Hacker News Reader

Comment by dmix | Hacker News Reader

I'm currently reading up on formal logic in order to fully understand Probability Theory by E. T. Jaynes [1] and I've found it difficult to find a good logic book. There's a thousand on Amazon and it's hard to tell which ones are written a century ago w/ re-released dates with old notation or missing some new ideas. An "original publishing" data feature on Amazon would be a godsend.

I've seen a few well reviewed Symbolic logic books but not sure if that fits the criteria. Mostly digging into boolean algebra atm.

[1] https://www.amazon.com/Probability-Theory-Science-T-Jaynes/d...

View on HN

I second the recommendation of checking the Teach Yourself Logic guides by Peter Smith, but I will give some more specific recommendations:

- Greg Restall - Logic (and he also gives some recommendations of his own that are worth checking out). If you use this, check the errata in Restall's website.

- Richard Jeffrey - Formal Logic: Its Scope and Limits (Smith has a textbook that is modelled largely after this, but Jeffrey's book is more to the point and fun to follow along)

- Daniel Velleman - How to Prove It, which has tons of good exercises for practicing symbolig logic in the context of the construction of proofs, and introduces some mathematics along the way.

Thanks, I wasn't familiar with the "Teach Yourself" series, that's going to help me on my path learning math/stats/probability theory.

I've used forall x (calgary remix), both the winter 2018 and the spring 2020 editions, and I found it quite good. It teaches truth functional and first order logic, and it's probably accessible to most high school graduates. I've seen it taught with good results to students of different academic backgrounds as a 'math' course for general education. I've linked it below and it's free (as in beer and as in freedom) so it's definitely worth glancing at, along with the rest of the open logic project.

http://forallx.openlogicproject.org/

https://www.logicmatters.net/tyl/

As a mathematician, I find this surprising - formal logic and probability theory don't usually have much to do with each other. Is the issue that you have trouble with formal mathematical notation?

This reminds me of my wish to see a unified theory (and notation) that encompasses formal logic and probability. On their own, each is lacking. A unified system, for example, in which we have all the tools of first order logic and Bayesian probability would be very powerful.

If you look at the probability book I mentioned it uses an algebraic boolean logic notation and basic set theory to build a formal theory of inference. The first chapter is an condensed introduction to the formal logic and notation used - which is what I was struggling with following some of the more complex equations.

Having read half a book on logic and subsequently learned basic set theory has already helped read the first portion of the probability book. But then I also got really into formal logic, I found it really fascinating as a programmer and I think every person should learn it (with plenty of applications to regular life), so I decided to take a deep dive into it. The venn diagrams visualizations are what helped me the most.

I think one of things that held me back initially was my background as a programmer, it made reading the logic set notation challenging, ie the plus signs meaning disjunctions and primes negation conjunctions.

Ah, I see. That's interesting, I'll see if I can find a bootleg copy and have a read =). Logic is a beautiful, deep subject - all the best with your studies!

Here's the first 3x chapters, which I presume was for some course:

https://bayes.wustl.edu/etj/prob/book.pdf

I've heard amazing things about it and it hasn't yet disappointed (the little I've read). Worth the $60 I spent on Abebooks for it (used) but the full copy is also on ThePirateBay if you want to see a longer preview.

Try this one:

https://www.cambridge.org/core/books/handbook-of-practical-l...

Reply on news.ycombinator.com