Getting Problems Right Isn't the Same as Understanding Physics: The Origins and Boundaries Behind Formulas

Getting a problem right shows that you found a workable solution under the given conditions; it cannot by itself prove that you understand the physics you used. Change the diagram, change one condition, and what truly tests understanding is often not calculation but whether you can clearly explain: what question is the formula answering? Where does it come from? Which conditions, once changed, force the reasoning to start over?

Equal signs in physics are not all the same kind of thing. Some are experimental laws or fundamental principles, some are definitions, some are constraints derived from geometry, and some hold only within approximate models. Recognizing a formula's identity before discussing fluent use makes it harder to mistake a memorized conclusion for a universal law.

Electricity and Magnetism: A Unified Theory Changed How Knowledge Is Organized

The common classification of fundamental interactions includes the strong interaction, the weak interaction, the electromagnetic interaction, and gravitation. The electric and magnetic forces of everyday life both belong to the electromagnetic interaction; they are not two unrelated fundamental forces.

In classical electromagnetism, the electric field and the magnetic field are different parts of the same electromagnetic field. Special relativity further tells us that when an observer's inertial reference frame changes, the components of the electric and magnetic fields mix with each other. An isolated point charge has only an electrostatic field in its own rest frame; in a relatively moving frame, the charge is in motion, and the field description generally contains both an electric and a magnetic field. The observer has changed, but the physical interaction has not switched to another one.

This does not mean that electric and magnetic fields can be mixed and calculated together in any problem. On the contrary, when using formulas one must specify the reference frame, and fields calculated separately must also be understood as combining. The value of a unified theory is that it returns phenomena once memorized separately to a single structure.

Potential Energy: Why We Don't Have to Calculate Work Along the Entire Path

If a force is conservative, the work it does from A to B depends only on the starting point and the endpoint, not on which path is taken in between. It is precisely because of this property that we can rewrite the problem of work done along a path as a comparison of the potential energy of two states:

$\Delta U=U_B-U_A=-W_{\mathrm{cons},A\to B}$

Potential energy is not a hidden "store of motion" that an object has at a certain position, but a quantity describing the state of an interacting system. Its zero point can be chosen arbitrarily for convenience; adding the same constant overall does not change potential-energy differences, nor does it change the work done by a conservative force. Non-conservative forces such as friction generally cannot be described by a globally single-valued potential energy.

Near the ground, when the vertical height difference is very small relative to Earth's radius,$g$the variation of can be neglected, so we have$U=mgh$. If we discuss the Newtonian gravitational field produced by a point mass$M$, the exact form of the potential energy is$U=-GMm/r+C$. Therefore$mgh$is not the definition of potential energy, but a result in an approximately uniform gravitational field.

Potential energy can also directly tell us where the force points. In the one-dimensional case, a conservative force satisfies

$F_x=-\frac{dU}{dx}$

In three dimensions this is written$\mathbf F=-\nabla U$. The negative sign indicates that the force points in the direction of decreasing potential energy. The formula is not merely a convenient substitute for calculation: the slope of the potential-energy curve itself describes the force.

Why Simple Harmonic Motion Is So Common

The restoring force of a spring is often written as$F=-kx$, but a great many small oscillations in nature do not require an actual ideal spring. Take an ordinary, smooth, non-degenerate stable equilibrium position$x_0$as an example, and let$\xi=x-x_0$. At equilibrium the first derivative of the potential energy is zero; if$U''(x_0)>0$, then within a sufficiently small range of displacement, the Taylor expansion is

$U(x_0+\xi)=U(x_0)+\frac12U''(x_0)\xi^2+O(\xi^3)$

Considering only the conservative restoring force, temporarily ignoring damping and external driving, and neglecting higher-order small quantities, the force is approximately$F\approx-U''(x_0)\xi$. Let$k=U''(x_0)$; Newton's second law gives

$m\ddot\xi=-k\xi,\qquad \omega_0=\sqrt{\frac{k}{m}}$

This is simple harmonic motion. It is universal not because springs are everywhere in the world, but because near many stable equilibria a smooth potential energy is approximately a parabola over a small range. After the displacement becomes larger, higher-order terms may no longer be negligible; if the equilibrium point is degenerate, with$U''(x_0)=0$, this conclusion also cannot be applied directly. The range in which a model holds is itself part of the formula.

The Same "2" May Also Come from Completely Different Places

In an inertial reference frame, Newton's second law relates the net force to acceleration:$\sum\mathbf F=m\mathbf a$. And in circular motion,$v=\omega r$is the kinematic relation between speed and angular velocity; it itself does not explain why an object turns. Uniform circular motion also has an acceleration directed toward the center. Looking over a very short time interval, the direction of the velocity turns through a small angle$\Delta\theta$, giving$|\Delta\mathbf v|\approx v\Delta\theta$; and since$\Delta\theta/\Delta t\approx v/r$, taking the limit yields$a_c=v^2/r$. The magnitude of the velocity is constant while its direction changes, so there is still acceleration.

The "2" in a movable pulley is also worth questioning. Assume the rope is inextensible, its mass is negligible, and the movable pulley is supported by two parallel vertical rope segments. When the pulley moves downward by$x$, each of the two load-bearing rope segments lengthens by$x$; to keep the total rope length unchanged, the free end must move upward by$2x$. The factor of two in the velocity relation comes from the rope-length constraint.

In force analysis,$2T$may appear again: the two rope segments each pull the pulley upward with tension$T$, so the net force is$2T$. This is the vector sum of two forces, and it is not the same thing as the "2" in the displacement ratio. If the total supported weight is$mg$and the system is in equilibrium, then one can write$2T=mg$; if the system has acceleration, one must return to$\sum F=ma$. When the supporting rope segments are no longer parallel or the pulley-block structure changes, the rope-length constraint also changes.

When you see a coefficient, it is best to ask where it comes from: from a measured law, a definition, a geometric constraint, or the combination of forces? Different sources come with different conditions; when the conditions change, you cannot simply copy it just because the formula looks similar.

Feynman and Landau: Two Reading Rhythms

Reading Feynman's Lectures on Physics, one often encounters an approach that unfolds from concrete phenomena and problems; reading Landau and Lifshitz's textbooks, one often feels that the expression is more concise and the theoretical structure more concentrated. This is a summary of the reading experience, not a ranking of the two physicists, nor does it mean that every volume and every chapter of theirs follows a single style.

Once a theory matures, it can often be compressed into a short and elegant logical chain; but when learning it for the first time, the reader needs to see how the problem arises, why the concept is defined this way, and at which step the approximation is introduced. The two orders address different difficulties: course textbooks and exercises help build terminology, models, and technique; Feynman is suitable when you "can calculate but cannot explain clearly" and want to try another route of questioning; once the foundations are solid, reading Landau can show how many conclusions are organized into a more compact theoretical structure.

The Feynman Lectures need not be read from the first page to the last to count as having been read. You can first choose a topic you are studying and that has left you with questions, and read with specific questions in mind; after closing the book, try to retell the key reasoning in your own words. The official online edition also lists the rough scope of the three volumes: Volume I mainly deals with mechanics, radiation, and heat, Volume II mainly with electromagnetism and matter, and Volume III with quantum mechanics. It can supplement angles for understanding, but it does not replace systematic practice and a complete course path.

The next time you encounter a formula, you can ask in turn: What quantity does it describe? Is it a law, a definition, a corollary, or a constraint? What assumptions were used in the derivation? Where do the symbols and coefficients come from? If you change a certain condition, which step will fail first? Only when you can answer these questions does the formula truly become your own tool.

References: [CERN: The Standard Model and the four fundamental interactions](https://home.cern/science/physics/standard-model/); [MIT OpenCourseWare: Relativistic transformation of electromagnetic fields](https://ocw.mit.edu/courses/8-033-introduction-to-relativity-and-spacetime-physics-fall-2024/mit8_033_f24_lec11.pdf); [OpenStax: Conservative forces and potential energy](https://openstax.org/books/university-physics-volume-1/pages/8-2-conservative-and-non-conservative-forces); [MIT OpenCourseWare: Small oscillations near stable equilibrium](https://ocw.mit.edu/courses/8-01sc-classical-mechanics-fall-2016/mit8_01scs22_chapter23.pdf); [Caltech: The Feynman Lectures on Physics online edition](https://www.feynmanlectures.caltech.edu/).