Good False Theories

Good False Theories

In what does the goodness of false philosophical theories consist? How can a theory be good and yet false? To be good a theory must have certain attributes: clarity, simplicity, interestingness, the ability to solve problems, integration with other theories, and explanatory power. Generally, it avoids mystery and provides a reduction of some sort. To be true, a theory must state the facts, capture how things actually are, correspond to reality. These two properties are not mutually entailing: theories can be good but not true and true but not good. In physics quantum theory is true but not good, and Descartes’ vortex theory is good but not true: quantum theory lacks a clear interpretation and is full of mysteries, and it turns out there are no vortices controlling celestial movements. In philosophy we have cases in which the theory has many attractive features but is clearly not true, and in some cases true but lacking in other desirable qualities. I will mention two theories that illustrate these points.

Phenomenalism is a good theory by the criteria stated: it is clear, simple, interesting, problem-solving, reductive, and non-trivial. It tells us what a material object is without invoking such concepts as substance and matter, and it links objects with our experience of objects, thus refuting skepticism. The trouble is that it is clearly a false theory: conditionals about sense experience are neither necessary nor sufficient for statements about physical objects. There would be objects even if there had never been any sense experiences. And this is perfectly obvious: therefore, phenomenalism is false, despite its theoretical virtues. This is typical of philosophical theories. Notice that rejecting phenomenalism leaves us theoretically bereft: now we have to face skepticism again, and we have to make sense of concepts like substance and matter. We are left with a true theory that isn’t a good theory, i.e., one that lacks the theoretical virtues we seek. Often, the theory we are left with has mysterious elements (like Newton’s theory of gravitation—true but “occult”).

The second example is truth-conditional theories of meaning: these are formally well-defined, they reduce meaning to truth, and they don’t introduce concepts that defy clear treatment. Tarski-style semantics is a case in point. The trouble is that such theories are manifestly false, because truth conditions are insufficient for meaning—sameness of reference is not the same as sameness of meaning. But if we abandon such theories, we find ourselves in murky waters with no workable theory to guide us. It is much the same with other philosophical issues: theories of moral value, theories of number, theories of mind, theories of necessity, etc. The true theory seems not be a good theory, and the good theory looks far from the truth. This is the characteristically philosophical dilemma: implausible reduction versus mysterious anti-reduction. The good seems to be the enemy of the true, and vice versa.

What should we do? Keep looking for good theories but bear in mind that true theories are often not very good. Platonism in mathematics is arguably true, but it leaves a host of problems in its wake—not least how we can know mathematics. Moral realism makes us nervous (rightly so), but moral anti-realism makes us angry—because it is so far from morality as we intuitively understand it. We can see it’s not true, but we have no alternative that leaves us free of worry. The only reason we believe the theories we do is that we want to avoid the alternative. Would anyone willingly be a materialist unless the alternative were worse? Truth and goodness don’t march in step in philosophy: the better the theory the less true it is apt to be. It would be great, theoretically, if the universe were as Berkeley describes, or if there were nothing but physical particles; but these theories have the disadvantage of falsehood. And the same goes for all the other theories that have occurred to philosophers throughout the ages.[1]

[1] I can’t think of a single theory in philosophy that I think is both true and good: the truth always seems problematic, and the good always seems unrealistic. In science, by contrast, truth and goodness generally go together—as with Darwin and Copernicus. It is rare in science to find a theory that is true but theoretically deficient, or false but theoretically sound.