Reading a primary source

Lesson 8 was a genre with a fixed shape and a stated method. This one is a genre with no shape at all: a letter, a minute of a meeting, a deposition, a press release, a photograph with a caption.

What they have in common is that none of them was written for you.

The finding this lesson is built on

In a study published in 1991, Sam Wineburg gave eight university historians and eight very strong high-school seniors the same eight documents about one small event, the confrontation on Lexington Green on 19 April 1775. Eyewitness accounts, diary entries, a deposition, a textbook excerpt, and a passage from a novel. He asked them to think aloud.¹

Two results, and the second is the one to build a method on.

The historians ranked the textbook dead last for trustworthiness, below the novel. Three of the eight students put it at or near the top. One called it "the facts". Another called it "straight information" and "objective". The textbook contradicted primary accounts from both sides.¹

And the behavioural finding, in Wineburg's own words: "In almost every opportunity to do so, historians, when first given a document, looked immediately to its attribution, whereas students usually began with the first word in the upper left and never stopped until they reached the last word on the bottom right."¹ That opening phrase is his and it's worth keeping. He is reporting a strong tendency, not a law.

That sentence is the lesson. Everything below is machinery for doing what the historians did.

Why the attribution comes first

A document is not a window onto an event. It's an act, performed by a person, for a purpose, at a moment.

Read from the attribution outward and every sentence changes meaning, because you now know what the sentence was for. Read from the first word and you are taking a purposeful act as a report.

This is also why the textbook came last. A textbook hides its author, states no purpose, and cites nothing. That absence reads as objectivity, and it is the opposite: it means you cannot perform the first move on it at all.

Notice that this puts something before understanding, which isn't what lesson 7 told you. Adler's rule 9 says you may not judge until you can say "I understand". Here you judge the source before you have read a line. The two are not in conflict, and the reason is worth having. What you judge first isn't the claim. It is what kind of act the document is, and that is part of understanding it rather than a verdict on it. You still may not say the author is wrong until you have read him.

Adler puts something like this order in his own book, and it's easy to miss because it comes so early. His rule 1 is to classify the book by kind and subject matter, and you do that from the title page, before the first page of text. Rule 9, which opens the third stage, criticism, is eight rules further on. Sourcing is not rule 1 exactly, since rule 1 asks what kind of thing this is and sourcing asks whose act it was. But it sits in the same place in the order and for the same reason: both settle what you are holding before you weigh what it says.

Four moves

The four are Wineburg's three plus one the Stanford curriculum adds.¹

Sourcing. Who made this, when, for whom, and why? What was their position, and what did they stand to gain or lose by how it came out?

Contextualization. What else was happening? And the sharp version of the question, which is the one that does work: what could this author have known? Not what do we know now.

Corroboration. What do other accounts say, and where exactly do they diverge? The divergence is usually narrower than it looks, and locating it precisely is most of the job.

Close reading. What does the language itself do? Which words are doing legal work, which are terms of art, which would a modern reader misread?

Two accounts, one morning

Here are two real documents about the same quarter of an hour, written one day apart. Both are free: Parker's deposition at the National Archives, and Pitcairn's report. Open them.

The uncontested part first, because corroboration needs somewhere to stand. On 19 April 1775 a British detachment marched from Boston to destroy colonial military stores at Concord. It reached Lexington around sunrise. A body of Lexington militia was drawn up on the Common. Firing occurred. Eight Americans were killed, seven of them Lexington men and one a Woburn man taken on the road. One soldier of the 10th Regiment was wounded and the horse of the major commanding the advance companies was hit in two places. The column went on to Concord.

Who fired first isn't settled, and this lesson is not going to settle it. That is not a hedge. It's the condition that makes the two documents teachable.

The first document

I, John Parker, of lawful age, and Commander of the Militia in Lexington, do testify and declare That on the 19th Instant, in the Morning, about one of the Clock, being informed that there were a Number of Regular Officers riding up and down the Road, stopping and insulting People as they passed the Road, and also was informed that a Number of Regular troops were on their March from Boston, in order to take the Province Stores at Concord, ordered our Militia to meet on the Common in said Lexington, to consult what to do, and concluded not to be discovered, nor meddle or make with said Regular Troops (if they should approach) unless they should insult or molest us and upon their sudden Approach I immediately ordered our Militia to disperse and not to fire. Immediately said Troops made their appearance and rushed furiously, fired upon and killed eight of our party, without receiving any Provocation therefor from us.²

Source it. John Parker, captain commanding the Lexington militia company, swearing at Lexington on 25 April 1775, six days after the morning. Sworn before three Justices of the Peace for Middlesex, and then certified by a notary public of Charlestown attesting that those men really were His Majesty's Justices.

That notarial certificate is the thing to notice, and it is the kind of thing a reader skips. It exists because the paper was going overseas, to readers who did not know a Middlesex justice from anyone. The depositions were taken by order of the Massachusetts Provincial Congress, whose covering address says the purpose in its own words: "it being of the greatest importance that an early, true, and authentick account of this inhuman proceeding should be known to you, the Congress of this Colony have transmitted the same". They were printed at Worcester, and the Congress put its set aboard a fast ship that reached England ahead of General Gage's own dispatches, which had already left Boston.²

So this isn't a witness statement. It is one exhibit in a brief, assembled at speed by a body that was a party to the quarrel, for a jury three thousand miles away.

Notice also the one thing Parker swears that doesn't help him, and it is in his own first clause: he "ordered our Militia to meet on the Common in said Lexington". In April 1775 that assembly is itself the fact a Crown lawyer would have wanted, and it is in the document because Parker put it there.

Be careful with the rest of the same sentence, though. That he ordered them "to disperse and not to fire" isn't a concession. It is the exculpation the deposition exists to deliver, and reading it as an admission against interest is the mistake this lesson exists to prevent.

Close-read it. Five words a modern reader will misread. Regulars means the British standing professional army, as against part-time militia; it's a category before it is an insult. Insult in 1775 is wider than today's word and reaches assault as well as affront, so in Parker's conditional it can reach being set upon, and not only being spoken to rudely, and which of the two he had in mind is not decided by the word. The same word is at work two clauses earlier, where officers are said to be "stopping and insulting People" on the road. Molest means to interfere with, with no sexual sense. Meddle or make with is an idiom meaning to have anything to do with. And of lawful age is a formula establishing competence to swear, not a fact about Parker.

Notice what the document doesn't say. Parker does not say he saw the first shot.

The second document

When I arrived at the head of the advance Company, two officers came and informed me, that a man of the rebels advanced from those that were assembled, had presented his musket and attempted to shoot them, but the piece flashed in the pan. On this I gave directions to the troops to move forward, but on no account to fire, or even attempt it without orders; when I arrived at the end of the Village, I observed drawn up upon a Green near 200 rebels; when I came within about 100 yards of them, they began to file off towards some stone walls on our right flank. The Light Infantry, observing this, ran after them. I instantly called to the soldiers not to fire, but surround and disarm them, and after several repetitions of those positive orders to the men, not to fire, etc. some of the rebels who had jumped over the wall fired four or five shots at the soldiers, which wounded a man of the Tenth and my horse was wounded in two places, from some quarter or other, and at the same time several shots were fired from a meeting house on our left. Upon this, without any order or regularity, the Light Infantry began a scattered fire, and continued in that situation for some little time, contrary to the repeated orders both of me and the officers that were present.³

Source it. Major John Pitcairn of the Royal Marines, commanding the light infantry companies at the head of the column, writing from Boston Camp on 26 April 1775 to General Thomas Gage, his commander-in-chief and the governor of Massachusetts. His first line says why: "As you are anxious to know the particulars ... agreeable to your desire". So: a subordinate's report, written at his superior's request, by the officer whose own conduct is in question, and written after his superior's own version of the affair was already in print.³

That is the mirror of Parker's position, one day later, and it's why these two are worth putting side by side. Neither is the disinterested one.

Close-read it. Rebels is the Crown's legal description of subjects in armed resistance; like Regulars it classifies before it abuses. Presented his musket is drill language for bringing the weapon up to the aim, and the piece flashed in the pan means the priming powder burned and the main charge did not go off, so the man pulled the trigger and no shot came. A misfire, not a figure of speech. The whole British claim of a first attempt rests on those five words. The Tenth is the 10th Regiment of Foot. Light Infantry were the flank companies sent ahead. File off is a drill term for moving away in column rather than fleeing. The meeting house is the town's church, which stood on the Common and served as its hall.

And now the two clauses a reader skims.

The first is an incident most summaries leave out, and it happened on the road before Pitcairn reached the village. Two officers "came and informed" him that a man of the rebels had presented his musket at them and that the piece flashed in the pan. Pitcairn was riding up at the time and he tells you so, which is the sourcing question of this lesson asked of the British document and handed to you by its author.

Keep it separate from what follows. His claim about the Common, that men who had jumped the wall fired four or five shots, is his own, from about a hundred yards, and doesn't lean on those two officers at all.

The second is "from some quarter or other". Read it carefully and see what it attaches to. Not the four or five shots, which he assigns without hesitation to men who had jumped the wall, but his own horse's two wounds, which he cannot place. The comma is doing that work, and an eighteenth-century comma is not always load-bearing, so read the clause twice before you lean on it. He is certain about the accusation and uncertain about the injury to himself. Notice the unevenness, and don't turn it into a verdict. A man writing fast about a confused few minutes would produce it, and so would a man who had already decided which part of the story mattered.

Corroborate

Now the move, and the discipline is to say where exactly they diverge rather than which is right.

They agree that the militia formed on the Common and that firing then happened, and neither contradicts the other about the rest of the frame: the march on the stores at Concord and the eight dead are in Parker, the wounded man of the Tenth and the major's horse are in Pitcairn, and each lets the other's detail stand. Two interested accounts written a day apart, from opposite sides, leaving all of that alone. The uncontradicted material is the strongest thing either document gives you, and a reader who goes straight to the disagreement throws it away.

And notice what is not agreed, though it looks as though it should be. Each commander states that he ordered his own men not to fire. Parker swears he told the militia to disperse "and not to fire"; Pitcairn writes of "positive orders to the men, not to fire". Neither document says a word about the other's order. Two parallel claims, each made by the man it exonerates, are not corroboration, and moving them into the agreed column is the commonest way this move goes wrong.

They diverge on one clause, and it's a square contradiction. Parker says the troops fired "without receiving any Provocation therefor from us". Pitcairn says some of the rebels "who had jumped over the wall fired four or five shots at the soldiers". Both cannot be right.

What the sourcing move buys you is not a way to dissolve that. It's a precise statement of what each man was in a position to know. Parker swears he had ordered his company to disperse and not to fire, and his sentence is a claim about his own company's conduct, sworn six days later into a legal brief. Pitcairn was mounted at the head of the advance, writing seven days later to the man who needed a version for London. His account of what happened on the Common is first hand; the earlier musket that flashed in the pan, out on the road, he has at second hand from two officers he does not name.

Neither says, in those words, that he saw the first shot. No third account settles it. The disagreement is one clause wide, precisely located, and unresolved.

That's a result. A reader who arrives at a precisely located disagreement has done something a reader who arrives at a verdict has not.

The move that separates a historian from a fact-checker

Here is a third document, and it's the one worth the lesson.

In 1826 Sylvanus Wood of Woburn, aged seventy-four, swore an affidavit about that morning. He had heard the bell, taken his gun, and gone in haste to Lexington, about three miles off. Among other things he said:

There was not a gun fired by any of Capt. Parker's company within my knowledge. I was so situated that I must have known it, had any thing of the kind taken place, before a total dispersion of our company. I have been intimately acquainted with the inhabitants of Lexington, and particularly with those of Capt. Parker's company, and with one exception, I have never heard any of them say or pretend that there was any firing at the British from Parker's company, or any individual in it, until within a year or two. One member of the company told me, many years since, that after Parker's company had dispersed, and he was at some distance, he gave them "the guts of his gun."⁴

One sentence earlier Wood has Parker's men "dispersing and leaping over the wall" as the British second platoon fired. That is the same wall Pitcairn says four or five shots came from. It settles nothing, and a document of 1826 couldn't settle it anyway, but it is the kind of overlap corroboration exists to find, and it's a reason to read past the sentence you came for.

Source it before you weigh it.

The date. 17 June 1826, which is fifty-one years and two months after the morning it describes. The man is seventy-four and is reporting speech word for word.

The moment. Lafayette's American tour of 1824 to 1825 had set off a wave of Revolutionary reminiscence and monument-building, and with it a public dispute between Lexington and Concord over which town had begun the war.

Who collected it, and this is the part that changes the document. Wood's affidavit was printed in a book by Ezra Ripley, minister of Concord, "with other citizens of Concord", whose subtitle states the thesis on the title page: that it was at Concord that "the first regular and forcible resistance was made to the British soldiery". He was answering Elias Phinney's History of the Battle at Lexington of 1825, which had argued that Lexington men returned the British fire, and so fired the first American shots of the war. Phinney was making Lexington's case as squarely as Ripley made Concord's, and both books want reading at the same arm's length. He printed a run of affidavits sworn in 1826 and 1827: four of them, three taken before Nathan Brooks, Esq., justice of the peace, of Concord. Three pages earlier Ripley prints Brooks as a witness on his own side, affirming that a Lexington colonel had never given him the idea of resistance and firing by the militia. So the officer who took three of these four oaths was himself a party to the argument they were collected to settle. Every one of the four makes the same negative point, and three of them add, against Ripley's own interest, the single instance they had heard of.⁴

So an affidavit that reads on its face as testimony for Lexington, sworn by a Woburn man who had fallen in with Parker's company that morning, was gathered by the Concord party, to take the priority away from Lexington. Two things keep that from being a discreditable story. Ripley put his thesis on his own title page rather than hiding it, and he printed the exceptions; on the page after Wood's affidavit he even prints Wood's claim to have taken the first British prisoner and then argues against it in his own voice. None of that makes him dishonest. It does mean his candour and his interest run through the same pair of hands, and a reader who takes the first should see the second. And a rivalry between two towns over which of them began the Revolution was an ordinary and public thing in the eighteen twenties, argued in print by respectable men, rather than a scandal.

What follows is not that the document is worthless. It is that it's evidence of something other than what it is about. As testimony about 1775 it is late, it was solicited by an interested party into a live dispute, and it quotes speech across half a century.

As evidence about 1826 it's first-rate: it shows what the Concord party was trying to establish and how the memory of that morning had been organised by then.

And read the last two sentences together, because they show something Ripley did not need. Wood states the general denial, and then, unprompted, gives away the one instance that cuts against it, and dates it "many years since" rather than to the recent dispute. A collected affidavit that volunteers the exception to its own point is a more interesting document than a clean one, and a reader who stops at the denial misses it.

That's the move. A fact-checker asks whether the document is reliable. A reader asks what it is reliable about, and those are different questions with different answers.

Where this course's own reading stops

Two things you are entitled to know, because lesson 8 asked you to check a course's sourcing and this is the lesson where ours is thinnest.

The Wineburg study was read through Wineburg's own later account of it, a short article he wrote for historians in 1992, and not through the 1991 journal article. Everything attributed to it here is in that account.¹

And the British document above is the weaker half of the pair, on provenance. Parker's deposition is held by the National Archives and its transcript is published by them. Pitcairn's letter is in a manuscript collection this course could not reach, and was printed in a 1932 edition nobody here has opened. The transcription used above is a web one that names no source, and two others like it don't agree with each other about spacing and headings.

What we could do is check that the text was not invented. General Gage's printed Circumstantial Account of an Attack that happened on the 19th of April 1775 tracks this report almost clause for clause: the militia beginning "to file off towards some Stone Walls" on the flank, the major calling to the soldiers not to fire but to surround and disarm them, the four or five shots from men who had jumped the wall, the man of the 10th, the horse wounded in two places.

Read the dates before you take that for reassurance. The Clements Library, which holds Gage's papers, dates that broadside to 21 April, which is five days before Pitcairn wrote.³ So it's not a second witness and it is not derived from this letter. The likeliest explanations all run the other way: that the broadside's Lexington paragraph descends from an earlier report of Pitcairn's to headquarters, or that both descend from Colonel Smith's report, which Pitcairn's own closing line points at, or that Pitcairn wrote this letter with the printed account at his elbow.

And the match is weaker evidence than it first looks, because a transcription made from the printed account would match it just as well. What it does establish is that a document of this wording was moving out of that headquarters within days of the morning. That is less than corroboration of the events and less than proof of the manuscript, and a lesson on corroboration would be teaching it wrongly if it claimed either.

So: the Parker passage is verified to the archive. The Pitcairn passage is verified only to a transcription whose provenance is attested at second hand. That's a real difference in what you should let each one carry, and it is exactly the kind of thing the sourcing move exists to surface.

Does any of this transfer?

A fair question, since sourcing a 1775 deposition is not obviously a general reading skill.

Avishag Reisman's 2012 study is the evidence that it is. Two hundred and thirty-six eleventh-graders across five San Francisco high schools, six months, eighty-three lessons built on documents. Gains on four measures: historical thinking, factual knowledge, general reasoning, and general reading comprehension.⁵

That last one is why a primary-source lesson sits in a general reading course rather than in a history one. This course read that study at abstract level and not further, which is a limit on how much weight to put on it.

What people get wrong

"A textbook is more reliable than a document from the period." The historians in Wineburg's study put the textbook last, below a novel, and the reason isn't that textbooks are bad. It is that a textbook gives you nothing to source: no author, no purpose, no position. What reads as neutrality is the absence of the information you need to judge.

"Primary sources are unbiased because they were there." Being there's a reason to read someone. It is not a reason to believe them. Both men above were there, both were interested, and both had a superior to satisfy.

"Bias means useless." It means the document is evidence of something specific. An interested account tells you what its side needed on the record, which is often a more answerable question than what happened, and sometimes a more useful one.

"Contextualizing is background colour." It decides what the author could have known. Pitcairn writing on 26 April did not know what the Provincial Congress had sworn on the 25th. Wood in 1826 knew about a dispute that didn't exist in 1775. Neither of those is scene-setting; each one changes what the document can mean.

Practice

Connections

Lesson 7 asked you to understand before judging. This lesson looks like an exception and is not: what you judge first is what kind of act the document is, which is part of understanding it. The verdict on what it claims still waits.

Lesson 8's methods section is doing the same job as sourcing here. Both ask what the thing in front of you can support, before you ask what it says.

Lesson 3 explains why a document from another period is hard, and it's not the words. It is the relations that everyone at the time supplied without noticing: what a notary's certificate was for, why a meeting house was on the Common, what the Provincial Congress was.

Lesson 10 needs this, because a question worth asking will usually have a document behind it as well as a book, and because the four moves are where the terms come from.

Go deeper

• Captain John Parker's deposition, 25 April 1775. The manuscript and a transcript, at the National Archives. Unrestricted for access and use, and worth looking at as an image: the handwriting is part of the source.
• Major Pitcairn's report to Gage, 26 April 1775. Free, and read it against the caveat above about its provenance.
• Wineburg, "Probing the Depths of Students' Historical Knowledge". Free, short, and his own account of the study this lesson rests on.
• The Digital Inquiry Group's document lessons. Free and openly licensed, built on this research, and the place to go for more documents already prepared with questions.

Sources

1. Sam Wineburg, "Historical Problem Solving: A Study of the Cognitive Processes Used in the Evaluation of Documentary and Pictorial Evidence", Journal of Educational Psychology 83(1), 1991. Read through Wineburg's own account of it in "Probing the Depths of Students' Historical Knowledge", Perspectives on History, 1992; the journal article itself was not fetched, and nothing here goes beyond that account. Eight historians and eight strong high-school seniors, eight documents on the Lexington confrontation, the textbook ranked last for trustworthiness below a novel, the students' phrases "the facts", "straight information" and "objective", and the quoted behavioural finding about attributions. Sourcing, contextualization and corroboration originate with this study; close reading is the fourth move the Stanford curriculum adds.
2. Deposition of Captain John Parker, Lexington, 25 April 1775, National Archives, identifier 595246, Papers of the Continental Congress, Record Group 360. The passage above is verbatim from the Archives' own transcript. The 1775 document is public domain and the transcription is a work of the United States Government; the record is marked unrestricted for both access and use. The Provincial Congress's covering address, "To the Inhabitants of Great Britain", 26 April 1775, is quoted from the printing of the depositions in Peter Force's American Archives, fourth series, volume 2. Note that Force and the Archives transcript differ in small ways: Force modernises spelling and capitals, prints the jurat and the notarial certificate, and omits the word "molest".
3. Major John Pitcairn to General Thomas Gage, Boston, 26 April 1775. The passage above is verbatim from the transcription at Teaching American History. The provenance is attested only at second hand, as the body of this lesson says: the manuscript is in the Gage Papers at the Clements Library and was printed by Allen French in 1932, and neither was reachable here. A faithful transcription of a public-domain manuscript carries no independent copyright, so the 1775 words are safe to quote; the site's editorial matter is theirs. The printed account compared with it above is Thomas Gage, A Circumstantial Account of an Attack that happened on the 19th of April 1775 (Boston, printed by John Howe), which the Clements Library's catalogue of the Gage papers dates 21 April 1775; the phrases quoted from it here were read from the American Antiquarian Society's transcription, and research/SOURCES.md records both.
4. Affidavit of Sylvanus Wood, 17 June 1826, printed in Ezra Ripley, A History of the Fight at Concord (Concord, 1827), 53 to 54, Internet Archive, a library-sponsored scan marked not in copyright. Both the 1826 affidavit and Ripley's 1827 volume are public domain. Ripley's subtitle carries his thesis, and his book answers Elias Phinney's History of the Battle at Lexington of 1825. Transcriptions differ in small ways: a modernised text drops Wood's own footnote qualifying his count of the militia, which is worth knowing before quoting that count.
5. Avishag Reisman, "Reading Like a Historian: A Document-Based History Curriculum Intervention in Urban High Schools", Cognition and Instruction 30(1), 2012, 86 to 112. Read at abstract level. 236 eleventh-graders across five San Francisco high schools, six months, 83 lessons, with main effects on historical thinking, factual knowledge, general reasoning and general reading comprehension.