Originally posted by Tyger
The only problem with this picture is that Gravity doesn't act between two bodies the way that other forces do. It acts on the space around a massive body. So the picture of the spin-2 graviton may not even apply in Nature. I think we're still looking for a competent quantum theory of Gravity and it would be unwise to close our minds to any viable possibilities.
First you need a mind open enough to want to understand the current state of theory and be honest about how it reflects on your own ideas.
Full QG isn't needed to understand gravity's behaviour at lower energies where GR, and hence it's linearization describing a self-interacting massless spin-2 field - the graviton- applies. This means that whatever gravity's true nature, the graviton will remain an indispensable concept.
More generally, the modern concept of particle itself holds up only as a group theoretic approximation: Briefly, particles are defined in terms of their mass and spin which label - and this is their real significance - the representations of the poincare group under which their states must transform to preserve invariances related to the local geometry of spacetime. However, it is only in minkowski space that poincare symmetry holds globally. Since curved spacetimes are only locally flat, the particle concept only holds as an approximation in regions over which curvature varies sufficiently slowly.
Let me describe in schematic terms how the gravitational field is viewed in terms of spacetime curvature and it's relation to the particle viewpoint at lower energies where GR holds.
Begin by representing the connection between spacetime curvature and energy density in general relativity schematically by the relation
(spacetime curvature) = G x (mass-energy density)
where G is the gravitational constant which thus characterizes how much spacetime curvature would be created in the presence of some given mass-energy density. We can understand why mass-energy density, rather than just mass-energy, appears in the above relation by noting the basic inverse square law (curvature has units of inverse distance squared) behaviour of gravitational force it implies, and then observing that if a planet's density were increased by gradually shrinking it, it's inhabitants would feel themselves growing heavier as they approached the planet's centre of mass.
Can this be taken further to build a more operational view of gravity? No, not directly, because quantum theory is needed to understand the current picture of fundamental interactions. But the main idea - and I think you've seen this before - can be understood by observing that two people on roller skates throwing a ball back and forth will roll away from each other (as a result of momentum conservation). Somewhat similarly, electrons are mutually repelled as a result of exchanging photons, and likewise, gravity is the result of graviton exchange.
However - as you already know - the analogy doesn't work as well in this latter case because gravity is attractive, the explanation of which is - as I mentioned - not intuitive because of it's quantum theoretic origin. Repeating it, elementary particles are classified in terms of spin, mass and charge, and only massless even-spin particles can mediate long-range attractive forces between like charged bodies: suffice it to say that gravitons are spin-2 and, like photons (which are spin-1, consistent with electromagnetic repulsion between like charged bodies), are massless (Actually, since theories with bosonic gauge fields like EM and GR exist as classical theories, this can be seen from looking at the interaction potentials in the corresponding classical lagrangians, it's just that spin itself is a purely quantum theoretic concept which confuses this intuitive classical aspect of the picture).
The reason that gravitational and electromagnetic fields - whose quanta as you know are the graviton and photon respectively - grow stronger as their sources are approached is also quantum theoretic. Again, briefly, because of the time-energy uncertainty principle, the distance from a source that field quanta can travel decreases with their energy. So, for example, the strength of the gravitational field grows stronger as a gravitating body - like a planet say - is approached because gravitons of progressively higher energies are encountered (analogous statements hold for the electromagnetic field).
So what's the connection between the classical notion of spacetime curvature, and the quantum view of the gravitational field as a continuum of gravitons? The answer is that because gravitons - unlike photons - interact with each other, they assemble themselves in a way that's in fact governed at lower energies by general relativity, so spacetime curvature is merely a large-scale structural quality of the resulting gravitational matrix.
Although this picture must - as you know - break down at very high energies, it provides no real foundation for your view of gravitational attraction as paradox.