|Image credit: A. P.|
Winter, fair as always, does not lay down the law about what a combination is, or is not, but looks for what famous players, or early sources, or readers think a combination is. Most definitions, however, seem to concentrate either on (a) sacrifice of material; (b) forcing continuations to reach a winning (or drawing) position; and (c) the use of more than one piece. (c) would imply that every time someone wins a pawn by putting more pressure on it than the opponent can rebuff it is a combination, while (a) and (b) would mean, for example, that exchanging a rook for a knight in order to simplify into a winning pawn ending is a combination. Of course, all of the above are good, even winning, moves (or plans) -- but are they really "combinations"?
I believe Irving Chernev and Fred Reinfeld had the best definition I've seen. In their view, it is not the power of pieces that is combined, but that of the tactical motifs (pins, skewer, overloading the defender, exploiting a weak back rank, etc.).
Indeed, the more such motifs are combined together, one after another, the deeper and more shocking and absurd the combination's moves seems at first glance -- and the more we understand and enjoy it once we analyze its motifs. In their words (from the book above, the 1988 reissue of the 1949 original by Faber & Faber):