The use of FSI, a question of efficiency.

[Elexi]

On some of the discussion above, Dr Conti's current wordpress article is useful, if at a slight tangent:

https://gianfrancoconti.wordpress.com/2 ... riorities/

[blaurebell]

But drills do not have to be behaviorist. You can design your drills based on cognitive theories, and test those theories using inputs and outputs.

Of course behaviourism is simplistic, I'm not denying that. And of course you're right that its application ignores the real world situation of trying to avoid any kind of torture at all cost. That doesn't change that from a common sense perspective learning something new is essentially creating new habits. That's also why more recent cognitive theories like enactivism in the end returned to the notion of habit, because it's simply so important. Behaviourism gave the notion of "habit" a bad rep, but we should rethink that because it's damn powerful in practice. To be honest, I also don't really know what you mean by "cognitive theories" in general, since there are many different cognitive theories and most of them are still the matter of some debate if not even fist fights. Am I right in assuming that it's another premature attempt by some linguists to take a hotly debated piece of philosophy and turn it into some kind of law of nature? I was especially amused by taking the philosophy course on Theory of Action and the linguistics course Pragmatics at the same time. Basically in the linguistics course I had to learn everything by heart that we tore to pieces in the philosophy course. It gave me a somewhat sceptical outlook on linguistics in general I suppose.

As for substitution drills, I really don't know why anyone would want to replace any of that by teaching scripts. All that "meaningful content" as in scripts and relevant vocabulary just gives people the impression that they have a higher level of proficiency than they actually do. Asking "Where is the train station" doesn't help you much if you don't understand the reply. Great to keep people motivated by actually being able to say things, but pretty pointless all in all. I think I get a lot more out of "The girl throws the towel" "The girl burns the towel" and "The girl uses the towel" than "Where is the train station?" or "Do you take credit cards?".

Substitution drills practice a generalised structure behind all of it instead of teaching just phrases that will fail you once someone deviates from the script. That's the wrong kind of "meaningful" content. Meaningful is to teach people the structure of a language so that they can use them in whatever way that fits their situation if they have the right vocabulary. Ideally you use such drills at a stage when you still need to internalise the basic structures and then later move on to the more complicated exercises where each part of the sentence changes a lot. Especially with highly inflected languages having drills that require all parts of the sentences will just lead to unnecessary frustration, since you can and will make mistakes in every single structure of the sentence until you have mastered each grammatical part of it. Too much of a learning curve. So, a workbook like that might work with Spanish, but will have you tearing your hair out with Russian! I once used a Russian textbook that introduced one of the cases and one verb form at the same time. All subsequent exercises required both. That book almost went out of the window

[Stefan]

I feel like there are several discussions going on at once. OP raised the point about "just using the language" instead of using a course but from what I can tell, it was quickly disputed and you seem to have moved on to discuss drills. I'm not sure if this adds to the discussion so I deleted it twice but I find it interesting so here goes..

People tend to forget that FSI is a lot more than just the audio. FSI German is 30 hours (according to fsi-language-courses.net) but in real life, they expected you do spend 800-960 hours. Finnish? That's 1540-1760 hours. Chinese or Japanese? That's 3080-3520 hours of which half was spent in the target country. All numbers fetched from Lessons learned from fifty years of theory and practice in government language teaching (1999).

One of the most important things was to experience the language:

There is no substitute for simply spending time using the language. Segalowitz and his colleagues have pointed out how crucial to reading ability is the simple fact of doing a lot of reading (e.g., Favreau and Segalowitz 1982). Our experience at FSI indicates unequivocally that the amount of time spent in reading, listening to, and interacting in the language has a close relationship to the learner’s ability to use that language professionally.

This is what they have to say about drills:

Pattern practice—drill—is a technique that continues to be useful for FSI learners, when used in concert with the various communicative, experiential, and task-based approaches. It is valued not only at the early stages of our students’ learning, but at the more advanced as well, as review. In training programs with time-specified outcomes, such as at FSI, the automatization of basic grammatical structures and communicative routines is essential for efficient learning.

McLaughlin argued this point nearly twenty years ago. As he explains in a more recent work, “[t]he acquisition of a cognitive skill [results] from the automatization of routines or units of activity. Initially, the execution of these routines requires the allocation of large amounts of mental effort (controlled processing), but repeated performance of the activity leads to the availability of automatized routines in long term memory. The result of this process is that less and less effort is required for automated routines and the learner can devote more effort to acquiring other sub-skills that are not yet automated” (McLaughlin 1987:149). In order to perform higher order communicative skills—such as participating in social conversations (see lesson 10) and other such job-related uses of the target language—our students must produce spontaneously and accurately the relevant grammatical structures and routines of the language.

Interesting point about how they teach at FSI today (i.e. 1999):

In our half century of language education at FSI, we have moved from “teaching the textbook” to “helping the learner to learn,” from a strict diet of sentence-based pattern drills to a range of “communicative activities,” from using predominantly teacher-developed materials to a heavy emphasis on authentic or “found” materials and realia.

But with an interesting twist:

But for our colleagues and friends in the great research institutions, it seems to us that at least part of your tasks ought to be to seek answers to why some of the things we have described are the way they are. Why did learners learn almost as successfully in the early days of the long histories of FSI, the Defense Language Institute, Georgetown’s English Language Institute, and other comparable institutions, as they do today, despite the clear increases in the field’s understanding of teaching and learning? Do the curriculum and teaching techniques, in fact, not really matter?

I apologise for the wall of text.

[DaveBee]

Why did learners learn almost as successfully in the early days of the long histories of FSI, the Defense Language Institute, Georgetown’s English Language Institute, and other comparable institutions, as they do today, despite the clear increases in the field’s understanding of teaching and learning? Do the curriculum and teaching techniques, in fact, not really matter?

I'd say no.

The vital ingredient is, as always, perseverance. Sticking with the program, whichever program you've chosen, and achieving that end-goal.

If you stick with them, they all work.

[jsega]

A few words on behalf of FSI (this refers to FSI Spanish Basic Course).


1. It starts with a conversation, which you are not really supposed to memorise, but be able to repeat fluently, like in an Assimil course.

"The student should learn the basic dialogs by heart. If they are committed perfectly to rote memory, the
drills will go easily and rapidly. Roughly half of the estimated ten hours that are spent in class on each
unit should normally be devoted to the basic dialogs."
-location 7/698 in FSI Spanish Basic Volume 1 Student Text PDF, 1st paragraph

I'm not sure where you're getting that you need only be able to repeat it because that's not the instruction given in the manual. I mean one can do whatever he or she wants but we're talking about how it was designed to be used.

[reineke]

Apply conditions that help learning

"Learning a language involves both deliberate and incidental learning, and these both involve the conditions of repetition, retrieval (recall), varied meetings and varied use, using visuals, and deliberate attention. Good learning also involves avoiding interference which makes learning more difficult.

Is there a best method for learning a language? Unfortunately, the answer is no. Language learning can occur through all kinds of methods. What is most important is that good principles of learning are applied. You may have heard of methods like grammar-translation, aural-oral, the silent way, Suggestopaedia, the communicative approach, and TPR (Total Physical Response). These all have their supporters and attackers. Each method has its strengths and weaknesses and there is no research that shows the superiority of one method over another. It is fine to follow a particular method although it is not necessary. What is important is that there is a balance of learning opportunities across the four strands of meaning-focused input, meaning-focused output, language-focused learning, and fluency development. It is also important that the learning conditions of repetition, retrieval, varied meetings and varied use, deliberate attention, and deep processing occur. The most useful items to learn should get the most attention and you need access to material which is at the right level for you."

I don't think that Nation or anyone else has nailed it 100%, but this approach is broad, reasonable and offers plenty of maneuvering room for most language learners.

I have included a link to Paul Nation's free book in the Master link thread.

What you need to know to learn a foreign language

and

¿Qué necesitas saber para aprender un idioma extranjero?

http://www.victoria.ac.nz/lals/about/staff/paul-nation

[jsega]

I'm curious, has anyone tried just cutting up all the audio from FSI with audacity and creating Anki cards with it? I mean I assume listen/repeat audio courses would be easy to cut up with the Analyze > Sound/Silence Finder option in Audacity. So it wouldn't require that much set up considering how much time you're spending with the course otherwise.

I think I may try this....

[reineke]

I thought you were trying to find reasons to drop FSI or at least make it more palatable?

I feel like there are several discussions going on at once. OP raised the point about "just using the language" instead of using a course but from what I can tell, it was quickly disputed and you seem to have moved on to discuss drills. I'm not sure if this adds to the discussion so I deleted it twice but I find it interesting so here goes..

People tend to forget that FSI is a lot more than just the audio. FSI German is 30 hours (according to fsi-language-courses.net) but in real life, they expected you do spend 800-960 hours. Finnish? That's 1540-1760 hours. Chinese or Japanese? That's 3080-3520 hours of which half was spent in the target country. All numbers fetched from Lessons learned from fifty years of theory and practice in government language teaching (1999).

One of the most important things was to experience the language:

There is no substitute for simply spending time using the language. Segalowitz and his colleagues have pointed out how crucial to reading ability is the simple fact of doing a lot of reading (e.g., Favreau and Segalowitz 1982). Our experience at FSI indicates unequivocally that the amount of time spent in reading, listening to, and interacting in the language has a close relationship to the learner’s ability to use that language professionally.

This is what they have to say about drills:

Pattern practice—drill—is a technique that continues to be useful for FSI learners, when used in concert with the various communicative, experiential, and task-based approaches. It is valued not only at the early stages of our students’ learning, but at the more advanced as well, as review. In training programs with time-specified outcomes, such as at FSI, the automatization of basic grammatical structures and communicative routines is essential for efficient learning.

McLaughlin argued this point nearly twenty years ago. As he explains in a more recent work, “[t]he acquisition of a cognitive skill [results] from the automatization of routines or units of activity. Initially, the execution of these routines requires the allocation of large amounts of mental effort (controlled processing), but repeated performance of the activity leads to the availability of automatized routines in long term memory. The result of this process is that less and less effort is required for automated routines and the learner can devote more effort to acquiring other sub-skills that are not yet automated” (McLaughlin 1987:149). In order to perform higher order communicative skills—such as participating in social conversations (see lesson 10) and other such job-related uses of the target language—our students must produce spontaneously and accurately the relevant grammatical structures and routines of the language.

Interesting point about how they teach at FSI today (i.e. 1999):

In our half century of language education at FSI, we have moved from “teaching the textbook” to “helping the learner to learn,” from a strict diet of sentence-based pattern drills to a range of “communicative activities,” from using predominantly teacher-developed materials to a heavy emphasis on authentic or “found” materials and realia.

But with an interesting twist:

But for our colleagues and friends in the great research institutions, it seems to us that at least part of your tasks ought to be to seek answers to why some of the things we have described are the way they are. Why did learners learn almost as successfully in the early days of the long histories of FSI, the Defense Language Institute, Georgetown’s English Language Institute, and other comparable institutions, as they do today, despite the clear increases in the field’s understanding of teaching and learning? Do the curriculum and teaching techniques, in fact, not really matter?

I apologise for the wall of text.

It sounds like a paradox, right? It goes against everything you have been taught on your teacher training course, doesn’t it? How can grammar drills ever improve fluency? It’s learning grammar by rote! It’s what language teachers used to do in the 50’s!

In actual fact, as I shall argue below, verb conjugation drills can indeed enhance the ease and speed at which L2 learners produce spoken and written output. To understand why, let us look at how L2 output is produced.

How L2 output is produced

Every time an individual produces linguistic output, they will have to translate the idea / message they are trying to convey (or proposition, as cognitive psychologist call it) into language. This process happens at very high speed with native speakers (Anderson, 2000). However, with novice target language speakers the process is much slower, especially when they are producing complex sentences which pose a heavy cognitive load on Working Memory. Imagine an intermediate student of average ability wanted to convey the following, in French:

Yesterday we went to the cinema with them (feminine). The film was great but the cinema was packed.

First, the student will have to retrieve all the vocabulary they need from Long-Term Memory. At this point Working Memory, that can only contain 5 to 9 digits at any one time for around 15-to-30 seconds (without rehearsal), will be already stretched in terms of storage capacity. Whilst holding the words in Working Memory, the brain has to ensure that each lexical item is arranged in the correct syntactic order and that the rules of tense and agreement are applied correctly. This requires a number of cognitive operations some of which involve sub-operations (e.g. ‘they arrived’ in French require the perfect tense; however it is a verb requiring the verb ‘to be’ as an auxiliary; also ‘they’ being plural I will have to add ‘-s’ to the ending of the verb).

If a student has not automatized verb formation and the application of the various rules of tense and agreement involved, the process will be extremely cumbersome and may lead to error. For instance, a typical intermediate student of average ability will master most of the vocabulary in the sentence above. However, they may struggle with the translation of ‘we went’ (being a verb requiring the auxiliary Etre in the Perfect tense); they may be undecided as to whether to use eux, ils or se for ‘them’; moreover, they may have problems deciding if ‘Was’ should be translated using the imperfect or the perfect tense; etc. Each and every decision has to be taken whilst simultaneously the brain has to rehearse every single vocabulary item in Working Memory. This is an ominous task for an intermediate learner and the slightest interference can cause Working Memory loss, leading to processing inefficiency errors.

It is easy to see how, in this and other linguistic contexts, the automatization of verb formation would contribute to speed up the process of output production; one less cognitive operation to worry about.

The objection often made by the critics of verb drills is that the students do not learn the target verbs in context. However, such criticism stems from a misunderstanding of the role of these activities. Verb conjugation drills do not aim at the acquisition of verbs; their goal is the automatization of one aspect of verb acquisition: verb formation. Hence, they are only a small but very important cog in a very complex mechanism. Unlike in behaviourist L2 pedagogy models, verb conjugation drills should not dominate our lessons. Not at all. They only make sense as a means to support the development of communicative competence. Their merit lies in focusing Working Memory’s attentional systems solely on verb morphology in a ‘stress-free’ environment, where there is no communicative pressure nor high levels of cognitive challenge. Hence, they should be used with discernment..."

https://gianfrancoconti.wordpress.com/2 ... n-fluency/

[jsega]

I thought you were trying to find reasons to drop FSI or at least make it more palatable?

Using it with Anki would potentially do that for me.

My questioning of FSI materials here has been an attempt at finding some "truths" (as close as one can get of course, it's ongoing) that at least make sense to me. This is regarding FSI specifically, but also language learning in general where it applies. This thread has given me much to think about.

[reineke]

Let me put a bee in your bonnet:

Irregular before regular – maximizing explicit grammar instruction by inverting traditional instructional sequence.

In most coursebooks and schemes of work adopted by UK MFL education providers, the exceptions to a given grammar structure are usually taught after the dominant rule governing that structure has been imparted. In the present post I argue that in many cases inverting the teaching sequence may have a more beneficial impact on acquisition. The rationale for this approach is rooted in the way the brain forms and revise the L2 Interlanguage system.

When a learner is taught a grammar rule, the brain creates a cognitive ‘structure’ that s/he will consolidate through much receptive exposure and production. As already discussed in my post on how L2 grammar ‘rules’ are acquired, when a grammar structure is in the process of being automatised, the brain tends to be extremely circumspect in accepting as ‘correct’ – and consequently ‘learnable’ – any use of that structure which does not match the declarative knowledge (or mental rule representation) stored in Long-Term Memory which refers to it. This is particularly true of the final stage in L2 grammar structure acquisition – Andersons’s (2000) Strengthening process. During this stage, the brain needs to be particularly impervious to any alteration to the rule system referring to that structure in order for that system to be stable and avoid encoding ambiguity. For any successful cognitive restructuring of an existing grammar rule to occur two conditions must be met:

The grammar rule one wants to restructure must be fully acquired for any exception to it to be incorporated; only then will the brain be more likely to ‘see’ the exception to that rule as a separate subsystem which does not pose any ‘threats’ to the dominant rule system;
The exception to the rule must be processed by the brain numerous times in salient and meaningful contexts; this entails that exceptions to a given rule which do not occur frequently in the language processed in classroom or out-of-the-classroom L2-based activities are less likely to be internalized as they will be ‘masked’ so to speak by the dominant rule.
Let us look at an example: teaching the Passé Composé in French. Coursebooks normally begin with the verbs forming this tense with ‘Avoir’ and after a few lessons move on to the ‘Etre’ verbs. Whilst some of the more able and focused learners can cope with this, in my experience many learners cannot. Very often, teachers may believe students have acquired mastery over the two sets of rules based on their learners’ ability to perform successfully at cloze tasks or other mechanical grammar activities. However, in less structured activities (e.g. spontaneous speech) errors in this area will be usually rife.

Issues in acquiring the exception to the dominant Passé Composé rule are exacerbated by the fact that very few of the verbs requiring Etre are high frequency verbs, hence the students do not usually receive great exposure to them when processing classroom or naturalistic French input. This will make restructuring of the ‘have + past participle’ rule more difficult.

In this case, teaching the ‘Etre’ verbs before the ‘Avoir’ ones is a more effective strategy; once acquired the exception (Etre + past participle) through extensive modelling and practice, the learners will find it easier to learn the dominant rule due to the very frequent occurrence of ‘Avoir Verbs’ in classroom or naturalistic target language input.

The same applies to any other grammar structure where the exceptions to the rule do not occur very frequently in the instructional or naturalistic target language input. Think about irregular past participle such as ‘reçu’, ‘vecu’, ‘su’, etc. which are notoriousy less easy for students to acquire than ‘pris’ or ‘fait’, for instance.

https://gianfrancoconti.wordpress.com/2 ... sequences/