How the Brain Learns to Read

by David A. Sousa

  • The second phase is the phonological stage, when the brain begins to decode the letters (graphemes) into sounds (phonemes).

  • The third phase is the orthographic stage, when the child is able to recognize words quickly and accurately.

  All of these phases activate several different brain circuits, which, over time and with practice, eventually converge in a specialized area of the left hemisphere. This area is now referred to in the scientific literature as the visual word form area. Let us see how this amazing process occurs.

  READING IS NOT A NATURAL ABILITY

  Humans have been speaking for tens of thousands of years. During this time, genetic changes have favored the brain’s ability to acquire and process spoken language, even setting aside specialized areas of the brain to accomplish these tasks, as we discussed in Chapter 1. Consequently, the brain’s proficiency at hearing and quickly remembering spoken words is natural, though no less remarkable. Remember that children begin to learn words before their first birthday and during their second year are acquiring them at the rate of 8 to 10 per day. By the time they enter school, they have a well-developed language system consisting of an active vocabulary of about 3,000 words and a total mental lexicon of over 5,000 words. At some point, the child’s brain encounters the written word and wonders, “Hmm . . . What are those symbols? What do they mean?”

  Speaking is a normal, genetically hardwired capability; reading is not. Writing was born about 5,000 years ago in an area of the Middle East known as the Fertile Crescent. That time span is just a blip in evolutionary terms, and hardly enough time to develop specialized brain circuits for reading. Thus, reading is a relatively new phenomenon in the development of humans. As far as we know, our genes have not incorporated reading into their coded structure, probably because reading, unlike spoken language, has not emerged as a survival skill over the relatively brief time that humans have been reading. And yet we do read. How does that happen? What parts of the brain must be recruited to carry out this exquisitely complex process of matching sounds to corresponding lines and squiggles? Because no areas of the brain are specialized for reading, it is probably the most difficult cognitive task we ask the young brain to undertake. If reading were a natural ability, everyone would be doing it. But in fact, according to the National Institute for Literacy, nearly 40 million adults in the United States alone are functionally illiterate. The Canadian Council on Learning estimates that about 12 million Canadians have low literacy levels.

  The Plasticity of Neural Networks

  Although Frith’s model is helpful, the real question is how a brain with no innate reading center learns to accomplish this complex task. How are brain regions not designed to read coopted to undertake that challenge? Psychologists and neuroscientists have wrestled with these questions for years. Now, thanks to the growing bank of thousands of brain images and other scans, as well as a large inventory of case studies, some new ideas have emerged about how the brain adjusts to the challenge of interpreting visual signals from print into sounds that have meaning.

  We already mentioned that writing appeared in our culture too recently to have made alterations in our brain through the slow process of biological evolution. Yet the fact that we can read must indicate that our brain makes adjustments in its capabilities that do not require changes in the genetic code. In other words, momentous events in our culture, like the invention of writing, can apparently cause important cerebral adaptations to occur as a result of cultural learning. These adaptions occur because of the brain’s plasticity—that is, its ability to adapt to significant changes in its environment.

  Neuronal Recycling

  Neuroscientist Stanislas Dehaene (2009) has proposed a new term, neuronal recycling, to describe the taking over of a brain region, initially devoted to a different function, by a cultural invention—in this case, writing. “Recycling” refers to retraining brain areas that performed an ancient function in our evolutionary past to carry out a new and more useful function in our present culture. This recycling, Dehaene asserts, does not totally undo the preexisting predispositions of these brain regions, but works around them. After all, the brain’s plasticity is still constrained by its extensive cerebral networking, genetic biases, and other factors that limit its degree of adaptability.

  At what age, and how, does this neuronal recycling for learning to read begin? Renewed emphasis in recent years on improving the basic cognitive skills of students has increased pressure to start reading instruction sooner than ever before. In many schools, reading instruction starts in kindergarten. Current researchers do not identify a definite age at which the brain can begin to learn to read. More important than chronological age is the degree and pace of brain development, both of which can vary widely among children of the same age. Some children make the transition from spoken language to reading with relative ease, once exposed to formal instruction. For others, reading is a much more formidable task, and for some, it definitely becomes the most difficult cognitive task they will ever undertake in their lives.

  Answer to Test Question #2

  Question: Learning to read, like learning spoken language, is a natural ability.

  Answer: False. Unlike spoken language, the brain has no areas specialized for reading. The skills needed to link the sounds of language to the letters of the alphabet must be learned through direct instruction.

  EARLY STAGES OF READING

  Intelligence generally does not play a critical role in learning to read. Three sources of evidence indicate this. First, studies of children who learn to read before entering school do not indicate a strong relationship between IQ and early reading. Second, studies have shown that IQ is only weakly related to reading achievement in Grades 1 and 2. Finally, children who have difficulty learning to read often have above-average IQs (Kortteinen, Närhi, & Ahonen, 2009; Paloyelis, Rijsdijk, Wood, Asherson, & Kuntsil, 2010; Shaywitz, 2003). It appears then that, to a large degree, learning to read is independent of intelligence. This is an important point because some teachers mistakenly assume that children with problems learning to read are of lower ability and will also have difficulty in other subject areas. Such a presumption can lead to lower expectations and less challenging work for those children.

  “To a large degree, learning to read is independent of intelligence.”

  Before children learn to read, they acquire vocabulary by listening to others and by practicing the pronunciation and usage of new words in conversation. Adult correction and other sources help to fine-tune this basic vocabulary. Because the ability to read is strongly dependent on the word forms learned during this period, a child’s beginning reading will be more successful if most of the reading material contains words the child is already using. The phoneme-grapheme connection can be made more easily. Reading, of course, also adds new words to the child’s mental lexicon. Consequently, there must be some neural connections between the systems that allow the brain to recognize spoken words and the system that recognizes written words.

  Multiple Lexicons. We should mention here that there are several different types of lexicons in the brain of a proficient reader. The mental lexicon generally refers to the store of familiar words. But we also have an orthographic lexicon for visual recognition of letters, graphemes, and morphemes. For instance, the orthographic lexicon might deconstruct the word island into is + land. Our phonological lexicon, which stores how words are pronounced, would tell us that this word is pronounced eye-land. No doubt we also maintain a grammatical lexicon that contains the rules of plurals and sentence structure. Finally, to arrive at meaning, each word is associated with certain properties. For example, an island is an isolated piece of land, surrounded by water. This information is kept in a semantic lexicon. These various dictionaries communicate with each other to provide the information needed to make reading successful.

  Learning to read starts with the awareness that speech is composed of individual sounds (phonemes) and a recognition that written
spellings represent those sounds (the alphabetic principle). Of course, to be successful in acquiring the alphabetic principle, the child has to be aware of how the phonemes of spoken language can be manipulated to form new words and rhymes. The neural systems that perceive the phonemes in our language are more efficient in some children than in others. Just because some children have difficulty understanding that spoken words are composed of discrete sounds doesn’t mean that they have brain damage or dysfunction. The individual differences that underlie the efficiency with which one learns to read can be seen in the acquisition of other skills, such as learning to play a musical instrument, playing a sport, or building a model. To some extent, neural efficiency is related to genetic composition, but these genetic factors can be modified by the environment. Nonetheless, being aware of sound differences in spoken language is crucial to learning to read written language.

  Phonological and Phonemic Awareness

  Phonological awareness is the recognition that oral language can be divided into smaller components, such as sentences into words, words into syllables, and, ultimately, syllables into individual phonemes. This recognition includes identifying and manipulating onsets and rimes as well as having an awareness of alliteration, rhyming, syllabication, and intonation. Being phonologically aware means having an understanding of all these levels. In children, phonological awareness usually starts with initial sounds and rhyming, and a recognition that sentences can be segmented into words. Next comes segmenting words into syllables and blending syllables into words.

  Phonemic awareness is a subdivision of phonological awareness and refers to the understanding that words are made up of individual sounds (phonemes) and that these sounds can be manipulated to create new words. It includes the ability to isolate a phoneme (first, middle, or last) from the rest of the word, to segment words into their component phonemes, and to delete a specific phoneme from a word. Children with phonemic awareness know that the word cat is made up of three phonemes, and that the words dog and mad both contain the phoneme /d/. Recognition of rhyming and alliteration is usually an indication that a child has phonological awareness, which develops when children are read to from books based on rhyme or alliteration. But this awareness does not easily develop into the more sophisticated phonemic awareness, which is so closely related to a child’s success in learning to read, especially if such awareness begins in preschool settings (Phillips, Clancy-Menchetti, & Lonigan, 2008). Nonetheless, reading programs that emphasize phonological and phonemic awareness have proved to be successful in schools, especially with children displaying some difficulties when learning to read (Bailet, Repper, Murphy, Piasta, & Zettler-Greeley, 2011).

  Phonemic awareness is different from phonics. Phonemic awareness involves the auditory and oral manipulation of sounds. A child demonstrates phonemic awareness by knowing all the sounds that make up the word cat, /k/ /a/ /t/. Phonics is an instructional approach that builds on the alphabetic principle and associates letters and sounds with written symbols. To demonstrate phonics knowledge, a child tells the teacher which letter is needed to change cat to can. Although phonemic awareness and phonics are closely related, they are not the same. It is possible for a child to have phonemic awareness in speech without having much experience with written letters or names. Conversely, a child may provide examples of letter-sound relationships without ever developing phonemic awareness. In fact, simply learning these letter-sound relationships during phonics instruction does not necessarily lead to phonemic awareness (SEDL, 2001).

  The terms phonological awareness, phonemic awareness, and phonics have different meanings, but they can be easily confused. Table 2.1 defines these terms. In this book, I will refer mainly to phonemic awareness because many of the research studies on learning to read focus specifically on phonemes.

  Phonemic Awareness and Learning to Read

  Beginning readers must learn the alphabetic principle and recognize that words can be separated into individual phonemes, which can be reordered and blended into new words. This enables learners to associate the letters with sounds in order to read and build words. Thus, phonemic awareness in kindergarten is a strong predictor of reading success that persists throughout school. Early instruction in reading, especially in letter-sound association, strengthens phonological awareness and helps in the development of the more sophisticated phonemic awareness. One thing has become increasingly clear. The discovery of phonemes is neither innate nor automatic. It results only from the explicit teaching of the alphabetic principle. Beginning readers must learn the code before they can decode written words! We have known for at least three decades that even struggling adult readers can fail to detect phonemes in words. Phoneme recognition does not arise spontaneously (Morais, Cary, Alegria, & Bertelson, 1979).

  Table 2.1 Definitions of Terms Related to Speech and Reading

  SOURCE: Yopp and Yopp (2000).

  Numerous research studies over the past two decades have established a strong positive link between phonemic awareness and success in early reading. About 70 to 80 percent of children are able to learn the alphabetic principle after one year of instruction. For the rest, additional study is needed (Shaywitz, 2003).

  “Beginning readers must learn the alphabetic code before they can decode written words.”

  Sounds to Letters (Phonemes to Graphemes)

  To be able to read, the brain must memorize a set of arbitrary lines and squiggles (the alphabet). The brain does not yet know the system or logic of writing, but it begins to recognize patterns based on visual characteristics, such as curvature, orientation, and shape. For example, it notices that some symbols seem to be mirror images of each other (e.g., b and d, p and q), and that some are symmetrical (e.g., H, M, and O). It also detects that some symbols occur more often than others (e.g., a, e, r, and s), and that some letter combinations are more often at the end of the string (e.g., -ed, -ing, and -tion). This is what we referred to at the beginning of the chapter as Frith’s pictorial stage. Some of this visual letter/word recognition may occur before the formal teaching of reading. How many sight words are remembered will vary greatly among children.

  In the phonological stage, the brain ceases to process whole words and begins to identify which symbols, or graphemes, correspond to the phonemes already stored in the mental lexicon. Many European languages use abstract letters (i.e., an alphabetic system) to represent their sounds so that the words can be spelled out in writing. The rules of spelling that govern a language are called its orthography. How closely a language’s orthography actually represents the pronunciation of the phoneme can determine how quickly one learns to read that language correctly. Some languages, like Spanish, Italian, and Finnish, have a very close correspondence between letters and the sounds they represent. This is known as a shallow (or transparent) orthography. Once the rules of orthography in these languages are learned, a person can usually spell a new word correctly the first time because there are so few exceptions.

  English, on the other hand, has a poor correspondence between how a word is pronounced and how it is spelled. This is called a deep (or opaque) orthography. It exists because English does not have an alphabet that permits an ideal one-to-one correspondence between its phonemes and its graphemes. Consider that just when the brain thinks it knows what letter represents a phoneme sound, it discovers that the same symbol can have different sounds, such as the a in cat and in father. Consider, too, how the pronunciation of the following English words differs, even though they all have the same last four letters in the same sequence: bough, cough, dough, and rough. This lack of sound-to-letter correspondence makes it difficult for the brain to recognize patterns and affects the child’s ability to spell with accuracy and to read with meaning. Eventually, the brain must connect the 26 letters of the alphabet to the 44-plus sounds of spoken English (phonemes) that the child has been using successfully for years. Table 2.2 illustrates the complexity of English orthography, compared to some other related languages. There are
more than 1,100 ways to spell the sounds of the 44-plus phonemes in English.

  Table 2.2 Language Sounds and Their Spellings

  Looking at Table 2.2, one can easily understand the plight of students whose native language is Italian or Spanish when they are faced with reading English. Because there is such a close sound-to-letter correspondence (shallow orthography) in their native language, reading and writing unknown words is quite easy. There are very few exceptions in their language’s rules of spelling. But English is rife with spelling irregularities, and this poses a significant challenge to these and other English language learners.

  Alphabetic Principle

  The alphabetic principle describes the understanding that spoken words are made up of phonemes and that the phonemes are represented in written text as letters. This system of using letters to represent phonemes is very efficient in that a small number of letters can be used to write a very large number of words. Matching just a few letters on a page to their sounds in speech enables the reader to recognize many printed words: for example, connecting just four letters and their phonemes /a/, /l/, /p/, and /s/ to read lap, pal, slap, laps, and pals.

  Despite the efficiency of an alphabetic system, learning the alphabetic principle is not easy because of two drawbacks. First, the letters of the alphabet are abstract and thus unfamiliar to the new reader, and the sounds they represent are not natural segments of speech. Second, because there are about 44 English phonemes (this number varies slightly, depending on the counting method) but only 26 letters, each phoneme is not coded with a unique letter. There are over a dozen vowel sounds but only five letters, a, e, i, o, and u, to represent them. Further, the child needs to recognize that how a letter is pronounced depends on the letters that surround it. The letter e, for example, is pronounced differently in dead, deed, and dike. And then there are the consonant digraphs, which are combinations of two consonants, such as ch, sh, and ph, that represent a single speech sound. There are also three-letter combinations, called trigraphs, such as tch and thr. With more practice at word recognition, the reader must work toward fast and accurate word recognition in order to increase reading fluency.