|
 |
Comparing language portfolios | ||
Tags: CEFR Language Levels | Fluency | ||
Share with:
| ||
 Language Learning Forum : General discussion |  | |
| litovec Tetraglot Groupie Switzerland lingvometer.com Joined 5133 days ago 42 posts - 60 votes  Speaks: German, Russian, French, English | Message 9 of 18 17 April 2015 at 5:31pm | IP Logged | ||||||||||||||
An algorithm on a tree, here is a relatively short description (on 4 pages) with formula and pictures
That's a good question. I interpret it as an "effective" number of languages contrary to a "nominal" number of languages. Imagine that New York declares independency tomorrow and call their language "New Yorkian". Nominally, a citizen of a new country would speak 2 languages: English and "New Yorkian", "effectively" only one. 50 years later the languages would go different paths, thus, "effectively" it would be then more than 1, however, still way less than 2. Another interpretation: if learning Chinese fluently gains 1 unit, then with 1.3 you have gained 30% of Chinese equivalent )
It's like comparing IQ of 157 to IQ of 132. IQ numbers are also not exact and depend on test, they show rather the direction. LQ (a number what a lingvometer produces) in the same way could also deviate depending on algorithm, they should, however, show the direction.
I'm absolutely agree with you, however, they don't ask it in IQ tests if a test person was born in professors family or had a fancy book and give (or subtract) extra points for it. They test an achievement that already took place and the same principle is applied in lingvometer. How, she/he got there is up to the tested person. 1 person has voted this message useful | |||||||||||||||
| Serpent Octoglot Senior Member Russian Federation serpent-849.livejour Joined 6599 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 10 of 18 17 April 2015 at 7:06pm | IP Logged | ||||||||||||||
|
I got 3.44.
I think one downside of the system is that people who learn related languages already take their free lunch into account. For example, my comprehension in Romanian is B1-B2 but I listed it as A2 based on my active skills. Now I listed all my passive levels and I got 4.45 :) Did you deliberately exclude Esperanto btw? 2 persons have voted this message useful | |||||||||||||||
| James29 Diglot Senior Member United States Joined 5377 days ago 1265 posts - 2113 votes Speaks: English*, Spanish Studies: French | Message 11 of 18 17 April 2015 at 7:24pm | IP Logged | ||||||||||||||
|
Thank you for posting the article. That is interesting.
The article suggests that "proficiency level" can be measured with a precise numerical value between 0% and 100%. It also says the precision of this value holds true even when it is compared between different languages. Am I reading the article correctly in this regard? This is not possible mathematically. Proficiency level can be compared and people can be ordered in order of proficiency (in the same language), but when you try to measure the difference between people the math does not work. What would the measuring unit be? 1 person has voted this message useful | |||||||||||||||
| robarb Nonaglot Senior Member United States languagenpluson Joined 5061 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 18 18 April 2015 at 12:03am | IP Logged | ||||||||||||||
|
Of course you can't map proficiency to an exact number, and the numerical difference
between two proficiency scores isn't very meaningful. The CEFR scale also cannot be measured in a universal way. It's just a simplification to map each CEFR level to a number between 0 and 100 to represent what "percent of the way to C2-level proficiency" you've reached. The precise values are necessarily fudged so that the output of this calculator is a rather rough measure. Still, it's a VAST improvement over "how many languages do you speak?" For that, good job. I was aggressively conservative in the assessment of my level as I like to interpret CEFR by the actual meaning of the guidelines rather than the far more lenient standards it would generally take to pass a typical exam, and I only credit myself with a level when I'm sure. So I ended up scoring 2.83. Seems low, but then again if 1 point is the equivalent amount of learning that it takes to go from nothing to C2 in Mandarin, then maybe that's not that low after all. It gives one pause to think how much less learning a polyglot like me actually has in each language relative to the amount of learning that differentiates zero from a native speaker. A major weakness in the approach is that it considers the knowledge shared between languages to have a tree structure according to language family relationships. In reality, you can have some long-distance connections through loanwords or chance grammatical similarities. Also, sibling languages within a family are treated as equivalent: in the example, Serbian, Croatian and Slovenian are treated as three nodes underneath Western South Slavic. But in reality, Serbian and Croatian could be grouped together in a subfamily and are much more similar to each other than either is to Slovenian. Without a full measure of similarity among all languages, some degree of information is lost by putting everything in the structure. 2 persons have voted this message useful | |||||||||||||||
| outcast Bilingual Heptaglot Senior Member China Joined 4951 days ago 869 posts - 1364 votes Speaks: Spanish*, English*, German, Italian, French, Portuguese, Mandarin Studies: Korean | Message 13 of 18 18 April 2015 at 3:22am | IP Logged | ||||||||||||||
|
I appreciate the OP for bringing this matter here and be open to both praise and critique.
I think what you are essaying to constitute with your work, and the concurrent value you are attempting to quantify, are most intriguing and certainly a topic of much debate and controversy. Of course people should in its totality take this with a big grain of salt, lest it set off another wave of "instant indignation" on the internet, but it is nonetheless fascinating to ponder. However, I would have to say, pace your mathematics, that if this calculator were a human being and I explained to "it" that I speak English and Spanish at a native level (and both with university-grade lexicon and grammar), and it replied to me "you speak 1.63 languages", I would look askance at it, evincing cheeky incredulity at this assessment. I believe I am pretty effective in those two languages. The problem therein lies that the initial assumption made seems that I either learned English and then Spanish, or viceversa. That's not how it happened. So I guess it is a matter of semantics. 1.63 "effective languages"? I don't think so. 1.63 "LQ"? Ok, I'll accept that, whether it is a valid measurement or not I'm fine with it, since everybody else presumably is measured with the same touchstone. 2 persons have voted this message useful | |||||||||||||||
| tarvos Super Polyglot Winner TAC 2012 Senior Member China likeapolyglot.wordpr Joined 4709 days ago 5310 posts - 9399 votes Speaks: Dutch*, English, Swedish, French, Russian, German, Italian, Norwegian, Mandarin, Romanian, Afrikaans Studies: Greek, Modern Hebrew, Spanish, Portuguese, Czech, Korean, Esperanto, Finnish | Message 14 of 18 18 April 2015 at 9:19am | IP Logged | ||||||||||||||
|
3.88 for me. And that's with pretty well-rounded skills, I averaged towards the middle
for passive/active in the cases of languages like Danish/Afrikaans. 1 person has voted this message useful | |||||||||||||||
| litovec Tetraglot Groupie Switzerland lingvometer.com Joined 5133 days ago 42 posts - 60 votes Speaks: German, Russian, French, English | Message 15 of 18 18 April 2015 at 10:24am | IP Logged | ||||||||||||||
It's a lot! If you take an average citizen of a multilingual country, they have a LQ below 2. You could also find authors of the books of the kind "Learn N languages in M hours" that are below that. It could be frustrating getting out of the n languages a number that is usually way below n, but I think it shouldn't be :)
Kind of yes, it wasn't there in the classification I used and I have a hard time embedding it in a tree structure. So, I don't have a nice solution for Esperanto at the moment.
If you as a let's say Francophone, add Spanish at A1 level (that would be a free lunch without even touching Spanish, right?), you get nothing. Even if you assume the free lunch to be the A2 level (that would be an exaggeration in my opinion) you get only 0.01 points extra. Rounding to 1 digit, you get nothing. Even if I add a bunch of languages at A1 level that a French would get for free, then it still a 0 points improvement. In this regard, the system is "free-lunch-proof" (if I understood your notion of the free-lunch correctly).
True.
It's true, I see it, however, as a minor issue. It has little impact on the score since the level of the subfamily is already deep enough. Moreover, if you add together one of the three with something outside Western South Slavic, it has no effect. 1 person has voted this message useful | |||||||||||||||
| Serpent Octoglot Senior Member Russian Federation serpent-849.livejour Joined 6599 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 16 of 18 18 April 2015 at 2:37pm | IP Logged | ||||||||||||||
Nope, I mostly mean what you get for free when you actually learn the language. So if it had actually taken me the same effort to understand Romanian as Portuguese (which was my first modern Romance language), I'd have listed it as B1, but since my active skills are low and I got the passive ones through the Romance languages and Russian, I opted for A2. So kinda the opposite of what you thought. I already subtract a bit while entering, and then you subtract more. Basically, to me it's strange to take into account related languages but not the passive/active levels. And nobody is truly A1 without prior study/exposure, so while the effort may be minimal, it's still needed (and A1 isn't that hard even in an unrelated language). 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. |
