Last year we talked about Sudoku (Great new way to waste time catches on fast), and followed that up with a blog about Kakuro (Do you do or do you don't
'doku?). There, we said that: “We at Dewey Manor don't
think Sudoku really counts as a mathematical game, since it doesn't require
players to carry out any mathematical operations,” and “[Kakuro] is really a
test of logic, not of arithmetic.” So we assigned Sudoku and Kakuro the Dewey number **793.73 Puzzles and puzzle games**, and not **793.74 Mathematical games and recreations.**

However, in the comments on the first blog, Thom Hickey and Dennis McGovern disagreed with us. Dennis (writing under the handle of “LCDennis”) said:

I disagree with the other residents of Dewey Manor in that I think Sudoku really counts as a mathematical game, even though it doesn't require players to carry out any mathematical operations. The popular version of Sudoku (i.e., the one described in the 1st 20 Google results) is played using numbers (mathematical symbols). I think of number games as mathematical recreations, and I agree with the LC subject heading (sh85082131) which has Number games as a 450 ref. to Mathematical recreations. To my way of thinking Guardian's Godoku (which doesn't have much literary warrant) is another game: a logical game using letters that should also class in 793.74.

He also pointed us to the Wikipedia article on Sudoku and to web pages on Sudoku and related puzzles by the Australian mathematicians Anne Street and Diane Donovan.

These things have way of coming back to bite you, and we have just dealt with assigning Dewey numbers to the LCSH Weekly List containing “Sudoku.” (Weekly List 2005 no. 37
-- the list with “Kakuro” is still in our future). And we’ve had a change of heart: we’ve
decided that Sudoku is a game involving Latin squares, a topic within **511.64 Permutations and combinations**, while Kakuro involves **519.77 Integer programming**. The Wikipedia even has an article on the Mathematics of Sudoku.

**510 Mathematics** is a large topic, and now includes a lot more than traditional topics like algebra, calculus, geometry and statistics: it includes many other formalized
and axiomatised systems, such as **511.3 Mathematical logic**, **511.6 Combinatorics**, and **519.7 Programming** (which is not the same topic as the mathematics of computer programming, but includes linear programming and nonlinear
programming). Both Sudoku and Kakuro fit within this broader concept of mathematics, and belong in **793.74 Mathematical games and recreations.**

The links to the previous posts aren't quite right. Here's the first: http://ddc.typepad.com/025431/2005/08/great_new_way_t.html

--Th

Posted by: Thom | June 20, 2006 at 12:06 PM

Thom -- Thanks, I got a bit confused in setting up the links, but I've corrected them in the article, and I think they are right now.

Posted by: Giles Martin | June 21, 2006 at 09:44 AM

Thanks for the shoutout to NCFC the other day, Giles! But of course it was always going to be The Sudoku Question that would finally wake me from post-Manor reverie. Here's my take on the latest news: Noooooooooo! Don't believe the hype! Sure, some people write about "the mathematics of sudoku" just as you can imagine other people writing about "the mathematics of chess." That doesn't make either sudoku or chess mathematical recreations. The class number in question is not being assigned to works on this "mathematics of sudoku"; it's being assigned largely to works that comprise lots of suduko puzzles. And the bottom line is: When you solve a sudoku puzzle in the normal way, you don't do anything that would normally be construed as mathematical. Sure, you can analyze the setting or solving of sudoku (or chess) puzzles mathematically. That doesn't make the act of solving sudoku (or chess puzzles) in the normal way a mathematical recreation. Sudoku is a logic puzzle, and logic puzzles go in 793.73. To cite "Mathematical logic" as part of mathematics is a red herring. Mathematical logic does not approximate the whole of logic. If it did, then presumably we would be want to class all logic puzzles in 793.74. But it doesn't and we don't! The distinction between logic and mathematics may well be a "fine line" (as Thom says), but so long as it's a line that Dewey continues to draw (in this case between logic puzzles and mathematical puzzles), then how can we possibly class sudoku -- a puzzle which requires no knowledge of any mathematical concepts to solve -- as a mathematical rather than a logical recreation?

Posted by: Jonathan | June 21, 2006 at 10:24 AM

Agreed. I've been playing Sudoku almost non-stop ever since a friend introduced it to me last Christmas. The game would play just as easily if you used sets of letters, shapes, or colors instead of numbers. It's a logic game.

Posted by: Kaffinator | June 23, 2006 at 12:23 PM

I think that Kakuro is much more interesting than sudoku beacuse that shapes change all the time!

Posted by: kakuro | October 07, 2006 at 05:25 AM

I disagree because to do Sudoku you have to use addition to complete each square. That means its a mathmatical problem!

Posted by: Free Jigsaw | January 12, 2009 at 07:51 PM

I don't think sudoku is a real math game. You just place numbers in a logical sequence.

Posted by: Domo Sudoku | December 24, 2009 at 08:35 AM