Does Kant justify intuitions existing without understanding?

I would have just placed this in a comment, but the comments are spilling over. I am probably missing something here, but I just don't see the problem. Kant does seem to use "intuition," "understanding," and "concept" in evolving ways, but I don't think, for starters, "understanding" and "concept" are interchangeable.

Intuitions that are unmediated sensations already contain, or are formed within, space and time, sort of primitive concepts. But certainly they can be perceived without being connected into judgments, which is the active task of "understanding" things and compiling knowledge.

The reverse is not possible. We can't have "empty" concepts without some material to conceptualize, or at least they provide no knowledge. But we can and do have "blind" intuitions, as animals or infants might, and even react, to a flash of pain maybe, without then adding a predicate, knitting together concepts, and forming a judgment. This latter is a human/spiritual faculty or set of faculties, not, in Kant's thinking, a property of sentience.

The example of your dream, of course, is loaded with overactive concepts mauling misplaced intuitions. But when we read a book, say, we may have an unmediated representation of the white paper under the letters, but we aren't engaging that whiteness with concepts, until we write about it, that is. As I say, maybe I'm missing something here, which I often do.

As to why he doesn't collapse intuition and understanding, well then he sure wouldn't be Kant. It is essential that he both interrelates and distinguishes them to answer "how a prior synthetic concepts are possible," which is more or less to answer how scientific knowledge is possible.

For Kant, intuitions do not mean only one thing. For instance, every sense-datum that is received by the faculty of sensation (Sinnlichkeit) is directly called an 'empirical intuition' (empirisichen Anschauung). In the form of empirical intuitions, they are not yet appearances, for in order there to be an appearance, empirical intuitions have to be 'molded' under pure intuitions, firstly by space (Raum) and then time (Zeit). Only then they would become appearances (Erscheinungen). However, these appearances necessarily go under the categories of the understanding (which I believe what you meant with concepts are categories) in the faculty of understanding (Verstand). Therefore, whether they are the products of imagination (Einbildungskraft) or empirical intuition which is received from the thing-in-itself (Ding an sich), there is nothing in mind which does not get molded by the concepts.

The only answer I could muster after reading various positions on this is, that it is contested.

There are two forms of interpretation - conceptualist and non-conceptualist. Former thinks Kant cannot establish that there do exist such intuitions, and therefore the difference between concepts and intuition is merely that of difference in the working of them. The latter, of course, does think that Kant is saying that humans do have a cognition (used here in the broader manner) of appearances lacking in concepts.

Both of them cannot be justified by looking into the CPR only, Kant in my opinion does not do justice to this doubt (as far as I know, at least).

There are multiple levels in which concepts fall short in determining and embracing intuition.

1. There are sensible representations that are unconscious, "obscure", no appercieved (this is inherited by leibniz's petite perception), yet they are given sensible representations. This is the strongest case of a sensible representation without a concept for it. Of course, even here, this representations are susceptible of being appercieved and even determined by judgment but, as I hope I'll show, intuition can be given without concepts in the sense that there can be undetermined experience, i.e. unknown things can be experienced.

2. Something we don't know happens, for example, a "sercio" hits you, you do experience the sercio hitting you but you have no idea of what a sercio is, still you have a sensible representation of it "that hard thing that hit you".

3. Now of course you can go on and study the "sercio", discovers its physical and chemical properties. By then you will have a better knowledge of the "sercio". Still, something of a contingency will remain and your concepts will never embrace and determine the "sercio" completly because of the multiplicity of sensation. The only way to have complete knowledge over something would be to build such a knowledge from top to bottom (intuitive knowledge) but human empirical knowledge is in it's constitution from the bottom up, and as such always lacking the first principle of determination (the inner determination) of the thing in question.

So there you have three perspectives under which intuition is non conceptualized. 1) it's not appercieved, 2) it's undetermined 3) it's undeterminable. You can of course think of it, which means that is part of experience and as such the process of conceptual determination is legitimate. Yet that can't win the overabbundance of multiplicity of intuition, which is coessential with it not being the product of a subject.

Kantian intuition is a mental 3D-image of a particular object. Like that chair in the living room. After it made a few appearances from different sides, you can close your eyes and enjoy the view of its intuition suspended a few feet above the floor and slowly rotating counterclockwise.

I'm sure Immanuel Aspergerovich would have explained it in his own words... Actually, I'd rather not be taking any chances, so... yes, concepts!

A concept of a chair doesn't pertain to any chair in particular (yeah, you can put that one down). In fact, the concept is not of a chair, but of chair-ness. That's why we cannot visualize the concept itself (we have no use for it consciously, and it's probably huge and scary). If we need to use an abstract char in simulation, we can use the concept to quickly instantiate an inner intuition of one, usually in hot pink leatherette and green stars on it.

Also, I found this treasure -- an actual Kantian-English lexicon, and it looks extremely decent.