Timeline for Why would one number theorems, propositions and lemmas separately?
Current License: CC BY-SA 4.0
20 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 10, 2019 at 14:05 | vote | accept | Jakub Konieczny | ||
| Aug 7, 2019 at 2:07 | comment | added | HJRW | Wasn’t method 2 used by Euclid? | |
| Aug 6, 2019 at 2:43 | comment | added | Jonny Evans | There's also the fact that, using method 2, Theorem n is the n-th theorem in the text, which is the conventional bijection used by humans everywhere for counting. While a fantastic idea, method 1 is not an obvious thing to do (I had never thought of it until a collaborator suggested it). | |
| Aug 5, 2019 at 22:32 | comment | added | F Zaldivar | According to the review MR0125113: " One ...criticism applies to this book as well as to a large part of contemporary mathematical production: the various statements are called by different names, such as Lemma, Theorem, Proposition, Corollary; the first three are numbered independently of each other, while the numbers assigned to corollaries are functions of several variables; in addition, numbered formulae have their own separate numeration. The strain placed on the reader by this partial ordering is obvious, but apparently readers seek vengeance on other readers when they turn into authors" | |
| Aug 5, 2019 at 21:22 | answer | added | Timothy Chow | timeline score: 4 | |
| Aug 4, 2019 at 3:05 | comment | added | Mike Shulman | You can make the numbering totally non-volatile by using fractional numbers when adding something in the middle. Or, as early programmers in some languages did with line numbers, start out by numbering things with multiples of 10 so that there are plenty of integers available in between. (-: | |
| Aug 4, 2019 at 2:34 | comment | added | Iosif Pinelis | Some journals require using method 2. | |
| Aug 3, 2019 at 21:46 | comment | added | paul garrett | Yes, echoing @JoeSilverman's comment, in a pre-computer setting, having the numbering be only locally volatile for edits is important. | |
| Aug 3, 2019 at 20:16 | comment | added | Joe Silverman | Try writing a book (or long article) without a computer, and certainly without any way of auto-cross-referencing. Then you realize that you want to add a lemma, or an example. Using Method 2, the required changes are much less than if you are using Method 1. Another way to accomplish this, as is done for example in Hartshorne, is to label using A.B.C, where A is the Chapter number (in upper case roman), B is the section number, and C is the item number in that section. LaTeX, of course, has an easy way to number in that fashion. | |
| Aug 3, 2019 at 14:21 | comment | added | Lucian | You can also number them according to the chapters and paragraphs where they occur; i.e., the various theorems occurring in the third chapter are T3.1, T3.2, T3.3, etc.; similarly for propositions and lemmas, numbered as P3.1, P3.2, P3.3, etc., or L3.1, L3.2, L3.3, etc., respectively. | |
| Aug 3, 2019 at 11:40 | comment | added | J Fabian Meier | What about hashing the theorem content? Something like theorem 1987568324010. | |
| Aug 3, 2019 at 8:39 | comment | added | Pietro Majer | In the numbering method 1, maybe I would antepone the number: 1-Thm. , 2-Prop., 3-Thm., 4-Lem. , etc. | |
| Aug 3, 2019 at 6:28 | history | became hot network question | |||
| Aug 3, 2019 at 6:04 | comment | added | Robert Furber | @MikeShulman Most people in the present day, for sure. But it's also used (annoyingly) in some pre-LaTeX textbooks, such as Berberian's Baer *-rings. | |
| Aug 2, 2019 at 22:56 | comment | added | Mike Shulman | I have always assumed that most people using method 2 haven't really thought about it and are just letting LaTeX get away with its default behavior. To make LaTeX use method 1 you have to explicitly tell it to use the same counter for all results. | |
| Aug 2, 2019 at 22:52 | history | made wiki | Post Made Community Wiki by Todd Trimble | ||
| Aug 2, 2019 at 22:33 | comment | added | François G. Dorais | If a short paper has three main results, the second of which has a long proof with five lemmas, then Thm. 1, Thm. 2, Lem. 1-5, Thm. 3 makes total sense. | |
| Aug 2, 2019 at 22:33 | answer | added | Bjørn Kjos-Hanssen | timeline score: 13 | |
| Aug 2, 2019 at 22:32 | comment | added | Fedor Petrov | The reader may quickly count the theorems in your paper. | |
| Aug 2, 2019 at 22:26 | history | asked | Jakub Konieczny | CC BY-SA 4.0 |