We appreciate that the relationship between a book and its reader depends on hundreds of contextual factors. The same book may be perfect for a 5-year-old in one context and a 7-year-old in another context. However, levelling is a necessity.
A broad levelling framework is required so that buyers can find the right books. Even loose guidelines are better than total anarchy.
Idea and Grammar
Our basic thesis is that books become harder on two broad dimensions – idea and grammar. BICW places books on a grammar complexity ladder while correcting for idea complexity and context.
From a mathematical perspective, levelling is a process of placing elements on a linear scale. In general, this can be done in two broad ways – absolute and comparative.
Absolute levelling would involve giving every book a value, for example, 23.54 and 27.66 and thus concluding that the second one is harder than the first. Here’s a simple example – suppose we take all the words used in a book and measure how many of these are from which Fry list and create a weighted score, then that would be an absolute score. This method has two obvious flaws – (a) Non-Fry words that our children are aware of, like samosa, bazaar, Nani, Thatha . . . may cause the score to spike. (b) The measurement is only for words used, and not the ideas presented. Suppose there is a book that deals with a difficult topic, like death, with very very simple words. Does that make the book simpler and, therefore, appropriate for kindergarten?
BICW levelling uses a comparative method. For example, our process tells us that Treasure Island > Gruffalo, but doesn’t assign a score to either of them. We do not focus on giving the book a “difficulty number.”
If you’re feeling particularly nerdy today, do look up the Elo Chess Rating system. In our early days, we were heavily influenced by the mathematics and philosophy behind the Elo system: it awards points based on comparisons.
Trial and Error
The bedrock of BICW levelling is trial and error. We use the same levels as BLPS, a program for schools that encourages reading. BLPS has been levelling books since 2010. They take books to class and use the reaction to understand which of the two books in a pair is easier/harder. BLPS works with more than 500 classrooms every year, so they get ample opportunity to hone the levelling. This approach also makes the levelling process more human and less algorithmic.
A Work in Progress
We believe that experimentation is the most pragmatic way to make levelling work. There are books we expected to perform at a certain level but turn out to be easier. And others, harder. A core tenet of experimentation is that the experiment may fail. So, we also constantly re-check our levelling. There are instance when we have re-levelled a book based on new feedback from classrooms.
Level – Age Matrix
As per our experience, BICW levels correspond to the following age recommendations –
|Level||Regular Readers||Advanced Readers|
|L1||Age: 5||Age: 3|
|L2||Age: 6||Age: 4|
|L3||Age: 6||Age: 4|
|L4||Age: 7||Age: 5|
|L5||Age: 8||Age: 6|
|L6||Age: 9||Age: 7|
|L7||Age: 10||Age: 8|
The art of reading should tread ground familiar enough to be engaging, but still slightly unfamiliar. The reader should aspire to understand harder grammar and imbibe new ideas.
Taking a page out of Goldilocks, the ideal book is like the perfect bowl of porridge: too hot and it discourages interest, too cold and it turns stagnant; but just right and it succeeds.