Introduction - Social Semantics: The Search for Meaning on the Web - page 2

and - although radical on the surface - actually ends up being a reactionary fad
when viewed within the context of a larger landscape outside academic philosophy
and cognitive science. In particular, computer science - with the exception of the
peculiarly anthropomorphic line of research of artificial intelligence (AI) - does
not seem to care about embodiment. In his One Hundred Billion Lines of C

,
computer scientist-turned-philosopher Brian Cantwell Smith notes that in artificial
intelligence, debates over representation tend to frame the debate as if it were
between “classical” logic-based symbolic reasoners and some “connectionist” and
“embodied” alternative ranging from neural networks to epigenetic robotics (1997).
Smith then goes on to aptly state that the kinds of computational systems discussed
in artificial intelligence and philosophy tend to ignore the vast majority of existing
systems, for “it is impossible to make an exact estimate, but there are probably
something on the order of 10 11

++

, in the world.
And we are barely started. In sum: symbolic AI systems constitute approximately
0.01% of written software” (1997). What Smith fails to mention is that the same
small fraction likely holds true of “non-symbolic AI” computational systems such
as robots, artificial life, and old-fashioned connectionist networks (an exception may
soon be made for the machine-learning that runs phenomena such as advertising and
search on the Web). As raw statistics of deployed systems by themselves hold little
intellectual weight, no doubt a philosopher could argue that the vast majority of
computational systems may have no impact on our understanding of representation
and intelligence. In other words, what the vast majority of the planet is doing with
computation and representation - which is increasingly focused on the World Wide
Web - is simply intellectually uninteresting. In this topic we argue otherwise.
Although one can easily deny that anything resembling digital representations
exists 'inside the brain,' it is much harder to argue that there are no digital
representations on the Web. As one clicks from web-page to web-page, it seems that
the Web is nothing but a vast network of digital representations. The thesis of this
topic is that the wide class of computational systems outside of those traditionally
considered by artificial intelligence or philosophy presents what Cantwell Smith
calls a “middle distance” where questions of representation (and perhaps even
intelligence) come to the forefront in a peculiarly obvious manner and are likely
more tractable than they are for humans, given the relative complexity of computers
and humans (Smith 1996). At the present moment, with all the totalizing attraction
of a black hole, computational systems the world over are becoming part and
parcel of the World Wide Web, described by Tim Berners-Lee - the person
widely acclaimed to be the 'inventor' of the Web - as “a universal information
space” (1992). We further argue that not only may the Web reveal general insights
about the nature of representation, but its unique historical status as the first actual
universal information space may prompt an entire re-thinking of semantics. When
asked to consider this hypothesis, Michael Wheeler - a philosopher who is well-
known for his Heideggerian defense of embodiment - surmises that “the power
of the Web as a technological innovation is now beyond doubt” but “what is less
well appreciated is the potential power of the Web to have a conceptual impact on
cognitive science” and so the Web may provide a new “fourth way” in addition to the

- one hundred billion lines of C ++