Reading Primary Sources - Dr Robert A. Hatch - How to read primary texts - Using primary history documents - primary sources

University of Florida

Primary sources raise questions, perhaps first among them: What is a primary source? Other questions are less obvious. They represent a spectrum of concerns that range from critical reading skills to the nature and limits of historical knowledge. These issues are discussed below. The purpose of this brief introduction to Primary Sources is to provide useful working definitions about historical sources and practical tips for reading primary documents. This short essay ends with selected readings from various historians on the importance and use of primary sources.

Primary Sources

From the outset it must be admitted that the common distinction between primary and secondary sources is problematic -- however useful, it may be entirely artificial. By tradition, a primary source represents firsthand testimony or direct evidence concerning a given topic. Primary sources are often associated with a particular event, with those who participated or witnessed an event, or those who directly authored thoughts on a given topic, whether as a philosopher, scientist, fiction writer, journalist, or someone writing a letter or diary. Here historians are concerned about several issues, most importantly, the value of the source. This negotiable quality often involves authenticity, credibility, usefulness and the authority, reliability, and status of the author. Some sources are valuable due to dumb luck; by chance the document survives, by chance the author scribbled about an expected or little-known event.

Because 'the value and usefulness of a source' is open to question (as discussed more fully below) it is important to discuss the various categories and traditions associated with documents themselves. By tradition, different kinds of documents have carried different degrees of value and prestige, for example, historians have different views of the relative merits of handwritten manuscripts (unpublished diaries; journals; research notebooks; letters; notary, baptismal, legal, and other handwritten records such as census reports) and contemporary printed, often synonymous with published, documents (printed materials ranging from broadsides, leaflets, and newspapers to monographs and books).

The status of these documents, and their relative positions in various hierarchies, have changed over the last several centuries. Take a few minutes to jot down as many specific kinds of 'primary sources' as you can. Compare your list with the list at the very bottom of this page. Clearly 'primary sources' have changed over time. With these changes come new challenges in interpretation. But before approaching that problem, it is useful to provide some simple distinctions concerning these categories.

Primary sources fall into several categories and physical formats which sometimes overlap. The vast majority of printed or published texts include books and monographs. By current definition, a book is composed of leaves of paper, parchment, or other materials joined together (sewn, bolted, glued, etc.) whether it is printed, published, handwritten, or blank. A more recent invention, the monograph is traditionally described as a work written on a single topic (hence mono-graph) that is systematic. In practice, at least during this century, monographs are often shorter and sometimes more specialized than books. The reader can now see the useful limits of defining these objects. Specific examples (associated with a given time and place) are often more useful than abstract definitions.

Since the seventeenth century, there has been a marked increase in the number and kinds of other printed and published sources now commonly known as serials. Examples of serials include newspapers, periodicals, scholarly journals, and magazines. Generically, as the term suggests, serials are published or circulated on a regular time schedule. The 'serial' component suggests that publication commenced at a given time and appeared 'periodically' at a regular 'daily' (jour-nal), weekly, monthly, or quarterly (4 x year) interval.

Good examples for our purposes include the Philosophical Transactions of the Royal Society of London and the Journal des Sçavans. But we should be careful about our assumptions. Newspapers may have begun as broadsides or single sheet printings designed for the general reader (some seem to have been intended to be read aloud for the non-literate). From the outset, Henry Oldenburg published the Philosophical Transactions privately. We know little about how many copies of the PT were printed, much less where they were sent or who read them. Indeed, it is possible to ask if they were widely read at all, much less widely understood. In France, the Journal des Sçavans appeared irregularly during its first decades, and it appears to have had an identity crisis regarding its audience. The editors also seem to have been at odds with the journal's sponsors. As always, context is important, and readers of primary texts should be cautious about transporting contemporary assumptions into the past.

Manuscripts are often described as 'unique documents' which may be either handwritten, duplicated by a mechanical tracing device, typed, or otherwise electronically reproduced. Since the 19th century definitions have changed concerning manuscripts. For example, the term 'manuscript' has changed as it applies to individual written works, personal papers, correspondence, archives, and so forth. Today the term generally applies to a body of records, personal papers, or to a collection of manuscripts conserved by an individual or institution other than the original author.

Why Read Primary Sources?

Some historians have argued that primary sources provide important insight not only into the past but into the writing of history itself. Primary sources, it is argued, make history come alive, they immerse the reader in a network of issues and a variety of perspectives rather than a single vision or secondary summary. At their best, primary sources challenge conventional interpretations by presenting not only the full complexity of issues but sometimes a series of stark contradictions. Primary sources problematize historical interpretations.

More troublesome, if primary readings represent a puzzle, the puzzle inevitably has missing pieces. The challenge then becomes two-fold. The reader must not only attempt to understand the text as the author intended it (if this is humanly possible) but must also make sense of the text despite lost or limited information. To these difficulties, there is the additional problem of 'hidden' or unsuspected meanings. The 'meaning(s)' of a text inevitably invites discussion of context and sub-text.

Arguably, the most perplexing problem with primary sources is that the 'meaning' of a text is not black and white. It is difficult to imagine a text that is not open to multiple interpretations and hence scholarly negotiation. What an author writes is seldom transparent or a simple matter of historical documentation or textual fact. Here historical texts not only raise problems with our assumptions about 'objectivity' and 'honesty' but the very possibility of establishing a single defensible 'reading' of the text. To be sure, experience and rigorous research can help distinguish between a 'good reading' and others that are less defensible, appropriate, or satisfying.

But it often happens that the most important and telling elements of a primary text are absent, marginal, or easily glossed over. Interpretation is inevitable and often a dicey affair. Interpretation prompts the reader to come to grips with the text as well as the broader intellectual, cultural, and social contexts. And the task does not end there. Understanding the context (the surrounding circumstances) of a text often directs the reader to the difficult question of the 'subtext' of the author, that is, to the author's reasons, purposes, goals, motives, and, dare we say it, the author's intent. Hence: Text, Context, Subtext. Each becomes an inter-textual component of any sophisticated and nuanced reading.

The critical yet sympathetic reader approaching an historical text is thus drawn into the historical context (the broader contemporary circumstances) as well as the author's subtext (intellectual assumptions and personal motives). There is ample room for debate about what historians can know about such matters. Scholars disagree about what reading strategies are appropriate and what value there is to such an enterprise.

Enthusiasts claim that primary documents invite the reader to join a debate, to appreciate different perspectives, to interpret things from different points of view. Because human affairs are complex, so the story goes, primary sources prompt the reader to reconsider the difficulties of interpretation and the pitfalls of inference and generalization. Some historians imagine that primary sources offer an opportunity to reflect on how we read and write 'history', they argue that primary sources insist on a dialogue between the past and present. Other historians counter with the more modest claim that interrogating historical texts sharpens our awareness about the negotiable limits (or perhaps the very possibility) of historical knowledge.

Whatever theorists may claim, at a practical level of historical research primary texts require strong reading skills. Reading, of course, is a ritual for thinking, it is an invitation to interpret, to exercise critical and creative thought. Strong reading skills balance critical analysis and imaginative insight. Historians aim to justify their interpretations by displaying detailed (hardheaded) analysis and sensitive (often empathetic) appreciation of cultural difference. Somewhere between there may be a path connecting 'past and present.'

Practical Concerns in Reading Primary Sources

As we have noted, critical reading skills are important in coming to grips with primary sources. Arguably, one text is just like another (at least in Pan-Textual-Land, where an author's intent might count for little). But historians are generally less interested in theoretical constraints than making sense of documents. This, of course, may be a completely naive enterprise. Without question, it is risky business. Making sense of a document means making assumptions and risking unfounded inference. There is ample room for error. But historians have few alternatives. So we proceed with caution and humility -- and perhaps a hint of optimism.

Much of what historians would most like to know about the past is forever lost. But we are often at a loss for other reasons. Because the author imagines a different audience, critical assumptions are habitually left unstated. Important contemporary ideas and beliefs (perhaps unknown or unsuspected by later readers) are deemed too obvious for mention. Arguably, the most commonplace notions are thoughtfully omitted as redundant, too well-known to identity for a contemporary audience. For the historian, what is 'absent' is often most telling. For the author, contextual omissions lighten the load, they limit the banal while offering nuance for those in the know. For the historian, primary historical texts invite discussion of things absent. Readers must be careful about what they "bring to the text" in their interpretation. Understanding the historical context is critical. Readers must aim to see the world as the author and the audience saw it, how it was, how they wished it might be.

Like readers of any text, readers of historical texts wisely consider the author's assumptions and personal motives. The assumption here is that writers have a subtext, a web of ideas and beliefs, stated or unstated, rational or not. To complicate matters, writers themselves are often unaware of the full measure of their subtexts, their deepest beliefs are often not fully articulated or easily verbalized. This claim enters into difficult territory. But the present author would wager that most writers are hard pressed to provided a detailed map of their reasons and motives for writing. The more so for writing to their assumed audience.

If the above outline makes sense, the challenge for readers of primary historical texts is to unravel (and then weave back together) the text, context, and subtext. At risk of simplification, the text "then and there" is somehow not the same text confronted "here and now." The first difference, we might suppose, is between the "past" and "history" -- arguably a signature problem with events and texts. In sum, and without apology, a single sentence: The historical text is usefully approached by situating it in context (imagining a contemporary audience) in relation to the subtext (imagining the author's reasons and motives). More complicated yet, what applies to past authors applies to each reader.

As suggested above, finding historical answers in textual 'foreign countries' can be difficult. For all that, the interpretation of primary sources is at the core of the historical enterprise. And if there is no Royal Road, the journey requires careful preparation and a ragbag of skills. In practice, framing questions to unravel the text--to explore 'subtexts'--is often an exercise in imagining the text in a number of different 'contexts', imagining the author faced with unspoken possibilities, unplanned contingencies, and unknown consequences. These games of imagination aim at divining any number of possible meanings that can be supported by the text. Texts are always negotiable.

In the end, clear answers to important questions may not be possible. Historians are thus invited to provide good 'readings' rather than indisputable answers. Good readings conform to the text, they are defended by critical analysis and creative insight. Like any good theory, good readings of primary texts provide a relatively simple way to account for the greatest complexity. The best readings account for the widely agreed "facts of the matter" as well as the most subtle interlinear hints. Most perplexing, the best readings often address issues unidentified for centuries, frequently enough, good readings were not provided by contemporaries, they may not have been anticipated by the author. These are several of the more exciting possibilities in reading primary sources. Texts are continually re-read and re-negotiated.

But enough Thin Theory & Gee Whiz! Consider the following questions as a practical guide to reading primary texts. The issue again returns to good reading skills. The Mantra may look familiar.

1. Text: What is the argument? Specifically, what is the author's thesis? The objectives? How does the structure of the text inform the reader about the thesis and objectives? Does the author seek to persuade, convince, to identify problems, to provide a solution? What are the forms of evidence used by the author? Are they effective? For whom? Are important facts or perspectives omitted? What is left out? Can you improve the author's case? Could new evidence be adduced? Could the text be better organized? What are the 'facts' of the text (stuff most readers might agree upon) and can they be verified? Is the primary text consistent with what you have read in other primary texts? In other 'secondary sources'? What does the text seem to reveal that is not central to the author's purpose (for example, what may seem unremarkable to the author may be revealing to a modern reader.) Is the author credible -- to whom? Explain. Does the author attempt to be objective, impartial, neutral? Does the author consider alternative positions and perspectives? Does the author acknowledge prejudice or personal interest? Is there an ax to grind? Are 'opponents' mentioned, either by name or by school or by tradition? Provide examples of an interpretation made by the author and compare it with what you think represents an 'indisputable fact' found in the text. What arguments and evidence do you have for your claims? How could you make your interpretations more persuasive? Be specific.

2. Context: What kind of document is it? A diary, manuscript treatise, letter, printed document? Was it published--when and where? Who is the author and what position, role, reputation, status, did the author have at the time of writing? Is the author well-known today? At the time of writing? Who is the intended audience? Who read this text at the time? What were the responses? What was to be gained and what were the risks in writing this text? How is this document related to other primary documents known to you, particularly from the same time period? Does this document square with what you know from secondary sources? What evidence do you have for your claims about the text? Be specific.

3. Subtext: What does the text assume on the part of the reader? What has the author knowing assumed (consciously, rationally, purposefully), what has the author silently presupposed (perhaps unconsciously, or perhaps to spare the audience from the obvious)? What intellectual assumptions has the author made? Do you think the author aware of these assumptions? What personal motives are there in writing this piece, that is, why was it written? Can you distinguish between the author's purpose and the author's motives, that is, between reasons stated and unstated, and statements that suggest other unspoken factors (causes vs. effects; reasons vs. actions; statements that may betray distinctions crudely distinguished by conscious vs. unconscious)? Was the purpose of the writing to convince, persuade, motivate? Was the writing a call to action, to change ideas, beliefs, behaviors? Is the structure based on defining and presenting a problem/solution? Is there a school or genre or tradition to which the author belongs? Is there an ax to grind? Is the author straightforward in presenting information and arguments? What rhetorical devices are employed? What does the author assume? What preconceptions (which may not be perfectly clear to the author) seem evident? Identify these unstated aspects (as you interpret them) and other assumptions, presupposition, and possible motives not stated in the text. What evidence and arguments do you have for your interpretations and your claims? Be specific.

4. It's-Your-Text: Themes & Patterns: Primary texts sometime work together to form a pattern and a set of themes. Here the usual concern is to look for similarity and difference. Primary texts can be made to speak to continuity and discontinuity, to coherence and correspondence, to chaos and contradiction. As you read texts from the same time period dealing with similar topics, identify recurrent themes as well as differences of assumption, approach, and opinion. If there are distinct differences, are there similarities in the difference? How do the texts present and defend their views? Are there similarities in method and approach? Why are some texts more compelling than others?

Be prepared to explain your views and defend your interpretations using the text itself. Be prepared (with exact page numbers) to read the relevant passages outloud and have your arguments ready. Be self-conscious about your own assumptions and motives. Be clear about your own methods and approach. What criteria would you use to evaluate the 'truth value' of a primary text? How would you judge the relative credibility of two different sources? Can a primary source be 'valuable' without being 'credible' or 'reliable'? What can be said about the reliability of an author? What can be said about the 'truth-value' of a text? Be clear: What does this ongoing exercise tell you about 'writing history' and the enterprise of 'telling the truth'?



As a means of further focusing your views on the place of primary sources in writing history, consider the following excerpts from a prominent historian early in this century, James Harvey Robinson. The following excerpts have been derived (and modified) from Robinson's essay 'The Historical point of View' published in Readings in European History, Vol I, (Boston: Ginn, 1904): 1-13.


The Sources of History

It is clear that all our information in regard to past events and conditions must be derived from evidence of some kind. This evidence is called the source. Sometimes there are a number of good and reliable sources for an event, as, for example, the decapitation of King Charles I of England in 1649, or for the march of Napoleon into Russia. Sometimes there is but a single, unreliable source, as, for instance, in the case of the burial of King Alaric in a river bed. For a great many important matters about which we should like to know there are, unfortunately, no written sources at all, and we can only guess how things were. For example, we do not know what the Germans were doing before Julius Caesar came into contact with them and took the trouble to give a brief account of them. We can learn but little about the bishops of Rome (or popes) before the time of the Emperor Constantine for few references to them have come down to us.

Few, however, of those who read and study history ever come into contact with the primary, or first-hand sources; they get their information at second hand. It is much more convenient to read what the modern historian Edward Gibbon has say of Constantine than to refer to Eusebius, Eutropius and other ancient writers from whom he gained knowledge. Moreover, Gibbon carefully studied and compared all the primary sources, and it may be urged that he has given a truer, fuller, and more attractive account of the period than can be found in any one of them. His Decline and Fall of the Roman Empire is certainly a work of the highest rank; but, nevertheless, it is only report of others' reports. It is therefore not a primary but a secondary source.

The Problem of Secondhand Knowledge

Most of the historical knowledge current among is not, however, derived from even secondary source such as Gibbon and similar authoritative writers, it comes from the reading of textbooks, encyclopedia stories, dramas, and magazine articles. Popular manual and articles are commonly written by those who know little or nothing of the primary sources; they are consequently at least third hand, even when based upon the best secondary accounts. As a matter of fact, they usually patched together from older manuals and articles and may be four, five, or six removes from the original source of knowledge.

It is well known that the more frequently a report passes from mouth to mouth the less trustworthy and accurate does it tend to become. Unimportant details which appeal to the imagination will be magnified, while fundamental considerations are easily forgotten, if they happen be prosaic and commonplace. Historians, like other people, are sometimes fond of good stories and may be led astray by some false rumor which, once started into circulation, gets farther and farther from the truth with each repetition. [...]

Questions to Ask about a Historical Work

One of the first questions then to ask upon taking up an historical work is, Where did the writer obtain the information? Has the writer simply copied his statements from the more easily accessible works in a familiar language, however unreliable and out of date they may be; or, dissatisfied with such uncertain sources, has the writer become familiar with the most recent researches of the distinguished scholars in the field, in whatever language they may have been written; or, still better, has the historian made a personal study of the original evidence which has come down to us of the events and conditions which are under discussion? [...]

The Necessity of Using Primary Sources

No improvement in the methods of historical instruction in our high schools and colleges bids fair to produce better results than the plan of bringing the student into contact with the first-hand accounts of events, or, as they are technically termed, the primary sources.

This term may perhaps call up in the minds of some the vision of a solitary stoopĦshouldered, spectacled enthusiast, engaged in painfully deciphering obscure Latin abbreviations on yellow parchment. But it is a mistake to conclude that the primary sources are always difficult to get at, dull, and hard to read. On the contrary, they are sometimes ready to hand, and are often more vivid and entertaining than even the most striking descriptions by the pen of gifted writers like Gibbon or Macaulay.

The best secondary authorities stand to the sources somewhat as the description of a work of art or of a masterpiece of literature stands to the original. Just as we cannot afford to ignore the picture itself, or the great poem or drama, and confine ourselves to some one else's account of it, so in our historical work we ought to grasp every opportunity of examining for ourselves the foundations upon which history rests.

It may, of course, be urged that the trained historians, after acquainting themselves with the people and the circumstances of a particular period, can make better use of the sources than any relatively unskilled student. But, admitting the force of this argument, there is, nevertheless, so much to be learned from a study of the original accounts that cannot be reproduced by the most skilled hand, that no earnest student or reader should be content with secondĦhand descriptions when primary sources are available.

The sources are unconsciously molded by the spirit of the time in which they were written. Every line gives some hint of the period in which the author lived and makes an impression upon us which volumes of second-hand accounts can never produce. The mere information, too, comes to us in a form which we do not easily forget. The facts sink into our memory. One who actually talked with Attila, or who witnessed the capture of Jerusalem by the crusaders, is clearly more likely to excite our interest than a writer of our own day, however much the modern may know of the king of the Huns or of the first crusade. It makes no great impression upon us to be told that the scholars of Dante's time had begun to be interested once more in the ancient learning of the Greeks and Romans; but no one can for get Dante's own poetic account of his kindly reception in the lower regions by the august representatives of pagan literature, Homer, Horace, Ovid, and Lucan, people 'with eyes slow and grave, of great authority in their looks,' who 'spake seldom and with soft voices.'

Moreover, the study of the sources enables us to some extent to form our own opinions of the past, so that we need not rely entirely upon mere manuals, which are always one, and generally two or three, removes from the sources themselves. When we get at the sources themselves we no longer merely read and memorize; we begin to consider what may be safely inferred from the statements before us and so develop the all-important faculty of criticism. We are not simply accumulating facts but are attempting to determine their true nature and meaning.

The power to do this is not alone necessary to scholarly work; it is of the utmost importance as well in dealing with the affairs of everyday life. To take a single illustration: one cannot fail to see from a study of the sources that Luther was exceedingly unfair to his enemies and ascribed their conduct to evil motives when they were acting quite consistently and according to what they considered the truth. His opponents, on the other hand, treated him with equal unfairness and proclaimed him a wicked and profligate man because he refused to accept their views.

We meet precisely the same unfairness nowadays, as, for instance, in the case of a municipal election, where each party speaks only evil of the other. It is, however, not so hard to look impartially at the motives and conduct of people who lived long ago as it is to be fair-minded in matters which interest us personally very deeply. By cultivating sympathy and impartiality in dealing with the past we may hope to reach a point where we can view the present coolly and temperately. In this way thoughtful, historical study serves to develop the fundamental virtues of sympathy, fairness, and caution in forming our judgments.

For a longer version of the J.H. Robinson text, see:

A Tentative List of 'Primary Sources' -- Add Others

Handwritten (manuscripts):
Diaries, Journals, Personal Notes, Calendars, Research Notebooks, Laboratory Notes; Drafts of Letters, Copies of Letters; Drafts & Copies of Published and Unpublished works; Private Papers; Laundry & Grocery Lists? - Minutes & Notes from private societies and public institutions;
Notary, Baptismal, Legal, and other handwritten records);

Contemporary Printed Documents (usually 'Published'):
Broadsides, Leaflets, Newspapers, Journals, Monographs, Books;
Published Minutes & Papers of scientific institutions;
Royal or Parliamentary Commissions & Committees;
Legal, Governmental, Patent, & Commercial Documents;
Contemporary textbooks, autobiographies & biographies

Other Stuff (use your imagination here):
Sketches, Drawings, Illustrations, Diagrams, Pictures, Photographs; Photographic Slides, Film, Video, Computer Files, Audio Records, Tapes, CDs; Objects (telescopes, instruments, equipment) Collected Samples (rocks, leaves, insects), Mummies, DNA ... {You get the idea}