Now, this may seem like a bit of a long shot, but I think there's basis enough for this question to be valid.
In measuring volume and axises of movement, we're forced to rely on our perception. Perception, in turn, depends on our eyes and our brain to decode the visual signals received by our eyes. As most of us know, sometimes when the eyes are pitted against a paradoxical image, for instance images with conflicting colour spectrum, merging patterns and other stuff like that, the brain fills in the gaps and we in turn get visual anomalies -- essentially optical illusions -- the cause being that the signals are overpowering, stressing the brain into making and remaking conclusions about the perceived image, or related to our cognitive ability, with the brain making unconscious inferences to perceived images. In more rare, but often documented, cases the brain can make similar omissions or alterations for other types of processed information; audio, motion, sense etc.
So, let's hypothesise, for the sake of the argument, that the third dimension is a product of faulty perception. The cause of the faulty perception, in that case, would primarily be the lack of a third dimension. For one, our physiology changes -- our eyes are no longer spherical, but instead flat toroids, limiting our ability to focus right in front of us but giving us about the same all-around fov for the remaining 2 dimensions. It's evident right from the start that there would be a visual conflict involved, as no matter where you looked you'd be staring straight into infinity; possibly a self-replicating infinity, if the string-theory is added to the equation. The brain would have no other choice but to omit the actual image, and fill in the gap for the missing dimension off hand, to keep you living and breathing, as even something as simple as balance becomes an issue when vision overextends the ammount of dimensions available to the matter you yourself occupy.
Obviously, the third dimension is accounted for in mathematics, and visual representations have long since made their way into the world of computers. I'm not going to debate that part, as math may well be used to account for lots of things that will never actually see the light of day here in the real world, much like computer generated graphics are able to present us with myriads of items and scenarios that permanently belong to the realm of science fiction: in this context, there may very well be a third dimension, possibly even a 59'th one. But in the real world I personally believe that there's no way of knowing for sure whether or not the third dimension really exists.
Thoughts?