-The Indo-Iranian family of languages belongs to the great Indo-European group and its ancient unity is regarded as certain and well defined by linguists (Manfred Mayrhofer, 1994, Bernard Sergent, 1995). This large family is sometimes called the Arya family or Aryan, because its members called themselves Arya.

According to Martin Haug (1862), the separation of Iranians and Indians is a consequence of a religious division, but it has been pointed out that geographical dissemination of Indo-Iranian on large territories also played a role in its evolution in two main subgroups (Herrenschmidt, 1987). These two subgroups are Indo-Aryan and Iranian.

Through religious texts we know two of the ancient Indo-Iranian languages : through Vedas, Sanskrit in the Indian world, and through Avesta, Avestan in the Iranian area. But there is a third one, Old Persian, inscribed on stone in Western Iran and quite different from Avestan. According to J.P. Mallory (1989), perhaps, the settlement of Indo-Iranians in the Near-East is older to what was thought. On the basis of the vocabulary from Mitanni (Upper Euphrates) found in the Hittite archives, he thinks that Indo-Aryan and Iranian were already independent before the middle of the second millennium BC. André Martinet (1994) even suggests 2000 BC. 'The earliest Aryan texts show two languages already separated and their writers established in Iran and India respectively or, rather, in that part of India which borders on Iran' (Jules Bloch, 1965), precisely where the famous civilization of Indus has been brought to light.

There are many questions about the geographical dissemination of Indo-Iranians and its chronology, and several different theories which try to explain their surprising mobility. The process of contemporary settlement of both Indo-Aryans and Western Iranians in Western Asia, i.e. in the North of Irak (Mesopotamia), Syria and Asia Minor, where the Kurdish dialects are spoken nowadays, is one of these major questions.

The linguistic connexion of ancient Cretan and Indo-Iranian adds a new problem to this large field of research. It is not impossible that the two registers of language in Linear A roughly correspond to the Indo-Aryan Vedic on the one hand for religious texts on libation tables, and to the ancient Iranian on the other hand for economic and administrative purposes on clay tablets.

Hubert La Marle, 2008

> the sacred bull in ancient Iran and in Crete