180 ifie Subject Areas part Il + Age Changes in Spec ming: that i, they are highly familiar with yy, ir ieee vas a result, they are free £0 use their atten, ed by RAN ee gs, In ado, ss-year-olds who ae Deni tional resources 10 think about OFT’ from six-year-olds who are Ho. Fluency only Teer nares must fe in mporans aE TT sent or preschool teacher) must present comes from lots of repetition. his same person might also engender 4 letter names to children i in id in daily interactions an n ii 6 in them as well (Goodman, 1991). a A Oe to note pee i TRAN and reading nt only Hol fr eters But alo fy that the lictive relati rapidly naming colors, num bers, and objects (Compton, 2003). Thus, the connection is not purely due to repetii s rte sue of io letter names may use this information ‘second possibility is that i adams, 1990; Treiman, Tincoff, & Richmond-Welty, beret Teas east AleE sound like how a letter is pronounced. For exam. cand it is pronounced “buh.” Children may exploit this similar. “dcorrespondences. Thus, the key to the process may not be leter wn of the connection between letter names, pronunciations, ‘and the alphabetic principle (i.¢., that there is a systematic correspondence between letters ang sounds), Rebecca Treiman and her colleagues have provided considerable support for the idea that letter names help children come to appreciate the alphabetic principle (Treiman et al, 1996; Treiman, Weatherston, & Berch, 1994). To illustrate one key find- ing, many preschoolers spell words that start with w using y. The name of y is /wei/ and this name is close to the “wuh” sound made at the beginning of words like woman, The story for phonemic awareness is similar. Children cannot appreciate the al- phabetic principle until they can “hear” the component sounds of words and map these sounds onto letters and groups of letters (i.e., graphemes). Early in the preschool period, are also fast and accurate in ee ees information (as assessed by RAI ple, the name of B is ity to induce symbol-sound con names per se, but the recognitio children cannot say which words rhyme and whether a string of words all start with the ‘name initial sound. To be able to perform these tasks and map sounds onto print, two things have to occur. First, they need to have segmented mental representations of words la 1999). In such representations, a sound of word is stored as an interconnected _ pat component sounds. At first, the phonetic representation of a word is stored as an undifferentiated eprese (ee. bat ash tea eg Se bat 28 Pout), Over time, it sCh se Sapien Reading : 181 Nevertheless, the information as a whole». ofthe differences between successful ang ieee ces ction between peorital Causes and distal causes ae oe proximal causes or eeee Sisrenser include sentenve/sna’ ded to earlier. The three 21 aoe sefl children) Rae MO Ae found o be ETON. leer knowleage (successful children) have greater Sentencelstory ooOd Teaders at the end of first yemic awareness than children wh ry Memory, letter knowledge, and mpetent in reading at the 8 tentative explanation eT, We return to the dis- tine fs the ability to consciously reflect on the extensive practice recognizing letters, writ 8 letters, ar is Further, they have had many stories read to them, eee forstories, which in turn help them to reme, m ne Out invented spellings, them, a 0 create mental schemata Soros Caper neo i Cae Te deal and asked to express themselves on numerous occasions, Win sy on Et cesfl children have an extensive productive vocabulary, practice wats ea i * , Stories read pile, and regular conversational encounters? Because icy come fester Ghilren who come from high-SES homes tend to (a) be expened iat os (have parents who have advanced language skill and who intentional sees tto lees and writing, and (c) attend high-quality preschools and Kinderganere oan engage children in language games (e.g., rhymes and Songs) and emergent literacy activi- tes (Sénchal & LeFevre, 2002: Snow et a, 1998), Further evidence of the role of SES comes from several additional sources, Studies of disadvantaged children show that these children have considerable diffcuty -hearing™ the individual sounds in words as first graders (Adams, 1990: Wallach, Wallach, Dovier & Kaplan, 1977). One could interpret ths finding as evidence that there isa higher incidence of reading disability in disadvantaged children than in advantaged children, or that these shildren have not had the relevant formative experiences described above, The later seems more likely, given the success of various tutoring programs that have brought many disad- rai s 996), Relate. Jers about ten minutes will have been ex- |. Note that the aver- (Adams, 1990). le, up to the point a nicely with the re- promote important factors. however. is ss are not only pre- Teams to read! That ) show’ more phone are just starting182 pant « Age Changes in Specific SHS Avees 1998; Stanovich, 1988). Mor a) 2003; Snow et al. t none trad a at eect awareness, and insight into the alphabetic principle wd oi vee AS ho ad. TNY 8 a much ba ag is a large part of ai well i the first grade only if noun ee ssi pons children learn to de (Adams, 19 Goodman, 1991) s ie whe a ea snowball” over time (Stanovich, 1986). these studies is that reading skills tend t0 Precocious Readers et that most preschoolers vigorously resi nt en naa eg eal nen pildren are below the age of five (Feitelson, or Srl), There are, however, documented cases ‘of precocious reat 10 five when they ee reading ewe ‘y few experimental programs that have 1371 Ole eel on Four and ve- year-olds (Feitel#on ch al, 1982; Fowler, 1971). If we combine this evidence with the fact that it is often straightforward to teach {pondisabled) working-class and nee ereagssix-yea-olds to read (Snow etal 1998) the data as a whole can be interpreted in two Ways: ‘According to the reading readiness view, data 3 4 nerological basis o being ready and iling to read. According to a motivational ther age ends reflect normative trends iP reading motivation. ath egard to the reading readiness views it WS noted earlier that well-controlled srudis have found that age alone is not a Very £008 predictor of responsiveness to reading srction when factors related wo knowledge and experience are taken into account tadams, 1990; Bryant, MacLean, Bradley, & Crossiand, 1990; Stanovich & Siegel, 1994) Early studies (e.g. Fowler, 1971; Morphett & Washburn, 1931) showed that it was a child's renal age, not his or her chronological age that mattered (©-B- 4 two-year-old who could ent fag-year-old questions on an 1Q test could lear to read better than & five-year-old wvho eould only answer four-year-old questions). Moreover, mental ag¢ and intelligence Mere soon, replaced in later studies with better predictors such as leter knowledge and It is common for t¢* being taught 0 read, d have general phonological processing. Children are clearly not born with knowledge of letters and they q n iuctive vocabulary (through experience) in order to create described earlier (Met- ng links between ould reflect a lack of precocious readers als who leam to read er, 1971). reflect the fact that it , knowledge of letter dge, however, must beChapter 7 Beginning Reading 183 est that only the neurons in the frontal lobe cong; fancy (Byes, 2001), $0 the primary tempo; see meer the time it takes for synapses to form after repeated sain elated 10 m tation eg. the sight of a given letter). Conve camer : sel, itis possi : iiss and decoding are subtended by areas inthe Fron bl that skis ike phonemic aware. i is i (Shaywiwz et al, areas take some time to finish their development, iywitz et al., 1998). If these plain why itis hard to teach children below the there may be two kinds of factors that ex- read: (1) experiential factors bstantially beyond in- ‘ally toread as preschoolers. Developmental Models and Trends In the previous a ‘we examined reading devel which children enter the first grade, In this section, we shail e i Y In particular, this section begins with a description of some oe in performance. Then, a developmental model of early decoding skill is described, The general developmental trends can be understood with reference to the model of proficient reading described earlier. In the earliest phases of reading development, children have very litle orthographic knowledge but considerable knowledge cf meaning-sound re- _Iations (i.e., oral vocabulary). Moreover, as mentioned above. successful readers also enter first grade with phonemic awareness, syntactic awareness, the ability to ‘apidly name let- ters, and knowledge of the alphabetic principle. Early reading development consists of putting all of this entering knowledge together with written words; that is, it can be char- acterized as the progressive acquisition of letter-sound correspondences and well as letter- meaning correspondences. Because word recognition is slow and effortful in the beginning, beginning readers rely heavily on prior context to help them guess words rather than read them phonetically. ‘With lots of repeat . word recognition eventually becomes automatic, thereby dicating thatthe con ns between jeanings are firmly established. Whereas in reading, they are no kup systems to the lopment from infancy to the point at increase these skills allows readers to at- sion strategies (see it can also be cast in to pronunciations). junce new words that4 ges in Specific Sublect Areas Par Il + Age Chan ag (je. making educated 2Uesses aboy jag by predicting a 7 share letter clusters)- So jeuess). For familiar words. however, they tend tg ves based on context automatic. The gos! iS (0 become a reader who shifs reading because 1015 ches to relying primarily on sight reading. from relying mostly 0? © ore they can allocate most of their cog. When readers acquire af e316 ‘ord meanings and creating mental modes, nitive resources 10 t2skS such ‘ve reading and repeated €XPOSUe to the Sight vocabularies are een printed words, their meanings, spellings, iD ec the sight of a word triggers rapid retrieval or the lat. ns. n a Ea ongitudinal suudies with young children that extend preschool riod to the point at which many children are fluent readers in the i ests that the connection-forming process proceeds through thitd grade, Bhri (1995) OEP ohabetic, full alphabetic, and consolidated alpha- a Sete wed To denote the fact that (a) words consist of leters and ol ee function as symbol for phoneme and phoneme blends in words. During the realphabetic phase, “beginners remembeh how to read sight words by forming con- are araverween selected visual atibutes of words and their pronunciations or mean- rey and storing these associations in memory” (P 118). For example, they may see the ine iter gin dog and associate it with real dogs, oF see the TS humps in m and aa acl with real camels, This phase is called pre-alphabetic becase children are not really focusing on letters and their association to phonemes. This is also the phase a ao retten engage in reading environmental print such as stop signs and fast-food ‘inbols During the partial alphabetic phase, “beginners remember how to read sight lords by forming partial alphabetic connections between only some of the letters in vveitea words and sounds detected in their pronunciations” (p. 119). For example, they n, associate these two letters with their names (not spoon based on prior encounters (e.g., the last few ddly, they might peChapter 7 * Beginning Reading 185 Individual Differences in Reading Skills If we look at the reading skills of children in any grade, well below average, some read at grade level, and some h section, we will explore some of the variables that mi differences. The first class of variables relates to cogniti of variables relates to demographic factors we find that some children read lave above-average ability. In this ight account for such individual ive processes and the second class Good Readers versus Poor Readers Core Differences. With respect to cognitive processes, there are a number of ways in which good readers could conceivably differ from poor readers. On the one hand, they could differ in terms of general processing factors such as intelligence, working memory capacity, perceptual ability, rule induction, and metacognition. On the other hand, they could differ in terms reading-specific processes such as word recognition, use of context, phonemic awareness, and comprehension strategies. It turns out that significant differences have been found between good and poor readers for all of these variables (Stanovich, 1980, 1986, 1988). The question is, however, which of these variables seem to most clearly dis- tinguish good readers from poor readers. Careful reviews of the literature have revealed three particularly important differ- ences between good and poor readers. First, good readers are better than poor readers at recognizing words automatically (i.e., they do not have to pay attention to the decoding process). When word recognition is automatic, a reader can focus his or her attention on higher-level sentence integration and semantic processing (in the same way a skilled dri- ver can drive a car and have a conversation at the same time). However, automatic recog- nition is most important in the first and second grades because most high-frequency words ‘are automatized to adult levels by the third grade (Stanovich, 1980). Beginning in the third grade, it i and poor readers emerges: units (de Jong & van der operate on information between good and logical representation fading by (a) providing ‘a more stable code for jovich, 1980). to be a major differ- however. this proposal uch if not more than they encounter an un- much more heavily on 1, Context only exerts doctored or degraded. older. poor readers doPart Il + Age Changes in Specific Subject Areas not show a similar kind of decreasing reliance on context presumably because the so much trouble recognizing and deciphering a word. Of course, this is not to say th text is irrelevant to good readers. As mentioned earlier, context serves as an im, backup system to the connections between text and meanings. Y have ‘at Con. Porta, Explaining Core Differences, Why are good readers better at automatic and fast reco. nition and pronunciation of words than poor readers are? From what we know about a nature of skill acquisition and the formation of associations, it seems cléar that good reag. ers have had considerably more practice at recognizing and pronouncing words than readers, A likely cause of practice differences could be the fact that children are groupe by reading ability starting in the first grade. Children, in higher groups are given more op. portunities for practice than children in lower groups are and the initial gap between groups widens with age (Stanovich, 1988). However, it would also appear that there are fairly stable individual differences in the core phonological processes of reading ability (Stanovich & Siegel, 1994; Torgesen & Burgess, 1998). One study, for example, showed that the longitudinal correlations between phonological processing abilities in successive years in school ranged from a low of 6210 ahigh of .95 (mean = .81). Hence, it would appear that children's relative ranking in terms of their phonological processing skills remains fairly constant over time. In further suppor of this claim, another study of disabled readers showed that children who scored at the tenth percentile for phonological skills at the start of school ended up no higher than the thirtieth percentile by the end of fifth grade, even after being provided with remediation services (Torgesen & Burgess, 1998). The stability of individual differences and apparent intractability of severe reading problems has led many to wonder whether reading skills have a neuroscientific basis. In what follows, we will look at several lines of neuroscientific research to see what they can tell us. To begin with, studies of brain-injured adults have revealed a number of distinct (but sometimes co-occurring) deficits in their reading skills. At one time, the typical ap- proach was to classify collections of deficits in terms of syndromes (e. dysgraphia versus dyslexia without dysgraphia). Since the 1960s, information-processing and other psychologi unis have prompted investigators to. subdivid ing problems in of writtenovaj 1 > beginning Keaaing 187 parietal lobe, the right (terminal) Part of words is neglected when there is damage to the left parietal lobe. Attentional dyslexics can read individual letters quite well (e.g., A), but have significantly more difficulty when they have to read individual letters that ne foghed by other letters in the visual field (e.g., the A in “K A L”). The few patients who have had attentional dyslexia have had large tumors that ‘occupied posterior regions of the left hemi- sphere and extended into subcortical structures. The rarity of the disorder makes the pre cise location of damage difficult to specify at present, In contrast to visual word-form dyslexias, central dyslexias are thought to include reading processes that occur after the initial visual processing of words (McCarthy & War. Fington, 1990). The two major types of central dyslexias include reading by sound (surface dyslexia or phonological reading) and reading by sight vocabulary (with a corresponding inability to sound out words). Surface dyslexia is a problem because of the abundance of words that defy the rules of regular letter-sound correspondences (e.g., yacht, busy, sew; erc.). Because of their overreliance on pronunciation rules, surface dysiexies are likely to pronounce irregular words in predictable ways (e.g, sew as “sue”) and also pronounce phonologically regular pseudowords quite well (e.g. “blean"). Anatomically, surf dyslexia has been associated with a wide range of lesion locations, but usually tends to i volve damage to the temporal lobes in conjunction with damage to other areas. The second type of central dyslexia consists of being able to read using one’s sight vocabulary but losing the ability to read by sound. In contrast to surface dyslexics, such patients are very poor at reading pseudowords that obey the spelling-sound rules of theie . language (e.g., “blean” or “tweal”). In addition, patients with the latter disorder may also have difficulty reading function words (e.g. if for ete.). grammatical morphemes (e.., -ed ot ~ing), or abstract words (e.g., idea), and they sometimes make semantic errors as well. The co-occurrence of pronunciation difficulty and difficulty with abstract words has been called deep dyslexia. Case studies of such patients reveal no consistent pattern of local- ization (Warrington & McCarthy, 1990). ‘The late Norman Geschwind tried to summarize the literature on acquired dyslexias in the form of the anatomical model presented in Figure 7.4 (e.g., Geschwind & Galaburda. 1987). As i that a written is first registered in the st i then relayed to the (a) perceiving syntactic pro- related to re- tract-concrete ynd redundant at least some ically regular NTE CTBT, a188 Part Il + Age Changes in Specific Subject Areas Parietal Lobe Occipital Lobe Frontal Lobe Forehead Back of Head FIGURE 7.4 Brain Areas Associated with Reading ProcessesChapter 7 * Beginning Reading 189 Problems and mechanisms that are unique to developmental dyslexia (Raynor & Pollatsek, 1989). Further, it was noted earlier that deep dyslexia has not been associated with dam, age to particular sites in the brains of adults. If deep dyslexia in adults is, in fact, similar to developmental dyslexia in its manifest symptoms, itis not clear why developmentalists would expect to find abnormalities in particular sites in the brains of children, Notwithstanding these caveats, Shaywitz and colleagues (1998) have used MRI tech- nology to determine whether different patterns of activity can be observed in the brains of dyslexic children and nondisabled children when they read. Focusing on the regions iden- tified by Geschwind (see Figure 7.4), they found that brain activations differed signiti- cantly between the two groups, with dyslexic readers showing relative underactivation in the striate cortex, angular gyrus, and Wemicke’s area, and relative overactivation in Broca’s area. Whereas these authors suggest that this pattem of activation might be a “signature” for dyslexia, it should be noted that involvement of left frontal areas is also indicative of task difficulty and working memory (Barch, Braver, Nystrom, Forman, Noll, & Cohen, 1997). To show that the aforementioned pattern realy is a sign of functional disruption in the reading circuitry, these researchers would need to demonstrate at least two additional things in future studies: (1) that the pattem is not observed in nondyslexic children who are given a task that is hard for them and (2) that the pattem in not found in nondisabled children who are just beginning to read (who might have as much trouble pronouncing new words as older dyslexic children): With respect to PET scan studies of normal adults, Posner, Peterson, Fox, and Raichle (1988) and Petersen, Fox, Snyder, and Raichle (1990) have found that passively’ looking. at real words (e.g., board), pseudowords (e.g.. “floop*), nonwords (e.g., “jvjfe"). and strings of letterlike fonts all activate the same portions of the occipital lobe. However, real words and pseudowords also activate portions of the occipital lobe that are not activated by non- words and false fonts (the extrastriate cortex). Recall that damage to occipital regions often produces the visual word-form dyslexias that were described earlier. Whereas presentation of real words and being asked to define words activates regions of the left frontal lobe, pre- sentation of pseu and extrastriate 3 poral lobes and. activation for crepancy betw