§T. Greer (@Scholars_Stage) — A Persona Analysis

T. Greer is the Director of the Center for Strategic Translation and writer of The Scholar's Stage, an essay blog he has been running for more than a decade — first at scholars-stage.org, now mostly at scholarstage.substack.com. He lives in Virginia. By day he reads and translates Chinese Communist Party documents; in his off-hours he writes long, citation-heavy reads on grand strategy, American conservatism, and ancient history, with bylines in Foreign Affairs, American Affairs, Palladium, City Journal, the LA Review of Books, and the NYT. The 100-tweet, 60-reply slice analyzed here covers late February through April 2026 — a window dominated by the second-Trump Iran war, the slow-burn debate over a "techno-nationalist" American elite, and his own essay "China and the Future of Science."

§I. Core Worldview & Mental Models

Greer's worldview is best described as historicist realism. The historicist part: he reaches reflexively for analogues from late-Han China, Suez 1956, the Seven Years' War, the Pacific War, the WASP establishment of 1870–1930. Almost no claim he makes is delivered without a historical reference inside three sentences. The realism part: he treats interests, demography, and institutional capacity as the load-bearing variables, and treats sentiments — including his own tribe's sentiments — as variables that need to be argued down before policy can be argued up.

A representative move: when defenders of Trump's NATO posture float exiting Finland, he answers in pure realist register: "The U.S. remains in NATO. A NATO with Finland is more lethal and defensible than a NATO without it. By a large margin… Any nation who hates us more than Turkey should go fend for themselves" (Apr 26, 167L). Then in a reply: "we live in a world where we did include the Baltics, and have to live in that world, not the one Mersheimer wishes we had chosen instead. To me it is really that simple" (to @ConservaWonk, Apr 27).

His most-reached-for mental models in this corpus:

• Telos and trendline. Borrowed from his reading of Ken Jowitt on Leninist systems, applied to China: "the Chinese system has a telos…Either it hurtles towards some clearly defined goal or things start to fall apart" ("China and the Future of Science," scholarstage.substack.com, Mar 2026). When critics push back on his China-rise claims, he answers: "you guys are missing the point. As with AI development the real story here is about the trendline" (Mar 21, 39L, replying to himself).

• Structural inevitability vs. contingent implementation. He told Jordan Schneider on ChinaTalk in April 2025: "there was no path back to 2016 at this point" — the trade reckoning was structural; only its choreography was Trump's. He uses this same hammer on Trumpism in 2026: at six years in, "Failures and frustrations can no longer be passed onto the old bogeymen" (Apr 19, 42L).

• Cui bono done historically. Asked whether the present U.S. setback in Iran is comparable to Suez, he answers: "the American equivalent drops the United States from unipolar superpower to one of two superpowers… Consider where France was after the Seven Years War. Or Britain after the American revolution. Both far more devastating defeats than this" (Apr 8, 167L + 31L reply). This is his characteristic move: borrow the energy of a historical-decline claim, then de-escalate it by citing harder cases.

• "Historical inevitability is an intellectual drug." This line, dropped in passing in an exchange with @W_T_Han about a (fake) Diplomat-style "350 years of survey data" tweet, is Greer's anti-determinist hedge against his own determinism: "We must not confuse the fruits of prolonged effort and conscious choice for the natural state of things" (Apr 11, 39L). He keeps both moves in tension — China is on a trendline; outcomes are contingent — and he does not appear to have resolved the contradiction.

His intellectual DNA is genuinely unusual. From the corpus and essays you can triangulate four distinct lineages stacked in one head:

1. Cold-War realism, in the Kissingerian register. He defends Kissinger's actual record on Iraq against young right-wingers who have remade Kissinger into a non-interventionist (Apr 19, 171L; he goes so far as to send the @Polit_eurOpines / @TrueConnallyite contingent to "look up the op Eds he wrote in 2004").
2. Classical Chinese political thought. When his wife reports that "Xunzi is the only one in this book who speaks any sense," he counts it a marriage-vindicating moment (Mar 22, 249L). His "Making Sense of Chinese History" reading list is one of the longest-circulating things he's written.
3. Mid-century American narrative nonfiction. His Apr 2 list of writers all born within one generation — Mailer, Manchester, Capote, Baldwin, Edward Abbey, Matthiessen, Wolfe, McPhee, Talese, Didion, Halberstam, Caro, Sheehan, HST, Rhodes — is presented as a moral reference point, not a curiosity (137L). "The very best of their work is as good as any poem or novel in our language." He returns to Robert Caro and David Halberstam-style biography as his template for what serious writing should look like.
4. Contemporary techno-conservative & national-conservative discourse. He reads James Poulos's Palladium essays the day they go up ("explicitly hearkens to Thiel's 'The Straussian Moment'… Will be pondering this one for some time," Apr 28). His own American Affairs essay (cited but not directly quoted in tweets — "intothefuture45" calls it "one of the best pieces of essay writing I've read for a long time," Apr 21) sits in this scene.

The blind spots worth naming:

• AI capability uncertainty. His AGI/Hormuz challenge — "if America had AGI, how would it help us open up the Straits of Hormuz?" (Apr 13, 293L) — generated 81 replies, many of them substantive (cyber dominance over Iranian C2; persistent satellite tasking; clone-IRGC-commander deepfake ops; cheap PV/battery making the strait less critical). He took none of them seriously enough to update; his next-day follow-up was "Answers to this question universally uninspiring" (Apr 14). The framing — "what could AGI do that we cannot do already?" — assumes the bottleneck is doctrine, not the cost-curve. This is a pattern: he is incurious about technological capabilities he has not personally validated.
• Reasoning by anecdote about AI text. His most-shared AI take of the window — "Increasingly humans are trained on AI writing, not the other way around" (Mar 23, 1583L) — is sourced to "a professor tells me" and survives on its rhetorical inversion, not on data. Asked for details, he supplies one (bullet points in short answers). Replies disagreeing on first-hand grounds get no response.
• Moral vs. prudential arguments collapse together. When he says the AI safety crowd's failure to commit political violence proves they "did not actually believe their own claims" (Apr 8, 32L), he is using a prudential test ("if you knew apocalypse was coming, you'd act") to draw a sweeping moral conclusion. Several sparring partners (e.g. @AntiDisentarian) push back on this; he does not concede.

§II. China & Grand Strategy

This is the day-job domain, and the corpus shows the day job in full force. Three commitments structure his China writing.

1. China is on a scientific telos and the West is in denial. "China and the Future of Science" is the load-bearing essay, and he tags 22 specific accounts under it on Mar 22 (Cowen, Tooze, Burja, Ben Reinhardt, Caleb Watney, Lyman Stone, Bonnie Glaser, Reihan Salam, et al.) — itself a portrait of who he considers the relevant audience for serious China discourse. The argument, distilled from the essay: "For 66 of the 74 categories tracked, a majority of the institutions that are now publishing the highest-impact science are Chinese." When pressed in tweets — "what about citation cartels? what about classified research?" — he answers with the trendline frame and refuses to litigate the levels.

2. Iran is a sideshow. This is his single most distinctive 2026 take, and it is unpopular in his own ideological neighborhood. His "On Bombing Iran" (scholarstage.substack.com, Mar 2026) frames the strikes as a maximalist negotiation gambit that the U.S. lacks the resources to follow through on. The tweet voice sharpens this further: "if the end result of all this is a taxation regime over Hormuz transit jointly administered by Iran and the United States.... lol" (Apr 8, 121L). When Trump-aligned voices try to claim the Iran war as a Chinese setback, he answers: "In the realms where China is most ambitious, Iran is a side show" (Mar 21, 67L). The whole rhetorical scaffolding is: every dollar and every analyst-hour spent on the Persian Gulf is a dollar and an analyst-hour not spent on the Pacific.

3. China-watching itself is broken. This is the through-line of his ChinaTalk appearances and a quieter critique inside the corpus. On the podcast he names the "My Friend From Beida Problem" — Western experts befriend liberal Chinese elites unrepresentative of the actual party leadership and trust their account of CCP intent. The same skepticism shows up tonally in tweets: he reposts Howard Wang's primary-document research with "Howard's research is always worth reading" (Apr 22, 44L) but is conspicuously cool toward more speculative China commentary. His own American Affairs essay (April 2026) is part of the same project: build the techno-nationalist policymaking class on documents, not on Beida coffee chats.

On the IR field as a whole. He thinks the discipline has failed: "In almost all cases journals articles and books in international relations would have been better if they had either been written as straight histories, and to a historian's standard, or if instead they had given the topic over to an econometrician. As things stand the field has the worst vices of both the historians and the statisticians but none of the merits of either. What a disaster" (Apr 17, 63L + 62L). This is consistent with his treatment of the Cowen/economics passage on Mar 28 (446L): "Economics was once the study of economic markets but has become synonymous with 'econometrics'… the tool has swallowed up the original object of study." Fields that mistake their methodology for their object are, for Greer, a recurring civilizational failure.

On Trumpism's foreign-policy structural problem. Six years in, he says the act has run out: "Trumpism has a serious structural problem… It is an oppositional program. Its rhetorical posture and emotional force is predicated on opposition to The Establishment. But here is the catch: at this point Trumpism is the Establishment" (Apr 19, 277L thread). This is the political-theory complement to his Iran-is-a-sideshow argument: Trumpism has won the wars it knew how to fight, and now has to govern, and is bad at it. On ChinaTalk (April 2025) he gave the operational version: Trump "pits these people and factions against each other, then act[s] as the kingmaker" — a management style that solves principal-agent problems but produces "multiple people supporting the same policy for conflicting reasons."

§III. American Conservatism & The Question of Elites

If China is the day job, this is the obsession. Almost every essay he has self-promoted in the past year is some version of the question: who governs America, and where does a serious governing class come from?

The intellectual armature is in "35 Theses on the WASPs" (Jan 2026) and the companion "Book Notes: The Technological Republic" (Dec 2025). The argument across both: America's last real governing class was the Eastern Establishment of 1870–1930, built on Civil War unity, Second Industrial Revolution wealth, and a self-conscious commitment to "the integration and greatness of the American nation." It was reproductive — "it built schools, endowed universities, and founded literal dynasties." Today's tech elite, by contrast, has the wealth but neither the schools nor the dynasties nor the political coalition. Karp's Technological Republic is, in his read, "a series of TED talks sloppily sewn together by the ChatGPT of 2023" — the right intuition (techno-nationalist elite) and the wrong execution (no actual content).

This essay-stack drives a lot of the tweet voice in the window:

• "What, we are talking about the ideology of Palantir's CEO and his desire for a stronger techno-nationalist leadership class? I might have written an essay about that" (Apr 22, 52L) — followed immediately by his own essay link.
• His Harvard tweet (Mar 30, 389L): the anger isn't really about Harvard; it's that "a Harvard degree bestows great benefits on those who hold it—and those are no longer commensurate with what it takes to get the degree…the purpose of Harvard is not to choose America's future elites, but to train them. It is on that latter count it must be judged" (Mar 31, 66L).
• His Pacific War tweet (Apr 6, 1086L): the Toll quote about U.S. naval aviator rescue policy is implicitly an elite-formation argument — "Smart leaders protect and reward their best people," in a sympathetic reply he lets stand.
• His Charlie Kirk essay ("Bullets and Ballots," Sep 2025) is part of the same project from a different angle: Kirk as "one of the most effective institution-builders and coalition-crafters in the United States." Greer does not think bombast is what builds movements; institutions are.

The Trumpism-as-establishment thread is a coda to the elite question. Trumpism inherits the establishment's position but not its project — no schools, no successor class, no theory of the country that survives Trump. That is what makes it structurally weak in his view, even though he agrees with several of its diagnoses.

There is also a quieter thread on Europe. "The Euro-American Split (I): Dread Possibility" (Feb 2025) predicted that Cold War instincts would die with the people who held them: "Cultures change when people with new ideas replace the people with old ones." His April 2026 NATO defense is in some tension with that essay — there he is descriptively neutral about the split; on the timeline he is prudentially hostile to abandoning Finland — but he keeps both registers active.

§IV. Actionable Principles, Reading, and Learning Practice

Greer is unusually open about how he works, and the corpus contains more concrete rules-to-live-by than most accounts of his size. The recurring principles:

• "From my copybook." His Feb 8 tweet (170L) opens with that phrase, then quotes Paradise Lost on Satan's destructiveness. The implication: he keeps a literal commonplace book of passages he wants to remember. This is the practice underneath the constant allusiveness.
• Read the primary thing, not the summary. Implicit across his work; explicit in his attack on China-watching that "data analysts" can't answer "why reduce Uighur births" because they don't read party ideology directly (ChinaTalk, Taiwan episode). Same energy directed at the field of IR (Apr 17) and at Steinbeck readers who treat Grapes of Wrath as social realism (Mar 29, 102L: "The novel does not accurately describe dust-bowl Oklahoma… Better to think of the novel less as a piece of social realism and than as a slice of the American mythos").
• AI writing should be short. "If you are using AI to write, all length signals is disrespect for the reader" (Apr 25, 66L). The companion principle: "I have more tolerance for long and discursive writing when I know the writer had to spend a large amount producing it." This is a standard he applies to himself — his essays are long because he wrote them, and the labor is the meaning.
• Math is worth doing seriously. The Seth Braver linear-algebra recommendation is his single highest-engagement tweet of the window (Mar 28, 1809L), and it's not framed as a hot take — it's a genuine personal endorsement after a personal study program. He asks a separate question about Spivak vs. Apostol for a calculus second pass (Apr 6, 23L). A China analyst grinding through linear algebra in his late thirties or forties is the most under-the-radar thing about him, and it is a real practice, not a posture.
• Generational thinking. His Mar 28 reflection on the Cowen passage uses the same lens: anthropology and economics each reduced themselves to a method ("ethnography," "econometrics") within a generation, and it changed what the field was for. The lesson he draws and applies to his own work: protect the object of inquiry from the tool.
• Sequential, not parallel, work. When he announces a long review essay — "also reading this book. Plan a very long review essay" (Apr 20, 280L) — it is sequential reading, not parallel curation. His Substack averages roughly one essay a month, which is the cadence the practice produces.

§V. Rhetorical Style — What Makes His Tweets Work

The single most striking pattern in his engagement data is that his domain expertise is not what gets liked. The top five posts in the window:

1. The Seth Braver linear-algebra recommendation (Mar 28, 1809L) — a personal study report.
2. The Ukraine/demographics tweet (Mar 28, 1897L counting reply: "decisively settled the question 'can two countries in demographic decline sustain bloody wars of attrition?'") — a structural insight derived from one observation.
3. "Increasingly humans are trained on AI writing" (Mar 23, 1583L) — a rhetorical inversion of a familiar discourse.
4. The Lauren Oyler / Jia Tolentino "rereread the takedown" recommendation (Apr 25, 1095L) — pure literary signal-boosting.
5. The Pacific War / aviator rescue thread (Apr 6, 1086L) — a moral-historical observation passed forward.

Not one of these is about China or Trump foreign policy. The voice his audience rewards most is the voice of the well-read generalist passing forward a useful observation, not the voice of the China specialist. His own China posts — including the announcement of "China and the Future of Science" (194L on Mar 21) — get respectable engagement from his obvious target audience but rarely break out. He knows this; the Mar 22 tag list (22 named accounts) is the move of someone who has accepted that his China audience is vertical and has to be summoned by name.

The rhetorical devices that recur:

• Quote, then extend in a self-reply. Almost every long thought in the corpus is one tweet quoting someone else, then a chain of his own follow-on tweets that read as the actual essay. He uses Twitter as a serialization platform for paragraph-length argument, not as a one-liner format. The "Trumpism has a serious structural problem" thread (Apr 19) is canonical: the focal post is the thesis (277L); the three replies are the argument (44L, 42L, 67L); engagement on the focal post outpaces the argument-tweets by an order of magnitude. Most readers stop at the first sentence.
• The dry "lol" as period-equivalent. "if the end result of all this is a taxation regime over Hormuz transit jointly administered by Iran and the United States.... lol" (Apr 8). "Only Trump can go to Tehran?" (Apr 8). He uses "lol" the way a 1950s columnist used a one-line ironic kicker.
• The historical analogue as de-escalator. He likes inviting alarm and then disciplining it: "Consider where France was after the Seven Years War. Or Britain after the American revolution. Both far more devastating defeats than this" (Apr 8). This is the move that makes him sound calmer than the rest of the timeline without committing him to optimism.
• Self-quote as citation. When the discourse touches a topic he has written about, the move is to quote himself with "I might have written an essay about that" or the essay link in a reply. This is highly effective at producing a serialized essay-feed effect on the timeline.
• Diplomatic in posts, sharp in replies. The reply corpus runs noticeably more clipped and more cutting than the post corpus. Examples: "Read the thread you responded too bud" (to @SimplySatiracal, Apr 16, after a misreading); "Yes the Vietcong was famed for its financial systems" (same thread, sarcastic); "Useless professor, biased, racist and worthless" gets ignored, but he will spend two replies dismantling someone he respects who is wrong (the Iraq-memory thread with @whsieh and pSmith). He does not flame strangers; he reserves the heat for people who should know better.

What the thread replies to his viral hits reveal about how the audience hears him:

• The AGI/Hormuz thread (Apr 13, 81 replies) is dominated by AI/tech specialists who do take the question seriously and answer it on capability grounds — cyber dominance, persistent satellite tasking, IRGC deepfake ops, cheaper energy making the strait less critical. He treats those answers as evasions. The audience that engaged hardest on this post is not, broadly, his audience — they came over from AI Twitter and gave the steelman, and he was already on the next post.
• The Pacific War thread (Apr 6, 14 replies) is mostly his own audience supplying additional anecdotes (pilot rotation, Shattered Sword damage-control argument). One reply pushes back ("This doesn't really come up all that much in the two books I've read") — Greer's response is uncharacteristically warm ("He talks about it quite a bit—my guess is you just forgot. I'll find page numbers later"), suggesting the responder is someone he respects.
• The AI/ESL thread (Mar 23, 37 replies) splits into two camps: people supplying confirming anecdotes ("Around 5th grade my daughter's writing sounded like AI") and people skeptical of the data ("seems like an easy anecdote to confabulate"). He engages neither side; the skeptical replies are unanswered.

§VI. Contrarian & Hidden Takes / Evolution & Tensions

Greer's contrarian takes against his own ideological neighborhood are unusually direct, and worth listing:

• Against the conservative anti-NATO right. "Any nation who hates us more than Turkey should go fend for themselves" (Apr 26, 167L) — defending NATO with Finland against a Trump-aligned questioner. Then, in private-feeling reply mode: "But the point is that we have already committed ourselves to the Baltics — something very hard to defend without Finland" (Apr 27, to @ConservaWonk). He is willing to lose followers over this.
• Against the new-right rewriting of Kissinger. "Some young guy told me this weekend that we never would have gotten into Iraq if America had stuck to a 'Kissigerian foreign policy.' I suppose he never learned what Kissinger's actual position on Iraq was" (Apr 19, 171L). Then: "young politicos have a gap in their education when it comes to history of the more recent past" (Apr 19). A generational rebuke aimed at exactly the demographic that retweets him.
• Against AI doomers. "The fact that such violence did not take place I took as strong evidence that the AI doomers did not actually believe in their predictions" (Apr 8, 32L). Uncharitable but not flippant; he has thought about it and means it.
• Against "AGI as super-weapon." He frames the AGI-Hormuz challenge as an empirical question and treats every reply as inadequate (Apr 13–14). He does not engage the strongest counters.
• Against Tyler Cowen's AI-economics optimism. "Interesting how this is essentially the opposite thesis of @tylercowen's new book" (Apr 24, 35L). Cowen is on his Mar 22 tag list, so this is intramural. He respects Cowen as audience; he does not buy his big call.
• Against Trumpism-as-permanent. "We are in our sixth year of Trump governance… much of what the Biden administration did was closer to Trump's model than to Obama's or Bush's" (Apr 19, 67L). The implicit corollary, which he does not state on-timeline but which is explicit in the American Affairs essay neighborhood: the next governing class is being chosen now, and Trump personally is not going to choose it.

The central tension in his thinking, surfaced repeatedly across the window, is between:

• The structural-determinist register: trendlines, telos, "no path back to 2016," demographic pyramids settling questions. This is the Greer who told ChinaTalk the trade reckoning was inevitable.
• The voluntarist register: "Historical inevitability is an intellectual drug. We must not confuse the fruits of prolonged effort and conscious choice for the natural state of things" (Apr 11, 39L).

He does not resolve this. The pragmatic resolution he seems to operate by: trends are real but which side of a trend you end up on is determined by conscious effort. China is on a scientific trendline because the CCP made it so. The U.S. could be on a different one if its governing class made the choice — but it doesn't have one yet, which is the whole problem.

The evolution over the window is the Iran war moving from his explicit forecast in "On Bombing Iran" (Mar 1, 2026) — the strikes will become an unaffordable distraction — to confirmation by April that he was right: Trump is now floating a joint Iran/U.S. Hormuz tax regime, the strikes-and-sanctions cycle has hardened, and his own line ("we cannot escalate our threats against this regime far beyond what we now are doing") is being borne out in real time. He is not gloating, but the tweet voice has tightened: more "lol," fewer hedges. His dated prediction from the essay — that maximalism without resources produces a long-term escalatory drag — is in early validation as of late April 2026.

What he might say after three drinks that he won't say on-timeline (synthesized from interviews, replies, and tonal inversions): that most of his MAGA-adjacent readers don't know enough recent history to be useful interlocutors and that he is keeping company with them as a cost of the project, not a benefit; that the AI safety crowd's silence under their stated beliefs is genuinely contemptible to him and not just rhetorically convenient; that he thinks Tyler Cowen, whom he respects and tags, is going to be wrong about AI in a deep way; that he is doing self-study math because he expects to need it and does not expect his guild to.

§VII. Network Graph

The cleanest single window into Greer's network is his Mar 22 self-promotion tweet for "China and the Future of Science." He tagged: @deanwball, @DouthatNYT, @Ben_Reinhardt, @mattparlmer, @tylercowen, @palepurshankar, @ShivshankaMeno, @makosloff, @adam_tooze, @elyratner, @calebwatney, @SamoBurja, @timhwang, @reihan, @aphysicist, @mualphaxi, @tszzl, @ByrneHobart, @shirleyzeyu, @BonnieGlaser, @lymanstoneky. That is the inner audience-of-record for serious China and grand-strategy discourse, and several of them turn up across the corpus as conversational partners (Cowen, Stone, @aphysicist, @SamoBurja, @mualphaxi).

His inner circle as recoverable from the reply corpus:

• @yuanyi_z — the only handle to which he says something genuinely vulnerable: "I am so beaten down by all these takes that it is he's to respond to them anymore; thanks for waging the good fight" (Apr 22). That is friendship.
• @deanwball — short familiar exchanges; uses the bare first name "Dean" (Apr 11). Shows up in the tag list.
• Howard Wang (@howardgwang) — "Howard's research is always worth reading" (Apr 22, 44L). The one China researcher he amplifies unreservedly.
• @mengde — a private aside: "might call you to talk more about this" (Apr 9). The phone-call register is its own signal.
• Jordan Schneider / ChinaTalk — the venue he repeatedly returns to for long-form podcasting; his main public Q&A platform.
• @SamoBurja — a recurring quote partner. The Pacific War / rescue tweet is built off Samo's frame.

His sparring partners — the ones he disagrees with by name and engages seriously:

• pSmith / Mr. pSmith — extended disagreement about the Iraq War's place in the recent right's memory (Apr 19, "Mr. pSmith's review of Leap of Faith had this line about how no major American politician prior to Trump was against Iraq…how can anyone believe that?"). He is not flattering pSmith here, but he is treating him as someone whose error matters.
• @panickssery — they exchange real arguments on the Harvard / canon threads (Mar 30, Mar 31). Greer responds substantively, not dismissively.
• @ConservaWonk — the long Baltics-and-Finland argument (Apr 26–27). Greer keeps replying because he thinks the argument is winnable.
• @RichardHanania — quoted at one remove, but the tone toward Hanania's circle is collegial-curious rather than antagonistic.

Whom he ignores. The crank replies. The "Russia-Ukraine is about denazification" responses in his demographics thread (Mar 28, 8L) get zero engagement from him. He does not flame; he scrolls past.

What he amplifies. Real-document research (Howard Wang, Emily Feng), book recommendations, individual essays in Palladium / American Affairs / American Conservative, niche-but-good educational content (Seth Braver, Meleteon for live humanities seminars). Almost no AI tooling, almost no engagement-bait repost.

What he ignores. Daily news churn, viral memes, cryptocurrency, sports. The Persian Gulf war is all over his feed because it is consequential, but he is not retweeting hot takes; he is responding with paragraphs.

§VIII. The Essay He Keeps Rewriting

It is the question of what produces a serious governing class, and what to do about its absence in present-day America.

The same essay shows up under several titles. "35 Theses on the WASPs" is the historical version: it tells the story of how the Eastern Establishment was built, and what it actually believed, and why it stopped reproducing itself. "Book Notes: The Technological Republic" is the negative version: Karp wants the same thing the WASPs had but cannot specify it, cannot build it, and is in any case unwilling to commit to controversial positions himself. "The Making of a Techno-Nationalist Elite" (his April 2026 American Affairs piece) is the constructive version. "Bullets and Ballots" on Charlie Kirk is the same essay applied to populist conservatism: Kirk built institutions; that's why he mattered; the alternative is bullets. "Did Taiwan 'Lose Trump?'" is the same essay applied to a foreign-policy question: Taiwan's fate will be decided by an emergent MAGA-aligned governing class whose members have not yet formed views; the project is to cultivate them. The China essays are, at one remove, the same essay too: a serious governing class is what built modern China's scientific dominance, and it is exactly what the U.S. lacks.

The reading curriculum behind this obsession, drawn from corpus citations and essay endnotes:

• Digby Baltzell, The Protestant Establishment (1964) — the historical anchor for "35 Theses."
• Robert Caro, The Power Broker, and Caro generally — the standard for what serious biography should look like.
• Richard Norton Smith, On His Own Terms (Rockefeller biography) — same family of source.
• Helen Andrews (cited via Twitter thread) — contemporary intellectual companion on WASP-era nostalgia.
• Godfrey Hodgson, "The Establishment" (Foreign Policy, 1973) — origin point of the term he is contesting.
• Ken Jowitt, New World Disorder: The Leninist Extinction — for the telos concept and the framework underneath all his China-system writing.
• Joseph Fewsmith, Rethinking Chinese Politics — main contemporary China-political-science reference.
• Ian Toll's Pacific War trilogy — referenced repeatedly as the standard for narrative military history; underwrites the elite-formation framing of the rescue-doctrine tweet.
• Mailer, Manchester, Capote, Baldwin, Abbey, Matthiessen, Wolfe, McPhee, Talese, Didion, Halberstam, Caro, Sheehan, Hunter S. Thompson, Richard Rhodes — his canonical generation of American narrative non-fiction. He thinks of this as a peak, not a curiosity.
• Sources of the East Asian Tradition (Columbia anthology), with Xunzi as his particular touchstone.
• Paradise Lost; the Iliad; the Shuihu Zhuan / Liangshan Marsh tradition — the literary copybook he draws from when he wants to dignify a historical observation. The Grapes of Wrath tweet (Mar 29, 102L) names all three together, which is the closest you'll get to a self-portrait of his reading mind: "There was a Trojan War. There were bandits in Liangshan Marsh. But you don't read either of those tales for their historical details. That is not the point. So here, I think, so here."

The unifying instinct: he reads for what produced — and what could again produce — a country that takes itself seriously enough to govern itself well. The China project, the WASP project, the techno-nationalist project, even the Charlie Kirk project — these are the same project, looked at from different windows.