Yeah, they are both true, they just kind of sound opposite without being opposite.
If you find somebody who plays pool, billiards or the like, you can ask them to show you backspin tricks and stuff like that. The way that a ball spins on a surface is not determined by its translational motion, they can be different. Pool players use the resulting friction to curve the cue ball around obstacles or ensure that after a collision the cue ball rolls into a more favorable position.
The rolling-without-slipping behavior $\omega =v_\text{trans}/R$ is enforced by the friction, in the sense that the friction will eventually bring it to this behavior if nothing else intervenes. Once that equilibrium has been reached, the friction has nothing else to do, it has brought the system exactly where it wants it.
So simultaneously the friction is the cause of the behavior, and the friction doesn't do anything once the behavior is established.
Just to give a totally non-physics example of this, you could punch someone every time that they say a swear word until they don't say it anymore. Once this system reaches the steady state, “the reason I don't say that word anymore is that I don't want to be punched.” / “But I have observed you for many days, getting punched does not seem to be a big problem for you!” / “Right, because I stopped saying that word!”
This is something that happens very often with all sorts of constraints, constraints very often have forces that only apply if your motion is trying to violate the constraint. So if one is trying to find out when a car falls off of a loop-de-loop, one would indirectly look for a vanishing normal force, the normal force enforces the constraint that the car does not fly outward through the loop-de-loop surface, when this vanishes it implies that the car is just barely holding on and any lower speed falls off of it.