U.S. Intellectual History Blog

Analysis or Synthesis?

During his tenure leading the American Historical Association, William Cronon has been using his “From the President” column, which appears monthly in Perspectives on History, to explore the various ways in which people seek to understand the past. He’s been comparing and contrasting the habits of professional historians with other people who study the past, such as popular and amateur historians. Cronon’s overarching objective, it seems to me, is to convince professional historians to make their scholarship more accessible to a larger public, without losing the scrupulous disciplinary standards which we’ve been conditioned to since our first graduate seminar. This is a laudable goal, even though, as Ben Alpers rightly pointed out a few months ago, in response to one of Cronon’s editorials, “the story we tell ourselves about academic history appealing to a mass audience is to a very great extent a myth.”


Cronon’s latest column in this series, “Breaking Apart, Pulling Away,” compares and contrasts analysis-driven history with synthesis-driven history. He introduces the subject as follows:


Analysis or synthesis: which should we prefer?

Is it better to explore tightly bounded specialized topics by asking small unasked questions that can be answered as rigorously as possible, combining previously unknown primary documents and technical arguments in original ways whether or not they ultimately matter very much? Or is it better to range widely across the historical landscape, borrowing insights from secondary sources to make large claims, relying even on documents everyone already knows to pursue big familiar questions which however unanswerable, we all recognize to be undeniably important?


Cronon loves his false choices. Of course good historians should closely analyze their subjects in original ways. And of course we should also put our subjects in a broad enough context so as to make our work relevant, which requires synthesis. But Cronon points out that where we’re trained more to do the former, the public is hungrier for the latter. “The Big Questions of history are often what members of the public most want historians to discuss. Yet Big Questions are precisely what our training has taught us to be wary about tackling—and what the sharp knives of our colleagues make us fear we would be unprofessional even to ask.”


Really? Do we still live in this world of hyper-specialization? This is not my understanding of recent historiographical trends. Is my understanding peculiar? Perhaps intellectual historians  tend more to Big Questions than others? Even those of us who focus on intellectual biography usually write about individuals who ranged widely–thus requiring us to be capable of synthesis. Two of my favorite recent intellectual biographies–Daniel Geary’s Radical Ambition: C. Wright Mills, the Left, and American Social Thought and Jennifer Burns’s Goddess of the Market: Ayn Rand and the American Right-could only have been written by historians able to synthesize a great deal of postwar US history.

I’ve been thinking about this issue of synthesis in my own work. I just wrapped up a chapter in my culture wars book on the New Left, broadly conceived to include not only the student antiwar movement and counterculture but also the Black and Chicano Power movements as well as the women’s and gay liberation movements. I am extremely confident that my take on the topic is original, since my purpose in writing this chapter is to set the stage for the culture wars of later decades. I am also fairly sure that the mix of intellectual history I bring together in my analysis will strike most readers as fairly original. And yet, because my topic is very broad, because I am asking and seeking to answer Big Questions, I am necessarily reliant upon the historiography of others. In short, I synthesize. And I am very comfortable with this. In fact, I can’t imagine it any other way. Is this strange? I’d love your thoughts.

25 Thoughts on this Post

  1. It’s not at all strange. To the extent that there is a professional prejudice against synthesis, it needs to be overcome. David Graeber’s Debt has had to take a similar approach in anthropology. I like your point about how intellectual history, even intellectual biography of individuals, often requires synthesis because its subjects’ intellectual endeavors range widely.

    Analysis and synthesis: I’m reminded of C.S. Lewis’s remark (about another such pair of concepts) that it’s like asking which blade of a pair of scissors is more necessary.

  2. “Cronon loves his false choices.”

    Which is why I gave up reading his monthly pontifications after the first couple. It’s as though he’s auditioning for the NYT editorial page the way he sets out these false dichotomies and tries to come up with some reasonable, well-meaning solution.

    A more interesting question is when this dichotomy originated. Or maybe it’s always been there. If Cronon were writing this in the eighteenth century, he’d probably be casting this in terms of the antiquarians vs. the historians (distinct categories until Gibbon finally fused them). I’d say historians don’t read enough Momigliano, but as I haven’t read him, I can’t cast that stone. What I can say is that it seems historians seem condemned to repeat all the battles of the past, even though they really should know better.

  3. I agree the choice between analysis and synthesis is a false dichotomy and further an exercise in the kind of minutia of which Cronin is critical. Both are skills that compliment one another and are required for professional work so why artificially create this contest? I’m currently reading Thomas Bender’s, “A Nation Among Nations” which beautifully combines analysis and synthesis taking a very broad approach to discuss a very specific thesis. It is a very provocative work and one that is probably read only by academics and students but its subject is one that is critical to America’s world view. I would presume that Cronin’s real lament is that works such as this don’t get a larger audience and that the thesis doesn’t get a broader debate.

  4. Thanks for the comments. Varad, I think you’re being unfair to Cronon, who himself admits to setting up false choices. In fact, it was probably unfair of me to write, “Cronon loves his false choices” immediately following the passage I quoted from him since his first sentence following that particular passage is: “These are false choices, of course.” As a friend pointed out to me, I come off as a snarky critic in this instance, which was not my intention. Cronon intentionally puts up false choices as a straw man. I think it’s effective and I don’t have any problems with it as a rhetorical device. My only contention is that it seems to me more and more historians are comfortable with synthesis now than they used to be, which is great but does not ensure that our synthetic work will be read by a non-academic public (as Paul points out in his comment above, Bender’s work is an example of synthesis but is not read by non-scholars). But my contention here was tentative, as I admit. On twitter, Rob Townsend responds, “I think the discipline is as specialized as ever, but also more open to synthesis than 30 years ago.” This seems right.

  5. I admit to being puzzled at Andrew’s claim, and at the unanimous agreement with it among the commentators. I find nothing “false” about Cronon’s dichotomy. Indeed, as one who works in history departments but was not trained or socialized in that particular discipline, I find it to be beyond question that historians have a preference for archival work, primary documents and “engaging the literature.” Indeed, in my experience this preference is so strong that it might better be labeled a fetish or a prejudice.

    I will certainly grant that Cronon’s dichotomy is slanted: “small, unasked questions” answered by reference to “primary documents and technical arguments” are not necessarily projects that don’t “matter very much.” And lots of books that take on “big questions” are self-indulgent wankfests. But if one takes a look at any publishing catalog in history, he/she will see vastly more books that fit into Cronon’s first category than his second. (An awful lot of books published these days have subtitles that end with something like “…in Franklin County, 1820-1860.”) And as a generalization, historians treat Doris Kearns Goodwin, David McCullough and the like with such disdain that I don’t see how anyone can seriously say that the profession appreciates this kind of work.

    Finally, and most importantly, I don’t believe that a person would be allowed to write a dissertation in a history department along synthetic lines, nor is it likely (or possible?) that a person would impress a hiring committee and get a tenure-track job with the kind of book that exemplifies the second category. The profession appears to reserve books that “range widely across the historical landscape, borrowing insights from secondary sources to make large claims, relying even on documents everyone already knows to pursue big familiar questions” for senior scholars who have already “proven” themselves as historians by having previously written the other kind of work.

    I would therefore argue that, yes, these are different kinds of books and, yes, the contemporary configuration of the historical-academic complex clearly privileges the narrower, more rigorous form over the broader, more accessible one. (I don’t think that “primary” or “analytic” versus “synthetic” is the best way to name these styles, but that is an issue for another day.) And I am puzzled as to why others do not see something that seems–to me, at least–so obvious. I wonder if what puts people off is not Cronon’s actual cleavage, but his question of which is “better.” I–along with, I suspect, the other commenters–don’t see the need to make that choice. And I don’t know why Cronon would either. Though he seems to side with the “big picture” books, I cannot imagine that he would say that there is something illegitimate about the narrower ones. In fact, the kind of books that he lauds are substantially dependent upon the other sort for their existence. I see the relationship as more of a symbiotic one. But as long as two sorts of work exist within the same field, one style will always be “up” and the other “down.” At the moment, big books with sweeping claims are “out,” and I suspect that Cronon’s implicitly disparaging description of the other sort of project stems from, as Paul said, his frustration at the lack of respect given the “synthetic” works.

  6. Thanks, Mike. I knew this topic would generate a comment from you! You’re definitely write that the two types of scholarship are symbiotic. And you’re probably right that more narrow but rigorous books are still “up.” But would you say this is the case for intellectual history?

    Think about former USIH blogger David Sehat’s Big Question book, “The Myth of American Religious History,” which began as a dissertation. My first book, “Education and the Cold War,” began as a dissertation–quite a bit of it was broad and synthetic.

    And what about your book, “A Commercial Republic: Democratic Capitalism in American Thought,” which began as a dissertation and is contracted to be published by Kansas?

    I could go on. If intellectual history is making a comeback in the broader field of US history–I know, I know, big “if”–and if intellectual history is much more open to Big Question scholarship, then perhaps synthesis is on its way back up? Wishful thinking? Most of the books with titles that end with “…in Franklin County, 1820-1860” are social history. I think social history is in slow decline. And, another can of worms is now open.

  7. I think it’s important to realize that we’re talking about three dichotomies, not one or two. On the one hand, we have the analysis/synthesis distinction: as the title of Cronon’s post puts it, “breaking apart, putting together.” Or as we’ve seen in other debates on this blog, “splitting vs. lumping.” On the other hand, we have a distinction between focused studies that rely for their punch on close attention to primary sources, usually new archival sources; and broader studies that pull widely across time, space, and previous work in the discipline to make a point (that’s more like three distinctions, but it’s manageable as one). On the other, other hand, we have a distinction between small questions and big ones, which often appears in debates about public relevance. While we’ve been treating these as prototypically correlated, all three choices are independent of each other.

    It should be obvious that you can write a narrow synthesis and a broad analysis, just as you can answer big questions with analysis and small questions with synthesis. I think what we’re really after here is the relationship between the second and third distinctions, and that we’ve been slowly trying to push out the analysis/synthesis framework under which Cronon tries to, ahem, lump them. So: can you answer big questions with narrow studies and small questions with broad ones?

    I say, sure. The archetypal example is microhistory, but in our own field, think of Veysey’s Emergence of the American University: answers big questions, but gets its results by using all the damn primary sources he can get his hands on, doesn’t cover more than about forty years and that in a fairly carefully defined space of actors. Perry Miller (who did, it’s true, lump like a fiend) answered big questions through Very Close Attention to primary documents. Perhaps the era has passed when that was possible? Or perhaps the era has merely passed when that could be done with *new* archival sources that aren’t recent. This may get at a lot of what Cronon is asking. As for the other side of the equation, it is absolutely possible to write a very broad, synthetic work that ranges across time, space, and source base without answering questions of major significance. This is often a flaw of bad biographical writing (although it doesn’t have to be, if you set out to answer a small question). Cf. the thread on Robert Caro.

    I have little insight into what “the field” encourages or discourages in the aggregate, but as Andrew notes, it’s definitely possible to publish well-received, broad scholarship in one’s early career. Andrew points to books that came from his, Mike’s, and David Sehat’s dissertations. I’ll point also to Sarah Igo’s “The Averaged American,” which employed the triple-case-study model but which I think we can agree was big, broad, and well-received. As for the question of social and intellectual history and their fates, I have even less to offer.

  8. “…you can answer big questions with analysis…” Well, put, John. But is this often the case? Other than in famous microhistory books like Ginzburg’s “The Cheese and the Worms”? Sure, plenty of the authors of the narrow social histories that proliferated in the 70s, 80s, and 90s made grandiose claims about how their subject matter revealed universal truths. But were their claims ever convincing? Or were these merely overwrought attempts to claim the relevance of their work?

  9. I must apologize: I don’t have an answer to that question, since I’m not nearly well-read enough in the literature. I should certainly be loath to suggest that The Cheese and the Worms is something that can be done more than rarely, however, and I imagine the “grandiose” microhistories you reference are very weak to the extent that they try to stand on their own. But it’s not so difficult for a narrow study to put itself in league with a big-questions literature, even when the core of the argument is focused. It is also often possible to pose new questions with such research by problematizing widely shared assumptions or conclusions. So the question is, how partial can answers to big questions be before they become small? And did those grandiose histories satisfy a lower (but sufficient) bar than the one they set for themselves?

  10. Andrew, if I wasn’t worried before I read this post about 1) passing my qualifying exams and 2) writing a good dissertation, I sure am now. Thanks.

    I mean, I have spent the last two weeks absolutely immersed in the historiography section of my reading list, where this very dichotomy that Cronon posed and that you picked up emerges repeatedly as one of the leitmotifs of the literature (along with causality v. meaning, explanation v. understanding, event v. thought, etc., etc.).

    Having read this far (just finished Hayden White this morning and am headed into the home stretch with Novick, Haskell, Megill, Wickberg and Klein), I asked myself, “Could I sit down and write a cogent, nuanced, thorough response to this blog post and the comment thread?” And I decided that, No, I could not. I find this conclusion distressing. I mean, I just read this stuff — read it, took notes on it, wrote precis on it — and yet if you asked me right now to give you a history of how this dichotomy has been deployed and what historians have had to say about it, I’d be up a creek for sure. (I will say that Hofstadter’s essay in Fritz Stern’s volume, as well as Mink’s essay, are probably where I would begin if I had to sit down and answer this question right now. But I might not get very far.) Fortunately, I’m not taking my exams until Fall of 2013, so I have some time to figure this stuff out. I’m just frustrated that more of it didn’t “stick” on the first pass; I feel like I have to go back and read it all over again now.

    And then there’s the dissertation. I think John’s taxonomy of dichotomies was helpful here — though Mike’s comment, John’s comment, and yours are like a hat-trick of horrible possibilities. “Look at how many ways the wheels can fall off the cart!” As I mentioned in an earlier post, I have changed my dissertation topic. Won’t go into that here — I’ll post on it later — except to say that considerations of whether my new periodization/topic/approach would be part of the resurgence of synthesis or the death-spiral of analysis did not enter into my calculus. And perhaps that’s for the best.

    Anyway, your post was a jarring reminder of just what a big hill I am climbing in getting ready for exams — all so that I can tackle the Mt. Everest of the dissertation.

    Good times.

  11. @LD—Your concern is bringing out my caring paternal instinct. As such, here goes: Just remember that there are NO RIGHT ANSWERS in the humanities, and no answers are ever thorough enough. In historical monographs there is evidence and plausibility. And in historiography the key is to remember who is in conversation with whom. Whether these works are analysis or synthesis can be beside the point, in some ways. In a field/qualifying exam situation you just have to show that you’re aware of the conversations between authors about theory and/or evidence. Your inspectional reading of all these works will bring out those points. And I’m sure your exam readers will be tolerant of differences of opinion. Hang in there! – TL

  12. Tim, you are very kind. Seriously. If it’s all the same to you, I’ll take your remarks as friendly advice from a fellow academic, rather than expressions of fatherly concern. I was mostly just expressing my exasperation with my learning process, which isn’t fast enough to suit me. In that regard, my examiners will probably prove more tolerant than I am. We’ll see.

    As to there being “no right answers” in the humanities — well, maybe, but I’m pretty sure there are some wrong ones. And while my examiners might be tolerant, they will definitely be tough. And I have a whole year to worry about it. Everybody wins!

    • I apologize for the word paternal. I certainly didn’t mean anything fatherly, esp. in any age/maturity way, toward you. It’s been one of those days here. …On your point about there being wrong answers, well, yes—and no. It depends on the mood of the person reading your piece! 😉 For what it’s worth, I think that tolerant and tough can stand together. I’m 100 percent confident you’re going to knock ’em dead. – TL

  13. LD: From my perspective, the only thing you have to worry about is that you worry too much. You’re damn smart and you obviously work like a demon. Chill out. If it makes you feel any better there were certainly moments during my time prepping for comps when I felt helplessly ignorant of the material. I think this is typical.

  14. “As to there being ‘no right answers’ in the humanities — well, maybe, but I’m pretty sure there are some wrong ones.”

    Oh, there are definitely wrong answers in the humanities. Some off the top of my head: the Civil War wasn’t about slavery, the American Revolution was conservative, the Earl of Oxford wrote Shakespeare’s plays, there was a “Burning Time” during the early modern period in which millions of witches were burned at the stake, the Renaissance was the result of the Ottoman conquest of Byzantium, Homer was a woman, parents were indifferent to their children before the industrial age, etc.

  15. I think John Gee is on the right path when he suggests that we fit the analysis/synthesis binary within larger frameworks of lumping/splitting, primary/secondary, and so forth. As I suggested above, the issues aren’t particularly new, as they were being debated in the seventeenth and eighteenth centuries. The difference now is that all of these dichotomies fall under the rubric “history,” whereas back then they described distinct endeavors that were felt to have no bearing on each other. The people who did the research (antiquarianism, philology, archaeology – everything you might consider “source criticism” of one kind or another) weren’t historians, and the historians (those who wrote narratives about the past) weren’t researchers. Nowadays no one would be considred a historian who wasn’t a researcher. DIY is part of the job description, even if though the research can take myriad forms.

    I do wonder what the German historicists said about this divide. In a lot of ways it’s their fault, if fault is the term, that we still have these debates, since they’re the ones (with a big assist from Gibbon) who fused synthesis and analysis and made it our SOP – standard operating procedure.

    I also wonder if this goes in cycles. If so, that may explain why Mike and Andrew can have such diametrically opposed perspectives yet both be correct. Andrew’s right that historians seem to be more comfortable with synthesis than they were in the previous few decades. Yet Mike’s also right that if you take a look, you’ll find lots of narrow, even trivial studies with “subtitles that end with something like ‘…in Franklin County, 1820-1860.'” But the vogue for the social, demographic, and cliometric history that produced libraries of these airless tomes hit its peak in the late 1970s and early 1980s, and it’s been on the decline ever since. As that trend has gone into eclipse, synthesis and narrative have been gradually returning to the light. But the impression that historians do a lot of tedious number crunching lingers, both because they did, and because some of them still do that sort of thing because it’s what they were trained to do.*

    There was a time in my field when the cutting edge was histories of provincial academies, book distribution networks, and the organization of salons. It’s enough to make one channel Benedict Cumberbatch in Sherlock and bellow, “Boring!” I suppose the world is an immeasurably better place for our knowing that pornography was rampant during the Enlightenment, but then what?

    “Who cares about stupid towns in Texas? Where is width and depth? We (I) want oceans, not wells.

    I scribbled that in the back of my copy of Structuring the Past: The Use of Computers in History, an AHA pamhplet by Janice Reiff about the subtitular subject published in 1991. I did so in 1993, when I was a junior in college and read Reiff’s little book for my historiography proseminar. My attitude has not altered one jot in the nineteen years since. Admittedly, my research drills a well: what people in the eighteenth century thought about classical Sparta. But it does so to reach an ocean: how people think about the past.

    Newton said that whatever others might think of him, in his own eyes all he was doing was playing on the seashore with a smooth pebble here and pretty shell there, “whilst the great ocean of truth lay all undiscovered before me.” We are all of us playing along the seashore, but we must never forget that ocean along which we toil.

    • *Not all the social history was lousy. It gave us Thompson’s Making of the English Working Class and Olwen Hufton’s The Poor of Eighteenth-Century France the latter being one of my half-dozen or so favorite works of history. But it also gave us The Cheese and the Worms, the history equivalent of the Pied Piper of Hamelin, Natalie Davis’ hugely overrated Society and Culture in Early Modern France, the history equivalent of no there there, and Lynn Hunt’s equally overrated Politics, Culture, and Class in the French Revolution. The latter two especially seem to embody the principle of ’70s and ’80s social history that if the past is a foreign country, then it has to be one of those dull, boring ones. But I suppose that’s a better approach than something like Piero Camporesi’s Bread of Dreams, which comes out of the same milieu as Ginzburg, and treats the past as a speculative fiction in which both the author and his characters have been dropping acid.

  16. Okay, it really irked me that after reading 5,000+ pages of historiography in the past week, I felt that I had nothing to add to this conversation. So I thought over what I have read so far, looked back at my notes, and, while I am not ready to present a paper on the subject, I at least have an idea of what I would say.

    What makes Cronon’s dichotomy (admittedly) false is the assumption that someone who borrows “insights from secondary sources to make large claims, relying even on documents everyone already knows” hasn’t also addressed “small unasked questions that can be answered as rigorously as possible.” He pretends that “big idea” synthesis of the kind promised/practiced by Armitage, Kloppenberg, et. al.–or, for that matter, by David Sehat or Susan Pearson or Andrew Hartman–doesn’t build its evidentiary authority on close reading and fine-grained analysis of documentary sources. Further, the only way that such matters of close investigation may not “ultimately matter very much” is if they are constructed as cul-de-sacs of antiquarian curiosity rather than as avenues into a much broader, more complex past context.

    As I mentioned above, I would begin with Hofstadter on “History and the Social Sciences.” Here is his observation, and I think it bears repeating:

    The answers to small questions sometimes shed bright but narrow beams of light on th larger problems of human behavior and the social process, and it is not unthinkable that they may have some important cumulative rsult. But there is no reason to think that the answers to all such questions will add up to a comprehensive view of society or of historical processes. The historian stands in somewhat the same relation to his pile of monographs. Monographs are useful, but when put together they do not yield comprehensive answers to the comprehensive questions. No synthesis — at least no “scientific” and commonly acceptable synthesis–can be reached through sheer addition. (in Fritz Stern, Varieties of History, 369)

    Indeed, the debate over “anlaysis v. synthesis” is but one front in that older, larger debate about where/whether history fits in the “social sciences,” and in what sense historical knowledge can be “scientific.” This is where Mink comes in, with his explication of historical epistemology as offering a way of knowing and an object of knowledge (they often go together, don’t they!) that “science” — even social science — cannot encompass. And this is where the Wingspread conferences comes in as well, since one of the main questions at issue there was not just “social history v. intellectual history” (w/ social history on the ascendant, it seemed) but also “social science v. historical inquiry” — and, yes, analysis v. synthesis.

    On that point, Rush Welter’s essay “On Studying the National Mind” puts the matter nicely:

    …good intellectual history requires an extraordinarily broad awareness of the intellectual orientation of the whole society or region in which any given statements were made; generalization of this sort is an indispensable preamble to studying intellectual events….The obverse is also true: particular studies of discrete groups can have little meaning unless they serve to illuminate a more general perception of historic phenomena….Only competent understanding of the larger context of events will lend appropriate meaning and significance to ay particular study, no matter how carefully conducted

    And then I’d probably go on to invoke Hayden White, but I think I’ve run out of comment space.

    Anyway, I feel better about being able to answer this question. :-/

    • Russ Welter, “On Studying the National Mind,” in New Directions in American Intellectual History, edited by John Higham and Paul K. Conkin (Baltimore: Johns Hopkins UP, 1979), 65.

    • And all the typos/misspellings/grammatical errors in the above comment are mine. Sorry. Thinking faster than I type, and apparently reading historiography faster than I think!

  17. “Okay, it really irked me that after reading 5,000+ pages of historiography in the past week.”

    5,000 pages? Seriously? How is that even possible?* That’s 750 pages a day. My eyes would be bleeding after three days if I tried that. Heck, maybe two days. I know I said I want oceans, but this is a deluge; it’s enough to drown. No wonder you seem a tad dyspeptic, LD. Trying to swallow so much in so little time would give anyone indigestion.

    • *I read Harry Potter and the Deathly Hallows in a day, which is 750 pages, so it is possible to read that much in a day. But that’s Harry Potter. 750 pages of historiography would be another matter entirely, I’m sure.

      NB: The * belongs at the end of the next sentence, so move it there mentally. Thanks.

    • That was an intentionally hyperbolic statistic. I could as well have written “1,000,000+” — it feels about the same.

      But, in the interest of historical accuracy, I went back and added it up. In the past seven days, I have read 1,446 pages of historiography, taken 56 pages of notes, and written five precis of 1,000-1,500 words each.

      So it was about 200 pages a day of reading, plus the note-taking, plus the writing. I am assuming that this is a normal workload for someone getting ready for exams — but if it isn’t, I don’t want to know. Seriously.

    • 5,000 pages is just plausible enough that it’s hard to tell. If you’d said “1,000,000+” there’d have been no doubt. It’s kind of like when Mitt Romney said he’d bet $10,000. Probably a joke, but given that he could easily afford it, maybe not. That gives us an easy rule of thumb: the more zeros, the more exaggerated.

Comments are closed.