It's pretty simple once you understand how the function fits together. The function selects a range of cells that has a marker cell at the top left. The position of that top/left cell is defined by the first three arguments.
The first gives the start point, the 2nd and 3rd arguments (rows and columns) set how far to move before you set the top left corner of the selected range.
The height and width arguments set the size of the selected range.
So J14, is offset by zero rows and zero columns, in your example. Which is to say, the selected cell starts at J14, moves zero rows down and zero columns across, and so stays at J14.
However, the function also selects a range of height 1. Fine, that's easy enough. A single cell is a range of height 1.
But it also has a width of -5. At this point you need to think like a computer. -5 columns to the right of J14 is actually 5 columns to the left.
It selects the range F14:J14, because that range is 5 columns wide.
The height and width will default to the height and width of the original selected reference.
A more intuitive example might be:
The function starts at cell A1. Then it moves 2 cells down (ROWS) and three cells to the right (COLUMNS). The top left cell selected is D3. Then it selects a range that is 4 cells high (HEIGHT) and 5 cells wide (WIDTH) from cell D3. That is to say, D3:H6.