We often like to think of religions as clear-cut things, with distinct boundaries between them. In reality, they’re often anything but; they borrow and copy and change and blend, in a process broadly referred to as syncretism.
You see this with particular frequency in polytheistic religions, because those can be remarkably flexible — much more so than monotheism. While as an author you might design a fictional pantheon of precisely seventeen gods, no more and no less, most real-world examples are far less fixed in nature. Which means that when they come into contact with outside deities, there are multiple options for how to deal with them.
This works in part because the relationship between mortals and the divine tends to be different from what you may assume. The relevant question to ask is often not “do you believe in that god?” but “do you worship that god?” If my enemies on the battlefield worship a different deity of war than I do, the question is not which of us is venerating something that doesn’t exist, but which of our deities is more powerful (and thus able to grant us victory). Becaus of this, when a pantheon changes, it doesn’t require a fundamental change in theology — only a change in practice.
In this framework, adopting a new deity is easy. We see this over and over again around the Mediterranean, with lines of influence and wholesale borrowing connecting different pantheons. Cybele appears to have started out in Anatolia, and from there spread to Greece and eventually Rome; Isis followed a similar path, but originated in Egypt; I’ve also previously mentioned Mithraism, which came from Iran and Zoroastrianism, and likewise wound up being very popular in Rome. It didn’t matter that these gods weren’t indigenously Roman: they held appeal for worshippers, and so they were worshipped. (Frequently in mystery cults, as it happened, which may owe something to the long-standing association of the east with magical power and wisdom that was considered “ancient” even by the standards of the ancient world.) On the other side of the planet, the Mexica, the core ethnic group of the Aztecs, erected side-by-side temples to their patron deity Huitzilopochtli and Tlaloc, a god probably inherited from the Toltec people.
Mind you, this raises the question of what counts as the “indigenous” faith of any particular place. Going back to Rome, many of their gods, myths, traditions, and iconography of course map pretty closely onto Greek those; those which don’t are often traceable back to Etruscan sources. Thus the Roman Minerva is connected to both the Greek Athena and the Etruscan Menrva . . . because where there’s a resemblance, there’s a chance to declare that you’re just using different names for the same divine power. So widespread was this tendency, in fact, that it has its own name: interpretatio graeca for the practice of explaining foreign religion in terms of Greek gods, or interpretatio romana for doing the same with Roman analogues.
And once you do that, differences of religion become not so much an ideological gulf as a semantic and practical one. Because of this — and unlike monotheistic missionaries preaching the scriptures of a revealed faith — the Romans often operated on a principle that might be summed up as “why go to the trouble of wiping out local religious practices when you can co-opt them instead?” Much of what we know about Gaulish religion is filtered through a very heavy Roman perspective, such that we’re told the goddess Sulis was their name for Minerva, and Lugus was the same as Mercury. So pervasive was this tendency, in fact, that some gods have whole lists of “provincial epithets,” tracing how foreign gods were slotted into the existing framework of the colonizers’ faith.
If you expect tidy delineations of what god has responsibility for what, this tendency will give you a headache. Culture A has a sea-god they call Gara, who also has power over storms. Culture B has a sea-god they call Ushio, who is also the god of the underworld. Culture C has a storm-god they call Lekinas, who is also associated with lust and artistic inspiration. When Culture A pulls a Rome and conquers the neighboring B and C cultures, they wind up with a version of Gara who rules over the sea, storms, the underworld, lust, and artistic inspiration — all of which he probably shares with other deities in his pantheon.
But sometimes what’s being borrowed isn’t an entire god, whether wholesale (as with Isis) or as an analogue to an existing god (as with the Gaulish gods). We also get syncretic with other aspects of a religion, such as the symbols, stories, and practices. This shows up in monotheistic religions as well; Catholicism may be a single faith, but looks distinctly different in Ireland, Mexico, and the Philippines, due to the influence of local customs in each place. Which isn’t surprising; it’s hard to truly stamp out long-standing habits and assumptions, and even if you try, they’ll often find cracks through which to leak back into the new ways. Festivals take place at the same time and have the same foods and dances, but they’re in honor of different things. The offerings once made to a pagan god instead get left at the altar of a Catholic saint.
And sometimes the borrowing gets even larger, to the point where it’s whole faiths being combined together. You see this particularly with the various African diaspora religions: because the Atlantic slave trade threw together people from disparate groups and combined that with varying efforts by slave owners to Christianize the people they chained, the result is a wide variety of traditions that blend these different elements together into new wholes. Santería, Candomblé, Haitian Vodou, and more bear the stamp of history on their beliefs and practices.
In short: real religion is often messy. It changes over time, adopting new elements from the influences around it, and it defies any effort to fix it down to a nice clean system distinct from those around it. Fictional faith doesn’t have to be like that . . . but there’s so much vitality and interest in letting things get a bit complicated.