Fun with UEB Indicator Scopes - SusanJ/UEB_back GitHub Wiki

Indicators are the braille equivalent of markup. UEB has more than 50 indicators most of which come in sets of four: three start indicators and one terminator. The three start indicators in a set each have different scopes. The symbol version applies only to the next symbol; the word or sequence version applies only to the next item; the passage version applies to several, usually at least three, words or sequences. The three start indicators also have different means of being terminated. The symbol one is always terminated implicitly after its immediately following braille symbol. The word one is terminated several different ways depending on the situation. It is often terminated implicitly by the end of the word or by other context specific to the type of indicator but can also be terminated explicitly by the terminator when required. The passage one is always terminated explicitly by the terminator.

This implementation of indicators differs from what is used in XML-based markup languages. XML markup tags come in sets of two: a start tag and a matching end tag by which the effect of the start tag can always be terminated independently of the size of the run of affected text.

The purpose and use of indicators in UEB amd other braille systems is essential to compensate for the small number of braille cells in contrast to the large number of print characters in even simple print documents. However, braille indicators is too big a subject to cover in a single short article. As the title suggests, the focus of this article is indicator scoping rules. Those braille indicators used to represent different print typeforms make a good introduction to this topic because thier scoping rules are similar to those of other braille indicators but, unlike more specialized braille indicators such as the numeric indicator, they have print analogues that should make examples easier to follow.

[Although this Wiki is intended primarily for persons familiar with braille, it is probably worth pointing out that braille typeform indicators aren't intended for rendering braille source in different way as, for example, the way markup in HTML source is used by a browser. This is because the nature of tactile reading necessitates consistency in the rendering of the braille cells so physically different renderings aren't used. In other words, any typeform (as well as other) indicator symbols are simply part of the braille text to be read and interpreted by the braille reader. The intent of having different versions of an indicator is to facilitate reading by minimizing the clutter associated with the unavoidable use of indicators.

Although it is not correct braille to not use a passage indicator wherever appropriate, using several word indicators instead of a passage indicator simply adds clutter; it doesn't lead to an incorrect interpretation. Although there are probably fewer situations in informal braille such as email where a passage indicator could be used, in my opinon avoid passage indicators in informal braille is entirely acceptable.

On a related note, I've been asked by persons unfamiliar with braille why bother with typeform markup if it doesn't affect the appearance of the braille. There are of course many answers. One is that special typeforms sometimes have a semantic purpose. Another is that there are various reasons why a braille user might want a facsimile transcription of a print document.]

The differencs between the use of XML-style typeform markup and UEB typeform indicators is illustrated in the following examples of bold typeform markup. The first four examples show HTML-like source followed by a typical visual rendering.

(1) A <b>v</b>ery nice day. could be rendered as: A very nice day.

(2) A <b>ve</b>ry nice day. could be rendered as: A very nice day.

(3) A <b>very</b> nice day. could be rendered as: A very nice day.

(4) A <b>very nice day.</b> could be rendered as: A very nice day.

Next the same four examples are shown in pseudo-uncontracted braille where the letters SS, WW, PP, and TT represent the Symbol, Word, Passage and Terminator braille indicators respectively. (Most braille indicator symbols use several braille cells; UEB bold typeform symbol indicators each use two braille cells.)

(1) A SSvery nice day.

(2) A WWveTTry nice day.

(3) A WWvery nice day.

(4) A PPvery nice day.TT

Although the use of different versions depending on the scope of a braille indicator is intended to make reading braille easier for a human, it adds to the difficulty of automating the backtranslation of braille to print as well as automating the translation of print to braille. (Other indicator-related issues also add to the difficulty of automating backtranslation and translation.)

Scoping rules for typeform indicators affect the automated backtranslation of braille to print primarily because of the need to devise and implement algorithms for recognizing which start indicators have implicit scopes and for determining where to insert the desired corresponding explicit print terminating markup. Braille symbols for the typeform start indicators and termination indicators can of course be directly replaced by the corresponding print markup.

Scoping rules affect the automated translation of tagged print source to braille in at least two ways. In the first place, the amount of affected text has to be taken into account in order to determine which version of the corresponding braille start symbol to use. (For certain other types of indicators the nature of the affected text also has to be taken into account.) Second, in those cases where a word-scoped indicator is the appropriate one, it is necessary to determine whether or not an explicit termination indicator is required since it is not correct braille to use a termination indicator when not required.