Why this post?
This post aims at providing language educators interested in my approach with an up-to-date description of the MARSEARS instructional sequence faithful to the ‘letter’ and ‘spirit’ of EPI, as opposed to the many adaptations and hybridizations floating on social media these days, which may be as or even more effective, but often substantively misrepresent the original framework.
The MARSEARS sequence
As can be seen in figures 1 & 2 below, a typical unit of work lasts a term and consists of five sub-units: four of which deal with new material and consist of MARSEA (Modelling, Awareness-raising, Receptive processing, Structured Production, Expansion and Autonomy) and two of RS (Routinization and Spontaneity) which aim at consolidating and automatizing that material, as well as interweaving it with materials from previous units (e.g. in Term 2 in year 8, you will interweave core vocabulary and grammar from Term 1 in year 8, and if you can, from Terms 1,2 and 3 in year 7). The holes in between units can be used for revision and/or fluency training. Important: I recommend that the structured production phase (in green in figure 1) last 1.5 or 2 lessons with average-ability learners rather than one, as indicated in the picture.
Figure 1 : a macro-unit is subdivided in five sub-units. The first four sub-units include the MARSEA sequence, whilst the last sub-unit include the RS phases which culminate in the spontaneity assessment task. The ticks refer to the recycling of the target items (the material included in the sentence builders). Items 5 to 8 refer to the new items, whereas items 1 to 4 refer to the ones taught in the previous unit.
Figure 2 : the MARS EARS CYCLE at KS3, adapted to contexts where one teaches 20 to 24 lessons per term
1 – Modelling – the target sentence pattern(s) is (are) modelled through sentence builders or any other means. I prefer sentence builders, occasionally preceded by a round of two of flashcards to sensitize the students to any new vocabulary included in the target sentence builder. Activating prior lexical or morpho-syntactic knowledge before the modelling of the target construction will obviously help, should the target sentence builder include previously encountered material (for instance: if you are teaching the perfect tense of verbs requiring the auxiliary ‘Avoir’ in French, you may want to activate their knowledge of the present indicative of AVOIR and the meaning of the verbs you have included in the sentence builders).
Figure 3 – A sample German sentence builder from http://www.sentencebuilders.com
2 – Awareness-raising – the learners’ attention is drawn to specific features in the model sentences, e.g. SSC (symbol-to-sound correspondence), phonotactics (e.g. liaison), intonation patterns, grammar and syntax. This is snappy pop-up grammar and pop-up phonics awareness-raising aimed at sensitizing the learners to specific linguistic features in the input, so that when they encounter them multiple times in the next (receptive) phase they are more likely to pay attention to and eventually internalize them.
3 – Receptive processing – in the first part of this phase, the focus is on sentence-level processing only, at least with beginner to pre-intermediate students. The teacher engages the students in the repeated processing of the model sentences through a variety of enjoyable snappy activities mostly through the aural medium. The input is flooded with multiple occurrences of the target pattern and the instructor is deliberate in targeting the whole range of listening and reading sub-skills: phonological and graphological processing (both at phonemic and syllabic level); segmenting; lexical retrieval, parsing (grammar and syntax), and semantic (meaning) processing. The activities include a balance of focus-on-form (phonics, grammar, syntax and function words) and focus-on-meaning activities. Scripted listening (i.e. listening whilst reading) activities enable the students to develop decoding (SSC) and segmenting skills, key for vocabulary acquisition through the aural medium.
In the second part of this phase, the teacher proceeds to work on connected texts, which include current and previously studied material. The texts usually include narrow listening and narrow reading texts, i.e. near-identical texts where the discourse structure is identical but some of the lexical items differ. The highly-patterned texts are flooded with the target features and contain at least 90 % comprehensible input. Input-enhancement (both acoustic and visual) draws attention to the target features.
The input-flooding, input-enhancement, repeated processing and thorough-processing elicited by the tasks included in the instructional sequences throughout this phase, allow for multiple encounters with the target input, which is likely to result in implicit learning. Throughout the phase the teacher will EXPLICITLY and repeatedly point the students’ attention to the target phonological, grammatical and syntactic features thereby ensuring that IMPLICIT and EXPLICIT learning work in synergy. Deep-processing activities can also be embedded in this phase to elicit the greater cognitive investment that according to much research (e.g. Leow, 2015) results in longer-lasting learning.
Important: the listening and reading activities staged in this phase are not your typical receptive activities. They are Listening-As-Modelling (LAM) and Reading-As-Modelling (RAM) activities which are designed to model speaking and writing as part of your effort to convert input into output. This doesn’t mean you won’t stage any traditional Listening and Reading comprehensions. You will do both. For obvious reasons, though, with beginner-to-intermediate learners you will stage listening-for-learning (LAM) tasks prior to engaging in listening-for-testing ones.
Fig. 4. MAR
4 – Structured production – in the first part of this pushed-output phase, the beginner-to-pre-intermediate learners engage in chunking-aloud games/tasks (e.g. Sentence stealer, Sentence chaos, Mind-reading, Lie-detector) which elicit repeated processing of the target sentence patterns. These games, too, will involve an alternation of focus-on-meaning and focus-on-form activities. These games aim at enhancing decoding skills, phonotactic and articulatory fluency. More importantly, chunking-aloud games, being fun and inclusive, foster learner willingness to communicate and participate, two key pre-requisites for the attainment of oral fluency.
In the second part of this phase, the learners engage in retrieval practice, in highly structured oral (e.g. role plays) and written communicative activities and less structured information-gap, reasoning-gap and opinion gap communicative tasks . Initial individual or pair-work written activities – carried out using worksheets or digital resources – help transitioning from receptive to productive retrieval and pave the way for a series of engaging and enjoyable oral retrieval practice activities (some of which, not all, are described here). The latter involve group work and are totally student-managed, typically consisting of competitions between two players with a referee/coach (equipped with an answer sheet) providing instant positive or negative feedback and awarding points. The fact that these activities are totally student-managed means that the teacher can devote his/her attention to monitoring learner performance whilst moving around the classroom. This enables him/her to gather a lot of observational data and engage in the provision of formative feedback.
Fig. 5 – From Modelling to Structure Production
5 – Expansion – this phase is about learning explicitly the morpho-syntactic patterns that have already been processed many times over. The teacher may decide to do this deductively (explicitly teaching the grammar); through guided discovery (leading the students with guiding questions to working out the grammar rules) or totally inductively (the learners, given a number of sentences exemplifying the target rules infer the rules by themselves). Receptive and productive activities from the previous phases (in which the target rule is task-essential) can be employed here.
6 – Autonomous recall – this is where short achievement tests are staged. These are snappy, easy-to mark, low-stake assessments aimed at ascertaining whether the students have attained at least receptive mastery of the target input (or productive if you are dealing with groups of higher attainers). As far as I am concerned, I prefer to make the first three assessments (whether form- or meaning-focused) receptive in nature and the last two productive. The rationale: the ability to perform productive retrieval emerges later than receptive. Also, throughout the structured productive phase in each sub-unit I have usually already obtained a good idea of where my students are in terms of productive retrieval in non-exam conditions. A grammar assessment component, initially receptive and later productive, can be included with the right groups at the end of each sub-unit.
Fig 6: Expansion and Autonomy
7-Routinization/Fluency training – in this phase, a number of truly engaging and enjoyable pushed-output activities are staged which aim at consolidating the target material and speeding up its receptive and productive retrieval. The activities, based on Paul Nation’s research into fluency training, typically involve: (1) repeated processing, (2) task-repetition; (3) pre-task priming; (4) pre-task planning; (5) incrementally challenging timed constraints. As happens in the previous phases, one will alternate tasks which focus on negotiation of meaning (‘Messengers’, ‘Dictogloss’, ‘Five’, Detective and Informants’, ‘All for one and one for all’, ‘Secret sentences’, ‘View and Recall race’) with others which focus on form (e.g. ‘Chain reading’; ‘Chain dictation’, ‘Tongue twisters’, ‘Fast and Furious’, ‘Puzzle race’). Very important: since automatization means making what is already known more easily and rapidly retrievable, the language used in this phase should include only familiar lexis and grammar.
8 – Spontaneity – In this phase communicative tasks (e.g. Oral picture-description tasks; Interviews; Role-plays) require the students to produce UNPLANNED output under time constraints to simulate R.O.C. (Real Operating Conditions). One of these tasks can be used too for assessment purposes. With lower-ability learners, the assessment could include some planning time and even a short priming task.
Fig :7 Fluency and Spontaneity
Wow! This is long !
An understandable reaction is to consider this too long and time-consuming. The answer: fluency can’t be achieved using the traditional PPP approach or short instructional sequences of a couple of lessons, as textbooks purport to do. Students need to receive substantive structured receptive and productive practice to be able to attain that degree of proceduralization of the input which will ensure long-term retention. Much textbook-based teaching fails in this regard and, consequently, more than often teachers at KS4 (15-16 years old) have to reteach what was taught at KS3 (years 11 to 13) all over again wasting valuable time (now THAT is time-consuming!). Also, going ‘slower’ means being more inclusive thereby increasing the chances of a higher future GCSE and A-level uptake. Fortunately, at KS3 one can afford to go slower, as the focus should be on fostering fluency development and a passion for language learning, rather than manufacturing cohorts of exam takers.
Note that at KS4, the EPI teacher will be able to go faster, having built strong foundations at KS2 and 3, as opposed to flimsy declarative knowledge, as often happens.
In this post I have provided an outline of the MARSEARS sequence faithful to the true letter and spirit of EPI. Teachers are, of course, very free to adapt it to their own contexts always hopefully cognizant of the fact that they can’t hope to create fluency at KS3, unless they teach exceptionally gifted learners, in three or four lessons or by stopping at the structured production phase as some advocate.
The aim of this sequence is to be inclusive and create durable learning. Each phase in the sequence primes the next one. So, whilst the challenge increases gradually, the students get to the next phase prepared and more confident. Building learner can-do attitude is key in the design and delivery of an EPI instructional sequence, as self-efficacy is one of the strongest, if not the strongest, predictors of language learning success.
Obviously, each MARSEARS sequence must not be considered in isolation but in relation to the curriculum as a whole, always ensuring that the core items you have identified as the non-negotiables (or Universals, as I call them), the must-learn phonics, vocabulary and grammar, are constantly recycled across contexts, consistently with the Transfer Appropriate Processing principle and with what we know about memory decay and proactive/retroactive interference (Bjork’s law of disuse).
Finally, the key to successful teaching is ensuring that the students – all of them – enjoy learning and succeed at it on a daily basis. Unless this happens, students are unlikely to develop a passion for language learning. Traditional PPP, especially when consisting of masses of grammar learning, however well-planned and systematic, is unlikely to be inclusive and exciting for the average teenager.
To find out more about the approach, do get hold of the best-selling book authored by myself and Steve Smith, “Breaking the Sound Barrier: teaching learners how to listen”. or attend my upcoming workshops
One thought on “The real MARSEARS: how an EPI sequence TRULY unfolds (the real ‘Conti method’)”
Reblogged this on The Language Gym.