"Self-interest and sympathy were always the same argument."
Ask most people what Adam Smith believed and they will hand you the same two images: a butcher who sells you meat because it profits him, and an “invisible hand” that turns that self-interest into public good. It is a tidy story, and it is not the book Smith spent his life on. The work he kept revising until the year he died was not about markets at all. It was The Theory of Moral Sentiments, published in 1759, and its argument runs close to the opposite of the caricature: human beings are built to care intensely about how others see them, and that caring, not calculation, is the hidden machinery underneath everything we call right and wrong.
Smith came to the question as a professor of moral philosophy at the University of Glasgow, in a field that tried to account for human beings whole — economics did not yet exist as a discipline of its own. What he proposed was a capacity he called sympathy, and he meant something more exacting than feeling sorry for someone: the imaginative act of placing yourself inside another person’s situation to construct what they must be feeling. Do that enough times, across a lifetime of exchanges with parents, friends, and strangers, and something durable forms — an internalized judge Smith called the impartial spectator, or “the man within the breast,” whose imagined approval or disapproval is what moral judgment actually feels like from the inside.
Seventeen years passed before Smith published An Inquiry into the Nature and Causes of the Wealth of Nations, the 1776 book that gave the world its line about strangers getting their dinner not from the butcher’s benevolence but from his self-interest. Read on its own, that sentence sounds like a rejection of everything the earlier book argued. But Smith’s claim there is narrower than it looks: you do not need to love your butcher for a single transaction between strangers to go through. The same book warns that tradesmen’s private meetings tend to end “in a conspiracy against the public,” and The Theory of Moral Sentiments calls justice “the main pillar that upholds the whole edifice of society.” Even the phrase that supposedly carries his whole theory of markets barely appears — the “invisible hand” shows up only twice in Smith’s mature published work, once for a self-absorbed landlord whose estate spending happens to distribute resources evenly, once for a merchant who invests at home for selfish reasons and incidentally strengthens the domestic economy. It was never Smith’s general theory of anything. It was a narrow illustration, used twice.
That gap between the two books is not a modern misreading — it has a name. Nineteenth-century German scholars called it the “Adam Smith Problem,” the claim that his two major works flatly contradict each other, sympathy governing one and self-interest the other, as though Smith had swapped philosophies mid-career. Later scholarship largely dismantled the idea: the books were never describing rival forces. Self-interest gets dinner on the table, but in Smith’s own account it operates inside a framework of internalized moral rules, built out of sympathy and backed by society’s demand for justice — never given license to justify fraud or theft. The version of Smith enlisted to argue that markets do the moral work for you on their own is not supported by how rarely, or how narrowly, he actually used the phrase supposed to prove it.
None of this is in dispute among historians of economic thought as a matter of record — the publication dates, the decades of revision, the mere two appearances of the “invisible hand” are simply what the texts say. What remains genuinely open is a finer question: how seamlessly the two books cohere as a single system, and how much continuity of vocabulary and method actually runs from the psychological language of the earlier work into the more technical language of the economic one. That is a live scholarly conversation, not a closed case, and it is worth remembering that the tidy resolution — Smith was consistent all along — can flatten a real and still-unsettled question almost as much as the caricature it replaces.