From day one, however, the duchess could not be pleased… She was consumed with petty matters; although the government was paying for Blenheim, she counted every penny.2
⚠ That clause is the case. Although the government was paying for Blenheim, she counted every penny.
By 1711 her grumbling has broken her friendship with the queen and she is dismissed from court. ⚠ On leaving she "emptied the apartment of every fixture down to the brass doorknobs."
Then twenty years of litigation. ⚠ The duke dies without ever spending a night in the palace. Marlborough leaves an estate of over £2 million — more than enough to finish it — and the duchess still withholds Vanbrugh's wages and the workmen's, and has the architect dismissed.
Vanbrugh died in 1726, locked out of the palace by the duchess, unable to set foot in his greatest creation.3
⚠ This is what makes the case diagnostic rather than merely unpleasant.
She was not protecting a hoard. ⚠ The Treasury was paying, and after Marlborough's death there was £2 million available. The cheeseparing had no financial function at either stage.
Which forces the reading Greene gives:
For the Duchess of Marlborough, money was a way to play sadistic power games. She saw the loss of money as a symbolic loss of power.4
⚠ And the corpus should note what this does to the Sadist type two pages earlier. Greene's definition there has the sadist believing "paying for something gives them the right to torture and abuse the seller" — ⚠ but she was not paying. The entitlement came from controlling the disbursement, not from being the source of the funds.
That is a broader and more common condition: ⚠ anyone with sign-off authority over money they did not earn has the same licence available to them.
With Vanbrugh her contortions went deeper still: He was a great artist, and she envied his power to create, to attain a fame outside her reach. She may not have had his gifts, but she did have the money to torture and abuse him over the pettiest details—to ruin his life.5
⚠ Greene offers this second and it explains more than the first.
Sadism-with-money predicts abuse of anyone she paid. ⚠ Envy-of-the-maker predicts abuse of Vanbrugh specifically — and the case's pattern is specific: the workmen sued for wages, but it is the architect who is dismissed, locked out, and dead at the end of it.
⚠ And the mechanism is a substitution. She could not compete on the dimension that mattered to her — the corpus recorded the same structure at Law 37's Diane de Poitiers, who changed the dimension of comparison rather than competing on age. ⚠ The duchess did the opposite: she stayed on a dimension she could not win and used the one lever she had to degrade the other party's performance on it.
Vanbrugh's twenty-year nightmare is what that looks like.
⚠ Finally, too, posterity had the last word: Vanbrugh is recognized as a genius while the duchess is forever remembered for her consummate cheapness.6
Greene includes the long-run scoreboard and the corpus should mark it as unusual. ⚠ Most of this book's accounting stops at the actor's lifetime.
And it identifies the specific asymmetry: ⚠ she spent twenty years degrading his working conditions and he produced the building anyway — *"the man who took his place finished Blenheim in a few years, following Vanbrugh's designs to the letter."*
⚠ The design survived the persecution intact. What she destroyed was his life, not his work — and the work is the thing that carries the verdict.
⚠ 🚩 SECONDARY WITHOUT PRIMARY · 🚩 SINGLE SOURCE — no historian, no biographer, no source, for one of the best-documented feuds in English architectural history. ⚠ Sarah Churchill's own memoirs and Vanbrugh's letters both survive, and neither is cited.
🚩 ⚠ The account is one-sided by construction. The duchess's grounds — that Vanbrugh was over-spending public money on a project whose funding was genuinely and repeatedly interrupted — are reported only as her "grumbling" and "petty matters." Greene's own text concedes that "the funds became harder to procure from the government", ⚠ which makes at least some of the counting a response to a real problem. [CONTESTED] on the motive attribution.
🚩 ⚠ The psychological reading — she envied his power to create — is an attribution of private motive with no source, and it is the page's most explanatory claim. [PLAUSIBLE — needs corroboration].
🚩 The brass doorknobs detail is exactly the kind that circulates because it is good and is given with no provenance.
⚠ Greene's lesson is unusually specific and worth separating from the story:
The powerful must have grandeur of spirit—they can never reveal any pettiness. And money is the most visible arena in which to display either grandeur or pettiness. … And if you want to meddle in the work of creative people under your hire, at least pay them well. Your money will buy their submission better than your displays of power.7
⚠ That last sentence is the chapter's most cynical and most operational.
It concedes that meddling will happen and relocates the question from whether to how you purchase tolerance for it. ⚠ Payment buys submission; abuse buys resistance and litigation — and the duchess got twenty years of the second because she declined to spend on the first.
⚠ And it is arithmetically true in the case. The wages she withheld were a fraction of the £2 million estate, and the litigation, the delay and the doubled construction time cost more than paying would have. Her pettiness was not even cheap.
You control the money on a project and you are unhappy with how it is being spent.
Check whether you are protecting funds or exercising a licence. ⚠ The duchess was disbursing the Treasury's money and then her husband's estate. Sign-off authority over money you did not earn confers the same feeling as ownership and none of the discipline — and the tell is whether your scrutiny scales with the actual exposure.
Notice if your attention is on one person. ⚠ Generalised cost control is a policy; a twenty-year campaign against the architect is something else. If the intensity is concentrated on the person whose contribution you could not have made, the money is not the subject.
Price the friction, not the line item. ⚠ Blenheim took twenty years instead of ten. Delay, litigation and replacement are real costs and they are invisible in the ledger she was keeping — the same denomination error the corpus recorded on the Bargain Demon.
And take Greene's cynical clause seriously if you are going to meddle. ⚠ At least pay them well. Interference is tolerable to people who are being paid properly and intolerable to people who are not — and you are buying tolerance either way, in money or in years.
Strongest evidence. A twenty-year outcome with dated turning points (1704 grant, 1705 start, 1711 dismissal from court, 1726 Vanbrugh's death), a decisive structural detail (the government was paying), and an arithmetic result — the withheld wages were trivial against a £2 million estate and the delay cost more.
Tension — the duchess had a real grievance the chapter concedes in passing: the government funding was genuinely interrupted.
Tension — Greene's two motives predict different behaviour, and the case's specificity supports the second (envy of the maker) over the first (sadism with money).
Tension — the Sadist type as defined requires the abuser to be paying, and she was not.
🚩 [POPULAR SOURCE] · 🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY — no historian, for a feud documented in both parties' surviving papers. 🚩 [CONTESTED] on motive; [PLAUSIBLE — needs corroboration] on the envy reading.
Open questions. Would paying well actually have bought submission? ⚠ Greene asserts it and the case cannot test it, because she never tried. And Vanbrugh's response to twenty years of abuse was to keep designing — the palace was completed to his drawings — which is at least weak evidence that his compliance was not the variable she was failing to purchase. She may have been buying nothing at all.
⚠ This is Transgression II and, with Pizarro, one of only two cases in the chapter arguing the title. Seven of nine argue strategic generosity.
⚠ And it is the worked case for the Sadist two pages earlier — which it does not fit. The definition requires payment to confer the licence and she was disbursing other people's funds. The corpus should record the correction: the licence comes from control of disbursement, which is far more widely distributed than ownership.
Against Lorenzo and Aretino later in the same chapter, the contrast is the chapter's own design: ⚠ three relationships between money and an artist — Lorenzo buys reputation with patronage, Aretino makes artists his partners in gift-giving, and the duchess buys twenty years of litigation. Same period, same instrument, three outcomes.
⚠ And against Law 30's effortless mastery: Vanbrugh's labour was made maximally visible and maximally contested, and the corpus's finding there was that visible labour reads as competence to a qualified audience. ⚠ Posterity was the qualified audience, and it read the building.
Creative collaboration — the patron who cannot make the thing. Collaboration as Marriage and the Third Rail treats creative partnerships as governed by what each party can and cannot contribute, and identifies the failure mode as one party asserting on a dimension where they have no standing.
She envied his power to create, to attain a fame outside her reach.
The insight neither produces alone: the chapter frames the duchess as a sadist who happened to have an architect. ⚠ The collaboration frame identifies the specific structure: a patron holds the money and the maker holds the capability, and the arrangement is stable only while each stays on their own dimension. The duchess crossed — she began adjudicating design — and, having no standing there, could only assert through the lever she did hold. ⚠ Which gives a diagnostic that generalises well past architecture: when a funder's interventions shift from scope and budget to choices, the relationship has already failed, because the only instrument available for enforcing an aesthetic judgement you cannot argue for is payment, and using it that way converts a collaboration into a hostage situation.
Operations — the cost that is not in the ledger. Small Team Velocity Advantage locates outcomes in structural properties — friction, coordination load, rework — rather than in the line items that get counted.
Blenheim took twenty years instead of ten.
What the pairing produces: the chapter's verdict is moral — pettiness costs you in reputation. ⚠ The operational frame makes it arithmetic and therefore harder to dismiss: the delay, the litigation, the workmen's suits, the architect's replacement and the duplicated management were all real expenditures, and they exceeded the sums being defended by a large multiple. ⚠ She was not trading reputation for money — she was losing on both. Which is the same denomination error as the Bargain Demon, at institutional scale: the ledger she kept had one column, and the costs she generated were entered in a column she was not maintaining. Nobody was keeping the twenty-year account, which is why it ran.
Sharpest implication. ⚠ Although the government was paying for Blenheim, she counted every penny — so the cheeseparing had no financial function, and the licence came from controlling the disbursement rather than owning the funds. That condition is far more widely distributed than ownership. And the second motive explains more than the first: ⚠ envy of the maker predicts abuse of Vanbrugh specifically, which is what the case shows — she stayed on a dimension she could not win and used her one lever to degrade the other party's performance on it. The design survived intact and was built to his drawings; what she destroyed was his life.
Generative questions.