Linguists classify languages into families based on similarities in their forms, their grammatical structures, and their vocabularies. For many centuries, people have recognized some obvious language classes: the similarities between French, Italian, Spanish, and Portuguese are clear to anyone who knows the languages, so they're classified as parts of the same language family (known as the romance languages). Other groups (such as Germanic) are equally obvious.

But in the seventeenth and eighteenth centuries, linguists began to notice some similarities not just within but across these family groups. So, for instance, the Germanic, the Romance, and Hellenic (Greek), and the Indo-Iranian language groups are all distinct, but the words for "father" look suspiciously similar: English "father," German "Vater," Latin "pater," Greek "patêr," Sanskrit "pitar." In the middle of the eighteenth century, a scholar named Sir William Jones explained these cognates by proposing an Ur-language now called Indo-European or Proto-Indo-European (IE or PIE), from which all these language families descended. By working out the laws by which grammatical structures and sounds shift over time ("p," "b," "v," and "f," for instance, can be shown to be related), linguists have demonstrated that all the languages in this group are descended from a common parent.

Nearly all the languages of Europe belong to the IE group, with only a few surprising exceptions (such as Finnish and Hungarian). A number of the languages of India (Sanskrit, Pali, Panjabi, Gujarati, Bengali, Hindi, Urdu, Nepali) are more distant cousins.

The European IE languages can be further subdivided into romance, Germanic, Hellenic, Slavic, and other groups.

From the Guide to Literary Terms by Jack Lynch.
Please send comments to Jack Lynch.
Note: This guide is still in the early stages of development.
Three question marks mean I have to write more on the subject. Bear with me.