关于√2无理性：是否存在非反证法证明及更优证明方法？

Great questions—let’s break these down one by one with concrete, easy-to-follow examples.

1. Are there proofs of √2's irrationality that don’t use proof by contradiction? #

Absolutely! The classic contradiction proof is everywhere, but there are several direct or constructive-style proofs that skip assuming the opposite first:

• Geometric Infinite Descent
  Imagine an isosceles right triangle with integer-length legs a and integer hypotenuse b (so 2a² = b² via the Pythagorean theorem). On the hypotenuse, mark a segment of length a from one vertex, then draw a perpendicular from that mark to the adjacent leg. This creates a smaller isosceles right triangle with legs b - a and hypotenuse 2a - b—both positive integers, and smaller than the original triangle’s sides.
  You could repeat this process forever, generating infinitely many smaller integer-sided triangles. But positive integers can’t shrink infinitely—this is a direct demonstration that no such integer pair (a, b) can exist, so √2 can’t be written as a ratio of integers. Unlike the classic contradiction proof, this builds a concrete infinite sequence to show impossibility, rather than starting with a false assumption.

• Continued Fraction Expansion
  The continued fraction representation of √2 is [1; (2)]—an infinite repeating sequence of 2s after the initial 1. A core property of rational numbers is that their continued fraction expansions always terminate (e.g., 3/2 is [1; 2], a finite sequence). Since √2’s continued fraction is infinite, it must be irrational. This is a direct proof: we use a defining property of rationals to rule out √2 being rational, no contradiction needed.

• Prime Factor Counting (Non-Contradiction Framing)
  Instead of starting with "suppose √2 = p/q", we can rephrase the argument to be direct: For any integers p and q (q ≠ 0), p² = 2q² is impossible. Why? By the Fundamental Theorem of Arithmetic, the exponent of 2 in the prime factorization of p² must be even (since squares have even exponents for all primes), while the exponent of 2 in 2q² is odd (even from q² plus 1 from the extra 2). Since prime factorizations are unique, these two can’t be equal. This avoids explicitly stating "suppose √2 is rational" and instead directly proves no integer ratio can satisfy the equation.

2. Are there "better" proofs than the classic contradiction method? #

"Better" depends on what you value—simplicity, intuitiveness, generalizability, or extra insight. Here are a few candidates that outperform the classic proof in specific ways:

• Prime Factor Counting Proof
  This is incredibly concise, relies on a foundational number theory result (the Fundamental Theorem of Arithmetic), and easily generalizes to prove that √n is irrational for any non-square integer n (just replace 2 with n and look at the exponent of a prime that divides n an odd number of times). It’s a workhorse for extending the result beyond √2.

• Geometric Infinite Descent
  If you’re a visual learner, this proof is far more intuitive than manipulating equations. It turns an abstract number theory problem into a concrete geometric construction, making the "impossibility" feel tangible rather than just a logical contradiction.

• Continued Fraction Proof
  This proof does more than just show irrationality—it gives you a way to generate the best rational approximations of √2 (like 3/2, 7/5, 17/12) from the repeating continued fraction. It connects irrationality to a deeper property of √2’s structure, giving you extra mathematical insight alongside the proof.

• Minimal Element Argument
  Suppose there exists a smallest positive integer q such that √2 q is integer. Let p = √2 q, so 2q² = p². Then p is even, so let p = 2k—substituting gives 2q² = 4k² → q² = 2k². This means √2 k is integer, but k = p/2 < q (since p = √2 q < 2q because √2 < 2). This contradicts q being the smallest such integer. While this uses contradiction, it’s more focused on the well-ordering principle (every non-empty set of positive integers has a smallest element) and feels more structured than the classic proof, making it easier to follow for some.

内容的提问来源于stack exchange，提问作者zeraoulia rafik