What youtube has to say about the science of reading
I watched a few videos on “the science of reading” on youtube. Much of the material is about the education of children and those with reading difficulties, clearly a valuable task but not something that I am directly interested in. There is some consensus about the use of phonics and the value “phonic tasks” and the value of knowledge in reading. The role of domain specific knowledge in reading fluency is interesting to me, but I’m not sure how well it generalises to other settings. The point of listening comprehension preceding reading comprehension is interesting, and may apply to more “skilled” reading.
There’s a bit of “corporate educational fusion” going on here in that the videos are produced by companies who sell tools.
This vidoe considers reading education for those learning basic reading It separates out word cognition and language cognition as both necessary for reading proficiency.
It has a model of a number of skills and “interwining” of these skills in reading moving towards automaticity in tasks. Tasks include letter recognition, word recognition, sound recognition, gist of articles, and knowledge (broadly defined).
It claims that phonics universally beneficial and crucial for some based on research. Underlying skills of letter recognition and linking and deconstructions of phones is emphasized. Argument that this should be used before reading.
It argues that language comprehension is overlooked for young readers. Argument that introduction vocabulary on topics can be necessary to understand topics, with the concept of “knowledge-rich” curriculum.
Listening comprehension vs reading comprehension. Reading aloud is suggested as modeling of proficient reading.
It argues that morphology can aid comprehension.
Necessity of “unlearning” classic whats of doing things to improve reading.
Prior knowledge as a tool to motivation and leading skills, responsibility of building background knowledge (me: extrinsic reading). Concept a layered structure of knowledge content is created. “Prior knowledge like velcro.”
The concept of “decodable” (following phonemic rules) and “non-decodable” words is introuduced
Again some corporate influence going on here. There are a number of popsci books on reading which the lecture introduces.
This has a similar model to the above with word/phonemic understanding combined with language knowledge and the “braid model” of reading.
A neuroscience model of the brain. I am suspicious of neurobabble - i.e a neuroscience framing which is unnecessary and could just as well be framed in terms of a psychological model. But, they do have fmri scans showing sequential activation of different areas of the brain, so there is something there.
The talk claims the brain is not designed to read. Reading happens on the left side of the brain. There are claims of localisation of reading tasks in different parts of the brain. Claim that the part of the occiptal (visual word form area) lobe used for reading also recognised faces. Meaning, phonology and visual elements are separated and localised to different regions on the brain.
There are a view different claims:
Phonetic mapping is necessary for finding meaning.
Angular girus maps letters to phonemes
There is a separate context system necessary to reading
This is a slightly more technical discussion of psychological models of reading acquitsiiton.
The lecture talks about orthographic elements of reading, the mapping of writing to pronunciation. It proposes three types forms of learning that go on during reading: paired-associate learning, statistical learning, orthographic learning: types of learning and claims that these can be distinguished by performance of learners and neuronal activity.
Statistical learning is defined as individuals deriving rules for pronunciation themselves. The ability to read nonsense words indicates the existence of statistical learning.
Paired-associate is the mapping of a word to a sounds. The speaker argues that it can take an order of 20 repetitions to learn to word
Orthographic mapping referes to knowing the rules for pronunciation of certain sounds and applying this rule. The speaker claims that it takes only 3-4 exposures for orthographic mapping to work.
Thie is a claim that irregular words are still learned with mapping but with an adjustment.
test of word inefficiency: comparing common words to rarer and nonsense words.
There are separate phoneme tasks. analysis- breaking down words; and synthesis: putting together phonological parts.
People aren’t sure what skills (thing that is learned) underly phonetic *tasks* (things that you can do to phonemes). Claim that the task of segmenting phonemes is not correlated with the actual activity of segmenting phonemes.
Synthesis
Much of the science of reading understandably focuses on early acquisition and reading education. I am not necessarily interested in this, but having a vague grounding feels useful. I am more interested in expects / semiexperts / people who have things to do with reading rather than general education.
What is interesting here: the role of knowledge and the creation of layers of knowledge for learning is interesting, because it can apply more generally.
It is also interesting that a form of reading education has been simplified and spread throughout the learning community.
Read next: Emotional barriers to academic reading