Introduction

The history of education in ancient North Africa is often narrated through a convenient threshold: before Rome there would have been practical knowledge, royal culture, religious activity and scattered literacy; with Rome there would have appeared the school, the curriculum and the social prestige of grammar and rhetoric. The convenience of this threshold is also its weakness. It mistakes the greater visibility of Roman sources for the historical birth of education, and it risks reducing earlier regimes of learning to a preface whose only function is to prepare the Roman school.

The present article adopts a different starting point. It asks not when education began, but how forms of learning were organised, evidenced, institutionalised and socially valued across successive historical configurations. This shift matters because the sources do not support a single evolutionary line from Saharan technical knowledge to Carthaginian writing, from Carthaginian writing to royal scholarship, and from royal scholarship to Roman schooling. They support, rather, a layered history in which several regimes of transmission coexisted, disappeared, survived through translation, or were re-embedded in new institutional fields.

The central hypothesis is therefore that Roman rule did not invent education in North Africa. It reorganised the conditions under which certain competences became socially convertible. Latin literacy, grammatical culture and rhetorical training acquired enhanced value because they could be used in law, civic administration, advocacy, municipal distinction, ecclesiastical authority and literary production. Their expansion, however, depended on households, cities, teachers, materials, fees, time, mobility and gendered expectations. The result was neither universal schooling nor cultural replacement, but selective expansion under conditions of inequality.

This article accordingly treats education as a social history of knowledge. It brings together the history of ancient schooling, urban history, epigraphy, material culture and historical sociolinguistics. The aim is not to reconstruct a complete map of schools in Roman Africa. The evidence does not allow that. The aim is to identify the mechanisms by which learning moved from household, temple, workshop, scribal environment and court into more recognisable urban trajectories of elementary literacy, grammar and rhetoric. Such an analysis must remain attentive to absence as well as presence: destroyed Punic archives, elite literary filters, uneven epigraphic survival and ambiguous artefacts are not neutral gaps; they shape the object that historians are able to see.

Four questions organise the inquiry. First, what kinds of knowledge transmission can be attributed to pre-Roman North Africa without confusing technical or ritual competence with school institutions? Secondly, what do Carthaginian written culture and the mediated survival of Punic books reveal about learning environments? Thirdly, how did the Numidian and Mauretanian courts concentrate multilingual scholarship while keeping access socially narrow? Finally, how did Roman urban schooling transform literacy, grammar and rhetoric into instruments of mobility, discipline and social differentiation?

By answering these questions, the article proposes a model of educational diachrony. Diachrony does not mean a continuous institutional genealogy. It means a controlled comparison of successive forms of learning according to their material supports, linguistic values, social access, institutional settings and degree of formalisation. This model makes it possible to recognise both continuity and rupture without exaggerating either.

1. Corpus, Evidence, and Method

1.1. A Controlled Corpus

The study is based on a purposive corpus rather than a quantitative census. Its first component is literary and autobiographical: Pliny the Elder on the fate of Carthaginian books, Sallust on Punic books associated with Hiempsal, Livy on Sophonisba, Augustine on childhood and schooling, and the bilingual scenes of daily school life preserved in the Hermeneumata Pseudodositheana. These texts do not speak with the same degree of evidential force. Some transmit institutional memory; others narrativise childhood, kingship or ethnic difference. Their value depends on the question asked of them.

The second component is epigraphic. Funerary commemorations of children, inscriptions referring to pupils, teachers or literate professions, and the wider Latin, Greek, Punic and Neo-Punic epigraphic environment provide evidence for the social value of writing and instruction. Epigraphy is indispensable because it moves the analysis beyond famous authors. It is also socially selective, since it represents households and institutions able to commission durable signs.

The third component consists of material evidence: tablets, styli, inkpots, writing implements, school-related representations, calculative tools, urban contexts and iconographic traces. Such objects are essential for reconstructing the material economy of schooling, but they must be handled cautiously. An artefact associated with writing is not automatically a school object; a child represented with a scroll may signify status rather than classroom attendance.

The fourth component is modern scholarship on Roman education, Punic culture, North African urbanism, children in antiquity, Latinisation, bilingualism and imperial power. This scholarship allows the article to avoid the older diffusionist model in which Rome appears as an active civilising centre and Africa as a passive receiver. Recent work instead encourages analysis of local agency, uneven acculturation, multilingual practice and the social conditions under which language competence became useful.

1.2. Three Regimes of Evidence

The first methodological distinction separates direct evidence of schooling from weaker forms of inference. Direct evidence includes a named teacher or pupil, an explicit school scene, an autobiographical memory of instruction, a pedagogical text, or an exercise intended for learning. Such evidence remains relatively rare, but it anchors the reconstruction.

The second category consists of institutional proxies. Administrative writing, commercial calculation, temple accounts, legal careers, civic inscriptions, multilingual mediation and literary production do not prove the existence of a school in every case. They do show demand for competences that usually require some form of training. Proxies therefore reveal the social need for learning without documenting the classroom itself.

The third category is contextual evidence of knowledge transmission. Technical representations, agricultural expertise, royal reputation, translated books or religious formulae demonstrate forms of learning, memory and practice. They should not be converted automatically into schooling. Their importance is different: they show that knowledge circulated before and outside formal educational trajectories.

This hierarchy prevents two symmetrical errors. The first is sceptical reduction: since no school register survives, one would deny education. The second is inflation: since writing, royal culture or technology existed, one would infer a general school system. A defensible history of education in ancient North Africa must resist both errors.

1.3. Learning, Schooling, and Institutional Reconfiguration

The expression learning environment designates settings in which knowledge was transmitted with some regularity but without sufficient evidence of stable curricular stages: households, temples, workshops, shipyards, scribal offices, commercial networks, military contexts and courts. Schooling is used more narrowly for a recurrent trajectory involving pupils, specialised teachers, texts, exercises, expectations and stages such as elementary literacy, grammar and rhetoric.

Institutional reconfiguration names the process by which pre-existing competences are not simply preserved or destroyed but reorganised under new conditions. Roman civic life did not erase Punic, Greek or local practices. It ranked them differently. Some competences became more convertible into public authority, legal standing, ecclesiastical influence or literary distinction, while others remained local, domestic, ritual or occupational.

The concept of linguistic capital is used descriptively. It does not project a modern market of languages onto antiquity. It identifies the unequal convertibility of linguistic competence into social opportunity. Latin, Greek, Punic and local Amazigh languages did not occupy fixed or mutually exclusive domains; their value varied according to place, profession, legal status, religion, mobility and institutional recognition.

2. Pre-Roman Knowledge Transmission

2.1. Technical Knowledge beyond the School Model

The oldest forms of knowledge relevant to this history are not school forms. The rock art of the central Sahara and the debated representations of vehicles or chariots point toward practical intelligence embedded in gesture, observation, correction and repetition. Spruytte’s experimental reconstructions, together with the cautious analyses of Camps and Hachid, show that harnessing, balance, animal training and vehicle movement involved complex know-how. Yet such complexity is not proof of schooling.

The significance of this evidence is methodological. It obliges the historian to recognise knowledge where written texts are absent, while also refusing to transform technical competence into institutional education. The history of learning is wider than the history of schools; the history of schools cannot be deduced from every trace of learning. This distinction is crucial for ancient North Africa, where the archive is uneven and often mediated through later languages and imperial frames.

Technical knowledge was probably transmitted through apprenticeship, family practice, occupational routines and repeated participation in tasks. It is thus part of the social history of knowledge but not part of a school system. The article’s diachronic perspective begins here: not with an origin, but with a contrast between regimes of transmission.

2.2. Carthaginian Writing and Practical Knowledge

Carthage offers a more formalised but still difficult case. Its political institutions, maritime networks, commercial reach and ritual life required writing, calculation, formulaic competence and linguistic mediation. Punic and later Neo-Punic inscriptions demonstrate the public, religious and administrative importance of writing. They do not, however, provide a literacy rate or reveal a homogeneous school network.

It is therefore more accurate to speak of Carthaginian learning ecologies. In commerce, knowledge concerned counting, contracts, lists, weights, routes and multilingual negotiation. In temples, it involved ritual formulae, dedicatory language, accounts and specialised vocabulary. In administration, it involved scribes, officials, envoys and archives. In agriculture, it moved between practice and written codification, as the later Roman interest in Mago’s agricultural work indicates. These domains show structured learning, but their institutional forms remain partially hidden.

The destruction and redistribution of Carthaginian written culture after 146 BCE profoundly shaped the historical record. Pliny’s notice concerning the distribution of books and the translation of Mago’s treatise into Latin is not merely an anecdote of loss. It shows that Punic knowledge survived through conquest, selection and translation. Survival was therefore also transformation: books could endure only by entering the libraries, languages and classificatory regimes of others.

The Periplus of Hanno raises a similar problem. What survives is not simply a Punic original but a Greek transmission that claims to preserve a Carthaginian expeditionary record. It confirms the mobility of knowledge across languages while reminding us that transmission is never neutral. The same applies to Punic passages and Carthaginian figures in Latin literature: they demonstrate encounter and appropriation, not transparent ethnography.

2.3. Women, Children, and Elite Visibility

Questions of gender and age reveal the selectivity of the archive. Elite women such as Sophonisba may indicate access to language, music, dynastic protocol and cultivated performance. They cannot be used to infer general female education. Literary traditions often transform women into figures of beauty, loyalty or tragedy; the historian must separate the representation of aristocratic exception from evidence of ordinary practice.

The same caution applies to children. When evidence appears, it often concerns households able to leave inscriptions, royal families represented in literary sources, or pupils who entered elite memory through authors such as Augustine. The silence surrounding rural children, enslaved children, girls outside elite contexts, artisans and pastoral communities does not prove the absence of learning. It indicates the social asymmetry of preservation.

3. Royal Knowledge in Numidian and Mauretanian Courts

The Numidian and Mauretanian kingdoms occupied a crucial position between local traditions, Punic writing, Hellenistic culture and Roman power. Their courts could attract books, teachers, physicians, envoys, artists and specialists. Yet courtly learning remained selective. It concentrated knowledge near dynastic power and did not necessarily diffuse it to the wider population.

Hiempsal II is important because Sallust refers to Punic books attributed to him or transmitted through his authority. Whatever the historical reliability of Sallust’s ethnographic account, the citation demonstrates that royal written traditions could pass into Latin historiography. It also shows the political use of knowledge: history, genealogy and geography became resources through which African kingdoms could be described, claimed or reclassified by others.

Juba II offers the clearest example of scholarly kingship. Educated within Mediterranean elite networks, ruling Mauretania under Roman hegemony, and writing in Greek on geography, natural history, antiquities, theatre and art, he embodied the conversion of knowledge into dynastic prestige. Caesarea functioned as a centre where local, Hellenistic, Egyptian and Roman resources were assembled. Scholarship here was not detached from power; it was a mode of governing, legitimising and mediating between worlds.

The intellectual splendour of a court must not be mistaken for educational democracy. Courts display concentration, not distribution. They show what expertise can do when supported by wealth and authority, but they leave unresolved the question of access outside dynastic and aristocratic circles. This is why the present article treats royal culture as a distinct regime of learning rather than as a transitional stage toward mass schooling.

4. Roman Urban Schooling and Literacy

4.1. Cities, Households, and Educational Costs

Roman rule multiplied the environments in which schooling could become socially recognisable. Cities such as Carthage, Madaurus, Mactaris, Calama, Thagaste, Hippo Regius and Caesarea offered denser networks of teachers, texts, patrons, legal practices, churches, administrative needs and literary ambitions. This expansion was not the result of a central public education system. It emerged from a social ecology in which households, municipalities, teachers, patrons and professional markets interacted.

The household was the decisive mediator. A child’s entry into schooling required money, time, materials, an escort, clothing, light, writing implements and, at advanced stages, the possibility of travel or residence in a larger centre. Education involved direct cost and opportunity cost. A family had to decide whether a child could be withdrawn from work long enough for literacy to become grammar, and for grammar to become rhetorical competence.

Funerary inscriptions commemorating pupils or children who died before completing their studies do not yield statistical rates of schooling. They do reveal the symbolic value of education within households that could afford commemoration. A pupil’s death could be represented as the interruption of a trajectory. Schooling thus entered family memory and public display as an investment in possible status.

4.2. Literacy, Grammar, and Rhetoric

The curriculum of Roman schooling was cumulative but not bureaucratically uniform. At the elementary level, pupils learned letters, reading, writing and arithmetic. The practices of repetition, vocalisation, copying and calculation formed a bodily as well as intellectual discipline. Augustine’s memories of elementary learning show both the importance of these exercises and the violence that could accompany them.

The Hermeneumata Pseudodositheana preserve rare bilingual scenes of a privileged pupil’s day. They connect the household, the escort, the street, the teacher, writing implements, greeting rituals and classroom seating. They do not describe every child’s experience. Their value lies in revealing the material and social density of schooling: a pupil did not simply learn letters; he moved through a network of servants, parents, teachers, objects, gestures and hierarchical relations.

Grammar represented a more selective stage. Pupils encountered authorised texts, learned vocabulary, syntax, mythology, metre, paraphrase and commentary, and acquired the cultural references through which educated speech became socially audible. Rhetoric was still more selective. It transformed textual competence into public performance through declamation, argument, legal reasoning and persuasive speech. It could lead toward advocacy, teaching, administration, preaching or municipal distinction.

Advanced education was therefore a filter. Many could perhaps begin elementary instruction; fewer could continue to grammar; fewer still could reach rhetoric and convert it into profession or authority. The educational trajectory narrowed as it became longer, more costly, more urban and more dependent on family strategy.

4.3. Discipline, Memory, and the Educated Body

Ancient schooling formed bodies as well as minds. It organised waking, movement, posture, voice, silence, writing gestures, memory and obedience. Memory should not be dismissed as mere mechanical repetition; in a world of limited access to books, memorisation was a technology of possession. Yet the same pedagogical order could legitimise fear, beating, humiliation and coercion.

Augustine’s testimony is indispensable because it gives access to the child’s experience of education as ambivalent: desired by parents, celebrated by society, painful to the child, and productive of later cultural authority. His account should not be universalised, but it prevents a purely celebratory narrative of classical schooling. Education was a resource, but it was also a regime of discipline.

The pedagogical relationship joined affection, hierarchy, violence, emulation and performance. Teachers corrected voice and language; pupils memorised and repeated; families condoned or encouraged correction because educational success promised future honour. Discipline thus participated in social reproduction: it trained children to recognise authority while giving some of them tools to exercise authority later.

4.4. Material Infrastructures of Learning

Schooling required things. Tablets, styli, inkpots, scrolls, rulers, satchels and calculation devices were not decorative details. They were conditions of possibility. A wax tablet had to be obtained; a stylus had to be used correctly; a text had to be stored and carried; a pupil had to possess time and light. Material culture reveals education as a chain of practical arrangements rather than an abstract transfer of knowledge.

This perspective also changes the analysis of inequality. A child living far from a teacher, lacking an escort, needed for labour, unable to purchase writing materials, or unable to travel to a city for advanced instruction faced obstacles that no formal curriculum could erase. The material history of schooling is therefore inseparable from the social history of access.

The schoolchild statue from Madaurus and similar representations must be interpreted with restraint. They show that pupil identity could be materially represented in Roman Africa; they do not automatically identify a specific child or prove a complete educational institution at every site. Objects become evidence only when placed within findspot, iconography, comparison and social context.

5. Language Hierarchies and Social Inequality

5.1. Latin, Greek, and Linguistic Convertibility

Roman North Africa was not linguistically homogeneous. Latin, Greek, Punic, Neo-Punic and local Amazigh languages interacted unevenly across domains. Roman schooling did not create multilingualism, and it did not simply erase existing languages. It made certain languages and registers more institutionally profitable. Latin opened privileged access to law, civic administration, epigraphy, advocacy and many forms of public advancement. Greek retained prestige in philosophy, medicine, scholarship and eastern intellectual networks. Punic and local languages persisted in social, religious, regional and domestic settings.

The key issue is therefore not language replacement but unequal convertibility. A language became capital when an institution recognised it and rewarded it. Latin could be converted into legal competence, administrative employment, public inscription or ecclesiastical authority. Punic might remain important in local memory, ritual or community practice while losing access to the most prestigious Roman civic pathways. The hierarchy was negotiated in practice rather than dictated by a single coherent language policy.

Families understood these values. Investing in Latin grammar or rhetoric was not merely a cultural preference. It could be a strategy for advancement. At the same time, such investment required resources and was unequally available. Language hierarchy and social inequality therefore reinforced each other.

5.2. Mobility and Social Reproduction

Education could enable mobility. African authors, advocates, bishops, teachers and rhetors show that schooling created pathways into Mediterranean intellectual and institutional life. Apuleius, Tertullian, Cyprian, Augustine and Martianus Capella demonstrate that Roman Africa was not a passive periphery; it produced forms of thought and writing that transformed Latin literature, Christian theology, philosophy and rhetoric.

Yet the existence of brilliant trajectories does not reveal the ordinary reach of schooling. Exceptional careers prove possibility, not prevalence. The same institutions that allowed some children to rise also reproduced inequalities of wealth, gender, city, legal status and family capital. Schooling was not outside society; it was one of the mechanisms through which society selected and ranked.

The concept of expansion must therefore be qualified. Roman North Africa experienced an expansion of visible schooling, literate careers and rhetorical culture. It did not experience universal access. The more advanced the stage, the stronger the filters. Education could be liberating for some and excluding for many.

6. Discussion: Diachronic Reconfiguration

The preceding analysis suggests that continuity and rupture are inadequate if treated as mutually exclusive categories. Pre-Roman North Africa possessed knowledge regimes: technical expertise, Punic writing, ritual and commercial literacy, philosophical circles, courtly scholarship and translated books. Roman rule did not erase these layers, but it reorganised the field in which they could acquire value. Some resources were appropriated, some translated, some subordinated, some preserved in altered form, and some displaced by the higher public value of Latin schooling.

Institutional reconfiguration is the most precise description of this process. It accounts for the persistence of earlier competences without denying the new power of Roman urban structures. It also explains why the school became a privileged site of linguistic formation, bodily discipline and social selection. A child entering the Roman educational trajectory did not merely acquire skills; he or she entered a hierarchy of languages, texts, behaviours and opportunities.

The model also clarifies the limits of the evidence. A history of ancient education cannot be built by adding examples until they appear to form a system. It must ask what each example proves. Technical images prove embodied knowledge, not schools. Punic inscriptions prove writing and social value, not enrolment. Royal scholarship proves elite concentration, not diffusion. Augustine proves a powerful trajectory and a memorable experience, not a demographic norm. In each case, the evidential regime determines the claim.

Future research should extend this model through a georeferenced corpus of inscriptions naming pupils, teachers, grammarians, rhetors, pedagogues and literate professions; through systematic reassessment of school-related artefacts by context; and through comparison with Egypt, Gaul and the Iberian Peninsula. Digital mapping may reveal clusters and silences, but only if every dataset includes a confidence scale and an explicit account of preservation bias. Quantification must be governed by source criticism rather than replacing it.

Conclusion

Ancient North Africa did not pass from ignorance to education with Roman conquest. Long before Roman civic schooling became visible, the region contained multiple environments of learning: technical, religious, scribal, commercial, philosophical and courtly. These environments did not form a single educational system, but they produced skills, memories, texts and practices through which knowledge circulated.

Roman rule changed the conditions of visibility and value. Urban networks, household investment, teachers, texts, civic careers, legal institutions and ecclesiastical opportunities gave literacy, grammar and rhetoric a stronger social convertibility. Schooling became a route toward participation and distinction, but also a mechanism of selection. Wealth, gender, geography, legal status, family strategy and opportunity cost continued to regulate who could begin, who could continue and who could convert learning into authority.

The most accurate history is therefore neither a narrative of Roman origin nor a romantic story of uninterrupted continuity. It is a diachronic history of reconfiguration. Earlier forms of knowledge were not simply abolished; they were re-ranked, translated, appropriated or marginalised within new structures. Roman education in North Africa must be understood as a selective expansion of schooling within a multilingual and unequal society.

The article’s contribution is methodological as well as historical. By distinguishing direct evidence, institutional proxies and contextual traces, it offers a controlled way to write the history of education where the archive is fragmentary. By treating language competence as institutionally convertible rather than naturally valuable, it connects schooling to power without reducing it to domination. And by replacing diffusion with negotiated multilingual hierarchy, it restores North African actors to their role as makers, mediators and transformers of Mediterranean knowledge.