by Abraham Nemeth
The California Transcribers and Educators of the Visually Handicapped (CTEVH) held its annual conference on April 26-28, 2001 in San Francisco.
What follows is what I prepared to say when I participated in panel discussions concerning the Unified Braille Code (UBC). We were limited to a 15-minute presentation.
I have been a CTEVH member since 1958. I have been intimately associated with the Unified Braille Code project (UBC) ever since its inception in 1991. That code has two clearly defined aspects: its literary aspect and its technical aspect. But examples of the code as it has emerged have been nearly exclusively limited to the literary aspect of that code.
The UBC proponents are eager to point out with pride that one can hardly tell the difference between what has been written in the UBC and what has been written in standard Grade 2 braille. Their point seems to be: Don’t be afraid; UBC is not that bad! A very natural question that arises, at least in my mind, is “If there is so little difference, why bother to change? Why commit hundreds of thousands of dollars, perhaps millions, and why plan such an elaborate and comprehensive retraining program to institute a change that is hardly noticeable?” The answer, of course, lies in the technical aspect of the UBC, which bears no relation to the braille codes with which we are familiar in this country, and very little of which has been shown to us.
The technical part of the UBC is based on the use of upper numbers. In general, numbers in a literary context are widely separated and are not, for the most part, associated with adjacent letters. Therefore, upper numbers with their associated number signs in a literary context are easily tolerated. But in mathematics, numbers and letters are generally crowded together. This entails such frequent use of the number sign and also the letter sign that the resulting transcription is hardly tolerable since these braille signs constitute a constant interruption to the smooth flow of the notation in which they are embedded. I have pointed this out in the several papers that I have written regarding the UBC; many of you have seen some of those papers.
One of the examples I used to make this point was a multiplication example that might be encountered in an eighth- or ninth-grade elementary algebra course. This example is set out in five lines of braille. In the UBC implementation, this example requires 27 number signs and 18 letter signs. Since there are 6 plus signs and 4 minus signs in this example, each of which is a two-cell symbol beginning with dot 5, there are also ten dot-5 prefixes in addition to the 27 number signs and 18 letter signs. It is important to point out that this example will not fit across a standard 40-cell braille line even when it begins at the left margin. The five braille lines in this example contain 130 braille characters. The 55 braille characters for the number signs, the letter signs, and the dot-5 prefixes constitute 42 percent of the total braille characters in this example.
On the other hand, this same multiplication problem, if presented in the Nemeth Code, contains no number signs, no letter signs and no dot-5 prefixes. It easily fits across a braille line with a two-cell indentation, as is customary for such problems, and can be further indented to accommodate an identifying number if need be.
Despite my repeated requests, no one championing the UBC has shown me how to fit the UBC example on a standard 40-cell braille line. The use of upper numbers makes the UBC massive beyond reason.
The craftsmen of the UBC have turned a deaf ear to all the protests aimed at them concerning the mischief that upper numbers create. All the reasons they adduce concerning the advantages of upper numbers with a fervor that borders on paranoia are nothing but rationalizations. I believe that the real reason for espousing the use of upper numbers is entirely political. The proponents know that they could not possibly enlist the participation of our overseas colleagues if they did not advocate upper numbers.
Braille As You Like It Ad Hoc Consortium Proposes a Different Universal Braille System (UBS)
Unable to persuade the project committee to budge from its entrenched position, a group of us formed an e-mail listserv which we call “braille_zylx.” (I devised the “zylx” part; it is bad Grade 2 braille for “as you like it.”) Our goal on this list is to produce a counter-code based on dropped numbers. This counter-code will be a near clone of the Nemeth Code. I have already posted several chapters on the list. We have close to 100 subscribers on this list; and the code, which we call the “Universal Braille System” (UBS), is being developed with the advice of this group.
Everyone is welcome to join the “As You Like It” listserv. To subscribe, send a blank e-mail message to: majordomo@telepath.com. Leave the subject field of your message blank. In the body of the message write the following: “Subscribe braille-zylx.” That’s all there is to it.
One of the archives that I have prepared is called “A UBC Anthology.” It is a collection of all the papers and letters regarding the UBC, including e-mail messages, that have come to my attention since the inception of the project to about January 2000. It fills about a ream of paper in print. Although it is too voluminous to permit individual distribution, I have made it available for inspection by anyone who cares to read it.
Another document that I have prepared is a transcription into braille of the first chapter of a freshman calculus book. First, I transcribed the chapter into the Nemeth Code; then I converted the Nemeth Code version into the Unified Braille Code; and finally, I converted that into our Universal Braille System. Page after page, the two versions appear side by side in simulated braille; the UBC version on the left-hand page and the UBS version on the right. Each pair of facing pages begins and ends with the same word or notational element for purposes of comparison. The Nemeth Code version filled 87 braille pages. This was expanded in the UBC version to 97 braille pages, an increase of about 11.5 percent. Accompanying all this is a photocopy of the printed text from which the original transcription was made. Figures were omitted, although references to them were retained.
I have also prepared a statistical table comparing four important features in both codes. For each version, these features are: the number of number signs per page, the number of letter signs per page, the number of punctuation indicators per page, and the number of braille lines per page. It is of interest to note that, in the entire braille volume, there are 155 punctuation indicators in the UBS version. Compare that with the 55 number signs, letter signs, and dot-5 prefixes concentrated on just five braille lines in UBC notation. All of this material is available for your inspection, either in the volume described above or in an interpoint version which takes up less shelf space. There is no privileged information in any of these documents, and they may be freely copied and disseminated.
Technical Aspects of the UBC Cause Confusion, Misalignments, Inconsistencies, Difficulties Representing Superscripts
Apart from numbers and letters, the core symbols in mainstream mathematics are the signs for plus, minus, equals, fraction line, and two parentheses. Although I have no statistical evidence, my long experience tells me that these six signs collectively constitute more than 90 percent of mainstream mathematical notation, apart from numbers and letters. However, all the operation signs in the UBC mimic those in the Braille Authority of the United Kingdom (BAUK) math code (the code used in England) except that the prefix has been changed. They are all two-cell symbols.
This adds to the massiveness of the notation that the UBC creates.
By contrast, the UBS generates no letter signs, very few number signs, and the six symbols to which I have alluded above are all one-cell symbols.
In the UBC, the use of upper numbers can cause misalignment. A number like 1234 requires one number sign and no letter signs; it occupies 5 cells. On the other hand, a number like 2a3b requires two number signs and two letter signs and occupies 8 cells. Regarded as 4-digit hexadecimal numbers, these cannot, in their raw form, be aligned for addition.
To circumvent this problem, UBC has devised an “alignment mode” in which the unfamiliar dot-6 or French number system is used. (Coincidentally, the 63 symbols in the alignment mode are a 100 percent clone of the British computer braille code.) Consequently, this “Unified Braille Code,” which was ostensibly devised to simplify and to eliminate multiple representations, now contains two number systems.
Some alignment problems entail floating-point numbers in which superscripts are required. However, the UBC alignment mode makes no provision for representing a superscript.
In UBC, enclosure symbols are also sometimes required. But alignment mode makes no provision for these enclosure symbols either. It has been proposed that symbols that do not participate in a current mathematical expression be “deputized” to serve as a superscript symbol or an enclosure symbol for the duration of the current example.
In the absence of any guidelines for selecting suitable candidates to be deputized, each transcriber will make his or her own choice about which particular symbol to “deputize.” So much for uniformity. It is unclear by what incantation a chicken can be turned into a fish. In addition, a transcriber’s note requiring at least two lines of braille for identifying to the reader what symbols have been chosen, plus a skipped line above and another below this transcriber’s note would add four braille lines to the transcription. This clumsy solution occupies the low end of the “elegance” spectrum.
Some proponents of the UBC claim that hexadecimal alignment is incredibly rare. We are told not to throw out a whole code on that account. Hexadecimal computations are not as rare as one might think. Every introductory textbook in computer science has a chapter on hexadecimal arithmetic. And programmers who work in assembly language must continually address calculations involving hexadecimal arithmetic. Saying that something is incredibly rare is a form of denial. Blind people are not exempt from dealing with rare situations; when they occur they must be dealt with, and blind people need the tools to deal with them.
Representing numbers with two symbols which involves utilizing a separate number symbol, can lead to other subtle unintended consequences. Consider the word “34-cent” as in a thirty-four-cent stamp. The first character is the number sign which initiates both numeric mode and Grade-1 mode. The next two characters are then correctly interpreted as the digits 3 and 4. Then comes the hyphen. This causes numeric mode to be turned off, but Grade-1 mode remains in effect. As a result, the next three characters are interpreted as “c subscript t,” rather than as “cent.” To avoid this, a Grade-1 terminator must be inserted either before or after the hyphen. This terminator is a two-cell symbol formed by dots 5-6, 3. Now the last three characters can be read as “cent.” It takes a two-cell Grade-1 terminator to permit the “en” contraction which — after all of this —saves just one cell! Whenever a word is composed of a number and a word, for example in an expression like “3-dimensional,” where the number and the word are joined by a hyphen, we have the same situation.
You will no doubt be told that such situations can be remedied by changing or adding a few rules. Remember, young children will have to cope with all these UBC idiosyncracies. Upper numbers have already been given so many booster shots and have been so bandaged up that they are on life support. It is time to pull the plug.
Some people may tell you that switching to lower numbers is too radical a departure from English Braille Grade 2. On the contrary, such a switch would be a non-event. Braille users are already doing so privately both in reading and writing. Lower numbers have long since become a de facto numbering system, and no one would be inconvenienced by the switch.
Some maintain that a general-purpose code like the UBC cannot be as efficient as a subject-specific code. The UBS proves otherwise. That code preserves the current literary braille code far more than does the UBC. The UBS does not abolish sequencing and it does not eliminate any contractions as does the UBC, and it is extremely efficient. Any seasoned transcriber will tell you that, if lower numbers were used everywhere, he or she could transcribe anything from Chaucer to quantum mechanics, or from Hamlet to nuclear physics using the Nemeth Code or its near clone, the UBS.
There are many other problems with the UBC that I cannot present in the limited space which “The Braille Forum” can allocate to the topic. There are basic problems with fractions and there are basic problems with level indicators. With regard to the UBC, my identity as the creator of the Nemeth Code is a handicap. Regardless of the logic of my position, there will always be people who will dismiss it on the perception that I am defending personal turf. So it is up to you to assess the true weight of the logic behind my position.
Where is the UBC constituency? The proponents that I know of can be counted on the fingers of one hand. No one else has made a supporting public statement either verbally or in writing to my knowledge. At the 1995 ICEB London Conference, no one submitted a paper in support of the UBC in spite of the fact that UBC was the principal topic of discussion at that conference.
I harbor no enmity to anyone. I consider the people who support the UBC my friends and colleagues. It is their code, not its sponsors that I oppose. I am ready and willing to contribute whatever abilities I have to the formulation of a uniform braille code; but that code cannot be based on upper numbers, given all the difficulties with them that I have set before you above.
Shall we stand by and agonize over the broken spirits of one young blind child after another who finds it impossible to cope with this UBC in any meaningful way?
Many presidents, the most recent being Bill Clinton, have repeatedly admonished the Congress that if certain bills come to their desks, they will veto them. We, the readers, writers, and transcribers of braille, must admonish BANA that if this UBC comes to us after adoption, we will ignore it. We are organized in many organizations, to make the blind children for whom we work our beneficiaries, not our victims. I urge all who care about braille, about literacy, about the future of children who are blind and visually impaired, to make clear to BANA at this time that we reject the UBC.
For More Information
Should you wish to contact me on this subject, here is my identifying information: Abraham Nemeth, 24111 Civic Center Drive, Apt. 409, Southfield, MI 48034; e-mail anemeth@ece.eng.esumr.edu.
To subscribe to the Braille-As-You-Like-It listserv, simply send a message to majordomo@telepath.com, following the instructions above.
Final Thoughts
Editor’s Note: We are honored that Dr. Nemeth chose to share his thoughts with readers of “The Braille Forum,” and I am especially indebted to Winifred Downing, who so graciously shared the original article with me and made many helpful editorial suggestions.
As one who considers myself a “survivor” of college-level calculus, and as a person who has been reading braille for fewer than 10 years, I found many of the concepts included in the article above to be challenging to say the least. Yet I felt it important to include many of these very communicative examples, and I hope that I have done Dr. Nemeth's original content no disservice.
For an interesting look at the origins of the Nemeth Braille Code for Mathematics and Science, and the history of the Unified Braille Code and the Unified English Braille Code Projects, I recommend the recently published “Braille Into the Next Millennium,” available in print, braille and on audiocassette from the National Library Service. In addition, see “Where Shall We Go with the Unified Braille Code?” by Christopher Gray in the July “Braille Forum.”
At its 40th annual convention in July 2001, the American Council of the Blind adopted the following resolution:
Resolution 01-27
WHEREAS, the adoption of a Unified English Braille Code (UEBC) has tremendous potential for creating the availability of standard English Braille which can be freely shared among all English-speaking nations; and
WHEREAS, the International Council on English Braille (ICEB) has proposed to adopt such a code during the current quadrennium; and
WHEREAS, premature adoption of the UEBC would result in the failure to provide adequate opportunity for Braille users in North America to examine, comment upon, and verify the capacity of the code to express advanced technical material in a comprehensible manner;
NOW, THEREFORE, BE IT RESOLVED by the American Council of the Blind in convention assembled this 7th day of July, 2001 at the Polk County Convention Center in Des Moines, Iowa, that this organization vigorously oppose the adoption of the Unified English Braille Code if proposed for adoption at the 2003 quadrennial meeting of the ICEB; and
BE IT FURTHER RESOLVED that this organization strongly urge that the ICEB postpone a final decision concerning adoption of the UEBC until its 2007 quadrennial meeting; and
BE IT FURTHER RESOLVED that this organization not endorse the UEBC until satisfied that wide-ranging input has been received, considered, and evaluated from consumers in North America; and
BE IT FURTHER RESOLVED that the ACB urge the board of the ICEB to consider papers prepared by Abraham Nemeth and Christopher Gray, and review the proposals of the Universal Braille System (UBS), before making a decision on a unified code.