
We construct an ndimensional BornInfeld type gravity theory that has the
same properties as Einstein's gravity in terms of the vacuum and particle
content: Namely, the theory has a unique viable vacuum (maximally symmetric
solution) and a single massless unitary spin2 graviton about this vacuum. The
BI gravity, in some sense, is the most natural, minimal generalization of
Einstein's gravity with a better UV behavior, and hence, is a potentially
viable proposal for low energy quantum gravity. The GaussBonnet combination
plays a nontrivial role in the construction of the theory. As an extreme
example, we consider the infinite dimensional limit where an interesting
exponential gravity arises.

We construct BornInfeld (BI) type gravity theories which describe treelevel
unitary (nonghost and nontachyonic) massless spin2 modes around their
maximally symmetric vacua in four dimensions. Building unitary BI actions
around flat vacuum is straightforward; but, this is a complicated task around
(anti)de Sitter backgrounds. In this work, we solve the issue and give details
of constructing perturbatively viable determinantal BI theories. It is
interesting that the GaussBonnet combination, which is a total derivative in
four dimensions, plays an important role in the construction of viable BI
theories.

We show that, in the weak field limit, at large separations, in sharp
contrast to General Relativity (GR), all massive gravity theories predict
distancedependent spin alignments for spinning objects. For all separations GR
requires antiparallel spin orientations with spins pointing along the line
joining the sources. Hence total spin is minimized in GR. On the other hand,
while massive gravity at small separations ($m_gr \le1.62$) gives the same
result as GR, for large separations ($m_gr>1.62$) the spins become parallel to
each other and perpendicular to the line joining the objects. Namely, the
potential energy is minimized when the total spin is maximized in massive
gravity for large separations. We also compute the spinspin interactions in
quadratic gravity theories and find that while at large separations GR result
is intact, at small separations, spins become perpendicular to the line joining
sources and antiparallel to each other.

We explicitly confirm the expectation that generic Lovelock gravity in D
dimensions has a unitary massless spin2 excitation around any one of its
constant curvature vacua just like the cosmological Einstein gravity. The
propagator of the theory reduces to that of Einstein's gravity, but scattering
amplitudes must be computed with an effective Newton's constant which we
provide. Treelevel unitarity imposes a single constraint on the parameters of
the theory yielding a wide range of unitary region. As an example, we
explicitly work out the details of the cubic Lovelock theory.

In this thesis massive higher derivative gravity theories are analyzed in
some detail. Oneparticle scattering amplitude between two covariantly
conserved sources mediated by a graviton exchange is found at treelevel in $D$
dimensional (Anti)de Sitter and flat spacetimes for the most general quadratic
curvature theory augmented with the PauliFierz mass term. From the amplitude
expression, the Newtonian potential energies are calculated for various cases.
Also, from this amplitude and the propagator structure, a three dimensional
unitary theory is identified. In the second part of the thesis, the found three
dimensional unitary theory is studied in more detail from a canonical point of
view. The general higher order action is written in terms of gaugeinvariant
functions both in flat and de Sitter backgrounds. The analysis is extended by
adding static sources, spinning masses and the gravitational ChernSimons term
separately to the theory in the case of flat spacetime. For all cases the
microscopic spectrum and the masses are found. In the discussion of curved
spacetime, the masses are found in the relativistic and nonrelativistic
limits. In the Appendix, some useful calculations that are frequently used in
the bulk of the thesis are given.

We construct all the unitary cubic curvature gravity theories built on the
contractions of the Riemann tensor in D dimensional (anti)de Sitter
spacetimes. Our construction is based on finding the equivalent quadratic
action for the general cubic curvature theory and imposing ghost and tachyon
freedom, which greatly simplifies the highly complicated problem of finding the
propagator of cubic curvature theories in constant curvature backgrounds. To
carry out the procedure we have also classified all the unitary quadratic
models. We use our general results to study the recently found cubic curvature
theories using different techniques and the string generated cubic curvature
gravity model. We also study the scattering in critical gravity and give its
cubic curvature extensions.

We give an exact solution of the quadratic gravity in D dimensions. The
solution is a plane fronted wave metric with a cosmological constant. This
metric solves not only the full quadratic gravity field equations but also the
linearized ones which include the linearized equations of the recently found
critical gravity. A subset of the solutions change the asymptotic structure of
the antide Sitter space due to their logarithmic behavior.

We construct all the bulk and boundary unitary cubic curvature parity
invariant gravity theories in three dimensions in (anti)de Sitter spaces. For
bulk unitarity, our construction is based on the principle that the free theory
of the cubic curvature theory reduces to one of the three known unitary
theories which are the cosmological EinsteinHilbert theory, the quadratic
theory of the scalar curvature or the new massive gravity (NMG). Bulk and
boundary unitarity in NMG is in conflict; therefore, cubic theories that are
unitary both in the bulk and on the boundary have free theories that reduce to
the other two alternatives. We also study the unitarity of the BornInfeld
extensions of NMG to all orders in curvature.

We develop techniques of analyzing the unitarity of general BornInfeld (BI)
gravity actions in Ddimensional spacetimes. Determinantal form of the action
allows us to find a compact expression quadratic in the metric fluctuations
around constant curvature backgrounds. This is highly nontrivial since for the
BI actions, in principle, infinitely many terms in the curvature expansion
should contribute to the quadratic action in the metric fluctuations around
constant curvature backgrounds, which would render the unitarity analysis
intractable. Moreover in even dimensions, unitarity of the theory depends only
on finite number of terms built from the powers of the curvature tensor. We
apply our techniques to some fourdimensional examples.

We derive and study the equations of motion of the BornInfeld extension of
New Massive Gravity for globally and asymptotically (anti)de Sitter spaces,
and show that the assumptions of the nullenergy condition and holography (that
bounds the cfunction) lead to two simple cfunctions one of which is
equivalent to the cfunction of Einstein's gravity. We also show that, at the
fixed point, the cfunction gives the central charge of the Virasoro algebra
and the coefficient of the Weyl anomaly up to a constant.

We define various BornInfeld gravity theories in 3+1 dimensions which reduce
to Horava's model at the quadratic level in small curvature expansion. In their
exact forms, our actions provide z>(infinity) extensions of Horava's gravity,
but when small curvature expansion is used, they reproduce finite z models,
including some halfinteger ones.

We present a threedimensional gravitational BornInfeld theory which reduces
to the recently found New Massive Gravity (NMG) at the quadratic level in the
small curvature expansion and at the cubic order reproduces the deformation of
NMG obtained from AdS/CFT. Our action provides a remarkable extension of NMG to
all orders in the curvature, and might define a consistent quantum gravity.

We give an explicitly gauge invariant canonical analysis of linearized
quadratic gravity theories in three dimensions for both flat and deSitter
backgrounds. In flat backgrounds, we also study the effects of gravitational
ChernSimons term, and include the sources, compute the weak field limit as
well as scattering between spinning massive particles.

We find the propagator and calculate the tree level scattering amplitude
between two covariantly conserved sources in an Antide Sitter background for
the most general Ddimensional quadratic, fourderivative, gravity with a
PauliFierz mass. We also calculate the Newtonian potential for various limits
of the theory in flat space. We show how the recently introduced three
dimensional New Massive Gravity is uniquely singled out among higher derivative
models as a (tree level) unitary model and that its Newtonian limit is
equivalent to that of the usual massive gravity in flat space.