How Does Gravity Actually Work? (It's Not a Pull)

General relativity describes gravity through the geometry of space and time. Newton's force model remains useful, while Einstein's theory explains effects that the older model misses.

The YouTube player loads only when you press Play. Playing connects to YouTube, which may use cookies and storage. See Privacy.

Gravity and a falling ball

A dropped ball accelerates toward Earth. Newton describes this with a gravitational force. Einstein describes free fall as motion along a natural path through spacetime.

Spacetime combines three spatial dimensions with time. The two descriptions can predict almost the same everyday motion, despite using different concepts. The video's phrase “nothing pulls” concerns Einstein's description, rather than a claim that Newtonian calculations stop working.

Galileo's ramps

In the early seventeenth century, Galileo uses inclined planes to study accelerated motion. A gentle slope slows the motion enough for measurement. The experiment makes a rapid fall easier to investigate.

For uniform acceleration from rest, successive equal time intervals give distances in the ratio 1:3:5:7. Adding those distances gives totals of 1, 4, 9, and 16. Total distance therefore grows as the square of elapsed time.

Rolling on a ramp differs from unrestricted free fall because rotation also uses energy. The useful connection is the pattern of constant acceleration. The ramp's acceleration need not equal Earth's vertical free-fall acceleration.

Without significant air resistance, objects fall with the same acceleration under the same conditions, regardless of their ordinary mass. The famous Leaning Tower story has uncertain historical status. The experimental work does not depend on that story being true.

Newton's falling Moon

Newton connects falling objects with orbital motion. The Moon lies about 60 Earth radii from Earth's centre. An inverse-square law reduces gravitational acceleration there by a factor of about 3,600.

Near Earth's surface, an object released from rest falls nearly 5 metres in its first second. Dividing by 3,600 gives about 1.4 millimetres. This estimates the Moon's inward deviation from a straight tangent over a second.

The Moon does not lose 1.4 millimetres of orbital height every second. It moves sideways while gravity changes its direction. Its path curves around Earth instead of continuing along a tangent.

The cannonball drawing develops that idea. Faster horizontal motion carries the ball farther before it reaches the ground. At an appropriate orbital speed, Earth's surface curves away beneath the falling path.

Newton's 1687 law connects all masses with gravitational attraction. Doubling separation reduces the force to one quarter, if the masses stay unchanged. The same framework helps explain planets, tides, and comets, and still supports many spacecraft calculations.

The problem with Newton

Newton's law predicts a force but does not supply a physical mechanism that carries it through empty space. Earth and the Sun have a separation near 150 million kilometres. The question concerns how their interaction relates across that distance.

Newton himself objects to treating unexplained action at a distance as a complete account. His concern does not erase the predictive success of the law. It separates a useful mathematical rule from a deeper physical explanation.

Mercury supplies a quantitative problem. Its orbit's closest-approach direction slowly turns. After known Newtonian effects, an extra shift of about 43 arcseconds per century remains.

An arcsecond is a small angular unit, equal to one three-thousand-six-hundredth of a degree. The discrepancy is small, but repeated measurements make it scientifically important.

The hammer and the feather

In 1971, Apollo 15 astronaut David Scott drops a hammer and a feather on the Moon. They reach the surface together within the demonstration's precision.

The Moon has no substantial atmosphere to slow the feather. On Earth, air resistance strongly changes a feather's motion. The comparison separates gravitational acceleration from an additional force caused by air.

The result illustrates the universality of free fall. Einstein treats that shared response as a clue about gravity's nature. The historical order still matters: his ideas precede the lunar demonstration.

Einstein's falling box

In 1907, Einstein considers an observer in free fall. A scale inside the falling box reads near zero because it no longer supports the observer. The observer and nearby loose objects follow almost the same fall.

Inside a small freely falling laboratory, objects appear weightless. Gravity has not disappeared from the universe. Over a sufficiently large region, differences in gravitational acceleration produce tidal effects that reveal the field.

Now imagine a box accelerating through otherwise empty space. Its floor accelerates toward a released ball. An observer inside can interpret that relative motion as a downward fall.

This local connection has the name equivalence principle. It relates acceleration and gravity within suitable limits. It does not make every extended gravitational field identical to one uniformly accelerating box.

Why light bends

Imagine a light beam crossing the accelerating box. During its flight, the box moves upward by an increasing amount. Relative to the box, the beam follows a curved track.

The equivalence principle therefore suggests that gravity affects light's path. Light does not need a rest mass for this effect. The full prediction requires the geometry of spacetime.

The eclipse of 1919

During a total solar eclipse, the darkened sky makes stars near the Sun easier to photograph. Astronomers compare their apparent positions with positions when the Sun lies elsewhere.

General relativity predicts a deflection near 1.75 arcseconds for light passing close to the Sun's edge. A simpler Newtonian corpuscle calculation gives about half that value.

The 1919 observations favour Einstein's prediction, with substantial measurement uncertainty. “Roughly two arcseconds” describes the scale, rather than an exact value for every star. Later observations provide much stronger tests.

Curved spacetime

Einstein presents general relativity in 1915. Matter and energy affect spacetime geometry. Freely moving objects follow geodesics, which are the natural paths defined by that geometry.

A geodesic need not look straight on a spatial drawing. A planet's orbit can curve around a star while remaining a free-fall path through spacetime. The drawing shows the path's spatial projection.

The theory accounts for Mercury's extra 43 arcseconds per century. That is a specific numerical success, beyond a general claim that space can bend.

John Wheeler's familiar description links matter's influence on geometry with geometry's influence on motion. It is a compact guide to the relationship. Einstein's equations give the precise version.

Why you feel heavy

When you stand, the ground prevents your natural free fall. Contact forces support your body. A scale measures that support force.

In free fall, the scale loses the force it normally exerts against you. This explains weightlessness without requiring gravity to become zero. The difference is between supported motion and free motion.

Gravity and different clock rates

For slow motion in a weak, nearly static gravitational field, the time part of the geometry gives the leading Newtonian acceleration. This is the limited meaning behind the video's statement that most of the effect is “in time”.

Compare stationary clocks at different heights near Earth. The lower clock ticks more slowly relative to the higher clock. Moving clocks require an additional motion correction.

Compared with a stationary clock very far from Earth, the surface difference is about seven parts in ten billion. The comparison idealises Earth's field and specifies the clocks' motion. It is not a statement that every clock nearer any mass always loses against every distant clock.

Pound and Rebka test gravitational frequency shifts with gamma rays across a Harvard tower about 22 metres high. Their work begins in 1959. Modern atomic-clock experiments resolve a corresponding height effect across a millimetre-scale sample.

GPS and relativity

GPS satellites carry clocks that require both gravitational and motion corrections. Their higher position produces a gain near 45 microseconds per day relative to surface clocks. A microsecond is one millionth of a second.

Their orbital motion produces a loss near 7 microseconds per day. The combined difference is therefore about 38 microseconds per day. Both effects matter; quoting only the gravitational term gives an incomplete comparison.

Light travels more than 10 kilometres during that daily timing difference. This shows the scale of range errors that an uncorrected system can develop. Actual position error also depends on the navigation calculation and satellite geometry.

The axle picture and its limits

The video shows two wheels joined by an axle. If one wheel turns more slowly, the axle changes direction toward that side. The picture suggests how a difference across space can alter a path.

An object does not literally contain two time wheels. A point-like object also falls, so its two sides cannot provide the full mechanism. The exact explanation uses geodesics and the variation of spacetime geometry.

The analogy helps connect clock-rate differences with motion under weak-field conditions. It should not replace general relativity with a universal rule that objects simply seek slower clocks.

Gravitational waves

Changing mass distributions can produce travelling changes in spacetime geometry. These have the name gravitational waves. They stretch and squeeze separations in a characteristic pattern.

On September 14, 2015, LIGO detects waves from a merging pair of black holes. The source lies at a distance corresponding to travel over roughly 1.3 billion years.

The detector signal corresponds to extremely small changes in its long arms. The video's proton comparison places the motion near a thousandth of a proton's width. It is a scale comparison, not a direct image of a moving proton.

The weakest force and an open question

A small magnet can hold a paperclip against Earth's gravity. This familiar example illustrates the relative weakness of gravity compared with electromagnetic forces in that situation.

General relativity describes many large-scale phenomena extremely well. A complete, experimentally established quantum theory of gravity remains unavailable. The question concerns how gravity and quantum physics fit together at the smallest scales.

Recap

Galileo measures accelerated motion. Newton connects falling objects and orbits with one force law. Einstein explains free fall and clock effects through spacetime geometry.

What this means

Free fall and supported motion differ in what an observer feels. Relativity explains that difference while retaining Newton's law as a useful approximation under suitable conditions.

FAQ

Does the Moon continually approach Earth by the quoted amount?

No. The number describes inward deviation from a tangent, not a repeated loss of orbital height.

Does free fall remove every sign of gravity?

No. Tidal effects can reveal gravity across a sufficiently large region.

Do GPS clocks need only a gravity correction?

No. Their motion also changes their rate relative to the chosen surface reference.

Is the axle a literal model of an atom?

No. It is a visual analogy with limited scope.