Stuff in my head

Tomayto Tomahto

I Guess it just boils down to Semantics…

How many times has that been the bewildering footnote to an epic duel between two empassioned geeks on the Twitter?

You often hear stuff like that after the realisation dawns three hours and countless 140 character bile nuggets later:

“…we mean kinda the same thing but we’re describing it differently”.

That’s the nature of semantics in a nutshell. With all the best will in the world they are still inherently arbitrary, and prone to ambiguity in the same way any subjective thing is. Even more so when the communication of that meaning is written or indirect. We try to apply rules and contextual hooks to help lean towards a certain semantic wrapper or definition, but humans are notoriously crap at agreeing on things.

If you’ve been working on the web for a while you’ll no doubt have been involved in a rather extended debate about what particular tag to wrap your content in. Should this be a <section> or an <article>? Is this really a list? At some point someone has to make a call. And that call is based on personal choice. And it’s that moment where the true value of semanticity comes into question. If something means something to you and something else to someone else, then who is right?

We can’t police semantics, it’s impossible to validate against free will. We can only try to get to a point where meaning is close enough to be of some value to us or the machines that need to work with it. And we have that with the new semantic wrappers in HTML5 which are a big step in the right direction - structurally helpful but still baggy enough to allow ambiguity to slip in. As a set of best practices they work well and certainly aid the readability of markup by removing a lot of the need for extra class names etc.

However, the point I’m building up to is really who are semantics for these days? Molly Holzschlag gave a great précis of the rationale for the initial drive for semantics over the weekend on Twitter. Primarily because assistive technologies were so cacky, the web community basically said screw you we’ll try and fix this from the inside. It dovetailed nicely with the standards movement too, and us lot who built the web, did our bestest to make sure the stuff we made conveyed as much meaning as possible. The problem was it was our meaning, which could be spot on, or could be complete nonsense.

We’re in a very different place now technologically. Having our markup machine readable through semantic markup is no longer a necessity. Look at Apple’s Reader technology in Safari. It happily sifts through the shittiest of markup and manages to deliver a pristine version of an article sans clutter. Similarly for events inside Mail app, without any markup whatsoever it figures out things that look like dates or events and provides calendaring functionality off the back of it. Search engines are pretty clever too these days - and they couldn’t give a flying fuck about what your content is wrapped in.

So what i’m getting at is this.

Machines are smart and getting smarter daily. They’re also impartial and use hard data instead of opinion to inform their choices. We already trust them with categorising and contextualising so much of our content already, and yet somehow we still have that distrust and slightly holier than thou attitude about controlling stuff that - dare I say it  -perhaps isn’t that important any more.