Philosophers have tried to explain how science finds the truth by using new developments in logic to study scientific language and inference. R. G. Collingwood argued that only a logic of problems could take context into account. He was ignored, but the need to reconcile secure meanings with changes in context and meanings was seen by Karl Popper, W. v. O. Quine, and Mario Bunge. Jagdish Hattiangadi uses problems to reconcile the need for security with that for growth. But he (...) mistakenly insists that all problems are mere contradictions and artificially separates rigid from flexible aspects of meanings. In order to resolve the conflict we must (1) replace the quest for rigid terms with techniques for improvement, (2) use plausible arguments to uncover confused meanings, (3) use frameworks to choose problems and to regulate meanings, and (4) employ a bootstrap approach that uses frameworks to improve meanings and refined meanings to improve frameworks. (shrink)