[text] [DRAFT] New info and clarification on figcaption

DRAFT for discussion. Note that the central 
question addressed below is whether the user need 
for access to certain kinds of information is met 
if alt is dropped in the presence of figcaption. 
I understand there may be additional questions, 
for instance regarding rendering, that may be 
valuable to discuss; but would like to see where 
we are on the core question of user requirements.

- Judy

[draft starts]

This mail presents new information objecting to 
the figcaption exemption, and proposes a 
different evaluation of certain rationales.

The HTML Co-Chairs' decision on alt validation, from Maciej
http://lists.w3.org/Archives/Public/public-html/2011Apr/0451.html
asked the question



>== Should it be permitted to omit alt when the 
>image is in a figure with a figcaption? ==

and concluded that



>The presence of figcaption makes missing alt conforming.

The decision explored a variety of rationales for 
and against a "figcaption exemption," e.g., that 
alt can be omitted in the presence of figcaption. 
Summarizing, these rationales were evaluated as follows:

* Strong objection to figcaption exemption:
- the use cases objecting to the title exemption 
were considered overall to also constitute 
relevant objections to the figcaption exemption;

* Weak objection to figcaption exemption:
- lack of figure/figcaption support in current 
assistive technologies (this was evaluated as 
weak because no specific products were mentioned 
and because assistive technologies may support 
figure/figcaption in the future);

* No objection to figcaption exemption:
- title and figcaption are non-redundant mechanisms;
- a placeholder alt value could be provided when figcaption provides a caption;
- saying that one can drop alt in the presence of 
figcaption complicates the usage of alt;

* Weak support for figcaption exemption:
- [nothing listed];

* Stong support for figcaption exemption:
- excessive requirements will be perceived to 
lower the seriousness of all requirements and 
requiring information that authors are likely 
to     consider duplicate information is likely 
to be perceived as an excessive requirement (this 
was evaluated as strong although no evidence was presented);

* Overall evaluation of rationales:
The single strongly-evaluated rationale 
supporting the figcaption exemption -- that 
requiring alt in the presence of figcaption might 
be considered excessive, and then that excessive 
requirements might be perceived to lower the 
seriousness of all requirements -- was apparently 
evaluated to override all other strongly- and 
weakly-evaluated objections listed above, even 
though no evidence was presented for this doubly-speculative assertion.


Following is new information and a proposed 
re-evaluation of certain rationales.

1. THERE IS GREAT VARIATION IN THE KINDS OF 
INFORMATION THAT FIGCAPTION WILL CONTAIN, AND 
MANY FIGCAPTIONS WOULD BE INAPPROPRIATE AS ALT SUBSTITUTES

Some figure captions are terse, contain info that 
is relevant for a minimal visual description, and 
would be acceptable to substitute for an alt. For instance:
http://www.sciencefriday.com/program/archives/201104297#innerdl
shows an image of the Allen Telescope Array, with 
an alt of "Array" (should be better), a title of 
"Allen Telescope Array SETI Institute image" and 
a <p></p> below the image which displays as a 
figure caption. In HTML5, the same text would 
presumably be appropriate for figcaption, and also for alt.

However, the following figure caption meets the 
conventions for and is typical of figure captions for scientific publications:
http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1000612#pbio-1000612-g001

>TMRM or MitoTracker-stained PC12 cells (A–C) or 
>hippocampal neurons (D) were treated for the 
>indicated times with either vehicle or the 
>listed compounds and mitochondrial morphology 
>was quantified by blinded comparison to 
>reference images (B–D, means ± S.E.M.). (A) 
>Representative epifluorescence images show 
>formation of interconnected mitochondria upon 
>treatment of PC12 cells with forskolin/rolipram 
>(forsk/roli, 20 µM/1 µM, 3 h). In the 
>representative experiment shown in (B), rapid 
>mitochondrial elongation by 50 µM forskolin is 
>not affected by 100 µg/ml cycloheximide to 
>inhibit protein synthesis. Long-term (20 h) 
>forskolin or a cell-permeant cAMP analog (200 µM 
>cpt-cAMP) promotes mitochondrial fusion in PC12 
>cells (C) and hippocampal neurons (D); summaries 
>of three independent experiments with 20–30 cells per condition are shown.
>doi:10.1371/journal.pbio.1000612.g001

Breaking this down, the figure caption contains 
detail on how the cells were treated; and, for 
each of the individual images that are part of 
this composite image, some interpretation of what 
the individual images demonstrate. Such figures 
typically include details about how the samples 
were processed and how they were statistically 
analyzed (sometimes non-redundant with the text), 
with little detail about their visual appearance 
since the authors assume that the reader can see 
the images. Yet information that is visually 
discernable is essential to evaluating the 
scientific conclusions of the research presented, 
and an important initial aspect of that is 
identification of the kinds of information presented.

An appropriate alt for this image, in addition to 
being terser, would contain a crucial bit of 
information that is visually discernable but not 
mentioned anywhere in the figure caption above, 
nor easily findable within the text of the 
article: "Epiflorescent images of interconnected 
mitocondrial, with mito-length scores from three 
experiments." A blind user would be deprived of 
important information otherwise obtainable 
through alt if they had to rely solely on the 
figure caption, though this text is appropriate as a figure caption.

There is great variation in the kinds of 
information that figure captions contain, 
impacting their appropriateness as alt 
replacements. Even carefully composed figure 
captions for images used to enhance the theme or 
subject matter of page content are sometimes 
inadequate to provide a blind user with even 
minimal information about the image, and would 
not be appropriate substitutes for alt. For 
instance, an image accompanying an article 
highlight for the April 2011 PLOS Biology Journal
http://www.plosbiology.org/article/browseIssue.action?issue=info%3Adoi%2F10.1371%2Fissue.pbio.v09.i04
is accompanied by the following figure caption:

>On the functionality of reef diversity: Existing 
>experimental knowledge suggests that in diverse 
>ecosystems, given the redundancy in the 
>ecological roles of species, the efficiency of 
>ecosystem processes stabilizes after a certain 
>number of species is reached, and that the 
>functioning of such processes is not 
>considerably impacted when few species are lost. 
>A new field study shows evidence of the 
>contrary. 
><http://dx.doi.org/10.1371/journal.pbio.1000606>Mora 
>et al. (e1000606) show the existence of 
>accelerating relationships between biodiversity 
>and ecosystem functioning in coral reefs 
>worldwide and conclude that the functionality of 
>the most diverse reef systems is the most 
>impaired by the deleterious effects of human populations worldwide.

Appropriate use of alt is discussed in the Lady of Shallot examples
http://dev.w3.org/html5/alt-techniques/#images-enhance
and, in for the example above, would advise an 
alt at least imparting the information that the 
image shows a reef, for instance: "A reef showing diverse species."

Given the variation in information that is 
included in figure captions, it seems inaccurate 
to generalize that all figcaptions are 
appropriate substitutes for alt, as the decision 
that "the presence of figcaption makes missing alt conforming" would imply.

Depending on the type of image and the type of 
publication, figure captions may be either 
concise or verbose. When information contained in 
figcaption is detailed and complicated, it is 
more similar to that supplied by the current 
longdesc attribute than to that supplied by 
alt.  alt, on the other hand, is normally brief, 
and identifies the image rather than fully 
describing it, especially when the image is 
complex.  Permitting figcaption to take the place 
of alt will in some situations result in more 
information being delivered to the user than the 
user needs or wants. The user should be able to 
access the information in figcaption, but not be 
deprived of the type of information they would normally receive through alt.

In addition, figcaption may not fulfill the needs 
of assistive technology users, particularly blind 
or visually impaired users/screen-reader users, 
as it may _omit_ information that the user needs. 
Figure captions describe images that users _can_ 
see. In contrast, alt and longdesc attributes 
identify and/or describe images that users 
_cannot_ see. The two audiences are different, 
and as such may require different approaches for 
image description. For instance, in scientific 
publications, information presented in figure 
captions will often state the scientific 
principles being illustrated, but not describe 
the illustration nor necessarily even identify 
the image since many authors assume that their 
audience can identify and discern information 
that is presented visually. To adequately support 
the needs of blind or visually impaired users may 
require description of visually discernable 
information that sighted users would object to as 
redundant if presented in a visible description via figcaption.

One could solve the problem of varying 
appropriateness of figcaption as an alt 
substitute by further complicating guidance for 
how to use alt, thereby making alt required with 
figcaption in some situations and not in others. 
However, given the range and variations in types 
of visually discernable information, and the low 
burden for providing alt, it is likely to cause 
more burden to enumerate and require authors to 
learn the rule variations rather than to simply 
continue to require alt in the presence of a figure caption.


2. THE DOUBLY-SPECULATIVE RATIONALE THAT 
(paraphrasing) "ALT REQUIRED WITH FIGURE CAPTION 
MAY BE PERCEIVED AS EXCESSIVE WHICH MAY LOWER THE 
SERIOUSNESS OF ALL REQUIREMENTS" SHOULD BE 
EVALUATED AS "WEAK" RATHER THAN "STRONG" SUPPORT FOR A FIGCAPTION EXEMPTION

The doubly-speculative rationale that...

>excessive requirements will be perceived to 
>lower the seriousness of all requirements and 
>requiring information that authors are likely to 
>consider duplicate information is likely to be 
>perceived as an excessive requirement

should be downgraded from "strong" to "weak," 
particularly given that alt in many cases takes 
considerably less effort to write than does 
figure caption; and additionally that once one is 
writing a figure caption, one is already focusing 
on a descriptive task for a given image, 
therefore writing an alt for that image is even easier than usual.


3. THE SPECULATIVE RATIONALE THAT (paraphrasing) 
"A FIGCAPTION EXCEPTION FOR ALT MAY COMPLICATE 
AUTHORS' UNDERSTANDING OF HOW TO USE ALT" SHOULD 
BE EVALUATED AS "WEAK" RATHER THAN AS "NO" OBJECTION

Advising that the "no objection" evaluation of 
the speculative rationale that a figcaption 
exception for alt may indeed complicate authors' 
understanding of how to use alt (evaluated as 
constituting no objection) is at least as worthy 
of consideration as the doubly-speculative 
rationale that the seriousness of all 
requirements would be lowered if authors were to 
perceive alt as an excessive requirement, and 
that it should therefore be upgraded at least to "weak objection."


4. [CHANGE PROPOSAL text to be added, including
- a summary description of the change;
- a summary of the rationale;
- proposal details including spec text;
- positive and negative impact, including on conformance.]

###

--
Judy Brewer    +1.617.258.9741    http://www.w3.org/WAI
Director, Web Accessibility Initiative (WAI), World Wide Web Consortium (W3C)
MIT/CSAIL Building 32-G526
32 Vassar Street
Cambridge, MA,  02139,  USA

Received on Monday, 9 May 2011 15:20:15 UTC