Don't teach your child to read
This text is part of: "Problem of Schooling" by Piotr Wozniak (2017-2024)
Introduction
The entire system of compulsory schooling rests on the foundation of reading. Learning to read begins in the first days at school at the age of 4-8, depending on the country. Allegedly, without school, the entire planet would collapse under the burden of illiteracy. The opposite is true. The psychological damage inflicted on children with early reading and early math lays a foundation for mental health problems later in life. In this text, I explain why we should immediately stop teaching kids to read (unless they want to be taught). The mountain of evidence is overwhelming. All it should take to convince you is a bit of open mind.
Compulsory schooling must end.
The main things I want you to know:
- all healthy children will learn to read sooner or later
- there is no educational advantage to early reading (unless it comes naturally)
- unwanted help in learning to read is usually harmful
- most of dyslexia today is educational dyslexia, i.e. a result of schooling
- today, computers are the best tool for learning to read
Reading instruction discourages reading
My work over "I would never send my kids to school" was inspired by children depressed before the school start at the end of summer vacation (2016). Over the last couple of years, I noticed that children start hating school earlier and earlier. These days, it is not unusual for a kid to say "I hate school" before she goes for her first class. This happens despite the fact that in Poland, we raised the age of entry from 6 to 7 in recent years. I managed to identify the main culprit of school hate at this early stage: reading instruction. Due to the popularity of YouTube and computer gaming, children are less motivated to learn to read. They can start exploring the world early, long before they read their first sentence. While children find new ways to learn, adults keep living the old myth that reading is the gateway to all knowledge. This creates a vicious feedback loop. The more kids like electronic media, the less they like the print, the more the adult world panics, and the more pressure is put on children to learn to read early.
Reading might be the area where the difference between schooling and free learning can be seen most painfully. It is the area where the pain of schooling is greatest, and yet the area where it is almost impossible to free the mind from dogmas inculcated at school.
Then the second myth kicks in. If reading is hard, it is allegedly due to neglect in the early years. Children learn the letters of the alphabet at the ever earlier ages and are pushed to read long before they are able to make any good use of the simplest texts. The earlier they start learning, the more problematic it is, he more they hate learning, so the adults keep tightening the screw of early reading. This is sheer madness that provides the basis of school hate that is likely to begin earlier and do much more damage than in prior years. This produces a conflict of generations: children do not want to learn to read, parents get more and more obsessed with reading.
However, school hate is not the only side effect of the race to early instruction. There is an epidemic increase in various diagnoses associated with learning disability. It is not that our children got more stupid. The increase is almost entirely associated with the increase in the pressure for early education. In case of reading, we have an epidemic of educational dyslexia (reading difficulty induced by the stress of schooling). A vast majority of dyslexia diagnosed in the first years of school is a result of the pressure to read. Educational dyslexia can be remedied, but often is not. The scars of early schooling can stay with a student for life.
When I say, "Do not teach your child to read", I do not want parents to give up on reading. I just want them to save a great deal of time and stress, by letting the child learn on her own. Today, learning to read is easier than ever, and it should never be a source of worry or hurry.
The optimum strategy is to wait. We should wait until children ask for help, or perhaps more likely, until they learn reading on their own. The experience of democratic schools or unschooling indicates that reading may arrive very late. Peter Gray and others estimated the average onset to be 9 years of age. For example, learning to read at the age of 13 or later stands at 7% chance in this estimate (for unschoolers). However, the arrival of new interactive technologies might lower that seemingly pessimistic delay. I will explain how reading emerges naturally, and why this emergence leads to better reading skills.
If this text is too long, and you want a 4 min video summary, check Dr Peter Gray's opinion. You can also check more readable Seven Principles of Learning to Read Without Schooling.
How do we learn to read?
The ignorance of the art of reading in society is biblical. I blame school for this. If you ask an average citizen about how we learn to read, he will point to school, exercises, reading books, phonics, learning syllables, chanting the alphabet songs, etc. In the US, some may point to the whole-word method. Very few people can point to the essential building blocks in the process of learning to read.
There is only one central and valid effort that leads to the ability to read: it is the quest for understanding the world in all contexts where visual patterns of print stand in the way to comprehension. All parallel learning processes (e.g. learning the sounds of individual letters) may assist the central thread, but their deployment must occur on demand where all subservient processes fit in the main goal of comprehension. All artificial approaches employed at school or in individual reading instruction may weave individual threads suboptimally. This may result in suboptimal architectural changes in the way cortical wiring is laid down for efficient reading.
All readers tend to converge on their ultimate expected reading fluencies given a reasonably painless learning process. We can then hope that all errors in the art of learning to read can be remedied with subsequent training (i.e. primarily a great deal of reading). However, all departures from the optimum are wasteful. On one hand they take additional time from the learning process, and require additional learning in the recovery period. In the end, we cannot be sure that all changes are reversible. We cannot be sure that suboptimal wiring leaves no permanent mark. There are no doubts when self-orchestrated learning process occur. There are many swimming styles, but we all tend to arrive at a variant of freestyle. The shorter the path, the likelier the optimum wiring.
In terms of the concept network of the brain, the art of reading aims primarily at connecting visual pattern recognition maps with semantic maps. This means that we want to decode the meaning behind the print.
However, the process inevitability involves concept maps needed for speech generation. This means that while learning to read, we also usually learn how to convert text into sounds.
Here are some primary directional associations involved in reading:
- visual→semantic: visual pattern (e.g. DOG), and the semantics (the concept of a dog that lights up in the concept network); this is the only association essential for reading
- visual→audio: visual pattern (e.g. DOG), and the pronunciation of the word DOG, which can then be associated with the semantics; this association may assist reading in some contexts
- visual→motor: usually based on semantic intermediary (see semantic→motor below)
- semantic→audio: used in speech generation, not essential in reading, develops as the byproduct of the process
- semantic→visual: visualization of print; a byproduct of reading
- semantic→motor: writing skills, typing skills, speech production; a byproduct of reading
The second association (visual→audio) is of lesser importance and may be skipped in dyslexic or fluent readers. In reading with comprehension, all we care is a link between visual patterns and semantics (so called semantic mapping). If the child cannot read aloud, he may be branded dyslexic. However, as long as the child knows what she is reading about, she got the core art of reading. As patterns, sounds and semantics are closely interlinked, "understanding without reading" is rare and disappears over time (in a healthy learning process). In a dyslexic, it can persist for life.
The prime importance of visual→semantic decoding can be seen in reading without comprehension where a student can convert text into speech without understanding the meaning of the read text. The echoes of erroneous learning trajectory can be seen in readers who need to vocalize in order to comprehend. Their brain decodes the sounds before it can decode semantics. It is as if the student read to himself only to decode the meaning from sound with a bit of delay. The skill of converting text to sound is also useful in life. Many TV anchors excel at that. It is hard to learn to read without the ability to vocalize. However, the ability to understand without reading and without the ability to vocalize characterizes many dyslexics. If this "disability" survives till adulthood, we have clear evidence of the errors in wiring the concept maps involved in reading. Some communication channels are clearly hidden from mainstream processing.
.jpg)
Figure: Brain structures involved in reading. Visual text input travels from the retina to visual areas (in pink). Letter concepts are decoded in the Visual World Form Area (VWFA)(in red). Semantic concept maps are dispersed in various areas of the cortex (not marked in the picture). The remaining areas in blue, yellow and green are often involved, but are not necessary for decoding the meaning. In reading Braille, instead of the visual cortex, tactile areas are involved, but the output signals may converge in VWFA as well. Source: Pegado F, Nakamura K and Hannagan T (2014) How does literacy break mirror invariance in the visual system? Front. Psychol. 5:703
Knowledge is the key to reading
The key to effective language processing is world knowledge and semantic decoding. It is not the speed of decoding characters, words, or photo-capturing paragraphs. It is not the speed of rushing through strings of text. This linear race is what kids mostly learn at school in reading classes. What matters is the prompt construction of meaningful semantic trees in working memory.
As reading relies on knowledge, there must be a vast bank of knowledge to rely on, and a vast bank of interests to uncover. Vast knowledge is missing in little children, and this may be the main stumbling block. The problem is less associated with the cortical wiring needed to make all the necessary associations. Robert Pondiscio explains the role of knowledge in this lecture.
The optimum trajectory towards efficient reading does not go through early reading, but through self-dosing of the reading challenges in proportion to sensory development, world knowledge, interest, etc. In short, kid's brain powered by the learn drive is the best optimizer. It will, sooner or later, discover the value behind the world of written word and master individual skills in proportion to the needs.
If the kid picks his own channels of information, he might actually gain more. Watching videos might bring more knowledge of the world and more benefit than slow text decoding. Once brain networks and skills catch up, so will the flow of information.
Early reading does not offer much advantage in that area. It won't' make people faster readers. The difference will level in a matter of years. Early reading might alter brain network to facilitate lexical processing, or serialized reading, but these are not limiting factors to fluency.
Babies, toddlers, and kids that cannot read, all soak in knowledge and get ready for being fluent at reading. It is wasteful to make kids go slowly through lengthy books while their reading skills are poor. At the same time, they can capture knowledge at much greater speed and volume through other channels: communication with parents, peers, video, computer games, and more. The channel of the written language will open up in proportion to development. It will be highly individual. It will depend on intellectual capacity, disposition, interests, and many other factors. It will be surreptitious. Minor events can bring major passions and the skill of reading may appear seemingly overnight.
Problems at school
There are a couple of problems that children face at school when learning to read under the supervision of an adult. Stress and coercion usually come from insufficient error tolerance, insufficient room for error correction, excessive cognitive load, etc.
Error tolerance
In the process of learning, the tolerance of error is extremely important. In neural networks, achieving higher accuracy may come at a cost that is increasing exponentially with a drop in error rate. The cost of a misspelling maybe negligible. The quest for accuracy should be proportional to goals. The accuracy of speed-reading or incremental reading may be low, but the flow of new value is very high. It is important to let children seek their own optimum without a pressure to be accurate, or to meet any other benchmarks.
In 100 bad habits learned at school, I explain that in learning, the brain needs to be treated as a concept network as opposed to a tape recorder. Pattern recognition should rule the reading. In that process, semantics is the king. Teachers often insist on the primacy of "reading with comprehension", while the only goal should be comprehension. Substituting "pony" for "horse" is an example taken from the whole language approach. Such an error indicates good decoding of the semantics and should not be subject to systematic eradication.
In the same way as it is the case with neural networks, concept network learning can be compared to an incremental tuning of an old cathode-ray tube TV set. Therapists may intervene early to remedy dyslexia. Each time perfection is demanded, stress, harm, and unlearning are likely. Peter Gray argues that once the pressure to read is taken away, the symptoms of dyslexia may ease, and natural learning to read may proceed naturally, even though slowly.
In the ideal model of perfect education, a perfectly adaptable human brain uses the best channels of communication, and controls knowledge acquisition by the brain's learn drive system. In this model, optimal conceptualization in the concept network of the brain would produce a highly applicable model of reality. In the case of reading, the model of the printed messaging system.
Using the tools of compulsory schooling, children are forced into letteracy instead of conceptualizing the true nature of the universe. On a smaller scale, the same happens in learning to read. Letteracy undermines comprehension, while ultimately, in most cases, leading to readers of comparable fluency.
Error correction
Adults who read fluently do not fully understand how error correction in reading works. Some clues come from artificial intelligence. It is also helpful to watch little children read on their own on condition nobody is disturbing or providing guidance, and the child volunteers to read aloud. Reading aloud is probably rare in pristine voluntary environments due to lesser emphasis on connecting print with speech production.
Today, it is also possible to see error correction in action in some speech analysis software, esp. speech-to-text applications. On the first run through a sentence the brain or intelligent software takes the first take on the likely message. The brain then looks for inconsistencies based on grammar, word knowledge, knowledge of the world, context, etc. In case of software, analogous process may be based on sheer statistics of machine learning.
At first the brain fixes the errors that are most glaring or most obvious. The new approximation is closer to the actual message and the process can repeat until a stopping criterion is encountered. In humans, this usually is the realization that we know the meaning of the message. On occasion, the stopping criterion may also be the realization that some errors or comprehension gaps cannot be fixed due to insufficient knowledge or due to errors or imperfections in the message itself.
This process was insightfully described by Ken Goodman in the 1960s as "psycholinguistic guessing game". Sadly, Ken's ideas have been dragged through mud of slander as a result of reading wars.
The emergent phenomena are rarely understood or taken into account at school. Instead, a child is reprimanded for errors, and set on a fast journey towards educational dyslexia. For more on school's war against heuristics see: 100 bad habits learned at school.
Skill emergence
The incremental assembly of memory patterns is not a thing we can investigate in real time. We lack the tools to investigate processes that span multiple networks and take milliseconds.
However, we can amazingly see it in SuperMemo in time-lapse. If we ask a child to spell a difficult word at the very early age, she will fail. The assembly process will be incomplete, and the end result will not be falsifiable by separate memory patterns (e.g. visual representation of the word). If we ask the same child the same question 2 months later (without hints), we may get a wrong answer again. However, if the answer differs, it will reflect the creative process terminated differently.
Here is the sequence of spelling efforts collected in the span of 4 years for the word "horse" (duplicates shown in parentheses). This illustrates the brain's effort to assemble the correct spelling of the word over time. The correct spelling emerged on the 24th review after 3 years and 2 months.
roqest, hplle, hszz, holre, hold, hsrs, honhtd, hofk, hofda, hotq, horda, hsee, hols, horss (2), hores (2), hors, herss, horse
The assembly process occurs in the brain independently of SuperMemo. Spaced repetition does not work for kids due to childhood amnesia (and other factors). In here, SuperMemo only plays a role of a clock triggering recurring review (research value only). However, once a correct memory forms, review in SuperMemo makes sense as it verifies the correctness of the consolidated pattern. Interestingly, this process is also characterized by frequent sudden emergence. When "readiness" is achieved, review becomes flawless. This illustrates that there is no "reading readiness" per se. Readiness refers to each individual memory pattern taking part in learning (see: Jigsaw puzzle metaphor).
In contrast, at school, the kid is likely to see the word "horse" and is asked to write it down many times on a page as in a classic Prussian drill. In those drills, memorizing the correct spelling is not a hard thing, however, due to the immature memory substrate, suboptimum memories are likely to form. Those will later need to be remolded via generalization.
To maintain such a pattern, repeated drills will be needed to sustain it. Lack of practice will result in forgetting, which assists generalization. A kid may struggle with spelling his own name again and again, and even 3 years of drills will bring no result after just 1-2 month break given to let forgetting do its job.
While the brain attempts to correct the error of schooling, school attempts to preserve its gains. This back and forth ends when an unwanted memory pattern solidifies, or the class runs out of time and moves on to the next level of learning.
This process is wasteful. It consumes child's time, brain energy and affects passions. If toxic memories form, we actively work to discourage schoolwork, writing, or reading, or worse.
Cognitive load
Reading seems extremely easy to readers. Once all basic memory blocks are formed, juggling words and sentences is a breeze. However, before those building blocks are formed, the reading process is like assembling a complex jigsaw puzzle. The earlier the stage of the learning process, the bigger the size of the jigsaw puzzle to assemble to extract meaning. The problem is that before long term memories stabilize, the puzzle needs to be assembled in working memory.
It is a not a set on the table that allows of systematic resolution of mini-puzzles and keeping the big picture in perfectly reliable "puzzle on the table" memory. If there are too many pieces in working memory, cognitive load may result in an increase in error rate. It will also accelerate fatigue. For high cognitive loads, a child may get mentally tired literally in minutes. If the entire process is not spontaneous, and is being pushed by an adult from above, the tolerance for errors may be forcibly limited. The time for error correction may be reduce. Those factors compound complexity, and hinder learning. Fatigue sets in much faster. However, if fatigue was the only problem, it could easily be resolved by a bit of rest (or sleep). However, in high cognitive loads, the assembly of the puzzle may be suboptimum. Wrong memories are formed. Those hinder the process in the long run. Worst of all, if the whole process involved stress, the child may start developing toxic memories, i.e. association of individual memory blocks or mental challenges with anxiety or fear. This is how educational dyslexia develops.
The problem is compounded in whole language approach in where the jigsaw puzzle gets more complex and needs many approaches from different directions before an easy memory block can be found and consolidated. Little wonder that proponents of phonics want to equip kids with the set of universal master keys to memory. In natural learning, all the necessary keys do form spontaneously. Cognitive loads are less, but the trajectory is extended (more attempts are needed at each step).
In addition, in multilingual children, learning to read in more than one language can be a nightmare (unless it is voluntary). Not only two giant separate memory structures need to be built. They also need to be largely independent. Those memory sets tend to be entirely separate at first (for the sake of computational complexity), and yet they need to integrate with the same semantic set of concept networks. For that to happen, the brain needs to be equipped with the mode switch, which tells the networks how to route signals depending on the language mode. As concept map activation is not a process that can be extinguished in milliseconds, switching between languages may be a taxing process. While the brain needs to set up a new fire, the embers of the "wrong one" are still attracting attention. A newly trained reader will laboriously switch between languages as if this was a transition from swimming to running in duathletes.
It easy to recognize the problem of excess cognitive load. For example, if the student gets corrected at the start of the word or sentence, stumbles seconds later in another place, is asked to repeat the process, and appears to have already forgotten the starting correction, we know the load is excessive. The child is not dumb, defective, forgetful or dyslexic. He simply got too many irons in the fire. The same kid, given complete freedom, may be entirely automatic on the exact same word or sentence just 1-2 years later without ever trying or being taught.
Unschooling
In addition to neuroscience, unschooling and democratic schools bring a great of evidence that illustrates the natural emergence of reading in self-directed education.
Democratic schools
In democratic schools, children usually learn to read sometime in between the ages 6-13. Some may need longer depending on their interest, and the degree of dependence on text. Learning can occur with some assistance from a tutor, from a friend, or on one's own, in various proportions, depending on the child. Very interesting observations came from the extreme case of the Hole in the Wall experiment in which children could learn almost exclusively from a computer, or from volunteers contacted via the Internet. Peter Gray collected a great deal of anecdotal evidence from unschoolers who learn to read in various ways using comic books, cereal boxes, street signs, etc. In all those learning environments, we need to provide a decodable link between visual print patterns and the meaning (semantics). The link can be provided by human assistance, or it can come from any other source (e.g. short messages in TV adverts that combine text and visual or audio feedback).
Freer Speckley
In democratic schools, with no pressure and no structured instruction, virtually all kids learn to read, however, Freer Speckley is a case that left the confines of the Summerhill school without the ability to read (at the age of 15). Freer was a self-confessed problem kid who lost his mom early, but he claims the actual reason he did not learn to read was his interest in carpentry and pottery, and the lack of pressure from school or his father. He attended Summerhill in the years 1955-1963, i.e. long before the availability of interactive computer technology. After leaving school, Freer hitch-hiked twice around the world, and the only times he was bothered by his lack of writing skill was when he had to fill out forms when crossing borders. He would use illegible script and get away each time. In the end, he visited 90 countries and held various jobs in many of those.
He ultimately learned to read on his own. When he was a teacher of English conversation in Japan, he was asked to spell "caravan", and this was a tipping point that made him decide to learn to read. He learned the skill in 3 months on his own. Now he got a degree in economics and uses his traveler's past to work as a consultant in many countries around the world. Today, Freer says that he would be worried if kids took the same path.
The case of Freer shows that there is no guarantee that every child will learn to read. A bit of motivation and some contact with printed matter feedback are needed. More importantly though, even if the child shows no interest in reading, he will not be trapped by the alleged critical period window. Adults can learn to read too. They can do it on their own or with instructions. All they need is a bit of motivation. With a nearly universal access to the Internet, cases like Freer are far less likely these days.
You can read about Freer here or listen to his story in his own words here.
The myth of the critical period for reading is so powerful than it biases the reasoning of those who encounter evidence such as Freer's case. When I spoke about Freer to a reading expert, she remarked "in every field, there are geniuses, and he might be just one in a million". Clearly, Freer was inconsistent with the critical period model and this is what instantly drew the attention of the expert. Instead, she should read a message that is far better documented and more important. Of hundreds of cases of children in democratic schools, we do not have even one that would not learn to read when the need arises (if you heard of any, please let me know). Ironically, Freer who was diagnosed as super-smart by our expert, would have probably suffered years of jeers at school as a "reading retard".
Jack did not read
There are millions of illiterate people on this planet. They usually lack conditions or motivation to learn to read. However, before the arrival of computer games and the internet, opportunities to learn reading on one's own were limited. Carpentry or metalwork where the hobbies that kept boys away from the books at Summerhill. In his book, Alexander Neill documented another earlier case of a boy who left the school without the ability to read. Despite Neill's description, the kid was not dyslexic. He learned to read later in life:
There was Jack, a boy who could not learn to read. No one could teach Jack. Even when he asked for a reading lesson, there was some hidden obstruction that kept him from distinguishing between b and p, 1 and k. He left school at seventeen without the ability to read. Today, Jack is an expert toolmaker. He loves to talk about metalwork. He can read now; but so far as I know, he mainly reads articles about mechanical things and sometimes he reads works on psychology. I do not think he has ever read a novel; yet he speaks perfectly grammatical English, and his general knowledge is remarkable. An American visitor, knowing nothing of his story, said to me, “What a clever lad Jack is!”
Dyslexia in free learning
What is most remarkable about learning to read in free learning is that there are no documented cases of dyslexia emerging in free environments.
Even in reference to school, John Taylor Gatto wrote:
In 26 years of teaching rich kids and poor, I almost never met a “learning disabled” child; hardly ever met a “gifted and talented” one, either. Like all school categories, these are sacred myths, created by the human imagination. They derive from questionable values we never examine because they preserve the temple of schooling
Robert Pondiscio observed a different thing:
I never had a single student who couldn’t “read.” Put a piece of text in front of them and they could all (some with greater fluency than others certainly) verbalize the words in front of them, or “decode.” What they couldn’t seem to do consistently and competently was to discuss or answer questions about their reading. They “read it” but they didn’t “get it.” They could decode, but not comprehend
Peter Gray documented several cases of children whose reading problems dissipated once they were taken away from coercive environments: How Dyslexic Kids Learn to Read When Removed from School. Danny Greenberg reported no cases of dyslexia in Sudbury Valley School in nearly six decades of school's existence. This inspired Je'anna L Clements to research the topic and write What If School Creates DYSlexia?. My own reasoning about educational dyslexia is derived from the concept of toxic memory, which itself can be studied best with spaced repetition. The reason for the value of spaced repetition is that it ruthlessly exposes all poorly formulated learning material. If misemployed, SuperMemo can also lead to toxic memory. The analogy to educational dyslexia is direct.
Professor Joe Elliott is a critic of the use of the term dyslexia as he can points to lack of ways to detect symptoms of a neurological pathology that would differentiate dyslexia from a sheer difficulty in reading. His crystal-clear reasoning finds most opposition not among scientists, but among teacher lobby and parents with their sacramental "you should spend a day with dyslexic children" (spoken to an expert who has devoted 4 decades to the subject). In this light, throughout this text, dyslexia should be understood simply as a reading difficulty.
Optimum learning
For centuries now, reading seemed like a gateway to wisdom. This is why literacy has become a central point of early education. Debates on the optimum learning method have been waged for decades. Today, however, we can cool down. Kids can learn about the world using a variety of methods. Long before they can read, they can learn from YouTube. They are also fortunate in that their mom has access to ChatBots, and can verbally answer vastly more questions that kids of past geniuses could ever hope to have answered by their parents. Finally, with speech recognition and synthesis, they can chat with AI on their own at younger and younger ages.
Reading is natural
Proponents of direct instruction and phonics often mention the fact that humans have evolved without print therefore our brains are not equipped with tools needed to learn reading "by osmosis".
In reality, a healthy brain is equipped with all that is needed in the process of recognizing print and assembling it into meanings that match the spoken description of the model of the world. The model of the world we first build interactively and then we add its spoken description by learning to speak.
Absurdity of the evolutionary claim can best be revealed when we observe that we evolved without cars, and yet most of kids learn to drive in simulation games. They do it easily and with pleasure. We evolved without smartphones or computers, and yet the elderly pick the smartphone skills at ease (given undamaged faculties). Kids learn how to program by just watching other kids' program on YouTube. Check the case study of Renata in this text to get to know a mom who learned to read on her own at 4, and daughters branded "dyslexic" or "special needs" who also learned largely on their own. See Peter Gray reports (e.g. "How children learn to read" video).
Reading is not a high-level cognitive skill. Kids with learning disabilities are also capable of learning to read. Kids with Down Syndrome can learn to read despite a serious organic handicap. Probably more than half of those children improve their reading to the point of reading stories (see Pearl read a book). Their reading ability often goes well ahead of what might be expected by their general knowledge level. The Rule #1 is the same for healthy kids and the DS kids: "Have fun!".
Children with hemispherectomy show how adaptable the brain is. If half-a-brain is enough to read, given early adaptation, why so many parents worry their kid will never read? Why so many teachers claim that without their help children will remain illiterate? It turns out that the main stumbling block on the way to fluency is school. It discourages at best and leads to educational dyslexia at worst.
Ironically, proponents of coercive schooling claim that without compulsory schooling, illiteracy would thrive. The opposite is true, school puts many people off from reading for life.
Whole language
Whole language approach works through "discovery of meaning through experiences in a literacy-rich environment" (source). Wikipedia makes it less glamorous: "teaching children to use context cues to guess the meaning of a printed word".
The whole language approach to reading is explicitly described as "discredited" on Wikipedia. This assertion is substantiated by an extensive list of references.
However, whole language indeed should play a dominant role in learning to read. Both in natural learning and in the whole language approach, pattern recognition is the key. Individual letters and/or sounds are rarely strung together. Patterns in words are recognized (e.g. "printd word" is easily recognized as "printed word"). Written context and non-language context also play a role. For example, "c AT" is easily read as "cat" if it is accompanied by a picture.
Supporters claim that learning to sound out words is boring. They are right. Opponents says that taking away the cueing habits is essential for fluent reading because fluent readers do not need cues, and recognize words faster than pictures. They are wrong because they discount the generalization that is an essential part of conceptualization in all forms of learning. Kids may use cues today and become fluent readers tomorrow by enhancing fast pathways and discarding those that slow them down.
Phonics
Phonics is based on teaching the associations between letters and sounds. Printed text is then decoded serially. Phonics is the simplest way to quickly push a kid into reading. It dominates coercive schooling around the world. Phonics seems like an obvious way to go in phonetic languages (e.g. Italian, or Polish) where there is a nearly one-to-one correspondence between letters and sounds. However, phonics is a straight path to reading without comprehension. A child can become a robot who can decode any text, even when she has no idea what she is reading. A typical symptom of troubled reading is vocalization. It develops when the brain attempts to discipline the serial decoding process. A disruption in vocalization is clearly perceptible (unlike a break in silent reading). This provides extra incentive to automate reading even if comprehension does not follow fast enough. Natural readers do not vocalize, and they are also less likely to hear a voice in their mind while reading. This arises from the fact that whole language allows of non-linear progression through the text, incl. skimming known of speed-readers.
Phonics seems superior
Phonics seems easy. You can teach a 2-year-old to recognize letters. Soon she can sound them out. This provides some groundwork for self-learning. By the age of 4, the kid might learn to read on her own. The same kid at school is less likely to fall into the dyslexia spiral. In Poland, where phonics is the way to go, by the age of 10, most kids read well.
However, usually this rosy scenario does not work out. The other end of the prognostic spectrum is educational dyslexia.
A little kid forced to learn phonics will get discouraged. In the UK, if the child does not master reading by the age of 7, he is likely to enter a spiral of dyslexic feedback loop (see Harriet Pattison's lecture)(minute 4:27):
- the less you read, the more they work on you
- the more they force you, the less you like reading
- the less you like it, the less you read
- the less you read, the worse you read
- the worse you read, the more you fear the ridicule of peers
- the more you fear ridicule, the more you avoid all situation where reading plays a role, etc.
This comment complements it:
As a dyslexic person I can't explain how traumatic it was for me always being in special classes, working longer and harder than everyone else and feeling attacked and drilled by every teacher. Never a moment peace. Geuss what, none of it helped. None of it. What it did was give me mental health issues as a child. Only when I got to high school did things get better because I was old enough to say no. My test scores jumped up and I got my first A in English. I went on to get a B in in English A level. Be kind, be gentle and don't force kids to do ten times more work than every other kid. Test scores mean very little in the real world and they will find ways around their dyslexia. Encourage them to be creative, explore words and they will learn
When in a downward spiral, the diagnosis of dyslexia is a relief. The child feels the relief. The parents and teachers feel the relief too. The whole world got more reasons to study "brain problems" of kids who are slow to read. Billions of dollars go into a fake problem. Biological origins of slow reading may account for less than 1% of dyslexics (see the reasoning of Frank Vellutino).
The alternative is to let the brain figure phonics out on its own. On the way, adult help might be requested. That's great. If not, we should celebrate the child's self-reliance.
The whole language approach that us based on natural 3-cue system described by Ken Goodman is a slow process. However, this is a natural process that involves no frustration. Frustration is the basis of educational dyslexia.
For most kids, phonics is excruciatingly boring. In Poland, it takes 1-3 years of drills to get to the goal of reading. Polish is largely phonetic. English is way harder. In English there are 1,100 sound combinations to tackle (see: Geography of dyslexia). In tougher languages, kids need either extra curiosity in learning, or extra time to learn. The same knowledge may be gained by playing interactive computer games involving text. With pleasure.
Any kind of pushing or even encouragement can result in lesser enthusiasm for reading. That's a shot in the foot!
I said that in Poland, at 10 most kids read. However, by adulthood, 70% of them show serious comprehension deficits. The learn to read early, and they quit reading in adulthood en masse.
Dehaene champions phonics
Stanislas Dehaene is an award-winning neuroscientist. He came up with an important term of neuronal recycling, which is pivotal for understanding the conceptualization process in the brain.
Dehaene specializes in studying the brain as it learns, in particular, how it learns to read. He came up with a great number of novel ideas about the reading process. Few people know more about how individual brain structures cooperate to produce the magic of reading.
However, Dehaene still errs in his claim that phonics is the best way to learn to read. I think his key error is to reason in terms of schooling. If we are to force kids to learn early, and we are to pick a better method, he might be right that phonics will result in less frustration and less risk of generating dyslexia due to a buildup of toxic memory.
However, Dehaene does not seem to take into an account the concept of natural learning where a kid does not ever seek strategies. He may learn some phonics intuitively but is rather involved in pattern recognition where not only whole-word strategy is involved, but also whole-phrase or even whole-context learning takes place. Computational complexity of such a learning job is significantly higher. No wonder teachers would never opt for anything similar in the classroom. However, in natural learning the kid will cut the complexity by instinctively using Goodman's 3 cue system.
A free-learning child may easily read "cat" in the context of cats and react with bepuzzlement in the face of the same word without context. Again, the brain plays the game of slowly tuning in the TV set signal until a clear picture emerges (see: generalization). A powerful side effect is the ability to understand without reading, i.e. comprehension based on pattern recognition without the ability to read or even recognize individual words. This is an excellent strategy that leads to intelligent reading optimized for comprehension in prevalent reading contexts. The conceptualization will produce different outcomes for reading from YouTube, comic books, Wikipedia, or programming code. In other words, programmers will read code better than ordinary fluent readers, but may turn out slow in reading plain text.
Dehaene observes that phonics competence predicts the speed of learning to read. This is correct. However, at the very basic level, letter-sound associations are only a microscopic fraction of knowledge needed for reading. This knowledge is acquired naturally and early while in contact with synchronized visual and phonic representations of words and sentences.
Dehaene claims that the adult whole-word reading is an illusion.
The whole-language approach has nothing to do with how our visual system recognizes written words – our brain never relies on the overall contours of words, rather it decomposes all of its letters and graphemes in parallel, subliminally and at a high speed, thus giving us an illusion of whole-word reading
I do suffer from this "illusion" indeed, but if you peek at the reading experiment later, you can draw your own conclusions without the need to study brain images.
Brain imaging shows faster progress in kids who study with the use of phonics. They may show faster changes in the brain. However, this does not imply the strategy is optimum. Brain scans do not say a word about wiring optimality. They don't say a word about the speed of individual pathways, cross-interference, and resulting preferences that affect fluency and comprehension.
Dehaene worries:
Experiments even suggest that the whole-language method may orient learning towards the wrong brain region, symmetrical to the visual word form area in the right hemisphere!
I say that if many areas of the brain are involved, increase in fluency will result in trimming and consolidation based on performance reward. Why worry?
It may be easy for a swimming instructor to suggest optimum forms of motion that maximize swimming speed. Olympics swimmers have an edge over your average natural swimmer, but we have little impact on how brain wires its circuits in learning. Those who read naturally and with pleasure turn out best readers in the long run. See also: educational dyslexia.
Metaphorically, the suggestion to use phonics is similar to asking a student in a walking class to learn to bike only because bike will get him faster to the neighboring town. Biking will not prevent learning to walk but does not assist walking much per se. What is worse, if a kid develops toxic memories of educational dyslexia, he may fear stepping off the bike, and will continue reading serially, letter by letter, word by word, with difficulty and with anxiety. The road to natural reading may be blocked!
Emily Hanford champions phonics
Emily Hanford is a journalist who started a campaign against the "harms" of Ken Goodman's theory of whole language learning. Hanford's campaign in support of phonics commits an error of judgement derived from the claim that "fluent readers do not use cueing, and what is worse, cueing is characteristic of struggling readers". She observes correctly that a skilled reader reads the printed word "book" faster that he can recognize a book from a picture.
The erroneous conclusion is that cueing is unwelcome in the learning process as it might lead to the development of wrong reading habits. However, this error stems from the old educational misconception, which claims that learning has the form of imprinting the "correct" neural pathways in the brain, while fluency comes from "strengthening" those "correct" pathways and making them more "efficient/faster".
I reality, learning is based on the conceptualization process in which generalization is based on a constant optimization of pathways dependent on the performance reward. This is why cueing should be seen as a prop assisting the early stages of setting up the neural framework for associating printed patterns with semantics.
If you consider that in fluent reading, you need to recognize 50,000 word-patterns in an instance, the job of reaching fluency is a monumental memory job. Conceptual props like phonics are helpful in early stages, but the same goals can be achieved using different props, e.g. computer games. In the end, semantic mapping is the brain's goal (mapping text to meaning). Orthographic mapping (associating sounds-letters-meaning) is a superset needed in a wider range of language skills, and grapheme-phoneme mapping may be seen as a prop or as one of the ultimate goals of learning, or anything in between. The brain of the reader will optimally make the best choices in all those mapping permutations and shortcuts.
Metaphorically, Hanford's reasoning is akin to discouraging 2-ball juggling due to the fact that it might establish bad habits needed in 20-ball juggling. Unlike in swimming, there aren't many reading styles with different optimization trajectories. Best readers show similar saccadic patterns with pattern recognition based on photographic mini snapshots of text. As a result, the fluency is more a matter of practice than strategy.
It is also important to emphasize than in her criticism of 3-cue strategy, Hanford uses examples, where the strategy itself is erroneously employed. For example, a book that juxtaposes pictures with text may be missing a vital ingredient of learning: the need to decode text to obtain the meaning. If the meaning is instantly extracted from the picture, the learning may not occur. Unlike an artificially prepared learning book, ordinary comic books may play a better role. In a comic book, the focus of the author and the reader is on the story. Learning occurs as a side effect. Such learning may take a multiple of minimum time needed to register the equivalent well-schooled progress, but those are the hours that the child will never consider schooling. Those hours are fun.
Syllable method
There are 15,831 syllables in English (the list). No wonder, learning individual syllables is not popular and is considered part of phonics.
The case is different in phonetic languages, where the number of syllables is less, and the rules of pronunciation can be derived via generalization from a smaller set. In Poland, the syllable method is all the rage, heavily marketed, and a pretty solid competitive business that involves the diagnosis of dyslexia, therapy, courses for therapists, and a sale of therapeutic educational tools and toys.
Opponents claim that the method is responsible for the torment of many children and the explosion of dyslexia in Poland. Proponents claim that the method saved many children with reading difficulty. The truth is in the middle. Well-used with an expert, the method may complement simple phonics and whole language with pleasurable reading. In the hand of an empathetic expert, all roads lead to Rome. All kids learn to read. However, most "experts" and parents resort to a degree of coercion and this is invariably harmful.
As always, distilled skills detached from the need for world knowledge tend to be boring (unless well compensated by an expert hand). Experts work with syllables in contexts, but parents often limit the drills to syllables only (occasionally enhanced with pictorial associations). What is worse, many parents believe that starting at 2 is a form of dyslexia prevention.
For how the syllable memorization does not work, I have evidence from SuperMemo again.
Many words in Polish begin with przy- (e.g. przygotować, przyglądać, przyprawa, przygoda, przydatny, etc.). It stands to reason then that learning the syllable "przy" makes sense (esp. that "przy" makes a separate word as well). The syllable is useful. I can show in data a failure to learn in 23 successive trials. The child is asked to read "przy" and fails each time in 23 trials in the course of therapy sessions spanning 2.5 years.
And then an incredible turnaround happened. The same child who cannot read "przy", decodes correctly the word "przydatny" (useful) and is suddenly able to read "przy". On the same day, suddenly, "przy" is easy and comes to mind instantly. I verified this with a bit of amazement. It worked each time. Fail at "przy", and success after reading "przydatny". This proves how futile learning can be when it occurs without context (in here, the very limited context of a specific word).
Reading wars
Reading wars started at the dawn of compulsory schooling. Horace Man was an empathetic man (despite fathering all evils of today's school). He considered learning the alphabet as a boring chore, so he justly proposed the whole-word method. His opponents attacked viciously, and thus reading wars started.
When Ken Goodman described his theory of whole language (1967), the wars intensified with people like Frank Smith promoting natural ways of learning. Three cues approach, or balanced approach became popular. Tom Durrie's book The end of intelligence as strongly impacted by Frank Smith's philosophy.
To ask "Is phonics better than whole language?" is not much smarter than to ask: "Which is more important: the brain or the heart?"'. We need both.
We should rather see the war as the battle between schooled way or natural way. We all know the differences between learning a spoken language the natural way and the schooled way. At school we can learn English in a few months. But if the same if achieved in computer games, the outcome is usually vastly superior (esp. in younger kids). What's most important, natural ways never produce a guy claiming "I have no talent for languages". This is a toxic memory equivalent of a dyslexia in the area of languages. The same guy may speak his native tongue fluently in the same way as a dyslexic may be surprised that reading in a foreign language may turn out easy.
Reading wars are largely about what to force kids to learn first: phonics, sight words, cues, or else. They have nothing to do with the pleasure of learning. They are a product of a deficit in educational empathy
Optimum strategies should be obvious. Kids should learn on their own at their own pace using their own methods.
Whole Language, besides being more consistent with a philosophical commitment to active learning, is also “the best phonics program there is,” in the words of Jerome Harste of Indiana University. Why? Because it’s easier to decode a word when you already know what it means. By virtue of insisting that meaning should come before skills, Whole Language isn’t just more enjoyable but also more effective
See: Reading wars are over: Whole language vs. Phonics
Reading networks
At the root of reading wars and passionate campaigns (as that of Emily Hanford) is poor understanding of concept networks. The simple world of linear learning is still attractive to educators who see education in terms of the assembly line that filles up the empty brain bottles with knowledge (as it was poetically observed by Holt).
Horace Man, John Dewey, Ken Goodman, Frank Smith, and Tom Durrie came from very different walks of life, but rooted their understanding of the process in meticulous observation of children as they learn.
The same conclusions can be drawn by understanding the feedback loops in a conceptualizing brain in the process of adaptation to the environment.
Reading is neither a bottom up nor a top-down process. It is a process of multiple feedback loops occurring both at the level detectable consciously and underneath. The meaning, the letters, the words, the sounds, or the computation will dominate depending on the context, e.g. the level of novelty, level of difficulty, etc. However, the goal is to maximize the analysis of content and minimize the effort of the decoding process that becomes automatic in advanced readers. In fluent reading, vast knowledge network involved in reading is amplified by the knowledge of the world. This networked and instinctive interplay should be the model of the ideal process in which reading emerges from the effort to find meaning.
Ken Goodman identified the "three cuing system" that is involved in the art of early reading. The three systems are: (1) the graphophonemic system (phonetics), (2) the semantic system, and (3) the syntactic system. However, the system of cues is vastly extended in the modern multimedia world. Today, learning to read is easier than ever, even if the incentives keep dimishing in childhood.
Reading wars come from the fact that in the learning process there are multiple networked streams of activity in the brain: top-down, bottom-up, decoding, guessing, skipping, multi-level pattern recognition, semantic feedback, conceptualization, audio cueing (e.g. when reading on YouTube), and many more. Those phenomena are poorly understood by the proponents of learning strategies.
Proponents of phonics want to see a solid groundwork in orthographic mapping, which is a good foundation for reading. Hence its popularity. If skipping and guessing is disallowed, the strategy becomes a form of coercion. Balanced literacy is a way to come close to the natural learning process, but proponents of phonics will call it whole language due to insufficient emphasis on phonics, while proponents of natural learning will still see it as phonics due to an unnaturally high injection of phonics instruction. There is not a single name that would provide an optimum balance because such balance does not exist. The balance is set by each child's brain on its own. It differs from task to task, and from minute to minute.
The war is over in that we know how reading works and how learning should proceed, however, the battles and skirmishes will continue as there is plenty of warriors who cannot wrap their minds around the complexity of the networked processes in a conceptualizing brain.
You are a whole-word reader
All proficient readers use rich pattern recognition when they asymptotically approach their maximum reading speed. That speed is determined by comprehension and reasoning about text.
The following simple experiment shows that you are a whole-word reader too.
Example 1: do you understand this?
chin eseandj apan esecitiz ensm ustlearnth ousan dsofcha racte rsand cha racterc ombin atio nstofun ctio ninso ciety
Example 2: do you understand this?
chineseandjapanesecitizensmustlearnthousandsofcharactersandcharactercombinationstofunctioninsociety
Example 3: do you understand this
Chineeze amd Japenese cityzens musst learrn thouousandz of harakters end harakter kombynetions to funktjon yn sociyety
Some readers are pretty good at decoding Example 2. However, there is always a hint of bad habits of schooling (incl. phonics drills). Perhaps an exception should be made for those who known Thai (see: Thai needs no spaces).
For a vast majority of good readers, Example 3 is easiest. This proves that all reading is based on pattern recognition. We know more or less how words look like. Kenneth Goodman called it a recognition based on "graphophonemic cues". The brain fills the remaining gaps with syntax, semantics, and general knowledge of the world (pragmatics).
Best way to learn to read
To learn to read, the student needs to see print in contexts when its meaning is important to achieve goals. The difficulty of print must match current reading skills. This means it must start from simple words like GO, STOP, YES, PLAY, etc. This cannot be a science book. This cannot be a "See Spot Run" booklet either. The science book is too complex. While the whole-word booklet has no meaning for the reader as it may include just 10% of the vocabulary used by the child.
Over the last decades, a perfect tool has been developed to assist reading without the cost of instruction, without stress, and with the pleasure of productivity. That tool is interactive computer technology. Today a 3-year-old can browse through mom's phone. Later on, she can play with a tablet, watch YouTube, or play simple computer game. Most games include print. This may not be as easily available in languages other than English, but even this dominance drawback is gradually being remedied by rapid expansion of gaming into non-English markets.
While a kindergartener is being bored to death in alphabet drills, a free kid equipped with a tablet will have regular interactions with all the letters of the alphabet. While direct instructions may create an illusion of fast progress, it is the prime way to discourage young kids to the concept of school and learning in general. In the meantime, a free kid will develop slowly and systematically without ever throwing a frown.
It is hard to say how many kids today benefit from computer technology because it is hard to separate the benefits of computers from the impact of direct instructions at school. While kids are bored with drills, they do translate them to tangible benefits while working with computers. On occasion, such feedback may actually result in the love to learn to read. More often, a reverse phenomenon happens, school discourages reading and the child subconsciously avoids tasks that involve reading. Computer games that involve reading get deprioritized. Educators swear that natural reading methods as described in the presented text are harmful fiction.
In addition, parents often fear harms of screentime, and a tug of war begins (see: Dangers of imposing screen time limits on children). Parents limit screentime, school discourages reading, the child craves gaming more, and drifts towards more soothing entertainment, which gives games its bad reputation as thwarting development. Prof. Manfred Spitzer scares the world with a prospect of digital dementia (see: The morbid myth of Digital Dementia).
How long should parents wait if they see no progress? How can they withstand complaints from school, family, or even other children? The waiting period is unpredictable. It depends on interests and the motivation. The rule of thumb is that until the child specifically deems reading beneficial, progress may be negligible. As reading relies on knowledge, there must be a vast bank of knowledge to rely on, and a vast bank of interests to uncover. This is rare in little children, and this may be the main stumbling block (as opposed to cortical wiring needed to make all the necessary associations). If there is no interest, one can just observe how the kid interacts with technology and if print puts any obstacles on the way, and if obstacles get tackled efficiently. If so, learning to read is a matter of time.
While parents may be reluctant to wait in the face of social pressures, it is worth to remember that the child's brain choses its learning strategies optimally. It picks the best sources, methods, and time. If reading is slow to come, it is simply not a priority at that given stage. If you have doubts, read Optimality of the learn drive.
Peter Gray reported an exemplary case of a late unschooled reader in the words of her father:
My daughter did not read until she was 10 and not to her satisfaction until near 11. We read to her every day when she was little, she saw us reading, and she had plenty of resources. Books, papers, magazines, cereal boxes and her favorite – computer games – were always available. Even hearing words from the TV and movies helped. She made connections from seeing the printed word and hearing the words spoken; she had context. She not only read at a college level by 12. She has become a writer and grammarian since
Optimum age to learn to read
When less informed educators push for early reading programs in preschool, they hope to raise a more literate generation, but instead expose kids to the risks associated with early instruction. Such moves are opposed by experts. Lillian Katz (b. 1932) is an erudite educator and a prime expert on early learning. She keeps warning governments around the world that compulsory early reading is a very bad idea (see: Lively Minds: Distinctions between academic versus intellectual goals for young children).
Optimum age for reading is not much different from the optimum age for learning multiplication or learning how to use a map. Optimum age is as absurd as the optimum time to learn the Beer's law in physics. Some people learn Beer's law at 3, some never learn it. In the same individual, Letter A may be familiar at the age of 1.5, while the distinction between U and Y can be made stable only at 7. Should that smart kid learn to read at 1.5? No! That would sure lead to toxic memories (unless under zero coercion). Perhaps at 7? No! Why wait? Knowing A may accelerate learning B.
As long as it comes naturally, early is good but late is also nice, and always better than never.
Natural interest in reading or in geography may come at any minute and the trajectory is unpredictable as is the learn drive. One movie can make the kid interested in biking, and then in mechanics, or woods, or anatomy. From there, further opportunities determine the radiation and the differentiation. This process cannot be easily controlled. Siblings of similar age, sex and personality, and in the same household may diverge wildly (e.g. one may read at 4, the other at 11). Parents can try many tricks to coax kids into reading. As long as these are not coercive, I see little harm. The ultimate outcome is also hard to predict. Late readers may become bibliophiles, while early readers may change allegiance and become programmers.
The learn drive is unpredictable and overrides are unwelcome. Child's choices are the best guidance. If a kid loves to read books at 2, why not. If he hates it at 10, why force it. The areas of strength will develop elsewhere and will reflect the demands of the environment. If parents do not provide a shining example, interaction with other adults may help. Or perhaps that lack of example is a reflection of societal demands. Kids of reticent parents may be more reticent, but listening is a virtue and a strength too. That's what Larry King would tell you.
Peter Gray's research shows a span of ages from 3 to 15 where most kids learn to read in democratic schools. Harriet Pattison found even a bigger range: from 1.5 to 18 (that 1.5 might be just the ability to read the language of emoticons). The actual numbers are secondary. Kids differ and their needs differ. The conceptualization process reflects the environment and the chaotic impact of events on trajectories. The moment we observe kids to read is often sudden and easily associated with specific events, which have a form of major passions associated with printed matter.
For independent minds, the first good sign is an urge to produce texts (e.g. to text family members or to comment on YouTube). Reading usually follows suit. It may also be the other way around. Reading a comic may come years ahead of the urge to preserve memories or to communicate in writing.
Critical period
A harmful myth says that if you do not master reading early, you may never become a fluent reader, or might even be doomed to dyslexia.
Baby brain
A baby brain is a tangle of connections. It is a mess. It needs to be shaped. Lots of that shaping occurs via synaptic pruning. Contrary to popular belief, the child does not predominantly keep sprouting new connections as a result of new experience. The opposite happens: unused connections get thrown away to make the brain more efficient. Adult brain is bigger and yet it may carry fewer neurons in many of its vital structures. Some heavily pruned areas lose half of their brain cells, and most of their axons.
Critical period is not just a simple "use it or lose it" phenomenon. It is a specific period of brain rewiring in which neural darwinism affects the functionality of the cortex. It is not that unused areas of the cortex atrophy. Just the opposite, it is the time when neural competition results in prolific invasion of new connections in critical areas needed to process sensory inputs.
Cortical wiring
The myth of critical periods for reading originates from the Nobel-winning research by Hubel and Wiesel on critical periods in the development of sight in little kittens. If cats are deprived of vision in babyhood (e.g. with their healthy eyes obstructed), they may lose sight capacity for life. The research shows that portions of the sensory cortex that are not used in development may be invaded by functionality of a complementary organ. For example, cortical areas devoted to an obstructed eye may be taken over by neurons responding to the unobstructed eye. When kittens are artificially deprived of vision in one eye, in critical period, even for a day, there is an increase in the width of cortical columns in the primary visual area of the cortex associated with the non-deprived eye (layer IV). At the same time, there is little change in the lateral geniculate nucleus that sends signals from the eye to the cortex, even though LGN is little used in the period of deprivation.
In addition to identifying an ideal model for critical period research, Hubel and Wiesel also illustrated the importance of cortical proximity in the efficiency of processing in cortical computation. For stereo vision, the input from the eyes is interwoven in tiny areas of cortical surface.
As critical periods refer primarily to the sensory cortex and the early long-distance wiring, we are perfectly capable of learning using various cortical resources later in life. Critical periods will primarily encompass periods of rapid dynamic change and specialization (e.g. monocular deprivation can lead to a complete eye dominance within 10 days). On the conceptualization axis from plasticity to stability, the critical period is limited to the most plastic extreme, and rapidly progresses towards largely irreversible stabilization that closes the period. This illustrates that fact that a global conceptualization curve is just a resultant of multiple parallel processes at various areas of the brain that depend on age and use. The neuro-ontogenic process begins at 2-3 weeks of pregnancy with the formation of the neural tube. The sensory conceptualization beings at the time when the first sensory signals get wired into the brain (e.g. proprioception, auditory signals, smell, taste, etc.). It is important to note that sensory and motor areas of the cortex may differ in the expression of plexin receptors, which provide for different modes of stereotyped pruning (i.e. pruning of axonal collaterals in response to use). Most sensory critical periods terminate early in childhood and underlie the potential for further learning later in life. The stabilization phase of a critical period may be reversed experimentally by disruption of perineuronal nets. This indicates that the mechanisms of stabilization might be analogous to those that extend over the lifetime, with the timescale as the main differentiator between the two.
White-matter development
While dendritic sprouting provides for adaptation throughout life. White-matter development will stagnate in early adulthood. We can say that the range of new things we can learn in our nineties is largely set in the first two decades of life. That's no reason to worry. This means that whatever you are able to learn at 20, you will probably still be able to learn at 100 (assuming excellent health). The speed of learning may not be the same, the ability to generalize may be less, but all the areas of the brain that are usable and well-connected at 20, should provide a similar range of opportunities at 100.
It is very important however, that the childhood is rich in experiences. Ideally, in youth, we should explore all potential environmental niches we might encounter in adulthood. Kids drive in that direction naturally. However, adults tend to limit the range of exposure. In the mind of the designers of the Prussian school system, wide exposure to multiple book subjects in a classroom is a great way to develop. In reality, playing in the woods, playing computer games, or playing football might provide for a wider basis for establishing a healthy brain architecture.
An interesting case can be made about the early reading instruction. The timing and the method may affect the ultimate proficiency. Luckily, most kids will become proficient readers by just reading a lot. We should only worry about those who were discouraged to read at school.
Figure: White-matter development model. The relative developmental trajectory for pruning and myelinization may affect the ultimate density of white fibers in brain areas associated with reading proficiency. The readers who proceed fast with myelinization may end up with a lower plateau of proficiency. See: Dual-process model of white-matter development
Cortical bridging
When Polish researchers studied brain activity in 7-year-old children, they compared readers and non-readers. They compared children using various writing systems (incl. ideograms, right-to-left, etc.). They noticed that in exposure to printed matter and speech, many areas of the brain light up in a very similar fashion. In learning to read, they observed however one crucial difference. In reading kids, the areas corresponding with phonological processing and semantic memory were "connected" with the visual areas responsible for decoding text.
If visual processing is unaffected, if semantic processing is normal, all the necessary white matter tracks shall be in place to be utilized in the process of learning to read. For that bridging process, we will not observe a critical period window; however, we can expect a significant slowdown in adulthood. Proponents of early reading are right. It is better to master reading before teen years are over. However, we should never impose timelines on the learning process as they all undermine optimality.
Optimum exposure
If the above understanding of critical periods is correct, the optimum way to raise babies might be naked in the woods with all the sensory experience having a chance to provide optimum adaptation. In practical terms, we should rather keep children maximally free in environments that are not much different from target environments with a degree of change that matches similar changes in adulthood and reflect the speed of brain's adaptability. Can the speed of change and the environments be optimized computationally? I doubt it. We should rather live along with our kids and try to make their lives free and fun (see: Optimization of behavioral spaces in development).
In 100 bad habits of schooling, I mention several habits that are associated with linearization of the thought process. I hypothesized that early reading may also have a contribution to that change in thinking modes. Not only do we have reasons to make sure the brain imaginatively navigates rich 3D spaces using parallel processes. Harriet Pattison (and myself too), documented cases where little children explicitly complained about the serialization of thought occurring in reading instruction. Children use a finger, a ruler, rhymes, music and dance to coerce the mind into a rhythmic progression through the text. This is quite different to the way proficient readers read (let alone increaders as in How to read a book in an hour). This is also different from the way artificial intelligence tackles speech recognition.
Again, if the dive into reading is voluntary, I see no reason for worry. I see no reason why linear and parallel modes of thought could not coexist as separate modes of action. After all, we all tend to be creatively parallel in problem solving only to regress (to some degree) to linearity (e.g. as it happens to my mind when I write this text in a mode akin to ChatGPT). After a bout of reading, a kid will sure feel the need to play some basketball or just to daydream.
Due to the brain's adaptability, it is hard to make grave errors in the process of unconstrained development. Freedom alone is a solid guarantee that a healthy child will live up close to its maximum potential.
Stabilization
Due to the interference from prior memories and skills, some forms of learning may get harder in adulthood. In particular, a solid baggage of toxic memories from school may be a major obstacle. However, there is no age limit on learning to play chess, ride a bike, learning a new language or learning to read.
Critical periods may result from a natural process in which the competition between neurons results in specific wiring that is hard to undo. It is also possible that critical periods are supposed to terminate learning altogether. If so, this is called imprinting. For example, a baby goose may imprint the identity of its mother. All new mother candidates may be rejected later on. This can famously be used to train geese to follow human-made flying objects playing the role of a mother goose leading a flying flock. Imprinting was studied by Konrad Lorentz (a genius worth knowing).
In the course of the life, the brain undergoes stabilization. This is the basis of the claim "it is hard to teach an old dog new tricks". However, the well-explored and well-used areas of the cortex may retain their excellent learning capacity till 100 and beyond. We age too fast to extend that claim beyond the age of 100. My genius friend Tom Durrie is 94 (in February 2024). I have not yet finished reading his fantastic book ("End of intelligence") published in Mar 2023, and his new book about the school of the future is ready (Jan 2024). Teenagers do not sprint that fast!
Adult illiteracy
In the 1600s, wise monks would claim that "even 20-30% of the population are smart enough to read". That was considered an optimistic prediction at the time when literacy was closer to 10-15%. Today, most literate nations claim 100% literacy (acc. to UNESCO). Humans always tend to underappreciate the intelligence of others (taken as average). Today, we still hold child's intelligence in contempt. We force kids to read even though they all can do it on their own (see: Child is always right).
Researchers who investigated illiterate adults, e.g. guerrilla fighters who had little access to printed matter, noticed that learning proceeds in proportion to motivation, and takes over the same cortical areas as in those who learned reading as children. This shows that the necessary wiring is already in place. This stands to reason as there are only two primary components involved: knowledge of the world (which all healthy people develop nicely), and the ability to recognize visual differences between individual letters of the alphabet. Naturally, we also need the underlying network of tracts laid between the involved areas of the brain. Deficit in those networks would result in disabilities far more crippling than illiteracy. This implies that all healthy illiterate adults can learn to read if they only choose so.
A notorious case of a teacher who could not read is only an example of how social pressures and toxic memories can be a far greater stumbling block than just developmental neurobiology (see the case of John Corcoran). The allegedly missed critical period, and the alleged dyslexia volatilized as soon as the motivation to undertake actual work was in place. For clarity, learning to read is not much different from learning to speak in terms of vast resources of memories involved. A fluent reader of English knows thousands of patterns involved in the process. These are not just letters, words, or portions of sentences. These are cultural habits associated with particular forms of writing. See: No critical period for reading.
Figure: According to Dr Stanislas Dehaene, adults may be slower to learn, but they can learn to read even in retirement. The gramps on the picture seems to write better than most children
Age provides a disadvantage in second language acquisition, esp. in phonology. This is less a matter of neural disadvantage due to critical periods. This is more a matter of motivations. All students are pragmatic. If bad pronunciation does not prevent communication, they may tend to ignore errors. Errors are less tolerable in grammar, or vocabulary use.
Adults learn vocabulary better and faster. It is possible then need to "reactivate" their brains after a longer period of neglect. There is a definite positive feedback loop between learning and the learn drive. However, there might also be a deeper going feedback between learning, learn drive, sleep, creativity, and the neurotrophic impact of learning on overall brain plasticity.
Kids learn via generalization. This is very effective. The well-schooled adults try to rationalize, understand the rules, memorize the rules, etc. Those shortcuts affect the outcome. We may get faster to the goal, but we do that idiosyncratically. This is particularly visible in phonology. When there is a missing gap to fill out by the brain, the child will improvise, and provide a natural filler. I, for example, fall back on my Polish, on English spelling, the way a word might be spelled in Polish, etc. As a result, my way of filling gaps will be idiosyncratic and provide a specific phonological profile than an expert might recognize as the native speaker of Polish. A child will fill the gaps using its own world knowledge shaped by its environment. It matters if this is an English-speaking world. The kid and adult may improvise to the same degree, but the effects of that process will differ widely leading to possibly a false conclusion: kids learn better.
Adults may be less flexible, less motivated, more self-aware, stingier about their time, more "programmed" by rationality, etc. Adults are usually well-schooled, which is a major disadvantage.
The same idiosyncrasy occurs in writing. Are my texts strange? Are they difficult? They are SuperMemo-biased and second-language biased. They are easy to recognize.
Memory stabilization makes it harder to learn via generalization, but also provides a richer learning toolset. Adults learn differently.
Selective dyslexia
My own case helps me illustrate that dyslexia is a complex phenomenon that goes way beyond the art of reading.
If I was to pass any test based on comprehension of texts typical of Polish literature class, I would be in trouble. Despite being a proficient reader, I might struggle even with texts for 12-year-olds. The reason is a form of aversion developed at school. There is a constellation of psychological factors, motivation, toxic memory, prior knowledge, short-term memory, and more.
Selective dyslexia illustrates the harms of compulsory schooling. I insist that freedom and rebellion were the main factor that helped me survive school relatively unscathed. However, my "dyslexia" is the best example of damage I could identify and hypothesize about. Interestingly, for similar reason, I would rather not do well on an IQ test. In both cases, instead of solving problems, my brain is busy with a single question "What are you doing here? Why are you doing it?". I hypothesize that my brain might be recovering towards a semantic brain typical of creative children. This is the type of brain that violently reject anything that is not serving a pragmatic cause. I can become passionate about the dumbest school material as long as it serves my research. Otherwise, the brain is getting "cranky" and dysfunctional. If refuses to co-operate on anything that does not serve a greater purpose.
Conceptualization optimum
When discussing all learning processes in the perspective of lifetime, we need to remember that the brain is a dynamic set of neural networks (a concept network). We need to understand the properties of those networks to draw conclusions about the global process and its timelines.
Concept networks
Neural network theorists may find it hard to translate their knowledge into the realities of the brain because the brain is a smarter kind of a set of neural networks that we call a concept network.
In a concept network, learning of individual nets proceeds along principles that are unique to a biological network. The underlying properties of individual synaptic connections will change overtime along the two-component model of memory. At the same time, at the higher level, we will observe the formation of concepts, i.e. neural net assemblies representing individual macroscopics concepts that live in an active mind (also at the level of subconsciousness). At the very top of the concept hierarchy are concept neurons that are highly specialized neurons representing concepts along the one-cell-one-concept principle (see: grandmother cell). For simplicity, I use the term conceptualization to refer to the entire process of learning in a concept network to differentate it from various forms of learning in different types of artificial neural networks. As capacity of concept networks is not infinite, we can theorize that all concept network proceed from a plastic state to a stable state along the conceptualization curve as illustrated in the picture.
Figure: Hypothetical course of learning and conceptualization in a fixed-size concept network. The naïve network begins the learning process at high plasticity (in red). As individual concepts form, they are consolidated and stabilized. The overall stability of the network keeps increasing (dark blue). The speed of conceptualization (in orange) is a resultant of plasticity and stability. It reaches its theoretical maximum somewhere on the way from the random graph stage to a sparse representation stage. This is the time of a large supply of concepts that may be subject to generalization, and a good balance between stabilization and forgetting. The overall problem solving capacity of the network (light blue) is negligible at first, and tends to saturate with network stabilization. Large number of well-stabilized concepts makes it harder to find new plastic network nodes for further conceptualization. The maximum capacity of the network depends on its size. Speed of learning in spaced repetition at older ages seems to indicate that the size of the concept network of the human brain is high enough to provide for lifelong learning without noticeable saturation. See: Conceptualization theory of childhood amnesia and How much knowledge can human brain hold
Lifetime learning
When trying to research conceptualization in the course of the entire life, I met with a constraint of incomplete data. We all know that older people learn slower, however, I hypothesize that it is largely a matter of needs and lifestyle. Nonagenarians such as Tom Durrie show no sign of slowdown. Unfortunately, Tom does not use SuperMemo to provide data for my research. Even if he chose to learn for the sake of science, we could only draw conclusions by comparing his data from a wider age span, e.g. from 50 to 100. My own data of my own learning process becomes increasingly valuable. I am far from being the oldest user, however, I own the oldest spaced repetition learning dataset in existence (well backed up in case of my premature death). Thus far, my hypothesis stands firm. There are no detectable signs of slowdown with age. My brain capacity is far from being used up. I am happy to report that I can still learn new tricks in ways that do not differ much from younger ages (see: How much knowledge can human brain hold).
Nevertheless, the theoretical course of learning in concept networks is extremely valuable when we consider that the brain is actually a set of subnetworks, called neuronal circuits, that each is a concept network with its own capacity and the same statistically predictable course of learning in life.
Critical periods in concept networks
The previously discussed critical periods, and the phenomenon of imprinting can also be explained by means of the conceptualization curve. A critical period is nothing else than a period in which a concept network moves from its plastic to its stable state at a very fast rate. In critical periods, we do not care much about the shape of the conceptualization curve. We are primarily interested in the position of the learning peak. We want to know when the process begins, and when it reaches the critical phase beyond which changes are hard or impossible. For many sensory experiences, the conceptualization explodes at birth, and the end of the critical periods is determined by the network capacity and the plasticity of the network that may have a genetically determined parameters. Highly plastic networks, e.g. the wiring of sensory cortex, may rush through the conceptualization curve at a high rate for the sake of survival. for example, we want a baby to be able to visually discriminate objects as early as possible. This first step in visual learning determines a lifetime of learning based on visual stimuli, incl. the acme of them all: face recognition. Recognition of print is probably not that much different in complexity than facial recognition. We need to be able to read ugly handwriting, blurred inscriptions, partially covered billboards, and then be able to disentangle complex meanings and metaphors of fiction.
Readiness vs. motivation
The relevance of the independence of individual concept sub-networks is that for a given cognitive phenomenon to emerge, all individual underlying subnetworks must be ready in place. In psychology, it may be described as readiness (e.g. readiness to undertake learning to read). However, in the case of reading, we need a second component of readiness. Not only do we need necessary networks such as the ability to recognize different shapes in print. We also need an output from knowledge valuation network that says: "I need to know how to read". That part we can call motivation. Unlike sensory readiness, or network readiness (i.e. connectivity between individual areas of the cortex), the motivation is an unpredictable component. As in the case of Freer Speckley, interest in carpentry may compete with the interest in reading. That's the beautiful part of brain's adaptability that skills emerge in proportion to needs. It is the school system that harmfully interferes in this process.
Saltatory readiness
If we want to simulate the readiness curve for reading, we need to plot all individual conceptualization curves for individual subcomponents in the process, and include the last and most complex conceptualization curve of world knowledge. Superposition of those curves is not additive. It is not a logical disjunctive relationship (operation OR), but a conjunctive relationship (operation AND). In other word, the superposition may be flatlined until all components are in place, and then raise along the world knowledge curve and reach the stabilization point at the moment when most of unknown words may be read in approximation with the help of the rules of phonics. As we can see it in the case of Freer Speckley, even solid knowledge of the world may not suffice to complete the process if the output of the valuation of reading is insufficient. It all depends on the individual.
School hate
A huge and independent concept network that is a inseparable mix of readiness and motivation is the cortical store of world knowledge. World knowledge is necessary to associate words with meaning (readiness), and to associate meanings with value (motivation). The need to learn to read and the efficiency of the process is closely intertwined with the understanding of the surrounding world. In the meantime, we do a double whammy intervention in that complex process. While a child wants to explore the world (e.g. via YouTube), we ban the electronic media to provide room for the process of learning to read. We rip the carpet from under the child's feet. At the same time, the secondary hit is in the authority of the adult world. A child who instinctively understands the power of knowledge is being conditioned to never trust the adult guardian who is imposing abstract tasks that have no grounding in the knowledge valuation network. This is the key component of school hate. The effort to make children read is the first powerful introduction to school hate that is unlikely to be unlearned at later stages of education by illustrating the wonderful marvels of the world in the context of compulsory schooling. If learning is the greatest pleasure, compulsory schooling is the denial of what is most pleasurable (by violating the Fundamental Law of Learning). Coercive learning is a form of cruelty.
By forcing children to learn to read, in addition to school hate, we hurt reading (see: educational dyslexia), we hurt adult authority, and, most of all, we hurt the process of acquiring world knowledge.
Simulating superposition
Many skills conceptualize as a resultant of multiple conceptualization sub-processes relevant to individual circuits or sub-skills. Superposition of conceptualization curves may therefore seem likely to assume complex shapes that reflect the complexity of the underlying process. In real life measurements, we do not seem to observe that. This in part comes from problems with measuring implicit sub-processes. However, we often see exponential explosions in performance, which may look like sudden emergence of a skill. In case of reading, a child may resist learning, hint at a disability, seem to be confused only to suddenly appear to be able to read. To better understand the superposition, I decided to separate most salient sub-conceptualizations of the art of reading, and plot their hypothetical course on the timeline of age. Despite combining very disparate processes, my performance curve for reading turned out to support sudden emergence with rapid progression. If the inflection point is missed by the observer, we may have an impression of zero-to-one emergence. The following figure is based entirely on wild suppositions and should be treated only as the illustration of emergence:
Figure: Exemplary theoretical conceptualization curve for the ability to read derived from a super-position of disparate conceptualization sub-processes. Despite being a resultant of a number highly varied processes, in this example, the performance curve seems to explode exponentially at around the age of 9. Conceptualization of visual discrimination might be the only significant component of the process that is subject to a critical period (light blue). The motivation component will be a derivative of the knowledge of the world. This component is the least predictable as all children show different interests. The acquisition of reading pattern memories will proceed in parallel with the process of learning to read, and will constitute a positive feedback in the process, where each pattern adds to the performance, and motivates further reading that adds to fluency, and further pattern acquisition. The same refers to meta-knowledge of reading, which will largely reflect the child's environment. Knowledge (in orange) must precede reading (in red), otherwise the tolerance for reading without comprehension may increase. The presented graph has been simulated for illustration only. It may differ vastly from the actual course of learning in a particular child
The figure above shows an example of the conceptualization process for the ability to read. Despite being a resultant of a number highly varied processes, the performance curve seems to explode exponentially at around the age of 9 (in this example). We can mark that age as the age of readiness, and it will differ vastly between individuals. The "age of emergence" will depend on the exposure to print and the progression of skills but may by apparent as soon as 1-5 months after the age of readiness.
Conceptualization of visual discrimination might be the only significant component of the process that is subject to a critical period (light blue). This might indicate that the only critical aspect of reading is to make sure we do not keep our babies blindfolded early in life.
The motivation component will be a derivative of the knowledge of the world. The motivation component is the least predictable. All children show different interest. A child interested in stories may wish to hear book reading at an earlier age. A child interested in carpentry (see Feerer Speckley) may fall out of the presented chart altogether. In the presented example, the progression of reading was most tightly linked to the world knowledge. For that reason, other factors did not seem to significantly affect the shape of the performance curve. We need knowledge for motivation, and for a substrate of memory associations between the visual (print) and the semantic (knowledge).
The acquisition of reading pattern memories will proceed in parallel with the process of learning to read, and will constitute positive feedback in the process, where each pattern adds to the performance, and motivates further reading that adds to fluency, and further pattern acquisition. The same refers to meta-knowledge of reading, which will largely reflect the child's environment. This knowledge will start with simple memories such as "We use print to pass messages", or "We can know the thoughts of dead people", or "There are many languages with their own codes for sending written messages".
Knowledge (in orange) must precede reading (in red), otherwise the tolerance for reading without comprehension may increase. This is exactly what happens at school! For a 3-year-old to read the Constitution, as neatly demonstrated by Sanger, we need to artificially accelerate the learning of (1) the recognition of reading patterns (e.g. by memorizing words, syllables, and/or characters), and (2) the vocalization of those patters. Vocalization is not included in the graph as it does not form the necessary part of the art of reading. Arguably, this form of reading is missing the most important component: translating the text to the semantics, which is hindered by the need to understand the government, history, tough words (e.g. abridged), etc. In other words, it is possible to accelerate the conversion of text to speech, but if that acceleration precedes the build up of the knowledge of the world, the effort is suboptimal (e.g. involving unnecessary costs). It is very hard to predict how such an acceleration might affect the wiring of long-distance cortical connections. In theory, we might form a preference for learning from texts, which again may turn out as a useful skill, but might not be optimum for the population in general. Equally well, due to high plasticity and low myelinization, the gains and losses of that early reading may gradually dissipate over time (as indicated by the research by Sebastian Suggate).
Early instruction
Early instruction is likely to be harmful. Reading needs to follow comprehension. Larry Sanger succeeded in making his children read at very young ages. However, when a child reads the US Constitution at 3, the reading is not likely to be semantic. The jury is out on if those early successes will translate into value in later life. Sanger invented Wikipedia. Children of people like Sanger are likely to turn out fantastic even if the educational strategies are suboptimal. There is a great value in documenting and researching cases like this. However, there is also a substantial danger. Parents may be inspired to attempt to copy-cat the solution and fail in a single variable that will result in more harm than benefit. It is a bit like running marathon, it is easy for a pro to think that everyone can run a marathon. Only when dozens fail with injuries, we realize that there are too many variables to keep in mind when suggesting strategies for beginners.
Headstart myth
Proponents of early reading and phonics make the same old mistake about the math of the developmental trajectory. They say "if you start early, you build up skills exponentially and it only gets better. If you experience a false start or late start, frustration will only build up from year to year and you spiral in the opposite direction".
The error is two-fold. Firstly, exponential growth plateaus in proportion to the volume of reading. Late starters will plateau at the level comparable to early birds, or may even rip a comprehension benefit if they started naturally.
Secondly, the downward spiral of discouragement occurs only in conditions of coercive school where those who are left behind at first stay further behind later. Unschoolers do not get discouraged because they do not compare their performance to others.
In the world of coercion, a subset of students will fail entirely. In the world of freedom, all students get to their best potential level.
Early start makes sense only in cases where it is born out of intrinsic motivation.
Dr Sebastian Suggate devoted a great deal of his research to studying early reading (see: Children learning to read later catch up to children reading earlier).
Suggate remarked:
Teaching children to read from age five is not likely to make that child any more successful at reading than a child who learns reading later, from age seven
Suggate data indicate that there is no detectable advantage in learning to read at 5 in the light of the reading proficiency at 15. Suggate noticed that "language development is a better predictor of later reading than early reading".
John Taylor Gatto did not need research. He could see it with his own eyes:
David learns to read at age four; Rachel, at age nine: In normal development, when both are 13, you can’t tell which one learned first—the five-year spread means nothing at all. But in school I label Rachel “learning disabled” and slow David down a bit, too. For a paycheck, I adjust David to depend on me to tell him when to go and stop. He won’t outgrow that dependency. I identify Rachel as discount merchandise, “special education” fodder. She’ll be locked in her place forever
If you still think starting late is a serious handicap, consider that the only PhD president, Woodrow Wilson started reading at 10. It was not a handicap for him.
Are earlier readers faster readers? The advantage will be melting over years. In the end, the speed of reading is largely irrelevant. See "speed reading on steroids".
Effect on motivation
We have known for ages that early reading instruction can lead to problems. The idea of school dyslexia emerged in the 1970s, but we have always known that early training is a formula for less motivation for reading. When a fashion for early instruction started spreading in the 2000s, Neil Swidey of the Boston Globe decided to investigate. In "Rush, Little Baby" he wrote:
A classic study in the 1930s by noted researcher and Illinois educator Carleton Washburne compared the trajectories of children who had begun reading at several ages, up to 7. Washburne concluded that, in general, a child could best learn to read beginning around the age of 6. By middle school, he found no appreciable difference in reading levels between the kids who had started young versus the kids who had started later, except the earlier readers appeared to be less motivated and less excited about reading. More recent research also raises doubt about the push for early readers. A cross-cultural study of European children published in 2003 in the British Journal of Psychology found those taught to read at age 5 had more reading problems than those who were taught at age 7. The findings supported a 1997 report critical of Britain's early-reading model
Effect on reading difficulties
David Elkind in "Much too early" reported:
Evidence attesting to the importance of developmentally appropriate education in the early years comes from cross-cultural studies. Jerome Bruner reports that in French-speaking parts of Switzerland, where reading instruction is begun at the preschool level, a large percentage of children have reading problems. In German-speaking parts of Switzerland, where reading is not taught until age six or seven, there are few reading problems. In Denmark, where reading is taught late, there is almost no illiteracy. Likewise in Russia, where the literacy rate is quite high, reading is not taught until the age of six or seven
Reading prodigies
The legend says that the Supreme Eternal Leader of North Korea: Kim Jong-Il was a miracle baby. He supposedly learned to walk at 3 weeks old and was able to talk at 8 weeks. He was reading as a baby. We cannot ask Kim about details because the Eternal Leader is dead. However, his magnificent son Kim Jong-Un could read at an even younger age. I would e-mail Kim if I could hope for an answer, however, few people would use him as an example of: "early readers, are good leaders".
We have dozens of documented cases of reading in children as young as 2-3 years old (see Larry Sanger in this text for a detailed example). Harriet Pattison had a case of 18-months-old able to recognize patters in a way that resembled reading. We should celebrate early reading if it comes naturally. Sadly, many a parent see similar examples as a way to follow, which is a formula for an educational disaster. If you have a kid with special abilities, check Larry Sanger's manual (in references), and give up as soon as your results diverge from the vision.
Reading before speaking
If a kid can read the word "go" at the age of 2, the parent will proudly announce the readings skills to the world. If the same kid fails to build sentences in speech, the same parent can bemoan "my kid does not speak". And then an even more enticing observation arises: "my kid could read before he could speak". A myth is easily born. We tend to remember the extreme and then mold it for years. No parental anecdote is reliable. Naturally, reading before speaking is possible. If a deaf kid learned to read, had his hearing fixed, and then learned to speak, we would have an easy proof. This also shows that reading converts texts to meaning (semantic mapping). It does not need to generate words, sounds or speech as it is usually demanded at school (orthographic mapping).
The old truth still holds true: kids differ. Moreover, different environmental stimulations will cause different passion patterns. The same kid taking a course in aikido might become a mathematician, while taking dancing classes might turn her into a world class sprinter. Brain development pathways are highly unpredictable.
Larry Sanger
If you are skeptical about prodigious early reading feats, check out a true case of Larry Sanger. Larry is the inventor of Wikipedia, and one of the greatest minds I ever met. Larry is a guy with horizons spanning beyond the Oort cloud. He is passionate about homeschooling. I disagree with Larry on many points, but that's exactly how it should be. Nothing sparks inspiration and passions better than different points of view. He soaks knowledge about the universe and shares his experiences in his blog. He openly reports on homeschooling of his two boys, and I found lots of inspiration for this article in his texts.
Reading at two
Larry Sanger introduced his kids to reading very early. His kid would read the US Constitution at the age of 3 (now developing EncycloSearch). For comparison see: Caleb Anderson reading at 2 (now in college). Or Caleb Green who read 100 book in a day at the age of four.
Larry wrote a fantastic handbook on early reading using his experience.
I was impressed but very skeptical. The old argument is that there is no point in learning to read texts that do not form fully connected semantic constructs in the child's brain while reading. Many adults do not know the word "abridging" or "the right to assemble", "redress grievances", etc. This is an argument against teaching to read early. Let the semantics guide the process. 3-year-old would probably never voluntarily pick up the Constitution. When Larry says he used no coercion, he sure can be trusted. He is a type of an erudite who can radiate his conviction and it would not be difficult to convince a child to follow his lead or set up rules for systematic progress.
Larry's handbook of early reading contains a few things that I strongly disagree with. He encourages early reading. I prefer to see reading emerge in a child who can understand the world and read to further that understanding. Larry recommends his Reading Bear and insists that it should not be a problem for every parent or educator to dedicate 15 min. per day to that process. Not every parent will have the charisma of Larry. Not every child will show the readiness or interest in reading. The recommendation "it should be easy" is dangerous because many will try, and their failure will inflict harm on children. The entire toxic school system is based on the assumption "it should be easy" (see: educational empathy). Larry's case should come with a caveat: "Don't try this at home, unless your kid asks you to achieve similar feats"!
Love of reading
Larry has a solid background in classical knowledge, literature, history, and philosophy. To him, reading is above oxygen in Maslov's hierarchy of needs. I am totally not surprised that he introduced his kids to reading early and that he promotes early reading. However, with all his success, he may underappreciate dangers related to early reading. His instruction manual says clearly that kids can start their immersion in reading as soon as they get ready and get to like it. That liking part may easily be missed by ambitious parents. Larry set a high bar for others and there may be many overzealous copycats with lesser pedagogical talent, or kids with lesser predisposition for early reading. Kids forced to learn is a formula for slowing down development.
I would like to explain that, in addition to inventing, Wikipedia, Larry's greatest legacy in the area of early reading will not be the high reading achievement of his kids, or a population of early readers, but a meticulous analysis of what is possible. Human brain will never cease to amaze. Larry has his own contribution to the field.
As for early readers, I would like to explain why I believe that brain science suggests their added advantage may not be as high as it seems at first.
Benefits of early reading
For a parent, a kid that can read gains an early independence from adults, from indoctrination, from manipulation, from rigid philosophies of the adult world. Early reading is an ticket to early thinking about the universe. This should result in freer and more mature young mind.
There are two main problems with this picture. Early in development, kids learn things that are far removed from the abstract adult way of thinking. They need to master basic things like moving objects in 3D space. Lego bricks might serve that purpose better than a book. Secondly, slow and unconstrainted brain development is an advantage in the long run. All form of acceleration may backfire. Early reading based on a well-instilled passion is great for development. Early reading with any degree of coercion may go well beyond the optimum push zone.
The key advantage to early reading is the entry into the world of text, which is the best means of communicating knowledge in many areas of learning. In incremental learning, there are various sources of knowledge: text, video, images, sound, formulas, etc. For soaking in general knowledge, text might cover most of the areas as the best medium, but the optimum proportion drops fast with a drop in age. This will vary in some domains and between students with different cognitive preferences. Easing kids into text processing early seems important for establishing that strong learning anchor with no fears and no biases.
It is pretty unlikely that early readers gain an advantage other than knowledge. Later readers seem to catch up fast. Early readers do not seem to gain a lasting advantage in visual processing, decoding symbols, text pattern recognition, semantic processing, eye movement, etc. In other words, I am not aware of any neural adaptations that would make early readers better readers in the long term. All kids will plateau at their own specific level and use reading as just a tool for the rest of their life. It is the conceptual computation that keeps getting better from year to year. Reading is just a source of extra data for processing.
The main value from early reading comes from knowledge and establishing good reading habits. In that light, letting the kid decide when to enter the realm of books seems like a safe default strategy. Moreover, some developmental milestones can be reached by kids at very different timelines. Kids who spend their days on the football field will focus on motor skills first. Those who love videogames will do great with a subset of motor skills and visual processing. There is no inherent danger in being monothematic early in life other than honing down one's preferences, which may not always be to parent's liking.
For a toddler or a preschooler, the world of the unknown is so vast that all forms of learning are welcome. The parent only needs to make sure that the kid does not waste her time, e.g. commuting to daycare. Otherwise, all forms of learning are useful. The parent needs to provide the environment. The rest will be taken care of by the learn drive.
Modern media
A child who gets his general knowledge from YouTube, as opposed to books, will soak in equivalent model of the world. Her knowledge tree will be different for personal reasons. There might not be much difference in learning about Ancient Greece from a book or from YouTube. There might be a world of difference in learning another topic. For example, see birds of paradise dance from "Our Planet" on Netflix. This seems like an acme of educational art. The dance cannot effectively be described in words.
In the end, the child will choose her best path using her learn drive. Individual differences will be leveled by generalization. Most of the effort will fall victim to forgetting. General world knowledge skeleton is all the kid needs in those early years.
Parents should study Larry's book, see if they can emulate his achievements, and never be worried if the job appears harder than it was the case for Larry and his kids. One of my teacher friends teaches reading to first graders in Poland. They are aged 7. Those who stayed behind, she was obliged to drill hard after classes. She considers this the greatest mistake of her young career. Today, when a kid lags behind, she just does not tell the world. Let them have as much space and time as possible.
Conclusion
If kids start to read early, their advantages will wane over time (see: Headstart myth). By adulthood, both early and late readers are both likely to fulfill their best potential, esp. if the decision to read was taken voluntarily.
See also: Peter Gray under attack from Larry Sanger
Reading without comprehension
Early reading may condition reading without comprehension. A kid whose brain isn't primed for reading, isn't too curious about what's hiding behind words and letter. If this is the case, most of mental energy and processing is focused on decoding text. The mere success of uncovering the words hidden behind combinations of letters becomes the goal of reading. The story becomes secondary. This is how fast reading with low comprehension may develop. The brain develops tolerance for accepting meaningless messages. This kind of tolerance is welcome and necessary early in the process. However, tolerance for meaninglessness must be balanced with clear gains in semantics that arrive at the same time. This is necessary for minimizing frustration and retaining the enthusiasm for learning. This is what makes kids good language learners.
The key difference that makes for good reading is engagement. If learning is self-directed, it is always goal oriented. There is a need to decode a message behind a spoken word or behind a written word. It is hard to produce wrong conditioning in speech. We do not ask kids to repeat long sentences. We rarely do speech instruction. Reading is different. Early reading is particularly different as it rarely comes from within. It is usually a parent who prompts the kid to decode written texts. If early reading comes in group instruction, dangers of wrong conditioning get magnified.
Dr Sebastian Suggate research shows that early reading correlates with lower comprehension when comprehension tests are taken at later ages.
Dangers of early instruction
Psychologists agree that starting too early may put a kid off for life. All acceleration programs are inherently dangerous. Tutoring always injects a bit of the learn drive override. If it goes beyond the push zone, development will be slowed.
An open-minded loving parent, e.g. Larry Sanger, is far less likely to bring that scary outcome. Nobody in the world can read the kid better than a dedicated parent. Then it takes a special kind of kid who is ready to read early. The key is motivation, and a special match parent-child may provide the necessary ground. There is also an issue of alternative options. When a parent sits down with a happy kid to teach reading, he cannot have a tablet hanging around, or there is a risk the device would turn out more attractive than the proposition of the parent. A dedicated parent knows best how to motivate the kid, and how to make sure motivations do not get overridden by a stronger attraction characterized by higher learntropy.
In an institutional setting, early instruction leads to anxiety, which provides a risk of educational dyslexia (source of the quote):
“We see many of our kindergartners struggle with anxiety about school because they know they are expected to read,” the teachers wrote in their letter to school officials. “A significant body of research exists showing the negative consequences to children’s emotional well-being when they are forced to read before their developing brains make sense of it”
Educational empathy
One of the main problems of education systems around the world is poor understanding of the child's needs. Adults build education to their idea how it should look like. They lack educational empathy. That deficit is natural. The brain changes over time. The older child may have little empathy for a younger child. An older child may also be surprised with her own deficits in learning (e.g. as documented on family videos).
Can you read this sentence: "Sdrawkcabdaeruoynac"? If you are Finnish or Italian, you might be able to decode it using school methods, but it still means nothing. This is how a typical child feels when reading in Latin alphabets early in its journey.
I am tempted to say that education systems should be designed by brain experts, and brain experts will tell you that only self-directed education makes sense. Free learning based on the learn drive is the optimum.
As for the sentence above is "Sdrawkcab daer uoy nac?" any easier?
What if I told you that you need to read it backwards? You know phonics, you know the rules, and yet a mere reversal of the order makes it so hard for your brain. Feel for the kid!
Computers instead of pre-school
The learning environment in preschool is made pleasant and cozy. Walls are painted with fun pictures, shelves are full of toys and a nice carpet lets kid huddle around the teacher in a small group, who will often make it all more fun with a teddy bear companion, and an occasional talking puppet.
Lessons follow a plan designed by experts. Step by step, kids are getting introduced to the alphabet and numbers. Teachers and caregivers are often specially trained. Kids sound out phonics and recite numbers. Teachers go out of their way to keep the group enthusiastic.
The major problem with his picture is that it is all based on an inefficient artificial adult-centric design. The true reward from learning is part of the learn drive. As long as kids want to learn, they will be richly rewarded. In early education, kids overwhelmingly do not want to learn because what they are being served does not satisfy their inner learn drive. Instead, all the environment, sweet-talking teachers, and talking puppets are supposed to sweeten the deal. They are supposed to pleasantly cover up for an inherently unpleasant thing: learning things kids do not want to learn. A kid with good defenses will protest. He will fidget, mutter or fall asleep. He will be considered disruptive or branded ADHD. Others, with assistance of learned helplessness, will accept their fate and plod along. After a while sweet voice of a teacher will cause cognitive dissonance in kids: it is hard to hide that even the most patient educator sooner or later gets enough of the exercise. Facial expression cannot lie. Kids will keep answering, with a bit of masked intimidation. The show must go on.
Figure: Preschool environment looks neat and satisfying. However, it largely meets the adult dream of the perfect childhood. When kids are let free to pursue their creative needs, there is not need for neatness. Play is the best way to learn, incl. learning to read and count
Contrast this with a preschooler's home. The kid is banging computer keyboard playing some mindless computer games. Driving the same car over and over again, through the same obstacle course, only to get to the next level. The kid ignores mom calling from the kitchen "stop playing those stupid games!". He is not bothered by his hunger, by a disruptive sibling, by awful mess on his desk or even by a sticky spot caused by a soup spilled on the keyboard. All those unpleasant components in his environment are compensated handsomely by the fact the kid is pleasing his learn drive. What to an adult is a mindless game, to a kid may be a next, and totally new variant of a challenge. While doing seemingly repetitive stuff, the kid learns numbers. How else would he recognize his current level? He also masters letters of the alphabet and their place on the keyboard. He even picks up a lot of reading skills: Start, Play, Level, Game over, Install, etc. Even 3-year-olds can figure those out on their own.
In this new world saturated with computational tools, experts should refocus from preschool curricula to immersive playful computer environments when kids can master all essentials by never even discovering the surreptitious teaching programs masquerading as fun computer games. A few minutes per day with a good application could replace the whole day of drilling in preschool. Naturally, the social component of preschool needs to be played out separately, e.g. in a neighboring playground.
Reading Bear
Sanger is a designer of a adorable early reading initiative called Reading Bear. However, even the best piece of educational material must be combined with voluntary play on the part of the child. No education works without the learn drive. I believe Sanger is wrong by saying:
In 2012, professional advisers to reading teachers state that, by the end of first grade, a good student “has a reading vocabulary of 300 to 500 words, sight words and easily sounded out words”. This, I think, is a travesty. At the end of first grade, students should be able to read whatever words they can comprehend, meaning (of course) many thousands of words. I’m using Reading Bear and other resources to teach my second son to read. He is 25 months old. We started seriously when he was around 18-20 months. He’s up to presentation #11, capable of getting 13-15 out of 15 on the quizzes on that material [...]. If my two-year-olds can do this, your six-year-olds can do it. Teachers, know that Reading Bear is complete and 100% free, you have no excuse. I understand that your districts impose a lot of top-down control of curriculum. But you still have some freedom. Reading Bear’s presentations are 15 minutes long, and you can definitely take that much time out of your busy day for this program [...]. If all you do is show Reading Bear’s presentations for 15 minutes each day for the 180 days of first grade, you’ll be giving your students the gift of literacy
This is a good hearted and concerned advice. However, it carries the risk of projection. Kids differ wildly. Even if they have parents with smarts and determination of Larry's, they may simply show little interest in Reading Bear. If they have already been infected with a bug of computer games or free roaming on YouTube, or just prefer Lego, or playing football outdoors, it would be unwise to make a hard push for reading for reasons explained in this article. Larry used feed time to do most of his reading, but the child needs to be consulted. Perhaps instead of reading, she prefers to watch Curious George?
The situation in the classroom is even worse. When kid is enslaved, and shows no interest, his memory will retain nothing of the class. All he will learn is resentment for Reading Bear, reading, the teacher, the school, and education in general. See: 100 bad habits learned at school
15 min. of instruction time is not little. Some kids love spending time with parents and cherish every moment. Others have only a small window of tolerance for instruction. If that window happened to be the said 15 minutes, the kid would learn nothing else but reading at the cost of all the learning in the world. If there is only 15 minutes to economize, it would better be spent on things that the kid is passionate about.
The problem gets worse for parents with lesser skills or availability of time. The real dangers begin when the push is made in institutions such as schools. This is where kids are conditioned to hate instruction and they are bound to come with negative preconceived notions about instruction software, even if it is as well thought through as Reading Bear.
Last but not least, reading instruction is not equivalent to the gift of literacy. Literacy must come from within. If it is not powered by the learn drive it is not likely to last. If so, reading itself should be kid's own initiative. The best a parent can do it to explain or stoke up the enthusiasm. All kids who love Reading Bear should use it. All the rest needs to be given space to make their own choices.
Software developers should pay utmost attention to stay away from meticulous step-by-step instruction reminiscent of schooling and move in the direction of self-directed play where kids can learn and explore, preferably on their own, without supervision and interference. If software attempts to smuggle school, it is already treading on a dangerous ground. Educational Minecraft is a good example. Even standard Minecraft can be a failure if it just set in the context of a classroom.
The appeal to the teachers should not tell them what to teach but help them suggest resources where kids can explore the world on their own, voluntarily, incl. the world of written word.
Parents as spoilers
Computer technology accelerates learning and development in children. Parents eager to contribute to tutoring can interfere with this process. When they step in with reading instruction, they can do tangible harm. Not only can they discourage reading. They can also discourage self-directed learning. They can even discourage the use of technology in learning as presented in the anecdote.
Both Olaf and Filip are smart and both mastered counting and alphabet early, in play, in drills with parents, and in computer games. They both know computer keyboard well as they both prefer computer games and YouTube over books.
They both opposed all forms of instruction in early reading. Between ages 2 and 5, apart from the learning the alphabet, they did not make much progress until an interesting discovery by Filip. Nobody knows exactly how, but Filip discovered a text2spech application on the net. He was totally blown by the fact that he could paste texts to the app and have an avatar read them aloud. This is how Filip "mastered" reading in one evening. He would run around the house, picking books, ointments, boxes, food packaging and his mom's cosmetics. He would meticulously type labels to find out what all those inscriptions mean. In 2-3 days, he became pretty fluent in typing and "reading". It was just a few days later that Olaf heard of Filip's discovery and followed the same path. The fascination was the same, behavior was the same, and the increase in typing fluency was equally stunning. One of the first things Olaf typed on his own at Google was "PBS"! He was searching for "PBS kids website". A typical reading program would make him practice Dog, Cat, Mat, while what mattered to him was PBS. Self-directed learning always works better as the learn drive provides best hints on what to learn to make learning last.
One day, Filip's mom hit on an idea. She pulled out a children book and asked Filip to go through the book, page by page, and "read" using speech synthesis. The idea was to read the book and gradually accelerate by memorizing how simple words look (e.g. "the", "bear", "tree", etc.). The problem was that Filip wasn't interested. To him "Clotrimazole" seemed more interesting than "Papa Bear". After a short tug of war, Filip rebelled. He refused to use the application and has not tried it since (over a month now as I type these words).
In the meantime, Olaf was spared that early intervention. His enthusiasm for "reading" with speech synthesis waned gradually. However, he got the app ready in case some important things showed up in his on-line experience. Very often, instead of typing, he just pastes some YouTube titles to "read".
Inspired by Larry Sanger, I asked Olaf for a little experiment. I asked him to please go with me through a few lessons of Reading Bear. It never seems difficult to ask kids for a change to their routine. They will gladly take anything new, esp. from a stranger who is untainted with the "teacher stamp" (i.e. the aura of someone who wants to enforce instruction). This type of a request resides comfortably within the push zone. However, after just 6 minutes, the kid explained unpassionately, eloquently and honestly: "It is boring. I do not learn anything. I would rather go back to my computer games". The fault is not with Reading Bear. Software and videos, to my adult eye, are excellent. Well designed and with utmost attention and concern for child's need. The problem resides in kid's brain. For Olaf the abstract theme of letters and sounds blending into words is interesting for a while until it become repetitive. It is interesting for a few combinations, and then it becomes routine. They taste of drilling.
Perhaps it is easier to convince a two-year-old to play reading games? Perhaps the virus of videogames is to blame? Perhaps a bit of a push from a parent would be helpful? In Olaf's case it does not matter anymore. His learn drive clearly gives added value to the alternative. Parental assistance might make sense; however, this short experiment told me that Larry Sanger's point of view is an optimistic projection from his own experience. This projection can backfire if other parents try to hit the same benchmark. I insist they can try but they should not try too hard.
I still wanted to know more about Olaf's motivations. I asked if he wants to learn to read. He explained that he would love to, and that reading is very important however he added "It is too much. I can't remember anything. I have a small brain. My memory gets destroyed". Blaming simple cognitive overload, I asked "perhaps I will come again tomorrow, and we try another 6 minutes". He was clearly not enthusiastic. His mom later reported that he agreed to do some more lessons, but instead of reading he kept guessing, he was fidgeting impatiently and talking all the time instead of listening. That would instantly make one suspect ADHD, and his non-top talking might not be a bad thing. Perhaps just a natural defense against learning he did not want to undertake. When his mom moved the lessons to late evening to reduce fidgeting, the kid would just instantly get sleepy and ready for an early night.
Olaf and Filip progress slowly. In both families, there is no talk of early reading. The silent assumption is that Filip will return to speech synthesis at some point and both kids will learn reading on their own. They both resist drilling and tutoring. They hopefully got their early ticket out of "home school" with Filip's invention. Until the actual school time arrives, they are free to learn on their ownEducational dyslexia
Reading instruction is a waste of time (unless asked for by the child). Proponents justify early reading by a claim that it will result in a bit of extra reading in early years. That extra reading is of little benefit due to the high level of recycling of memories in children (see: Childhood amnesia). Two kids who learn to read at the ages of 4 and 9, will read at comparable level at the age of 15 (see: Headstart myth). The world knowledge of those kids will also be comparable. The kid who learns to read may actually stay behind due to having less time for actually learning about the world. There is a major danger in early reading though. Premature instruction in conditions of stress will result in educational dyslexia.
Toxic memory
I write elsewhere about toxic memory. Early reading instruction is an excellent example of how learning drills, that are not a product of the learn drive, can produce toxic memories that can be hard to undo.
Kids are conditioned to decode words, put them to sentences, and move on to the next paragraph, often with vocalization or sham comprehension. They develop high tolerance for rushing through texts, running the words and grammar through their working memory without ever decoding the semantics.
Toxic memories are easily eliminated by forgetting. In SuperMemo, they can be more persistent due to the inherent power of the program to combat forgetting. When kids learn to read on their own, e.g. while playing computer games, they almost never form toxic memories. Anxious memories never stick. They get obliterated before they move to the parasitic, let alone toxic stage. In contrast, kids in early reading programs can quickly develop the whole host of memories that will show up as educational dyslexia, loss of focus, reading without comprehension, and, in the worst case, in combination with fear conditioning, dislike or hate of reading.
Mechanism of dyslexia
Educational dyslexia is reading difficulty induced by rigorous reading instructions combined with the stress of schooling. The mechanism of educational dyslexia is analogous to other form of toxic memory. When a child associates reading with stress, reading will cause displeasure regardless of the context. Sensitive children with a slow developmental trajectory are at particular risk. Educational dyslexia is most likely to arise in early reading instruction when the brain is developmentally not ready for reading.
The mechanism:
- optimum learning is guided by the learn drive (see: Optimality of the learn drive)
- in optimum learning, all pieces of knowledge are combined like a jigsaw puzzle (see: Jigsaw puzzle metaphor)
- the critical part of that puzzle is valuation and semantic distance (i.e. the length of the semantic pathways the brain needs to traverse to connect pieces of knowledge)
- an easy and reliable measure of the optimality of learning is the pleasure of learning
- to master reading, this gradual buildup of knowledge takes months or years as there are thousands of pieces of knowledge to connect
- when reading happens seemingly overnight, it is a result of long months of implicit buildup of knowledge that occurs each time a child is in contact with texts
- at school, the semantic distance and knowledge valuation are ignored. Students are supposed to do what they are told, and they are told to master skills as prescribed in the curriculum or a specific textbook
- if the brain cannot make a connection between pieces of knowledge (i.e. it cannot traverse the semantic distance) there is stress involved (see: decoding failure penalty)
- if the connection fails, the input (i.e. a question, a word, a syllable, etc.), instead of being connected with the right output (e.g. sound, meaning, etc.), will be associated with stress
- if the reading procedure brins up a great deal of unpleasant associations (called toxic memory), it will lead to anxiety, and a dislike of reading (or even fear of reading instruction)
- the state of anxiety makes learning hardly possible. Outwardly, the inability to make progress in simplest reading tasks will manifest as dyslexia
- if reading activates a great deal of toxic memories, defense mechanisms may be triggered (e.g. avoidance of reading, avoidance of school, avoidance of school friends, etc.). Further learning becomes impossible
- the best remedy against dyslexia is an immediate cessation of coercive instruction. A child may need months of "detox" to recover interest in reading (see case studies elsewhere in this text)
Precocity paradox
A common myth says that kids who achieve things early in life will achieve more and get further. Sadly, the opposite is closer to the truth. Unless delays are a result of an organic problem, slow developmental trajectory serves the brain well. We admire kids who walk at 6 months old but forget that kittens can do it at 3 weeks, which does not apply kittens are smarter. The smartest kids are usually the slowest because the reason for developing their nimble brain is prolonged cortical maturation. In smartest kids, cortical thickness peaks at 12, which is around four years later than average. Later talkers are often people of special talent (Einstein was one). Later talking is usually followed by late reading. Delays may also come from specific passions for numbers, for visuals, for music, for motion, etc. Early passions serve human genius, but do not always serve progress at school (e.g. in reading).
In early school days, boys lag behind girls, which deepens the separation gap between sexes. Girls are considered smart and dutiful. This results in further training in obedience by means of school grades (see: Dangers of being a straight A student). Boys are considered slow and unruly, which often leads to ADHD or dyslexia. Developmental delays are frequently a reason for extra effort in speech therapy and reading drills. In those cases, adult efforts to help often backfire. Instead of better speech or better reading, we end up with stuttering or dyslexia.
For more see: Precocity paradox
Figure: Precocity paradox explains why early acceleration may also result in an early stagnation. Slow and rich brain growth will prolong a set of unwelcome side effects of neurogenesis such as childhood amnesia. This may lead to an illusion that early academic training improves long-term developmental prospects. In reality, early acceleration may be a result of the crystallizing effect of stress on the brain structures. Kids who bloom late may bloom better. The best way to assist a healthy brain development is freedom and access to rich environments
As long as passions thrive, the parent can be pretty sure the kid will efficiently use available resources to climb towards her peak. However, there is also a simple metric than can provide reassurance. If the kid falls into a very bottom percentile in performance, e.g. the size of vocabulary, the parent will naturally worry. However, the same kid may show steady progression in percentile ranks. If so, we may bet on the size of special talents in the future. Vocabulary measurements are equally useful in speech delays and in reading delays. Interestingly, a semantic brain allows of one more interesting observation. After 1-2 years of reading, kid's reading vocabulary approximates her spoken vocabulary. A semantic brain will struggle to read a word it does not know. The exercise in phonics allows of reading without comprehension, but in natural learning, phonics develops in parallel with word recognition, and is not particularly useful in reading unfamiliar words. This fact is more pronounced in English than in phonetic languages.
Figure: Speech delays often spur parents or educators to seek speech therapy. However, it is not unusual for a seeming disability to self-correct. A laggard may grow well ahead of its peers as in the presented example. Similar metrics may be used to reassure parents, and prevent frivolous push for therapy. Uneven or slow development underlies the precocity paradox
Case studies
The following examples show how kids learn to read on their own, how school obliterates the ability to read, and how freedom restores the ability to learn to read. You can also see how dyslexia may easily develop in bilinguals in a language that is subject to coercion as opposed to the language that develops naturally. I am in direct communication with all involved parents and children (in case you had any questions or doubts).
Case study: Power of freedom
A single story of Renata and her three daughters is enough to convince skeptics to the mechanism of educational dyslexia. If science does not speak to you, this anecdotal evidence will. Renata is a single mom who I know personally. She was kind to lay the entire truth to me, incl. medical data that some parents or kids might find embarrassing. I have got to know more about her kids that many a kid in my own family. Her beautiful daughters all suffered from being labeled with learning disability at one stage or another. Today I watch them thrive in areas too many to mention.
That one story should be enough to convince you that coercion in learning is doing immeasurable harm to society.
All three daughters took entirely different paths that demonstrate the same thing: free children learn best. As a young mom, Renata was convinced that the path to reading goes through regular drills with a primer (elementary textbook). In Poland, the primer by Falski is most popular. Generations of kids learned reading with the same book. Renata was convinced that a nice teacher is the key to success. This is surprising because Renata learned to read at the age of 4 all on her own.
In the end, it turned out that school is a formula for dyslexia. In contrast, self-learning is a formula for reading with pleasure.
Here are individual stories in Renata's words:
Irena (b. 2008): It was a great disappointment when it turned out that my first daughter started avoiding reading after 3 months of school. The lady was racing with the program, completing subsequent letters in a phonics class, and Irena developed a mental block. She couldn't for the life of her understand why "py-i-es" would mean "pies" [dog]? The phonic method completely destroyed the enthusiasm and willingness to read. The desire and enthusiasm for learning in general was also destroyed more with the first failures that came, and the first punishments in the form of public reading aloud in the classroom. I started teaching my daughter at home using the syllable method, but it didn't go well. It was effective enough for her to more or less associate the letters with sounds. And then salvation came. The movie "Frozen" was in theaters and my aunt had just bought a book with stills from the movie. After two days, my daughter proudly announced that she wanted to read the book to me. And she read it. I was stunned. It turned out that she had been reading under the coverlet at night, and she had read it so many times that she had almost memorized it. Today she devours book and always has a book in her hand
Gabby (b. 2010): My second daughter's experience with learning to read at school was worse because she was suspected of having dyslexia. Learning to read was hard work and did not bring any results, and due to the pressure at school, my daughter began to stutter, became withdrawn and unhappy. It was a pain to read aloud in class. The third grade passed, my daughter's reading was very poor and there was no hope that anything would improve. Halfway through fourth grade, I moved my daughter to homeschooling (pandemic helped make the decision). For half a year she read literally NOTHING. I didn't force her because we were all under a lot of stress due to the pandemic and I didn't want learning to be an additional stressor for her. After half a year, during a visit to the library, she became interested in Donald Duck comics and liked them so much that over the next year she read all the Donald comics in local libraries. After a year of adventure with Donald, she started reading normal books. "Dyslexia" and stuttering disappeared
Hanna (b. 2013): I have a completely different experience with my youngest daughter, whom no one has ever taught to read. She learned to read completely on her own, using her own method, before she even went to school. She read whole words and whole sentences at once. And here, just like with older daughters, motivation was key, my daughter loved singing and wanted to sing karaoke with YouTube. This was a great method because it required reading at a pace and rhythm. She became proficient very quickly. School failures in learning to read result from lack of motivation and school oppression in the form of bad grades and being forced to read in public when the child is not yet ready for it. Children want to demonstrate their skills. However, they want to do it on their own schedule. When my youngest daughter learned to read, she announced that she wanted to read fairy tales aloud to us in the evenings. It gave her real pleasure
In all three cases, the breakthrough came from a factor that generated a motivation to read: movie "Frozen", comic "Donald Duck" and karaoke. Those contexts cannot be engineered at school. This is why so many kids, instead of enthusiasm, end up with a diagnosis.
The story would not be complete if I did not add that Hanna is a young genius, however, in kindergarten she was diagnosed with autism. The fact that she was in a special needs group spared her the early drills in phonics and reading!
Hanna was diagnosed with autism. She started talking late, she had echolalia for several years, i.e. she repeated fragments of cartoons she watched from memory, often several months later. For example, on the bus, Hania spoke to a passenger: "I'm in my stomach and you're sitting on me, move a little!". Only I knew that it was a text taken straight from a cartoon. Nowadays this has disappeared, but in turn she watches the same fairy tale or movie many times and learns it by heart. She loves to read regulations, encyclopedias, atlases, etc. She has some fixations, for example, she studied the solar system furiously for half a year. When we went to the zoo, she could act as a guide and had detailed knowledge about animals. She has difficulties in ordinary relationships because she is in charge and gets offended when the children "don't want to play the way I want". She doesn't understand social rules, she has to be taught it like any other subject at school. In her monologue about mushrooms, she got very angry when she remembered that I picked a toadstool. "Toadstools cannot be picked and that's it! You can't joke about this"
It is impossible to predict the direction Hanna will take, but it is clear her future is bright as long as it is never interfered with by the educational system of coercion. Genius is sensitive and is best left alone.
Figure: This angel spent the whole day in the library in the English books section. This is Hanna (b. 2013) in September 2023. Despite a tentative diagnosis of autism, she learned to read "completely on her own"
Case study: Johnny can't read
Johnny is 10 and he does not read. He was late to talk. He speaks little and not too correctly. His mom, Natalia, in desperation, considered taking Johnny to a school for kids with special needs. Ordinary school was a nightmare.
I suggested to Natalia that there is a better alternative. I suggested she homeschooled Johnny. I declared that even if she does not do anything, Johnny will read in 4 years. You may think that 14 is pretty late to read, however, I know Johnny will need a long detox period from his school experience. Moreover, having a remote deadline takes a way one of the most toxic components of school life: meeting milestones on deadline. This destroyed millions of kids around the world. I know that Johnny got a loving mom, and all he needs is some peace and time to find things he loves most in life. If those things involve any form of reading, Johnny will read. My declaration is pretty unusual because I hardly spoke to Johnny. Due to his speech delays, Johnny is not too talkative and shy. Knowing that I am a friend, he agreed to talk to me, but his defensive posture made him answer "I do not know" to most of my questions. That's not a problem. I know that each time we talk, Johnny will build trust and he will tell me more. I will then know more about his trajectory.
In Poland, homeschooled kids need to pass annual exams. In an ordinary school, such an exam may be a nightmare and horrible stress for the kid. However, we have a new player in the field of homeschooling: School in the Cloud (original Polish name: Szkola w Chmurze). It offers student-friendly exams with warm empathetic teachers. Students who suffer from any form of anxiety may choose a written or oral form of the exam. I do not know a single case of a student unhappy with his exams even though the mere concept of an exam brings sensation of horror for many kids affected by school experience.
Natalia trusted my word and agreed to homeschool Johnny. I mentioned scientific evidence, but my job was easier than that. I could list hundreds of similar cases that were rescued by School in the Cloud. The school offers online learning, free consultations, and voluntary psychological advice. Its educational platform is simple, easy to use, and devoid of all the excesses of Polish curriculum. Moreover, it offers test exams that kids treat almost like a computer game. It offers the basic minimum in the attractive form with students free to expand their interest in any direction they dream. Most importantly, the school takes an active role in defending their students against oppressive and not always legal actions of authorities. Obstacles are thrown by school inspectorate, local authority and the education department. School's exponential growth was halted by the minister of education, Przemyslaw Czarnek. Right-wing Czarnek branded the school "mafia" and complained of the "pathological avoidance of school duties". Czarnek wanted to crush the school with a new bill (Lex Czarnek 2.0), but the bill was vetoed by President Duda. Czarnek then cut school's subsidy to 20%. That put school's survival in question. Luckily, Czarnek was voted out of office in October 2023. Johnny is now in safe hands. He can be sure that nobody in the school will complain about his slow development. He will have all the time to learn the way he likes. Johnny passed his first exams in June 2023 at ease. Mom was extremely proud.
However, soon Natalia's anxiety returned. She did not have to worry about school and exams, but she still worried that Johnny does not read. The demands of the curriculum will keep increasing. Schools are required to be lenient with slow kids for the first 3 years. Natalia was again haunted by the demons of the Prussian school system. In her mind, Johnny should be reading already.
When working with Johnny, she noticed very slow progress. However, what she could not stomach were moments of regress. She is fearful when the kid shows poor memory (poor [asemantic memory]] is a norm in healthy development). This can be very frustrating. As Johnny does not speak too clearly, she tries to correct his errors. Speech therapy at school was too stressful. But learning with mom can also be frustrating. When Natalia insisted Johnny corrected his pronunciation of the word "microscope", Johnny was brought to tears. His nimble younger sister used it to mock her brother. Mom started questioning her own fitness to be a mother. Her health and well-being are at risk. In my eyes, this lovely family suffers from only one problem: culturally imprinted belief that kids need to read at 7. Coercive school system has a talent for destroying people and entire families!
Most of society seems to think that all delays are a sign of retardation. My prediction is opposite. Johnny is a talented guy who is more likely to be pretty far on the brain development trajectory projected by the precocity paradox.
A fantastic thing happened at Christmas 2023. Johnny got his dreamed of microscope (whose name he still mispronounces). When he spent 2 days straight using the device with passion, Natalia realized she got a little scientist at home. Johnny would microbiologically inspect all surfaces at home. He also decided to breed paramecium for observation. In Polish schools, paramecium is notorious for being a metonymy of boring school that requires knowledge of useless stuff. For Johnny, paramecium is now associated with discovering the world.
After Christmas Natalia told me that this was her best time in life. She was happy to tears.
This is naturally no happy ending. School will have its demands. Natalia will find no peace until Johnny reads. Later she will probably worry about further exams. Mom's anxiety never sleeps.
I promised I will keep talking to Johnny. I declared he will read in 4 years. I make this promise public in this text to be sure we use this case as a clear closure stamp on the wrong approach to education.
Is 4 years too long you think? I intentionally avoid deadlines. If reading happens earlier, we will celebrate. Deadlines are a prescription for anxiety and frustration.
Compare a similar but tougher case of Heather who was diagnosed with dyslexia, dyspraxia, ADHD, and bottom 3% IQ. After leaving school at 8, she took a longer detox break from reading. She learned to read at 14, and then she finished college. See her story in the words of Dr Harriet Pattison here(minute 20:48). Never mind the napping professor (nappers live longer).
Case study: Dyslexia in bilinguals
I keep studying the case of a fantastically smart and stubborn Olaf. He is an example of a rare case of dyslexia in one language (Polish), and fluent reading in another language (English). He passed his exams in Polish with difficulty, while he can read fluently in English. It is easy to guess that he was subject to coercion only in the language in which he cannot read. That's the effect of compulsory schooling.
Here is a slightly abbreviated text, which was posted by his parents at Peter Gray site:
Our child does not have formal diagnosis of dyslexia [...] but meets all the criteria. He was late to speak (2 years behind his peers). At the age of 10, he still could not sign his name and used a short nick instead. [...] At 6-7, our child could not break the barrier of counting beyond 13. He was very stressed with the fact. When we decided to stop asking, within about 5-6 months, using computer games, he learned to use large numbers, and simple arithmetic right in time for school. The situation with reading was much worse. He never liked books because he loves computers. Somehow, at the age of 9, he was able to install Linux and Windows unassisted while showing very little detectable ability to read. We have evidence of comprehension without being able to convert text to speech. While learning to read for exams in Poland, our son showed all signs of dyslexia. Reading seemed impossible. He managed to pass the exams when we asked him to stop reading in English (his first language), and only learn the phonics of Polish. In Polish, it is possible to robotically sound out words without comprehension. All he needed to do is to learn the sounds, and to stop using English to eliminate the interference between differently sounding characters. Whatever he learned for the exam, he forgot within 1-2 weeks, and was clearly stressed when we tried to maintain those little gains in his memory. At 9.5 years old, we stopped all forms of instruction, reminders, nagging, or questions. At the age of 9.8, the big shock arrived! The kid announced: "Stop reading for me. I know how to read". A bit of testing verified child's claim, he could read in English at 50-60% accuracy and correctly interpret all sentences. We have a great deal of evidence that all attempts at instructions did more harm than good. Our son learned to read from his daily use of the computer. At 12, Olaf reads books fluently in English. His reading in Polish is 5 years behind school requirements
Here are the conclusions drawn from Olaf's experience:
- Unschooling may delay reading
- Unschooling increases intolerance of all forms of instruction
- Unschooling leads to great passions and love of learning
- Coerced reading may have a tragic impact on the ability to read
As Olaf has no formal diagnosis of dyslexia, which requires a complex procedure, I asked an independent expert to provide an evaluation of Olaf's reading ability. In a dyslexia diagnostic test, Olaf scored 53 out of 63. Even after advancing to grade 6 in Poland, he is unable to read even simple texts in required Polish despite reading fluently in English.
Compare: Bilingual boy with dyslexia in English only
Epidemic of dyslexia
With my understanding of the role of toxic memory and the mechanisms of educational dyslexia, I would never bother a dyslexic kid with learning. I have no experience of working with children diagnosed with dyslexia. However, I talk to them all the time. I talk to them each time I get a chance. Normal intelligence is part of the diagnosis. Dyslectic kids are predominantly smart and curious. I am yet to see one that would look like a child with a neurodevelopmental deficit. Many get the diagnosis just to get an easier life at school! As nearly all kids in my country are subjected to coercive phonics, I have no qualms in positing that a vast majority of cases in Poland are educational dyslexia. If the mechanism is clear, if coercion is omnipresent, if many actors show vested interest in diagnosis, there is no escape from the conclusion that coercive schooling is the main culprit in the epidemic of dyslexia.
Mental health epidemics
School contributes to epidemics of dyslexia, dyscalculia, dysgraphia, ADHD, Asperger, autism, and more.
Here is the mechanism:
- coercion results in forming toxic memories
- persistent coercion results in learning difficulties
- teachers seek medical diagnosis as an excuse for lack of progress of students
- parents seek medical diagnosis as a relief from coercive pressures and for disability benefits
- psychiatrists driven by their empathy in the light of a pathological coercive system of schooling stretch the criteria to provide relief
The entire school system is a machine that is driven by vested interest and convenience of all actors. Children are the victims.
Some scenarios:
- A creative child that fidgets in class provides grounds for the diagnosis of ADHD. Medicated child is more obedient and ready to suffer through the boredom of class. See: Confusing creativity with ADHD
- A creative and highly sensitive child becomes socially awkward because of her departure from normative conceptualization. A diagnosis of Asperger's provides all the excuses parents and teachers need to keep the child in the system of coercive schooling. See: Trading genius for Asperger
- A creative child refuses instruction in phonics or in sight words. A diagnosis of dyslexia makes it easy for the child to get extra funding and a green light to pass exams on privileged rules. See: Educational dyslexia
- Coercion generates learned helplessness which is the first step to depression, addiction or suicide. See: Optimization trap of coercive schooling
As most kids lose most of their creativity in the first 3-5 years of school due to coercion, compulsory schooling is exacting an astronomical toll on the intelligence and mental health of society.
Summary
- all children learn to read naturally when they discover the need
- as we increase the pressure to read early, children hate school more than ever
- most children learn to read on their own between ages 3 and 18
- the average age to read for unschoolers is 9
- early reading is an advantage only when it is voluntary
- adults retain the ability to learn to read throughout life
- there is no optimum age to learn to read
- late talking and late reading often come hand in hand and may be a good sign (see: precocity paradox)
- reading develops slowly but naturally in environments rich in print
- learning to read at school is hard labor for the brain
- we should not care about fast learning, but about optimum learning
- the emergence of reading usually reveals itself suddenly
- the key to reading is the conversion of printed matter to meaning (sounding out is optional)
- school often teaches phonics and linear reading with poor comprehension
- school shows insufficient tolerance to errors that are a necessary part of learning to read
- in democratic schools, all children learn to read when the need arises
- little children employ semantic learning and their metacognitive skills are poor
- schooling conditions tolerance for asemantic learning, which leads to fast forgetting, weak generalization capacity, and the hate of schooling
- minor changes in context can result in major changes in recall
- pleasure of learning plays a role in recall even if it is not directly associated with the content of knowledge
- even a weak association with school can lead to turning off the consolidation signal (the learning does not occur)
- at school, instead of learning, kids often learn to dislike learning
- early in life, self-directed free learning is the most efficient form of learning
- early academic instruction may cause harm by forming toxic memories
- even a tiny does of coercion can ruin fantastic learning opportunities
- one memorable event may count for more than dozens of repetitions for school
- learning phonics is ok as long as it is voluntary
- reading wars are all about optimizing coercive learning
- when children are made to read before they understand the world, they naturally increase their tolerance for reading without comprehension
- epidemic of dyslexia is associated with the increase in the pressure for early reading (see: Educational dyslexia)
- only 1% of dyslexia has biological origins
- best evidence for educational dyslexia comes from bilingual children (see: Bilingual boy with dyslexia in English only)
- compulsory schooling should be blamed for the epidemic of dyslexia (and other learning disabilities)
- in cases of slow progress in reading, the most important thing to do is to take away the pressure
- when reading is fluent, incremental reading might be the best training tool in efficient, non-linear, semantic reading
Further reading
Wozniak
Most of the ideas mentioned in the above text have been covered in the following texts at SuperMemo Guru:
- Toxic memory: stress at school can result in dyslexia, dyscalculia, stuttering, and more
- Reading wars are over: Whole language vs. Phonics: children settle the reading wars with free learning
- Educational dyslexia: most of dyslexia observed in coercive schooling is a result of coercion
- Asemantic curriculum: curriculum that makes no sense to a kid will not be retained in memory
- Problems with special-needs education: teacher lobby generates special needs which generate special needs lobby
- How school turns off memory: coercive school poisons all associations formed while learning
- Precocity paradox: late to talk, late to read? you may have a genius at hand!
- Semantic learning and Asemantic learning: semantic brains reject knowledge of no measurable projected use
- Optimality of the learn drive: why we can be sure that the child is always right in her educational choices
- Brain is a perfectly adaptable device: why it is the brain, not the teacher who is right
- Childhood amnesia: why early instruction in declarative knowledge is a waste of time
- Tunnel vision of school letteracy: in the era of YouTube, pushing phonics is doubly discouraging
- Hole in the wall: kids learn great even if all they have is a computer
- Accuracy in reading is costly: school perfectionism takes away all the fun of reading
- 100 bad habits learned at school: learning in a coercive school result in forming an alien human being
- Why kids hate school?: kids love some aspects of school, but they universally hate coercion in learning
- Special needs education: how teaching to read may power special needs lobby
- Dual-process model of white-matter development: brain development mirrors erroneous reading strategies
- Gray: Mechanics of educational dyslexia: democratic schools know of no dyslexia
- Daycare misery: giving up your kid for the sake of "education" will slow down her brain development
- Schools contribute to the dyslexia epidemic
- Can coercion cause dyslexia?
- Hole in the wall
Others
- Peter Gray: How children learn to read (4 min. video)
- Seven Principles of Learning to Read Without Schooling
- How dyslexic kids learn to read: Pioneering effort by Dr Peter Gray
- Much too early: David Elkind's opinion on early reading instruction
- How Learning to Read Changes the Cortical Networks for Vision and Language
- Children learning to read later catch up to children reading earlier: Dr Suggate explains why early instruction is pointless
- Larry Sanger: How I taught my toddler to read: Larry Sanger's report on how his 2 year old could read
- Can babies read?: more about Larry Sanger kids
- Rush little baby: about precocity paradox
- On Teaching Reading and Spelling by Alfie Kohn
- Specific reading disability (dyslexia) by Frank R. Vellutino (and others)
- Rethinking Learning to Read, by Harriet Pattison
- Reading: A psycholinguistic guessing game: historic paper by Kenneth Goodman
- Thought, Language and Reading Readiness: How early reading produces disabled readers
- Sam Blumenfeld - Dyslexia: Man-Made Disease (1993)
- I was a teacher for 17 years, but I couldn’t read or write
- Link found between resilience to dyslexia and gray matter in the frontal brain
- Don’t Teach Kids How To Read, Teach Them Why
- How Jessica learned to read fluently on her own by the age of 6