Indeed, Hume, perhaps the most famous compatibilist, denies the existence of free will in his Treatise, only advocating compatibilism later, in the Enquiry Concerning Human Understanding. It certainly seems to me that he doesn’t actually change his mind; his early position seems to be “this thing people call free will is incoherent, so we should talk about things that matter instead,” and his later position seems to be “people won’t stop talking about free will, so I’ll call the things that matter free will and reject the incoherent stuff under some other label (indifference).”
So his opinions kind of did change over that time period, but only from “I reject these words” to “alright, if you insist, I’ll try to salvage these words”. I’m not sure which policy’s best. The second risks arguments with people who don’t know your definitions. They will pass through two phases, the first is where the two of you legitimately think you’re talking about the same thing but the other is a total idiot who doesn’t know what it’s like. The second phase is perhaps justifiable umbrage on their discovering that you are using a definition you totally just made up, and how were they even supposed to know.
The former position, however, requires us to leave behind what we already sort of kind of suspect about these maybe-not-actual concepts and depart into untilled, unpopulated lands, with a significant risk of wheel-reinvention.
Indeed, Hume, perhaps the most famous compatibilist, denies the existence of free will in his Treatise, only advocating compatibilism later, in the Enquiry Concerning Human Understanding. It certainly seems to me that he doesn’t actually change his mind; his early position seems to be “this thing people call free will is incoherent, so we should talk about things that matter instead,” and his later position seems to be “people won’t stop talking about free will, so I’ll call the things that matter free will and reject the incoherent stuff under some other label (indifference).”
So his opinions kind of did change over that time period, but only from “I reject these words” to “alright, if you insist, I’ll try to salvage these words”. I’m not sure which policy’s best. The second risks arguments with people who don’t know your definitions. They will pass through two phases, the first is where the two of you legitimately think you’re talking about the same thing but the other is a total idiot who doesn’t know what it’s like. The second phase is perhaps justifiable umbrage on their discovering that you are using a definition you totally just made up, and how were they even supposed to know.
The former position, however, requires us to leave behind what we already sort of kind of suspect about these maybe-not-actual concepts and depart into untilled, unpopulated lands, with a significant risk of wheel-reinvention.