I have read some of the books suggested in What are the prerequisites to Jech's Set theory text?, so I have some beginning experience with transfinite recursion, ordinals, cardinals, order types, and the axioms in ZFC.

My question is this: How much formal logic should I know before reading Jech's Set Theory (In particular, how much model theory should I know; I see that in the table of contents of Jech's Set Theory there is a section with "Models of Set Theory")? Will the text "Mathematical Logic" by Wilfrid Hodges and Ian Chiswell suffice? Should I first read a book solely about model theory first as well? What are your suggestions? Thanks in advance!


At the very least, you should understand the statements of the completeness theorem, the incompleteness theorems, and the Lowenheim-Skolem theorem. This involves all of the basic definitions of formal logic and model theory; models, theories, satisfaction, elementarity, elementary substructures, Skolem functions, and more. Of course, as it is in any area of math, it's easy to trick yourself into thinking that you understand these things, when you really don't. You can get a more thorough understanding by reading the proofs, which make up the essentials of many logic texts.

I don't know the Chiswell-Hodges book, but glancing at the table of contents, it looks like it doesn't quite cover all of the prerequisites. (It may however be a very good book for what it does cover.) Enderton's book (on logic) is considered a classic, and it has approximately the right content, including the relevant definitions from model theory.

General model theory is a good thing to have an understanding of when going into set theory, but don't think it'll make the "Models of Set Theory" chapter of Jech a breeze; models of set theory are pretty counter-intuitive at first glance, and require a lot of thinking about on their own.

I agree with Arthur Fischer in the question you linked to; Jech is not a great book for a budding set theorist to learn the field from. It's more of a reference than a text. However, I think the above remarks apply to essentially any introduction to modern set theory.

  • 1
    $\begingroup$ Except its lack of exercises, I find Halbeisen's book to be very well written. $\endgroup$ – Asaf Karagila Mar 10 '15 at 14:32
  • $\begingroup$ @Paul McKenney Thank you for the helpful post. Also, when you say that Jech's text is a reference, is it similar in style with Rudin's texts, or is it even less motivated and even more of a reference in flavor? Also, thanks to your suggestion, I've decided to read Enderton's "Intro to Math. Logic" book first before heading in any direction toward higher level set theory. $\endgroup$ – Metric Mar 10 '15 at 20:12
  • $\begingroup$ @AsafKaragila Are you referring to Halbeisen's Combinatorial Set Theory? That's the only book I see from this author on amazon. Is this book near the level of Jech, or more in the middle of elementary and advanced? Thanks. $\endgroup$ – Metric Mar 10 '15 at 20:14
  • $\begingroup$ @Nectric: Yes, and there is a free copy on the author's homepage. It's fairly advance, although Jech covers way way more topics. But this book should help you get acclimated to set theory. $\endgroup$ – Asaf Karagila Mar 10 '15 at 20:35
  • 1
    $\begingroup$ @Nectric: I'd say Jech's book is actually fairly self-contained; but its scope is far too large for it to be anything other than a reference. Plus, its treatment of certain topics (forcing in particular) is not what I would want for a beginner. $\endgroup$ – Paul McKenney Mar 10 '15 at 22:11

I don't have any formal background in math, logic or philosophy. I come from law. Recently, I got interested in foundations of mathematics and ultimately in philosophy of set theory. I picked up some logic from the book "Computability and Logic" by Boolos and Jeffrey. I think one advantage of taking Jech as the primary source of self-study of the mathematical aspects of Set Theory is that you gradually get used to his style of exposition and notation. That makes things smoother. Of course, it's a very comprehensive and rather heavy book. But it's not necessary to start a chapter and master it end to end. We can decide to look at all the results but study only a few proofs or parts of some proofs on first reading. We can always come back to that chapter later. This I think basically combines elements of using it as a text as well as reference.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.