Reader Bob Hunt makes two bold assertions:
"Conscience is a function of the intellect that guides us in choosing to act for good or for evil."
"It is unconscionable to demand that a person act contrary to his or her certain conscience."
Both of these assertions are incorrect on a number of grounds.
First, Hunt asserts that Conscience is a function of the intellect; the obvious interpretation is that that matters of conscience are rational and supportable by reason. This is not the case; worse yet, if it were true, it would invalidate his second assertion. If matters of conscience were indeed rational, then merely asserting one's "certain conscience" would be unnecessary. One should invoke the actual underlying reasons, not the fact of the conclusion. It is well-understood that people's conclusions are not perfectly reliable: they do not always reason correctly, they may they may be ignorant of relevant facts, or they may have false-to-fact beliefs.
Normative principles are fundamentally subjective, i.e. they are preferences. We cannot establish an objective normative principle syllogistically, because the fundamental premises of a syllogism are subject to the Universal Philosophical Refutation: they can always be rationally denied or contradicted. (Fundamental premises are by definition not themselves syllogistically established.) We cannot establish objective normative principles evidentially, because they are categorically unfalsifiable.
Normative preferences are rational only in the sense that they have a causal history: why and how someone forms a normative preference is a matter of scientific truth (albeit difficult or intractable to actually determine); the content of that preference, however, is not truth-apt: it just is.
Let's unpack Mr. Hunt's second assertion: "It is unconscionable to demand that a person act contrary to his or her certain conscience."
Since he says, "It is unconscionable..." he is stating a matter of conscience, and therefore a subjective normative preference. He states a normative preference regarding "demands", which implies some sort of social context; demands are meaningful (except for trivial cases) only among at least two people, and a demand (as opposed to a request) in this context implies some sort of coercion. He explicitly references compulsion, rather than prohibition. Finally, the preference is unqualified, even in the original context: he's stating a preference with universal application.
Mr. Hunt has to make the distinction between compulsion and prohibition; it would be wildly counter-intuitive to demand an exception from socially established prohibitive norms on the basis of one's certain conscience. I might (hypothetically, of course!) believe that it is ethically compulsory to murder my neighbor, but very few, I suspect, would exempt me from laws against murder by virtue of the fact of my "certain conscience."
All (coherent, reasonable) normative principles objectively specify some set of actions (in objectively specified circumstances) and a set of alternative actions; the preference privileges one set as "good" and deprecates the alternative as "bad". The only substantive difference between a compulsion and a prohibition is which of these sets are well-defined. For example, a prohibition against murder defines and prohibits the actions that constitute murder, but does not define the actions that do not constitute murder. Contrawise, the compulsion that I pay my income taxes defines and compels the set of actions that constitutes paying my taxes, but does not define the actions that constitute not paying my taxes.
The problem arises when both the preferred and deprecated actions are well-defined. I am, for instance, compelled to stop my car at a red light, but I am equally prohibited from driving my vehicle into the intersection. When all the relevant actions — preferred and deprecated — are well-defined the distinction between compulsion and prohibition becomes meaningless.
And we see in medical ethics in general that the preferred and deprecated actions are well-defined: either the physician (or other medical professional) denies the treatment or he performs the treatment; if Mr. Hunt objects to compulsion, we can simply prohibit physicians from denying treatment. So Mr. Hunt's objection to compulsion in medical ethics is therefore meaningless; Mr. Hunt is obviously not arguing that medicine should be subject to no social norms whatsoever.
Furthermore, even in situations where the distinction on the basis of good definition is relevant, the objection to compulsion is also wildly counter-intuitive. We are compelled in a meaningful sense to pay our taxes, to serve on juries, to testify when subpoenaed. We are compelled to fulfill our contractual obligations. And, most importantly, most of us are compelled to do our jobs, to comply with the reasonable orders of our superiors, to fulfill the standards of our professions. And it is precisely against those who might have some personal conscientious objection to an obligation that the compulsion is directed: A compulsion that allows exceptions based on individual characteristics is no compulsion at all.
I must admit to some sympathy for Mr. Hunt's second assertion: I do in fact find compulsion objectionable to a degree, and I admire (at least in theory) the radical anarcho-libertarian society depicted in And Then There Were None. But Russell is writing a fantasy; in the real world it's at best confused to object in principle to compulsion.