I read this Robert Kagan piece "America's Dangerous Aversion to Conflict" with initial interest, followed by growing boredom. Analogies to the 1930s are such a stock part of conservative foreign policy rhetoric that it's a wonder there's no app for them. Some wording about Nixon-Kissinger foreign policy being "fatalistic" also struck a tinny note. (That policy was in fact quite activist and creative, which is basically the opposite of fatalistic.) I recently expressed my disagreements with the ill-considered and poorly argued non-interventionism at Reason, but Kagan's essay is a good reminder that neoconservatism has its own formidable blinkers, as Jacob Heilbrunn points out.
What the U.S. needs is to be smarter than its enemies, or at least not manifestly dumber, and to have what's sometimes called, a bit pretentiously, "grand strategy." That strategy should include, sometimes, making common cause with enemies, against worse or more pressing enemies, and thus making judgments about which enemies can potentially be given a more benevolent status. Insofar as the U.S. is now working with Iran, or will do so, that reflects a belated recognition that sometimes governments that hate each other still have reason to work together. In the aftermath of 9/11, that recognition went by the wayside, with unfortunate consequences spilling down to the present.
Inasmuch as we do live in a period remotely similar to the 1930s, that's all the more reason why we need to be building up alliances, de facto and de jure, including in some cases with traditional enemies. If there's anything that's "fatalistic," it's assuming that this can't be done.
UPDATE: "Obama Should Play Nixon and Go to Iran."
UPDATE: "2014 Is Not 1931," though it is true, as has been noted, that Chamberlain wasn't PM in 1931 either.