session, whether associating a character with a syllable or a spoken word would lead to neural tuning for print, and whether tuning would differ for words versus syllables. They tested native (Swiss-)German speakers and let half of them learn the association between a character and a syllable and the other half between a character and a (German) word. Characters were either in Japanese (Hiragana) or Chinese. The results showed a training-related increase of the N170 that was specific for the trained characters. However, associating characters with syllables or words led to a similar increase. The results demonstrate that neural tuning for print can develop very rapidly, and suggest that the mechanism behind such rapid development is phonological association or visual discrimination learning, rather than learning the meaning of the character.

Xia et al. (2017) investigated whether cognitive subcomponents of reading were associated with specific patterns of cortical thickness in different areas of the reading network. They found specific areas that were associated with phonology-related skills (including phonetic representation, phonological awareness, orthography-phonology mapping skills), sentence comprehension, morphological awareness, and orthography-semantic mapping, respectively. The authors see this as evidence that the reading network of typically developing children consists of multiple subdivisions, which thereby provides neuroanatomical evidence for a multi-componential view of reading.

Hämäläinen et al. (2017) investigated in Finnish and American children, whether neural mechanisms underlying processing familiar speech sounds would differ from unfamiliar ones and whether these mechanisms would be associated with reading skills. They used a change detection paradigm contrasting ba-da-ga and ata-at: a. No differences were found in the change detection responses between the Finnish and American children. The change detection for the unfamiliar phoneme length contrast (ata-at:a) in the American children, however, correlated with reading skills, with larger responses being associated with poorer reading skills. The Finnish children also showed associations between change detection and reading skills, but the pattern was less clear. The authors conclude that speech perception is indeed linked to reading skills, but that the association is stronger for uncommon speech sound contrasts.

Taken together, the five papers demonstrate how measures of neural mechanisms that underlie reading can inform our understanding of reading-related processes. Neural mechanisms of visual-orthographic processing were shown to differ between some learning strategies (Zhao et al., 2017), but not others (Brem et al., 2017), and to differ depending on the degree of automaticity (Sanchez-Vincitore et al., 2017). Reading skills were also shown to be associated with the neural mechanisms underlying speech perception, and, interestingly, the most prominent association was found for unfamiliar speech sound contrasts (Hämäläinen et al., 2017). Finally, one paper showed that reading can be considered as being supported by several subcomponents, each of which was associated with a specific pattern of the underlying neural substrate (Xia et al., 2017).