09/04/2026
Study probes grade 4 reading crisis
By Lufuno Masindi
Makhanda classrooms are some of many in South Africa where grade 4 learners cannot read for meaning. This is according to a four-year study conducted by Gadra Education in partnership with Rhodes University, and aided by trained fourth-year B.Ed students.
While the results show that Makhanda is outperforming national benchmarks, Kelly Long, Gadra's primary education programme manager, cautions that the findings also reveal persistent challenges. "We need to face it — language is an issue, and we're letting our children down."
According to Long, the fact that the study has been reliably replicated for four consecutive years indicates that the data-collection, capture, and analysis methods are valid and that the results are trustworthy. They now have a large data set to work from to understand trends and what is happening.
Drawing on a dataset of over 4 000 learners tested across all 25 of Makhanda’s primary schools — including private schools — the study confirms that the town is well above the national average. The Progress in International Reading Literacy Study (PIRLS) 2021 assessed grade 4 reading comprehension in South Africa, testing all 11 official languages, and found that 81% of grade 4 learners cannot read for meaning (below the international benchmark of 400 points), meaning only 19% can. By contrast, Makhanda’s city-wide results over four years show between 37% and 45% of Grade 4 learners able to read for meaning, more than double the national rate.
Though this is heartening, Long points out that there is a long way to go until we see equity in the system. Makhanda’s quantile system reflects inequity in the education system. Long explained this by drawing on the huge attainment gap between wealthy and poor schools. The research showed that learners in private schools and quantile five learners (wealthiest schools) are doing exceptionally well, with averages of around 90% able to read for meaning, compared to quantile three schools which average between 31% and 52% able to read for meaning.
Language of teaching and learning further highlights these disparities. “We have 22% of our learners learning in isiXhosa who can read for meaning by the time they get to grade 4, compared to 52% of children going to English quantile three schools, very few of whom speak English in their homes,” Long asserted.
To better understand whether language was a determining factor, the researchers analysed four years of data, isolating all quintile three schools and comparing results between those using English as the language of learning and teaching (LOLT) and those using isiXhosa. In isiXhosa LOLT schools, the proportion of learners reading for meaning is clustered in the low-to-mid twenties (24% in 2023; 22% in 2024 and 2026; 25% in 2025). By contrast, English LOLT schools show substantially higher achievement across all four cohorts (45% in 2023, 34% in 2024, 46% in 2025, and 51% in 2026). “Our children who are learning in isiXhosa are not doing as well as those who are learning in English from grade 1, even for children who are speaking isiXhosa in their homes,” Long said.
“Even when attempting to control for socioeconomic status, English learners are doing a lot better,” Long said, challenging the common blame on poverty alone. Infrastructure, management, and class sizes (27-37 in isiXhosa vs 34-40 in English) are on par, she noted, forcing a hard look at how African languages like isiXhosa are taught as academic tools.
Experts have long tied poor African-language outcomes to under-resourced schools, but Makhanda's data flips the script. "What it's showing us is that we have a responsibility to develop African languages properly," Long said. She critiqued rushed programmes such as Mother Tongue-Based Bilingual Education (MTBBE), where complex terms like "photosynthesis" are reduced to lazy adaptations like "ifotosinthesesi" or “iKrusti” to refer to the earth’s crust — not true linguistic growth.
Fluency metrics underscore the dysfunction: English readers typically comprehend at 60 words per minute, but isiXhosa learners can speed through 60 words per minute yet still fail to grasp meaning, suggesting flawed teaching methods. "They're barking at the text without understanding," Long said, advocating for morphology-focused approaches over traditional phonics.
This report not only informs local policymakers but also contributes to national discussions on equity and literacy. For Makhanda's educators and communities, it's a call to build on strengths while tackling inequities head-on. As we move forward, sustained partnerships and evidence-based strategies will be key to ensuring every child can read for meaning by grade 4.
Kelly Long, Gadra's primary education programme manager involved in conducting the Grade 4 reading for comprehension study. Photo:Lufuno Masindi
