Most Counterfactuals are False

SPEAKER

Alan Hajek

ABSTRACT

Counterfactuals are all the rage these days. They figure in influential philosophical analyses of many important concepts, such as causation, perception, and rational decision. Science freely traffics in counterfactuals. They are also earning their keep in the social sciences, especially in psychology, history, and the law. And we use counterfactuals nonchalantly in daily conversation. Nevertheless, I argue that most counterfactuals are false.
I focus on two strategies for showing a counterfactual of the form ‘if X were the case, then Y would be the case’ to be false: appealing to

indeterminism—in particular, chanciness; and to
indeterminacy—in particular, imprecision.

Both are strategies for securing the truth of ‘counterfactuals’ of the form ‘if X were the case, then Y might not be the case.’ These ‘might not’ counterfactuals, I argue, are incompatible with the corresponding ‘would’ counterfactuals.

I consider, and reject, a number of rival positions:
- most counterfactuals are indeterminate;
- they have context-dependent truth values;
- the ‘might not’/’would’ clash is merely pragmatic; and
- there is no such clash at all.

I concede that some counterfactuals are true in virtue of necessary connections between antecedents and consequents. But such counterfactuals are rare, and do little to offset the preponderance of false counterfactuals.

How, then, does our practice of uttering counterfactuals survive? Close to the ordinary but false counterfactuals that we utter are counterfactuals that are true but not ordinary—e.g., ones with probabilistic consequents. They support our practice when the standards for asserting counterfactuals are forgiving, as they typically are on the street. However, the street is not always forgiving; even when it is, falsehood is merely tolerated rather than eradicated; and we philosophers are not always on the street.