BOOK EXCERPT
Developing Ocular Motor and Visual Perceptual Skills: An Activity Workbook
Kenneth Lane OD, FCOVD

Chapter 3
Eye Movements and Reading

"The control of eye movements during reading can be considered to involve temporal and spatial decisions" (Rayner, 1983)

Introduction

This chapter is one of the most important chapters in this book. You must have a clear understanding on how ocular motor exercises help a child with reading difficulties. This requires a good knowledge of eye movements and reading.

We have already discussed the complexity of the human brain. It is staggering to think that the cerebral cortex alone has about 10 billion neurons and 1 million billion connections or synapses. Counting one synapse per second, we would finish counting 32 million years from now. If we consider the number of ways in which circuits or loops of connections would be stimulated, we would be dealing with a 10 followed by at least a million zeros. There are 10 followed by 79 zeros, give or take a few, particles in the known universe (Edelman, 1998). Considering how complex the brain is, one of its most complex assignments is reading and one of the most important parts of reading is moving the eyes across the page of print in perfect harmony to be able to encode print. Encoding enables the brain to form a visual or nonvisual code of the word and place it in working memory (Krueger, 1993). In decoding, the letter components are accessed and compared against target letters in memory and the word is remembered.

It is important to remember that the brain can't handle all the visual information available to it. Three-quarters of the visual information available to the brain when we read is ignored (Smith, 1994). When we read, we don't take in large amounts of visual information. The eyes move across the page in a series of quick movements called saccades and pause to take in visual information called fixations (Figure 3-1) (Ygge, 1994). Because only a small fraction of the retina has heightened acuity, the point of fixation must constantly be moved about to allow detailed visual processing. Moreover, this must be done so that each change of gaze places the eye at convenient locations on the target (words). For example, in reading, the eyes must accurately jump from word to word so that when each item is fixated, it can be processed rapidly and the next eye movement can be planned. This requires that each individual eye movement be guided by detailed visual information obtained from locations peripheral to the current point of fixation (Moore, 1999). As I mentioned before, we do not take in a large amount of words as we move our eyes across the line of print (Figures 3-1 and 3-2). The average good reader does not look at more than one word per fixation until he is in the tenth grade (Morris, 1973). Even college students don't take in more than 1.11 words per fixation. A child in first grade only takes in about 45% of a word (Vogel, 1995). The amount of visual information available to the brain during the fixation is called the perceptual span or the span of recognition. The perceptual span is the region around the center of vision within which some aspect of visual detail of interest is used in reading (Rayner, 1983). The perceptual span for letter information in words lies asymmetric with respect to the fixation point and extends farther to the right than the left. For Israeli readers, it is the opposite (Rayner, 1983). A skilled reader has an average perceptual span of four characters to the left of fixation and 15 characters to the right (Solan, 2001). Word recognition and processing for comprehension occurs only seven to eight characters to the right of fixation. Information from eight to 15 characters to the right of fixation helps to direct subsequent saccades (Vogel, 1995). While the central foveal areas are processing information for word recognition, parafoveal (peripheral) retinal areas and their corresponding cerebral centers are analyzing word shape and length information to help direct subsequent saccades (Vogel, 1995). Regressive eye movements during reading occur 15% of the time. These are eye movements in the opposite direction (to the left). Most are only a few characters and typically reflect some text confusion or comprehension problem, or perhaps a “recheck” or “double check” confirmation. Children learning to read and poor readers make excessive numbers of regressions (Figure 3-3). Normally, approximately 10% to 15% of all saccades (or fixations) are actually regressive in nature. Uncommon words are refixated more than common words (Ciuffreda, 1995). The average first grader makes 52 regressions per 100 words, while the average college student only makes 15 (Vogel, 1995).

When we pause to take in information as we read, this is called a fixation. The average child in first grade pauses 224 times per 100 words, while the average college student pauses 90 times per 100 words. The average first grader has an average fixation time of .33 seconds, while the average college student pauses for .24 seconds (Vogel, 1995). Because of this, the limit that most people can read is 250 words per minute or about four words per second (Smith 1994). In normal reading, fixations average about 250 msec. During this short interval, visual information is extracted from the printed material (Lovegrove, 1990). Fixation durations are influenced by properties of the text such as word length (Hoffman, 1995). Other factors that can influence longer fixation durations include: low-frequency words, technical words, shorter words, certain grammatical elements, words at the beginning of a new line, words that are misspelled, and regions of the text with important information. Shorter durations of fixation are often caused by the final word of a line, fixations before a regression, and in regions between sentences (Garzia, 1994). The brain is very busy during the fixation. Not only must it encode information but it must decide when to move the eyes and how far to move them for the next fixation. To give you an idea of what happens during the fixation, let's suppose the fixation lasts for 250 msec. We know that much of the visual information necessary for reading can be acquired beginning at about 50 msec into the fixation (Vogel, 1995), leaving the remainder of the fixation (200 msec) to complete programming the next eye movement and for higher level linguistic processing (Rayner, 1983). Processing of information available during the fixation is not completed by the end of the fixation and the onset of the next fixation is not triggered by the completion of processing of information (Rayner, 1983). In other words, the visual processing might be completed but the brain is still digesting information from one fixation to the next fixation. Language aspects of the text must begin having their influence on processing within about 100 msec after the onset of the fixation. This is called textual influence threshold. By about 100 msec, the brain has an idea when the next saccade will start, and after 100 msec stimulus changes will not affect when the next saccade will start. This is called the saccade deadline. The time when the brain centers have become fully committed to the time of the next saccade is called the point of no return, and is estimated to occur at about 30 msec prior to the saccade onset (Rayner, 1983). Contrary to what some people think, all the visual information available to the foveal area is scanned and processed. Good readers do not skip over parts of words when they read. If one letter in a child's foveal vision is masked, his reading speed decreases by 50% (Stanovich, 1993). Children must be taught to scan each letter in a word and every word when they read.

We know that during the fixation the time of the next saccade is determined, but how does the brain know how far to move the eyes for the next fixation? The brain uses a combination of peripheral visual information and knowledge of language patterns to know how far to move the eyes (Rayner, 1983). Visual information such as word length patterns is acquired at least 12 to 15 character positions to the right of the fixation point and specific letter and word shape information no further than 10 character positions to the right of the fixation point (Evans, 1990). There is an optimal landing position for the eyes within a word (Fischer, 1993). The most likely landing spot is near the center of a word. In fact, the location tends to be between the middle and beginning of the word. The preferred viewing location in a five letter word is the second letter and for a ten letter word, it is the fourth (Rayner, 1983). The fixation is less likely to be of the word “the” or on a blank area (Garzia, 1994). If the eyes miss the optimal landing location in the word, the penalty is in the order of 20 msec slower reading speed for every letter that the reader is away from his optimal location (Richman, 1992). Accuracy in saccadic eye movement is obviously a very important component in reading.

Laboratory studies have shown that after the presentation of a visual stimulus, the stimulus continues to be “seen” for some time. The visual response to a stimulus outlasts the actual duration of the stimulus. This continuation of a response after the removal of a stimulus is known as visible persistence and can last up to 300 msec. (Lovegrove, 1990). As you can imagine, if this happened when we read, it would cause considerable problems. The brain cannot allow the visual image of one fixation to continue into the next. If this occurred, the two individual visual inputs may be seen but we would not know which was from the first fixation and which was from the second fixation (Lovegrove, 1990). This, of course, does not happen with most readers. However, what does the brain do to overcome this? The answer is that there are two parallel visual subsystems that operate from the retina to the visual cortex. One is called the transient (M cells) and the other is called the sustained (P cells) system.

Retinal images are sampled twice by the visual system (Solan, 1994). It is sampled first by our peripheral vision to get the gross overall view of an object or upcoming word in a sentence. It is then sampled again to extract detailed information from the object or word. The two systems originate in the retinal ganglion cells (Solan, 1994). The system that is involved in the detailed analysis is called the sustained system (P cells) and is also called the parvocellular system. This is active during the fixation. The sustained system's role is an identification of shape, patterns and the resolution of fine detail. P cells comprise 80% of the retinal ganglion cells and are concentrated at the fovea. They have small receptive fields and are more responsive to low temporal (slow movement) and high spatial frequency (close together or detailed information) (Solan, 1994). Visual acuity and color vision are principally sustained system's functions (Garzia, 1990).

The visual system that operates with the sustained system is called the transient system (M cells). It is also called the magnocellular system. M cells comprise 10% of the retinal ganglion cells, are distributed evenly across the retina (Solan, 1994), and have large receptive fields (Bassi, 1990). The transient system is thought to be involved in the perception of motion, depth, brightness discrimination, the control of eye movements, and the localization of targets in space. It seems to function to accomplish a quick global analysis of a visual scene. It performs a global analysis of the incoming stimulus, breaking the field into units and regions and coding the position and movement of objects in space (Williams, 1990). Two of the primary functions of the transient system is that it carries motion detection and ocular motor control information (Garzia, 1990).

The role of the transient system in reading is critical. Previously, I have mentioned that the sustained system during the fixation is processing detailed information about the text. It is during the fixation that word encoding and decoding occur. The words are identified and the visual information is used for comprehension. When we pause to fixate during reading, both visual and cognitive information is used. Cognitive information is held in our subconscious to help in word identification and comprehension. When we move our eyes from one fixation to another by saccadic eye movement, the cognitive information continues; however, the visual information is terminated. If it isn't terminated at the end of the fixation, you will have visual information from one fixation smearing into the next fixation like an after image, making the text appear superimposed or overlapping (Lovegrove, 1990) (Figure 3-4). This is called visible persistence and can greatly interfere with reading. The reason why this does not happen is that the stimulation of the transient system generated by the start of the saccade inhibits (suppresses) the visible persistence of the sustained system from the pervious fixation (Solan, 1998). A deficit that affects the timing of either system will interfere with the processing of the second fixation and could lead to superimposition of successive inputs (Solan, 1994). This is what often happens with children with reading disabilities. In fact, over 75% of children with a reading disability manifest a transient defect (Lovegrove, 1990). Since inputs from successive fixations would be superimposed, disabled readers would do better in reading that does not require eye movements.

Transient deficits can cause the following problems:

1. Readers may only see parts of words.
2. If they do not know which fixation the information came from, they would know very little about the spatial arrangement of the letters and this could lead to reading errors and word or letter reversals.
3. It would be very difficult to learn any systematic grapheme to phoneme (phonics) rule if the appearance of the graphemes was in some way unstable.
4. The disabled reader may make repeated errors in different readings of the same word (Lovegrove, 1990).
5. The disabled reader may experience perceptual grouping deficits.
6. Readers may suffer an inability to selectively attend.
7. Readers may require larger time intervals to make accurate temporal judgments.
8. They may require more time to alleviate attention across visual space without eye movements.
9. Readers may skip lines during reading.
10. Readers may have to use a finger to help keep place during reading.
11. They may complain of words appearing to move on the page (Garzia, 1993).

It is important that we understand the relationship between the sustained and transient systems. It gives us a much better understanding as to why we do certain activities. This will become clearer when you follow the pathways of the two systems in the brain. For this description, I will use P cells to describe the sustained system and M cells to describe the transient system. These two parallel processing pathways in the visual system are relatively independent. These pathways transmit information from the retina to the dorsal lateral geniculate nucleus to the visual cortex and then to higher cortical levels with little or no cross-talk between them (Steinman, 1996). Although the principle function of the lateral geniculate nucleus is to relay ganglion cell information to the visual cortex, less than 20% of the synaptic input to the lateral geniculate nucleus is retinal in origin. The majority of the afferent neurons is extraretinal, midbrain, and brain stem; therefore, M and P ganglion cell information is influenced by nonvisual inputs (Solan, 1994). This is another reason why we do a lot of nonvisual activities that involve lower brain areas, such as motor activities.

Information travels faster along large, highly myelinated axons, causing information from the M pathway to reach the visual cortex faster than information from the P pathway (Steinman, 1996). Receiving information first from the M system and then from the P system allows the visual system first to quickly locate objects and then identify them (Steinman, 1996).

It seems obvious from the previous discussions that the proper functioning of the transient system is critical for normal reading skills. It also seems obvious that anything that can improve the transient system would be beneficial. Research has shown that the transient system is sensitive to short wavelengths (e. g., blue) (Solan, 1994) and may perform more efficiently when stimulated by this wavelength. One research paper showed that blue overlays significantly improved reading comprehension in 70% of children identified as reading disabled (Solan, 1998). Positive results were obtained by gray overlays but they were not as successful as the blue overlays (Solan, 1998). It has also been shown that while blue enhances transient function, red reduces it (Richman, 1992). Does this mean that the Scotopic Sensitivity Syndrome exists? Also called the Irlen Syndrome, this claims that some children have difficulty processing full-spectrum light efficiency and that certain colored tints may improve their reading efficiency. This syndrome consists of eyestrain, photophobia (preference to reading in dim light), problems in visual resolution (blurred print, unstable text), restricted span of focus (only small areas of print seen), difficulties with sustained focusing (print blurs unless the reader puts a lot of effort into keeping it clear), problems in depth perception (difficulty judging distances), and handwriting (Garzia, 1990). The problem is there is no scientific evidence of a Scotopic Sensitivity Syndrome. All of the symptoms of this syndrome listed above are usually caused by visual problems such as ocular motor, convergence, or accommodation. What is important to know is that there does seem to be improvement in some children with a blue tint, but not necessarily other tints.

This chapter was devoted to eye movements and reading. It is extremely important that you have an understanding of eye movements and reading and also the sustained and transient systems. This enables you to understand how some visual procedures improve a child's overall reading efficiency.

Tips for a Successful Activities Program

There are no exercises that have been proven to enhance the transient system. Because of this, I recommend that we do exercises that affect the functions of the transient system and hope by doing this that the transient system is improved. Therefore, I recommend the following:

1. Do activities that work with the peripheral visual system.
2. Do a lot of visual scanning activities.
3. Do figure-ground activities; for example, hidden pictures, etc.
4. Try a light blue filter and see if this helps the child's reading performance. This can be obtained from an office supply store, usually as a page separator or file cover.
5. Do a lot of ocular motor activities.

References

Bassi, C. J. (1990). Clinical implications of parallel visual pathways. Journal of the American Optometric Association, 61(2), 98-109.

Ciuffreda, K. J. (1995). Eye movement basics for the clinician. New York: C.V. Mosby.

Edelman, G. M. (1998). Building a picture of the brain. Daedalas, 127(2), 37-69.

Evans, B. J. W. (1990). Review of ophthalmic factors in dyslexia. Ophthal Physiol Opt, 10, 123-132.

Fischer, B. (1993). Saccadic eye movements of dyslexic adult subjects. Neuropsychologia, 31(9), 887-906.

Garzia, R. P. (1994). Vision and reading II: Eye movements. Journal of Optometric Vision Development, 25, 4-26.

Garzia, R. P. (1993). Vision and reading I: Neuroanatomy and electrophysiology. Journal of Vision Development, 24, 4-51.

Garzia, R. P. (1990). Visual function and reading disability: An optometric viewpoint. Journal of the American Optometric Association, 61(2), 88-97.

Hoffman, J. E. (1995). The role of attention in saccadic eye movements. Perception and Psychophysics, 5, 787-795.

Krueger, L. E. (1993). Detection of letter repetition in words and nonwords: The effect of prior knowledge of repetition location. American Journal of Psychology, 106(1), 81.

Lovegrove, W. J. (1990). Experimental evidence for a transient system deficit in specific reading disability. Journal of the Optometric Association, 61(2), 137-146.

Moore, T. (1999). Shape representations and visual guidance of saccadic eye movements. Science, 285, 5435.

Morris, H. F. (1973). Manual for the edl/biometrics reading eye II. New York: McGraw-Hill, Inc.

Rayner, K. (1983). Eye movements in reading. New York: Academic Press.

Richman, J. (1992). Annual review of the literature 1991. Journal of Optometric Vision Development, 23, 3-37.

Smith, F. (1994). Understanding reading. Mahwah, NJ: Lawrence Erlbaum Associates.

Solan, H. A. (2001). Role of visual attention in cognitive control of oculomotor readiness in students with reading disabilities. Journal of Learning Disabilities, 34(2), 107.

Solan, H. A. (1998). Eye movement efficiency in normal and reading disabled elementary school children: Effects of varying luminance and wavelength. Journal of the American Optometric Association, 69(7), 455-464.

Solan, H. A. (1994). Transient and sustained processing. Journal of Behavioral Optometry, 5(6), 149-154.

Stanovich, K. E. (1993). Understanding and teaching reading: An interactive model. American Journal of Psychology, 106(3), 456.

Steinman, B. A. (1996). Vision and reading III: Visual attention. Journal of Optometric Vision Development, 27, 4-28.

Vogel, G. L. (1995). Saccadic eye movements: Theory testing and therapy. Journal of Behavioral Optometry, 6(1), 3-12.

Williams, M. C. (1990). Perceptual consequences of a temporal processing deficit in reading disabled children. Journal of the American Optometric Association, 61(2), 111-121.

Ygge, J. (1994). Eye movements in reading. Oxford, England: Elsevier Science, Inc.