I worked on this and published it at: http://www.scivillage.com/thread-2580.html
This should work, but it would be more useful to evaluate it and the feedback.
.x:before{ content:attr(title); display:inline; }
<ul> <li><a href="#"><span class="x" title="Homepage"></span></a></li> <li><a href="#"><span class="x" title="Contact" /></a></li> </ul>
(I shortened the class name to reduce markup)
Search engines should ignore HTML tags with empty values when searching for keywords, this should mean that it ignores what is written in the title attribute. (He assumes that the value is what matters if it is empty, then there are no dots checking the attributes)
It has been suggested that it is possible to deny the existence of a closing tag in HTML5 due to limited stringency, however, there are counter suggestions that end tags are still required.
I would suggest not using it directly in tags (anchor), as they can be used for Sitemaps (using #), so that means they would like Title Spidered.
Although it is possible that he might suggest that any content in the headline should inflate keywords with hidden elements, I cannot confirm this.
source share