And Sherbert Introduction To Real Analysis

Working Through Bartle & Sherbert Without Losing Your Mind

I picked up the third edition of this textbook when I was a graduate student because that's what everyone told me to pick up. Three years later I still use it as a reference. It's not a fun read. It doesn't try to be. That's kind of the point. The book is structured around building real analysis from the ground up. It starts with the real number system and completeness, moves through sequences and series, then topology of the reals, continuity, differentiation, and the Riemann integral. The later chapters cover metric spaces and a proof of the fundamental theorem of calculus in that broader context. That's the table of contents at a glance. What matters more is how the problems are set up.

Getting And Sherbert Introduction To Real Analysis

The book is widely available in used copies, which makes sense since the content doesn't change meaningfully between editions. The third and fourth editions are nearly identical in structure. If you're looking for a PDF, there are plenty of places that host it, but I won't link to anything sketchy. Buy a used hardcover for fifteen dollars or grab a legitimate digital copy if you have the budget. Either way works. The content is the same. Here's what most people miss about this book. The exercises are where the actual learning happens, and they're not optional fillers. Chapter 3 alone has problems that will make you sit on a single question for forty-five minutes before you figure out you were approaching it backwards. I spent two full days stuck on Problem 14 in Section 3.2 because I kept trying to construct a sequence when the problem wanted me to use the supremum property directly. The hint in the back doesn't help unless you've already spent time wrestling with it. That's intentional. The book assumes you'll struggle and builds the rigor through that struggle. Another thing nobody tells you upfront: the notation in the first five chapters is deliberately conservative. It looks dense on page one because Bartle and Sherbert refuse to hand-wave anything. They define everything before using it. This means Chapter 1 reads slowly, but by Chapter 4 you'll be moving fast because the machinery is already built. Don't skim the first chapter. The completeness axiom and the archimedean property show up in proofs later without fanfare, and if you didn't internalize them early you'll be guessing at steps that should be obvious.

I ran into a specific issue working through the metric space chapter. The textbook proves that every convergent sequence is Cauchy and then states the converse for complete spaces. But it doesn't explicitly walk through constructing a completion of a general metric space until the exercises. I hit a wall when a professor asked me to demonstrate that the rationals are incomplete using an explicit Cauchy sequence with no rational limit. The standard answer is the sequence approximating sqrt(2), but the book's exercise variant uses a recursive sequence defined by a_n+1 = a_n/2 + 1/a_n starting from a_1 = 1. It converges to sqrt(2) and every term is rational, but proving the limit isn't rational requires the standard irrationality proof, which the book leaves as a separate exercise. I ended up just writing out the full proof on scratch paper and verifying each step against the exercise number. Took about twenty minutes once I stopped second-guessing myself. The Riemann integral chapter is the heaviest part of the book for most students. The definition using upper and lower sums is clean, but the integrability criteria and the relationship between continuity and integrability take several sections to develop properly. Chapter 6 spends real time on the connection between differentiation and integration, and the proof of the fundamental theorem is split across two sections because the authors want to establish uniform continuity first. This is correct pedagogy but it feels slow if you're coming in with some calculus background. Push through it. The payoff is in the exercises at the end of the chapter, where you get to actually work with improper integrals and see where the basic theorem breaks down. One counter-intuitive insight: the book treats sequences and series before topology, which is the opposite of how many modern texts organize things. Some instructors prefer the topology-first approach because it makes convergence definitions more natural. Bartle and Sherbert do it the other way around, and honestly it works better for self-study. You learn to manipulate sequences concretely before abstracting the notion of open sets. The tradeoff is that when they introduce open and closed sets in Chapter 4, you're already familiar with convergence and the topological definitions feel like a reformulation rather than a foundation. That's fine. It's still rigorous.

The book has genuine limitations. It doesn't cover Lebesgue integration at all. If your program requires measure theory, you'll need a second text, probably Rudin or Folland. The exposition is also fairly traditional and dry, which means it won't hold your attention on a bad day. There are no sidebar discussions or motivation paragraphs. It's definitions, theorems, proofs, exercises. For some people that's exactly what they want. For others it's suffocating. I'd recommend pairing it with a video lecture series if you're studying alone. There are decent MIT OpenCourseWare lectures that follow this book's structure closely. Also, the problem difficulty is uneven. Some sections have fifteen straightforward applications of a theorem followed by one problem that requires a construction you haven't seen before. The earlier chapters are gentler. By Chapter 7 the gap between routine and hard problems widens noticeably. Don't abandon a problem set after one failed attempt. The kind of thinking this book develops doesn't come from. It comes from sitting with a page of scratch paper and watching a proof reveal itself slowly. For anyone actually working through this book, my practical advice is to keep a separate notebook for definitions and theorems written in your own words, not copied verbatim. When you can reconstruct a statement without looking at the book, you actually understand it. The exercises are non-negotiable. Skip them and the reading is mostly decorative. The book is about thirty dollars new but worth every penny of a used copy. Read it straight through once, then go back and do every problem in the chapters you find weakest. That's the method that works.