Video: If you don't know the word for shoelace, you are not fluent

Open Forum
21

If you teach a novice the word shoelace he/she won’t be anymore functional in that language because of that, I agree, but if you ask language learners if they know certain words like shoelace, hovercraft etc. the more they know of these lower frequency words the better they will likely be in that language.

Another point I wanted to make is that I think the example of a person who has this enormous vocabulary but cannot string together sentences is more an anectote than a typical case. Normally, if we have acquired words and have trule internalised them we also tend to speak better and better.

22

I have lived in lots of different countries over the past 30 years, and met lots of expats in those countries who have studied extensively, have large vocabularies, and can hardly say anything at all in an actual conversation. Adding shoelace, or other rare words, to their lists would make no difference. There are two things they had in common: (1) It was their first foreign language, so they had never learned the skills I talk about in the video; (2) They believed that learning more words would compensate for point 1.

In contrast, I have seen time and again people who have already learned to speak a foreign language, and therefore gained the skills I mention in the video. Since these skills transfer across languages, these people can usually start a new language and then speak very early on (albeit with limited vocabulary).

Therefore, a first time language learner who knows the word shoelace will, at least from the cases I have seen, be far less functional, than an experience language learner with a far more limited vocabulary.

In short, I believe that for a first-time language learner, the list of known words is less important then helping them gain conversational skills in which they can express everything they wish to say.

Once people have got over that conversational hurdle, then - sure - fill up their vocabulary reservoir as fast as you can.

23

@Friedemann: “…but if you ask language learners if they know certain words like shoelace, hovercraft etc. the more they know of these lower frequency words the better they will likely be in that language…”

I can see your point, Friedemann. But the problem is, I just don’t think you can isolate any given low-frequency word (or even small list of words) such as “shoelace” or “hovercraft”, etc. The exact group of known low-frequency words is surely going to vary from one person to the next?

Now you knew the word “shoelace” in English, Norwegian, French, Mandarin. Very good. But in theory it’s perfectly possible that in each case you just happened to learn this word early on - maybe even in your very first lesson. (I’m not saying this actually was the case - I have no way of knowing that. But it’s possible.)

It’s equally possible that there are other words (maybe even slightly higher frequency ones) which you don’t know in some/all of these languages.

If someone could trip you up with a given word, would that mean you suddenly stopped being fluent in all of these languages? (And yes, I know that you really can speak fluently in English, Norwegian and Chinese, because you once did a multilingual interview with Steve in these languages!)

But the word “shoelace”…!?

Well, I didn’t know it in German - and yet I have passed C1 level exams in the language!

24

@ J_for_Jones - “But in theory it’s perfectly possible that in each case you just happened to learn this word early on - maybe even in your very first lesson”

The question is one of probability, and can be described using Bayes’ theorem. Imagine that we have a person, we want to know if they are at a C1 level in language X, and we give them the shoelace test and they pass it. Prior to giving them the shoelace test, based on what we previously knew about their ability in the language, we estimated that they had a certain probability of being at a C1 level, which we call P(C1). What we want to know is the probability of them being at a C1 level given the fact that they passed the shoelace test - let us call this P(C1 | pass). Bayes’ theorem states that this is

P(C1 | pass) = P(C1) * P(pass | C1) / P(pass)

where P(pass | C1) is the probability of them passing the shoelace test if they were at the C1 level, and P(pass) is the probability of them passing the shoelace test regardless of their level. The important part of this equation is the term P(pass | C1) / P(pass). Since P(pass | C1) is greater than P(pass), i.e. somebody at the C1 level is more likely to know the word for shoelace than most learners, the term P(pass | C1) / P(pass) is greater than 1, and therefore P(C1 | pass) is greater than P(C1), i.e. the probability of them being at a C1 level is greater after giving them the shoelace test than it was prior to giving them the shoelace test.

Assume now that they failed the test, we now have the equation

P(C1 | fail) = P(C1) * P(fail | C1) / P(fail)

where the quantities mean the same as before, but with reference to them failing instead of passing. The probability of somebody failing the shoelace test at a C1 level is lower than the probability of a random learner of language X failing it, therefore the term P(fail | C1) / P(fail) is less than 1, and therefore P(C1 | fail) is lower than P(C1), i.e. the probability of the person being at a C1 level is lower after failing the test than it was prior to the test being given.

Let us instead ask whether or not they are at the A1 level. We have the same equation as before, but now it looks like

P(A1 | pass) = P(A1) * P(pass | A1) / P(pass)

where P(A1 | pass) is the probability of the person being at an A1 level given that they passed the shoelace test, P(A1) is the probability of them being at the A1 level before the test, P(pass | A1) is the probability of them passing the test given that they are at the A1 level, and P(pass) is the probability of them passing regardless of their level. In this case, P(pass) is greater than P(pass | A1), i.e. the A1 learner is less likely than the average learner of the language to know the word for shoelace, and therefore P(pass | A1) / P(pass) is less than 1 and P(A1 | pass) is less than P(A1), i.e. the probability of them being at an A1 level is reduced by the fact that they know the word for shoelace.

In the above reasoning, I have assumed that we are only considering groups of learners of language X and not the general world population as a whole. If instead we considered the world population as a whole, things are slightly different, but the general reasoning holds.

Sorry for bringing mathematics into this, but I find this stuff interesting, and I know other people do. People who don’t find it interesting should just ignore this post. Anything who sees anything wrong with my reasoning should of course tell me.

25

@Colin

I’m sorry, you lost me after the first sentence! :smiley:

But you are kidding, of course? (Please tell me that you are kidding!)

UPDATE
I wasn’t trying trying to be rude, or anything. It’s just that my eyes totally glaze over when I hear things like “Bayes’ theorem”…! (Maths was never really my thing - sorry!) Ah well, it has to be better than “Clug’s Theorem”, I guess! :smiley:

26

How about the last sentence -

“Anything who sees anything wrong with my reasoning should of course tell me.”

27

I hate getting into mathematics or formal logic to talk about these thing, but:

The problem with the shoelace test is that it relies on abductive logic. It accepts a single point as “evidence” of something, which is not reliable.

Let’s use cats as an example of where abduction goes wrong:

Cat’s have tails. My dog has a tail, so my dog is a cat.

Which is just like saying:

Functional speakers know the word shoelace. Charlie Brown knows the word shoelace, so Charlie Brown is a functional speaker.

28

BTW Anthony, are you the guy who has a PhD in Computer Science? (Or am I thinking of someone else?)

29

Yes, I do have a PhD in Computer Science! You are well informed!

30

Well, I was just thinking that I read it somewhere (maybe it was on your website?)

I still think that your fluency in Czech is (even) more impressive though! :wink:

31

@ J_for_Jones - Your post was not rude. My post was to an extent a joke, in that I did not expect anybody to really read it in detail, but the actual reasoning was serious. I will explain what I mean in a better way.

Imagine you have two dice, both with six sides. On one dice (I will write the singular as dice as this is standard usage, even though dictionaries dictate that I should use die), five of the sides have a 6 on them, and the other one has a 1. The other dice is normal, with only one side that is a 6. You put both dice in a bag and randomly pull one out and throw it without looking at it. At this point, the probability of you having the dice with with five 6s is 50%. You then see which side ends facing upwards (and for some reason you don’t see the other sides), and this side is a 6. It should be clear here that this does not prove that you have the dice with the five 6s, because you could easily have the dice with the one six and just happened to have got the 6. However, it should also be clear that you are more likely to have the dice with the five 6s. The reason is that having the dice with the five 6s makes you more likely to get the six than having having the dice with the one 6.

Similarly, if you have a group of learners of language X and you do not know anything about any of them, but you do know that they are all at different levels in the language, and you pick one at random and give them the shoelace test, if they pass, then they are more likely to be at a higher level in language X. This is because people at a higher level in language X are more likely than the average to know the word for shoelace. The person who you picked could easily be a complete beginner who knows 100 words and just happened to learn the word for shoelace the day before and will forget it the next day, but this is less likely than them being an advanced learner who knows 10,000 words in the language.

32

@ AnthonyLauder - I don’t see any similarity between your two syllogisms and the problem of the shoelace test. The reasoning in the first one is

  • Cat’s have tails.
  • My dog has a tail,
  • so my dog is a cat.

This has nothing to do with probabilities and is not an example of abductive logic [edit: at least not in the way that I first read it, but having read it a few more times, I can see how it could be]. It is an example of bad deductive logic. The same is true of

  • Functional speakers know the word shoelace.
  • Charlie Brown knows the word shoelace,
  • so Charlie Brown is a functional speaker.

This reasoning is nothing like the reasoning I gave in my previous post. If you want to put my reasoning into a syllogism then it would look like this

  • Learners of a language who know the word for shoelace are more likely than the average learners to be at a high level in a language because according to Bayes’ theorem…
  • Charlie Brown knows the word for shoelace
  • so Charlie Brown is more likely to be at a high level in a language than the average learner
33

My opinion on the subject of the shoelace test can be summarised as follows:-

Q. Can you know if somebody is functional in a language from whether they know the word for shoelace?
A. No

Q. Can you learn anything about how good somebody is in a language from whether they know the word for shoelace?
A. Yes, but only in terms of probabilities.

Q. Is this a reliable test?
A. Probably not.

34

@Colin: “…It should be clear here that this does not prove that you have the dice with the five 6s, because you could easily have the dice with the one six and just happened to have got the 6. However, it should also be clear that you are more likely to have the dice with the five 6s. The reason is that having the dice with the five 6s makes you more likely to get the six than having having the dice with the one 6…”

Yes, that is indeed a much nicer and more elegant explanation than the original post :wink:

I think I follow your reasoning: at the point where you draw the dice out of the bag it’s a complete 50:50 whether you get the “strong dice” or the “weak” one. But having thrown a six it’s still more likely to have come from a “strong dice”.

Well, okay, I can see that this is a very good point. But how about if there were 10 “weak dice” in the bag, and only one “strong dice”? (It may be the case that there are more average learners than advanced learners out there…?) In that case you would be starting with only a 1:10 chance of having the “strong dice”, right? So that would have to boost the odds of a six coming from a “weak dice”, right?

In general terms I think you would need to test people with a much larger sample of low-frequency words than just two or three. If you tested people with a 1000 low-frequency words, then those who knew (let’s say) 50% of them would almost certainly be very much more likely to have a high level than those who only knew (let’s say) 5% of them.

The problem with using just one word (like shoelace) is that there is just such a complete randomness about the result. It is a word which a person could easily learn in his/her first lesson - or it may only crop up many years later in a online bust-up with Herr Dr Friedemann Rohr!

35

(double post, oddly enough)

36

I agree with everything you write here. Testing people on a larger number of low-frequency words would be a much more reliable test. This would be similar to throwing the dice a few times. If it comes up 6 four times in a row, then the probability of having a “strong dice” is very high.

By the way, regarding your example with the 11 dice in a bag, if you pull one out, throw it, and get a 6, the probability of that dice being the one “strong dice” is 33% (unless of course I made a mistake in my calculation).

[edit: …which actually shows how good your point about there being more average learners than advanced learners out there is. If you had a group of learners, most of whom are at a low level, and a couple are at a high level, and you chose somebody at random, and they passed the shoelace test, then depending on the actual numbers, it could easily be that the probability of the person being a low level learner who just happened to get the answer correct could easily be higher than the probability of having a high level learner who got the answer correct because they know 20,000 words.

I have heard of similar reasoning in poker where somebody starting with a good hand will try to start with a large bet to scare off all the people with crappy hands because if there are too many people with crappy hands, the chances of one of them getting lucky when the later cards are dealt and winning is really high.]

37

I agree that learning vocabulary lists alone won’t get you to fluency because one needs context to truly absorb vocabulary. Having said that I still believe that quizzing people about words, medium, to low frequency is a very good proxy for proficiency.

Vocabulary rules!

38

@Friedemann: “…I still believe that quizzing people about words, medium, to low frequency is a very good proxy for proficiency…”

I actually agree - but only insofar as one would be testing people on a fairly large sample of medium/low-frequency words. If you test with just one (or even several) such words, there is a high chance of getting a freak result, IMO.

(BTW Thanks to this darned thread I am now the proud ‘owner’ of the words “Schnürsenkel”, “Luftkissenboot”, “Pfau” und “Fußleiste”. I doubt whether I’ll be using them anytime soon, however! :-D)

39

In my view and in my experience it is possible to be fluent in a language and to not know or forget some of the simplest and even commonest words in the language.

40

Be fluent in a language and not know the common words? That’s unpossible!