12. As far as we know, Earth is the only planet on which life has evolved, and all known life forms are carbon-based. Therefore, although there might exist non-carbon-based life on planets very unlike Earth, our scientific estimates of the probability of extraterrestrial life should be generated from estimates of the number of planets like Earth and the likelihood of carbon-based life on those planets.

Which one of the following general principles most strongly supports the recommendation?
(A) There is no good reason to think that unobserved phenomena closely resemble those that have been observed.
(B) A scientific theory that explains a broad range of phenomena is preferable to a competing theory that explains only some of those phenomena.
(C) It is preferable for scientists to restrict their studies to phenomena that are observable and forego making estimates about unobservable things.
(D) A scientific theory that explains observed phenomena on the basis of a few principles that are independent of each other is preferable to a theory that explains those same phenomena on the basis of many independent principles.
(E) Estimations of probability that are more closely tied to what is known are preferable to those that are less closely tied to what is known.

OA is .