languages no longer used in conversational speech and, as a rule, known only from written texts.
In some instances dead languages that no longer serve as a means of oral communication are preserved in their written form and are used to fulfill the needs of science, culture, and religion. For example, Latin ceased to be the conversational language of the Roman Empire in the fourth century; in different regions of the empire, Latin underwent linguistic changes that gave rise to new languages—the Romance languages. However, in the Middle Ages Latin remained the language of science, religion, and culture. Latin continued to be the language of religion and some sciences into the 20th century. Medieval Greek is the official state language of Greece (katharevousa—literally, “pure”), as opposed to the colloquial popular language (demotike).
The sounds of an artificially preserved language are pronounced according to tradition; for example, in Russia it is usual to pronounce Latin words in accordance with the German tradition and Greek words in accordance with the Western European Latinized tradition or a Russian modification of the Byzantine pronunciation (after the 1860’s used only by the church).
In other instances, when an ancient language was used primarily in spoken form, it has disappeared completely; however, interest in it is renewed when written texts in that language are discovered. Decipherment makes it possible to read and comprehend ancient texts, and dead languages are made accessible to scientific study. Thus, with the aid of a parallel bilingual text (the Rosetta Stone), the English scholar T. Young in 1814 and the French scholar J.-F. Champollion in 1822 deciphered the Egyptian language. In 1915, the Czech scholar B. Hrozný deciphered the Hittite written language. In 1963, the Soviet scholar lu. V. Knorozov decoded the written language of the Maya Indians, which consists of ideograms, phonetic symbols, and radicals.
Often the original name of a deciphered language is unknown, and it is given a conventional name; for example, in 1907–08, the German scholars E. Sieg and W. Siegling deciphered two closely related Indo-European languages, which were given the conventional names Tocharian A and Tocharian B, according to the sites where the texts were found. One of the dialects of ancient Greece (the island of Crete) was conventionally named Linear B (in accordance with the type of writing system). Sometimes the writing system and the individual, more or less clear lexemes of a dead, undecipherable language permit a certain group of texts to be given a particular name; for example, Linear A refers to the ancient, undecipherable texts found on the island of Crete.
The word stocks of dead languages are limited and often contain lexemes of unknown meaning and lexemes that occur in texts only once (hapax legomena).
Attempts have been made to revive dead languages. In Avignon (France) and the Vatican, collections of new Latin terms are published; in such cases the new vocabulary is borrowed or is translated from living languages using the lexical stock of the dead language.
M. L. VOSKRESENSKII