by James Rucker
When you want the answer to a question, where do you start? I, for one, am unashamed to admit that I check Wikipedia first when I’m unfamiliar with the topic. However, when I am familiar, I check more specialized sources. Perhaps I’ll know of a historian who has written about it, or maybe of a relevant news article or or memoir. Either way, we can recognize these as different kinds of information sources. But how are they different, exactly? And how does this inform our reading of them? To answer these questions, today we’ll discuss the central distinction employed by historians, between primary, secondary, and tertiary source documents.
A good rule of thumb for classifying a document as a primary, secondary, or tertiary source is to consider how many steps removed from the topic the source is. Let’s pretend that on June 22, 1902, Teddy Roosevelt rode a unicycle along the streets of New Orleans. Many people saw him do it. If someone were to interview any of these witnesses, that interview would be a primary source (since it is a single step removed from the alleged unicycling). If a historian were to read several such interviews and write an article describing how they think it took place, this would be a secondary source document (since it is two steps removed). If yet another historian reads many such secondary source articles and books, which themselves were written in response to primary sources, and then synthesizes that knowledge into a textbook or encyclopedia entry then we would refer to that as a tertiary source document for the event (since it is three steps removed).
At this point, I want to direct your attention to three titles we have at Linebaugh which capture this distinction: Reporting the Revolutionary War, edited by Todd Andrlik, a selection of primary source newspaper snippets from the revolution; 1776, written by David McCullough, a secondary-source monograph on the Revolutionary War itself; and Encyclopedia of the American Revolution, edited by Harold E. Selesky, a tertiary source.
Right on the cover of Andrlik’s Reporting the Revolutionary War, we have a tagline that seems to say it all: “Before it was History, it was NEWS.” Most people would categorize a newspaper article as a primary source (even though many reporters are not eyewitnesses to events they describe), and the tagline makes the important point that history is usually written retrospectively. Historians know how the events in question turned out, which gives them an advantage when deciding what events are significant. Journalists can be blind at times to whether current affairs will be of any consequence. For example, on August 13th, 1776, the London Chronicle announced the Declaration of Independence, merely saying, “Advice is received that the Congress resolved upon independence the 4th of July; and, it is said, have declared war against Great Britain in form” (Andrlik 197). It is buried in a page of unrelated stories, less than an inch or two from announcements of local marriages and bankruptcies.
On the other hand, a memoir (say of Teddy Roosevelt recalling his unicyclical antics years later) does have the benefit of hindsight, but it still isn’t a secondary source because it is limited to his personal perspective on the event. Taking these two examples together will highlight an important trait of primary sources: They are limited in scope, either by perspective or by time.
Primary sources like these do provide an important advantage to the curious and critical reader: They present a perspective in a more or less pure form. With secondary or tertiary works, you not only have the perspectives of the witnesses, but those of the historian as well, and, of course, with a perspective comes its biases and assumptions. It is important to say that bias does not make a source useless. In fact, once harnessed, an identified bias makes a source incredibly useful. Historians don’t merely want to know what people did, they want to know why. Uncovering a person’s bias can help immensely when pinning down their motivations.
If primary sources provide pure perspectives, why read secondary or tertiary sources at all? Well, because sometimes an eyewitness is mistaken about what happened, and perhaps another witness can provide a more accurate account, or at least cast some much needed doubt on our first interviewee. Cross-referencing different perspectives on the same event is important work not only because it helps to identify who is mistaken about events in question, but also because it can show how the perspectives exist in conflict or consensus. In an ideal universe, someone could consult only primary sources, but there simply is not enough time to read them, let alone learn all the languages necessary. Historians must rely on each other to provide context for the events they study in depth. How do we understand the significance of Roosevelt’s speculative unicycle ride? Was his presidency popular at the time and so his unicycling was done with a light heart? Or did he just declare war and such frivolity was in poor taste?
This is where secondary sources like McCullough’s 1776 come into play. Rather than merely quote primary sources, the historian attempts to synthesize them into a single account. Central to the work of historians is the distinction between testimony and evidence. Testimony is the set of claims a primary source makes, while evidence is any information that can be inferred from the document, even if it is information the source’s testimony disagrees with. As in everyday life, where we can choose not to believe someone because their testimony is too inconsistent, incoherent, or fragmentary, historians can do the same to their sources. Any source can be dishonest or misinformed, so no particular source’s testimony is given absolute authority. Therefore the bulk of the work is in source criticism, where you read “against the grain” and “between the lines” of the text to glean information which it doesn’t wish to divulge. Rules and regulations are an excellent place to see divergence between testimony and evidence. For example, if New Orleans in 1905 passed a law banning unicycle use in the streets, we would not want to believe that unicycles were therefore never used there. Quite the contrary, we would instead have to infer that people rode their unicycles in the street so often that it became a problem, otherwise, such a law would not exist. So while the testimony of the rule is that unicycling through the streets is illegal, punishable, and therefore something the authorities are putting a stop to, the rule is itself a piece of evidence of the opposite. Even though secondary source documents do not typically represent direct experience of what they describe, they do tend to represent the professional authority that comes from having been trained to read texts critically in this way.
As I said at the beginning, I’m unashamed to consult a tertiary source like Wikipedia when approaching a topic which is foreign to me, but I wouldn’t dream of citing one in a paper of my own. Note that this isn’t exclusive to Wikipedia. The problem with Wikipedia isn’t so much that it is edited by the general public; the problem is that it is a tertiary source. No tertiary source is acceptable to cite as an authority in a formal setting. A tertiary source, like an encyclopedia, may frequently be compiled by a professional historian; however, they will tend to have less familiarity with a given topic in the book than another historian would who has focused more exclusively on it. A tertiary source, therefore, is a more general information source. It does not tend to represent cutting edge research, nor can it be expected to provide detailed information.
Now that I have established the difference between these three sorts of sources, I need to qualify it. Whether a document is primary, secondary, or tertiary is not intrinsic to the document itself, but is entirely relative to what questions the historian chooses to ask it. For example, Titus Livius (also known simply as Livy) wrote a massive history of Rome called From the Founding of the City. It is a secondary source (and probably not a very accurate one at that) for studying the early Roman Republic. However, it is just as much a primary source as Virgil’s Aeneid when studying the literary culture in the court of Emperor Augustus. In fact, every secondary source document for a time in the past is a primary source document for the time in which it was written. A book like Livy’s demonstrates that the distinction between primary, secondary, and tertiary is not between exclusive categories, but regions along a spectrum. Some sources are more primary than others, some less. Roosevelt’s memoirs of his fictional frivolities would be more primary than an article in the New Orleans Bee which happens to recount the tale, too. On the other hand, Wikipedia and the World Book are more tertiary than our Encyclopedia of the American Revolution.
Before I conclude, I’d like to take a moment to step back and talk about the purpose of my blog entries more generally. I want to write about how to get the most out of both history books and news articles (primary sources of the present). Hopefully interested readers will find my posts helpful when cultivating their critical reading skills. Of course, the blog is mainly intended to point people to various resources that the library provides, so I intend to inform the readers about various materials and services that RCLS has available. However, I also want my entries to feel like a resource in themselves. I am drafting one more entry right now about the Historical Research Rooms at Linebaugh and Smyrna. After that, each month I will discuss a different types of history writing, exploring the strengths and weaknesses of their methods. Different kinds of historians utilize different sources or utilize the same sources in different ways. They ask different kinds of questions and expect different kinds of answers. After all, historians, just like the people they write about, have motivations. Knowing why they write and what they think is important will hopefully help my readers better understand an author’s argument and be able to evaluate it more thoroughly.