Braille Reading:
Effects of Different Hand and Finger Usage
M.J.C. MOMMERS
Dr. Mommers is with the Institute for Educational Sciences, Catholic University, Nijmegen, The Netherlands.
Abstract: Reviews the literature on hand and finger usage in reading braille. Describes an experiment using 25 blind students to test whether they read faster and/or more accurately with: the left index finger rather than the right one; the left middle finger rather than the right one; the index finger rather than the middle finger; and, with spontaneous hand usage (usually both hands), whether the student read faster and/or more accurately than with the index finger of one hand.
Most researchers believe than braille reading is very similar to reading print. But there is one very obvious difference: the way in which the written word is perceived. While sighted persons perceive with two eyes when reading, just as blind persons read mainly with both hands, for sighted persons with normal reading ability, coordination of information received by each eye is no problem, though it has been found that one eye is very often dominant.
Little is known, however, about the analogous process of acquiring and coordinating information when reading braille with both hands. Most braille instructors generally favor a technique in which the pupil begins a line with the left index finger and moves it forward to the middle of the line, reading the right half of the line with the right index finger; the left finger, meanwhile, searches for the beginning of the following line. This method is probably based on practical experience and observation of hand movements in experienced braille readers. But observation of the reading habits of blind children shows that most of them move both index fingers, held close together, forward across the whole line. The hand movements of these children thus differ from what is considered to be the most efficient technique. Furthermore, individual differences exist, even among pupils who have learned braille from the same teacher.
On the basis of these considerations we searched the literature for information about the relative contribution of each hand in reading braille. At the back of our minds was the question, "Does an obvious division of labor exist between both hands or does each hand play an equal role? Is one hand dominant or not?" The results of this literature search are given here, followed by more thorough discussion of Hermelin and O'Connor's experiments (1971), because they based their work more explicitly on a theory. Finally we give the results of our own research, which may be regarded as a replication of Hermelin and O'Connor's work.
Literature on Hand and Finger Usage
The question of the function of separate hands was raised in the early years of empirical research into the psychology of reading braille. Theodor Heller (1904, p. 87) comments:
During the initial teaching of reading it is mainly the right hand which is involved. This hand executes the genuine reading movements while the left hand tries to fixate the lines and to point out to the right hand where to start its movement.... Once the blind person knows the characters sufficiently well, the left hand too participates in reading. This hand, however, reads neither as quickly nor as continuously as the right one; one can often notice that the arm abruptly changes position. Because of the greater difficulties that naturally go with movements of the left hand, it is especially fitting for a slow, analytic way of touching, while the right hand which moves rapidly over the lines gives the blind person global images of the single characters, even if only fleetingly.
Heller's Studien zur Blindenpsychologie was first published in 1895 in the eleventh volume of Philosophischen Studien edited by Wilhelm Wundt. In the introduction Heller paid tribute to the help he had received from the Leipziger Institute, but his views were probably based only on observation. Grasemann (1917) and Bürklen (1917), on the other hand, carried out more systematic research and both came to the conclusion that the left index finger rather than the right index finger should probably be regarded as the better reading finger.
A report from the Educational Publications Bureau of Research and Education of the American Foundation for the Blind suggests, on the basis of data compiled by the Uniform Type Commission on the Mechanics of Reading Raised Type, that the right hand is the better reading hand. Whereas only 19 percent of readers who use only the left index finger for reading were among the best readers, 32 percent of those who read with the right hand were among the best readers (Maxfield, 1928, p. 47).
Maxfield (1928, p. 47) felt that when a child is unable to learn braille with both hands, it is probably best to train the right hand, unless the child is very obviously left-handed.
Villey (1931) pointed out that hand usage in braille reading varies enormously and that it can possibly be traced back to local teaching traditions and to individual differences.
On the basis of an experiment in which a right-handed and a left-handed individual learned how to read braille in similar circumstances, J.M. Smith (1934) maintained that in right-handed people, the left hand functions more as a sensory organ and the right hand as a motor organ. The righthanded person reads faster with the left hand and the lefthanded person reads faster with the right hand.
In her research group Fertsch (1947) had three very obviously left-handed subjects, one of whom read best with the right index finger, one with the left index finger, and the third could read equally well with both index fingers. This result does not confirm Smith's findings.
Fertsch's research had, however, a much broader framework. Her starting point was: If a blind person reads with both hands, which hand plays the most important role? Which of the fingers used contributes most or least to the total process? Is reading speed faster in individuals who assimilate the meaning mainly through the reading finger of the left hand, or is it faster in those who mainly use the right hand, or is the reading speed faster in those who can read more or less equally well with both hands?
Her sample was composed of pupils from the third to the eleventh grade. She used only pupils who had an average to high score in an intelligence test and who, in a test for reading comprehension, also achieved a score that placed them in either the best third or the weakest third. This was done to obtain two groups which obviously differed greatly in reading ability.
She used the same criterion as Grasemann (1917) to group the children on the basis of hand preference in braille reading. If, when reading two comparable tests with the left hand, the pupil's reading time differed by more than 20 percent compared with the time when he read with the right hand, then the pupil was classified as being right-handed, and vice versa. If the time difference was less than 20 percent, then the pupil was put into the group that showed no particular preference for the left or right hand.
In Table 1, taken from Fertsch (1947), results from three other studies are shown beside hers. (It must be pointed out that data from research by Holland, which had another aim, were also analyzed by Fertsch.)
Examining the last two columns, Fertsch noticed that the distribution over the three distinct groups differed considerably in the good and weak readers. Few good readers had a preference for the left hand, while the percentage of weak readers with no hand preference was very low.
On the basis of a thorough analysis of motion pictures she came to the additional conclusion that pupils who had no hand preference in normal reading with two hands, read a greater number of braille dots with independently functioning hands than the other two groups did, and that, on average, their reading speed was faster. The group of good readers read an appreciable number of the braille cells with independently functioning hands-the right hand touching about twice as many cells as the left hand did. Weak readers held right and left fingers close together so that very few letters were read with independently functioning hands.
On the basis of practical experience Stocker (1970) concluded that people who become blind at a later age, on the whole perceive better with one index finger than with the other when they are beginning to learn braille. She maintains that if the individual is left-handed, generally speaking, he perceives better with his right index finger and vice versa (p. 7). Kimura (1973) states in an article about the asymmetry of the human brain, that the motor and sensory pathways of the two hands are almost completely crossed and that each hemisphere functions mainly with the opposite hand (p. 76). But at the same time he points out that no data are available concerning dominance with respect to the perception of point patterns [braille] (p. 78).
In a study of 40 Swedish and Danish adult braille readers, Lothman (1975) established that there is a distinct difference in hand usage between those born blind and those who become blind at a later age. Of those born blind, 16 read with both index fingers, two with the right index finger, and one with the left index finger. Of those who became blind later, 13 read with the left index finger, four with the right index finger, and only two with both index fingers. Those who used the left hand only could give no special reason for it. These data do not allow any conclusion to be reached-except that further research is necessary.
Hermelin and O'Connor's Experiments
The experiments done by Hermelin and O'Connor (1971a, b) on hand usage in braille reading differ from others in the literature. Their aim was not so much to help improve the way braille is taught, but to test hypotheses derived from neurological theory, based on the following considerations: It is generally accepted, that in right-handed people the parieto-occipital area (at the back and rear side of the head) of the left cortical hemisphere is important for the ability to read. It is also well-known that the right hemisphere contributes more to non-verbal, visual, and spatial functions than does the left. Kimura (1966) demonstrated this functional asymmetry. She showed that subjects were more accurate in reporting the number of dots presented tachistoscopically to the left than to the right visual field.
Table 1. Survey of the results of four studies on hand preference in braille reading.
Grasemann Bürklen Holland Fertsch
Total Good Poor
readers readers
Number of readers 31 66 66 63 33 30
No preference 22.5% 23.0% 16.5% 31.5% 48.5% 13.0%
Prefer right hand 29.5 32.0 57.5 40.0 33.3 47.0
Prefer left hand 48.0 45.0 26.0 28.5 18.2 40.0
Conversely, letters were identified better when they were presented to the right visual field, that is, the left hemisphere. Braille reading material consists of dots in specific spatial arrangements. The braille symbols are constructed as spatially arranged items, and spatially arranged, non-verbal items are processed in the right hemisphere, while reading is a verbal task that must be handled mainly in the dominant left hemisphere. The question arises as to whether a blind reader reacts to braille in the same way as a sighted person does to normal script. In other words, does the dominant left hemisphere play the leading role in the case of the blind reader, or are the stimuli mainly processed in the right hemisphere, as is the case in spatially arranged non-verbal items?
On the basis of these considerations, Hermelin and O'Connor assumed that in the first instance the stimuli of braille are treated by the brain as spatially arranged items, which can be analyzed more efficiently by the right hemisphere. This process probably takes place before or during the verbal decoding of the material in the left hemisphere.
If these assumptions are true, then right-handed braille readers can be expected to read faster and more accurately with the left rather than with the right hand. To test this hypothesis, Hermelin and O'Connor studied 14 blind children between the ages of eight and ten years. The children were asked to read equally difficult passages under four different conditions: with the left index finger, with the right index finger, with the left middle finger, and with the right middle finger. The sequence of conditions and passages was determined by a balanced design. Both reading time and number of mistakes were used as criteria for reading ability.
Before the study began, they checked to see whether the subjects were left- or right-handed. Twelve children were predominantly right-handed and two were ambidextrous. They found that left-handed reading was significantly faster and more accurate than was right-handed reading. The effect was more marked with time (p = 0.001) than with error scores (p = 0.025), and more evident for the middle fingers than for the index fingers (p = 0.05).
Both researchers did a similar experiment with 15 totally blind adults aged between 25 and 65. In this case they used research material composed of separate letters, arranged under one another in columns. The letters had to be read from top to bottom. Six different letter cards were presented in three different ways to the subjects, namely, directly in front of the midline, and to the left and to the right of it. The letters had to be read with both the left and right middle finger. Consequently there were six conditions. The sequence of conditions and presentation of the letter cards was determined by a balanced design. The position of the cards in relation to the body and the hands seemed to have no significant effect. Eleven of the 15 subjects made fewer mistakes with the left middle finger than with the right middle finger. If one used the number of letters correctly read as a score, no significant difference in relationship to speed was found. If the letters read correctly, as well as those read incorrectly were used as a score, then it was shown that reading was faster with the right hand than with the left hand. Teachers in the schools for the blind were rather surprised to observe that righthanded blind children could read faster and more accurately with the left hand than with the right. They had expected the opposite.
Replication of Hermelin and O'Connor's Research
The number of subjects who took part in Hermelin and O'Connor's experiments was limited, which affects the reliability of their results. It was decided, therefore, to carry out a similar experiment in the Netherlands with right-handed blind children. We started out with the following hypotheses.
1)The student reads faster and/or more accurately with the left index finger than with the right index finger.
2) The student reads faster and/or more accurately with the left middle finger than with the right middle finger.
3) The student reads faster and/or more accurately with the index finger than with the middle finger.
4) With spontaneous hand usage-which is nearly always with two hands-the student does not read faster and/or more accurately than with the index finger of one hand.
The third and fourth hypotheses had no connection with Hermelin and O'Connor's assumptions, but they were formulated on the basis of practical experience. Moreover, it was shown in the preceding research that pupils who read with only one hand generally performed as well as those who read with both hands. This was contrary to the prevailing opinion of the teachers (Mommers, 1977).
The experimental situation presented the opportunity of investigating, under more-controlled circumstances, the effect of use of two hands.
Study Design
In order to test the hypotheses, the student had to read under five different conditions: with the left index finger, with the right index finger, with the left middle finger, with the right middle finger, and with spontaneous hand usage. To allow a counterbalanced design, the number of subjects had to be a multiple of five (Kirk, 1968). It was found possible to include 25 totally blind, right-handed children in the research. They varied in age from 71/2 to 12 years. Whether the subjects were dominantly left- or right-handed was established by an adapted version of a lateral dominance test (Harris, 1958), which uses ten tasks to trace the dominant hand, e.g., Let us see how you throw a ball. How do you clean your teeth? How do you use a pair of scissors?
In their experiment with the children, Hermelin and O'Connor used reading texts composed of separate sentences. In the second experiment with adults they used separate letters. Given the difficulty of compiling five reading texts made up of equally difficult sentences, use was made of five word lists in our replication (Mommers, 1977). Because braille reading becomes more complicated when the reader has to move constantly from one line to another, the words were not printed in columns one under another, but beside one another. In addition to the five word lists, a list of numerals between 1 and 9 was used. Like words or letters, numerals are also verbal stimuli (Bakker, 1972, pp. 17, 67). Since the influence of memory is probably negligible while reading this numeral list, no parallel forms were designed. The numerals also were printed in lines beside one another. Preference was given to numerals above separate letters, because the first nine letters of the braille alphabet correspond to the nine numerals. They were all preceded by the numeral sign, which made it easier to localize the dots of the braille cell.
The sequence of conditions and lists that had to be read was determined by an incomplete balanced design. Only two conditions were taken down at each sitting to avoid student fatigue. Each time the word list had to be read first and then the numeral list. Two to three days elapsed between each sitting, and the experiment was always administered by the same person. In every condition, instruction was given by using a word list similar to the five word lists of the experiment. analysis of variance, no cases showed statistically significant differences.
Everything was taped and scored later.
Dependent Variables conditions.
To test the hypotheses, the following scores were used as
1) The number of words read in two minutes, both correct and incorrect.
2) The number of mistakes in the first 50 words.
3) The number of numerals read in one minute, both correct and incorrect.
4) The number of mistakes in the first 50 numerals.
The number of words or numerals read in the given time by each child was used as a measure of reading speed. Accuracy was measured by the number of mistakes made in the first 50 items, irrespective of the amount of time needed to read these items.
Results
a) Difference in speed and accuracy while reading with one and with two hands. From observation of the subjects, 22 children were shown to spontaneously use both hands while reading braille. Three children used one hand just to indicate the line, and then read with the other hand. Two subjects used their left hands exclusively and one used the right hand exclusively.
From Table 2 it appears that in both the word lists and numeral list, the average speed with spontaneous hand usage was clearly faster than even the fastest index finger. The differences were not statistically significant (p = .005). [For the words t = 7.02, and for the digits 4.35. DF = 24.]
Table 2. Mean number of items read (speed) and number of mistakes (accuracy) in spontaneous hand usage and with the fastest index-finger (N = 15)
words numerals
speed mistakes speed mistakes
Spontaneous
hand usage 69.4 1.8 88.8 0.4
Fastest
index finger 57.7 2.0 75.6 0.1
The number of mistakes made was minimal and the differences did not appear to be significant. On the basis of these data, the relative hypothesis was confirmed with regard to reading speed, but not to accuracy.
b) The effect of the use of different fingers.
The first and second hypotheses are relative to left- and right-handed reading. They form the essence of the replication. Table 3 shows that in the word lists, the average speed of the left index finger is faster than that of the right index finger, and vice versa for the numeral list. The differences in the middle fingers appeared to be similar. When tested by analysis of variance, no cases showed statistically significant differences.
Table 3. Mean under four different conditions (N = 25) of the number of words read in two minutes, and the number of numerals read in one minute.
words numerals
Left index finger 51.8 65.3
Right index finger 46.9 66.3
Left middle finger 31.9 42.2
Right middle finger 29.0 44.2
Table 4 gives the average number of mistakes under four conditions. More mistakes were made with the right finger than with the corresponding left finger, but these differences were not statistically significant.dependent variables:
Table 4. Mean under four different conditions (N = 25) of the number of mistakes in the first 50 words and in the first fifty numerals.
words numerals
Left index finger 2.8 0.7
Right index finger 4.5 1.2
Left middle finger 7.5 2.2
Right middle finger 10.0 2.8
On the whole, even though a trend indicating better performance while using the left hand does appear, considerable individual differences were also found.
A number of pupils performed obviously better with the right hand. Consequently, performance of the left and right fingers was individually compared. Tables 5 and 6 give thesedata.
Table 5. Comparison of the number of pupils who read faster with left or right fingers (N = 25)
words numerals
f % f %
Left index faster 17 68% 9 36%
Right index faster 8 32 14 56
Index equally fast -- -- 2 3
Left middle faster 16 64 10 40
Right middle faster 9 36 15 60
Middle equally fast -- -- -- --
About two-thirds of the pupils read faster with the left than with the right fingers when reading the word lists. But the number of pupils who read fastest with the right fingers was larger in the numeral list. With regard to accuracy, on every occasion there were pupils who read better with the left fingers than with the right fingers.
Table 6. Comparison of the number of pupils who read more accurately with the left or right fingers (N = 25)
words numerals
f % f %
Left index more accurate 17 68% 5 20%
Right index more accurate 1 4 4 16
Index fingers equally accurate 7 28 16 64
Left middle more accurate 14 56 12 48
Right middle more accurate 10 40 5 20
Middle fingers accurate 1 4 8 32
On the whole, the superiority of left fingers is most obvious and consistent in the word lists. Even though the hypothesis in relation to left-handed as opposed to right-handed reading is not confirmed in this research, it may be concluded that, in general, there is a trend for left-handed reading of words to be faster and more accurate than reading with the right hand. However this trend is open for discussion in the case of the numeral list.
Since the middle fingers are hardly ever used while reading, it was to be expected that reading with the index fingers would be significantly faster.
Discussion
On the basis of these results a number of comments can be made.
1) It is striking that differences in the use of left and right hands stand out more clearly when reading words than when reading numerals. A probable explanation is that recognition is not only an easier task, but that the processes involved are different. For to recognize words, selection has to be made from a greater collection of characters than is the case with numerals. There are only ten digits, and they are composed of the four dot positions of the upper part of the braille cell. The speed at which numerals are recognized is more than twice that of word recognition, while the number of mistakes made is not even one-third of that made with word recognition. On the basis of experimental research, Millar (1975) argues plausibly that recognition of braille letters by blind children may be influenced by both verbal and tactile characteristics, and that this is also dependent on the number of letters that can be retained in the short-term memory. If this number is more than four, then the verbal aspects play a relatively bigger role than is the case with three or less letter signs.
In other words, characters assimilated either verbally or haptically are also influenced by the number retained in the short-term memory. In the present experiment, the subject had to remember only one sign from the digits list, while most of the words were made up of three or more characters. These results thus support Millar's findings.
In further research into the effect of hand and finger usage in braille reading, the interaction established by Millar between the number of signs that must be retained and the way in which the information is assimilated must be taken into account.
2) It is more difficult to understand why Hermelin and O'Connor found significant differences in a smaller group of children. The fact that they used sentences instead of words may have had an effect; this does not seem likely. If Hermelin and O'Connor's assumptions are correct, then it is to be expected that the more the material to be decoded shows an obvious verbal character, the greater the role of the left hemisphere in the decoding process. The difference between reading with the left and right hand should be smaller when reading sentences than when reading separate words.
The fact that the pupils were, on average, a little older in the replication may have had an influence. In the older pupils the reduced difference between left and right hand could be a result of training and practice; from the first grade, teachers encourage reading with two hands.
Another influence on performance could be from the students' use of the braillewriter. Especially in the first years of learning, students are often urged to check every written word by touching, and because of the way the braillewriter is made, these checks are nearly always done with the left hand. 3) In the design both of Hermelin and O'Connor's experiment, as well as the replication, we suspect that inadequate attention was paid to several variables that at a later stage must be regarded as relevant. This neglect may have hidden other possible influences. It is known that hemisphere dominance of language functions develops gradually during maturation; at first the symmetry is directed to the left (Dumont, 1971, p. 56). Bakker et al. (1976) have, however, shown that with sighted children, the significance of this dominance in learning to read is different in the beginning and later phases. Good reading achievement at the beginning may be the result of either left or right hemisphere dominance. But at a more advanced stage (after the third grade) the ability to read well becomes more and more associated with left hemisphere dominance. With regard to reading development, girls are more advanced than boys during the primary school period. In order to get a clearer picture of the influence of hemisphere dominance, a distinction must be made in the research design between the different developmental stages of boys and girls.
A second weak point is that hand preference appears to be an unreliable indication of language dominance. Kimura (1967) demonstrated that by using the so-called dichotic listening test to ascertain ear preference, it is possible to establish reliably in which hemisphere language is represented.
4) That reading with two hands is, on average, faster than with one hand, corresponds to practical experience. Is it possible to draw a parallel here with visual reading, when research has shown that if one confines the field of vision to one or two letters, reading speed slows down appreciably? If braille is read with two hands it increases the field of observation. However, research with the Optacon has shown that reading with two index fingers does not lead to an increase in reading speed (Hill, 1973). An explanation of this should probably be looked for in another direction. Reading is not reducible to one function. In the reading act, different phases can be differentiated. In the first phase, emphasis is probably on the ambient aspects of perception, such as establishing the general layout of the text and discovering the length of words and sentences. The second stage is more focal, in which the sensorial pattern formed on the basis of information obtained in the first phase is tested. The difference between what an organism perceives and where it does this (Van Galen, 1974, p. 47) could be useful for further research into the differentiation of the functions of the hands in braille reading. This, however, necessitates a completely different research design.
5) Another aspect that has received very little attention in the discussion about hand and finger usage in braille reading is the possible influence of writing direction. On the basis of his research into the importance of left or right in perception, Zwaan (1966) concludes that in the static field of vision, the material is observed, assimilated, organized, and reproduced from left to right (p. 55), and that this is a function of the prevailing writing direction (p. 159). It is an open question as to what extent similar influences exist in finger usage in braille reading. Using sighted first- and fifth-grade children, Bakker and Van de Kleij (1975) did research into lateral and directional symmetry in observing tactile stimulation of the fingertips. It is true that their results indicate that reading is less dependent on spatial and tactile skills than spelling is, but it could be otherwise in the case of braille reading.
6) This replication was carried out in order to obtain more reliable data about the effects of different hand and finger usage in braille reading. It is extremely clear that in general, reading with two hands gives better results. It is also evident that reading with the index fingers is faster and more accurate than reading with the middle fingers. With respect to the preference for the left or the right hand, no great individual differences seem to exist, even though a slight preference appeared to exist for the left hand. This last point seemed to be so complicated that teachers should not apply too much pressure in this respect, but should allow the pupils to follow their own inclination.
References
Bakker, D. J. Temporal order in disturbed reading. Rotterdam: University Press, 1972.
Bakker, D. J., & Kleij, P. C. N. Van der. Lateral and directional asymmetry in the perception of tactile stiniuli, presented to the fingertips: Relations to age and sex, reading and spelling. Amsterdam: Paedologisch Instituut V.U., 1976.
Bürklen, K. Dos tastlesen der blindenpunktschrift. Leipzig: Verlag von Johann Ambrosius Barth, 1917.
Dumont, J. J. Leerstoornissen. Oorzaken en behandelingsmethoden (Learning disabilities. Causes and remediation.) Rotterdam: Lemniscaat, 1971.
Fertsch, P. Hand dominance in reading braille. The American Journal of Psychology, 1947, 60, 335-349.
Galen, G. P., Van. Ambient versusfocal information processing and single-channelness. Nijmegen: Psychologisch Laboratorium, K.U., 1974.
Grasemann, P. Das Tastlesen. In K. Burklen Dos tastlesen der blindenpunktschrift. Leipzig: Verlag von Johann Ambrosius Barth, 1917.
Harris, A. J. Harris tests of lateral dominance. New York: The Psychological Corporation, 1958.
Heller, Th. Studien zur Blindepsychologies Leipzig: Separat-druck aus Wundt, Philosophische Studien, XlBd, Heft 2 und 3, pp. 10-85, 1904.
Hermelin, B., & O'Connor, N. Right and left handed reading of braille. Nature, 1971a, 231, p. 470.
Hermelin, B., & O'Connor, N. Functional asymmetry in the reading of braille. Neuropsychologia, 1971b, 9, 431-435.
Hill, J. W. Limited field of view in reading lettershapes with the fingers. In F. A. Geldard Conferences on cutaneous communication systems and devices. Princeton, N.J.: The Psychonomic Society, 1973.
Kimura, D. Function asymmetry in the brain in dichotic listening. Cortex, 1967, 3, 163-178.
Kimura, D. The asymmetry of the human brain. Scientific American, 1973, 228, 70-78.
Kirk, R. E. Experimental design: Procedures for the behavioral scierices. Belmont, Calif.: Brooks-Cole Publishing Company, 1968.
Lothman, M. Research on braille reading. In Conference Report 'The Louis Braille British Conference on research into reading and listening by the visually handicapped.' London: 1975.
Maxfield, K. E. The blind child and his reading. New York: American Foundation for the Blind, Inc., 1928.
Millar, S. Effects of tactual and phonological similarity on the recall of braille-letters bv blind children. The British Journal of Psychology, 1975, 66, 193-201.
Mommers, M. J. C. Lezen met de vingers. Onderzoek bij blinde kinderen naar de samenhang van verbale intelligentie, de haptische waarneming en de leeftijd met het lezen van brailleschrift. (Reading with the fingers. Investigation into the relations between verbal intelligence, haptic perception and age with reference to braille reading by blind children.) 's-Gravenhage: Staatsuitgeverij, 1977.
Smith, J. M. The sensory function of the non-preferred hand. Journal of Experimental Psychology, 1934, 17, 154-159.
Stocker, C. S. Modern methods of teaching braille 1. Teacher's manual. Louisville, Ky.: American Printing House for the Blind, 1970.
Villey, P. Psychologie de la lecture tactile. Journal de Psychologie, 1931, 28, 214-249.
Zwaan, E. J. Links en rechts in waarneming en beleving. Left and right in perception and experience.) Utrecht: Erven J. Bijleveld, 1966.