The intuition behind deontology is one almost everybody shares somewhere. Most people think there are things you simply may not do to a person, whatever good would follow — that a doctor may not kill one healthy patient to save five who need organs, even though the arithmetic favours it. Deontology is the attempt to give that intuition a systematic foundation.
Kant's version is the most developed. He argued that morality cannot rest on consequences, because consequences are contingent and outside your control, whereas moral worth must lie in something fully yours: the principle you act on. What makes an action moral is acting from duty, out of respect for the moral law, rather than from inclination or expected benefit.
The categorical imperative is his test. Act only on a maxim you could at the same time will to become a universal law. Take lying to obtain a loan: universalise it, and promising collapses as an institution, since nobody would accept promises. The maxim destroys the very practice it depends on — so it fails. Note what this test does: it identifies principles that are self-undermining when generalised, not principles with bad results.
His second formulation is more intuitive and does much of the work in practice: act so that you treat humanity, in yourself and in others, never merely as a means but always also as an end. This does not forbid using people — you use a taxi driver as a means to travel — but forbids treating them merely so, in ways that bypass their rational agency. Deception and coercion are the paradigm violations, because both work by preventing someone from choosing on the basis of the truth.
Deontology is not only Kantian. Rights-based theories give individuals claims that others may not override for aggregate benefit. Ross argued for a plurality of prima facie duties — fidelity, gratitude, non-maleficence, justice — none absolute, which must be weighed by judgement when they conflict. What unites these is structural: they impose constraints on action rather than goals to maximise.
The objectionsWhere the theory strains.
Catastrophic rigidity. The famous case: Kant argued you may not lie even to a murderer asking where your friend is hiding. Almost nobody accepts this, and it exposes the cost of refusing to weigh consequences at all. Most contemporary deontologists adopt threshold views on which constraints can be overridden by sufficiently severe outcomes — which is more plausible and arguably concedes the utilitarian's central point.
Conflicting duties. If duties are absolute, what happens when two conflict? A promise to be somewhere and an obligation to help at an accident cannot both be met. Ross's prima facie duties handle this by allowing judgement, at the cost of a decision procedure.
Where do duties come from? Utilitarianism has a clear answer about what grounds morality. Kant's derivation from rationality alone is powerful but contested — critics argue the universalisation test either rules out too little, since maxims can be described narrowly, or too much, depending on how they are framed.