The whole conception was that during that stage, there were no civilians inside the area in which we were fighting – there was no such thing. It doesn’t happen. The situation on the ground is very clear – the preliminary fire, which isn’t directed at houses, signals very clearly [to the Palestinian population] that we are going in; this is after the leaflets and all those things. We have our zone, which is totally open, there are no other forces there and we know where the next battalion is located. During the first stages we would still go on the [two-way radio] to get authorization [to fire] from our commander – we say, “We spotted a [missile] launch, we have a coordinate, do we have authorization to fire?” After we validate that launches were made in our direction, or in the direction of Israel, from within the built-up area – then there is authorization to fire at them, and then we just open fire on our own. In the beginning, they tried to define priorities, and [we were encouraged] to avoid directing high-arc trajectory fire into built-up areas. But in practice, when we were in the field and we had a combat chopper or some other aircraft up in the air, we used it first thing [to fire at the site from which the launch was detected]. But at times when [air support] wasn’t available due to constraints – we employed [the artillery] straight away.

After you detect a launch, you go on the radio with a senior officer, get authorization to fire and then employ the artillery division. First an explosive shell is fired. We identify the spot where it landed – I don’t know if I could say whether it was right on target – but if we saw that it landed in the same area from which we saw the rocket being launched, then we authorize the artillery guys to fire, say, five shells.

The first shell is for calibrating [the mortar’s] range? It’s not calibration because we don’t exactly pinpoint the location [where the shell lands], but [if] we see it hit the spot and the deviation isn’t serious than we authorize five shells on that spot.

What’s the range of a mortar’s deviation, how precise is it? The standard deviation range, that isn’t due to an error in aim, is up to 400 meters. Less than 400 meters, it’s not an error in aim; it’s a possible deviation that gets fixed. In practice, [the mortars] are usually precise at a range of 50 to 100 meters.

Did you have any of those new GPS-fitted mortars? Our crew didn’t, we heard that the battalion tried to get hold of some but didn’t manage. I know that generally those are very, very precise, with a 20-meter range of deviation.