We show that the interlayer tunneling I-V in double-layer quantum Hall states displays a rich behavior which depends on the relative magnitude of sample size, voltage length scale, current screening, disorder and thermal lengths. For weak tunneling, we predict a negative differential conductance of a power-law shape crossing over to a sharp zero-bias peak. An in-plane magnetic field splits this zero-bias peak, leading instead to a "derivative" feature at VB(B || ) = 2πhvB || d/eφ0, which gives a direct measure of the dispersion of the Goldstone mode corresponding to the spontaneous symmetry breaking of the double-layer Hall state. PACS: 73.20.Dx, 14.80.Hv, 73.20.Mf Through a series of experimental and theoretical papers, it is now well-established that a bilayer of twodimensional (2d) electron gases (2DEGs), when the layers are sufficiently close, exhibits an incompressible Quantum Hall (QH) state at a total filling fraction ν = 1, [1][2][3], even when interlayer tunneling is negligible. [4]. The QH state is stabilized by the Coulomb exchange interaction, which leads to spontaneous interlayer phase coherence, with the layer phase difference φ, the "phason", as the corresponding Goldstone mode. Because the layer index may also be viewed as a pseudo-spin-1/2 degree of freedom, this state is also known as the QH FerroMagnet (QHFM), and the phasons as (pseudo-)magnons.The first direct experimental evidence for this mode was obtained recently by Spielman, et al. [5]. For bilayers with small separations d, they observed a giant and spectacularly narrow zero bias peak in the inter-layer tunneling conductance at low T , a feature that naturally emerges from our study of the QHFM. We find rich variations of the non-linear tunneling conductance G(V ) = dI/dV as a function of interlayer voltage V , tunneling strength ∆ 0 , disorder, temperature, and applied parallel magnetic field B . Two particularly appealing and testable predictions of our theory are: (1) a finite-bias feature in G(V ) at a voltage V B =hω(Q = 2πB d/φ 0 ) which tracks the collective mode dispersion relation ω(Q) as a function of B (φ 0 = hc/e) and (2) the appearance of a power-law negative differential resistance at higher bias in cleaner samples. For clarity, below we first summarize different regimes and results, and then give a brief sketch of the calculations.We focus here on voltages eV below the QH charge gap Ω, where the only available bulk excitations are phasons, described in the absence of impurities by the Lagrangianwhere v is the pseudo-magnon "sound" velocity, ρ s is the phase stiffness, [4] ℓ = hc/eB is the magnetic length, Q is the in-plane "magnetic wavevector" Q = 2πB || d/φ 0 due to applied B || = B ||ŷ , and ω = eV /h. The only other low-energy modes are edge states, which carry the in-phase, in-plane current at low temperatures. For weak tunneling, characterized by a "naive" dimensionless coupling δ 0 ≡ ∆/ρ s ≪ 1, the tunneling current density J(V ) can be computed in a controlled perturbation theory (but see below). There are thr...