Various further attempts have been made in the last hundred years to nail down what scientific thinking does and to prove that it is a truly reliable way of knowing, but they have all come with insoluble conundrums of their own.

The logical positivists, for example, tried to bypass Hume’s problems with the terms in scientific laws and to place the burden of meaning and proof onto whole propositions instead. A key point in the logical positivists’ case is that all meaningful statements are either analytic or synthetic. Any statement that does not fit into one of these two categories, the positivists say, is irrelevant noise.

Analytic statements are those whose truth or falsity is determined by the definitions of the terms they contain. For example, “All bachelors are unmarried men” is an analytic statement. If we understand the terms in the sentence we can immediately verify, by thinking it through, whether the statement is true.

Synthetic statements are those whose truth or falsity we must work out by referring to evidence found in the real world, not in the statement itself. In science, the needed evidence is found in human observations of the real world. “All substances contract when cooled” is a synthetic statement—not quite a true one, as observations of water and ice can show. So is “If a creature is a whale, then it is a mammal.”

The logical positivists aimed to show that discussions between scientists in all disciplines can be made rigorously logical and can therefore lead us to true knowledge. They intended to apply their analytic–synthetic distinction to all statements in such a rigorous way that any statement made by anyone in any field could be judged by this standard. If the truth or falsity of a statement had to be checked by observations made in the real, material world, then it was clearly a synthetic statement. If the statement’s truth value could be assessed by careful analysis of its internal logic, without reference to observations and measurements made in the material world, then the statement was clearly an analytic statement. Idea exchanges that were limited to only these types of statements could be logically sound. All other statements were to be regarded as meaningless.

The logical positivists argued that following these prescriptions was all that was needed for scientists to engage in logically sound discussions, explain their research, and size up the research of their fellow scientists. This would lead them by gradual steps on to true, reliable knowledge of the real world. All other communications by humans were to be regarded as forms of emotional venting, empty of any real content or meaning.

              

                                                                 Rudolf Carnap.

Rudolf Carnap, especially, set out prove that these prescriptions were all that science needed in order for it to work and to progress in a rigorously logical way toward making increasingly accurate statements about the real world—generalizations that could be trusted as universal truths.²

                   

                                                                 Willard V.O. Quine.

But the theories of Carnap and the other positivists were taken apart by later philosophers such as Willard Quine, who showed that the crucial positivist distinction between analytic and synthetic statements was not logically defensible. Explaining what makes an analytic statement (e.g., “All bachelors are unmarried men”) analytic requires that we first understand what synonymous terms like bachelors and unmarried men are. But if we go into the logic carefully, we find that explaining what makes two terms synonymous presupposes that we first understand what analytic means. In short, trying to lay down precise rules for defining the difference between analytic statements and synthetic ones only leads us to reason in circles.³

Quine’s reasoning, in turn, was further critiqued and refined by later philosophers like Hilary Putnam. As Putnam eventually put the matter:

“… positivism produced a conception of rationality so narrow as to exclude the very activity of producing that conception” and “… the whole system of knowledge is justified as a whole by its utility in predicting [future] observations.”⁴

In other words, logical positivism’s rigid way of talking about thinking, knowing, and expressing ends up in a logically unsolvable paradox. It creates new problems for all our systems of ideas and doesn’t help with solving any of the old problems.

We can see that most of the laws that have been formulated by scientists really do work. They guide us toward ways of living that get results. Why they work and how much we can rely on them—that is, how much we can trust science—are a lot trickier to explain.

---