A study in the use of embedded readings to improve the accessibility and understanding of Latin literature at A Level

Introduction

Context and motivation for research

Part way through the second school placement of my PGCE teacher training course I began to teach Latin unseen translation, where students are given an unfamiliar passage to translate, to a Year 13 cohort of just two students. The translation element of the course was proving challenging for both students – they struggled with language comprehension, and the A Level language course seems not very accessible to students with low prior attainment. Furthermore, my second placement school is an Upper School, so students start Latin from scratch in Year 9. Because of the time constraint those who choose take GCSE do the Eduqas Latin exam, which requires that slightly less grammar and syntax needs to be covered than the OCR syllabus (for example, the ablative absolute construction is absent from the Eduqas examination, but present in the OCR). For the small number carrying on in the 6th form, this results in a linguistic shortfall which has to be made up before beginning the OCR A Level specification. This seems to be a challenge even for very able Latinists, and more so for my two students, Alice and Michael (not their real names), and had presented an added barrier to their facility in grasping and translating complex language structures, both in unseen translations and the demanding set texts. Simply ‘practising’ by going through texts together was not proving helpful, as once my oral prompts during this process were removed and they were left on their own they reverted to guesswork instead of applying their knowledge and logic to the text. Similarly poor results in translation were evident from practice literature papers which they had taken before February half term, even though these were prepared texts – while the language paper asks students to tackle short unseen translations, the literature component calls for the extensive preparation of a much longer text, from which students must translate and analyse selected passages in the examination, as well as answering a broad essay question. However, despite the fact that Alice and Michael had been through their set text in detail with their teacher and made their own translations, it appeared that, when presented with a passage from the set text in exam conditions, the language looked just as intimidating to them as if they had not seen it before, and they were equally unable to produce a sensible attempt at translation. I therefore became interested in finding ways for them to approach texts which would minimise the ‘fear factor’ and allow them to access meaning without being put off by complex syntax.

Choice of research question

During one of our PGCE faculty sessions we had been introduced to the technique of embedded reading, also known as tiered reading, which broadly involves the simplification of a text in order to create versions with differing levels of complexity, which can then be used to improve comprehension of the original. I was very intrigued by this method, and wondered if it could help my two Year 13 students to feel more confident in reading complex and challenging Latin. I hoped to use aspects of the technique to give students the tools to deconstruct and simplify texts themselves, in conjunction with other techniques for helping unseen translation. However, on researching the method it soon became clear that it could be more comprehensively and usefully applied to set texts. As Year 13 was the only year group currently reading their set texts, it made perfect sense to use this method with them, with the hoped-for side effect that it would also improve their confidence when approaching unseen texts. As I began to think about how I would prepare the text, I realised that the tiered reading approach could also prove effective for the literary analysis element of the A Level literature exam, as stylistic points such as word order and choice of particular vocabulary and syntax ought to arise organically as a result of comparing the original text with the simplified versions. I hoped, therefore, that this would give them easier access to and deeper understanding of the literary criticism, by allowing them to actually see, rather than be prompted to imagine, how a text could look in simpler form, and therefore why the author wrote it in the way they did.

Summary of method and findings

At this point in the term they were part way through reading their prose text, a selection from Tacitus’ Annals 1, so it was decided that having observed their usual method of going through the text in a more traditional way (partially preparing each section in advance before going through it in the lesson and then going back over it to identify stylistic points), I would then take over for the final third of the prescription (Ann.1.46-49). As I did not have to deal with side-issues of discipline or significant lack of motivation, I could focus primarily on my preparation and effective delivery of the text. However, the proximity of the exams also meant that this had to be a very focused study, and could not take any risks.

The study is naturally also limited by there being only two students, as well as by the fact that there is no objective measure of how effective the method had been in comparison to their previous teaching. However, the responses which I received from Michael and Alice, as well as my observations during the lessons, indicated that, as far as their own perception was concerned, the effect of the method had been very much what I had hoped for, and in line with the small amount of other research which has been undertaken on the use of embedded reading in Classics to date. The main downside was the lengthy preparation required on my part, which must be weighed up with the benefits of using this technique.

Literature review

To my knowledge there is a limited amount of scholarly literature on the use of embedded readings, and even less that is specific to the Latin classroom and Latin literature in particular. Much of what does exist is not in published articles but in blog form on the embedded readings website, which I shall refer to frequently. As such, I have found it appropriate to divide this literature review into two parts. The first will deal briefly with the history of teaching Latin literature in schools, before covering various approaches which have been taken to making it more accessible to school-age students. The second part will give a theoretical and practical overview of the method of embedded reading and its applications in English literacy and modern foreign languages, before examining in detail the small body of research which has so far been undertaken on its potential for use in improving the accessibility of Latin texts.

Latin literature in schools and reading approaches

The 1966 issue of the journal Didaskalos shows a particular interest in the way Classical literature should be taught in schools, following the publication of the textbook Aestimanda (Balme & Warman, 1965). The aim of the book was that classical authors should be treated ‘as literature rather than as a matter for linguistic exercise alone’ (Balme & Warman, 1966, p. 46). Until the second half of the 20th century the priority in teaching literature had been linguistic understanding, an approach which Sharwood Smith criticised as being ‘incurious about the nature of the text as a piece of literature’ (Sharwood Smith, 1977, p. 50). Other articles in the 1966 Didaskalos show reaction against this. Doughty criticises the approach of Aestimanda, but stresses that

the teacher should never forget that everything depends ultimately upon his ability to create a situation where a powerful first response to poems can be a reality. Without the initial response, the rest is so much verbal abacus (Doughty, 1966, p. 26).

These critical assessments point to a sea-change in the reasons for and methods of teaching Classical literature at school level, a very positive change, but with the inevitable result that the amount of original text which can be covered is smaller, as time spent on literary analysis and appreciation has replaced time spent on language work.

This means that, although the attitude towards teaching Classical texts has firmly shifted towards literary appreciation rather than linguistic exercise, there is a gulf in difficulty between set texts and the confected Latin which most students are used to reading in their course books (most commonly the Cambridge Latin Course), such that in practice understanding the language of the set texts demands a great deal of attention, and they are not easy for lower ability students to access. One consequence is that literature exams have, for some, become an exercise in rote learning rather than translation – anecdotal evidence suggests that learning set texts off by heart is seen by many students as preferable to translating them in the exam. These texts certainly are not accessible for students to ‘read’, since, as mentioned above, reading courses such as the Cambridge Latin Course use confected texts which are much more predictable in terms of syntax and word order than most original authors. Therefore, as much of the literature allocation is still spent grappling with understanding the language, it is difficult to elicit any kind of ‘first response’, the importance of which Doughty emphasised, other than bafflement.

This difficulty is noted by Davies (2006), who defines ‘The Problem’ with literary texts as being that the average student depends on certain assumptions based on word order and verb placement, which are confounded when they approach a work like the Aeneid (Davies, 2006, p. 173). Davies’ solution is to retrain students using a reading approach, wherein they are taught to read by the line or half-line:

This encourages them to read the Aeneid as it was meant to be read, rather than treating it as a puzzle created by Vergil for them to solve. The students also begin to experience the text directly, rather than worrying about writing down and then memorising the right translation. In addition, if they can learn to read the Aeneid, rather than solve and translate it, they will begin to see for themselves many of the poetic effects that make the poem great. (Davies, 2006, p.174)

Being able to ‘read’ Vergil, or other difficult authors, is something all Classicists aspire to (or ought to). However, while Davies (2006) does suggest some activities to aid this, such as themed vocabulary learning and identifying agreement, it is difficult to see how this type of ‘retraining’ could be achieved over a short period of time. My project aims, by means of the method of scaffolding discussed below, to achieve much the same outcomes which Davies (2006) identifies, namely allowing students to feel able to ‘read’ the set text without ‘translating’ it at the same time and also to independently notice some of the literary devices.

Embedded reading

Embedded reading is a method developed by two modern language teachers, Laurie Clarcq and Michele Whaley. Clarcq's introductory article on the method defines it as follows:

An embedded reading is three or more scaffolded versions of a text. It is designed to prepare students to comprehend text that the students perceive to be beyond their capability. Embedded readings provide information in the target language in a way that actually develops the students’ reading skills. Using embedded readings not only allows students to acquire language, it provides a framework for improving reading abilities. (Clarcq, 2012, p. 21)

Embedded reading is very similar to what is sometimes termed ‘tiered reading’, a method of improving literacy which has largely been researched in the context of English literacy in both native speakers and students with English as a Foreign Language, aiming to reduce the intimidation factor of literary texts for students with below-average literacy. For example, an article by Moss, Lapp and O'Shea (2011) describes the results of using this method of teaching Shakespeare, a classic example of the kind of text which is very hard even for highly literate students to access, in a US high school. They identify a mismatch between the language students use at home and the academic English which they are required to tackle at school (Moss et al. 2011, p. 54). ‘Tiered texts’ are described in very similar terms to Clarcq's definition of embedded readings:

Students begin with an easy-to-read text aligned with their entry-level background, academic, and topical knowledge. As their bases of knowledge and language expand, they are able to read, discuss, and write about more difficult texts on the same topic, using their newly acquired topical and academic vocabulary. (Moss et al., 2011, p. 54)

In the context of the project in question, conducted by O'Shea at a High School in San Diego with the text of Romeo and Juliet, the first and most basic tier consists not of ‘texts’ in the strictest sense, but of graphic material, a rap and Zeffirelli's film version (Moss et al., 2011, p. 56), with the aim of this tier being to pique interest and give a broad knowledge of the plot of the text in question. In the second tier ‘The Romeo and Juliet Rap’ is analysed in more detail as a literary text, introducing the terminology and types of analysis that the students will need. The article reports that students were then able to tackle scenes from the original ‘with minimal teacher assistance’ (Moss et al., p. 59), although it does not explain how successfully the students are now able to deal with the difficult linguistic aspects of Shakespeare.

This model, then, while it has the same aims and structure as embedded reading as described by Clarcq (2012), is aimed more at developing broad knowledge of plot and character, along with understanding of the tools and terminology of literary analysis, than at aiding comprehension of difficult linguistic features. A different focus is seen in Clarcq's (2012) demonstration of a tiered reading of a story about an ant, where the tiers are purely textual, and where each version is literally ‘embedded’ in the last, using the same language but with fewer elements and simpler linguistic structures. Using this model Clarcq (2012) demonstrates how a reading can be created either from the top down, or from the bottom up. The basic aim of this method is to improve literacy in a foreign language, and this is where the idea of specifically linguistic tiered readings comes into its own. Clarcq (2012) is a Spanish teacher, and developed the method from seeing how this building up of levels of complexity, with a simple version written by the student within it, helped that student to access the language, after previously being completely disengaged. Not only content, but also specific vocabulary and syntax can be introduced this way to improve reading comprehension.

Very little work has been done on using this method for Latin. In his blog Todally Comprehensible Latin, US High School teacher Keith Toda has two posts on using embedded readings (Toda, 2014a and 2014b) with the stories in Stage 10 and onwards of the CLC, which he describes as becoming ‘long and actually quite complex grammatically and vocabulary-wise for students’, adding that ‘what ends up happening is that students go from reading the stories to now translating/decoding them’, thereby (he claims) losing sight of the initial aim of the CLC as a reading course. He gives an example of an embedded reading of a story in Stage 29, created by first rewriting the story in its simplest possible form, and then creating a version in between this and the original. He also gives an example of enodatio, a type of embedded reading which involves putting the original into English word order, with Aeneid, Book 1, lines 419-429 and 437. Toda acknowledges that this could invite criticism from those thinking that students should learn to read left-to-right, but counters this by saying ‘my primary concern is establishing meaning FIRST.’ My own rationale behind using the method was similar – it would be ideal if students could read Virgil or Tacitus from left to right straightaway, but since most cannot, giving a re-ordered text first is a useful aid to them being able to fluently read the original.

Toda's second post discusses a problem identified in his first, which is that if students do not find a story that interesting in the first place, they will not want to read it for a second time in a more grammatically complex version. He addressed this problem by giving the stories a different ‘twist’ each time. However, for my purposes, the aspect of repetition was not a hindrance but a direct advantage, as motivation was not an issue, but familiarity with the events of the text was. It depends, therefore, whether the purpose of using embedded readings is to motivate students to read by giving them texts which are less intimidating, or to give them easier access to a text which they have to read and remember. Toda's (2014) outcomes are nevertheless much the same as those which I hoped to achieve:

1. 1) Due to the scaffolding nature of the stories, it greatly lowers students’ affective filters regarding the reading itself. Handing students the original text probably would overwhelm them if it were too difficult or too long.

2. 2) Because students are reading multiple scaffolded versions of the same story, they are already familiar with what they are reading, thereby, they can anticipate vocabulary and language structures

3. 3) Due to the re-reading, students are getting plenty of meaningful/contextual repetitions of the language. (Toda, 2014a)

The only attempt to study in a more academic context the value of this approach with original Latin texts is an article by Sears and Ballestrini (2019). Similarly to Davies (2006), Sears and Ballestrini (2019) identify a basic distinction between reading and translation: ‘Translation, at its core, [is] not a comprehension activity’ - i.e. once students have translated a text, they usually do not know what happened in it. They identify tiered reading as a way of bridging the gap between the texts students are used to reading and those which they are expected to translate in exam. In the middle section of the study, the authors detail the process of creating a ‘top-down’ reading of a Latin text on the subject of the myth of Apollo and Daphne – word order is rearranged first, before syntax and vocabulary are progressively simplified. Sears and Ballestrini (2019) conclude that this method reduced the intimidation factor for of a long and difficult passage, reduced what they describe as the ‘drudgery’ of literary analysis by allowing it to be introduced more gradually, and engendered a deeper and longer-lasting understanding of the text. Therefore, although the amount of research and information on using embedded reading for Latin texts is small, it all points towards the same kinds of outcome, in terms of its intention for making texts feel less intimidating and helping to deepen student understanding. I hoped to replicate the same outcomes in my use of the technique, but with a slightly heavier focus on using the different versions to help with literary criticism.

Methodology

Because of the slightly unusual nature of my research, in that I had only had two students to work with, and a large part of the project consisted in working out how best to prepare the texts, my methodology will be divided into two sections: the first will discuss why I chose to conduct the research in the way I did, and the second will give an account of how I prepared the texts, and the reasons for doing so.

Method of research and data collection

Action research, in the formulation of McNiff, begins with the identification of ‘an issue that needs attention or investigation’ (McNiff, 2014, p. 25). At a wider level the issue which I focused on was the difficulty of approaching original Latin authors from a background of having read mainly confected texts which are designed for ease of reading rather than for maximisation of stylistic effect, such as those in the Cambridge Latin Course. More specifically, this problem was highlighted in the case of two A Level students at my school whose level of language competence was lower than average for this stage in their study of Latin. I therefore chose to use a different and relatively untested approach to try and make a linguistically challenging text (a selection from Tacitus Annals 1) more accessible for them.

In terms of my methodology, there were, naturally, a number of limitations to my research. Firstly, a group of two students is an extremely small research sample, and because both were of a similar level, there would be no opportunity to compare the responses of students of different ability to the method. Furthermore, the small number of students removed any potential advantage of using an anonymised questionnaire to collect data, a methodology which can increase the honesty of student responses (McNiff, 2014).

Secondly, the timing of the project was not ideal: the section of the text which I was doing with them (Annals 1, 46-49) was their final section of set text, and with the research taking place over two weeks either side of the Easter break, by the time we finished the text they would have only a couple of weeks remaining before study leave, and thereafter the start of their A Level exams. However, while the lower ability of these students presented some challenges, it also released the pressure somewhat in regard of their examinations – neither was aiming for a very high grade, and it was likely that any intervention would help rather than hinder their chances. Moreover, neither was relying heavily on a grade in Latin for their plans after school: Alice already had an unconditional offer to study Maths at a good university, and Michael was hoping to enter the Royal Navy. Still, I did not want to place them under any extra stress at this point by making them feel as though they were being experimented on.

For this reason, I decided to take quite a low-key approach to introducing the project, just giving them a brief introduction at the start of the lesson sequence, saying that I was researching a different way of doing set texts, but not stressing the research element too heavily. I explained the method to them roughly as follows: ‘Basically, the idea behind this is that instead of trying to translate the text at the beginning, you build up to it by reading a simplified version first, and then hopefully that makes the original feel more accessible when you do translate it.’ I avoided going into detail about the exact aims of the method, as I wanted to see if they would identify them when I interviewed them afterwards, without having been told what I wanted them to say beforehand. I also decided not to formally interview them at this stage on their feelings about the accessibility of the text, although in hindsight this would have been helpful as a benchmark to compare with how they felt about it afterwards. However, having observed the lessons in which they read the rest of the prescribed sections with their usual teacher, I was able to note how that process worked, the length of time it took to get through a certain amount of text, and the level of confidence which the students had with the text covered.

I also spoke informally to both students in our unseen translation lessons about their difficulties with this aspect of the course: in particular, I was interested to get their own views on what I had observed, which was that they were actually quite confident with parsing, and able to translate when prompted with which word to go to next, but utterly baffled when left to work out a sentence on their own. Alice in particular was excellent at identifying grammar and syntax, but felt that she just panicked when there was no-one reminding her to do it. Interestingly, in an individual lesson with Alice I had tried out the method of cutting up the words of Latin sentences and asking her to arrange and translate them. She completed the exercise without too much prompting, but then, when I put up the original on the Interactive Whiteboard, telling her that she had already translated it, she remarked with a slightly crestfallen expression, ‘Oh, I didn't get the order right then.’ I explained to her that she had got the order right, and it was Ovid who had played around with it! I was quite astonished that she had not understood this concept until now, and this gave me more reason to hope that they could approach Tacitus with much greater ease if his language was put in a more familiar format.

For the reasons given above, then, my data collection was done solely during the lessons themselves, and in an interview that took place during the lesson directly after finishing the text. I decided to conduct the interview as an informal chat during the lesson and not as a separate interview in a more formal occasion, as I felt this would make the students more comfortable and able to express their opinions. There is still the danger that the students may not have felt comfortable to voice any negative opinions to me, since I now had a good relationship with both after teaching them unseen Latin – English translation for several weeks. I did consider also asking my mentor to also interview them, but since she had been teaching them both since they were in Year 9 we felt that this would be more of a problem with her, as they would be less likely to voice positive opinions for fear of giving the impression that they preferred the new approach to her method. I also ruled out the idea of asking them to type out anonymous written responses, as I felt that this was too much of an imposition at a pressured time, and moreover helpful responses were more likely to arise during an informal chat. Moreover, Michael is a fairly blunt individual, so I felt that he at least might feel comfortable voicing criticism.

Process of preparing readings

I will use Tacitus Annals 1, Chapter 46, the first section of the prescribed text, to demonstrate how I prepared the text. My first step was to make a literal translation, as follows:

at Romae nondum cognito qui fuisset exitus in Illyrico, et legionum Germanicarum motu audito, trepida civitas incusare Tiberium quod, dum patres et plebem, invalida et inermia, cunctatione ficta ludificetur, dissideat interim miles neque duorum adulescentium nondum adulta auctoritate comprimi queat. ire ipsum et opponere maiestatem imperatoriam debuisse cessuris ubi principem longa experientia eundemque severitatis et munificentiae summum vidissent. an Augustum fessa aetate totiens in Germanias commeare potuisse: Tiberium vigentem annis sedere in senatu, verba patrum cavillantem? satis prospectum urbanae servituti: militaribus animis adhibenda fomenta ut ferre pacem velint. [Tacitus Annals 1.46]

But at Rome, since it was not yet known what had been the outcome in Illyricum, and the discontent of the Germanic legions had been heard about, the fearful citizen body began to accuse Tiberius because, while he was messing around with the senators and people, weak and unarmed bodies, with fake hesitation, the soldiery meanwhile was discontented and could not be quelled by the not-yet-full-grown authority of two adolescents. They thought he ought to go and set his imperial majesty against men who would yield when they saw a princeps of long experience, who also held the ultimate power of both punishment and generosity. Indeed, Augustus in his tired old age had been able to visit the German provinces so many times: was Tiberius, in the prime of life, to sit in the senate quibbling over the words of the senators? Urban servitude had been amply provided for; emoluments now had to be applied to soldierly minds so that they would be willing to bear peace.

My next step was to establish what were the main points of content which I wanted to establish, and the context which this version could be used to discuss, so in this I wanted to be able to go over the rebellion in Illyricum, recap what Tiberius’ earlier actions had been, and make sure the students knew the identity of the ‘two young men’ and what they were doing. I would then write a summary of the passage in Latin which would convey the basic outline of the events. This part of the process, then, was different to that described by Sears and Ballestrini (2019), as it used a ‘bottom-up’ approach, therefore becoming an exercise in composition for the teacher.

While the main point of this version is to establish content, I also wanted to retain some of the original vocabulary and syntax, so that this could start to become familiar to the students. However, these are not genuine embedded readings, as most of the text which I created is not contained in the original in exactly the same form. Below is the version which I created -words underlined are genuine embedded readings, i.e. exactly the same as the original, while words in italics are semantic cognates of words in the original, but with different accidence and syntax. (These indications are for the purposes of this article and were not in the version given to the students.)

at Romae cives nondum cognoverunt quod in Illyrico accidisset, sed de rebellione in Germania audiverant. cives Tiberium incusabant quod, dum ille Romae cunctetur, milites dissideant. duo iuvenes non poterant milites comprimere. cives dicebant: ‘Tiberius debet ire et milites comprimere, neque in senatu manere.’

But at Rome the citizens did not yet know what had happened in Illyricum, but they had heard about the rebellion in Germany. The citizens began to accuse Tiberius because, while he was hesitating in Rome, the soldiers were discontented. Two young men were not able to suppress the soldiers. The citizens were saying: ‘Tiberius ought to go and suppress the soldiers, and not remain in the senate.’

Syntax was heavily altered for ease of reading, such as the removal of ablative absolutes in the opening sentence, change of historical infinitive to imperfect (incusare), substitution of collective singular for plural (miles dissideat), and the conversion of indirect speech to direct speech. Whole sentences were left out or paraphrased both as being non-essential in terms of meaning, as being complex, and as being interesting to introduce in the next version for layers of interest and meaning. In terms of vocabulary, some words were substituted for more familiar ones for ease of reading (accidisset for exitus fuisset, rebellione for motu, poterant for queat). Others were substituted because I wanted to talk specifically about the choice of a specific word in the final version (iuvenes for adulescentes, manere for sedere).

The second version was created top-down, similar to that of Sears and Ballestrini (2019), by taking the original text and removing, re-ordering or adapting in order to make the text simpler. The main difference in my approach to word order compared to Toda, who recommended putting the text in English word order (Toda, 2014a), is that I rearranged the text into the basic Latin word order with which the students, having studied Latin for a number of years, were now familiar. This allowed for greater stylistic impact, as when a verb is in initial position in a Latin sentence it is usually something that can be mentioned in stylistic analysis – putting the text in English word order would reduce the effect of what the author had done. By showing the text as it might have been written in straightforward Latin (my indifferent composition skills aside), I hoped to allow the students to see why the original was a stylistic improvement, rather than just unnecessarily confusing.

I will now set out each sentence of Versions 2 and 3 side by side, with bold text indicating where changes other than word order have been made. Below each I briefly explain my thought process in preparing the text as I did.

at Romae nondum cognitum est qui exitus fuisset in Illyrico, et cives de motu legionum Germanicarum audiverant.

at Romae nondum cognito qui fuisset exitus in Illyrico, et legionum Germanicarum motu audito, …

Here the main change was to replace the impersonal ablative absolute and supply cives as the subject. The change in word order would make translation easier as well as allowing for discussion of the placement of cognito … audito in the final version.

itaque cives trepidi Tiberium incusabant quod, dum patres et plebem cunctatione ficta ludificetur, interim milites dissideant, neque auctoritate duorum adulescentium comprimi queant.

trepida civitas incusare Tiberium quod, dum patres et plebem, invalida et inermia, cunctatione ficta ludificetur, dissideat interim miles neque duorum adulescentium nondum adulta auctoritate comprimi queat.

As well as changes to word order again, the historic infinitive incusare was replaced with an imperfect, a couple of phrases were removed to reduce the complexity, and collective singulars were replaced with plurals to prompt discussion.

cives dicebant: ‘Tiberius ipse debet ire et maiestatem imperatoriam opponere militibus. milites, ubi principem longa experientia vident, cedant.’

ire ipsum et opponere maiestatem imperatoriam debuisse cessuris ubi principem longa experientia eundemque severitatis et munificentiae summum vidissent.

Oratio obliqua, something which the students struggle to pick up, was replaced by speech marks and an introductory phrase. I wanted the students to go away from these passages with the impression of people in the city talking to each other, or of Tiberius thinking out loud, without being lost in the sea of accusative-infinitive. The tricky future participle, cessuris, was replaced by a periphrasis, and one difficult phrase was missed out.

‘Augustus, quamquam senex erat, totiens in Germanias commeare potuit. sed Tiberius, qui annis viget, sedet in senatu, verba patrum cavillans.’

an Augustum totiens in Germanias commeare potuisse: Tiberium vigentem annis sedere in senatu, verba patrum cavillantem?

Direct speech was continued, and the variatio of fessa aetate and vigentem annis was replaced in order to bring the focus of discussion onto this aspect.

‘satis prospectum est servituti civium. nunc fomenta adhibenda sunt militibus, ut pacem acciperent.’

satis prospectum urbanae servituti: militaribus animis adhibenda fomenta ut ferre pacem velint.

Again, direct speech was continued. This sentence had not been included in the basic version, as I found it too difficult to simplify and inessential for the sense of the passage. Consequently, I hoped that we would spend some time discussing it. Aside from direct speech and rearranging the word order to emphasise juxtaposition of servituti: militaribus I hoped to draw attention to the use of the phrases urbanae servituti and militaribus animis, and to the interesting ut ferre pacem velint (so that they would be willing to bear peace).

My intention was that each sentence of the second version would be brought up on the PowerPoint, and we would ‘translate’ it carefully together, before I brought the original up alongside it, and asked them to see if they could not read the original, with the scaffolding which had been provided by the middle version. This version, then, had to have a very similar meaning to the original, and contain most of the same vocabulary, in order to remove barriers to comprehension when they saw the original, just as Davies’ (2006) approach of preparing certain aspects of the text beforehand was intended to do. However, I also needed the middle version to contain differences which would automatically prompt stylistic discussion, in the hope that the students would begin to identify this for themselves.

Findings and discussion

The prescription of Tacitus was taught over five lessons, with each of the four prescribed chapters, including all three readings and close analysis, being covered in one hour-long lesson. The fifth lesson was used to finish off the very end of the text which we had not been able to quite finish, to hold an informal discussion with the students, and to go back over the text to identify broader themes for essay questions. I have no data, so my findings are based on my own observations in the lessons and the students’ responses in the informal interview which was held immediately after finishing the prescription.

Usually the students would prepare vocabulary and try to identify verbs in each passage in preparation for the lesson, so my aim had been that instead of this I would give them the basic version of the text to prepare beforehand, so that they could read it independently and also have time to learn the key vocabulary within it which I had glossed. However, I did not create the versions enough in advance, and therefore had to give them the basic version to translate at the start of each lesson. This ate into lesson time, but did have the benefit of allowing me to see how easily they were able to translate this basic summary. My intention had been that it should be simple enough to allow them to ‘read’ it without help from me, but to my disappointment they still needed some pointers and vocabulary in order to be able to get through it at a quick enough pace. Although the text itself was certainly very simple, I had slightly underestimated the difficulty with translation that Michael and Alice have. Nevertheless, with a little help from me they processed the text quickly, and we were able to discuss context and content without the difficulty of the translation getting in the way, as intended. Encouragingly, in the third lesson (Tacitus, Annals 1, 48), Michael remarked, after finishing Version 1, ‘Was that the whole chapter? … Oh, it's not too bad.’ This was the effect I had been aiming for, that the students would realise that underneath all the ‘foliage’ the basis of each chapter was actually quite simple. Furthermore, when I got their views afterwards, they implied that they had found the basic version pretty simple, indicating that the effect on the way they perceived the text had been the desired one, in that they had not felt intimidated by it, even though in practice I had still needed to help them with it.

I had chosen to tackle the middle and final versions of the text simultaneously so as to get the benefit of comparison for noticing literary effects. Therefore, I put each sentence of the middle version up on a PowerPoint, along with the relevant section of the basic version and vocabulary. The corresponding part of the original would be brought up alongside once the students were familiar with the middle version. My aim was that we would ‘translate’ the middle version together, and this would hopefully then allow the students to read the original without having to translate. This was working quite successfully by the end of the sequence, but it took quite a while for the students to grasp the idea that the meaning of the original was basically the same as the version they had just translated, and only the expression was different. It would perhaps have been helpful to explain this clearly at the outset.

I had feared that a negative consequence of tackling the final version one sentence at a time in this way might be that the general shape and flow of the whole might be lost. I therefore asked the students about this during the interview, but they responded that they had not found this to be a problem, as they had essentially been doing it once sentence at a time before anyway. Moreover, with the new method they also had the overview provided by Version 1, so they in fact felt that they had a better grasp of the content of the whole section. When asked in the interview about whether the Latin felt more approachable, Michael commented that he thought it was ‘[be]cause when you read the original you've… kind of … already read it before, so it doesn't, like, seem so hard… and you already know, like, what's happening.’ It appears, therefore, that a combination of scaffolding the content and the Latin did contribute towards the desired effect of making the original version more accessible.

In terms of literary analysis, my impression of how the lessons went was mixed. On one hand, the students did still need some guidance, but in general a broad question by me: ‘So what's different about this and the original and what's the effect?’ was enough to elicit interesting responses, with only occasional extra guidance of what sort of things to look out for being required. The thing which the difference between the versions highlighted most effectively was word placement, which the students were able to identify and discuss without prompting. Missing words, choice of words and variatio were all also successfully picked up on, although guidance was still required on the latter. It was encouraging that during the interview, Alice, without being asked specifically about the literary criticism, commented that ‘[It] helps you see what Tacitus is doing, you know, opposed to the simpler one, um, like it makes you see why he puts the words in different places and that kind of thing.’ This indicates that it being able to have a visual point of reference for things such as word order is helpful to students who may not be naturally attuned to how Latin word order can be manipulated, as had been evident in my discussion of Ovid with Alice (see above). Drawing arrows here there and everywhere had not been enough to really bring home the effect of having words ‘out of place’.

Another positive result which emerged from the interview was Michael's first observation that doing the text this way helped him remember it better, because instead of reading it once in a rather fragmented way we had read it three times. However, as with all of the answers in interview, it is difficult to ascertain whether this was just the students’ perspective, or whether it had had a measurable effect on how well the students could remember the text after a period of time. In the fifth lesson of the sequence, when we went back over each section to note some more general themes, the students remembered the content well, but I had no way of quantifying whether there was any appreciable difference compared to if they had prepared the text in the usual way. A potential method of comparison would have been for them to do another mock examination, which could then be compared to one which they had previously done on the other section of the text. However, I decided not to make them do a paper this close to their exams, as it should now be their choice what their revision would focus on.

It was disappointing that the students did not have any suggestions as to how the method could be made better – it is possible that they were afraid to offer any criticism, although as noted above I do not think Michael would have held back too much. When prompted, he did agree that it might have been better to prepare the first version beforehand, but did not think this had been too much of a problem. From my own perspective, I think I should have placed more focus on memorising the key vocabulary which was glossed in each version, as both students consistently relied on the glosses. If I had asked them to learn at least the vocabulary in the basic version beforehand I could have had fewer glosses and therefore replicated better the effect of reading the text. I also wish I had given them at least one or two sentences of the original to work out completely for themselves without prompting, both in language and style – it would have been interesting to see if they could have arrived at the same results without any guidance other than the middle version, but the pressure of time led me to prompt a little too readily, therefore negating some of the benefit of giving the scaffolding of the middle version. On the whole, however, I was pleased with how the lessons went, and how the students responded.

Conclusion

The results of my study, though limited in scope and reliability, were much along the same lines as the other research which has been undertaken on the subject. The use of embedded readings did appear to make the text being studied less intimidating, easier to translate, and easier to remember, in accordance with the benefits identified by Clarcq (2012a, p. 21) and Toda (2014a). Furthermore, Sears and Ballestrini's (2019) contention that the method should make the literary criticism less of a chore seemed to be backed up by my own observations in the lessons, as the students were able to pick up on stylistic points for themselves in the process of going through the text, rather than having to go back over it afterwards. The responses of the students were positive, and indicated that they had gained what I had hoped they would from the exercise. A comment made by Michael, however, best sums up the results of this study: ‘I know it's more work for you, Miss, but I think it was really useful.’ It was, indeed, a lot of work for me to prepare the texts, and this is perhaps the main disadvantage of using this method for set texts, which in the Latin syllabus change every two years. Further and more measurable research would be required to see whether the benefits outweigh the amount of work which has to go into preparing the texts. It would also be interesting to see whether there would be an equal or any positive reaction from students who do not struggle with translation to such an extent. With a high-achieving student, it may be that this method causes boredom, creates confusion where there was none before, or makes the experience of the text less immediate. However, I certainly feel that the method has potential for being used with a particularly complex text, or with a small part of a text to illustrate stylistic aspects.