Dorothy Sayers Was Wrong: The Trivium and Child Development

Myth: The trivium (grammar, logic, and rhetoric) corresponds to stages in child development.

The origin of this fantastic claim, a nearly ubiquitous presupposition of classical education literature and the central organizing tenet of many a school’s curriculum, can be traced back to Dorothy Sayers’ essay, “The Lost Tools of Learning.” So foundational (dare I say creedal) is the significance of this text for the Classical School Movement, that many schools require applicants for teaching positions to submit an essay on it.

Sayers, a celebrated Christian apologist and medievalist, draws from her knowledge of medieval universities and cathedral schools to map out a programmatic alternative to schools of her day, presumably even the English public schools (think posh prep schools like Eton or Rugby, schools built on a humanist model that included the study of the classics) at which her audience, Oxford students and faculty in 1947, had been educated. The “syllabus,” or “educational theory” to which she intends to return—she specifically invokes the idea of a retrogression—is that of the trivium and quadrivium, the “syllabus” of the Middle Ages. But wait, there’s less! Forget the quadrivium—that’s just subject specialization—she says, belying her claim to actually advocate for a return to the Middle Ages.

Sayers’ foil, what she’s reacting against, is domain specialization, instruction in “subjects” like physics or European history, that teaches students the facts, what to think rather than how to learn. Insert the trivium. For Sayers, grammar, logic, and rhetoric are not so much subjects as programs, ways of thinking, cognitive substructures that guide how we learn. “The Trivium is by its nature not learning but a preparation for learning.” So, there is a grammar of history, a grammar of science, etc., and this grammar is inculcated in children through methods that transverse subjects.

These methods correspond to three stages of child development. The youngest stage in which students pick up information through repetition and imitation is the grammar stage. Herein students master “observation and memory.” Next, students enter a phase where they can master “discursive reasoning” through training in formal logic. Lastly, students blossom into the rhetoric stage where they have recognized the limits of their knowledge and imaginatively seek to expand it in a somewhat free and wide engagement with different subjects. This, however, does not result in a mere superficial factual knowledge, because, having been trained in grammar and in logic, such a student can discern the interrelation of all knowledge. Sayers even provides general age categories which correspond to the stages of development. Coincidentally, these ages fit neatly with elementary school, middle school, and high school, respectively.

Not only is Sayers’ program anything but a return to medieval educational theory, there is not a shred of evidence that her take on developmental stages has any grounding in reality. Firstly, Sayers’ program is a complete anachronism; medieval educational theorists never conceived of the trivium in terms of developmental stages that roughly correspond to a student’s age. In her way of apology, she anticipates a potential objection to her proposal, namely that it is completely ahistorical, and then dismisses it with little argument (quintessential apologist that she is): “Let us now look at the mediaeval scheme of education—the syllabus of the Schools. It does not matter, for the moment, whether it was devised for small children or for older students, or how long people were supposed to take over it.“ Really? How could it not matter? There is, after all, a huge difference between university students and small children. How can she be advocating for a return to the Middle Ages, when the trivium was not devised for and, with the exception of grammar, was not taught to small children? Latin was not a native tongue for anyone in the Middle Ages. As the language of instruction, it had to be taught to fluency before any discursive or abstract treatment of the trivium could be approached. Much of the trivium was, therefore, college-level material. A student might begin broader studies in the trivium and other liberal arts at the preparatory level for the university, but even the baccalaureate was often limited to instruction in grammar and logic.

This is not the only lacuna in her account of the trivium. Her account of the trivium is rife with anachronisms and mischaracterizations. While she conceives of medieval education as a handle for a “particular pedagogical theory,” medieval education was anything but a monolith. What was precisely denoted by “grammar” was itself in flux during the Middle Ages and subject to reform. What grammar was not, however, was a “stage” that operated as a cross-curricular organizational principle, at least not in the sense in which Sayers advocates. There never was, for example, a “grammar of history” or a “grammar of science.”

While Sayers is accurate in presenting grammar as a faculty (she calls it a “tool”) as well as a “subject,” she divorces grammar from its curricular content with slight of hand. She writes, “[M]ediæval education concentrated on first forging and learning to handle the tools of learning, using whatever subject came handy as a piece of material on which to doodle until the use of the tool became second nature.” Sayers is at pains to present the curricular content as merely incidental to the “tools,” to separate the materia from the forma. Here she does the most violence to her subject. In the Middle Ages, instruction in grammar most often followed a set curriculum with fixed instructional materials, foremost of which were the works of Donatus and Priscian. Early medieval scholar and theologian Cassiodorus defined grammar thus: “grammatica est peritia pulchre loquendi ex poetis illustribus oratoribusque collecta” (grammar is the skill of speaking beautifully gathered from the illustrious poets and orators). With this definition any discussion of medieval education is, therefore, a discussion of what texts belonged to the poetis illustribus oratoribus since these are the source whence the expertise is collected. While the medieval scholastics certainly valued the classical authors much differently than did the Renaissance humanists, their choice of texts was not incidental to the art they were teaching, and the choice of texts used as exemplars was the cause of no small amount of dispute throughout the ages. Sayers’ attempt to superimpose “medieval education” as a “pedagogical theory” onto a model of child development gives short shrift to the actual, and essential, curriculum. The texts that we read are not incidental to the “tools” we acquire.

In Sayers’ quaint manner of dismissing potential objections by simply naming them, she writes, “My views about child psychology are, I admit, neither orthodox nor enlightened. Looking back upon myself (since I am the child I know best and the only child I can pretend to know from inside) I recognize three states of development.” Whether or not her views about child psychology are “orthodox” or “enlightened” is immaterial. They are worse; they are purely anecdotal. It would be one thing, if Sayers were drawing from examples from classical and medieval texts in relating how children of different ages apprehend information and skills unique to their age—it would certainly be more conformant to the general ethos of classical educators—but she’s not. She is talking about herself, her own experience. And yet classical educators talk about whether students have entered the “logic phase” as if they were appealing to timeless wisdom.

Structuring a diachronic curriculum progression around putative stages of development that correspond to the trivium may, for all I know, be pedagogically advantageous. If that is the case, however, classical schools should demonstrate it to be so on its own merits and drop any pretense that idea has historical precedent or is in any way self-evident. If it’s not particularly advantageous, it could just be getting in the way of what should be the curricular focus, namely content. Sayers was right when she identified domain specialization as a problem in modern education. She took the wrong turn, however, when she separated the “how” of learning from the “what” of learning.