0810.0771/ch1.tex
1: \chapter{Introduction to quantum gravity}\label{ch1}
2: 
3: \section{Classical gravity, physics of the large}\label{sec:Classical gravity, physics of the large}
4: 
5: Gravity, omnipresent and inescapable...
6: \\
7: \\
8: Unlike the other three fundamental forces of nature its reach is universal.
9: All objects and substances in the universe are sensitive to the gravitational pull.
10: Besides being mere slaves to the will of gravity, matter and energy also play a more
11: proactive role, since everything inside our universe acts as a source for gravity.
12: \\
13: \\
14: In our daily lives we are only confronted with the passive side of gravity.
15: If we jump, the gravitational pull of the earth will inevitably let us fall back down again.
16: The only effect by which gravity reveals itself is by dictating the way we move, never
17: do we experience our role as sources of gravity. More generic, in no microscopic or mesoscopic
18: experiment does the gravitational pull between the objects play an important role. The reason for this is clear,
19: gravity is an extremely feeble force when compared to the other fundamental interactions.
20: Although we usually take this fact for granted, it might strike one as strange that the weakest
21: force of nature dominates the motion of objects on the scales relevant in our everyday life.
22: The fundamental reason behind this is that the source of gravity only comes in one flavor, matter
23: and energy are always positive causing gravity to be always attractive. If gravity would have had both positive and
24: negative charges similar to electromagnetism, gravity would not have played any role in our everyday lives, since
25: it would have been overshadowed by the other forces.
26: \\
27: \\
28: One of the key insights that enabled Newton to formulate his theory of gravity is
29: the realization that gravity is not only important at scales familiar from our experiences,
30: but it is also the relevant force at solar system scales and even beyond. He realized that
31: the motion of a falling apple is similar to the trajectory of the moon, both are caused by the tug of gravity.
32: \\
33: \\
34: Even though Newton's theory was tremendously successful, as it explained the motion of the planets
35: with unprecedented accuracy, Einstein felt uneasy. He embarked on a historic quest to construct a more aesthetic
36: description of gravity, an endeavor that turned out to be one of the pinnacles of human ingenuity in recent modern history.
37: One of his motivations to construct a
38: more elaborate theory was to harmonize Newton's theory with the principles
39: of his own theory of special relativity. Another motivation that greatly influenced Einstein's work when constructing his
40: theory of gravity was the equivalence principle, inertial and gravitational mass were measured to be the same
41: with remarkable accuracy.
42: From these incentives and a few other rather philosophical arguments he developed
43: the theory of general relativity. So by combining aesthetic reasoning with known results from experimental physics
44: he found a geometrical theory that was seen to describe the real world.
45: In so doing he extended the validity of gravitational theory to
46: the largest scales possible. In particular, general relativity has allowed us to compute corrections to
47: Newton's theory that are vital for the study of cosmology. Furthermore, Einstein's description of gravity has
48: been tested and confirmed to be valid for the largest distances, masses and velocities that we can measure.
49: 
50: \section{Quantum gravity, physics of the small?} \label{sec:Quantum gravity, physics of the small?}
51: 
52: What about the converse regime? Does gravity really become weaker and weaker when we study nature
53: on increasingly small scales? The simple empirical answer is yes, the gravitational interaction is so
54: incredibly weak that it has only been tested down to millimeter scales. It was found that at these scales Newton's law
55: still holds implying that gravity indeed becomes negligible for the extremely small.
56: The theoretical expectations are more interesting however.
57: \\
58: \\
59: We know that the physics of systems on small distances is well described by the laws of quantum mechanics.
60: One of the many peculiar features of quantum theory is that it connects small and large scales by
61: virtue of Heisenberg's uncertainty principle. To probe physics on increasingly small scales one needs progressively
62: larger momenta. Since a large momentum implies large energy one expects gravity to become very relevant at the
63: very tiny scales, contrary to the naive extrapolation of the classical theory.
64: The scale for which the probe gravitational field becomes large is the Planck scale $L_P\simeq 1.6\times10^{-35}m$.
65: At this scale the energy needed to resolve the microstructure needs to be so concentrated that a black hole would form.
66: \\
67: \\
68: Quantum mechanics is, unlike gravity, a theory of probabilities. We know that all other forces and all matter fields
69: satisfy its probabilistic laws, so why should gravity be an exception? To avoid the coexistence of classical and
70: quantum theory, gravity should be quantized too.
71: \\
72: \\
73: Why have we not yet been able to accomplish this? What sets gravity apart from the other forces of nature? What
74: makes it so hard to unify gravity with the laws of quantum mechanics?
75: The reasons are plentiful, one essential fact that makes
76: the analysis of general relativity very hard in general, also on the classical level, is that it is highly
77: nonlinear. This nonlinearity is much more severe than in the other interactions of the standard model
78: as the action is not even polynomial in its fundamental field, the metric.
79: This has dramatic consequences for the quantization of
80: gravity by perturbative methods. For the most straight forward methods it basically implies that there is
81: an infinite number of interaction vertices that have to be taken into account
82: \footnote{With a suitable (non tensorial) field redefinition of the metric it is however possible to write the Einstein action in a polynomial form. This implies that it is possible do perturbative quantum gravity with a finite number of interaction vertices.}.
83: Therefore we can conclude that the standard formulations of gravity are not easily treated by perturbation theory.
84: Another glaring problem that exemplifies the tension between gravity and perturbation theory is the absence of a
85: natural dimensionless
86: coupling constant to define a perturbative expansion. Instead, the coupling constant of gravity, Newton's constant
87: $G_N$, has dimensions of inverse energy squared. Subsequently, the natural parameter of the perturbation expansion
88: is $G_N E^2$ and we see that the coupling of gravitons increases with their energy. At the point where
89: the gravitons reach the Planck energy the coupling becomes strong, $G_N E^2 \sim 1$, which inevitable leads
90: to a breakdown of the perturbation expansion. In contemporary terms we say that the dimensionful nature of
91: Newton's constant makes general relativity nonrenormalizable as a quantum field theory.
92: \\
93: \\
94: Several points of view can be taken regarding the nonrenormalizability of gravity. The most popular stance
95: is that the problem comes from an inherent mismatch between the principles of general relativity and quantum theory.
96: According to this attitude, a resolution for a quantum theory of gravity can only be found in a modification of
97: the physical principles behind either quantum mechanics, general relativity or both. The most popular candidate for
98: such a scenario is string theory, where gravity is found to be compatible with quantum theory only if it is
99: accompanied by a plethora of extra fields and dimensions. Gravity by itself is viewed as a mere low energy effective
100: theory, and the exact harmonization with quantum theory happens only upon considering the dynamics of the
101: fundamental strings.
102:  \\
103:  \\
104: A perhaps more conservative attitude is to suppose that gravity and quantum mechanics are not fundamentally incompatible
105: per se, but that the standard perturbation theory simply is an inadequate tool for the quantization of gravity.
106: Precisely this philosophy is an inspiration for the models we present in the present thesis.
107: \\
108: \\
109: The construction of a nonperturbative formulation of gravity is a far from trivial task however.
110: For example, even though the other field theories of the standard
111: model are considerably simpler than gravity, we cannot solve the path integrals exactly.
112: Often, the path integrals are merely a helpful tool to set up a perturbative
113: description of the physical problem at hand. Although extremely successful in QED, it is not an adequate scheme
114: to study physics in strong coupling regimes such as confinement in QCD.
115: The computation of quantities that go beyond perturbation theory is often
116: very difficult, but even worse, it is mostly unclear whether there exists a nonperturbative definition of a path
117: integral at all! In more than two dimensions there are few methods that enable one to address
118: the nonperturbative existence of path integrals. In most cases the nonperturbative definition of a
119: path integral in field theory is only possible by defining it as a limit of a discrete theory.
120: Although mathematically more rigorous, it is in practice not a very convenient tool to compute concrete
121: amplitudes, analytical methods are largely unavailable. Nonetheless, the enormous growth in computing power over the recent
122: years has transformed lattice quantum field theory from a mathematically nice idea into a serious competitor in the arena
123: of theoretical physics. In particular, the study of QCD has benefitted a lot from these developments. Among the successes are
124: the calculation of realistic values for meson and baryon masses from first principles. Such formidable achievements are
125: currently beyond reach of other methods.
126: \\
127: \\
128: In this thesis we investigate simple gravitational models that are based on the method known as Causal
129: Dynamical Triangulations. In spirit the scheme is a succinct gravitational analogue of lattice QCD. It is a natural
130: method to define the path integral by a lattice regularization. What remains is a finite statistical sum that,
131: similar to lattice QCD, lends itself perfectly to computer simulations. One distinguishing feature that sets Causal
132: Dynamical Triangulations apart from other discrete attempts is that a genuine causal structure is imposed on
133: the quantum geometry from the outset. The results of the simulations are encouraging,
134: in four dimensions a well behaved continuum limit seems to exist and there is compelling evidence that a
135: classical spacetime superimposed with small quantum fluctuations emerges from the nonperturbative path integral.
136: Despite the intriguing results these numerical methods have to offer, the understanding is far from complete
137: and inherently restricted by computer power. Furthermore, the statistical model is very complicated and
138: has so far resisted attempts at a solution by analytical methods.
139: \\
140: \\
141: In two dimensions the situation is much better however, the pure gravity model can be explicitly solved and many interesting
142: results can be obtained. Of course one might contest that two dimensional gravity is an oversimplified model
143: as it does not possess some of the essential difficulties of four dimensional gravity such as a dimensionful
144: coupling constant. Nevertheless it still contains some vital characteristics that set gravitational theories
145: apart from any other. Issues such as diffeomorphism invariance, background independence and the Wick rotation
146: are as relevant for the two dimensional model as they are for its higher dimensional analogues.
147: \\
148: \\
149: Besides being interesting from a pure quantum gravity point of view, it might also be regarded as a minimal form of string
150: theory. Particularly, the two dimensional model of Causal Dynamical Triangulations might shed some light on the
151: role of causality on the worldsheet of a string. A tantalizing indication that this might indeed
152: be consequential is that the results of two dimensional Causal Dynamical Triangulations
153: are physically inequivalent to the outcomes of two dimensional Euclidean quantum gravity.
154:  \\
155:  \\
156: For the purposes of this thesis we primarily view two dimensional quantum gravity as
157: an interesting laboratory were nonperturbative aspects of quantum gravity can be studied in an exactly solvable setting.
158: Before presenting our original contributions, we first discuss some general remarks and present
159: the known results of two dimensional Causal Dynamical Triangulations in chapter \ref{ch2}.
160:  \\
161:  \\
162: In chapter \ref{ch3} we start the discussion of our first generalization by reviewing the
163: previously established relation between Euclidean and Causal dynamical triangulations.
164: It is discussed that imposing causality has the important consequence that the spatial topology
165: of the geometries in the path integral is fixed. Additionally, we recall that in Euclidean Dynamical
166: Triangulations, as for example defined by matrix models,
167: the quantum geometry is highly degenerate in the sense that the spatial topology fluctuations dominate
168: the path integral.
169: \\
170: \\
171: In the remaining sections of chapter \ref{ch3} we show that this situation is not as black and white as
172: is discussed above. In an original contribution we demonstrate that one can allow for spatial topology
173: change in two dimensional causal quantum gravity in a controlled manner. We argue that the topology fluctuations
174: are naturally accompanied by a coupling constant reminiscent of the string coupling.
175: Upon taking a suitable scaling limit we show that the quantum geometry is no longer swamped by the topology
176: fluctuations. Surprisingly, we are able to compute the relevant amplitudes to all orders in the coupling and
177: sum the power series uniquely to obtain an exact nonperturbative result!
178: \\
179: \\
180: In chapter \ref{ch4} we return to the ``pure'' model of Causal Dynamical triangulations. In this chapter we
181: extend the existing formalism by studying boundary conditions that lead to a path integral over
182: noncompact manifolds. We begin by recalling that a similar mechanism is familiar from non-critical string
183: theory where the noncompact quantum geometries are known as ``ZZ branes''. Further we show that
184: a space of constant negative curvature emerges from the background independent sum over noncompact spacetimes.
185: Fascinatingly, we can compute the quantum fluctuations and are able to show that they are small almost everywhere
186: on the geometry! The model is a nice example of how a classical background can appear from a background independent
187: theory of quantum gravity.
188:  \\
189:  \\
190: To conclude, we tackle the problem of \emph{spacetime} topology change in chapter \ref{ch5}.
191: Although we are not able to
192: completely solve the path integral over all manifolds with arbitrary topology, we do obtain some
193: results indicating that such a path integral might be consistent, provided suitable causality
194: restrictions are imposed. As a first step we generalize the standard amplitudes of causal dynamical
195: triangulations by a perturbative computation of amplitudes that include manifolds up to genus two.
196: Furthermore, a toy model is presented where we make the approximation that the holes in the manifold
197: are infinitesimally small. This simplification allows us to perform an explicit sum over all genera
198: and analyze the continuum limit exactly. Remarkably, the presence of the infinitesimal wormholes
199: leads to a decrease in the effective cosmological constant, reminiscent of the suppression
200: mechanism considered by Coleman and others in the four dimensional Euclidean path integral.
201: 
202: 
203: 
204: 
205: 
206: 
207: 
208: 
209: 
210: 
211: %Nonpertubative enquiries of realistic quantum field theory are not hopeless however, a principal example
212: %that demonstrates this nicely is lattice QCD.
213: 
214: 
215: 
216: 
217: \begin{comment}
218: 
219: \section{Quantum gravity and its problems}
220: 
221:   SMALL SCALES -> QM+GRAVITY
222:     - gravity more complicated theory ->  GR nonlinear (non polynomial 2 order) / maxwell linear i.e.~gaussian weight in PI
223:     - no superposition principle GR
224:     - GR background independent -> action only needs the metric field
225:     - QED need metric field (background)
226:     - can treat GR analogous to Gauge theory by linearizing around a background
227:     - metric gauge potential for coordinate transformations
228:     - difference with ordinary gauge theory gauge potential is not equal to the connection
229:     - similar to gauge theory try to perturbatively quantize
230:     - no dimensionless coupling (since no elementary mass compare Newtons gravity law with electrodynamics point source)
231:     - no natural perturbation parameter
232:     - Gn dimensionful -> expansion parameter $Gn*E^2$,important reason why perturbative linearized gravity non renormalizable
233:     - \cite{tHooft:1974bx} \cite{Goroff:1985th} \cite{van de Ven:1991gw}
234: 
235: It should be noted that these points of view are not necessarily mutually exclusive.
236: 
237: 
238: \subsection{physics of the small}
239: 
240: -   gravity dominant force large scales surprising since it is the most feeble forces of the four fundamental forces
241:     of nature
242: -   compare strength of EM with GR
243: -   no plus/minus charges in gravity + graviton massless -> gravity dominant force large scales
244: -   gravity is completely unimportant for microscopic physics
245: -   The physics of the small -> the world becomes subject to the laws of quantum mechanics
246: -   physics no longer deterministic but governed by probability
247: 
248: It is the only fundamental force of nature that
249: -   everything gravitates
250: -   Newton's insight why we fall same reason why the planets rotate around the sun
251: -   gravity dominates the large
252: 
253: \subsection{Why general relativity?}
254: 
255: -   Einstein unhappy -> Newton gravity not beautiful
256:     - not compatible with the priciple of special relativity
257:     - GR removes action at a distance -> graviton massless-> gravity travels with the speed of light
258:     - equivalence principle -> intertial mass = gravitational mass
259:     - EM force described by fields -> find field of gravity
260: -   GR created on philosophical grounds-> pinnacles of human thinking
261: -   turns out essential in the description of physics it corrects Newton's laws
262:     that ar valied of the large to the extremely large ->
263:     -largest distances
264:     -largest velocities
265:     -largest masses,energies
266: 
267: \subsection{quantum mechanics -> small = big}
268: 
269: -   quantum mechanics tells us that to probe very small scales we need very high energies
270: -   planck scale -> black holes
271: -   quanta also gravitate-> combined system qmechanical
272: -   Unite QM with GR
273: 
274: \section{Quantum gravity and its problems}
275: 
276:   SMALL SCALES -> QM+GRAVITY
277:     - gravity more complicated theory ->  GR nonlinear (non polynomial 2 order) / maxwell linear i.e. gaussian weight in PI
278:     - no superposition principle GR
279:     - GR background independent -> action only needs the metric field
280:     - QED need metric field (background)
281:     - can treat GR analogous to Gauge theory by linearizing around a background
282:     - metric gauge potential for coordinate transformations
283:     - difference with ordinary gauge theory gauge potential is not equal to the connection
284:     - similar to gauge theory try to perturbatively quantize
285:     - no dimensionless coupling (since no elementary mass compare Newtons gravity law with electrodynamics point source)
286:     - no natural perturbation parameter
287:     - Gn dimensionful -> expansion parameter $Gn*E^2$,important reason why perturbative linearized gravity non renormalizable
288:     - \cite{tHooft:1974bx} \cite{Goroff:1985th} \cite{van de Ven:1991gw}
289: 
290: \subsection{Non perturbative methods}
291: 
292: - 2 attitudes:
293:       1) popular: QM+gravity needs unification with other forces
294:       -> change fundamental degrees of freedom with automatic
295:       ultraviolet cutoff and one can still do perturbation theory e.g. string-theory .
296:       2) conservative: GR is fine perturbation theory is ill defined
297: 
298: -   here take attitude 2
299: -   nonperturbatively well defined: \cite{Witten:1988hc}\cite{Witten:2007kt}
300: -   asymptotic safety:\cite{Weinberg:1980gg}
301: -   renormalization group , background field method: \cite{Lauscher:2005xz}
302:     review: \cite{Niedermaier:2006ns}
303: 
304: \subsection{Causal dynamical triangulations}
305: \end{comment}
306: 
307: \begin{comment}
308: \subsection{Length scales and range of validity}
309:   GR EXPERIMENTALLY GOOD AT SOLAR SYSTEM SCALES:
310:     - classical GR very successful
311:     - computes corrections to Newton, Newt only valid small velocities, small masses
312:     - experimental successes GR -> perihelion precession mercury, bending light sun gravitational lensing,
313:       gravitywaves pulsars strong field test GR,
314:     - comparison of gravity to QED solar system
315:     - gravity terribly weak but....
316: 
317: 
318:   GR EXPERIMENTALLY NOT GOOD BIGGEST SCALES?
319:     - GR-> essential for cosmology ->recent cosmological observations -> classical gravity modified in the far infrared?
320:     - GR -> valid big scales (solar system ed)  not succesfull at the \emph{biggest} scales? galaxy,cosmological scales?
321: 
322:   GRAVITY AT SMALL SCALES?
323:     - GR strong big mass -> Gravity not important at small scales
324:     - so feeble -> strength only measured centimeter level
325:     - comparison of gravity to QED nucleus
326:     - gravity much weaker
327:     - even smaller scales -> gravity strong again -> momentum/position duality of QM
328:     - gravity strong -> mass very big or length scales small
329:     - inherent length scale in GR strong -> planck scale -> scale way beyond accelerator scales
330:     - planck scale-> gravity so strong -> black holes
331:     - same energy scales QM and classical gravity?
332:     - Tension: QM probabilistic <-> gravity deterministic
333: 
334: \subsection{Non perturbative methods}
335:   NON PERTURBATIVE QUANTUM GRAVITY
336: 
337: asymptotic safety:\cite{Weinberg:1980gg}
338: renormalization group , background field method: \cite{Lauscher:2005xz}
339: review: \cite{Niedermaier:2006ns}
340: nonperturbatively well defined: \cite{Witten:1988hc}
341: 
342: \end{comment}
343: \begin{comment}
344: \subsection{Causal dynamical triangulations}
345: 
346: introduced in: \cite{Ambjorn:1998xu}
347: coupling to matter: \cite{Ambjorn:1999gi}\cite{Ambjorn:1999mc}\cite{Ambjorn:1999yv}
348:                     recently\cite{Benedetti:2006rv}{Benedetti:2006rv}
349: relation to Euclidean theory \cite{Ambjorn:1998xu} \cite{Ambjorn:1998fd}\cite{Ambjorn:1999fp}
350: summary:\cite{Ambjorn:1999nc}
351: generalizations \cite{DiFrancesco:1999em}\cite{DiFrancesco:2000nn}
352: higher dimensional formalism: \cite{Ambjorn:2001cv}\cite{Ambjorn:2000dv}
353: three d computer simulations: \cite{Ambjorn:2000dv} \cite{Ambjorn:2000dj}
354: three d analytical    \cite{Ambjorn:2001br}\cite{Ambjorn:2002nu}\cite{Ambjorn:2003sr}
355:                     recently \cite{Benedetti:2007pp}
356: CDT solves conformal factor problem and continuum interpretation of Wick rotation,
357:                     \cite{Dasgupta:2001ue} \cite{Ambjorn:2002gr}\cite{Dasgupta:2002jm}
358: lecture notes: \cite{Loll:2002xb}
359: four d results computer sims:   \cite{Ambjorn:2004qm}\cite{Ambjorn:2004pw}\cite{Ambjorn:2005db}\cite{Ambjorn:2005qt}
360:                         \cite{Ambjorn:2005jj}\cite{Ambjorn:2006jf}
361: 
362: 
363: topology change \cite{Loll:2003rn}\cite{Loll:2003yu}\cite{Loll:2005dr}\cite{Loll:2006gq}
364: 
365: lorentzian AdStwo \cite{Ambjorn:2006tt}\cite{Ambjorn:2006hu}
366: Euclidean AdStwo \cite{Ambjorn:2004my}\cite{Ambjorn:2007iq}
367: \end{comment}
368: 
369: \begin{comment}
370:     -   outline CDT approach
371:         - CDT nonperturbative background independent
372:         - define PI over geometries by statistical Sum over piecewise linear ones
373: 
374:         - analogous to lattice QCD to get access to nonperturbative information of the theory
375:     -   discuss results four dimensions CDT
376:     -   discuss topology change
377:     -   argue lower dimensional models -> solvable, diff invariant, background independent, non topological theory (anomalies)
378:     -   liouville/matrix models vs CDT
379:     -   outline of the thesis
380: 
381: Quantum gravity is a term that can have rather variable connotations. In this thesis we formally define the quantization of
382: gravity by a path integral over all Lorentzian geometries. Where we use the term Lorentzian geometry to describe a
383: smooth manifold M equipped with a metric tensor field gmunu with Lorentzian signature \emph{modulo diffeomorphisms}.
384: Factoring out diffeomorphisms is known to be exceedingly difficult, making it a formidable obstacle to
385: solve gravitational path integrals by continuum methods. We do not go into these details here since we use the
386: inherently discrete and coordinate free method of dynamical triangulations to define a gravitational path integral.
387: \end{comment}
388: