Two people can look at the same fact — you were born to these two people, in this house, in this year — and draw the arrow in opposite directions. The ordinary arrow: parents, genes, and circumstance produced you; you are the output. The reversed arrow, the one that sounds like a trick: your arrival is the point from which "parents," "circumstance," and even "before" first get organized into a line at all — so in a real sense you are what makes them count as your beginning, not the other way round.
Posed like that, the reversed arrow reads as a parlor trick — flattering, unfalsifiable, the kind of thing that feels profound at 2 a.m. and evaporates by breakfast. It deserves better, because underneath the mysticism there are three genuinely separate, checkable questions, and the honest answer to each is different, more interesting, and more limited than either the ordinary or the reversed slogan admits.
Question one: who actually shapes whom, developmentally?
For most of the twentieth century, developmental psychology ran on an unexamined assumption: influence flows one way, parent to child. Richard Bell's 1968 paper reread the existing socialization literature and found the data didn't support that assumption — much of what looked like "parenting shapes the child" was at least as consistent with "the child's own characteristics shape the parenting." A temperamentally difficult infant gets handled differently than an easy one; a curious, high-energy toddler pulls a different kind of attention than a withdrawn one. The correlation between parenting style and child outcome had always been read in one direction only, with no principled reason to.
That reframe hardened into a research program. Sandra Scarr and Kathleen McCartney's 1983 theory of genotype → environment effects distinguished three ways a child's own heritable traits shape the environment it receives: passive (parents who share genes with the child also shape the home, so gene and environment are correlated from birth), evocative (the child's traits evoke specific reactions from others — a sunny baby gets smiled at more, a colicky one gets a more frazzled household), and active (as children grow, they select and construct the niches they occupy). Robert Plomin and C. S. Bergeman pushed the empirical test further in 1991, using twin and adoption designs to ask a blunt question: how much of the variance in supposedly "environmental" measures — parental warmth, parental control, the emotional tone of a household — is itself under genetic influence, mediated through what kind of child evokes what kind of parenting? The answer, replicated across many such measures since, is: a real, non-trivial share of it. The standard instrument that assumes the arrow runs one way — a questionnaire asking "how warm are your parents" — is partly, quietly, measuring the child who is answering it.
None of this says a child invents its parents. It says something narrower and better-supported: the parents you experienced were, in a measurable proportion, co-authored by the temperament you brought into the relationship before you had any language for what a "parent" even was. The line from them to you was never one-way traffic to begin with.
Question two: whose record is it, anyway?
Set biology aside and ask about the story instead — the account of who your parents were and what they made of you, the one you actually carry around and tell. Here the science is older and, if anything, more unsettling. Frederic Bartlett's 1932 remembering experiments — subjects retelling a folk tale, "The War of the Ghosts," across repeated retellings — showed memory isn't a filed transcript; it's a reconstruction, rebuilt on each retrieval to fit the rememberer's current schema, smoothing out what doesn't fit and importing details that were never there. Elizabeth Loftus's decades of misinformation-effect research push the same point further: post-event information, even a leading question's choice of verb, measurably alters what a person later reports having witnessed — and, in a subset of her studies, entirely fictional childhood events could be implanted and then sincerely "remembered" as real. At the neurobiological level, Karim Nader, Glenn Schafe, and Joseph LeDoux's foundational 2000 work on memory reconsolidation shows that recalling a memory briefly returns it to a labile, rewritable state before it re-stabilizes — meaning the physical trace of "what happened" is measurably altered by the act of remembering it, not only by the passage of time.
Layer developmental timing on top of that. The capacity to hold a continuous, dated, first-person narrative — the thing that lets an adult say "when I was four, my father..." — isn't present at birth. It comes online gradually across early childhood, which is exactly why almost no one retains verifiable episodic memories from before roughly age three or four, a robust and much-studied phenomenon (childhood amnesia; Patricia Bauer's developmental work traces its onset directly to the maturing of the narrating self). Whatever happened to you before that narrator existed reaches you only as secondhand testimony — photographs, family stories, your parents' own retellings — assembled after the fact by the very observer whose existence the story is supposedly explaining. Dan McAdams calls the resulting structure "narrative identity": not a record you consult, but a story you keep authoring, revising its meaning (rarely its raw facts) every time your present circumstances give you a reason to look back.
Put the two findings together and the reversed arrow gets a second, independent, well-evidenced leg: the "before you" that supposedly precedes and explains you is, as far as you can ever actually check, a document with your own fingerprints on every page, written and rewritten from wherever you currently stand.
Question three: what does physics actually license here?
This is the question that tempts people into real overreach, so it's worth being exact. John Wheeler's delayed-choice thought experiment, and its real laboratory versions — Kim, Yu, Kulik, Shih, and Scully's 1999–2000 delayed-choice quantum eraser, and Jacques and colleagues' 2007 experimental realization of Wheeler's original proposal — arrange things so that an experimenter's choice of how to measure a photon is made, in the lab's own clock time, after the photon has already, in the ordinary sense, passed through the earlier part of the apparatus. The measurement choice does correlate with which pattern the data show once everything is compared afterward.
Here is the part popular treatments routinely mangle: nothing is sent backward in time, no signal or usable information travels from the later choice to the earlier photon, and no experimenter can use this to change a past event — the no-signaling theorem of quantum mechanics forbids exactly that, and every delayed-choice experiment respects it. What actually happens is subtler and more interesting: the pattern only appears when you later sort the full data set by which measurement was chosen, comparing complete records after the fact. There is no moment at which the past changes; there is a moment at which what counts as "the relevant past" for a given later question gets fixed by the asking of that question. Wheeler's own phrase for it — "the present elects the past from among various possible pasts" — is exactly this narrower, licensed claim, not retrocausal magic.
That distinction, real and replicated but much narrower than the pop-science version, is precisely the shape the ODTOE formal answer takes as well.
The proven part: time is born, not entered
ODTOE's core claim, in one line, is that realized reality is a function of an observer applying an operator to a field of open configurations: R = Ô(Ψ). The question of who comes first — the operator or what it operates on — is exactly the same chicken-and-egg shape as "did I make my parents or did they make me," just moved one level down, into the mathematics of the observer itself: the operator Ô is needed to actualize a configuration, but the configuration seems needed first to define the operator. A 2026 composite theorem in the ODTOE corpus dissolves that exact shape for the observer-operator by exhibiting an explicit, one-directional chain: a symmetric field with no observer yet in it, disturbed by a small fluctuation, filtered by a stability selection that picks out one orientation, closing into a fixed configuration — and only then, as a consequence of that closure rather than a precondition for it, does an ordering (a "before" and "after") and a minimal spatial structure appear. Nothing in the chain looks forward to the observer it will produce before that observer is actually produced.
That result licenses something real but narrower than it sounds: the split itself — the operation that lets a biography read as a line running from parents to you — comes into existence together with you, not before you. It does not license erasing or rewriting what the split then labels "past." The same corpus proves the opposite, in a separate theorem this composite result explicitly builds on (the V∗ theorem on temporal asymmetry): whatever lands in the "past" half is conserved, never overwritten, by construction. So the honest formal statement mirrors the honest physics statement above: the observer's birth does not rewrite what came before; it is the moment at which "before" and "after" first become a meaningful pair of labels to apply to anything at all — including the two people you call your parents.
The unproven part, labeled as such
There is a further, tempting step, and the corpus itself refuses to call it proven. A separate, explicitly flagged conjecture asks whether something like a goal-shaped gradient in the observer's own coherence can select which of the many mathematically self-consistent life-lines actually gets realized — whether, among every biography that is logically available, the one that plays out is nudged toward one that fits the observer's own developing sense of direction. If true, this would be the closest formal cousin to "you chose this particular story, these particular parents, on purpose." The corpus states plainly that this piece is a conjecture, not a theorem: it comes with an explicit list of what it does not claim (no rewriting of the conserved past, no uniqueness of the resulting life-line) and a short list of falsifiable predictions rather than a proof. That is the correct place to leave it — an open, testable research question, not a settled answer, exactly as the honest reading of quantum delayed choice and the honest reading of child-effects research both refuse to overclaim past what their data show.
So: did you create your parents?
Not in the sense that would actually settle a bar bet — nothing here licenses the claim that your birth caused your parents' atoms, or their earlier life, to exist. Every line of evidence above explicitly conserves that: the twin studies still need parents with real genomes to correlate against; the memory research still requires that something happened, prior to your retelling of it; the physics forbids sending information backward; the proven ODTOE theorem proves the past is unwritten-over, not unwritten.
But three separate, well-evidenced, much narrower claims survive. Developmentally, a measurable share of the parenting you received was evoked by the temperament you brought to the relationship, not supplied to you ready-made. Narratively, the meaning — and, more unsettlingly, a good deal of the remembered content — of who your parents were and what they made you is something you are still actively authoring from wherever you currently stand, not a transcript you're allowed to consult. And structurally, per the one part of ODTOE's own account that has actually been proven rather than conjectured, the temporal architecture that makes "they came first" a sentence with any content at all is born together with you as an observer, not handed to you beforehand.
Why the reframe earns its keep anyway
None of that is decoration. Narrative therapy — Michael White and David Epston's founding 1990 formulation of it — is built almost entirely on the clinical observation that re-authoring the story of where you came from, without changing a single fact in the historical record, measurably changes what a person can do next. Robert Neimeyer's work on meaning reconstruction after loss makes the same point from a different clinical door: it is the narrated relationship to the past, not the past itself, that moves. That is the actual, checkable payoff of the reversed arrow — not a retrocausal loophole, but a documented lever. You did not create your parents. You are, right now, still the only place from which the sentence "my parents made me who I am" gets written, and rewriting it is one of the few genuinely evidence-backed forms of authorship you were never told you had.