Introduction and background

Upon first thinking of potential projects and areas of research, I wanted to focus on an aspect of Latin which crosses over with another field of education. In my first placement, I was struck by the relative differences between practices taught in a Spanish classroom and that of Latin. The Spanish class revolved around adding carefully chosen bits and pieces of language to an existing story, with the final aim of creating phrases and sentences from a bank of existing student knowledge about the language. I had never seen this practice used in a Latin class before I started this project. I was familiar with ‘prose composition’ exercises, which usually require an English sentence to be directly translated into Latin. This kind of composition exercise is part of the GCSE and Latin A level examinations in England, and is therefore widely taught (The International Baccalaureate also includes Latin composition, requiring more creative composition, but this is not as widely taught in the UK).

Prose composition is said to be useful for reinforcing students’ knowledge of Latin vocabulary and grammar, as it requires them actively to recall words and phrases when faced with their English equivalent. In my brief experience of teaching, I have both observed and led lessons using this kind of prose composition exercise several times, and I am always struck by a specific kind of fear it seems to induce in students, which appears to hamper their ability to complete the exercises in the first place. This fear seems to be rooted in a general impression that Latin is extremely difficult to write fluently, perhaps given the lack of practice students are given to do so. This fear may also be due to a kind of latent perfectionism – with many grammatical rules and intricacies to follow, writing Latin can seem overwhelming to students of all levels. While planning this project, therefore, I wished to aim for a different kind of exercise. Instead of a sentence to be translated, I wanted to give students more freedom to write their own Latin, using existing Latin as a template. This, I thought, might provide two key benefits. The first relates to comfort and fluidity in the exercise: with more freedom to write their own material, students might be more engaged in the class and more keen to write Latin. This is partly drawn from personal experience, and partly from research I will mention later in this essay. The second relates to personal creativity as a helpful part of the process of learning; if students can create their own Latin to help themselves understand aspects of morphology or other parts of language, it would surely be helpful at any stage of learning.

As will be discussed later in this essay, this aspect of Latin learning is not a new idea, but there is little thorough research regarding the expansion and changing of an existing Latin text for pedagogical means; the majority of evidence that it is useful is largely anecdotal. For instance, Buczek (Reference Buczek2017) provides a concise overview of his approach to changing and developing Latin narratives in collaboration with his students, incorporating characters and scenes sourced from the coursebook. Schwamm and Vander Veer (Reference Schwamm, Vander Veer, Lloyd and Hunt2021) describe a similar activity based online, taking inspiration from ancient myth to allow students to create their own narratives around characters and legends of antiquity. Hunt notes that this creates opportunities ‘not just to recall the ancient mythical references, but also to project and negotiate their own thoughts and ideas as young people of today’ (Hunt, Reference Hunt2022, 149). Thus, these exercises have grounding in the context of studying Latin in a new and creative way, but lack solid research to act as a pedagogical foundation. With this in mind, I wanted to focus on the correlation between Latin composition and its potential to strengthen and reinforce students’ understanding of specific grammatical rules. This is not unheard of in linguistic education: within French, German and Spanish studies, for instance, free composition acts as an essential part of assessment of linguistic skills at all levels. This is presumably due to the requirement of linguistic flexibility required to practise a living and commonly spoken language, as opposed to a ‘dead’ language whose linguistic patterns are unchanging, and in which creative writing is rare. However, I think this flexibility can exist in Latin – if students’ creativity can be harnessed, a compositional exercise might be able to reveal students’ grammatical mistakes and misunderstandings. Applying this idea to a class which was still grasping the basics of the language, I thought that composition might be an interesting way to investigate students’ usage of certain grammatical rules. Thus, with this project, I aim to introduce two separate methods of Latin composition within the environment of a low-level class, aimed towards helping students to experiment with Latin morphology in a way which also encourages creative freedom.

Clarification on types of composition

To avoid confusion, I wish to clarify two phrases I will be mentioning plentifully in this project. ‘Prose composition’ is an exercise involving direct English-to-Latin translation, as in GCSE and A level examinations. ‘Creative composition’ is an exercise which tests students’ ability to transfer their own ideas into another language. These two exercises are derived from the same general exercise of composition, which revolves around calling Latin words to mind and constructing Latin grammar and sentences which make sense. However, the two exercises test different aspects of students’ understandings of Latin. Prose composition aims to test grammatical knowledge through reverse translation. This form of exercise commonly has a correct answer which students are expected to produce, and it is therefore fairly restrictive and grammatically rigorous. In contrast, free composition aims to create meaning as a form of personal expression. In this kind of exercise, the meaning of composed Latin is prioritised over the grammar used. Since there is no exact answer to be sought by students, the aim of these exercises is often loose, relying on students’ initiative to be bold and express themselves in their text. In my opinion, if a student does not feel motivated to take part in creative composition, they will likely not benefit from the exercise.

For many Latin educators in the UK educational system, creative composition has been viewed purely as an interesting exercise, by which a student could take to reaching the proper meaning of the sentence (while adding their own pieces of flair, perhaps). Within this project, I wish to avoid this. Instead of telling students to creatively translate a narrative into Latin, I will allow and encourage students to warp and change the narrative I give them. In turn, this freedom will permit students to approach grammar and morphology in new ways, and to push themselves to write things they want to say correctly, instead of writing what they have to say correctly. This, I think, will allow students to take control of the language in a way that the conventional practice of Key Stage 3 Latin rarely allows.

Teaching sequence

I conducted research for this project at a school anonymised as School A, which uses the Cambridge Latin Course (CLC) series of textbooks for its students throughout Key Stage 3 (students aged 11–14), and switches to Taylor's Latin To/Beyond GCSE at Key Stage 4 (students aged 15–16). Students are encouraged to choose prose composition over comprehension within their exams (there is a choice offered between the two), both at GCSE and at A Level. Lessons are one hour long.

Within this project, I wanted to provide students with two separate opportunities to compose their own stories, based around relevant grammar they had learnt recently. For the purposes of this project, for both Lesson 1 and Lesson 2, I wanted a story which revolved around action: none of the upcoming parts of the CLC were sufficiently exciting enough to base a story on. I also wanted a story with more open-endedness, so that students could be more creative and bolder with their ideas of where the story might go next. To this end, I decided to write my own short story about a female gladiator (called Attica) facing a lion in the arena, with spectators cheering on (this story can be found in full in Supplementary Appendix 1). I wanted to include a female character – partly because the CLC at this point lacks female characters in general, and partly to encourage conversation about the existence of female gladiators. The story ends on a cliffhanger with an open-ended question placed at the end: quid deinde accidit? [What happens next?].

After a brief starter activity based on Latin-to-English translation, ensuring that students were comfortable with the verb sum and its singular uses, I showed this story to the class on a projector screen. I offered additional vocabulary where students had not seen it before (for numquam [never] and gladius [sword], for instance) as well as a full translation of the ending question, which I thought students might struggle to grasp. I purposefully scaffolded the story around relevant grammar to allow students to remind themselves of the grammar in focus, and I left the story as open as possible at its end to allow students to think of their own ending.

When I was confident that students had all grasped the cliff-hanger and the beats of the story, I asked them to think about what could happen next in the story for five minutes. I gave encouragement and cold-called students for some of their ideas, then told them that they would be writing their own endings to the story in Latin. I asked for two to four sentences of their own Latin, or more if they wished. As part of this process, I wished to offer as wide a range of vocabulary as students wished; I showed a list of related vocabulary on the projector screen at the front of the class (with words glossed such as timebat [was afraid], fugit [ran away from] etc.). I also gave out dictionaries around the classroom so that students could look up any word they could think of, with the goal of total freedom in their own compositions. I showed a brief example of my own creation of what this task might involve for students, modelling grammar and creative expectations. I also anticipated grammatical difficulties with morphology and general sentence structure, given that this was their first real attempt at Latin composition, and encouraged students to help each other and ask for help if they felt they needed it.

When planning Lesson 2, I took a different approach. In learning from Lesson 1, I was pleased that every student attempted and leaned into the creative aspect of the exercise, but I felt that some students were carried away by their own creative ambition, losing sight of attempting to construct correct grammar. With this in mind, I wanted to try another lesson using the CLC as a basis, with less support for unnecessarily extended vocabulary usage and more of a focus on morphology and relevant grammar. I considered this approach a more restricted version of the exercise within Lesson 1, with the ultimate goal of comparing a more creative approach with an increased focus on trying to write good Latin. For Lesson 2, therefore, I wanted to focus on a particular aspect of grammar to be practised in composition. For this purpose, I decided on uses of the perfect and imperfect tenses, building on the narrative contained within the CLC's model sentences (in Stage 6). These model sentences cover a brief story: a dog attacks two people, and another person intervenes to help them out. This story has enough drama that changes can be easily worked in. This lesson took more time to set up.

Firstly, I took students through the model sentences and spent the rest of the lesson cementing students’ knowledge of how both past tenses worked individually. In Lesson 2 proper, having refreshed students’ memories of the new grammar, I asked them to compose Latin within the framework of adding something to an existing story. This time, however, I asked students to think of a new character they could add into the story and act within it. I was intentionally vague in how this new character should interact with the action, because I wanted a degree of creative freedom. This was intended as a new twist on a now-familiar exercise, and I was impressed with the breadth of characters students thought up. Instead of providing dictionaries, I restricted extra vocabulary to a short list of potentially useful words, and encouraged students to use words they already knew. I also showed my own character, Santius Clausus, interacting with the story in a few short sentences, modelling what their work should look like.

Findings

Data for this project was gathered in two primary ways. The first and most significant of these was work completed within the lesson itself, which I copied from their books. This work can be found in Supplementary Appendices 2 and 3. The second of these took the form of a focus-group discussion loosely based around questions I asked openly. I held two focus-group sessions with the same five students, with one session taking place after each lesson. I have included some particularly notable extracts of students’ work within the discussions of data below for ease of access. Regarding the work completed in the first lesson, I was impressed with how much text students managed to write in a short time-frame of 30 minutes; these five students were not exceptions. Only a few in the class of roughly 20 were not able to complete two to four sentences, as I had stipulated in the introduction to the exercise. Of the five monitored students in the class, all engaged with the exercise, and the general mood of the class was one of enthusiasm and excitement, primarily driven by the realisation that students could use any words they wanted.

In the post-lesson focus group, students reported that the lesson was ‘fun’ and ‘easy,’ and appreciated the broad range of vocabulary they could use. Two students seemed to think that the purpose of the lesson was to learn the vocabulary used in the exercise. Some students took their lexical freedom as far as possible, asking what the word for ‘chocolate cake’ was in Latin, while others stuck fast to the vocabulary provided on the projector screen (most of which existed at some point in the CLC; e.g. currit [runs], decidit [falls], fugit [ran away] etc.). Broadly, most students tried new words and put new phrases together which had not appeared in their textbook before. The data collected from students’ classwork represents this creative spirit faithfully: Alissa and Florian clearly drew as much knowledge as possible from the dictionaries they were given, using interesting-sounding words such as cerebrosus [enraged], erripit [drew out] and prosilit [leapt forth] to describe varied and floridly violent scenes (disconcertingly, two of the five compositions include decapitation, though this may be my fault for staging the exercise around a gladiatorial fight).

Throughout the lesson, these students and many others were constantly asking me for the Latin word for specific phrases which they were not able to find in the dictionary. I tried to remind them of grammar rules where possible, but tried to avoid writing their composition for them using as-yet-unknown cases and forms of grammar. For instance, when asked ‘How do I say “Attica's sword was sharp?”’ I did not want to introduce the genitive case solely for this purpose. Instead, I suggested an alternative and simplistic way of phrasing this: ‘Attica had a sword’. ‘The sword was sharp’. This annoyed students who wanted to write in Latin what they were thinking in English; usually, these students would then decide to write something else entirely. Despite their irritation, I saw this moment as a useful learning opportunity for students: by encouraging them to rephrase their desired statement, I pushed them to make choices based on the grammar points I wanted and to use vocabulary that they were familiar with. In other words, instead of relying on complex explanations, I created an environment where students were stimulated to stretch their own knowledge to reach the creative limits of their stories.

In terms of morphology, students broadly had a good grasp of noun- and verb-endings. I had intended the various forms of sum to be the primary focus of this exercise, but in the event, students hardly used these words at all. Perhaps I did not sufficiently clarify this before the exercise began. In any event, this is not a major issue for this project; students used the morphology of nominatives and accusatives in interesting and unexpected ways, and this will serve well as the sole focus of this study. Indeed, looking at the results of students’ work even for Lesson 1's exercise, it is startling how clearly students’ trains of thought can be understood and their individual issues with morphology diagnosed. This seems to me to bear out Hunt's suggestion that compositions can act as a method of diagnosis for a teacher, and also – potentially – as a method of self-revision for students to use themselves (Hunt, Reference Hunt2022, 129). As I had expected at the beginning of this process, the most common mistake made within Lesson 1's exercise was the mixing-up of nominative and accusative endings. This is visible particularly with the work of Florian and Iannis. Florian mixed up both case and gender several times (e.g. perterritus turbam est [it is a frightened crowd]) but pushed for a strong sense of narrative. Given her history studying Latin before Year 7, and having observed her behaviour in other classes, I expect her mistakes come from an impatience which, though it might fuel her creative drive, undermines her attention to detail within morphology.

In contrast, Iannis’ work is short and unambitious, aiming to complete the exercise given and nothing more. Iannis’ sentences largely consist of short actions, all but one of which include an error (e.g. Attica fugit leo. Attica paret gladius [Attica fled the lion. Attica prepares the sword]). During the lesson, Iannis and her friends sitting nearby were talking loudly about how they could tell their respective stories, and what might happen in them. I took this to mean that progress was being made within the narrative of the story, and I was surprised to find that there is not much action in Iannis’ phrases. Clearly, Iannis did not take special care in approaching grammatical and morphological correctness, and her lack of deviations from her set sentences show that little experimentation took place either. However, Iannis’ work in Lesson 2 was a step beyond her work in Lesson 1, and I take this as a sign of increased comfort within Latin composition. Iannis’ remark in the focus-group discussions is telling of her attitude within Lesson 1: ‘I thought I had a good story set up but it was a lot harder to put it into words.’

Looking at Alissa's work (Figure 1), there are a few nominative-accusative errors, but not many. Alissa's work is carefully put together and interesting in that her word-order is closely correlated to grammatical correctness. When creating simple sentences in a subject-object-verb order, her nominative and accusative endings were perfect: Attica leonem capit. Attica gladium erripit [Attica captures the lion. Attica draws her sword]. However, when this stable structure faltered, her sense of the correct endings did too: leo fugit Attica [the lion runs away from Attica] In this sense, Alissa had clearly absorbed correct sentence structures and individual morphologies she had seen within the CLC, and this was serving her creative purposes well. Clarissa's shorter piece of work seems to stem from an English sense of narrative, using longer phrases connected with the useful et [and] as well as extremely anglicised uses of the word ad [towards], culminating in the strangely poetic line ad surgit, et subito currit ad [she gets up and suddenly runs towards it] In conventional Latin, ad [towards] must be connected to an accusative noun, and cannot exist on its own. I failed to remind students of this aspect of grammar before the exercise began, but I would never have thought to do so, since the English word ‘towards’ similarly does not exist without an attached noun.

Figure 1. Alissa's work in Lesson 1.

Out of all the students in question, Alissa and Leila were the only two to use ad [towards] with an accusative noun. Alissa's work is the only one who used ad as it would be used in a textbook. Leila's usage of ad makes sense, albeit with unusual word order and the wrong gender in Atticum. While Clarissa used ad on its own, presumably to indicate a sense of motion (Figure 2), Florian and Iannis attached ad to a noun without applying the accusative case.

Figure 2. Clarissa's work in Lesson 1.

The realisation that a simple preposition, which I had considered so straightforward to use in Latin, is a potential roadblock for some students, sheds some light on the more common problems found by students within seemingly straightforward and familiar Latin. Doubtless every student in this project could translate ad as towards or to if given an on-the-spot test, but if the inner workings and governance of its rules are not fully understood, I can see other problems developing along the line of Latin language progression. Thus, revealing these issues at this early stage is a useful effect of this exercise, further supporting the idea that composition could act as an early diagnostic tool.

Also intriguing to note is the relatively wide misunderstanding, construed by multiple students, that Attica was male. Leila's error of Atticum (as opposed to the correct Atticam) is perhaps attributable to this, although the additional error of leonum [lion] (instead of the correct leonem) suggests a broader misunderstanding of differing declensions having varied accusative singular endings. Indeed, Leila's work contained no correctly formatted accusative nouns, but still functioned as an impressively dynamic narrative, with reactions from the crowd structuring the movement and action of her story. Leila also included two correctly formed instances of sum, which I was pleased to see functioning properly within her writing. Where Lesson 1 was largely a creative-writing exercise in Latin, within which students used morphology they were already fairly familiar with, Lesson 2 held more of a focus on writing Latin while consciously trying to include and experiment with perfect and imperfect tenses. Similar to Lesson 1, I made sure that students were familiar with relevant grammar as well as the narrative they were inserting their own creation into, but even so this resulted in a more agitated and uneasy lesson than I had hoped. Students were not comfortable enough with the new grammar to feel as creatively able as they had within Lesson 1. Despite this unease, I was very impressed by students’ abilities to work with the new knowledge they had only recently met.

Alissa's work was surprising in its difference from her work in Lesson 1 (Figure 3). When, during the lesson, I asked about the des within the name Bob des Ingens [Bob the Huge] she said that ‘there needs to be a Latin word for “the,” so I can put in “Bob the Huge.”’ I was fascinated by this anglicisation (not least because des is such an odd choice for a fictional Latin word for ‘the’), especially as Alissa had seemed so sure of Latin's unspoken rules in Lesson 1's exercise. One of these rules, which I repeat as a mantra in lessons, is that Latin has no word for ‘the’. To invent this word, instead of working around its absence, seemed to be out of character for Alissa, and perhaps speaks to the frustration other students felt when Latin would not obey the meaning they wanted to convey.

Figure 3. Alissa's work in Lesson 2.

However, besides this intriguing anomaly, Alissa's grammar was perfect, and her usage of imperfect and perfect verbs (along with associated endings) was excellent. Less had been written than in the previous exercise, but from the perspective of morphology, Alissa was confident in both exercises. Her explanation of tenses after the second focus-group was short and a little vague; I am not sure she could explain it to another person, but she used both tenses very well and seemed to grasp them both within written Latin. Her written explanation of the tenses reflects this: ‘the imperfect tense continues for a long period of time however the perfect tense does not go on for much longer.’ It is only worth mentioning in passing that Alissa maintained the level of gruesomeness found in Lesson 1's stories, with the unfortunate Bob the Huge being slain by a dog (alongside Florian's work, who wrought vengeance upon poor Grumio and Clemens through the ominous character of the ‘Dog King’).

Interestingly, the work of most students in the class, besides Iannis and Clarissa, was less productive in Lesson 2 than in Lesson 1. There are no doubt other variables at play here, including the difficulty latent in experimenting with a relatively new piece of linguistic information. It is also possible that an overreliance on the new grammar, as well as students’ unfamiliarity and lack of confidence using it, led to a reduced sense of creative freedom and enthusiasm. I think this was the case in Lesson 2; this might also have been compounded by a reduced allowance of new, exciting vocabulary to use. This was clearly a concern of students: when asked what they thought of Lesson 2's exercise, most students seemed to comment on access to vocabulary. Clarissa asked for a dictionary, Alissa asked for more vocab to learn for future composition exercises, and Iannis asked for a structured set of key words to use. For Alissa and Florian especially, this seemed to make up a large part of the appeal of writing Latin. It seems to me that, after the story of Attica was followed by a textbook-based contractual exercise in grammar, as opposed to the open-ended universe of the dictionary, their hunger for creating exciting stories was dampened. This, in turn, seemingly hindered students’ abilities to experiment with writing Latin and trying new bits and pieces of the language. Particularly for Leila, this seemed to be an issue: ‘There was one bit when we were writing our own stuff where it was harder to structure it all, because we didn't have as many words to use.’

Within the objective test of writing an explanation of imperfect and perfect tenses, I was pleased with students’ answers, which were largely acceptable and sensible. Some students gave details about the tenses’ functions in English, and some gave Latin examples, but almost all grasped the sense of imperfect as a continuous or incomplete action, along with perfect as a single and complete action. Florian's explanation (Figure 4) reveals that she has not fully understood the difference in meaning between the imperfect and perfect tenses, but she does correctly identify each tense and give good examples of each tense within a Latin context.

Figure 4. Florian's explanation of past tenses.