Semantic waves – repack!

One of the “flagship” aspects of EAL pedagogy is Jim Cummins’s BICS/CALP theory, which posits that whilst BICS (Basic Interpersonal Communication Skills) is an earlier stage of second language acquisition characterised by a learner being able to speak the language for socialising (for instance, in the playground, at lunchtime and in the classroom), CALP (Cognitive Academic Language Proficiency), which is the language they will need to carry out academic tasks in the classroom. BICS usually takes up 1-3 years to develop, whilst for CALP it is 5 to 7 years.

Cummins has also provided a diagram where he categorised the different tasks expected of students in a classroom according whether they required BICS or CALP type of language. See below:



As one can see above, any task in a lesson can be classified as cognitively demanding or undemanding and with context embedded or not embedded. It also is almost self-explanatory that cognitively undemanding tasks are on the BICS side of the spectrum and those more cognitively demanding are CALP. Whilst D is to be avoided as such activities result in no learning on the part of an EAL learner, A is where a more heavily supported EAL (new to English) learner would start with tasks given to them becoming gradually more abstract with less context embedded and thus more academic.

However, how exactly this journey would look like (from heavily contextualised tasks to more abstract and academic ones) is a rather difficult task and less discussed. This is precisely where the notion of Semantic Waves is so helpful!


The following discussion reports largely on two separate articles on Legitimation Code Theory and Semantic Waves. The first is Maton, K.’s 2014 A TALL Order? Legitimation Code Theory for academic language and learning and the other Macnaught et al.’s (2013) Jointly constructing semantic waves: Implications for teacher trainingAll other comments are my own.

The notion of semantic waves is part of the Legitimation Code Theory (LCT). This theory challenges the very notion of ‘knowledge’. Karl Maton, who is behind LCT says that whilst knowledge is obviously central to modern societies, there are few theories as to what knowledge actually is. Maton holds the view that many treat knowledge as homogeneous and neutral, and ignore, for instance that there are “relations within knowledge” (p.172 in Van Krieken, R. et al. (2014) ): for instance, knowledge can vary internally based on whether it’s concrete or abstract, whether it’s context-dependent or context-independent or whether it’s based on personal experience or whether it’s specialist. Maton calls this state of affairs ‘knowledge-blindness’ (2013: see his book Knowledge and Knowers: Towards a realist sociology of education) . LCT therefore is a theoretical framework for analysing underlying knowledge practices. This framework/toolkit is composed of five chief dimensions: Autonomy, Density, Semantics, Specialisation and Temporality. For the purposes of the present blog post, we are interested in the Semantics dimension and its Semantic Waves.

The LCT’s dimension of semantics distinguishes between its aspects:

  • semantic gravity (SG) – which is about how much meaning is related to its context. SG+ (strong semantic gravity) will mean that meaning is more dependent on context than SG- (weaker semantic gravity). Thus, for instance, naming a specific animal would be SG+ whilst talking about an event from history would be, by comparison, SG-. However, the same historical event could be SG+ than talking or writing about the process of photosynthesis (SG-, then)
  • semantic density (SD) – this concept relates to how highly condensed a meaning is. Much like above, SD+ is a stronger semantic density whilst SD- is weaker.Maton (2014’s A TALL Order? Legitimation Code Theory for academic language and learning) provides an example of what such density actually means. The word ‘gold’ is normally understood as a bright yellow metal. However, in the field of Chemistry, this could additionally mean an atomic number, atomic weight and more. In other words, depending on the context in which a word is used, it can take on additional meanings.

These two aspects of the LCT can therefore be used to create a semantic profile for a piece of writing or any discourse within a classroom. Please see the graph below, which I am adapting from Maton’s 2014 journal article (the same as referenced above):

semantic profile


As we can see here, such semantic profiles allow to analyse language used in a discourse over time. The text A would have been one with weak semantic gravity (abstract, little dependence on context) and with strong semantic density (thus highly academic language). Line B, on the other hand is a text highly dependent on context (strong semantic gravity) and with words exhibiting simple meanings (low semantic density). A and B represent what Maton calls a semantic flatline. However, line C is a semantic wave, and one can observe how the language is less academic and more context based at the beginning (stronger semantic gravity and weaker semantic density), but then changes to more academic (weaker semantic gravity and stronger semantic density) to then fall back to the type of less academic language used at the beginning.

We also need to remember that the example of the semantic profile (the bell curve) with just one wave is just that, an example, and there will be several waves, typically, in a student’s writing or, indeed, our own (teacher’s) discourse. The wave can also start at a completely different place on the vertical axis – for instance, it might very well be that in a practical subject such as P.E. or Food Technology a teacher’s or student’s discourse would start with SG+/SD- and then move towards SG-1/SD+. Definitely read Maton’s article (linked to above) on this as it has more than this blog post can possibly cover. As such, semantic waves are an excellent way to illustrate that different subjects come with different academic literacies and that their language varies – which is what I have recently written about in my article for Innovate My School titled Academic literacies across the curriculum.

Unfortunately, what happens in classrooms all too often is what Maton terms a ‘down escalator’ semantic profile (see just below), meaning that more complex language and meanings are expressed by teachers to their students in more concrete and simpler meanings.



Now, explaining more complex academic language in source texts through the use of more concrete terminology and simpler language is not a problem in itself – teachers do need to make texts more accessible to all learners, EAL and not EAL alike. The trouble lies in the fact that if this is where the process stops, students are not learning how to use this new understanding again, this time independently, for academic writing (or other forms of discourse), which is precisely what is required of them at the GCSE examinations stage in KS4. Put very simply, if we only simplify the academic language, but do not teach them how to go back to the academic once they understand it, we do our students a disservice by now equipping them to write academically. They might have the understanding of the topic or content, but cannot express it on the level that is required for them if they are to be awarded higher grades. This is of course for subjects across the curriculum and certainly not limited to English.

You would have noticed the image of a backpack at the beginning of this article. This is because what students need to do – and we need to help them with that – is to first unpack a concept and once that concept is cognitively internalised, they need to be able to repack it using academic language.

Let’s go back to Cummins’s theory now and EAL learners. Teachers’ ability to use semantic waves effectively is crucial for for EAL learners if they are ever to move beyond the Quadrant A and B towards the Quadrant C, where the context is reduced and language is more abstract (=academic). In other words, EAL learners need the access to weaker semantic gravity and stronger semantic density. If we continue to only simplify content for our EAL learners and not provide them with density (= wider range of academic language), we are putting them in a vicious circle where they’re stuck with very basic language unable to actually express themselves. For all we know, of course, they might be able to do that very well in their first language.




A natural emergent question is, of course, how to take the semantic waves and use this awareness in our teaching. McNaught et al. (2012) provide some rather very specific examples of this can be done. In their article Jointly constructing semantic waves: Implications for teacher trainingthey name the Teaching and Learning Cycle (TLC), originally developed in Sydney, Australia, as a possible approach. This framework consists of three main steps as shown below:


Deconstruction involves contextualising content and analysing model texts. This is a great time for checking students’ prior knowledge. Here, the structure of a text can be analysed. In addition, the way individual language items (power words and power grammar) build up the entire body of the text (power composition) can be discussed. Both modelling and using metalanguage do mirror, of course, what NALDIC in the UK see as components of the distinct EAL pedagogy (see their Working Paper 5 for more). Additionally, in the UK, our own United Kingdom Literacy Association (UKLA) certainly shares the view of the the usefulness of a shared metalanguage in their booklet on grammar:

“Why is it useful? Because to be able to speak about language, to have a technical language with which to discuss and describe language, a metalanguage, is as useful as it is to have a technical language with which to speak about any other area of human endeavour. To have technical vocabulary at one’s command brings a sense of mastery and clarity to the person so equipped.” (p.8, Richmond, J., 2015, English Language and Literacy 3 to 19: Principles and Proposals: Grammar and Knowledge About Language)

Joint Construction is about teacher-led collaborative writing. Collaborative, of course, immediately makes me think of collaborative learning approach, which is very much a part of EAL pedagogy (see Collaborative Learning Project and British Council EAL Nexus page on Collaborative activities). Here, students are to use the knowledge they’d gained in the deconstruction stage and construct together another text. Of course, by doing so they are already changing the semantic density and gravity towards the less concrete and more academic type of writing.

The first two steps lead the students to the Independent construction. This model is somewhat similar to the Gradual Release of Responsibility Model (see, for instance, Effective Use of the Gradual Release of Responsibility Model by D. Fisher)



Further, Macnaught et al. divide the joint construction stage into two sub-stages:

(1) Bridging – which is about revising shared knowledge (or allowing students to access their prior learning). This is a preparation stage before the text is analysed. The scholars use an example from a biology lesson on the immune response, and show a diagram, which compares the humoral immune response to cell-mediated response. I am adapting it below, and you can see how this looks strikingly much like (in fact, that’s what it is!) a graphic organizer – which is something that I, as an EAL Coordinator and Teacher, go by in pretty much every lesson. Can’t live without it! Please note how even at this early Bridging stage the academic language is already taught (locate “activate” and “gives rise to”, for instance in the key visual/graphic organizer below.

The image below can be found at the following address, and originally comes from a Pearson Publishing book. It’s the same as the one seen in the journal article by Macnaught et al.


As we can see, not only is the language made apparent here, which allows for the teaching of grammar, too, if required (e.g. Helper T cell activates active and memory helper T cells). Of course, modelled here can be the fact that we say “cell activates” but “cells activate”. In addition, the chain of cause and effect are clearly visible here and the arrows are a very important part of the graphic organiser, which allows to see the organisation of different elements to learnt and how they relate to each other – in other words, content and language are already taken care of at this stage.

Of course, this way of approaching a lesson bears very strong resemblance to Bernard Mohan’s Knowledge Framework. Please see this slide from one of my own teacher training sessions last year where you can see how the very simple text on the left (of the EAL new arrivals sort, clearly) is constructed similarly to a piece of text from Maths: graphic organisers are instrumental in organising the information (types of triangles) and it is only the academic language that needs to be taught – such as, in the text on the right as made up of. All that would be required is providing students with the language required on the graphic organiser. For that, see the image below, where I am taking the same graphic organizer categorisation tree and differentiating it to include the language.




Now, here there are now arrows as there is no specific order in which students might talk or write about it, i.e. it doesn’t matter whether they first talk about scalene or isosceles triangles or whether they talk first about sides or angles of isosceles triangles. But the language to talk in an academic way about it made available to them. More than one option can be made available to them (e.g. are made up of = consist of, are composed of…), but a decision would need to be made whether in this bridging stage this wouldn’t take too long. Providing students with more options could be left for later.

(2) Text Negotiation – this is where the writing will start. This is where the teacher facilitates and mediates suggestions from students as the text is gradually written. In essence, the teacher leads the construction of the text and made the differences between how strong semantic gravity and density are explicit.

Macnaught et al. present a semantic wave on the basis of biology teachers’ discussion of how they would present to students an explanation of the connection between antigens B and T cells. They start in the middle of the semantic scale first as they identify that “antigens are molecules or proteins”. But when they next talk about immune response” (using abstract and technical term), the density is stronger and gravity is weaker and so the semantic wave goes up. A little later, however, when they decide to explain to students that “activates means to switch it on”, the wave goes almost all the way down – as “switching on” is as context-dependent as it gets: relating the concept to the idea of switching a light or switching a device on, which students would easily recognize from their everyday life. But then, when they say “this involves the activation of…” the wave goes sharply up. For a full analysis and better idea of what the entire analysis looks like – and the entire wave – please read page 57 in their article.

The teachers used a number of phrases to precisely “unpack” the meanings for students. For instance, they used a number of words related to ‘war’ (e.g. like your infanty, barrier, artillery, a sniper, lines of defence) to illustrate to students the relationship between the body and the pathogens. Many of us will recognise this very strategy we use to unpack meanings for students in this way. What we need to do, however, is ‘repack’ it. That’s why the joint construction stage – showing students how to move from the simpler to more academic (more dense but with weaker gravity) language is required.

Thus, the writers suggest, in our classrooms, we need to:

  • collaboratively enable students to move down the semantic scale first – by using strategies such as analogies and gesture (using context)
  • subsequently use power grammar to move back up the semantic scale so they are able to write academically
  • acknowledge that individual linguistic items, if understood and made explicit to students, will enable them to create power composition



Power words are the words that are absolutely essential for the understanding of the topic. They might also be the kind of words – usually content-based – that are specifically required and essential for the purposes of writing exam papers. More simply, they are key words that mainstream teachers very often put on the board or next to the board at the beginning of their lessons. Macnaught’s observed in her study that whilst Austrialian teachers certainly did generate such lists, often through brainstorming them with their learners,they would create lists. The trouble with such lists or sometimes mindmaps, Macnaught et al. write, is that they are lists! Rather, they should be reorganised into chains of cause-effect relationships. Neither mind-maps nor lists will allow to show such relationships.

Ensuring that relational representations (Macnaught et al.’s term), such as graphic organisers with arrows and links between the power words / power phrases are made explicit means that the grammar that needs to be used to explain these relationships can naturally be taught. The scholars write that showing the links between terms in a larger group of terms (power words) allows for their meanings to be more condensed. Simply put (or taking the semantic wave lower down here!), if we teach how various key words (power words) are connected with each other, we will lock our students into having to use more academic language simply because they will need to be talking about the links between these particular elements relating to each other. And that’s how, in English, you will teach your students to avoid writing/saying ‘is nice’ and will get ‘acts in an enchanting way’, and in Geography instead of ‘travels’ you will get ‘design an itinerary’ instead.



Power grammar, therefore, is linked to power words. One of the aspects of academic power grammar that Macnaught et al. mention is nominalisation. Do, please, imagine me nodding my head very vigorously as I was first reading this. This is certainly recognised within the EAL pedagogy! Indeed, in the UK, Graham Smith, the director of EAL Academy (I myself am one of their associates) has spoken on numerous occasions about the positive effects of teaching nominalisation to EAL learners – which has been found to increase their attainment.

For those of you who might struggle with the concept of nominalisation: it is the process of changing another part of speech into a noun, for instance, create becomes creation or substitute becomes substitution. It would also entail the understanding that the verb start could be nominalised into formation or creation. Therefore, a sentence:


When we started/created this group, a really special bond between us resulted.

could be changed into

The creation of this group resulted in a really special bond between us.


Nominalisation features very heavily in academic writing across subject domains. At an EAL TeachMeet in the summer of 2014 in Morden, Graham Smith suggested that teachers should be trained in teaching nominalisation and concluded by saying that there has been a considerable impact of teaching it in Islington where this was introduced under his guidance. The result: considerably higher A*-C results in English and Maths when compared to the rest of England.

Macnaught et al. suggest that the students could identify groups of verbs in their original texts and rewrite these sentence so that nominalisation is used. This is, of course, a directed activity related to text (DART) – focusing on language itself whilst studying subject content – and I have promoted the DARTs approach for over a year now with a number of teachers. DARTs are another EAL-pedagogy approach and type of activity. Yet again we see how EAL and semantic waves (and LCT) overlap.

For inspiration, have a look at the activity here:

Macnaught et al. suggest other approaches to teaching power grammar to students. Linking words are suggested to be made available to students for the Joint Construction stage. However, this does not mean your bog-standard connectives – not necessarily, anyway. Rather, they are also verbs linking concepts: such as allowed, enabled, lead to. In the study conducted, this actually led to one of the students asking if they could “thesaurus” the word allowed as they were getting tired of repeating the same word and wanted to find an alternative (synonym) for it! The scholars suggest that students are not just given one pre-defined option to follow, but are given a number of alternatives to choose from! I wholeheartedly agree! This is reminiscent of our EAL substitution tables where you can give your students a number of options to build up their academic vocabulary. See below for an example:


Your learners – including EAL learners – need only to know one of the options above to understand what to do. Trust me, this has been tested on numerous occasions with new arrival EAL learners. It works.



This leads us to power composition. The previous stages (Bridging, Text Negotiation) would normally be followed by a Review of a text. Teaching power words and power grammar, initially individually, allows students to see, eventually, how it is impossible to show the complexities and linkages between different elements (power words) without using the grammar (language). Thus, composition relies on their ability to understand and be aware of the language they use (such as nominalisation or linking words). This will allow them to string their individual items into a coherent, linked up whole. Without the awareness of the language, it will simply not happen.



I was very excited to be guided towards the research into Legitimation Code Theory and in particular its Semantic Waves part. The repacking aspect of our lessons is certainly something that often is missing from our teaching – it’s quite tempting to fall into the trap of unpacking/simplifying the language or content for EAL learners (and other learners, too, for that matter) without repacking it again! Maton’s and Macnaught et al.’s papers provide fantastic framework and practical ideas to get you started on the analysis of the kind of teaching and the kind of language that is used in your classroom – with EAL learners and all your learners who have the right to be taught literacy across the curriculum. As all our subjects come with different language requirements, it is important to analyse what happens in our individual classrooms or, at the very least, in our subject departments.

Let’s remember that Semantic Waves and Semantic Profiles can be used to analyse the language that your higher achieving and lower achieving students use in their writing as much as to analyse your own talk. Any text or discourse within the process of teaching can be analysed and thus our own practice involved.

I love how much of the semantic profiling approach overlaps with EAL pedagogy and its elements such as graphic organisers, substitution tables, DARTs. It also addresses the need (as identified by Cummins long time ago) to teach EAL learners less context-dependent and more abstract academic language if they are catch up with their native English-speaking peers. To me, it means we are definitely right when we say that what works for EAL learners works for everyone in our classrooms: language is an essential and integral element of any learning and creation of new knowledge in any subject and any academic domain we teach.

Let me just leave you with this: if you leave out or ignore language teaching in your classes, regardless of which subject you teach, your students are not likely to be able compose high-level academic texts, which is the only way they will be able to show you what they know in an organised, understandable fashion.


Print Friendly, PDF & Email

Leave a Reply

Your email address will not be published. Required fields are marked *

Please confirm you are human: *