In the summer of 1944, an eighteen-year-old from South Hadley, Massachusetts, enlisted in the United States Army and shipped out to the South Pacific. South Hadley was a paper-mill town on the Connecticut River, Catholic and working class in large parts, the kind of place where a boy grew up knowing which families were Irish, which were French Canadian, and which were Polish, and knowing that the distinctions carried weight. Walker Francis Connor spent the last stretch of the Second World War in a theater where Americans fought under one flag while the war itself proceeded as a contest of peoples who understood themselves as peoples. He came home, went to the University of Massachusetts at Amherst on the strength of his service, then to Georgetown, and finished a doctorate in political science in 1961. He spent the next five decades asking one question that his discipline believed it had already answered. What is a nation?
The discipline he entered had an answer, and the answer was a habit. Political scientists in the 1950s and 1960s used “nation” as a synonym for “state.” They spoke of international relations when they meant relations between governments. They called the United Nations by its name without noticing that almost none of its members were nations. They labeled the new countries of Africa and Asia “nation-states” the day the colonial flag came down. And they had a theory, confident and well funded, that explained where history was going. Modernization theory held that roads, schools, factories, newspapers, and radios might dissolve the old attachments of tribe and dialect and pull villagers into the common culture of the modern state. Karl Deutsch (1912-1992) had given the theory its scientific apparatus, measuring social mobilization and communication flows. The developmental economists supplied the money and the optimism. Ethnic identity was a residue. It might be melting.
Connor counted. For his 1972 article “Nation-Building or Nation-Destroying?” in World Politics, he surveyed the 132 states in existence as of January 1, 1971, and found that only a small fraction of them, about a dozen, could be called ethnically homogeneous. Nearly a third contained a largest group that did not even form a majority. The “nation-state,” the basic unit of the field’s vocabulary, described almost nothing on the map. The article became one of the most cited in the study of nationalism, and its title posed the charge as a question. What governments called nation-building was state-building, and state-building, when it required a Breton child to stop speaking Breton or a Kurdish child to learn that he had always been a Turk, could be nation-destroying. The central authority saw integration. The minority saw erasure. Both were looking at the same school.
The deeper argument turned modernization theory against itself. The theorists assumed communication produced assimilation. Connor argued it could produce the opposite. A road connects the village to the capital, and now the villager can see how the capital lives. A school teaches literacy, and now there is a readership for the nationalist newspaper. A radio broadcasts in the state language, and the listener notices it is not his language. Urbanization throws a dispersed people together in the same tenements, where they discover each other. The tools of integration double as tools of mobilization. The question was never how much communication occurred. It was who spoke to whom, in whose language, under whose institutions, and who noticed what.
Connor had made a version of this case earlier. His 1967 World Politics essay “Self-Determination: The New Phase” appeared at the high tide of decolonization, when the transfer of sovereignty from European empires to new governments was celebrated as the fulfillment of national self-determination. Connor asked whose self had been determined. A colonial territory could become sovereign without becoming a nation. The borders were European borders. The peoples inside them had not been consulted about each other. Independence answered the imperial question and left the national question standing. The subsequent history of Nigeria, Sudan, Ethiopia, Burma, and dozens of other states did not embarrass the essay.
His most durable contribution may have been the least glamorous kind of intellectual work, the discipline of terms. His 1978 essay bore an exasperated title: “A Nation Is a Nation, Is a State, Is an Ethnic Group, Is a…” The literature used its crucial words interchangeably, and Connor believed the confusion was not cosmetic. A state is a legal and territorial organization, with borders, institutions, and a claim to authority. A nation is a group of people who believe themselves ancestrally related. Nationalism is loyalty to the nation. Patriotism is loyalty to the state. The two loyalties can coincide. They do not have to. A Basque could be a passionate nationalist and a reluctant citizen of Spain. A government that mistook the second loyalty for the first, or mistook silence for consent, was reading its own press releases. Connor warned that a quiet minority might be quiet because it lacked organization, opportunity, or safety, and that forced assimilation, once national consciousness had spread, tended to confirm the nationalist’s claim that his people were in danger. The policy meant to end the identity advertised it.
He gave the field a word, ethnonationalism, and admitted the word was redundant. In his usage, nationalism already meant loyalty to an ancestrally imagined people. But “nationalism” had been stretched to cover state patriotism, protectionism, flag-waving of every kind, so he added the prefix to return the term to its object. The nation, in his definition, was a self-differentiating ethnic group. An ethnic group might persist for centuries as a fact visible to outsiders, a matter of language, religion, custom. It became a nation when a substantial share of its members grew conscious of themselves as one people and believed, whether or not the genealogists could confirm it, that they shared descent. The belief was the thing. Members of a nation are usually mongrels by any objective measure. The myth of common ancestry, and Connor used “myth” to mean a shared account of origins rather than a simple lie, did the political work.
Western Europe supplied his most awkward evidence, awkward for his opponents. Ethnic conflict was supposed to be a disease of the developing world. Britain, France, Spain, and Belgium were the finished products, the models the new states were supposed to become. Connor looked at the finished products and saw Scottish and Welsh nationalism, Breton and Corsican movements, Basques and Catalans, Flemings and Walloons. In 1975, scholars gathered at a conference on ethnic conflict in the Western world, and the recurring question from the floor was why these movements had appeared now. Connor’s answer was that most of them had been visible for years. The surprise in the room interested him as much as the movements did. The scholars had a theory that told them ethnonationalism could not survive in advanced industrial societies, so they had not seen it. Britain remained a state containing four national identities. Spain’s stability under Franco proved coercion, not consensus. France, the most successful centralizer in Europe, had built its unity in part by grinding down Occitan, Breton, and Alsatian over generations, which demonstrated Connor’s point rather than refuting it. Durability is not homogeneity. A multinational state can last for centuries on legitimacy, bargaining, habit, or force. Lasting is not the same as being one people.
The obvious question was why the belief in shared ancestry carried such power, and Connor’s answer, developed most fully in “Beyond Reason: The Nature of the Ethnonational Bond” (1993), embarrassed a discipline built on interests and institutions. He read the speeches of nationalist leaders across centuries and continents and noticed that the successful ones spoke a single dialect. Blood. Ancestors. Mothers and fathers. Brothers in arms. The motherland, the fatherland, the sacred soil where the forefathers lie. Scholars filed this language under rhetoric. Connor filed it under evidence. The leaders knew their audiences better than the analysts did. The family metaphor converts millions of strangers into figurative kin, and kinship is the one moral bond people honor without running a cost-benefit analysis first. A man does not calculate whether his brother is a good investment. Nationalism extends that intuition to the imagined national family, which is why men accept for the nation sacrifices no rational-choice model can price. Connor called the bond nonrational rather than irrational, and the distinction carried his argument. Irrational means foolish. Nonrational means operating outside calculation, the way love of family does. A theory of politics that admitted only interests might underestimate nationalism every time, and, he noted, it had.
This led him to attack what he called the economic fallacy. Nationalist movements were routinely explained by poverty, regional inequality, or competition for jobs. Connor pointed out that the wealthy regions rebelled too. Catalonia and the Basque Country were among the most industrialized parts of Spain. Flanders outproduced Wallonia. Quebec was not Bangladesh. Rich minorities resented subsidizing the poor center; poor minorities resented exploitation by the rich center; opposite conditions produced the same demand. Economics could tell you when grievance sharpened. It could not tell a man why he was a Basque. Class explained conflicts within the question. It never explained the question.
His first book took the argument into the empire that had staked everything on the opposite bet. The National Question in Marxist-Leninist Theory and Strategy, published by Princeton in 1984, ran past six hundred pages and examined how Marx (1818-1883), Engels, Lenin (1870-1924), Stalin (1878-1953), and their successors had understood the nation. Marxism held that class was real and nation was superstructure, a byproduct of capitalism that socialist development might dissolve. Lenin grasped the tactical value of national grievance and promised self-determination, and the state he founded then built a deep contradiction into its bones. The Soviet Union drew internal borders around nationalities, gave them republics, schools, official languages, historians, and administrative elites, and waited for the identities to wither as theory required. The identities did not wither. The republics gave them addresses. Connor’s argument, published five years before the Berlin Wall opened, was that communist rulers had confused the suppression of nationalist activity with the disappearance of national consciousness. In 1991 the Soviet Union came apart along the lines of its own republics, the borders it had drawn around identities it expected to outlive. Yugoslavia came apart with blood. The book acquired the reputation of prophecy, though Connor had predicted no date. He had described a load-bearing error and waited.
Ethnonationalism: The Quest for Understanding, also from Princeton, followed in 1994, collecting the essays of three decades into the clearest statement of his position. The book’s architecture was a course of correction: first fix the vocabulary, then dispel the illusion of homogeneity, then retire the economic explanation, then face the emotional core. He positioned himself between the field’s two camps and refused both. Against the primordialists, he accepted that mass national consciousness was largely modern, made possible by print, schooling, and mobilization. Against the constructivists, he denied that elites could invent a nation from nothing. Leaders can select symbols, curate history, and name enemies, but their appeals work only when they touch beliefs and memories people already find binding. Ernest Gellner (1925-1995) explained nationalism as a functional requirement of industrial society; Connor thought function could not explain fervor. Benedict Anderson (1936-2015) showed how print let strangers imagine a community; Connor pressed on what the strangers imagined, which was that they were relatives. Anthony D. Smith (1939-2016), closest to him in spirit, traced the ethnic pasts that supplied nations their symbols; Connor kept his eye on the present tense of the question, whether living people felt themselves one ancestral family now. In “When Is a Nation?” (1990) he added the corollary that a nation’s birthday is unknowable, since nationhood requires mass awareness and the masses leave thin records. And he noticed that the question is one the nation’s own members cannot hear. A nation experiences itself in mythic time, ancient regardless of the documents. The historian proves the flag is recent; the believer knows the people are old. Both are right about different things, which is why the proof changes nothing.
The limits of the work were the price of its edges. His definition of the nation was ethnic all the way down, which left the United States, Switzerland, and India as puzzles, political communities with emotional depth but no plausible myth of single descent. His likely reply, that loyalty to a constitution should be called patriotism rather than nationalism, defended the vocabulary while conceding the phenomenon. Critics also observed that he explained the intensity of the national bond better than its distribution, since not every Basque is a Basque nationalist and every person carries several identities whose rank shifts with circumstance. And his attention to the felt reality of kinship left underexamined the institutions, censuses, and schoolbooks through which the feeling gets manufactured and maintained. Yet the critics worked inside his distinctions. Anyone who wanted to argue about nations and states first had to separate them, and Connor had done the separating.
The career that produced this work never settled into the usual channels. Connor taught at Nasson College in Maine, then Rensselaer Polytechnic Institute, SUNY Brockport, and Trinity College in Hartford, and held visiting appointments at Oxford, Cambridge, the London School of Economics, Harvard, Dartmouth, Pomona, Central European University, and the Institute of Political Science in Warsaw, among others, along with a fellowship at the Woodrow Wilson International Center for Scholars and a term as scholar-in-residence at the National Foreign Assessment Center, the analytic arm of the CIA, which had its own reasons to want a better theory of why multiethnic states crack. He spent his final institutional years at Middlebury College in Vermont, as Visiting Distinguished Professor and Scholar-in-Residence from 1998 to 2003 and then as a visiting scholar through the 2010-2011 academic year, teaching seminars and winter-term courses and running a faculty colloquium. He never held a chaired professorship at a research university. The field he helped found took institutional shape in the 1980s, with its journals and associations, a generation after his first articles, and in 2003 Daniele Conversi edited Ethnonationalism in the Contemporary World: Walker Connor and the Study of Nationalism, a volume of scholars across half a dozen disciplines engaging his work. The University of Nevada named him its Distinguished American Humanist for 1991-1992. The University of Vermont named him its Distinguished American Political Scientist in 1997.
In 1951 he had married Mary Lyon. She worked with him on the research and the writing for more than sixty years, an uncredited collaboration of the kind that sustained many careers of that generation, and when her health failed he became her caregiver. They raised three children, Peter, Joan, and Daniel. Mary died in 2014. Connor lived in Belmont, a village in the Green Mountains, in a house with a view he had earned. He loved jazz, antiques, art, travel, and college basketball. Colleagues remembered a correspondent of wit and generosity who read the manuscripts of strangers and wrote recommendation letters for the young, a warmth at odds with the austerity of the prose. He supported Doctors Without Borders. He died at home on February 28, 2017, at ninety.
The man who watched the Soviet Union confirm his book left one insight that governments keep relearning at expense. A state can build roads, print passports, staff schools, and declare its people one nation. None of it proves the people agree. Modernization can sharpen difference. Education can awaken memory. Repression can turn a dialect into a cause. And underneath the institutions runs the conviction that made Connor’s subject the most powerful form of human association in the modern world, the belief of millions of strangers that they are family, a belief that answers to no calculation, which is why the calculators keep getting it wrong.
In 2019, I did a series of videos on Connor’s book, Ethnonationalism: The Quest for Understanding.
Notes
Core facts, including Connor’s birth in South Hadley, his 1944 enlistment, service in the South Pacific, education at the University of Massachusetts, Georgetown doctorate in 1961, teaching appointments, Middlebury positions from 1998 to 2003 and through the 2010–11 academic year, and death on February 28, 2017, at age ninety:
Middlebury memorial: https://www.middlebury.edu/announcements/memoriam/2017/03/memoriam-walker-f-connor
The New York Times obituary, hosted by Legacy: https://www.legacy.com/us/obituaries/nytimes/name/walker-connor-obituary?id=20287641. It confirms his appointments at Nasson College, Rensselaer Polytechnic Institute, SUNY Brockport, and Trinity College, as well as his numerous visiting positions, work at the National Foreign Assessment Center and Woodrow Wilson International Center for Scholars, editorial-board memberships, awards, and residence in Belmont, Vermont.
The count of 132 states, using January 1, 1971, as the baseline, appears in footnote 2 of “Nation-Building or Nation-Destroying?” in World Politics 24, no. 3 (1972): https://www.cambridge.org/core/journals/world-politics/article/abs/nationbuilding-or-nationdestroying/777F7F9A644F44F1D7A87AEC81FB715E.
The commonly cited breakdown is that only 12 of the 132 states, or 9.1 percent, were essentially homogeneous, while in just under one-third the largest ethnic group constituted less than a majority. I used “about a dozen” and “nearly a third” to remain cautious.
Sources for scholarly assessments of Connor’s status as a founder of the field, the growing influence of his two major books, and his intellectual position in relation to Ernest Gellner, Benedict Anderson, Anthony D. Smith, and Karl Deutsch include:
The obituary in Nationalism and Ethnic Politics: https://www.tandfonline.com/doi/full/10.1080/13537113.2017.1311139
Daniele Conversi, “Introduction: Why a State Is Not a Nation, and Whether Economics Really Matters: Walker Connor 50 Years On,” Nations and Nationalism (2018): https://onlinelibrary.wiley.com/doi/abs/10.1111/nana.12441
“A Modernist Beyond Modernization Theory: Walker Connor and His Time,” Nationalities Papers: https://www.cambridge.org/core/journals/nationalities-papers/article/abs/modernist-beyond-modernization-theory-walker-connor-and-his-time/6F539BD7EB1450257AA76E49AD6D3D99
Conversi’s encyclopedia entry on Connor: https://onlinelibrary.wiley.com/doi/abs/10.1002/9781118663202.wberen289
Wikipedia, for his Fulbright chair at Queen’s University in Kingston and several other appointments.
The conference was the event that produced Milton J. Esman’s edited collection Ethnic Conflict in the Western World (Cornell University Press, 1977). Connor’s chapter, “Ethnonationalism in the First World,” reportedly recounts the episode.
Several passages are extrapolations rather than documented biographical facts. The South Hadley opening accurately describes a paper-mill town on the Connecticut River, but its references to Irish, French Canadian, and Polish families are reasonable inferences about western Massachusetts mill towns rather than facts established about Connor’s childhood. The description of the town as substantially Catholic and working class is likewise inferred from its social history, not from a source concerning Connor’s family.
Connor’s use of the GI Bill is an inference based on his 1944 enlistment and postwar attendance at the University of Massachusetts.
The phrases “a view he had earned” and the closing rhetorical passages are interpretive language rather than sourced claims.
The statement that Connor “never held a chaired professorship at a research university” is an inference from his recorded career path. His obituaries do not list such a chair.
On Christmas night, 1991, the red flag came down the Kremlin flagpole and the state that had promised to outlive the nations of the earth stopped existing. In Belmont, Vermont, a village in the Green Mountains with one general store and a pond, a sixty-five-year-old political scientist without a chaired professorship had spent seven years explaining, in a six-hundred-page book most of his discipline had not read, why this might happen and why the rulers of that state could not see it coming. The empire dissolved along the borders of its own internal republics, the lines it had drawn around identities its theory said were dying. Walker Connor had been right, in print, since 1984, and in the articles since 1967. There was no prize for this. The field that could have given him one was only then coming into existence, in part because he had been right. So the question of that winter is the question Ernest Becker (1924-1974) spent his last decade asking. What does being right buy a man? What is the currency, and what is it convertible into, and how long does it hold its value after he is dead?
Becker’s answer, in The Denial of Death and Escape from Evil, runs as follows. Man is the only animal that knows it will die, and the knowledge is unbearable, and so every human society is at bottom a machine for making the knowledge bearable. The machine is the hero system: a shared structure of meaning inside which a person can perform acts that count, earn a standing that outlasts him, and feel himself a being of value in a world of value rather than a heap of matter with an expiration date. The hero system can be religious, national, scientific, commercial, familial, criminal. It works only while it is not seen as a hero system. Becker called the necessary blindness the vital lie. A man inside a working hero system does not experience himself as denying death. He experiences himself as doing his job, raising his sons, serving his people, getting the words right.
Connor is a rare kind of subject for this frame, because his life’s work was the anatomy of the largest hero system of the modern age, and he performed the anatomy from inside a smaller one he never named. The essay can therefore run on two levels at once, and it should, because the levels explain each other.
Take the findings first and restate them in Becker’s terms. Connor found that the nation is a group of strangers who believe they are ancestrally related, that the belief resists all evidence about actual descent, that nationalist leaders across every century and continent speak the language of blood, ancestors, and family, that members experience the nation in mythic time as ancient regardless of what historians prove about the recency of the flag, and that people accept death for the nation in ways no calculation of interest can price. He called the bond nonrational. Becker supplies the reason it has to be. The nation is an immortality project. It converts the terrifying arithmetic of one man’s seventy years into membership in a body that was here before him and continues after him. The ancestors are not decoration. They are the point. A nation is a contract among the dead, the living, and the unborn, and the contract’s hidden clause is that the living get to stop being afraid. This is why the family metaphor recurs from Bismarck to Ho Chi Minh, and why Connor found it everywhere he looked. The family is the first immortality vehicle any human knows, the primal evidence that something of you continues in bodies that are not yours. Nationalism scales that evidence up to millions. And it is why the bond cannot be audited. A hero system submitted to cost-benefit analysis is already dead. Connor spent fifty years telling his discipline that the analysts kept underestimating nationalism because they modeled men as interest-calculators. Becker would have said the analysts were doing something more desperate than miscalculating. They were protecting their own hero system, in which man is rational, history has a direction, and the men with the regression models stand at the front of it.
Now take Connor’s sacred words one at a time, because each of them is sacred elsewhere too, and means something else there, and the differences are the demonstration.
Begin with blood, the family, the ancestors. In the Granite Mountain vault outside Salt Lake City, Mormon genealogists keep microfilmed records of the human family running back centuries, because in their hero system the dead can be baptized and sealed to their descendants, and family is the literal architecture of eternity, an organization chart of heaven. For the foster child who aged out of the system at eighteen with his belongings in a trash bag, family is the machine that failed, and the sacred thing is the found kin, the friends who show up, and he might tell you blood is the least of it, and his hero system makes loyalty holy in proportion to its distance from biology. For the Bedouin who can recite his patriline for twelve generations, the recitation is the deed to everything: honor, water rights, the right to speak. For the startup founder who tells forty employees at the all-hands that we are a family here, family is a solvent for overtime, and everyone in the room knows it, and the word does its work anyway. For the Serbian conscript on a hillside above Vukovar in the autumn of 1991, the ancestors are a summons, six hundred years of them, and the summons overrides everything his own short life might prefer. Same word. Five heavens. Connor’s contribution was to see that the fifth case, the deadly one, runs on the same fuel as the first four, the human refusal to be a discrete organism with a stop date, and that the nationalist version is the most powerful because it requires no vault, no recitation, no equity package. It requires only the belief, available to every peasant, that the strangers are kin.
Take the second sacred word, precision, the exact term, because this was Connor’s own. His most famous title is a complaint about vocabulary: “A Nation Is a Nation, Is a State, Is an Ethnic Group, Is a…” He built a career on the insistence that nation and state are different things, that nationalism and patriotism are different loyalties, and that a field using its crucial words interchangeably cannot think. Consider what precision means across hero systems. For the Talmudist, precision is devotion; the difference between two near-synonyms in a verse can carry a law, and God is in the distinction, and a career spent on the meaning of a single word is a career spent well. For the hospital pharmacist, precision is the wall between the ordinary day and the dead child, a decimal point as a moral event. For the jazz musician, and Connor loved jazz, precision is nothing like the metronome; it is time-feel, the note placed a shade behind the beat, exactness in the service of swing, and a player who is merely accurate is the worst thing you can be, which is stiff. For the artillery officer, precision is other people’s death arriving where intended. For the diplomat, precision is often the enemy, and the sacred skill is the phrase both sides can sign because it means two things. Connor’s precision belonged to the Talmudist’s family with one difference. His God was not in the text. His faith, and it should be called that, held that the world has joints and that words can be made to cut at them, and that a man who gets the words right has done something permanent no matter where he teaches. This is the hero system of the scholar stripped to its minimum, and it is worth noticing how well suited it was to his position. A man at Nasson College and SUNY Brockport could not compete in the currencies of the center, the grants, the graduate students, the Cambridge chair. Conceptual precision is the one scholarly currency that requires no institution. It can be minted alone, at a kitchen table, with a wife checking the references. Becker teaches that a man’s hero system tends to be the one his terrain makes possible. The periphery made Connor a purist the way the desert makes anchorites.
The third sacred word is evidence, and here a scene carries the argument. In 1975, scholars of comparative politics gathered to discuss ethnic conflict in the Western world, and the question that kept coming from the floor, in the accounts Connor gave of it, was some version of: why now? Why has Scotland erupted now, why Brittany now, why the Basques now? And Connor’s answer, delivered by a man who had been publishing on this for eight years, was that there was no now. Most of the movements were decades old. The Basques had not changed. The audience had. Inside modernization theory, which was not merely a research program, the data had been invisible. It should be said plainly what modernization theory was, in Becker’s terms, for the generation that held it. It was the hero system of postwar American social science: the faith that development runs one direction, that communication integrates, that the parochial attachments melt under literacy and asphalt, and that the American present is the global future, with the political scientist as the cartographer of salvation. A Ford Foundation grant was not just money. It was a role in the redemption of the underdeveloped world. Inside that system, a persistent Breton nationalist was not evidence. He was noise, a lag, a residue scheduled for dissolution, and the scheduling was the theory. Connor’s counting, twelve homogeneous states out of 132, was the kind of act Becker says a hero system cannot forgive, because it does not attack a finding, it attacks the immortality project of the people holding the findings. This is why being right early earned Connor so little for so long. The scholars at that conference were not fools and were not liars. They were believers, and the vital lie was doing what it is for.
The Marxist-Leninist case, the subject of his 1984 book, gives the same collision at imperial scale, and adds a refinement, because there the sacred word is history. For the Leninist, history is a court with a verdict already written: class is real, nation is costume, and the future has been assigned. For the nationalist inside the Soviet system, history is the ancestors, the graves, the mythic continuity that makes a Lithuanian a Lithuanian regardless of what the textbook says. For the professional historian, history is chronology and documents, and he can prove the national costume was stitched recently. Connor saw that the three do not share a subject. The Soviet state ran one immortality project, the classless future, while administering the raw material of another, the national pasts, and it institutionalized the second to manage it, drawing republics around peoples it expected to outlive. When the first project lost its believers, the second was standing there with borders, anthems, and educated elites, fully provisioned. Becker’s law covers this: a state can suppress a hero system’s activity and cannot suppress the need it answers, and when the official channel to immortality silts up, the traffic reroutes. Seventy years of scientific atheism and socialist schooling produced, on the far side, processions carrying icons and flags. The need had never gone anywhere. Only the vehicle had been impounded.
Connor’s case tests Becker at Becker’s weakest joint. Becker held that the hero system works only unexamined, that the vital lie dies of exposure. Connor spent half a century exposing the national lie, in the neutral sense he insisted on, the myth of common ancestry believed by populations that are demonstrably mixed. He published the anatomy in the leading journals. The field absorbed it, taught it, made it canonical. And nationalism, over those same fifty years, did not weaken by a gram. The Soviet successor states, the Yugoslav wars, the devolution of Britain, the return of national politics across the rich world after Connor’s death, all of it proceeded as if the anatomy had never been published. The resolution is that Becker’s law holds at the level of the person inside the system and fails at the level of the library. Proving to a believer that his flag is recent does nothing, because, as Connor saw, the historian’s chronology and the believer’s mythic time run on different planes, and the belief was never a factual error to be corrected. It is a need with a vehicle. You can total the vehicle. The need walks away from the crash and finds another. Connor’s work is therefore best read as a confirmation of Becker that Becker himself might not have predicted: the diagnosis of the illusion, distributed at scale, leaves the illusion intact, because the illusion is not made of information.
Connor’s immortality project had four foundations. The first was the work, the distinctions themselves, nation from state, nationalism from patriotism, built to outlast him the way a stonecutter’s joints outlast the stonecutter, and they have; nobody in the field can argue about nations without using his separations, including the people who reject his conclusions, which is the deepest form of citation. The second was the field, nationalism studies, which took institutional shape in the 1980s with journals and associations, a generation after his first articles, so that the man without the chair became a founder, and the 2003 Conversi volume performed the canonization in the standard rite of the scholarly hero system, the festschrift, which is a funeral a man gets to attend. The third was the mail. Connor built his church by correspondence, decades of letters to young scholars on other continents, comments on manuscripts, recommendation letters remembered for their humor, an apostolic succession run through the post office of a Vermont village, and in Becker’s terms this was heroism perfectly scaled to its terrain, the periphery scholar’s answer to the seminar he did not run. The fourth was Mary. They married in 1951, she worked inside the books for sixty years without a title page, and when her health went he put down the work and nursed her until she died in 2014, and Becker, who wrote that the romantic solution makes the beloved into the divine, might have noted that Connor’s version was the undeluded kind, service rather than worship, the immortality of two people woven into one production so tightly that the credits cannot be separated, which may be the point of the arrangement.
He died at home in the mountains in the winter of 2017, at ninety, three years after her. Set the two heavens side by side and the essay can close. The nationalist’s heaven is mythic time, the unbroken people, the ancestors at your back and the unborn ahead, and it asks of the believer only that he never look at the stitching. The scholar’s heaven is the footnote, the distinction still in use in arguments conducted by people not yet born, and it asks the opposite, that he look at everyone’s stitching forever, on the wager that one or two of his own seams were sewn straight. Connor knew the first heaven was made of myth, said so for fifty years in the plainest terms his discipline would print, and never once talked as if the knowledge licensed contempt for the believers, because he understood, before Becker gave him the vocabulary he never used, that the believers were doing what men do with the fear, and that he was too, at a kitchen table, with the references checked, getting the words right.
Nasson College closed in 1983. The campus in Springvale, Maine, went through bankruptcy, and the buildings were sold off piecemeal, and today the institution where Walker Connor began his teaching career exists as a set of court records and an alumni association without a college. Hold that fact against another. In the same postwar years, the Center for International Studies at MIT ran on Ford Foundation and CIA money, the Social Science Research Council’s Committee on Comparative Politics, under Gabriel Almond (1911-2002) and later Lucian Pye (1921-2008), set the research agenda for a generation, and Princeton University Press published the committee’s sequence on political development in uniform volumes, a series with the confidence of an encyclopedia. One side of the discipline had contracts, graduate students, and a client in Washington. The other side had Springvale. Pierre Bourdieu (1930-2002) built a sociology for reading exactly this kind of asymmetry, and Connor’s career may be the cleanest American case of it, a forty-year experiment in what a man can and cannot do in a scholarly field when he holds the wrong positions and the right arguments.
Bourdieu’s terms first, briefly. A field is a game with stakes its players believe in. Positions in the field are defined by holdings of capital, and capital comes in kinds: economic, social, cultural, and the symbolic capital of recognition, which in a scientific field means the authority to say what counts as knowledge. Position shapes strategy. Players rich in the field’s capital defend the orthodoxy that made their holdings valuable. Newcomers with good inheritances play succession, extending the orthodoxy in approved increments toward the chairs. Players at the margin face different odds, and for them heresy can be the rational wager, because subversion, redefining what the game is about, is the only move that might revalue their holdings. And a field has a degree of autonomy, measured by its power to refract outside demands into its own terms. A field that takes its questions from its funders has surrendered the measure.
Postwar American comparative politics was a heteronomous field wearing the costume of an autonomous one. The state needed a science of the new countries, needed to know how Nigeria and Indonesia might be kept stable and pro-Western, and the foundations translated the need into research programs, and the programs produced modernization theory, which taught that development runs one direction and that communication, literacy, and asphalt dissolve parochial identity into the culture of the state. The vocabulary tells the story of the clientele. “Nation-building” was an administrator’s word. It named a service the field offered its patron: instructions for manufacturing loyalty. The word assumed the product could be manufactured, and the assumption was the orthodoxy, and the orthodoxy was load-bearing, holding up not a theory but a revenue model. In Bourdieu’s economy of symbolic goods, the doxa of a field is whatever must remain unquestioned for the field’s capital to hold its value. Ask whether nations can be built by governments, and the Princeton series becomes a shelf of consulting reports for a procedure that does not work.
Connor asked. Look at what he held when he asked. A Georgetown doctorate, 1961, respectable and peripheral, a Jesuit school without a placement machine, no Cambridge network, no committee patron, a first job at a college that no longer exists, then Rensselaer, then SUNY Brockport, then Trinity in Hartford. In the field’s terms he was capital-poor in every kind except one. He could not run a research shop, could not place students, could not repay patronage. What he could do was mint the one scholarly currency that requires no institution behind it, the exact distinction, and he minted at a kitchen table with his wife checking references, a domestic mode of production in Bourdieu’s sense, the unwaged labor of Mary Connor subsidizing an output the field priced as one man’s. And he placed the coinage in the center’s own vault. “Self-Determination: The New Phase” ran in World Politics in 1967, and “Nation-Building or Nation-Destroying?” ran there in 1972, and World Politics was the orthodoxy’s house journal. This is a move Bourdieu’s model predicts and prizes: the heretic must be consecrated somewhere or the heresy is noise, so he spends his small capital buying space in the temple. The 1972 article counted the world, twelve homogeneous states out of 132, and told the committee that its central term named a process observed almost nowhere, that the “nation-state” was a fiction covering the map, and that the tools of integration were operating, in case after case, as tools of mobilization. The article was cited immediately and for decades. Citation is the field’s cheapest coin. The chairs stayed where they were.
A department chair is not a prize for being right. It is an investment in the field’s reproduction. The professor holding it will train the graduate students, and the students must be placed, and they can be placed only if their vocabulary is the vocabulary the hiring departments trade in. Now run the calculation from inside a hiring meeting, and the scene can be built from the logic, because the logic is what the meeting is. It is 1974, a research department with a development line to fill, and the file on the table shows two World Politics articles of unusual force. The senior man who built his career on political integration turns the pages and says the work is sharp and the work is narrow, one idea pressed hard, where is the theory of development, and a colleague says the students need training in the survey methods and this man does none of that, and someone asks what he might teach at the graduate level, and someone answers, a seminar against the field, and there is mild laughter, and the file closes. No conspiracy occurred. Every judgment was locally reasonable. That is Bourdieu’s point about how fields defend themselves: the orthodoxy does not need to censor the heretic when it can simply find him, in perfect good faith, not a fit. Hiring Connor meant endowing the negation of the enterprise the department sold. Fields do not buy their own short position.
Bourdieu found in the French academy two currencies circulating at once, academic power, the control of committees, appointments, and reproduction, and intellectual prestige, the recognition of peers and readers, and found the two held by different men and often inversely. Connor became an American illustration drawn with a ruler. The prestige accumulated, visiting appointments at Oxford, Cambridge, the LSE, Harvard, Dartmouth, Warsaw, the Wilson Center, a Fulbright chair in Canada, and the power never came, because the visiting circuit is exactly how a field honors a man it will not let govern. The visitor gets the lectern and the dinner. He gets no vote on the curriculum, no students of his own, no line to fill. The circuit converts a dangerous position into a decoration. And one appointment on his list shows the structure by exception. The National Foreign Assessment Center, the CIA’s analytic arm, brought him in as scholar-in-residence, and the exception is legible in field terms: intelligence is a market where being wrong carries a price, so it buys predictive accuracy without asking whether the seller is orthodox. An analyst who spent 1979 reading Connor on the Soviet nationalities was not defending a paradigm. She had accounts to reconcile, Soviet censuses that kept finding Uzbeks where the theory expected Soviet men, and Connor priced the anomaly correctly. The academy’s autonomy had become, on this question, the freedom to stay wrong longer than the government could afford to.
The value of any capital depends on the state of the field, and when the state changes faster than dispositions do, the players caught holding the old currency suffer hysteresis, the lag of a habitus trained for a game that has ended. Between 1989 and 1991 the empirical world executed a margin call on modernization theory. The Soviet Union, seventy years of literacy campaigns, industrialization, internal migration, and unified schooling, the most sustained nation-building program ever run, came apart along the administrative borders of its nationalities, precisely the fault lines the orthodoxy had scheduled for dissolution and precisely the ones Connor’s 1984 book, The National Question in Marxist-Leninist Theory and Strategy, six hundred pages from Princeton, had mapped. The crash revalued every portfolio in the field at once. Twenty years of Connor’s holdings, illiquid, cited but unconvertible, repriced in a single news cycle. The scholars who had asked, at the 1975 conference on ethnic conflict in the Western world, why the Basques and Scots were erupting “now,” were asking the same question in 1991 about Lithuanians and Croats, and this time the man who had answered it since 1967 was not correcting the field’s surprise. He was the obvious thing to read about it.
But a repriced heretic in an old field is still a heretic, and the more instructive move is the one that followed, because it is the move Bourdieu’s model names as the subversion strategy carried to completion. When you cannot win the game, you change the stakes, and when you cannot change the stakes, you found the game next door. Nationalism studies condensed into a field in the years around the crash, the Association for the Study of Ethnicity and Nationalism at the LSE in 1990, the journal Nations and Nationalism in 1995, syllabi, conferences, a canon. A new field must perform its autonomy by declaring founders, since founders are the proof that the field has its own past and is not a province of some older empire, and the founding generation is recruited, as a rule, from the heretics of the adjacent fields, the men whose capital the old game refused to honor. In political science Connor’s distinctions had been an attack. In nationalism studies the same distinctions were infrastructure. Nation against state, nationalism against patriotism, ethnic group against nation: every seminar in the new field opened by teaching his separations, including the seminars that went on to reject his conclusions, and there is no deeper consecration in a scholarly field than becoming the thing students must learn before they are permitted to disagree. The identical asset, moved across a field boundary, converted from heterodox capital to founding capital. Bourdieu might have enjoyed the purity of it: the value was never in the coin, it was in the game the coin was spent inside.
The rite came in 2003. Daniele Conversi edited Ethnonationalism in the Contemporary World: Walker Connor and the Study of Nationalism, contributors across political science, sociology, history, anthropology, and linguistics, the festschrift, which is the scholarly field’s sacrament of consecration, the moment the community certifies that a man’s name has become a position on its map. Watch the editor’s task as a scene in the sociology of the ritual. Assembling a festschrift means soliciting essays from the consecrated to consecrate, and every acceptance is a position-taking, a scholar filing his claim of descent, and the table of contents becomes a genealogy the new field writes for itself, with Connor at the trunk. Then run the arithmetic that the ritual politely omits. The tribute volume appeared from Routledge when its subject was seventy-seven years old and held the title Distinguished Visiting Professor at Middlebury College, an undergraduate college in Vermont, his base since 1998. The new field could issue symbolic capital in any quantity. It could not issue the other kind retroactively. No festschrift converts into the graduate students never trained, the school of doctoral descendants never founded, the forty years of committee votes never cast. Connor built his succession through the mail instead, decades of letters and manuscript comments and recommendation letters remembered for their humor, an invisible college run from a village post office, which is what the apostolic function looks like when a man has prestige without an apparatus.
So answer the question the career poses, why the prophet had no chair, and answer it without psychology, because the field answers it structurally. He had no chair because chairs are instruments of reproduction and his work was an instrument of rupture. Because the field that should have hired him was heteronomous, its agenda mortgaged to patrons who were buying nation-building, and he had proven the product defective. Because his capital was of the kind that peers can honor and committees cannot spend. Because he was right too early, and in a field, early is a solvency problem: the market can stay orthodox longer than a heretic can stay employable, and the empirical world does not always schedule its confirmations inside a career. And because, when the confirmation came, the rational structure of fields did what it does, not repenting but rerouting, chartering a new game next door where the old heresy could be honored as a founding, at an age when honor was the only currency left to pay him in.
One coda, and it belongs to Bourdieu’s late reflexive turn, the demand that the sociology of fields be aimed at one’s own position. Connor never read his situation in these terms in print, and yet his scholarship contains the homology, exact and unremarked. His life’s argument held that a state can build every apparatus of integration, borders, schools, passports, anthems, and still fail to make the people inside feel themselves one nation, because the apparatus and the belief run on different planes. His career demonstrated the same theorem with the terms exchanged: a discipline can withhold every apparatus, the chair, the students, the line, and fail to prevent the recognition, because in a scholarly field the apparatus and the authority also run, in the end and after long delay, on different planes. The state cannot manufacture the nation. The department cannot, forever, unmake the knowledge. He proved the first proposition with thirty years of evidence. He was the second proposition, tested on his own body, from Springvale to the festschrift, and both propositions reduce to a single line that could stand over the field theory of his case: institutions command everything except the thing they exist to command.
Prematurity in Discovery
Between 1972 and 1989 political scientists cited “Nation-Building or Nation-Destroying?” and changed nothing. The article ran in World Politics, the house journal of the field it attacked, and it became one of the most cited pieces in the study of political development, and the study of political development proceeded as before. Graduate students read it. It appeared on syllabi and in footnotes. The development economists kept their escalator, the one that carried traditional men up into modern nations, and the article that said the escalator sometimes ran the other way sat in the literature without touching the machinery. The finding was in print. It was not in the field’s mind. A discipline can hold a fact in its hand for two decades and never once let the fact into the room where it thinks.
Gunther Stent (1924-2008) gave this condition a name in the same year Connor supplied its best social-science example. Stent was a molecular biologist at Berkeley, a man who had watched the birth of the double helix from close range, and in December 1972 he published an essay in Scientific American on why some correct discoveries arrive dead. He called them premature. His definition was exact and it was not the loose one people reach for when they say a thinker was ahead of his time. A discovery is premature, Stent wrote, when its implications cannot be connected by a series of simple logical steps to the canonical, generally accepted knowledge of its day. The trouble is not that the finding is wrong, or unfashionable, or poorly argued, or resisted by jealous men. The trouble is structural. There is no path. The accepted knowledge offers no chain of small accepted moves that reaches the new result, and so the result, however solid, has nowhere to attach. It floats. It gets cited the way a strange coin gets passed around a table, examined, and pocketed by no one, because no till in the room is built to take it.
Stent’s own case was Oswald Avery (1877-1955). In 1944 Avery and two colleagues at the Rockefeller Institute showed that the substance carrying heredity in pneumococcus was DNA, not protein. They published in a first-rank journal. Biologists knew the paper. And the paper lay near-inert for close to a decade, because the canon held that DNA was a monotonous molecule, four bases in dull repetition, the tetranucleotide picture associated with Phoebus Levene (1869-1940), a molecule too simple to carry the information of life. Proteins were the complex ones, so genes had to be protein. There was no accepted route from “DNA transforms the bacterium” to “genes are made of DNA,” because every accepted step said DNA could not hold enough. The result was correct and unconnectable at once. The steps came later. Erwin Chargaff (1905-2002) broke the tetranucleotide picture by finding the bases in unequal ratios. Hershey and Chase in 1952 traced heredity to the DNA of a virus. Watson and Crick in 1953 showed a structure that could carry information. Only then did Avery become connectable, and once connectable he became obvious, and once obvious his nine years in the cold read as a curiosity rather than a scandal. Gregor Mendel (1822-1884) is the older instance, thirty-four years between his paper and its rediscovery, unconnectable to a biology that had no concept of the discrete hereditary particle his numbers required.
Connor is the same structure moved into political science, and he fits Stent’s definition more cleanly than most laboratory cases, because the canon he faced was not a single wrong belief but an interlocking system that turned his finding into a thing the field could not reach.
Name the canon. Modernization theory rested on a picture of development as a single ascent. Traditional societies were parochial, their loyalties fixed to village, clan, tribe, and dialect. Modern societies were national and integrated. The passage from one to the other ran one direction, powered by the forces Karl Deutsch had measured, social mobilization and communication, the roads and schools and newspapers and radios that pulled men out of the local world and into the wide culture of the state. Ethnicity in this picture was primordial, a residue of the traditional stage, and residues dissolve. The dependent variable everyone cared about was integration, and integration was assumed to be where the escalator went. Every accepted proposition in the system pointed the same way. More communication, more integration. More schooling, more common identity. More development, less ethnicity.
Now set Connor’s finding against that. He said the same forces could produce the opposite result. A road connects the village to the capital and lets the villager see how differently the capital lives. A school teaches the literacy that creates a readership for the minority’s own newspaper. A state broadcast in the state language tells the listener his language is not the state’s. The tools of integration double as tools of mobilization, and modernization, rather than dissolving ethnic identity, can manufacture ethnonational consciousness where none had reached the surface before. He counted the world to prove the canon described almost nothing on the map, twelve homogeneous states out of 132.
Read that against Stent and the prematurity is exact. To connect Connor’s result to the canon a scholar needed a picture in which one process yields opposite outcomes according to conditions. The canon had no such picture and no place to put one. Its architecture was the one-way escalator, and Connor was reporting passengers walking down it, and there was no accepted step, none, that led from “modernization” to “stronger Basque nationalism,” because every accepted step led away from that conclusion by construction. The finding could be filed and could not be used. A field cannot metabolize a result that contradicts the premise required to state the result in the field’s own terms. So it did what fields do with the unconnectable. It cited the coin and pocketed nothing.
Here Connor departs from Avery, and the departure is the part of his case that teaches something Stent’s laboratory examples cannot. Avery escaped prematurity because the connecting steps got built inside his discipline. Chargaff, Hershey and Chase, Watson and Crick supplied the bridge, and the bridge was theoretical and experimental work done by biologists in laboratories. Connor got no bridge of that kind. The event that ended his prematurity was not an advance in political science. It was the collapse of the Soviet Union between 1989 and 1991, and the breakup of Yugoslavia behind it, the most sustained nation-building program in history coming apart along the internal borders of its nationalities, the fault lines the canon had scheduled for dissolution and Connor’s 1984 book had mapped. History ran the experiment the field never ran. The world supplied the missing step.
This places Connor at a joint Stent’s cases do not reach, the joint where prematurity meets the paradigm crash Thomas Kuhn (1922-1996) described. Two routes lead out of the premature state. In the first, someone builds the connecting bridge and the finding attaches to a canon that still stands, which is Avery. In the second, the canon itself falls, and the finding is not so much connected as freed, because the thing it could not be connected to has stopped existing. Connor took the second route. He did not become reachable because political science constructed a path to him. He became reachable because the escalator broke in public, on television, in a single news cycle that repriced every claim in the field at once, and in the rubble his distinctions were the materials lying closest to hand.
In Stent’s biology the connecting steps come from inside, from experiments a discipline can schedule and fund and run. In the study of nations the crucial experiment was run by reality, on reality’s timetable, and no scholar could hasten it. A social science whose decisive tests are performed by history is a science condemned to wait, and waiting outlasts careers. Connor published the finding at forty-one and forty-six. The confirmation arrived when he was sixty-three and sixty-five. Twenty years passed between the correct result and the event that made it legible, and nothing a good man could do at his kitchen table could close that gap, because the gap was not made of missing arguments. It was made of missing history, and history is not a variable the researcher controls.
Stent paired prematurity with a second idea, uniqueness, and it sharpens the reading. He argued that scientific discoveries are not unique. If Watson and Crick had not found the helix in 1953, someone finds it within a year or two, because the discovery sits in the logic of the field and the field is closing on it. Artistic works are unique. No one else writes the Ninth Symphony if Beethoven does not. The test tells you which kind of thing a finding is: could another have made it, and soon. Connor’s finding passes as science, not art, and the proof arrived in 1983, when the insight surfaced in force. Benedict Anderson’s Imagined Communities, Ernest Gellner’s Nations and Nationalism, and the Hobsbawm and Ranger volume The Invention of Tradition all appeared that single year, a cluster too tight for coincidence, the field arriving at the station as the canon that had blocked the platform began to give way. The insight was discoverable and got discovered many times over. Connor was earliest, by fifteen years, and earliest is the position that falls into the prematurity gap. The 1983 authors reached the same country as the gap was closing and were heard. Connor reached it while it was sealed and was filed. Same territory, different arrival times, and the arrival time set the reception.
The premature discoverer pays twice, and the second payment is the crueler one. The first is neglect while the canon holds. The second comes after the canon falls, when the finding turns obvious and re-derivable, and the man’s priority reads as a footnote rather than a rescue, because absorbed knowledge always looks as though it was always known. Connor escaped the second punishment more than most, and the escape was a field accident rather than a Stent effect. Nationalism studies condensed into a discipline just as his prematurity ended, and a new field needs founders, and it reached back and named him one, which is why his distinctions are taught before students are permitted to disagree with them. The prematurity ended in the journals. The credit followed, late, by a separate route.
A correct social science finding can lie fully published and fully inert for two decades because the field has no path to it, and while it lies there governments keep acting on the canon that made the path impossible. Modernization theory was not an academic ornament. It was a policy science with clients, and its clients ran nation-building programs on the assumption Connor had counted out of existence in 1972. The Soviet collapse ended the prematurity inside the discipline. It did not end it inside the state. In 2003 the United States launched a project in Iraq under the name nation-building, on the premise that a people could be manufactured by building the apparatus of a state, roads and schools and a constitution and an army, the premise Connor had falsified before the men running the project were born. So the frame closes on a distinction Stent’s biology never had to draw. A finding can leave the premature state in the library and remain premature in the cabinet room, because absorption into a field and absorption into a government are separate events with separate clocks, and the second clock, the one that spends lives and money, runs slower than the first. Connor got his footnote. The policy got its escalator. The two were reading different literatures, and the gap between them is where the cost fell.
