Laura Marsh reviews Nicholas Ostler's book of that title in The New Republic:
While English is the most widely-spoken lingua franca in history, so-called common or working languages can be much less pervasive. Elamite, for example, was the submerged administrative language of the Persian Empire in the sixth century B.C.E. All official documents were written down in Elamite, but they were both composed and read out in Persian, the language of the illiterate ruling class. Then there is Pali, the language of Theravada Buddhism. No longer used in everyday conversation, Pali is written in different scripts in Sri Lanka, Cambodia, and Burma, and sounds different when read aloud by Thai and Burmese speakers. The identity of the language is almost obscured by its profusion of forms.
Pali is a tantalizing case for Nicholas Ostler, because it suggests to him the possibility of a “virtual” language. A “virtual language” would not be read or spoken itself. It would allow the user to understand what is being written or said without learning the original language—in much the same way that “virtual reality” allows the user to have an experience of something without actually doing it. Pali is not “one language” in the concrete sense that it has one set of words, but those who know any of its forms can access exactly the same information. Yet on closer inspection this is not because it is a “virtual language.” It is because the differences between its forms are largely superficial. However the words are pronounced or written down, they mean the same thing. It is one language after all.
Despite this setback, Ostler has faith in a virtual system, which he claims will revolutionize global communications, and make foreign language learning a thing of the past.