Suppose you make a threat against one of your opponent’s pieces; perhaps your knight attacks his bishop, and the bishop is loose. He can respond either by moving the bishop or by guarding it. If he guards it, you now have a second offensive focus besides the bishop: the guard itself. Undermining it can become your immediate task or can hover as an ongoing issue in the game. With time, more such issues come up: at any given moment several of your pieces may be aimed at pieces, pawns, and possible mating squares belonging to your opponent. A critical question in every position is whether those points you have under attack can be made vulnerable; they all may have guards, but can the guards be captured, or blocked, or driven or lured away?
This section is devoted to methods for achieving those aims: ways of removing or disabling the pieces that guard targets in the enemy camp. It covers material that has been labeled in a wide range of ways in the literature, including terms such as destruction, deflection, diversion, decoying, damming, drawaway, driving off, breaking communication, blockading, overloading, attraction, interference, interception, and obstruction. While some of those terms have useful meanings, on the whole the proliferation of jargon is unfortunate. All of those devices can be assimilated under the heading of “removing the guard”; we can then subdivide this theme into four methods—four ways to loosen a piece or square you would like to take. (a) You can capture the guard; (b) you can attack the guard (i.e., threaten it so that it becomes obliged to leave its square); (c) you can take something else the guard protects (distracting it, or demonstrating that it is “overworked”); or (d) perhaps you can interpose something between the guard and its protectorate, interfering with the defensive work the guard is trying to do.
Each of those themes gets a chapter of its own in this section. We have seen most of them in earlier chapters as ways to create forks and other tactical strikes; they are the methods we have used to loosen targeted pieces or to loosen the square needed to impose a fork or pin, etc. Here we focus on them as ways of loosening pieces and mating squares so that you can capture or occupy them straightaway.
Our first theme involves capturing the guard: simply taking it, so that its protectorate becomes loose. We begin with simple examples like the position to the left. In this section, as in the others, we will be dwelling on a single technique to guide our searches; in this case it will be to ask what enemy pieces you threaten (and then investigating whether the threat can be made to work). Here, then, you start by noticing that your rook attacks Black’s bishop, which is protected by a pawn and rook. And your queen attacks Black’s knight, which is protected by...his bishop. In those observations lie a tactical idea: play both captures in succession. First comes RxB, removing the knight’s guard. Black’s recapture RxR momentarily wins the exchange but replaces the bishop with a rook that no longer can guard the knight on g4. Now White plays QxN, winning two pieces for a rook.
Black has a defensive idea here, by the way, that doesn’t quite work but is important to notice. He can reply to RxB by taking the offensive with Nxf2. Now Black has picked up a pawn and he attacks two pieces at once; his knight (which was a goner anyway) attacks White’s queen, and his rook (and also his f7 pawn) attacks White’s rook. The move fails to impress, though, because White can keep pressing his rook forward with RxR+, not only taking a rook but inflicting a check that Black has to fend off. Now White is the one who attacks two pieces at the same time. Black can play RxR, but then White has another move he can spend taking Black's knight or moving his own queen to safety. So Nxf2 doesn’t help Black here; he ends up trading rooks and losing a piece for a pawn. But the idea behind Nxf2 is valuable: when one of your pieces is taken (as happens to Black here when White plays RxB), you might be able to profit by delaying or forfeiting your recapture and instead playing a threat of your own elsewhere. If Black’s pieces were arranged a little differently, the idea would have helped him to reduce his losses here.
Now return to the correct sequence (1. RxB, RxR; 2. QxN) and see what can be learned from it. You already know from prior studies that it is dangerous to send a piece out onto the board loose (i.e., with no protection). Even if it sits on a square that is not being attacked, such a piece easily can become the subject of a fork and get taken for nothing. Here Black avoided that problem—in part. All of his pieces were protected; but they were not protected equally well. The bishop was protected by a pawn, and so was very safe in the sense that it could not itself be taken profitably. The knight, however, was protected not by a pawn but by a fellow piece. It often is hazardous to venture a piece into enemy territory with no better protection than a fellow piece farther back on the board. This position shows why. A knight advanced to close range, like Black’s knight here, is easy for White to attack. It then becomes only as safe as the piece that defends it—Black’s bishop. Even if that bishop is well-protected by a pawn, the pawn provides no security against the risk that White can throw a piece at the bishop—even with a sacrifice, as where White uses his rook in this case—and thus cause an exchange that leaves the knight loose.
All this shows why the better practice usually is to avoid planting a knight in enemy territory until you have created a suitable home for it: a square protected by a pawn. It would be too much to state this as a rule; like everything in chess, it depends on the position, and sometimes the benefits of putting your knight in a riskier position outweigh the costs. The important thing is just to be aware of the dangers involved in protecting pieces with pieces.