That is not in fact how science works.

Science only welcomes new ideas when they offer a possible solution to a problem with the existing paradigms. For this type of theory, if you had data showing the earliest known societies were unusually advanced, or if you had artifacts that suggest more wide spread early societies, these would be facts wanting an explanation.

You can't just say something like "maybe all language was singing until 20,000 years ago" and claim it's scientific until proven false.