K.D. Harrison, who looks like a US marine but is a professor of linguistics at Swarthmore College, has penned this book as an appeal to the world to understand the loss to human knowledge involved in the extinction of small languages. As he explains in the beginning of the book, the 100 largest languages in terms of population cover over 80% of the world's speakers, whereas the entire bottom half of languages in such a ranking covers only a very small percentage. This means that there is a vast amount of languages that are spoken by small, isolated groups, and the increasing interconnectedness of the world population through developments in communication and information technology threatens the survival of these languages: popular languages crowd out less popular ones.
Harrison himself is a specialist in Turkic Siberian languages, all very rare and small ones, but his defense of preserving small languages applies to all of them. Yet, as he admits, the issue is less straightforward than it seems. The first problem is that the vast majority of the speakers of small languages are illiterate, and that it is a known fact of linguistics that non-written languages tend to vanish much faster than written ones. However, creating a script for a language kills the oral traditions of that culture by fixing them forever at a given point, which is a hard thing to ask of an anthropologist, and which we may not have the right to do.
A second problem is that it is not an evident thing that small languages are worth preserving in the first place. Harrison clearly sees this argument coming, and the greatest part of the book consists of an attempt to provide various reasons why small languages can be, if one looks at it purely from a practical non-romantic standpoint, worth keeping alive. Using various case studies from his own research, he shows that a lot of things that used to be taken for granted in linguistics as 'constants' of all human language use where in fact only aspects of popular languages, but lacking in certain rare ones entirely. This is well demonstrable in counting systems, which Harrison spends a chapter discussing, as although use of numbers seems something that would be universal in all cultures, it is in fact wildly variant even to the degree of noncommensurability. The recent discussion about the Piraha people, who appear to not count at all, makes this even more relevant. In a similar way, Harrison provides examples of certain linguistic assumptions about word use that were proven wrong by 'discovery' (by Western researchers) of rare languages.
There are nevertheless some things missing that would have really added to Harrison's discussion of the problematic. The first and most obvious thing is the political and social contexts of use of rare languages. In many cases, users of small languages, at least in the younger generations, do not wish to use their own language any more since it is strongly socially and often even politically disadvantageous to do so. Many minority languages in various nations are discriminated against or considered not to exist, and there is additionally the pressure of wanting to appear 'modern' or 'civilized' instead of speaking some backwater language the old folks use. A discussion of the various ways governments as well as minor language speakers have dealt with this would have been useful both for judging the value of preserving small languages (a political question) as for scientific purposes, but is entirely missing.
Another thing that would have really added to the book is a more in-depth discussion of the Sapir-Whorf hypothesis and its relation to the usefulness of preserving small languages. Harrison does mention the hypothesis (which states that language determines thought, i.e., that some thoughts can only appear or be expressed in certain languages and not in others) and that its radical form is usually rejected, but does not otherwise go into it. This is a major gap, as it seems to me this is central to the issue. If anything can be expressed in any language, there is not much point in keeping small languages around, if only for efficiency reasons and the way it divides people. But if, on the other hand, certain things can be at least expressed far better in one language than in another, there can be strong reasons to want to keep small languages around, even if just for the sake of poetry and literature. The more strongly the Sapir-Whorf hypothesis appears valid, the more reason there seems to be to politically support the use of small languages.
With these two considerations outside Harrison's discussion, the book is more of an emotional appeal combined with a series of case studies to pique the interest of the common reader. This is interesting enough as is, but unsatisfactory to resolve the real question: what does it mean for us when languages die?