Dewey is a very interesting one. He and I are have a lot in common; I’m perhaps more similar to him than any other noted philosopher: We’re both interested in ethics as a practical guide to behavior that can readily adapt to any circumstance. We share a sense that the big work to be done is on the societal level, not the individual level. We both seek broader discussion of ethics within our respective societies and are very active in education. We have similar criticisms of deontological/rule-based ethics systems (“Thou shall not kill” etc). We both strive to integrate actual human psychology into our thinking instead of assuming some ideal behavior. We also share a strong interest in “decision procedures”, i.e. how we do or should make decisions, which goes beyond the more basic matter of which decisions are good. We seem to have similar intuitions on what constitutes wellbeing. Dewey even seemed fond of utilitarianism (it “represented progress in ethics” - link).
I will skip further discussion of our similiarities here (but feel free to ask for elaboration on any of these points in the comment thread). I’m also going to skip over my concerns with his views on meta-ethics and philosophy of science. Instead, I’m going to discuss his views on means and ends and his idea of experimental ethics, the latter of which I think has some merit for certain circumstances but is less good in others and is downright frightening in a specific class of circumstances- irreversible processes.
Regarding ends & means (also referred to as intrinsic goods and instrumental goods), I support Dewey’s criticism of the then-traditional breakdown, but I’m not sure I support his substitute. In the then-traditional breakdown, ends and means are both things happening to the individual. Anderson offers as an example of an end “getting to a lake” with a corresponding mean of “jumping across the ditch that’s in the way” (link). I think Dewey is right to criticize the traditionalists where they claim that this means need not be valued.
The utilitarian’s breakdown, as I see it, functions much better: The means here would be both getting to the lake and being at it, while the ends would be the wellbeing experienced from both getting there and being there (including the wellbeing of both the individual in question and all other individuals who are affected). I’m not sure if Dewey would accept this breakdown, and I’m less sure if Dewey would accept the utilitarian’s only valuing the ends (wellbeing) and not the means. This passage in particular perplexed me:
“Although hedonism fails as a theory that gives us a fixed end, it does contain a methodological insight. Nothing is good that cannot be desired. All desire contains an element of enjoyment or liking. Hence, pleasure can be seen as a sign of the good, as evidence of what is valuable. Nevertheless, what makes desire a sound guide to the good is the fact that it incorporates foresight and reflection on the wider consequences of acting on it, not just that it incorporates a liking of its object.” (link)
Never mind that hedonism, whether egoist or utilitarian, seems relentlessly focused on pleasure as the ultimate, fixed end. (Egoistic hedonism recommends maximizing one's own pleasure; utilitarian hedonism recommends maximizing everyone’s pleasure.) What confuses me most is how intangible Dewey’s notion of the good is. Does he actually think that action based on “foresight and reflection” is intrinsically good? What if such action lead to horrendous outcomes, like, say, genocide?
Regarding experimentalism, I think there are some circumstances where this approach can be successful. I'm imagining very new and exotic circumstances like wireheading. If (and it quite possibly doesn't) my utilitarianism recommended large-scale wireheading ASAP, I would instead recommend a Dewey-esque experimental approach, sacrificing some total utility for the assurance that we weren't doing anything that we didn't realize was, for whatever reason, not a good idea.
It is only in these unusual, extreme cases where I’d recommend the experimental approach. In most ordinary circumstances, utilitarianism works quite well. Further, the experimental approach is attrocious with irreversible processes. IPs include such things as biodiversity/species extinction or permanent reduction in carrying capacity. Because they can’t be undone, these are not matters to experiment on. If we’re going to do them, we better be sure they’re the right thing to do. Note that utilitarianism is not always great on these either. While it probably places more weight on preventing humanity’s extinction than anything else, it is indifferent to extinction in general, especially the extinction of non-sentient species. However, I have at least some inclination to prevent their extinction in case we might later decide to value these, even beyond biodiversity's potential usefulness to sentient beings.
So, I’d recommend experimentalism for new, reversible moral dilemmas, caution for irreversible processes, and utilitarianism for everything else. Where experiment or caution might substantially reduce total expected utility, I would side with utilitarianism.