Working Through Taylor's Classical Mechanics
I've been helping grad students and upper-level undergrads work through John R. Taylor's Classical Mechanics for about eight years now, mostly at office hours or on forums when people get stuck. The book is solid. It's not perfect, but it's one of the better bridges between introductory physics and graduate-level theory. The table of contents runs roughly like this: Conservation laws and simple harmonic motion early on, then a deep dive into Newtonian mechanics, followed by oscillations and waves, nonlinear dynamics and chaos, the calculus of variations and Lagrangian mechanics, two-body central force problems, rigid body rotation, non-inertial frames, special relativity, and finally a section on Hamiltonian mechanics. The later chapters shift tone noticeably. Taylor moves from computational, example-driven explanations to more abstract derivations. Students often don't realize they're in a different book by chapter 7. I want to address something most people miss about this text. The Lagrangian chapter isn't just a reformatting of Newtonian mechanics with fancier notation. Taylor treats generalized coordinates as genuinely independent variables, and that philosophical shift is where students hit their first wall. I watched a student spend three weeks trying to force a pendulum problem into Cartesian coordinates because she didn't trust that swapping to angle theta would actually simplify things. It would have taken twenty minutes if she'd just committed to the coordinate change. The Lagrangian formalism rewards that commitment. It punishes hesitation.
Another counter-intuitive thing: the calculus of variations section is actually the most practically useful part of the book for anyone doing computational physics later. The Euler-Lagrange equations that come out of it are structurally identical to the numerical routines you'll write for trajectory optimization. If you understand how Taylor derives those equations from first principles rather than just memorizing the form d/dt(dL/dq_dot) = dL/dq, you'll save yourself real headaches when you get to simulation work. I learned that the hard way. Early in my graduate studies I tried to code a simple orbital propagator using a brute-force Newtonian approach and it failed on eccentric orbits because I hadn't internalized the constraint structure that variational methods make explicit. Now, the book has weaknesses that nobody talks about enough. The treatment of chaos in chapter 6 is competent but shallow. The Poincaré sections are introduced without enough geometric intuition, and the bifurcation analysis feels tacked on rather than integrated. If you're serious about nonlinear dynamics, you'll need supplemental material. I recommend pairing it with Strogatz for the conceptual side and Tabor for the deeper mathematical treatment. Taylor gets you to the doorway. He doesn't walk you through it. The Hamiltonian chapter is similarly uneven. The canonical transformation material is glossed over, and the generating function formalism appears almost as an afterthought. This is where the book stops being self-contained. You'll find yourself going back and forth between Taylor and Goldstein, which is fine if you have the time, but it means this isn't a single-volume solution for anyone preparing for comprehensive exams.
If you're working through it on your own, here's a practical note about the problem sets. Taylor writes problems that range from straightforward plug-and-chug to genuinely challenging derivations that require insight you haven't been given yet. The hardest problems in chapters 3 and 4 often need you to combine conservation laws in non-obvious ways. I usually tell people to attempt each problem for at least forty-five minutes before looking at any solution. The learning happens in the struggle, not in the checking. That's my standard recommendation across every mechanics text I've encountered. The book is available through most university libraries and commercial retailers. The third edition from University Science Books is the one most people use. It's expensive new, around ninety dollars, but used copies in decent condition run much lower. Don't buy the first edition if you can help it. The errata in that one are significant, particularly in the relativity chapter where some problem parameters were changed without updating the solutions. There's no official solution manual published by the author, which frustrates self-learners. What exists online is fragmentary and sometimes wrong. I've seen posted solutions for Chapter 7 problems that contain sign errors in the angular momentum terms that propagate through three subsequent lines. Always verify against the derivation in the text itself before trusting an external answer key.
The relativity chapter deserves a specific mention because it's handled well compared to most undergrad texts. Taylor doesn't shy away from four-vectors or the stress-energy tensor at an introductory level. But he doesn't develop the differential geometry machinery that appears in a full GR course, so don't expect this to prepare you for Wald or Carroll. It prepares you to not be completely lost when you encounter those books later. For people asking about whether to use this as a primary text or a supplement, my answer depends entirely on what you're coming from. If you've finished Kleppner and Kolenkow, Taylor is a natural next step. If you're coming directly from introductory physics at the Halliday Resnick level, you'll find the pace aggressive. The first fifty pages assume comfort with vector calculus and differential equations that many students haven't fully developed yet. I've seen students try to power through and end up confused not because the physics is hard but because the mathematical prereqs weren't solid. Spend a week refreshing your multivariable calculus before diving in. It'll save you weeks of frustration later.
Get the Full Details
