W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > April to June 2004

Action item: Bug 265: Title to differentiate linked text

From: Sailesh Panchang <sailesh.panchang@deque.com>
Date: Fri, 25 Jun 2004 13:12:12 -0400
Message-ID: <018401c45ad7$8ea8eb80$a501a8c0@deque.local>
To: <w3c-wai-gl@w3.org>
# 265

Summary: Accept title attribute as valid duplicate link text differentiator?
JAWS has a checkbox (read or don't) but doesn't read both).
Question: Is it acceptable for the title attribute of the link to be the sole distinguishing characteristic, or does the link text itself always need to be unique to fulfill this requirement?
The HTML specifications describe the title attribute as, "This attribute offers advisory information about the element for which it is set." HTML techniques for WCAG 1.0 also recommend use of title on a link to differentiate it from other links. This recommendation is valid and may be continued. Accordingly the title attribute may be used to qualify the linked text so that when both are read in conjunction they uniquely identify the resource. The title attribute on the anchor <a> element is useful   To differentiate two or more links that share the same linked text but point to different resources. Sighted users who are able to mouse over the link will be able to read the additional information contained in the title as a tool tip or in the status bar depending on the browser.

Requiring the screen text to be different  will impose a limitation on the content author's freedom of expression when other mark-up can be used to distinguish the links. 

Text to speech technologies today  read the  screen text (or linked text)  by default. In its absence they read the title attribute on the anchor tag. Some screen readers give the user the option to set what should be spoken by default: the title or screen text or the longer of the two. This seems to suggest that the user is able to  determine if links have title attribute and decide what should be spoken. Not all links need title set on them and expecting the user to check the title for every link is impractical, if at all allowed by the screen reader.
Makers of screen readers too cannot set the title to be spoken by defaultwhen in most cases it is the screen text that should be read. So the solution is for screen readers to be able to read screen textfor the link and the title if they are different.
In a sense screen readers are able to  do this. When there is no linked text ( or linked text is null), it reads the title as in the case of   an invisible skip navigation link with  only a title attribute and  no screen text.
Note: As of now, no text to speech software  reads linked text and title if they are different. 
Jim Thatcher and Sailesh Panchang
Received on Friday, 25 June 2004 13:12:13 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:47:30 GMT