SPARQL is purpose built for the RDF world where you're mixing and matching a zillion different vocabularies, all of which have to be painstakingly declared and name spaced every time.
For most of us working not on the "semantic web", we typically only have 1-2 vocabularies, which is our data model, and SPARQL is super clunky to use.
For most of us working not on the "semantic web", we typically only have 1-2 vocabularies, which is our data model, and SPARQL is super clunky to use.