'I go to school, I have entered, I have said, "Greetings, Master," and he has kissed me [and] returned my greeting. My secretary slave has handed me tablets, writing case, model, in my place I clean the (surface of the tablets?), I copy the model as instructed; as I have written, I show [my work] to the teacher; he has corrected it and smoothed it over; he orders me to read. After I have been ordered I have given [the text] to another. I learn glosses, I have recited (?) them. But immediately, a fellow pupil has dictated to me ... while this is going on, the little boys go, on the teacher's orders, to a separate place, and one of the older boys has provided syllables for them; others return in order to the assistant, and write names; they have written verses, and I have taken dictation in the first group. Then as we sit down, I go through commentaries, language, the art (of grammar). When I have been called to read I listen to expositions of the reading, interpretations, the (grammatical or historical?) persons ...' (CGL III 639-40, 646 trans. Morgan 1998, 66 n. 33)

Stories like this, themselves schooltexts dated to about the third century CE and originating probably in Gaul, give us the most vivid picture we have of education and assessment in the Roman world. In these days of increasing assessment of everyone and every step of education, it may be mildly diverting to consider how our predecessors assessed and were assessed, and with what implications for Roman education as a whole.

The 'classroom' in this story is informally structured; in other stories there is more than one teacher, one who listens to pupils reading and a subordinate who gives out and corrects writing exercises (Dionisotti 1982). Children all seem to work on their own, though they talk to each other (and fall out). Except when a teacher dictates grammatical information, there is no sense of the group as a focussed 'class', and this impression is reinforced when in some stories, pupils arrive and depart at different times. This style of organization seems to be particularly characteristic of the earlier stages of education; at a later stage, when they begin to learn rhetoric, pupils take turns to declaim to the whole class, and the teacher gives out his corrections for the benefit of the whole group (Quintilian I.O. 2.2).

Even among the rich, it was not assumed that children should be sent to school at all - exposing them too young to competition, or the influence of their peers, might be damaging. Quintilian is keen to counter this fear (1.2.1ff):

'It would be folly to shut our eyes to the fact that there are some who disagree with [Quintilian's own] preference for public education owing to a certain prejudice in favour of private tuition. These persons seem to be guided in the main by two principles. In the interests of morality they would avoid the society of a number of human beings at an age that is especially liable to acquire serious faults: I only wish I could deny the truth of the view that such education has often been the cause of the most discreditable actions. Secondly they hold that whoever is to be the boy's teacher, he will devote his time more generously to one pupil than if he has to divide it among several.'

Quintilian's response is that children can pick up bad habits just as well at home, if their parents or teacher are not moral. It is the duty of everyone, whether at home or at school, to make sure that moral standards are upheld (shades of recent debates about the relative roles of home and school in the production of juvenile delinquents). But it is noteworthy that all the emphasis, here as elsewhere, is on the best possible development of a notional individual pupil under his (rarely her) teacher. There is no indication that it might be useful, let alone necessary, to compare a pupil's progress with that of others - neither for the teacher, nor for the family nor the pupil himself. Apparently, it will simply be clear to everyone concerned whether the pupil is growing up a credit to his family and community or not.

In schools, virtually all the forms of assessment which modern education takes for granted are absent. As far as we know, teachers did not test the aptitude of potential pupils before they took them on. (The exceptions might be professional philosphers, if they regarded entry into their group of disciples as a privilege to be earned.) In general, if you could pay, you could get an education. Nor was there any means of testing aptitude in any abstract sense, at any point: no intelligence tests. Apart from the immediate test of reading out your work to a teacher, we hear nothing of tests in the classroom. There were no end of year exams, nor any point when the teaching of the last few weeks was summed up. The absence of tests may be linked with the fact that at no time in the Roman world was there any legal requirement to be educated at all, nor any specifications about the age at which children should enter or leave school. Some Stoic philosophers recommended that literacy be taught as early as three; Quintilian settles for seven; others suggest as late as ten. No-one stipulates the best age to stop basic schooling, but since under the Empire people were taxed as adults from the age of 14, that is one likely stopping point. So there was no incentive for anyone to consider whether there were things that children should have learned at any particular age, let alone to devise tests to find out if they had done so. The idea of age cohorts, so ingrained in modern practice, was not associated at all, as far as we can see, with the kind of education that involved learning to read and write. The place where it did feature was in athletic training, where boys of the same age were grouped together to compete.

This does not mean that pupils' activities were not assessed in any way. In the school scene above, there seems to be a good deal of immediate overseeing of pupils' activities. They are given an exercise to do, show the results to the teacher and are congratulated or corrected. The standard is evidently in the mind of the teacher - and Roman grammarians and rhetoricians were certainly proud of their (self-)perception as expert 'guardians of language' (Kaster 1988). (No source on Roman education mentions the idea of work being taken away by the teacher to be assessed. At a later stage, though, pupils are certainly expected to do 'homework' - to read a set text in private - and it might be clear later whether they had read it or not.)

As far as we know, there were no exams in schools. There were certainly no 'passing out' exams - no such thing as obtaining a formal qualification in a subject to show that you had been educated, or to what standard. (In the later Empire there were increasingly formal entrance requirements to the imperial civil service, the closest thing to exams in this period (Marrou 1975, 310ff).) The place which comes closest to an institutionalized practice of assessment seems to have been the rhetorical school, where pupils declaimed to each other: on which Quintilian is moved to say (2.2.9ff.):

'I strongly disapprove of the prevailing practice of allowing boys to stand up or leap from the seats in the expression of their applause. Young men, even when they are listening to others, should be temperate in manifesting their approval. If this be insisted upon, the pupil will depend on his instructor's verdict and will take his approval as a guarantee that he has spoken well ... For if every effusion is greeted with a storm of ready-made applause, care and industry come to be regarded as superfluous. The audience no less than the speaker should therefore keep their eyes fixed on their teacher's face, since thus they will learn to distinguish between what is praiseworthy and what is not ...'

Here, Quintilian comes as close as anyone ever does to describing a formal practice of assessment.

The absence of formal assessment in education was made up, or its place taken, by, on the one hand, a good deal of informal interest and interference by parents, other relatives and family friends, and on the other, a highly organized and institutionalized series of competitions for the young in both literary and physical disciplines.

That adults should take an informal interest in the young, including their education, was a well-established convention by the later Roman Republic, especially in aristocratic life. It was common for an established orator to take a young man, a family friend or relative, under his patronage and let the youth follow him around and watch him perform in court. Much younger children also attracted advice: Pliny (Ep. 4.13.1ff.), Ausonius (Praef. 1) and Jerome (Epp. 25, 29, 34, 37, 41, 45, 107, 108, 127) are three who have left us letters of guidance to children (in the case of Jerome, a girl) about their education. In the Cena Trimalchionis, Trimalchio shows off his young son to his dinner guests (Sat. 46):

My little boy is growing into a follower of yours already. [He is talking to a guest.] He can do simple division now; if he lives, you will have a little servant at your side. Whenever he has any spare time, he never lifts his nose from the writing board ... He has stuck a heel in his Greek now and begins to relish Latin finely, even though his master is conceited and will not stick to one thing at a time....

If there were no examinations, there was plenty of competition, and it came in two principal varieties: athletic and literary/musical. Athletic competitions seem to have been institutionalized for boys in their teens, while musical ones seem more likely to have been restricted to adults or ephebes. Some were open to all citizens or to anyone, and some were in practice restricted to professionals. But the many lists of victors which survive on stone show that competitions for the young were often a large part of the proceedings. Their importance is attested, too, by the fact that if a well-educated young man died young, his literary prowess, along with his preternatural virtue, sagacity and charm, often featured in his epitaph:

'Weep when you see me, Dioscorus, son of Greece,
Wise in the Muses and a new Heracles.' (Bernand 1969, no. 82, from Karmouz)

'My fatherland is Lycopolis; I am Elemon
whom fate cut off in his twenty-first year;
servant of Phoebus and the Muses, I was known to everyone.' (Bernand 1969, no. 74)
Athletics featured especially in the most distinctive and colourful, if marginal, form of educational assessment in the Roman world: the competitions of the Spartan agoge. Assessment was built into the Spartan system from the moment of birth, when the Ephors inspected the baby and, if it showed signs of weakness or disability, put it out to die (Lyc. 16.1). Survivors were taken from their mothers at the age of seven, to be communally educated and tested rigorously at every stage for strength, obedience, endurance, cunning and ruthlessness. Boys and young men competed in age cohorts, frequently in public and in religious contexts, at all kinds of games and military skills. The most notorious, by the Roman period, the festival of Artemis Orthia, pitted younger and older boys against each other, while fellow-worshippers and tourists from all over the Empire watched with horrified avidity (Pausanias 3.14.8ff; p. 49ff; Kennell 1995, 49ff.). How the modern school sports day has declined from its roots.

Towards a sociology of Roman education

The absence of examinations, and the presence of competition, in Roman education, has some interesting sociological implications. I have discussed the features of competitive systems in more detail elsewhere (Morgan 1998, 74ff.), but in summary, I take it that examinations, broadly, qualify some people to do something and disqualify others. They tend to locate people in groups by subject and achievement, and qualifications so obtained last throughout life, even if they are superseded. Examinations often have a competitive element - when a fixed percentage, for instance, gets the top grade, or fails - or when the thing for which the exam is supposed to qualify people is in short supply.

Competition has some features in common with examination - it often claims, for instance, to be upholding an absolute standard. In other respects it is rather different. Competitions tend to rank rather than qualify participants, and rank them singly rather than in cohorts. The results typically last only until the next competition, when the winner is either invited to be tested again, or is not retested but is replaced by a new winner.

We can conjecture some of the effects which the existence of competitions, but not examinations, in Roman education may have had on its participants. The pupil would have had a degree of freedom in what he learned, but a corresponding degree of anxiety: he could never be sure that what he learned would be what the cultural group to which he aspired would appreciate (Burt 1992, 33). The lower his social status, the less access he would be likely to have to information about what the culture-group valued and the more likely he might be to play it safe (for some modern comparisons, see e.g. Anyon 1981; DiMaggio 1982); Willis 1977). Thus in accounts of elementary schooling, and also in surviving schooltexts, we find pupils beginning with the most central canonical authors: Homer, Virgil, Menander, Euripides and Terence.

Teachers must have felt some of the same effects of competition. The absence of a curriculum would give teachers freedom but also responsibility: they would have to judge what reading, and what writing or speaking exercises would best serve their pupils' social, as well as intellectual interests. If they got it right, we can expect them to have acquired proportionately high kudos and more pupils. Quintilian, for instance, as a highly successful teacher, is able to claim that he can make not only good orators but good citizens, good men and even rulers of the world (1 pr. 10; 2.20.4ff; 12.1.26-8, 2.6-7, 11.1).

A competitive educational system gives a society, or a particular culture group, a high degree of control over who enters that group (Little 1990). The social benefits of this for the controlling group are obvious. Roman elites, whether at local or empire-wide level, were conservative and self-perpetuating. Though in practice there was a good deal of social mobility, especially correlated with wealth, elites sought to maintain themselves and present themselves as a stable group. The criteria for belonging to the educated Roman elite cannot for practical purposes have been the same among the aristocracy of Rome and in small provincial towns in Egypt or Britain, but the competitive system allowed each dominant group to define it at their own level. And competition made it possible for the culture group, in the shape of anyone from local magistrates to the emperor himself, to decide who had excelled in competition sufficiently to qualify for entry to the socio-cultural elite (for a modern comparison, see Alba & Moore 1978). Education worked as a force for socio-cultural stability, or at least slowed or controlled the rate of change: there was never any threat of large groups of examined and qualified, officially cultured Romans waving their qualifications and demanding to join the elite. If education had produced such qualified cohorts, as it did for at least a generation in Britain in the mid-twentieth century, Roman education might have been a force for social change. But there is no sign that any ruling group in antiquity considered the possibility of challenging the social order through education.

Teachers and assessment

I have touched on the freedom and responsibility of teachers in the business of assessing their pupils. In some contexts, however, and increasingly in the later Roman Empire, teachers themselves were subject to assessment.

During much of the hellenistic and earlier Roman periods, there was little or no practical assessment of teachers. Anyone could walk into a town, sit down in the market place and claim to be a teacher. If they attracted, and kept, pupils, the claim would be regarded as substantiated. Philosophers and rhetoricians reckoned their reputation by how many pupils sought their company and how much they could charge, and they were vigorous in promoting the value of what they had to sell. Ps.-Plutarch tells a story of the Greek philosopher Aristippus, who demanded a thousand drachmas for teaching a child. "Heracles!" the father responded. "I can buy a slave for that!" "Then you will have two slaves," retorted Aristippus, "your son and the one you buy." (De lib. educ. 4f-5a)

Suetonius reports the interesting case of the freedman Marcus Verrius Flaccus, who enhanced his reputation by instituting a novel form of competition in his school, which Suetonius says was very successful. As a result, he gained imperial patronage, as a result of which he was able to charge vast fees, thereby increasing his prestige exponentially (Gramm. et rhet. 17.1-2, transl. Kaster):

Marcus Verrius Flaccus gained fame especially from the character of his teaching: for he made a general practice of pitting students of similar ages and attainments against each other in competition, to give their talents a workout, and would propose not only the subject for their compositions but also a prize for the winner - typically, some old book that was attractive or rare. Because of this renown he was also chosen by Augustus to teach his grandsons, and so he transferred his teaching to the Palatium[presumably the imperial palace on the Palatine hill], taking he whole school with him - though on the condition that he accept no more students thereafter; he taught in the atrium of Catulus' house ... receiving 100,000 sesterces a year.

Since Flaccus was a grammarian, his innovation was probably to import into grammatical teaching a practice something like the one of which we found Quintilian disapproving earlier, from the rhetorical stage of education.

With official imperial interest in teachers came new forms of assessment. In the first place it was associated with municipal and city posts in philosophy, rhetoric and grammar (Dittenberger, 577; MDAI (A) 1907:278; IG 12.9.235). Under the Roman empire these posts increased in number until most large or aspiring towns and cities had one (Kaster 1988, 233ff.). We do not know what the system of appointments was, but by its nature it probably involved assessment of a number of candidates, and quite likely some form of competition, perhaps composition and performance of a speech or poem. Towns and cities which had gymnasia must also have assessed their athletic instructors, who were appointed at public expense (Marrou 1975, 110ff.).

From the 301CE onwards, there was also much closer assessment, at least in theory, of what type of teacher a teacher was. Diocletian's price edict fixed different salaries for different types of teacher on an ascending scale of expertise: teacher of letters, 50 denarii per pupil per month; teacher of arithmetic, 75; shorthand writer, 75, Greek or Latin grammarian or teacher of geometry, 200; orator or sophist, 250 (7.66-71). We do not know whether, how or by whom this regulation was enforced - and there is evidence that other parts of the edict were wholeheartedly ignored, so it may never have been a burning issue in practice - but it meant that at least in theory, a teacher had to be able to define him or (possibly) herself, and prove the definition right.

There was one brief period during which the assessment of teachers was carried out in a more hostile spirit, with significant effects. In 362 the Emperor Julian, as a vigorous and vociferous convert from Christianity to paganism, issued an edict forbidding Christians to teach the traditional pagan disciplines, mainly on the grounds that it would be immoral, since the literary subject matter of education - Virgil, Terence, Horace, Cicero and their kind - contained references to gods in whom they did not believe (Cod. Just. 13.3.5; Ep. 36). This edict is likely to have been directed at grammarians and rhetors in public 'chairs' in towns and cities, and it was effectively enforced: Christians were stopped from teaching and lost in the process a valuable public platform and status symbol.

The practical effects of this measure did not last long, since Julian was killed in 365, and no more pagan emperors succeeded him. But in Christian circles it stirred up a debate about the relationship between Christianity and pagan culture which bitterly divided the church authorities until Augustine achieved a pacifying compromise two generations later. A compromise - insisting that the forms, techniques and wisdom of pagan writers could only benefit the young, whatever their background, religion or expectations - which is still the basis of our commitment to the subject today.

Teresa Morgan

Oriel College, Oxford


Alba, R. and Moore, G. 1978, 'Elite social circles.' Sociological Methods and Research 7, pp. 167-78

Anyon, J. 1981, 'Social class and curriculum knowledge.' Curriculum Inquiry 11, pp. 3-42

Bernand, E. 1969, Inscriptions Métriques de L'Egypte Gréco-Romaine, Paris: Annales Littéraire de l'Université de Besançon 98

DiMaggio, P. 1982, 'Cultural capital and school success.' American Sociological Review 47, pp. 189-201

Dionisotti, C. A. 1982, 'From Ausonius schooldays?' Journal of Roman Studies 72, pp. 83-125

Kaster, R. 1988, Guardians of Language, Berkeley: UC Press

Kennell, N. 1995, The Gymnasium of Virtue, Chapel Hill: UNC Press

Little, A. (1990), 'The role of assessment re-examined in the international context.' in Broadfoot, P, Murphy, R. and Torrance, H. eds, Changing Educational Assessment, London: Routledge pp. 9-22

Marrou, H. I. 1975, History of Education in Antiquity, London: Sheed and Ward

Morgan, T. 1998, Literate Education in the Hellenistic and Roman Worlds, Cambridge: CUP

Willis, P. (1977), Learning to Labour, Farnborough: Penguin

CUCD Bulletin 29 (2000)
© Teresa Morgan 2000

Bulletin ContentsCUCD HomeRoyal Holloway Classics Dept