President Joe Biden’s actions over many months suggest that Israel can determine when and where the United States goes to war in the Middle East. That is unacceptable, and the next American president must change this dynamic.

In one framing, the past 12 months have witnessed a remarkable display of America’s might and resolve in the Middle East—especially relative to our principal adversary in the region, Iran. Since October of last year, Israel has severely degraded Iran’s two most important affiliates in the area, Hezbollah and Hamas. Iran has lashed out directly only twice, with one ballistic-missile assault in April and another this month—both largely neutralized by U.S., Israeli, and allied air and missile defenses.

The United States, in contrast with Iran, has backed its principal affiliate in the region, Israel, to the fullest extent. It has shipped billions of dollars of military equipment and munitions to Israel over the past 12 months, on top of the roughly $3.8 billion it already provides annually; shared sensitive intelligence to allow Israel to target Hamas’s senior leaders and recover its hostages; and repeatedly deployed its own troops to defend Israel from assault. Israeli Prime Minister Benjamin Netanyahu’s assertions that Israel stands alone are at once laughable and insulting.

[Read: The choice America now faces in Iran]

Yet few people in the Middle East, or at home, would view the United States as particularly strong at the moment in the region. The reason is that the Biden administration has made abundantly clear over the past year that it has chosen not to dictate the terms of its own Middle East policy. It has repeatedly allowed Netanyahu and the rest of Israel’s leadership to do so instead.

In April, Israel conducted an air strike in Damascus on a facility adjacent to the Iranian embassy. The United States received no warning about the strike; Biden and his advisers were caught unaware. The strike killed seven Iranian officers. Then Iran and its affiliates in the region launched a barrage of missiles at Israel. But the United States and several of its partners—most notably Jordan, France, the United Kingdom—helped blunt the attack with a coordinated display of air and missile defenses.

With that, a Rubicon had been quietly crossed. Israel had always boasted that a generous supply of U.S. arms allowed Israel to fight its own fights, and that no American soldier had ever been asked to fight Israel’s battles for it. But America has tens of thousands of troops semipermanently garrisoned in the region, in part to respond to contingencies involving Israel, and by interceding to thwart the missile attack, American troops were fighting directly on Israel’s behalf.  

The situation in April repeated itself this past week, when Israel dramatically escalated its military offensive in Lebanon. No one should mourn the late Hezbollah Secretary-General Hassan Nasrallah. But Israeli aerial and ground assaults have displaced more than 1 million Lebanese, and America was once again forced to commit its troops, including two Navy destroyers in the eastern Mediterranean, to repelling an Iranian ballistic-missile attack. This is now a pattern: Israel escalates the conflict, Biden and his team do nothing to stop it, and America follows Israel into war.

Reasonable people can and will argue that the killing of Nasrallah and the destruction of Hezbollah are in America’s interest. But America’s leaders should be the ones making the decisions here, not Israel’s. Instead, and at each step of the conflict over the past 12 months, Biden and his advisers have ceded questions of strategy to Israel, in part by giving Israeli decision makers the benefit of the doubt at every major juncture. Previously established boundaries, such as the demand that Israel not march into Rafah this past summer, have been ignored as soon as Israel crosses them.

Neither the Trump administration nor the Obama administration behaved this way. As different as they were, each administration owned its Middle East policy and dictated policy to Israel, not vice versa. The Trump administration killed Iranian General Qassem Soleimani (which surely delighted Israel) and also unilaterally announced a withdrawal from Syria (which surely did not delight Israel). The Obama administration, meanwhile, negotiated the nuclear deal with Iran over strenuous Israeli objections, because it assessed the deal to be in U.S. interest.

This is the way things are supposed to happen. A superpower does what it understands to be in its interest, and its partners in the region adjust. The Biden administration, by contrast, is acting like a dog that has decided that its own tail should wag it.

Franklin Foer’s account of the Biden administration during this crisis makes for maddening and essential reading. Biden and his advisers are consistently confused as to why a strategy regarding Israel made up of all carrots and no sticks isn’t affecting Israeli decision making. The president is surprised and upset by an Israeli prime minister who is ungrateful for American support and consistently does what is in his own interest without regard for his patron in Washington, D.C.

Biden, alone among major Democratic politicians, has a strong and warm relationship with Netanyahu. He seems to think that this rapport, along with the U.S. president’s powers of political persuasion, will somehow trump Netanyahu’s well-established and well-documented pathologies, which have frustrated American policy makers since James A. Baker. President Clinton famously asked, after meeting Netanyahu in 1996, “Who is the fucking superpower here?”

Netanyahu doesn’t care about his friendship with Biden, or even about Israel’s dependence on the United States. He cares only about his near-term political interests. Everyone but Biden can see this.

[Read: Lebanon is not a solution for Gaza]

Many Arab American voters are fed up, and understandably so. Some of them see Donald Trump as a stronger leader than Biden because, let’s be honest, when it comes to the Middle East, he appears to be a stronger leader, or at least more assertive about U.S. interests. And the average voter can reasonably doubt that Biden’s vice president will be much different from him as president. This war could thus cost the Democrats a Senate seat in Michigan, which has a large Arab population, as well as the presidency itself given how narrow Kamala Harris’s lead in the state remains. Just yesterday, an American citizen from Michigan was killed in Lebanon. Yet when Palestinian Americans in Gaza or Lebanese Americans in Lebanon are killed, the response from their president is little more than a shrug, as if to say, What else can we do?

That was, in fact, the conclusion that Biden’s team reached last month, according to Foer’s reporting:

Over the course of two hours, the group batted ideas back and forth. In the end, they threw up their hands. There was no magical act of diplomacy, no brilliant flourish of creative statecraft that they could suddenly deploy.

With this president, they may be right. Biden has made clear that his Middle East policy will be decided in Jerusalem, not Washington.

But Israel is not going to stop. As Thomas L. Friedman once observed, Israel’s mentality has always been: If I am weak, how can I compromise? Yet if I am strong, why should I compromise?

An American president has to be the one to say “enough.”

But it will probably not be this American president.

Ever since the Supreme Court overturned Roe v. Wade, Democrats have considered abortion a winning issue and have been eager to talk about it. Emphasizing reproductive rights helped the party achieve victories in the 2022 midterm elections and has generated enthusiasm for Kamala Harris’s presidential campaign.

But some abortions Democrats would rather not discuss: those that occur in the final months of pregnancy. Democrats tend to brush off questions about whether these abortions should be restricted, either by denying that their policies would allow abortions late in pregnancy or by pointing out that these abortions are rare, implying that they are therefore not worth our moral concern.

In the recent vice-presidential debate, Tim Walz sidestepped a question about a relatively permissive abortion bill he signed into law in Minnesota. And in the presidential debate before that, when Donald Trump pointed out that Roe had allowed for abortions in the seventh, eighth, and ninth months of a pregnancy, Kamala Harris plainly said, “That’s not true.”

[Read: Trump and Vance are calling their abortion ban something new]

It’s true that third-trimester abortions are rare. But they do happen. Representatives from the CDC, the pro-abortion-rights Guttmacher Institute, and the anti-abortion Charlotte Lozier Institute told me that national data simply aren’t available. But Colorado, which is home to clinics that perform third-trimester abortions, recorded 137 third-trimester abortions in 2023. That’s only one state—eight other states, plus Washington, D.C., have no restrictions on third-trimester abortions. Just a few minutes from my office building in D.C., a clinic offers abortions up to nearly 32 weeks. In nearby Bethesda, Maryland, a clinic performs abortions up to 35 weeks’ gestation.

Those who support such expansive abortion laws tend to argue that third-trimester abortions are the result of a devastating medical diagnosis. In many cases that’s true, but it is not always the situation. The D.C. clinic I mentioned above confirmed by phone that it performs abortions for any reason. Data on the reasons women have later abortions are also scarce. But when The Atlantic’s Elaine Godfrey interviewed a doctor who specializes in late abortions, he estimated that about half of his patients have healthy pregnancies. Of course, some of his patients are in serious distress for other reasons; some are victims of sexual assault, or are teenagers who didn’t realize they were pregnant. This leads to another logical flaw in how the pro-abortion-rights crowd tends to frame its argument.

The group complains that people are overly focused on exceedingly rare third-term abortions. But abortions after a pregnancy from rape or incest are also comparatively rare, and abortion-rights supporters still push opponents of abortion to take these rare scenarios into account. Discussions about third-trimester abortions should therefore be fair game as well.   

Downplaying third-trimester abortions isn’t necessary for Democrats to protect reproductive rights, and could well alienate the plurality of voters best described as abortion moderates. The grim reality of later abortion is simply too much for most Americans to countenance—and reasonable policy makers should listen to them.

Most Americans believe that third-trimester abortions should be restricted. If Democrats want a platform that truly reflects majority opinion, they should address the question of what to do about later abortions, and adopt a position that protects abortions in the first trimester while limiting second- and third-trimester abortions to pregnancies with fetal abnormalities or maternal health crises.

Democrats keep dancing around the fact that, under Roe, states were not required to restrict later abortions. Under Dobbs, which superseded Roe, they still aren’t; they can choose to ban the procedure or allow the abortions without limits. Of course, the fall of Roe means that more states are banning abortion altogether.

[Read: The abortion absolutist]

But the fact remains that Americans are broadly uncomfortable with third-trimester abortions. A 2023 Gallup poll found that although more than two-thirds of Americans believe abortion should be legal in the first trimester, just 22 percent think it should be legal in the third. And a 2021 Associated Press poll found that just 8 percent of respondents believe that third-trimester abortions should be legal in all cases.

When Democrats hammer home just how rare later abortions are, they’re making an important point: More than 90 percent of American abortions take place in the first trimester. A reasonable platform would adopt the Western European standard, in which abortion is legal for any reason in the first trimester, but later procedures are restricted except in cases of devastating maternal or fetal medical diagnoses. Preserving women’s right to choose does not require Democrats to adopt an extreme position that allows for abortion at any stage of pregnancy, no questions asked.

Special Counsel Jack Smith’s recent filing to the D.C. District Court in the Trump v. United States presidential-immunity case both fleshes out and sharpens the evidence of Donald Trump’s sprawling criminal conspiracy to overturn the 2020 election. To understand the filing’s larger significance as well as its limitations, we must first review a bit of recent history.

In its shocking decision on July 1 to grant the presidency at least presumed immunity from criminal prosecution for all official acts, the Supreme Court’s conservative majority showed once again that it was intent on immunizing one president in particular: Donald Trump. The Court majority’s decision, delivered by Chief Justice John Roberts, was explicit. It held, for example, that Trump’s alleged efforts to pressure then–Vice President Mike Pence into voiding the 2020 election results on January 6 constituted “official conduct” from which Trump “is at least presumptively immune from prosecution.” That presumed immunity, the Court contended, would disappear only if the prosecution could convince the courts that bringing the case to trial would pose no “dangers of intrusion on the authority and functions of the Executive Branch.”  

The Court thus remanded the case back to the D.C. District Court to decide the matter, along with the question of whether Trump is actually immune to the rest of the charges against him. How, though, could the prosecution of a president or former president over an “official act” fail to intrude on presidential authority? Seemingly, anything pertaining to Trump’s contacts with the vice president as he presided in his constitutional role as president of the Senate—as well as Trump’s contacts with the Department of Justice, which the Court also singled out and which the prosecution, significantly, felt compelled to omit from its revised indictment—deserves, as the Court sees it, virtually ironclad protection, a powerful blow against the entire January 6 indictment.    

Although the sweeping outcome of Trump v. United States took most legal commentators by surprise, its protection of Trump was completely predictable given the Court’s previous conduct regarding the January 6 insurrection. The refusal of Justices Clarence Thomas and Samuel Alito to recuse themselves from any matter related to the insurrection, despite their own conflicted positions—Thomas due to the direct involvement of his wife, Ginni Thomas, in the subversion; Alito because of his flag-waving support of Trump’s election denials—has received the most public attention concerning the Court majority’s partisan partiality. But another set of telltale signs becomes apparent after a closer tracking of the Court’s decision making.

Almost as soon as the case against Trump came before D.C. District Court Judge Tanya Chutkan, the Supreme Court played along with the Trump lawyers’ efforts to delay the trial until after the November 2024 election. First, after Chutkan ruled against Trump’s absolute-immunity claims in December 2023, Special Counsel Smith asked the Supreme Court to expedite matters by hearing the case immediately, not waiting for the U.S. Court of Appeals to rule on Trump’s appeal of Chutkan’s decision. The Supreme Court refused. Two months later, though, when the appeals court ruled against Trump and set a new trial date, the Supreme Court dragged its feet for as long as possible before announcing that it would take up the case after all. It then set the date for oral arguments as late as possible, at the end of April. This meant that even before hearing the case, the Court made it highly unlikely that Trump’s trial would proceed in a timely manner, effectively immunizing Trump until after the election.

Although radical in its long-term reconstruction of the American presidency, the ruling more immediately affirmed and extended the Court’s protection of Trump from prosecution. By remanding the case to the D.C. Circuit Court to decide what in the indictment constitutes official (and, therefore, presumably immune) conduct, the justices guaranteed that no trial would occur until after Election Day. After that, meanwhile, should Trump win the election, no trial would occur at all, because he would certainly fire Smith and shut down the proceedings.

Smith’s filing tries to slice through the Court’s security shield regarding the insurrection. Skillfully quoting from or alluding to language in the Court majority’s own opinion, the filing demolishes the notion that Trump’s activities, culminating on January 6, deserve immunity. Outwardly, Smith’s filing respects the Court’s dubious ruling about the immunity of official presidential acts. Legally, Smith had no choice but to operate within that ruling, a fact that sharply limited how far his filing could go. But even though it never challenges the conservative majority directly, the filing makes a case, incontrovertible in its logic and factual detail, that the core of Trump’s subversion involved no official actions whatsoever. It persuasively argues, with fact after fact, that Trump was the head of an entirely private criminal plot as a candidate to overthrow the election, hatched months before the election itself.     

In remounting his case, Smith has taken the opportunity to release previously unknown details, some of which he says he doesn’t even plan to present at trial, that underscore the depravity as well as the extent of Trump’s criminal actions. Consider, for example, Smith’s telling of Trump’s reaction to the news from one of his staff, at the height of the violence on January 6, that his tweets attacking Pence had placed Pence’s life in extreme danger. “So what?” Trump reportedly replied. He had clearly intended for his tweets to reach the mob at the Capitol. His nonchalance about the vice president’s life epitomizes the lengths to which he would go to complete his coup d’état.

But the real force of Smith’s filing is in its tight presentation of the evidence of a criminal conspiracy in minute detail, dating back to the summer before the 2020 election, when Trump began publicly casting doubts on its legitimacy should he not be declared the winner. “The only way they can take this election away from us is if this is a rigged election,” he told the Republican National Convention in his nomination-acceptance speech in August 2020.      

From that point forward, Trump was at the center of every effort to keep him in power, even once he was fully aware that he had no grounds to contest Joe Biden’s victory. There were his private operatives sowing chaos at polling places and vote-counting centers, the scheming to declare victory on Election Night before the results were in, the bogus legal challenges, the fake-elector fraud, the plot to deny official certification by Congress on January 6, and finally the insurrection itself. “It doesn’t matter if you won or lost the election,” one witness reports Trump saying. “You still have to fight like hell.”

The crucial point to which the filing unfailingly returns is that none of Trump’s actions listed in the revised indictment, even those that the Court cited as “official,” deserves immunity. As Smith makes clear, the Framers of the Constitution deliberately precluded the executive branch from having official involvement in the conduct of presidential elections. The reason was obvious: Any involvement by a president would be an open invitation to corruption. To make the case that any such involvement falls within a president’s official duties would seem, at best, extremely difficult.       

It is here that Smith turns the Court’s Trump v. United States ruling to his own advantage. Concerning specific charges that Trump’s speechmaking contributed to the insurrection, the Court allowed that “there may be contexts in which the President speaks in an unofficial capacity—perhaps as a candidate for office or party leader.” Quoting from an earlier Court decision, the ruling then states that determining these matters would require that the district court undertake “objective analysis of [the] ‘content, form, and context’” of the speeches in question, a “necessarily fact-bound analysis.” Likewise, regarding the allegations apart from Trump’s supposedly official communications and public speeches, the justices enjoined the district court, on remand, to “carefully analyze” those charges “to determine whether they too involve conduct for which the President may be immune from prosecution.”     

Citing those exact phrases as the Court’s standard of inquiry and proof, Smith then offers evidence that every count in the revised indictment concerns either technically official conduct undeserving of immunity or unofficial conduct involving Trump’s private actions as a candidate and not his official duties as president. These actions include his efforts to pressure state officials, preposterously presented by Trump’s defense attorneys as official inquiries into election integrity. They include his conversations about elector slates, about which the president has no official duties. They also encompass all of his speechmaking about the allegedly crooked election, up to and including his incitement at the January 6 rally at the Ellipse, which was not an official function.

Above all, Smith nails down a matter that the Court’s opinion went out of its way to declare “official” and presumably immune: Trump’s efforts to pressure Pence into declining to certify Biden’s win. Although the filing acknowledges that the Court had held that these conversations between Trump and Pence about “their official responsibilities” qualified as “official,” it rebuts the presumption that those discussions therefore qualify as immune. The filing observes that the discussions did not concern Pence’s duties as president of the Senate “writ large,” but only his distinct duties overseeing the certification of a presidential election—a process in which a president, whether or not he is a candidate for reelection, has, by the Framers’ considered design, no official role.   

Here the logic of Smith’s argument cuts to the quick. By the Court majority’s own standard, as stated in its Trump v. United States decision, the presumption of immunity for official actions would disappear only if a prosecutor could demonstrate that bringing criminal charges against a president or former president would not present “dangers of intrusion on the authority and functions of the Executive Branch.” Because certification of a presidential election, the subject of Trump’s “official” pressuring, involves neither the authority nor the functions of the executive branch, the immunity claims concerning that pressuring are therefore groundless—according to the Court majority’s own logic.           

The rest of Trump and Pence’s interactions do not even qualify as official, Smith shows. In all of their other postelection, in-person conversations and private phone calls, Trump and Pence were acting not in their capacities as president and vice president but as running mates pondering their electoral prospects, even after Biden had been declared the winner. If, as the Court itself has stated, context is important with regard to speechmaking, so it is important with regard to communications between the top officials of the executive branch. To be sure, Smith allows, Trump and Pence “naturally may have touched upon arguably official responsibilities,” but “the overall context and content of the conversations demonstrate that they were primarily frank exchanges between two candidates on a shared ticket”—strictly unofficial conduct.

In all, by recasting the case against Trump in view of the Court’s immunity decision, Smith has drawn upon that very ruling to establish that none of Trump’s actions in connection with January 6 cited in the revised indictment is immune from prosecution. And in doing that, he has further discredited an already discredited Supreme Court.

Unfortunately, important as it is with respect to Smith’s specific case, the filing cannot come close to undoing the damage that Trump v. United States has wrought, with its authorization of an authoritarian American regime. The very fact that Smith had to omit from both his revised indictment and his filing Trump’s nefarious but official dealings with the Justice Department, including his brazen hiring and firing of top law-enforcement officials on the basis of who would do his personal bidding, shows how fearsomely the Court’s immunity decision has constrained the special counsel. There was a great deal more criminal behavior by Trump and his co-conspirators, as laid out in detail in the House January 6 committee report, that Smith could not touch because the Court has effectively immunized it as “official” activity under the executive branch’s authority.

These limitations show all over again how the Court has given the president absolute license to rule like a tyrant, against which even the ablest special counsel is virtually powerless. Nothing in Smith’s filing alleviates Justice Sonia Sotomayor’s judgment in her forthright dissent in Trump v. United States that the decision empowers the president, acting in his official capacity, to order the assassination of political rivals, to take a bribe in exchange for a pardon, to organize a military coup with impunity: “Immune, immune, immune.” That Smith managed to outsmart the Court as much as he did is a remarkable feat that could have important results—but only if Kamala Harris succeeds in winning the presidency.

On the basis of their past decisions, it is reasonable to expect that both the D.C. district court under Judge Chutkan and the U.S. Court of Appeals will rule in favor of Smith. Trump v. United States would then go once again before the Supreme Court. This will happen if Harris wins the election, because a Justice Department under her administration would almost certainly allow Smith to remain to continue prosecution of Trump. What, then, would the Court do? Would it uphold those decisions and throw Trump upon the mercy of a D.C. federal jury? Or would it strike those decisions down, thereby redoubling the disgrace it earned the first time around?  

The only way the Court can avoid that dilemma is if Trump wins the election, an outcome that its conservative majority would now have all the more reason to desire. But what happens if, as seems highly possible, the election leads to litigation, much as the 2020 election did, only this time the Court is left to make the final decision? Will the Court then intervene as Trump’s enabler once again, installing him as a constitutionally tainted president, allowing him to kill the indictment against him, and to pardon those convicted of violent crimes in the attack on the Capitol whom he calls “hostages”? The Court, in Trump v. United States, claimed that it was protecting the sanctity of the presidency, but if it aids Trump in his attempt to escape justice for his January 6 insurrection, it will further seal its illegitimacy while also sealing MAGA’s triumph—and, with that, the majority of Americans, not to mention the rest of the world, will pay a crushing price.

Cam and Dan Beaudoin’s three-decade-old problem began when they were kids. Dan would follow his big brother around. Cam, who’s about three years older, would distance himself. Dan would get mad; Cam would get mad back. Although their mom assured them that they’d be “best friends” some day, nothing much changed—until about three years ago, when a fight got so bad that the brothers stopped talking to each other completely. Dan left all of their shared group chats and unfriended Cam on LinkedIn.

But the brothers, who didn’t speak for about a year and a half, started to understand the gravity of this separation: They were on track to live the rest of their days apart, silently stewing. Dan reached out to Cam, and he told me they agreed: “We don’t want to be in our 70s and 80s wishing we had a relationship.” So they tried something that most siblings never do: They decided to go to therapy together.

Sibling therapy is a relatively unusual practice, which is perhaps surprising considering how common childhood sibling conflict is. It’s so common, in fact, that many siblings assume that the relationship will work itself out naturally. But sometimes it doesn’t. And the tenor of the bond can affect people for their whole lives, either protecting them through hardship or leaving them vulnerable to it. One study tracked participants for 30 years and found that closeness with a sibling in childhood (but not with a parent) predicted their level of depression by age 50. Another concluded that the warmth a subject felt from their sibling at age 23 predicted lower levels of depressive symptoms in middle adulthood, while the hostility they felt predicted anxiety and depressive symptoms. A sibling relationship “is like a shadow on us,” Geoffrey Greif, a University of Maryland School of Social Work professor and co-author of Adult Sibling Relationships, told me. “It is one that always is there, whether it’s the most wonderful relationship in the world or the most troubled.”

Still, psychologists tend to overlook the influence of siblings. Roughly 80 percent of Americans have at least one, but a 2012 review of literature between 2008 and 2018 found that fewer than 3 percent of close-relationship studies looked at sibling ties. Most of the research focuses on parent-child bonds, spouses, or peers, Shawn Whiteman, a Utah State University researcher who co-authored that paper, told me. And few resources or training materials exist for clinicians who might be interested in sibling work; as a result, few practice it.

But when I spoke with clinicians who do, they told me that sibling relationships seem to be getting more public attention. Erin Runt, a therapist who works with siblings, thinks that despite the recent increase in family members becoming estranged—or perhaps in response to it—a counter-impulse is brewing: a realization that severing a painful connection doesn’t always end in healing. “We have a generation of people who are more and more interested in repairing relationships,” she said. Perhaps they’re just accepting that as hard as they’ve tried, they still haven’t outrun the shadow.

Many people think of parent-child relationships as the most formative ones—the source of attachment styles, daddy issues, tragic flaws. But siblings tend to model how to interact with peers, to share and to compromise (or not), to fight and make up (maybe). They’re the ones who can make a difficult home situation much better or much worse, Runt told me. They form their identities in relation to each other: what the other excels at or struggles with, how others—perhaps, especially, parents—compare them. And siblings compete for resources growing up. “You’re kind of always assessing,” Runt said. “What are they getting versus what I’m getting? How am I treated versus how they’re treated? What version of Mom or Dad or other family members are they getting, versus what I get?”

Karen Gail Lewis, a Washington, D.C., therapist and the author of Sibling Therapy: The Ghosts of Childhood That Haunt Your Clients’ Love and Work, had already been practicing for about 15 years when she started to understand just how much those dynamics might have shaped her clients. A deeply depressed man who began seeing her said that he’d already tried other therapists who hadn’t helped. So she asked him to come in with someone who knew him well. “I assumed he would bring his wife,” she told me. He brought his brother.

Lewis realized that without asking her clients specifically about their siblings, she might have been missing a key to their problems. She believes now that many people feeling stuck in a friendship or romance are to some degree stuck in patterns from early siblinghood—what she calls a “laboratory for learning everything you need to know in your adult relationships.” Runt told me she sees this in her work, too—perhaps one sibling felt they had to take care of the other, for instance, and then in adulthood they’re always trying to solve others’ problems.

[Read: What it’s like to visit an existential therapist]

Discussing siblings in individual therapy can be illuminating, but coming in together might be the only way to mend a rift—whether between pairs or a whole group. (Lewis has worked with a family of 10 siblings.) Sibling therapists told me that many clients go decades before seeking help; busy with their new family, perhaps they ignore pain stemming from the family they grew up with. Then their kids get a little older, and as they get closer to the end of their life, they start to reflect. “I have had siblings who have not talked to each other for five, 10 years,” Lewis told me. “Now they have more energy to be angry.” Other times, some event will force them together, such as a parent getting sick and needing care, or dying and leaving logistics for the siblings to tackle. Perhaps the siblings want their own kids to have a cousin relationship, or one child is getting married and invites an aunt or uncle to the wedding. Sometimes a person just realizes they don’t want to die before reconciling with their sibling.

People often reach this conclusion in midlife. Lewis works largely with clients in their 40s to 60s, though she does see a set of sisters in their 90s. She referred to this as the “hourglass pattern” of a sibling relationship: a lot of contact in the beginning of life, then a period of distance—then a reunification before it’s too late.

The Beaudoin brothers, who are in their 30s, saw going to therapy together as a final effort to rebuild their relationship. “We really only had one shot,” Dan said. Lewis, they told me, essentially served the role of a parent wrangling her squabbling little boys. She wasn’t afraid to jump into their arguments. She would cut off their tangents or call them out on deflecting blame, and then she’d pull them back to important questions about, say, their communication patterns. (Sometimes sessions do get heated, Lewis told me, but other times siblings are so distanced that they’re too polite. “They talk from their head,” she said, in which case she tries to “get that heat out.”) Through that process, Cam and Dan finally started to see the problems with their own interactions—the same issues that had plagued them for so many years. Dan, a talker, would exasperate Cam when he went on for too long; Cam would shut down, and his silence upset Dan. With Lewis, the brothers developed hand signals and code words to quickly show when a conversation is feeling too emotionally intense—and also the language to say “I don’t want to talk about that,” Cam told me.

[Read: The great cousin decline]

Sibling therapy isn’t all about hashing out current conflicts, though; it also involves going back in time. Because siblings typically grow up in the same home, they tend to assume they had similar childhoods. But often, that’s not true. “You may have had completely different experiences based on how skilled your parents were at parenting at that point in time, what stressors were involved, what money looked like, and also your birth order,” Runt told me. (Studies suggest there’s no consistent effect of birth order, but that doesn’t mean that it doesn’t influence dynamics—just that it likely does so in different ways for different families.) Sibling therapists ask questions to see where narratives match and diverge, what Runt calls “comparing notes.” What was your sister’s role? Who was your mom’s favorite? What was it like at the dinner table? That context can explain a lot about a sibling’s behavior.

The goal of this work isn’t necessarily to become best friends. Sometimes it’s to survive family gatherings or to come together to care for a sick parent; other times, it’s about letting go of pain, even if you don’t keep in touch after you leave the therapy room. You could think of sibling therapy, in those cases, as harm reduction. Not all siblings can or should be in each other’s lives. But the ones who manage it get something unique from one another, however fraught their time together is: Even if a sibling didn’t have the same childhood as you, they’re probably likelier than anyone else in the world to understand where you came from. They might be the only one who can tell you, with firsthand knowledge: “Your experience is real,” Runt said. Without that, “we’re the only one who knows that and can validate that for ourselves. And that’s a lonely place to be, right?”

A lot of Americans are in a lonely place, and most of the proposed solutions focus on friendship: cultivating relationships with the people we’ve chosen to hang out with and build our adult lives with, rather than the ones we got stuck with at birth and might see only at Thanksgiving. But Robert Waldinger, a psychiatry professor at Harvard Medical School and the lead author of the study finding that sibling closeness predicted depression at age 50, told me that many siblings know they can count on one another when needed, even if they’re not emotionally close—giving one another what he called “instrumental support.” In a more recent study, he asked middle-aged participants: “Who could you call in the middle of the night if you were sick or scared?” They could list as many names as they wanted—and some people didn’t have a single one to write down. But many, he said, listed a sibling.

The siblings I spoke with said that their relationships hadn’t been completely transformed by therapy so much as subtly shifted, enough to delicately start again. Cam and Dan are working on keeping expectations for each other realistic: They don’t need to talk all the time, or agree on all the things that once tore them apart. But they’re moving ahead—not just for themselves, but to show their kids what a healthy sibling relationship can look like. “We have to build that,” Dan said. “We have to be the ones to model it.”

​​When you buy a book using a link on this page, we receive a commission. Thank you for supporting The Atlantic.