Another point is that even if you did want to prepare, eg by learning an old form of the likely language--you wouldn't be able to speak it, but you'd be better-equipped--you wouldn't know which one. A thousand years ago, Google suggests (ok, actually Quora, but I couldn't find better numbers) there were c.2 million speakers, out of a total population of at least 200 million - in other words, all languages with over 1 percent of the population, and possibly (taking a higher estimate) with as little as 0.5 percent of the population speaking it, criteria met (former) by Chinese, Spanish, English, Arabic, Hindi/Urdu, Bengali, Portuguese, Russian, Japanese, Vietnamese, French, Portuguese, Indonesian, German, Marathi, Telugu, Turkish, Hausa, Tamil, Swahili, Nigerian Pidgin, Tagalog, Punjabi, Korean, and Javanese, and, for the latter, in addition to all of the languages of the former, Amharic, Bhojpuri, Burmese, Gujarati, Italian, Farsi, Kannada, Lingala, Malayalam, Thai and Yoruba.<p>You can't even rely on languages being somewhat similar (French, Spanish, Portuguese and Italian are close enough that, given the amount of change, you probably can get away with learning just one of them; if you were really reductionist you could take one Indo-European language, one Turkic language, etc). Even being maximally reductive, the number of languages you need to learn is still 9, 10 if you go by the upper estimates (the Tai-Kadai family, represented by Thai, being optional). That's a lot, especially given that this metric lumps languages as separate as English, Bengali, and Farsi together, or Hausa, Arabic, and Amharic (and Maltese, for what it's worth). And that's just to have a decent shot at understanding a thousand-year-old version of the lingua franca!