NAVIGATING A PAPER SEA: PRIMARY RESEARCH
The National Archives are huge - and perhaps daunting. Why submerge into that vastness? Why spend hours trawling through box after box of ancient papers, most of them of absolutely no interest? I have been doing that for more than three decades, and I consider it time well spent - and rather enjoyably spent, too. I hope to convince you that similar efforts, though perhaps on a smaller scale, are well worth your time, too. What makes primary research - as opposed to research in published books -- so interesting, or rewarding or, indeed, so much fun? I wish I could do that kind of research more often, and at greater length. A good friend of mine used to talk about researching as a form of refreshment, and I would agree. Nothing else gives you the satisfaction of coming so close to actually knowing what happened - of why and how, for example, that extra set of 40mm guns was installed in June 1944, or how well your ship performed in service, or, for that matter, why she was built or equipped the way she was. And nothing else provides the same kind of surprises, when you find out that what you thought you knew wasn't quite so - or when your eyes are opened to something you didn't imagine had happened (or had existed) at all. We think we have a good idea of the world of, say, the 1940s, but dipping into contemporary documents often shows that our understanding is limited or badly flawed. When you read a written history, you are seeing that world through later eyes. When you read contemporary documents, you begin to see it as those living at the time did. You are moving into what is, for you, a new world.
When you go through libraries of published material you discover that a lot of what you wanted to know either isn't there at all, or is there in contradictory versions. You generally don't know where the author found the more interesting things he published, or, unfortunately, what he invented or assumed. I found that out by about 1970. Existing reference books were impressive, but I wanted a lot more. Knowing that I would be visiting Washington in the spring of 1970, I wrote to the National Archives in the naive expectation that it had a set of formal books, which I thought should have been produced for ships' commanders, summarizing their characteristics - for example, their schemes of protection (with expected immune ranges) and their speed and maneuverability. In return I received a letter from Harry Schwartz, then head of the Navy and Old Army section. Nothing that formal existed. However, if I visited, he would be glad to show me what there was. I had no idea that this brief visit would be the beginning of a lengthy research career. Above all, I came to want to know why things had happened - why ships and their weapons and other equipment had been conceived and designed as they had been, and why and how they had been modified. What larger considerations were involved? And, overshadowing such questions, what ideas had gone almost all the way to becoming ships or weapons, but had not quite made it? What would the various cancelled ships and armament and other equipment have been like? I already knew a bit about some of them, from what had been published, but I wanted more. I still do, more than thirty years later.
I feel particularly fortunate to have seen considerable material produced by the preliminary designers who were responsible for the main features of most ships. Among the materials Mr. Schwartz showed me were sets of Research Memoranda produced by the designers for their own use - things like notes on why a design had turned out as it had, or on a new scheme of protection, or on the U.S. Navy's interest in gyro-stabilized ships, for example. Such memoranda go beyond particular ships to suggest wider concerns within the design community. They also gave an idea of just how designers approached their projects.
As you research, you gain a sense of what I think of as the overall shape of the collections of documents you use - you get to recognize what you should expect, and what is extraordinary. For example, I have seen numerous sets of preliminary design books, which are associated with particular designs. Typically there is a book of relevant written documents (such as the initial request for the design) plus other books of calculations (including one book devoted entirely to calculations of survivability after damage). The first book generally includes one or more 'spring styles' prepared to give some idea of what the ship would be like. If you did not expect the particular sorts of calculations involved, you might imagine that there was something special about having a separate book on how a ship would survive battle damage. If the full set of data were not included, you would wonder how serious the design project was. Going through masses of preliminary designs, you naturally try to guess which ones were important (likely to become real ships) and which ones were really intended to show that some clever idea was actually pointless. I remember a 1956 study of a protected missile frigate (DLGN, not FFG) which was clearly intended to show that someone's request for it was ludicrous.
You also learn that practices evolved over time. The set of preliminary design books Mr. Schwartz showed me, which ended about 1922, generally provided only one volume per design - but provided multiple volumes for different versions of what ended up as the same ship. It later turned out that different series of these books were in very different places, both in the Archives and under continuing Navy control. Many such books have recently been released at College Park. Others have disappeared; unfortunately no one seems to have maintained a consistent list of preliminary designs.
Reading book after book teaches you how ships were conceived. When I began, I spent many happy hours in the General Board files, which included the formal Characteristics (staff requirements) to which ships were designed. The natural conclusion was that the Characteristics came first, the designers conforming to what was asked of them. However, as you read book after book of design documents, you find that the process went the other way. Characteristics were actually a summary of the features of the design developed to meet earlier informal requirements. You know that because the book includes both sets of details, as well, often, as notes of back-and-forth negotiation of design characteristics. This perception, that successful design is a negotiated process, is something I have learned from archival material, not from any formal write-up. It carries lessons relevant to the current defense procurement process.
Material from the files of the General Board often illuminates what happened, for designs executed between about 1910 and 1940. But there are still surprises, and often they make you realize what has been lost when documents were destroyed. Not long ago I happened upon a design history of the Alaska class, ending some time in 1940. The ship described had two triple turrets and one twin, because the hull forward was relatively narrow, and it seemed a triple would not fit in No. 1 position. The file included a plea from the shipyard (New York Shipbuilding) to adopt three triples for simplicity of design - and the adamant refusal by the design chief, then-Captain Cochrane (later Bureau chief). At that point the ship had a low bridge like that of a Baltimore or Cleveland class cruiser. Whatever transformed the ship into what we know as the Alaska came later, during Contract Design - but the records involved seem to have been destroyed. We know from a very different file that New York Ship was allowed to revise the hull lines of the Cleveland class, but we don't seem to know who redesigned the Alaska - and who approved that (the answer may be in the main BuShips correspondence files; I haven't looked).
When you see documents in the archives, you are looking at the navy and its ships from the navy's point of view, not from the point of view of a researcher with specific interests or specific questions in mind. You are forced to see things in that more authentic way, because the documents you see fit together that way. For example, when you gain access to box after box of ship design documents, you are looking at the navy in an integrated way, class alongside class, often not at a specific class or a particular ship. Even when you think you are looking at the documentary record of, say, one heavy cruiser, you are seeing lessons from the last few classes and notes on what might be better in the next design. You are also almost inevitably seeing how different parts of the navy interacted, and you realize that points of view within the navy were different. The ordnance officers from the Bureau of Ordnance had very different priorities from, say, the machinery men of the Bureau of Engineering and the hull design and overall design men of Construction and Repair. When a ship was in one stage of design (say, Contract Design, the stage at which plans were produced for bidders) another ship was being built, and another was being conceived by the Preliminary Design group. You get a sense that ship design was a stream of continuous work, from which particular designs were selected as needed. If you are really lucky, you also get a sense of the design projects which did not become real ships, because they illuminate what was done.
You get something more, a sense of contingency, of how people made choices at the time they made them. In retrospect, almost everything seems inevitable, and a lot of history is written that way. But when you read what people wrote at the time, you become aware that events, from the way a ship might be equipped to the way a war might be fought, were anything but obvious at the time. I don't know of any other way really to understand and appreciate that sort of uncertainty. You can read about how people felt, for example as World War II loomed closer and closer, but if you read a mass of documents (the right ones, of course) you come to appreciate the crucial assumptions they made. I was once shocked to read, in a General Board paper written in the early 1930s, that in rejecting British treaty proposals (for smaller battleships) U.S. officers casually pointed out that if the British did not like the U.S. views, the United States could and would simply outbuild them. That seems obvious in retrospect, but at the time I had absorbed the British bias of much existing naval literature, and imagined the United Kingdom to be the greatest shipbuilding country on earth (it was, but we had greater industrial potential). A U.S. Navy naval architect I know once told me that he liked my presentations because they carried a sense of the surprises experienced by people at the time - they were not simply about how X had to develop into Y, because my reading of the record showed that would not necessarily have happened - often the line between two known historical points turned out to be anything but straight.
Exactly what you see may seem to make sense, but you find that you want to know more about context. Exactly what did that memo mean when it referred to available funds? Why was there suddenly interest in, say, bullet-proof plating one year and not before? What was that particular lesson derived from, say, the 1935 Fleet Problem? How did this particular kind of ship figure in the navy's idea of what a future war would be like? What was that idea in the first place? The more you look, the more interesting the questions become, and the deeper your knowledge becomes.
The best thing is the surprises. It is one thing to know for sure that, say, a particular battleship had 12.1 inch belt armor, but that is hardly a great surprise. It is quite another thing to find, as I did recently, that the British came close to dropping the treaty limit on cruisers in 1938 (i.e., to building something in the Alaska class well before the outbreak of war). Just about every book I have written has brought me research surprises, things I never would have suspected. For example, one of the big surprises in U.S. files was that the categories defined by the U.S. Navy were not nearly so clear-cut as I might have imagined. For example, when I was working on amphibious ships, I thought there was a clear distinction between point-to-point transports (AP) and attack transports (APA). It turned out that several ships were completed as APs and then used exclusively as APAs. In one case, the ship's prospective CO, observing that his ship was being equipped to carry numerous landing craft, wrote OpNav asking for the more powerful air defense he thought he needed. No, he was told, don't worry, you'll never be in a landing. When his ship was completed, she served only in amphibious landings, filling out a division of attack transports, and being used the same way. I'm still not sure what bureaucratic or fiscal rule demanded the gross mis-designation of that particular ship, but it probably originated in negotiations fixing the numbers of merchant hulls taken over by the navy for particular purposes, probably agreed at the Joint Chiefs of Staff level.
For me, all this began with that visit to Washington in the spring of 1970. Mr. Schwartz, who has long since been retired, showed me the book listing and very briefly describing the contents of Record Group 19, the series of documents the Archives then had from the Bureau of Construction and Repair (C&R), predecessor (along with the Bureau of [Steam] Engineering) to the Bureau of Ships which produced the World War II and postwar fleets (it in turn was succeeded in 1966 by the current Naval Sea Systems Command). Mr. Schwartz took me into the stacks, where I saw hundreds of boxes of 'history' files, correspondence arranged ship by ship (in this case for the period 1927 - 1940), and by subject code within the series for each ship. He also showed me a box or two from one of the hundreds of entries in the C&R book, which contained books of records concerning preliminary designs, often including potted design histories. I had had no idea of how complex the files were, or of how much there was (and a great deal had been destroyed over the years). Later I discovered the printed documents produced for the fleet, such as the General Information Books, which were envisaged, in effect, as operating manuals for ships. Much later I discovered the files of the Bureau of Ordnance, which I am still working through (much of which has only recently been released to researchers). And of course a great deal of fresh material has been released over the years. This mass explains why the National Archives building in downtown Washington, the one I visited to meet Mr. Schwartz, has now been joined by a much larger building (Archives II) in College Park, Maryland - which is now close to overflowing. Of course most of what is stored in both buildings has nothing to do with ships, or with the U.S. Navy, but even the naval material is massive. And that is not to mention material under U.S. Navy control, for example at the Washington Navy Yard.
What I really remember from that first trip was how much I regretted that I could not spend a lot more time rummaging through the Archives, because it seemed that everywhere I turned I was being surprised. I also much regretted that I could not easily reproduce and bring home a lot more material, even at a dime a page for reproduction. I was very lucky that my professional life brought me frequently to Washington, and hence to the Archives, though unfortunately not frequently enough. I was fortunate enough to spend considerable time in the complementary files of the Operational Archives at the Washington Navy Yard and elsewhere (recently, on an extended basis, in the archives of the Naval War College). I learned that I could get the best results by combining what I learned in different archives. Doubtless others will have different combinations of archives in mind, but the point is that once you have the habit of research, you keep looking for additional places to look. For example, the close relationship between the U.S. and Royal Navies means that there is a lot of British material in College Park and a lot of U.S. material in England. The surprise is that often what should be in one place is actually in the other. For example, College Park has a great deal of British ordnance and radar material, including manuals, which cannot be found in the main British collections. On the other hand, if you are interested in the postwar U.S. Navy, it is very illuminating to read British reports of information-gathering missions to the United States. It was equally illuminating to read British accounts of U.S. work on amphibious warfare during World War II. Much more recently, I have been as lucky that new technology, in the form of a digital camera, makes it possible to bring home far more material than I could have imagined in the past, in a more accessible form.
You start with some specific questions. In 1970 a lot had not yet been published, and a lot of what is now easily available was still classified. Exactly what, for example, was the armor arrangement on board a particular battleship? What was it supposed to protect against? How fast were various ships when they ran trials? Your first discovery is that the archives (indeed, all archives) are arranged not around the questions you have but rather around the documents which have survived, which were created for other purposes. In American archives, that often means arrangement by the office which ended up sending its files to the archives, which was not always the office which created the material in the first place. Foreign archives are often arranged more specifically by document. From your point of view, the chance that something surprising is buried in a U.S. archive is much greater.
Your first job is to figure out what sets of documents are likely to give you the answers you want. That often involves wading through a lot of false leads. You may find the files you want in surprising places. For example, it turns out that the trials reports, which do exist, survived because they had to be submitted to the legal arm of the navy (the Judge Advocate General or JAG) in order to justify payment (which is probably why this file does not seem to include trials reports for ships built at navy yards). There may have been copies in the Bureau of Ships files, but I never found them. I found out by accident that they were in the JAG files: someone writing in Warship International had cleverly used this source for USS Atlanta, discovering in the process that all the talk about her 40 knot speed was rubbish. It turned out, too, that trials reports were an excellent source of descriptions of ships' machinery, including their generators. I'd never have guessed that in advance, but I was glad to find out. In this process, incidentally, you become very appreciative of authors who actually did documentary research and therefore provided references to files you would like to search.
As you go through files, you learn something more. You begin to understand the process which produced the ship in which you are interested. You can't help it. Much of what you go through is correspondence files. If, for example, you want to know when a battleship was modified during World War II, you almost inevitably begin with the Bureau of Ships files on battleship modifications. When you read the message ordering, say, the addition of a quadruple 40mm gun, you see where it came from and who had to be informed - and there are sometimes surprises. You may read a cover slip explaining that the ship asked for three such mounts, but that stability precluded more than one, or there may have been permission to add more when they became available - and you had not thought before about bottlenecks in production. In some cases lighter weapons were installed with the understanding that something specific would replace them at the next refit, or when it became available. You may find out what had to be removed to compensate for that extra mount. You may also become curious as to exactly why the ship was in the yard when she was - was that scheduled, or was there some emergency? But this is a file ordering things, and approving them. What actually happened? Maybe you are lucky enough to have a photograph taken just after the ship was refitted, with those little circles showing changes. If not, you start trying to get reports of what was actually done at shipyards (those files do exist, at least in large part). You probably also find that many official files are misleading, because whoever put them together made mistakes. I am thinking here of the impressive Bureau of Ordnance Armament Summaries, which I have often used. Recently on the 'Steel Navy' website Rick E. Davis highlighted the problem, when he began to look into which Fletcher class destroyers had received the full 'anti-Kamikaze' armament upgrade. His list, based on photographs, was rather different from the one I had compiled from the Armament Summaries. I have no doubt that his was the correct one.
In effect Mr. Davis was showing that the archives, and primary material, is more than written. Photographs are in themselves archival sources, and often attention to them can both make up for gaps in the written record and can suggest further lines of inquiry. You would not, for example, wonder about the origin of the foremast platform that distinguishes USS North Carolina from USS Washington by reading BuShips files - but the difference leaps out at you from a photograph. As Chris Wright has demonstrated in Warship International, once you understood that there was a question, you could often dig up an answer in the endless memoranda in the files. Doing so, of course, would be possible only once you knew what sorts of files to search through.
A lot of the time you have more questions than answers, partly because a great deal of the paper you want has long since gone missing. That is generally not for any sinister reason. Offices are often sloppy. Sometimes a paper is not in the files when they are boxed up to go to the archives. At other times someone suddenly decides that files of a given type are expendable, even though much the same sort of files were preserved by someone else. That is sometimes heartbreaking. Many preliminary design files for significant ships appear to have been destroyed. Once I was researching in current files in a Navy office, and discovered to my horror that they generally threw out older files (some of which I badly wanted) when they ran out of filing cabinet space. On the other hand, you learn that large bureaucratic machines like the U.S. Navy produce an incredible amount of duplicate paper, and a lot of it survives - or at least a lot of very helpful references often do. The gaps hurt, and they complicate the historical picture, but that's life - it helps make research interesting and creative. The sad part is that the further back you go, the less paper was generated. The pre-typewriter era, where I try not to go, must have been the worst; copies were made by clerks, and there were not so many of them.
As a researcher, I am happiest when I am surprised, and archives files often give me that pleasure. When I was working on a book on U.S. submarines through 1945, I thought that the story of the earliest ones was quite straightforward. However, something bothered me, and at literally the last minute I looked at one more file (Bureau of Navigation, pre-1910). It contained material which I had not even suspected, and I had to rewrite an important part of the book at a rather late stage. Going through this file in turn changed my understanding of the structure of the navy of that era. Later writers had emphasized the revolution wrought early in the Twentieth Century when the General Board and then the Office of the Chief of Naval Operations (OpNav) were created. Later BuNav was actually the navy's personnel office. The papers I was reading from 1903-1908 showed that it was much more a predecessor to OpNav than I had imagined, often a clearing house for the fleet officers to ask for changes in their ships and in ongoing designs. I'd never have guessed, had I not gone dredging in the files.
I was once looking at something (I forget what) and it mentioned the embarrassment of structural weakness demonstrated in an accident involving the bow of a Brooklyn Class cruiser. Naturally I had never heard of that, but within the ship design and construction community, it had been notorious - and I was able to find a prewar file which described the accident and included photos.
A lot more than ship design files has survived. During World War II, for example, every ship filed a war diary and after-action reports, some of them very illuminating. It was the excerpts from after-action reports in Bureau of Ships files which showed that some of the ships classified as point-to-point transports served as attack transports, in squadrons of attack transports, late in World War II. Many files include excerpts which were clearly the basis of later decisions to modify ships. For example, files on the small LCI(L), a personnel transport, include lessons learned from amphibious operations which helped justify modifying such craft as small gunboats (one surprise was that the modifications, which might look like forward-area improvisation, were actually standardized and were carried out by small shipyards in California).
The Bureau of Ships produced detailed War Damage reports, classified for many years but now released, which give an excellent idea of what combat was like. Cominch (Admiral King's office) produced detailed lessons-learned papers on every major Pacific action and on all the amphibious assaults, including those in European waters. They often included comments (sometimes stinging) on ships' reports. The series ended with a 1945 'message to an Ensign in 1955' summarizing war experience as then understood. The Pacific Fleet produced four large reports on lessons learned concerning materiel, including not only ships and aircraft but also ordnance and radar. It turned out that other files provided background on these reports. For example, the OpNav Underwater Warfare Division reports at the Navy Yard explained what the submarine portions of the lessons-learned report meant, and by extension helped explain the initial postwar U.S. submarine development programs.
Many other kinds of files in the archives provide remarkable insights. For example, when I was at College Park a few years ago, Richard Franks happened upon a manuscript internal history of the prewar U.S. Navy signals intelligence organization. He kindly gave me the reference. The history just happened to show that the U.S. Navy changed its strategy for war against Japan after its analysis of the 'take' from the 1930 Japanese maneuvers showed quite clearly that the Japanese understood, and could counter, the existing U.S. war plan. This intelligence was why the then Chief of Naval Operations told President Hoover that he could not effectively threaten the Japanese when they invaded Manchuria. I think it is also fair to conclude that the navy suddenly became aware of how insecure its own communications practices were. This is not the sort of thing which found its way into less classified histories, but it made a tremendous difference. Several intelligence files at College Park analyzed the effect of radio intelligence on the German U-boat campaign. They provided significant insights into the compromise of the German cypher system which I had never seen anywhere else. In the process, they gave me a much better idea of what it meant to use code-breaking information against the U-boats. The fact that some of the files were written by Royal Navy historians attests to the closeness of cooperation between the U.S. and Royal Navies at the time.
U.S. files include attache reports and intelligence files. Many of those assembled before about 1940 were remarkably complete, because so many navies regarded the United States as harmlessly neutral. For example, in the mid-1930s a U.S. officer visited the factory of the Italian Officine Galileo company, which made fire control computers. They told him, he said, everything he could have wanted to know - but unfortunately he was not a gunnery specialist, and most of it went over his head. He did retain their brochures (which I found in a different file), and he mentioned some points which had stuck. One of them was that the Italians had presented the Germans with two complete sets of fire control equipment. The Germans had clearly been finding it difficult to build modern fire control computers, and photographs and descriptions of the equipment on board the Bismarck suggest kinship to the Italian computer, good descriptions of which can be found elsewhere. Other sources from other archives make it clear that the Italian system in turn was derived from a British system produced by Barr & Stroud, the rangefinder makers (and rejected by the British Admiralty in favor of an Admiralty device). Taken separately, none of this is particularly shocking. Taken together, you get 'HMS Hood was sunk by a German fire control system derived from a British one.' When the Admiralty not only rejected Barr & Stroud but also cleared the firm for exports, it was setting up a spectacular own goal two decades later. Incidentally, U.S. sources also show that all World War II Japanese fire control computers were derived from Barr & Stroud's, so there were actually multiple own goals.
The U.S. intelligence files include many comments on what smaller navies were doing, and on how well the ships they were buying performed. If you are curious about the international naval arms market (as I am), they are not a bad starting point. I certainly found them useful when writing about British firms trying to export cruisers. Again, none of the files has the complete story - it has what a smart U.S. naval officer talking relatively freely with local naval officers was able to uncover. Sometimes he did better, sometimes worse - and some countries were of greater interest than others. But the files in question add enormously to what we know, particularly for navies whose historical files are anything but easy to obtain.
There are also extensive files of the printed documents the U.S. Navy produced, things like handbooks and collections of tactical doctrine and lessons-learned summaries of recent operations (including the Fleet Problems). When I started, such things were quite rare. However, in recent years a classified OpNav library has surfaced at College Park (now declassified), and many Bureau of Ordnance publications are now available. They tell you, for example, how to operate and how to service a radar or a gun mount or a missile system of the 1960s - many of which are aboard HNSA ships. Again, the collection is incomplete, but it is a lot better than anything available a decade ago. Again, reading these handbooks, you get a sense of context and you can imagine further lines of inquiry. At the least, it is wonderful to be able to see exactly what much of the equipment and weaponry of the fleet, from the 1920s onward, was like. You'd be surprised about some of the questions this sort of material raises, because you begin to understand both the potential of the equipment and its limitations. Both must have impacted tactics, but the connections were so well understood at the time that you have to draw many of the connections yourself - no official writer would have seen any need to do so. The collection at College Park includes many publications on fleet tactics, from the 1920s on (some relevant material is also in the Navy Department Library at the Washington Navy Yard). Incidentally, the archive at the Naval War College is a wonderful repository for papers on U.S. Navy tactical thinking through the interwar period.
All of this is aside from the pleasure you should get by connecting directly to those who created and operated the ships you now preserve. There is something special about holding a file which the CO of your ship wrote, describing what he did in his own words. Can you, or should you, devote two or three decades to burrowing in the various archives? Not unless you have something larger in mind. But even a few days can be very rewarding. You owe that pleasure to yourselves.
I should add some practical tips. The National Archives issues you a research (identification) card, good for a year, based on any standard kind of identification you present. Their view is that they are there to connect the public with its documents, to help as much as possible. In return, they want you to be careful not to damage what they have, so that it remains available to future generations of researchers. That means, for example, that you cannot use a pen, and that they try to make sure that everyone in a research room follows their reasonable rules. Too, they can only provide a limited number of documents at a time - in the main archives, four or five 'pulls' per day. Each 'pull' can be up to a trolley-load of documents (up to 24 boxes), all from the same group of records. That is not bad, but you can be disappointed if a whole trolley load turns out to be useless (I have had that experience, and I'm sure it is not unique). So you begin by spending time trying to guess in what records you will find what you want. That is not a bad way to discover the remarkable range of documents involved. College Park has, in addition to its documentary archive, separate photo, microfilm, and cartographic (including ship plans) rooms, and you can pull files from all three more or less simultaneously (much of the microfilm is on open shelves, so the pull times are irrelevant there). Finding aids (descriptions of documents) vary considerably in quality.
Your goal in an archive is to find documentary information and to extract information from it in some usable way. That generally means copying what you can find, to at least some extent. At one time that meant notes on index cards. The National Archives still provides cards (you provide the pencil). Most of us cannot hope to spend so much time in the Archives that such techniques will gain us much. We are therefore reduced to finding relevant documents and then trying to reproduce them so that we can think about them later on. For many years that meant photocopying or, in extreme cases, ordering masses of microfilm. Now there is something much better in the form of digital cameras. In effect, a reasonably good digital camera is a hand-held Xerox machine, capable of taking up to 2500 or even 3000 shots on a good day (good meaning that you come upon that many items you want). 'Reasonably good' means that it is auto-stabilized, so that you don't need a tripod. It also means auto-focussed (otherwise your eyes would soon become useless). I would recommend a model whose viewing screen can fold out, so that you can hold the camera pointing down at the document while you are sitting, looking across at the screen. You have to look because you want to make sure that the camera sees the whole document you are shooting. The alternative, in which you have to look down through the camera, will preclude rapid photography, and jumping up and down will ruin your back (I have seen many researchers doing that, and I always wonder whether they are getting much out of the time they are spending). You have to be able to disable any flash the camera has (Archives reasonably understand that flashes ruin their documents), but I have not had much trouble shooting in the available light. Cameras often allow you to set the resolution of the photos you take, and you will want to experiment to see what resolution lets you read the results while packing as many shots as possible onto a digital card (I get about a day's worth of shots on one card).
Another point is worth making. If you are taking notes, or copying only a very few documents, a lot depends on whether you fully understand what you are seeing. Otherwise what may seem to be unimportant pages or items may assume considerable significance late in the work, when you cannot hope to go back to see what you have missed. Since you are learning from what you read, it is most unlikely that you will enjoy that sort of certainty. Often you see something in a later document that illuminates what you should have seen in the one you had last week, or lest month, or last year. The less the effort involved in copying further pages, the better the chance that you will collect entire documents, and that you will be able to go back (using what you already have) to see what you later realize mattered greatly. That has certainly been my own experience. Being able to go back as many times as I like has paid off enormously. I have also found that the ease of photography encourages me to copy documents which seem interesting but perhaps secondary, only to find much later that they were invaluable.
For example, I was once in the Royal Navy historical archive, and I came upon a booklet of war game rules dated July 1913. I casually photographed it. Later I realized that the war game rules reflected what the Royal Navy thought its guns could do, hence could contribute deeply to my understanding of contemporary British tactical thinking. I would certainly not have made the effort, in an earlier time, to photocopy such a booklet, because I would have been looking for much more significant papers to copy.
I find that a single camera memory card is good for about 2600 shots at the settings I use, and I run through one or two sets of batteries a day. Compare the cost of that to the ten cents or more that you have always spent on Xeroxing, and you see the difference. If you add in the difference in time spent the improvement is spectacular - I doubt that anyone using the copier takes home more than a few hundred pages of copies each day. I know that history is about quality, not quantity - but, as the Soviets used to say, sometimes quantity has a quality all its own, and in recent years I have certainly felt the improvement in my own work due to easier access to a larger mass of material.
I'd add that you can photograph documents you can't photocopy, for example when they are bound into large books, such as ledgers. In recent years I have done a lot of research in the Brass Foundry near London, where the British keep the 'Covers' in which design and construction data are collected. The 'Covers' are essentially large scrapbooks, into which documents were pasted by fitting them into holders in the large book. I cannot imagine being allowed to flip over a Cover in order to photocopy it. Photography has been relatively easy. I should add that I bought my camera when I began work on a book on naval fire control. Chris Wright warned me that the Archives would not allow me to photocopy the tightly-bound books of annual U.S. exercises, and that the scanners at College Park were unreliable, 'so bring your camera.' I bought my first digital camera for that job, discovering in the process that it was far less expensive than I had imagined. The camera proved so valuable that within a year I had worn it out - literally. Its shutter button stopped working, while I was at College Park. Now I always carry two cameras, although, fortunately, I have not had a similar failure.
Moreover, the photo card preserves the order in which you have shot, whereas the pages are easy to mix up. I also find it a lot easier to organize a lot of documents inside a computer than to keep order among thousands of physical pages. Even microfilm is more difficult to use than computer images, although it tends to be more readable. On the other hand, you can't quite pencil into the bottom of a photograph where you got it (although you might put a card with that information in the picture). I never print the documents; given my technique, there are far too many. Instead, I read them off the computer screen (I bought a larger screen to make that easier). It is relatively easy to split the screen so that I can read a document on one side while typing opposite.
There is, as usual, a down side. It is difficult to impossible to hold a camera so that the image it captures is lined up exactly; what you get is distorted. If for some reason you badly want a precise copy of a drawing or a photograph, you are in the land either of copy-stands (the National Archives has some) or of scanners. In either case your shooting rate will be a fraction of the rate I have cited, but then again you are looking for something reproducible rather than just for information. I have used both methods, and both work, sometimes spectacularly well. Their only drawback, besides slow speed, is the mass of stuff you have to carry around - but that is slowly declining as electronics improves.
I hope that I have convinced you: go out and enjoy something that is open to you, but quite possibly untouched for too long. I know that some of you already use the archives extensively, because you have posted numerous documents on the HNSA web-site. I hope that many more of you do the same, not just for old naval publications, but for the raw material of naval history - the tons of documents in the National Archives and similar collections. I'd add one thing. You may imagine that somehow it takes a degree in history to appreciate what is there. I don't know how many of those who use the National Archives are academic historians (I'd say professional, but I count myself a professional naval historian but hardly an academic one). I do know that the British found, to their surprise, that the most interested and committed researchers in their own archives were interested amateurs, who wanted to know, rather than simply to complete their academic projects. Knowing is what this is about.