Tallis

Aristotle preferably if that's possible.

Let's start with his Ethics. What we call the Nicomachean Ethics, for example, is mentioned under this name in Cicero's works and Porphyry (234-305) knows both the Nicomachean and Eudemus' Ethics. The Great Ethics is a compilation of the former two, made after Aristotle's death.

All MSS of the Eud. Eth. contain Books V-VII of Nic. Eth., which points to the fact that this was already the case in Antiquity (or that all MSS are based on a single root MS containing this). This shows the bad nature of Aristotle's transmission, which is unlike most authors.

Now the Nic. Eth are based on six MSS, Laurentianus lxxxi. 11 (10th c.), Parisiensis 1854 (12th), Marcianus 213 (14th), Riccardianus 46 (14th), Marcianus 214 (14th) and Marxianus Append. iv. 53 (14th). We also have Moerbeke's latin translation which was used by Aquinas.

All these manuscripts have different readings, which mean different sources. We can track many of these sources down to ~250 BC Alexandria, where most scientific editions have been made in Antiquity. Their different errors tell us that many parallel traditions existed.

But there's more: Aspasius (2nd AD) commented the Nic. Eth, as well as Alexander of Aphrodisias, in the same century. Heliodorus did a paraphrase of the Nic. Eth., and all were in possession of slightly differing versions of the same text.

The fact that the MSS contain so many errors and differences is actually a proof for the validity of the MS tradition we have. It shows that early on, these works already had many versions, which means we have multiple sources, and multiple proof of Aristotle's works.

Yes, and this is all a matter of speculation: people are citing hypothetical texts to validate the existence of hypothetical parallel transmissions that "must date" from a hypothetical certain time.