Writing in Latin: an experiment in composition among early-stage Latin learners

Abstract

In the United Kingdom, especially since its re-introduction into GCSE exams by the coalition government of 2010, Latin composition attracts strong opinions. Indeed, Latin teaching methodologies altogether are highly debated. Traditional methods of grammar-translation are avoided by reading courses because of their supposed elitist nature, yet they are still used by many practitioners, and this is typically where prose composition is seen. This study investigates the use of composition in the teaching of Latin to a group of Year 7 students who usually follow a reading course, to see if writing Latin can be of any benefit to students who otherwise would not write any Latin. There is a great deal of literature on the topic, both in favour and against the pedagogical uses of composition. The aim of this study was to implement techniques from the literature into the students’ lessons, and to see what the outcomes were of this new skill.

Rationale

School X is a mixed secondary state grammar school with approximately 1000 students. It is a selective school whereby students sit the 11+ to gain entry, and a Catholic school. Its local authority is Slough, and its latest OFSTED report was outstanding in September 2011. The proportion of students gaining Maths and English GCSE grades 5–9 is 97%. It has a Free School Meals eligibility of 2.4% and a Pupil Premium eligibility of 3.6%. The focus class of this study is a mixed-ability Year 7 class of 28 students. There are three students with SEN needs and one student with EAL. At school X, all students study Latin in Year 7 and they can elect to continue studying in the following years.

The focus class follows the Cambridge Latin Course (Cambridge Schools Classics Project, 1998). Being in Year 7, they are all in their first year of Latin study, and at the time of these lessons they had completed one full term of Latin lessons, with only one hour-long lesson per week. During this term, students had missed several lessons due to other school activities such as trips and extra-curricular days. Students had therefore received limited opportunity for class-time work and learning, as well as home learning. This lack of frequency may have been the cause of their frequent struggle with retrieval and recall, as well as it taking a considerable amount of time at the start of the lesson for students to be in Latin lesson mode.

Students follow the first book of the Cambridge Latin Course (Cambridge Schools Classics Project, 1998), and prior to my teaching them, had used the reading-only approach. They had not written in Latin beyond grammatical notetaking. I was interested to see what the response to writing in Latin would be among this class. I wondered if writing in Latin would cause students to make new connections to Latin, and to see if they could appreciate and have any benefit from variation to their cognitive processes as they learn Latin. Lessons were planned to build up the compositional tasks, from re-working exercises to groups composing mimes and this was based on the literature.

The focus class is a mixed-ability class, although all students have passed the 11+ exam to get into this academically selective school; therefore, ability was likely higher than a typical Year 7 class. I chose this topic to see if the variation in the use of Latin would challenge them, cause them to manipulate the language, and improve their retention. The aim was to see if students would improve their retrieval and recall, and maybe even their overall fluency in Latin. Prior to the series of compositional lessons, students struggled to recall the Latin for even the simplest of words which they had covered multiple times previously through their coursebook's repetitious nature. I was curious to see if there was any difference in this as a result of their compositional lessons, and whether this helped them overall in Latin.

Literature review

A survey of thoughts on composition

Composition is a loaded and divisive topic within Classics, and much like Classics itself, it often cannot be mentioned without devolving into an emotional and political argument. The literature reflects this. Saunders (1993, 385) writes in an understated manner ‘Latin prose composition … is a topic with few fence-sitters among teachers of classical Language’, and she is not wrong. One of the strongest opinions against composition is its supposed traditionalist, elitist and old-fashioned nature. For some authors, it has no place in modern education for this reason. It is scolded excessively (Ball and Ellsworth, 1989, 1992) as a pedagogical method with an elitist motivation, a tool for pupil humiliation and the preserve of private schools. Hunt (2022) states that this elitist argument stands, stating that there are few teachers outside of very well-funded schools and universities capable of teaching composition, and Beard believes that composition often condones snobbery and drives the gatekeeping of class privilege (2013). However, pecuniary objections are contested as being false and proxy arguments about something else (Pearcy, 1998). Hunt (2022) argues that the claim of composition being a means to humiliate pupils does not stand post 1970. Newman (1990) is dismissive entirely of the elitism objections as being an unimportant discussion within the grand discourse, which does not sooth tensions of the emotional and political quagmire.

Latin has been seen through a political lens since the educational upheaval of the 1960s when the reading-only Cambridge Latin Course (Cambridge Schools Classics Project, 1998) was born, chiming with the anti-elite zeitgeist (Gay, 2003), and the later removal of O levels and the Oxford and Cambridge Latin entry requirement. The political view of Latin stands to the present day. Composition is more recently seen as a neo-traditionalist, borderline right-wing social-justice weapon of the Conservative government since 2010 (Hunt, 2018). These lenses, however, are often absent. Saunders (1993) states that composition itself is not elitist, but attitudes about it can be.

Discussion of composition often cannot escape discussion of what the purpose of studying Latin is. For some (Coffee, 2012; Beard, 2013; Hunt, 2016), the purpose of studying Latin is to be able to read ancient texts, with composition cited as a tool for this (Coffee, 2012) or not (Beard, 2013; Hunt, 2016), and opponents justify its removal from study on this basis (Saunders, 1993). However, not all students will go on to read original Latin, and with this narrow goal, students’ learning suffers (Saunders, 1993; Pearcy, 1998; Carpenter, 2000). For some, composition is a tool that can enable different goals of Latin study. These can be goals in their own right of increased Latin proficiency and historical and cultural knowledge (Pearcy, 1998; Carpenter, 2000; Coffee, 2012). Coffee (2012) and Hunt (2022) state that the secondary or utilitarian benefits of learning Latin can be a goal of Latin study too: improvement of English, brain training, literary appreciation, increased modern language ability, and increased knowledge of language structures are worthy aims of Latin study. Therefore, the goal of Latin study and the goal of teaching prose composition can depend on the level of the student and the goal of the student.

Although composition is surrounded by a thick haze of political, social and pedagogical discussion, some arguments are less valid than others. Composition can slot into a Latin learner's goal, which does not have to be universal.

What is composition?

Composition takes many different forms in the literature, falling into the prescribed composition category which is typically seen in traditional British education (Hunt, 2022), or the free composition category which is increasingly more common in the United States (Hunt, 2022).

Prescribed composition takes many forms, and composition's critics focus on this category. A typical method is turning disconnected sentences into Latin (Hunt, 2022). It is partly thanks to this that composition has fallen out of favour due to its content lacking relevance to anything (Dugdale, 2011) and overtly being a grammatical challenge (Ball and Ellsworth, 1989). Surprisingly, Hunt (2016) defends this method as it confines the composition to the syllabus. However, it does not show students all the possibilities of what you can do with Latin (Dugdale, 2011). Although this form of composition seems old-fashioned, it had its contemporary critics, being called needlessly pedantic and uninteresting (Derry, 1940). Recently this format is still deemed disengaging as an exercise in churning out someone else’s thoughts into Latin (Minkova and Tunberg 2004). Another method is to take a translation of an ancient author, and then try to write in Latin in the author's style (Matz, 1986; Maltby and Blecher, 2014). This method supposedly gets away from the robotic nature of prescribed sentences and gives the student the added interest of historical context. Unsurprisingly, Ball and Ellsworth (1989) criticise this practice as being both pointless and elitist. Another form is turning contemporary works of literature into Latin. Although Beard (2013) denounces this practice such as in Terrence Rattigan's The Browning Version where The Lady of Shallot is tediously and pointlessly turned into Latin, Lord (2006) shows it has many positives. Lord uses famous passages such as Mandela's Long Walk to Freedom. Lord does this as they recognise that modern students have a broad education and wide sense of social issues, raising compositional relevance and interest. Kershner (2019) gives students pop songs to be put into Latin in the artist's style.

On the other hand, free composition allows students to write without restraint, and there are various methods. Among the more popular forms are re-working exercises: these are asking or answering questions, summarising a text, writing titles for passages, simplifying complex sentences, setting a phrase in a different grammatical construction, or changing vocabulary to alter a text's meaning (Davisson, 2000, Minkova and Tunberg, 2004; Gruber-Miller, 2006; Minkova, 2009). These are free because there is no prescribed English to be put into Latin, but they offer limiting boundaries so as to not make the task intimidating. Holke (2019) endorsed a political slogan graffiti project as it was relevant to his students’ point of study, and Dugdale (2011) also practises these shorter-form free compositions. Continuous prose is also a method of free composition, and the literature shows the importance of guiding students to this point (Davisson, 2000; Benneker, 2006; Gruber-Miller, 2006; Lord, 2006; Dugdale, 2011). Free composition is flexible and diverse, but it is important for students to be guided.

Although authors tend to be in one of the two camps, Newman (1990) takes a level-headed view and endorses free and prescribed composition, stating that both practices inform each other and improve Latin. Both free and prescribed composition have their varieties, benefits and drawbacks. The content, method and limitations of the composition contribute to the success or failure of the compositional activity and it is the limitations placed on it by the teacher which decide its success or failure.

Reasons to teach prose composition

There are many voices in favour of teaching composition in the literature. Composition is said to improve overall proficiency in Latin, with some stating that it aids reading Latin (Pearcy, 1998; Coffee, 2012; Holke, 2019), that reading and writing are interdependent skills (Ball and Ellsworth, 1992) and even that writing continuous prose shows students that Latin is more than something to be decoded and aids fluency (Gruber-Miller, 2006; Hunt, 2022). Hunt (2022) cites DeKeyser (2015) who says reading and writing are different skills that need to be taught differently, so writing Latin may not improve reading.

Composition is another tool by which grammar and syntax can be learned (Matz, 1986; Pearcy, 1998; Davisson, 2000; Fogel, 2002) and reading-only approaches are criticised because they fail to equip students with grammar proficiency (Sharwood Smith, 1977; Gay, 2003). However, Hunt (2022) states that as reading-comprehension methods have more success, the argument that composition aids grammar and syntax learning becomes less valid. Harklau (2002) says a lack of writing in the Modern Languages classroom means students lack key syntactical learnings for language. Conversely, Lo and Hyland (2007) state that in their English as a Second Language classroom they saw a fall in accuracy during compositional activities. Free composition can allow students to create personal connections to grammar and explain it to themselves (Gruber-Miller, 2006; Sinclair, 2018).

Composition for beginners shows disagreement. Derry (1940) (as may be suspected from the year of her writing), Pearcy (1998) and Gruber-Miller (2006) agree that composition is useful for the novice. Ball and Ellsworth veto composition in the first two years of Latin study because it tests multiple things at once (1992); this shows poor understanding of the possibilities of composition. However, Hunt (2016) states free composition is better suited for university study because of its difficulty and accrued knowledge required.

Students in England have the choice of a compositional part of their GCSE and A level exams, and this may be one of Hunt's (2022) utilitarian reasons for writing in Latin. They also must engage with ancient authors in these exams, and composition is one way to achieve this (Hunt, 2022). Composition helps students to appreciate an author's nuances (Matz, 1986; Newman, 1990; Pearcy, 1998; Fogel, 2002; Gruber-Miller, 2006), and Lee (1964) states that composition achieves this goal to a higher level than any other method. Ball and Ellsworth (1989) say that composition teaches students inauthentic Latin, and Hunt questions the correlation between composition and improved literature reading (2016 and 2022). Batchelor states that student confidence in their ability increases by writing in Latin (Batchelor, 2018), which is crucially important for success in any subject (Deans for Impact, 2015).

Finally, composition is cited as being an important tool for communication. Although Ball and Ellsworth (1992) state that communication only happens in living languages, other authors say that composition is an aid to communication (American Classical League and The American Philological Association, 1997; Davisson, 2000; Dugdale, 2011), especially in a language classroom which often lacks an oral component (Lee, 1964). Interestingly, composition is more commonly practised in America than in England, with it being placed within the first standard of Latin Learning (American Classical League and The American Philological Association, 1997). Even in Modern Languages classrooms writing has been marginalised as a form of communication (Harklau, 2002) despite literacy's centrality to communication (Grauberg, 1997). Natoli (2018) says that students want to communicate in Latin, thus rendering them active participants in their learning. This is a change from Wilding's benefits of Latin composition (1955), stating that the purpose of education is to learn to communicate concisely, at which the Romans were experts. Clearly the reasons to write in Latin have changed over time.

Overall, in my opinion, the list of reasons to teach prose composition outweigh the reasons against teaching it. Often the problems cited can be remedied.

The importance of creativity in composition

The literature shows that for composition to be beneficial and successful, creativity is key. Composition affords students self-expression that they do not find in a reading-only course (Barrett, 2020) and students are proud of themselves for writing their own ideas in a foreign language (Benneker, 2006). Derry (1940) and Dugdale (2011) state that students ought to be able to write about things of contemporary interest. Conversely, Matz (1986) says that a historical setting for composition increases student motivation, and others state it is important that students are connected to the culture of the language (Spinelli, 1998; Gruber-Miller, 2006). Either way, writing is another modality for students to express themselves (Harklau, 2002) and their interest is raised when participating in creative activities (Sinclair, 2018), due to the creativity and flexibility they might not otherwise get (Reinhard, 2012). Creative tasks can daunt some students who usually perform well (Lo and Hyland, 2007) and some students can struggle with being creative (Morgan, 1994). Generally younger students are more enthusiastic about creative tasks (Morgan, 1994). This creativity gives students ownership over their learning, giving authorial control which promotes engagement and motivation (Dugdale, 2011) and a sense of personal involvement in work that is not artificial (Morgan, 1994); this raises a self-determined motivation and a belief in the ability to achieve (Deans for Impact, 2015; Gray and Macblain, 2015). Morgan (1994) warns that students can use this as a forum for talking about highly personal and off-topic issues. This can be addressed by the teacher if it happens.

Creative composition is cited as particularly beneficial to weaker students within Modern Languages teaching. Grauberg (1997) states that they get a means of personal expression, and Lo and Hyland (2007) state that composition is liberating and confidence-boosting for lower prior-attainers as composition is not seen as a test. They also say that the more able students are challenged in their perspectives of what could be done with the target language. Indeed, composition can enable students to learn grammar outside of their (possibly) usual stylistic vacuum (Fogel, 2002; Holke, 2019) and they can even learn to write in different registers (Harklau, 2002).

Creativity helps students engage and take ownership.

How to implement composition in the classroom

Some authors provide methods of how to effectively implement composition in the classroom.

Scaffolding is shown as one of the keys to success. Authors recommend building up task difficulty (Davisson, 2000; Gay, 2003; Gruber-Miller, 2006) from simpler exercises to more difficult ones. Re-working exercises are recommended as an initial activity (Davisson, 2000; Minkova, 2009; Barrett, 2020) so as to avoid students being intimidated by writing and this takes account of students’ working memory (Husbands and Pearce, 2012; Fordham, 2017). Support is recommended in the form of using the textbook as inspiration (Dugdale, 2011) as it can give grounding to those struggling. Placing limitations on students is important (Dugdale, 2011; Reinhard, 2012). Although students like flexibility, they do need some confines within which to work, and students need to be informed of what they are and are not able to express in Latin. Lord (2006) states that modern students, compared to those of yesteryear, require more grammar and syntax help, and therefore she glosses unfamiliar vocabulary and syntax. Sinclair (2018) advises giving students easy phrases appropriate to their current stage of progress. Lo and Hyland (2007) even advise giving students examples of good writing and the process of getting to that point. For students not to waste time searching for vocabulary, it needs to be limited (Lord, 2006). A controlled vocabulary list can be given (Benneker, 2006) or students can be instructed to use the words they have already learned (Dugdale, 2011). Whatever the tasks are, they ought to be sequenced to build in difficulty (Gruber-Miller, 2006). Ways to support students are therefore to be tailored to the needs of each class.

A few authors show that the student voice ought to be taken into consideration for successful composition. Student age has an impact on the willingness to participate, with Morgan (1994) stating that younger children are more interested than older children in being creative in the classroom. Student reactions ought to be monitored when being asked to write in Latin and be creative (Barrett, 2020). Lord (2006) recommends that students write about things of relevance to their own world, since it is common when learning other languages to do it through everyday vocabulary.

A variety of compositional tasks is advocated by many authors. Harklau (2002) states that the reason for this is that in our modern lives we experience multiple modalities of written communications, while Husbands and Pearce state that a range of techniques is a vital part of effective pedagogy (2012). Deagon (2006), on the other hand, states that variety is good for all the different cognitive styles that may be present in the classroom, as it can allow for diagnosis of errors which may be hiding. Dual coding, whether it be with songs, videos, cartoons or pictures is recommended as it allows for students to make everyday links to Latin (Reinhard, 2012) and because we are better at remembering things if we see them in more than one way (Fordham, 2017). Variety is also provided through different groupings of students in compositional tasks, as is increased participation and enthusiasm (Sinclair, 2018). Enthusiasm can be increased through making a creative task the reward for good class work (Reinhard, 2012) although this may reduce enthusiasm for the main part of the lesson. On the whole, a variety of compositional tasks seems to be beneficial.

Finally, there is not a huge amount in the literature as to how one ought to measure the results of compositional tasks. Student reflection surveys are recommended, as well as the teacher's own observations (Sinclair, 2018; Holke, 2019). Quizzing before and after a compositional project may offer some data (Holke, 2019) as diagnostic tools and summative assessment, as well as taking in the project itself. Common sense ought to be the best approach.

Lesson programme

Lesson 1: Start of CLC Stage 6. Perfect and imperfect tenses (See Supplementary Appendix).

Aims: To introduce students to the perfect and imperfect tenses. Students to be able to identify each tense and differences. Students to be able to translate tenses and explain differences in meaning.

Lesson 2: Stage 6 CLC. First story and further work on perfect and imperfect tenses (see Supplementary Appendix).

Aims: Students to be able to identify perfect and imperfect tenses. To recall meaning of each. To understand meaning and to be able to translate within passage of text. To be able to produce written perfect and imperfect tenses in correct person.

Lesson 3 – Start of mime project. Students to plan mime storylines and write out vocab list. Using perfect and imperfect tenses (see Supplementary Appendix).

Aims: Students to demonstrate knowledge of perfect and imperfect tenses. To be done via building a short story for a mime performance. Demonstrate knowledge of vocabulary acquired.

Lesson 4 – Students to compose their mime stories in full following planning and vocabulary finding (see Supplementary Appendix).

Aims: Students to put into practice their newly learned tenses by writing their mime stories in full. To use the tenses appropriately to convey the meaning desired.

Lesson 5 – Students performing their mimes (see Supplementary Appendix).

Aims: Students to demonstrate their compositions and use of perfects and imperfects by performing their mimes to their class.

Evaluation

Creativity and freedom

By and large, students enjoyed choosing their own subject matter and plots (Derry, 1940; Benneker, 2006; Dugdale, 2011; Barrett, 2020). There were high levels of engagement, motivation, and storyline variety. One group even made their own dummy character by stuffing clothes. There were a couple of issues with this, however. Although it was explained to students that they were to use the vocabulary that they knew, and to use the word bank provided (Benneker, 2006) or the textbook for help, some students wished to tell stories of wizards poisoning the emperor or late-night pizzeria fights (see Supplementary Appendix). While this showed genuine interest in self-expression in Latin (Grauberg, 1997; Harklau, 2002; Lo and Hyland, 2007; Minkova, 2009; Natoli, 2018) students lacked the vocabulary to do so. Students were set for homework to fill their own word banks on their planning sheets to aid their team writing the stories next lesson. Although it was not suggested anywhere in the literature, I settled on providing students with an online word study tool and advised them to choose the word form they recognised most so that students could write about what they wanted.

Some students were finding it difficult to be creative across the series of lessons. When completing a storyline under a cartoon, where some of the sentences were started with a prompt, some students struggled to understand that they could choose the words they wanted to complete their sentences. Again, a word bank was provided for help (Lord, 2006; Benneker, 2006). However, the structure that was provided seemed to limit students to thinking there was only one appropriate word to put in the gap, despite being told they could use whatever word they wanted so long as it made sense. The following extension task to this was to write short sentences of own choosing in Latin. Some students found this a struggle, stating they did not know what to write about (Morgan, 1994; Lo and Hyland, 2007). These students were guided to write about whatever they wanted, or to use the book for character inspiration (Grauberg, 1997). Some however really enjoyed the freedom and wrote a few nice sentences manipulating tenses, such as ‘femina canem ambulavit [the woman walked [sic] the dog]’, ‘canis erat ferociter [the dog was brave[ly] [sic]]’. When put into groups to write their mime stories, I did not see students struggling with creativity, possibly indicating that group work is a solution to this problem.

Placing limitations on creativity was a difficult task. Vocabulary lists were provided to prevent too much time being dedicated too looking up vocabulary, or the dreaded internet search. The lists were intended to guide subject matter. Largely this worked, however I am not sure how much it influenced student subject matter. There were surprisingly few ‘How do I say this?’ questions, and often students were asking something which was already provided on the sheet. Sometimes, I did provide vocabulary when asked because time was ticking by, and to allow students to express themselves, despite explaining that students needed to consider if they already had the tools to write what they wanted (Dugdale, 2011). This was mostly taken on board, but students did need help in rethinking how to use their acquired knowledge to express in Latin what they wanted and write intuitively rather than making a word for word translation into Latin of pre-written English, which I was trying to avoid (Hunt, 2016). Something which did surprise me was that one student, looking at the word bank, commented that they could spot similarities between English words and Latin words, and asked if English words came from Latin. Although the word bank was limiting, it also may have increased students’ awareness of the relevance of Latin and to make personal connection to it (Morgan, 1994; Grauberg, 1997; Gruber-Miller, 2006).

These planning sheets were effective in helping the students reach their end goal of writing a short mime story (Reinhard, 2012). It gave students a story structure to follow and length guide while choosing their own subject matter. The cartoon compositional task the previous week was in the same story structure as this sheet, which in turn was built on the story format of the pugna story from CLC Stage 6. I am not sure if these demonstrations of story structure over the course of the lessons had any effect on students’ accurate uses of perfect and imperfect tenses. A lack of familiarity with writing in Latin did not seem to be a barrier to the students’ enthusiasm to write in it, so perhaps the series of composition tasks, building in difficulty and freedom, was effective in increasing student engagement with the language (Dugdale, 2011; Sinclair, 2018). If doing this again, an improvement I would make would be to give several scenarios, settings or characters. This would help struggling students and provide more limitation for stories.

Accuracy

Lo and Hyland (2007) report that levels of accuracy fell during compositional tasks in the Modern languages classroom. Between students’ usual reading course and their new format of written work, it was hard to find direct comparison between errors made in their usual reading course and errors made when using this new writing skill, and if these errors were down to lack of familiarity alone. Errors were generally such that I have seen in students’ work and assessments prior to the study. Students sometimes failed to use singular and plural verbs appropriately. Another typical error was failing to put the object of a sentence into the accusative. Something which I wasn't expecting was for students to use a mixture of the present, perfect and imperfect tenses throughout their final piece. Although I tried to labour the point over the course of the lessons by emphasising their plots should be narrated in the past tense and modelling a good example (Minkova, 2009; Lo and Hyland, 2007), lots of groups started off in the past tenses, but began to use the present tense as their writing went on. I gave some verbs in the past tenses such as to be in the word banks, but I did list most verbs in their present form, as part of the objective was for students to recognise what tense they needed, and to manipulate the verb into the right tense, to align with their current stage of learning the perfect and imperfect tenses.

The students who wrote short sentences of their own choosing, using the perfect and imperfect tenses as an extension task, showed a surprisingly large amount of accuracy, and it was students from across the class who partook. The only error was one student writing ‘senex es sedetbulat [sic] [the old man you're was-walk-sitting]' which still shows an attempt to use the imperfect. On the whole, grammatical errors did not demonstrate a complete lack of understanding of the tenses and errors weren't wildly unexpected or prevalent. What was unexpected was the creativity and sheer variety of stories that students produced for their groups to perform (Dugdale, 2011). Students were able to see the possibilities open to them through Latin (Pearcy, 1998) and did not view writing in Latin as some insurmountable task; they were able to use Latin in an active manner to communicate with each other (Coffee, 2012). When it came to students performing their mimes to each other, what few errors were remaining didn't appear to be a barrier to communication with their audience, who knew what was happening during each and every performance (Gruber-Miller, 2006). Ultimately, each student was manipulating verbs to be in either the perfect or imperfect tense. The compositional task alone did not provide enough feedback to monitor students’ understanding of these tenses, so it was incorporated with other summative assessments such as games of noughts and crosses to allow students to demonstrate their knowledge in other formats (Derry, 1940).

Scaffolding and feedback

Getting the scaffolding right for this study as well as feeding back to the students was challenging. Building up the tasks for the students, whilst also wanting them to write their mimes and fitting this all into five lessons was tough (Davisson, 2000; Gay, 2003, Gruber-Miller, 2006). The first task of swapping tenses was an introduction to writing in Latin, and may have been overly simple for some students, but these students were offered an extension. Most students did have errors. The first task of tense-swapping verbs between perfects and imperfects I had thought might not be challenging enough; however, most students had errors. In fact, the first two lessons contained dual coding in the form of cartoons. Although I had thought this would allow students to make connection to the Latin (Harklau, 2002), some students failed to realise that the sentences beneath were supposed to describe the action in the picture which was intended to guide their composition.

I could have made more use of modelling throughout this study. Perhaps modelling how to fill out one of the planning sheets would have been useful for the students to see, and it might have highlighted for them the importance of using the perfect and imperfect tenses (Minkova, 2009), although this was indicated to students by the questions on the planning sheet. I could have expanded on the modelling of writing the stories. I showed myself writing Latin under my English story, however some student interaction in this demonstration would have benefitted their mimes, especially for those who were struggling to commit their stories to paper. Although surprisingly these students were the most able. This was probably because they were either too concerned with perfection or had written more off-the-wall stories, so perhaps the lower prior-attainers really did feel the benefit of free composition (Lo and Hyland, 2007).

Feedback was difficult. I found I was very busy during the lessons with students who had their hands up for help and found that I was not able to give live feedback to students as they composed in a manner I would have liked. With the second cartoon sheet, I did find it was better as students were used to writing in Latin and more used to the perfect and imperfect tenses, and I was able to offer a higher percentage of help. I think, however, I could have done some whole-class feedback as it would have probably been beneficial for all. I felt I also struggled with feedback as the literature suggests multiple drafts and corrections, something I did not have the time for (Matz, 1986). I did take books in after students had completed writing their stories before their performance lesson, and errors here were not numerous. I did do whole-class feedback the following lesson and this seemed to right most of the errors. In future, I would build in feedback into a project like this, and to consider what scaffolding is appropriate for the students.

Composition form and lesson sequence

The sequence of compositional lessons, both of themselves and within the curriculum map, worked well. The end goal of performing a mime was based on students being set homework before the study to colour in mime masks as they studied the Roman theatre. Students were desperate to use these masks as soon as they had done them; therefore I thought that this project would be motivational for them, to be working towards their own drama festival (Dugdale, 2011). Some students’ plays did match to their masks, and some chose to keep their stories about the slavery theme of CLC Stage 6.

The preparation for this project was an undertaking. For the lesson with cartoons and gap fills, I chose a famous family from Springfield as a familiar set of characters for the students and a readily available source of domestic images (Lord, 2006). On the whole I found that there were pictures readily available that I wanted; however I dedicated perhaps too much time photoshopping characters into the room of the house I wanted for students to be able to describe them. The literature indicated that dual coding (Fordham, 2017) was beneficial for students and so I was keen to do it. However, it took a lot of time and I even considered using an AI tool to get the images I wanted. It took far too long to create this resource, and had I been working a full timetable this would have had large knock-on effect on my workload. It made me think that the prevalence of composition in schools does depend on department resources and capabilities (Hunt, 2022), although I do think that this is not something unique to composition and the Latin classroom. In future I would have to be less precious about getting the exact images I wanted.

Compositional tasks were designed to build in difficulty as students became more familiar with the idea of writing in Latin (Davisson, 2000). One thing which I think worked fairly well was using the course book as a springboard for composition, so that it did have relevance for students (Dugdale, 2011). Cartoon strips included images and characters from the books as well as scenes that students had encountered previously. One thing which I am not sure sank through to the students, for whatever reason, was that uses of the perfect and imperfect tenses ought to occur at different times in their stories. I tried to highlight this during whole class translation of the pugna passage in Cambridge Latin Course Stage 6. Students were able to identify the perfect as a ‘short action in the past’ and the imperfect as a ‘long action in the past’ when questioned. Also, on their planning sheets there was guidance for which tense to use and where. However, there were mixed results across their mime stories in using their past tenses in this manner. Writing in continuous prose was a real departure from the norm for them. Although they all found meaning in doing so, it was perhaps a bit unrealistic to expect nuances such as this (Gruber-Miller, 2006; Hunt, 2022). The focus class is a particularly theatrical class, so producing mimes was motivating for all and added an element of familiarity to an otherwise alien task (Lord, 2006; Dugdale, 2011).

One of this study's aims was to see if student recall and retrieval could be improved by adding variety to their reading course with minimal teacher contact time (Deagon, 2006). Group work shows the collective knowledge rather than testing individuals’ knowledge, and I did not wish to quiz individuals to avoid taking up lesson time and for students to feel like they were being tested (Lo and Hyland, 2007). Overall, however, students found the project meaningful and enjoyable, and all contributed to their groups’ success of producing a good chunk of quality written Latin. It would have been interesting to see if there were noticeable differences in their work in following lessons.

Conclusions

I started this project with a view to seeing if adding variation to a reading course through composition could improve a class's retention and recall, and Latin fluency. It is difficult to measure results, since these lessons were the last that I taught these students. However, all students did achieve success in writing in Latin, and using the perfect and imperfect tenses. It would have been better if more of the literature had shown how to measure the results of compositional activities and how to measure the benefits of writing in Latin. If I were to do this project again, I would allow for composition to be spaced out across more lessons, and to build in feedback and modelling more effectively. I suspect that there will always be similar issues of student level of proficiency limiting compositional scope (Hunt, 2022); however, some of the workarounds I tried proved effective. Most effective were the vocabulary word bank, planning sheets and motivation of students to perform their own work. Overall, I have learned much during this study about how students process and produce Latin. I really had not anticipated the high level of enthusiasm there was from students to write in Latin from those who had never written in Latin beyond note taking, and who were so early on in Latin learning. I would be interested to see how I would implement composition with different age groups and abilities.