As the quality of cosmological data continue to improve, it is natural to test the statistics of primordial fluctuations: are they Gaussian or non-Gaussian? I review a model which generates non-Gaussian adiabatic fluctuations from inflation. Current investigations suggest that there may possibly be a non-Gaussian signal in large angle cosmic microwave background anisotropy data. Statistics of microwave anisotropies could thus serve as a powerful probe of the very early Universe.

Strongly coupled gravitational systems describe Einstein gravity and matter in the limit that Newton's constant G is assumed to be very large. The nonlinear evolution of these systems may be solved analytically in the classical and semiclassical limits by employing a Green function analysis. Using functional methods in a Hamilton-Jacobi setting, one may compute the generating functional (`the phase of the wavefunctional') which satisfies both the energy constraint and the momentum constraint. Previous results are extended to encompass the imposition of an arbitrary initial hypersurface. A Lagrange multiplier in the generating functional restricts the initial fields, and also allows one to formulate the energy constraint on the initial hypersurface. Classical evolution follows as a result of minimizing the generating functional with respect to the initial fields. Examples are given describing Einstein gravity interacting with either a dust field and/or a scalar field. Green functions are explicitly determined for (1) gravity, dust, a scalar field and a cosmological constant and (2) gravity and a scalar field interacting with an exponential potential. This formalism is useful in solving problems of cosmology and of gravitational collapse.

A Green's function method is developed for solving strongly-coupled gravity and matter in the semiclassical limit. In the strong-coupling limit, one assumes that Newton's constant approaches infinity. As a result, one may neglect second order spatial gradients, and each spatial point evolves like an homogeneous universe. After constructing the Green's function solution to the Hamiltonian constraint, the momentum constraint is solved using functional methods in conjunction with the superposition principle for Hamilton-Jacobi theory. Exact and approximate solutions are given for a dust field or a scalar field interacting with gravity.

Hamilton-Jacobi theory for general relativity provides an elegant covariant formulation of the gravitational field. A general `coordinate-free' method of integrating the functional Hamilton-Jacobi equation for gravity and matter is described. This series approximation method represents a large generalization of the spatial gradient expansion that had been employed earlier. Additional solutions may be constructed using a nonlinear superposition principle. This formalism may be applied to problems in cosmology.

Recent advances in observational cosmology are changing the way we view the nature of time. In general relativity, the freedom in choosing a time hypersurface has hampered the implementation of the theory. Fortunately, Hamilton-Jacobi theory enables one to describe all time hypersurfaces on an equal footing. Using an expansion in powers of the spatial curvature, one may solve for the wavefunctional in a semiclassical approximation. In this way, one may readily compare predictions of various inflation models with observations of microwave background anisotropies and galaxy clustering.

Recent advances in observational cosmology are changing the way we view the nature of time. In general relativity, the freedom in choosing a time hypersurface has hampered the implementation of the theory. Fortunately, Hamilton-Jacobi theory enables one to describe all time hypersurfaces on an equal footing. Using an expansion in powers of the spatial curvature, one may solve for the wavefunctional in a semiclassical approximation. In this way, one may readily compare predictions of various inflation models with observations of microwave background anisotropies and galaxy clustering.

Hamilton-Jacobi theory provides a natural starting point for a covariant description of the gravitational field. Using a spatial gradient expansion, one may solve for the phase of the wavefunction by using a line-integral in superspace. Each contour of integration corresponds to a particular choice of time-hypersurface, and each yields the same answer. In this way, one can describe all time choices simultaneously. In an interesting application to cosmology, I compute large-angle microwave background anisotropies and the galaxy-galaxy correlation function associated with the scalar and tensor fluctuations of power-law inflation.

The nature of cosmic time is illuminated using Hamilton-Jacobi theory for general relativity. For problems of interest to cosmology, one may solve for the phase of the wavefunctional by using a line integral in superspace. Each contour of integration corresponds to a particular choice of time hypersurface, and each yields the same answer. In this way, one can construct a covariant formalism where all time hypersurfaces are treated on an equal footing. Using the method of characteristics, explicit solutions for an inflationary epoch with several scalar fields are given. The theoretical predictions of double inflation are compared with recent galaxy data and large angle microwave background anisotropies.

Cosmological inflation provides the simplest and most promising mechanism for generating fluctuations for structure formation. Using powerful Hamilton-Jacobi methods, I will describe (1) how to compute density fluctuations and cosmic microwave anisotropies arising from inflation, and (2) improvements of the Zel'dovich approximation describing gravitational collapse. I compare these results with the latest cosmological observations.

Using Hamilton-Jacobi theory, we develop a formalism for solving semi-classical cosmological perturbations which does not require an explicit choice of time-hypersurface. The Hamilton-Jacobi equation for gravity interacting with matter (either a scalar or dust field) is solved by making an Ansatz which includes all terms quadratic in the spatial curvature. Gravitational radiation and scalar perturbations are treated on an equal footing. Our technique encompasses linear perturbation theory and it also describes some mild nonlinear effects. As a concrete example of the method, we compute the galaxy-galaxy correlation function as well as large-angle microwave background fluctuations for power-law inflation, and we compare with recent observations.

Hamilton-Jacobi theory provides a natural starting point for a covariant description of the gravitational field. Using a spatial gradient expansion, one may solve for the phase of the wavefunction by using a line-integral in superspace. Each contour of integration corresponds to a particular choice of the time-hypersurface, and each yields the same answer. In this way, one can describe all time choices simultaneously. As a demonstration of the formalism, I will compute large-angle microwave background anisotropies and the galaxy-galaxy correlation function associated with scalar and tensor fluctuations of power-law inflation.

Beginning with a relativistic action principle for the irrotational flow of collisionless matter, we compute higher order corrections to the Zel'dovich approximation by deriving a nonlinear Hamilton-Jacobi equation for the velocity potential. It is shown that the velocity of the field may always be derived from a potential which however may be a multi-valued function of the space-time coordinates. In the Newtonian limit, the results are nonlocal because one must solve the Newton-Poisson equation. By considering the Hamilton-Jacobi equation for general relativity, we set up gauge-invariant equations which respect causality. A spatial gradient expansion leads to simple and useful results which are local --- they require only derivatives of the initial gravitational potential.