There are two different ways to apply the distinction between dialectical and demonstrative argument, and it’s important to be aware of the difference. I draw the terms dialectical and demonstrative argument from Alasdair MacIntyre in Three Rival Versions of Moral Inquiry (pages 88-9), who in turn takes the distinction from Boethius‘s De topicis differentiis and ultimately from Aristotle’s Topics. The key point is that dialectical argument argues to first principles, and demonstrative argument from first principles.
But what are those first principles? Are they first principles for knowledge in general, or merely first principles within a single paradigm? In the Three Rival Versions passage MacIntyre does not distinguish between the two. When René Descartes makes a dialectical argument to first principles they are clearly the former, and I suspect that that is what the medievals were talking about as well. Their first principles are intended as secure foundations for all human knowledge.
But many first principles are something more modest: they are first principles of a paradigm. I noted last time how Thomas Kuhn describes paradigms as “universally recognized”, but that universal recognition is only within a given discipline (or even subdiscipline). The principle of evolution by natural selection is a first principle that is (effectively) universally recognized by biologists, for example – it is the central tenet of modern biology’s paradigm – but it cannot be said to be universally recognized by the world at large.
It is worth remembering that such first principles can conflict. In my limited exposure to the discipline of public health, it seems to be a first principle that individuals make choices that harm themselves as well as the wider society, and that we would therefore do well to regulate those choices for people’s own benefit. (At a conference I helped organize on urban health in poor countries, it surprised me how even in that context, much of the discussion was not merely about obvious externally imposed health risks like poor sanitation, but about risks like obesity and smoking that individuals pose to themselves.) The first principles of economics, on the other hand, typically assume that individual decision-making is a matter of rational self-interest, and that the more individual preferences we can satisfy the better; economists build their models on such assumptions. These first principles of eonomics and of public health do not quite contradict each other, but they are in tension with one another, and for this reason economists and public health practitioners stand a strong chance of clashing when shared decision-making happens. The gulf may not be as wide as that between biologists and fundamentalists, say, but there is a basic similarity between the two gulfs – in both cases, the two sides are deriving their substantive conclusions from entirely different sets of first principles.
The first principles, in these cases, are assumptions, taken for granted – and as Kuhn notes, it is that very taken-for-grantedness that allows the detailed work of the discipline to be done. I take this to be the truth underlying popular astrophysicist Neil deGrasse Tyson’s otherwise ignorant attack on philosophy: philosophy rightly makes it a business to question and challenge taken-for-granted assumptions, and such challenges get in the way of the day-to-day business of normal science. (They may be important for revolutionary science, when paradigms are overthrown, but Kuhn notes that the vast majority of scientific progress does not happen in the course of such a revolution.)
Moreover, these assumptions are not always articulated. Kuhn points out how once a discipline takes on a paradigm, its only works accessible to nonspecialists tend to be introductory textbooks. In such a case it is the job of a textbook to make dialectical arguments: to persuade its reader rationally, give the reader good reason to believe, that the paradigm’s first principles are justified.
Unfortunately, too many textbooks simply state the first principles and expect their readers to believe them – as do professors in introductory courses. It is assumed that the point of the course is simply to prepare students to learn what follows demonstratively from the first principles. I recall my introductory macroeconomics class at Cornell, where the professor on the first day simply stated that we would assume the goal to be pursued was maximizing GDP (per capita). He mentioned there have been a lot of criticisms of the use of GDP in this way, paused, and then said: “but we’re going to use it anyway.” And that was that. No discussion of what the criticisms of his first principles even were, let alone how they might be answered. In short, no dialectical argument at all, merely proceeding straight to demonstrative. The assumption in such cases is that the demonstrative arguments within the paradigm are so important that the dialectical arguments can simply be skipped.
To be fair, much of life works this way. We do not start out with argument. Long before we are able to reason and make arguments, we are immersed in a worldview we simply absorb, one we take for granted. This is the pre-philosophical world of “prejudices” in the Gadamerian sense, of “intuitions”, of Aristotle’s phainomena, indeed of common sense. We are in this world before we get to the first principles of any paradigm. Thus first principles are not first in a chronological sense – there is much that comes before them – but rather in the logical sense that demonstrative arguments follow from them. One may be convinced of the consequences of a first principle without having any understanding of the first principle itself.
The problem is that these starting points, the prejudices and intuitions, are unreasoned – or at least not fully reasoned. We have not thought them through back to their logical roots. Our thought depends logically on assumptions we may not even have thought of. That is why I have argued that in cross-cultural philosophy, assumptions can be as important as argument: we can see the assumptions that other cultures don’t share with us, and in the process learn about unreasoned assumptions we didn’t even know we had.
[EDIT: The sentence above "When René Descartes makes a dialectical argument to first principles they are clearly the former" originally said "clearly the latter". Thanks to Jim Wilton for the correction.]