Who doesn’t love a good puzzle?

This collection of problems is intended to appeal to several audiences: students at all levels who are interested in languages, teachers who present courses in linguistics or computer science, and anyone with a general interest in language. Many of the problems have been used in the North American Computational Linguistics Olympiad, an annual high school competition.

The problems are designed to develop reasoning about languages and linguistics in general with an emphasis on computational thinking. Demonstrating the analytic problem-solving skills and techniques used by workers in computational linguistics is a great way to introduce students to the topic.

No computer science or linguistics experience is necessary for enjoying this book or for solving the problems. There are 50 of them, rated with one to five stars for level of difficulty. This rating gives readers a general guide to how hard the problems are and how long it might take to solve them. Of course, this will likely vary, depending on the reader’s background and interests. The reader may also improve as he or she works through the problems, since the same techniques work on many of the puzzles.

Many involve reasoning inductively from specific examples to learn general rules about a language and then applying those rules in new cases. By comparing words, phrases, or sentences in a foreign language to their English equivalents, you can deduce patterns that allow inferences about meanings, syntax, or usage.

As a simple example, consider the words “wer” and “bag,” meaning “creek” and “loincloth” respectively, and “walawer” and “balabag,” meaning “little creek” and “little loincloth.” What is the word for little leg if leg is “takki”? One can induce from these and other examples that the diminutive is made by adding “Xala” to the beginning of the word, where X is the word’s initial letter. If this is correct, then “talatikki” means “little leg.” Of course, induction isn’t always either that simple or correct.

Other topics include parse trees or finite state machines, which will be familiar to anyone with a computer science background. Even Braille and the Incan knot system are used in problems. Part of the book’s appeal comes from the use of many real languages that are non-European with non-Roman alphabets and symbols.

Answers are provided for all the problems. Most (but not all) of the answers include explanations showing how to reason through to the solution. It would have been more satisfying and enlightening to have explanations in all cases, especially when the given answer seems wrong or runs counter to the intuitive answer.

For example, one problem involves several foreign language sentences and their English translation. One sentence in this problem is “A ka kumun,” which is translated as “It’s sour.” Later, the reader is asked what the English is for “kumun.” The answer, given without explanation, is “summer.” That’s either an error, since “sour” seems the obvious answer from the context, or it’s a rather interesting twist or idiom that’s worth explaining. Perhaps since summer is hot, it causes things to spoil and turn sour?

Anyone who enjoys puzzles or language will probably find this book appealing. There’s no sequence to the problems, so readers can skip the ones they find too easy, too hard, or simply uninteresting. Readers will be able to learn to count in Irish, French, or Russian, or learn a few words of Tadaksahak, Sahaptin, or Tanna.

The problems are interesting enough to accomplish the author’s goal to motivate students or anyone with an interest in linguistics or computational linguistics.

And if you like this book, note that it is actually the second volume in a two volume set. Volume 1 [1] was published in 2013 as well.