Visual walkthrough — Spin — intrinsic angular momentum

---

Step 1 — What "angular momentum" even means (a picture, not a formula)

WHAT. Angular momentum is a number (really a little arrow) that measures how much turning something carries. For a real spinning object it points along the spin axis, and it is longer when the object turns faster or is heavier.

WHY start here. Every symbol below — $\vec{S}$, $S_z$, $|\vec{S}|$ — is a piece of one arrow. If you don't picture the arrow first, the letters are noise. We will treat spin as an arrow living in ordinary 3D space for now, because that picture makes the algebra visible — but see the caveat below before you trust the picture too far.

PICTURE. Look at the figure. The blue arrow is the angular-momentum arrow. Its length is one fact about it; its shadow on the vertical axis is a different fact. We will spend the whole page on the tug-of-war between "length" and "shadow."

The letter $\vec{S}$ (with the little arrow on top) is the whole spin arrow. Plain $S_z$ is just its up–down shadow. Keep those separate — half the mistakes on this topic come from mixing them.

---

Step 2 — The rules the universe hands us

WHAT. Nature gives us a starting rule about angular-momentum arrows. Because space has no special axis (it is isotropic — the same in all directions), the rule cannot single out $x$, $y$, or $z$: it must come as a matched set of three, one for each pairing of axes. This set is the cyclic triple $[S_x,S_y]=i\hslash S_z,[S_y,S_z]=i\hslash S_x,[S_z,S_x]=i\hslash S_y.$

Reading the symbols:

• $S_x,S_y,S_z$ — the shadows of the arrow on the three axes (left–right, front–back, up–down).
• $[A,B]$ — the commutator, shorthand for $AB-BA$: "does the order of doing $A$ then $B$ matter?" If it's zero, order doesn't matter; if not, the two things interfere.
• $\hslash$ — a fixed tiny number of nature (Planck's constant divided by $2\pi$), the natural chunk-size of angular momentum. Units: joule-seconds.
• $i$ — the imaginary unit, $\sqrt{-1}$; here it is just bookkeeping that keeps the algebra consistent.
• "Cyclic" means: read the labels round a loop $x\to y\to z\to x$. Each line is the one before with the labels shifted one place along that loop. No axis is special — that is isotropy made into algebra.

WHY this tool and not another, and why all three lines. We could try to guess how spin behaves. Instead we borrow the exact rule that Orbital Angular Momentum already obeys, because spin turned out to obey the same algebra. We need the whole cyclic set, not just the first line, because in Step 3 we will build the total length ${\vec{S}}^2$ out of all three axes; only with all three commutators can we prove the length has a fixed, definite value.

WHAT IT SAYS IN PICTURES. Every one of these commutators is non-zero. That means: you cannot pin down all three shadows at once. Measure the up–down shadow sharply and the other two shadows blur.

---

Step 3 — Length: deriving $|\vec{S}|=\hslash \sqrt{s(s+1)}$ from the ladder

WHAT. The squared length of the arrow is the sum of its three shadows squared — Pythagoras in 3D: ${\vec{S}}^2=S_x^2+S_y^2+S_z^2.$

• Each $S_x^2,S_y^2,S_z^2$ is a shadow squared (always $\ge 0$).
• Adding them gives the full squared length, by the 3D Pythagorean theorem.

Because ${\vec{S}}^2$ commutes with $S_z$ (Step 2), a state can have a definite squared length and a definite shadow at once. We do not assume the $s(s+1)$ answer — we derive it. Here is the whole argument in three clicks (the figure carries them).

Click 1 — rewrite ${\vec{S}}^2$ using the ladder operators. Recall (from Step 4's helpers, which we introduce a moment early because we need them here) $S_{\pm }=S_x\pm i{S}_y$. Multiply them out: $S_{-}{S}_{+}=(S_x-i{S}_y)(S_x+i{S}_y)=S_x^2+S_y^2+i[S_x,S_y]=S_x^2+S_y^2-\hslash S_z,$ using $[S_x,S_y]=i\hslash S_z$. But $S_x^2+S_y^2={\vec{S}}^2-S_z^2$, so $\boxed{{\vec{S}}^2=S_{-}{S}_{+}+S_z^2+\hslash S_z}$

• Each symbol: $S_{-}{S}_{+}$ is "step up then step down," $S_z^2$ the shadow squared, $\hslash S_z$ the leftover.

Click 2 — evaluate on the top rung. Call the top rung of the ladder $m_{\max }$, where the shadow is $S_z=m_{\max }\hslash$ and there is nowhere higher to climb, so $S_{+}$ gives zero. Apply the boxed identity to that top state: $S_{-}{S}_{+}$ contributes nothing (because $S_{+}$ already killed the state), leaving ${\vec{S}}^2=(m_{\max }\hslash {)}^2+\hslash (m_{\max }\hslash )={\hslash }^2{m}_{\max }(m_{\max }+1).$ That is where the $m(m+1)$ shape is born — straight from "step up then down = length minus $S_z^2$ minus $\hslash S_z$."

Click 3 — name the top rung $s$. Define the spin quantum number as the top-rung shadow, $s\equiv m_{\max }$. Then ${\vec{S}}^2={\hslash }^{2}s(s+1),|\vec{S}|=\hslash \sqrt{s(s+1)}.$ Because ${\vec{S}}^2$ commutes with every component, this same length holds for every rung, not just the top — the length is one fixed number for the whole ladder.

WHY the length beats the naive $s\hslash$. The extra "$+s$" inside the root is the $\hslash S_z$ term in the boxed identity — the unkillable contribution of the sideways shadows $S_x,S_y$, which Step 2 forbade from both being zero. The naive guess ignores that term.

PICTURE. The figure walks the three clicks: the ladder identity, the top-rung cancellation of $S_{+}$, and the resulting $s(s+1)$ length compared with the short naive $s\hslash$ bar.

---

Step 4 — Shadow: why $S_z$ comes in equal steps of $\hslash$

WHAT. The up–down shadow is not free to be anything. It is quantized: $S_z=m_s\hslash ,m_s=-s,-s+1,\dots ,+s.$

• $m_s$ — the magnetic quantum number for spin: a counter that ticks up by 1 each rung.
• Multiplying by $\hslash$ turns the pure count $m_s$ into an actual amount of angular-momentum shadow.

WHY equal steps — the concrete reason. The ladder operators we already met: $S_{+}=S_x+i{S}_y(\text{step up}),S_{-}=S_x-i{S}_y(\text{step down}).$ Compute their commutator with $S_z$, using the cyclic set of Step 2: $[S_z,S_{+}]=[S_z,S_x]+i[S_z,S_y]=i\hslash S_y+i(-i\hslash S_x)=\hslash (S_x+i{S}_y)=+\hslash S_{+},$ and identically $[S_z,S_{-}]=-\hslash S_{-}$. Read this as an instruction: applying $S_{+}$ raises the $S_z$-shadow by exactly one $\hslash$, and $S_{-}$ lowers it by one $\hslash$. (Proof of the "raises by $\hslash$" reading: if $S_z\psi =m_s\hslash \psi$, then $S_z(S_{+}\psi )=(S_{+}{S}_z+\hslash S_{+})\psi =(m_s+1)\hslash (S_{+}\psi )$ — the new state has shadow one $\hslash$ higher.) That is the concrete "why," not a hand-wave.

WHY the ladder must stop (this pins down $s$). A shadow can't be longer than the whole arrow, so $S_z^2\le {\vec{S}}^2$. Hence the climb up must hit a top rung $m_s=+s$ (where $S_{+}$ gives zero) and the climb down a bottom rung $m_s=-s$ (where $S_{-}$ gives zero). Top and bottom sit symmetrically about zero, and the number of $\hslash$-steps between them, namely $2s$, must be a whole number. That whole-number condition is what allows $s=\frac{1}{2}$: from $+\frac{1}{2}$ you take one step down to $-\frac{1}{2}$ and you're done — two rungs.

PICTURE. The figure is a literal ladder. For $s=\frac{1}{2}$ there are only two rungs: $m_s=+\frac{1}{2}$ and $m_s=-\frac{1}{2}$. The $S_{+}$ / $S_{-}$ arrows hop between them; there is nothing above or below.

---

Step 5 — Put length and shadow together: the tilt angle

WHAT. Now combine the two numbers into geometry. Drop the spin arrow, its shadow, and the axis into one right triangle. The angle $\theta$ from the $z$-axis satisfies $\cos \theta =\frac{S_z}{|\vec{S}|}=\frac{m_s\hslash }{\hslash \sqrt{s(s+1)}}.$

• The numerator $S_z=m_s\hslash$ is the adjacent side (shadow along $z$).
• The denominator $|\vec{S}|$ is the hypotenuse (the full arrow).
• $\cos \theta =\frac{\text{adjacent}}{\text{hypotenuse}}$ — that's the definition of cosine on this triangle, and it's the tool that converts "shadow over length" into an angle.

WHY cosine and not something else? We want the angle between the arrow and the vertical axis. Cosine is the one trig ratio built from adjacent-over-hypotenuse, and here "adjacent" is precisely the vertical shadow. So cosine is the natural translator from our two numbers into a tilt.

Plug in the electron ($m_s=+\frac{1}{2}$, $s=\frac{1}{2}$): $\cos \theta =\frac{1/2}{\sqrt{3}/2}=\frac{1}{\sqrt{3}}\approx 0.577,\theta =\arccos \frac{1}{\sqrt{3}}\approx 54.7^{\circ }.$ Here $\arccos$ is simply the question "which angle has this cosine?" — it undoes cosine.

WHY it can't be $0^{\circ }$. Straight up would need shadow = length, i.e. $\cos \theta =1$. But the shadow $\frac{1}{2}\hslash$ is smaller than the length $\frac{\sqrt{3}}{2}\hslash$ (Step 3's sideways tax). So the arrow is stuck leaning — never vertical.

PICTURE. The right triangle drawn to scale: hypotenuse $\frac{\sqrt{3}}{2}\hslash$, vertical side $\frac{1}{2}\hslash$, and the $54.7^{\circ }$ opening. The spin-down case is the same triangle flipped below the axis at $180^{\circ }-54.7^{\circ }=125.3^{\circ }$.

---

Step 6 — The magnetic moment: why the arrow can be pushed

WHAT. A spin arrow that carries charge also acts like a tiny bar magnet — a magnetic moment ${\vec{\mu }}_s$: ${\vec{\mu }}_s=-g_s\frac{e}{2m_e}\vec{S}.$

• $e$ — the electron's charge magnitude; $m_e$ — its mass.
• $\frac{e}{2m_e}$ — the classical "charge-loop" conversion from angular momentum to magnetism.
• The minus sign — the electron is negative, so its magnet points opposite to its spin arrow.
• $g_s$ — the spin g-factor, a fudge that experiment forces on us: $g_s\approx 2$, not $1$.

WHY the extra $g_s\approx 2$? A classical spinning charged loop predicts $g=1$. Measurement (and later the Dirac Equation) shows the electron's spin is twice as magnetic as that guess. This "$g=2$" is a genuinely quantum surprise — you cannot get it from a spinning ball.

The $z$-shadow of the magnet (the only part a vertical field feels): ${\mu }_z=-g_s\frac{e}{2m_e}{S}_z=-g_s\frac{e}{2m_e}{m}_s\hslash \approx \mp \frac{e\hslash }{2m_e}=\mp {\mu }_B,$ where ${\mu }_B=\frac{e\hslash }{2m_e}$ is the Bohr Magneton, nature's natural magnet unit.

PICTURE. The figure shows both electron states as little bar magnets: spin-up electron → magnet pointing down (minus sign), spin-down electron → magnet pointing up. Two orientations, opposite magnets.

---

Step 7 — The payoff: two spots in Stern–Gerlach

WHAT. Put the electron in a magnetic field that gets stronger as you go up — a field gradient $\frac{\partial B_z}{\partial z}$ (this symbol means "how fast $B_z$ changes as $z$ increases"). The energy of a magnet in a field is $U=-{\mu }_z{B}_z$, and force is minus the slope of energy: $F_z=-\frac{\partial U}{\partial z}={\mu }_z\frac{\partial B_z}{\partial z}=(\mp {\mu }_B)\frac{\partial B_z}{\partial z}.$

• ${\mu }_z=\pm {\mu }_B$ has only two values (Step 6).
• So $F_z$ has only two values — one pushes up, one pushes down.

WHY exactly two spots and not a smear. A classical bar magnet could point any angle, giving ${\mu }_z$ any value between $-{\mu }_B$ and $+{\mu }_B$ — a continuous smear on the screen. But quantum $m_s$ has only two rungs (Step 4), so ${\mu }_z$ jumps between two values. Discrete rungs → discrete spots. The ladder is literally imaged on the detector.

PICTURE. The beam enters, splits cleanly into an upper and a lower spot, nothing in between. The two forces are labelled by the two $m_s$ rungs.

---

Edge & degenerate cases (never leave a gap)

---

The one-picture summary

Everything above, compressed: one cyclic rule → a length, a ladder, a tilt, a magnet, two spots.

Recall Feynman retelling — the whole walkthrough in plain words

Start with a single arrow that measures "how much twirl." Nature hands us one matched set of rules — a cyclic triple, because space has no favourite direction. Those rules say you can't know which way the arrow points in all three directions at once — pin the up–down direction and the sideways ones blur. But one special quantity, the total length, stays perfectly sharp even while a shadow is known (the sideways twists from $S_x^2$ and $S_y^2$ cancel exactly), so a spin state carries two labels: a length and a shadow. That unkillable blur has a price: the arrow is always a little longer than its up–down shadow, so it can never stand perfectly straight. To find the exact length, use two "climbing" operators: stepping up then down equals the length minus the shadow-squared minus one $\hslash$ of shadow, and on the top rung the climb-up gives nothing — that little algebra hands you $s(s+1)$ for free. The climbing shifts the shadow by exactly one $\hslash$ each hop, and the climb must have a top and a bottom. For an electron there are exactly two rungs — a little up, a little down. Because it carries charge, each arrow acts like a tiny magnet, twice as magnetic as a classical spinning ball would be. Slide it through a magnet that's stronger up top: the two rungs feel two opposite pushes, so the beam splits into exactly two spots. Count the rungs, count the spots. (One honest warning: the arrow is a picture of the algebra — a true spin-$\frac{1}{2}$ object needs a $720^{\circ }$ turn to return, not $360^{\circ }$. Trust the algebra, use the arrow as a map.)

Recall

Why can the spin arrow never point straight up? ::: Its shadow $\frac{1}{2}\hslash$ is smaller than its length $\frac{\sqrt{3}}{2}\hslash$, because the sideways components can never both be zero. Where does the $s(s+1)$ come from? ::: From ${\vec{S}}^2=S_{-}{S}_{+}+S_z^2+\hslash S_z$ evaluated on the top rung (where $S_{+}$ gives zero), leaving ${\hslash }^{2}s(s+1)$. Why does $[{\vec{S}}^2,S_z]=0$? ::: The commutators $[S_x^2,S_z]$ and $[S_y^2,S_z]$ are equal and opposite, so they cancel exactly. What forces the ladder of $S_z$ to have a top and bottom rung? ::: A shadow can't exceed the arrow's length ($S_z^2\le {\vec{S}}^2$), so the climb up and down must terminate — giving symmetric rungs from $-s$ to $+s$. What is $[S_z,S_{+}]$ and what does it mean? ::: $[S_z,S_{+}]=+\hslash S_{+}$; it means $S_{+}$ raises the $S_z$-shadow by exactly one $\hslash$. Tilt angle of the $m_s=0$ rung of a spin-1 particle? ::: $\theta =90^{\circ }$ — the arrow lies flat in the equator, zero shadow but full length $\sqrt{2}\hslash$. Why discrete spots instead of a smear in Stern–Gerlach? ::: ${\mu }_z$ takes only the discrete values set by the two $m_s$ rungs, so the force takes only two values. What makes the electron's magnet twice as strong as classical? ::: The spin g-factor $g_s\approx 2$, an irreducibly quantum/relativistic result.

---

Connections

• Stern-Gerlach Experiment — Step 7 is this experiment.
• Orbital Angular Momentum — supplies the same cyclic commutator rules (Step 2).
• Quantum Numbers — $m_s$ is the fourth label built in Step 4.
• Bohr Magneton — the unit ${\mu }_B$ born in Step 6.
• Dirac Equation — explains the $g_s\approx 2$ of Step 6.
• Pauli Exclusion Principle, Zeeman Effect, Fine Structure — downstream consequences of this two-rung structure.

Concept Map