<p><pre><code> To present a theory is to specify a family of structures, its
models; and secondly, to specify certain parts of those
models (the empirical substructures) as candidates for the
direct representation of observable phenomena.
— van Fraassen. The Scientific Image. Page 64. 1980.
</code></pre>
This is Wittgenstein's family resemblance concept applied to Popper's notion of "conjectures" in "the system of reference by means of which we interpret an unknown language".<p><pre><code> It is not only agreement in definitions but also (odd as it
may sound) in judgments that is required.
— Wittgenstein's PI, somewhere deep in there around 200.
</code></pre>
The "unknown language" is itself the outward bounds of possible sense data for which we must cultivate a scientific description framework — or, if you will, RDF (Resource Description Framework).<p>From the perspective of a Hilbertian mathematical ideal, all conjectures are a kind of "paradise" for which we must apply method and tools. Strictly speaking, linguistic items serve as the material edifice in which we make experiment and theorizing about "propositions" and "mathematical objectivities" insofar as their existences are palpabable (by Occam's Razor, and various other First Principles) which would appear repugnant to reason to flout, ignore, or misplace.