Five long reads on things we assume are settled: what belief is, why squares charm, what makes a father, whom we can judge, and who wins by moving fast.
Picked from 25 new articles across 14 sources · arrow keys for other days
Today in 30 seconds
30s left
A philosopher and a law professor argue religious credence is chosen and evidence-proof, and that this psychology explains Jefferson's wall. story 1 ↓
Camillo Sitte's 1889 case for small enclosed squares and off-centre monuments, buried by modernism, is winning architects back. story 2 ↓
Force an infanticidal male mouse to live with his pregnant mate and he cares for the pups; Asif Ghazanfar traces that flexibility to men. story 3 ↓
Edward Abel Smith still answers calls from Robin Garbutt, a convicted wife-killer thrice rejected by a watchdog, and can't say if he's guilty. story 4 ↓
Michael Fitzpatrick traces Japanese inventions that went nowhere, from contactless cards to neural networks, and asks whether slow adoption was the wiser course. story 5 ↓
AI is solving maths problems that stumped mathematicians for decades, and the field is turning to Terence Tao for guidance.
1/5Aeon· Neil Van Leeuwen & Jacob Eisler· Nov 3
Two scholars ground church-state separation in the psychology of religious belief
Neil Van Leeuwen and Jacob Eisler argue that religious conviction is a different mental attitude from factual belief, and that this difference explains why democracies keep religion out of policy yet protect its private practice.
Why it matters · It replaces a familiar slogan with a testable claim about the mind, showing how psychology can bear on constitutional design and why "belief" may name more than one thing.
Know this first
Context from the editor, not the article · 28 sec
1 · The First Amendment's two halves
The US Constitution's First Amendment bars Congress from establishing a religion and from prohibiting its free exercise. So the state may neither impose a faith nor stop people practising one. Many democracies have comparable rules.
2 · Content versus attitude
Philosophers separate what a mental state is about from how a person relates to it. You can hope, doubt, imagine or believe the same idea, such as rain tomorrow. The content stays fixed while the attitude changes.
3 · Cognitive science of religion
This field uses experiments and cross-cultural studies to examine how religious ideas are acquired, held and acted on. It treats religious thought as a natural psychological phenomenon, without trying to settle whether the ideas are true.
2/5Works in Progress· Nov 3
The 1889 book that explains why some city spaces feel good
A Works in Progress essay revives Austrian architect Camillo Sitte, who argued that beloved cities depend on small, enclosed, irregular public spaces more than fine buildings, and says builders today need his forgotten rules.
Why it matters · It changes what you notice in any city: why a famous square can feel bleak, why a plain street can charm, and why clearing space around a monument can diminish it.
Know this first
Context from the editor, not the article · 18 sec
1 · Nineteenth-century city planning
Planners of the 1800s favored regular street grids, long straight boulevards and major buildings set in the middle of open squares. Paris under Haussmann, who cut wide avenues through the old city, is the best-known example.
2 · Modernist planning
From the early twentieth century, modernist planners separated buildings from streets, setting towers and slab blocks in open space with room for greenery, parking and roads. Brasília, built from scratch, is the classic example.
Brussels's Grand-Place: corner streets continue the line of each side, so the exits nearly vanish from view.1 / 4
3/5Asterisk· Nov 3
Why Few Male Mammals Parent, and What Turns Them Into Fathers
Asif Ghazanfar argues that fatherhood is not a fixed trait. Hormones, brain circuits, gene activity and social setting can move male animals, from mice to men, between neglect, infanticide and devoted care.
Why it matters · It treats a behaviour we file under character or culture as an adjustable biological state, and shows how far animal evidence can, and cannot, explain human fathers.
Know this first
Context from the editor, not the article · 29 sec
1 · The mating-versus-care trade-off
Time and energy spent guarding young cannot be spent finding new mates. Evolutionary biologists therefore ask when care pays off, weighing its costs against the chance that the offspring being protected actually carry the caregiver's genes.
2 · How lesion studies work
To learn what a brain region does, researchers damage it in an animal and watch which behaviours weaken or vanish. This shows the region matters for that behaviour, though usually as one part of a wider circuit.
3 · Hormones and behaviour
Hormones such as testosterone, estrogen and prolactin travel through the blood and act on many tissues, including the brain. They can change neural activity and which genes are switched on, making some behaviours more or less likely.
4/5Longreads· Edward Abel Smith· Nov 3
A writer befriended a convicted wife-killer and lost trust in his judgment
Author Edward Abel Smith found a murder file by accident, then spent five years visiting the convicted husband. He now leans toward innocence, calls him a friend, and no longer trusts his own read of people.
Why it matters · It tests a belief most of us hold without examining it: that long, close acquaintance with someone makes us better judges of their character, rather than worse.
Know this first
Context from the editor, not the article · 31 sec
1 · How wrongful-conviction claims work
Once ordinary appeals fail in England, Wales or Northern Ireland, a convicted person can apply to the Criminal Cases Review Commission. This independent body can send a case back to the Court of Appeal, usually only if the conviction might realistically be overturned.
2 · The Post Office Horizon scandal
Horizon is the Post Office's accounting software. Its bugs produced phantom cash shortfalls, and hundreds of subpostmasters, who run local branches, were prosecuted for theft or false accounting. Court findings and campaigners exposed the faults from 2019 onward.
3 · Majority jury verdicts
5/5Noema· Michael Fitzpatrick· Nov 3
Japan's failure to copy Silicon Valley may be its best defense
Michael Fitzpatrick argues that Japan's slow institutions, which squandered its lead in contactless payments, electric cars and AI, may also be shielding jobs and social cohesion as Tokyo courts Silicon Valley.
Why it matters · It challenges the assumption that adopting technology fastest is the only way to win, using a rich country that has lived with the costs of moving slowly.
Know this first
Context from the editor, not the article · 20 sec
1 · The Black Ships
In 1853, U.S. Navy Commodore Matthew Perry sailed armed ships into Japanese waters and demanded access to a country that had restricted foreign contact for over two centuries. Japanese writers still use the phrase for unwelcome outside pressure to change.
2 · Japan's lost decades
Japan's 1980s boom in stock and property prices collapsed around 1991. Years of weak growth, flat wages and financial strain followed, leaving firms and workers wary of risk. The period is widely known as the lost decades.
Image: Noema
Japan's government is courting Silicon Valley to rescue its economy. In Noema, the Tokyo-based writer Michael Fitzpatrick argues that the country's long failure to copy that model may be its most valuable asset.
Why would a government consult meteorologists about the weather but not clergy about God? Neil Van Leeuwen, a philosophy professor, and Jacob Eisler, a law professor, argue in Aeon that the answer lies in psychology: religious "belief" is a different mental attitude from ordinary belief.
The gap in the usual story
In 1802 Thomas Jefferson wrote that government's powers "reach actions only, and not opinions", describing a "wall of separation between Church and State". The authors say that can't be literally right. Schools shape beliefs about history, and fire officials correct false beliefs about unsafe buildings.
So religious belief must be special, but not because of its subject. Buddhism has no supreme deity, superstitions are supernatural without being religious, and much morality is secular. The authors instead look at how an idea is held, separating 1 from 2.
Three differences
Choice. Religious credence can be chosen, and people switch faiths for romance or social reasons. Factual belief can't. Offered $1 million to sincerely believe lions are vegetarians, you could only pretend.
Evidence. Factual belief bends to evidence: find your laundry dry and dirty, and you stop thinking you started the machine. Faith is meant to survive without proof. Cross-cultural studies, including in Iran and the US, find people less confident in God and angels than in germs and oxygen, though all are invisible.
Scope. Factual belief is always on; even someone pretending concrete is a soft bed lies down carefully. Religious credence is 3, active mainly in sacred settings. In one economic game, 100 per cent of Moroccan shopkeepers chose the most charitable option while the call to prayer was audible, falling to 50 per cent in the following 20 minutes. Another study found Bible Belt residents start pornography subscriptions at the same rate as others, except on Sundays.
Language seems to register the split. Experiments by Van Leeuwen and colleagues find English speakers use "believes" for religious attitudes and "thinks" for factual ones, with parallels in Thai, Mandarin and other languages.
From mind to constitution
The First Amendment to the US Constitution bars establishing a religion and protects its 4. The authors justify each half separately.
Establishment fails because democracy runs on argument. A quarantine based on disease risk can be challenged with evidence by anyone. A policy imposing religious practice persuades only those who already hold the credence. Believers can still be privately motivated by faith, the authors say, but public justification must rest on contestable claims.
Free exercise works because of compartmentalisation. People who believe in demons don't perform exorcisms on strangers in the supermarket. Since credences can be set aside in public, the state needn't challenge them in private.
What remains unsettled
The authors admit conspiracy beliefs and quasi-religious ideologies are hard to classify. The essay also doesn't say what follows for believers whose faith does drive their public conduct, which is where the bargain would be tested hardest.
Governments consult meteorologists about storms and economists about recessions. So why, ask Neil Van Leeuwen and Jacob Eisler in Aeon, don't they keep clergy on hand to find out what's going on with God? The authors, a philosophy professor raised in the Christian Reformed Church and a law professor who regularly attends synagogue, both support separating church and state. They argue that the principle only makes full sense once you see that religious "belief" is a different mental attitude from ordinary belief, one that empirical research can help describe.
The hole in Jefferson's wall
In 1802 Thomas Jefferson wrote to the 1, a group of Connecticut Baptists, that government's legitimate powers "reach actions only, and not opinions". The First Amendment, he said, had built "a wall of separation between Church and State".
The authors say this can't be taken literally, because governments shape beliefs all the time. Schools teach astronomy and history. Fire officials who learn that people wrongly think a building is safe set out to change their minds. The law even attaches duties to belief: a psychotherapist who believes a patient is dangerous must report it.
So the wall can't cover belief in general. Its defenders must assume religious belief is somehow distinctive, but the obvious candidates all fail. God can't be the mark, because Buddhism has no supreme deity. The supernatural can't be, because non-religious superstitions involve it too. Morality can't be, despite the biologist Stephen Jay Gould's suggestion: many religious beliefs aren't moral, many moral beliefs aren't religious, and every government presupposes some morality, such as that murder is wrong.
All these answers look at what the belief is about, and the authors think that is the mistake, since any idea can be held religiously or not. They ask instead how a person holds it. They call the two attitudes 2 and 3. The everyday word "belief", they note, is 4: it stretches from where the ice-cream store is to the conviction that "all is one".
Three differences
Choice. A congregant at the 5, an evangelical movement studied by the anthropologist Tanya Luhrmann, told her: "There is always a choice to believe what it is." People also change religions for romance or other social reasons. Factual belief doesn't work that way. Offered $1 million to believe, in a matter-of-fact way, that lions are vegetarians, you could imagine it or pretend to, but you couldn't simply do it.
Evidence. Factual belief yields to evidence, despite biases and misremembering. If you find your laundry dry and dirty an hour later, you stop thinking you started the machine. This shows up early in life: Alison Gopnik and colleagues describe babies forming and testing mini-theories, and Paul Harris and Melissa Koenig find that children favour sources that have been accurate before.
Faith, by contrast, is meant to withstand absent or contrary evidence. Hebrews calls it "the conviction of things not seen". Doubting Thomas, who demands proof, is the model of what faith is not. Cross-cultural studies, including in Iran and the US, find people less confident in God and angels than in germs and oxygen, though all are invisible; the authors presume the gap reflects a felt lack of evidence. Emily Liquin, S Emlen Metz and Tania Lombrozo found that people felt less need for explanation and accepted more mystery in religious beliefs than in scientific ones. 6, who defend doctrine with claimed evidence, are in the authors' view outliers, and claiming evidence differs from being moved by it.
Scope. Factual belief is your default map of the world, always on. Even if you pretend a concrete slab is a soft bed, you lower yourself onto it carefully. Religious credence is 7, guiding symbolic actions such as prayer and ritual far more than commuting or feeding the dog. The authors cite several findings:
One study found Bible Belt residents start pornography subscriptions at the same rate as people in less conservative regions, except on Sundays, when the rate is significantly lower.
Another found charitable giving among religious and secular people not significantly different, again except on Sundays.
The experimental economist Erik Duhaime found that 100 per cent of Moroccan shopkeepers in Marrakesh's Medina chose the most charitable option in an economic game while the call to prayer was audible, falling to 50 per cent in the following 20 minutes.
The philosopher Daniel Dennett made the point bluntly: people do things with God watching that they wouldn't do with their mothers watching. The authors add that a religious rule demanding round-the-clock devotion is itself a clue, since a norm would hardly be needed if people already behaved that way.
The resulting picture is a two-map mind. One map is involuntary, evidence-responsive and guides action everywhere. The other is chosen, resistant to evidence and activated in sacred settings.
People already half-know this
People say "I personally believe" about religion, but nobody says it about where their bicycle is. Van Leeuwen and colleagues have found experimentally that English speakers tend to use "believes" for religious attitudes and "thinks" for factual ones, with parallels in Fante (spoken in Ghana), Thai, Mandarin and 8 (a creole of Vanuatu). "Believes" correlates with seeing an attitude as voluntary, held for reasons such as loyalty and expressed symbolically. "Thinks" correlates with seeing it as evidence-based and guiding practical moves such as investing.
From psychology to constitution
The two halves of the First Amendment each get a separate justification.
The ban on establishing a religion rests on evidence. Every government action rests on some belief: quarantining travellers during an Ebola outbreak rests on the factual belief that they could spread disease. Because that belief answers to evidence, anyone can contest it in argument, and the authors see such reasoning as the heart of democracy. A policy imposing religious practice can only persuade those who already hold the credence, so deliberation has no grip on it. Hence the distinction between private motivation and public justification. Abortion politics may be driven by faith, the authors note, but its advocates still try to argue in terms that reason and evidence can contest.
Free exercise rests on compartmentalisation. Since believers can set credences aside in public (people who believe in demons don't typically perform exorcisms on strangers in the supermarket), the state has no need to challenge those credences. They can stay protected as private matters, alongside other questions of privacy and autonomy.
The bargain is religious freedom in exchange for no established faith. The authors argue that it works only because religious credence sits at a different level from factual belief, and they hear a foreshadowing of this in Jefferson's phrase "between man and his God".
What it leaves open
The authors concede that conspiracy "beliefs" and quasi-religious ideologies are hard to categorise, and they don't resolve them. Their free-exercise argument also depends on compartmentalisation being the default, and the essay doesn't say what follows for believers whose credences do drive public behaviour. Nor does it address people raised in a faith who may never experience it as a choice. The behavioural findings are cited briefly, without details of their methods, so readers can't judge from the essay how much weight each will bear.
Remember
Asking how an idea is held, rather than what it concerns, sidesteps failed definitions of religion based on God or morality.
The Sunday and call-to-prayer findings suggest religious conviction can switch on and off with setting, even among sincere believers.
On this argument, private religious motivation is acceptable in politics; only the public justification must be open to evidence.
People love the streets of Verona and Prague, and the easy explanation is their beautiful old buildings. But, as a Works in Progress essay points out, people also love the streets of Osaka and Hong Kong, whose buildings are mostly nondescript. Its explanation comes from The Art of Building Cities, published in 1889 by the Austrian architect Camillo Sitte.
The principles
Enclose the space. Sitte argued that good squares work as outdoor rooms, where people feel calmer and sheltered. The trouble is that streets punch holes in the walls. In a 1, such as the Grand-Place in Brussels, a single street enters at each corner and continues the line of one side, so the walls appear to overlap and hide the exit. Barcelona's Plaça Reial goes further, bridging six of its seven entrances. By contrast, the Place de la Concorde in Paris has nine roads and buildings on one side only.
Keep it small. Sitte's rule of thumb was that a square should measure no more than twice the height of its main building. The essay argues that St Peter's Basilica looks stubby at the end of its vast square, while the smaller Pantheon towers over its tiny one.
Don't worry about symmetry. Pedestrians can't see a plan. The neat octagon of the Place Vendôme is lost at ground level, while the wonky pentagon of Verona's Piazza delle Erbe is barely noticeable.
Put monuments at the edges. Statues in the center block routes and look lost. A bronze David over five meters tall, set in the middle of a square in Florence, left tourists thinking it was life-sized, Sitte reported, because nothing nearby gave it scale.
Tangle great buildings with ordinary ones. Of Rome's 255 churches, Sitte found only six freestanding. When Haussmann, the planner who remade Paris, demolished the narrow street leading to Notre Dame, he meant to improve the view. The essay argues he made the cathedral look smaller and lonelier by removing the ordinary buildings that gave it a frame and a sense of scale.
Forgotten, then revived
The book hit Europe hard. Within a year, expansion plans in Germany and Austria were being redrawn. In Helsinki, a Sitte-style district gets "old town" effects from the same tall apartment buildings and wide streets as the grid beside it.
Then modernist planners, who treated public space as whatever was left over after placing buildings, dismissed him as a romantic. The essay says their approach has since been repudiated almost worldwide. 2 designers, who favor traditional walkable towns, built Seaside in Florida and Poundbury in England under his influence.
The evidence is Sitte's observation and the essay's own judgment rather than measurement. Its practical claim is that mainstream attempts at traditional streets often look awkward because the craft must be relearned.
An essay in Works in Progress opens with a puzzle. People love the streets of Verona, Bruges and Prague, and the easy explanation is their beautiful old buildings. But people also love the streets of Osaka, Taipei and Hong Kong, whose buildings are mostly modern and nondescript. Architecture alone can't explain the appeal.
The essay looks for the rest of the answer in The Art of Building Cities, published in 1889 by the Austrian architect Camillo Sitte (pronounced "zitter"). Sitte attacked the repetitive grids and sweeping boulevards of his day. Those now look rather attractive beside the motorways and concrete plazas that followed, which the essay argues makes him more relevant today than he was then.
Squares as outdoor rooms
Sitte's first principle was enclosure. People feel calmer and more protected in physically enclosed spaces, which they experience as outdoor rooms, and each such room can develop its own character. A city without them would be like an extreme open-plan house, with kitchen, bedroom and bathroom bleeding into one another.
The difficulty is access. A square is a room with five or ten large doors left permanently open. Old towns had two main fixes. One is the 1, where each corner admits a single street that continues the line of one side rather than cutting in diagonally. In Brussels's Grand-Place, this means the two walls at a corner appear to overlap from almost anywhere in the square, hiding the exit. The other fix, which Sitte preferred, is to bridge the entrances. Barcelona's Plaça Reial has seven entrances, six of them bridged over, so it feels like a cloister. 2, rows of columns carrying a roof or upper floor, achieve this simply.
London's Trafalgar Square shows the opposite: it was designed to improve carriage flow, and you can generally see two or more busy roads from wherever you stand. The Place de la Concorde in Paris has nine roads and buildings on only one side.
Streets follow the same logic. Curves and T-junctions close the view. Manhattan's Broad Street curves and ends at the Federal Hall National Memorial, so it reads almost as a room. 42nd Street is a canyon, what Sitte called "parallels vanishing at infinity".
Keep squares small
Sitte's rule of thumb was that a square's dimensions should not exceed twice the height of its principal building. To take an illustrative case, a 30-meter church would get a square no more than about 60 meters across. Beyond that, buildings shrink visually and stop acting as walls. At Madrid's Plaza de la Armería and Budapest's Heroes' Square, the essay says, monumental buildings dwindle to doll's houses. Taller buildings would restore the enclosure but feel made for giants.
Even Bernini's St Peter's Square fails this test, the essay argues: impressive in drone footage, exhausting on foot, with the Basilica looking small and stubby. The Pantheon in Rome is a smaller building than St Peter's, but its tiny square makes it tower over you, even though the street has risen 1.3 meters since Antiquity and swallowed its entrance steps.
Symmetry is for maps
Sitte denied that harmony requires regular plans. The octagonal Place Vendôme in Paris looks striking from a satellite, but pedestrians rarely stand at the point where its balance is visible. Verona's Piazza delle Erbe is a wonky, elongated pentagon, yet nobody can see all five sides at once, so its flaws exist mostly "on paper". Sitte thought the irregularities even helped, arousing visual interest and creating recesses for benches or statues.
Move monuments to the edges
The essay offers an analogy: you wouldn't stand an ornament in the middle of your carpet or in front of the fireplace. You'd put it against a plain wall. Sitte applied this to statues. In the center they block routes and look isolated, and on an axis with a major building they block the view.
Florence supplies both cases. In the Piazza della Signoria, statues line the sides, and the replica of Michelangelo's David stands out against plain masonry. A bronze David at the center of the Piazzale Michelangelo stands over five meters tall, yet Sitte reported that tourists left thinking it was life-sized. We judge size by comparison, and against the sky there is nothing to compare it with.
Buildings need neighbors
Sitte applied the same reasoning to major buildings, which he thought should be entangled with 3 architecture, the ordinary everyday buildings of a place. Of Rome's 255 churches, he counted 41 adjoining other buildings on one side, 96 on two, 110 on three and two on all four. Only six stood free, and two of those were modern Protestant churches.
Nineteenth-century planners ran the experiment in reverse through 4, the demolition of ordinary buildings around monuments. Haussmann cleared the narrow Rue Neuve Notre Dame to create the large square in front of Notre Dame. The essay argues this backfired. The old street framed the facade, gave it scale and hid it until you were close enough for it to loom. Now it looks smaller and lonelier.
Sitte proposed the reverse for Vienna's Votive Church, stranded on a "desolate monster plaza": build ordinary city over most of it, leaving three small squares with framed views. It never happened. The contrast is visible in Padua and Washington DC, a 5 reversal, meaning the shapes and the background swap roles: in Padua buildings are molded around spaces, in Washington spaces around buildings.
Fall and revival
The book landed, in the words of its English translator Charles T Stewart, "like a demolition bomb". Within a year, it had prompted redrawn expansion plans in Germany and Austria. Helsinki switched from a neoclassical grid to curving streets and turbine squares in the early twentieth century. The newer district has "old town" effects despite the same tall apartment buildings and wide streets.
Then modernism took over. Modernists treated public space as whatever remained after placing buildings, and preferred parallel blocks to the 6, which lines the edge of a city block with buildings. The critic Sigfried Giedion called Sitte "a kind of troubadour".
The essay says modernist planning has since been repudiated almost worldwide. 7 designers, part of a movement favoring traditional walkable towns, built Seaside in Florida and Poundbury in Dorset under Sitte's influence. Practices such as Ash Sakula and Peter Barber work in a similar vein. Mainstream attempts, the essay says, are often awkward, because this is a craft that has to be relearned.
What stays open
The case rests on Sitte's observations and the essay's own judgment of famous places, not on measurements of how people use or feel about them. It also never returns to Osaka, Taipei and Hong Kong to show how the principles apply there, and says little about fitting enclosed, irregular spaces around modern traffic.
Remember
Our sense of a monument's size depends on nearby comparators, so clearing space around it can make it look smaller.
Helsinki suggests layout alone can create charm: its Sitte-style district uses the same tall apartments and wide streets as its grid.
Plans and drone shots mislead: Sitte judged a space from where pedestrians actually stand.
Only about 4–5% of male mammals care for their young. In an essay for Asterisk, Asif Ghazanfar, who has studied the biology of fish, rodents, monkeys and humans, asks what turns a philanderer into a father. His answer is that no single thing does. Fatherhood, he argues, emerges from a system of genes, neural circuits, hormones, social context and environment, and a shift in any of them can move a male between care and killing.
Why mammals philander
The first parent was probably a male fish. A male already defending a territory can guard eggs laid there at almost no extra cost. Life on land ended that bargain. With internal fertilization, a male could not see his sperm reach the eggs. Viviparity, growing embryos inside the mother's body, added a long gap before birth, during which she could mate with others. Waiting for young who might not be his is a poor use of time and energy.
The switch
Male house mice normally try to kill pups, even their own. Yet force one to live with his mate after mating and he cares for the pups born about 18 days later; the longer the stay, the more care. Prairie voles evolved toward a similar result: a 1, a lasting preference for one partner, raises confidence in paternity, and the fathers huddle, groom and protect.
The underlying circuit seems widely shared. Damaging the 2 (MPOA), a small part of the 3 at the brain's centre, drastically reduces parenting in rodent mothers and in California mouse fathers. What varies is the trigger. California mouse fathers eat the placenta, then keep caring through exposure to their mate's urine, which works better than other females' urine, even when she is absent.
African striped mice do not need a mate at all. Sexually naive males raised alone care for unrelated pups and show far more MPOA activity; males raised in bachelor groups ignore or kill them. A gene called 4 dampens the MPOA, and boosting it makes males less caring or outright infanticidal. A month of isolation reverses course, lowering Agouti and raising care.
Humans
Testosterone is the best-studied link. A study that tracked hundreds of men in Cebu City, Philippines, over time found that single men with the highest levels were likelier to later partner and have a child. Levels then fell with partnering and fatherhood, most steeply among men with newborns, and the most hands-on fathers had the lowest.
Culture complicates this. Among the Aché of Paraguay, several men can count as a child's fathers, providing food and protection but little holding or play. That undercuts the assumption that care tracks certainty of paternity. Causation also remains unsettled: castrating male rats stops pup-killing without creating caregivers, and MRI cannot yet reliably resolve the human MPOA.
Asif Ghazanfar, who has studied the biology and behaviour of fish, rats, lemurs, several monkey species and humans, opens his Asterisk essay with advice for his single brother: get a dog, because walking one advertises that you would make a good father. The serious claim underneath is that male parental care is rare, flexible and not the product of any one cause. Fatherhood, he argues, emerges from a system of gene expression, neural circuits, hormones, social context and environment, with no component privileged over the others.
Why fathers are rare
Only about 4–5% of male mammals care for their young. Ghazanfar treats this as an economic outcome, with time and energy as the currencies. The logic is clearest in fish, where male care is common.
Most fish simply spawn and leave. The next most common strategy is a male guarding the eggs. If parental care means any behaviour that improves an offspring's chances of survival, the very first parent was probably a male fish. The reason is cost. A male already defending a territory that attracts females can protect eggs laid there with little extra effort, and the eggs do not hurt his chances with other females; they may even help. Females, meanwhile, keep growing throughout life, and bigger females produce more of their expensive eggs. Even male guarding is conditional: bluegill sunfish decide whether to guard based on how many eggs are present.
Life on land broke this bargain. Eggs and sperm would dry out in air, so land animals evolved internal fertilization, along with either hard-shelled eggs or 1, keeping the embryo inside the mother's body. A male could no longer watch his sperm reach a batch of eggs, so paternity became uncertain. Viviparity also created a long gap between fertilization and birth, during which a female can mate with other males, sometimes precisely to blur paternity. With the mother already committed to gestation and milk, waiting around for young who may not be his looks like a poor bet. Mating and moving on looks better.
A behaviour that can be flipped
The typical male house mouse tries to kill any pups he finds, including his own. He cannot recognize them as his, and killing them lets their mother return to fertility sooner. Yet if experimenters force him to live with a female after mating, he cares for the pups born about 18 days later. The effect is dose-dependent: the longer he stays with his pregnant partner, the more care and the less aggression he shows. Ghazanfar reads proximity as somehow working on the male's biology to register paternity certainty.
Evolution ran a similar experiment with the 2, a lasting preference for one mate. Neural systems for mating were tuned so that a specific partner became rewarding and others lost their appeal, which raises confidence in paternity. Two near-identical voles show the split. Montane voles behave like typical mammals: males mate and leave. Prairie voles, in a harsher habitat with less food, pair bond and raise fewer young together, with fathers huddling, grooming and defending.
The circuit and its triggers
Is the fathering circuit new, or latent in most males? Ghazanfar suggests evolution finds it easier to build similar brains in both sexes and activate them differently by context. A key node is the 3 (MPOA), part of the 4, a structure in the middle of the mammalian brain. Damaging it drastically reduces parenting in rodent mothers, and in fathers of species like the California mouse.
In mothers, the trigger is the hormonal swing of pregnancy and birth, much of it from the placenta. Males get there by other routes. In Mongolian gerbils and Campbell's dwarf hamsters, paternal care is linked to a drop in testosterone after birth. Pheromones, chemical signals detected by the 5 in the nasal cavity, may also push males toward care, as may a newborn's smell and sounds.
California mice combine several of these. The male helps induce birth by licking the female, eats the placenta, and starts caring at once. As those signals fade, his mate's urine sustains the behaviour. Urine from a virgin, another lactating female or a familiar non-mate works less well. A couple of daily doses is enough, even if she is absent.
The mouse that needs no mate
The African striped mouse removes the mating requirement entirely, in work published in Nature in 2026. After weaning, males may live alone, in bachelor groups, or with family, depending on population density and resources. Solitary males lick, groom and warm pups; those from bachelor groups show no care and are often infanticidal.
Recreated in the lab, the difference appeared in sexually naive males caring for unrelated pups: isolated males had far greater MPOA activity than group-housed ones. A major influence is 6, a gene whose protein reduces MPOA activity when expressed there. Adding more shifts behaviour a step: caring males become indifferent, and indifferent males become infanticidal. And it reverses. A month of isolation makes a group-housed male more caring and lowers his Agouti expression.
Humans make it harder
The Aché, an Indigenous people of Paraguay's tropical forests, believe any man who had sex with a woman in the months around conception, up to the birth, contributes to the child's paternity. There are primary and secondary fathers. They offer protection and food rather than holding or play, and children who have them survive better. This cuts against the assumption that care tracks certainty of paternity.
Testosterone is the best-studied biological link. Marmosets, 7 whose group members share infant care, offered early evidence: fathers who carried infants most had the lowest levels. Early human studies found married men had lower testosterone than single men, but most were 8, comparing different groups at one moment, so lower testosterone might simply make marriage likelier.
A 2011 9 study, following the same men over time, addressed this with hundreds of men in Cebu City, Philippines. Single men with the highest testosterone were later likelier to partner and father a child. Levels then fell with partnering and with fatherhood, most steeply among men with newborns, and could not be explained by age or stress. The most hands-on fathers had the lowest levels, and a follow-up found levels shift as time with children changes.
Brain imaging shows mothers and fathers engaging a similar caregiving network, in gay and straight fathers alike, and its connections strengthen with time spent as a child's sole carer. In one study, men with larger testes did less fathering and showed less activity in a reward area when viewing pictures of their child.
What stays open
Ghazanfar concedes cause and effect are hard to separate. Castrating male rats around puberty ends their pup-killing without making them caregivers. MRI cannot yet reliably resolve the tiny human MPOA, and experiments manipulating human relationships are off limits. He also suspects behavioural flexibility is common but hidden, since biomedicine studies so few species. His conclusion that falling testosterone "facilitates" fatherhood goes beyond what correlational human data can show, and the essay reports few effect sizes. Asked why he had children and his brother did not, he can only guess: perhaps a summer near Lahore at age 12, falling for a one-year-old cousin.
Remember
Paternal care probably began as a cheap add-on: territorial male fish guarding eggs on ground they already defended.
In striped mice, a month of isolation can make a bachelor-group male more caring by lowering Agouti in the brain.
Human testosterone seems to work both ways: high levels predict later partnering and fatherhood, then fall with hands-on care.
English juries are first asked for a unanimous verdict, but a judge may later accept a majority, such as 10 of 12, after a minimum period of deliberation. It carries the same legal force as a unanimous verdict.
Image: Longreads
A voicemail reached Edward Abel Smith earlier this year as he drove his daughters to a playground: congratulations on The Times naming his new book its Book of the Week. "Bloody well done, my friend!" The caller was Robin Garbutt, serving life in Wealstun Prison, Yorkshire, for murdering his wife, Diana.
In this Longreads essay, Smith describes what five years of calls and visits have done to his certainty, and to his faith in his own judgment.
The case
Diana was beaten to death in bed in 2010, in the flat above the post office the couple ran in a North Yorkshire village. Robin said an armed robber had struck, and that he then ran upstairs and found her. No one saw a robbery, and paramedics noted 1, the stiffening of a body after death, suggesting she had died hours earlier. Ten of 12 jurors convicted. Smith stumbled on the file in 2021 and felt contempt. Then he noticed how many villagers publicly backed Robin, and wrote to him.
The doubts
Robin has challenged the verdict through the 2, the independent body that decides which claims of wrongful conviction reach the Court of Appeal. By Smith's account, it accepted his points three times and rejected him anyway.
Motive. Prosecutors said he was stealing from the Post Office. Later-disclosed records showed the spending patterns were normal for the branch. The accounts also came from 3, the faulty software behind hundreds of wrongful prosecutions of branch operators.
Opportunity. A stomach-contents analysis put the death between 2:30 and 4:30 a.m., but an independent pathologist found the method flawed. A customer testified to hearing a woman call "Robin" at about 6:45.
Weapon. TV footage showed the wall where the pole was later found standing empty the day after the murder. An officer's DNA was on the pole; Robin's wasn't.
Against this stand 79 seconds, the gap between the safe opening and Robin's emergency call, into which his robbery story must fit. Smith re-enacted it at home. It is possible, he concludes, but only just. A fourth application has waited two years.
The friendship
Smith finally suggested to Robin that he might be guilty. Robin sounded deflated and never answered the question. Two weeks later his only copy of his case files arrived on a USB stick. Smith can't say whether that was innocence or calculation.
Smith now leans toward innocence without concluding, and says the relationship has clouded rather than sharpened his instincts. He second-guesses his reads of colleagues, friends and family.
The account has limits: the essay never gives the CCRC's reasons or returns to the rigor mortis. It ends with his daughter asking whether the man on the phone is his friend, with Robin listening. "Yes, darling," he says.
Earlier this year, as Edward Abel Smith drove his daughters to a playground, a voicemail played through the car speakers. The Times had named his new history book its Book of the Week. "Bloody well done, my friend!" said the caller. He was Robin Garbutt, phoning from Wealstun Prison in Yorkshire, where he is serving life for murdering his wife, Diana.
In this Longreads essay, Smith, an author and investigative journalist, offers no verdict. He gives an account of what five years of letters, calls and visits have done to his certainty, and to his confidence that he can read people at all.
A file found by accident
In 2021 Smith was at the National Archives researching a possible sequel to his Ian Fleming biography when a filing error and boredom led him to recent criminal files. One held the judge's summing up in Diana Garbutt's murder. She had been beaten to death in bed, in the flat above the post office she and Robin ran in Melsonby, a North Yorkshire village of fewer than 800 people.
On the morning of March 23, 2010, Robin called emergency services to report an armed robbery, saying he had then run upstairs and found Diana dead. Nobody in the village saw signs of a robbery on a busy morning. Paramedics noted 1, the stiffening of a body after death, which suggested she had died many hours earlier. The judge made his view clear, and 10 of 12 jurors convicted by a 2. Smith says he felt "nothing but contempt."
Then he found a website run by Robin's family, listing supporters who included a striking number of villagers. He began writing through 3, a service that prints a letter and delivers it to a prisoner's cell; for 26 pence more, it includes a sheet for a handwritten reply. At their first meeting, Robin told him: "I won't try to persuade you either way." Smith liked him immediately. He also noted that this was precisely how a guilty man would want to come across.
Three applications, three rejections
Robin's only route to appeal runs through the 4, the independent body that screens claims of miscarriage of justice and decides which reach the Court of Appeal. He has applied three times, each time targeting a different part of the prosecution case. By Smith's account, the CCRC accepted the substance each time and refused him anyway. Robin calls it Whac-A-Mole.
Motive (2015). Police said Robin had been stealing from the Post Office and killed Diana when she found out. Post Office records for his six years there, previously undisclosed, showed the supposedly suspicious spending patterns were normal for the branch. The financial records also came from 5, the faulty software behind hundreds of wrongful prosecutions of 6. Smith is careful here: Horizon doesn't exonerate Robin, but the motive may rest on unreliable data. His own digging is mixed. The couple were just getting by, yet spent over £5,000 on retreats, a US holiday and Glastonbury tickets.
Opportunity (2017). An expert used Diana's stomach contents to place her death between 2:30 and 4:30 a.m., a window that left Robin the only possible killer. An independent pathologist found the method flawed, and the CCRC accepted that this evidence should be removed. Robin had also served about 70 customers from 4:30 a.m. that morning, seemingly cheerful, and one friend testified to hearing a woman's voice call "Robin" from the stairs at around 6:45. The jury did not accept his account, for reasons Smith says have never been fully explained.
Weapon (2019). The pole used in the killing turned up on a wall near the post office days later. Time-stamped TV footage showed that wall empty the day after the murder, contradicting the claim that Robin left it there that night. Robin's DNA was not on the pole. The DNA of the officer who found it was, and also at the scene, in a property he said he had never entered. Smith has heard darker theories, but favours incompetence. The trial judge himself cited a "regrettable lack of professionalism," and a clump of unidentified hair was bagged as evidence and then lost.
A fourth application has sat with the CCRC for two years. Each quarterly update letter says, in some variation, that the case officer is "familiarising myself with the issues."
Seventy-nine seconds
Against all this stands one stubborn interval. The safe was remotely unlocked at 8:35:54 a.m., and the 999 call arrived at 8:37:13. In those 79 seconds, on Robin's account, a masked gunman appeared, Robin bolted the door, filled a holdall, let the robber out the back, ran upstairs, found Diana and phoned.
Smith measured the distances and re-enacted the sequence at home, managing it in the time. Investigators and former police officers told him it was possible, but only just. Both camps read the tightness as proof: believers say a liar would invent something more plausible, doubters that it makes no sense. "Impossibility," Smith writes, is the only thing that would have settled it.
The question he finally asked
Eventually Smith put it to Robin directly: perhaps the CCRC was right, and he was guilty. Robin was audibly deflated, talked about the dread of waiting, and never actually answered. Smith's first reaction surprised him. He worried about having hurt Robin, which is when he realised this might be a friendship.
Two weeks later, an envelope arrived holding Robin's only copy of his case files on a USB stick, including a 240-row spreadsheet cataloguing witness recollections. "A man with nothing to hide," Smith writes, "or a man who had calculated, very precisely, that transparency was his best remaining move." He could not tell which.
The files held leads never run down: boxer shorts in the couple's bin with an apparent rust stain, resembling a pair a neighbour said he was missing, and never tested; a tip to 7, the anonymous crime hotline, about a local man boasting he had "bashed her around the head," whose timing Smith cannot establish; and allegations that Diana's now-dead stepfather had abused her in childhood, and that she had recently resolved to go to the police. None resolves anything.
What it cost
Smith now leans toward innocence without concluding. Making a Sky documentary, Murder at the Post Office, he settled on framing Robin as "contested," then was relieved when Robin approved, then suspicious of that relief. Learning about Diana through her letters and friends has left him feeling responsible to her, and wary of Robin.
The sharper claim is about judgment. Smith assumed that years of watching Robin answer hard questions would hone his instincts. He says the reverse happened: the relationship clouded his vision, and he now second-guesses reads of colleagues, friends and family.
The essay has limits worth noting. It is one engaged writer's account, and the author's other work centres on contested convictions. It doesn't give the CCRC's stated reasons for rejecting Robin, or return to the paramedics' rigor mortis observation. North Yorkshire Police gave no comment. Smith's last scene needs no resolving: his daughter asks if the man on the phone is his friend while Robin listens in silence. "Yes, darling," he says.
Remember
Knowing someone longer doesn't necessarily help you judge them; Smith says intimacy degraded his instincts and spread doubt to other relationships.
Discrediting an investigation isn't proving innocence: as Smith notes, a flawed institution is not automatically a wrong one.
The same fact can serve both sides: Robin's implausibly tight 79-second story reads as a liar's slip or an honest man's truth.
About a decade ago, at the British Embassy in Tokyo, Boris Johnson, then London's mayor, pitched the city's contactless Tube cards as "light-years ahead" and asked, "Who else can boast that?" Fitzpatrick, in the press seats, pointed out that Japan had had the technology for years. Johnson's hosts had been too polite to correct him, and many didn't know themselves. Japanese consumers, loyal to cash, largely ignored the invention until it took off in Europe and elsewhere in the 2000s.
Invented here, scaled elsewhere
Fitzpatrick sees this pattern everywhere. Toyota showed him electric and self-driving cars 30 years ago. Japanese scientists pursued 1 in the 1980s while Western researchers doubted them. Others collected the rewards.
His explanations overlap. Under the "2" theory, Japan's standards and habits evolved, like island species, into forms that don't travel. Tokyo University's prestigious startup factory has one part-time PR rep, who speaks no English. Decisions crawl through 3, the slow trust-building done before anything is formally decided. After the 1991 crash, firms and workers retreated into safety.
Disruptors also pay a price. Takafumi Horie, an internet founder whose hostile 2005 takeover bid rattled the establishment, was later convicted of securities violations over falsified results and served about 21 months. Startups, Fitzpatrick says, took the hint.
The emergency
The alarm is justified. Japan's share of global 4 fell from about 18% in 1995 to about 3.8% in 2025. GDP per person dropped from second in the world to 38th. Real wages have been flat for 30 years.
So Tokyo is welcoming investors such as Peter Thiel and Marc Andreessen, who opened his first overseas office there this summer. It is also pressing universities to produce global startups. Fitzpatrick calls this a "sham": founders take grants that require the word "global," though most don't speak English.
The turn
Then he asks whether the plodders were right. In 1996, Japan's education minister politely refused Apple's school technology, which looked backward at the time. Sweden has since reversed its own digital push after literacy fell, returning to paper and banning phones in schools. Japanese firms have long shielded workers from shareholder demands, and men still guide traffic around roadworks where a light would do.
Fitzpatrick's target is 5, the belief that societies must adapt to whatever gets invented. Japan's friction, he suggests, shows they can choose.
The argument is better as provocation than proof. Its evidence for the upside is largely anecdotal, and it never weighs preserved jobs against stagnant wages. And Japan didn't really choose to ignore its cashless lead. It failed to notice it, which is habit rather than sovereignty.
In an essay for Noema, the Tokyo-based writer Michael Fitzpatrick spends most of his length cataloguing how Japan has failed at technology capitalism, then asks whether those failures amount to protection. His hedged answer is that Japan's slowness may be one of the few working defenses against 1, the belief that once a technology appears, societies must adapt to it.
A gaffe at the embassy
The essay hangs on a scene from about a decade ago. Fitzpatrick, raised in an Irish republican home, is sitting in the British Embassy in Tokyo, a Georgian-revival building of damask and mahogany whose charm, he says, is its refusal to update. Boris Johnson, then mayor of London, is on a trade mission, pitching the "light-years ahead" contactless cards used to pay for the Tube. "Who else can boast that?" he asks.
Fitzpatrick answers from the press seats: Japan has had this technology for years and did much to create it. Johnson, thrown, asks whether this is true. Heads bob. "Why was I not told this?"
Fitzpatrick draws three conclusions. Johnson's staff and the embassy hadn't done their homework. Many Japanese hosts didn't know their own country's record. And etiquette would have let the error stand rather than embarrass a guest. The deeper irony is that Japanese consumers, trusting only cash, largely ignored the technology until it took off in Europe and elsewhere in the 2000s.
Invented here, scaled elsewhere
That, Fitzpatrick argues, is the national pattern. Toyota put him in one of its first commercial electric cars 30 years ago, and in a self-driving vehicle the same year. The computer scientist Noriko Arai told him 15 years ago that a "mad scientist" dream of humanoid robots had put Japan ahead in AI. In the 1980s, while Western researchers doubted 2, Japanese scientists were pursuing them. "We say Japan was ahead at the time," the AI researcher Yutaka Matsuo told him. "But perhaps that doesn't mean a lot now." American giants and a fast-moving China took the rewards.
He offers several overlapping explanations:
The fullest case is Takafumi Horie, known as Horiemon. In 2005 his internet firm Livedoor used a loophole in off-hours trading to buy 29.6% of the radio broadcaster Nippon Broadcasting in one go, reaching about a third, in a bid to control Fuji TV. The government changed the law to close the gap, and the bid failed. A year later Horie was arrested on a separate matter: Livedoor had reported a pretax profit of about 5 billion yen when, prosecutors said, it had lost about 300 million. He served about 21 months of a 2½-year sentence. Whatever the merits of the charges, Fitzpatrick says, the startup community took the message that this style of capitalism would be punished here.
The emergency and the cure
The economic alarm is real. Japan's 6 slipped from second in the world to third in 2010 and fourth in 2023. Its share of global nominal GDP fell from about 18% in 1995 to about 3.8% in 2025. GDP per person dropped from second to 38th. Real wages have stagnated for 30 years, and the yen is near a 40-year low.
So Tokyo is turning to the American model. Prime Minister Sanae Takaichi cheerfully received the investor Peter Thiel and his trade mission. Marc Andreessen, who runs a major venture capital firm, opened his first overseas office in Japan this summer. Fitzpatrick frames 7, a U.S.-led pact on AI and chip supply chains, as a shotgun wedding whose vow is: join us or become irrelevant. AI rules have been made as loose as America's, and many see that as an invitation to U.S. firms.
Universities have been drafted as startup factories. Fitzpatrick thinks this is a category error, asking slow, hierarchical institutions to behave like fast, unstable ones. Founders take grants for "global startups" while skipping the global part, because funding requires the slogan and most don't speak English. He calls this a "sham" and a "8," and he notes The Economist's view that the startups-first plan looks misguided.
The turn
Then he changes sides, or nearly. In 1996 Apple pressed Japan to bring its teaching technology into schools, and the education minister politely declined. Apple's CEO Michael Spindler asked when to try again. Hideki Onda, then at Apple, told him: when the 1960s floor indicator above the ministry's elevator goes digital. It looked Luddite at the time. But Sweden, after a drop in student literacy, has reversed its push toward digital learning, banned phones in schools and gone back to paper.
An unnamed Japanese tech journalist supplies the essay's thesis: Silicon Valley "is optimized to maximize returns," not wellbeing, and history may judge this era as producing "technological pollution." Countries with long traditions, he argues, should propose other ways to use technology.
Fitzpatrick adds the labor case. In a 2006 government survey, nearly 70% of responding firms called employees important stakeholders and 90% planned to maintain long-term employment. Japan, he says, is "a massive job creation scheme": men guide traffic around roadworks outside his house where a traffic light would do. Societies should choose which technologies to adopt, how quickly and for whose benefit, a position he links to UNESCO's 2021 call for human oversight of AI. Against that stands Johnson's closing line at the embassy: "Once the toothpaste is out of the tube, folks, that's it."
What the essay leaves open
The argument is stronger as provocation than as proof. Its evidence for the upside is a Swedish reversal, a two-decade-old survey, a street anecdote and a friend's opinion, set against hard numbers on the downside. Nor does it weigh 30 years of flat wages against the jobs preserved. Its central example also cuts both ways: Japan didn't deliberate and reject cashless payments so much as fail to notice what it had. Sovereignty means choosing, and much of what Fitzpatrick praises is habit. He half admits this, opening with the embassy's continuity as a "fantasy" and closing by praising it, along with the old joke that when the world ends you should move to Vienna, where everything happens 20 years later.
Remember
Japan's inventions often stalled at home first: consumers ignored contactless payment until it succeeded abroad.
Fitzpatrick's strongest diagnosis is institutional: universities and firms treat storytelling and marketing as peripheral, so strong research stays invisible.
Slowness protects only selectively: the same inertia credited with saving jobs coincides with 30 years of flat real wages.
Isolation. The "3" theory holds that Japan's standards and business habits evolved, like island species, into forms that don't survive elsewhere. Weak English and overengineering compound this.
Contempt for storytelling. Tokyo University's prestigious startup factory has one part-time PR rep, who doesn't speak English. "What use is a mere 'story' when you have specs!?" he was told. Fitzpatrick speculates that this echoes the feudal classing of entertainers and storytellers as 4, literally "non-humans."
Pace. Decisions take months or years of 5, the slow groundwork of building trust before anything is formally decided.
Trauma. After the 1991 crash, Japanese retreated into safe jobs and investments. A 2024 Bank of Japan survey found many firms had curbed investment and built up cash after the crisis and later shocks, though Fitzpatrick concedes that doesn't explain the whole inward turn.
Punishment. Shuji Nakamura, the inventor of the white LED, originally received a corporate bonus of 20,000 yen, roughly $200.
