Language performance disparities in South African education: a deep dive into Tshivenda success
Often, a comparative analysis based on disaggregated data reveals the essentials required for significant inflection points in the life of a nation. They sharpen the lens of policy, as the comparison will ask the difficult question of what drives the differences observed in performance. The medicine for differential performance in the South African education system is not heroic, podium-based statements but digging deeper into disaggregated data. Therein lie the answers for policy, but this seems to have attracted statesman profanity.
In this regard, I have been putting out information disaggregated by race on education that reflects the significant prowess made by Whites and Indians and the regressive nature of Black and Coloured education outcomes. I further worked on the subject by language and pointed to the fact that English speakers perform much higher than Afrikaans speakers. When you do the analysis by race and language, White English speakers perform at twice the level of other English speakers. This means that Black, Indian, and Coloured English speakers dilute the performance of White English speakers. There was no debate on this matter.
But blood was on the floor when I pointed to the fact that Tshivenda speakers perform better than any other Bantu language speakers. The groups were abuzz, frothing profanity and chanting the rise of tribalism. So, in fact, you come to the sad conclusion that presenting factual differences is taboo. Africans are so prone to being provoked by African differentiated outcomes analysed by language group.
So, the question is: if Africans refuse to look in the disaggregated mirror, how else will they ever solve the problems they face? Whether you are ill or not, you should be vaccinated because the other African was vaccinated—this seems to be the vile logic that came out of the vitriol spewed at the science of difference and the source of difference.
Was it natural that Tshivenda speakers should not perform better, as this is reserved for English speakers or Whites, where no one ever complained about the analysis? It was only when this was about African Black difference that vitriol was spewed on facts from the same study. Of course, when I circulated the reading materials, the Africans went quiet.
It is unclear whether the Africans had read the material and were preparing a natural rejoinder, or if they simply did not know what to do with the yellow bus that had been delivered to their doorstep. The dog stopped chasing the yellow bus once it came to a halt.
Vusi Mathebula, a former employee at Statistics South Africa, sent me a message and said there is a raging debate about my statement on the educational prowess of Tshivenda language speakers compared to other Bantu language speakers. So engaging and often provocative were the comments that I called for a webinar to discuss this hypersonic topic, which some claimed promoted tribalism.
I have been “dodging” the podcaster Penuel of Konvo Show over the past three years. He was eager to host me, and when he finally did, he placed me in hot water. When I called him, we had a good laugh about the protestations.
Ntate Mathata Tshedu led the charge and wrote in response to Professor Itumeleng Mosala (Tower) regarding his question on what the anger was all about. He wrote: “You may remember, about a year ago, when Pali wrote glowingly about Bop and unleashed a similar response.” I am now here to reassert the scientific basis of my observations, and I am hosting a webinar at 7pm on Monday, February10th to take on challengers.
In 1992, a number of South African academics and I, coming from Bophuthatswana Statistics, where I was the director, found ourselves converging at the University of Swaziland for an international conference on linguistics. My interest in the conference was toponymy (the study of place names) and onomastics (the study of proper names) because these are crucial in managing statistical operations in their applied sense, especially in understanding cultural economic geography—a subject I have always been keen on. I was probably the only statistician among these language luminaries.
Professor Tsietsi Mohapi, an extensive author in African languages, was there, and we spent a lot of time together. Whether that predicted our being recipients of the Unisa Chancellor’s Calabash Awards 32 years later is difficult to imagine. But last year, in November, at the Chancellor’s Awards ceremony, I reminded him about the lecture on Tshivenda in relation to other African languages.
The linguistics professor who delivered this seminal lecture had a distinct pictorial representation of how Tshivenda is a root language that projects to the north through its rich Shona roots, then to the south through its strong Sotho and Nguni roots, and to the east through its Kalanga and Tswana roots. He pointed to Venda as a versatile language that enables Tshivenda speakers to find no difficulty in speaking other Bantu languages. The reverse is not true for other Bantu language speakers, who struggle to transition, especially between Sotho and Nguni, or even within Sotho and Nguni.
Fast forward to 2016, I was pleasantly surprised by the results of performance in sciences. This took me back to the Swaziland international linguistics conference. Further inquiry pointed to competence and mastery of languages as a good predictor of competence in numeracy and mathematics, but even more importantly, this competency as a stronger predictor in programming.
In the report titled Education Series Volume III: Educational Enrolment and Achievement, 2016, I, as the Statistician-General, revealed that “Among languages spoken at home, English and Tshivenda speakers had higher chances of higher educational achievement, with 2.609 and 1.47 times greater odds, given that all other variables are held constant.” I analysed the recent 2022 Census Report by language group, and Tshivenda speakers perform at least twice as well as Xhosa speakers, 2.1 times better than Sesotho speakers, 2.2 times better than Zulu speakers, 2.6 times better than Swati speakers, and 3.5 times better than Ndebele speakers.
I hope this conclusive observation—from Venda as a root language, to multilingualism enabling proficiency in mathematics, but more importantly in programming, and the conclusive, uncontested results comparing performance in different languages—settles the missives of those who profaned the armour of tribalism while ignoring the one etched in facts. I will meet you at the webinar on Monday.
Dr. Pali Lehohla is a Professor of Practice at the University of Johannesburg, a Research Associate at Oxford University, a board member of the Institute for Economic Justice at Wits, and a distinguished alumnus of the University of Ghana. He is the former Statistician-General of South Africa.
BUSINESS REPORT