Spacetime Metric — Season 1 — 03-general-relativity

Transcript

Cold open (≈ 90 seconds)

Watch an apple fall off a table.

Half a second from the edge of the wood to the kitchen floor. You have seen this happen thousands of times. You know exactly what is supposed to happen, and exactly when. There is nothing surprising about it.

Now ask the question that nobody asks at the kitchen counter.

Why does it fall.

For most of human history, the answer was that there was a force. Something invisible, attached to the Earth, reaching up through the air and pulling the apple down. That is how Isaac Newton wrote it in 1687, and the equation he wrote for it is so accurate that NASA still uses it to fly probes around the solar system. It is not wrong. But it is not what is actually happening.

What is actually happening — according to the picture Einstein wrote down between 1907 and 1915, and that the rest of physics has been testing against for the better part of a hundred years — is that there is no force on the apple at all. The apple is doing the most natural thing an object can do. It is following the straightest possible path through the geometry of spacetime in the room. And the geometry of spacetime in the room is not flat. The Earth, sitting six and a half thousand kilometers underneath the kitchen floor, has bent it.

The floor of the kitchen is geometry. The apple is just rolling downhill on a hill you cannot see.

This is Lecture Three. General relativity. The lecture in which geometry stops being a fixed background and starts being something that responds to what is in it.

Where we are in the story (≈ 2.5 minutes)

Two lectures behind us. Quick recap, so the new ideas land on top of something solid.

In Lecture One we built the metric. We started from the Pythagorean rule on a flat sheet of paper, watched it fail when we curled the paper into a sphere, and ended with the general line element — ds^2 = g_{munu} dx^mu dx^nu. We said the metric is the local rule for converting coordinate steps into physical distances, and we said the rule can vary from one patch of the surface to the next. The metric, in that lecture, was the geometry.

In Lecture Two we extended the idea from space to spacetime. Three dimensions of space, one of time, glued into a single four-dimensional object by Hermann Minkowski in 1908. We met four-vectors — position, velocity, momentum — each of them a quantity that carries one time component and three spatial components and transforms cleanly when you change observers. And we met the stress-energy tensor, T_{munu}. The bookkeeping device that takes a small patch of spacetime and tells you how much mass-energy is in it, how much momentum is flowing through it, and how much pressure and shear it carries. The stress-energy tensor is how spacetime knows what is in it.

So at the end of Lecture Two we had two objects.

The metric on the one hand. The stress-energy tensor on the other hand. The geometry, and the contents of the geometry.

In every lecture before this one, those two objects have lived on opposite sides of the room. The metric was the stage. The stress-energy was the performers on the stage. They did not, in the conversation up to this point, speak to each other.

This lecture is the lecture in which they speak to each other.

Einstein's equations, the equations we are going to write down before the half-hour mark, are a single statement. The statement is that the geometry on the left and the matter-energy on the right are not independent. They are the same problem. The metric tells the stress-energy how to move. The stress-energy tells the metric how to bend. Neither side gets to choose without consulting the other.

That is what we are after. How geometry responds to mass and energy. And which solutions of that response have been measured.

The equivalence principle, plain language (≈ 4 minutes)

Einstein's path into general relativity begins not with an equation but with a thought experiment. He had it in 1907, while he was still working as a patent examiner in Bern, and he later called it the happiest thought of his life.

The thought is this. Imagine you are sealed inside a room. You cannot see outside. There are no windows. You are asked to perform any local experiment you like to determine whether the room is sitting on the surface of a planet — gravity acting downward — or whether the room is floating in deep space inside a rocket that is accelerating upward at exactly the strength of that planet's surface gravity.

You drop a ball. It falls to the floor. Same rate, in both rooms.

You shine a flashlight horizontally. The beam appears to bend very slightly downward. Same bend, in both rooms.

You weigh yourself on a bathroom scale. The scale reads the same in both rooms.

Einstein's claim — and this is the equivalence principle, in its plain-language form — is that no local experiment, performed inside the sealed room, can distinguish "sitting on a planet" from "accelerating in deep space." Locally, the two situations are physically identical.

That sounds like a small observation. It is not a small observation. Read it the other way around. If "sitting on a planet" and "accelerating in deep space" are locally indistinguishable, then gravity is not a force that some objects feel and others don't. Every object in the room — the ball, the flashlight beam, the researcher, the air — is being affected the same way by the same thing. And the thing is not a force pulling on each object individually. The thing is the geometry of the room itself.

The reason every object falls at the same rate in a gravitational field is not that gravity pulls equally hard on each one. It is that there is no pulling. The geometry of spacetime in the room has the property that the natural straight-line path of every object goes downward. The objects are not being acted on. They are coasting. They are following the straightest possible path that the local geometry permits. The floor is the only thing exerting any actual force, and it exerts it upward, on the objects' feet, to keep them from following their natural geodesic path through the floor and into the Earth.

The standard analogy for this — and you have seen it in every popular-relativity book published since the 1960s — is the rubber sheet. A heavy ball depresses a stretched fabric. A smaller marble rolling nearby curves toward the heavy ball, because the fabric is sloped. The marble is not being pulled. It is rolling on a sloped surface. The slope is the geometry.

The analogy is useful and it is also slightly wrong, and it is worth knowing how. The rubber sheet shows you space being bent by a mass. What general relativity actually says is that spacetime is bent — three dimensions of space and one of time, woven together — and that most of the apparent "pull" of gravity comes from the time component, not the space component. An apple falls because the geometry of time near the Earth is such that the apple's natural worldline through spacetime curves toward the Earth.

The space gets bent too, but for everyday objects moving slowly compared to the speed of light, the bending of time dominates. Light, by contrast, moves at the speed of light, so for light, both the time-bending and the space-bending contribute. That is why Eddington's 1919 measurement of starlight bending past the sun gave a number twice the value you would have gotten from a naive Newtonian-plus-equivalence-principle calculation. The factor of two is the space-curvature contribution. The rubber sheet does not show you that.

But the load-bearing idea is correct. Mass and energy bend spacetime. Objects coast along the geodesics — the straightest possible paths — of the bent spacetime. What we call gravity is what coasting through bent spacetime looks like from the inside.

That is the equivalence principle. That is the conceptual foundation of general relativity.

Now we have to write down the math.

Christoffel symbols — how to walk a straight line on a curved surface (≈ 5 minutes)

Here is the next problem. If geometry is bent, what does it even mean to walk in a straight line on it?

On a flat sheet of paper, "straight line" is easy. You point in some direction. You keep walking. You do not turn the wheel. The line you trace is straight.

On a curved surface, "straight" is harder. If you keep your feet pointing exactly forward and you do not turn, the curve of the surface itself will tilt the direction you are facing, from one step to the next. Stand on the equator, point yourself due north, and start walking. After a few thousand kilometers, the surface of the Earth has rotated the meaning of "due north" underneath you. You did not turn. The geometry turned you.

So if you are going to write down a rule for how to walk a straight line on a bent surface, you need a rule that tells you, at every point, how much the surface itself is rotating the local notion of "direction." That rule, in general relativity, is called a Christoffel symbol, and it is written Gamma^lambda_{munu} — capital Greek letter gamma, with three indices.

Three indices is a lot of indices. Let's name them. The bottom two — mu and nu — say "I am moving in direction nu, and I am tracking how much my direction-of-pointing in the mu axis gets tilted." The top one — lambda — says "and the tilt I am tracking is in the lambda direction." It is bookkeeping. Three labels, because at every point on a four-dimensional surface you need to track how each of four directions tilts as you step along each of four other directions, and the tilt itself is a quantity that points in some direction. Three labels handle all the cases at once.

You do not have to compute Christoffel symbols by hand to follow what they do. Their job is one sentence. A Christoffel symbol is the correction you have to make to "keep going straight" when the geometry under your feet is rotating your local sense of direction.

On a flat plane, every Christoffel symbol is zero. The geometry is not rotating you. You can ignore them entirely. The straight line you draw is the straight line Euclid drew in 300 BCE.

On a sphere, the Christoffel symbols are not zero. They encode the fact that the surface tilts your local north as you walk along the equator. If you write down the geodesic equation — the equation of motion for an object coasting freely under no forces — and you plug in the Christoffel symbols for a sphere, the geodesic equation tells you that the natural paths are great circles. The plane from New York to Tokyo flying over Alaska. That is what "coasting" looks like on a sphere.

In general relativity, the geodesic equation written with Christoffel symbols is the law of motion. An object with no forces on it — not propelling itself, not being pushed by anything external — follows the geodesic of the local spacetime metric. The apple. The Earth, going around the sun. A photon, going past a galaxy cluster. None of them are being acted on. All of them are coasting, on the geodesics of the bent spacetime they happen to be in. The bending is encoded in the Christoffel symbols. The Christoffel symbols are computed from the metric g_{munu} by a specific formula involving how the metric changes from point to point.

The Christoffel symbol is, in other words, the first derivative of the metric. It tells you how the metric is changing from one patch to the next. If the metric is constant — same rule everywhere — the derivatives are zero, the Christoffel symbols are zero, and there is no gravity. If the metric is varying — different rule at different patches — the derivatives are nonzero, the Christoffel symbols carry the variation, and what we call gravity emerges from that variation.

That is the first piece of mathematical machinery you need. We need one more piece before we can write Einstein's equations.

The Riemann tensor — actual curvature, not just coordinates (≈ 4.5 minutes)

There is a subtlety the Christoffel symbols do not handle on their own. Christoffel symbols can be nonzero on a flat surface, if you happen to be using bad coordinates. Polar coordinates on a flat plane, for example, give you nonzero Christoffel symbols even though the plane is not curved. The Christoffel symbol mixes two things together: the actual bending of the geometry, and the awkwardness of the coordinate grid you chose to describe it.

Distinguishing those two — actual curvature from coordinate weirdness — is exactly the job of the next object in the toolkit. It is called the Riemann curvature tensor, and it is written R^rho_{sigmamunu}. Four indices, which is even more bookkeeping, but the idea behind it is again one sentence.

The Riemann tensor measures how much a vector gets rotated when you carry it around a tiny closed loop.

That is the test. Take any vector — any arrow representing some physical quantity, a velocity, a momentum, anything that has a direction. Carry it around a small loop without turning it. If you come back to where you started and the vector is now pointing in a different direction than it did when you set out, the geometry inside the loop is actually curved. If the vector comes back exactly as you sent it, the geometry inside the loop is flat — regardless of what coordinate grid you happened to draw on it.

The Riemann tensor is the number that quantifies that mismatch. Zero Riemann means flat. Nonzero Riemann means curved. And the four indices are the bookkeeping for "which vector did I carry, in which directions did I form the loop, and how did the answer come out."

You can compute the Riemann tensor from the Christoffel symbols, and the Christoffel symbols from the metric. So the whole chain of curvature information starts at the metric and propagates outward. The metric determines the Christoffel symbols. The Christoffel symbols determine the Riemann tensor. The Riemann tensor tells you, at every point, how much the geometry is actually curved — not curved in a coordinate sense, curved in a sense that no choice of coordinates can flatten away.

Now we have everything we need. The Riemann tensor has too many indices for everyday physical statements — most of physics works with two-index summaries of it — so general relativity uses two contractions of the Riemann tensor that boil it down. The first is the Ricci tensor, written R_{munu}, which collapses two of the four indices and keeps the most physically relevant part of the curvature. The second is the Ricci scalar, written R, which collapses everything down to one number per point — a kind of average curvature at that point.

A useful one-sentence summary: the metric g{munu} tells you the geometry, the Christoffel symbols Gamma^lambda{munu} tell you how to walk straight in it, and the Riemann tensor R^rho{sigmamunu} — together with its boil-downs R{munu} and R — tells you how curved the geometry actually is. That is the entire mathematical apparatus of general relativity, named in plain language.

We have the geometry side. We had the matter side already, from Lecture Two — the stress-energy tensor T_{munu}. Now Einstein has to write the equation that ties them together.

The Einstein field equations, symbol by symbol (≈ 4.5 minutes)

Here it is. The equation Einstein arrived at in November 1915, after eight years of work on the equivalence principle, and that he published in its final form on the 25th of November of that year in the Sitzungsberichte of the Prussian Academy of Sciences in Berlin.

$G{munu} = 8pi T{munu}$

Read it once. Geometry on the left. Mass-energy on the right. Equals sign in the middle.

The version on screen uses what physicists call geometric units — a convention in which Newton's gravitational constant and the speed of light are both set to one. In SI units the equation looks busier, with a factor of 8pi G / c^4 on the right where the 8pi alone sits in the geometric form. The physics is the same. The geometric-units form is cleaner to read out loud, so we will use it.

Let's name every piece. Slowly. Same discipline as Lecture One, when we named every piece of ds^2 = g_{munu} dx^mu dx^nu.

The thing on the left, G{munu} — capital G, two indices — is the Einstein tensor. It is built out of the Ricci tensor and the Ricci scalar that we met two minutes ago, in a specific combination: G{munu} = R{munu} - tfrac{1}{2} g{munu} R. You do not have to memorize that combination. You have to know what it represents. The Einstein tensor is the curvature of spacetime, repackaged in a form that is automatically consistent with energy and momentum being conserved. That second clause is the load-bearing part.

The reason the Einstein tensor takes the specific Ricci-minus-half-times-metric-times-scalar form rather than just being the Ricci tensor on its own is that this is the unique combination whose mathematical structure guarantees that anything written equal to it on the right-hand side will have its own conservation laws satisfied. Einstein went through several wrong versions before finding this one. The wrong versions did not have automatic conservation. The right version does.

So the left-hand side reads: the curvature of spacetime, in a conservation-respecting form.

The thing on the right, T{munu}, is the stress-energy tensor. We met it in Lecture Two. Every cell of T{munu} at a point tells you something about what is at that point: how much energy density, how much momentum, how much pressure, how much shear stress. If the point is in empty space far from any mass, T{munu} is zero. If the point is inside the sun, T{munu} has a large energy-density entry and several smaller entries for the pressure of the solar plasma. If the point is in the cosmic microwave background, T_{munu} has tiny entries that reflect the very dilute radiation filling the universe.

In between the curvature on the left and the matter-energy on the right sits 8pi. The factor of eight pi is, in this unit system, just a constant. It is what falls out of the calculation when you demand that Einstein's equation reduce to Newton's law of gravity in the limit of weak fields and slow speeds. The 8pi is the conversion factor between "how much spacetime curves" and "how much mass-energy bent it." A bigger pile of mass-energy on the right produces a bigger curvature on the left. The constant of proportionality is 8pi — in these units.

Read it as a sentence. At every point of spacetime, the curvature of spacetime is determined by the mass-energy content at that point — with a fixed constant of proportionality. The geometry is not a stage. The geometry is set by what is on it.

That is the central content of general relativity. Ten equations, packed into the index notation as one. Solve them for a given mass-energy distribution and you get a specific g_{munu} — a specific metric — that tells you the geometry of spacetime around that distribution. Every prediction of the theory falls out of that.

We are going to bring in Kip Thorne now. The person who, with Charles Misner and John Wheeler, wrote the standard graduate textbook this entire derivation lives in. He is going to tell us about the two cleanest solutions of these equations — the ones that have been tested against observation — and what has actually been measured.

Thorne on the solutions that have been confirmed (≈ 7 minutes)

There are two solutions of the Einstein field equations I want to walk you through, because between them they cover most of the territory the equations have been tested against.

The first is the Schwarzschild solution. Karl Schwarzschild wrote it down in December 1915, less than two months after Einstein published the field equations themselves, while Schwarzschild was serving in the German army on the Russian front. He sent it to Einstein, who presented it to the Prussian Academy on Schwarzschild's behalf in January 1916. Schwarzschild died of an autoimmune disease that May. He never saw what his solution would do to physics.

The solution assumes the simplest case you can write. Empty space outside a single, non-rotating, spherically symmetric mass. No other matter anywhere. No rotation. Just one point of mass, sitting at the center of a vacuum. You write down that ansatz on the right-hand side of the field equations, and you solve for the metric on the left. The line element you get has a specific, exact form — and it has two physically distinct features that have been measured.

First, time runs at different rates at different distances from the central mass. The closer you are to the mass, the slower your clock runs, compared to a clock far away. That gravitational time dilation is not a theoretical curiosity. It is the effect we correct for in every GPS satellite. The atomic clocks on board the GPS constellation, twenty thousand kilometers above the Earth's surface, run faster than the clocks at the ground stations, by about thirty-eight microseconds per day, exactly as the Schwarzschild solution predicts. If we did not correct for that, GPS positioning would drift by about ten kilometers per day.

Every smartphone in the world depends, quietly, on the Schwarzschild metric being correct.

Second, the solution has a critical radius. At a specific distance from the center, called the Schwarzschild radius, the geometry of the solution becomes degenerate in a particular technical sense. For an object whose mass is concentrated inside that radius — for an object dense enough that there is no longer any physical surface outside the Schwarzschild radius — the geometry has the property that no light, no particle, no signal of any kind that originates inside that radius can escape to infinity. The Schwarzschild radius is the event horizon of a black hole. The Schwarzschild solution is the description of the spacetime geometry around a non-rotating black hole.

When John Wheeler started using the term black hole in lectures in 1967, he was talking about exactly this object. The vacuum exterior of the Schwarzschild solution.

The classical tests. Schwarzschild predicts that light passing close to a massive body — the sun, for example — should be deflected by a specific angle: 1.75 arcseconds for a ray grazing the solar limb. Arthur Eddington measured that deflection during a total solar eclipse on the 29th of May 1919, from the islands of Príncipe and Sobral, and the value he reported was consistent with the Einstein prediction and inconsistent with the Newtonian prediction by a factor of two. The 1919 measurement is now over a hundred years old.

The light-bending prediction has been re-measured at radio frequencies using Very Long Baseline Interferometry to a precision of about one part in ten thousand, and it agrees with the Schwarzschild prediction at every measurement.

Schwarzschild also predicts a specific shift in the perihelion of Mercury — the small annual rotation of the long axis of Mercury's orbit around the sun. Newton's gravity, after all the perturbations from the other planets are subtracted, leaves an unexplained residual of about forty-three arcseconds per century. Schwarzschild's solution predicts exactly that excess, with no free parameters. It was, in 1915, the first quantitative success of the theory. Einstein wrote in a letter at the time that the discovery had given him heart palpitations.

And in 2015, exactly a hundred years after Einstein's field-equations paper, the Laser Interferometer Gravitational-Wave Observatory detected the first direct signal from the merger of two stellar-mass black holes — designated GW150914. The signal was a sweep, in frequency and amplitude, that matched the numerical-relativity prediction for two Schwarzschild-class objects of about thirty-six and twenty-nine solar masses spiraling into each other and merging into a single black hole of about sixty-two solar masses, with about three solar masses radiated away as gravitational waves during the merger itself. The match between the data and the predicted waveform is one of the strongest single-experiment tests general relativity has ever passed.

That is the Schwarzschild solution. One mass, in vacuum, in empty space. The cleanest exact solution of Einstein's equations. Confirmed by GPS, by light bending, by perihelion precession, by gravitational-wave inspirals.

The second solution I want to walk you through is on the opposite end of the scale. The Friedmann-Lemaître-Robertson-Walker solution.

This is the solution you get if you assume the universe is, on the largest scales, homogeneous and isotropic — the same density everywhere, the same in every direction. Not on the scale of a kitchen, where there is an apple and not an apple. Not on the scale of a solar system, where there is a sun and not a sun. On scales of hundreds of millions of light-years, where the lumpiness averages out and the universe looks, to a first approximation, like a uniform fluid.

If you write down that assumption on the right-hand side of the field equations, the geometry on the left-hand side has a specific form. There is one scale factor — call it a(t) — that tells you how stretched the universe is at every cosmic time. The line element is essentially the flat-space metric, but with the spatial part multiplied by a(t) at every moment. As a(t) grows, the distance between any two galaxies that are otherwise just sitting still grows along with it. That is what cosmologists mean by the expansion of the universe. Space itself is being stretched. The galaxies are not moving through space.

The space between them is getting bigger.

This solution was worked out independently in the 1920s and 1930s. Alexander Friedmann in Petrograd published the first version in 1922. Georges Lemaître in Louvain published an independent version in 1927 — and in the same 1927 paper, he gave the first derivation of what is now called Hubble's law, two years before Edwin Hubble's observational paper. Howard Percy Robertson and Arthur Geoffrey Walker refined the mathematics to its modern form in the mid-1930s. All four names get attached to the solution, in the order they entered the literature.

The classical tests. The FLRW solution predicts that the universe should be expanding — and Hubble's law is the observed evidence that it is. The recession velocity of distant galaxies is proportional to their distance, exactly as a uniform expansion would predict. The solution also predicts that, if the expansion is real, there should be a leftover thermal glow from the dense, hot early universe — a cosmic background of microwave radiation, with a near-perfect blackbody spectrum at a temperature of a few degrees above absolute zero.

Arno Penzias and Robert Wilson measured exactly that radiation in 1964 from a horn antenna at Bell Labs, with a temperature of about 3 kelvin, and earned the 1978 Nobel Prize in Physics for the measurement. The FIRAS instrument on COBE, launched in 1989, measured the spectrum to be the most perfect blackbody ever observed in nature, with deviations smaller than one part in ten thousand. The cosmic microwave background is the FLRW solution speaking to us at a temperature of 2.725 kelvin.

So we have two exact solutions of the Einstein field equations. One — Schwarzschild — describes the local geometry around a concentrated mass, and it has been confirmed by GPS, eclipse measurements, perihelion precession, and gravitational-wave inspirals. The other — Friedmann-Lemaître-Robertson-Walker — describes the global geometry of a homogeneous universe, and it has been confirmed by the redshift-distance relation, by the cosmic microwave background, and by the abundances of the light elements produced in the first three minutes after the Big Bang. Both are mainstream physics. Both have been tested. Both work.

What general relativity says, in those two confirmed cases, is that the metric is what the matter-and-energy distribution dictates. The geometry is not a free parameter. You put mass-energy on the right of Einstein's equations. The metric comes out on the left. The universe gets to choose the metric only by choosing what mass-energy is in it.

Hossenfelder, on what the math admits versus what the universe permits (≈ 4 minutes)

Lecture One ended on a discipline. Hossenfelder's discipline. A solution of the equations is not yet a piece of physics; a piece of physics is a solution that the universe has been independently observed to permit. That distinction is going to come back in every Block-B lecture and every Block-D lecture in this series, so the most useful thing I can do now is restate it in Hossenfelder's own published words, in exactly the form she wrote it on her Backreaction blog on the 21st of November 2020.

Any space-time will solve the equations of General Relativity, provided you assume suitable mass and energy distributions. The real question is whether required distributions are physically reasonable.

Read that against everything we just walked through. Thorne gave you two solutions of Einstein's equations whose required mass-energy distributions are boringly physically reasonable. The Schwarzschild solution requires one compact mass. The FLRW solution requires a uniform fluid. Both are sources we know how to find in the universe. We can point at the sun. We can point at a galaxy. We can point at the cosmic microwave background. The mass and energy that those solutions require on the right-hand side of Einstein's equations is mass and energy that the universe has been observed to actually contain.

That is why those two solutions are mainstream-confirmed. The math admits the solution. The universe also permits the implementation. Both halves of Hossenfelder's two-question test are satisfied.

What is coming in Lectures Four and Five — the next two lectures in this season — is what happens when the math admits a solution but the required distribution on the right-hand side stops being something the universe is known to deliver. In Lecture Four we will meet Miguel Alcubierre's 1994 warp metric — a specific exact solution of Einstein's equations that describes a localized region of spacetime moving faster than the surrounding spacetime. The metric on the left-hand side is mathematically clean.

The stress-energy distribution required on the right-hand side involves negative energy density at macroscopic scale — and that is something the universe, as far as we know, does not provide on the scales the solution requires. In Lecture Five we will meet the Morris-Thorne wormhole — a Kip Thorne paper, written in 1988 with his graduate student Michael Morris in the American Journal of Physics, in part because Carl Sagan asked Thorne what kind of geometry might let a character in his novel Contact travel between stars. Thorne worked out the answer. The geometry exists. The metric solves Einstein's equations.

The required source — again — is negative energy density at macroscopic scale at the throat of the wormhole.

Hossenfelder's distinction is the editorial spine of those two lectures. The Alcubierre solution is real theoretical physics. The Morris-Thorne solution is real theoretical physics. Both are in peer-reviewed journals, written by credentialed numerical relativists, citable by DOI. And both are gated, at the engineering level, by whether the right-hand side of Einstein's equations can be physically sourced at the scales the solutions demand.

That is not a critique of those lectures. It is the content of those lectures. The discipline of separating "math admits a solution" from "universe permits an implementation" is what makes the metric-engineering question a real research question rather than a daydream. The Schwarzschild metric is mainstream because its source is mainstream. The Alcubierre metric is exotic because its source is exotic. The work, when it gets serious, is the work of finding out whether the exotic source can be made non-exotic.

What this lecture buys you (≈ 3 minutes)

Let me tell you what you can now do that you could not do at the end of Lecture Two.

You can read G{munu} = 8pi T{munu} and say what each side carries. Geometry on the left, in a conservation-respecting form. Mass-energy on the right, in the bookkeeping form we built in the last lecture. A constant of proportionality in between, set by the requirement that the theory reduce to Newton's law in the slow-and-weak limit. That is one equation. Ten equations packed into the index notation. Solve them, and you get a specific metric. That metric tells you the geometry. The geometry tells the matter where to coast. The matter tells the geometry where to bend. Neither side moves without the other.

You can name two solutions of these equations that have been tested against observation, and one observational test of each. Schwarzschild — confirmed by GPS time dilation, by Eddington's eclipse measurement of light bending past the sun, by the perihelion precession of Mercury, by the LIGO observation of black-hole inspirals. FLRW — confirmed by Hubble's redshift-distance law, by the cosmic microwave background, by the abundances of the light elements left over from the first three minutes of the universe.

You can hold the editorial distinction that this entire series rests on. The Einstein equations admit many solutions. The universe permits only those whose required mass-energy distributions can be physically sourced. The two solutions we walked through today — Schwarzschild and FLRW — are mainstream because their sources are mainstream. The two solutions we will walk through next — Alcubierre and Morris-Thorne — are exotic because their sources are exotic. The metric-engineering research program is the program that asks whether, under any boundary conditions, the exotic sources can become non-exotic. That is the question this lecture series is built to engage.

The discipline of the question is exactly the discipline Hossenfelder named, applied to every concrete proposal we will meet in Blocks B, C, and D.

Three blocks ahead of you in Season One. Block B starts in the next lecture — exotic solutions of Einstein's equations, written down by credentialed general relativists, published in peer-reviewed journals, and gated at the engineering level by the sources they require. Block C, the quantum vacuum and the only place in nature where negative energy densities have been experimentally measured at macroscopic scales — the Casimir effect, the dynamic Casimir effect, the vacuum's structure as a place that can be acted on. Block D, the engineering proposals — patents filed and attested, evaluations conducted and reported, replications attempted and either confirmed or null. Throughout, the same discipline.

The metric on the left. The required source on the right. The question every time: has the required source been independently demonstrated, and at what scale.

Closing (≈ 90 seconds)

Watch the apple again.

It is not being pulled. It is coasting. It is following the geodesic of the spacetime metric that the Earth, six and a half thousand kilometers underneath the kitchen floor, has bent. The bending is encoded in g_{munu}. The amount of bending is set by Einstein's equations. The Earth's mass goes on the right-hand side. The kitchen's geometry comes out on the left-hand side. The apple, with no force on it of any kind, falls because the straightest possible path through the bent geometry of the room runs downward through the floor.

That is what happens at every scale of the universe. The Earth coasts around the sun on the geodesic of the metric the sun has bent. The sun coasts around the galaxy on the geodesic of the metric the galaxy has bent. The galaxies in the cosmic web coast through the geometry of the universe on the geodesics of the metric that the average density of mass-energy in the universe has bent. Same equation. Same constant of proportionality. Different sources. Different metrics.

Lecture Four picks up the same equation and asks a harder question. What happens when somebody writes down a metric that solves Einstein's equations cleanly — but whose source, on the right-hand side, is something the universe is not known to provide. Alcubierre's 1994 paper, in Classical and Quantum Gravity. The warp drive, as a solution of general relativity, with the engineering question of the source held open.

A short bridge before we leave the kitchen. Forbes's framing in Lecture Zero — that the deep principle is simpler than the popular coverage suggests — is the editorial frame the series adopts. We do not adopt his specific physical identification ("it's Coulomb's equation") as a load-bearing claim; that identification is not what the last three lectures have walked. What we have walked is the vocabulary — the metric on a flat sheet, the Lorentz transformation, the Einstein field equations — and the vocabulary is tractable for a careful undergraduate.

Salvatore Pais — who is a credentialed physicist — put the discipline more carefully in Hard Truths Podcast number one: "The simpler the mathematics, the more important the physics — especially if you have a good understanding of what the problem is." That is the credentialed restatement the series carries forward. Three lectures of vocabulary buys you, now, the right to ask the harder question Lecture 4 takes up.

Same equation. Same listener. New geometry.