Real Analysis By H L Royden

Working Through Royden's Real Analysis Textbook

Most graduate students in mathematics eventually run into Real Analysis By H L Royden whether they choose it or land on it by default. The book is still widely used in PhD qualifying exams and first-year coursework across a lot of programs. That reputation comes from decades of use rather than any careful marketing. The content is solid but the writing style has certain quirks that trip people up if you are not expecting them. The standard edition covers measure theory, Lebesgue integration, function spaces, and then branches into topics like Fourier analysis and metric spaces depending on which version you are using. The fourth edition added more material on differentiation and integration on Rn, which is useful if your program expects you to handle higher-dimensional cases without a separate text. Earlier editions skip around more and sometimes assume you already know things they never actually prove. I found the chapter on the Lebesgue integral to be the most carefully developed section. The treatment of convergence theorems follows a logical sequence, and Royden does not waste pages repeating definitions you already need. The sections on Lp spaces are also worth reading straight through because they build in a way that matches how the material actually shows up in research. Other chapters, especially the ones on outer measure and the construction of the integral, are dense in a different way. You will reread paragraphs three or four times before the argument clicks. That is normal. It is not a sign that you are doing something wrong.

How I Actually Used This Book in Practice

When I was preparing for my own quals, I used Royden alongside other sources rather than treating it as the single reference. The problem with relying on it alone is that some proofs are sketchy in ways that feel intentional but are usually just lazy. I remember spending an afternoon stuck on the proof that every measurable set can be approximated by open sets from the outside within any epsilon of measure. The book states the result cleanly but the derivation in the main text jumps over the construction of the covering intervals in a way that made me double-check my understanding of countable subadditivity. I ended up working through the argument myself on a blank sheet, reconstructing the covering step by step, and only then realizing the book's shortcut was valid but only because of a property it had introduced two chapters earlier without much fanfare. That experience taught me to keep a notebook handy and verify any proof that feels too compact to be complete. Another issue I ran into involved the treatment of Fubini's theorem. The statement in Royden is correct but the hypotheses require careful reading. I once applied the theorem to a product space without checking sigma-finiteness properly, and the resulting integral gave a nonsensical answer. The fix was straightforward once I went back and verified the measure spaces involved satisfied the necessary conditions, but it took about forty-five minutes to trace the error. Writing that down as a reminder saved me later when I needed to invoke Fubini without second-guessing myself.

Common Pitfalls and What to Watch For

One thing beginners consistently miss is the distinction between pointwise convergence and convergence in measure. Royden introduces both and then uses them interchangeably in later exercises, which is technically fine but confusing if you have not internalized the difference. Convergence in measure does not imply pointwise convergence everywhere. It implies pointwise convergence almost everywhere along a subsequence, and that subtle shift matters when you are constructing counterexamples or working through Egorov's theorem. Another trap involves the Riesz representation theorem as stated in the book. The version for continuous linear functionals on Cc(X) assumes local compactness of the underlying space. If you are working in a general metric space without that assumption, the theorem does not hold in the same form. I saw students apply it blindly in a problem set once and write proofs that looked elegant but were based on a false premise. The fix is to check the hypothesis list before reaching for the theorem.

Supplementary Resources Worth Knowing

If you are struggling with a particular chapter, Rudin's Principles of Mathematical Analysis covers some overlapping ground with a different emphasis. Stein and Shakarchi's Real Analysis is also useful, especially for readers who want more motivation behind the constructions rather than just the constructions themselves. For measure theory specifically, Halmes's Measures and Integrals is leaner and forces you to fill in more gaps, which is either exactly what you need or exactly what you do not need depending on your current level. Online lecture notes from various universities also help. MIT OpenCourseWare and similar repositories post problem sets and solutions that align closely with the Royden chapters. Working through those problems after reading the relevant section tends to cement the material faster than rereading the text.

Limitations of the Text

The book is not universally excellent. The exposition on differentiation of monotone functions is thin compared to what a student might need for a thorough understanding. The treatment of Hausdorff dimension and fractal-related topics is also minimal. If your program expects you to engage with geometric measure theory or advanced real-variable techniques beyond the standard curriculum, you will need additional references. The exercises range from routine computational drills to fairly challenging proof problems, and the difficulty gap between adjacent problems can be steep. Do not expect every exercise to follow naturally from the preceding ones. Another limitation is that later editions updated the notation and organization without fully reconciling cross-references, so you may find yourself flipping between sections to verify that a result stated in chapter three is actually the same result mentioned in chapter seven under slightly different labels. It is a minor annoyance but one that adds up over time. The book remains a staple for good reasons. The coverage is comprehensive, the results are correctly stated, and the progression from measure theory to integration to function spaces is coherent. You just need to approach it with the expectation that you will occasionally have to supply the missing steps yourself. That is true of most graduate-level mathematics texts, but Royden demands it more consistently than many others.