Pattern identification and language learning

A few weeks ago I read an article reporting on a scientific study which had found that students who are good at abstract pattern matching tasks perform better in Hebrew language lessons. (Unfortunately I don't have access to the original journal paper.)

Now, we have two ways of interpreting this outcome: the fatalistic and the optimistic.

The fatalistic interpretation

A fatalist will say that it is proof positive of the existence of talent, and that those who do not have this talent are doomed to failure at learning languages.

The optimistic interpretation

The optimistic interpretation is to say that the successful learners are succeeding despite the teaching, and that this study is, by shedding new light on what the successful learners actually do, showing what teachers should be doing if they want to be successful.

"Rules" are a bit out of fashion in language teaching.  I've always said that this is because people have been teaching incorrect rules, rather than that rules are inherently unsuited to language teaching.  A linguistic "rule" is (or should be, at least) nothing more than an observed pattern emerging from real usage.  If successful learners are those who can identify patterns, then we must assume that after identifying these patterns, they learn them.

So I would suggest that the logical conclusion is that we should be teaching these patterns to students, rather than relying on them identifying them.

An example pattern

When I was learning French (my first foreign language) at high school, I noticed the distinction between the "long conjugations" and the "short conjugations" of the present tense (the long conjugations are the nous and vous forms, and the short conjugations are the rest), and I noticed that in irregular verbs, these forms were almost always regular.  (Is there any verb other than être that has irregular long conjugations?)

So while my classmates were attempting to memorise the irregular verb tables by rote, I was saving myself time and effort by only memorising the conjugations that didn't arise out of bog standard, regular conjugation.

Now it could be that the advocates of the discovery method are right, and that part of my success was down to the fact that I worked this out for myself, but I doubt it.  And even if that's true, is it fair to trade off the success of the majority against the success of a lucky few?  It is obvious that being told this would have reduced the effort required by my classmates to learn their irregular verbs.

Teaching patterns vs teaching rules

The problem with most of the "rules" traditionally presented in grammar books is that they are more strictly ranked and regimented than they are in real life.  Real patterns in real language can't be so neatly packaged by tense.

Take one of the patterns in Spanish taught by Michel Thomas:
The third person plural conjugation of a verb is the third person singular conjugation plus N, except in the preterite.

Now that's not how he taught it, but that's the concise description.  When he taught it, he simply used it in the present, and then got the students to apply it in the other tenses.

One of the reasons some people don't like this pattern is that "except" bit, but this really isn't a problem, because the pattern also holds for the second person singular: third singular +S, except (again) in the preterite.

It's regular, it's predictable.  Even the "exception" is regular in that it's an exception for both 2S and 3P.

This is the sort of pattern that I suspect all those successful learners are finding, and this is the sort of pattern that made Thomas such an effective teacher.

So let's find those patterns, and let's teach them.

The obligatory car analogy...

When you want to explain something complex to a non-expert, there's no tool more useful than a good analogy. Sadly, there's are few tool more open to abuse than the humble analogy, and in a great many cases, the subject of this abuse is the humble automobile. There is a rule of thumb on the internet that says you should never trust a car analogy.

The danger in analogy is that it comes to what appears to be a logical conclusion, even when the analogy is false, but thankfully we've had the conceptual tools to analyse logic since at least Ancient Greece. Heck, even the words “logic” and “analogy” come from the Ancient Greek language.

Regular readers will know that I'm not a fan of the “learning/acquisition” distinction, or the school of thought that says that rules don't matter, and that the only way to “acquire” is through exposure.  Well, recently I was reminded of that particular school-of-thought's own pet car analogy, and I would like to dismantle it here.

Grammar, they tell us, is unimportant. Do we need to know how a language works in order to speak it? Well, they say, consider a car: do you need to know how the engine works in order to drive it?

The reasoning seems persuasive to those who are predisposed to listen, but as with all analogies, the problem lies in the equivalence of the analogised items.

Is “how a language works” analogous to “how the engine works”? Certainly not – it is analogous to “how the car works”. Some commentators would suggest that the engine is how the car works – I would like to argue against this.

To a driver, a car is not the engine. From the very beginning, the goal of the engineer has been to abstract away features that the driver shouldn't have to think about and turn the engine into something of a “black box” – you read the instrumentation, manipulate the controls, and then the car responds in a consistent and predictable way based on what you tell it to do. The driver does not need to know what “RPM” means to recognise when they're over-revving the engine – revolutions-per-minute, cylinder cycles... irrelevant – but the driver does have to be told that over-revving is a bad thing, and has to learn the “rules” of reading the needle and listening to engine noise to avoid doing it.

The acquisition crowd are not, I hope, suggesting that you could put someone in a car with no knowledge of the steering system, gearbox, speed controls and indicator and just let them get on with it. The end result of this would at best nothing, at worst a seriously damaged car. OK, so you're not going to destroy someone's brain by throwing them into a language at the deep end, but if they can't even start the language's “car”, they're never going to get any useful feedback at all.

So we have three elements in the target of our analogy:

• The car as a whole
• The car's control system
• The car's engine

The question is, is the grammar the “control system” or the “engine”. Quite simply... urm... possibly maybe both....

Grammar as Control Sytem

Most of the grammar of a language is unambiguously “control system”, as the speaker must directly manipulate it in order to make himself understood.

Consider the spark-plugs in a diesel engine. Wait... a diesel engine doesn't have any spark-plugs. But this doesn't matter – this makes practically no difference to the driver. The accelerator works the same as the accelerator in a petrol engine with its spark plugs, and pressing it down harder makes the wheels spin faster. “The car”, as an entity, is operated identically – as far as the driver is concerned, it “works the same way”.

But let's look at a grammatical distinction, and for the sake of the argument I'll take the use of articles. English has them, Polish doesn't. If articles were like spark-plugs, that would mean that the article is entirely irrelevant to the manipulation of the language, but this is patently false. If you don't correctly manipulate the article, your sentence is wrong.

So a great many grammar rules are undeniably part of the control system.

Grammar as engine

Grammar as a whole has been a very expansive and extensive field of study – in fact, I'm led to believe that grammar originally meant the description of a whole language. Grammar today usually means “everything except vocabulary, pronunciation and spelling”, so a lot of stuff gets caught up in it which may be considered “engine”. Historical changes, derivational morphology (the etymology of the word presuppose is of very little use to the average learner of English) and distinctions like that between reflexive and impersonal/pronominal pronouns in the Romance languages.

But to use these few examples as a reason to throw out all conscious description of grammar is hugely short-sighted.

My car analogy

So that's their car analogy disproven, and I'd like to replace it with one of my own.

To ask someone to learn a language without grammar is like putting someone in the driving seat of a car without drumming the words “mirror, signal, manoeuvre” into their heads, and without telling them never to cross their hands at the wheel.

On the other hand, a lot of grammar-heavy teaching is like teaching someone to drive by carrying out the exact same manoeuvre 20 times in superficially different (but functionally identical) locations, then moving on and doing the same thing with a different manoeuvre.

This is not what any good driving teacher does.

A driving teacher takes the beginner to a safe, simple environment (eg an empty car park) and teaches the basic rules of operating the vehicle. The learner won't even be allowed to start the engine until they've started building the habit of checking all mirrors. Then they will learn to start and stop. A bit of controlled speed, then a bit of steering. The complexity increases steadily, and the instructor chooses increasingly complex environments so that the learner has to apply and combine the rules in ever more sophisticated ways. Rules are introduced gradually, as required, and then applied and manipulated as the situation demands. The teacher initially picks routes that don't require turning across traffic, then picks a route with a safe across-traffic turn, then adds in crossroads, traffic lights, roundabouts, filter lanes etc one by one. But these features are never treated as discrete items to learn individually – they are elements of one continuous whole that must be practised in the context of that whole.

This is what a good driving teacher does, and this is what a good language teacher does. Listen to one of Michel Thomas's courses¹ and you'll see that's exactly what he does: an increasingly complicated linguistic environment, and no language point ever treated in isolation beyond its basic introduction.  That's proper teaching, and it's all built on grammar and rules.

1I mean a course that he himself planned and delivered, not one of the courses released after his death.

Guardians of Grammar

I came across a link before Christmas that I found quite interesting.  It was a Guardian guest article by an applied linguistics lecturer, Dr Catherine Walter.  The headline was bold and clear: "Time to stop avoiding grammar rules", so of course I was interested, being firmly in the pro-grammar camp myself.  The subhead went for the jugular: "The evidence is now in: the explicit teaching of grammar rules leads to better learning"

Excellent, I thought, everything I've ever said has been vindicated.  Does this mean the tide will now turn in the teaching world?  Sadly not.

The article was something of a disappointment.  Despite its mention of meta-analyses, it provided precious little evidence and instead went down the road of discussion views and hypotheses... views and hypotheses which I personally agree with, but do nothing to convince those currently holding the opposite view.

That said, I still find it difficult to see how this sort of logic fails to convince people:

most English language learning takes place in countries where English is not the predominant language: a foreign language situation. Much of the thinking leading to strictures against grammar teaching has taken place in countries where English is the predominant language: a second language situation. The enormous difference in exposure to the target language makes arguments based on exposure or emergence much less plausible in the foreign language situation.

ie the leading lights of the ESOL/EFL world are working in an atypical environment (and of course Walter one of them) -- their students have plenty of opportunities for practice outside of class.  Why are so many teachers so quick to accept the pronouncements of people whose teaching environment and student base is so radically different from theirs?

(This argument alone goes beyond the article's boundaries of exposure vs training and drills into the fundamental identity crisis of the Communicative Approach: in a class with a shared native language, isn't English inherently a barrier to communication, not a facilitator of it?  Surely the communicative imperative is broken if the language gets in the way?)

She also talks about "chunking" -- a central pillar of the lexical approach.  She says:

But the best estimate is that there are hundreds of thousands of chunks in English; learning enough of these to have an appropriate chunk to hand in a given situation is not a quick or trivial job. With much less time and effort, learners can acquire grammar for putting together comprehensible phrases and sentences that can serve them on the long journey towards more native-like proficiency.

...which is perfectly correct.  If a chunk is composed of regular grammatical features, then the chunk can be understood as a construction until the learner has seen it enough times to identify it as a chunk.  Learning some of the underlying grammar rules of a language makes the learner capable of dealing with a lot larger a subset of the language than learning the same number of chunks.

But Walter doesn't go far enough, because she doesn't point out the inconsistency of those who espouse both "learn by induction" and "learn by chunks".  Learning by induction is supposed to be the "natural" way -- it's how babies do it.  But it is readily demonstrable that babies learn grammar before they learn chunks -- they cannot repeat a large chunk until and unless they have internalised the component language.  OK, you will find babies and older children who have incorrectly generalised two frequently co-occurring words as units, but even if you consider that chunking (the alternative interpretation is that they've mistakenly identified the two words as being single ones), the vast majority of early language appears to be unchunked.

If the justifications and theoretical underpinnings of the approach are inconsistent, why have any faith in it?

Just what is applied Linguistics?
A lot of teachers will be likely to reject Walter's views because she's an academic, a linguist.  Sadly, the term "applied linguistics" is somewhat opaque, because applied linguistics could involve so many things -- from forensic linguistics ("was this confession written by the accused?") right through to producing better instruction manuals for washing machines.  But no, applied linguistics is used almost exclusively to mean one thing and one thing only: language teaching.  Most "applied linguists" are genuine, honest-to-goodness teachers, but ones that do research as well (many universities make a lot of money by running summer EFL courses out of their Applied Linguistics departments).

Applied linguistics as a field is so far from ivory towers that one author named his textbook "An Introduction to Applied Linguistics: from Practice to Theory" (Alan Davies, Edinburgh University Press), in order to highlight that all language teaching theory starts in the classroom, and all theory is an attempt to describe and understand successful teaching.

So Walter isn't an outsider -- she isn't just someone who doesn't understand the reality of teaching, which is one of the criticisms most often aimed at academics who attempt to give advice on how to teach.  It didn't pop up in the comments thread explicitly, but there was a rumbling...  One commenter, Espoolainen, noted the lack of "chalkface examples".

The collective noun for "anecdotes" is "an internet"
We all know that the plural of "anecdote" is not "data", but yet anecdote is what the internet thrives on.  We don't want figures, we don't want proof, we want one or two stories with real protagonists.  It doesn't matter to many of the commenters on the article that the uncontrolled variables in a single example make it meaningless -- that's what people want to hear, but they don't call it "anecdote", they call it "chalkface examples" and kid themselves on that it's not really anecdote.  (On the other hand, I'd bet good money that if Walter had given any genuine examples, 101 commenters would have jumped on her for using anecdotes in her article.)

One of the other commenters linked to a paper by the National Research and Development Centre for adult literacy and numeracy(NRDC) which was proclaimed by the very next commenter as "more useful than the original article".

The report is full of "chalkface" examples.  One chapter is called "take 40 teachers: ESOL teachers' working lives", and the next "take 40 classrooms: teaching and learning strategies in the classrooms observed", followed by "Telling cases: ten classroom case studies".

That's a plurality of anecdotes, isn't it?

Meta-analyses and literature reviews
The problem with arguing any case in terms of figures, evidence or science, is that for every published paper you can cite to support your case, your opponent will be able to cite one that states exactly the opposite.

Good academic practice relies on a thorough and complete overview of all available data, which is nigh-on impossible for any one person to do, and a humungous waste of time for everyone to do.  This is where meta-analyses and literature reviews come in.  Someone (more often a group of people) sits down and sifts through all the published papers they can get their hands on and try to work out what they all mean when taken together.

This means eliminating studies with unreliable methodologies, checking whether the conclusions were truly warranted and balancing the volume of evidence on both sides.

This is the sort of material that Walter claims backs her up, and when pressed she quoted the following sources:
Norris, J. M. & L. Ortega. 2000. Effectiveness of L2 instruction: a research synthesis and quantitative meta-analysis. Language Learning 50/3: 417-528.
Gass, S. & L. Selinker. 2008. Second Language Acquisition: an Introductory Course (Third Edition). New York: Routledge/Taylor.
Spada, N. & Y. Tomita. 2010. Interactions between type of instruction and type of
language feature: a meta-analysis. Language Learning 60/2: 1-46.
Spada, N. & P. M. Lightbown. 2008. Form-focused instruction: isolated or integrated?
TESOL Quarterly 42: 181-207

Now let's look at the NRDC study.  In the introduction they say:

The five NRDC Effective Practice Studies explore teaching and learning in reading, writing, numeracy, ESOL and ICT, and they set out to answer two questions:

1. How can teaching, learning and assessing literacy, numeracy, ESOL and ICT be improved?
2. Which factors contribute to successful learning?

Even before NRDC was set up it was apparent from reviews of the field (Brooks et al, 2001; Kruidenier, 2002) that there was little reliable research-based evidence to answer these questions. Various NRDC reviews showed that progress in amassing such evidence, though welcome where it was occurring, was slow (Coben et al, 2003; Barton and Pitt, 2003; Torgerson et al, 2003, 2004, 2005). Four preliminary studies on reading, writing, ESOL and ICT were undertaken between 2002 and 2004 (Besser et al, 2004; Kelly et al, 2004; Roberts et al, 2004; Mellar et al, 2004). However, we recognised the urgent need to build on these in order greatly to increase the research base for the practice of teaching these subjects.

So what we've got is a group saying that they were essentially starting from scratch; that they were incapable of performing a meta-analysis with the existing research.  That makes this a single, solitary study, with 500 students (not a lot if you're building a national strategy on it).  And there's absolutely no mention of Norris and Ortega's meta-analysis, either to rule it in or rule it out of their work.  There are two possible interpretations of its absence: either they didn't know about it or they didn't like its conclusions.  If it's the latter, in my book that's just dishonesty.  If you disagree with something, don't ignore it: explain why you are disregarding it, so that readers can judge for themselves.

The NRDC report is pretty much what Walter complained of in the article:

Each approach has been defended with carefully structured arguments, and some approaches have been embraced enthusiastically by ministries of education around the world.

Arguments, not evidence, define policy.  That's what she's saying, and it's all too often true -- just Google the term "policy-based evidence"!

Taking off the blinkers of English
Dr Walter came so close to making an important point when she said:

There is a problem with English: it is a morphologically light language. It doesn't have many different verb endings, and its nouns only inflect for plural. If the language under discussion were Polish, with its three noun genders and seven cases, the idea that teaching grammar rules wasn't necessary would probably not even occur.

Why did she not expand on this?  I came across a masters thesis on the net a while ago on the teaching of Finnish and it took as a given that conscious grammar study was required, because decades of figures showed conclusively that immigrants really never quite "got" the language without a bit of explanation.

The problem with Finnish is often claimed to be its complexity, but the author of the thesis was more interested in the difference between Finnish and the immigrants' native languages.  Just look at the very first sentence of the abstract:

Research has shown that explicit instruction with a focus on forms is needed in learning a very different language.

Difference: that's what we should be looking at.

Consider that one of France's top-selling language products is Assimil, a course that makes a big deal out of learning by assimilation.  We'll leave aside the fact that there's still a fair bit of explicit grammar explanation in the books and instead look at the situation in terms of differences.  French is anything but a language isolate, being part of a broad continuum of Romance languages.  When you consider that Italian and Spanish are two of the most popular languages for study in France, it's easy to argue that Assimil got its reputation by teaching languages with a very low "difference" from the learner's native tongue.  Even German, the other big neighbour, isn't too different from French in many respects.  (French may not mark case in nouns, but its pronoun system still makes a distinction between subject, direct object, indirect object and possessive that maps closely enough to German's nominative, accusative, dative and genitive to give the learner a head-start.)

ESOL has justified itself generally by dismissing any learner errors as unimportant, but it can be readily seen that the most common errors are caused by fundamental differences between languages, and the most obvious example is the use of articles, because most learners have some kind of problem with them.  Speakers of article-less languages either leave them out or add them in an arbitrary and meaningless way (NB: arbitrary, not random: most are consistent in their misuse) and speakers of Romance languages tend not to be able to chose between an indefinite article and the number one.


Even if English isn't very complex, it is still different in several fundamental ways from any given language, and if that's a problem for Finnish, it must be a problem for us too...


...and it is, because these problems never go away.  There's no magic tipping point where the Spanish speaker has had enough exposure to English to tell the difference between "a car" and "one car", or where a Polish speaker suddenly can distinguish between "cars" and "the cars".  It just doesn't happen.

Walter's cardinal sin
But Walter's biggest mistake was trying to make it sound as though her conclusions were based on some piece of brand-new research.  It grabs the attention, but it undermined her argument when she got called on it.

Why not simply point out that the evidence has been available for years to those who were willing to listen?

I know it's a hard argument to sell, but it's the only intellectually honest one.  Giving an explanation of why orthodoxy is so slow to change won't necessarily have an immediate effect, but in the long term it should open peoples critical faculties.

Why orthodoxy is so slow to change
The human lifecycle is a fairly predictable thing.  We are born. We go to school.  If we do well at school, we'll probably end up in university.  After university, we walk out into the world looking to do things differently and better than those who came before us.  But nobody listens to us and we end up as the least important employee in our companies.

Over the next decade or two, the genuinely gifted start to rise to prominence in their careers.  It's only now that they can start to impose "their way" on the world, but their information is over a decade out of date by this point.  Their view of the mainstream is in reality a river that ran its course a long time past.

This leads them to reject new research that supports the old orthodoxy that they have already "proven" wrong in their heads.

Would any of us in our 30s or 40s pass our university exams if we went back now?  Probably not -- our information is hopelessly dated (unless you're a mathematician, perhaps).  So who are we to tell others in our field what to do?

Don't they remember what it was like to go into that first job and have everything you knew about your field rubbished by people who just didn't understand the new stuff?  Why do we visit this same humiliation upon the next generation?

Everyone understands grammar.

I've often heard it said that grammar is difficult or that normal people don't understand grammar.  Some teachers believe that grammar would be useful in the language classroom, but only if the students came in understanding grammar in the first place -- they don't want to waste time teaching theoretical grammar before they can start teaching the target language.
And these are good arguments but for one small detail:
everyone understands grammar.
It has been observed that in cloze tests (passages of texts with blanked out words) native speakers will have a notion of the word-class of the correct answer before they know what the specific word is.
Take these three simple examples:

1. I like ____.
2. I want a ____ one.
3. Don't ____ me.

It's impossible to know what the original word was that the author intended, but at the same time, it's pretty obvious that number 1 is a noun, number 2 an adjective and 3 a verb.  If you were to put these sentences in front of any literate English speaker, they could give you any number of possible words for each example that fit the categories.

So they have an internal concept of word-classes, even if they don't have an explicit awareness of it.  This means that there's very little work for the teacher to do -- all you've got to do is point out what they already know.
Michel Thomas tried to do this a different way, and it's a way that tends to draw a lot of flak.  On his recorded courses, you can hear him say that he doesn't like the traditional definition of "person, place or thing", and then he gives a couple of examples of abstract nouns, which he says aren't really things.  He then goes on to introduce his prefered rule -- a noun is anything you can use "the" before -- and a couple of examples, using both abstract and concrete nouns.

It is quite trivial to prove that his rule is wrong -- "the white house"; adjective after "the"; wrong.  "John"; most proper nouns don't take "the"; wrong.  And yet for all it's inaccuracy, this "rule" seems to work better than most.  Why?  Because Thomas isn't trying to teach a new concept -- he is merely trying to evoke the concept that the learner already has.  In fact, he is doing exactly the opposite of defining, and I would argue that he is doing so on purpose -- if you already know about the concept, thinking too much will override your instinctual understanding, so thinking should be avoided.

What Thomas did was quite subtle.  He took the traditional (correct) rule, but said he didn't like it, freeing the student from feeling inadequate for not understanding it.  He then presented the case that causes consistent real problems for students, the abstract noun.  All he basically said in informational terms was "these are nouns too", but he worded it in such a way as to say "it's not your fault if these don't make sense in the old rule", swatting away any confusion and guilt or inadequacy.  His final comment, about "the", is valueless out of context -- it is factually incorrect, but he uses it to reinforce the concepts already clarified by the old rule and the abstract examples.  He ties them into one bundle.  The student never writes down the rule, the student never memorises the rule, in fact, regardless of what he says, he has not given a rule.  All he does is evoke the learner's internal concept of nouns and label it with the word noun.  I'm pretty confident that the average student coming out of a face-to-face class with Thomas would have forgotten he'd even said it, but they'd be able to label nouns pretty acccurately if asked to.

It took Thomas a few minutes to teach "noun", "verb" and "adjective" that way, less time than it takes me to explain what he did.

Any concept that exists in the student's native language has the potential to be taught in a similar way.

The secret is not to take too long, and not to get hung up on the technicalities.  All you need to do is evoke the concept and stick a label on it.

And if you think about it, that's what we do as teachers every day -- it's called "vocabulary".

Counterintuitive, perhaps, but sometimes it's easier to start with the harder material...

In general, whenever we teach or learn something new, we start with the easy stuff then build on to the more difficult stuff.  But this isn't always a good idea, because sometimes the easy stuff causes us to be stuck in a "good enough" situation.

When I started learning the harmonica, I learned to play with a "pucker technique", ie I covered the wholes with my lips.  The alternative technique of "tongue blocking" (self descriptive, really), was just "too" difficult for me as a learner.  So for a long, long time, the pucker was "good enough" and tongue blocking was too difficult for not enough reward.  It limited my technique for a good number of years, and now that I can do it, I wish I'd learnt it years ago.

The same block of effort vs reward happens in all spheres of learning.  If you learn something easy, but of limited utility, it's far too easy to just continue along doing the same old thing, and it's far too difficult to learn something new, so you stagnate.  Harmonicas, singing, swimming, skiing, mathematics, computer programming; there's always the temptation to just hack about with what you've got rather than learn a new and appropriate technique.

This problem, unsurprisingly, rears its ugly head all too often in language learning, but with language it has an altogether insidious form: the "like your native language" form.  If you've got a choice of forms, one is going to be more like your native language than the other, and this is therefore easier to learn.  Obviously, this form is going to be "good enough", and the immediate reward to the learner for learning the more difficult form (ie different from the native language) isn't enough to justify the effort.  However, in the long term, the learner who seeks mastery is going to need that form in order to understand language encountered in the real world.

The problem gets worse, though, when you're talking about dialectal forms.

Here's an example.  Continuous tenses in the Celtic languages traditionally use a noun as the head verbal element (known as the verbal noun or verb-noun).  I am at creation [of] blog post, as it were.  Because it's a noun, the concept of a "direct object" is quite alien, and instead genitives are used to tie the "object" to the verbal noun.  In the case of object pronouns, they use possessives.  I am at its creation instead of *I am at creation [of] it.  Note that the object therefore switches sides from after to before the verbal noun.

Now in Welsh, the verbal noun has become identical to the verb root, and is losing its identity as a noun.  This has led to a duplication of the object pronoun, once as a possessive, once as a plain pronoun -- effectively I am in its creation [of] it.  This really isn't a stable state, as very few languages would tolerate this sort of redundancy, and the likely end-state is that the possessive gets lost, and the more English-like form (I am in creation [of] it) will win out.  In fact, there are many speakers who already talk this way.

But for the learner, learning this newer form at the beginning is a false efficiency.  There are plenty of places where the old form is still current, so unless the learner knows for certain that they'll be spending their time in an area with the newer form, they're going to need the conservative form anyway.  To a learner who knows the conservative form, adapting to the newer form is trivially easy, but for someone who knows only the newer form, the conservative form is really quite difficult to grasp.

So teaching simple forms early risks restricting the learner's long-term potential.  So while you want to make life simple for yourself or you students, make sure you're not doing them or yourself a disservice.

Overgeneralising and undergeneralising in general...

In English, we have two ways to talk about nouns in a general sense.  In normal speech, we say things like cats are vicious little creatures -- i.e. we use an indefinite plural.  In some very formal prose, you'll see instead the cat is a vicious little creature --i.e. a definite singular.

The existence of the second is probably just a case of "translationese" -- it arises in lots of translations of Latin works, and I believe it is used that way in most of the modern Romance languages (French, Italian etc).  Unfortunately this isn't easy for me to verify, as I have no idea whatsoever what to look for in the index of my grammar books.

Bizarrely, this fundamental (and straightforward) element of language seems to have been overlooked in the classical grammar models, so there is no common label for it (hence me not being able to look it up!).  This means it is often overlooked in teaching, too.  Many beginners' courses pass it by, and even when it comes up, you're not likely to get more than a little box-out mentioning it.  It's not really "taught" in the same way as other grammar points.  I suppose the reason for this goes back to the very basics of the structuralist view of grammar, which values form over meaning, and too often simply gives a few short sentences explaining usage after drilling form.

But we've been moving away from structuralism for quite some time now.  The in-general/universal has been marooned by the incoming tide, as functional and communicative approaches have picked up on the link between form and meaning in the noun and article for specifical and truly indefinite cases, but they've not integrated the general/universal with it.

This underemphasis of the general/universal is particularly noticeable in Gaelic.  It's not a subject I've seen come up often at all.  I read it in one book and one book only, and I don't believe I've ever heard it discussed ever in classes.  According to the book (well, my memory of it -- the book's 100 miles away), the general/universal in Gaelic is the definite singular. (The cat is a vicious little creature, the lion is a noble beast etc.)  And yet....

When you study the genitive in Gaelic, it may be pointed out to you that while "describer nouns" in English always stay singular even when representing a plural concept (for example "biscuit" in "biscuit tin", "tooth" in "toothbrush"), this isn't the case in Gaelic genitives, which have both singular and plural forms.

So I was giving a talk in a classroom debate, and I mentioned "teenage pregnancy" which I rendered as "leatromach nan deugairean" -- "pregnancy [of] the teenagers".  Genitive, plural.  After the class, I started asking myself if that was right, thinking of the general/universal rule.  Now I'm too confused and I'll just have to ask one of my teachers to try to clarify....

When teaching grammar, stick with the uncontroversial stuff...

I was in a grammar class for my Gaelic course today, and we were looking at noun declensions.  For one set of questions, we were using the word fàinne (ring).  Unfortunately, this word is masculine in some dialects and feminine in others.  The question stated that the word was masculine, but my partner for the exercise (a native speaker) has used it all her life as a feminine word.  She declined it perfectly correctly in each case -- as a feminine word.  The explicit instruction to decline it as masculine was ignored because she already has 100% intuitive command of the word.

Correct completion of the task therefore required that she stop dealing with the words as "language" and start thinking of them as some kind of mechanical logic puzzle.

The problem with the task is that it became counter-intuitive.  When teaching grammar, we need to employ as much pre-existing knowledge as possible.  Grammar teaching for natives has to start with forms they know, because you are not actually teaching "grammar", you are teaching "grammar awareness", and that simply means making them consciously aware of things they already know intuitively.

So you have to pick the most uncontroversial examples, the most universal and unchanging.

Work with your students, not against them.

The importance of phonology

OK, so I promised this a while ago, and I've let myself get distracted by a few other points in the interim, but I'll try to draw them in and show how they are related to the teaching of phonology in general.

In my posts 4 skills safe and 3 skills safe, I argued that the division of language teaching into the traditional 4 skills of reading, writing, speaking and listening was trivial, superficial and of very little pedagogic value.  Instead, I suggested that we should look at individual skills of syntax, morphology and phonology, and that we could add orthography as an additional, more abstract skill (Lev Vygotsky described reading and writing as "second-order abstractions").
vowe
Phonology often gets very little attention in the classroom, as it is seen as a sub-skill of speaking, and speaking's "difficult".  But phonology is fundamental to many languages.

If you haven't already, you might want to take a look at my posts In language, there's no such thing as a common error, and Common errors: My mistake!  In the first post I described a particular common error in written English (might of instead of might have, could of instead of could have etc) and in the second I expanded on the mechanisms that cause this "error", with the aim of showing that this wasn't an "error", but in fact a change in grammar, analogous to changes that have occurred in other languages.  What I didn't focus on there, but which is extremely relevant here, is that this change in grammar is pronunciation-led -- ie the phonology of English has caused this change in grammar.  The prosody of English has led to 've being always weak, and it has lost the link to the related strong form have.

And of course the change in the Romance languages that I mentioned in the second post is also led by phonological patterns.  If you look at any language whatsoever, many grammatical rules have arisen from mere matters of pronunciation.

The archetypal example is the English indefinite article -- a/an.  You may well be aware that like most other Indo-European languages in western Europe, this evolved out of the same root as the number one.  But the modern number one is a strong form and has a diphthong.  A/an is a clitic and always weak, so split off (completely analogous to 've and have).  This weak word /ən/ then lost its [n] before consonants, simply because it's easier to say that way, and retained it before vowels again because it's easier to say like that.  (And if you'll indulge a slight digression, that brings us back to would've etc, because you'll often hear woulda before a consonant and would've before a vowel.)

If you look at the Celtic languages, one of the trickiest parts of the grammar is the idea of initial consonant mutations.  Lenition in Modern Irish is a bit inconsistent (probably due to the relatively large number of school-taught speakers against native speakers), but the three mutations in Welsh are fairly systematic, with mutated forms usually only differing from the radical in one "dimension" of pronunciation.

These sorts of rules become very arbitrary and complex when described purely in terms of grammar, whereas when considered physically, they make a lot more sense.

Let's go back to a/an and take a closer look.  We all know the rule: a before a consonant, an before a vowel, right? Wrong! It's actually: a before a consonant phoneme, an before a vowel phoneme.  To see the difference between the two, fill in the following blanks with a or an:

I want __ biscuit.
I need __ explanation.
He is __ honest man.
I have __ university degree.

Now it's not a difficult task for a native speaker, because you wouldn't normally have to think about it: honest may start with the letter H, but you know intuitively that you don't pronounce it, so you write an without thinking.  Similary, university may start with the letter U, but you know intuitively that it starts with a y-glide sound (like "yoo", not "oo"), so you write a.

I have seen quite a few English learners write "an university" or "a honest man" because they are either trying to work from a grammatical rule in isolation from pronunciation, or because they simply pronounce these words wrong.  In the case of honest, the problem is compounded if the student can't pronounce H, because if he follows the rule correctly on paper, he undermines the phonological basis for the true rule.

It follows, then, that we cannot teach grammar without considering phonology.  (And anyone who has succeeded in understanding the French liaison rules can tell you categorically that this is true.)

But how does phonology affect us in other ways?

Phonology and the ease of vocabulary learning

It may seem trivial, but for his PhD thesis, an Australian teacher of Russian demonstrated that it is easier to learn foreign words that are possible in your native language than ones that aren't.  EG the word brobling with first-syllable stress is easy, brobling with second-syllable stress is a bit harder, grtarstlbing with lots of consonant clusters that can't occur in English is very difficult.  He then took a massive leap of logic that I'll examine later in greater depth.

This corresponds with what a lot of teachers believe, but few teachers have the time or patience to implement: that it's easier to teach phonemes one at a time and reuse them in different words.  Again I'll come back to that when I start discussing techniques.

For now, though, I'll simply suggest that it's easier to learn words that are made out of familiar "blocks" than ones that aren't.  It follows from this that good teaching of phonetics (whatever that means) is a prerequisite to vocabulary learning.

Phonotactics: the "crisps" problem

My high school had an exchange programme running with a school in France.  Teenagers are naturally curious beasts, and when my big brother and sister first went on one of these exchanges, the class discovered how funny it was to get the French people to say crisps (UK English for what the French and Americans call chips).  Very few of the French kids could actually pronounce it, because they were using French phonemes with a northern accent (the school was near Lille).  The French P is unaspirated (unlike English) and the French S is quite slender and hissy.  As a combination of sounds, French SPS is difficult, nearly impossible -- the P either gets lost in the hiss or one of the Ses gets cut short.  The English combination is physically much easier.

Similar problems occur in other places.  Spanish people find wants quite difficult to say, because Spanish T is not compatible with Spanish N or S due to the method of articulation.  NTS in Spanish needs the tip of the tongue to be in two different places at once -- the alveolar ridge for N and S and the gumline for T.

The problem is that many books will tell us that T, D, B, P etc are sufficiently similar in English and Spanish, French or whatever that we can use them equivalently, but this is only true for each phoneme in isolation.  Once we start trying to combine them, the differences start to accumulate.

Which brings us back to:

Grammar again - and how writing suffers for it

If you cannot pronounce the inflectional affixes in a language, your grammar suffers.  Many, many Spanish learners of English drop their -s and -ed suffixes because of the problems of incompatible sounds.  They replace it's with is.  These mistakes filter through from their pronunciation into their internal model of grammar and eventually into their writing.  But it's easy to ignore this, because most of the time they correct their own writing mistakes with their declarative knowledge, and on the few occassions where they don't correct it, the teacher simply tells them the rule again, but never attacks the root cause of the problem: if they learned to pronounce English [t] and [d] phonemes, most of the difficult sound combinations would become much, much easier, their internal model of the grammar would be built up to incorporate these non-syllabic morphemes (and there are no non-syllabic morphemes in Spanish as far as I know, so it's a totally new concept to them edit (2-feb-2014): Spanish has at least one non-syllabic morpheme: plural S after a vowel.) and they would write natural based on their procedural knowledge of the grammar..

And finally...
Allophones and comprehension

Apparently there are certain accents that are considered "hard" in some languages. Now I'm not implying that there is no such thing as a hard accent, but I do believe that most of the difficulties stem from the teaching, not from the language.

In Spain, the accent of Madrid is considered quite difficult to understand.  The reason for this is that the madrileño accent tends to lenite (weaken or soften) its non-intervocalic consonants.  The classic is the weaking of D to /ð/ (roughly equivalent to TH of then).  There is little physical similarity between the English D and ð as is clear from their technical descriptions: /d/ - voiced alveolar plosive; /ð/ - voiced dental fricative.  But the Spanish /d/ is a voiced dental plosive, which the description shows is quite similar to /ð/.  Basically, the soft D in Madrid is basically an incomplete hard D -- the tongue doesn't quite go far enough to touch the teeth and stop the sound, but instead it hisses slightly.

Now, if understanding language is a reflective act (as I claim here and here) then we understand sounds by considering what shape our mouths would be in if we were to make the sound we hear (something suggested by the concept of mirror neurons).  The soft and hard Ds in Spanish are not "soundalike" allophones at all, but they have a similar shape, which is different from the English D.  To me it seems clear that physically learning the Spanish hard D shape would result in better comprehension of the similarly shaped soft D in a way that simple hearing it won't accomplish.


Conclusion

All in all, it seems to me that phonology is an intrinsic component of language, and that the system of a language falls apart when phonology is not given the proper support throughout the learning process.

As for how to teach phonology, I have my own views, but I'm currently reading up on some alternative opinions so as to give a more balanced write-up of the options available.

Learning from mistakes...?

I've heard it said many times that the best way to learn a language is to make mistakes and be corrected, because this correction somehow "personalises" the learning.  But to me, this is a logical absurdity.

Here's an example from my personal experience.  In Spanish, there is a special word hay that is equivalent to the English there is.  You therefore do not translate there is verbatim (which would give you *allí está).  In Spanish, they also have a word "demás" which is used for other(s), meaning the rest of a group (not other as in different).

Now I often mistranslate the others (the rest of the guys) as *los otros, but I never mistranslate there is as *allí está.   This means I can say hay correctly despite never being corrected and I keep getting los demás wrong despite fairly frequent correction, and conceptually one is no more difficult than the other.

This is only one example, but in general when someone corrects my Spanish, it's for one of a closed set of mistakes that I make all the time.  Being corrected seems to have absolutely no direct effect on my errors.

The only effect is when I subsequently choose to work consciously to eradicate that mistake.  Or perhaps more accurately, when I consciously work to learn the correct form, because more often than not, an error isn't the result of learning something wrong, but actually an indication that you haven't learnt it at all.  I get hay right because I learned it early on, I get los demás wrong because I never really learned it, and even now I'm only "aware of existence" -- I still don't feel I've learned it.

This isn't to say that errors and correction are valueless, not at all.  Corrections have no special ability to make language stick, but they at least indicate a gap in your knowledge and they often give you a starting point for filling that gap, so you certainly should listen to and take note of any corrections you're given.  What you shouldn't do is ascribe magic powers to corrections and believe that they are a substitute for other ways of learning -- this will only slow down your progress while making conversations in your target language far less enjoyable for both parties.

Why is this idea so persistent?

There is a small set of mistakes that people only make once.  For example, when a Spanish learner tries to say she's tired (cansada) but instead says he's married (casada), it is an instantly memorable situation and unlikely to happen again.  Of course it's a situation that's quite embarrassing, which might be compounded if she now mistakenly says she's pregnant (embarazada).  And she'll never do that again.  Similarly a Spanish speaker suffering a cold is unlikely to forget that being constipated is not the same as having a bunged-up nose (constipado in Spanish).

So these mistakes certainly do lead to better recall of the correct form, but this is not because of how correction works as a generally applicable learning strategy, but rather a consequence of the vocabulary in question.  Saying translación (an archaic variation of translación - movement/transfer)instead of traducción (translation) doesn't have the same comedic value (and in fact isn't likely to obscure the meaning in context) so doesn't have the same potential to stick.

Most often it is these extreme cases, these outliers, that are used to convince us of the efficacy of the technique, but don't be fooled: learning the correct form from the word go is far more effective than making it up as you go along and picking up a catalogue of corrections.

Why I chose to study grammar

No matter what language you're learning, and no matter how complicated its grammar seems to you, one thing holds true for any human language you might study:

There is hardly any grammar.

How so?  Go into a bookshop or library and compare the size of the biggest grammar book with the biggest dictionary.  And don't forget that big dictionaries are printed on thinner paperstock than grammar books.

Juan Kattan Ibarra's Modern Spanish Grammar has 472 pages, and is pretty comprehensive.
Collins' unabriged Spanish dictionary has a whopping 2208 pages, in a smaller typeface than the grammar book and with a printed area approximately equal to two pages of the grammar book, and formatted to reduce white space to an absolute minimum.  In terms of raw text, a comprehensive dictionary is about 20 times as big as a good grammar book.

On top of that, each grammar point in a grammar book needs a couple of pages of explanation and multiple examples, where a word gets a couple of inches in each half of the dictionary.

Really, there's loads of words in any language, and hardly any grammar.

So why not get the grammar early on?  It's quick and it is universally useful.  Very few people can pass a single day without using the majority of the verb tenses and noun cases available in their native language.

Vocabulary is a different matter.  When did I last say "robot"?  I can't remember.  "Lamb"? Roughly four weeks ago.  "House"?  About 2 weeks ago.

So if I study vocabulary, I not only have an almost never-ending task (the dictionary I mentioned above has 315,000 references), but I also find myself unable to remember words because I don't use them enough.
But if I study grammar, I can cover all the basics really quickly, and those basics can be used every single time I have a conversation, and they will stick.

The best part, though, is that once you know grammar, learning words is easier, because you can use them and understand them in various natural contexts, because grammar can change both the form and meaning of words.

Right now, I'm mapping out the grammar of Polish in order to teach it to myself and a friend, and when I'm done, I expect to know less than 50 words.  But I can learn more words later.

"Say what sounds right."

Bad advice has an annoying habit of sounding like good advice, and this little phrase really is something of a wolf in sheep's clothing.  It's definitely appealling when someone points out that that's what we do in our own language.

I wouldn't argue that my end-goal isn't to be able to simply say what sounds right, but I just can't see how that end-goal affects my learning path: nothing will never "sound right" until I've learnt it, so how can I learn by what "sounds right"?

The consequences of this are not insubstantial, because if I've not learned it yet, what's going to sound right to me?  What sounds right is something that I have learned, but this will be out of context.

Take for example the verb "start" in English.
You can "start something".
You can "start doing something".
You can "start with something".
You can "start by doing something".

If you learn only two of these, then only those two will "sound right".

Saying "what sounds right" traps you into what you know and stops you expanding your language.  What you need to do is stop and think, and use the appropriate form, even if it isn't familiar enough to "sound right" yet.

Here's another example.

French has a feature called "liaison" -- certain final consonants are silent but reappear when followed by a vowel, but only if the two words are tied together syntactically.

In the word "vous", the S is normal silent, but in the phrase "vous allez" (you go), it has a /z/ sound.
Now, the past participle of to go is "allé", which is pronounced identically to "allez", so when you ask "êtes-vous allé" (have you gone), if you go by what "sounds right", you might pronounce that /z/.  But in this question "vous" and "allé" are not syntactically bound together and liaison should not occur.

The idea of "what sounds right" reaches a very messy conclusion in Scottish Gaelic. A single syllable consisting of a schwa before a noun can be one of three things: "the", "his" or "her".  "his" always causes initial lenition (soft mutation of the first consonant) of the following noun, "her" never does.  As "the", this form can occur before masculine and feminine nouns in certain cases, and causes initial lenition, and only with certain letters.

Many teachers suggest that you learn noun gender by "what sounds right", by agreement with the article, but your ear will be exposed to the various case-inflected forms and possessives, so what sounds right might not be "the boy" at all, but "her boy" or "his boy".

The 3 sources of confusion in vocabulary.

There are various pieces of advice on the internet regarding how to learn vocabulary, but most writers set out to write their advice with the goal of convincing you that their way is best.  This means that they skip the weaknesses in their chosen method, and they attack other methods on rather simplistic, superficial grounds.  I would be happy to do the same thing, as it would really stoke my ego to know that people were doing stuff because I said so.  But that wouldn't be particularly useful, so I'm going to try to avoid telling you how to learn vocabulary.

Instead, what I want to do is arm you to make an informed decision on techniques yourself.  It's a topic I may revisit later, but for now I want to focus on the reasons vocabulary items become confused.  As per the title, I break confusion down into three main categories: confusion by form; confusion by function and confusion by co-occurrence.

Confusion by form
Confusion by form is the simplest.  If two words sound and/or look alike, they are very easy to confuse.  English is full of great examples -- classic spelling mistakes between homophones such as "bough" and "[take a] bow", "aloud" and "allowed" etc.  Most languages aren't quite as bad as English for having homophones that are written differently, so the confusion is normally between similar words, not identical ones.

But there's more to it than that.  Have you ever been trying to think of a word and you've got a sort of shadow of the word in your head?  The features of words that are easiest to remember are the first syllable and the number of syllables, as well as which syllable is stressed.  So similarity doesn't rely on just having similar letters all the way through -- it can occur on just the first syllable, or the words may both share a particular rhythm.

We can consider so-called "false friends" as simply a special case of confusion by form, as the only difference is that the confusion occurs across languages, rather than within a language.


Confusion by function
Confusion by function can be split into two subcategories: function of grammar and function of concept.

Confusion by function of grammar is fairly simple but has extremely absurd results.  Some people can't believe it exists, but keep your ears open and you will hear it at some point.  A word gets dropped in that fits the grammatical category of the place it is in the sentence, but it makes no sense.  Anecdotally, I'd say that I've seen it occur mostly with verbs.  So for example someone might say "I like drinking books".  It's rare, but it happens.

Confusion by function of concept is where two words mean something similar.  The obvious example would have to be that high school language class favourites: pets.  I could never remember whether un cobaye was French for a gerbil or a hamster, and still don't.  I don't suppose I would have recognised the difference between the two animals anyway, so the concepts in my brain were very similar: little fluffy critter.

But the two types of confusion by function can combine to create even bigger confusion.  While we do occassionally see completely nonsensical statements like the "drinking books" example, learners will quite often substitute a verb with a similar meaning.  So instead of saying "I didn't say anything" they might say "I didn't speak anything", of instead of saying "open the door" they might say "close the door", or even "close the window" (I had massive problems getting my open and close and door and window right in Welsh).

Confusion by co-occurrence
Finally we come to the one that is the most complicated and troublesome: confusion by co-occurrence.  Basically, the brain likes to associate things with each other.  We have salt and pepper and we have bread and butter, but we never have pepper and salt or butter and bread.  When things appear with each other a lot, they start to stick.  So when we're learning vocabulary, we can accidentally trick our brains into linking particular words more strongly than it should simply by having them appear next to each other a lot.  This is one of the biggest risks with word lists -- in a list, each word is forced to co-occur with the rest of the words on the list, particularly those directly before and after.  Reading the same list multiple times is almost guaranteed to create confusion by co-occurrence.

Multidimensional confusion
I've already demostrated how the two types of confusion by function combine to make a bigger problem, and of course all different forms of confusion can combine in this way.

I used to confuse my oats and my hazelnuts in Spanish.  Both are food and go together in my breakfast bowl for a sort of home-made muesli (confusion by function of concept).  They were stocked less than a metre apart in the local supermarket and I kept them next to each other in my kitchen cupboard (confusion by co-occurrence).  Oats is "avena" and hazelnuts are "avellanas" (confusion by form). Even when I came back to Scotland, I still couldn't get the word right.  Ironically, it is only when I started using this example that I became able to make the distinction correctly.

Lessons to be learnt
A learning technique cannot eliminate the confusions completely, but it must try to minimise them.
When evaluating a vocabulary learning strategy, or devising your own, look out for the 3 types of confusion.  When you're learning, be mindful of your mistakes and what they tell you about how you're learning.  If you can't think of a word, or keep getting it wrong, the chances are it's down to one of these types of confusion, and you should be able to refine or alter your technique to help you remember that word correctly, which over time should help you avoid making the confusion with new vocabulary in the future.