One of the most powerful selling points of Linked Data is the AAA concept: Anybody can say Anything about Any Topic. This concept is nice because it gives a sense of freedom. “I do not have to work with data or models that do not completely satisfy my requirements, but I can use parts and and add my own to it.” Indeed that is exactly what is beautiful about it. No one can provide information that works well for everybody. Because of cultural differences, point of views, different use cases it is necessary that different sources describe the same topic.
However.. the AAA concept should not encourage people to see it as a license to desist from applying standards. After all, standards in general are created to warrant interoperability. The whole purpose of creating Linked Data is to provide interoperable data! Both syntactic and semantic.
What I see is that syntactic interoperability in Linked Data is not an issue, people apply these standards fine. The issue is semantic interoperability. The whole point of creating semantically correct Linked Data is sometimes neglected or ignored, and often misunderstood. Does it matter? Yes! Because if you don’t apply proper semantics, you create messy Linked Data and people who use it can’t rely on its meaning. That’s pretending to provide something interoperable but in reality you provide something that only you understand. Plus you are throwing away half of the potential of Linked Data. Simply a waste of technical resources.