The term “decisive reason” is used in two distinct senses in the contemporary literature, and the conflation between them is, I think, a source of recurring confusion in arguments about practical reason.
The first sense, which I will call permissibility-decisive, is this: a reason is decisive against an act if it makes the act impermissible. To say there is a decisive reason against φ-ing is to say that φ-ing is forbidden.
The second sense, which I will call weight-decisive, is this: a reason is decisive in favor of an act if it outweighs all competing considerations. To say there is a decisive reason for φ-ing is to say that φ-ing is best supported by reasons, all things considered.
These are different claims. The first is about deontic status; the second is about the balance of reasons. A claim can be true in one sense and false in the other.
An example
Consider a familiar case. I have promised to meet a friend for coffee. I have, separately, a strong inclination to stay home and read instead. Reading would be more pleasant for me; meeting the friend would honor a promise.
Is there a decisive reason against staying home?
In the permissibility-decisive sense, the answer depends on whether breaking the promise is impermissible. I think it is, in this case — the promise was a small one, but a promise nonetheless — so there is a decisive reason against staying home in this sense.
In the weight-decisive sense, the answer depends on whether the reasons against staying home outweigh the reasons for. Here the answer is less clear. The friend will be mildly disappointed but not seriously harmed. The reading would give me genuine pleasure. The promise was small. It is at least open whether the weight of reasons clearly favors going.
These two answers can come apart. The act might be impermissible (permissibility-decisive reason against) without being clearly outweighed by reasons (no weight-decisive reason against). Or, less commonly, an act might be permissible while being clearly outweighed by reasons in favor of an alternative.
Why this matters
The two senses come apart because permissibility and reason-weighing are governed by different principles. Permissibility is, on most accounts, a threshold property: above the threshold of acceptability, an act is permissible; below it, forbidden. Reason-weighing is, on most accounts, continuous: reasons can outweigh one another by larger or smaller margins.
When we say something like “there is decisive reason to φ” without specifying which sense we mean, we open the door to equivocation. The interlocutor who hears “weight-decisive” concludes that φ-ing is what the reasons support, all things considered. The interlocutor who hears “permissibility-decisive” concludes that not-φ-ing is forbidden. These are different conclusions, with different implications for action.
A proposal
I propose that when we use the term “decisive reason” in technical contexts, we specify which sense we mean. The convention might be: “decisive-p” for permissibility-decisive, “decisive-w” for weight-decisive.
This is, admittedly, an aesthetic uglification of the prose. But it would prevent a class of confusions that I have seen in the literature with some regularity, and that I have myself made more than once.
Alternatively, we could simply refrain from using the term where the sense is ambiguous, substituting “most reason to” for the weight-decisive sense and “required” or “impermissible to not” for the permissibility-decisive sense. This is wordier but clearer.
An observation
I will close with a meta-observation. The conflation I have identified is, I think, a special case of a more general phenomenon: in normative ethics, our technical vocabulary frequently elides distinctions that turn out to matter. “Decisive reason” is one example; “should”, “ought”, and “must” are others; the distinction between “is best” and “is required” (which I have discussed elsewhere) is another.
These elisions are partly historical — we inherit a vocabulary that conflates what later authors come to distinguish. They are partly functional — the elision serves to gloss over disagreements that would otherwise block discussion. They are, in my view, mostly worth correcting, even at the cost of less elegant prose. Clearer concepts make for better arguments, even when they make for clumsier writing.