Chapter 2: Developing Reading Recognition

Thank you for visiting Learning to Read in the Computer Age. The book will be taken down on December 31, 2014. We recommend that you now use the 2014 multimedia book Universal Design for Learning: Theory and Practice, which can be accessed at no cost.

Back to Chapter 1 Forward to Chapter 3

Skilled readers recognize words and other markers in text automatically. They may slow down or pause to problem-solve when they come across new vocabulary, a novel sentence construction, or unfamiliar content, but generally their recognition of the symbols, sounds, and patterns in a page of text requires virtually no conscious effort. Good readers have a prodigious capacity to recognize the letter a or the word prodigious instantaneously in different type sizes and fonts, and in different linguistic contexts, text structures, and media. Once we have learned to read fluently, we imagine that recognition is simple because it comes so easily. But our ability to identify text elements quickly and accurately under many different conditions is a remarkable achievement that baffles the most powerful computers. IBM's Deep Blue can play chess as well as the world champion, but no computer can begin to match the recognition skills of even a novice reader.

Recognition and the Brain

These feats of recognition are accomplished mainly by the "recognition systems" in the rear of our cortex. Neurological and cognitive research has produced some findings about these systems that inform our understanding of how reading should be taught.

One is that the recognition process does not occur in a unique place. No single neuron or cluster of brain cells has the job of recognizing, for instance, the letter A. PET scan studies show that the processing necessary to recognize an A is distributed—color is processed in one place, shape in another, orientation and location in still others. Thanks to this distributed processing, recognition systems operate like an efficient committee. Different brain areas take on small parts of a task and carry them out simultaneously. This is the brain's version of what computer scientists call "parallel processing," and its great advantage is speed. Having a number of systems perform sub-tasks at the same time (in parallel) is faster than having one system do them one after the other (serially), especially when the work to be done is complex. At the same time, distributed processing underlies some of the differences in how students learn recognition. Learning is never just one process, and teaching should not be either. Individuals' varying abilities to identify color, shape, or location means they face different recognition problems and will respond differently to particular learning materials and teaching techniques.

Perhaps even more important to teaching this skill is the fact that recognition is rooted in classification. Cognitive scientists have shown that our ability to recognize the letter H depends on understanding the general idea of H, the essential elements that define a category we might call "H-ness." Color, size, and whether the symbol is ink on paper, carved wood, or molded plastic are irrelevant. What matters—what makes an uppercase H an uppercase H —are the two generally parallel lines connected by a crossbar somewhere near the middle. The following symbols are somewhat different from each other—

—but our understanding of H-ness lets us identify them all as uppercase Hs, and tells us that these similar symbols —

—are not Hs. They fit another of our letter recognition categories, the one that defines uppercase A.

Once thoroughly mastered, these categories make future recognition flexible and fast, but learning them is a demanding task, and automatic, accurate recognition requires considerable practice. PET scans show dramatic differences between a brain learning to recognize a category and one that has learned it, supporting the distinction between developing a skill and using it. Recognizing that a particular stimulus fits into an established category is mainly accomplished by the recognition systems; learning a new recognition category is a whole-brain activity, often involving the strategic and affective systems along with the recognition systems.

Teaching Recognition

Authors' Note

Traditional approaches to teaching reading would suggest that this chapter should be about teaching word analysis and the next chapter about comprehension. But both neurological research on learning and growing recognition of the importance of balanced, holistic reading instruction suggest that analysis—or recognition —and comprehension cannot be separated. Recognition skills are required at all levels of reading from small patterns (such as a letter) to larger patterns (such as an author's style). Similarly, strategic skills are needed to decode words as well as to make meaning from text. Teaching recognition and teaching strategy development do call for different methods, however. Our separate chapters on recognition and strategy speak to these differences in instructional approach.


The most effective teaching of recognition skills includes a repertoire of strategies that engage all three brain systems and take individual differences into account. Because all three systems work closely together, especially during learning, skilled teachers address them all at once. For the sake of clarity, though, we will discuss these systems separately, while keeping in mind that good teaching and good software always weave them together. Involving the three systems means:

  • highlighting patterns to be learned (recognition systems—learning what);
  • providing opportunities to act on and manipulate patterns (strategic systems—learning how); and
  • engaging learners in activities that support pattern learning (affective systems—attention and motivation).

Highlighting Patterns

Children viewing letters, words, and sentences often need help recognizing repeating patterns. Many teaching techniques help children identify patterns by making them more noticeable. Printed instructional material uses color, style, size, or font to make target patterns stand out visually from surrounding letters or words. This highlighting technique can emphasize a pattern as small as a single letter or as large as a group of sentences. In addition to locating patterns, it can show relationships between text elements. For example, the relationship between the vowel and the final e in the silent e pattern can be illustrated through typestyle or color as in: fate, same, pale.

Teachers call attention to written patterns by pointing at them (whether or not they are already emphasized in the printed text), writing them on the board, and underlining or circling them. Similarly, they emphasize the patterns of spoken language by changing pitch, loudness, or rate of speech as they talk or read aloud. Both teachers and printed material can also highlight patterns by isolating, grouping, and explicitly categorizing words or letter groups that share them. For example, a teacher may select words that end in at and highlight the features that they have in common by presenting them as a group.

Taken together, these techniques accommodate students' learning differences. Students not particularly sensitive to differences in color or shape may grasp patterns that are isolated and grouped; students who have trouble with written symbols may learn more readily from hearing patterns emphasized in speech. Good teachers use multiple techniques, though in practice almost any group lesson is likely to suit some students better than others.

Manipulating Patterns

Seeing or hearing highlighted patterns supports the brain's recognition systems. Manipulating patterns—building them, changing them, playing with them— brings the strategic systems into action. Collaboration between these systems is critical to learning. The skill-building activities that engage the strategic systems provide essential support to the recognition systems, developing the solid, sure knowledge of patterns that eventually makes recognition automatic.

To involve students in pattern manipulation, teachers commonly model activities such as isolating, grouping, and categorizing patterns, and then ask students to follow suit. To teach the silent e pattern, for example, teachers often show the contrasting sound of the vowel with and without the final e by grouping words: hat, at, and slat versus hate, ate, and slate. They then ask students to read a new set of words aloud with and without the final e.


Mastering pattern recognition requires persistent practice, especially by students who do not pick up patterns readily. Appealing, meaningful activities support students' affective brain systems, and motivate them to engage in the necessary practice. Some teachers draw patterns from students' own reading materials or written work, making practice more personal and meaningful (Routman, 1991). Pattern recognition activities can also be embedded in challenging, structured, appealing games. For example, students can play a modified game of bingo using words that demonstrate the pattern being learned (Bos & Vaughn, 1994).

Using Computers to Support Recognition


Even when teachers do their best to use a balanced "whole-brain" approach to teaching recognition, too many children fail. The main reasons are not hard to determine. Printed teaching materials come in only one format, meeting the needs of some students but not others. There is rarely enough time to give individual students enough personal attention and opportunity for supported practice. Teachers are torn between helping those having the most trouble and holding the interest of those ready to move on. Students who don't "get it" right away often face social difficulties that make it harder to learn.

To address these problems successfully, reading software must reflect the same kind of whole-brain, individualized approach that good teachers use. In other words, it must:

  • highlight patterns;
  • provide opportunities for meaningful practice;
  • motivate students to learn and practice; and
  • address individual differences.

Computers can supplement a teacher's efforts to meet the needs of particular students by providing support that is tailored to individual differences. The flexibility we have identified as the essential quality of computers and the guiding principle of universal design is key. Computers can highlight patterns in an almost unlimited number of ways: what is being emphasized can be changed, and so can how the emphasis is applied (for example, through color, sound, size, animation, grouping, or repetition). The flexible responsiveness of computers makes them excellent tools for skill-building practice, since they can simultaneously reflect, guide, and support students' choices. Teachers can customize activities for individual students. For instance, all of the students in a class might be given different levels of the same phonics game, some working with ten letter-sound combinations while others work with only one. Everyone in the class can use the same material, but each student works at an individually appropriate level. Finally, computers can stimulate engagement through interesting games, through practice in authentic contexts (such as text from literature and the World Wide Web), and simply through giving students the power to influence what happens on the screen. As the success of many computer games shows, having the ability to control an electronic environment seems innately satisfying and fascinating.

No existing programs realize the full potential of the technology to teach recognition. None is as flexible, complete, and engaging as future software is likely to be, and some programs solve particular pedagogical problems more successfully than others. But several offer substantial help to reading teachers. Keeping our essential criteria in mind, we will look at some specific applications to see what they do well and to identify opportunities for improvement. We will first describe programs that reinforce sounds, letters, and letter-sound correspondences, and then consider a few that can address the recognition of larger textual patterns.

Teaching Sounds, Letters, and Letter-Sound Correspondences

Research shows that sounds (phonemes), letters (graphemes), and letter-sound (phoneme-grapheme) correspondences are the three primary language patterns fundamental to developing word recognition skills (Adams, 1990; Anderson et al., 1985). Most reading software applications are designed to teach children to recognize these three kinds of patterns.

Sound Patterns: Teaching Phonemic Awareness

Phonemic awareness—the ability to recognize the underlying separate sound elements of spoken language—is an essential component of learning to read and has been the focus of much research (Adams, 1990; Chall, 1967; Richek et al., 1989; Snow, Burns, & Griffin, 1998). Reading curricula today devote considerable attention to building phonemic awareness.

In classrooms and tutorial settings, phonemic awareness instruction depends largely on interaction with a teacher. Teachers emphasize phonemes and segment words orally, helping students to hear the component sounds, blend these sounds, and ask the students to repeat or experiment with sounds. Printed materials such as textbooks, workbooks, or activity sheets provide limited support, since they can only represent sounds with letters, not produce them orally. Media such as audio and video cassettes can emphasize sound segments, but their instructional value is limited because children can only listen passively to recorded examples.

Computer software, on the other hand, is both alive with sound and interactive. Taking advantage of the sound capacity of CD-ROMs, many programs use songs, poems, and other audible exercises to emphasize the sound patterns of our language. For instance, Bailey's Book House includes an activity called "Read-A-Rhyme" (see Figure 2-1) in which the student can complete a nursery rhyme by selecting one of four word choices that sound alike. The program then speaks the rhyme aloud using that word and shows a matching animation. The animation links meaning to the sound and provides an entertaining and relevant "reward." Children can repeat the activity as many times as they want.

FIGURE 2-1. Bailey's Book House (Edmark)

In addition to being patient, computers have the capacity to exaggerate or alter sounds to make patterns easier to recognize. According to Merzenich et al. (1996) and Tallal et al. (1996), the transitions between sounds are too fast for many children with language-related disabilities. Slowing the sounds down makes them more obvious and recognizable. Their work on altering sounds for students with language-learning impairments has recently been embodied in a program called Fast ForWord. Digitized sounds can be modified in many controlled ways. It is possible, for instance, to slow sounds down without lowering their pitch.

The best early reading software programs let children match, sort, play with, and even change sounds. Active manipulation, rather than mere listening, engages strategic systems along with recognition systems, and helps students develop essential skills. One activity in Curious George Learns Phonics (see Figure 2-2) uses a matching game to highlight sound patterns and build phonemic awareness. A word, such as bib, is voiced and shown as a picture. The student is asked to select a match for the first sound of the word from among other sounds hidden behind randomly opening shutters and doors of a house. The student is then asked to match middle sounds and final sounds. By drawing attention to each of the sounds and making recognition of them important for success, the game emphasizes that the pattern of sounds in a word is actually composed of several segments.

Two screenshots from the software Curious George Learns Phonics
Two screenshots from the software Curious George Learns Phonics

FIGURE 2-2. Curious George Learns Phonics (Houghton Mifflin Interactive)

The home version of Curious George Learns Phonics pairs the letter with the sound, but the soon-to-be-released school version first presents the sound without the letterform. It will be among the few available programs that support phonemic awareness by presenting sounds without the visual cues of letters. While some students clearly develop phonemic awareness by working with letters and sounds together (Shefelbine, 1995), some experts suggest that very beginning readers benefit most from work with sounds alone (Grossen, 1997; Yopp, 1992).

Visual Patterns: Teaching Letter Recognition

Another essential component of reading proficiency is alphabetic understanding, or knowing that words are composed of letters (Adams, 1990; Anderson et al., 1985; Morrow, 1993). The capacity of the computer to represent letters in many ways makes it an excellent tool for supporting visual pattern recognition systems in students with diverse learning styles. Graphical manipulations of color, shape, size, and proportion emphasize the essentials of a letter's shape and help learners grasp the features that define A or H or K as categories. Presenting letter shapes with letter names and pictures of mnemonic cues enriches children's associations while emphasizing these key features.

In the example shown in Figure 2-3 , the child can "morph" any letter into an animal or object that begins with that letter. After seeing an animation and hearing the pictured word pronounced, the child can use the rainbow slider tool to step through the morph frame by frame.

Two screenshots from the software Reading Recipes/Grandma Ollie's Morphabet Soup
Two screenshots from the software Reading Recipes/Grandma Ollie's Morphabet Soup

FIGURE 2-3. Reading Recipes/Grandma Ollie's Morphabet Soup (Corbitt Design)

While television or video can also present rich animations, computers allow children to control what happens when. This element of exploratory play is far more engaging than passively watching someone else's sequence. Classic letter recognition programs such as Muppets on Stage, Stickybear Early Learning Activities, and A to Zap! give children feedback that reinforces learning when they select letters. The Letter Machine, a letter recognition activity in Bailey's Book House (see Figure 2-4), responds to student letter choices by speaking the letter's name, showing uppercase and lowercase letters, and presenting an alliterative sentence using the letter, and an animation of the sentence. This activity highlights the pattern and supports engagement by providing playful outcomes to actions.

Screen shot from the software Bailey's Bookhouse

FIGURE 2-4. Bailey's Book House (Edmark)

Reader Rabbit 1 (see Figure 2-5) is one of many programs that use customizable matching and sorting games to help develop automatic letter recognition. In the activity shown, for instance, children can sort words by their first letter, middle letter, or final letter, practicing the specific skills they most need.

FIGURE 2-5. Reader Rabbit 1 (The Learning Company)

The abundance of structured letter recognition programs should not stop teachers from developing their own activities using application software. "Homegrown" products are highly motivating, awakening children's creativity and pride of ownership. One easy and compelling possibility is to help students compose multimedia alphabet books. Using a program such as HyperStudio or ClarisWorks, teachers can create a simple template and model the construction of pages that show a letter, a picture, and a word. Students can produce their books without using the keyboard at all, drawing their letters and pictures with the mouse instead. They can also record themselves saying the letter, naming the picture, and reading the word. (HyperStudio lets students record sound directly in the program.) These electronic books can be shared on screen as personal slide shows, or printed in color or black and white. Printed books can also be colored by hand, laminated, and shared in a class library.

Any word processing program can be used to highlight individual letters in the context of connected text. Teachers can select targeted letters and emphasize them by changing font, size, typestyle, and color. They can ask students to search the text for other examples of target letters and similarly modify their appearance, an exercise that reinforces letter recognition skills through active practice.

Phonics: Teaching Letter-Sound Pattern Recognition

In addition to recognizing individual sound segments and individual letters, children must learn the patterns of relationship between sounds and letters: in other words, they must learn phonics (Adams, 1990; Chall, 1967, 1996; Morrow, 1993). While English is more phonically regular than it seems, its patterns of letter-sound correspondence are numerous, complex, and hard to learn. Teachers often do not have time to reach the children who have the most difficulty mastering this skill.

Like letter recognition, phonics has received substantial attention from the software design community. The best phonics activities highlight key features of letter-sound correspondences and provide opportunities for structured practice and exploration in an electronic environment.

For example, in one Kid Phonics 2 activity, students can click through the alphabetical index of more than 200 items stocked by the general store (shown in Figure 2-6), selecting anything from an antelope to a zucchini.

FIGURE 2-6. Kid Phonics 2 (Davidson & Associates). Reproduced courtesy of Davidson & Associates, Inc. ©1996.

After they choose, the program takes students to the Word Builder Ranch, where they build the word for the item by picking sound cards in the right order (see Figure 2-7). The same letter or letter combination can represent different sounds depending on context, so students must click on a card to hear the sound before selecting it.

FIGURE 2-7. Kid Phonics 2 (Davidson & Associates).
Reproduced courtesy of Davidson & Associates, Inc. ©1996.

Every time a new sound is added, the computer highlights and speaks each phoneme of the growing word. When the word is finished, the phonemes are voiced separately and slowly and then blended together both visually and aurally. The program encourages students to construct a sentence around their word and then turns their sentences into headlines for the Wild West Times, giving the activity a contextualized goal.

Cautions and Suggestions Regarding Phonics

Using computers to support phonics learning is challenging. The central problem is that patterns of correspondence between letters and sounds are not one-to-one, as are those between letters and their names, but many-to-many. The letter a can represent a variety of sounds, depending upon context (h_t versus h_te, for instance). And there are multiple ways to spell the vowel sound of raid, rate, etc. Teaching phonics is ultimately about teaching the conditional probabilities of real language, in context. It is not about the sound of a but which sound of a, when.

This central problem leads to others. Given the complexity and variety of letter-sound correspondences, effective phonics instruction must be systematic (Shefelbine, 1995) and at least potentially comprehensive. No existing software program offers comprehensive and systematic coverage and a flexible approach to teaching. Integrated learning systems such as Waterford Early Reading Program from Electronic Education follow a thorough scope and sequence, addressing every sound-symbol correspondence in a prescribed order. Such programs are expensive and lack the flexibility needed to accommodate varied teaching styles and individual learner needs. Some CD-ROM-based programs offer a flexible sampler of well-designed, engaging phonics and word recognition activities, but since they are not comprehensive, they cannot address the contextual problem of sound-symbol correspondence. Other CD-ROMs are comprehensive but provide only a limited set of activities.

Add to these difficulties arguments among researchers, teachers, administrators and even politicians about how, why, when, and whether to teach phonics, (Chall, 1967, 1996; Lemann, 1997; Shefelbine, 1995; Waldman, 1995) and it is easy to understand why software developers might skirt certain instructional issues to avoid controversy. They have not consistently or convincingly answered questions such as these: Do you provide vowel pronunciations in isolation, or only in the context of words or syllables? If representing an "open" syllable that is going to be closed when combined with the next syllable, do you speak the vowel as long or as short?

We believe that the best way to minimize these shortcomings is to take time to assemble a mix of software resources to suit the instructional approach of a classroom or school. The software chosen should support a balance of extensive exposure to meaningful connected text and attention to the relationship between the sounds in speech and the patterns of spelling (Adams, 1990; Shefelbine, 1995). If an integrated learning system is already in place, teachers should evaluate its appropriateness for them and their students. They might use it selectively to reinforce particular sound-symbol correspondences, or extensively with students who need systematic, step-by-step instruction.

We recommend choosing CD-ROMs with flexible designs that can accommodate teacher preferences and varied student needs. Some programs, for example, are "half full" of content, enabling teachers to add their own material. Similarly, word processors that talk (such as ClarisWorks, Write:OutLoud, or ULTimate Reader), and software with sound recording capabilities (such as Microsoft® Word or HyperStudio) can serve as tools for designing activities using students' own words from writing or reading, and for developing activities around particular phonic patterns. (Authors' note: CAST developed ULTimate Reader for Universal Learning Technology.) These more "personal" materials can be especially engaging and appropriate. Combining resources and using different programs and activities at different times to support students' various developmental needs is the most effective approach to using technology.

Word Recognition

Most software programs support whole word learning and word part learning in similar ways. Many include word parts and whole words in the same activities. Since most of the techniques used to teach these related skills are the same, a separate discussion of using computers to teach whole word recognition is not necessary.

It is important, though, to consider how the designers of word recognition software select words for study. Most systematic programs and many good "sampler" programs choose high-frequency words or members of common word families. The best programs include word lists in the teacher's manual and explain the rationale behind their selection. These lists and explanations help teachers integrate the programs with classroom reading activities, especially if the software management systems let teachers select particular words from the list for individual students or add their own words. Programs that have no particular rationale for choosing words are less useful for teaching word analysis. Many "edutainment" programs share this weakness. They study words randomly chosen from a story or theme often unrelated to frequency, regularity, word family, or any other meaningful category.

Recognizing Larger Patterns in Text

Good readers automatically recognize and use textual patterns. Typographic elements such as headings, indentations, typestyle, punctuation, and white space all convey information about the organization, structure, and meaning of a text. Grammar and syntax provide patterns of language structure that communicate meaning to skilled readers. Such readers are alert to the clues to understanding embodied in the structure of sentences, the topic sentences of paragraphs, and the order of paragraphs in a chapter. Good readers recognize style, context, purpose, and difficulty level. They understand how all these pieces fit together; they see the pattern of these patterns. They also recognize what they don't understand and when they must work at making sense of a passage.

Poor readers recognize few of these patterns. Some must allocate so much attention to word analysis that they have little cognitive space left to consider larger textual patterns. Others find these larger patterns just as difficult to recognize as the smaller ones. To make meaning, students with language-based learning difficulties need the same kinds of supports at textual levels as they do when learning sounds, letters, and words.

Remember that learning and knowing are functionally different. The skilled reader's pattern recognition systems work automatically and require little help from other brain systems. But learning to recognize textual patterns involves the whole brain. Effective teaching must address recognition, strategic, and affective systems together. Here too we can judge software on the basis of how well it highlights textual patterns, provides opportunities for practice and exploration, and provides motivational support.

Using Computers to Support Textual Pattern Recognition

Though software and curriculum designers have developed relatively few programs supporting textual pattern recognition, this aspect of reading comprehension has received a certain amount of attention. Printed materials such as elementary school textbooks highlight parts of sentences or paragraphs to emphasize key components and to support comprehension. Of course this kind of typographical emphasis is not flexible and cannot be customized for different students or lessons. Electronic text, on the other hand, can be infinitely modified as learning progresses.

In traditional reading comprehension software programs, graphic cues selectively emphasize elements of text. In the example in Figure 2-8, an activity in Tomorrow's Promise helps a student learn to draw inferences. An incorrect answer to the question makes the relevant part of the text change color. When the correct answer is selected, the answer and highlighted passage text colors match. In essence, the changing format of the text acts as a scaffold to support recognition.

Screenshot from the software Tomorrow's Promise
Screenshot from the software Tomorrow's Promise

FIGURE 2-8. In this activity from Tomorrow's Promise, graphic cues emphasize parts of the text. When the student answers incorrectly, a phrase in the passage changes color to indicate information relevant to the answer. In this example, "[the] sun was barely up" has changed color.

Teachers can use any word processor to emphasize textual patterns. For example, Figure 2-9 shows a passage digitized from a printed edition of Little House on the Prairie. Words that indicate the sequence of events are highlighted for students using boldface typestyle, alerting them to markers that denote sequence. Once the text is captured in electronic form the teacher can easily change what is emphasized depending on the instructional focus or the needs of different students.

FIGURE 2-9. Words that highlight the sequence of events in this passage are highlighted using boldface to emphasize a textual pattern.

While it takes time, creating passages for pattern study has the advantage of serving the specific needs of different students. When text content and textual patterns are germane to a group, learning is likely to be meaningful and engaging. Once text has been scanned and digitized or downloaded from a digital source, it can be used and re-used to make different points for different students. The effort invested in building a library of "emphasized-pattern" texts will pay dividends over time.

Using Computers to Support Practice and Exploration of Textual Patterns

The best examples of traditional comprehension software use color and style to highlight key portions of text, but few programs support active exploration and practice with graphic and syntactic patterns.

Teachers and students can fill the gap to some extent by using text from sources such as the Internet with outlining programs and word processors to practice and explore textual pattern recognition. In the example in Figure 2-10, text from a World Wildlife Fund Web site is used to help students focus on the significance of subheadings. Students act on the text by selecting, copying, and pasting the subheadings from the Web page directly into outlining software or an electronic visual organizer.

FIGURE 2-10. Web pages can provide text for students to explore textual patterns. World Wildlife Fund {}.

©1998 Netscape Communications Corp. Used with permission. All rights reserved. This electronic file or page may not be reprinted or copied without the express written permission of Netscape. "Whales in the Wild" copyright 1994 by WWF—Worldwide Fund for Nature (known in the United States and Canada as World Wildlife Fund), Gland, Switzerland.

By isolating and manipulating the subheadings, students focus on their meaning and significance for the whole passage. Using outlining tools, students can explore relationships between key passage concepts and connect the concepts to their own experience. Software tools can support this exploratory process. For example, Inspiration (see Figure 2-11) allows students to display ideas in a linear outline or in a visual organizer, typed in shapes connected by lines and arrows.

Screenshot of an outline view from the software Inspiration
Outline View
Screenshot of a diagram view from the software Inspiration
Diagram View

FIGURE 2-11. In Inspiration, the student can toggle between outline view and diagram view.
Editing in either view updates the information in both the outline and the diagram formats.

Switching back and forth between these modes can help students with different learning styles grasp patterns that organize text. A similar effect can be produced by using a word processor to create the text outline and a drawing program to create the visual organizer. Students can explore the text further by filling in information under each subheading; they can use what they already know about the topic, make predictions, and gather information from the original text source.

Some Concerns and Expectations

The programs described in this chapter make valuable use of computer capabilities to help students learn recognition skills. They use sound, color, and motion to highlight patterns and make them interesting. To a greater or lesser degree, they allow material to be customized and responsively guide students' efforts. The best of them can simultaneously engage the brain's all-important recognition, strategic, and affective systems.

But even the best have significant limitations. Sampler CD-ROMs like the Reader Rabbit series exploit the computer's potential to build engaging and innovative environments for learning and practice but are seldom comprehensive or systematic. Because these samplers do not present enough patterns or enough patterns in sequence to constitute a systematic instructional program, they are useful mainly as supplements to a thorough classroom instructional program, or for practice at home.

Large, expensive systems published by Jostens Learning, Computer Curriculum Corporation, Lexia Learning Systems, and Electronic Education are comprehensive and provide a sufficient instructional scope and sequence of patterns, but too many of the activities are unimaginative, pedagogically inadequate, and repetitive. For now, technology seems to offer a choice between economy and impact on one hand and comprehensiveness on the other.

An even more important problem is the inadequate flexibility and responsiveness of most existing programs. Although computers have the potential to respond to individual learning differences, very few programs realize this potential to any meaningful degree. None can match a teacher's ability to recognize the needs of each student and respond by shaping the content, difficulty, approach, or feedback accordingly. While some programs track errors and base subsequent activities on a student's performance, many merely make students repeat activities that they have failed. Some programs provide on-screen "tutors" that support children on request (Microsoft® My Personal Tutor, for example), but few allow teachers to choose alternative modes of presentation, emphasis, action, or motivation for students having difficulty. Most existing programs either give students more of the same or offer an extremely narrow range of other options.

As the power and sophistication of computers increase and we continue to learn more about how they should be used in education, we believe that software will transcend these limitations. Let's Go Read! An Island Adventure uses computers' speech recognition capabilities to support student reading in a new way. In one activity (see Figure 2-12), the student flies a simulated plane through clouds with words written on them. Students read the words out loud into a microphone. Through speech recognition, the computer "hears" what the student says and compares it with the target word. When the student reads a word correctly, the corresponding cloud disappears.

FIGURE 2-12. Let's Go Read! An Island Adventure (Edmark)

Other programs will no doubt similarly exploit new capabilities. Most importantly, we hope they become more responsive to individual needs.

As our discussion has made clear, few programs are designed to teach larger textual patterns such as syntax, grammar, or style. The same capacity that multimedia lends to teaching word recognition—emphasizing common patterns, opening those patterns to manipulation by students, and supporting practice—can easily be applied to these textual patterns. The current state of pedagogy rather than the current state of technology may explain the relative paucity of such programs. Because textual patterns are variable and some (like narrative structure or style) are subtle, identifying them may be conceptually difficult. We have few models of classroom materials that teach textual patterns effectively. Perhaps the imbalance also reflects the old assumption that reading is mainly a matter of deciphering words.

We believe that technology can support students in learning these higher level patterns, and that this domain represents an area for considerable growth. Computers open possibilities for malleable "emphasized texts" that could highlight different structural elements and show how they work by showing what happens when they change. Like the flight simulator described in chapter one, they could help students learn by guiding them through various textual "terrains." Students can also create their own emphasized texts to demonstrate their strategic skills.

Making Choices

The same general principles apply to evaluating existing recognition software, devising your own activities, or thinking about what improved software should be like. Look for:

  • programs that simultaneously engage recognition, strategic, and affective systems
  • programs or combinations of programs that provide the best mix of comprehensiveness and flexible, engaging activities
  • programs that are consistent with the pedagogical approaches being used in your classroom
  • programs that highlight patterns in various and flexible ways
  • programs that can be tailored to individual learning styles and actively respond to students' choices
  • programs that keep track of each student's performance, and provide that information to you and the student in a helpful fashion
  • programs that include (or help you develop) activities that engage students' existing interests

Back to Chapter 1 Forward to Chapter 3