Reading information can also leave working memory if either of the following occurs:
• Too many minutes have elapsed since the information was last activated. For instance, just after reading the first sentence of a lengthy paragraph, the reader gets involved in a brief conversation not related to the reading with another child. When the reader returns to the text, the previously read material has faded, and the child must start again from the beginning.
Figure 2.4 Reading and memory interact in several ways. Reading that is interesting to the reader will pass through immediate memory into working memory for conscious processing. New reading may activate long-term storage to retrieve related gists already learned. Reading information can fall out of working memory by fading or by being displaced by additional information.
• Additional information is placed in working memory, and its capacity is exceeded. This can happen if the child is reading faster than he can comprehend, thus giving the brain insufficient time to form the necessary gists and then clear out the individual words. In this case, the words are removed from memory before the gist is generated, and comprehension suffers.
As more demands are placed on working memory during reading, other difficulties may arise. A young child who is just learning to read using a book with advanced vocabulary words will have trouble with comprehension because so much memory capacity is being used trying to decode unfamiliar words. Research studies on the pace at which people read have offered further insights into how memory affects comprehension (e.g., Carretti, Borella, Cornoldi, & De Beni, 2009; Fedorenko, Gibson, & Rohde, 2006). For example, the studies found that readers spent more time reading the topic sentence of a paragraph. This may indicate that working memory is exerting more effort to generate and ensure the retaining of this gist because it is most useful in understanding the rest of the passage, and for creating the gist of the entire paragraph. The research studies also demonstrated, to no one’s surprise, that readers remembered best those ideas and concepts that were referenced repeatedly, as well as those that were linked through cause and effect (SLC, 2000). On the other hand, readers spent the least amount of time reading the details of a paragraph and, thus, had difficulty recalling them. They were able to recall, however, details related to a humorous or vivid event, probably because such events evoke emotion, a powerful memory enhancer.
Memory and Comprehension
Reading without comprehension is an unfulfilling endeavor. Why go through all that practice and devote all that neural energy required to rewire parts of the brain if there is no reward at hand? The magic of proficient reading is that it allows us to form visions in our head of places we haven’t been or could not even exist, to discover the thoughts of those long gone, and to share our thoughts with others. Fortunately, nature has provided us with a visual processing system that allows us to create mental images as well as a memory system that can remember these images and thoughts for years to come. In what ways does reading comprehension influence the processes of memory storage, consolidation, and recall?
Gist formation is one model that helps explain some aspects of recall, but it does not account for many of the characteristics of memory recall that we all have experienced. A reader’s ability to comprehend gists is largely dependent on that individual’s past experiences and the mental networks that have evolved as a result of those experiences. We all use these networks to help us interpret the world and to predict situations occurring in our environment. Mental networks containing memories of our past experiences are important in helping us to comprehend text. Readers use these memories to interpret cause and effect, to compare and contrast, and to make inferences about the author’s meaning. Information that does not fit into our memory networks may not be understood or may be understood incorrectly. This is one reason why readers may have problems comprehending text on a subject in which they have no experiences even though they understand the meaning of every word in the text. Memory networks are greatly influenced by an individual’s culture. Thus, young readers who were not brought up in the United States may have a difficult time reading and answering questions about George Washington.
Memory networks store not just information, but also images. Our visual lexicon contains thousands of images from our past encounters. Some are vivid, and some are blurry, depending on how many times we recalled them. Recalling an image or a memory strengthens the neural pathways containing the elements of that image, thereby making it easier to recall and more intense. If I told you that I am now going to read you a story about a cat, chances are your brain instantly created a mental image of a cat, most likely one you know. You might even see its color, and hear its meows, or sense the softness of its fur. These images are not only essential to understanding language but are important components of reading comprehension.
Modifying Our Memory Networks
Our memory networks are created through repeated experiences with events, people, and objects that we encounter in our world. When we encounter a new experience, our networks can be modified in any of the following three ways (Figure 2.5):
• Accretion: The learner incorporates the new information into an existing schema without altering that schema. For example, suppose I visited a public library, and all that I experienced there fit into my long-held schema of a library as a place with just print material and a card catalog. As a result, I did not alter my library schema in any appreciable way.
• Tuning: The learner realizes that the existing schema is inadequate to accommodate the new information and alters the existing schema to be more consistent with the new experience. For example, when I visited a modern public library and realized that the card catalog was replaced by a computer database, I had to modify my library network to accommodate this experience.
• Restructuring: The learner realizes that the new information is so inconsistent with the existing schema that a new schema has to be created. For example, now my ability to access the print information at the local public library directly from my computer at home any time of the day or night has forced me to create a new schema.
WHAT DOES NEW RESEARCH REVEAL ABOUT READING?
Although studies using brain imaging and other types of scans have helped researchers understand more about how the human brain learns to read, there are still limits as to what these technologies can do. For example, they can detect which brain regions are involved in reading, but they cannot track reading progress directly in an individual child’s brain. By examining a large collection of images and data, researchers can make educated inferences about how we think such development occurs. However, it is difficult to know precisely whether activity in a brain region is due solely to a particular stimulus. Nonetheless, much progress has been made so far, and new findings are appearing regularly. Table 2.4 describes what researchers have surmised about the brain and reading, based on the information available at the time of this printing.
Figure 2.5 This illustration shows how a stored experience from a memory network helps us to interpret new information. As the new information is processed, the experience can be returned to long-term memory unchanged or altered to accommodate the new experience. In some cases, the new information cannot be accommodated by the current network, so a new one is created.
Reading Pathways
Using functional imaging scans, neuroscientists have discovered the neural mechanisms that are activated during reading. As would be expected, the scans have shown that all readers use neural circuitry dedicated to visual processing, because the curves and lines of the alphabet need to be visually analyzed to distinguish one letter from another. In addition to the visual processing area, Shaywitz (2003) and other researchers (Dehaene, 2009; McCandliss, Cohen, & Dehaene, 2003) noted that other areas of the brain were involved in reading. However, which of these areas the brain used was dependent on how skilled the person was at reading. Apparently, beginning readers use different neural pathways than skilled readers, most likely because certain br
ain regions need to be modified to detect the different shapes of letters and to make the sound-to-letter correspondences. Here is what the researchers found.
Pathways for Beginning Readers
Before learning to read, when a child sees a word, the left-hemisphere language centers do not show unusual activity (Table 2.4). Rather, regions in the right hemisphere activate, most likely processing the visual picture, or snapshot, of the word. This could represent the pictorial stage that we mentioned earlier in this chapter. During this prereading time, the child’s brain memorizes the snapshots from the shapes of the letters, similar to how it would recognize faces. For instance, we pointed out earlier that young children who spot and respond to a McDonald’s restaurant sign are probably reacting to the Golden Arches logo and a mental picture of the word McDonald’s, not to its phonemes. Nonetheless, the pattern-seeking capabilities of the brain are constantly at work, and there is evidence that this right-hemisphere region may visually differentiate strings of consonant letters (Maurer, Brem, Bucher, & Brandeis, 2005). At about the age of 6 or 7 years, the brain of a beginning reader responds more actively to printed words than to geometric shapes or meaningless letter strings—that is, nonwords (Maurer et al., 2006).
Visual Word Form Area. By the age of 8, the brain is aggressively recruiting areas in both hemispheres, especially in the visual recognition system, to deal with the challenge of increased reading. A clearly defined area of high activity emerges in the left hemisphere at the boundary of the occipital and temporal lobes (Parviainen, Helenius, Poskiparta, Niemi, & Salmelin, 2006; Yeatman, Rauschecker, & Wandell, 2013). That would place it slightly to the rear of the left ear. This region is officially called the left occipitotemporal area. However, scientists have suggested that it be called by a less technical term: the visual word form area, abbreviated as VWFA (Cohen et al., 2000). The term emphasizes its role in the visual analysis of letters and their shapes. With more reading practice, the VWFA attaches meaning directly to whole word forms, thereby allowing the reader to significantly increase fluency and comprehension (Glezer, Jiang, & Riesenhuber, 2009). Broca’s area, a specialized region for language located behind the left temple (see Chapter 1), is also significantly involved in this word analysis. Studies show that the activation in the VWFA and other language areas in the left-hemisphere area is in direct proportion to the child’s skill at phonemic awareness, or the ability to mentally manipulate the basic sounds of language (Turkeltaub, Gareau, Flowers, Zeffiro, & Eden, 2003). Once again, we see the impact of phonemic awareness on reading acquisition.
At this point, when the child’s brain now sees the word dog in print, the visual word form area records the three alphabetic symbols and activates the left-hemisphere language centers to provide and match the duh-awh-guh phonemes to their appropriate letters, duh = d, awh = o, and guh = g. This developing sound-to-letter relationship is the alphabetic principle in action. Eventually, a mental image of a furry animal is conceptualized in the mind’s eye, adding meaning to the symbolic representation of d-o-g. With more frequent exposure to the word dog, the VWFA will eventually attach the meaning to the word’s letter string, thereby increasing reading speed and comprehension.
Table 2.4 Comparison of Eye Movements of Novice and Skilled Readers
SOURCES: Häikiö et al. (2009); Rayner el al. (2001).
Children at Risk for Reading. Given the substantial amount of cerebral reorganization that takes place during the acquisition of reading, it is no surprise that difficulties can arise along the way. Imaging studies of poor beginning readers show variations in the expected development of the sites responsible for learning to read. For instance, although most brain activation associated with typical beginning readers occurs in the left hemisphere, a more bilateral pattern is observed in poor readers (Bach et al., 2010; Yamada et al., 2011). It seems that both the visual recognition areas and the frontal lobe of the right hemisphere are more engaged than in typical readers. These findings may indicate that the brain recruits right-hemisphere regions to compensate for difficulties that are occurring during the modifications of the left-hemisphere reading areas.
Pathways for Intermediate Readers
Learning to read enhances a child’s spoken language capabilities. Between 10 years of age and the beginning of adolescence, Broca’s and other language areas undergo significant modification in addition to the visual recognition system. Visual analysis occurs faster. With repeated encounters of the same word, the child’s brain makes a neural model—called a word form—and the child can read this word far more quickly. Just seeing it activates all the necessary components at once, mostly in the left hemisphere, without any conscious thought on the part of the reader. As more word forms collect in this cerebral region, reading becomes more fluent, and reading skill levels rise dramatically. Broca’s area now plays only a minor role in assisting this region. The more skilled the reader, the more quickly the VWFA responds to seeing a word—in less than 200 thousandths of a second (200 milliseconds), much faster than the blink of an eye (Dehaene, 2009).
Pathways for Skilled Readers
If the intermediate reader continues to read regularly, then by early to mid adolescence, the visual form word area reaches full maturity. As the reader acquires new vocabulary and more sophisticated rules of syntax, and as semantics play a greater role in establishing comprehension, additional brain areas are required and recruited. Scanning images reveal complex interactions and connections to numerous networks, mainly in the brain’s left hemisphere. Becoming an expert adult reader results in activity changes that are not present in accomplished child readers. Imaging scans reveal that there are similar but not identical patterns of regional brain activity when adult and child skilled readers perform identical reading tasks at comparable levels of performance. Children activate some neural regions during reading that adults do not activate, or may activate less, and vice versa. These differences in activity patterns probably reflect the increase in the efficiency of brain processing as the child matures into an adult (Schlaggar & Church, 2009).
Figure 2.6 reflects the most recent model of the neural networks required for skilled reading, and how researchers believe they interact. This model is considerably different from the linear model that scientists accepted until just a few years ago. At that time, scientists believed that visual input went from the visual cortex to the brain’s reading center in a region called the angular gyrus, located in the left parietal lobe at the top left side of the head. For many years, this structure was believed to be responsible for interpreting shapes in the environment, so it made sense to assume it also interpreted the physical forms of letters. From there, according to the older model, information passed through Wernicke’s and Broca’s areas for interpretation, and finally to the frontal cortex for comprehension. It was a fairly simple linear route.
But more recent and extensive imaging research indicates that the process is much more complex. Letter analysis actually occurs in the VWFA, while several other word processing functions are operating in parallel. Perhaps a dozen or more neural sites are involved. The newer model helps explain how a skilled reader can acquire the visual images of words, extract their roots, determine syntax, and ultimately find meaning in just a fraction of a second. Brain regions must be operating simultaneously and constantly exchanging information with other sites.
Figure 2.6 This model illustrates the various brain networks currently thought to be involved in reading. Visual input goes first to the visual word form area (solid dark blue areas) and is then distributed to regions spread across the left hemisphere. The medium blue areas (connected by solid arrows) are mainly used for gaining access to word pronunciation and articulation. The checkered regions (connected by dotted arrows) help in establishing meaning. All these areas (except the visual word form area) likely play other roles in processing spoken language. The diagram shows that reading results from collaboration between language and visual areas of the brain.
SOURCE: Ad
apted with permission from dehaene (2009).
It is important to note that most of these areas are not exclusive to reading, and may carry out other tasks. The VWFA, however, generally responds only to written words and not to spoken words. Some areas contribute to spoken language processing while others process visual information. Signals can move in both directions between sites. Learning to read tends to be linear at the very beginning, progressing from phonemes to morphemes to graphemes. However, it becomes more bidirectional as reading skills, especially comprehension, develop and expand. By showing how complex the reading process is, researchers realize that the model is not static and will probably evolve as newer findings emerge. In fact, some neuroscientists doubt whether we will ever fully understand how the reading brain truly works. Nonetheless, we have a lot of information to date, certainly enough to recognize what educational strategies are more likely to help children learn to read as well as what problems can arise during that process. We will discuss these topics in the following chapters.
Reading Is Universal. Surprisingly, images of an adult reading brain are substantially the same regardless of the native language or culture of the individual in the scanning device. Even for Chinese and Japanese readers who use a pictorial script, or for those who read from right to left, the location of the VWFA in the left hemisphere is virtually identical. It is genuinely remarkable to realize that despite the vast differences among the nearly 7,000 languages on this planet, and the many forms in which those languages are expressed in writing, we all still summon the same brain areas to help us identify and comprehend the written words of our language.
How the Brain Learns to Read Page 8