According to the 2001 census, Indias literacy rate for the population, aged seven and above, was 65.4 per cent. What does this number really mean? Can 561 million people, that this rate implies, read a newspaper headline in their own language? Not really. What it means is that households across India reported 65.4 per cent of its members to be literate, when the census fieldworker showed up. The literacy rate is a perceptual number people perceived to be literate. It is not an accurate indicator of the proportion of readers in the population.
What if the national census actually tested for reading ability? We did just that, with a sample of around 20,000 people drawn from 3,200 randomly chosen households in four Hindi states Rajasthan, UP, MP and Bihar.
First we followed the census approach. Then we asked every household member, aged seven and above, to read a simple text in Hindi, of 35 words, that a student in class three would be expected to read. Those who could read it correctly, at their own pace, were marked as readers. Those who could read only parts of it, or took recourse to sounding syllables before putting together words, were classified as early-readers. The rest were nonreaders, who could not read at all.
The census approach gave us a literacy rate of 68.7 per cent in the sample. The reading test, in sharp contrast, resulted in 26 per cent readers, 27 per cent earlyreaders and 47 per cent non-readers. Even if one were to club the readers and early-readers, at best 53 per cent could be considered to be readers. The census method, thus, led to an overestimation of the literacy rate, in the Hindi states, by a whopping 16 per cent. Why is there such a big overestimation?
First, in the 1990s, the National Literacy Mission did a remarkable job of drumming up interest in literacy and started off millions of people on the path to literacy. Once someone acquires beginning alphabetic knowledge, that person becomes literate in family and self-perception, for life, and therefore, in the census. The 1990s added nearly 100 million perceptual literates, permanently to the census. Many of them, however, never quite attained functional reading ability or relapsed quickly into nonreading in a lifelong sense.
Second, our data show that 90 per cent of those who completed first grade, were automatically reported as literate. First grade completion is now very high among children aged 6-14, because enrolment itself, nationally, is over 93 per cent. So whether a child can read or not, if you can get her to enrol and complete first grade, she immediately joins the ranks of the literate. Yet, our testing found that, at first grade, less than 1 per cent were actually able to read a simple paragraph, 27 per cent read it like an early-reader, and 72 per cent could not read at all. Even after the completion of grade five, 26 per cent could not read at all and only 12 per cent could read it comfortably.
These two reasons explain why the literacy rate is galloping but not the ability to read. For the latter to improve, national policymakers would need to draw upon innovative strategies that can make lifelong reading, inescapable at a mass level. One such strategy that we have been advocating for national policy adoption, is Same Language Subtitling (SLS). Essentially, SLS is the idea of adding karaoke-like subtitles to film song-based content on TV, in the same language as the audio. SLS is well-researched and proven to improve reading ability, is cost-effective, and causes automatic and lifelong reading.
SLS allows a school-going child to pick up emerging reading skills in school and right away practise them at home. This constant interplay of school learning and home practice of an essential skill, such as reading, deserves more policy attention. In the census we have 260 million so-called literate people who cannot read. National progress ultimately depends on their ability to read, not our ability to call them literate.
The writer is an adjunct professor at IIM, Ahmedabad.
What if the national census actually tested for reading ability? We did just that, with a sample of around 20,000 people drawn from 3,200 randomly chosen households in four Hindi states Rajasthan, UP, MP and Bihar.
First we followed the census approach. Then we asked every household member, aged seven and above, to read a simple text in Hindi, of 35 words, that a student in class three would be expected to read. Those who could read it correctly, at their own pace, were marked as readers. Those who could read only parts of it, or took recourse to sounding syllables before putting together words, were classified as early-readers. The rest were nonreaders, who could not read at all.
The census approach gave us a literacy rate of 68.7 per cent in the sample. The reading test, in sharp contrast, resulted in 26 per cent readers, 27 per cent earlyreaders and 47 per cent non-readers. Even if one were to club the readers and early-readers, at best 53 per cent could be considered to be readers. The census method, thus, led to an overestimation of the literacy rate, in the Hindi states, by a whopping 16 per cent. Why is there such a big overestimation?
First, in the 1990s, the National Literacy Mission did a remarkable job of drumming up interest in literacy and started off millions of people on the path to literacy. Once someone acquires beginning alphabetic knowledge, that person becomes literate in family and self-perception, for life, and therefore, in the census. The 1990s added nearly 100 million perceptual literates, permanently to the census. Many of them, however, never quite attained functional reading ability or relapsed quickly into nonreading in a lifelong sense.
Second, our data show that 90 per cent of those who completed first grade, were automatically reported as literate. First grade completion is now very high among children aged 6-14, because enrolment itself, nationally, is over 93 per cent. So whether a child can read or not, if you can get her to enrol and complete first grade, she immediately joins the ranks of the literate. Yet, our testing found that, at first grade, less than 1 per cent were actually able to read a simple paragraph, 27 per cent read it like an early-reader, and 72 per cent could not read at all. Even after the completion of grade five, 26 per cent could not read at all and only 12 per cent could read it comfortably.
These two reasons explain why the literacy rate is galloping but not the ability to read. For the latter to improve, national policymakers would need to draw upon innovative strategies that can make lifelong reading, inescapable at a mass level. One such strategy that we have been advocating for national policy adoption, is Same Language Subtitling (SLS). Essentially, SLS is the idea of adding karaoke-like subtitles to film song-based content on TV, in the same language as the audio. SLS is well-researched and proven to improve reading ability, is cost-effective, and causes automatic and lifelong reading.
SLS allows a school-going child to pick up emerging reading skills in school and right away practise them at home. This constant interplay of school learning and home practice of an essential skill, such as reading, deserves more policy attention. In the census we have 260 million so-called literate people who cannot read. National progress ultimately depends on their ability to read, not our ability to call them literate.
The writer is an adjunct professor at IIM, Ahmedabad.