What do you think?
Rate this book
Puzzles in Logic, Languages and Computation: The Red Book (Recreational Linguistics)
This book collects engaging problems created for students competing in the Computational Linguistics Olympiad. It offers both linguistic and computational puzzles with their solutions. Recommended for anyone fascinated by the phenomena of human language.
Hardcover
First published February 10, 2012
2 people are currently reading
31 people want to read
About the author
Dragomir Radev
6 books2 followersDragomir R. Radev is a Yale University professor of computer science working on natural language processing and information retrieval. He previously served as a University of Michigan computer science professor and Columbia University computer science adjunct professor.
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 of 1 review
September 5, 2014
This is a great collection of linguistic puzzles written in English. A lot of them come from North American Computational Linguistics Olympiad (NACLO), few are original. Puzzles vary in level of difficulty, as indicated by number of asterisks; several of the puzzles are related to topics in computational linguistics, and there's an index at the end of the book which lists these topics with proper terminology, which is quite useful if you want to look them up and to read more about them. There's also an index which lists puzzles by languages.
All problems can be solved only using logic, but some basic knowledge of linguistic or computer science will certainly come handy. Knowledge of random trivia can help too; for instance, in puzzle 3, the reader matches Greek phrases with their English translations; (similar to this one which is also included in this book)
but if you know the meaning of "Philadelphia", then you will probably recognize "adelphos" in the Greek phrases, and if you ever saw the TV commercial of the yogurt "Oikos" then there's another word you already know, but this is stripping off the fun from the puzzle... of course, it's another story if you already speak Greek.
Puzzles from NACLO are the most interesting and complex; they involve less well-known languages such as Dyirbal, Aymara and Manam Pile, where the reader usually has to figure out basic grammatical rules and some particularities about the language; observation skills are quite important, one often has to make bold guesses too and see if they fit the observations.
Puzzles on English language are more related to CS, concerning parsing, sentence similarity and text processing.
I was pleasantly surprised to find out the book even contains a puzzle on spectrograms!
Puzzles from other sources vary in quality. For instance, number 47, also available here .
Readers are given three sentences in English and their Russian translations
John killed Mary — Джон убил Мэри
Mary killed Sam — Мэри убила Сэма
Sam killed John — Сэм убил Джона
then they are asked to translate sentences like "John killed Sam" into Russian.
If one doesn't know any Russian (or similar Slavic language's) grammar, then the main question is probably why only the middle sentence has "убила" instead of "убил". There are few possible guesses, but even if one guesses right (the verb in the past tense ends in 'а' if the subject is female), it cannot be verified using only the given information. Furthermore, if the given sentence is instead "Mary killed John - Мэри убила Джона", one may guess it ends in 'а' for phonetic reason because 'Дж' is a foreign consonant. After all, phrases like "in Canada" and "in Russia" are respectively "в Канаде" and "в России", but "in France" is "во Франции", since "France" starts with a consonant cluster.
Personally, I find some puzzles like this one above sloppy and disappointing; (the next puzzle by the same author was much better and on par with the NACLO ones) However, these examples are few, and as the Chinese people say, "few spots on a jade cannot hide its beauty" :)
Oh well, that was long... hope this review will be useful for someone! So, in short, this book is one of the best compilations of linguistic puzzles I've seen, and I would recommend it to anyone who enjoys recreational linguistics or who would like to try logic puzzles of a non-mathematical nature.
All problems can be solved only using logic, but some basic knowledge of linguistic or computer science will certainly come handy. Knowledge of random trivia can help too; for instance, in puzzle 3, the reader matches Greek phrases with their English translations; (similar to this one which is also included in this book)
but if you know the meaning of "Philadelphia", then you will probably recognize "adelphos" in the Greek phrases, and if you ever saw the TV commercial of the yogurt "Oikos" then there's another word you already know, but this is stripping off the fun from the puzzle... of course, it's another story if you already speak Greek.
Puzzles from NACLO are the most interesting and complex; they involve less well-known languages such as Dyirbal, Aymara and Manam Pile, where the reader usually has to figure out basic grammatical rules and some particularities about the language; observation skills are quite important, one often has to make bold guesses too and see if they fit the observations.
Puzzles on English language are more related to CS, concerning parsing, sentence similarity and text processing.
I was pleasantly surprised to find out the book even contains a puzzle on spectrograms!
Puzzles from other sources vary in quality. For instance, number 47, also available here .
Readers are given three sentences in English and their Russian translations
John killed Mary — Джон убил Мэри
Mary killed Sam — Мэри убила Сэма
Sam killed John — Сэм убил Джона
then they are asked to translate sentences like "John killed Sam" into Russian.
If one doesn't know any Russian (or similar Slavic language's) grammar, then the main question is probably why only the middle sentence has "убила" instead of "убил". There are few possible guesses, but even if one guesses right (the verb in the past tense ends in 'а' if the subject is female), it cannot be verified using only the given information. Furthermore, if the given sentence is instead "Mary killed John - Мэри убила Джона", one may guess it ends in 'а' for phonetic reason because 'Дж' is a foreign consonant. After all, phrases like "in Canada" and "in Russia" are respectively "в Канаде" and "в России", but "in France" is "во Франции", since "France" starts with a consonant cluster.
Personally, I find some puzzles like this one above sloppy and disappointing; (the next puzzle by the same author was much better and on par with the NACLO ones) However, these examples are few, and as the Chinese people say, "few spots on a jade cannot hide its beauty" :)
Oh well, that was long... hope this review will be useful for someone! So, in short, this book is one of the best compilations of linguistic puzzles I've seen, and I would recommend it to anyone who enjoys recreational linguistics or who would like to try logic puzzles of a non-mathematical nature.
Displaying 1 of 1 review