Matti Sintonen

The paper sketches an account of explanatory practice in which explanations are viewed as answers to explanation-requiring questions. To avoid difficulties in previous proposals, the paper uses the structuralist account of theory structure, arguing that theories are complex and evolving entities formed around a conceptual core and a set of intended applications. The argument is that this view does better justice to theories which involve a number of different kinds of theory-elements to give narrative explanations. Theories are, among other things, devices which can be used to turn explanation-requiring questions into a form which allows assessment of potential answers. Evolutionary theory, both in Darwin's and the modern synthetic forms, are used as examples. The view advanced is that modern evolutionary theory need not have a unique core to which other theories serve as subcontractors.

In what follows, I want to discuss two particular—though broad—topics that have been raised by recent advances in cognitive science and science studies. First, the role of creativity in scientists’ self-understanding has changed dramatically through centuries and, with help from our friends in cognitive science, it is now possible to go beyond the so-called scientific imagination. I shall also suggest that creativity requires persistence over a long period. In our times of immediate gratification, this is an increasingly difficult mental disposition to promote. Secondly, although discovery and creativity are seemingly drifting apart, it turns out that discoveries have a historical depth which creativity lacks. That is, discoveries are characteristically products of often long historical processes.

Jaakko Hintikka's Kantianism in philosophy of logic and mathematics is known to go further than Kant's own, for he argues that mathematical reasoning involves the "language-games" of seeking and finding. Therefore, logic mirrors the structure of this activity. But Hintikka also pushes the Copemican Revolution further to epistemology and philosophy of science. He agrees that "reason has insight only into what which it produces after a plan of ist own", but gives the idea a new logical turn. Kant thought that reason imposes certain architectonic constraints on the possible outcome of inquiry, but Hintikka's interrogative model of inquiry also emphasizes the activity of and therefore the strategy in, putting questions to Nature.

The paper explores two ways in which the logic of questions might aid in the understanding of explanations. First, the "logic" of question-answer sequences imposes constraints on what answers are acceptable for an inquirer. Secondly, there are field- specific type-requirements built into questions. There is always more to a question than meets the potential answerer's ear. It is argued that, since there are nonepistemic presuppositions of why-questions, there are no interesting necessary and sufficient conditions for all explanations. Explanations are a family of concepts tied together by the thin thread of the logic of question-answer sequences.

Doren Recker has criticized the prevailing accounts of Darwin's argument for the theory of natural selection in the Origin of Species. In this note I argue that Recker fails to distinguish between a deductive short argument for the principle of natural selection, and a non-deductive, long argument which aims at establishing that the principle has explanatory power in the various domains of application. I shall try to show that the semantic view of theories, especially in its structuralist form, makes it easy to distinguish between the two arguments and to explain how Darwin's long argument counts as one argument. I also raise a question about Recker's views on Darwin's mid-Victorian background, arguing that Newton's First Rule of Reasoning was not just a constraint on hypotheses involving unobservables, but a general request to keep conjecture and certainty apart