Proposal for Guideline 3.1 (“meaning”)

Draft 2005-05-312005-06-06
Submitted by John Slatin

Reworded proposals

The following are modified versions of SC in my original proposal.

• Delete L1 SC2 (A mechanism for finding the expanded form of acronyms and abbreviations is available.); delete current wording (The meaning of acronyms and abbreviations can be programmatically located)
  Rationale: Discussion of the proposed SC on the 2 June telecon made several points that seem to argue against treating acronyms and abbreviations in a separate success criterion: (1) acronyms are defined as words formed from the initial letters of other words. As such, familiar acronyms are likely to appear in comprehensive online dictionaries and would thus be covered under the proposed L2 SC1 (a mechanism for finding definitions is available). This would likely apply to acronyms whose original meaning has disappeared (e.g., "IBM" no longer = "International Business Machines"). (2) Unusual acronyms and abbreviations could be covered under L3 SC1 (adopted 26 May). (3) For some technologies, use of markup to identify acronyms and abbreviations and provide appropriate expansions is covered by GL 1.3. (4) Finally, there are languages for which acronym is not a meaningful concept. So I propose deleting the SC and instead providing appropriate advice in the Guide doc.
• Move L3 SC3 to L2 and modify as follows:
• [Updated 6 June] When content is divided into sections, the sections have descriptive titles.
  Rationale: The previous proposal says "Link text and section titles make sense when read by themselves or as a group." Discussion on the 2 June telecon raised questions about the distribution of responsibility between the author and the user agent. The user agent is responsible for extracting link text and headings from Web content. The author is thus responsible for providing the user agent something meaningful to work with. But link text is already addressed by GL 1.1 L1 SC1 (for graphical links, for example) and GL 3.2 L2 SC6 (for text links). So this SC need not concern itself with links. Use of markup to define semantically meaningful sections is covered under GL 1.3, but I'm not convinced that GL 1.3 requires that section titles be semantically appropriate. So I propose modifying the SC as shown above.

• L2 SC3

• [Updated 31 May and 6 June] One or more of the following supplements is available when text requires reading ability at the upper secondary level:

  1. a text summary that can be read by adults who have completed primary education.
  2. Graphical illustrations of concepts or processes that must be understood in order to use the content.
  3. A spoken version of the text content.

  Rationale for 6 June change: This is an attempt to make the SC more readable and understandable. Comments on the list indicated that the 31 May version was hard to understand ("One or more of the following supplements is available..." should be easier to parse than the 31 May wording ("One or more of the following are available as supplements when ..."). I hope "a text summary that can be read by adults who have completed primary education" is easier to understand than "a text summary that can be read by adults with the reading ability expected of native speakers who have completed fewer than seven years of school"!. Finally, I've substituted the names of education levels defined in the International Standard Classification of Education for the specific numbers used in previous versions of the proposals. Discussion on the list indicated that "10 years of education" means significantly different things across countries and even within individual countries. Thus the proposed terms, while unfamiliar, may be more useful in helping readers understand the (proposed) requirement.

• Rationale for 31 May changes: Discussion on the list suggested that the use of the term "alternative versions" in the original proposal might have caused confusion. Some people appeared to be concerned that graphical illustrations, etc., would be required to be presented instead of the text. I hope the new wording makes it clear that summaries, graphical illustrations, and spoken versions are meant to appear in addition to the text.

• L3 SC2

• [Updated 25 May] A mechanism is available for finding the correct pronunciation of any word whose pronunciation cannot be determined from context.

Rationale: Relieves the burden of requiring pronunciation support for all words in content. Raises new question of how to test whether pronunciation can be determined from context.

Proposed SC on which consensus was reached on 26 May

• L1 SC1: [Adopted] The primary natural language or languages of the content can be programmatically determined.

• L1 SC3 : [Rejected] A measure of the education level required to read the content is available.
• L3 SC1: [Adopted] A mechanism is available for identifying specific definitions of words used in an unusual or restricted way, including idioms and jargon.

Anything in the full proposal below that is not marked as having been adopted or rejected remains to be discussed. The updated items listed above appear below as well, so that they can be read in context.

<proposed>[Adopted 2 June] Guideline 3.1 . Make text content readable and understandable.

Level 1 success criteria for Guideline 3.1

1. [Adopted 26 May] The primary natural language or languages of the delivery unit can be programmatically determined.

   Guide to GL 3.1 L1 SC1

2. A mechanism for finding the expanded form of acronyms and abbreviations is available.

   Guide to GL 3.1 L1 SC2

3. [Rejected 26 May]A measure of the education level required to read the content is available.

   
   Guide to GL 3.1 L1 SC3

Level 2 success criteria for Guideline 3.1

1. A mechanism is available for finding definitions for all words in text content.

   Guide to GL 3.1 L2 SC1

2. A mechanism is available for identifying the natural language of each foreign passage or phrase in the content. Foreign passages or phrases are written in a language that is different from the language of the delivery unit as a whole.

   Note: This requirement does not apply to individual words or phrases that have become part of the primary language of the content. This is because "correct" pronunciation of such words and phrases might confuse or distract native speakers of the content’s primary language.

   Guide to GL 3.1 L2 SC2

3. [Updated 31 May] One or more of the following is available as a supplment to text content that requires the reading ability expected of native speakers who have completed at least nine years of school:

   1. A text summary that can be read by adults with the reading ability expected of native speakers who have completed fewer than seven years of school.

   2. Graphical illustrations of concepts or processes that must be understood in order to use the content.

   3. A spoken version of the text content.

   Guide to GL 3.1 L2 SC3

Level 3 success criteria for Guideline 3.1

1. [Adopted 26 May] A mechanism is available for identifying specific definitions of words used in an unusual or restricted way, including idioms and jargon.

   Guide to GL 3.1 L3 SC1

2. [Updated 25 May] A mechanism is available for finding the correct pronunciation of any word whose pronunciation cannot be determined from context.

   Guide to GL 3.1 L3 SC2

3. Section headings and link text are understandable when read by themselves or as a group (for example in a list of links or a table of contents).

   Guide to GL 3.1 L3 SC3

4. A mechanism is available to identify text that states important ideas or provides important information.

   Guide to GL 3.1 L3 SC4

5. For delivery units at the first or second level in a set of delivery units, text content can be read by adults with the reading ability expected for native speakers who have completed fewer than seven years of school.

   Guide to GL 3.1 L3 SC5

6. For delivery units below the second level in a set of delivery units, text content can be read by adults with the reading ability expected for native speakers who have completed fewer than nine years of school.

   Guide to GL 3.1 L3 SC6

7. Simplified graphical illustrations are available when text presents ideas or describes processes that users must understand in order to use the content.

   Guide to GL 3.1 L3 SC7

8. A spoken version of text content is available.

   Guide to GL 3.1 L3 SC8

9. Signed video is available for key pages or sections of pages.

   Guide to GL 3.1 L3 SC9 (in progress)

Changes following the 26 May 2005 WG call

• L1 SC 1: Adopted; "programmatically determined" replaces "A mechanism is available"

• L1 SC3: Rejected

• L3 SC1: adopted, with proviso to work on processes for testing "unusual" and "restricted"

Change introduced on 25 May 2005

• Not yet discussed by WG: L3 SC2 (pronunciation( has been changed so that it now requires a mechanism for finding the "correct pronunciation" of "any word whose pronunciation cannot be determined from context" instead of requiring pronunciation support for "all" words in text content. This still isn't quite right: it introduces a new problem of testing whether pronunciation can be determined from context. But pronunciation support is important and I think this is better than my original proposal requiring support for all words in the content.

Differences between draft of 21 May 2005 and draft of 30 April

• L1 SC1 now refers explicitly to the "primary natural language or languages" of the content. This responds to questions raised by Wendy and Tim Boland; the I18N Workign Group notes that the lang and xml:lang attributes may take comma-separated language identifiers for documents that have multiple primary languages, e.g., Canadian documents in which English and French have equal place. Note that identifying multiple primary languages still requires that language changes within the body of the content are identified when they occur (as per L2 SC2; we may need to promote L2 SC2 to L1 for this reason).
• L1 SC3 now requires a readability measure of the text content ("measure of the education level required to read the content") instead of calling for a description of the education level of the intended audience. I believe this addresses Jason's concern about the need to provide precise data about actual educational attainment within the target audience by refocusing the SC on the content instead of the audience.
• L3 SC1 changes "intended definitions" to "specific definitions" of words used in unusual or restricted ways. This avoids the problem of seeming to require a test of authorial intent, and I think addresss Tim Boland's concern.

• Brief summary of differences between this proposal and the current wording

Relatively small changes

• Several SC were rewritten to describe functional outcomes, as per LA decision and baseline analysis: L1 SC1, SC2; L2 SC1, SC2; L3 SC 1, SC2

• L2 SC1 (meaning and pronunciations) is broken into two SC and the SC about pronunciation information has been moved to L3

• L2 SC2 (idioms) has been moved to L3 and merged into L3 SC1. Rationale: L3 SC1 deals with words used in highly specific ways. An idiom is a word used by native speakers in a way that breaks the bounds of the dictionary definition, so I think that idioms and jargon both qualify as instances of the more general category.

Major changes

1. L3 SC3 (statement asserting that the following list of strategies for reducing complexity has been considered) is deleted.

A number of new SC have been introduced. The primary goal was to replace L3 SC3 with meaningful and testable success criteria that would promote readability and make understanding easier for people with a range of disabilities, including reading disabilities.The new SC are:

1. L1 SC3: requires a description of the education level of the intended audience for the content.

2. L2 SC3: requires one or more alternative versions (including optional non-text alternatives) for text content that requires education level at or above 10th grade (US), 10 years in school/upper secondary level international classification.

3. L3 SC2: requires pronunciation information. This one was originally included in L2 SC1; has been separated from the issue of definitions and moved to L3 because it seems significantly more difficult to provide.

4. L3 SC4: requires a mechanism for identifying most important points in text content.

5. L3 SC5: requires that text on first- and second-level pages is readable at 6th grade level (US)/end of primary education international classification L3 SC6: requires that text below second-level is readable at 8th grade level (US)/late lower secondary international classification

6. L3 SC7: requires simplified graphical illustrations of important ideas/descritions of processes (this is an option at L2, required at L3

7. L3 SC8: requires spoken-word version of text content (also an option at L2 that becomes a requirement at L3

8. L3 SC9: requires signed video for key pages or passages