These last days, I have received – like some of you I guess – two guidelines about how to do a master or a Phd Thesis in history of economics. That is too bad that I completed mine 10 years ago! Anyway, I am not sure that these pieces are really useful for it very much depends on what is the conception of history of economics you have and you intend to convey in your work. The best advice I can give – and that is the way I proceeded myself – is to look a recent Phd Thesis from someone you respect as a scholar and whom you feel shares your general methodological outlook, and use it as a kind of reference document. Receiving these guidelines made me wonder about the fact that in the blog, which is made by if not made for young (and not so young anymore) scholars there is, indeed, very little information about one organizes his research work.
There was a time, not so long ago, when the would-be historian of economics can safely go on the first day of his research the university library, inquire the librarian about where the complete works of Ricardo (by Sraffa) or the complete work of Marx (or Keynes or Hayek) were and very much sit here for the rest of his Phd Thesis. 
This is not the case anymore and it is now common and almost an obligatory requirement for a Phd student to have done some archival work. However, doing the Archives can be a very different – sometime nice, sometime quite painful – experience depending on which Archives you go. A few looks like that:
But quite often, you end up like this nice looking young researcher on the left. This image might look scary, but you may have gotten the wrong idea. Because for all the reassuring neatness and geometrical perfection of the archive below, the archivist may have done a very lousy job in cataloging it, meaning that you probably would have to open half of the drawers and go thoroughly through their content to get anywhere. Whereas, on the other hand, the half-smiling women might very well be the curator who knows in detail the content of the folders she is actually browsing for preparing the new catalog and you would get what you want in less than an hour!
More seriously, when looking forward to do an archive, it is very important to understand that the easier you get the information – by browsing on a digitallized catalog for example – the less chance you have to find something really new and unexpected. On the other hand, when you are inquiring about an archive on which you have very little information it is most important to keep a very open mind and to be a little stubborn even when the odds seems against you. Moreover, keep in mind that curators or archivists made mistakes and have limited knowledge of the content of their own archives. When one has been working for a long time in a specific archive, one often knows it better than its own curator.
Let me share a recent – last week in fact – experience I had in the Archives of the Museum of Science and Industry in Chicago. In the context of a research project we are conducting with Yann on the visualization in US economics, we became interested in the Vienna economist and philosopher Otto Neurath and its possible connections on the other side of the Atlantic. As one of the former directors of the MSI was cousin to the economist and philosopher Otto Neurath, we inquired about the possible existence of a correspondence between the two cousins. We were told by an assistant curator that there was no correspondence. Indeed, as I discovered last week, the same assistant curator had been already sollicited, two years ago, by another researcher for the same piece and after working on it for a few days was unable to find it. The same researcher who was preparing a book on Neurath e-mailed the MSI again a few months later because he would go to Chicago for a conference, asking if it was worthwhile passing by the Museum archives to give a try about the correspondence. As the assistant curator repeated to him that he was unable to find anything, the researcher decided not to come. Hence, when he answered to our query, the assistant curator was pretty sure that there was nothing to find.
Being in Chicago for another project, I decided to give it a try anyway and spend a few days, at least one, in the MSI archives. The first day, I came with my sole computer not knowing if there was anything worthy. Looking through the inventory (made by the same competent assistant curator mentioned above) I did find a few things, enough to spend the whole day there and too decide to come back two days later with a digital camera to save my findings. I did not find the correspondence though: at the letter “N”, there was no folder “Neurath”. However, I found a mention to a letter the former MSI Director received from Neurath in one of the documents I read. It was enough to convince that there might be something that have been missed. On the second day and after digitallizing what I had found , I decided to go ‘fishing’. What I did was fairly simple and obvious, instead of looking at the name of the person, I looked at the name of the institution he was in charge at the time (the Vienna Museum of Economy and Society). And, indeed, I found the correspondence: dozens of letters and a few important documents that were supposed not to exist. Later on (I came for a third day), I tried to force my luck and repeat the same procedure with another individual. But this time I end up in a dead endd: the box I wanted to look at was nowhere to be found (I was looking for the box with the letter P, but the boxes between M and R were missing). We looked with the assistant curator in the premisses, but nothing! However, instead of calling it a day I asked for another kind of documents (pictures) to spend the two hours that were left for my last day at the archives. The assistant curator took me to the location where the pictures were archived but we were unable to find any of interest for me. Out of sheer curiosity, I browsed the shelves and look at the name on the boxes that were around, just to find the correspondence box I was looking after a few hours before. Inside was indeed the piece of correspondence I expected and a few valuable documents.
To sum up this very long post:
– When there might be a chance you can get an important piece of information or documentation in an archive, but you are told that there is nothing by the archivist. Try to verify it by yourself and take the necessary few days to do it properly. If you find nothing, which happens quite often, you may have lose a few hours or a few days, but if you find something unexpected you may have gain an easy and good article or a chapter from your forthcoming Phd thesis, which may in turn launch or speed up your career.
– When you plan to do an archive, use all the time you have even if the odds are that you ain’t gonna get anything more.
– Keep an open mind, if you do not find something at the obvious location it does not mean it is not there, it might have been placed somewhere else because the archivist does not have the same logic as you (you are thinking names of individuals and he sees names of institutions, or the reverse) or because it have been misplaced (shit happens).
– Do not lose heart. Especially when working on a archive that has not been properly catalogued or arranged. Quite often, the first hours or even the first days are not very useful: you do not know how to begin, what you are really looking for, you do not understand the logic of classification (which is almost always different in each archive). You may have the impression that it is like finding a needle in a haystack, there is however one big difference: when the archives are classified and most of them are, it means that there is a logic to it, you just have to find it! – Needless to say that it is easier say than done.
History of Economics Playground Redux 
This is a very nice post!
I think, however, that it does not address much another important question, which is: “Why is it so important to do the archives?”. Obviously, those who have written the guideline for writing PhD dissertation thought that doing the archives is a subsidiary activity and that reading and interpreting the “texts” is what history of economics is about. On the other hand, the Center for the History of Political Economy at Duke, one of the main attempts to give the field more visibility, is a project which is built in close relation with the fact that Duke has this impressive collection of Economists’ papers and those are supposed to be one of the main attractions of the Center. All the visitors at Duke are expected to work on the archival materials to some extent.
I would argue that the archives are not only interesting for those who think that the most important things are the context or the practices. The large array of materials that are in the archives can be meaningful for those who are mainly interested in the texts, because the archives constitute some kind of texts themselves. They contain drafts, working papers and correspondences that may help give another interpretation of some published papers and help understand economic controversies.
I should stress, however, that generally, if you search the archives for something in particular, you have little chance to be satisfied with the result. Having a strong thesis and thinking that the archives are going to provide evidence will lead to disappointment (except if you are extremely lucky). History is like a puzzle and the archives only provide pieces of a larger picture the historian has to build for himself. In addition, the same piece can be used to build different pictures. In the story Loïc has told, it is possible that the elements he found useful and fascinating in the archives would appear uninteresting and uninspiring for other people’s work. As part of the same research Loïc is interested in, I have come across some fascinating correspondences that many people have probably looked at before, but did not take into consideration because they were simply not interested in the same issues Loïc and I are. In other words, the archives themselves have no utility if they can’t be integrated into one’s narrative. Part of our work as historians will be to convince other people that these materials are worth considering and that they help understand economic issues in a different way. This is far from being obvious and that’s why archival work can be simultaneously frustrating and fascinating!
These days, I’m working at the Duke library warehouses with Michael Thomas and we try to arrange Leonid Hurwicz’s papers so that future historians will be able to find some interesting materials out of it. Hurwicz, obviously, kept everything. He had tons of handwritten notes, teaching materials, reprints of various papers, brochures, referee reports and so on … all in all, there are 180 boxes full of these materials. Most of them are hardly identifiable and sometimes, skimming through these appear as a dull and unproductive work … but sometimes these materials begin to speak and yield some interesting issues: Why was Hurwicz so fascinated by weather reports? Why did he collect home improvement magazines? Hurwicz’s papers are too vast to be processed in a fully satisfying way but historians who will spend a lot of time studying them and will open the seemingly less interesting boxes may well discover some treasures there!
So that gives us two reasons for doing the archives: 1) to support an already existing thesis (in which case you’ll most probably find nothing, 2) to dig for gold…
3) given your well constructed historical argument, the archives can provide a doubts or support for the supposition that two people were in contact through correspondence or not. And if they were, what they were discussing.
(I think that is somewhat different than trying to prove a thesis in the archives, and lets not forget Popper, empirical evidence can only falsify, and at the end of the day, archival material are empirics to some extent)
I do not share your (Benjamin) sentiment. The archives are the bodily matter on which you can construct your argument, without them you are always on the mercy on some archive nerd that some correspondence or draft proves you wrong. Let me take an example: after a brilliant reasoning spiced with a handful of careful chosen citations, you come to the conclusion that Hume’s theory of flow mechanism had nothing to do with Cantillon’s view on international monetary flow, so far so good. But then, someone discovers in some remote archives a letter from Hume stating that he had read Cantillon in manuscript and that he was quite taken by what he says on international money movements.
Your brilliant hypothesis then needs some serious revising. It does not mean that the archive is all, because it does not speak by itself, but it means that the archives can often be what makes the difference between a brilliant but useless piece of reasoning and a brilliant and important paper in history of economics.
A bit like the much repeated thesis that Schumpeter’s view of economic development had absolutely nothing to do with biological analogies of evolution, until somebody found a letter by Schumpter to Frisch where he stated something really like “to think about innovations, I prefer the mutation analogy to the shock-to-the-pendulum analogy.”
Yet, this is just fragmentary evidence in a broader argument that remains to be narrated. Archives, like any other object, do not speak for themselves, they are raw material waiting for further interpretation (I am not arguing against anybody in particular here, I merely enjoy talking about archives).
I would disagree that the archives are the only “bodily matter on which you can construct your argument”, in fact, I would prefer to take Clement’s broader view and think of the archives as one source of raw material which needs to be considered (very seriously) within the broader narrative. That narrative can be constructed from other sources, but it is necessary to visit the archives within that.
On my own point: If you have a beautifully constructed thesis about Hume and Cantillon, and it is ruined by the ugly fact that Hume himself wrote to contradict your point, then the archives have provided serious meat to your topic, albeit not your theory. Therefore you need to re-think and re-work the whole paper, if not throw it out – as painful as that can be. Lets be honest, the argument posed was at worst plainly wrong, so at least we have learned that.
If an ‘archive nerd’ can find it, then as you said earlier, we should be that archive nerd to start with. That’s where the archives are so important to me, in supporting or rejecting some hypothesis, for which we can only find evidence in the archives. While they can provide brand new material (‘striking gold’ as Floris says), the archives need not, and probably should not – in my opinion – form the backbone of every paper.
Do you guys (Loic and Benjamin) really mean to give us “rules” for digging into the archives ?
What my very little archive experience taught me is that there are really no rules beyond 1) don’t panic, 2) Loic’s “keep an open mind” I totally agree on, and which should de facto exclude other rules. Each time you work on a new project, paper, you use archive differently. Sometimes it looks like what Benjamin says, you have a thesis, and you try to confirm or disconfirm it, somehow like you do empirical work. Usually you fails because you rarely find exactely the correspondance or assertion you look for (in fact, you are more likely to find it in other people’s archives, colleagues, opponents, editors, family, etc.). But if a nerd later comes up with a piece related to your subject a few years, it not so often prove or disprove your thesis. Often it’s not straight, you have to interpret it, and it gets a totally different meaning depending on the line of argument it’s used in. Watch for instance how different the interpretation of the Friedman-Alchian correspondance on evolutionary matters and the biological analogy in Friedman’s essay is in Mirowski and other Friedman specialists.
In other research projects, when the period/subject/person have not been much investigated, you start from scratch, so after a preliminary research you imerge yourself into archives, browse them lenghtly, generally and as exhaustively as you can in a few weeks or so, and try to transcribe in a paper the picture you got from them. In that case the archives genuinly are “the bodily matter on which you can construct your argument.”
And sometimes you unexpectedly unearth a set of material not directly related to you present research. You find it revealing of what was going on in the profession/society in such and such period, and you write an article on it. For instance a dispute where two big guns, in the privacy of their letter exchanges, clarify their scientific vision and smash up others’ in a way they never do in published material.
That’s three situations I’ve been confronted to. I guess there’s an infinity. But that’s enough to conclude that while sharing our archive experiences is very useful (such is for me Loic’s insistence on classification or the lack thererof), turning them into a set of rules or a general philosophy is less so.
And just because each research is different, the Hume-Cantillon exemple, at least the information Loic gives, does not make much sense to me. Finding a letter where Hume says he read Cantillon and “was taken by his argument” raises more issues than it solves, and may not even totally contradict the thesis that the two are independant if it’s the kind of “I was thrilled by what I read because it is completely in line whith what I’ve been thinking independantly” remark. Or it could be “I read Cantillon and it changed my mind” or” I read Cantillon and it turned my attention to something I had neglected” or else. Each interpretation give credit to different intellectual processes and you would then need more material to choose one. I don’t know the story here, but I’m not just trying to split hair. You often find the kind of “I’ve discussed with X”, “Y sent me his paper”, “I attended Z lecture” sentences you feel happy with at first in archives, only to realise that it “proves” nothing and gives little information. Actually, I’m very surprised to see how little archive research solve historical controversies.
With some exceptions – Kenneth Carpenter on the diffusion of the Wealth of Nation in France and in French – librarians do not write history of economics! It is quite amusing that people who have been trained as “post-(whatever)” historians of economics use a Popperian framework to talk about the use of archives!! I have doubts that any piece of archival material can itself constitute a revision of previous accounts in the history of economics. As Samuelson put it, it takes a theory to kill a theory! I would tend to agree with Clément that archives constitute some raw materials and that the role of the historian is to integrate them into a convincing narrative. But it does not mean that Loïc is necessarily wrong. Some papers can be built from archival materials, but on the other hand, there must be a reason, an intuition, a purpose that led you to some particular collection.
It’s like the chicken and the egg … there’s no need to ask what comes first. The thesis and the materials mutually stabilize (reinforce) each other. You begin with an intuition and find new materials that help articulate your framework, then you can interpret any new material in relation to this framework until you find a new framework and so on …
Some people disagree with Phil Mirowski’s Machine Dreams but no one would reject the wonderful archival work that has been done by the author. What is at stake is how satisfyingly these materials are put together. Someone may come with a more convincing narrative using the same materials as well as other sources, but there’s no arguing that Phi offered a strong reading of the materials in the first place. At least, that’s the way I see it.
EDIT: I saw Béatrice’s comment after posting mine and I agree completely.
Loic,
I think your post is wonderful, but I also think you begin with a false premise about previous historians of economic thought. Who ever just read the collected works and thought they had done their job?
Take two examples — Ross Emmett and then Karen Vaughn. Ross dug deep into the resources on Knight, and Vaughn with respect to Locke.
Now some people did history of thought as a secondary field and may have at times relied on searching through the collected works. But anyone whose primary research was in the field combined archival work with textual exegesis.
The difference is in the interpretative act and the requirement of contextualization that many of the younger generation now feel is primary to acts of rational reconstruction, or “getting it right” as far as the economic doctrine is concerned. But reading letters, early drafts, and other background information has been part of the historian of economics craft for sometime.
Just as contemporary economic historians such as Joel Mokyr and Avner Greif did not just do cliometrics, but also engaged in archival research, so even the older historians of economics might have focused on texts, they didn’t eschew archives in their research if they were going to make a serious contribution to the field.
On how one does “detective” work in the archives, like in most acts of interpretation, the glasses one wears brings into sharpe image some things, but often clouds others. Thus, the need for dialogue, criticism, reformulation, etc. This, btw, is why the discussions of hermeneutics were important in the 1980s and 1990s. That conversation is different from Stanley Fish.
To Pete: It does not seem to me that I was implying that we (my generation and the subesequent ones) are the first to do Archives. What I was saying is that nowadays (and I maintain that it is something recent, 10 years ago it was not there), it seems established in the history of economics’ profession that doing archives is a kind of must do for Phd students even for purely strategic reasons and that it does not really impact on your work. In a way not unlike, that if you were doing your PhThesis in history of economics in post 1968 France and not discussing what Marx had to say on your topic or the marxist interpretation, you were not a normal researcher. I acknowledge the fact that we are in this area as in many others building on the shoulders of giant. The difference I am pointing does not relate to individual methods of research, but to the way these methods are considered by the profession as the whole.
Loic,
OK, granted the expectations have changed. But I still think the people who were recognized as doing the “best” work, always engaged in that sort of scholarship. And since the jobs were always few and far between, those who got the jobs in history of thought practiced along those lines. It was the dabblers in history of thought, that didn’t use archives. Think of the scholarship that Dan Hammond did on Friedman — making extensive use of the archives on Friedman (that was actually published well over a decade ago), or Caldwell’s work on Hayek (again extensive use of the archives), let alone the numerous works on Keynes.
But I understand the point you are making about professional expectations in general. Textual exegesis doesn’t cut it anymore.
I also think there is a philosophical shift in the way the archives are viewed — though I think your discussion is actually more consistent with the older view ironically (perhaps because you got your PhD 10 years ago now!, don’t be sad I got mine 20 years ago this year — time flies when you are having fun).
BTW, where did you get these guidelines? How many students today are actually writing dissertations IN the field? What institutions are they doing this at? And, how is the placement record of these students in getting tenure-track appointments?
I am assuming that if someone wrote guidelines, there must be a significant number (which is great) and that they must be able to find employment after completing their degrees. This is all VERY encouraging signs to me, I would like to know the data and the experience.
When I was coming out (20 years ago), you needed to have a primary field that was identifiably in economics, and then history of thought was viewed as a teaching field. This explains the public economics work by Steve Medema, or the comparative systems work by myself in addition to our research on the history and methodology of economics.
I am very excited to learn of these new developments in the field.
Thinking that HE is similar across countries independent of the institutions in which HE is situated leads to confusion. Pete’s comments about HE dissertations is misleading. The US and France are not institutionally similar and so the job markets are not comparable. One can get caught up in “Americanization” of all scholarship and seeing that everyone uses Mas-Colell, but this masks the differences. Germany is still different, as is Italy, as is Japan. The earlier thread on countries represented at HES is important.
Loic’s arguments about archival work makes sense for those trained in historical research. But few who wrote on HE in the US two generations ago were so trained by their mentors. My work to create the Economists Papers Project at Duke in the mid-1980s, following my beseeching Arjo Klamer to “Go ASK them!”, was not the way to go about this kind of work at that time. I was not then, nor am I now, an economist, so I was quite unconstrained by the strictures which Pete identifies about making a better economics. Students then indeed were encouraged to have a model or theory of the problem situation, which they were encouraged to “test” by the historical work. Was there a Keynesian Revolution a la Kuhn?, was Development Economics an example of a Lakatosian Research Program?, etc.
That is why “Methodology” slopped over into “HET”, since where else could theories about how economics “worked” could be found? I don’t see that kind of stuff being done much anymore, which is another side effect of taking history more seriously.
Since they were too many interesting issues overlapping in the various comments to my previous post on the topic. I have chosen to provide a general response to all the others.
The archives are not a pretext for not doing the historical work. Whether you are interested in history of economics as culture, as economic history or as economic analysis, the archives will never do the job in your place. On the other hand, the archives or more precisely (and I was I am afraid too bold in the last post on this point) primary sources including books, podcast, published correspondence are, I maintain, the only body of evidence on which you can construct your argument. To go back to the Cantillon-Hume(imaginery) example, what I was saying is not that the piece of correspondence gives you an answer to whatever question you might ask about a possible intellectual relation between Cantillon and Hume monetary thinking, but 1. the fact that every interpretation that do not face the existence of the physical evidence of this piece because they did not know of its existence is subject to a revision and 2. the fact that even if your interpretation of the Cantillon-Hume relation is somewhat classical and déjà-vu, the discovery of a new piece of evidence gives it a new twist. So to respond to Beatrice, it can be either of each of the interpretations you mentioned but the fact remains that in any case, the correspondence piece is an unavoidable matter.
I totally agree with Beatrice when she says that sometimes you do not have a lot of secundary literature to begin with an autor, or an issue and the archives is all you’ve got. This was the case with the MSI stuff for example. And I would say that archive are much more than “raw materials”, they tell you personal things about people that are rarely put on the public place through printed matters, such as intents, dreams, projects and reading them is much more than simply classifying evidences, it more often than not brings fresh hypotheses and directions in which one should conduct or redirect one’s future research.
Roy,
I don’t disagree with you that scholarship that does not even concern itself with “making a better economics” is worthy in its own right. In fact, I find this sort of scholarship to be very interesting to read. I personally like intellectual history as an area of scholarship.
But I am ignorant of much of the particulars of different institutional arrangements in France, Germany, etc. with respect to the history of economic thought. I would like to learn more about the “market” for historians of economics in these other countries as compared to the marketplace for researchers/scholars in North America. I have been a visitor to the Max Planck Institute (Jena) as well as the LSE (twice), and as far as I could glean the pecking order in economics was strikingly similar to the pecking order in the US and Canada. I am not making a normative assessment, just a positive description. Perhaps my view is screwed up, or things have changed since I was there, or I am just ignorant of the landscape.
I want to learn from these folks about what is going on in the field. As I said in an earlier post, I was intrigued and impressed at the HES meetings about what I saw — heck I was trying to email Loic and Tiago during the meetings. I have my reservations about this or that, but I am moved to reaction because I think the quality of work is very high. That is independent of my own “passion” for trying to fix economics from a certain perspective. I can compartmentalize and appreciate historical scholarship for historical scholarship sake. And I think this particular post on archival research is very important for the issues it raises about historical scholarship.
It was surely well meaning for ESHET to provide guidelines on thesis writing in HET. (Alessandro where are you? You should know the inside story!) If you follow those rules you will create a piece of work, that looks like history, maybe talks and walks like history but that is just a bit dull. It’s like learning to paint by numbers. A craft, and history is craft work, is learnt by doing and by following (imitating) the masters. That is why the best historical work is not done by 20 year olds but by 50 and 60 year olds. So the intuition that Loic suggests we follow, is something you earn, not something you start with.
Reading the post and comments I was struck by how many times (nine) the term “interpretation” came up (and a few more times insinuated). I don’t known if my job is to interpret economics or even interpret economists. I would argue that we can practice other histories. We can use the record to reconstruct finances, networks and events. There is interpretation whenever there is text, but the archive need not be a store of interpretative traces.
I think of history of economics as an area studies, like sciences studies. In it fit dissenting economics, biography and intellectual histories, but also sociology, social history, political history, media studies, economic history.
I came across this sentence in Cristina Marcuzzo’s paper “Is HET a Serious subject?” in the first issue of the EJPE:
Does it make any sense to oppose archival work and “fact finding” (provided that the latter means anything at all)?