Why a 2-state photon is interpreted as spin 1? Both Ising spin and photon's polarization degree of freedom are used in quantum information as Qbit implementation.  They both have 2 level state systems, which means mathematically their state could be described by a vector in 2d vector space. But the physics on them are completely different.  We call Ising Spin spin-1/2 system and the photon spin-1.
Which settings in the theory bring this distinction?
 A: A particle in a relativistic theory is considered be an excitation of a quantum field whose hamiltonian, in the absence of interactions, possesses the symmetry of the inhomogeneous lorentz group (rotations and translations of spacetime). 
From group theory we know that eigenstates of the hamiltonian are states of definite symmetry with respect to the hamiltonian's symmetry group. In other words, we can assign to these eigenstates quantum numbers associated with their symmetry properties, in the same way that we can classify eigenstates of particles in a 1D reflection-symmetric potential as even ($+$) or odd ($-$). Since a particle is an excited state of a quantum field whose hamiltonian commutes with the inhomogeneous lorentz group, we should then be able to assign associated symmetry quantum numbers to particles.
For the inhomogeneous lorentz group it turns out spin (or total angular momentum) is one of these quantum numbers, and can take any half-integer value (take spin $j$ for example). Further, we can assign a "helicity" quantum number $m_j$ which tells you the projection of the particle's spin along the direction parallel to its momentum. 
For massive particles the helicity can take any value from $-j$ to $+j$, in integer steps. For massless particles, however, the structure of the lorentz group constrains helicity to either $-j$ or $+j$. So the helicities for any massless particle (excluding spin $j=0$, since $-0=+0$) constitute a qubit.
see Wigner (1939) 
