We study the spin-orbit coupling induced by the splitting between TE and TM optical modes in a photonic honeycomb lattice. Using a tight-binding approach, we calculate analytically the band structure. Close to the Dirac point,we derive an effective Hamiltonian. We find that the local reduced symmetry (D 3h ) transforms the TE-TM effective magnetic field into an emergent field with a Dresselhaus symmetry. As a result, particles become massive, but no gap opens. The emergent field symmetry is revealed by the optical spin Hall effect.PACS numbers: 71.36.+c,73.22.Pr, Spin-orbit coupling in crystals allows to create and control spin currents without applying external magnetic fields. These phenomena have been described in the seventies [1] and are nowadays called the spin Hall effect (SHE) [2,3]. In 2005, the interplay between the spin-orbit coupling and the specific crystal symmetry of graphene [4] has been proposed [5] to be at the origin of a new type of spin Hall effect, the quantum spin Hall effect, in which the spin currents are supported by surface states and are topologically protected [6,7]. This result has a special importance, since it defines a new class of Z 2 -topogical insulator [8], not associated with the quantization of the total conductance, but associated with the quantization of the spin conductance. However, from an experimental point of view, the realization of any kind of SHE is difficult, because spin-orbit coupling does not only lead to the creation of spin current, but also to spin decoherence [9]. In graphene, the situation is even worse, since the spin-orbit coupling is extremely weak. Deposition of adatoms has been proposed to increase the spinorbit coupling [10], and it allowed the recent observation of the SHE [11], but associated with a very short spin relaxation length, of the order of 1 µm.On the other hand, artificial honeycomb lattices for atomic Bose Einstein Condensates (BEC) [12] and photons [13][14][15][16][17] have been realized. These systems are gaining a lot of attention due to the large possible control over the system parameters, up to complete Hamiltonian engineering [18,19]. In BECs, the recent implementation of synthetic magnetic fields [20] and of non-Abelian, Rashba-Dresselhauss gauge fields [21] appears promising in the view of the achievement of topological insulator analogs. Photonic systems, and specifically photonic honeycomb lattices appear even more promising. They are based on coupled wave guide arrays [22], on photonic crystals with honeycomb symmetry [23], and on etched planar cavities [17]. A photonic Floquet topological insulator has been recently reported [24], and some others based on the magnetic response of metamaterials predicted [25]. In photonic systems, spin-orbit coupling naturally appears from the energy splitting between the TE and TM optical modes and from structural anisotropies. Both effects can be described in terms of effective magnetic fields acting of the photon (pseudo)-spin [26]. In planar cavity systems, the TE-TM effective field bre...