Is retro-causality a science fact?

[Soumya_M]

The "delayed choice" double-slit experiment seem to suggest that a measurement done, in the present, is able to change the past (history) of a particle. Does it imply that a cause arising in the future can change the past? If not what else does it mean?

In our common sense view of causality, we make two assumptions. These are: -
1. Cause precedes the effect in time.
2. Effect is dependent on the cause.

These assumptions agree perfectly with our everyday observations. We always see things of the past having an effect on the present and not the other way. So, causality seems to follow the so called arrow of time perfectly. But (as the delayed choice experiment suggests) could it be that our first assumption about causality is wrong?

But the question then is how can the past be changed if it is pre-determined? Or may be there are many pasts instead of just one, as we think! The observer chooses the past by our observing the present. In the words of Stephen Hawking "We create history by our observations, rather than History creating us."

However, I think the second assumption still holds good. That is, irrespective of the point of time at which events take place, they may be causally related simply by interdependence; i.e. the effect always depends on the cause and not the other way (unless we resort to some sort of destiny or "divine purpose").

[DrChinese]

...These assumptions agree perfectly with our everyday observations. ...

This is a variation on the argument of Hume. However, an assumption is still an assumption. The Delayed Choice experiments (and variations) would tend to cast doubt on the assumption.

[Steger]

The delayed choice experiment is entirely describable in your quantum interpretation of choice that does not contain retro-causality. Aharanov's "Time-Symmetric Interpretation" is an interpretation which sees retro-causality but in terms of the math he just suggests that the standard unitary operator used for time evolution can be split into two with a forward going and backward going operator from pre and post selection conditions. This kind of stuff is very interesting but all that it really says is that some variable
A can be written as B - C for some variables B and C. To the best of my knowledge NO interpretation of quantum (whether it incorporate retro-causality or not) makes any testable predictions beyond the standard copenhagen. Which really makes the discussion moot.