top of page

Phonics 2000 years ago - Quintilian...the accidental phonics genius with a toxic legacy.

The teaching of the skills to decipher ‘the small, black marks on white paper,’ (Dehaene, 2009, p1) that in objective combinations form written communication and assign meaningful and accurate language to them has troubled pedagogues since the inception of early writing. Five thousand years later, modern day teachers are in the enviable position of having far greater understanding of the pedagogical and cognitive progression that enables young children to develop into fluent readers. The road to this position has, nonetheless, been long, troubled and painful particularly in the case of modern English.

The teaching and learning of the ancient logographic writing systems relied on constant repetition and memory to enable reading and writing. A Sumerian schoolboy would repeatedly copy lists of symbols that represented syllables until memorised and then he would write out and memorise the nine hundred logographs along with syllable signs. He would also have to learn the correct pronunciation of all the logographs as well as the syllables (Kramer, 1963). He was not learning to decode because there was no code to learn and his only method of studying was memorisation. The only teaching technique was repetition of the task until accuracy was achieved and the only pedagogical approach was the threat of corporal punishment and the actuality of corporal punishment. So difficult to master was the language, that a Sumerian schoolboy would learn this syllabic and logographic code well into his late teens (Kramer, 1963).

Contrast this with the economy and efficiency of teaching reading with the ancient Greek alphabet that was, ‘ingenious in its simplicity and monumental in its impact’ (Enos, 2012, p. 5). By evolving a system of only twenty-four letters, each of which captured the discrete but essential sound of the utterances of their language and when blended together reconstructed the vocal patterns of everyday speech, the alphabet removed the cumbersome feats of memorisation. It was a code which required teaching, learning and practice and was constructed with such simplicity that it could be easily absorbed by a child. The grammatists of ancient Athens taught this exclusively until a child was seven years old. With one to one sound to letter correspondence language, once the twenty-four letter and sound representations had been learned and blending had been practiced to automaticity, literacy was merely a function of vocabulary. This prised literacy from the elite and handed it to whole societies through changing reading instruction, ‘from a craft skill…into an art of social power’ (Gelb, 1989, p 184).

Virtually every individual element of Roman education was inherited from the Greeks. However, the Romans cohered the Greeks pedagogical principles into a system that could be replicated across their empire as a tool to service public policy (Murphy, 2012). The greatest evolution was the reversion from tuition of the individual to the Greek ‘school’ method: the education of large groups of pupils in forms. In regard to the teaching of reading, this required a systemised approach and method. Roman education was dominated and sculpted by the pedagogue Quintilian whose Institutio Oratoria (1913) documented both approach and method. With its focus on producing worthy orators for public life, Quintilian separated a boy’s education into three distinct elements. The first required the acquisition of the basic language skills of reading and writing, followed by a period of rationalisation and practice along with the study of grammar which, when mastered, ensured that the student was ready to study the art of rhetoric. Although foreshadowing Piaget’s (1977) theories on stages of cognitive development, it differed in one crucial (and very modern) aspect: movement from stage to stage was dependent on mastery and not age. Quintilian was, perhaps, the forerunner of the mastery movement (Bruner, 1966).

It was in this first phase that the fundamentals of reading were learned. Classical Latin, like Greek (which was often learned first), being an almost wholly phonetic language with one to one sound to symbol correspondence and every letter pronounced (Allen, 1989) , decoding was relatively easy to acquire; macrons indicated long and short vowel sounds, which in Latin referred solely to the length of time the vowel was held and not to a different sound. Quintilian was insistent that sound to symbol correspondence was privileged over the learning of alphabet letter names (Van Nieuwenhuizen, Brand and Classen, 2014). Quintilian’s insistence that the learning of letter names in the initial stages ‘hinders their recognition…’ (1977, p46) was a remarkable assertion, which can only have been discovered empirically and was entirely at odds with the development of reading instruction following the demise of the Roman empire. For many hundreds of years until the development of linguistic phonics programmes (McGuiness, 1997), letter names were always learned initially, and, in some cases, spellings were taught before sounds. Indeed, the teaching of English reading became dominated by the learning and recitation of the alphabet, a pedagogy that almost certainly retarded decoding proficiency and created a toxic legacy for the emergent state education systems of the nineteenth century. Nonetheless, Quintilian’s discovery of this phonemic pedagogical veracity was crucial to the speed at which decoding was achieved by his students.

Quintilian separated reading into four pedagogical stages (Lannham, 2012): the learning of sound to symbol correspondence was followed by an almost obsessive fixation on syllables; once syllables could be decoded at speed then words were introduced followed by short sentences. This four-stage pedagogical approach remained in place for centuries and was referenced by Remigius of Auxerre in the ninth century (Lanham, 2012).

It is worth analysing this staged approach to the teaching of decoding by Quintilian because of the long-lasting resonance on the teaching of reading and its implications on the development of the teaching of the decoding of English. What is revealing, bearing in mind Quintilian had no recourse to cognitive science, is his attitude to the syllable and his insistence that syllables were required to be memorised. ‘There is no short way with syllable,’ he wrote (1977, p165). As a result of this relentless pedantry, pupils practiced reciting the letters of the alphabet in every possible combination before repeating the procedure with syllables and then with whole words to the point where Quintilian has been accused by Capella (1977, p105) of developing a ‘fatigue-march’ progression towards reading that would have ‘bored the gods’. Nevertheless he had stumbled across a cognitive keystone for reading: that the brain privileges syllable when reading (Dehaene, 2014).

But did Quintilian’s pupils really memorise all of the syllables? The answer is an emphatic ‘no’. Although he placed a huge emphasis on memory and memorisation, there is no way they could have. As an alphabetic language, the consonant, vowel combinations of Latin were myriad and formed tens of thousands of syllables, far exceeding the two to three thousand symbols that the human mind is capable of holding in long term memory (Willingham, 2009). So, what was happening, because Quintilian’s pupils certainly learned to read? They had to. His insistence that each stage required mastery ensured that no child, irrespective of age, could move onwards unless he had fully grasped the entirety of the previous stage. This fundamental tenet ensured that no pupil entered the second ‘grammaticus’ phase without being able to decode the language (Murphy, 2012). Thus, if they could not ‘remember’ the thousands of syllables required to read, no pupil would ever have progressed beyond the elementary stage of learning.

Although pupils were being taught to memorise syllables and were certainly trying to memorise the syllables, they were doing no such thing. They were decoding the letter sounds that they had spent so long learning (and, yes, memorising, because there were only twenty-four of them and they had sound to symbol correspondence) and then they were blending them together at speed to sound out the syllables. Because they were encouraged to practice this relentlessly, they became adept at it: to the point of automaticity. They were then compelled to practice the reading of longer polysyllabic words and then phrases, clauses and sentences.

During all of this relentless practice the assumption was that pupils were memorising. They were in fact decoding: initially at individual letter level, then at syllabic level and finally at polysyllabic level. For all the wrong cognitive developmental reasons, Quintilian had stumbled across the perfect way to teach the decoding of Latin: sound to symbol correspondence, followed by blending, followed by polysyllabic blending and word reading. Quintilian added a pedagogical piece de la resistance: an insistence on the avoidance of haste when reading. By demanding that reading be, ‘at first sure, then continuous and for a long time slow…’ Quintilian (1977, p14) gave his charges the space to practice decoding to automaticity. He seemed to have understood the importance of this much neglected and misunderstood stage on the road to reading fluency.

Furthermore, it should be remembered that Quintilian was teaching reading in classes of up to twenty boys and that all of these boys learned to read. Contrast this with the chaos that ensued when the education systems of England and Scotland moved from one-to-one tutoring of the rich to universal, compulsory, mass schooling and found the collective teaching of reading to mastery almost impossible, and Quintilian’s reputation as a master pedagogue is well-earned.

There is, however, a sinister caveat.

Quintilian’s genius bequeathed the teaching of reading English a lethal hangover. Latin syllables were decodable once letter sounds had been mastered. The memorisation that Quintilian demanded was in fact merely practice to mastery. But what if, as with English, single letters represented more than one sound and the same sound could be represented by a number of letter combinations? Without knowledge of how to decode those complexities, memorisation of syllables would be no more than that: a memorisation feat, and with the brain’s capacity for memorising symbols at about two thousand (Willingham, 2009) and over fifty thousand syllable combinations in English (McGuiness, 1999) a pedagogical perfect storm was brewing.

The perfect storm hit landfall with the English language.

The next reading ape blog will analyse how the teaching the reading English evolved.

This blog is number 3 in a series of blogs. You may be interested in reading:




Follow the reading ape on twitter @thereadingape

bottom of page