I refer, from time to time, to a concept called "Gygaxian Naturalism." I realize that I've never actually explained what I mean by this phrase. As I use it, it refers to a tendency, present in the OD&D rules and reaching its fullest flower in AD&D, to go beyond describing monsters purely as opponents/obstacles for the player characters by giving game mechanics that serve little purpose other than to ground those monsters in the campaign world.
This naturalism can take many forms. For example, OD&D often tells us that for every X number of monster Y, there's a chance that monster Z might also be found in their lair. In the case of the djinn and efreet, as another example, we find that they both can create nourishing food and potable beverages, as well as many other kinds of materials through the use of their innate powers. In AD&D, these sorts of things get expanded upon greatly, with the Monster Manual telling us how many females and children can be found in a monster lair and giving many creatures powers and abilities that don't serve a specifically combat-oriented purpose, such as a pixie's ability to know alignment, for instance.
The intention behind Gygaxian Naturalism is to paint a picture of a "real" world, which is to say, a world that exists for reasons other than purely gaming ones. The implication is that monsters have lives of their own and thus go about their business doing various things until they encounter the player characters. Exactly what they do is described by reference to game mechanics, whether it be the numbers of non-combatants in a lair or spell-like abilities that help the monster do whatever it naturally does when it's not facing off against an adventuring party.
A consequence of Gygaxian Naturalism is that it grounds D&D a bit more in a pseudo-reality. I don't mean to imply that it's realistic in any meaningful sense, only that its fantasy follows "natural" laws of a sort, much in the way that, for example, I know that there are squirrels and raccoons and rabbits in my neighborhood who go about their business when I'm not seeing them in my yard or chasing them away from my recycling bins. That's one reason why AD&D has stats for so many kinds of "ordinary" animals: you can't build a "real" world without stats for sheep and cows and horses and such, because you never know when the PCs might need to kill one.
The end result of this is that Gygaxian fantasy is a simulation -- a fantastical one, to be sure, but a simulation nonetheless. The downside is that it's a very specific kind of simulation and it carries with it a lot of assumptions and expectations that not everyone shares. I know many OD&Ders, for example, who don't like "naturalistic" orcs, preferring them instead to be spawned from black ooze that bubbles up from the mythic underworld that is the dungeon. Likewise, the tendency to provide stats for everything is a self-perpetuating one, reaching its zenith in 3e, in which the game almost literally did stat out every conceivable thing with which your character might interact. Needless to say, some find this to be too much, myself included.
Gygaxian Naturalism survived Gygax's involvement in the development of Dungeons & Dragons and formed one the most important, if often only sub-consciously, creative foundations of the game through its second and third editions. My read of the latest edition is that it largely rejects Gygaxian Naturalism without embracing the alternative offered by some interpretations of OD&D, instead opting for a different model altogether. I myself have drunk deeply from the wells of Gygaxian Naturalism, so it's second nature to me now. I won't go so far as to say that Gary's approach is inseparable from D&D, because that's certainly not the case, but I will say that it's so deeply ingrained -- even in OD&D, particularly if you add the supplements -- that, to remove it, is very likely to have the effect of creating a different game entirely -- certainly a different one that what has been called D&D for most of the game's existence.