Grammar-Translation: What Is It—Really—For Students?

This article began as a blog post with the title ‘Grammar-Translation: Not Really A Method…’ (Piantaggini, 2020). Of course, that choice of title was intended to hook the reader rather than actually make such a claim. Without a doubt, Grammar-translation (GT) is a language teaching method recognised for its historical significance (Musumeci, 1997), its role amongst other emerging language teaching methods (Shrum & Glisan, 2005), as well as its shortcomings (Richards & Rodgers, 2014). This article investigates the method, and its effect on teachers and students.

Although the blog post was based on my own ideas, research for this article has shown that others have had similar thoughts and findings about the method. I am certainly not alone in my observations. For example, I do not know a single Latin teacher who would disagree that GT is still the dominant paradigm in teaching in schools today. Surely enough, in a report on the 2018 National Latin Exam Survey given in the United States, the number of teachers primarily using GT (478) was over twice that of the next most-used ‘Reading Method’ option (202), and over 16 times that of the least-used ‘Active Latin’ option (27).¹ I find this report unsurprising. Still, efforts to support my own claims from their beginnings as musings within a blog post have, if anything, strengthened my position on what GT is for teachers and students, and what it is doing to them both. Of course, this article would not be complete without a brief history of language teaching, and I do mean brief, although we shall cover a great many years within a couple of short paragraphs.

GT's origin story can be traced to 5th century BCE with two contrasting philosophies. Musumeci (2009) characterises these philosophies as follows. While Plato took the nature (i.e. innate ability) side of the debate, Aristotle took nurture (i.e. learned behaviour). Plato would have the teacher lead out (educere) what was innately acquired, while Aristotle would have the teacher instruct (instruere) knowledge that students were to gain from experience, habits, avoidance of bad models, and error correction. Over time, Aristotle's tabula rasa view won that debate as the Christian Church adopted his philosophy as ‘itself the major source and foundation of institutionalised education from the early Middle Ages onward’ (Musumeci, 2009). From then it could be said that ‘Reading and writing Latin was the foundation of formal education, which was for elite males’ (Herman, 2017). The rest, literally, is history. In the shadows, however, there were some key players in our last 600 years advocating for a different approach to teaching Latin, a communicative approach² (Musumeci, 1997), but the pitfalls of the dominant language pedagogy were documented as early as the Middle Ages.

In the world's first universities, Latin was the language of instruction. Therefore, the ability to speak Latin was required. Musumeci (2009) reports that in fact, students had to pay a fine if it were discovered that they violated the policy. Letters home to parents have confirmed how these young scholars found it difficult to follow the ‘Latin-only’ rule. There were even some students called ‘wolves’ who reported on their peers for not speaking Latin. ‘Why was a Latin-only rule needed?’ you might ask. That is a good question. Above all, such a rule tells us that Latin was artificially imposed upon the students. By this time, it was no one's native language, and when there is a shared language that is intentionally avoided, the learning environment is at best a role play, and at worst submersion—the metaphorical drowning under lack of understanding in an immersion³ setting (Toda, 2019). Provided that the elite university scholars found role play compelling, the result would have been a meaningful communicative purpose to using Latin exclusively. If, however, students saw through that charade, the Latin-only policy would have lacked a meaningful communicative purpose. If Latin was the medium of instruction, but students found it difficult to speak, clearly, they were not learning it well. We certainly could place blame on those students. However, the more likely culprit was a problem with pedagogy. The low proficiency of the young scholars gives us insight into that failed pedagogy, the roots of which are GT.

So, GT has been around for a while, but what is it? There are three terms that must be defined before analysing it: approach, method, and technique. Shrum and Glisan (2005) cite Anthony (1963) with the definition of an approach being ‘a set of theoretical principles,’ a method being ‘a procedural plan for presenting and teaching language based on the approach’, and a technique being ‘a particular strategy for implementing a method’.

Technically, then, the GT method is a procedural plan for presenting and teaching language. Yet what approach is it based on? What are its theoretical principles? Without going down that path too far, Richard and Rodgers sum things up in a rather damning way:

‘…though it may be true to say that the GT method is still widely practised, it has no advocates. It is a method for which there is no theory. There is no literature that offers a rationale or justification for it or that attempts to relate it to issues in linguistics, psychology, or educational theory’. (Richards & Rodgers, 2014, p.7)

They go on to offer explanations for how the method has continued to be practised despite such criticism, which include:

1. (a) limited command of spoken [target language]⁴ of language teachers;

2. (b) the fact that this was the method their teachers used;

3. (c) it gives teachers a sense of control and authority in the classroom; and

4. (d) it works well in large classes (Richards & Rodgers, 2014, p. 7).

Clearly, reason (a) must resonate with Latin teachers, who rarely possess any spoken proficiency with the language.⁵ Reason (b) also seems to be a common story, and one that is not exclusive to teaching just languages. However, reasons (c) and (d) deserve more attention.

The astute reader, however, will have recognised by now that I still have not given GT a definition. There is a good reason for that. As a method, GT is a procedural plan for presenting and teaching language, yet methods are based on principles, of which GT has none. Given the definitions of approach, method, and technique, then, my original blog post title might be more apropos than not! Despite all that, I still do not intend to claim that GT is not a method. Let us instead describe the characteristics of GT even if the procedural plan - in regard to defined terms - must be left suspended like an ablative absolute. According to Richards and Rodgers (2014) the main procedural plan for teaching Latin via GT is that ‘presentation and study of grammar rules, which are then practised through translation exercises’ (Richards & Rodgers, 2014, pp.6–7). Other characteristics the researchers mentioned include ‘little or no systematic attention is paid to speaking or listening’ (Richards & Rodgers, 2014, pp.6–7), as well as ‘the sentence is the basic unit of teaching and language practice. Much of the lesson is devoted to translating sentences into and out of the target language’ (Richards & Rodgers, 2014, pp.6–7). This focus on the sentence, with the parsing of individual words, places very little emphasis on whole passages. Some sentences for translating follow no continuous narrative; instead they are used to highlight a particular grammar feature (e.g. first declension nouns). If passages are to be read, though, the procedure is to go piece-by-piece until all the pieces can be assembled. Based on those descriptions, the procedural plan of GT is quite simple in terms of the teacher's role, which includes the Aristotle influence of supplying knowledge, and then testing that knowledge. For the teacher, GT consists of presenting students with textbook grammar rules; but what is GT – really - for the student?

Memorising

I posit that the entirety of GT can be reduced to memorising (i.e. storing and recalling of knowledge). Maybe that comes as no surprise to you. Then again, reducing student success and achievement to one's ability to just memorise raises concerns. As mentioned before, I am certainly not alone in my observations. Richards and Rodgers (2014) also define GT ‘as consisting of little more than memorising rules and facts in order to understand and manipulate the morphology and syntax of the foreign language’ (Richards & Rodgers, 2014, p.6). Students commit grammar rules to memory, and then recall and apply those rules in order to translate into and out of Latin. Students are expected to derive meaning at some point during this process. To be clear, this is a fairly complex way to arrive at establishing meaning. Establishing meaning refers to making a form-meaning connection (e.g. canis means ‘dog’ or evanuit means ‘she vanished’). With GT, students not only must do this for themselves, consulting dictionaries and grammar notes, which accounts for a lot of schoolwork, but this process also places a high demand on cognitive load (i.e. short-term brain processes). To be sure, there are advocates for engaging students in demanding tasks and it is a common idea in education today, and the rigour of demanding tasks is often highly valued. However, such tasks have bigger outcomes than merely understanding something. With GT, the result of demanding tasks - the recall of grammar - can be considered the lowest level of Bloom's Taxonomy, that is, understanding (Bloom, 1956).

The memorisation required of GT, then, relies on a significant amount of declarative knowledge (i.e. grammar rules). We already know that the elite male scholars of the Middle Ages found this difficult (Musemeci, 2009). Therefore, we should not be surprised that students today still find it difficult not only to accurately retain, but also to recall and then apply grammar rules in an efficient and accurate manner. Memory, then, is both paramount to student success with the GT method, as well as something teachers have no practical control over. Consider how much time is spent on the teaching and learning of mnemonic devices and other tools to help recall grammar rules from memory. Students must first remember the remembering tools, then, they must apply what is recalled. While the former might be successful—largely based on whether a student has a good memory to begin with—it does not ensure the latter. For example, I often encounter adults able to recite amo, amas, amat, as well as agricola, agricolae, and even the occasional famous first-lines (e.g. Gallia est omnis divisa in partes tres), yet further discussion reveals these adults have almost no understanding of those words in context.

A lack of context and focus on the smallest units of a sentence informs how Latin is tested. ‘Vocabulary selection is based solely on the reading of texts used, and words are taught through bilingual word lists, dictionary study, and memorisation’ (Richards & Rodgers, 2014, p.6). Perhaps GT would be less exclusive if memorising were just one part of the curriculum. However, it turns out memorising is the basis for quizzes, tests, marks, and course grades, all based on identifying, producing, and the translation of Latin with impeccable accuracy. Richards and Rodgers (2014, pp.6–7) have this to say regarding accuracy:

‘Accuracy is emphasised. Students are expected to attain high standards in translation because of “the high priority attached to meticulous standards of accuracy which, as well as having an intrinsic moral value, was a prerequisite for passing the increasing number of formal written examinations that grew up during the century”’ (Richards & Rodgers (2014, pp.6–7), quoting Howatt (1984, p.132).

And there it is. In the teaching profession, we call this teaching to the test, which is considered to be an abhorrent practice these days. Back to the reasons Richard and Rodgers (2014) give for GT being still widely used (i.e. (c) it gives teachers a sense of control and authority in the classroom, and (d) it works well in large classes), teaching, testing, and grading accuracy sets very clear expectations for the teacher and students. The issue of those expectations being unreasonable, though, is seldom addressed. So, for the teacher, grading becomes a simple process even in large classes if marks are given on a correct/incorrect basis. There is very little evaluation needed. Compared to considering a student's preference when asked an either/or question, for example, and then responding based on the content of that student's message, whether a noun is ablative or nominative requires barely any thought on the teacher's part. Interaction or evaluation of a student's response, however, is far more subjective. In many ways, then, GT is comfortable for the teacher, and it is the teacher who has control. Since the goal is accuracy, students with good memories who maintain those exceptionally high standards validate the teacher's practices. Teachers, then, have more control over the direction of their program. In the highest-level classes, students fulfil the teacher's goal of producing students who meet the high levels of accuracy that has been valued for so long. Such a practice is self-fulfilling, however, since little attention is given to the students without good enough memories to live up to the gold standard of accuracy and excellence. This is exclusivity.

Exclusivity is increasingly becoming more a part of the discussion concerning Classics. Yet, social media platforms such as Facebook and Twitter are rife with red herrings downplaying pedagogy. In fact, in a personal exchange, I was told by an advocate for social justice that teachers were free to apply ideas to whatever method or teaching practices they so chose. That was surprising, to say the least. To think that teaching grammar in the manner described above is not a part of what excludes students is perhaps the biggest oversight such an advocate can make. The field of Classics is undergoing an audit of sorts at the moment. Granted, pedagogy alone cannot dismantle systemic racism, therefore content must certainly be addressed. However, eliminating oppressive representations in Latin texts while still using teaching practices that favour privilege can be said to maintain the status quo.

The status quo has been to discriminate against less-privileged students, of whom many are students of colour. Yet interactions on social media have taken a turn for the worse. One retort to the narrative of teachers pointing out GT's exclusivity is a false claim that it suggests less-privileged students of colour are not capable of learning with GT. That's preposterous. What makes GT exclusive does not have anything to do with intelligence or potential. Instead, GT's exclusivity—aside from requiring a good memory—is made worse by testing measures and assignments that favour privilege. GT's exclusivity is also a result of not meeting students where they are and moving forward from there. When students are disadvantaged from the start of their education, almost always due to poverty—which affects more students of colour—they enter the Latin classroom less-privileged. Therefore, the testing measures and assignments that favour privilege exclude these students. It's that simple. This next sentence? Not so simple, but the Ciceronian style seems fitting. So, as long as every Latin teacher can recognise that…

1. 1. …under certain conditions, using GT and its associated practices favouring privilege…

2. 2. …especially in certain communities with students of colour…

3. 3. …not at all because of intelligence or potential but because of poverty and denied access to high quality education…

4. 4. …excludes students with bad memories, even if they happen to be privileged…

…then we will be on a better path towards making Latin more comprehensible for all students by using more equitable practices.