Who Could Read and Write?

"I conclude that the overall level of literacy is likely to have been below 15%." — William Harris, on Italy, 1989¹

"Accordingly, it seems a safe statement that between five and ten percent of Judaean men in the years dividing Pompey from Hadrian were able to read books." — Michael Wise, 2012²

Summary

In the first-century eastern Mediterranean, "literate" did not name one skill.  A person might hear a text read aloud, read a familiar text slowly, copy a formula, sign a name, write a short letter, dictate a long one, or compose a book in polished Greek.  These are different abilities, often in different languages, and evidence for one does not by itself establish another.  This article sets out what the evidence shows about each, who had them, and what that does and does not tell us about the people behind the New Testament.

1. Reading, writing a signature, writing freely, and composing are separate skills. The papyri record men and women who could do one and not another. A village clerk named Petaus, in Egypt in the 180s, signed papers but "could not read what he wrote."³ A man kept from writing by the condition of his arm had his nephew write for him after he had "read (the document) clear through".³ That man was literate and physically unable to write; he is not evidence for a class of people taught to read but not to write. Literacy in Aramaic was not literacy in Greek, and owning a document was not writing it.
2. Every population figure is a model, not a count, and the models answer different questions. William Harris (1989) estimated literacy below 15 percent for Italy and below 5 to 10 percent for the western provinces, from schools, economy and inscriptions; he gave no figure for Judaea.¹¹ Meir Bar-Ilan (1992) put Jewish literacy in the Land of Israel "probably less than 3%" by weighting assumed rates for country, town and city, with zero in the countryside.⁴ Catherine Hezser (2001) said whether it was 3 percent "or slightly higher, must ultimately remain open."⁵ Among these four studies, Michael Wise (2012) alone started from named people. In the legal papers of mostly propertied Judaean families, 13 of 53 laymen who were parties to the documents signed in a practised hand, which he takes as the mark of a reader of books.² To reach "between five and ten percent of Judaean men" he assumed that such families were the top quarter of society and that nobody below them read books.²² His figure that "about 16% of Judaean adults" could sign is, on his own population inputs, the share of adult men; for all adults the same inputs give about 8 percent.²⁶ This article quotes these models; it does not adopt any of their numbers as a measured rate.
3. The surviving documents show a world run on other people's writing. In Egypt the formula for writing on behalf of someone who cannot write is, in the words of the standard papyrus lexicon, "of constant occurrence".⁷ A deed of AD 99 ends, "I, Heron, also wrote for my wife, who does not know letters."⁸ In Arabia in AD 125, one deposition of a Jewish landowner named Babatha carries the hand of a professional scribe, a man writing her subscription "because she does not know letters", her guardian's own subscription, and witnesses signing in Aramaic and Nabataean.⁸ Yigael Yadin, who found the revolt leader Bar Kokhba's letters, saw a different hand in each and judged that they were dictated; "Not one letter is signed by Shimeon bar Kosiba personally."⁹ Writing a name reached further down. In first-century Jerusalem, hundreds of the stone boxes that held the bones of the dead carry names scratched by relatives rather than by professional scribes, though whether those writers could do more than names is, as Hezser says, "difficult to determine".⁵
4. Schooling was an ideal for some; how far it reached is not known. Josephus says the law "orders that they shall be taught to read".¹⁰ That states an ideal in a defence of Judaism, not an enrolment record. The report that the high priest Yehoshua ben Gamla set up teachers of children "in each and every town" is told in the Babylonian Talmud in the name of Rav, a third-century teacher.¹¹ It may preserve a memory of a reform by a real first-century high priest, but it cannot show that every village had a school.⁵ In Wise's documents all ten women who were parties were "unable to sign".² Ten women from propertied families do not measure all women; together with the papyri and the ancient authors read here, they show that women's writing was rare, and they say nothing about whether women read.
5. Acts 4:13 calls Peter and John, and only them, ἀγράμματοι καὶ ἰδιῶται. The first word means "unable to write" in the legal formulas of the papyri and can mean "uneducated" in literature; the second means a layman. Specialists divide. Craig Keener and Craig Evans read the pair as "without scribal training"; Moulton and Milligan read "illiterate"; Thomas Kraus, in a study devoted to the phrase, accepts "illiterate" as possible "provided that other meanings are not excluded".^7121314 The main readings considered here share a portrait of men without recognised scribal credentials. Luke presents that judgement as accurate, though his narrative is not a record of their schooling. Whether it also says they could not read or write cannot be settled from the verse, and it says nothing about the other ten. The Gospels themselves place Peter, Andrew, James and John in boat-owning families that hired labourers (Mark 1:20), not among the day labourers; the fishing papyri of Egypt put such men "above-subsistence" in a trade full of leases and receipts, though in one typical lease of AD 161 the three fishermen "do not know letters" and another man wrote for them.¹⁵⁸
6. Origen's claim that the fishermen and tax collectors "had not even learned their first letters" is a third-century reading, not a record. The words are Origen's own, in an argument that their success came from divine power, though he says Celsus accepted the gospel's account of their lack of training.¹⁶ The same Origen held that Matthew, "once a tax collector", wrote the first Gospel "in Hebrew letters", and that Mark wrote as Peter guided him.¹⁷ He affirms both, without explaining their relation; later learning, writing in Hebrew and help from others are all open.
7. The tax-collector question does not resolve by occupation. One literate customs receipt does not show who wrote it. One semi-literate Egyptian village clerk does not set the skill of every tax collector. The trade made handling documents and accounts likely; how much it raised the chance of literary skill is unknown, which is not the same as zero. Hezser notes that Roman Palestine's administration was "rudimentary only" compared with Egypt's.⁵
8. Secretaries were real, and "using a secretary" covers different jobs. Cicero dictated one careful letter "syllable by syllable" to a slower writer rather than to Tiro, who took down "whole periods at a breath",¹⁸ and in a crisis he asked a friend to write letters in his name and "say whatever you think right."¹⁸ Josephus wrote his Jewish War "with the aid of some assistants for the sake of the Greek".¹⁰ An author composing aloud, a friend writing in his name and assistants polishing a book's Greek are three different things. Tertius wrote Romans (Romans 16:22), and Paul added lines in his own hand. E. Randolph Richards, the main specialist, offers his idea that Tertius took Romans down in shorthand only as a possibility that "deserves a future investigation", and calls one step in it "a tenuous assertion".¹⁹¹⁹ Knowing that secretaries existed does not tell us what one did in a given letter.
9. Note-taking was a real practice among educated students, and no early source read here says a disciple did it. Quintilian's pupils published his lectures from "such notes as my good pupils succeeded in taking down".²⁰ Arrian says he wrote down Epictetus "word for word, as best I could".²¹ Alan Millard calls notes by Jesus' hearers "quite feasible".²² Papias, who does say how he thought Mark and Matthew came to write, mentions no notes.²³ Possible is not attested, and silence is not disproof.

Verdict.  Reading books was probably a minority skill in Roman Judaea and Galilee.  The evidence does not give the same rate for reading, signing, writing letters and composing literature.  That judgement rests on models, on documents from propertied Judaean and Arabian families and on analogy, not on any count for Galilee, so no percentage should be quoted as known.  Writing reached far beyond those who could do it, through family members, village scribes and secretaries.  A population estimate is a starting point, not a verdict on any named person.  It gives a reason to seek positive evidence when a particular fisherman, tax collector or craftsman is said to have composed a long Greek text.  Choosing its words, writing it by hand and authorising a secretary are different claims.  It does not show that anyone named in the Gospels could not read, could not dictate, or could not stand behind a book written down by someone else.  Each authorship question has to be settled by its own evidence, as the sibling articles Who Wrote the Gospels?, Did Paul Write the Disputed Letters? and Oral Tradition and Memory try to do.

Part 1: Six skills, not one ladder

What the ancient words meant

Ancient Greek had no single word that matched the modern question "can he read?"  The papyri speak of a person who "does not know letters" (μὴ εἰδὼς γράμματα) or who is ἀγράμματος, "unlettered".  In documents these words have a narrow meaning.  Harris puts it plainly:  "In documentary contexts all Greek and Latin expressions concerning “knowing letters” refer to literacy in the narrow sense".¹  In literature the same words drift.  Harris again:  "Like the word illiterate, Greek agrammatos and Latin illitteratus seem to veer between the meanings “uncultured” and “incapable of reading and writing.”"¹  He concludes that "Such expressions have to be interpreted case by case."¹

The Gospels and Acts use the same vocabulary.  John 7:15 has the crowd ask how Jesus "knows letters" (γράμματα οἶδεν) "having never studied".  Acts 26:24 has Festus tell Paul, "your great learning (τὰ πολλά ... γράμματα) is driving you mad."  Nobody thinks Festus meant Paul knew too many letters of the alphabet.  So a reader cannot settle what any one passage means by translating γράμματα as "letters" and stopping.

The skills, separated

Chris Keith, whose Jesus' Literacy (2011) is the fullest study of the question for Jesus' world, calls literacy "a spectrum, running from simple skills like signing one’s name to advanced skills like composing intricate texts."²⁴  That is a useful start, but the evidence suggests something more like several separate dimensions than a single ladder.  A person could be high on one and absent on another.

Two more distinctions cut across the table.  The first is language.  Keith notes "a single person could hold different literate competencies in different languages."²⁴  A Jew could read Hebrew scripture and be unable to write Greek, or sign in Aramaic and need a Greek scribe.  The second is the difference between documentary and literary writing.  Hezser draws a "wide gap" between the "literary literacy" of "rabbis, Torah scribes, and some members of the Jewish aristocracy" and "the various types and levels of practical literacy which some merchants, artisans, and farmers may have possessed".⁵

Two traps in the vocabulary

Reading without writing.  Alan Millard argues that ancient societies had "less educated ones who could read but hardly write, some who were readers alone".²²  Harris, who first raised the possibility, judged the other way:  "There is, however, no especial reason to think that those who could truly read and truly not write were numerous."¹  Neither cites a count.  The question is open, and it matters, because a population of readers who left no writing would be invisible to every method that counts signatures.

"Craftsman's literacy" means two different things.  Harris coined it for a society, not a person:  "By craftsman’s literacy I mean not the literacy of an individual craftsman but the condition in which the majority, or a near-majority, of skilled craftsmen are literate, while women and unskilled labourers and peasants are mainly not".¹  Keith quotes that definition and then uses the term for an individual's skill set:  "Petaus’s ability to reproduce his formula and sign his name in Greek, however imperfectly, is a form of craftsman’s literacy."²⁴  Harris's term describes how literacy is spread across a population;  Keith's describes what one man could do.  Arguments that slide from one to the other, for example "the trade gave him craftsman's literacy, so he could keep accounts", are using the word in two senses.

Part 2: How many? What the estimates are

Four estimates, four definitions

The figures repeated in popular books ("3 percent", "10 percent", "90 percent illiterate") come from a few scholars.  Each answered a different question, with different evidence and different assumptions.

Harris never estimated Judaea.  He discussed it only to reject one scholar's claim that writing there "was an essential accompaniment of life at almost all levels".  Jesus' "Have you not read?" was put to Pharisees and chief priests:  "They presumably had read."  The "mirage of mass literacy in first-century Judaea ... begins to fade."¹  The often-quoted figure of 10 to 15 percent for the Roman Empire is a later summary of his work, including Hezser's,⁵ not a figure Harris gives for the empire as a whole.

How each figure was made

Harris argued from conditions, not counts.  He says plainly that "We shall obviously never know in a clear-cut numerical way how many people were literate".¹  His reasons are that mass literacy elsewhere has needed subsidised schools and strong incentives, and that "some of the vital preconditions for wide diffusion of literacy were always absent in the Graeco-Roman world".¹  His numbers are bands of plausibility, set against early modern England, where male literacy was 20 to 30 percent.

Bar-Ilan started from a rule for "A town in which there is only one who reads; he stands up, reads (the Torah), and sits down ... even seven times."⁴  He takes it from the late tractate Soferim, with a parallel in the Tosefta that speaks of a synagogue with only one reader.⁵  From synagogues seating more than 50, he inferred towns with literacy "approximately 1 percent (if not lower)".  He met the objection that the rule covers a rare case by arguing that "usually the Tannaim did not speak of rare cases", and he set the countryside at zero because "there are rules that reflect a zero literacy rate in the rural areas".⁴  He then assumed 70 percent of the population was rural with 0 percent literacy, 20 percent lived in towns at 1 to 5 percent, and 10 percent in cities at 2 to 15 percent.⁴  The arithmetic of his own inputs gives 0.4 to 2.5 percent.  The result is fixed by the inputs.  If rural literacy were 1 percent instead of zero, the total becomes 1.1 to 3.2 percent;  at 5 percent rural, 3.9 to 6.0 percent.²⁵  A rule written for a single reader shows that the rule's authors thought such a congregation worth legislating for.  Whether such congregations were common, and how literate the rest of the population was, the rule does not say.

Hezser gave two answers because she used two definitions.  For "the ability to read documents, letters, and “simple” literary texts in at least one language and to write more than one’s signature", Jewish literacy was "well below" Harris's level.  For "the ability to read a few words and sentences and to write one’s own signature only, Jews probably came closer to the Roman average rate."  And she adds that "exact numbers can neither be verified nor falsified and are therefore of little historical value".⁵  Her "ninety percent" figure comes with the word "likely", and it groups those who could only sign with those who could do nothing.⁵

Wise is the only one of the four who counted people, and his work has three layers that need to be kept apart.

1. What the documents record. Wise graded the handwriting of every named principal, witness and subscriber in the contracts and letters from the Judaean Desert caves. Of 69 principals, "twenty-nine of sixty-nine (twenty-nine of sixty-three discounting scribes)" could not sign: "Nineteen men and ten women".² That is 42 percent of all principals, 46 percent of those who were not scribes, and 36 percent of the 53 lay men. Thirteen of those 53 men wrote at level 3.² The documents mostly belong to the decades before and during the Bar Kokhba revolt, a minority to the first century BC and AD, and they come from families with property to register. Wise calls the people he counts "individuals"; the dissertation does not set out a rule for someone who appears in several documents, though only ten women are counted, which suggests people rather than appearances.
2. What the hand is taken to mean. Wise's step from a practised hand to a reader of books rests on Raffaella Cribiore's study of Egyptian school exercises: "because they copied what they read, their hands improved correspondingly", so a hand "affords a rough idea of how far any given writer has traveled down the road to reading."² He admits "An obvious element of subjectivity" in sorting hands into levels,² and no test by a second judge is reported in the pages read here. He also warns that a working scribe "might acquire a practiced hand simply by dint of dull, methodical reproduction" without literary reading, which is why he sets scribes aside.² The same can happen to a layman who signs often. In the other direction, a man who read well but wrote little would be graded low. So level 3 is a reasonable sign of long schooling, not a measurement of reading.
3. How the counts become a population figure. Wise pooled the 34 literate lay male principals with 156 male witnesses, though the witnesses "were mostly chosen as meeting the dual criteria of ability to sign and membership in the family".² In that pool of 190, 90 wrote at level 3, which gives "fully 47.4%".² He then multiplied: 65 percent of propertied men wrote at some level, 47.4 percent of those read books, and the propertied made up the top quarter of the population. Below that quarter, "We have no reason to think that they could read books."² That gives 7.7 percent of all men.² For all adults he halved it, since women "were more or less universally illiterate".² He calls the result "more of a thought experiment than a hard number".²

The model is sensitive to its inputs.  Using the principals alone for the level 3 share gives 6.2 percent of men.  The smaller elite of about 8 percent, which Wise also discusses, gives 2.5 percent.  If 1 or 3 percent of men below the top quarter read books, the total becomes 8.4 or 9.9 percent.⁶  None of these is a better estimate;  they show what the printed figure depends on.

Wise's signing figure has a further problem.  His population base is "one million", of whom "about one in five would be adult males ... (for every man, one woman and three children)", so "200,000 men".²  He then counts 65 percent of a quarter of 200,000, "32,500 men", and says this "would entail that about 16% of Judaean adults were signature literate."²  On his own inputs, 32,500 is 16 percent of adult men, and about 8 percent of the 400,000 adults if no women signed.  He halves the reading figure for women but not the signing figure.  This article therefore does not repeat 16 percent as a share of adults, and it does not count Wise's reading and signing figures as two results that support each other:  both come from the same counts and the same assumptions.

What the evidence supports, and how strongly

The four estimates are not four measurements of one quantity.  What the evidence does support, with the article's own confidence:

1. In the one counted group, many could not sign. Among the parties to the Judaean Desert legal papers, mostly propertied families of the early second century, 19 of 53 lay men and all 10 women could not sign, and 13 of the 53 men wrote a practised hand.²² Confidence is high for the sample as Wise classifies it. It is a selected group, chosen by what survived in the caves and by who had property to record, not a sample of Judaea.
2. Reading books was a minority skill. Every scholar read here agrees, including those who argue for a literate Jesus. Craig Evans writes that most historians put literacy "somewhere between 5 and 10 per cent", adding "I do not dispute this conclusion", and infers 10 to 20 percent among men;¹³ John Meier holds that "we are not to imagine that every Jewish male in Palestine learned to read".²⁶ The agreement is real, but it is agreement among models built on overlapping assumptions. The article judges book-reading to have been a minority skill, probably a small minority on these models. The models do not establish its size, and they do not give the same rate for signing or practical writing.
3. Signing was more common than reading books. This holds in Wise's sample and in Egypt, where Youtie found many who "at best could write their names and perhaps a few words in addition".³
4. Writing followed property and sex. In the samples it did. For slaves and the rural poor there is no direct evidence: Wise says he has "no means whereby to assess such missing categories as the slave population".² Their low literacy is an inference from analogy and from the lack of schooling, not an observation.
5. Nobody measured Galilee. Millard notes that Galilee "has yielded very little" written material, and puts that down largely to "the paucity of Herodian period material from anywhere in Galilee".²² Part 3 gives the state of the evidence.

A population figure is not a verdict on a person

A rate describes a population.  It is a reasonable starting point, a prior, for someone about whom nothing else is known.  If Wise's model were right, a Judaean man about whom nothing else was known would have had something like a 5 to 10 percent chance of reading books.  But the people the New Testament names are not unknown members of a population, and they were not picked at random.  A tax collector at a toll booth, a fisherman whose father employed hired men (Mark 1:20), a Pharisee whom Acts 22:3 says was educated in Jerusalem, and a doctor (if Luke was one, Colossians 4:14) each belong to a different slice of the population, and the rate for that slice is not measured.  Occupation, family and teachers can move the chance either way.  A trade that needed records makes some writing more likely;  the lack of a need does not show the lack of a skill.  People also learn in groups:  a man whose relatives or teacher could read is more likely to read than a stranger.  Writing also clustered where work needed it.  At the Roman fort of Vindolanda in northern Britain, in the western provinces where Harris puts literacy below 5 to 10 percent, the ink tablets of about AD 100 show, in the words of their editor Alan Bowman, "the immense number of individual hands represented. In the tablets discovered in the 1970s more than eighty different hands were identified and in the material from the 1980s there are several hundred more."  Requests for leave are "written in different hands, some of them quite good and coming, we may assume, from soldiers in the lower ranks."²⁷  Bowman also warns that all of this material "emanates from military circles", and that literacy grew "more untidy and haphazard the lower one goes down the social scale".²⁷  A regional band does not set the rate for a community whose work ran on writing, and it does not say which communities those were.  The prior says what to expect before the evidence about a person is weighed.  It does not replace that evidence, and it cannot by itself give a probability that a named person composed a book.

Part 3: What the documents show

Egypt: the formula of the person who could not write

Egypt preserves an exceptionally large body of everyday documents because its dry conditions preserved papyrus.  Contracts there routinely end with a subscription in another hand.  James Moulton and George Milligan, compiling the first dictionary of New Testament Greek from the papyri, found ἀγράμματος "of constant occurrence in the formula used by one person signing a deed or letter on behalf of another who cannot write".⁷  A few first-century examples show who did the writing:

1. A village scribe. In AD 91 at Bakchias, after a declaration by a man named Peteuris: "ἔγραψεν Ἀφροδίσιος γραμματεὺς τῆς κώμης [ὑπὲρ αὐτοῦ] μὴ εἰδότος γράμματα," "Aphrodisios, scribe of the village, wrote [for him], as he does not know letters." The words in brackets are restored by the editors.⁸
2. A trade group's scribe. In AD 16 at Euhemeria, "Maron, scribe of the cattle-keepers of Euhemeria," wrote for a man "because he does not know letters."⁸
3. A husband. In AD 99 at Theadelphia, a woman named Charition acted "with her husband as guardian", and he wrote: "Ἥρων ἔγραψα καὶ ὑπὲρ τῆς γυναικός μου μὴ εἰδυίης γράμματα," "I, Heron, also wrote for my wife, who does not know letters."⁸
4. A son. In AD 95 a woman named Aphrodisia, who had leased out an oil press, acknowledged her rent, and her son wrote "for my mother, who does not know letters", in a line partly restored by the editors.⁸

Three points follow.  First, owning property and being party to documents did not require writing:  the woman with the oil press was a lessor.  Harris notes the same pattern in Egypt, "lenders to be more literate than borrowers, lessors more literate than lessees", and adds that "the illiterates who were involved in the transactions just referred to were, in spite of their illiteracy, making use of the written word."¹  Second, the writer was usually someone close.  Hezser summarises Herbert Youtie's study of these substitute writers:  "in the large majority of cases, the person employed by the illiterate as a substitute signature-writer stands in a family or friendship relationship with him or her."⁵  Third, "does not know letters" in a Greek subscription is a statement about writing Greek, made for a legal purpose.  Youtie found that the line between literate and illiterate was drawn low:  Petaus defended a fellow clerk accused of being ἀγράμματος on the ground that he signed his papers, insisting on "the narrowest possible definition of the word, one that would cover both Ischyrion and himself, as it would cover also a great many others, who at best could write their names and perhaps a few words in addition."³

Petaus: what one case proves

Petaus was village clerk at Ptolemais Hormou in Egypt in the 180s.  Youtie's description of his signatures is blunt:  "The writing in all of them is stiff, awkward, uneven, kept on the line with obvious effort. Petaus is totally without skill as a writer. He is indeed not a writer at all in any proper sense, but a man copying a model or repeating it from memory."³  His practice sheet survives.  He copied his formula correctly four times, dropped a letter on the fifth, and repeated the mistake for seven more lines because he was copying his own copy.  "He was unable to introduce a correction at any point because he could not read what he wrote".³

Petaus proves that holding an office whose work was writing did not guarantee being able to read.  He does not prove that most clerks were like him.  He is famous because he is extreme.  He also belongs to second-century Egypt, and he was a village clerk, not a tax collector.

Judaea and Arabia: one document, many hands

The richest Jewish evidence comes from caves near the Dead Sea, where refugees of the Bar Kokhba revolt hid family papers about AD 132-135.  The archive of Babatha, a Jewish woman from Maoza in the Roman province of Arabia, contains documents in Greek, Aramaic and Nabataean.  Hezser counts "Out of the 35 documents associated with Babatha 23 are double documents written in Greek, Aramaic, and Nabatean."⁵

One deposition of October 125 shows how many people one document could involve.⁸

1. The body was written by a professional: "ὁ δὲ γράψας τοῦτο Θεενας Σίμωνος λιβράριος," "the one who wrote this was Theenas son of Simon, librarius."
2. Babatha's own statement was written for her: "Ἐλεάζαρος Ἐλεαζάρου ἔγραψα ὑπὲρ αὐτῆς ἐρωτηθεὶς διὰ τὸ αὐτὴν μὴ ε[ἰ]δένα[ι] γράμματα," "I, Eleazar son of Eleazar, wrote for her at her request, because she does not know letters."
3. Her guardian for the case, Judah son of Khthousion, "being present, subscribed" himself. Wise classes his hand as a "Practiced Hand".²
4. Witnesses signed in Aramaic and Nabataean script.

Two years later, a land declaration Babatha filed at Rabbath-Moab survives in a Greek copy.  Her sworn subscription is headed "ἑρμηνεία ὑπογραφῆς", "translation of the subscription", and ends "Ἰουδάνης Ἐλαζάρου ἐπετρόπευσα καὶ ἔγραψα ὑπὲρ αὐτῆς", "I, Judanes son of Elazar, acted as guardian and wrote for her."⁸  The original subscription was in another language and another hand, and the Greek copy translates it.

Babatha was a woman of property who litigated, registered land with a Roman prefect and kept her papers for years.  She did not write Greek.  The documents do not say whether she could read Aramaic.

The leader who did not sign

Yigael Yadin, who excavated the Cave of Letters, observed of the letters sent in the name of the revolt's leader to his commanders at En-gedi that "no two letters bear the same handwriting, so they must have been dictated in an over-staffed office".  He added:  "Not one letter is signed by Shimeon bar Kosiba personally."⁹  The observation is about one group of letters found in one cave.  Different hands show several writers;  "dictated" and "over-staffed" are Yadin's inferences, since a scribe might also have composed a letter from instructions.  A Greek letter from one of his officers, Soumaios, explains its own language:  "ἐγράφη δ[ὲ] Ἑλληνιστὶ διὰ τ[ὸ ἡ]μᾶς μὴ εὑρηκ[έ]ναι Ἑβραϊστί ἐ[γγρ]άψασθαι," "it was written in Greek because we have not found [the means] to write in Hebrew."⁸  The reading of that sentence has been revised, and what was lacking, a writer or something else, is not stated.  Either way, in the middle of a nationalist revolt, a Jewish commander's letter was written in Greek.

These are second-century documents of propertied families in Judaea and Arabia.  They show how writing worked there:  layers of hands, several languages, and professionals and relatives writing for people who could not.  They cannot show how many Galilean fishermen of AD 30 could read.

Before AD 73: names on bone boxes and storage jars

The legal papers above are mostly from the second century.  First-century Judaea left a different kind of writing:  short, informal, and often not by professionals.

1. Ossuaries. Jerusalem families of the period gathered the bones of their dead into stone boxes. The standard corpus of Jerusalem inscriptions counts "only about 600-650" inscribed ossuaries, "and most of these inscriptions only identify the name(s) of the deceased".²⁸ The boxes favoured families of means, but the corpus editors add that "the many crudely made and undecorated boxes found in the caves attest to the efforts by the humbler classes to provide themselves with a decent burial."²⁸ Hezser, who argues for low literacy, gives the share in L. Y. Rahmani's catalogue of the state collections as about 26 percent, "233 out of a total of 897 items", and says of the writers: "They were not written by professional scribes but probably by the deceased’s relatives or by a tomb custodian employed by them."⁵
2. A workman's list. One ossuary lid from Bethphage carries eleven Hebrew names with numbers. Hezser follows Eleazar Sukenik in taking it as the wage list of the workshop that made the boxes, "a private scrawl, written by a master workman for his own needs".⁵
3. Masada. Yigael Yadin's summary of the finds from the rebels' last stronghold, occupied from AD 66 to 73, reports "More than seven hundred ostraca of various kinds", mostly in Hebrew or Aramaic. "About half of the ostraca found are a kind of tag with single or a combination of several letters", which he linked to food rationing during the siege. Others are lists of names with a number beside each, and owners' names on store jars. A group of eleven tags, each with one name, "are all written in the same handwriting".²⁹
4. Two readings of the same finds. Millard takes this material to show that "many non-professionals wrote to some extent, keeping accounts, putting names on pots, ossuaries and other possessions".²² Hezser reads it more narrowly: "All of these handwritten names represented a non-professional writing practice which was very limited in scope. Whether the writers were able to write more than the names is difficult to determine." Many are crude, and "the writers may have been barely literate."⁵

The two scholars agree on the facts.  In first-century Judaea, writing a name or a label in Hebrew, Aramaic or Greek was not confined to professional scribes:  relatives scratched names on the boxes that held their dead, and a workman kept a list.  That is direct evidence from before the revolt for the lower end of the skill range, which the second-century legal papers and the Egyptian formulas show only by analogy.  It does not show how many people could do it, since one writer could serve a whole family, and the Masada tags show one hand serving a whole group.  It says nothing about reading books.  And it comes from Jerusalem and Masada, not Galilee.

Galilee: the missing evidence

Four different claims are often run together here, and they need to be kept apart.

1. Little first-century writing has been found in Galilee. This is true of the evidence published by the early 2000s. Mark Chancey's survey of the region's inscriptions found "very, very few Galilean inscriptions from the Early Roman period". For the lifetime of Jesus, "The only extant inscriptions definitely produced within Galilee ... are the bronze coins of Herod Antipas — and a lead market weight from Tiberias from 29/30 CE naming its agoranomos, Gaius Julius." For the rest of the century he adds a few items, among them an ostracon from Yodefat with Semitic letters from before AD 67.³⁰ This is Chancey's 2005 inventory, not a complete inventory of finds now known.
2. The absence has an archaeological cause as well as a social one. Millard notes that most inscriptions and documents of the period come "from Jerusalem and Judaea. Galilee, where Jesus gave much of his teaching, has yielded very little", and attributes this mainly to "the paucity of Herodian period material from anywhere in Galilee", where later buildings sit on top of earlier layers.²²
3. The absence does not measure Galilean literacy. The lack of surviving papyri cannot by itself show that Galileans did not write, and Egyptian or Judaean documents do not measure Galilean skills directly. They are analogies, useful for the kinds of writing that existed, not rates.
4. The analogies may point in different directions. Hezser warns that the Egyptian analogy may overstate the need for writing in Palestine, because Roman administration there "seems to have been rudimentary only", so "one may assume that much less scribes were needed than in Egypt."⁵ On the other side, Galilee lay on a main road (Part 4), and Josephus has the chief magistrate of Tiberias address a crowd early in the revolt "With a copy of the laws of Moses in his hands".¹⁰ That shows a book of the law in a Galilean city, not how many could read it.

Two later studies qualify the picture.  Scott Charlesworth's 2016 author abstract calls the inscriptional evidence "patchy and unrepresentative" and argues that counting inscriptions cannot measure the use of Greek.  He draws instead on bilingualism, cities, the economy and rural-urban contacts, concluding that knowledge of Greek was probably common.  That is an argument about language use, not a measured proportion able to read books.³¹  Pablo Ozcáriz Gil's 2025 author abstract reports an inscribed intaglio found at Magdala in 2011, with a name read as Nikoma(chou) and "a probable date for the intaglio of the 1st century CE".³²  A portable inscribed object found in Galilee does not establish local manufacture or its owner's ability to write.  Only the abstracts of these two studies are assessed here, so neither their detailed arguments nor the proposed dating has been independently checked.  Excavation since 2009 has also uncovered first-century synagogue buildings in Galilee itself, at Magdala on the lake shore and at Tel Rekhesh in the countryside (Part 4).  They show places built for public reading, not how many people read.

So the article's judgement on Galilee is a weak one:  restricted book-reading is a reasonable expectation from the regional models and the Judaean documents, but no representative Galilean sample establishes its rate.  Greek language use, the presence of writing and the proportion able to compose literary Greek remain different questions.

Part 4: Who got schooling

What Josephus says, and what kind of statement it is

Josephus, a priest from Jerusalem writing in Rome late in the first century, gives the ideal in two works.  In Against Apion 2.204 the law "orders that they shall be taught to read, and shall learn both the laws and the deeds of their forefathers".¹⁰  In the Antiquities (4.211), retelling the laws of Moses, he has Moses say that "the children too shall learn the laws first" (μανθανέτωσαν δὲ καὶ οἱ παῖδες πρῶτον τοὺς νόμους).³³  And in Against Apion he boasts that any Jew asked about the laws "would repeat them all more readily than his own name."¹⁰

These passages are positive evidence that teaching children the law, and in Against Apion teaching them to read, was a recognised ideal in Josephus' world.  Their purpose limits what they show.  Against Apion defends Jews to Greek readers, the Antiquities passage restates a law rather than describing practice, and knowing the laws by heart is not the same as reading them.  They do not say how many families met the ideal.  Craig Evans, who argues for a literate Jesus, accepts that Philo and Josephus "are painting idealistic pictures and perhaps have in mind affluent families", while arguing that the pursuit of literacy was not "limited to the upper class".¹³  Josephus' own education was exceptional by his own account:  "about fourteen years old, I won universal applause for my love of letters; insomuch that the chief priests and the leading men of the city used constantly to come to me for precise information on some particular in our ordinances."¹⁰  It is a self-portrait, and not a modest one.  He also says he needed help to write Greek well, a point taken up in Part 7.

The rabbinic school report: late, and still evidence

Many popular accounts say every Jewish village had a school.  The source is the Babylonian Talmud, Bava Batra 21a.  In a ruling about whether neighbours may stop a teacher from teaching children in a shared courtyard, the Talmud cites "Rav Yehuda says that Rav says" that the high priest Yehoshua ben Gamla deserves to be "remembered for the good".  At first fathers taught their sons;  then teachers were set up in Jerusalem, and later in each region for pupils of sixteen or seventeen;  "until Yehoshua ben Gamla came and instituted an ordinance that teachers of children should be established in each and every province and in each and every town, and they would bring the children in to learn at the age of six and at the age of seven."¹¹

The report has real weight and real limits.

1. For it: Yehoshua ben Gamla is usually identified with a real person: Josephus names Jesus son of Gamaliel as high priest in the 60s (Antiquities 20.213, 20.223).³³ A later tradition attached to a real high priest may preserve a memory of some reform he sponsored.
2. Against reading it as a first-century record: it is told in the name of Rav, a Babylonian teacher of the early third century, about a man who held office more than 150 years earlier. Hezser dates the passage this way: "Even if the text is tannaitic, it cannot be dated earlier than the end of the second or the beginning of the third century C.E., that is, hundreds of years after the events it describes purportedly happened."⁵ Its tidy three-stage history, and its interest in the legal question at hand, are the marks of a story that explains a later institution.
3. What it does not claim: the schools it describes teach boys Torah. It says nothing about writing, about Greek, or about girls, and even taken at face value it describes an ordinance, not how far it was carried out in Galilean villages by the 20s.

Hezser concludes more broadly that "Rabbinic ideals reflected in literary sources cannot be taken as historical evidence about the real situation", and that the schools the rabbis describe "are likely to have taught reading skills only", for reading the Torah aloud.⁵  Evans, from the other side, calls the rabbinic school material "idealistic and tendentious", and rests his case for Jesus on other grounds.¹³  Keith's judgement is a fair summary of what can be shown:  "There is no concrete evidence that literate education was typical for first-century Palestinian Jews (or the rest of antiquity). For those Jews who did receive an education, Torah reading was a prime focus, with writing instruction being reserved for later specialized scribal training."²⁴  John Meier thinks a village synagogue with "some educational program for Jewish boys is a likely hypothesis" for Nazareth;²⁶ that is a reasonable possibility, argued from Jesus' later skill, not a documented school.

Synagogues as places of reading

There is first-century evidence that reading aloud was a public institution.  A Greek inscription from Jerusalem, dated by its lettering to the Herodian period or the early first century, records that "Theodotos son of Vettenos, priest and archisynagogos ... built the synagogue for the reading of the Law and teaching of the commandments".²⁸  Luke 4:16-20 depicts the same practice in Nazareth.  Public reading spread the content of scripture to people who could not read it, which is why Keith stresses that "knowledge of what the texts say, even intricate knowledge, does not necessitate the ability to read them for oneself."²⁴  It also means that being able to quote scripture is weak evidence of being able to read it.

Excavation since 2009 has found such buildings in Galilee.  At Magdala, the only town on the western shore of the lake before Tiberias was founded, the Israel Antiquities Authority excavators describe a first-century synagogue of "two large rooms—a vestibule and a reading room—and one smaller room", the small room perhaps "used to store the scrolls".  A coin of AD 43 lay in the foundation of the main hall's mosaic, and the building was "probably deserted at the time of the Great Revolt" of 66.³⁴  At Tel Rekhesh in Lower Galilee, a Japanese expedition excavated a benched hall dated by its lamps to the first century, on a settlement of half a hectare.  It could seat "about 50 people", with the reader presumably in the centre of the room.³⁵  Takahiko Yamano, one of its excavators, concludes:  "The existence of a synagogue in a small rural settlement such as Tel Rekhesh, and not only in the urban centre of Magdala and the large town of Gamla in the Golan, indicates that the Galilean Jews sought a structure for meeting on the Sabbath as early as the first century CE."³⁵  He also reports that the excavators agree the hall "belongs to a part of the farmstead owned by a wealthy family".³⁵

These finds test one input in Part 2.  Bar-Ilan set rural literacy at zero.  A hall built for reading the law in a rural settlement implies that someone read there, so zero is unlikely as an exact figure for the countryside.  But the Rekhesh hall stood on a wealthy family's farm, and one reader to a congregation is the case the Tosefta legislates for.  On Bar-Ilan's own method a congregation of about fifty with one reader is about 1 percent, and 1 percent rural literacy moves his total from 0.4-2.5 percent to 1.1-3.2 percent.²⁵  The buildings show that public reading reached Galilean towns and at least some of the countryside.  They do not show many readers, and they fit the picture of readers concentrated in families of means.

City and country

Hezser's first reason for a low Jewish literacy rate is "the largely rural character of Palestine";  "Village-dwellers ... will have rarely needed writing. If the need arose, they could always go to the next town or city."⁵  Harris, comparing early modern New England, notes that "very small settlements and truly rural patterns of living are a hindrance to elementary education".¹  But Hezser herself warns that "A strict dichotomy between city and countryside ... should be avoided", and that "the large majority of the population in both rural and urban areas will not have been able to read literary texts at all".⁵  Galilee was not sealed off:  Millard notes that "the main route from Egypt to Damascus ran along the north-west shore of the Sea of Galilee".²²  But no inscription or document tells us how many people in its lakeside towns could read.

Men and women

The direct evidence on women is thin and comes from a narrow slice of society, but what there is points one way.

1. Wise's documents. All ten women who were parties to the legal papers were "unable to sign their names".² They come from propertied families, mostly in Judaea and Arabia in the early second century. This is Wise's classification, not a conclusion inferred here from a missing signature alone. His tables distinguish people signing in a different language from those he classes as unable to sign. The classifications have not been audited person by person, and they do not establish inability to read. Ten selected women do not give a rate for Judaea, still less for Galilee or the whole eastern Mediterranean.
2. Babatha. The best-documented of them was a landowner who went to court, and in Greek documents others wrote "for her ... because she does not know letters".⁸ The documents do not say whether she could read Aramaic.
3. Egypt. Harris cites Youtie's conclusion from the papyri that "illiteracy in women was traditional at all levels of society", and comments: "This is hardly true about the daughters of the very best families, but otherwise it cannot be faulted."¹ The Egyptian formula itself records women, like Heron's wife Charition, for whom husbands or sons wrote (Part 3).
4. The exception. One witness in the Babatha documents, a woman named Julia Crispina, subscribed in Greek in Wise's level 3 hand.² Tal Ilan, as Wise reports, proposed that she was a Herodian princess. The identification is Ilan's argument from her name and connections, not a statement in the document, but on any identification she is a woman who wrote Greek well, and she is the only one in the sample.
5. Schooling. Hezser notes that the rabbinic schools' aim, to train Torah readers, "excluded girls from the outset".⁵ That describes the schools in the rabbis' picture. It does not show that no girl learned at home. Meier's verdict is that "women were rarely given the opportunity".²⁶

The article's grade:  women's writing was rare in the families that left documents, and plausibly rarer below them;  that is moderately well supported for propertied Judaean and Arabian families and is an inference elsewhere.  Women's reading is almost unrecorded.  The same caution applies to slaves and peasants, for whom no direct evidence survives in Wise's sample.

Ordinary does not mean uneducated

Status and skill were related but not identical.  A craftsman (τέκτων, Mark 6:3) was not at the bottom of society:  Keith concludes that craftsmen "could be literate to a degree associated with their craft and were not the poorest of Palestinians."²⁴  And Wise found propertied men who wrote at an intermediate level for practical reasons.  "Lazar b. Joseph, writer of Mur 21, exemplifies such a person. He copied out his daughter’s ketubbah even though he possessed only limited graphic skill."²  Social standing sets a prior.  It does not settle what any one person learned.

At the same time, writing for a living did not carry high status.  Josephus reports that Herod's sons threatened to make their half-brothers "village clerks" (κωμῶν γραμματεῖς), "mocking them for their careful education".³⁶  Keith comments that village scribes "were already among the lower rungs of literate society".²⁴

Part 5: "Unlettered and ordinary": Acts 4:13

What the verse says

After Peter and John are arrested for preaching in the Temple, they speak before the high-priestly council.  Luke reports:  "Θεωροῦντες δὲ τὴν τοῦ Πέτρου παρρησίαν καὶ Ἰωάννου καὶ καταλαβόμενοι ὅτι ἄνθρωποι ἀγράμματοί εἰσιν καὶ ἰδιῶται, ἐθαύμαζον," "Seeing the boldness of Peter and John, and perceiving that they were unlettered and ordinary men, they were amazed".³⁷

Four features of the sentence matter before any word is defined.

1. It is about Peter and John only, not the Twelve, and not the disciples in general. Bart Ehrman names the two correctly, calls ἀγράμματοι "the ancient word for illiterate", and adds his own wider judgement: "Nothing in the Gospels or Acts indicates that Jesus’ followers could read, let alone write."³⁸ The verse supports a claim about two men; the wider claim rests on silence and on the population models, which are applied to the disciples through their trade (see "What kind of fishermen" below).
2. It reports what the council "perceived" or found out (καταλαβόμενοι). Luke uses the same verb for Peter's realisation in Acts 10:34 and for Festus' finding after a hearing in Acts 25:25, so it does not say how the council knew: from how the men spoke, from who they were known to be, or from inquiry. Luke does not correct the council's judgement, and the scene's force depends on the judgement being true, so the narrator treats it as accurate.
3. The council's surprise is at their παρρησία, their boldness in speech. The scene shows no reading or writing. That makes the contrast "bold, though untrained" natural, but it does not tell us which kind of training the council had in mind.
4. The words come as a pair, and the pair has to be read together.

The case for "unable to read and write"

The strongest case for this reading is the papyri.  Moulton and Milligan argued in 1915 that "The great frequency of ἀγράμματος, invariably in this sense," that is, a person "who cannot write", "suggests that the sneer in Ac 4 is intended to picture the Apostles as “illiterate,” and not merely “unversed in the learning of the Jewish schools” (Grimm)."⁷  Literary Greek has the literal sense too.  Epictetus compares a lazy student to an ἀγράμματος who says "tell me what to write" when given a name to spell.³⁹  Craig Keener, who prefers the other reading, begins by granting that ἀγράμματος "literally means “illiterate.”"¹⁴  Thomas Kraus wrote a study devoted to the phrase.  Its 1999 abstract argues that the documentary papyri show illiteracy and non-specialism "without any social or negative impact", and that they "make the terms ‘illiterate people’ and ‘laymen’ acceptable".⁴⁰  The abstract of his revised version (2007) says that "'Illiterate' may be accepted as the possible meaning of άγράμματος in Acts 4.13, provided that other meanings are not excluded by this."¹²  His full argument from the papyri could not be read for this article, so only what the abstracts say is reported here.

The case for "without scribal training"

The strongest case for this reading is the word's broader use and the pairing.  Keener goes on:  "The term more broadly indicated simply lack of formal education."  Here, he suggests, it may stand in contrast to γραμματεύς, "scribe", "and hence mean one without scribal training in reading and interpreting Scripture."¹⁴  Evans makes the same contrast with the scribe and says that the comments in John 7:15 and Acts 4:13 "should not be taken to imply that Jesus and his disciples were illiterate. In fact the opposite is probably the intended sense, as most commentators rightly interpret."¹³  Harris states the general rule that literary uses of the word "veer" between "uncultured" and "incapable of reading and writing", and that "Such expressions have to be interpreted case by case."¹¹

Luke's and John's use of the related noun γράμματα fits this, without settling it.  In Acts 26:24 Festus tells Paul, "τὰ πολλά σε γράμματα εἰς μανίαν περιτρέπει," "your great learning is driving you mad."  In John 7:15 the crowd asks of Jesus, "Πῶς οὗτος γράμματα οἶδεν μὴ μεμαθηκώς;", "How does this man know letters, having never studied?"  These show that the noun can mean learning in these writers.  They do not show what the adjective means in Acts 4:13, since a word family can carry both senses.

The second word, ἰδιώτης, means a private person or a layman, as against an official or an expert.  Josephus uses the feminine for a woman who was not of royal birth (Jewish War 1.479).³⁶  Paul calls himself "ἰδιώτης τῷ λόγῳ", untrained in speech (2 Cor 11:6), although he plainly could read and write.  A layman can also be illiterate, so the word does not exclude the first reading.  Keener adds that in rhetorical contexts the word marked a layman in rhetoric, and concludes that "Plainly, the apostles lacked formal advanced schooling, especially in rhetoric".¹⁴  Kraus's abstract warns against loading that onto the verse:  "in Acts 4:13 nothing is said about rhetoric, social class or the like."¹²  Neither specialist's full reasons on this point could be compared here, so the article takes no side on whether rhetoric is in view.

Weighing them

The main readings compared here overlap in their portrait of Peter and John as lacking recognised scribal credentials.  That shared portrait does not make literacy and formal schooling identical:  skills differ by language, can be acquired outside school and can be lost.  The verse can be graded in two steps.

1. The minimum. In the readings compared here, Luke presents Peter and John as men without recognised scribal credentials and treats the council's judgement as accurate. That is a secure statement about his portrait, rather than a full biographical record of their education. The verse names no school or institution. Keener notes that "formal rabbinic ordination was almost certainly not mandatory in this period"; that does not establish that no formal training existed.¹⁴
2. Beyond the minimum. Whether the council also meant that the two could not read or write cannot be settled from the verse. The literal and documentary sense supports it; the pairing with "laymen" and the speech scene fit "untrained" at least as well; and specialists who have studied the phrase disagree. A Galilean fisherman with no scribal training might well also have been unable to write, but that is an inference from the population models, not something the verse says.

Verdict on the verse:  secure that Luke portrays Peter and John as lacking recognised scribal credentials;  open whether he also means they could not read or write;  silent about the other ten, and about Greek.

What kind of fishermen

The population models are applied to the disciples through their trade, so the trade matters.  Bart Ehrman states the usual picture:  "From the Gospels we learn that the disciples of Jesus, like him, were lower-class peasants from rural Galilee. Most of them—certainly Simon Peter, Andrew, James, and John—were day laborers (fishermen and the like)".  His summary:  "Lower-class, illiterate, Aramaic-speaking peasants from Galilee."³⁸

The Gospels he cites describe the four fishermen differently.  Mark, usually dated the earliest, has James and John leave "their father Zebedee in the boat with the hired men" (μετὰ τῶν μισθωτῶν, Mark 1:20), and has Jesus go to "the house of Simon and Andrew" in Capernaum (Mark 1:29).  Luke has Jesus teach from "one of the boats, which was Simon's", and calls James and John Simon's "partners" (κοινωνοί, Luke 5:3, 5:10).³⁷  On these texts the four came from families that owned boats and paid day labourers.  They were not the day labourers.

The papyri show what such families did.  John Kloppenborg of the University of Toronto, a critical historian of Christian origins, compared the Gospel fishing stories with the documentary evidence for fishing in Roman Egypt:  leases of fishing rights, boat taxes, tax rolls and daily logs of catches.  He concludes:  "We should imagine fishermen as belonging to the “above-subsistence level”, though certainly not “upper class”."¹⁵  Their trade was bound up with paperwork and officials:  "fishing required constant and continuous interaction with agents of the lessor and local administrators."¹⁵  Three limits come with this.  Kloppenborg himself asks, "Were Galilean fishermen like those we know from Egypt? Probably not in all respects."¹⁵  K. C. Hanson, whom he quotes, holds that boat owners and hired men alike "were ‘peasants’ in the broad sense".¹⁵  And Kloppenborg warns that Matthew and Luke tend "to shift the social and cultural registers of Jesus and his immediate entourage upwards",¹⁵ so Luke's boats and partners carry less weight than Mark's hired men.

A trade that ran on documents did not make its members writers.  The fishing lease Kloppenborg calls "Typical",¹⁵ from Oxyrhynchus in AD 161, rents a year's catch to three men and ends:  "Σαρᾶς Φιλοξένου ἔγραψα ὑπὲρ αὐτῶν μὴ εἰδότων γράμματα," "I, Saras son of Philoxenos, wrote for them, since they do not know letters."⁸  The first of the three lessees is named Philoxenos, so the writer may have been his son;  the document does not say.

So the evidence moves Peter, Andrew, James and John up from the bottom of the population to a middle slice:  owners of boats in a trade that put leases, receipts and tallies in front of them every season, and families that could find a writer when they needed one, as the Oxyrhynchus fishermen did.  It does not show that any of the four could write, and for Peter and John Acts 4:13 still points the other way.  The population rates of Part 2 bear on this slice with less force than on day labourers, by an amount nobody has measured.

Ehrman's further sentence, "Nothing in the Gospels or Acts indicates that Jesus’ followers could read, let alone write",³⁸ is about the Galilean disciples, and for them the texts are indeed silent apart from Acts 4:13.  The wider circle those books describe is different.  Joseph of Arimathea is "a respected member of the council" (Mark 15:43), whom Matthew and John call a disciple (Matthew 27:57, John 19:38).  Joanna is the wife of "Chuza, Herod's steward" (Luke 8:3).  Acts adds "a great many of the priests" (Acts 6:7), believers "from the party of the Pharisees" (15:5), the Levite Barnabas (4:36), Manaen, "brought up with Herod the tetrarch" (13:1), and Paul.³⁷  None of these verses says that a named person read.  They place followers in the groups where, on every model in Part 2, reading was concentrated.  When the Jerusalem leaders needed a letter, "the apostles and the elders" sent one in their own names (Acts 15:23), and the text does not say who held the pen.  These are the narratives' own claims, and a reader who doubts Luke will weigh them less.  But for the question whether the movement could produce writing, the unit is the group, not one fisherman, and the group as its own sources describe it included people from the reading classes.

Origen's use of the verse

In the third century the Christian scholar Origen answered the pagan Celsus, who had called Jesus' followers "the very wickedest of tax-gatherers and sailors" (τελώνας καὶ ναύτας τοὺς πονηροτάτους).⁴¹  Celsus' charge, in the words Origen quotes, was about their trade and character.  Origen answers:  "who is there on seeing fishermen and tax-gatherers, who had not acquired even the merest elements of learning (as the Gospel relates of them, and in respect to which Celsus believes that they speak the truth, inasmuch as it is their own ignorance which they record), discoursing boldly ... would not inquire whence they derived this power of persuasion".⁴¹  The Greek is ἁλιεῖς καὶ τελώνας μηδὲ τὰ πρῶτα γράμματα μεμαθηκότας, "fishermen and tax collectors who had not even learned their first letters", and the word Crombie renders "ignorance" is ἰδιωτεία, lack of training.¹⁶  Keith's English presents this phrase as part of Celsus' accusation.²⁴  The wording is Origen's;  what Origen says about Celsus is that Celsus believed the disciples' own account of their ἰδιωτεία.  Whether Celsus himself went further than "tax-gatherers and sailors" cannot be checked, because his book survives only in Origen's quotations.

Three things limit what Origen's sentence proves.

1. It is a later reading of the same Gospel texts, as Origen himself says: "as the gospel records about them". The word he picks up, ἰδιωτεία, is the noun of Acts 4:13's ἰδιῶται. Nothing in the passage suggests a source beyond the Gospels and Acts, though the passage cannot show he had none.
2. It serves an argument. Origen's point is that if Jesus had chosen men "wise" by popular standards, his preaching would look like a philosophical sect; their success without "the arts of Grecian dialectics or rhetoric" shows divine power.⁴¹ He has a reason to state their lack of schooling as strongly as he can.
3. The same Origen held that the first Gospel was written by Matthew, "once a tax collector, afterwards an apostle", in Hebrew letters, and that Mark wrote "as Peter guided him".¹⁷ So Origen did not think his phrase ruled out a role for these men in writing Gospels. He does not say how he reconciled the two. He might have meant lack of Greek schooling only, or thought of later learning, of writing in Hebrew, or of help from others, as with Peter and Mark. The sentence shows how a learned third-century Christian read Acts 4:13. No independent information about a disciple's schooling can be identified in this passage.

The same chapter has a second detail that matters for Part 6.  Origen says that of the twelve apostles "Matthew alone was a tax-gatherer", and that Levi "was not of the number of the apostles, except according to a statement in one of the copies of Mark's Gospel."⁴¹  The Greek says "some of the copies" (τινα τῶν ἀντιγράφων).¹⁶

Part 6: The tax collector, as a test case

What the texts say

The Gospel of Matthew calls the tax collector at the booth "Matthew" (Matt 9:9) and lists "Matthew the tax collector" among the Twelve (Matt 10:3).  Mark and Luke tell the same call story of a tax collector named Levi (Mark 2:14;  Luke 5:27).  Whether Levi and Matthew are one man, and whether the apostle Matthew had anything to do with the Gospel that bears his name, are questions for Who Wrote the Gospels?.  This article asks a narrower question:  what does the trade of toll or tax collector imply about a man's ability to read, write and compose?

What a toll collector wrote

The surviving receipts from Egypt show what the job produced.  An ostracon from Thebes dated 10 September 84 reads, in the editors' regularised text:  "Φατρῆς καὶ μέτοχοι τελῶναι ... Πουορενούβει χαίρειν. τετελώνηται ἐν Τωὺτ φοινίκων ἀρτάβαι λ," "Phatres and partners, tax collectors ..., to Pouorenoubis, greeting.  Duty has been paid at Toou on 30 artabas of dates."⁸  The editors record where the writer spelled phonetically:  κὲ for καί, τετελώνητε for τετελώνηται, φυνιξ for φοινίκων.  That is the genre:  a name, a formula, a commodity, a quantity, a date.

Two cautions limit what this shows.  First, the receipt does not say who wrote it.  "Phatres and partners" were a company of lessees, and any of them, or a clerk they paid, may have held the pen.  Second, Egypt is not Galilee.  Hezser points out that in the Roman province of Arabia, as Benjamin Isaac argued from Babatha's tax papers, tax collection "was simple and crude", and Palestine's administration "seems to have been rudimentary only".⁵  The Egyptian paperwork may overstate how much writing a Galilean toll collector needed.

The strongest case on the other side is Millard's.  Taxes "had to be assessed and collected", and at Jericho "the tax-collector Zacchaeus promised to repay four times any sum he had taken wrongfully, a promise which implies he had accounts he could check (Lk. 19.1-10)."²²  That is a fair inference about the job:  someone in a toll business kept accounts.  It is not evidence about whether the collector kept them himself.

Two arguments that prove too much

"He was a tax collector, so he could write a Gospel."  A toll collector plausibly needed to read and write numbers, names and short formulas, in Greek or Aramaic.  That is a long way from composing a narrative of 18,329 words, the length of Matthew in the SBL Greek text.⁴²  The trade raises the chance of documentary literacy.  Whether it also raised the chance of literary skill is not measured.  A man who handled records daily, mixed with officials and earned well may have been more likely than a fisherman to have had more schooling, or to have gone on learning;  equally, the job could be done with formulas and an assistant.  An unknown effect is not a zero effect, and it is not a demonstrated one either.  Whether the Greek Gospel of Matthew is a translation, and whether its author used Mark, are questions about the book's composition, handled in the sibling article.

"Petaus was a clerk and could not read, so tax collectors could not."  Petaus was a second-century Egyptian village clerk, and he is notable because he was unusual.  One extreme case cannot fix the skill of every holder of a different office in a different province.  Arguments that assign a whole occupation one "rung" of literacy from a single case use Keith's individual sense of "craftsman's literacy" as if it were a measured level for that occupation.

What survives of the argument

1. The trade made it likely that a toll collector handled documents and numbers. Who wrote them, the collector or an assistant, is not recorded for any first-century Galilean.
2. Origen's grouping of "fishermen and tax collectors" as unlettered is a third-century rhetorical reading of Acts 4:13, by a writer who also credited the tax collector with a Gospel (Part 5).
3. Neither the trade nor the receipts can show that a particular tax collector composed a literary book. Nor can they show that he could not dictate, take notes, or stand behind a book written down by someone else. Writing in his own hand, composing aloud and authorising another's draft are separate claims, and each needs its own evidence.

Part 7: Secretaries, dictation and notes

Five different jobs

"He used a secretary" covers several different jobs.  E. Randolph Richards, whose The Secretary in the Letters of Paul (1991) remains the main study, built a classification from Greek and Latin private letters and placed the roles "on a spectrum composed of four basic categories, ranging from more author-controlled to more secretary-controlled roles."¹⁹  The list below is this article's own working list, adapted from Richards' spectrum with translation added.  It is a set of questions to ask of a document, not a finding that every job is attested in every case.

1. Taking dictation word for word. The author chooses every word; the writer records them.
2. Taking dictation fast, or in shorthand. Still the author's words, but the writer may smooth what he hears.
3. Editing. The writer corrects grammar and style from a draft or from dictation.
4. Translating. The writer puts someone's words or ideas into another language.
5. Composing on instructions. The writer drafts the letter from directions, and the author approves or sends it.

What matters for any one document is which of these happened, and that is seldom recorded.

What the ancient cases show

Dictation was common among the educated, and the writer's speed mattered.  Quintilian, the Roman teacher of rhetoric, complains about "the luxury of dictation which is now so fashionable".  When we write by hand, he says, the slowness of the hand gives us time to think, "whereas the presence of our amanuensis hurries us on";  and "if the amanuensis is a slow writer, or lacking in intelligence, he becomes a stumbling-block".²⁰  Cicero tells his friend Atticus that he took special care with a letter to Varro:  "Consequently I did not dictate it even to Tiro, who usually takes down whole periods at a breath, but syllable by syllable to Spintharus."¹⁸  Both cases are jobs 1 and 2:  authors composing aloud, with the writer recording.

Authorised composition by someone else happened, among friends.  In 48 BC, ill and in crisis, Cicero asked Atticus to "write to Basilus and to others to whom you think it proper - even to Servilius - in my name, and say whatever you think right."¹⁸  That is job 5, done by a trusted equal, not by a hired scribe.  Another letter shows the risk:  "If they remark on the absence of my signet or handwriting," Atticus should give an excuse (si signum requirent aut manum).⁴³  So letters in someone's name could be composed by another, and recipients sometimes looked for the sender's seal or hand to check.  The sibling article Ancient Citation and Authorship Conventions treats these letters, and what later readers made of the secretary idea, in detail.

A trained author still used help for literary Greek.  Josephus, who had studied Greek grammar, says the habit of his native tongue prevented "my attaining precision in the pronunciation",⁴⁴ and that he wrote the Jewish War "with the aid of some assistants for the sake of the Greek".¹⁰  That is help with a book, jobs 3 or 4, not with a letter.  Richards thinks "their assistance extended beyond mere translation."¹⁹  Josephus does not say how far.  Neither his accent nor his assistants shows that he could not compose:  he says he composed, and he names the help as being for the Greek.

A leader's letters could pass through several scribes.  Bar Kokhba's letters, as Yadin noted, show different hands, and none bears his signature (Part 3).⁹  That supports several writers.  Whether each letter was dictated word for word or drafted by a scribe from instructions, the hands do not say.

Tertius and Paul's own hand

Romans 16:22 is the one place in Paul's letters where the writer names himself:  "ἀσπάζομαι ὑμᾶς ἐγὼ Τέρτιος ὁ γράψας τὴν ἐπιστολὴν ἐν κυρίῳ," "I, Tertius, who wrote the letter, greet you in the Lord."  Richards is confident of the basic fact:  "That Tertius is the secretary is certain."¹⁹  He then suggests that "If Tertius was a tachygraphist", a shorthand writer, "it may explain why he was used to record this long letter".¹⁹  One step in that argument is that Rome was the likeliest city to have such writers, and of that step Richards says, "Still this is, of course, a tenuous assertion because of the nature of the evidence".¹⁹  He ends by saying only that "the possibility that Tertius was a professional secretary deserves a future investigation."¹⁹

Paul also marks where he took the pen.  The secure cases are in letters nearly all scholars accept as his:  "The greeting is in my own hand, Paul's" (1 Cor 16:21);  "See with what large letters I am writing to you with my own hand" (Gal 6:11);  "I, Paul, write this with my own hand:  I will repay it" (Philemon 19).  Two more notes are in letters whose authorship is disputed:  Colossians 4:18 repeats the greeting formula, and 2 Thessalonians 3:17 adds "this is the sign in every letter;  this is how I write".  Whether those two are Paul's is argued in Did Paul Write the Disputed Letters?;  this article does not rest anything on them.

Chris Keith reads the own-hand notes as a display of status:  Paul shows he "“knows letters” (and Greek ones at that)" and is "able to write, but able to avoid it as well and have an amanuensis write the bulk of the epistle."⁴⁵  Keith infers that Paul's Greek handwriting went beyond a formula, "Since he penned greetings (i.e., not only formulae)".⁴⁵  The inference is reasonable but should not be pushed.  A closing greeting is itself a conventional form, and "large letters" (Gal 6:11) has been read both as emphasis and as the hand of someone who wrote slowly.  The promise to repay in Philemon 19 is a short legal clause in the first person.  Together they show that Paul could write short connected Greek in his own hand.  They do not show how much he wrote, or that he wrote the rest.

The same habit appears in ordinary letters outside the New Testament.  At Vindolanda, about AD 100, Bowman found that "although most of our letters were not written by their authors but by clerks or amanuenses, the authors normally add the closing greetings (which show some variety of expression) in their own hands."  An officer's wife, Claudia Severa, ended a birthday invitation "with three lines in her own hand".²⁷  For people who could write, a secretary for the body and a line or two in one's own hand was the normal way to send a letter.

What these cases prove, and what they do not:

1. Paul used at least one secretary (Tertius) and wrote some lines himself. The own-hand notes are the positive evidence that he could write. Richards adds that Paul's letters "that clearly evidence a secretary do not contain any illiteracy formulae".¹⁹ That absence fits a writer who chose to use a secretary, but by itself it would prove little: an illiterate sender need not announce that fact in every letter.
2. Three different things have to be kept apart even for Paul: directing a letter, writing it out, and choosing its words. The notes show the first and some of the second. They do not show which of the five jobs Tertius or anyone else did. Richards himself describes the full range as possible, "stretching from filling in the author’s detailed instructions to fleshing out the most skeletal of directions, and even to the complete assumption of the contents by the secretary."¹⁹ A range of possibilities is not a finding about a particular letter.
3. The secretary hypothesis can be tested, and it competes with other explanations. Where a letter names a writer or co-senders, or where the style differs from the undisputed letters, a secretary is one possible cause among others, such as a different author, different subject matter or a different occasion. It should be weighed on the same evidence as they are, by how much of the actual pattern it explains. Because the extent of a secretary's freedom is unknown in each case, the hypothesis is weak when it is the only support for an authorship claim, and it is equally weak as a reason to dismiss style evidence. Did Paul Write the Disputed Letters? makes that comparison letter by letter.

Notes taken by listeners

Some ancient students wrote down their teachers.  Quintilian says two books on rhetoric circulated under his name without his consent.  "One is a two days' lecture which was taken down by the boys who were my audience. The other consists of such notes as my good pupils succeeded in taking down from a course of lectures".²⁰  He does not disown them entirely:  his new books, he says, would repeat some of the same material, change much and add very much (eadem aliqua, multa mutata, plurima adiecta).⁴⁶  That is a published teacher's own report on notes of his lectures, and the strongest case here.  Arrian, introducing the Discourses of the philosopher Epictetus early in the second century, writes:  "whatever I heard him say I used to write down, word for word, as best I could, endeavouring to preserve it as a memorial, for my own future use, of his way of thinking and the frankness of his speech."²¹  That is Arrian's description of what he did, not a recording anyone can check against Epictetus.  Quintilian also says Cicero's own speaking notes were "afterwards collected ... by Tiro".²⁰

These are real cases.  They involve educated students of famous teachers in the Greek and Roman schools.  Millard argues that hearers of Jesus who kept accounts may have done the same:  "To imagine some of them opening note-books they carried for their day-to-day business, perhaps hung at the belt, and jotting down a few of the striking sayings they had heard ... is quite feasible."²²  Millard rejects the idea of shorthand records of Jesus' words, which "cannot be entertained in the light of current knowledge".²²

How much weight this bears:

1. Notebooks existed. Millard describes hinged wax tablets as "the everyday writing material for tax collectors, administrators, businessmen and scholars in Hellenistic and Roman times", and adds: "works of literature were excluded; except in the schoolroom, those belonged on rolls. Their function as note-books is well-attested." He then reports the ancient story that Plato's secretary copied the Laws "from unfinished drafts the philosopher had left on wax tablets".²² So finished books went on rolls, but drafts and notes, even of a long work, could sit on tablets first.
2. No early Christian writer read for this article says that a disciple took notes on Jesus. The silence has some weight, because two early writers do discuss sources. Luke's preface names earlier written accounts and what "eyewitnesses and servants of the word" handed down (Luke 1:1-2, αὐτόπται καὶ ὑπηρέται ... τοῦ λόγου)³⁷, and Papias says that Mark, "having become Peter's interpreter, wrote accurately whatever he remembered", and that Matthew "arranged the sayings in the Hebrew language".²³ Neither mentions notes taken at the time. The weight is limited: neither writer sets out to list every source, and a practice attested among Greek and Roman students is not evidence that it occurred in Galilee, nor that it did not.
3. Papias, early in the second century, said he valued what came "from a living and abiding voice" (παρὰ ζώσης φωνῆς καὶ μενούσης) more than "the things from books".²³ That is a preference between two sources he knew, written by a man whose own work, Eusebius reports, ran to five writings (συγγράμματα πέντε).²³ It does not say that no notes existed. The wider question of how reliably the sayings of Jesus were remembered and passed on belongs to Oral Tradition and Memory.

Part 8: Jesus, Paul and the Gospel writers

Depiction is not proof, and depiction is still evidence

The New Testament depicts people reading and writing.  Luke shows Jesus standing up in the Nazareth synagogue "to read" (ἀναγνῶναι), unrolling "the scroll of the prophet Isaiah", finding the place and rolling it up again (Luke 4:16-20).  Luke also shows Zechariah asking for "a small writing tablet" (πινακίδιον) and writing his son's name (Luke 1:63).  A depiction is an author's report, and it is positive evidence.  Its weight depends on the source's date, access to the event, genre and purpose, as well as any independent corroboration.  Where no second source survives, corroboration is unavailable;  that is not by itself evidence against the report.  The opposite error is to treat an event as false because it is unusual.  A low reading rate makes a particular man unlikely to read, not unable to.

Jesus

The evidence comes from three texts, of unequal weight, and from what the Gospels show him doing.

1. Luke 4:16-20 depicts Jesus reading from the scroll of Isaiah in a synagogue. It is the only text that shows him reading. Its weight is debated. Meier notes that Heinz Schürmann, who argues that Luke had a source independent of Mark here, still "admits that the description of Jesus reading from Isaiah (vv 17-21) is a later addition to the early form of the tradition", so that "the very part of the pericope that is relevant to the question of Jesus’ literacy is probably secondary".²⁶ Evans likewise notes that "Most scholars hesitate to draw firm conclusions from this passage".¹³ It remains positive evidence in Luke's report. Whether it preserves an event or serves Luke's narrative aims must be argued from the passage and its sources.
2. John 7:15 has the crowd ask how Jesus "knows letters, having never studied". It is part of the original Gospel, and it records a perception by hostile hearers, not a school record. Meier reads it as a reference to "his lack of formal education in Scripture under the guidance of some noted scholar", and calls it the firmest of the three texts, though only "indirect".²⁶ Craig Keener's commentary, in its second edition, says: "The issue here is not whether Jesus is literate or can recite Torah, but that he has never formally studied Scripture with an advanced teacher".⁴⁷ In the first edition, as Keith quotes it, the parenthesis was stronger: "This issue here is not that Jesus is illiterate (he is not)".²⁴ Harris allows that the question may mean "How does he know the scriptures?", but also reads it as implying that a man of Jesus' standing "would not be expected to know how to read".¹
3. John 8:6-8, where Jesus writes on the ground, is part of the story of the woman caught in adultery. Bruce Metzger, for the committee of the standard Greek New Testament, wrote that "The evidence for the non-Johannine origin of the pericope of the adulteress is overwhelming."⁴⁸ That settles where the story belongs in the text of John, not whether it is historical: Meier grants that "the possibility remains that the story may preserve a reliable tradition about Jesus", while Harris calls it "probably fictional".¹²⁶ Even if the story is old, it does not say what Jesus wrote, and Meier, following Raymond Brown, thinks the "much simpler possibility" is that he was drawing lines on the ground.²⁶ So the passage is weak evidence that Jesus wrote words, whatever its origin.

The scholars divide on what follows.  Meier concludes that Jesus "was literate, and his literacy probably extended beyond the mere ability to sign one’s name or to conduct basic business transactions (“tradesman’s literacy”) to the ability to read sophisticated theological and literary works and comment on them (“scribal literacy”)."²⁶  His reasons are not any single text, which he says "prove very little", but "converging lines of probability":  the place of scripture in Jewish life, the "consistent witness of many different streams of Gospel tradition about Jesus’ teaching activity", and John 7:15.²⁶  Evans argues the same way, from Jesus' title of rabbi, his disciples, his debates over scripture and his "Have you not read?":  "I find an illiterate Jesus harder to explain than a literate one."¹³  Keith concludes that "Jesus likely did not hold scribal literacy", while arguing that "the answer one received would depend upon which contemporary of Jesus one asked."²⁴

The article's grade:  the Gospels portray Jesus arguing from scripture with experts, and John 7:15 portrays surprise at his learning without formal study.  That is a statement about the surviving portraits;  their value for his actual schooling requires the historical arguments just discussed.  Whether he could read is a matter of indirect argument.  Meier's and Evans' case, that a teacher so engaged with scripture probably read it, is a reasonable inference, and Keith's caution, that knowing scripture well does not require reading it, is also reasonable.  The population models give reason for caution;  Jesus' occupation supplies no measured rate, and the specific evidence about him gives reasons in favour.  The article leaves the question open, with the reading of scripture more likely than writing.  No text belonging to the Gospels' original form shows him writing.

Paul

Paul is the clearest case of an author in the New Testament.  His undisputed letters argue from scripture in Greek, and he added lines in his own hand.  Keith opens his study of Paul's handwriting:  "That the apostle Paul was a literate individual is well-established and beyond doubt".⁴⁵  He used secretaries anyway, and that was normal for an educated man (Part 7).  Paul is the proof that using a secretary is not a sign of illiteracy.  He also shows the limit of what we know:  even for him, we cannot tell from the evidence how much of any letter's wording a secretary shaped.

The Gospel writers

The four Gospels are long narratives in Greek.  Whoever composed them, alone or with help, could sustain written composition over thousands of words, a skill the estimates in Part 2 make rare.  Their Greek is not all of one level.  Luke opens with a formal preface in a single long sentence (Luke 1:1-4).  Mark's Greek is plain:  Henry Barclay Swete, from Mark's vocabulary and constructions, concluded that its writer "was a foreigner who spoke Greek with some freedom, but had not been accustomed to employ it for literary purposes."⁴⁹  Ehrman, who calls the authors "highly trained", grants that "the Gospels are not the most refined books to appear in the empire—far from it."³⁸  So what the Gospels require is extended composition in Greek, not in every case polished literary Greek, and the bar differs from book to book.  That is a fact about the composers, whoever they were.  It is not, on its own, an argument for or against any traditional name.

Two opposite circular arguments need to be avoided.

1. Assuming the traditional author to prove his skill. "Matthew the tax collector wrote this Gospel, so tax collectors could write literary Greek" uses the conclusion as the evidence.
2. Assuming the skill level to disprove the author. "Fishermen were illiterate, so John could not have written a Gospel" turns a population prior into a verdict on a named person, and ignores the possibilities of later learning, dictation, a secretary, a translator or a disciple who wrote down a witness's account. Each of these is possible. None is shown by the bare fact that it was possible.

What the literacy evidence does change is the burden of proof.  A claim that a Galilean fisherman or toll collector personally composed a long Greek narrative is a claim that he was in a small minority, and it needs positive evidence.  A claim that such a man stands behind a book, written down or composed by someone else, fits the documented practice of the time, but also needs evidence that it actually happened.  The specific cases are argued in Who Wrote the Gospels?.

Graded findings

Objections and responses

1. "Jews were the people of the book. Every boy learned to read the Torah." The ideal is first-century: Josephus says the law orders children taught to read.¹⁰ The report of schools in every town is told in the name of a third-century teacher about a high priest of the 60s, and even at face value it describes an ordinance for boys, not its reach.¹¹⁵ First-century evidence shows public reading (the Theodotos inscription) and a high value placed on the law, which is consistent with most people knowing scripture by hearing it. Among the parties to Wise's Judaean documents, 19 of 53 lay men from propertied families could not sign.² Universal reading is not shown, though wider reading among pious families than the minimal models allow is possible.
2. "Harris showed literacy was 10 percent everywhere." Harris gave no figure for Judaea, warned that "We shall obviously never know in a clear-cut numerical way how many people were literate",¹ and gave different bands for different regions. His numbers are reasoned estimates, not measurements.
3. "Bar-Ilan proved it was under 3 percent." His figure follows from the inputs he assumed, including zero literacy in the countryside. Change that one input to 5 percent and his own method gives about 4 to 6 percent.²⁵ He estimated Torah-reading in Hebrew, not signing, documentary writing or Greek.
4. "Wise counted real people, so his 5 to 10 percent is the measured rate." Wise counted parties and witnesses in documents that survived in caves, mostly from propertied families of the early second century. His population figure adds a level-of-hand rule, a pooled sample that includes witnesses chosen because they could sign, an assumed elite of a quarter of society and zero readers below it.²² Change the elite share to 8 percent and the same counts give 2.5 percent; allow a few readers below the elite and they give 8 to 10 percent.⁶ He himself calls it "more of a thought experiment than a hard number".²
5. "Wise found that 16 percent of Judaean adults could sign." That is the sentence he wrote,² but his own inputs, 65 percent of a quarter of 200,000 adult men, give 16 percent of men and about 8 percent of adults if no women signed.²⁶
6. "Acts 4:13 proves Peter and John could not read." The word can mean that, and Moulton and Milligan read it so.⁷ It can also mean "without scribal training", which is how Keener and Evans read it, and the second word means "laymen".¹³¹⁴ The verse is secure only for the minimum both readings share. "Illiterate" is possible; Kraus accepts it as possible "provided that other meanings are not excluded".¹²
7. "Acts 4:13 only means they had not studied with rabbis, so they were literate." The second half does not follow. Not having formal training is compatible with being unable to read, and for Galilean workers the population models make that a real possibility. The verse settles neither way.
8. "Origen, the great Christian scholar, admitted the apostles were illiterate." The wording is Origen's own reading of "the gospel", in an argument where an unschooled apostle is the point.¹⁶ He also held that Matthew wrote the first Gospel in Hebrew letters and that Mark wrote as Peter guided him.¹⁷ His sentence is later testimony to how Acts 4:13 was read, not first-century evidence, and it does not tell us how he fitted the two views together.
9. "A tax collector had to be literate, so Matthew could write a Gospel." The trade required handling accounts and receipts, which someone in the business kept.²² Documentary writing of formulas and numbers is far from composing an 18,000-word book. The trade may have raised the chance of more schooling; by how much is not known.
10. "Petaus shows clerks were illiterate, so tax collectors were too." Petaus is one extreme case from second-century Egypt, from a different office. He proves that office did not guarantee skill. He does not set the rate for any trade.
11. "Any problem of style can be explained by a secretary." Secretaries existed and sometimes composed, as Cicero's letters show.¹⁸ But a secretary is one candidate explanation among several, and since the freedom any secretary had in a given letter is unknown, the idea can neither carry an authorship claim on its own nor be used to wave style evidence away. Richards himself presents his reconstruction for Romans as a possibility for "future investigation", with one step he calls "tenuous".¹⁹¹⁹
12. "There is no evidence the disciples could write, so the Gospels were not written by eyewitnesses or their circle." No early source read here says a disciple took notes, and Luke and Papias, who discuss sources, do not mention any.²³ That silence counts for a little, since they had occasion to mention notes. It does not show that no disciple wrote, or that none stood behind a book. The question has to be argued from the Gospels' own features and from the external testimony.
13. "Galilee was a backwater where nobody wrote." Very little first-century writing has been found in Galilee, and Chancey's survey lists almost nothing for Jesus' lifetime but coins and a market weight.³⁰ Millard attributes much of the gap to the state of the archaeological record.²² Missing papyri do not prove missing writers, and lakeside Galilee lay on a main road.²² Later finds and arguments for widespread spoken Greek deserve consideration too (Part 3), and first-century synagogue buildings now stand excavated at Magdala and at rural Tel Rekhesh (Part 4). None of these establishes a rate of book-reading or composition.
14. "Women in this period could not write." The evidence shows that women's writing was rare where it can be checked: ten women among Wise's parties, Babatha, the Egyptian formulas. It does not show a universal rule, and an elite woman, Julia Crispina, wrote Greek in a practised hand.² At Vindolanda in Britain, an officer's wife, Claudia Severa, added three lines in her own hand to a letter (Part 7). Whether women read is almost unrecorded.
15. "The disciples were day labourers, the poorest of the poor." Mark, the earliest Gospel, puts James and John in a family that owned a boat and employed "hired men" (Mark 1:20), and Simon and Andrew in their own house (Mark 1:29). Kloppenborg's comparison with the fishing papyri places such fishermen "above-subsistence" though "certainly not “upper class”", in a trade full of leases, taxes and officials.¹⁵¹⁵ The same papyri show the other side: in a typical fishing lease of AD 161 the three fishermen "do not know letters" and another man wrote for them.⁸ The trade made documents part of daily life; it did not make every fisherman a writer.
16. "Nobody in the early movement could write, so the Gospels cannot go back to it." Paul's undisputed letters show a writer inside the movement within about twenty years of Jesus. The Gospels and Acts also name followers from the classes where reading was concentrated: a council member, the wife of Herod's steward, priests, Pharisees, a Levite and a man raised with Herod Antipas (Part 5). Whether a given Gospel goes back to a given disciple still has to be argued from the evidence for that book.

Issues and unanswered questions

1. No representative Galilean sample. Every population percentage for Galilee is borrowed from elsewhere or modelled. The synagogue buildings at Magdala and Tel Rekhesh show that public reading had a place in Galilee, not how many people read. Selected later abstracts qualify Chancey's 2005 inventory below; a complete survey of post-2005 finds and the two studies' full arguments remains needed.
2. Wise's book was not read. This article cites his 2012 dissertation. The 2015 Yale book revises it, and its numbers and its signing sentence may differ. The substantial review in the Journal of Roman Archaeology (2017) was not read.
3. Wise's hand levels are untested here. The sorting of hands into four levels is one scholar's judgement, which he calls subjective; no second judge's sorting, and no direct test of whether practised signers read books, is reported in the pages read.
4. Kraus's study of Acts 4:13 was not read in full. Only the abstracts of the 1999 article and its 2007 revision were available. His detailed argument from the papyri may refine Part 5.
5. Reading without writing is unmeasured. Every method that counts signatures misses people who could read but not write. Harris thinks they were few, Millard that they were many;¹²² nobody has a count.
6. Language-specific rates are mostly unknown. Wise reports rates by language for his signers, but for the first century there is no way to say how many Galileans could read Aramaic as against Greek.
7. The reading of the Soumaios letter has been revised. The sentence on writing in Greek has been read differently by different editors, so what was lacking for writing in Hebrew is uncertain.
8. Specialist studies of tax collectors were not obtained. Fritz Herrenbrück's Jesus und die Zöllner (1990), John Donahue's 1971 article on tax collectors, and Pieter Sijpesteijn's Customs Duties in Graeco-Roman Egypt (1987) were not reached by open routes. The receipt evidence here is a single dated example, not a survey.
9. Women's reading is almost invisible. The documents record women who could not sign Greek. They are silent on whether such women read Aramaic or Hebrew.
10. The rhetoric question in Acts 4:13 is unresolved here. Keener reads ἰδιῶται as including lack of rhetorical training; Kraus's abstract says the verse says nothing about rhetoric. Neither argument could be compared in full.
11. The fishermen's standing rests partly on Egypt. The Gospels' hired men and boats are the only Galilean evidence for the four fishermen's families. The fuller picture of leases, taxes and above-subsistence income comes from Egyptian papyri. Asked whether Galilean fishermen were like those of Egypt, Kloppenborg himself answers, "Probably not in all respects."¹⁵ Galilean fishing documents, if any are found, would test the analogy.

Verdict

Reading, signing, writing and composing were separate skills in the first-century eastern Mediterranean, often divided by language, and the evidence for each has to be weighed on its own.

On rates, the best-supported conclusion is modest and qualitative.  Book-reading was probably a minority skill, concentrated among men of property, priests, scribes and officials.  The models suggest a small minority, but they do not measure its size or supply the same rate for practical writing.  Signing was more common in the documents studied.  The count of named people examined here, from propertied Judaean and Arabian families of the early second century, shows that many men and every woman among the parties could not sign.  Every population percentage, from Bar-Ilan's 3 percent to Wise's 5 to 10 percent of men, is a model built on assumptions about who was counted and who was left out, and this article does not treat any of them as known.  Wise's figure of 16 percent of adults signing is, on his own inputs, a figure for men.

On practice, the documents are clear and consistent.  Other people's writing reached nonwriters as well as writers.  Husbands, sons, village scribes, professional scribes and secretaries wrote for others, and authors from Cicero to Paul dictated.  Those jobs ranged from taking down an author's words to composing in someone's name, and the evidence of any one document has to say which.  Owning, hearing, dictating and composing were all ways of taking part in a written culture without holding the pen.

On individuals, the evidence sets priors, not verdicts.  Acts 4:13 portrays Peter and John as lacking recognised scribal credentials;  whether it also means they could not write is open.  On the Gospels' own account the four fishermen came from boat-owning families in a trade that ran on leases and receipts, which weakens the force of the population rates for them without showing that they wrote.  The tax collector's trade raises the chance that he handled documents;  its effect on literary skill is unknown.  The Gospels portray Jesus as a teacher debating scripture, with John showing surprise at his learning and Luke depicting him reading.  His actual reading is plausible but unsettled;  writing words is not established.  Paul read and wrote Greek and used secretaries.  Whoever composed the Gospels could write long narratives in Greek, alone or with help, though Mark's Greek is plain rather than literary.  The literacy evidence gives reason to seek positive evidence that a named fisherman or toll collector composed a long Greek book.  Composition, penmanship and authorisation are separate questions.  It cannot show that the people named in the New Testament were unable to stand behind the books attributed to them.  That question belongs to the evidence for each book.

Sources