Categorically speaking, the analysis of language, in a limited and antiseptic environment such as an experiment involving observer and subject, will almost always yield results expected, from inductive or deductive calculations, by the observer. It is the inherent nature of "ethnocentrism," a word originally used by Ludwig Gumplowicz in the 19th century to justify racial and class discord as a cause for war in order for a culture to progress.
Juxtaposing the above on to the field of anthropological linguistics, it is no wonder that results are speaking, not necessarily in tongues, but in categories. Categories of syntax, grammar, phonemes, morphemes and the like. When it wasn't enough to break down language into its components, the tendency then existed to do what every other field of scientific endeavor did in the last century, make it relative. Thus, the Edward Sapir-Benjamin Whorf theory. But like every other attempt to connect process (thought) with procedure (speaking), there would also be refutation. But in order to first do this, what little evidence exists to define a culture in terms of its language, there needed to be yet another division, hence, the "soft" and "hard" interpretations of Roger Brown, where that dialectic could just as easily be replaced by ching and chung.
There seems to be little emphasis on rather obvious components of the thinking-speaking-real world chain missing here; not so much in a strict Pavlovian take on the process, but from a Darwininian vis-a-vis BF Skinner approach. Without going into the details of the operant reaction experiments, it will suffice to say that conditioned response, behavior reinforcement and the environment play significant roles in whether there is any kind of influence on the process chain that constitutes a linguistic theory, whether relative of universal. To what degree conclusions can be drawn from over-simplistic experiments involving color terminology or spatial representations in various cultures might explain why the theories do not hold up to criticism.
To leave it on a metaphorical note, "the proof is in the pudding."