In discussing almost any enduring controversial issue in ethics we usually come up against fundamental differences about what ethics is. This is what a moral theory tries to explain. It picks a plausible or common moral intuition, one that most people think is important (at least, initially), and runs with it. Sometimes right off a cliff...
Metatheories: principlism and particularism
One way moral theories can be divided is according to whether they try to unify moral judgments or deny that possibility.
Principlism is the view that one or another principle, rule, law, etc. is at the heart of morality. Such a principle would unify our moral judgments, make them rational and coherent, and less subject to being thought ad hoc or contingent or situational.
Particularism is the view that there is no such overarching principle from which moral judgments can be derived. Moral judgments are situational and highly contingent. Conflicts of moral intuitions are not even in principle resolvable. Sometimes the very same thing will be both right and wrong depending on where you stand. (In fact, all of the time!)
Of the four main or most well known moral theories---deontology, consequentialism, virtue theory, and moral sense theory, the first two are principlist theories, the second two particularist.
We will focus on the two dominant and principlist theories for this session: deontology and consequentialism. The other two are important and revealing of the whole moral picture but in a different way. They will be the topic of a later session on moral theory.
Larson on utilitarianism (the most common form of consequentialism):
"Consequentialists hold that choices—acts and/or intentions—are to be morally assessed solely by the states of affairs they bring about." ---SEP
Good states of affairs are the ones consequentialism promotes. And "good" is cashed out in terms of pleasure, happiness, utility and similar things. Sentience, the ability to experience pleasure or pain, is an essential requirement on the part of actors for this theory to have any bearing. This may rule out some animals (arguably) and even some humans (in some states or conditions), all plants, and the natural world, in whole or in part, as proper objects of moral regard. If these have any moral status, it is only because their well-being impacts those who are sentient.
• Some consequentialists profess monism about the good: it is variously dubbed pleasure, happiness, utility. (Ultimately, consequentialists do not question the desirability of physical well-being and survival. Supposedly, these "goods" are self-evident. Problem: are they? If so, can we say by what measure without being circular?)
• Some consequentialists are not monists. They import other values in addition to those enjoyable by sentient beings: such as a program about how these enjoyments should be distributed. Ex., Bentham's dictum: "each to count for one, and none for more than one". Problem: aren't we diluting consequentialism here? Suppose someone has superpowers when it comes to enjoyment, wouldn't they be worth several of the rest of us put together?
• Some consequentialists, like Bentham, are hesitant to admit of purely qualitative differences in pleasures and pains. The fear is that once you admit there are higher and lower pleasures---where a little of one kind may outweigh in value a lot of another kind, we start (perhaps surreptitiously) introducing measures other than utility in order to make these ranking judgments. J. S. Mill, for instance, thought there were higher and lower pleasures. He thought most of us would choose to be a miserable human being than a happy pig if we had to choose. But if this is the case, why? That is, if pleasure is supposed to be all that counts? If you think human status is preferable, no matter how pleasant or unpleasant, how pure can your consequentialism be? Something about merely being "human" seems to trump having pleasure. What could that be if we have already admitted that pleasure or the absence of pain is all that counts? Whatever it is that is mucking up the purely utilitarian picture here, it does seem to be extra-consequential.
Virtues of consequentialism
We can say these things in its favor:
• it is conceptually simple: we all know what pleasure and pain are.
• it's naturalism-friendly: pleasure and pain are empirically accessible.
• it's scalable: it does seem to explain best why and how we make decisions affecting large groups of people: the suffering of more people seems to matter more than that of fewer. (The standard "lever" trolley problem illustrates this.)
Problems with consequentialism
How possible, let alone practical, is it to be always weighing pleasures and pains?
Besides the very basic problem of how we discriminate among the varying weights of quantifiable pleasures and pains, is it really true that utility or happiness, whatever those are taken to mean, is all that really counts morally?
There is a conceptual problem for at least the most common forms of consequentialism: if it is really the increased presence of pleasure that makes for a better world, where is the value in the container of the pleasure? Or does the container not have value or have it only in virtue of containing pleasure? Consider the possibility that we might be replaced by pure pleasure-experiencing machines or organisms optimally designed for pleasure and nothing else? After all, nothing else has value... If this thought is even mildly disturbing to you, you must have some additional item on your normative agenda besides pleasure or the absence of pain that you are trying to serve.
(As a normative theory, consequentialism is supposed to be telling us what we should be doing, not just what we actually do. If it is only describing, not prescribing, it is not a moral theory, rather a theory of behavioral or social science. Science does not (openly) claim to demand that the world be a certain way, rather it describes how it, in fact, is.)
Also, to the extent consequentialism in its utilitarian form---the form that takes the happiness of the greatest number as goal---is taken seriously, it may be very hard on common moral intuitions. It takes a very dim view of partiality, for example, the idea that some of us matter---and ought to matter---more to us than others: her children for a mother as opposed to some stranger's children, my friends and relations versus your friends and relations to me, my teammates or compatriots versus yours (assuming they are not the same), members of this philosophy meetup club versus one of those other philosophy clubs, etc. A consequentialist will tell you that this is what is good about consequentialism: it keeps us from playing favorites and isn't playing favorites unethical? Well, is it? Other theories may have a better answer here.
Do we ever act knowingly without regard to consequences (pleasurable or otherwise) because we think it is the right thing to do? If you think so, you may be more sympathetic to:
Deontologists are more concerned with what is right than what is good (in terms of consequences). What is right is that which comports with some rule or law or principle that is the ultimate source of moral imperatives and is not (directly) related to any aim or design we might individually or collectively have. The source or authority of the law or principle is essentially tied to what we take ourselves to be. Deontologists think that what makes for the kind of being for whom morality matters has to do with the fact that the being can reason its way to decisions about behavior or how the behavior feels. Humans, normally, are rationally capable. Merely being subject to pleasure or pain does not make you a moral actor. A mountain lion who kills and eats a small child didn't and could not have done wrong. We may (or may not) have good reasons for tracking down and destroying the mountain lion, but the reasons have nothing to do with the moral guilt of the animal. To the best of our knowledge at this time, only human beings---some of them some of the time---qualify as moral agents because they are capable of rationally assessing their behavior against a principle that applies only to rationally (of course) capable beings.
Rational autonomy is critical to deontology. An agent who is rationally autonomous is not a slave to instinct, consequences, impulse, passion, habit, tradition, inertia, peer pressure, affinity, aversion, environment, etc. Such an agent can oppose any and all of these if that's what the moral law requires. In fact, when the going gets rough is exactly when the moral law often requires opposition to what you might be inclined to do otherwise. Morality hardly enters the picture when what it requires of you is something you would be inclined to do anyway.
Deontology, as we can see, is at least as demanding as consequentialism but in a different way.
So what are these "moral laws" or the one big principle "to rule them all" we are duty-bound to uphold?
Immanuel Kant's is the most famous attempt at pinning it down: he suggests that the right thing to do is always that done from a motive that could be made into a law governing any rational being in the situation. Your motive has to be capable of being made a rule that could be applied, without exception, to any rational agent. So, if I lie to you in promising to pay you back next Tuesday what you are loaning me now, I am doing wrong because if I were to try to make my motive to exploit you for my own ends into a universal law, it would essentially destroy the very idea of a promise. Promises would be conceptual contradictions: no one would take them seriously. A promise would mean nothing. If my lying in this case works for my ends, it does so because I am expecting that the general rule against lying promises will be kept by others, just not by me except when it suits me. I am assuming others will do the right thing and leave me alone and free not to. I want to make an exception in my own case.
This exceptionalism is what a rational being, fully invested in being a member of a community of other rational beings, cannot tolerate. If I nevertheless make lying promises, I am disavowing membership in this rational community. Reason becomes purely instrumental to my ends, and not a defining characteristic of the kind of being I am. My actions do not flow from a correct understanding of what I am. This is what it means to be immoral from a Kantian or deontological perspective.
The importance of rational agency is captured in another formulation Kant gives of the moral principle: never act in such a way that treats another rational agent as only a means to some end and not also as an end. This means others should not be used for any purpose in which they themselves are not complicit. (Keep in mind, no one is allowed to use themselves this way either: thus, if someone is knowingly abusing themselves, you are not allowed to help them do so. We have moral obligations to ourselves as well as to others because we count as much as any of them do. Hence, Kant thought suicide no less immoral than murder.)
Problems with deontology
Deontology is unforgiving in a different way than consequentialism. There is the famous case of the would-be murderer at the door asking you for the location (which you know) of an innocent victim. Assume you, yourself, are not in any danger: should you lie or tell the truth? Since lying is categorically forbidden to a deontologist, you must tell the truth even to the murderer because not to do so would be to refuse to honor the rational autonomy of the murderer. People need correct information to make responsible decisions. Assuming the murderer is not stark raving mad (in which case, Kantian morality wouldn't apply to him because he wouldn't be rationally autonomous), he (a murderer usually is a "he") is an autonomous agent like any other and worthy of being respectfully treated like anyone else. If he does bad with the information, that's on him, not you. He can't very well act wrongly or rightly (thereby exercising his autonomy) if he has bad information to work with. He is owed correct information simply in virtue of what he is, a being capable of exercising reason.
But what about the consequences? Deontologically speaking, consequences are neither here nor there as far as the morality of an act. The specifically moral issue is whether lying is wrong. And it is. And you, as the potential liar, are the one who will be judged for the lying...
This line of argument may be unacceptable to you. It may make you want to go back to consequentialism. Or not. Or something else...
Some striking differences between consequentialism and deontology
• Strict consequentialism has something good or bad to say about every breath you take. It is good or bad depending on the consequences and everything has moral consequences you are not allowed to ignore. Deontology allows for the possibility that some acts may be morally neutral. Whether you do them or not is optional and can be dependent on non-moral (for instance, aesthetic) values you may have. So long as the acts don't involve ununiveralizable (exception-creating) motives or express disrespect for rationally autonomous being---your own or anyone else's, an act is permissible. Thus there are three classes of acts: required, forbidden, and optional. While, from a consequentialist point of view, if some act seems optional (inconsequential), you must be missing something. You are not looking closely enough at consequences. Acts are required or forbidden. For a strict consequentialist, there is little or no middle ground.
• Consequentialism is prospect-oriented or forward-looking. Deontology is about prior constraint and backward-looking. Consequentialism's seeming progressive structure seems to appeal to many naturalist philosophers with their science-based, positivist, let's-see-what-we-can-do-about-it-outlook. Consequentialism seems constructive where deontology has a distinct rear-view mirror stare concerned with what we became aware of first---motive, what's transpired---before we consider what may happen or have any power to influence. Ultimately, deontology seems concerned with maintaining a kind of abstract balance in the universe insofar as things really are in our control. (But notice that, even if we had zero control over the consequences of our material actions, we may still claim control over how we think about them.) Consequences in the real world have only a tenuous relation to the thorough-going control required for moral responsibility. Deontologically, what happened is just as significant as what may yet happen---in fact, may be the only morally significant thing. There are some deep assumptions about time here: one theory prioritizes the future, the other the past.
A criticism of both of these two major theories is that neither considers that the here and now may be at least as important as anything leading up to it or following from it... but we'll leave these thoughts for when we discuss the particularist theories.
There are several varieties of consequentialism and deontology. See the SEP articles below. The focus here has been on laying out the fundamental differences of intuition that separate the two.
Larry Alexander and Michael Moore on Deontology:
Walter Sinnott-Armstrong on Consequentialism: