Why is a lossy decomposition called lossy?

Why is a lossy decomposition called lossy?

What do we lose in a lossy decomposition?

There is a relation R. It is decomposed into two relation R1 and R2.

If R = (R1 JOIN R2) then it is a lossless join decomposition.

If R is a subset of (R1 JOIN R2) then it is a lossy join decomposition.

In a lossy join decomposition, after joining R1 and R2 we get more records compared to R.

What are we losing?