|
 |
Defining the Speaking Threshold Kernel | ||
Tags: Number of words | Speaking | ||
Share with:
| ||
 Language Learning Forum : Learning Techniques, Methods & Strategies |  | |
| robarb Nonaglot Senior Member United States languagenpluson Joined 4855 days ago 361 posts - 921 votes  Speaks: Portuguese, English*, German, Italian, Spanish, Dutch, Swedish, Esperanto, French Studies: Mandarin, Danish, Russian, Norwegian, Cantonese, Japanese, Korean, Polish, Greek, Latin, Nepali, Modern Hebrew | Message 9 of 28 12 October 2014 at 2:57am | IP Logged | |||||||||||||||||
|
Sorry for the double post. Let's try and see what it's like to build sentences using the kernel.
Example: Speaker A would like to play basketball with Speaker B. A: Veux-tu faire du sport? [gestures a basketball shooting motion] (no jouer and no basket) B: Quand? A ce... [I can't find a way to say "now" without maintenant, heure, moment, or temps, but the meaning should be clear when B hesitates here. A understands.] A: Dans quinze minutes, ça va? B: Ça va. Je suis en vacances, alors j'ai pas de travail. A: Ah non! Je n'ai pas mes chaussures! (careful, no oublier and no interjections of displeasure other than 'non') B: Alors on peut le faire le dimanche. Ce jour il fait trop froid quand même. (awkward sentences without aujourd'hui and demain, but the meaning is clear.) Translation: A: Do you want to play sports? [gestures basketball] B: When? This... [Couldn't find a way to say 'now,' but context makes it clear.] A: In fifteen minutes, OK? B: Sure. I'm on vacation, so I don't have work to do. A: Oh no! I don't have my shoes! B: Then we can do it on Sunday. On this day it is too cold anyway. [Cannot say 'today' or 'tomorrow'] I would suggest aujourd'hui, hier, and demain as core words instead of some of the lower-frequency nouns on the list, and probably also temps, so maybe we could do even better, but to keep the same list size we'd have to take off something that might be useful in other situations. When I'm a beginner in a language, I basically talk like this with the addition of grammatical and pronunciation errors. It's usually enough to get the message across, but for most sentences the first way I thought of was not allowed. Only in one case did we have to rely on gesture, and in one case we had to rely on the other speaker making a reasonable inference about what we were trying to say. I'm not trying to argue for or against the kernel here; I think this is a fair assessment of the level of expressivity you get using the wordlist s_allard posted. Edited by robarb on 12 October 2014 at 3:01am 4 persons have voted this message useful | ||||||||||||||||||
| s_allard Triglot Senior Member Canada Joined 5226 days ago 2704 posts - 5425 votes Speaks: French*, English, Spanish Studies: Polish | Message 10 of 28 13 October 2014 at 5:10pm | IP Logged | |||||||||||||||||
|
I've been busy lately, so I haven't had much time to respond to various posts. I'll try to address various points,
primarily those raised by robarb. 1. Why around 300 words and not some smaller number? The number is arbitrary of course, and one could make complete sentences with much fewer words. It's just that 300 contains a large enough number of key verbs, functional words such as pronouns, adverbs and connecting words with enough nouns to cover a variety of situations. One could argue for 250 or 500. I think 300 is a good threshold. 2. Why the focus on speaking and not on listening or reading? This does not say that we are not interested in listening. Of course, one has to listen to interact with native speakers. This says that the purpose of the list is to define a minimum set of productive vocabulary. I would think that speakers have a much larger receptive vocabulary, but the perspective here is from speaking. 3. Why search for transcripts of authentic conversations and not simply take a frequency list or for example or just use the more common transcripts of news programs? The answer is simply that we are not looking only at words. We are looking at the use of words in real interactions. This is what I call linguistic knowledge. Instead of looking at artificial dialogs or trying to imagine conversations, we look at the real thing. And we see all kinds of interesting stuff such as considerable repetitions, all sorts of conversation markers, contractions and abbreviations, incomplete sentences, non-standard constructions. 4. How to demonstrate what can be done with such a list? Rather than try to invent some dodgy conversations for the purpose of pointing out missing items, let's take some real conversations from the France Bienvenue web site. A : Alors Philippe, vous vous êtes levé à quelle heure ce matin ? P : Donc ce matin, je me suis levé à six heures et quart, pour commencer à huit heures. A : Oui. Donc ça veut dire quoi, là ? Vous partez à quelle heure de chez vous ? P : Je pars vers 6h55 – 7h de chez moi. A : Donc six heures et quart – 6:55, ça vous suffit pour vous préparer ? Qu’est-ce que vous faites alors ? P : Donc… A : Le réveil sonne. P : Le réveil sonne. Je prends un petit déjeuner. A : Vous vous levez tout de suite ? Ou vous restez cinq minutes, encore ? P : Tout de suite, parce que après, sinon… la flemme (1) de se lever. A : Oui. Vous rappuyez sur le réveil et vous repartez pour un tour (2)? P : C’est ça. Donc je prends mon petit déjeuner. A : Qu’est-ce qu’il y a au petit déjeuner ? C’est long ou pas ? I didn't check, but I think around two or three words in this dialogue are not in the 301-word list. Big deal. The really important thing here is: Are you the speaker able to construct all these or similar sentences spontaneously, pronounce them decently and engage the interlocutor? Here is another example from France Bienvenue: A : Donc Mélissa, vous nous avez raconté, là, toutes vos conditions de travail (1), tout ça et… bon, c’est quand même dur ! M : Oui. A : Donc vous gagnez de l’argent mais vous avez un projet ? M : Pas spé[cialement]… Bon là, oui, j’ai un projet pour… pour cet été : je pars aux Etats-Unis un mois. Donc c’est vrai que j’ai… A : Ouah ! Un mois ! M : Oui. Donc c’est vrai que j’ai travaillé… Ça fait un an que je travaille pour économiser. Donc ma famille me paye (2) une partie mais je finance aussi une partie du voyage. A : Oui. M : Donc je vais partir avec ma sœur et mon cousin. Donc… A : Là, ça va ? Vous avez tout ce qu’il faut ? M : Oui ! Là, il y a pas de souci. Depuis un an… A : Les billets sont payés, et tout… ? M : Tout est organisé. On sait déjà ce qu’on va faire. A : Alors, vous allez où ? M : Donc je fais… Donc je pars de Marseille. Je… J’ai une… J’ai un vol à… Je sais plus où c’est… En Allemagne, mais je sais plus la ville. Again, there is certainly a small number of words such as Marseille, Allemagne and a couple others that are not in my list. But look at everything else. This is not particularly complicated French. This is not Harry Potter in French. It's simply two people talking about an upcoming vacation. In terms of French grammar, there is nothing here that is not implicitly covered in the kernel. But how many learners can speak like this? 5. There are many words that are not on this list. At least 60,000. People can quibble about certain items. I would make some changes myself. I see, for example, that parmi is not there. Why not add a few more key words and make the list 315? I have no objections. This is why I have explicitly used the word 'threshold". 6. As I said in the OP I (we) have come to believe that the problem is not really words, it's the underlying knowledge. To try to equate levels of expressivity with just numbers of words is totally false. One can make some very sophisticated constructions with the words in this list using, for example, pronominal verb forms or the pronoun en. What we see in learners is poor control of: 1) phonetics, 2) word order, 3) morphology and syntax, 4) idioms and 5) meanings. The number of words learned doesn't deal with these major problems. The simple idea of the kernel is that it is a road map of what a learner should concentrate on for purposes of speaking. Anybody who has a good command of these 301 words knows a hell of a lot of French. And then they should build on that solid foundation. Edited by s_allard on 13 October 2014 at 6:31pm 1 person has voted this message useful | ||||||||||||||||||
| s_allard Triglot Senior Member Canada Joined 5226 days ago 2704 posts - 5425 votes Speaks: French*, English, Spanish Studies: Polish | Message 11 of 28 13 October 2014 at 5:26pm | IP Logged | |||||||||||||||||
This is a good point and the object of some thinking on my part. Should the nouns start with basic things such as common concepts of space, time, parts of the body, common objects, etc? That's part of the problem of defining the nouns that make up such an open category. The thinking behind this original list was to select some well- known objects or concepts that people are likely to encounter. One could argue that not everybody will need la baguette and that another word would be more useful. No problem, remove la baguette and insert the other word, 1 person has voted this message useful | ||||||||||||||||||
| robarb Nonaglot Senior Member United States languagenpluson Joined 4855 days ago 361 posts - 921 votes Speaks: Portuguese, English*, German, Italian, Spanish, Dutch, Swedish, Esperanto, French Studies: Mandarin, Danish, Russian, Norwegian, Cantonese, Japanese, Korean, Polish, Greek, Latin, Nepali, Modern Hebrew | Message 12 of 28 13 October 2014 at 6:29pm | IP Logged | |||||||||||||||||
What you call linguistic knowledge is why one must observe the language in use before being able to speak it properly. It's still not clear what this has to do with the kernel. Do you want to observe language in action in order to select what words should be in the kernel? If so, what are the criteria? Or do you want to observe language in action in order to document the multiple uses of the words in the kernel? If so, how do you use this to support learning?
Wow. The purpose was not to "demonstrate missing items." It was to concretely demonstrate that you can build meaningful communicative sentences using only 301 words! I just honestly pointed out a couple times when I felt it was limiting.
This is essentially the same conclusion I reached after using the words in the kernel to construct a "dodgy" dialogue. I achieved my communication goals despite one essential vocabulary gap and a couple nonessential ones. I was not arguing against the kernel, and I was most certainly not arguing that it is too limited! In fact I provided a direct, material demonstration of the claim that you can speak using the words in the kernel. Why argue against that? It was your original point. I don't believe you've addressed the question at all of why posting transcribed natural French conversations is relevant at all to evaluating or even illustrating the kernel idea.
I was building sentences using the words you posted and my own underlying knowledge of French. I have no difficulty producing pronominal verb forms or en- constructions. Not a lot of complicated constructions were necessary for the communication I demonstrated, but so what? If you're a native speaker, maybe you can build sentences that sound a bit more natural. But mine are perfectly comprehensible and probably more representative of what a learner with good but not yet native-like linguistic knowledge would be able to do. In contrast, I doubt you could use that vocabulary list to form a compact representation of the concepts "basketball" or "now." But I don't think you are claiming that all such things are covered (though they do seem to be with something like 1000 words, allowing for some paraphrasing). Also, you can't tell me now that a L2 user of French using the kernel to build French sentences is a "dodgy" idea. In the OP you said:
Building complete sentences. I believe I've shown that you can and it works.
Let me see if I am interpreting this correctly. Are you recommending that, for the purposes of speaking, learners should make sure they are comfortable using all 301 words in the kernel, and then they should concentrate primarily on linguistic knowledge (using whatever their favorite methods are), only adding further words to their productive vocabulary as they come naturally or as needed? Edited by robarb on 13 October 2014 at 6:37pm 2 persons have voted this message useful | ||||||||||||||||||
| Serpent Octoglot Senior Member Russian Federation serpent-849.livejour Joined 6393 days ago 9753 posts - 15779 votes 4 sounds Speaks: Russian*, English, FinnishC1, Latin, German, Italian, Spanish, Portuguese Studies: Danish, Romanian, Polish, Belarusian, Ukrainian, Croatian, Slovenian, Catalan, Czech, Galician, Dutch, Swedish | Message 13 of 28 13 October 2014 at 6:53pm | IP Logged | |||||||||||||||||
|
Anybody who knows 3000 words superficially (+some grammar) also knows a hell lot of French. I'm not saying they know more than the learner with a strong core of 300 words, the question is whether these strategies are equally viable, and whether either of them is preferable for the majority. I personally think that the usefulness of your approach is limited to multilingual environments, related languages and false beginners (say, any two of these factors). For most other learners it's simply unrealistic and/or pointless to polish their knowledge of the 300 words if they can hardly keep up with a normal conversation.
1 person has voted this message useful | ||||||||||||||||||
| s_allard Triglot Senior Member Canada Joined 5226 days ago 2704 posts - 5425 votes Speaks: French*, English, Spanish Studies: Polish | Message 14 of 28 13 October 2014 at 7:49pm | IP Logged | |||||||||||||||||
OK, maybe the word "dodgy" was a bit too strong for what I perceived as a contrived conversation that seemed unnatural to me. I take back my claim that this was done for purposes of pointing out missing items. That said, I really don't see the point of making up examples where there is so much real stuff. That's exactly why I like to work with transcriptions of real conversations. Rather than trying to invent things, I like to work with transcriptions of natural French conversations for two reasons. First, they illustrate how the language is really used. Everything is there, including, as I pointed out, all the features that are typical of spoken language plus the recordings with the phonetic component. Second, the transcripts are very useful for purposes of validation. I map the kernel list with its underlying linguistic knowledge against the transcripts and I see a good fit. I see lots of holes for content nouns, of course, but that's not a big problem. For example, in the following sequence, A : Alors Philippe, vous vous êtes levé à quelle heure ce matin ? P : Donc ce matin, je me suis levé à six heures et quart, pour commencer à huit heures. A : Oui. Donc ça veut dire quoi, là ? Vous partez à quelle heure de chez vous ? P : Je pars vers 6h55 – 7h de chez moi. Not only is nearly every single word from the list, but I also see the underlying grammar and particularly natural phrasing. A perfect combination of words and linguistic knowledge. From a learner's perspective, the question is: What does it take to be able to talk like this? To answer this and robarb's other question, I really do believe, as I have said, that anyone who knows these 301 words and the underlying linguistic content has a very solid foundation in French. How to acquire the underlying linguistic content is the difficult part. I hope nobody believes that it's a matter of learning 301 words in a month or that you stop at that figure and claim you can speak great French and pass a CEFR exam. Just mastering those 37 connector forms is a lot of work. 56 verbs may not sound like a lot when you think French has over 12,000 verbs but if you have those under your belt, you have a great base. I don't think that all this is really very different from what people do anyways. I've just put a figure and some specific words to the old idea that you can start speaking without knowing everything. But more specifically it addresses a problem that language teachers see all the time: learners who are relatively advanced but keep making basic mistakes because they glossed over some very fundamental skills. In French for example, you see this in grammatical gender, number agreements, past tense usage, the subjunctive mood and hypothetical statement constructions, among others. I would certainly not suggest trying to do nothing but work only on 300 words before going on to anything else, but I see this list as a kind of reference for things to look back at from time to time. I've implemented a similar strategy for high-proficiency Spanish whereby I've identified a number of very basic grammatical structures and idiomatic uses of very common verbs that I've neglected. Now I find myself reviewing things that I've seen years ago but never really mastered. I've even identified mistakes that I've been making for years. It's sort of a return to basics. 1 person has voted this message useful | ||||||||||||||||||
| s_allard Triglot Senior Member Canada Joined 5226 days ago 2704 posts - 5425 votes Speaks: French*, English, Spanish Studies: Polish | Message 15 of 28 13 October 2014 at 10:31pm | IP Logged | |||||||||||||||||
With all due respect to Jeffers, I think there was considerable hullabaloo about the idea of only 300 words for speaking. But all that is old hat. What I have noticed is that not many people seem to the object to the idea after all. Even the idea of passing the CEFR exams with 300 words has hardly caused a stir. Why the idea of threshold if people can start speaking from day 1. Isn't Bonjour, ça va? and Merci beaucoup speaking? Couldn't one start speaking French with "a couple dozen words"? I'm glad that I'm not the one saying that you can speak French with a few dozen words. The idea of a threshold is that around a certain number of words, e.g. 300, the learner is somewhat autonomous and can cover a wide range of situations beyond ordering a coffee in a café. 250 could also be a threshold. Maybe 400 is better. 300 is a figure that one finds in the literature, but it's arbitrary. One could also discard the whole idea of a threshold and say that there is simply a range from 48 (four dozen) to 1000 for the learner to start speaking. 1 person has voted this message useful | ||||||||||||||||||
| Ezy Ryder Diglot Senior Member Poland youtube.com/user/Kat Joined 4145 days ago 284 posts - 387 votes Speaks: Polish*, English Studies: Mandarin, Japanese | Message 16 of 28 13 October 2014 at 10:46pm | IP Logged | |||||||||||||||||
|
I think prof. Nation has been taking part in developing something similar. A set of IIRC ~200 items
(words, phrases, etc.) for traveling in a country. Maybe you'd find looking it up worthwhile. And the threshold might be a bit subjective, in my opinion. 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. |
