Latin and the Science of Reading

Over Spring Break, while working around the house, I listened to several episodes of the Educate podcast from American Public Media. Among them were two episodes (20 Dec 2019 and  27 Jan 2020) and an earlier audio documentary about American curricula to teach students how to read English. The episodes focus on how these curricula and the training of teachers tends to go against academic understandings of the science behind how reading works. Needless to say, the science of reading piqued my interest for teaching Latin, as did the mistakes that students were making due to the flawed teaching methods in use.

According to the APM reporting, the most popular American strategy for teaching students how to read English relies on the idea that skilled readers use three cues to determine (unknown) words in a sentence. While a debate about teaching phonics has caused the term “three cue-ing” to fade away, the “whole language” approach still often uses an abbreviation for the three cues: MSV. M is for “meaning,” or what the word means or what we hope is the meaning of a missing word–kind of like a “fill in the blank” problem. S is for “sentence structure” or whether the reader expects a noun, a verb, an adjective, etc. V is for “visual” appearance of the word. For example, “life” and “like” or audēre and audīre have similar visual appearances. In practice, this approach also emphasizes the first letter of words.

Although the three cueing idea seems to make sense to strong, adult readers, scholars of how reading is actually learned debunked it decades ago (Stanovich 1980). In reality, it teaches students how to be poor readers. Some of these bad habits resemble some of the bad habits evident in some of my middle school Latin students on a recent test:

  • Fixation on the first letter or two of a word. The prepositions ab and ad were, regrettably, often interchanged despite differences in their pronunciations and objects’ case. More generally than on this test, I often chastise students for being unable to distinguish words starting with p or v from each other. For example, pecunia and periculum are often confused, as are vita and vitium.
  • Sometimes this includes a failure to distinguish between parts of speech when students are asked to translate a sentence into English. The preposition propter was confused with the noun principium several times. Similarly, the noun vis and the verb vivere were confused.
  • Along with this fixation on the beginning of a word, there was often the failure to notice endings on words within sentences. While I suspect a failure to appreciate the importance of inflection is a factor, I do not think it is the only factor. Students can parse and then translate audēmus, “we dare,” in isolation, but then they translate it as if it were audent, “they dare,” in a sentence.
  • An impression of the entire word’s visual appearance can also mean students might skip over tense markers, such as the imperfect -bā- or the the reduplication that changes the present tense currīmus to the perfect cucurrīmus. As a result, students translate the verb with a seemingly arbitrary tense–in this test, present was rather common.

While I do not consciously teach students using the MSV cues, I have suggested students look for context clues when reading Latin. Needless to say, I do not want to work against myself and teach my students how to be poor Latin readers. Therefore, I looked into the science of reading. Essentially, people learn to read new words through a process called “orthographic mapping.” So what does that mean? and how does it happen?

First, in order to allay your fears that all this is not just another thought experiment (like three-cueing seems to be), I want to mention that this knowledge is based on experiments using fictional words and/or images and/or assessments like the Stroop color test (right) where the names of colors are written in different colors in order to see whether test subjects read the word or color of writing. Test subjects’ accuracy and speed are measured.

What the science of reading says…

In order to read texts, students need to learn a lot of words by sight. Once these words are known, they can more quickly recognize these words and focus on the meanings of these words. This process of recognizing, recalling, and reading the correct/commonly accepted spelling of words is called orthographic mapping.

Most scholars propose models that include visual clues, such as the appearance of words or accompanying pictures, as a first stage in learning to read. These pre-alphabetic readers rely on these visual clues (like the logo for restaurants) or the first letter of words to “read” names or words. Without these visual cues, students struggle to, or simply cannot, read these words or even recognize differences in spellings. Memories of new words are generally not durable and require significant practice for each word to become a sight word in long-term memory.

    • To me, this seems like students learning vocabulary through the appearance of words on a flashcard or based purely on an accompanying picture.

Most models focus on an understanding of phonics and pronunciation as crucial to becoming a more skilled reader. Indeed, as students become more acquainted with the alphabet and phonics, readers rely more on phonetic spellings than on visual cues or the visual appearance of words. Once a student transitions fully to this alphabetic stage, vocabularies increase greatly and words become sight words within long-term memory quickly. Indeed, a student may only need to be exposed to a word 1-4 times before it becomes a sight word.

  • “Stuart and Coltheart (1988) explain how beginners who possess full phonics skills can retain sight words in memory with minimal experience reading the words. It is because alphabetic knowledge provides learners with a basis for expecting specific connections between spoken and written words.” (Ehri 2005, 149)

Students teach themselves how to spell and pronounce new words by sounding them out (a process researchers call “decoding”) or by extrapolating from similarly spelled and pronounced words, like jump and thump (which researchers call “analogizing”).  While it does help if students know the meanings of words, they learn the correct spelling of a word through orthographic mapping and decoding. (Share 1999).

  • “More recently, Ehri, Nunes, Stahl, and Willows (2001) conducted a meta-analysis of experimental studies. They found that systematic phonics instruction boosted sight word reading, decoding, and reading comprehension more than other kinds of instruction including whole-word and whole-language instruction. Effects were especially pronounced in kindergarten and first grade.” (Ehri 2005, 147) Students trained to “read” whole words (as opposed to syllables) do not learn how to read as quickly.
  • “Seymour and Elder (1986) observed the word reading of children taught in a whole-word program. They found that when given a reading task, children produced only words they had been taught and were unable to read unfamiliar words. Typically they would refuse to read or would substitute known words for the unfamiliar words.” (Ehri 2005, 149)

As students learn more words as sight-words, they move from decoding and recognizing syllable units to simply recognizing and reading words as a whole. The memories of these sight words are secured through knowledge of how to pronounce them.

  • One way to encourage students to learn new words and their spellings is to do three things relatively simultaneously: someone read new vocabulary words to students aloud, students read the new words silently to themselves, and students are acquainted with the words’ meanings (Ehri 2013).

Another important element for orthographic learning is knowledge of what words mean (i.e. verbal semantic skills). Reading is learned more slowly by students who have a smaller vocabulary or difficulty understanding oral communication and/or figurative language. Even though researchers are still uncertain in what way semantic skills affect orthographic learning, it is somewhat obvious that students of a second language will not truly be able to read or understand a sentence containing an unfamiliar word or cultural concept, such as the Roman consul or Athenian boule.

In addition to the idea that students teach themselves how to spell and read words through decoding words’ pronunciations, there is the theory that students implicitly learn patterns and rules for what words are possible (i.e. orthographic patterns). For example, with Latin, eventually students may notice and learn that, when in is used as a prefix, such as with imponere, the n usually changes to an m. In the case of students learning multiple languages, the more students become familiar with the languages, the more easily the students can use these patterns to distinguish words in one language from words in another language.

Finally, it is extremely useful to note that these findings apply to phonologically complex languages, such as English, and phonologically transparent languages, such as Latin and German (Cunningham et al. 2002).

Possible ways to apply these insights in a Latin classroom

So how can we apply these wonderful insights in the Latin classroom?

  • Currently, I have only assessed students’ ability to pronounce words during the first test of their first Latin course. After that, students’ ability to pronounce Latin greatly decreases (and I begin to remind students we are not in math class when they encounter the preposition sine). I could continue to assess their pronunciation on subsequent tests.
  • Ask students to read texts aloud, or read the texts aloud to them, before we translate and/or discuss it for comprehension. While I do this regularly, I could be more intentional about doing it more frequently.
  • Correct students’ pronunciation of Latin and encourage students to sound out words syllable-by-syllable more often.
  • Teach students new vocabulary words along with their pronunciation. The publisher’s website for Wheelock’s Latin contains audio files with the pronunciation of vocabulary words, and Quizlet’s upgraded Teacher accounts allow you to add audio files and images to vocabulary sets. Soon, I plan to upgrade my quizlet account to maximize the effectiveness of this website for my students.
  • Apply more insights from the Spoken Latin practices
  • Assess students’ ability to spell with multiple choice options similar to those used in the experiments in Share 1999 and Cunningham et al. 2002. The choices were the correct word, a phonetically similar word, a word with two letters transposed, and a word with a visually similar letter substituted into a word.

How else can we apply this research to the teaching of (additional) languages?

Words Consulted

Intriguing Items I couldn’t consult due to paywalls


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s