152 explorations. The ones near the surface have substantial answers with edges. The ones at the bottom may not have answers in a form we would currently recognise.
Things we understand reasonably well.
7 questions
There are precisely stated questions that no algorithm can answer, and the proof does not depend on the technology. What remains open is which physical processes can be simulated, and whether the universe itself is computable.
Remote Oceania was settled from the west in a series of voyages ending about eight hundred years ago. Genomes of Pacific islanders, living and ancient, also record contact with Native Americans centuries before any European arrived. Two independent studies found it; they differ on when, with whom and how often.
The impact at Chicxulub is established, dated to the boundary layer, and a 2010 review of the evidence concluded it triggered the extinction. The Deccan eruptions were also underway. The remaining argument is about how much the volcanism contributed and whether the impact changed its behaviour.
The lightest elements were made in the Big Bang and most of the rest inside stars, but about half of those heavier than iron, gold among them, need conditions no ordinary star provides. One event in 2017 was seen in gravitational waves and in light, and showed those conditions producing heavy elements directly. How much of the total comes from that source is still argued, on this page as in the field.
The major divisions in Christianity are conventionally dated to 451, 1054 and the sixteenth century, and each time the doctrinal question came bundled with a question about who had the authority to answer it.
Noticing depends on attention, and attention is a limited resource allocated in advance. What falls outside it can go unreported entirely, even when people are asked about it directly. Whether it also went unseen, or was seen faintly and never reported, is still argued.
Cooperation is not a puzzle without an answer. Five mechanisms are understood well enough to be stated mathematically, each specifying the conditions under which helping pays. What remains argued is which of them explains human behaviour, and whether group selection explains anything kin selection does not.
Interesting questions with substantial evidence and real uncertainty remaining.
52 questions
The question sounds like a matter of fact and is largely a matter of definition, which is why it has not been settled. What is genuinely open is where viruses came from: out of cells, from something older than cells, or from both.
Human activity has left a distinct and worldwide layer in sediments and ice, and the case that it is stratigraphically distinguishable from the Holocene has been made in detail. A formal proposal to define an Anthropocene epoch was rejected in 2024, and the timescale still ends with the Holocene.
Decipherment has needed two of three things: a known language, a bilingual text, a large corpus. The famous unread scripts each lack at least two of those, and for the Indus signs there is a serious argument that they do not encode language.
Seismic hazard maps are quantitative and useful, and short-term prediction of a specific earthquake has never been demonstrated. The distinction between forecasting a rate and predicting an event is the whole subject, and where it is not made clearly, informal predictions fill the gap, as they did in L’Aquila before the earthquake of 2009.
Ancient DNA can be sequenced, but a sequence is not a working genome, and no mammoth cell exists to clone. What is feasible is editing a close relative’s genome towards an extinct one, which produces a hybrid with some of the traits and none of the lineage.
Machine learning applied to brain imaging can reconstruct the meaning of continuous language from a cooperative subject, once the decoder has been trained on that person’s own recordings. It produces paraphrase rather than transcript, carries over to a new person only with some of that person’s recordings, and can be resisted by the subject at will.
Information from one sense can be delivered through another. In people born blind, a region that specialises in reading takes up the work when letters arrive as sound. The team that showed it reads that as cortex organised by what it does rather than by which organ feeds it; another group, finding the same region in blind people responding to spoken grammar as well, argues that cortex without vision can take on a wide range of functions.
Memory is partly a reconstruction rather than a replay, which means it can be edited by what happens after the event, including by the questions asked about it. Under persistent suggestion a minority of people come to remember whole events that never happened, some of them serious. Estimates of how large a minority run from about 15 per cent to about 30, depending on what is counted as remembering.
In 2007 a group proposed that an airburst or impact 12,900 years ago contributed to an abrupt cold period, the extinction of North American megafauna and changes in human behaviour at the end of the Clovis period. A 2023 review of the evidence concluded the hypothesis should be rejected. Its proponents rejected the review, and continue to publish.
In a letter preserved under his name, a hostile emperor, trying to revive the old religion, ordered a high priest to set up hostels because the Christians were supporting the pagan poor as well as their own, and it was winning them converts. In the same sentence he says that no Jew ever has to beg, so the Christians were not the only ones looking after their poor.
Information is not an abstraction floating above physics. Erasing it has an unavoidable thermodynamic cost, which sets a floor under any computation that throws information away and is the standard answer to a paradox about the second law that had stood for more than a century.
Randomised trials of intercessory prayer for strangers have found no clear effect, and the largest prospective one found more complications in the group told they were being prayed for. What that settles, and what it does not, is worth being precise about.
The strong claim that language determines thought is not supported. A narrower one has been reported: the categories a language marks speed up the discriminations that cross them, and the effect shrinks or disappears when the language system is loaded during the task. The Russian blues demonstration did not reappear when another group repeated it.
No original manuscript of any biblical book survives. What survives is copies, in very large numbers, and comparing them is a mature discipline whose results are printed in the margins of ordinary Bibles.
Thermal expansion and melting glaciers are understood well enough to project. Antarctica is not: whether its marine-based ice can collapse rapidly is an open question whose answer separates a manageable rise from an unmanageable one, and it cannot currently be ruled out.
About 2.4 million species of animals, plants, fungi and other eukaryotes had been formally described by August 2026. Estimates of how many exist range from under 10 million to numbers that make the word “species” difficult to apply at all.
Thousands of planets have been confirmed around other stars, and the detected ones are a biased sample of a much larger population. Estimating the true number means correcting for what the methods cannot see, which is where the interesting arguments are.
Most of the seafloor has never been mapped at high resolution, and a far smaller fraction has ever been observed directly. The gap between “surveyed” and “seen” is the interesting part.
Vision is not a camera. Some of what reaches your awareness is filled in rather than measured from light, and the filling-in is invisible from the inside. How much of it is prediction is still argued.
A human body carries roughly as many bacterial cells as human ones, a figure that corrected a much-repeated estimate. What those bacteria do to health, mood and development is a field with many correlations, a few demonstrated mechanisms, and a large gap between the two.
DNA recovered from the teeth of Bronze Age skeletons carries Yersinia pestis, five thousand years old and lacking the gene that lets the later strains spread through fleas. The history of a disease can now be read directly rather than inferred from descriptions of symptoms.
The number of Christians is rising and the share of the world that is Christian is falling. Where they live is changing fast, and much of what else is claimed about it depends on which country the claimant is looking from.
Reversals are recorded in the rock and the current field is measurably declining, which is why the question keeps being asked. The decline has happened before without a reversal following, and no method exists for forecasting one.
A key part of the case for an undiscovered planet several times the mass of Earth is the alignment of a small number of distant objects. Whether that alignment is real depends on how the surveys that found them were pointed, and that question has been argued both ways in print.
Documented extinctions since 1500 are far fewer than in the five great mass extinctions. Measured as a rate against the background from the fossil record, vertebrate losses over the last century are up to a hundred times higher, and the comparison depends on assumptions that are stated openly and argued about.
Before Constantine, Tertullian and Origen both argue that Christians should not take up arms, though Tertullian shows that some did. A century after his conversion, Augustine is telling a Christian officer when and how he may. Both sides quote scripture, and both are on the record.
Between 1950 and 2002, four per cent of Catholic priests in the United States had allegations of child sexual abuse against them, from a study the bishops themselves commissioned. Inquiries in Ireland, Australia and France found comparable patterns. The institutional response, for decades, was transfer and silence. What follows is about the Catholic Church, because that is what the commissioned research and most of these inquiries examined.
A church manual, a defence addressed to an emperor, and a Roman governor’s report to his superior describe, between them, a small set of practices: a meal, a washing, a fixed day, and money for the poor.
Christian missions accompanied European colonisation for five centuries, sometimes as its instrument and sometimes as a witness against its cruelty. Both of those are in the record, in the same decades, occasionally in the same person.
A site in south-eastern Anatolia forced a rewrite of the standard sequence: large-scale construction came before farming, not after it. By the excavators’ count in 2012, most of its enclosures were still buried.
The disagreement is not between people who believe the book and people who do not. It is between traditions with different and published accounts of what kind of claim the book is making, and the differences are checkable.
For a few minutes the whole universe was hot and dense enough to fuse nuclei. The theory predicts how much helium, deuterium and lithium should have come out, using one free number. The helium and the deuterium match observation to remarkable precision. The lithium does not.
The island did lose its palm forest, and the standard account, that deforestation caused a population crash and societal collapse before Europeans arrived, has been tested against radiocarbon dates and found wanting. What replaced it is a story of adaptation followed by catastrophe from outside.
Redating of European sites places the disappearance between about 41,000 and 39,000 years ago, overlapping with modern humans for a few thousand years. The genome shows interbreeding. Why one population persisted and the other did not is not established.
More than twenty definitions of species are in use, and they disagree about real cases. A 2007 proposal reconciles most of them by separating what a species is from how one is recognised, which clarifies the disagreement without eliminating it.
A star’s ending is set mostly by its mass. White dwarfs and neutron stars are observed by the thousand. The line between a neutron star and a black hole is where nuclear physics is least certain.
The behaviour of lightning is documented in extraordinary detail, down to radio images of the first microseconds of a flash. What starts it is not established: the fields measured inside thunderclouds are lower than most proposed mechanisms require.
About two thirds of the genera of large land animals disappeared during the late Quaternary, continent by continent, at different times. A global analysis finds the pattern matches how long people had been present far better than it matches climate change, and the argument is not over, because the two are correlated in places.
Liquid water exists under the Antarctic ice sheet, in lakes and in water moving along the bed beneath it. A 2013 expedition drilled into one such lake with a microbiologically clean drill and found a working microbial ecosystem living on minerals in the dark.
The inquisitions were courts, staffed by clergy, tasked with identifying and punishing heresy. They kept detailed records, which is why historians can say what they did rather than what people said they did. The revision that followed those records corrected the numbers and did not make the institution defensible.
The consensus that people arrived about 13,500 years ago began to collapse in the late 1980s and 1990s and has not been replaced by a firm number. Footprints dated to the height of the last glaciation would put people in North America thousands of years before the route they were supposed to have used was open.
Human head lice and clothing lice are two forms of one species, and clothing lice live only in clothing. Dating their divergence gives a date by which the habit existed, and the 2010 estimate is of the order of a hundred thousand years, before the migration out of Africa.
Water carries a fingerprint: the ratio of heavy hydrogen to ordinary hydrogen, which depends on where and how cold it formed. Measuring that ratio in Earth, in asteroids and in comets has narrowed the candidates and has not closed the case.
A giant impact early in the solar system explains the Moon’s size, the spin of the Earth-Moon pair and the Moon’s lack of iron. Its standard version also predicts that most of the Moon should be made of the impactor, and the samples say otherwise. Several fixes exist and none is agreed.
Plastic was estimated in 2015 to enter the ocean at millions of tonnes a year. Measurements of what floats at the surface accounted for a small fraction of that. Later work closes the gap from opposite ends: one model puts the input far lower, while sampling below the surface finds far more plastic than the nets ever showed. Where the plastic is remains an open problem, with different consequences depending on the answer.
There are three defensible answers, and the outermost lies hundreds to thousands of times further out than the innermost. One is where the planets stop, one is where the Sun’s wind stops, and one is where the Sun’s gravity stops holding on. Spacecraft have passed the first and two have crossed the second. The third is tens of thousands of years away at Voyager 1’s speed.
Ancient DNA has revealed at least one human population unknown from fossils, and traces of others inferred from genomes alone. For hundreds of thousands of years several kinds of human lived at the same time, and the present situation, with only one, is recent.
Anyone reading this book in English is reading a set of choices. The choices are documented and several of them change what the passage is about.
Phantom limb pain is common after amputation and is produced by the nervous system rather than by tissue at the site. The leading explanation has been a reorganisation of the brain’s body map. The pain does track measurements of that map, but whether the map reorganises at all after amputation is now disputed, and so is where the pain starts.
Some placebo effects are physiological events with identifiable brain mechanisms, and effects on what patients report survive being told that the pill is inert. What they are responses to, and which conditions they touch, is narrower than the popular version and stranger than it.
A seventeen-year survey of the water column off California classed about three quarters of the animals it could assign as probably able to produce light; at every depth sampled these outnumbered the animals classed as unable, though up to 43 per cent at a given depth could not be assigned either way. What each of them is using it for is known in a few cases and assumed in most.
That he was crucified under Pontius Pilate is told by all four Gospels, and a hostile Roman source records the execution under Pilate too. What he did to arrive there is argued, and the accounts that describe it were written by people with an interest in where the blame fell.
Major unanswered questions.
68 questions
The Standard Model and cosmology contain numbers that no theory predicts. Whether they vary over time has been tested by quasar spectra, by the isotopes in a two-billion-year-old natural reactor, and by atomic clocks. Nothing has been established to vary, and the limits are tight.
Measurements on entangled particles are correlated more strongly than any theory in which each particle carried its answer with it could produce. In 2015 the experiment was finally done with the detection and locality loopholes closed at once, and the correlations held.
Christians were singing to Christ as to a god by about 112, and Paul, decades earlier, seems to quote an existing hymn about him. What the man himself said, in Aramaic, before others wrote it down in Greek decades later, is a harder question.
The charge is more than a century old and it has a real case behind it. So does the answer, which is that some of the oldest material in his own letters is material he says he was handed.
The story has a whole nation leaving Egypt and entering Canaan from outside. The archaeology of the highland settlements it would have produced reads the population as indigenous to the Levant, which leaves a genuine argument about what the story is and where it came from.
The expectation is in the earliest letters in the first person. Whether it was a mistake, a misreading, or a different kind of statement is an argument that starts inside the New Testament.
The mirror mark test was designed to detect self-recognition. Chimpanzees and orangutans pass it, and passes have been claimed for elephants, magpies and dolphins. When a small cleaner fish appeared to pass it too, its discoverers put the choice plainly: grant the fish self-awareness, or accept that the test measures less than it claimed. Critics took a third way, and said the fish had not passed.
Interference experiments have been pushed from electrons to metal clusters of thousands of atoms without encountering a boundary. Whether the limit is merely technical, or whether something physically destroys superpositions above some mass, is a question being decided by pushing further.
A biosphere exists inside the crust, in rock and sediment, running on chemistry rather than sunlight and, where energy is scarce, at metabolic rates so slow that its biomass turns over in centuries to millennia. Its extent and its total mass are estimated from sparse sampling, and the estimates are wide and have been revised sharply downward.
The legal turn is documented and fast. Why there were enough Christians by then for it to matter is the part nobody can settle, because the numbers do not exist.
Life appeared on Earth remarkably early, and we can reproduce several of the chemical steps that might have led to it. What we cannot do is explain how those steps joined into something that copies itself.
Magnetic sensing in migratory birds is demonstrated behaviourally beyond doubt. The mechanism has a leading candidate involving quantum-scale chemistry in the eye, and the sensory cells themselves have not been found.
The genome does not contain a plan of the body. Pattern arises from cells reacting to each other and to chemical gradients, and a mathematical account of how uniform tissue produces regular structures was published in 1952 and matched to mouse limbs sixty years later.
Estimates for the origin of language run from a sudden appearance fifty to a hundred thousand years ago to a capacity already in place half a million years ago. A 2013 review argues from Neanderthal anatomy, genetics and behaviour that the capacity is much older than the recent estimates allow. A 2019 review of primate vocal anatomy goes further, putting what it calls the dawn of speech more than twenty million years back.
Written history covers a small fraction of human existence, and what survives from that fraction is filtered by material, climate, politics and luck. The size of the gap can be bounded but not measured.
Expansion is accelerating, which was discovered in 1998 and was not expected. If the cause is a constant, the universe thins and cools forever. If it is a field that changes, several other endings become available, and recent survey data mildly prefer a changing one.
In thirty years the question moved from “do other planets exist?” to “which of the thousands we have found should we look at?” No detection of life beyond Earth has been confirmed.
Classical randomness is ignorance about details. Quantum randomness appears to be different in kind, and a 2018 experiment generated bits that no theory forbidding faster-than-light signals could have predicted, provided the measurement choices were made independently of the devices. The physics inside the devices is not assumed; the timing electronics and computers around them are trusted.
A systematic review of the evidence for lowered serotonin in depression found no consistent support for it. Separately, large analyses find antidepressants more effective than placebo by a modest margin. Both are true, and holding them together is the whole of the honest answer.
In quantum field theory the vacuum is the lowest energy state of a set of fields, not an absence. It has measurable consequences, including a force between nearby plates. It also has an energy density, and the value that theory predicts for it is wrong by some fifty orders of magnitude on a careful estimate and 120 on the simplest.
Twice in the Cryogenian period, glaciers reached sea level in the tropics. Whether the ocean froze over completely, or kept a belt of open water, is still argued, and the difference matters because the first version is hard to escape from and hard to survive in.
Variations in Earth’s orbit and tilt change how sunlight is distributed, and the record of past ice ages carries those periods clearly. What the orbit cannot explain is why the dominant rhythm for the last million years is one hundred thousand years, when the orbital forcing at that period is weak.
The New Testament regulates slavery and does not condemn it. What Christians then did with that, over nineteen centuries, runs in both directions, and the texts did not settle it.
The first three Gospels are consistent about the subject: a kingdom, mostly in stories, with an ethic sharpened as far as a demand to be perfect. What is argued is which sayings are his, and what he meant by the phrase that holds all of them together.
Jesus is mentioned by Roman and Jewish writers within a century of his death, and by Christians within about twenty-five years. What those texts establish, and what they cannot, are two different questions.
The tension is inside the New Testament, not between it and the modern world, and one of the sharpest pieces of evidence is a name the printed editions changed.
Faith is routinely defined, by its critics and by some of its defenders, as believing things without evidence. The documents do not use it that way, and one of them stakes the whole claim on an event it says would falsify everything if it did not happen.
The rejected texts were not hidden. Eusebius names several of them in a work of history, sets them below even the books he rejects, and explains why he is doing it. What is genuinely lost is a different set.
The claim that the apostles all died rather than recant is used as an argument, and the oldest texts cited here record deaths one at a time. What the earliest sources say is thinner and stranger than either the popular version or the dismissal of it.
The practice is on record before any Gospel was written, and the earliest descriptions of what is done are plain. What the bread is, after the prayer, is one of the questions that split Western Christianity, Protestant from Catholic and Protestant from Protestant, and the answers are written down in documents that say plainly what they mean.
We can now locate and artificially reactivate specific memories in mice. We still cannot say what a memory *is* in a way that explains why yours can change when you use them.
What we know about the inside of this planet comes mostly from earthquake waves passing through it, read against laboratory experiments. Mantle rock lies under the crust and has been sampled where tectonics lifted it to the surface or where a piece came up inside a diamond, and two continent-sized structures near the core have no agreed explanation.
Galaxies, stars, planets, dust and every particle in the Standard Model together make up roughly 5% of the energy content of the universe. The other 95% has a name, a measured quantity, and no identification.
A fast radio burst usually lasts a few milliseconds or less, and the first one found released, on its discoverers' estimate, as much energy as the Sun emits in about a month. The 2020 detection of one from a known magnetar in the Milky Way established that at least some have a plausible source, without explaining the rest.
Cutting the bundle connecting the hemispheres is done to relieve severe epilepsy, and it leaves the patient recognisably themselves. Classic experiments suggested the operation produces two separate centres of awareness. A 2017 study tested that directly in two patients and reported the opposite, and the field’s own verdict is that neither result settles it.
The council of 325 produced a creed and twenty canons, and its documents survive. Most of what it is popularly said to have decided, including the contents of the Bible, is not in them.
Comparing the genomes of living organisms lets us reconstruct features of a population that lived billions of years ago. The reconstruction is real, partial, and contested.
One of the most consequential speeches of the Middle Ages exists only in accounts composed years later by men who wanted different things from it. What the crusaders then did is far better attested, including what they did to people who were not Muslims, and including what the pope who called the fourth crusade said about it afterwards.
Human-made materials are being buried in enough quantity that some will fossilise, and with the chemical traces of the last century they make a layer distinct from what lies beneath it. Deliberate communication is much harder. The panel that designed markers for a nuclear waste site planned for languages and institutions that might not last ten thousand years, gave its markers a better than even chance of still standing in most of its estimates, and said a system that both survives and stays readable needs further study before it is built.
There is no date. There is a first-century argument about whether Gentiles had to keep the law, and a second-century dialogue in which the two are already separate things and one is writing against the other.
That the plates move now is not in doubt and is measured directly. When the process started is argued from indirect evidence in very old rocks, and proposed dates range from more than four billion years ago to less than one, depending on which signature is accepted as diagnostic.
The physics of warming is settled. Whether specific components of the Earth system have thresholds beyond which change becomes self-sustaining (and where those thresholds sit) is not.
The picture the word usually calls up is only partly in the texts, and the texts do not agree with each other closely enough to produce it. What is on the page is three different things, two of them flattened into one word by translation, and an argument about their meaning that was already old in the fifth century.
The figure changes across the collection, in a direction anybody can follow, and one of the clearest cases is a single event that two books attribute to two different agents. Most of the familiar picture, including Lucifer as a name for the devil, comes from outside the texts.
The documents show the church at Rome acting with unusual weight from very early, and they show other churches disputing what that weight was. Both of those are on the record, in the same centuries, and the disagreement has never closed.
Christianity inherited a strict monotheism, then worshipped a man and baptised in the name of Father, Son and Spirit. The doctrine is the answer worked out over three centuries to the question that left.
Charged particles from space hit the atmosphere constantly, and a rare few carry energies no accelerator on Earth can approach. Their paths are bent by magnetic fields on the way, so the sky they arrive from is not the sky they came from.
General anaesthesia is reliable, common and not understood at the level of consciousness. A person who cannot respond is not necessarily absent, and the index that can tell the difference in a laboratory is not what operating theatres use.
The universe has a finite age, so only light from within a finite distance has had time to arrive. That distance is not fourteen billion light years, because space expanded while the light was in transit, and the horizon is a limit on observation rather than on existence.
The New Testament was not selected at a single meeting. The earliest surviving list of exactly the twenty-seven books most churches now read was written in 367, and Christian traditions still disagree about the Old Testament and, at the edges, the New.
Ancient DNA has rewritten human prehistory twice in twenty years. The current picture is a braided, interbreeding set of populations rather than a line, and large parts of it are still blank.
None of the four Gospels names its author inside the text. The names we use are attributions made by the early church, and how much weight they carry is one of the longest-running arguments in the field.
Species richness rises towards the equator in almost every group on land and at sea. The pattern is among the most robust in biology and among the least explained: the leading accounts differ on whether the tropics make new species faster, lose them more slowly, or have simply had longer.
The commandment in Exodus forbids images. The tradition that kept that commandment also came to hold that God had taken a face. A council in 787 answered the collision with a distinction, in Greek, though the argument had another half century to run.
Bacteria and archaea have been evolving for billions of years, and apart from one event they show no tendency to evolve greater complexity. The eukaryotic cell appears once, apparently through one cell taking another inside it. Whether that step is hard, and why, is an open question with one quantitative answer on the table and a quantitative challenge to it.
Agriculture began separately in several parts of the world within about six thousand years of each other, after hundreds of thousands of years in which it did not. The skeletal evidence suggests it made individual lives harder. What made it happen, repeatedly, at that moment, is unresolved.
The palaces of Mycenaean Greece, the Hittite empire, Ugarit and the towns of the northern Levantine coast all fell within a few decades of 1200 BC. A drought lasting centuries is now documented in sediment cores from the same period, which gives the collapse a physical context without reducing it to one cause.
The Harappan civilisation was spread over about a million square kilometres, more ground than either Egypt or Mesopotamia, then dispersed over several centuries. Work on the river systems it depended on finds a weakening monsoon and the drying of the rain-fed rivers its heartland sat on. Whether the timing matches the cities’ decline is still argued.
What the texts say is short and checkable. What they mean is argued, the arguments are old, and the formal positions of the largest churches are published documents anybody can read. This question sets those three apart and leaves them apart.
Three of the Gospels share passages that sometimes agree word for word, and the fourth shares few of them. What the divergences mean has been argued since the second century; who came first and who copied whom was asked by Augustine, and has been argued in earnest since the eighteenth century.
One method reads the expansion rate off the infant universe, the other measures it directly in nearby galaxies. Both have shrunk their error bars. The gap between them has not closed, and that is either a mistake nobody has found or physics nobody has written down.
The evolutionary account of why ageing exists turns on selection that weakens with age and on evolved limits to somatic maintenance, rather than on any programme. The molecular account of what ageing is has a catalogue of processes that change with age, and no agreement on which of them cause it.
Dreaming is universal and measurable. There is still no agreed account of what it is for, or whether it is for anything.
Sleep is found in almost every animal and is dangerous to do in the open, which means it buys something substantial. Several candidate functions have real evidence behind them. Whether one of them is the reason, or whether the question is badly posed, is unsettled.
A second-century teacher proposed cutting the Hebrew scriptures out, and the church refused. That refusal shaped Christianity, and the way Christians then read those texts has a history with consequences.
Sexual reproduction halves the rate at which genes are passed on and requires finding a partner, so it needs a large compensating advantage. Several candidates have theoretical and experimental support. Which one does the work, and whether any is sufficient, is unresolved.
Oxygen accumulated in two great steps separated by nearly two billion years. What produced it is known. Why it rose when it did, and why it waited so long between steps, is not.
Evolution repeats itself constantly at small scales, which makes it partly predictable. It also depends on history in ways that can be tested experimentally, and the experiment that tests it best took twenty years and thirty thousand generations.
Fundamental questions where our understanding is extremely limited.
23 questions
There is no validated test for consciousness in a system unlike us, and the usual informal test, asking it, is unreliable on anything trained to produce human text. The approach a 2023 report calls the most promising derives indicator properties from theories of consciousness and checks systems against them, which converts an unanswerable question into a partly answerable one.
The texts present signs as evidence meant to persuade. The standard philosophical objection is not that miracles do not happen but that testimony could never be good enough to establish one. Hume aimed it at religion, but the argument is about the structure of evidence.
A 2003 paper by Nick Bostrom argues that at least one of three propositions holds: civilisations almost never reach the ability to run ancestor simulations, almost none that can do so choose to, or almost everyone with our kind of experience is simulated. It is a disjunction, it rests on assumptions that have been questioned in print, and its author doubts that any simple experiment could show which branch is true.
Brain activity can be measured before a person reports deciding to move, and can predict the choice a little better than chance. What that shows depends on what the activity is, and the classical interpretation has a serious alternative that is hard to rule out.
In special relativity, simultaneity depends on motion. An argument made in 1966 and 1967 turns that into a claim that past and future events are as real as present ones, which is widely discussed and not settled, because what follows from the physics depends on what reality is taken to mean.
Space appears flat to the precision anyone can measure, which is consistent with infinite extent and also with a very large finite volume. The distinction is not one that light can carry: no signal from beyond the horizon has arrived yet, and because the expansion is accelerating, from far enough away none ever will.
Combining gravity with quantum mechanics produces a length below which the usual ideas of distance appear to fail. Models inspired by the candidate theories predict observable consequences, and astronomical observations have already bounded the simplest of them past the Planck scale itself.
Quantum mechanics treats time as an external parameter. General relativity treats it as part of a dynamical geometry. Combining them produces a description in which time has no separate role, and whether that is a deep truth or a technical artefact is unresolved.
Two churches, two documents, two answers in opposite terms, both citing scripture. This is the clearest place to see what a denominational disagreement actually looks like from the inside.
The most precisely tested theory in science comes with a set of mutually exclusive stories about what it describes, and no experiment has yet decided between them.
Thirty-three physicists, philosophers and mathematicians at a foundations conference were polled on what their own theory is about, and no interpretation took a majority. A survey of more than a thousand researchers in 2025 found the same. Some candidate theories have been ruled out by measurement, which makes this a scientific dispute rather than only a philosophical one. What is left includes accounts built to make identical predictions, which no experiment can tell apart.
By the time Paul joined the movement, people who had known Jesus were saying he had been raised, and Paul later met two of them in Jerusalem. That the claim was made early is documented. Whether the claim is true is not the kind of thing a document settles.
Black holes are observed, weighed, photographed and heard. What happens past the horizon is where general relativity and quantum mechanics openly contradict each other.
Expansion is not motion outward through space. It is space itself growing between things. That resolves the question as asked, and leaves several harder ones standing.
Entire ecosystems run on chemistry instead of light. We found the first one in 1977, and we still do not know how much of the biosphere works this way.
Between the release of the cosmic microwave background and the first stars there is a long stretch with nothing to see. The signal that could open it has been claimed once and contradicted once, which is a useful demonstration of how this kind of result actually gets settled.
By Hawking’s calculation, a black hole radiates and eventually disappears. If the radiation carries no trace of what fell in, quantum mechanics is violated. If it does, there is no agreed account of how. Recent calculations suggest the information comes out, by a route that is not yet understood.
Calculations varying the fundamental constants suggest that many nearby values produce universes with no stars, no chemistry or no structure. The calculations are real and their interpretation is not settled, because the space of possible physics is not something anyone can sample.
The argument was set out in four clauses in antiquity, and philosophers now argue a newer, evidential form of it alongside the old one. No evidence has settled either the answers or the objection, which is why this page is about what each position actually costs.
We can describe in detail what the brain does when you see red. We cannot explain why any of that activity is accompanied by the experience of seeing red rather than by nothing at all.
Mathematics developed without reference to the physical world repeatedly turns out to describe it with extraordinary precision. Eugene Wigner called this unreasonable in 1960 and offered no solution. The proposed explanations since are philosophical positions rather than results.
The asymmetry of time is the most obvious fact about experience and one of the least explained facts in physics. The standard account pushes the question back to the beginning of the universe and leaves it there.
Every particle has an antiparticle, and the known laws produce them in pairs. Yet the universe is made of matter, with no sign of antimatter regions anywhere in the visible universe. The conditions an explanation would have to satisfy were written down in 1967. The known laws meet them in principle, and far too weakly in practice.
Questions where we may not even know what an answer would look like if we found one.
2 questions
Where the Hebrew Bible speaks of the dead, it mostly describes death as the end of awareness. A bodily resurrection appears late and becomes central in Christianity. The popular picture of a soul going somewhere is a third thing again, with antecedents in Greek philosophy and in later Jewish writing.
Physics can describe the early universe in remarkable detail and has nothing to say about why there is a universe to describe. It is not clear that any possible evidence would settle it.
Every passage in every exploration carries one of these.