Search

We present a new analytic time dependent solution of cubic string field theory at the lowest order in the level truncation scheme. The tachyon profile we have found is a bounce in time, a $C^{\infty}$ function which represents an almost exact solution, with an extremely good degree of accuracy, of the classical equations of motion of the truncated string field theory. Such a finite energy solution describes a tachyon which at $x^0=-\infty$ is at the maximum of the potential, at later times rolls toward the stable minimum and then up to the other side of the potential toward the inversion point and thenback to the unstable maximum for $x^0\to+\infty$. The energy-momentum tensor associated with this rolling tachyon solution can be explicitly computed. The energy density is constant, the pressure is an even function of time which can change sign while the tachyon rolls toward the minimum of its potential. A new form of tachyon matter is realized which might be relevant for cosmological applications., Comment: 18 pages, 4 figures

We compute, on the disk, the non-linear tachyon $\beta$-function, $\beta^T$, of the open bosonic string theory. $\beta^T$ is determined both in an expansion to the third power of the field and to all orders in derivatives and in an expansion to any power of the tachyon field in the leading order in derivatives. We construct the Witten-Shatashvili (WS) space-time effective action $S$ and prove that it has a very simple universal form in terms of the renormalized tachyon field and $\beta^T$. The expression for $S$ is well suited to studying both processes that are far off-shell, such as tachyon condensation, and close to the mass-shell, such as perturbative on-shell amplitudes. We evaluate $S$ in a small derivative expansion, providing the exact tachyon potential. The normalization of $S$ is fixed by requiring that the field redefinition that maps $S$ into the tachyon effective action derived from the cubic string field theory is regular on-shell. The normalization factor is in precise agreement with the one required for verifying all the conjectures on tachyon condensation. The coordinates in the space of couplings in which the tachyon $\beta$-function is non linear are the most appropriate to study RG fixed points that can be interpreted as solitons of $S$, $i.e.$ D-branes., Comment: 29 pages, 1 figure, JHEP3.cls

Commutative Yang-Mills theories in 1+1 dimensions exhibit an interesting interplay between geometrical properties and U(N) gauge structures: in the exact expression of a Wilson loop with $n$ windings a non trivial scaling intertwines $n$ and $N$. In the non-commutative case the interplay becomes tighter owing to the merging of space-time and ``internal'' symmetries in a larger gauge group $U(\infty)$. We perform an explicit perturbative calculation of such a loop up to ${\cal O}(g^6)$; rather surprisingly, we find that in the contribution from the crossed graphs (the genuine non-commutative terms) the scaling we mentioned occurs for large $n$ and $N$ in the limit of maximal non-commutativity $\theta=\infty$. We present arguments in favour of the persistence of such a scaling at any perturbative order and succeed in summing the related perturbative series., Comment: LaTeX, 22 pages, no figures

We perform a perturbative ${\cal O}(g^4)$ Wilson loop calculation for the U(N) Yang-Mills theory defined on non-commutative one space - one time dimensions. We choose the light-cone gauge and compare the results obtained when using the Wu-Mandelstam-Leibbrandt ($WML$) and the Cauchy principal value ($PV$) prescription for the vector propagator. In the $WML$ case the $\theta$-dependent term is well-defined and regular in the limit $\theta \to 0$, where the commutative theory is recovered; it provides a non-trivial example of a consistent calculation when non-commutativity involves the time variable. In the $PV$ case, unexpectedly, the result differs from the $WML$ one only by the addition of two singular terms with a trivial $\theta$-dependence. We find this feature intriguing, when remembering that, in ordinary theories on compact manifolds, the difference between the two cases can be traced back to the contribution of topological excitations., Comment: 17 pages, LaTex, no figures

We study the perturbative unitarity of non-commutative quantum Yang-Mills theories, extending previous investigations on scalar field theories to the gauge case where non-locality mingles with the presence of unphysical states. We concentrate our efforts on two different aspects of the problem. We start by discussing the analytical structure of the vacuum polarization tensor, showing how Cutkoski's rules and positivity of the spectral function are realized when non-commutativity does not affect the temporal coordinate. When instead non-commutativity involves time, we find the presence of extra troublesome singularities on the $p_0^2$-plane that seem to invalidate the perturbative unitarity of the theory. The existence of new tachyonic poles, with respect to the scalar case, is also uncovered. Then we turn our attention to a different unitarity check in the ordinary theories, namely time exponentiation of a Wilson loop. We perform a $O(g^4)$ generalization to the (spatial) non-commutative case of the familiar results in the usual Yang-Mills theory. We show that exponentiation persists at $O(g^4)$ in spite of the presence of Moyal phases reflecting non-commutativity and of the singular infrared behaviour induced by UV/IR mixing., Comment: RevTeX, 27 pages, 8 eps-figures; report number and references added; minor changes

The partition function of a two dimensional Abelian gauge model reproducing magnetic vortices is discussed in the harmonic approximation. Classical solutions exhibit conformal invariance, that is broken by statistical fluctuations, apart from an exceptional case. The corresponding ``anomaly'' has been evaluated. Zero modes of the thermal fluctuation operator have been carefully discussed., Comment: RevTex, 14 pages, no figures. To appear on Nucl. Phys. B

We calculate in dimensions $D=2+\epsilon$ and in light-cone gauge (LCG) the perturbative ${\cal O}(g^4)$ contribution to a rectangular Wilson loop in the (t,x)-plane coming from diagrams with a self-energy correction in the vector propagator. In the limit $\epsilon \to 0$ the result is finite, in spite of the vanishing of the triple vector vertex in LCG, and provides the expected agreement with the analogous calculation in Feynman gauge., Comment: DFPD 99/TH/13, RevTex, 15 pages, no figures

A rectangular Wilson loop centered at the origin, with sides parallel to space and time directions and length $2L$ and $2T$ respectively, is perturbatively evaluated ${\cal O}(g^4)$ in Feynman gauge for Yang--Mills theory in $1+(D-1)$ dimensions. When $D>2$, there is a dependence on the dimensionless ratio $L/T$, besides the area. In the limit $T \to \infty$, keeping $D>2$, the leading expression of the loop involves only the Casimir constant $C_F$ of the fundamental representation and is thereby in agreement with the expected Abelian-like time exponentiation (ALTE). At $D= 2$ the result depends also on $C_A$, the Casimir constant of the adjoint representation and a pure area law behavior is recovered, but no agreement with ALTE in the limit $T\to\infty$. Consequences of these results concerning two and higher-dimensional gauge theories are pointed out., Comment: RevTex, 28 pages, two figure files included

Classical vacuum - pure gauge - solutions of Euclidean two-dimensional SU(2) Yang-Mills theories are studied. Topologically non-trivial vacua are found in a class of gauge group elements isomorphic to $S_2$. These solutions are unexpectedly related to the solution of the non-linear O(3) model and to the motion of a particle in a periodic potential., Comment: 9 pages, revTeX, no figures; note added

The Schwinger model, defined in the space interval $-L \le x \le L$, with (anti)periodic boundary conditions, is canonically quantized in the light-cone gauge $A_-=0$ by means of equal-time (anti)commutation relations. The transformation diagonalizing the complete Hamiltonian is explicitly constructed, thereby giving spectrum, chiral anomaly and condensate. The structures of Hilbert spaces related both to free and to interacting Hamiltonians are completely exhibited. Besides the usual massive field, two chiral massless fields are present, which can be consistently expunged from the physical space by means of a subsidiary condition of a Gupta-Bleuler type. The chiral condensate does provide the correct non-vanishing value in the decompactification limit $L \to \infty$., Comment: CERN-TH/97-164, 35 pages, revTeX, no figures

A rectangular Wilson loop with sides parallel to space and time directions is perturbatively evaluated in two light-cone gauge formulations of Yang-Mills theory in 1+1 dimensions, with ``instantaneous'' and ``causal'' interactions between static quarks. In the instantaneous formulation we get Abelian-like exponentiation of the area in terms of $C_F$. In the ``causal'' formulation the loop depends not only on the area, but also on the dimensionless ratio $\beta = {L \over T}$, $2L$ and $2T$ being the lengths of the rectangular sides. Besides it also exhibits dependence on $C_A$. In the limit $T \to \infty$ the area law is recovered, but dependence on $C_A$ survives. Consequences of these results are pointed out., Comment: 30 pages, latex, one figure included as a ps file, an Erratum included

We obtain a formal solution of an integral equation for $q\bar q$ bound states, depending on a parameter \eta which interpolates between 't Hooft's (\eta=0) and Wu's (\eta=1) equations. We also get an explicit approximate expression for its spectrum for a particular value of the ratio of the coupling constant to the quark mass. The spectrum turns out to be in qualitative agreement with 't Hooft's as long as \eta \neq 1. In the limit \eta=1 (Wu's case) the entire spectrum collapses to zero, in particular no rising Regge trajectories are found., Comment: CERN-TH/96-364, 13 pages, revTeX, no figures

In 1+1 dimensions two different formulations exist of SU(N) Yang Mills theories in light-cone gauge; only one of them gives results which comply with the ones obtained in Feynman gauge. Moreover the theory, when considered in 1+(D-1) dimensions, looks discontinuous in the limit D=2. All those features are proven in Wilson loop calculations as well as in the study of the $q\bar q$ bound state integral equation in the large N limit., Comment: Invited report at the Workshop "Low Dimensional Field Theory", Telluride (CO), Aug. 5-17 1996; 16 pages, latex, no figures To appear in International Journal of Modern Physics A minor misprints corrected

Pure Yang-Mills theories on the $S_1\times R$ cylinder are quantized in light-cone gauge $A_-=0$ by means of ${\bf equal-time}$ commutation relations. Positive and negative frequency components are excluded from the ``physical" Hilbert space by imposing Gauss' law in a weak sense. Zero modes, related to the winding on the cylinder, provide non trivial topological variables of the theory. A Wilson loop with light-like sides is studied: in the abelian case it can be exactly computed obtaining the expected area result, whereas difficulties are pointed out in non abelian cases., Comment: 21 pages, revtex, one figure included

In the first order form, the model considered by Strobl presents, besides local Lorentz and diffeomorphism invariances, an additional local non-linear symmetry. When the model is realized as a Poincar\'e gauge theory according to the procedure outlined in Refs.[1,2], the generators of the non-linear symmetry are responsible for the ``nasty constraint algebra''. We show that not only the Poincar\'e gauge theoretic formulation of the model is not the cause of the emerging of the undesirable constraint algebra, but actually allows to overcome the problem. In fact one can fix the additional symmetry without breaking the Poincar\'e gauge symmetry and the diffeomorphisms, so that, after a preliminary Dirac procedure, the remaining constraints uniquely satisfy the Poincar\'e algebra. After the additional symmetry is fixed, the equations of motion are unaltered. The objections to our method raised by Strobl in Ref.[3] are then immaterial. Some minor points put forward in Ref.[3] are also discussed., Comment: 7 pages, plain \TeX

Following the general method discussed in Refs.[1,2], Liouville gravity and the 2 dimensional model of non-Einstenian gravity ${\cal L} \sim curv^2 + torsion^2 + cosm. const.$ can be formulated as ISO(1,1) gauge theories. In the first order formalism the models present, besides the Poincar\'e gauge symmetry, additional local symmetries. We show that in both models one can fix these additional symmetries preserving the ISO(1,1) gauge symmetry and the diffeomorphism invariance, so that, after a preliminary Dirac procedure, the remaining constraints uniquely satisfy the ISO(1,1) algebra. After the additional symmetry is fixed, the equations of motion are unaltered. One thus remarkably simplifies the canonical structure, especially of the second model. Moreover, one shows that the Poincar\'e group can always be used consistently as a gauge group for gravitational theories in two dimensions., Comment: 13 pages, plain \TeX

Complete two-loop calculation of a dimensionally regularized Wilson loop with light-like segments is performed in the light-like axial gauge with the Mandelstam-Leibbrandt prescription for the gluon propagator. We find an expression which {\it exactly} coincides with the one previously obtained for the same Wilson loop in covariant Feynman gauge. The renormalization of Wilson loop is performed in the $\MS-$scheme using a general procedure tailored to the light-like axial gauge. We find that the renormalized Wilson loop obeys a renormalization group equation with the same anomalous dimensions as in covariant gauges. Physical implications of our result for investigation of infrared asymptotics of perturbative QCD are pointed out., Comment: 24 pages and 4 figures (included), LaTeX style, UFPD-93/TH/23, UPRF-93-366, UTF-93-291

In the context of a Poincar\'e gauge theoretical formulation, pure gravity in 3+1-dimensions is dimensionally reduced to gravity in 2+1-dimensions with or without cosmological constant $\Lambda$. The dimensional reductions are consistent with the gauge symmetries, mapping ISO(3, 1) gauge transformations into ISO(2,1) ones. One of the reductions leads to Chern-Simons-Witten gravity. The solutions of 2+1-gravity with $\Lambda\le 0$ (in particular the black-hole solution recently found by Banados, Teitelboim and Zanelli) and those of 1+1-dimensional Liouville gravity, are thus mapped into 3+1-dimensional vacuum solutions., Comment: 9 pages (TeX manuscript)

We have shown that two of the most studied models of lineal gravities - Liouville gravity and a ``string-inspired'' model exhibiting the main characteristic features of a black-hole solution - can be formulated as gauge invariant theories of the Poincar\'e group. The gauge invariant couplings to matter (particles, scalar and spinor fields) and explicit solutions for some matter field configurations, are provided. It is shown that both the models, as well as the couplings to matter, can be obtained as suitable dimensional reductions of a 2+1-dimensional ISO(2,1) gauge invariant theory., Comment: TeX Manuscript, 30 pages