How Do Reading Structures Affect Reading Proficiency?
By: Curtis Kelly
There is something very important that we have discovered about the brain: its main form of processing the world is by predicting. Processing sensory input from zero is just too hard for that little organ, too wasteful of cognitive resources, so rather than sorting through millions of models for each bit of input, the brain relies on predicting instead.
If you are in Starbucks, you expect to see Starbucks drinks. The mental models for the Starbucks drinks are already activated in your brain even before you walk into the store. Then, when you see a person holding something small and white, with just minimal sensory input, your brain stops looking and tells you that it is a Starbucks coffee cup, filling in the rest of the details top down, i.e. brain to eyes. The cup you “see” is only partially seen by your eyes, with the rest being passed down to your sensory areas from the part of your brain that manages internal mental models. Your eyes catch a vague outline and your brain automatically fills in the rest. In that way, it can shift its cognitive resources to other things. The way our brains paint the world is so efficient, so fast.
In that particular situation, being in a Starbucks café, your brain was aware that an encounter with a Starbucks cup was likely. After all, that was exactly what you experienced many times before in other Starbucks shops. Your brain uses Bayesian inferencing to determine that, in a Starbucks shop, something small and white in someone’s hand is likely to be a cup, but is also aware that once in a while this might be a napkin or cell phone instead. Once you decide, your eyes confirm. Your brain did not predict the small, white object would be a rabbit, so if it is, you do a double-take during confirmation, derailing cognitive processing, and spending a lot of energy readjusting your perception. (To learn more about this amazing process, see the MORE video and our Think Tank on Predictive Processing.)
This skill in predicting allows your brain to instantly identify anything it is seeing, hearing, or experiencing with just a smidgen of sensory input. Interestingly, your brain does this for language as well. It predicts.
At the sentence level, the starting words in a sentence help the brain activate models that tell it what the following words are likely to be. If you hear a start like The boy ate…then the possibility of what is coming next is instantly reduced to a much smaller word set, mostly food (But if the next words are an unexpected 9, 10, 11, you again do a double take). In other words, grammar itself is a tool our brains have created to aid predictive processing. So are common phrases and idioms, like get out of hand, make a projection, and be careful. These chunks of language, that we gain automaticity over, help us predict and thereby reduce processing. If you are a language teacher, I think you understand exactly what I am talking _____.
Reading is also a prediction game, which is why certain structures and styles make this incredibly difficult task easier to accomplish. And some of the most important structures live at the discourse rather than word order level. After all, there is much more to being able to read a page of text than just being proficient in the language (as any of us who read research papers know). There are also much larger discourse and content structures that help your prediction machine. They help you pre-activate likely models of language and meaning, the same way being in Starbucks made you ready to encounter a Starbucks drink.
I don’t know if anyone has ever listed these larger reading structures in one place, so I will attempt to do so off the top of my head, knowing I will miss some. (In fact, some of the things in my list came from reading the contributions to this issue.) So, a list of reading structures that help us predict what we will read next, the process of filling in between the lines, include:
1) discourse structures (expository, narrative, etc.; paragraphs and topic sentences; transition words; titles & subtitles, and so on) that give us familiar frames to fill in, with signposts along the way. See my article on discourse structures here. And watch our MAIN Video.
2) genre, author’s writing style, or plot type help us guess where the piece is going (thanks, Nobuko and Amanda).
3) content that relates to prior knowledge that allows us to simulate the situation by activating the right models for the content (maybe this is the most important).
4) common phrases that allow chunking and automaticity to take place.
5) the related sound system that work with grammar to convey information (thanks Meredith, Harumi, and Jamie)
6) illustrations that help us visualize the content.
Do you see where we are going with this? It is common for teachers to teach reading by mainly having learners work on vocabulary or grammar (as I did), unaware of the importance of these larger, less visible structures. But what we now know about the brain tells us that just focusing on basic language proficiency is not enough. The larger structures are important too, especially because many of them are not shared across languages.
We at Think Tank Headquarters are ecstatic that we found contributors to teach us about these extra dimensions of reading. We guarantee that if you read this issue, you will come away with a different understanding of what reading is and how to teach it. Likewise, as you read on, pay attention to how all those factors in the list above affect your own comprehension and ease of reading. Go for ___!
Curtis Kelly (EdD.) is a professor emeritus of Kansai University, founder of the JALT Mind, Brain, and Education SIG, and producer of the MindBrainEd Think Tanks. He has written over 30 books and given over 500 presentations. His life mission is “to relieve the suffering of the classroom.”