Why “Stop Translating in Your Head” Is Probably Bad Advice for Most Adult Language Learners

“Try not to think in your native language.”

Almost every adult language learner has heard this advice at some point. It appears constantly in fluency videos, productivity-style language hacks, and motivational reels. The implication is usually the same: truly successful learners somehow bypass their first language and begin operating directly in the target one.

It sounds convincing — until you’ve spent enough time in actual classrooms.

A student of mine recently mentioned several such videos during class. He is intelligent, highly analytical, and also something of a perfectionist — the type of learner who becomes visibly uncomfortable after producing even a minor grammatical error. The videos had left him with the impression that mentally translating was evidence of failure.

So I asked him a question.

“When you say something simple like ‘I want more time,’ do you really think your brain should ignore the fact that Ukrainian already contains a perfectly usable conceptual equivalent?”

Why would it?

From a cognitive standpoint, that would be extraordinarily inefficient.

In fact, decades of bilingualism research suggest that bilingual speakers do not simply “switch off” one language while using another. Researchers such as Judith Kroll and Annette de Groot have written extensively about co-activation in bilingual processing: both linguistic systems remain active to some extent, and the brain continuously manages competition between them.

This becomes obvious in real classrooms.

Imagine a Ukrainian-speaking learner producing the sentence:

“I live here for two years.”

An English teacher with a Slavic background immediately recognizes the problem: the learner has transferred the structure directly from Ukrainian or Russian, where the present tense naturally covers situations extending from the past into the present. English conceptualizes this differently and prefers the present perfect: “I have lived here for two years.”

This is a genuine transfer problem.

But now consider another sentence:

“I want more coffee.”
“She opened the window.”
“We arrived late because of traffic.”

These map remarkably well across English and Ukrainian. The conceptual structures are highly compatible. Translation here is not an obstacle: it is cognitive scaffolding.

Why force learners to pretend otherwise?

Adult learners already possess an enormous and sophisticated linguistic system. The brain will inevitably use existing structures to interpret and produce new language. Trying to prohibit this entirely may – and will – actually increase cognitive load by forcing learners to suppress mechanisms that normally support processing.

Psycholinguistic research broadly supports this idea. Studies by Kroll, Dijkstra, and van Heuven, among others, suggest that lexical items from both languages remain partially activated even during monolingual tasks. The bilingual brain appears to regulate activation rather than eliminate it.

In other words, bilingual cognition seems to function more like traffic control than like an on/off switch.

Let’s face it: this aligns with what many experienced teachers intuitively observe.

The students who progress fastest are often not those who avoid translation completely, but those who gradually become better at distinguishing between transferable and non-transferable patterns.

Take the so-called “complex object” in English (a term that seems to be used exclusively by Eastern European teachers, primarily post-Soviet ones):

“I want him to call me.”
“She saw him cross the street.”

For many Slavic-language speakers, these constructions feel deeply unnatural at first because their native languages often express the same meaning through subordinate clauses instead: “I want that he calls me.” This is precisely where conscious restructuring becomes necessary. Learners need repeated exposure, correction, and practice because direct transfer fails.

The same happens with prepositional phrases. A student says:

“depend from”
“married on”
“discuss about”

These are not random mistakes. The learner is importing relational patterns from another linguistic system. Prepositions are notoriously language-specific because they encode conceptual relationships differently across languages.

Isn’t that the reason we need to have them stop translating???

No. Good teaching is not about declaring war on the first language. It is about helping learners recognise where transfer accelerates production and where it creates interference.

Teach what is hard: why teach what is easy (direct transfer) the hard way?

Another idea frequently missing from online fluency advice is automatisation.

People often talk about language as though fluency were primarily a matter of courage or mindset. Yet cognitive psychology has long distinguished between declarative and procedural knowledge. John Anderson’s ACT-R framework, for example, describes how repeated practice gradually transforms consciously accessible knowledge into faster, more automatic performance.

Most teachers see this constantly.

A student understands the rule for the present perfect perfectly during exercises. They can explain it. They can even identify mistakes in written sentences. Then conversation begins, processing speed increases, working memory becomes overloaded, and the rule collapses under communicative pressure.

Why?

Because the structure has not yet become proceduralized.

This is why repetition matters so much — although perhaps “repetition” is an unfashionable word nowadays. It is labeled ‘like in school, ew!’ by info-gypsies, and shunned. But…

Ask musicians whether scales matter.
Ask athletes whether drills matter.
Ask actors whether rehearsal matters.

Language is also a performance system involving motor planning, retrieval speed, prediction, and automation.

A learner who has repeated “I should have called earlier” fifty meaningful times will usually access it faster than someone who merely understands the grammar explanation intellectually.

And this is where most widely used, best-selling PPP coursebooks, claiming to be focused primarily on getting students to speak, in fact, fail them. They simply do not provide nearly enough meaningful (!) practice for the newly-discovered grammar to proceduralize. A few gap fills, a handful of multiple choices, and a small bunch of matches simply cannot achieve automation. This is exactly where a teacher tired of remedial work at every lesson rummages through their dusty treasure chests and fishes out the good old Side By Side and Streamline: old-fashioned-looking, they still provide PLENTY of repetitive pattern practice, their methodological premise being the Audio-Lingual approach, which is pretty straightforward:

Hear it => Understand it => Repeat it => Repeat it => Repeat it => Repeat it (X100) => Use it.

Repeat it. I mean, repeat, not only recycle. Parrot-fashion sometimes. But meaningful (the famous SBS pictures, oh yeah). This works, yes. But not only this.

Curiously, one of the most effective forms of practice is also one of the least discussed: inner speech.

Many learners already do this instinctively. They narrate their commute internally. They rehearse imaginary conversations in the shower. They silently argue with people who are not present.

And no, this is not “fake speaking.”

Researchers such as Lev Vygotsky — and later scholars studying subvocalisation and inner speech — argued that internal language plays a central role in cognition and self-regulation. Modern neuroimaging studies also suggest that silent verbal rehearsal activates several regions associated with overt speech production, including motor-planning areas.

The research here is still developing, and we should be careful not to exaggerate the claims. Still, based both on existing findings and on years of classroom observation, inner speech appears far more valuable than many learners realize. It has been shown that football spectators’ muscles contract in a way similar to the actual football players. One can (and should) practice the language inside their head. It is gratuitous, shame-free, and can be done everywhere. But most importantly… it works.

The same is true of massive input.

Stephen Krashen’s theories became controversial in some respects, but one core insight remains difficult to deny: people who read extensively and listen extensively usually develop stronger intuitive command of language.

After enough exposure, learners begin to anticipate collocations, sentence rhythm, and grammatical probabilities before consciously analysing them.

Oddly enough, the closest modern analogy may be large language models. AI, in household terms.

Human beings obviously do not learn language the same way neural networks do. The comparison should not be pushed too far. Yet there is a superficial similarity worth noticing: repeated exposure to enormous quantities of meaningful language gradually trains predictive intuitions.

Fluency starts feeling less like “constructing sentences” and more like recognising patterns quickly enough to use them in real time.

Which brings us back to the original advice:
“Stop translating.”

Perhaps the more useful goal is not to abolish the native language (we simply can’t), but to build such strong automatised patterns in the second language that translation gradually becomes less necessary on its own.

That is a very different process.

And, in my experience at least, a much more realistic one.

Why am I skeptical about PPP

PPP Revisited (2025 Reflection)

Back in 2011, I wrote the article below after confessing to a colleague that I was skeptical about the Presentation–Practice–Production (PPP) model. She was surprised — CELTA had drilled it into her as the (one and only) way to teach.

Reading my own words now, 14 years later, I see that much of my frustration with PPP is still alive in today’s TEFL debates. Teachers still wrestle with coursebooks that feel too narrow, with students who can’t “produce on demand,” and with methods that sometimes ignore the messy reality of acquisition.

So I’m republishing this post as it first appeared — raw, unpolished, and very much “me” from that time. I’ve added a short 2025 note at the end to say where I stand today.

This article was written when I mentioned my skepticism regarding PPP to a colleague of mine, and she was puzzled as CELTA had taught her to use PPP. It took, however, more than 3 months to write… Not that I was doing any research – this one is in no way a well-grounded scientific paper. It is, rather, some summary, more or less structured, of my thoughts about the subject. It’s for my school’s teachers, and anyone interested.

1.       PRESENTATION

a). What do we present? How do we decide what language exactly is needed by the student? One option is to be based on the course book. Course books have been designed by clever people, you know. Respectable institutions are working on the syllabuses, why bother and re-invent the bicycle?

Without trying to underestimate the immense work done by linguists on developing standards and designing coursebooks, I have to admit that I personally have to amend at least 40 to 70(!) % of ANY coursebook I have used (and I have used a lot of coursebooks).

• Too little attention is still given to many widely used patterns such as prepositional phrases, the “for-to” pattern, the complex object, cleft sentences, emphatic constructions – you name it. This leaves the students handicapped until very late, upper-intermediate or even advanced stages, and if you keep noticing that, despite their hard work for a year or so, your pre-intermediate students are struggling and become demotivated, think – isn’t it due to you following the standard syllabus, which leaves so many useful, even indispensable! –  things – until later?
• The same is true about vocabulary, prepositions, set phrases with articles, etc, etc.
• No mother tongue specifics are taken into consideration. Yet, what is easy for a French person may be beyond comprehension for a Russian, and vice versa.

>One way I try to deal with it is to introduce an array of useful patterns, normally taught at later stages, rather early, the result being SS slowly but surely acquiring more tools for expressing themselves, and the effectiveness improving.

> I try to make room for these additional patterns by omitting whole sections of a coursebook, such as too much work on narrative tenses, or reported speech, which can easily be simplified or done without (as we see real English used in real life). Some topics, some boring texts – lots can be, indeed, omitted harmlessly.

> I am still constantly faced with the ever-growing “syllabus tree”, with the additional patterns I need, since every topic will call for more and more essential constructions. That eventually becomes difficult to handle, and better coursebooks are, perhaps, needed to help us cope.

b). When we base our presentation on recorded or printed texts, coursebooks force us to focus on the pattern in question, plus, maybe, a handful of topical vocabulary. Look at the text, however. There’s, in fact, a wealth of useful and necessary stuff. The PPP lesson framework will simply allow no room for working with those. That will leave so many questions unanswered, so many hypotheses unverified, and so many patterns unacquired. Is that efficient? I doubt it.

c). Some more recent coursebooks are trying to cope with these deficiencies by highlighting additional patterns. Yet, even all of those taken together, that is not nearly enough to really move from level to level. The extra tidbits will be highlighted – but what about acquiring them?  Imagine we had to work with every single pattern we need using the PPP model. That will add at least several dozen lessons to the course book. Is that possible at all? We therefore have to invent more intensive ways of getting SS to acquire the necessary patterns.

____________________________________________

2.       PRACTICE.

It is assumed that practice will make the subsequent usage of the patterns automatic. If the practice is meaningful, the usage is supposed to be meaningful. And automatic. There are different ways to practice whatever:

• Receptive practice (listening and reading comprehension, highlighting the target structure)
• Written drills (gapfills, sentence transformations, etc)
• Oral drills (repetition, substitution, communicative drills)
• Writing paragraphs and longer chunks, etc.

However, armed with PPP as our main tool, we will need to drill EVERY SINGLE variant that a student may need. E.g., make an “I’m wearing… ” sentence with every single item of clothing they have learned, + repeat the same for other persons, singular and plural… Is that ever possible? Something like that used to be done back in the 60s as part of the Audio-Lingual method and was largely abandoned (as too immense a task). A course book today will offer a couple of exercises of 7 – 15 sentences (phrases) in each per pattern, plus a few in the workbook, plus another one online, which is not nearly enough even for “rough” practice using PPP. Additional materials will help, but how much time will it take to drill everything that needs drilling?  Will uttering/writing every possible variant once only be enough to make it automatic? The answer is, NO. What is the purpose of the practice exercises, then? In no way is it possible to do that within a 2-hour lesson. Every teacher has faced that.

“Why are some of my students coping so much better than others in the same group, then – PPP seems to be working pretty well for them? “ – you may ask. In my observation, those coping better are in reality at a different stage of their learning – eg, they’d been studying English for 4 years prior to enrolling in your group, but with fewer chances to practice their theory. What they are doing at your lessons is largely reinforcing and really practising the patterns. Others struggle.

____________________________________________

3.       PRODUCTION.

Production is supposed to result from practice. IT DOES NOT.

• We have all seen and felt that the practice stage cannot fit into a 2-hour lesson, and that even huge homework may not be enough for the SS to stop making errors in the pattern in question, and that there is a wealth of patterns in English which can possibly not be practiced “audio-lingually” or in written form to ensure their automation.
• Production results from acquisition. Acquisition does not directly result from learning.
•  Hence, production is not the result of learning (P – P – P fashion). (If puzzled – please consult the link. Then, come back).

Have you ever noticed sudden breakthroughs in your students? Sudden surges in understanding, insights, fluency rocketing overnight, as it seems, after months of “two steps forward – one step back” learning?  If not – keep a closer eye on your students – and you will.  On a graph, a learner’s progress will most often be shown as a series of stairs rather than a continuous curve or a slanted line. Slow-slow-slow – ROCKET! Then, slow-slow again…   Are those surges helped by constant practice? Yes and no. Rather, in my opinion, they are a result of very complex interactions of various cognitive and emotional mechanisms, of which logic and thinking are only a small part. And – EXPOSURE.

• PPP rushes the students into production far too early. It ignores the need to HEAR and SEE the target structures in comprehensible contexts, many, many, many times before acquiring them and becoming capable of producing them.
• I have found that PPP works best with students taking the same level a second time. They will have seen and heard the language over the previous year, but, as should have been expected, were struggling to use it. This time, having had enough exposure, they start to actually use the language as PPP refreshes the knowledge, makes it more systemic, boosts confidence, and fosters production.

Is PPP dead? It is not. Is it suitable for each and every student, each and every background and personality? It is not, either.

---

This was written in 2011, and I’d phrase it differently now, but I still stand by the point. What do you think?

Paper or plastic?

Paper or plastic?  The very first post on  my very new lessonplans site.

It’s a lesson based on a nice video I came across on youtube, with student’s worksheets and teacher’s notes.

Teaching TOEFL and IELTS essays – a technique students love

There has been a great  discussion of a problematizing approach to correcting students’ grammar on Scott Thornbury’s blog (An A-Z of ELT). Well – this post has little to do with the debate itself, but one commenter there mentioned a problem with written assignments: it is hardly useful for the learner to receive the corrected paper full of red marks and crossed out words. Not only is it immensely discouraging,  but it is also very brain-unfriendly. It’s doubtful that the student will benefit at all.

This remark made me think I could  write this post describing an approach that:
– teaches how to write better essays without intimidating or discouraging the learners
– is far more brain-friendly that the conventional “red ink method”
– results in significant improvement of the essays
– is loved by the students.
– does not require too much preparation to produce a good 2-hour lesson involving all 4 skills.

If you are too busy to read the notes below, you can simply download the file with all the 4 stages of the approach 🙂 The detailed notes are under the ‘more’ tag.

Weird jobs, funny jobs… A free downloadable and enjoyable lesson.

WEIRD JOBS!
download the lesson handout
Teacher’s notes.

 Level: high pre-intermediate – upper intermediate

Time:  60-90+ minutes as-is (depending on the level), or it could be split into 2 lessons and some controlled practice can be added.

Skills: speaking, reading, writing.

This is a lesson that emerged out of one of our English club sessions devoted to funny jobs. Therefore, it provides scaffolding to facilitate conversation rather than overt teaching. Anyway, it will be effective as a freer practice lesson after you covered the material necessary to talk about job duties, personal qualities, describing your experience etc. – i.e., the language useful to go through an interview. Alternatively, use it as an introductory task-based lesson prior to teaching those topics. Another option is to add some presentation and controlled practice tasks to make a PPP-lesson out of it (you may wish to split the lesson into 2 then).

As for the last task – making a mock job interview – there is a great and simple free simulation at http://bogglesworldesl.com/lessons/job_fair.htm that you can easily adapt to match this lesson since it is in the doc format.

The language areas involved are:

• Using present perfect to describe your experience.
• Using present perfect with superlative adjectives (the funniest job I have ever heard of).
• Patterns to describe work duties.
• A number of action verbs to describe a job (to test, to assist, to instruct, to procure etc).
• A model to name a job: noun + verb-er (a mule handler).
• Adjectives to describe personality.
• Patterns commonly used in a newspaper job ad.

Copyright note: I borrowed the list of crazy jobs from the Internet and since lots of sites have published it I guess it’s open source information now .

Hope you have a good time teaching this lesson and drop me a line about how it went.

Download the teacher’s notes

Catch a falling star…

On my club’s blog I’ve just published a free downloadable song lesson based on “Catch a Falling Star” by Perry Como. Love the song, and my students invariably fall in love with it. I’m not duplicating it here as they say it’s bad for the SEO :))). Feel free to visit the language club’s site. The current version is for classes of Russian students, but a no-Russian version will be published very soon. In any case, you can edit the doc to match your students’ L1, or simply remove all cyrillics 🙂

My English Speaking club in Moscow…

Been busy setting up a new enterprise – an English speaking club in Moscow, where I live. Almost up an running now. What do you think of  the site? 🙂 Thanx in advance for your comments – I’d greatly appreciate some feedback (although the site is in Russian, but there are quite a few pages in English in the past sessions category (Что было?) – A BUNCH OF MATERIALS TO MAKE GREAT CONVERSATION/CLUB LESSONS.

TOEFL 3-in-1 lesson material: essay + editing + vocabulary

At first I simply wanted to publish my recent TOEFL lesson pack devoted to teaching a toefl essay. Then I thought the material is in fact a 3-in-one lesson, and tried to describe he approach behind it – which resulted in a BIG post about all-in-one approach… having little to to with this particular material :). So I ended up with 2 posts instead of one. 🙂

This one is all about a TOEFL essay, in particular:

• reading a sample essay (high-level, but imperfect)

I took the essay from here and hope the author doesn’t mind 🙂

• evaluation, noticing, editing skills development
• academic  vocabulary focus – for upper internediate+ students
• paraphrasing skills.

DOWNLOAD the lesson materials here.

If I were writing a coursebook…

I keep wondering about why so many coursebooks still  work on listening in one lesson and reading in another one.

I love the many-in-one approach. Not only is it economical in terms of time spent on lesson preparation – it is also a lot more effective than  the “one task – one skill” one.

Here you can download a TOEFL essay + vocabulary + noticing/editing lesson, which is not 100% related to this post 🙂 – but is one example of multiple benefits extracted from one text.

This is my blog to date :)

I think I’ll  make this a sticky. Created it as a response to David Dodgson’s challenge here :). The challenge is also to reflect on these keywords… Well – let me think … I’ll post it soon!

This is where you can make your own: wordle.net