Joel Feinberg in an influential article ["Some Conjectures about the Concept of Respect," Journal of Social Philosophy 4 (1973) pp. 1-3] suggested that we can divide forms of respect into three different kinds:
(1) Respekt: This is the sense of 'respect' in which one respects a rattlesnake. Its object is the dangerous, and it is the opposite of negligent or contemptuous disregard.
(2) Observantia: This is deference to others; its object is someone who has a claim on us. Feinberg suggests (not entirely plausibly, in my view) that this is a sort of generalization of (1).
(3) Reverentia: This is awe and its object is the sublime or sacred; it is the sort of thing, for instance, that Kant means when he says that we should have Achtung, respect, for the moral law.
A key difference, it seems to me, between respekt and observantia is that observantia is intrinsically concerned with communication. You cannot actually defer to others except insofar as there is some kind of communication between you and them. This is one reason to think it doubtful that social deference can genuinely be a direct of wary regard -- observantia or deference already presupposes recognized social relations and communications, whereas respekt does not.
An interesting question is whether reverentia is also distinguished from respekt in this way. Sublimity, it seems, need not belong to something with which you can be in a social interaction; one thinks, for instance, of the sublimity of a thunderstorm or of the 'starry sky above'. (This is, it should be noted, something that has been denied; Thomas Reid, for instance, holds that sublimity does, in fact, involve, at least indirectly, a mind other than one's own.) But it is also the case that in at least some of these cases, like the thunderstorm, one could argue that this is a case of the sublime calling forth respekt, not reverentia. Indeed, if you look at many early influential accounts of the sublime, the reaction to it is almost entirely in terms of dread and fear, like you would expect for respekt. Thus respekt and reverentia both seem to be something one can have for the sublime, which distinguishes them from observantia: the sublime is not something to which one merely defers. This makes it tempting to suggest, for symmetry, that reverentia and observantia share, against respekt, this communication aspect. To be sure, the standard examples from Kant, the starry heavens and the moral law, don't immediately suggest this to us, but this is complicated by the fact that in Kant's account our ability to recognize these as sublime is at least linked with our capacity to think of these things in personal terms (i.e., in terms of designer and Lawgiver, respectively), even if the link is not straightforward; and historically they both have, in fact, been thought of in such terms. And it's certainly undeniably that awe can involve this communication.
