Abraham Kaplan in his book The Conduct of Inquiry sets up too types of explanations: (1) semantic and (2) scientific.  A semantic explanation is any kind of verbal answer to why or how questions. For example, Question: Why does the ocean roar? Answer: You'd roar too if you had crabs in your bed. Following John Dewey, Kaplan wrote "A semantic explanation...is the outcome of transactions: a sale: the sale is not made unless the customer actually buys it." (p. 328). So one answer to what would count as an explanation is whatever  the customer or consumer buys into it. This may sound facetious, but it is not intended to be. People can be convinced of almost anything. There are still people today would believe that the world is flat or that the economy will grow if only the government will cut spending (a job killer) and deregulate (an environment killer). But according to Kaplan, a scientific explanation is one that is not relative to the consumer. A scientific statement is true. Well, anyone who has studied the history of science knows that not all scientific statements have turned out to be true. Many scientific explanations are inferentially based, meaning there are no independent warrants for what is called the explanans (see Carl Hempel's Philosophy of Science). So in the end, since science is a community enterprise, what counts as an explanation is whatever the scientific community buys into.