The main job of historical linguists is to learn how languages are related. Generally, languages can be shown to be related by having a large number of words in common that were not borrowed (cognates). Languages often borrow words from each other, but these are usually not too difficult to tell apart from other words. When a related group of languages has been studied in enough detail, it is possible to know almost exactly how most words, sounds, and grammar rules have changed in the languages.
One of the earliest observations about language was by the Romans. They noticed that Latin and Greek were similar. However, they incorectly assumed that Latin came from Greek. The reality is that both came from Indo-European. More on that later.
There were lots of people looking at languages in the middle ages. However, most of them were trying to show Hebrew giving rise to all of the world's languages, specifically European languages. This never really worked, since Hebrew is not directly related to Indo-European languages.
When Europeans started travelling to India about 300 years ago, they noticed that Sanskrit, the ancient literary language of India, was similar to Greek, Latin, and other languages of Europe. In the late 18th century, it was first correctly theorized that Sanskrit and the languages of Europe had all come from the same language, but that that language was no longer living. This was the beginning of Indo-European. Since then, many languages from all over the world have been studied, and we are starting to get a good idea of how all the world's languages may be related.
Back to Languages and Linguistics.