Nimzo-Indian Defense: Sämisch Variation

After 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4, the Sämisch Variation answers the pin with 4.a3 — a direct, uncompromising question. The bishop must either retreat, giving up the pin without achieving anything, or capture on c3, damaging White's pawn structure permanently in exchange for the bishop pair. Where the Rubinstein Variation's 4.e3 postpones this decision indefinitely and the Classical's 4.Qc2 avoids it altogether, the Sämisch forces the issue immediately, on move four, before either side has developed much of anything else.

The trade that defines the whole structure

After 4...Bxc3+ 5.bxc3, White accepts doubled c-pawns on the queenside — a genuine, permanent structural weakness — in direct exchange for the bishop pair and a claim to greater central influence, since the pawn on c3 supports a future d4-e4 buildup more effectively than a knight on c3 would have. This is structurally reminiscent of the French Winawer's own doubled-pawn trade after 5.a3 Bxc3+ 6.bxc3, and the resemblance is not superficial: both arise from a bishop pin met with a3, both trade structural damage for the bishop pair and central space, and both produce a position where the side with the damaged structure needs active piece play to justify accepting it.

What Black gets in exchange for causing the damage

Black's compensation for handing White the bishop pair is concrete: the doubled c-pawns are a genuine long-term target, and Black's typical follow-up plans — a quick ...c5 to attack them directly, or ...b6 and ...Ba6 to pressure the light squares White's remaining bishop can't cover as effectively while blocked in — are designed to make White's structural concession matter well before the bishop pair's long-term potential can be fully realized. The resulting middlegames tend to be sharper and more concretely calculable than the Rubinstein's wider strategic latitude, since both sides know exactly what they're playing for from move five onward.

Why some Black players retreat instead of doubling the pawns

4...Bxc3+ isn't forced — Black can retreat with 4...Be7 or a similar square instead, declining to hand over the bishop pair at all. This sidesteps the whole structural argument the Sämisch is built around, at the cost of having spent moves on a pin that achieved nothing concrete once abandoned. Players who choose this path generally do so because they judge White's resulting central space and piece activity, without any accompanying structural weakness to attack, as more dangerous in practice than the doubled-pawn version most theory considers the main line — a legitimate, if less common, strategic preference.

• 4...Bxc3+ 5.bxc3 c5 — the main line, immediately pressuring the doubled pawns before White completes development.
• 4...Bxc3+ 5.bxc3 b6, developing more slowly and aiming pressure at the light squares instead of an immediate pawn break.
• 4...Be7, declining the trade entirely and avoiding the doubled-pawn structure altogether.

Failing to actually use the central space and bishop pair the doubled pawns were traded for — playing too passively and letting Black's ...c5 pressure simply erode the structural weakness without a fight — is where White most often squanders the whole point of 4.a3. Black's own error is mechanical rather than passive: attacking the doubled pawns by trading pieces without regard for whether each trade actually favors the attacker, when in fact some piece trades help White's remaining bishop and central pawns more than they help Black's structural pressure.

The variation is named for Friedrich Sämisch, whose broader body of work across multiple openings favored exactly this kind of early structural commitment — force the question, accept the consequences, and play a position both sides understand clearly from an early stage rather than one that stays ambiguous for a long stretch of moves. That philosophy shows up again in the King's Indian line carrying his name, where a comparable early pawn-move commitment (f3 rather than a3) similarly trades development speed for a clear, if demanding, long-term plan.

For White's more flexible, less committal answer to the same pin, see the Rubinstein Variation. The Nimzo-Indian Defense overview covers White's full range of fourth-move choices.