Backward time travel creates an immediate logical trap known as the grandfather paradox: if you could travel into the past and prevent your own grandfather from ever having children, you would never be born, which means you could never have traveled back to prevent it in the first place. Physics has two genuinely serious, entirely different proposed answers to this, and neither one involves anyone slowly fading out of old photographs.
The paradox’s classic form imagines a person who builds a time machine, travels to the past, and kills their own grandfather before their parent is ever conceived. If the grandfather dies too soon, the parent is never born, which means the traveler is never born either, which means there’s no one left to build the time machine and travel back in the first place. It’s a clean, closed logical loop with no obvious way out, and it’s been a staple of both serious physics and pop culture for decades. Back to the Future dramatized it by having Marty McFly slowly vanish from a family photograph as his own existence came under threat, a vivid image, but not one physicists actually predict would happen if the paradox were ever real.
The Many-Worlds Answer
The most widely discussed resolution comes from the Many-Worlds Interpretation of quantum mechanics, first proposed by physicist Hugh Everett in 1957. In this interpretation, every quantum event with multiple possible outcomes doesn’t collapse into a single result, it branches, with each outcome playing out in its own separate, equally real universe. Applied to time travel, the implication is straightforward: if you travel to the past and kill your grandfather, you haven’t erased your own history. You’ve stepped into a branch of reality that splits off from your original timeline at the moment you arrived. Your grandfather dies in that branch, and a version of you is never born there, but the timeline you actually came from continues on, completely undisturbed, exactly as it always was.
Under this reading, the traveler doesn’t disappear. They simply become a permanent resident of a new branch of reality they created by arriving in it, unable to return to the exact timeline they left. It’s an elegant solution to the paradox, and the underlying mathematics of quantum mechanics doesn’t rule it out. The catch is that the Many-Worlds Interpretation, despite having serious, longstanding support within physics, remains an interpretation rather than a directly tested theory. No experiment has ever detected another branch of reality, and there’s currently no known way to design one that could.
What a Single Timeline Would Require
If there’s only one timeline, with no branching allowed, the paradox has to be resolved a different way. This is where the Novikov self-consistency principle comes in, a proposal developed by Russian astrophysicist Igor Novikov in the 1980s to address time travel scenarios permitted by certain solutions to Einstein’s general relativity, specifically ones involving what physicists call closed timelike curves. Novikov’s principle states, in simplified terms, that any event that would create a paradox in a single timeline simply has zero probability of happening. The traveler can go back, but the universe will not permit them to succeed at anything that would erase their own ability to have traveled there.
That’s a compact way of stating the rule. What it actually looks like in practice, especially once you try to work out exactly how the universe would prevent the paradox without simply freezing the traveler’s gun mid-trigger, is much stranger, and considerably harder to formalize. That’s the specific problem computer scientist Doron Friedman set out to explore.
Testing Self-Consistency With a Computer
In 2016, Friedman, then at the Interdisciplinary Center in Herzliya, Israel, published a preprint on arXiv describing a computer program designed to explore exactly this question. It’s worth being clear about what kind of research this is: it’s a computer science paper using automated logical reasoning, not a physics experiment, and as of its publication it hadn’t been through formal peer review. Friedman built a simplified version of the paradox in which a son travels back in time specifically to kill his own father, before the son’s own conception, and then asked his program to search for any scenario that remained internally, logically consistent under those exact rules.
Working through thousands of possible variations, the program surfaced two solutions Friedman highlighted in the paper. The first requires the son to become his own grandfather: after killing his father, he fathers a child with his own mother figure, and that child eventually grows up to become the very father he killed, closing the loop through what Friedman himself described as an unavoidably strange, incestuous structure. The second solution is less unsettling but requires the father to also have access to time travel. In this version, before the son ever arrives with lethal intent, the young father briefly jumps forward in time, conceives his child in the future, then returns to his original moment in the past, where the son, arriving from later, kills him, unaware the pregnancy has already been secured.
Neither solution is elegant, and Friedman didn’t claim to have proven anything about real physics. What his program demonstrated is narrower and still genuinely interesting: that within a single, non-branching timeline governed by strict logical consistency, solutions to the paradox do exist, they’re just considerably weirder than simply preventing the traveler from pulling the trigger.
What’s Actually Established
Backward time travel remains firmly theoretical. No experiment has demonstrated it, and serious open questions remain about whether the physics that would permit it, closed timelike curves under general relativity, could ever be physically realized rather than existing only as a mathematical solution. What both approaches to the grandfather paradox share is a rejection of the cinematic version, the idea that reality would simply glitch, erasing the traveler on the spot. Whether through branching into the Many-Worlds framework or through the stranger, more constrained loops the Novikov principle and Friedman’s simulations both point toward, the mathematics consistently suggests the same thing: a working time machine wouldn’t let you break causality. It would just make sure you became a permanent, structural part of whatever consistent history resulted from your trip.
