> Much of my research began with a large claim about society (e.g., "consumer culture constructs identities") and I then searched texts for language that could illustrate that claim. Because almost any passage could be interpreted that way through a sufficiently flexible framework (and enough imagination), it was difficult to say what evidence would prove the claim wrong.

Doesn’t sound like you’re doing science, to me.

In fact, this is the crux of the whole reproducibility crisis in social sciences, is it not?