Playing With Words

It never ceases to surprise me how much social sciences research depends on fluency in a language. Some social science research strikes me as playing with words--putting a slightly different spin on a conceptual term and passing it off as a contribution to the literature. As an aside, isn't it interesting that we often refer to an existing body of research on the subject as 'literature'? Also, the starting point of most research projects is a "literature" review.
How does one separate nuanced research from a play on words? One tends to dismiss research that one doesn't understand as a play on words. I came across a journal article today and I couldn't decide whether it was brilliant or faff. I was able to grasp some of the author's arguments, but not all. This is a perennial problem with  purely theoretical research. With empirical research at least you are saying something concrete about a real-world phenomena. Which is why examples are so useful. They are an age-old teaching tool, but I think they should be used more in research as well. Cases where a new concept seems like a play on words would be helped if the researcher can give an example where his concept is applicable and another where it is inapplicable. I understand precision can sometimes come at the cost of nuance, but surely intelligibility must be of primary importance,  ahead of nuance.