University of Edinburgh

Tactile Graphics

CHAPTER SIX: Effects of picture format

6.1 Outline pictures

Kennedy is particularly interested in the perception of outline images. (Kennedy 1974, p 106ff) He describes various examples of tactile pictures, including some actually made of distracting and confusing elements (like shells). He concludes that drawings and particularly outlines must be kept simple. However, it appears to the writer that he oversimplifies the conclusions which he draws from some of his earlier small scale experiments, and overrides differences in the experience of his subjects. (Ibid, p 152) Also his statistical conclusions seem unproven, because he does not isolate and control against a number of possible causative factors. For example, he does not deal with the problem of scale in comparing a fork drawing with a diagnosis of 'a hand and arm' (for example). Neither does he take into account the atypical shapes given to some objects in the illustrations he used.

Kennedy's main conclusion, based on use of four plane outlines (hand; fork; man-with-arm-raised; flag) and four projective line drawings (face; cup; man-with-arms-crossed; table) was that there appears to be some deep-rooted human capacity to understand outline depictions of solid objects, a capacity shared by sight and touch. (Kennedy, 1974, p 152) Here again, the possibility of previous relevant experience with pictures is not discussed. The writer's contact with old and young congenitally blind and early blind people provides evidence that practical experience of pictures and conversation relating to pictures undoubtedly takes place, even if these experiences are generally casual and not pedagogically organised. They quite clearly could affect the results of experiments like those of Kennedy.

According to Kennedy's description, brain cells outside the main touch and vision pathways, but with access to both, deal with discontinuities in either vision and touch and accept lines as equivalents of the discontinuities. (Ibid p 112). Picturing was discovered by early people and not invented by them.

The writer believes Kennedy's interpretation (Ibid, p 154) to be too simplistic and to gloss over some other possible explanations for results which are after all generalisations from a rather small sample. It is quite plain from the writer's fieldwork on the development of touch perception in children up to the age of nine that the ease of interpretation of straightforward modes of tactile illustration has a lot to do with the degree of exposure to the crucial shape. For example, the writer included within a selection of tactiles used to assess the ability of 4 to 6 year-old blind children to identify tactile pictures without textual support, two sectional diagrams of fruit which were in fact produced for GCSE examination courses.

The first, of an apple, was correctly identified by children down to the age of three-and-a-half (the youngest children worked with). The researcher cannot remember any blind child failing to identify the subject of this picture, although not all of its components were understood by the younger children. It was found, however that the children were basing their diagnosis mainly on superficial characteristics such as the basic shape, and the presence of a stalk. In this diagram the bold presentation permits no likelihood of figure/ ground confusion and similar misunderstandings.

It was interesting that the comparable diagram of a rose hip was hardly ever identified by younger children without recourse to accompanying Braille text, although this diagram was similar in its style. The researcher believes that this was simply because the rose hip is not a fruit with which young children, and particularly young blind children are very familiar. They were aware of the existence of such a fruit, and perhaps had drunk rose hip syrup, but rarely if ever handled the fruit.

Pring, another researcher who uses mainly outline drawings similar to those of Kennedy, recognises a conflict between 'views expressed within the framework of visual perception' and the way the sense of touch operates. (Pring, 1987, p 38) The writer's classroom experience suggests that with the simple planar diagrams produced on German film (which gives very similar results to the Sewell board which Pring used) the ability to identify and interpret a picture depends very much on the form and complexity of the drawing. When this is done successfully without textual or teacher support it is often by the use of relatively superficial features of the drawing, and as Pring discovered, a simple confusion over components of a drawing can lead to early confusion between the stalk of a flower and the handle of a toothbrush (to use Pring's example). (Ibid , p 42).

There is a strong tendency for line drawings of this sort to rely heavily on artistic conventions established for visual art where the image is supposed to be related to the view of the real world perceived by the retina receiving focused light through the eyeball. (Reed and Jones, 1982, p 270; Gombrich, 1960, p 213) While some of these conventions may pertain to and be relevant to the hand's impression of the world, and others have for the time being an ease of manipulation which permits convenience in educational use, still other conventions are less relevant to the hand's sensations, and may be very difficult for a blind person to comprehend. It may in fact be a hindrance to general understanding that such comprehension should be attempted.

Kennedy's resonant phrase 'Lines depict visual discontinuities' is, however, useful as a descriptor of line drawing (Kennedy, 1974, p 132), and Kennedy in his work on visual art discusses the various types of discontinuities of colour, texture, orientation, illumination and the way in which they may be depicted and understood in visual line drawings. (Ibid, p 106ff)

The much earlier work of Merry and Merry (1933, p 148-163) has already been cited as an example of sweeping generalisations being made on the basis of responses to diagrams of poor quality. They also found that after practice their subjects identified fewer pictures than at the beginning. Among other shortcomings their investigation did not check the possibility of a drop in motivation, and they therefore drew unfavourable conclusions about the ability of the blind to understand haptic pictures. The Nuffield report on the teaching of Mathematics and Science to the blind (Fletcher, 1968) echoed this pessimism despite quoting in an appendix blind students who said that pictures are recognisable and interesting.

Millar (1975) was concerned mainly with the cognitive process but she also had blind children drawing people, which they did recognisably on many occasions and her work demonstrated the benefit of some teaching and experience to her blind subjects.

Kennedy and his collaborators presented a number of simple drawings to adults and recorded a high recognition rate. (Kennedy, 1982, p 318) Although their blind subjects tended to be less adept at this some could be better than their blindfolded sighted subjects.

Once again the previous experience of subjects was not known, and as Tobin points out, (1972) this experience is relevant to the interpretation of results. When making their own drawings of three-dimensional objects Kennedy's blind adults and children often used 'bulky line', 'enclosing line' or 'fold out' techniques to show depth. (Ibid, pp 320 ff) There was also some (apparently instinctive) use of thick and thin lines for near and far portions, although the educational backgrounds of people who used these techniques cannot be checked. Also from actual pointing and movement tests Kennedy found some suggestion of knowledge of perspective phenomena, but seemed unable to discover whether this was learned or intuitive behaviour. Kennedy fails to give a reason why this behaviour should be intuitive in congenitally blind subjects. Some of the adults also made use of additional graphics to indicate movement, including representations of footsteps, bending wheel spokes and a 'smoky spiral' (the author's term, not Kennedy's). Kennedy tries to differentiate between what is literal in these techniques and what are metaphors. (Unsuccessfully in the writer's estimation).

Kennedy also describes a system devised by Campbell and himself to show features such as slope and concavity. (Ibid, p 330) From the writer's educational experience with blind students this system, though logical, would appear to be an unnecessary complication and one which contains features which are themselves possible causes of further confusion. Kennedy only illustrates the use of this system by an ex extremely simple example, whereas the writer has designed and used many hundreds of tactile diagrams which have been found to be effectively interpreted by blind children and older students, and in many such diagrams Kennedy's code would be hopelessly submerged.

Kennedy's judgement is that a display can be so simple that it does not contain enough information to clarify lines, and yet it can be serviceable. The writer would agree that local difficulties can sometimes disappear if the over-riding message of a tactile picture emerges clearly, but this kind of experience cannot be universally assumed and it should certainly not lead the picture designer to ignore possible sources of confusion in picture components.

6.2 Format adopted in the present project

In the writer's own work simple outline drawings have been found to be insufficient for effective educational use except for very simple subjects such as plane geometry, and outlines well known to the pupils.

The hand and other touch sensors are equipped to receive three-dimensional information about the three-dimensional world, and making sensitive diagnoses from this information. Such information can include surface texture and assessments of movement, weight, and temperature, as well as details of gross shape. Accordingly it is found that tactile diagrams which include appropriate textural information, and strong indications of the third dimension, including concave and convex surfaces, will be helpful.

The eyes can assess the three-dimensional world from what is essentially a two-dimensional image, but in vision the two-dimensional image is enriched by light and shade, and contains modulations and gradients of colour which are often difficult to analyse and describe, but can nevertheless be interpreted and used by unsophisticated people and by young children.

To this is added perspective information, and sequential information during scanning. Although vision is often described as instantaneous in comparison with touch, this statement is only true in a very loose sense, because slight alterations of image obtained sequentially can yield important information about the shape and orientation of objects in the real world, and about their spatial relationships to one another.

In the writer's earlier work on biological tactile diagrams It was found that bold multi-layered relief with convex and concave shapes was needed to convey adequately the rich anatomical information that was sometimes required. (Hinton and Ayres 1986; Hinton 1988a p.26)

People with full vision often see and make use of silhouettes in conditions of bad illumination. Totally blind people never encounter such a flat two-dimensional image. The world encountered by the fingers always has information from the third dimension added to it. Even the suggestion or hint of this third dimension as in, say, a reduced-relief diagram may be straightforward and obvious in use.

Thus the reader may surmise that a structure is shown whole or sectioned according to whether it is depicted with a concave or flat surface. For the same reason concave or sunken channels are often used on a diagram to indicate the presence of a duct or tube, instead of a raised strip or a convex ridge. Although such sunken areas are still uncommon in tactile pictures they are often a more logical and helpful way of dealing with the subject matter, and if of sufficient size, no more difficult to finger-track or interpret.

Map-makers habitually use standard textures and symbols. Although standard techniques are used in the work being described where possible, and it is particularly important that textures for identical structures in a linked series are followed throughout the series, it is not always possible or desirable to adhere to a standard symbol. In such work, it is the diagram context which is important, and individual textures are interpreted in relation to what is found around them, (Hinton, 1990) perhaps with the help of Braille or other annotations. Thus, standard textures become less important, whereas appropriate and meaningful textures are vital to the understanding of the diagram and its components.

6.3 Line-of-approach and preliminary scanning of tactiles

The line of approach to a tactile diagram is important to the reader's clear understanding of the contents. In his survey of tactile map reading strategy Berla (1972, p 217) found that some systematic procedure was essential if important features were not to be overlooked. Berla did not, however, find that any single strategy was the sole answer, but that any of a number of possible approaches chosen by the reader's personal preference or sometimes by the specific demands made by a particular diagram would suffice. (Ibid, p 286)

Maps are generally, though not invariably, framed within a rectangular border which may also bear co-ordinates and indications of North, whereas the majority of the tactile displays designed by the writer are unframed. The reason for this is that the frame often contributes no useful information to the readers perception of the diagram, but simply adds to the clutter on the page. It may therefore help to confuse the blind user. The whole question of clutter is one of the bugbears of tactile diagram design. The problem is exacerbated by the presence of Braille on the page. Vital though this tactile code is, it is very bulky in comparison to print, and its very bulk may be the overwhelming consideration in tactile diagram design. The problem becomes particularly acute in Higher Education where the display may involve complex now charts and process diagrams. The designer is also entrapped by standard page sizes, and to a certain extent by what the reader's hands can comfortably span, so the space available is limited. Diagrams at that educational level may thus become multi-page presentations which may then require considerable duplication of information for the user's ease of reference.

Diagrams are generally provided with a Braille title and this is almost always at the top of the page, so any instruction regarding the line of approach to the diagram may naturally follow this title. In the absence of such an instruction the natural way to approach tactile diagrams is often from the edges of the page inwards.

As with vision the first cursory (hand) scan is important in discovering the meaning of what is on the page. It is possible to manipulate components of a tactile presentation to bring desired features into prominence, thereby helping (he reader attach importance to such features in seeking to understand the diagram.

There are three main contrasts which are available to the designer for this purpose. The first is high and low relief, with higher relief tending to be more prominent. The second is rough and smooth texture, and in general rough texture is the more prominent. Both of these two contrasts are pan of the stock-in-trade of most tactile diagram designers although the writer is noted for the variety of relief which he employs in tactile diagrams. The third contrast which is even more characteristic of the writer's personal style and has been consciously adopted following a great deal of classroom observation and research, is the contrast between concave, flat, and convex shapes. In this case the prominence comes from the novelty of a particular form amongst its surroundings. (Hinton, 1988b, p 10).

At the most basic level it is these contrasts which the writer is able to use to avoid the figure/ground ambiguity which was a source of confusion for Kennedy and Domander's blind subjects in their use of certain tactile line drawings. (Kennedy and Domander 1984, p 216).

In the writing of scripts for tape recorded notes, or in the preparation of oral comments to assist a blind student in the examination of such a tactile it is often useful to use such prominent features as a starting point, and also as a continuing point of reference for the location of other less obvious features of the picture. (Ricker 1981, p 297) It is important that a sighted teacher or picture designer keeps in mind always the sensations that the picture would evoke in a blind person. Some teachers find it necessary to blindfold themselves or shut their eyes, although with experience one learns to be aware of the needs of a blind reader even when working with one's eyes open.

6.4 Audio scripts for use with diagrams

The method which we generally adopt in preparing audio scripts to accompany the more complex diagrams owes a great deal to the work of Ricker at the University of Georgia. His paper (Ricker, 1981) gives further details of the experiences and reasoning which gave rise to his approach.

The summary of this approach which follows is reproduced with the author's permission.

  1. A brief preliminary overview of the diagram should precede detailed observation of the diagram. The individual elements of the diagram should be examined first, working from the outer to the inner part.
  2. Each component of the diagram should be located before it is described, and the description should precede the discussion of the component.
  3. The amount of information included in the script should be controlled carefully. Readers should receive enough information to form a mental image of the object or process depicted in the diagram. Additional information should be supplied by separate tapes, Braille or large print.
  4. The script should include a signal that informs readers when it is appropriate to stop the tape to obtain extra time for review.
  5. A summary should be included to help the readers review what they have just seen with their fingertips. While listening to the summary, readers try to conceptualise the information they are hearing rather than examine the diagram with their fingertips.
  6. A written glossary of pertinent words should accompany the diagrams. Ricker, (1981).

Where more than 16 labels were needed, the author's practice was to double the letters, starting at KK. Letters were then used in order from top to bottom of the diagrams. With rare exceptions, key letters did not correspond with the initial letters of the structures shown. Although it would be helpful if the connection between key letters and initial letters could be consistently achieved, in practice the scatter of initial letters makes this impossible, so it is best to use top-to-bottom key letters with an alphabetical listing. The Braille 'letter symbol' is needed on the key to avoid confusion with contracted Braille, but is not necessary on the actual diagram.

Consistent placing of key letters is important. Letters can be placed on the diagram components themselves when there is room, on the side nearest the margin of the page or alternatively can be placed beside the component. To be consistent, labels should be placed towards the left on the left half of a diagram and towards the right on the right half, and generally slightly above rather than below a structure. However, this is a rule of thumb and each label must be put where it will be most helpful to the students.

Where several small structures are so close together that there is not enough room for individual labels, two or more structures can be described in the key under one key letter. If these combinations are carefully chosen, the search for the structures can be a learning exercise in itself, and there need be no ambiguity for the student.

It is important to minimise all of these distractions in tactile diagrams, and to be aware that quite small tactual defects, virtually invisible to the eye, can be obtrusive to the touch, and can mar and confuse a tactile impression. At the same time in other contexts, the fingers can be aware of interesting features that the eye can overlook. (Hinton, 1984, p 30).

6.5 Distracting elements in tactile pictures

Kennedy (1982, p 314) has warned of possible confusion for blind users from distracting elements in tactual pictures as, for example in collages made from components like shells which already have a strong shape of their own. The writer has met similar problems, and also possible confusion from the use of textured sheet materials with a strong one-way pattern (eg; a wallpaper with a pattern of parallel ridges). Materials like these may have a use in a particular diagram context, but it must be recognised that they impose a strong pattern or trend of their own which may contradict or confuse the picture. This may not be a problem when a print pattern of a similar kind is used as an element in a visual diagram, because borders of patterned areas and delineating marks can be given greater weight to compensate for the strength of the pattern.

Many of the author's earliest tactile pictures were of biological subjects and were originally given strongly contrasting and sometimes very artificial textures, to ensure that picture components could be discriminated. It was quickly realised that it was desirable for such diagrams to be 'lifelike' as well, so that the blind user could make connections between the picture and the reality. To this end, textures were made less harsh and were chosen for their lifelikeness to the touch. (Hinton, 1988a, p 24).

It was realised that with the relatively strong relief which became a characteristic of the author's designs, natural forms could be used as picture elements. Raised border lines marking the limit of one tissue and the beginning of the next were frequently found to be unnecessary; a slight difference in relief level or texture would suffice. (Ibid, p 10) The omission of such edge lines removed yet one more distracting element from the tactile presentation.

6.6 Braille labels

With a few exceptions (generally diagrams of very simple layout) full Braille labels take up too much room to go directly on the diagrams. The best alternative is to put key letters on or near the structures and to have an alphabetical key on the facing page. (Hinton, 1988a, pp 9 and 10). This also helps with revision and testing of diagram comprehension. Where a series of diagrams share the same key and are bound up into a booklet, it is most helpful to the student if a separate copy of the key is bound to face each diagram.

The arrows which lead from the labels of printed diagrams always cause confusion in their tactile form, so they are best avoided altogether. Directional arrows are needed on flow diagrams of course, and these are satisfactory if they are prominent and carefully drawn. This is also true of the snakeskin-like tactile arrow symbol of Schiff, Kaufer and Mosak.

When using key letters in Braille, it is recommended that only the letters from K to Z are used, because the first to letter symbols are also used to denote figures in Braille. Where more than 16 labels were needed, the author's practice was to double the letters, starting at KK. Letters were then used in order from top to bottom of the diagrams. With rare exceptions, key letters did not correspond with the initial letters of the structures shown. Although it would be helpful if the connection between key letters and initial letters could be consistently achieved, in practice the scatter of initial letters makes this impossible, so it is best to use top-to-bottom key letters with an alphabetical listing. The Braille 'letter symbol' is needed on the key to avoid confusion with contracted Braille, but is not necessary on the actual diagram.

6.7 Microcapsule* diagrams

In recent years the diagrams produced on microcapsule paper by the stereocopier process have come to be widely used in education. In fact one of the principal speakers at the International Cartographic Association symposium held in 1988 at King's College, London produced user trial evidence to back his assertion that they were now the formal of choice for tactile maps. (Dacen and Coulson, 1988).

[*Please note that since this publication was written, the equivalent available technology would be Zychem or an equivalent.]

Their widespread use stems partly from the vigorous salesmanship of the makers of the copying machine, and partly from the speed at which diagrams can be produced. However, experienced diagram makers now find that it can take almost as long to produce a good-quality black and white ink drawing for reproduction in the stereocopier as it does to produce a master diagram of the same subject for thermoforming. It is also likely that the low-technology alternatives to the stereocopier for producing this type of diagram will be used more in future. (Ibid, p 9; Edman, 1989).

An advantage of this type of diagram is that lines, textures and Braille annotation are coloured in black, so that they are easier for a person with residual vision to see.

Disadvantages of this type of diagram include a rubbery surface texture which is found to be unpleasant by some users. These diagrams also smudge and deteriorate in use after a while. From the perceptual point of view the major disadvantage of this type of diagram is that all of the lines, areas and symbols on a diagram or map are raised to a constant height above the surface of the page so that there is no opportunity to provide the variety of levels and three-dimensional shapes which are possible with thermoformed diagrams.

In the opinion of the present writer the use of microcapsule diagrams should be determined by valid perceptual and educational considerations, and not by mere convenience to producers. The factors which would suggest their use in particular contexts would seem to be as follows:

(1) They probably provide the neatest and best option for producing simple, open-textured maps and graphs, particularly where simple geometrical shapes are shown. A comparable thermoform would probably be less tidy, or would take an inordinately long time to produce.

(2) Diagram components can be electronically stored for re-use in other diagrams. (Buultjens. 1988)

(3) They also function effectively as tactile overlays for the Nomad audio-tactile device (Parkes, 1988), where the more elaborate kinds of thermoform are less satisfactory when they interfere with the sensitivity of the touch-pad. In fact it may be that Nomad when developed to its full potential will make up for some of the shortcomings of microcapsule diagrams and allow diagrams of poorer structural quality to be used with greater success.

(4) Because of IT factors, they may be the best alternative for diagrams for use by blind students in Further and Higher Education at the present time, particularly in subjects where the diagram structure is relatively simple, but where the information carried is very varied and detailed.

However it is quite clear from me work of the present project, and from many conversations with experienced specialist teachers and technicians, that in many learning situations microcapsule diagrams are inadequate. There is also clear evidence from the present project and from other tactile diagram collections that the thermoform method will provide a solution to many diagram problems where the microcapsule method would utterly fail. Experience also suggests that the potential of both ways of making tactiles has yet to be fully exploited.

The criticism which is occasionally levelled at thermoforms by a few teachers is that they are bland and bleak in colour, and therefore dull and less helpful to users with some sight. It is interesting that this criticism is not levelled at thermoformed braille text, although admittedly the diagrams can be used by students with more vision than those who would want or need to read braille, and are even helpful to bright students with normal vision.

The writer's experiments with the colouring of thermoformed diagrams for use by people with some sight have so far proved disappointing. The two possible approaches are hand-colouring of the finished thermoform, and screen printing of the sheet before thermoforming, with careful registration of the print with the thermoform. Not only are technical problems found with both of these methods, but the results when apparently achieved are visually disappointing. More appears to be lost than is gained.

Some teachers get their partially sighted pupils to search for diagram components in order to colour them by hand with felt-tip pens. This can be a useful learning experience in its own right. However, in view of the problems noted above, and also me fact that the writer has never received complaints from partially sighted students, despite their readiness to draw attention to other shortcomings in diagrams, thermoformed diagrams in plain PVC appear to be the most useful product of this method of producing diagrams. Future colouring experiments may include rapid airbrush techniques.

6.8 Realism versus ease of interpretation

Anyone who makes biological tactile diagrams has to find an effective path between what appears to be realistic and what is easy to interpret by touch. When a student finds a diagram difficult to interpret the teacher can try the following:

a) Careful thought and questioning of the blind student may reveal obvious defects in the construction of the diagram which can easily be put right.

b) The teacher may need to redesign the diagram completely in order to produce something which is easier to follow. It may be necessary to make the diagram more crude, stylised or 'unreal'. On the other hand it may simply be a matter of putting less information on the page, perhaps by spreading the subject matter over two or more diagrams.

c) The teacher may give the student a more careful briefing, particularly if the student is not used to reading diagrams.

d) A preliminary diagram or series of diagrams may be given to students to prepare them for a more difficult one.

e) An audio file which instructs the student in both the exploration and interpretation of the diagram may be provided, perhaps in the style of those described by Ricker (1981) and summarised in Section 6.4 of this book.

f) Models, living organisms or other teaching materials may be brought in. Diagrams should be part of an overall teaching strategy involving many different media and plenty of concrete experience. It was clear from our classroom trails that most teachers are aware of the need to link the thermoformed diagrams to all sorts of practical experiences in the laboratory and outdoors. Museum services are sometimes a useful supplement to school resources.

We believe that biological diagrams should be as lifelike as possible. When textures and other features are used for labelling purposes. the student should not be given a misleading impression of the living organism or a false idea of its structure and function. This is particularly important when the student is trying to compare a diagram and the real object.

Ron Hinton
First published 1996
ISBN: 0901580775