Project Leaders: Dr Mark de Vos & Ms Kristin van der Merwe

The Postgraduate Strategic Early Literacy Imperative:

Linguistic and social partnerships in foundation-phase learning in African languages

• The importance of literacy
• The Crisis in South African Foundation Phase Literacy
• Tracing the outlines of the problem
• Useful literacy links
• How to apply: View call for applications which appeared on LinguistList

This project will host and provide scholarships for Postdoctoral, PhD, Masters and Honours students.  Prospective students may apply at any time to Dr Mark de Vos.  Apply now and join our growing group of dynamic and enthusiastic postgraduate students who are involved in this project.  

The importance of literacy

Literacy is probably the single most important skill for the individual and society at large. It is an empowerment tool that gives access to further education and life opportunities. Literacy “determines educational success” (Pretorius & Mokhwesana 2009:55) and is a significant predictor of success in life (IRA & NAEYC 1998). Furthermore, foundation phase literacy (grades one to four) is crucial to educational success as grade three literacy results are a good predictor of whether a learner will eventually graduate from high school (Snow, Burns, Griffin 1998).

The Crisis in South African Foundation Phase Literacy

•	South Africa scored worst of all countries polled.
•	The best South African performers were below the international average.
•	Only 13% of all South African learners achieve the minimum international benchmark and 1% achieve the advanced benchmark.
•	99% of IsiXhosa learners are illiterate in grade four.

Literacy is a central component of the economy, transformative democracy and an individual’s life competencies. In this context, it is crucial that foundation phase literacy be taught effectively and that there be structures in place that can identify reading problems as early as possible. It is then surprising that literacy provision remains an enormous challenge in South Africa. The statistics are horrifying.

In 2001, the national average for grade 3 literacy with just 38% (DOE 2003). Even in the province with the highest pass rate, only 39% passed the grade three literacy requirement (Pretorius & Mokhwesana 2009:55). Furthermore, subsequent research in 2004 showed that only 28% of grade sixes could read at or above their grade level (DOE 2005).

This literacy crisis is confirmed by the Progress in International Reading Literacy Study (PIRLS 2006) where South Africa was the worst performer out of 40 countries (including developing countries like Iran, Georgia, Macedonia, Indonesia and Morocco). “Not only did we come last, but we came badly last with a long tail of underachievement” (Pretorius & Mokhwesana 2009:55). The international average was 500 points and the best countries/territories (Russia and Hong Kong) got 565 and 564 points respectively. Even the countries which scored less than average were within 100 points of the average (only four countries scored less than 400 points). South African scored just 302 points! (PIRLS 2006:18).

When these statistics are broken down by language, the picture becomes even worse (PIRLS 2006:21). The best performers were Afrikaans and English: home-language Afrikaans and English speakers tested in their languages of instruction scored 364 and 458 respectively (PIRLS 2006:22). Note that both were well below the international average. However, scholars tested in African languages scored extremely poorly: the best performer, Setswana scored just above 250 points. IsiNdebele and IsiXhosa scored below 200 points. The remaining African languages scored between 200 and 250 points.

These statistics mean that only 13% of South African grade fours reach the minimum international benchmark of 400 points. This can be contrasted with 98% for Russia and the international mean of 94%. Only 1% of South African grade fours reach the Advanced International Benchmark. Most shocking of all is that only 1% of IsiXhosa, SiSwati and IsiNdebele learners reach the minimum international benchmark by grade four. This means that 99% of these learners are illiterate after three years of schooling.

Tracing the outlines of the problem

Because reading straddles “linguistic, neurolinguistic, cognitive, psychological, sociological, developmental and educational domains” (Pretorius & Mokwesana 2009:55), the roots of the problem are consequently diverse and multifaceted. However, the focus of the current project will be the numerous linguistic dimensions of the issue. Four inter-related linguistic components will be discussed here, namely: prestige and dialect, orthography, automaticity of word recognition and the lack of normative data.

Linguistic Dimension 1: Prestige and Dialect

First, there is the question of which dialect of the language has been standardized and how this relates to linguistic identity, prestige and orthography. Dialectal variation among indigenous African languages is much more prevalent than that within either English or Afrikaans. The result is that even speakers of a single language such as IsiXhosa may find that their spoken home languages are divergent from the orthographic standard. Nevertheless, we know relatively little about dialectal variation and there are very few dialect maps, known isoglosses, and descriptive studies of particular dialects.

Another issue relates to linguistic prestige. It is well known that the African language lingua franca in southern Gauteng is IsiZulu – nevertheless, it is not a prestigious dialect and students who speak Johannesburg IsiZulu sometimes claim not to speak IsiZulu because of the normative pressures associated with the rural KwaZulu prestige variety (De Vos, p.c.). This means that learners who are speakers of a non-standard variety may find it significantly more difficult to master literacy in the standard.

Linguistic Dimension 2: Orthography

The question of orthography also plays an important linguistic role. It is well known that all the indigenous African languages are agglutinative in nature – informally speaking, this means that a whole sentence can be expressed in a single word – and that consequently the nature of the “word” differs from the “word” in English (Guthrie 1948, Louwrens & Poulos 2006, Prinsloo 2009, Van Wyk 1995). This means that words are multisyllabic and much longer than their English equivalents: a clause consisting of monosyllabic words -- “the cat sat on the mat” -- is simply not as easy to do in a language like IsiXhosa. In turn, this has consequences for ease of acquisition or word recognition and automaticity (see Linguistic Dimension 3: Automaticity and vocabulary development). It also has implications for lexicography – and the paucity of research has resulted in poor dictionaries (Prinsloo 2009, Gouws 1990) which, in turn, are poor classroom resources.

This is complicated by the fact that languages differ according to whether they have conjunctive or disjunctive orthographies. The result is that orthographies may impose word boundaries that are different to the psychological realities of the speakers (a similar example occurs with the phrase “as well” in English, where speakers pronounce – and write – it as a single word reflecting a psychological reality different from orthographic reality). When orthographic and psychological word boundaries are misaligned, this may also negatively impact literacy acquisition.

Linguistic dimension 3: Automaticity and vocabulary development

As a learner gains proficiency in reading, recognition of individual words becomes automatic, considerably speeding up the reading process. This allows deeper and more critical reading as well as the use of contextual and semantic cues as “automated word recogntion frees mental resources for closer consideration of the meaning of a text and thereby allows readers to employ reading for the acquisition of new information and knowledge” (Verhoeven 2009:38). The importance of automaticity is paramount as there is a variety of evidence showing close relationships between automaticity, vocabulary size and text comprehension (Verhoeven 2009).

As implied by the previous section, orthographies, especially the conjunctive ones used for IsiXhosa and IsiZulu, cause an increase in the mean length of words making word recognition and automaticity more complex than for an English learner. The nature of the indigenous languages, compounded by conjunctive orthographies could result in a relatively high number of “orthographic neighbours” (i.e. words that differ by relatively few letters) which have been shown to inhibit reading speeds (Perea & Rosa 2000; Andrews 1997, Grainger 1992); to date these effects have not been explored in African languages. Even if we were to make progress in this area, there remains substantial study to be done in how South African learners link orthographic cues to contextual and semantic/inferential cues which have been shown to affect first and second-language learners differentially (Fitzgerald 1995).

In the South African context, there is an additional complication that many learners are second-language learners and/or the dialectal situation renders them thus, even if they are putatively learning within their first language (see Linguistic Dimension 1: Prestige and Dialect). However, there is very little research on how orthographic environment affects acquisition of a second language (Simon & Van Herreweghe 2010). Urquhart and Weir make the following claim about second-language learners but it is potentially also applicable to learners of conjunctive orthographies: “the additional processing time they require for word recognition may have a knock-on effect in other areas of the reading process” (Urquhart and Weir 1998:189). Additional anecdotal evidence that conjunctive orthography may slow down the reading process comes from English, interestingly enough. It has been shown that heavy NPs (i.e. multisyllabic and multi-word Noun Phrases) are a major impediment for second-language learners (Eskey & Grabe 1989:226). To the extent that this finding can be extended to African languages with conjunctive orthographies, it may shed light on aspects of the current literacy crisis.

Furthermore, while it is accepted that second-language learners read more slowly, very little is known about “emergent literacies” (Urquhart and Weir 1998: 190) and even more importantly, we have no idea about what exactly “slowly” means in this context or what a good reading speed for indigenous languages even means (see Linguistic Dimension 4: Normative Data).

Linguistic dimension 4: Normative data

Also of critical importance is the lack of adequate normative teaching and assessment resources. Regular, contextualized formative assessment is also critical for successful literacy acquisition. English teachers can draw on a deep body of international research on age-appropriate reading speeds, vocabulary size norms, word recognition norms and even tools to determine the reading ages of particular texts (e.g. the Flesch-Kincaid test is embedded within Microsoft Word). In contrast, none of these exist for any indigenous language (excluding Afrikaans). Even for Afrikaans, which has benefited from decades of apartheid, for half a century there was only a single literacy assessment tool developed in 1944 – it was only revised in 2002! (Esterhuyse, Beukes & Heyns 2002). The fact that the concept of the “word” differs parametrically between languages prevents the norms of one language being applied to another in an ad-hoc fashion; for example, what does a reading speed of 60 “words per minute" mean to a speaker of isiXhosa, which is written in a conjunctive orthography? That such norms and resources would be beneficial has been demonstrated: in a four-year intervention in an under-resourced school, word-recognition rates increased as teachers became more aware of what rates of achievement were possible – and achieved them earlier and earlier in the year (Pretorius & Mokwesana 2009).

Despite the fact that foundation phase literacy in general and literacy in African languages in particular is problematized there is a paucity of research on African language literacies. Pretorius and Mokhwesana (2009:55) talk of a “virtual absence” of research and point out that there is no journal on the African continent dedicated to the issue. They argue passionately that literacy in African languages will determine the fate of these languages: “the future of African languages in the 21st century relies to a large extent on the extent to which speakers of these languages willingly and effectively read and write in these languages. Investing in the future of the African languages starts with the development of reading in African home languages in primary schools” (Pretorius and Mokhwesana 2009:55).

Other dimensions

Although the project will focus on the under-researched linguistic-decoding dimensions of foundation phase literacy, this is not to deny the other axes of the problem. For instance, much research – including sociolinguistic research – has focused on the socio-economic axis of literacy: the family is a central site of the reproduction of power relations and families induct children into ways of knowing and ways of reading that affect their success in later life (Heath 1982). These ways of literate meaning-making are further refined and entrenched in the South African schools system (Prinsloo 2009) . For example, children from middle class families are exposed to a print-rich environment and are encouraged to not only read for content, but to adopt a critical subjectivity that questions the text; children from lower-income groups, are encouraged to read for content only and to regard the knowledge in books as being handed down from a higher authority, etc. (Heath 1982).

There is also an educational/developmental axis including school, teacher and classroom-based factors (Pretorius & Mokwesana 2009). These may include the strength of parent-teacher relationships, regular formative assessment, early interventions, good classroom management, effective communication between teachers, engaging in authentic reading tasks (Pretorius and Mokwesana 2009), infrastructure, the relatively large sizes of South African classes, lack of textbooks and libraries, teaching and reading resources, the relatively little time spent on reading in South African curricula and the quality of that time etc (PIRLS 2006), the quality of teacher qualifications (IRA & NAEYC 1998) and the central role of the teacher in making tactical, strategic, contextual decisions (Eskey & Grabe 1989:228—22).

Useful literacy links

• The Shine Centre
• Institute for the Study of English in Africa
• Programmes to increase literacy in Africa