Andrew Gelman posts an interesting graphic showing the adoption of various technologies from 1900 forward. See the post for the lineage on the graph and the details. Good graphic.
What caught my eye for topic maps was the rapid adoption of the Internet/WWW and the now well recognized failure of the Semantic Web.
You may feel like disputing my evaluation of the Semantic Web. Recall that agents were predicted to be roaming the Semantic Web by this point in Tim Berners-Lee’s first puff piece in Scientific American. After a few heady years of announcements of realization is just around the corner, the 21st century technology equivalent of the long retreat (think Napoleon).
Now the last gasp is Linked Data, the “meaning” of URIs is be determined on mount W3C and then imposed on the rest of us.
Make no mistake, I think the WWW was a truly great technological achievement.
But the technological progress graph prompted me to wonder, yet again, how is the WWW different from the Semantic Web?
Not sure this is helpful but consider the level of agreement on semantics required by the WWW versus the Semantic Web.
For the WWW, there are a handful of RFCs that specify the treatment of syntax. That is addresses and the composition of resources that you find at those addresses. Users may attach semantics to those resources, but none of those semantics are required for processing or delivery of the resources.
That is for the WWW to succeed, all we need is agreement on the addressing and processing of resources and not at all on their semantics.
A resource can have a crazy quilt of semantics attached to it by users, diverse, inconsistent, contradictory, because its addressing and processing is independent of those semantics and those who would impose them.
Resources on the WWW certainly have semantics, but processing those resources doesn’t depend on our agreement on those semantics.
So, the semantic agreement of the WWW = ~ 0. (Leaving aside the certainly true contention that protocols have semantics.)
The semantic agreement required by the Semantic Web is “web scale agreement.” That is everyone who encounters a semantic has to either honor it or break that part of the Semantic Web.
Wait until after you watch the BBC News or Al Jazeera (English), الجزيرة.نت, before you suggest universal semantics are just around the corner.
[…] may find it odd that I would cite such a resource on the same day as penning Technology speedup graph where I speak so harshly about the Semantic […]
Pingback by Nature Publishing Group releases linked data platform « Another Word For It — April 8, 2012 @ 4:21 pm