This dissertation discusses a broad range of problems
concerning the use of the SPARQL query language on the open,
publicWeb. It is motivated from seeing decentralisation of
data and infrastructure as an important social goal, and
SPARQL as an enabling technology to solve problems using the
Semantic Web. The dissertation makes contributions in
hypermedia, where RDF is used to create a format that can
tell humans and machines alike how to manipulate resources
on the Web; philosophy of science, where important
foundational problems around how to create valid knowledge
and objectivity are discussed; statistical methods that
better satisfies the requirements from philosophy of science
than current practice; how to improve developer efficiency
with novel programming paradigms; and finally how caching
infrastructure in the Internet may be used to make query
answering across the Web more robust.