# Consequences Are Always There
Published: 2026-04-28
Canonical: https://journal.xiaotianfanx.com/journal/consequences-are-always-there

On John Stuart Mill, “Utilitarianism,” J. J. C. Smart, “Defending Utilitarianism,” and Mozi, “Impartial Caring” from Readings In Classical Chinese Philosophy.



	According to J.S. Mill, originally, the utilitarian standard of the Greatest Happiness Principle is to pursue "the greatest amount of happiness altogether," given that a utilitarian will be "as impartial as a disinterested and benevolent spectator" to potentially sacrifice their own happiness for the greatest happiness of all. It is then provided a proof of the principle to establish that, allegedly, "there is nothing desired except happiness" as the sole end (on the means-ends dichotomy) for human desire or to be desirable indeed. It is noteworthy that in the course, the consequentialist gene is emphasized by the distinction Mill makes between the rule and the motive. Although the term "rule" is used here, this comparison evidently signifies the consequentialist interpretation—at least on reading Mill's words, as it is also equivalently explained by "the morality of the action."



	At first glance, one of the premises for the principle that initially alerted me, but to be reluctantly accepted for now, is the seemingly inevitable dissimulation/reduction of almost everything that is desired or desirable into happiness as the sole end.1 On one hand, one could confidently assert that some of their desire or pursuits should not fall into the categorization of happiness. However, it seems to me that Mill attempts to put forth everything that is desired and desirable as interchangeable with the word happiness itself. For example, in the case of virtue, Mill evidently reframes it in terms of it not being "originally part of the end, but it can become so"—where "end" refers to happiness, building upon the acknowledgement that "the ingredients of happiness are various, and each of them is desirable in itself." So, given any counterexample, it seems that he would probably always reframe those desired or desirable as happiness. Therefore, at least at this point, I find no point in getting too trapped in arguing against this premise itself. 



	Still, even if we tentatively accept the expression that says "what there is to be desired/desirable ≡ happiness," there is another thing that is noteworthy in the narrative—which is the interchangeable usage of "desired" and "desirable" as effectively an equivocation. Certainly, one of the usages of desirable can be describing something that can be desired. However, the more common usage of the word "desirable" points to another possibility: that is, to describe something that is worthy of being desired, and even, to some extent, should be desired. That said, I'd say it's bootstrapping when Mill arrived too early at the conclusion from saying "each person, so far as he believes it to be attainable, desires his own happiness"—which could be true in a psychological description sense—to say happiness itself is desirable if we value the second interpretation of the word "desirable." One of the clear pieces of evidence for this failure comes with Mill's analogy. When Mill says "the only proof that a sound is audible is that people hear it...the sole evidence that anything is desirable is that people desire it," the "audible-heard" connection is clearly a descriptive definition of the word "audible" itself, while "desirable" could mean "capable of being desired," but Mill needs it to mean "worthy of being desired." In other words, what is desired does not prove itself granted to be (objectively) desirable or, furthermore, as Mill attempts, the only thing that should be desired, or that we have a moral obligation to maximize it.



	Apparently, this links back to our previous discussions on metaethics. Particularly, I found the metaethical implications embedded in Mill's own argument to be interestingly divergent. On one hand, when Mill says "the standard of morality may be defined as the rules for human conduct by the observers of which an existence such as has been described might be to the greatest extent possible sacred to all mankind, and not to them only, but to the whole sentient creation," together with the normative leap of happiness from being desired to desirable that we observed just yet, it is very much clear that he is suggesting the utilitarian standards of maximizing happiness is an ought and normative claim that falls closer to objectivism, in principle. On the other hand, as we discussed a little bit earlier, the reframing approach taken by Mill to argue for the principle of treating everything that is desired as fundamentally happiness seems pretty much perspective-based or relative, as it depends on the various components of happiness that whoever may come up with to be their values. To my understanding, there is surely a tension between the principle itself and the approach Mill uses to justify one of the key premises of the principle—if the premise itself is backed by a somewhat perspectivist justification, then an objectivist principle relying on such a premise may be fundamentally undermined.2



	Before we dive into particularities, there is one more first-glance concern that I'd like to raise: When it comes to "the greatest amount of happiness altogether," where or what exactly is this "altogether"? In a typical trolley dilemma, where we have to decide whether to pull a lever to kill one person to save five who would be dead if we do nothing, the context (or the system) is well defined. In any similar scenario, a utilitarian may derive a reasonably calculated decision, taking for granted that we have already solved the problem of desire/happiness. However, when it comes to practice, there is never a certainly defined boundary to conduct such aggregational calculation of happiness upon, unless one inevitably admits that the calculation itself is incomplete from what utilitarian standards require—i.e. humanity as a whole, or that which even includes all sentient beings and Earth—given we still admit, for now, that the computational power over the globe or to be "utilized" in any "moral calculation" is yet to calculate for literally the entire humanity.3 Therefore, if, in principle, we factually cannot act upon the actual aggregated happiness, any action a utilitarian conceives to be conducive to any partially/locally aggregated happiness may, in reality, be detrimental to the actual aggregate happiness or utility—which is consequently against a full utilitarian ideal. I personally, for now, cannot see a way out for the utilitarians.



	Of course, there is a gap between what's in principle and what's practically possible. So that a utilitarian may claim that it is indeed only a normative guidance, acknowledging the realistic limitations.3 However, it seems to at least make the principle itself more contestable than sound in light of the realistic limitations in the first place—or, in other words, as an ultimate barrier. In fact, this concern may broaden into account of diachronic utilities4—not only should we consider whether or not the currently living humans, but also what about the future living ones, or all possible humans that could have been living? That said, the realistic limitations could at least present themselves in terms of the causal complexity and temporal uncertainty when things scale up rapidly without constraints. Nevertheless, someone may suggest that the boundary problem affects all theories equally.5 However, it seems that utilitarianism is especially vulnerable to this problem, as it indeed attempts to make a normative claim depend on the total consequence across all humans. Meanwhile, it is not necessarily true for other alternative theories to claim that each action is bound for all humanity. This makes it specifically problematic if (at least the metaethically strong version of) utilitarianism fails to address realistic limitations.





Is rule-utilitarianism better than or worse than act-utilitarianism? Why?



	Interestingly, Smart's rejection of the cognitivist stance and renouncing the attempt to prove act-utilitarianism altogether demonstrate a metaethical/epistemic humility and honesty that Mill doesn't. On the comparison between rule and act-utilitarianism, I won't really use terms like "better than" or "worse than." What Smart suggests is that rule-utilitarianism risks degrading into rule worship, becoming less utilitarian in the process—failing to deliver what utilitarianism initially promised. Therefore, act-utilitarianism sits closer to the original definitional core. But here's the problem—even with Smart's acknowledgeable concession—appealing to generalized benevolence as a sentimental attitude, or more realistically, as a shared strategy—the framework never escapes the question of defining consequences from each of our incomplete6 and biased perspectives.



	Metaethically, this move indeed weakens utilitarian morality from a normative appeal into what Smart treats as a strategy he hopes to share with his audience. In that case, Smart's response to the McCloskey dilemma should never be considered the act-utilitarian solution but a derivation Smart himself made using his proposed strategy, and someone else fully committed to act-utilitarianism too could derive different calculations leading to different decisions—due to different scopes of calculation, different information access, and different definitions of what counts as consequences or what it means for someone to be better or worse off. At its core, even if we grant generalized benevolence as a starting point, there are differences in defining or understanding what "general benevolence" itself means across cultures and among individuals—even if we all agree upon the practical strategy that act-utilitarianism suggests, we could still behave drastically differently and make distinct decisions in identical scenarios.7



	In fact, the common sense debate would be subjected to a similar reasoning. As common sense varies across populations, one person's utilitarian derivation could conflict not only with their own common sense but also with others' versions of common sense and their subsequent utilitarian derivations. I won't argue this shows utilitarianism is objectively false or bad, but I do think any individual who allegedly follows Smart's utilitarianism still need to reflect upon whether their own standards for calculating utility harmonize or jar with which common sense and whose common sense, if any, and then to acknowledge that none of those calculations can be objectively validated, even among people who follow the same kind of utilitarianism.



	That said, when discussions turn to demandingness, the first issue is obviously which standards or which perspective of utility becomes the benchmark for sacrifice. Even if we take for granted that everyone shares the same standards, it remains applicable only when everyone truly believes in those standards. Otherwise, utilitarians who believe in demandingness eventually become moral imposers on others—which actually leads us towards state consequentialist cases like Mozi's.





Do you think Mozi's notion of jian'ai is a kind of consequentialism? Why or why not? (Relatedly, should jian'ai be understood or translated as impartial care or inclusive care? Why?)



	If we consider consequentialism's scope of information and calculation, Mozi's state consequentialism arguably treats the scope as bounded by the state. This gives some evidence for interpreting jian'ai (兼爱) as "inclusive care"—there's a clearly defined scope of inclusion here. However, I still find "impartial care" more appropriate as it serves as a closer structural demonstration of the partiality-versus-impartiality contrast, which is the key mechanism driving Mozi's entire argument. In that sense, both translations may be appropriate from different perspectives: impartial care is closer to the ideal or core mechanism given the structural emphasis on 別 (partiality) as the problem, the logical opposite—impartiality—carries more argumentative weight, while inclusive care points towards the practical, applicable scope of the idea.



	Similarly, I'd say whether jian'ai is consequentialist presents valid arguments on both sides. On one hand, there's explicit calculation of benefits and harms, objectively measurable effects in Mozi's teaching. On the other hand, one key problem—at least from my perspective—is the appeal to Heaven's will, which makes jian'ai subject to heavenly command (potentially regardless of potential consequences), and poses the question whether we value jian'ai because consequentialist perspectives make it desirable, or because it is heavenly given? Even if we acknowledge the possibility that Heaven commands it because it has good consequences, this doesn't resolve the problem of why we consider it a desirable standard in the first place.



	Specifically, in the case of state consequentialism, I'm unsure if it's completely compatible with what we typically consider consequentialism, though it could constitute a kind of consequentialism. The problem is that typically, we assume consequentialist calculation is conducted on an individual level by each person in each case, keeping utility maximization in mind. But with state consequentialism, authority conducts the calculation for everyone, so that individuals are both relieved from—but also stripped of—the calculation, becoming to some extent "morally obedient" rather than genuine agents. Therefore, it turns what should be universal guidance for individuals into what is practically, on an individual level, a rule-based command. As the mechanism of consequentialist calculation becomes less significant or even eradicated for the individual, for someone living under state consequentialism, practice becomes effectively following whatever the state commands based on whatever calculation the state made, arguably degrading more severely than rule-utilitarianism does. 



	While these two problems—whether jian'ai is "heavenly given" and whether it's "state-defined"—exist on different levels, the former question bothers me most. A week ago or so, I had a debate—or less of a debate, more of a fact-check—with one of my old friends regarding our metaethical stances. As we know each other well enough, it's even easier to see my relativist stance against his objectivist/traditionalist one when we both found more reasons to hold our respective positions. But what still surprises me, or even physically disturbs me back then when merely looking at the phrase he literally used, is when his argument returns to the "heavenly given deity/dignity" of mankind. In that sense, if Mozi truly relies fundamentally on the heavenly aspect—not as rhetoric or heuristic but as genuine justification—there may be substantial incompatibility. To my relief (or misunderstanding), Mozi's case eventually appeared more appropriate to be interpreted as heuristic within that specific context. But even if we bracket the heavenly will—treating it as rhetorical rather than fundamental—the authoritarian structure remains troubling when the theory requires centralized calculation by state authority, ideological uniformity, strict rewards and punishments, obedience cascading up to the Son of Heaven. As a result, the question becomes whether we can extract the diagnosis and mechanism without the coercive enforcement.



	The diagnosis itself seems salvageable as partiality—caring more for one's own family, state, faction—generates systematic conflict. When I prioritize my group's interests over others, and others do the same, we end up in competition that produces the observable harms Mozi catalogs (warfare, theft, disorder). Meanwhile, the mechanism proposed—reciprocity and impartial treatment—also appears separable from state coercion—If I care for others parents knowing others will care for mine, if we treat each other's interests as equally weighty, coordination problems dissolve without requiring top-down enforcement. But this extraction faces a scaling problem still. Reciprocity works in small, repeated interactions where one can observe others' behavior and adjust one's own accordingly, and impartial treatment might be achievable among people who share context and can monitor each other. At the scale of states—or humanity as a whole—how do we sustain these mechanisms without the authoritarian infrastructure Mozi assumes? Perhaps the answer is we can't, at least not reliably, which is probably why Mozi reaches for centralized power.



	There's another path to impartiality worth noting, though it operates on a radically different level. Buddhist teaching on anātman—"no self"—suggests that the very category producing partiality (a bounded self with "my" interests distinct from "yours") is a conceptual construction. If the self is empty, if there is no stable entity to prefer over others, then, of course, impartiality becomes not a demand we impose on ourselves but a recognition of how things already are.8 This moves impartiality from a behavioral requirement (act as if everyone matters equally) to something closer to ontological insight (there is no "mine" to prefer in the first place). Eventually, whether this actually constitutes a foundation for ethics or merely a psychological technique for achieving impartial behavior—assuming we could achieve it at all—remains open, as I personally would regard any "no-self" state fundamentally different from regular human nature and thus probably suits another category of morality.



	Stepping back from the particulars of Mozi, Mill, and Smart, it seems to me that consequentialism might be less a substantive moral doctrine and more a style of justification. All three justify norms with the method—"judge actions by outcomes"—remains constant, while differing radically in what they're maximizing (pleasure, benevolence, state goods), how the calculation proceeds (individual agents, shared sentiment, state authority), and what scope they consider (all sentient beings, those we can reasonably calculate for, the bounded state). In essence, a Confucian reasoning from graded love, a Mohist from impartial care, a utilitarian from utility maximization—all are consequentialist in structure but divergent in substance. To my relativist stance, there's no view from nowhere to adjudicate between them, only the recognition that each framework is internally coherent when judged by its own standards.



	But, again, this raises another uncomfortable question, asking what the point of consequentialism is at all in that case. If it's merely a loose structure consistent with radically different values and if it doesn't constrain our actual choices beyond requiring we offer some justification by effects, then almost everyone is already a consequentialist of some kind to some extent, as most people already consider consequences from their own perspective when making decisions. The claim "you should be a consequentialist" becomes either trivially true (everyone already is) or question-begging (you should care about these consequences calculated in this way). Going back to the totality problem, if consequentialism requires calculating utility across all affected beings—humanity as a whole, all sentient creatures, possibly future generations—then it's practically impossible and thus fails as action guidance. But if it merely requires we consider consequences within whatever scope we can reasonably calculate, then it's no longer distinctively consequentialist. It's just what anyone reflective does anyway, regardless of their moral framework.





1 Besides my gut-level reluctance against its implicit justification for treating humans as means. ↩

2 Surely could be wrong or simply fail to see the possibility of compatibility between these two layers of the argument? We will see. In the worst case, though, it could be effectively saying: "Whatever you value, that's what I'll call happiness, so just maximize it." Although Mill indeed doesn't want it—he does argue for some pleasures to be higher than others (Socrates dissatisfied > fool satisfied), and there's an objective standard (general happiness) by which to judge actions. So, at the end of the day, the tension still seems to hold between the perspectivistically justified premise and the objective elements in the principle. Otherwise, to keep the objectivity intact, there may be some truly reserved standards for happiness itself. ↩

3 As we may consider Smart's response as a version of such, saying "we use imperfect methods, but they're the best we have," there is already a concession made in light of acknowledging we can never actually maximize utility but only act on our best estimates.  ↩

4 Although Smart also refers to the present and future total situations, what is emphasized more there is more like "before and after" to my understanding, rather than the long-term consequences I'd mention here. ↩

5 In fact, on balance, this is indeed a part of my consideration that eventually leads me towards a modest relativist stance. ↩

6 Recalling the boundary problem discussed earlier—we cannot actually calculate across all affected beings, but only make partial estimates that might contradict the full utilitarian ideal. ↩

7 In other words, again, a procedural universality versus a substantive universality—we might all use consequentialist reasoning (procedure) without converging on the same answers (substance). ↩

8 Though this might be another case of extracting what appeals to me from a tradition I don't fully inhabit. The connection here is speculative at best, and probably says more about my own discomfort with bounded selves than about what either Mozi or Buddhist ethics actually require. ↩