Moving a ceiling joist from one side of a rafter to the other has no structural impact.

Here's a hint if you get towards the eave and things get tight. Don't take out the ends of the old joist at the eave where it doesn't interfere with the potential finished ceiling. Just cut out the main span of the old joist. you can drill through the new joist, the rafter and end of the old joist and put through a bolt. This avoids the need of trying to yank out old nails and bang in new one in a confined space. Warning: this is only in a type of roof such as you are working with that has little load bearing on each individual rafter.

"Just cut out the main span of the old joist. you can drill through the new joist, the rafter and end of the old joist and put through a bolt."

You need more than one bolt and they ned to be spaced along the length of the joist to prevent a failure by twisting the fastener aunder the tension load the joist carries from the rafters.
The ceiling joist in a conventional framed roof are under tension from the roof load. As long as you make sure to provide adequate strength in the fasteners (size, number, and spacing) you can alter the joist.
Make a mistake and the roof will sag at the peak as the walls of the house are forced outwards.
If you do not know how to compute the loads present and design the fastener schedule, do not mess with this.

Not in the situation we are discussing. A garage with rafters over such as short span with no other configaration isn't going to twist. A middle joist could be removed completely in the middle of the roof and there are no forces to cause undue stress and any twist. A joist removed in the second or third positions at either end might be questionable but even then I can't see any twisting.

"A garage with rafters over such as short span with no other configaration isn't going to twist. "

The section bolted in is loading the bolts in shear since it is under tension. This causes the bolts to want to rotate about the long axis in the wood as the boards slide past each other.
If you exceeed the crush load of the wood the bolts will rotate and the joist will get longer transfering a thrust load to the walls.
The bearing area of the bolts is equal to their diameter and the crush is parallel to the fibers. This type of joint can easily fail if the fasteners are to close to the ends of either board, if fasteners are lined up stright wioth each other along the the boards, or if the wood compression is exceeded.
It usually shows up after heavy loading from snow (or anything that adds a prolonged load approaching the design limits).
Even headerin off ceiling joists can cause long term roof problems if the framing is not correct. Nails must not be loaded in tension (withdrawal).

Thanks for the input.

Greestar, I stopped and talked to a fellow framing a house and he said much the same thing about not to worry about it. He said not to bother puting in a joist but just to move the one that was there as nothing was going to happen in a small span. It took me all of 5 minutes to knock the joist out and reattach to the opposite side. It was easy to access at the ends. I spent more time knocking out the nails from the joist after I got it out. Sure enough when I put it on the other side to go and nail it I realized I could have left the nails and just reversed the direction of the joist and banged them in.