The truth is more subtle than this. Kant does not take the moral life to be one in which emotions and sentiments of every sort have no place and are at best a distraction. Rather, he distinguishes between two species of feelings and desires: those that are "pathological" (by which he does not mean what modern psychologists mean by "pathological" but rather something more like "desires we just happen to have because of heredity and environment, because of our physical natures and the way that external forces interact with them"), and those that are moral. It is the former that Kant finds suspect. Here is what Sullivan says:
However similar they may be to merely pathological feelings and desires in other respects, moral sentiments are radically different in one critical respect: They do not have their origin in empirical sources outside our own reason so that we passively feel them. Rather, they are the subjective effect of our prior recognition of the objective and absolute binding force of the moral law. It is just because they are caused by reason alone that they can be moral feelings.
Kant makes the contrast between the two this way: our desires are merely pathological when we “represent something to ourselves as good, if and because we desire it,” whereas when we experience moral desires, “we desire something because we represent it to ourselves as good” on the basis of a prior judgment of the moral law inside us. Morality does not obligate us because we first find ourselves interested in it; rather, we find we are interested in doing our duty because we first recognize that it is our duty.
A foundational moral sentiment is reverence/respect for the moral law itself. And it is clear that Kant thinks such reverence or respect motivates us--and our acts have moral worth if and only if this is our motive. But this raises the natural question, "Isn't someone who obeys the moral law out of respect for the moral law really just following a hypothetical imperative, that is, an imperative that is contingent on the sentiments you happen to have?" In other words, isn't respect for the moral law just another "inclination" among others, so that whether you are motivated to act on the moral law is going to depend on what your inclinations happen to be? Here is how Sullivan addresses this issue:
The only incentive that can motivate us to adopt and act on the motive of duty, Kant writes, is the reverence or respect we feel for the moral law. This is not a feeling we need to try to “will” ourselves to have, for it arises irresistibly from our unavoidable recognition of the moral law. In fact, it is because we all inevitably experience respect for the moral law that we can be said to have a “predisposition” to be morally good. Moreover it is an incentive of such “herculean strength” that it enables us to offset the influence of all the “vice-breeding inclinations”. Because it follows from our recognition of the Law of Autonomy, it is the only motive that can give our actions genuine moral worth.
For us, then, respect for the moral law is the subjectively necessary side of our consciousness of duty. As an emotion, respect resembles fear in that we recognize that the moral law may rightfully demand the denial of self-love; it is also like love in that we recognize that this law originates in our own reason and is something we impose on ourselves. Finally, it is like nonmoral emotions in that it is a feeling that gives us an “impulse to activity” on which to act. But it is not just a feeling of attraction or of aversion, for it is not just a feeling like sympathy or empathy or love, all of which can be understood in naturalistic terms.
So important a role does respect have that, in his third “proposition” about the human good will in the Groundwork, Kant defines duty as “the necessity to act out of reverence for the law”. In his defense of human moral agency in the third chapter of that work, he again discusses moral interest, to try to ensure that his readers do not interpret their moral agency as pathologically caused. As we saw in the last section, moral sentiments must not be invoked as causal explanations for any part of our moral life. Regardless of how an empirical psychologist might interpret respect, from the moral point of view it represents a ground that is not explicable in naturalistic terms. Because acting from duty means acting out of reverence for the moral law, that law remains the ultimate objective motive of moral actions.
There are several questions we can ask about this: First, does Sullivan get Kant right? Second, if so, are moral sentiments as described here distinct enough from "pathological" inclinations to provide an answer to how morality, not inclination, can be said to be our motive when we act morally? Third, even if we separate out moral sentiments from inclinations and concede that it is only the latter which are irrelevant to moral life and at best a distraction, is that enough to free Kant from the original challenge that the moral life is in some sense coldly dehumanizing?
With respect to the last question, the following anecdote from Kant's life, related by Sullivan, may be worth reflecting on:
An incident occurred about a week before his death that has often been used to illustrate how Kant guided his relationships with others by the disinterested interest of moral respect, which he nonetheless called the “courtesy of the heart.” Desperately weak, mentally unable to concentrate, and virtually blind, Kant insisted on rising and remaining standing until his doctor had seated himself. With great effort Kant then remarked that at least “the sense of humanity has not yet a abandoned me”.