|
 |
"Myths" of foreign language learning | ||
Tags: Learning Strategy | ||
Share with:
| ||
 Language Learning Forum : Learning Techniques, Methods & Strategies |  | |
| s_allard Triglot Senior Member Canada Joined 5431 days ago 2704 posts - 5425 votes  Speaks: French*, English, Spanish Studies: Polish | Message 41 of 72 03 May 2010 at 8:10pm | IP Logged | |||||
I could not agree more. Nowhere do I state that child learning is the model that adults should follow. I just pointed out that preliterate children AND illiterate adults do not use the writing system to derive rules for gender agreement. And, as I point out, uneducated speakers can make overgeneralizations that are corrected by the education system. My argument has nothing to do with how children learn; it has to do with an efficient learning strategy. My hypothesis, simply put, is that to master the French grammatical gender agreement system for spoken fluency, some, certain or many words (but not all) are best learned within context. If this is what children use, that's wonderful. Why not use it? On the other hand, I think that the adult strategy of all words ending in -e are feminine but for a few exceptions leads directly to that all too common observation that most adults find French very difficult to learn. They're just making things hard for themselves. 1 person has voted this message useful | ||||||
 Iversen Super Polyglot Moderator Denmark berejst.dk Joined 6704 days ago 9078 posts - 16473 votes Speaks: Danish*, French, English, German, Italian, Spanish, Portuguese, Dutch, Swedish, Esperanto, Romanian, Catalan Studies: Afrikaans, Greek, Norwegian, Russian, Serbian, Icelandic, Latin, Irish, Lowland Scots, Indonesian, Polish, Croatian Personal Language Map | Message 42 of 72 06 May 2010 at 11:21am | IP Logged | |||||
|
As far as I can see the latter part of this thread can be boiled down to the following questions: 1) do you remember items better if they are put into a 'lifelike' context or if they are boiled down to the bare essentials? 2) When do you give up looking for a rule and start learning single items instead?
In Spanish or Italian it is quite economical to associate -a with femininum and -o with masculinum. In Modern French there isn't such a simple rule, so either you have to learn single items or you have to use rules based on knowledge of other languages. And of course applying any rule will slow you down, so in the long run you will have to store single items in your memory, but in cases where the rules are fragmentary and complicated you can just as well start learning simple entities from the start. In French the relevant entity is a noun plus something that shows its gender. For me the optimal auxiliary item is the article if it is le or la, but I do see a raison to add a suitable adjective if it is l' . The situation I want to avoid is spending time on inventing or memorizing contexts that doesn't show strictly necessary information. Therefore I prefer using the same formulaic 'pseudo contexts' wherever possible, - articles, "quelque chose" (qc), "quelqu'un" (qn) etc. That's the way my memory functions best. But people who need more colorful and memorable and variated contexts should of course use those. However I fail to see how a context consisting of a long sentence can ever be more memorable than a context consisting of maybe 2 or five words. I simply don't understand how burying a target word in a long sentence with lots of irrelevant words can be a help, whereas adding an extra word to a short phrase to make it more memorable can't be a problem. I just don't feel the need. Learning single items consisting of 2-3 words (for instance verbs with their typical prepositions) is not a luxury, but a necessity. The question is whether you can find materials to do it in a systematic way. Good dictionaries will provide much of the material, and grammars also contain useful lists. And again there are two strategies: 1) try to standardize and shorten the phrases to concentrate the information, 2) spice it up with a superfluous and incidental, but memorable context. Nobody says that you always have to use one of these two methods, but different types of learners will probably find that the one or the other works better for them. Edited by Iversen on 06 May 2010 at 11:25am 1 person has voted this message useful | ||||||
| Cainntear Pentaglot Senior Member Scotland linguafrankly.blogsp Joined 6012 days ago 4399 posts - 7687 votes Speaks: Lowland Scots, English*, French, Spanish, Scottish Gaelic Studies: Catalan, Italian, German, Irish, Welsh | Message 43 of 72 06 May 2010 at 7:13pm | IP Logged | |||||
Everything I've read suggests that even where the patterns are very subtle or even obscure, the internal model of the native speaker still implicitly recognises these patterns, and if that. As I was saying earlier, I'm very interested in knowing the end-state of the native learner, and applying that knowledge to create a learning pathway that respects the internal model and presents the material in such a way as to support -- force, even -- the building of an analogous model of the language in the new learner's mind. That's why I'm not happy with the idea that any rule is ever too complicated to teach. If you say it is too difficult, in effect your are declaring the language unlearnable. Edited by Cainntear on 06 May 2010 at 7:15pm 1 person has voted this message useful | ||||||
| Iversen Super Polyglot Moderator Denmark berejst.dk Joined 6704 days ago 9078 posts - 16473 votes Speaks: Danish*, French, English, German, Italian, Spanish, Portuguese, Dutch, Swedish, Esperanto, Romanian, Catalan Studies: Afrikaans, Greek, Norwegian, Russian, Serbian, Icelandic, Latin, Irish, Lowland Scots, Indonesian, Polish, Croatian Personal Language Map | Message 44 of 72 06 May 2010 at 8:52pm | IP Logged | |||||
|
Any language contains myriads of words which you learn one by one ... so learning by items rather than by rules not only possible, but actually an essential part of language learning. Besides no language is entirely ruleless - the use of syntactical patterns is also an important kind of regularity.
However the point I made was that sometimes the rules are so complicated and filled with exceptions that it is easier just to start learning chunks of a few words like you learn words, i.e. itemized. And when you have learned enough items you may develop a 'feel' for the correct gender or case or whatever, even though the rules still would be so complicated that you would have problems trying to formulate them in words. 1 person has voted this message useful | ||||||
| Cainntear Pentaglot Senior Member Scotland linguafrankly.blogsp Joined 6012 days ago 4399 posts - 7687 votes Speaks: Lowland Scots, English*, French, Spanish, Scottish Gaelic Studies: Catalan, Italian, German, Irish, Welsh | Message 45 of 72 06 May 2010 at 10:46pm | IP Logged | |||||
A lot of academics talk of language learning as an application of Occam's razor -- learners develop a "theory" of what the target language is like based on the available data and refines the theory whenever they receive new data that doesn't match the theory. It's a very good model of how the process unfolds, and you can see this happening in infant language acquisition or in adult language learning. I spent a long time saying "los otros" before it clicked that everyone else was saying "los demás" to me. How's this relevant to the debate at hand? Most people fall into one of two schools of thought: 1) Learn complete, correct, explicit rules. 2) Learn by induction from examples. Or in terms of Occam's Razor: 1) Supply the correct theory to the student. 2) The student must construct his own theory. There is a middle ground that these two standpoints deny: Supply a partially-formed theory and allow the student to refine it based on induction from examples. This brings me back to "los otros". My initial theory was "other"="otro", which was what I'd been taught. Input didn't suggest I should throw out the rule, but that instead I needed to refine it, which is what I went on to do. And this is where the two camps tie themselves in knots, and in trying to avoid the centre ground, they go round the outside of the circle and meet. Group 1) The only complete, consistent, correct theory is the entire language, and you cannot teach the language in one go, so they teach subsets of the theory and limit it by saying "this only applies to the context in the given examples", which makes it closer to induction from context than they would like to admit. Group 2) Fully natural language with its wide variety overloads the brain of the learner, so the examples used for learning by induction are generally constrained to a single language point -- in identifying masterial that demonstrates the point, the teacher or course designer effectively supplies the theory in a slightly roundabout, cryptic way. And even though the teachers keep trying to avoid the middle ground, students still keep ploughing on through it -- you just can't stop students generalising. No matter how much you warn them that generalising based on translation is correct, they all do it. Some more than others, but everyone does it at some point. Why fight the urge to generalise? Why decry it as a weakness rather than harness it as a strength? Whenever there's a pattern, let's deal with the most general case and work the exceptions into that. Because, heck... if you're going to build your own theory out of the data you are going to overgeneralise to start off with. Starting off with an overgeneralised rule saves you carrying out that first step yourself. 1 person has voted this message useful | ||||||
| s_allard Triglot Senior Member Canada Joined 5431 days ago 2704 posts - 5425 votes Speaks: French*, English, Spanish Studies: Polish | Message 46 of 72 08 May 2010 at 8:22am | IP Logged | |||||
|
I totally agree with Iversen who, despite the differences we may have on various issues, seems to have understood what I was getting at. On the other hand, I would like to suggest that Cainntear is complicating something that does not have to be complicated at all.
What are we trying to accomplish? It seems to me that the goal here is for the learner to able to accurately and fluently produce spoken utterances with the correct syntax in terms of grammatical gender agreement. So, we are trying to devise an efficient learning strategy that will give us the best results. There would seem to be a two part process at work here. First the head noun must be identified as masculine or feminine, And then all the concomitant syntactic markers must be activated. Consider the following utterances: a) C'est un grand désastre naturel qui doit être analysé. b) C'est une grande catastrophe naturelle qui doit être analysée. We see how the orthography and of course the pronunciation of French reflect the fact that désastre is masculine and catastrophe is feminine. Let me stress here that the -e that is omnipresent as a written marker of the feminine adjective is not pronounced by itself. So, "naturel" and "naturelle" are pronounced the same. Here we see how spoken French gender marking differs so much from that of Spanish. And this is also how written French diverges from spoken French. In addition, this also explains why a contest called "dictée" can exist in French and not at all in Spanish. Native speakers never seem to have a problem producing the kinds of phrases given above. (There are some issues here with unschooled native speakers, but I would like to avoid that complication for the time being). On the other hand, English-speakers notoriously get tripped up and often make egregious mistakes because this system is so unlike English. Spanish speakers have a relatively easy time with this because Spanish syntax is quite similar. Again, I would like to stress that the problem is not only how to identify the gender of the head noun. It is also how to quickly make the necessary morphological adjustments. One could rightly say that if the user has a good command of gender syntax they simply drop the noun into the right sequence and on they go. However, learners of French have two problems. First, how to identify the right gender of the noun? We know that there are various subgroups where the gender is indicated by the word ending (not the vowel but the suffix). So if we see a noun ending in -euse or -tion, we know it is feminine. No problems here. We also know that trying to devise a rule that states that all words ending in -e are feminine with certain exceptions is extremely inefficient--if not outright counterproductive-- because of the very large number of exceptions. Again, let me point out that since the -e is not pronounced separately in French word endings, it is not useful as a audible marker. So, gageur and gageure, pair and paire, beur and beurre are pronounced exactly the same. This is very important because it is evident that preliterate children and illiterate adults can use the system perfectly without any idea of how words are spelled. What we do see, however, is English-speakers somehow trying to put some order in this system and make this rule of words ending in -e being feminine with a long list of exceptions. The end result is usually quite catastrophic because they end up with the wrong syntax. Again, let me stress that only one part of the problem is whether one should say UN or UNE cadastre and Un or UNE désasstre. The second problems is how to get all the syntax markers right orthographically and phonetically. A common phenomenon, for example, is to get the gender assignment right and the syntax wrong. I'm not sure how Occam's Razor applies in all of this. What I do know is that native speakers get it right nearly all the time and English-speaking learners, even at quite advanced levels, rarely get it completely right. Now, I would point out here that in my observation, learners who acquire the language spontaneously through extensive interaction with native speakers tend to have a good mastery of the system. The fundamental issue here is how to best learn the system. What I have suggested here, and I what I teach, is that for those words that are problematic--let's say the -e words, but there are others--it is best learn them in a meaningful context WITH appropriate syntax markers. As Iversen states, there is no need for long sentences. So, I would learn "une excellente idée" and "un beau musée", "un nouveau désastre financier" and "une nouvelle catastrophe financière" to highlight the contrasting syntax markers. The advantage of this approach, as I have stated a number of times, is that one learns the two facets of the system properly from the very start. I suggest that this is what native speakers do. Children learn "mon père", "ma mère", "mon frère" without any idea that these words end in -e, which is totally irrelevant. I'm not suggesting that one should not learn by rules or use a totally inductive approach to grammar. Neither am I suggesting that adults learn exactly like children. (Just on this last point, I don't, on the other hand, believe that the learning process is necessarily totally different in children and adults, but that is another debate). Why do native speakers rarely hesitate and usually always get the system right whereas English-speaking learners often stumble and get it wrong? It is my view that these learners are not able to spontaneously and accurately make the right gender assignment and the right syntax adjustments because they have learned them separately and, more importantly, they are confusingly using some incorrectly overgeneralized rule. Or, to put it more simply, they are all mixed up. Edited by s_allard on 08 May 2010 at 9:26am 1 person has voted this message useful | ||||||
| s_allard Triglot Senior Member Canada Joined 5431 days ago 2704 posts - 5425 votes Speaks: French*, English, Spanish Studies: Polish | Message 47 of 72 09 May 2010 at 6:12am | IP Logged | |||||
As I was trying to understand make sense of this debate, it seems to me that the crux of the matter is this idea of overgeneralizing by building a theory based on data on hand. If I understand this correctly, this could mean building the theory that all words in French ending in -e are feminine with certain exceptions. That would be the most general case and then there would be exceptions. I argue that such a theory leads learners to a dead end because it is based on an incorrect analysis of the data when the very data does not support this theory at all. I find it quite intriguing how someone could think that in French nouns ending with -e are feminine when even a cursory examination of the data demonstrates otherwise. On the other hand, I can see that beginners may latch on to this sort of idea because they have learned lists of words and just look at the spelling. Similarly, they may come to the conclusion that word ending in -r are masculine based on a few examples. In my opinion, this kind of rule making is flawed because it is based on imperfect data and probably comes from learning outside of context. This exactly why learners make certain mistakes in their target language. If you learn that "los otros" is the equivalent of "the others", that is to say with no reference to usage in context, then of course the tendency is to use "los otros" wherever "the others" is used in English. This is not overgeneralization based on the examination of the various usages of "los otros", as native speakers would do, but on the overgeneralization of a partial translation. This happens all the time when we learn a new language. This is why we get some funny translations made by people who just look up words in a dictionary or use translation software. For example, "to go out" can be translated by "sortir" in French. So, we can attempt to translate "Don't let the fire go out" by "Ne laissez pas le feu sortir". The problem with learning distinctions such as los demas and los otros outside of context (both mean "the others") is that one does not learn how to distinguish the two. And that is the problem. Other classic examples of these kinds of subtle distinctions would be ser and estar in Spanish and connaître and savoir in French. In Spanish the distinction between ser and estar is often taught (the explicit rule) in terms of permanent or durable characteristics (ser) versus temporary or non-permanent characteristics. Not a bad rule. But why do some many people have a problem? The complication lies in the application of the rule. Now, I'm well aware that there are other little rules to explain certain usages. And there are situations where one can use either ser or estar. And let's not forger regional differences in usage. Again, here my suggestion is to certainly use whatever rules that can one can apply and then learn by context for those particularly difficult or subtle usages. Actually, this is where collocations are important. Certain words go well together and should be learned as a set regardless of what kind of analysis one could make. 1 person has voted this message useful | ||||||
| Cainntear Pentaglot Senior Member Scotland linguafrankly.blogsp Joined 6012 days ago 4399 posts - 7687 votes Speaks: Lowland Scots, English*, French, Spanish, Scottish Gaelic Studies: Catalan, Italian, German, Irish, Welsh | Message 48 of 72 09 May 2010 at 12:46pm | IP Logged | |||||
I don't, as an adult, have time to study as much data as an infant does before I start speaking. My experience of context will naturally limit my knowledge of individual terms. Looking at it from this perspective, I overused "otro" in "los otros" not because I had learned "otro" wrong, but because I had not yet learned "los demás". I'd argue that this was not really an "error", but simply a gap in my knowledge. Overgeneralising allowed me to attempt to explore a wider area of language. I was understood when I said it, but I then started noticing the correct (untaught) form and I picked it up. Why Occam's Razor? Because this is a question of how to use limited data to maximum effect. The background of Occam's Razor comes from the philosophical standpoint of "nothing can be taken to be true that the data does not explicitly show", which is exactly equivalent to only using language you have explicitly encountered in the past. The argument was that theories and formulas necessarily rely on assumption and generalisation and are therefore inherently flawed. But we need to make generalisations or we limit ourselves greatly, so let's take the simplest explanation that matches the data and run with it until it is disproven. OK, so the grammarians know (almost) the full "theory" or "formula" of a language, but it's impossible to program that into a learner's brain, so instead, why don't we walk them through a controlled learning path, using specially selected input and/or output exercises to build a simplified theory and then continually refine it? This is how the brain likes to work anyway, and if the data is selected and paced appropriately, we can harness the natural process and make it more efficient. This is exactly what Krashen means when he talks about "n+1 input" -- trying to learn more than one thing at a time is a source of distraction, confusion and ultimately frustration; learning only one thing at a time is just more efficient. (I dislike Krashen not because of his starting point, but because of his conclusions, which are illogical, inconsistent and not supported by the evidence. In the case of N+1, the rest of his framework doesn't support a genuine "+1", so he takes "+15" and claims that 15 is 1...) Edit:
I'm sorry, that's an invalid comparison. My "los otros"/"los demás" thing was not a word-level translation error, it was a conceptual error. In my head, there was only one concept of "other"/"others", but I can assure you that I have very different concepts of "go out (=exit)" and "go out (for a fire)". Yes, some people will make the "los otros"/"los demás" error from a dictionary, at the word translation level, but even if the end result appears the same, the underlying processes are fundamentally different. Edited by Cainntear on 09 May 2010 at 12:54pm 1 person has voted this message useful |
|  | 
Printable version
|
| Copyright 2024 FX Micheloud - All rights reserved No part of this website may be copied by any means without my written authorization. |
