Working Through Folland's Real Analysis: A Practical Guide
Gerald Folland's Real Analysis: Modern Techniques and Their Applications is the book most graduate students end up using when they need a rigorous treatment of measure theory, integration, and functional analysis. It covers Lebesgue integration, Fourier analysis, distributions, and Banach space theory in a compact format that many find both efficient and frustrating at the same time. The second edition runs about 440 pages and assumes you already know basic point-set topology and can handle epsilon-delta arguments without hand-holding. If you come from a physics or engineering background, expect the first few chapters to move slowly while you adjust to the level of abstraction. The notation is standard but not forgiving.
Real Analysis By Folland
The text is organized into eight chapters plus appendices. Chapter 1 covers measure theory from scratch. Chapter 2 treats the Lebesgue integral and convergence theorems. Chapter 3 extends to signed measures and the Radon-Nikodym theorem. Chapter 4 handles $L^p$ spaces. Chapter 5 moves to functional analysis basics. Chapter 6 covers Fourier transforms. Chapter 7 discusses distributions. Chapter 8 treats Haar measure and abstract harmonic analysis. What makes this book useful is its emphasis on applications. Unlike pure measure theory texts, Folland connects the material to partial differential equations, probability, and signal processing. The exercises range from straightforward verification to problems that require genuine insight. Most students spend roughly two to three weeks per chapter working through the assignments if they are doing it seriously. I remember hitting a wall with the construction of the Lebesgue measure in Chapter 1 when I was trying to reconcile the outer measure approach with the Carathéodory extension theorem. The proof that the Lebesgue measure is complete and translation invariant is elegant but skims over some details about why the completion matters for applications. I spent about four hours reconstructing the missing steps before I could move forward. The workaround was to cross-reference with Halmos' Measure Theory for the foundational pieces and then return to Folland for the applications.
How to Actually Use This Book
Don't read it cover to cover. Work through the chapters selectively based on your needs. If you are preparing for qualifying exams, focus on Chapters 1 through 5. If you need distributions and PDE applications, skip ahead to Chapters 6 and 7 after building the measure theory foundation. The exercises are where the actual learning happens. Folland does not provide answers, so you will need to verify your work through discussion or by checking against known results. I typically spend one to two hours per problem set, sometimes longer for the challenging ones. The problems on uniform integrability and the Vitali convergence theorem are particularly good for building intuition. One thing beginners miss is that Folland assumes comfort with metric spaces and topological vector spaces. If you struggle with these prerequisites, spend time reviewing them before diving in. The appendices on complex analysis and topology are helpful but not sufficient on their own. I recommend having Rudin's Real and Complex Analysis nearby for supplementary explanations.
Get the Full Details
Common Pitfalls and Where the Book Falls Short
Folland's treatment of the Lebesgue integral is concise but dense. The jump from measurable functions to integrable functions can feel abrupt. Students often miss the distinction between $L^1$ convergence and almost everywhere convergence until they work through enough counterexamples. The book does not spend much time on pathological cases like the Cantor function or the Dirichlet function, which are important for developing intuition. Another limitation is that Folland does not cover stochastic integration or martingale theory in depth. If you need those tools for probability applications, you will need additional sources. The book also assumes familiarity with the Hahn-Banach theorem and the open mapping theorem, which are stated but not proved in detail. The chapter on distributions is thorough but briefly treated compared to dedicated texts like Strichartz or Schwartz. For someone needing a deeper understanding of tempered distributions and Sobolev spaces, this section will feel insufficient. I found myself referring to Evans' Partial Differential Equations for the analytical details and returning to Folland for the measure-theoretic foundations.
Practical Tips for Working Through the Material
Start with Chapter 1 and work through it methodically. Do not skip the exercises on sigma-algebras and measurable sets. These form the foundation for everything else. Spend at least a week on this chapter even if the material feels familiar from earlier coursework. For Chapter 2, focus on understanding the monotone convergence theorem, Fatou's lemma, and the dominated convergence theorem. These three results account for most of the proofs you will encounter in subsequent chapters. The exercises on simple functions and approximation are particularly valuable for building computational intuition. When you reach Chapter 4 on $L^p$ spaces, pay attention to the duality relationships and the Riesz representation theorem. These concepts appear throughout functional analysis and have direct applications to optimization and PDE theory. The chapter on Fourier transforms in Chapter 6 is more accessible than the distribution theory in Chapter 7, so some students prefer to work through Chapter 6 first for a change of pace.
If you are self-studying, plan for approximately twelve to sixteen weeks to complete the core material. This estimate assumes you are working through the exercises thoroughly and consulting additional sources when needed. Rushing through the book without doing the problems will leave significant gaps in your understanding.
When to Look Elsewhere
If your primary interest is probability theory, consider using Durrett or Billingsley instead. Folland's treatment of measure theory is general-purpose and not tailored to stochastic processes. For someone focused on harmonic analysis, Stein's Harmonic Analysis provides more depth and specialized techniques. Similarly, if you need a comprehensive treatment of Sobolev spaces and their applications to PDE, Adams' Sobolev Spaces or McLean's Strongly Elliptic Systems will serve you better than Folland's brief coverage in Chapter 7. The book remains an excellent reference and primary text for a first graduate course in real analysis. Its strength lies in the balance between theoretical rigor and practical application. Most students who work through it carefully find that the investment pays off in subsequent courses and research.
I typically keep a copy on my desk for quick reference when I need to verify a result about measure theory or functional analysis. The notation and conventions are standard across the field, so familiarity with Folland transfers easily to other texts and papers. The exercises remain useful even years after completing the material for qualification preparation or research reference.
Supplementary Resources
Several textbooks complement Folland well. Royden's Real Analysis provides more detailed proofs and additional examples. Rudin's Real and Complex Analysis offers a different perspective with greater emphasis on complex methods. For anyone working through Folland independently, having at least one of these as a secondary reference is advisable. Online lecture notes from various universities also fill gaps in the exposition. The MIT OpenCourseWare materials on measure theory and integration provide video lectures that clarify some of the more abstract arguments. These resources are freely available and can save considerable time when struggling with a particular concept. The appendix on the Riesz representation theorem is particularly useful for understanding the connection between measure theory and functional analysis. Students who skip this material often find themselves confused when the theorem is invoked in later chapters without full justification.
Working through Folland requires patience and consistent effort. The material is demanding but rewarding for those willing to invest the time. Most students who persist find that the depth of understanding they gain justifies the initial difficulty. The book remains a standard reference in graduate mathematics programs worldwide for good reason.
