PeterDonis
Mentor
- 49,326
- 25,360
pervect said:Distances as measured by yardsticks are frame dependent quantities, agreed. The time interval measured by a single, specific clock is, however, a frame independent quantity.
Unfortunately, I think these two statements taken together, while I see what you mean by them, are highly likely to increase the OP's confusion rather than decrease it.
The spacelike interval between two particular spacelike separated events is invariant, just as the timelike interval between two particular timelike separated events is invariant. (Here I am assuming the interval is being measured along a unique geodesic between the two events, to avoid a lot of technicalities that are out of scope for the OP at this point in the discussion.) A particular timelike interval has an obvious physical interpretation as the time measured by a clock that travels between the two events (along the geodesic between them). So it's easy to see how the invariance of a timelike interval is physically realized.
It's harder to see how the invariance of a spacelike interval is realized, because nothing moves on spacelike worldlines, so a measurement of any spacelike interval requires at least two timelike worldlines to be involved (one passing through each endpoint of the spacelike interval). This seems to introduce frame dependence, but it actually doesn't. It just means that we have to specify two worldlines instead of one. But worldlines are invariants, and if we specify two worldlines, say of the two endpoints of a ruler, and pick a pair of spacelike separated events, one on each worldline, that invariant spacelike interval will have a physical interpretation that can be described as "distance measured by the ruler". The interpretation won't be as intuitively obvious as the one for a timelike interval, but it will still be enough to support the invariance of the interval.
When the OP talks about "distance measured by the ruler" in two different frames, he is talking about two different spacelike intervals. Yes, these two intervals will have different lengths. But that doesn't mean the intervals themselves are not invariant. The length of each interval is invariant. They're just different.
I know this all is obvious to you, but I don't think it's obvious to the OP, and given how this thread has gone up to now, I think it's worth belaboring points like this to try to make clear what is going on.