The Peer Review Machine

Peer review was designed as a filter for error, but modern science has turned it into a test of endurance. A single paper can wait eight months, pass through three journals, and collect forty pages of contradictory demands before anyone reads its results. Defenders argue that slow scrutiny protects the literature from false claims, and they can point to retractions that slipped through hurried review. Critics answer that delay itself corrupts knowledge: urgent findings on disease or climate sit unread while reviewers debate footnotes, and young researchers learn to write for referees rather than readers. Proposed reforms fall into three camps. The first demands open review, where reports and names appear beside the paper, forcing reviewers to be fair and fast. The second proposes review after publication, letting the community test claims immediately while formal judgment follows. The third would pay reviewers and count their work toward promotion, recognizing that unpaid labor cannot carry a system publishing millions of articles a year. Each fix carries a cost: openness may silence junior critics, post-publication review may flood readers with unchecked claims, and payment may invite careless approvals. The deeper question is what review is for: certifying truth, which no two referees can guarantee, or improving drafts, which requires dialogue rather than verdicts.