                           CHAPTER 6
                                
                    A CLEARER UNDERSTANDING
                                 
  1.                                                   BRITAIN - A PERIOD OF CONSOLIDATION
       
       2.                                              AMERICAN - THE USE OF GRADE 1
       
       3.                                              1929 - 1932
       
       4.                                              A SURVEY OF THE FREQUENCY AND SPACE-SAVING OF THE
       CONTRACTIONS OF GRADE 2
       
       5.                                              THE 1956 BRAILLE WORKING PARTY
       
       6.                                              "A STUDY OF BRAILLE CONTRACTIONS", 1982
                                                       (A)  Literature Survey (Pt.2, 115pp.)
                                                       (B)  Questionnaire (Pt.2, Vol.2, pp.1-4)
       (C)     Surveys of Space-saving and Frequency of
            Contractions (Pt.1, Vol.2, pp.5-80)
            (D)     Experiments with Modified Grade 2 Braille Codes to
            Determine their Effect on Reading Speed (Pt.1, Vol.4, p.39;
            Material Used, Pt.1, Vol.3, pp.1-58)
            (E)     "The Limitations of Braille as a Medium for
            Communication and the Possibility of Improving Reading
            Standards" (Lorimer, J., 1978) (Pt.1, Vol.4A, pp.1-21)
            (F)     Analysis of Symbols, Meanings and Rules of
            Standard English Braille (Pt.1, Vol.4B, pp.1-16)
            (G)     Analysis of Errors (Pt.1, Vol.1, pp.31-34)
            (H)     Outlines of a Short Course to Improve Braille
            Reading Efficiency (Pt.1, Vol.4C, pp.1-47)
            
            7. THE WASHINGTON INTERNATIONAL CONFERENCE ON
       ENGLISH BRAILLE GRADE 2 (1982)
       
       8.                                              POST WASHINGTON
                                                       Capital Letters
                                                       Code Revisions
                                                       The BAUK Questionnaire
       
       9.                                              THE LONDON INTERNATIONAL CONFERENCE ON ENGLISH
       BRAILLE GRADE 2 (1988)
       
              10.                                                    OBSERVATIONS ON CODE DESIGN AND CHANGES IN
       ENGLISH BRAILLE FROM C1870 TO 1960
                                                            Circa 1870
     1895
     1905 Revised Braille
     1932
     1960
  
    1.   BRITAIN - A PERIOD OF CONSOLIDATION
  
  After revised braille had been accepted at the international conference
  held in Edinburgh in 1905, a time of consolidation followed.  The National
  Lending Library for the Blind, which supplied books in both braille and
  moon codes, had been in existence since 1862.  The standard of
  transcription was high for "volume by volume the books come in, are
  corrected at the library, shellaced and bound to wait on a cellar shelf until
  completed.  A report on each volume is sent to the writer, so that any
  mistakes are not repeated" (Outlook for the Blind, 1909, 2, p.173). 
  "Shellac" refers to a type of varnish applied to the pages which made
  them stronger and the dots less likely to be depressed.  The first braille
  copies were made by sighted transcribers then blind people made
  replicas either voluntarily or as a means of paid employment.  To
  maintain a high standard of teaching, the College of Teachers of the
  Blind was inaugurated in 1907.  Its examinations were recognised by the
  Board of Education and these included a test of a knowledge of braille.
  
  The settled state of the code was demonstrated by the fact that no
  meetings of the British Braille Executive Committee were entered in its
  minute book after 1905 until May 1916, though records of sub-committee
  meetings to tidy up the 1905 decisions still exist.
  
  In 1916 the National Institute for the Blind convened a meeting of
  representatives of educational societies, publishing houses, the home-teaching society and libraries, "to appoint a thoroughly representative
  committee (which shall in effect be a national committee) to deal with the
  whole subject of Uniform Type".  It became known as the National
  Uniform Type Committee and the first work to be considered was the
  changes contained in "Revised Braille for reading and writing, 1915" at
  the request of the American Commission on Uniform Type (see the
  previous chapter).  Besides the ensuing meetings and correspondence
  made necessary while the Americans were determining their code, a
  meeting was held in Britain in 1922 (Irwin, 1922, p.64) to the motion "that
  the present system known as Official Grade II does not adequately meet
  the requirements of the blind student, and that the present time is
  opportune for a more scientific adaptation of the Braille System".  In fact,
  the committee considered the timing most inopportune, for many
  Americans already thought Grade 2 was far too complicated and
  therefore differences between the two grades should not be increased. 
  It was decided that if so required, a scientific work could be prefaced with
  suitable symbols and their meanings.  No further code problems arose
  until Irwin put some American suggestions to a meeting of the British
  Uniform Type Committee in November, 1929.
  
  2. AMERICAN - THE USE OF GRADE 1
  
  Once the Commission on Uniform Type in the United States had
  accepted Grade 1 (1918) no time was lost in the production of text
  books for schools.  It was hoped that Grades 1 and 2 were sufficiently
  alike for exchanges of adult books to take place across the Atlantic.  In
  fact, it gradually became apparent from opinions given by library users,
  that whereas British people rarely read Grade 1, approximately 20% of
  Americans read Grade 2 and the trend seemed to be increasing (Irwin
  and Wilcox, 1929, p.1).  The Commission on Uniform Type for the Blind
  (1917, p.14) had been instructed to "continue its labors with the British
  National Uniform Type Committee with the view of reaching a complete
  agreement regarding the whole question of uniform type".  Having
  produced a number of text books for children the Americans were about
  to provide classics and more fiction for adult readers (Irwin, 1929,
  typescript, p.6), so it seemed the moment for the research department of
  the American Foundation for the Blind to undertake a more detailed study
  of the differences between the two codes.
  
  On first inspection the main differences seemed clear cut.  Grade 1
  used 44 of the 189 contractions included in Grade 2; two-space
  contractions, double letter signs and shortforms were not included; and
  the American version used capital letter signs wherever capital letters
  occur in inkprint.
  
  The American lexicographer, Noah Webster, devoted much of his life,
  particularly between 1789 and 1828 "to the zealous championing of the
  cause of American language, its spelling, its grammar and its
  punctuation" (McCrum, et.al., 1987, pp.240-241).  Webster's maxim was,
  "A good articulation consists in giving every letter in a syllable its due
  proportion of sound, according to the most approved custom of
  pronouncing it, and in making such a distribution between syllables of
  which words are composed that the ear shall without difficulty
  acknowledge their number", and school children were often made to
  chant columns of his spelling lists (ibid., pp.241-242).  It does not seem
  too far-fetched to suggest that his writings had an effect on American
  preference for syllabication in braille usage over a century later.  Syllable
  divisions are not always easy to determine so a dictionary is a useful tool
  for the transcriber.
  
  By contrast, in Britain contraction rules permitted the shortening of words
  so long as the clarity of meaning was not impaired.  This sequencing
  often resulted in contractions crossing syllable boundaries and
  sometimes a choice of contractions had to be made involving the need
  for yet more rules of usage.
  
  No exhaustive analysis of space-saving values had been carried out in
  the 1920s when Grade 1 was being used in America.  In the interests
  of uniformity, Irwin and Wilcox (1929) made a study of 91,564 words
  taken from four volumes, specifically to determine the space-saving
  effects of the use of Grade 2 over Grade 1.  They were particularly
  interested to know what the effect would be if Americans accepted the 44
  contractions and shortforms not included in Grade 1.
  
  The 91,564 words were treated as if written in a continuous line to
  obviate the effect of varying lengths of lines in the publications consulted. 
  When written in Grade 1 they took up 427,776 cells and the Grade 2
  version used 376,866 cells, that is, 11.9% less space than Grade 1. 
  This saving was caused not only by the differences between the space
  saving of the contractions, but because Grade 1 had to omit some of
  the contractions because of the syllabication rule (Loomis, 1936, pp.4-5). 
  For example, the following contractions were not permitted:
  
     AR   in   library, military,
     ER   in   serious, exterior,
     AND  in   candle, abandon.
  
  Because of these imposed restrictions in the use of Grade 1 the code
  was less flexible and the uncontracted letter strings were likely to have
  had a deleterious effect on the rate of reading.
  
  Irwin and Wilcox included tables showing in descending order the space
  saving value of each contraction and also its frequency of occurrence. 
  Both these measures are necessary for determining the relative value of
  contractions for inclusion in the code.  The extent to which they facilitate
  or hinder reading is more difficult to determine and more accurate results
  can only be provided by research.  The results showed that the code was
  even less economic than forecast, for it was found that though TO saved
  7% of all the space saved by Grade 2 over Grade 1, fifty contractions
  at the lower end of the table taken together did not save as much space
  as some individual contractions at the top of the table, and seven
  contractions or shortforms did not appear in the literature selected.  The
  study was sent to members of the (British) Uniform Type Committee with
  the hope that by a judicious addition of contractions to Grade 1 and
  some deletions from Grade 2, a better measure of uniformity might
  ensue.
  
  3. 1929 - 1932
  
  Irwin set out some general proposals for bringing the two grades closer
  together before the meeting of the British authorities held in July, 1929
  (British NUTC minutes).  Though not all members were enthusiastic, they
  agreed to set up a sub-committee to meet with the Americans.  The
  American Printing Press were so concerned by Irwin's activities that the
  director sent a cable to England, the gist of which was "We are not sure
  what you are doing in your effort to bring about braille uniformity, but
  whatever it is, please desist" (Irwin, 1955, p.50).  It did not reach him until
  his meeting with the British was concluded!
  
  The AAIB and AAWB were asked to bring the study before their members
  to enable them to discuss what further steps should be taken.  Publishers
  of braille magazines in America had helped readers to understand
  possible changes by gradually introducing Grade 2 signs in their
  publications before a conference was held in London in July, 1932,
  between representatives of the braille authorities of Britain and America
  (Irwin, 1932, p.138).  By contrast with past disagreements, the
  atmosphere was cordial and it would seem that a great deal of this co-operation was due to the careful preparation carried out by Robert Irwin,
  himself a blind man.
  
  The main decisions (British UTC Minutes, 1932) included the following:
  
  
  1. The dot numbers within the cell were changed from:
       
       
       1     2
  3     4
  5     6
  
  to
  1     4
  2     5
  3     6
  
  
          (This reversion to the French method was probably taken so as to
       be compatible with braille in mainland Europe.  It is suggested by
       the writer that the interim method may have been used so as to be
       the same as dot numbering for music.)
       
       2. The British signs for capital letters and italics were reversed, the
       new positions being:
       
                    Capital Letter Italics
                                l
                                
                        l        l
       
                    (This decision helped recognition of two-spaced compound
       contractions.  America continued to use capital letters wherever
       they occur in print, whereas in Britain the capital was not generally
       used.)
       
       3.           Britain agreed to drop the use of nine biblical words except in
       religious literature, and five new shortforms were added (see
       Appendix 1).
       
       4.           The Americans agreed to adopt all the contractions and
       shortforms not already included in Grade 1.
       
       The Americans had come a long way in accepting so many of the Grade
  2 rulings in the interests of uniformity, but they still differed over some of
  the British practices with regard to sequencing within words.  Rule 34
  was inserted in "Standard English Braille, Grades 1 and 2" (1932,
  National Institute for the Blind, p.18) as a compromise: "The contractions
  forming parts of words should not be used when they are likely to lead to
  obscurity in recognition of pronunciation, and therefore they should not
  overlap well-defined syllable divisions.  Word signs should be used
  sparingly in the middle of words unless they form distinct syllables. 
  Special care should be taken to avoid words of relatively infrequent
  occurrence."
  
  Lists of words showing preferred methods of contracting were appended,
  which is evidence of the problems involved in trying to use a code of
  comparatively few signs to represent the large vocabulary making up a
  language.  Despite these difficulties, after a century of endeavour, the
  English-speaking peoples on both sides of the Atlantic were united in
  having a code similar enough for an increase in the exchange of books,
  and any further code alterations could be decided together in a spirit of
  harmony.  Canada, Australia, and New Zealand would benefit as well as
  the peoples for whom English is a second language.  The conference
  was a significant occasion in the history of the development of the braille
  code.
  
  4. A SURVEY OF THE FREQUENCY AND SPACE-SAVING OF THE CONTRACTIONS OF GRADE 2
       
  Although a close degree of uniformity had been reached in 1932, a sub-committee of the (British) UTC undertook to look to the future to
  determine whether yet more improvements could be made.  A more
  detailed research was needed (British NUTC minutes, 28th January,
  1950), so the first task undertaken was a wide ranging survey of the
  frequency and space saving values of all existing Grade 2 contractions
  to see how far it is an effective system.  This was carried out by
  Lochhead and Lorimer, J. (1954).
  
  To fulfil its purpose a survey not only has to be carried out with accuracy,
  but must cover as wide a range as possible in proportionate amounts
  within the confines of the subject studied.  With these criteria in mind, the
  larger the number of words studied, the more accurate the study should
  prove to be.
  
  The study was to cover general reading so learned and scientific works
  were not included.  Twelve subject areas were covered ranging from
  nineteenth century fiction to modern day periodicals.  No calculators or
  computers were available at that time so the processing of half a million
  words was carried out by volunteer helpers who were all experienced
  braillists.  From this information the relative space saving of each
  contraction, and also the usefulness of "families" of contractions which
  have similar rules for usage, could be determined.  The results again
  showed that Grade 2 is a far from efficient system.  Nine contractions had
  been deleted and 5 new ones added in 1932 bringing the total to 185. 
  Of these, the first 14 contractions do almost as much work as the
  remaining 171 and the last 45 contractions between them only save 500
  spaces.  The survey was a model of careful planning and efficient
  execution.  A copy of the report was sent to the Joint Uniform Braille
  Committee in America (British NUTC minutes, 25th March, 1955).
  
  5. THE 1956 BRAILLE WORKING PARTY
  
  With the publication of a detailed rule book by the British UTC (1953),
  and the Lochhead and Lorimer survey showing strengths and
  weaknesses of the code, the NUTC felt in a strong position to make
  suggestions for possible improvements.  It was considered that "space
  saving was important, but that readability and ease of learning were more
  important" (British UTC minutes, 25th March, 1955).  A sub-committee
  produced 10 recommendations (ibid., 3rd June, 1955), which were sent
  to the Americans, who in turn sent some of their own which were less far-reaching.  They also suggested a joint working party for all the proposals
  to be discussed, so representatives from both countries met in London
  in 1956.
  
  The joint decisions (ibid., 30th November, 1956) were considered by
  representative bodies in Britain and North America.  Many of the
  suggestions were accepted in Britain, but in America, not only were the
  deletions suggested in 1956 approved, but further deletions had been
  approved since, thus breaking the "gentlemen's agreement" between the
  two countries.  It had been a democratic decision by the American public. 
  The main changes concerned the use of lower signs when used in
  sequence, the addition of four shortforms (bringing the total to 189), and
  the dropping of the use of a hyphen in three compound words.  America
  had agreed to most of the contents of Grade 2 in 1932.  The position was
  now reversed with Britain agreeing to American preference.
  
  6. "A STUDY OF BRAILLE CONTRACTIONS", 1982
  
  During the 50s and 60s some research studies on the braille code were
  carried out by members of the research committee of the College of
  Teachers of the Blind.  Meanwhile, a substantial amount of speculative
  and empirical braille research was accelerating in America into the
  perception of braille and methods of making the code easier to learn and
  to read.  The time had come for the CTB research committee to be
  disbanded and instead, the full-time Research Centre for the Education
  of the Visually Handicapped was set up at Birmingham University in
  1970.  Warwick University also had a research centre.
  
  Douce and Tobin (1976) considered that the time was ripe for a multi-variate approach "to generate the kind of information that would make it
  possible to formulate decisions about code modifications that would, in
  turn, make the learning of braille, and subsequent improvement in
  reading efficiency, easier".  A joint three year project was therefore
  undertaken by Warwick and Birmingham universities (Lorimer, J., et al.,
  1982, vol.1, p.2) with two main aims: (1) to explore the possibility of
  producing an alternative Grade 2 code that would be shorter and simpler
  to learn, but could be read as fast as the existing Grade 2 code and
  would require no more space, and (2) to investigate ways of improving
  teaching methods and materials.  Descriptions of some of the main
  studies follow.
  
  (a)     Literature Survey (Pt.2, 115pp.)
  An annotated bibliography was compiled by the writer during the first
  year of the study for the benefit of the research team.  It included works
  on the development of the braille code, research and education, but
  because of the time factor, parameters had to be set; approximately 300
  entries were included, most of them published between 1945 and 1980. 
  An author index was included and also a subject index compiled from the
  descriptors which had been included with each entry.
  
  (b)     Questionnaire (Pt.2, Vol.2, pp.1-4)
  The questionnaire survey was undertaken to help in identifying specific
  aspects of the braille contraction system which merited detailed study. 
  Three hundred and one responses were received from blind, partially
  sighted, and sighted subjects whose age range was 20-93 years, all of
  whom lived in the United Kingdom.  Information was received on such
  aspects as preferred grade of braille, opinions on space saving, specific
  problems in recognition of certain contractions, and opinions on
  punctuation and unit signs, and format.  In conclusion (p.4) "the general
  view seems to be that there is no desire for any change unless it can be
  very clearly demonstrated that there are some very significant
  advantages to be gained by modifying the system".
  
  (c)     Surveys of Space-saving and Frequency of Contractions
       (Pt.1, Vol.2, pp.5-80)
       The Gill corpus of 2,255,326 words was compiled at Warwick University
  using a selection of material already stored in the data bank.  This
  consisted of short runs which had been put into braille at the request of
  individuals, and also an almost equal number of words entered from
  novels, short stories and non-fiction.  Also included were the contents of
  the Kucera and Francis count (1967), often referred to as the Brown
  Corpus.  Numbers of pairs, three-letter and four-letter strings of letters
  were tabulated indicating whether they occurred at the beginning, medial
  and final positions in words, and a table in descending order of space-saving was also included.  Thus, by inspection, it was possible to
  determine which groups of letters would be the most economical to
  include in a projected improvement to Grade 2 braille.
  
  The Lochhead and Lorimer study (1954) containing approximately
  540,000 words, and the Kederis, Siems and Haynes (1965) count
  containing 291,000 words from children's literature, were already in
  existence, but it was felt that none of the four corpora was entirely
  satisfactory because of differences in the selection and proportion of the
  materials treated, the differences in the number of words analysed and
  also the variations in language and braille usage between Britain and
  America.  To obtain a more accurate result each contraction was reduced
  to occurrences per million words in each of the four counts and then the
  mean frequency was computed and tabulated in descending order.  A
  similar treatment was given to demonstrate the space-saving values.  A
  further table showed the space saving contribution made by  families' of
  contractions each containing contractions which have similar rules.  This
  knowledge is useful because keeping a whole family or deleting the
  whole or part of it may have a considerable effect on the use of space
  and the number of necessary rules.
  
  (d)     Experiments with Modified Grade 2 Braille Codes to
       Determine Their Effect on Reading Speed (Pt.1, Vol.4, p.39;
       Material Used, Pt.1, Vol.3, pp.1-58)
       In general, reductions in the contents of a code will decrease the number
  of necessary rules and make the learning task easier, whereas
  extensions will increase the learning load.  However, before changes are
  made it is necessary to estimate the probable effect of these changes in
  terms of space-saving, ease of learning and on rate of reading.  Findings
  of frequency and space-saving surveys had given information that might
  lead to improvements, but so far no measurements had been made of the
  effect of possible modifications based on this information in terms of
  words read per minute.
  
  During the first year of the study while other investigations were
  proceeding, 10 code modifications were tried out under controlled
  conditions in order to determine their effect on rate of reading.  It was
  realised that because of lack of experience with the experimental codes,
  subjects tended to read them more slowly than the Grade 2 version with
  which they were familiar, and therefore differences in the scores obtained
  should be regarded as maximal. However, they provided a useful
  indication of the effect of certain modifications on reading rate.
  
  The criteria for the experiments included the fiats that space-saving and
  reading rate should be at least equal to those for Grade 2.  The results
  of the experiments showed that about 70 of the useful contractions could
  be deleted, but to fulfil the terms of the criteria, about 10 new space-saving contractions of greater value could be added.  Statistical data
  supplied by other investigations in the study supplied further predictions
  (Vol.2, Ch.5, p.79).
  
  (e)     "The Limitations of Braille as a Medium for Communication and
       the Possibility of Improving Reading Standards" (Lorimer, J.,
       1978) (Pt.1, Vol4A, pp.1-21)
       This paper, written during the early stages of the project, was read at a
  meeting organised by the British Psychological Society.  Comparisons
  are made between accuracy, comprehension, and reading rates of blind
  and sighted children (Ashcroft, 1960; Williams, 1971; Lorimer, 1977). 
  Information is also included concerning intelligence and braille reading,
  the limitations of touch perception, and the application of rapid reading
  techniques for inkprint to braille reading in an attempt to improve braille
  reading standards.
  
  (f)     Analysis of Symbols, Meanings and Rules of Standard English
       Braille (Pt.1, Vol.4B, pp.1-16)
       It has long been realised that one of the main difficulties in reading braille
  has been the number and complexity of the rules.  Fewer contractions
  with attendant rules might make the code easier for some people at the
  learning stage, particularly adults who can already read inkprint yet have
  to learn a new set of symbols and need training in their perception. 
  Computer programmers would also welcome a simpler code.  An analysis
  was carried out to determine:
  
  1. the number of symbols, meanings, and governing rules;
       2. the number of rules which need not be known by beginners;
       3. how far rules for a revised code could be reduced without lowering
       the efficiency of braille.
       
  (g)     Analysis of Errors (Pt.1, Vol.1, pp.31-34)
  The Lorimer Braille Recognition Test (1962) was administered to 91
  subjects in the United Kingdom, of whom 64 were under the age of 20
  and 27 aged 20 and over.  The test included 178 of the 189 contractions
  of Grade 2.  No errors were recorded on 23 contractions and of the 23
  contractions most associated with error, 311 mistakes were recorded by
  subjects under the age of 20 and 111 errors by those who were 20 years
  and above.  This test had been standardised on a sample of 332 children
  but there is no standardised braille reading test yet available in Britain for
  use with adults.  These facts, though gained on a relatively small sample,
  can be taken as an indication of some of the reading difficulties of Grade
  2 braille.
  
  (h)     Outlines of a Short Course to Improve Braille Reading
       Efficiency (Pt.1, Vol.4C, pp.1-47).
       Based partly on rapid reading techniques involved for sighted readers
  (Fry, 1963; de Leeuw, M. and E., 1965) and research on rates of braille
  reading (McBride, 1974; Crandall and Wallace, 1974), Lorimer, J.
  planned a course intended to improve the rates of reading of pupils in the
  lower grades of senior schools.  The materials used are included in the
  report and also a table for converting reading time to reading rate so that
  replication of the experiment is possible.
  
  The Birmingham/Warwick study was the most extensive investigation of
  braille ever attempted and was bound to accelerate further studies of the
  code.  The report was sent to the Department of Health and Social
  Security, and members of the Braille Authority of North America (BANA). 
  The writer was present at the Toronto conference held in 1981 in
  preparation for the Conference of English Braille, Grade 2, held a year
  later.  The chairman praised the work of the university investigators and
  remarked (pers. comm.) that the enquiry was at least five years ahead of
  any other enquiry at that time.
  
  7. THE WASHINGTON INTERNATIONAL CONFERENCE
       ON ENGLISH BRAILLE GRADE 2 (1982)
       
  The basic assumption of the universities' study was that there should be
  one official code in general use.  Members of BAUK recognised the
  advantage of reducing the number of contractions to be learned without
  sacrificing space and it was hoped that such changes would induce more
  adults to want to learn the code.  However, a few members of the Braille
  Authority of the United Kingdom (BAUK, formerly British UTC) wanted an
  advanced grade as well, with many more contractions for the use of
  fluent, extensive braille users even though such people would be in the
  minority (Milligan, 1982, p.93).
  
  In 1978, the Braille Authority of North America (BANA) submitted the
  following recommendation for consideration by BAUK (proceedings of the
  Washington conference, 1982): "The Braille Authority of North America
  wishes to join with the National Uniform Type Committee of the United
  Kingdom to explore the possibility of devising a common Literary Braille
  Code for the English language.  The exploration should encompass both
  readability and computer implementation."
  
  At the Washington Conference (1982) delegates recognised that
  computer use of braille would make some code changes inevitable, but
  such changes should be regarded as "temporary, experimental
  guidelines, rather than amendments to the braille code" (Resolution 6,
  p.248).
  
  Amongst the research suggestions (Resolution 7, p.248) were: the
  development and field-testing of a contracted literary code based on the
  experience of users and teachers as well as on theoretical consideration
  and research findings; the achievement of uniformity in the form,
  placement, and spacing of unit abbreviations for coinage, weights, and
  measures; and attempts to persuade properly qualified persons to
  provide information about the structure and functions of the touch sense
  (with reference to braille reading).  These suggestions seem wide
  ranging and it was encouraging to the writer, who was present, to note
  the attempted welding together of expert theoretical knowledge with user
  experience.
  
  Possibly the most important resolution passed was the setting up of an
  international co-ordinating committee whose members were to be one
  representative from each of the participating nations.  Their mandate
  included the preparation of proposals for the establishment of an
  international authority on English literary braille.  This was a
  commendable step, for so far the only authorities making decisions had
  been those of Britain and America.  The activities of the proposed
  international authority will be further considered in Chapter 9.
  
  Another conference was planned to take place in London when definite
  proposals for changes to the code would be considered and braille users
  were to be fully consulted before any such changes were implemented. 
  The exception was the immediate acceptance by Britain of the American
  practice of using the AR contraction in preference to EA in the letter
  group "ear".
  
  8. POST WASHINGTON
  
  Capital Letters
  A joint investigation was carried out between the USA, Canada, and
  Britain to determine what effect the capital letter sign had on the
  comprehension and rate of reading.  The results were insignificant and
  at present each country is following its own preference regarding the use
  of capital letters.  Americans still prefer to use them where they occur in
  print and the British use them very rarely.  An ongoing enquiry under the
  auspices of BAUK is at present (1996) being carried out concerning
  "Capitalization in British Braille" (see Chapter 9).
  
  Code Revisions
  A large proportion of inter-conference time was spent in Britain on
  evolving a variety of new versions of the literary braille code system. 
  These may be consulted in "Essays in Code Design" (RNIB, 1986) where
  (p.5) the report is regarded as "the culmination of the research" carried
  out by Birmingham and Warwick universities.  They include:
  
    a simple contracted code intended for those who are unable or have
    not the time or inclination to master the intricacies of Grade 2;
         two codes intended for the average reader;
         and a very much simplified code which is the springboard to another
    which is greatly expanded for the advanced reader.
    
  These codes, together with the experimental ones tried out in the
  universities' study, demonstrate very clearly that there can be no one
  universal code suitable for all braille readers.  Varying abilities, interests,
  and circumstances need to be considered, preferably the codes should
  have the same meanings for the symbols that are common, and it would
  be impractical in terms of expense to have books in too many code
  versions.  The trials were useful exercises in demonstrating the
  difficulties for code design is a very complex exercise.
  
  The BAUK Questionnaire (1986)
  A questionnaire on braille code reform was circulated by BAUK in 1986
  to which there were in all 1627 replies, 1217 of whom were touch
  readers.  Braille is an emotive subject for blind people.  It is something
  they regard as their own and conservative feelings about change were
  expected, especially among the older members of the population.  On the
  whole, advanced readers and young readers were the more ready for
  reform.  About two-thirds of the braille using respondants wished the
  contents of Grade 2 to remain substantially unaltered but with some
  further contractions added (Lorimer, 1988).
  
  9. THE LONDON INTERNATIONAL CONFERENCE ON
       ENGLISH BRAILLE GRADE 2 (1988)
       
  The BAUK questionnaire (1986) had revealed that the majority of braille
  users who sent in replies did not want any major changes to the literary
  braille code.  The Washington braille conference (1982) had
  recommended the setting up of an international council on English braille
  which would be ratified at the London Braille Conference of 1988.  These
  two decisions led to a conference decision that any further improvements
  to the code should be delayed until the International Council on English
  Braille was set up.  Arrangements for the latter were to begin immediately
  following the conference.  However, a few changes were agreed
  involving the use of quotation marks, fractions, and unit abbreviations.
  
  10.     OBSERVATIONS ON CODE DESIGN AND CHANGES
       IN ENGLISH BRAILLE FROM C1870 TO 1960
       
  Because touch reading is such a slow medium, the choice of contractions
  to reduce the amount fingers must travel is of great importance.  In
  theory, the more contractions that are included, the greater the rate of
  reading, but there are other qualifying aspects.  Too many different
  symbols increase the learning load and number of rules, and of even
  greater importance is an understanding of the problems presented by
  embossed material.  These perceptual difficulties are complex and will
  be considered at greater length in Chapter 8.  More easily measured is
  the amount of space-saving contributed by each contraction leading to
  an increased rate of reading.
  
  Because "the Grade 2 code is not as efficient a communication medium
  as it might be" (Lorimer, J., Tobin, Douce, and Gill, 1982, p.121) it is
  obvious that some contractions save more space than others.  For
  example, "immediate" IMM saves 6 spaces whereas "your" YR saves 2,
  so it might be thought that "immediate" would be of better value in space-saving, but it is also necessary to know how often the contraction is likely
  to occur.  Even after calculating the mean values of four counts in order
  to gain greater accuracy of values of space-saving, Lorimer, J. added
  ("Study of Braille Contractions", Vol.2, p.49) that these "must be regarded
  as approximations, for there can be no absolute frequency or space-saving value for any contractions".  Using Table 17 (ibid., Vol.2, p.58) it
  may be seen that in a million words "immediate" saved only 865 spaces
  whereas "your" saved 3,134, the reverse of what might have been
  previously expected before frequency was considered.
  
  This aspect of space-saving was not always recognised in the past so
  Table 3 (below) was compiled by the writer to provide the information in
  a compact form.  It is realised that vocabularies change slightly over the
  years but even so the values shown in the table give some indication of
  the success of the added contractions.  A few of the contractions were
  given different symbols in ensuing revisions so, by adjustment, the
  contractions here show space-saving at the time when they were first
  included in the code.
  
  Nine religious contractions and five others were included in the British
  Literacy braille code at various times between c1870 and 1932.  These
  were not included after 1932 so there are no space saving values
  included for them in the 1956 and 1978 counts, and therefore do not
  appear in the table.  The table should therefore be regarded as indicating
  main trends in the year before 1932.
    Table 3
  
  Table to show the number of new contractions introduced into the British
  literary braille code between c1870 and 1960, which saved at least one
  space over Grade 1 per million words.
  (The space saving shown here does not include alphabetic, punctuation,
  composition or mathematical signs.)
  
  
  Total Number of Spaces
  Saved by Each
  Number of Contractions Introduced
  
  
  Contraction
  c1870
  1895
  1905
  1932
  1960
  
  
  
  Over 50 Thousand
  
  4
  
  -
  
  -
  
  -
  
  -
  
  
  40,000 - 49,999
  30,000 - 39,999
  20,000 - 29,999
  10,000 - 19,999
  
  3
  4
  3
  17
  -
  -
  -
  -
  -
  -
  2
  3
  
  -
  -
  2
  3
  -
  -
  -
  -
  
  
  
  9,000 - 9,999
  8,000 - 8,999
  7,000 - 7,999
  6,000 - 6,999
  5,000 - 5,999
  4,000 - 4,999
  3,000 - 3,999
  2,000 - 2,999
  1,000 - 1,999
  
  
  2
  1
  1
  2
  3
  5
  1
  7
  7
  
  -
  1
  -
  1
  1
  1
  2
  4
  4
  
  1
  1
  2
  4
  6
  5
  9
  11
  22
  
  -
  -
  -
  -
  -
  -
  -
  -
  -
  
  -
  -
  -
  -
  -
  -
  1
  2
  -
  
  
  
  900 - 999
  800 - 899
  700 - 799
  600 - 699
  500 - 599
  400 - 499
  300 - 399
  200 - 299
  100 - 199
  Under 100
  
  
  -
  -
  -
  1
  -
  -
  1
  -
  -
  -
  
  -
  1
  -
  -
  -
  -
  2
  -
  -
  2
  
  7
  -
  1
  6
  1
  1
  5
  1
  3
  8
  
  -
  -
  -
  -
  -
  -
  -
  -
  -
  -
  
  1
  -
  -
  -
  -
  -
  -
  -
  -
  -
  
  
  
  Total Number of New Signs
  
  
  62
  
  19
  
  99
  
  5
  
  4
  
  
    Circa 1870
  At once it becomes apparent that the choice of contractions included in
  the first version of the English braille code was excellent, for at least half
  of them saved over 10,000 spaces each per million words and very few
  must have saved less than a thousand spaces.  However, it has to be
  remembered that there was a certain freedom of individual choice
  concerning the use of shortforms so these cannot be included.  It is
  known that Armitage (1874) analysed passages from four works to
  calculate space saving, so it is surmised that he may have already used
  these values when the code was being adapted from the French version.
  
  1895
  The additions had been decided in committee over the intervening years
  since 1870 (BFBA minutes) so do not represent a full-scale revision of
  the code.  1870 had had first choice for the most space-saving
  contractions so a concentration on middle-range contractions was to be
  expected.  The worst feature was the inclusion of more contractions in
  the lowest space-saving range, which suggests that the choice was made
  without reference to frequency.
  
  1905 Revised Braille
  After revision in 1905 the new code became just over double the size. 
  This was partly caused by the addition of specific shortforms instead of
  the writer being left with the choice.  The decision made for less guessing
  by the reader and presumably speedier reading.  The transference of
  some meanings from simple to compound symbols to make room for new
  words is to be commended.  Altogether 13 simple upper wordsigns were
  added.  IT saves 10,515 spaces per million words, 9 words were added
  in the middle range, and 3 each saving less than a thousand spaces. 
  KNOWLEDGE occurs very infrequently and GO and US are so short that
  they were not worth being included, but other short words earn their
  place because of their frequency value.  Most simple wordsigns were
  planned to be represented by their first letter to help lessen the learning
  load.  It is not always easy to find a suitable replacement if this
  mnemonic help is to be retained.  On the other hand braille is limited by
  the number of signs available.  1895 had included such words as
  RECEIVE, DECLARE, REJOICE, and the 1905 version continued the
  trend, sometimes seeming to prefer a collection of similar words instead
  of finding ones of more value.  However, the committee members did not
  have the use of the extensive frequency lists that are available today. 
  1905 presented a great opportunity and some good decisions were made
  but insufficient knowledge and some subjective choices led to inclusion
  of some uneconomic contractions which remain in the code today.
  
  1932 "Standard English Braille"
  The Americans were hoping that in accepting Grade 2 there would be a
  reduction of some of the contractions (Irwin, 1931, p.33).  Only 9
  religious words were eliminated.
  
  1960
  Only four new contractions were introduced (British UTC, 1960).
  
  The BAUK questionnaire had indicated that a majority of braille readers
  preferred no further changes to contractions in words for the time being. 
  Vocabularies, codes and technical advance never stay still so the last
  chapter will include more information about the possible future of the
  English literary braille code.
  