In 1786, Sir William Jones suggested that similarities among languages such as Sanskrit, Greek, Latin, Gothic, and others were so striking as to suggest that they had sprung from a common, no longer existing source (see Lehmann's Reader, chapter 1) now called Proto-Indo-European. By the 19th century, August Schleicher's Family Tree had been proposed to model the relationships among the Indo-European languages as the branches of a tree (see Lehmann's Reader, chapter 8). Subsequently two new branches were discovered, Anatolian and Tocharian.
The tree above is very different from that of Schleicher. In the early part of the 20th century, Antoine Meillet suggested that Greek (Hellenic), Armenian, and Indo-Iranian were more closely related to each other than to any one of the other languages, and linguistic similarities among Celtic, Italic, and Tocharian are now thought to indicate a closer prehistoric community, while Germanic was isolated very early: only later, in northern Europe, did Germanic, Baltic, and Slavic speakers come back into contact.
Johanna Nichols' 1995 Indo-European and the Indo-Europeans, an English translation of Gamkrelidze and Ivanov's 1984 Russian original, suggests (pp. 349-350) layers of groupings beginning with two original groups labelled Area A and Area B. In Area A, an unseparated group, are the ancestors of Anatolian, Tocharian, and Italic-Celtic, while in Area B is another group, Germanic-Balto-Slavic together with Indo-Iranian-Greek-Armenian before subsequent splits (see our summary). Their update attempted to incorporate the discoveries of Hittite and Tocharian while at the same time accounting for interacting dialect zones that would have developed over time, creating ever new layers of areal interactions. Recent concerns for integrating the factors of time (the chronology of the languages) and space (the locations of the speakers during periods of their history and prehistory) must factor into re-evaluations of all models.
Views concerning the breakup of the initial language family into groups are often related to hypotheses about where the Indo-European Homeland was. One major view is that it was on the steppes of southern Russia north of the Caspian Sea, and that successive migrations to the west brought speakers into Europe, others south into Anatolia, and south or southeast into Iran and India. Another major view places the Homeland in Anatolia with migrations northwest, north, northeast, and east. Models of language history will also affect what is eventually accepted as the most likely reconstruction.