Reading Between the Lines: Neurocognition and Reading Acquisition in Remote Indigenous Australia

Language

Arguably the greatest barrier to English literacy acquisition for children in remote Indigenous communities is language (Prior, 2013). In Australia, many children living in very remote Indigenous communities in the NT speak English as a second or subsequent language, and many of these children have had little, if any, exposure to the English language before they attend school (Silburn et al., 2011). To gain a comprehensive understanding of why language barriers exacerbate difficulties when learning to read in English, it is necessary to understand how the cognitive processes associated with language learning facilitate reading acquisition.

When children learn language via exposure to the spoken word, they become attuned to the speech sounds of that language (Kuhl, 2004). Phonological awareness develops as children learn to discern between similar sound parcels and begin to identify patterns and regularities in their spoken language. Phonological awareness thus facilitates reading acquisition when children can identify links between sound segments and corresponding written word segments; when they learn to read, they begin to identify regularities in the written language that correspond with regularities in the spoken word (Aithal, Yonovitz, & Aithal, 2006). Evidence suggests that children who develop sound phonological awareness skills prior to starting school become better readers in later years (Schatschneider, Fletcher, Francis, Carlson, & Foorman, 2004).

For Indigenous children, the process of learning to read in English is inhibited because they are learning to read in a language that is not their spoken language. In instances where English is morphophonologically different to the child’s first language, as is the case for many Australian Indigenous languages, the child will have difficulty perceiving and distinguishing between sounds not present in their native language (Yonovitz, Yonovitz, Nienhuys, & Boswell, 1995). Furthermore, without a well-established, comprehensive understanding of English as an oral language, Indigenous children learning to read in English as a second language cannot rely on an English knowledge base to fill in their knowledge gaps as they learn to read (Aithal et al., 2006).

Due in part to differences in language and orthography between English and many Australian Indigenous languages and the difficulties associated with literacy learning in a second or subsequent language, linguists and educators have advocated for remote Indigenous children to receive a bilingual education where children can learn the fundamentals of reading and writing in their known language before they begin learning to read in an unfamiliar language (Nicholls, 2005). Although there is little available research regarding the transferrable literacy skills from children’s first language to English, there is some evidence to suggest that early delivery of literacy education in the child’s first language better develops English literacy skills for remote Indigenous children (for reviews, see Devlin, 2017; Silburn et al., 2011). In addition, a significant relationship between Aboriginal literacy proficiency and English literacy skill acquisition has recently been modeled using structural equation modeling (SEM), based on survey data of over 800 Indigenous people from NT and Western Australia (Byron Wilson, Quinn, Abbott, & Cairney, 2018). Hypothesized relationships between Aboriginal literacy, the incorporation of Aboriginal culture in educational curricula, and English literacy outcomes were examined; SEM indicated that Aboriginal literacy had a direct effect on English literacy, which was highly significant. An indirect effect of school culture and Aboriginal literacy on English literacy was also shown to be significant. Collectively, this research suggests that first language literacy can potentially scaffold English language literacy, suggesting that bilingual education is essential to improving English literacy skills for Indigenous children living in remote and very remote regions of the NT and Western Australia, whose first language is an Indigenous language.

The delivery of education in a child’s first language has a multitude of benefits beyond ensuring literacy acquisition in a language that is readily understood orally; it also ensures the vitality of Indigenous languages and promotes transmission of Indigenous culture (Disbray, 2016). While it is expected that when the child learns English literacy, skills developed in the child’s first language will be transferable, the neurocognitive mechanisms associated with cross-linguistic transfer of reading skills from an Indigenous polysynthetic language to an analytic language such as English are not yet fully understood. Furthermore, acquisition of literacy skills (in any language) is complicated by other social factors, such as health and socioeconomic factors. Discussion now focuses on these factors.

Health

For many Indigenous children living in remote communities, language barriers that obfuscate the reading acquisition process (in English or in the child’s native language) are exacerbated due to various health factors. In extreme cases of prolonged illness and infection, some early childhood health issues can have an irreversible impact on the development of essential neural processes associated with reading. One such health concern is otitis media, or middle ear infection, which is prevalent in many remote Australian Indigenous communities.

Complications associated with otitis media and conductive hearing loss in childhood are known to exacerbate difficulties in developing phonological awareness skills (Walker & Wigglesworth, 2001), and must be considered when researching phonological awareness in remote Indigenous populations, due to the high incidence of otitis media in remote regions (Williams & Jacobs, 2009). A recent investigation of the prevalence of otitis media in Australia revealed that 91% of Indigenous children aged 6 to 30 months living in remote communities across northern and central NT were affected, to varying degrees (Morris et al., 2005). In a follow-up review of otitis media, 31% of children were classified as suffering from acute otitis media, which is the highest recorded prevalence of otitis media worldwide (Morris et al., 2007). Severe cases of otitis media are associated with eardrum perforation, which can result in fluctuating or persistent hearing loss. Many Australian Indigenous children suffer from persistent bouts of otitis media from infancy well into late adolescence; by the time the child is 20 months old, they will have suffered hearing loss for an average of 32 weeks due to otitis media, and by the time the child is 20 years old, they will have suffered from complications associated with otitis media for a total of 32 months (Timms, Williams, Stokes, & Kane, 2014).

Given that otitis media and accompanying hearing loss most frequently present before a child is 3 years old, it affects the most critical stage of language development, when children are learning to recognize and distinguish different phonological sounds that are necessary for developing phonological awareness, and thus good reading skills. Evidence shows that otitis media and conductive hearing loss can have a deleterious effect on reading acquisition (Williams & Jacobs, 2009; Winskel, 2006, but also see Timms et al., 2014). Otitis media and associated hearing loss particularly affect literacy development when children start learning to read because they draw on knowledge of phonological regularities to decipher word sounds. If a child’s phonological processing infrastructure is underdeveloped due to complications associated with recurring hearing loss, that child may experience difficulty in analyzing distinct sound components of words and therefore have difficulty identifying and deconstructing these sounds as they appear in written text (Winskel, 2006).

It is important to note that some Indigenous children with severe hearing loss due to otitis media have been observed using compensatory strategies for their hearing loss, such as utilizing visual and contextual cues, including sitting closer to the teacher or away from distraction, lip reading, or looking to other children for cues (Lowell, 1995). Recommendations have previously been made to adapt the classroom environment to facilitate the use of compensatory practices by those children with conductive hearing loss (Price, 1981).

In addition to the direct effects of otitis media on reading acquisition, severe and/or recurring cases of otitis media often result in prolonged absences from school due to illness, which indirectly interfere with literacy instruction (Lowell, 1995). Comprehensive reviews of the impact of attendance on literacy acquisition have been conducted, although a review of the correlations between attendance and literacy is beyond the scope of the current article (but see Ehrich et al., 2010; Jorgensen, 2012).

SES

To date, there has been no investigation as to whether remote Indigenous children’s SES may directly affect the neurodevelopment of phonological, visuospatial, or memory pathways associated with reading acquisition. However, within the broader literature, there is evidence of a multiplicative relationship between SES and reading. Noble, Farah, and McCandliss (2006) investigated the effects of SES on the relationship between phonological awareness and reading. They tested age-matched children from elementary schools across New York, with a similar range of phonological awareness, from different socioeconomic backgrounds, to determine whether there was an interactive effect of SES and phonological awareness on reading ability. The results of their study indicate that SES amplified individual differences in the relationship between phonological awareness and reading ability for children with lower phonological awareness skills, but not for children with higher phonological awareness skills. Specifically, SES uniquely accounted for variance in performance on reading tasks only when phonological skills were lower than the national phonological awareness average; for the higher phonological awareness group, there was no unique effect of SES on reading ability. In addition, results of this study indicate that higher vocabulary levels may serve to buffer the effects of low phonological awareness in higher SES groups. This research provides evidence for a modulating effect of environmental factors on neurocognitive development. Lending support to their initial findings, a subsequent functional brain imaging study by Noble and colleagues found that SES of children from New York City also systematically modulates the relationship between phonological awareness and reading-related brain activity, thus highlighting the complex relationship between sociological factors, cognition, and neurobiology (Noble, Wolmetz, Ochs, Farah, & McCandliss, 2006).

Further empirical support for an interactive relationship between SES and reading ability comes from a study investigating the interaction between reading skill and brain structure of children from lower versus higher SES groups (based on parental education, conducted within the Chicago metropolitan area; Gullick, Demir-Lira, & Booth, 2016). Results show a differential interaction between reading skill and brain structure across SES groups, as measured using diffusion tensor imaging. Specifically, as well as showing white matter coherence in neural regions that typically support reading activity, regions typically associated with visuospatial processing show greater white matter coherence in higher readers versus lower readers from lower SES groups. The authors suggest such results may indicate that children from low-SES status groups rely more heavily on visuospatial processing mechanisms as a supplementary strategy when verbal strategies are insufficient or underdeveloped (Gullick et al., 2016). We note that there are multiple reasons why such complex brain interactions could be occurring, and such an interpretation would be strengthened by further research which demonstrates that white matter coherence in the visuospatial processing areas of the brain systematically covary with reading skill for lower SES groups.

Additional research is required to explore the multiplicative relationship between neurological structure and function, factors associated with SES, and cognitive behavior across SES groups. Collectively, such research could potentially lend support to the theoretical postulation that remote Indigenous children who have difficulties developing phonological awareness skills due to factors associated with SES may use visuospatial strategies to scaffold learning with regard to reading.

Culture

A child has the best opportunity to learn when education is contextually appropriate to, and builds from, their already learnt experiences (Fleer & Williams-Kennedy, 2002). Within many remote Indigenous communities, education is most valued when it teaches information that is culturally relevant to Indigenous beliefs and customs. For education to be contextually appropriate, children should be familiar with themes and materials referenced in educational instruction and educational resources. Standard Australian English literacy curricula and supplementary materials are often culturally and contextually irrelevant to that which is commonly experienced by Indigenous children who live in very remote regions within the NT; concepts and illustrations in story books commonly represent non-Indigenous cultural practices and themes and therefore can be difficult for Indigenous children to understand, comprehend, and contextualize (Jones, 2014; Simpson & Clancy, 2005). In addition, the actual practice of reading to children is unfamiliar to most Indigenous families living in very remote regions of the NT. Many Indigenous cultures traditionally transmit culturally significant knowledge and stories intergenerationally via oral storytelling sessions, rather than via written text (Farrant, Shepherd, Walker, & Pearson, 2014). Given that many Indigenous languages are predominantly oral languages, children may have limited access to story books at home, and culturally relative story books are generally not available to Indigenous children in many remote regions until they attend school (James, 2014; Maher & Bellen, 2015). This is concerning, given that early exposure to storybooks in the home environment helps to develop preliteracy skills that have been demonstrated to predict later reading success (Mol & Bus, 2011; Sénéchal & Young, 2008). However, just as parent–child reading has been found to facilitate reading acquisition, oral storytelling could also potentially scaffold reading acquisition for Indigenous children. A recent study by Farrant and colleagues (Farrant et al., 2014) has demonstrated that oral storytelling in language is a significant predictor of Australian Indigenous language development. Specifically, listening to oral stories and narratives increases an Indigenous child’s vocabulary. This work shows the potential opportunities for building literacy skills from Indigenous strengths. However, further research is necessary to determine whether oral storytelling and increased vocabulary are indeed predictive of reading for Indigenous Australian children.

Sociocultural research has previously highlighted ways in which Indigenous cultural educational practices are different from Western formal educational practices and has provided evidence of the benefits of including culturally appropriate pedagogy in schools. From a neurocognitive perspective, we can take this line of research further by considering how Indigenous cultural teaching and learning practices may influence an Indigenous child’s neurodevelopment. Specifically, we can consider whether Indigenous cultural practices serve to strengthen some of the neurocognitive processes associated with literacy learning. Discussion will now turn to those neural processes that may be enhanced by Indigenous teaching and learning practices.

With regard to visuospatial memory processes, early cognitive research by Kearins (1976, 1981) indicates that Indigenous children living in remote Australia perform better than age-matched non-Indigenous children on a variety of visuospatial memory tasks. Across all tasks, there was a significant group difference, with Indigenous children outperforming non-Indigenous children. Significant group differences were also found at each age level. Overall, such findings suggest superior visual-spatial processing abilities of Indigenous children in remote communities, compared with non-Indigenous children. Kearins also observed behavioral differences in task performance: Indigenous children made more deliberate, considered placement choices, tended to rely mainly on visual strategies, and rarely changed their minds, whereas non-Indigenous children were more likely to move and replace items and tended to rely mainly on verbal strategies (Kearins, 1981). In addition, Kearins (1981) noted distinct differences in the children’s conduct and demeanor: Indigenous children generally sat still and silent while undertaking the task, whereas non-Indigenous children tended to wriggle about, appeared more fidgety, and made exasperated utterances as they undertook the task.

Subsequent research by Kearins has demonstrated that central Australian Indigenous children living in urban environments perform similarly in visuospatial tasks when compared with remote central Australian Indigenous children (Kearins, 1986). One possible explanation for these findings, as offered by Kearins (1981, 1986), is that Indigenous and non-Indigenous cultures engage in different child-rearing practices, even in environments where traditional cultural practices are no longer taught. For example, Western children are taught with a considerable amount of verbal instruction, whereas Indigenous teaching instruction tends to focus more on visual techniques, such as observing then autonomously carrying out observed behaviors (Kearins, 1986). This explanation is consistent with other observed accounts of cultural education and traditional learning in Indigenous populations (Fleer & Williams-Kennedy, 2002; Simpson & Clancy, 2005) and of early documented observations of the enhanced navigational skills of Indigenous Australians (Lewis, 1976). Accordingly, another possible interpretation of Kearins’s findings is that the Indigenous children tested by Kearins performed better than non-Indigenous children at visuospatial memory tasks because they had developed enhanced visuospatial skills which can scaffold better memory performance. It is plausible that culturally specific early teaching and learning activities could potentially strengthen visuospatial neural networks during early childhood. It is noted, however, that Kearins’s results have not been consistently replicated (Drinkwater, 1976, 1978; Harris, 1977; Knapp & Seagrim, 1981).

In an attempt to replicate Kearins’s earliest findings, Drinkwater (1976) conducted variations of Kearins’s visuospatial array tasks. Overall, Drinkwater found no significant differences between Indigenous and non-Indigenous children in their performance on each of the tasks, although Kearins suggests this difference in results is largely due to the difference in population samples; Drinkwater (1976) tested Indigenous children from coastal Queensland, children who spoke Aboriginal English and attended an English-speaking school, which is a vastly different population from the more traditional central desert population tested by Kearins (1981, 1986). In a reply to Drinkwater (1976), Kearins asserted that her results were not deemed generalizable to other Indigenous populations from different geographical regions and should not be interpreted as such (Kearins, 1978). Employing Kearins’s (1981) design more closely, but using children from a remote community in north-east NT, Harris (1977) also failed to replicate Kearins’s (1976, 1981) results. Knapp and Seagrim (1981) conducted further testing to clarify these mixed results, using Indigenous children from the central desert region and non-Indigenous children from Canberra, but failed to produce evidence to support Kearins’s (1976, 1981) findings. Furthermore, Knapp and Seagrim (1981) suggest it is most likely that Indigenous and non-Indigenous groups employ both verbal and visual strategies when completing cognitive tasks, but concede that different cultural groups may use one strategy more or less than another group, and there will also be individual differences within groups in the reliance on different strategies.

Klich and Davidson (1983) investigated the conflicting evidence of Kearins (1976, 1981) and others by conducting further experiments that tested the visuospatial memory skills of desert-dwelling Indigenous children and rural non-Indigenous children, but using Drinkwater’s methodology. The results of their experiments demonstrated that Indigenous children performed better than non-Indigenous children (Klich & Davidson, 1983), which lends support to Kearins’s (1976, 1981) findings. In addition, when non-Indigenous children were primed to use visual memory strategies over verbal memory strategies, there was no longer a significant difference in performance. Klich and Davidson (1983) concluded that Indigenous children may indeed rely on visual memory strategies when verbal memory strategies are inefficient. Overall, evidence suggests that remote Indigenous children may have enhanced visual memory strategies, but the evidence is mixed. This may be largely due to differences in methodological approaches and different population demographics across samples.

Support for evidence of the superior visuospatial ability of remote Indigenous children has been demonstrated in an experiment investigating counting strategies for Indigenous children whose first and only language (Warlpiri or Anindilyakwa) contains a limited linguistic repertoire for counting words (Butterworth & Reeve, 2008; Butterworth et al., 2008). ² Butterworth and colleagues used three different enumeration tasks to examine how Warlpiri-speaking children from the Australian Central Desert region and Anindilyakwa-speaking children from coastal Groote Eylandt processed numbers without using numerical referents. For three separate tasks, results showed no statistical difference in counting and enumeration performance between age-matched Indigenous and non-Indigenous groups (Butterworth et al., 2008). The authors concluded that Indigenous children were able to use visuospatial strategies in place of verbal counting strategies to complete each of these enumeration tasks successfully. This provides further evidence that some Indigenous children are able to harness visuospatial strategies to complete tasks when other strategies are not available.

Drawing from the evidence provided by Kearins (1976, 1981), Klich and Davidson (1983), and Butterworth and colleagues (Butterworth & Reeve, 2008; Butterworth et al., 2011; Butterworth et al., 2008), it appears that some Indigenous Australian populations may demonstrate enhanced visuospatial processing skills, although mixed results have been found (Drinkwater, 1976; Harris, 1977; Knapp & Seagrim, 1981). Drawing from the results of these experiments, it is also plausible that remote Indigenous children may demonstrate superior visuospatial memory.

Although not yet researched within a neurocognitive framework, different lines of anthropological, ethnographical, and scientific research have documented the memory capabilities of remote Indigenous Australians. Early observations of enhanced navigational skills of Indigenous Australians were perceived to be associated with a psychophysical mechanism akin to a “dynamic image or mental ‘map’, which was continually updated [and] realigned at each change of direction” (Lewis, 1976, p. 27). This suggests that Indigenous navigation skills relied on enhanced visuospatial processing capacity and excellent visual working memory skills. This is evidenced by observations that when individuals (N = 12) were asked to point to geographical locations (average distance of 300 km), the average degree of error reported was 13 degrees, and average error in relation to the axis of travel when navigating and traveling between two points was 13.7 degrees (Lewis, 1976). Expert navigational skills of Australian Indigenous peoples are facilitated by songlines, which are defined as oral narratives of geographical and celestial landscape that document the environment (Norris & Harney, 2014). Such oral songs and narratives document the intrinsic relationships between people, culture, and country, and are transmitted intergenerationally according to strict rules and customs, to build a collective oral knowledge base of environment, history, spirituality, culture, law, as well as kinship relationships and obligations (Norris & Harney, 2014; Van Toorn, 2000). The practice of transmitting history and narratives intergenerationally has survived for many thousands of years, and the accuracy of Indigenous historical and geographical knowledge has been scientifically verified. For example, Nunn and Reid (2016) have empirically verified stories about rising sea levels and changing coastlines, which have been communicated intergenerationally for more than 7,000 years, as told by different Aboriginal groups across 21 Australian coastal locations. The enduring accuracy of such narratives not only provides evidence of superior long-term memory skills but also suggests excellent working memory when we consider that the receivers of knowledge must attentively listen to and follow the narratives and stories they are told so that they, too, can later communicate and share the information with future generations.

Overall, these bodies of research suggest that although remote Indigenous children may have difficulties developing phonological awareness due to a multitude of sociocultural factors including language, health, and SES, further research into Indigenous cultural teaching and learning strengths, such as enhanced visuospatial and memory processes, may offer new and unique educational opportunities with regard to reading acquisition.