Writing and the literary heritage of the Americas: What have we learned?

Norbert Francis
14 min readJul 16, 2022


The following review is an excerpt from: “Literacy and the language awareness hypothesis,” published in Writing Systems Research, v.11, n.2, pp. 176–187

With the growing emphasis on cross-language and cross-writing system research, we have begun to converge upon principles of literacy and literacy learning that are foundational, principles that should now apply across all cultures. Even in the case of historical research, in which experiments can no longer be carried out on learning, results from the study of the development of ancient writing systems suggest conclusions that are compatible with current interdisciplinary models.

Ancient Mayan Writing

Language and Literacy

For argument’s sake, taking ‘text’ to refer to both writing and units of oral discourse, the evidence strongly suggests that formal discourse of the oral tradition is also capable of fixing texts for reference and recovery. To one degree or another, authors on all sides of the debate on the effects of literacy accept this view. In addition, looking back in history, institutions of learning have presented ‘oral texts’ of this kind to students as a learning resource similar in many ways to how textbooks, print and electronic, are a resource today. Conceivably, the calmecac of the Nahuatl speaking Aztecs, to take one example, promoted memorisation in both narrative and expository format with effective instructional technique and by the fomenting of high motivation. These methods were probably aided by elaborate graphic representations, the famous codices, that served as context support for the oral tradition content (Boone, 1998; León-Portilla, 1996). As in similar pedagogical settings in other cultures, texts fixed in memory in this manner were available for reflection and commentary. The great epic poem and creation legend of the Maya, Popul Vuh, prior to the discovery of a colonial era alphabetic version, may have been passed down by the oral tradition and/or by texts composed in the pre-Columbian Mayan morphosyllabic script (Christianson, 2007).

On this last point, one of the objectives of our research project in Latin America has been to compile a corpus of oral tradition material for documentation and study. Over the many years of fieldwork, what has become evident is that among adults the most coherent and well-formed narratives include a good number performed by non-literate storytellers — children’s stories were evaluated separately. As other fieldworkers have also observed, the ability to execute a complex and self-contained narrative, epic poem or expository discourse emerges from exposure and instruction provided by cultural institutions and practices. This type of autonomous (non-conversational) discourse comprehension and expressive ability, including reflection upon structures of text not completely dependent on previous knowledge context, can develop without the participation of writing, even by learners who never acquire literacy (Finnegan, 2012; Gough & Bock, 2001; Gómez-Pulgarín, 2012). Again, regarding this aspect of the oral–written dimension, almost all researchers, our project included, have generally accepted these observations as intuitively valid based on selective field observation and theoretical argument from related learning scenarios.

Indigenous language writing

From the beginning of our project on bilingual literacy, summarised in Francis (2012), we found the concepts of the Olson (1991) literacy-metalinguistic paper useful in guiding the design of assessments. In turn, we called attention to the related idea of Secondary Discourse Ability (SDA) (Gee, 1996) as a way to bring together for observation the common features of written literature and oral tradition. The concept of SDA also highlights the continuities between literacy learning and its antecedents in pre-literacy language use (e.g. story telling), for example during early childhood. Independently, and around the same time, Gough and Bock (2001), began to make reference to this category to describe the same commonalities of text/discourse construction. SDA would not only be largely independent from the mental networks of both Spanish and Nahuatl (in our case), but would be largely nonlinguistic as a cognitive domain, if the model is correct. In the same way, aspects of metalinguistic awareness (MA) should also be non-linguistic, therefore ‘shared’ between the two language representations of the bilingual speaker. The question posed was: how are MA and SDA relevant to the study of literacy in the specific case of indigenous communities of Mexico where bilingualism is widespread in the school-age population? An important intervening factor in this case is the learning of Spanish as a second language. What continuities can be verified experimentally between community oral tradition practice (not yet appreciably eroded in the community) and elementary school literacy learning? Is the former an antecedent (in the sense of cognitive-linguistic support or foundation) for the latter? The parallel, historical, questions are about interactions across many generations.

The historical evidence

Nahuatl is only one of the American languages that played an important role in literacy development during the colonial period. Just as important for researchers and bilingual communities today are the historical precursors, during the epochs of the different pre-colonial civilisations, of this surge in literary activity. Thus, the discovery of a system-wide phonetic component of the Mayan script and its definitive deciphering came as a pivotal conclusion, rectifying theories that had taken us down the wrong track. It set the stage to properly unify the study of pre-Hispanic language and literature with contemporaneous developments in the Old World, in Western Asia and Egypt along with the Chinese invention of writing. The groundbreaking description of Mayan writing revealed interesting parallels with the analysis of the scripts adapted from morphosyllabic Chinese writing by the cultures that inherited the character system (Vietnam, Japan, Korea, the Zhuang people, among others). Handel (2019) is a recent summary of progress toward a complete account and decipherment of the ancient East Asian scripts, based on the principle that writing depends on language, a basic concept that has also served the analysis of modern adaptations of the standard Chinese characters.

The consensus today regarding the writing-language relationship, in particular the encoding of phonology systematically and completely, and the productive manner in which Mayan scribes represented this linguistic component, points to a high degree of metalinguistic awareness applied to the domain of writing and literacy. In this case, the script implements the connection at the point where the phonetic component (syllable unit in Mayan writing) is coupled with graphemes that represent content morphemes. It is not an exaggeration to call attention to the forthcoming unification of these language-writing projects as one of the most important research convergences in linguistics and anthropology internationally.

The research breakthrough overcoming decades of conceptual bias and entrenched schema that had denied the significance of the encoding-of-phonology hypothesis, has been outlined by Michael Coe (1992). All available evidence shows that among the pre-Colombian people the Maya had designed a complete writing system, more advanced for example than the largely pictographic methods of documentation of the Aztecs. The recognition of this achievement does not in any way disparage the elaboration of magnificently crafted Aztec pictorial documents or their various context-embedded communicative functions. The invention and widespread use of these context-dependent graphic systems stands as a parallel achievement to that of the Maya. In fact, the system of Aztec symbols can most accurately be described as representing an early stage of transition toward a complete writing system marked by its experimentation with the selective phonetic adaptation of pictographs (rebus principle). The examples of adapted pictograms, for example for place names, represented a potential early precursor to a syllabary or alphabet (Manrique Castañeda, 1988; Prem, 2008). In contrast, the Mayan morphosyllabic writing system has been shown to possess the full capability of high-fidelity transcription of any Mayan expository or literary work of the oral tradition of any length and complexity, including narrative and poetry. The resulting text could be reliably read by a literate speaker of the language without extensive previous knowledge and supporting contextual information. Extensive prior knowledge and context are essential when ‘interpreting the meaning’ of a pictographic display. Its use to support a fixed oral discourse already committed to memory would be an example of this relationship. In contrast, ancient Mayan writing encoded the actual linguistic constituents, of the syntax, morphology and phonology. To say that such a text was language-dependent means that the reader needed to possess linguistic competence in a specific language or specific languages of the Mayan linguistic family with which the orthography was aligned.

Writing System of the Maya

Revealing is the criticism of the findings of the Mayan literacy research in an article published in Reading Research Quarterly (Jiménez & Smith, 2008) making reference, coincidentally, to fieldwork in Puebla state in close proximity to our Spanish–Nahuatl bilingual project. Aside from the failure to clearly distinguish between pictographic systems and language-based writing, what the article reveals is a fundamental misunderstanding of the conclusions of the Mayan language-writing project. It effectively underestimates the importance of the historical achievement itself, the discovery of which today is widely recognised in the field. In taking issue with Coe’s proposal (p. 25) that ‘all known writing systems are partly or wholly phonetic, and express the sounds of a particular language’ Jiménez and Smith misleadingly argue for the need, in regard to the use of the term ‘phonetic,’ to ‘distinguish between systems that encode the phonemes of a particular language’ and those that cue other constituents (p. 36). But in the literature the term ‘phonetic’ refers to the general category of phonological pattern, not just to ‘phonemes’. For example in Chinese writing, characters usually consist of a semantic radical and a phonetic. No orthographic element or stroke pattern in characters maps onto the phonemes of Chinese; rather the phonetic corresponds to the syllable level units of the phonology (Pan et al., 2016). Similarly as in Mayan characters, the phonetic components map onto syllables. More precisely, it has been shown that Mayan writing was morphosyllabic, remarkably parallel in its design to the adaptation of Chinese characters by the East Asian cultures that received the influence of China. As Coe points out, the most interesting comparison is to the hybrid system of Japanese writing: for example, morphosyllabic kanji for content morphemes, purely syllabic (or moraic) kana for functional morphemes.

Coe differentiated between the qualitatively different design features of pictographic documents and morphosyllabic writing not out of a vague neglect to be inclusive, but rather because, precisely, morphosyllabic scripts allow for the full transcription of language as it is spoken; pictographic systems do not. For this reason, the work of decipherment went hand in hand with the successful reconstruction of the ancestral Mayan languages. As Breaking the Maya Code explains in detail, language recovery is a requirement of decipherment because the writing system must be aligned with a spoken language, with its phonology in particular. The research question now posed is whether further evidence can be found to support the conclusion that the invention by the Mayas of this advanced language processing technology allowed in turn for new capabilities in expression and creativity.

In this regard, it is important to reiterate that the alphabetic writing system that 16th-century Spanish missionaries invented for Nahuatl was rapidly adopted by the surviving educated elite of the former indigenous empire, appropriated as well for independent expression and legal and political advocacy. The literary, ecclesiastical and documentary purposes of written Nahuatl were prominent for approximately 100 years, as in the seminars celebrated at the famous Colegio de Tlatelolco, before receding in prestige and utility, relative to Spanish, during the seventeenth century (Fountain, 2015; León-Portilla, 1996; Rosas Xelhuantzi, 2018).

Tira de la Perigrinación

Learning how writing systems represent speech

The debate on the defining characteristics of pictographic-type documentation and language based writing directly engages the discussion of the language awareness hypothesis. Discounting the importance of the discovery of the link to phonology in the case of the Mayan script suggests an approach to studying writing that deemphasises this link in general. The same approach might then be extended to the study of literacy learning that also questions the importance of language awareness, including phonological awareness. Understanding how children learn the mapping between their native language and the writing system of their culture is also relevant to Olson’s literacy hypothesis. ‘Language’ and ‘writing’ cannot be reduced to ‘meaning’. In the mastery of skilled reading and writing, awareness of language helps the learner focus attention on the constituent patterns of words and phrases. ‘Holistic’ theories of literacy, popular in the educational field for many years, failed to clearly distinguish among the components of reading ability because they underestimated the importance for learners of mastering the system of how graphemes map onto the sound patterns of language. The failure could be traced to an ill-defined emphasis on top-down meaning construction while deemphasizing the language form-writing form correspondences essential for efficient decoding. One of the errors of early ‘whole language’ models of reading was to assume that processing of written text is heavily dependent on background knowledge and prediction, thus minimizing the role of phonological awareness and accurate word identification.

More directly to the controversy, Olson’s literacy hypothesis is based on the growing consensus today in the psycholinguistics of literacy that full writing systems represent the morphological and phonological patterns of the language, the language with which a specific orthography is aligned. That even in orthographies of deep irregularity, readers do not by-pass phonology when decoding for meaning in silent reading. Historical examples suggesting that early civilisations experimented with visual representations that directly linked symbols to concepts and images (by-passing speech) does not contradict the principle of language form-writing form correspondence. Where pictographic-type systems emerged, if they survived and evolved to represent language, the representation of morphology and phonology in the units of the writing system became part of their design. Recognising that language awareness also develops in the oral tradition does not support the attempt to minimise the unique contribution of literacy in attaining advanced text-based language proficiencies, necessary for modern scientific methods of inquiry. In addition, this understanding of the literacy hypothesis (Olson & Oatley, 2014) is compatible with the evidence that language awareness begins to emerge in childhood prior to literacy learning. It is also compatible with the observation that adults, who do not read or write, but who have been immersed in the oral tradition secondary discourse of their culture also develop corresponding levels of language awareness. Along the same lines, the repeated finding from different fields of study that literacy depends on language, including its sub-component domains, undermines the attempt to relativise the distinction between full, language-based, writing and pictographic-type communication. The principle of language-literacy dependence in turn helps us avoid confounding language competence and language awareness, useful for example in the research on the interactions and interfaces among the components that participate in reading ability — the linguistic subsystems, the visual system, working memory, and so forth. Lastly, the competence–awareness distinction does not contradict the principle of language-literacy dependence. Just as language awareness builds upon language competence, literacy (as does advanced SDA) also builds upon language competence.

After-school homework club: Puebla

Understanding context

Returning to the problem of processing contextual information, Olson addressed the difference mentioned above about the two kinds of context in a series of experiments with young children, on the so-called ‘say-mean’ distinction. The thought problem is about a precursor of school literacy. Advanced literacy, and cognitively demanding language comprehension in general, is sometimes characterised as ‘decontextualised’. We can see how the term lends itself to misunderstanding, but the idea of decontextualising comprehension tries to describe an important aspect of language awareness. Prior knowledge schemas are obviously important in processing new, textual, information. At the same time, new information can also conflict with, ‘challenge’, background concepts and culture-specific predisposition, such that accurate and reliable decoding and ‘close reading’ become even more important for comprehension. In this case, the reader (or listener) needs to apply ‘bottom-up’ type comprehension strategies to be able to decouple text meaning (what the words ‘say’) from a highly context-dependent meaning based on prior knowledge frames. The reader/listener then holds ‘the wording’ and preconception-based understanding up to scrutiny and comparison. In this study, Olson underscores the mental confrontation that metalinguistically aware readers engage in, comparing ‘literal meaning’ and interpretation driven by potential bias or by regulating authority. The ‘careful reading of words in their context’ is facilitated by language awareness (p. 56). Successful comprehension depends on both access to relevant prior knowledge and accurate decoding, as truly interactive models of reading propose (Stanovich, 2000). In our study of reading repair and self-correction/revision of compositions, an analogous process came into play: after committing an error, the most successful literacy learners prioritise precision in subsequent word identification, engaging language awareness strategies.

Redefining the concepts of writing and literacy for the purpose of blurring distinctions parallels the redefining of language with the same end in mind. There is nothing wrong in general in the use of the informal, broad, or metaphorical sense of ‘literacy’ to mean ‘ability’ or ‘knowledge’, as in ‘computer literacy’. Similarly, in the comparative study of animal communication systems, we make reference to the ‘language’ of bees and prairie dogs. Authors sometimes use the term ‘language’ to refer exclusively to relationships of meaning or to kinds of (non-linguistic) information system. But in the discussion of actual empirical research, it is necessary to distinguish between the broad versus narrow sense in these instances. Even intuitively, in ‘read the thermometer’ versus ‘read the report’ it’s clear to most people that we are talking about different kinds of skill and knowledge. Again, there is no reasonable objection in general to extending the reference or meaning range of words. Then we would simply assume that different questions and problems are being considered in the broad sense of ‘literacy’ and ‘language’, here, not relevant to the topic of this discussion: advanced literacy for higher-order classroom language proficiency and concept learning.

A similar research problem to the one of school-based literacy learning in general is that of second language learning for academic purposes, in the case of Spanish as second language in Mexico or English as international language. Here again clarity on the use of terms such as ‘reading’, ‘writing’ and ‘bilingual’ is important for better understanding the defining characteristics of high-stakes learning in school. Interestingly, the language-writing correspondence theory helps us better understand the differences between initial literacy learning in the first language and in the second language. This research problem of cross-language literacy with which we began the discussion now deserves new attention in light of the progress that has been made in the field.


Boone, E. H. (1998). Pictorial documents and visual thinking in postconquest Mexico. In E. H. Boone, & T. Cummins (Eds.), Native Traditions in the Postconquest World (pp. 149–200). Dumbarton Oaks Research Library and Collection.

Christianson, A. (2007). Popul Vuh: Sacred Book of the Ancient Mayas. University of Oklahoma Press.

Coe, M. (1992). Breaking the Maya Code. Thames & Hudson.

Finnegan, R. (2012). Oral literature in Africa. Open Book Publishers.

Fountain, C. (2015). Transculturation, assimilation, and appropriation in the missionary representation of Nahuatl. In K. Zimmermann, & B. Birte Kellermeier-Rehbein (Eds.), Colonialism and Missionary Linguistics (pp. 177–197). De Gruyter.

Francis, N. (2012). Bilingual Competence and Bilingual Proficiency in Child Development. MIT Press.

Gee, J. P. (1996). Social linguistics and Literacies: Ideology in Discourses. Taylor & Francis.

Gómez-Pulgarín, W. E. (2012). Rasgos Lingüísticos en Relatos Míticos Tikuna: Una Caracterización. (Master’s thesis). Universidad Nacional de Colombia

Gough, D., & Bock, Z. (2001). Alternative perspectives on orality, literacy and education: A view from South Africa. Journal of Multilingual & Multicultural Development, 22, 95–111.

Handel, Z. (2019). Sinography: The Borrowing and Adaptation of the Chinese Script. Brill.

Jiménez, R. T., & Smith, P. H. (2008). Mesoamerican literacies: Indigenous writing systems and contemporary possibilities. Reading Research Quarterly, 43, 28–46.

León-Portilla, M. (1996). El Destino de la Palabra: de la Oralidad y los Códices Mesoamericanos a la Escritura Alfabética. Fondo de Cultura Económica.

Manrique Castañeda, L. (1988). Atlas Cultural de México: Lingüística. Secretaría de Educación Pública.

Olson, D. (1991). Literacy as metalinguistic activity. In D. Olson, & N. Torrance (Eds.), Literacy and Orality (pp. 251–270). Cambridge University Press.

Olson, D., & Oatley, K. (2014). The quotation theory of writing. Written Communication, 31, 4–26.

Pan, J., Song, S., Su, M., McBride, C., Liu, H., Zhang, Y.,…Shu, H. (2016). On the relationship between phonological awareness, morphological awareness and Chinese literacy skills: Evidence from an 8-year longitudinal study. Developmental Science, 19, 982–991.

Prem, H. (2008). Cohesión y diversidad en la escritura náhuatl. Itinerarios, 8, 13–41.

Rosas Xelhuantzi, T. (2018). Colonización lingüística y subversión decolonial nahua en el siglo XVI. Revista NuestrAmérica, 6, 73–89.

Stanovich, K. (2000). Progress in Understanding Reading: Scientific Foundations and New Frontiers. Guilford.



Norbert Francis

Norbert Francis works on problems of language and culture, research in Latin America and East Asia. norbertfrancis501@gmail.com