- From: Joe Roeder <Jroeder@nib.org>
- Date: Fri, 30 Jan 1998 14:28:47 -0500
- To: Matthew Lye <mlye@trentu.ca>
- Cc: WC3-WAI-IG <w3c-wai-ig@w3.org>
I have been following this thread and would offer a comment. On Friday, January 30, 1998, you wrote: >My thinking is that a person familiar with the context in which 'RAM' is >being used might find that really irritating. But I gather that you are >thinking of a mark-up to browser to screen reader chain, rather than >direct parsing of the mark-up by a text-to-speech browser. In that case >I almost agree with you, except that for people reading the acronym >visually, R space A space M would be less than optimal. Thus, I would >suggest that the HTML 4.0 specification should instruct that browsers >support optional acronym spacing. Most screen readers have tools for dealing with acronyms/abbreviations by either spelling out them out or making a verbal substitution. Even without these automatic features, the user usually has a "spell word" command to apply to the text which can accomplish the same objective. In my experience, there are no universal rules here. Some abbreviations/acronyms are more understandable when parsed (spelled out) and others are not. When people read text aloud, they will usually choose one or the other according to which result is more understandable. For the most part they are following conventions rather thatn firm rules. For example, everyone will pronounce "ROM" as a word but spell out "IBM". Good screen readers have pronunciation exception dictionaries which are fine for applications that have a limited number of abbreviations/acronyms on the screen. When dealing with documents or web pages, however, where an abbreviation/acronym will be encountered just once, or at most, only a few times, it is not practical to take the time to add a new pronunciation exception to the dictionary. Parsing or spelling out every abbreviation/acronym is not the solution because this impedes the flow and rhythm of the spoken text and becomes tedious. Pronouncing every abbreviation/acronym as though it were an authentic word is not the solution either, because some letter combinations are too difficult to discern this way. Your suggestion of having an optional parsing built into the browser will probably be beneficial for those screen readers that don't already have features for dealing with abbreviations/acronyms and unnecessary for screan readers that do. Most important is that this be an option or it may short-circuit the screen readers functions that are already designed to handle them.
Received on Friday, 30 January 1998 14:36:28 UTC