The first is meant to apply generally, even across beings, so that humans’ reflective preferences are always prioritized over nonhumans’ revealed preferences. We can break ties with nonhumans’, but that will be so rare that it practically won’t matter.
The second means that sometimes we will prioritize revealed preferences over reflective preferences, and so sometimes the revealed preferences of nonhuman animals over the reflective preferences of humans.
The “optionally” part just means that if a particular being has both revealed and reflective preferences about something, we could use those particular reflective preferences and ignore those particular revealed preferences, although others’ revealed preferences may take priority. You could imagine that you have “true preferences”, and both revealed and reflective preferences are ways to try to measure them, but reflective preferences are always more accurate than revealed preferences, not that they’re more important. So, it’s like saying we have two measures of some individuals’ welfare (both revealed and reflective preferences) and we just prefer to use the strictly more accurate one (revealed preferences) when both are available, but it doesn’t mean the welfare of those for whom only the measure that’s less accurate in humans (revealed preferences) is available matters less.
The first is meant to apply generally, even across beings, so that humans’ reflective preferences are always prioritized over nonhumans’ revealed preferences. We can break ties with nonhumans’, but that will be so rare that it practically won’t matter.
The second means that sometimes we will prioritize revealed preferences over reflective preferences, and so sometimes the revealed preferences of nonhuman animals over the reflective preferences of humans.
The “optionally” part just means that if a particular being has both revealed and reflective preferences about something, we could use those particular reflective preferences and ignore those particular revealed preferences, although others’ revealed preferences may take priority. You could imagine that you have “true preferences”, and both revealed and reflective preferences are ways to try to measure them, but reflective preferences are always more accurate than revealed preferences, not that they’re more important. So, it’s like saying we have two measures of some individuals’ welfare (both revealed and reflective preferences) and we just prefer to use the strictly more accurate one (revealed preferences) when both are available, but it doesn’t mean the welfare of those for whom only the measure that’s less accurate in humans (revealed preferences) is available matters less.