To me a word is like a person you get to know and who is going to help you learn the language. You know lots of people without knowing them in detail. The more often you meet them in different situations the better you get to know them. It is the same with words. The task in language learning is not only meeting words or friends for the first time but also getting to know them through frequent exposure. If you focus your efforts on trying to know a few words very throughly, then you won’t have the time to expose yourself to the words you have already met. We constantly need to see and hear even the most basic words that we are always getting to know better and better.
When I see a yellow word go by, and recognize it more easily than I did previously, I mark it with a lighter yellow, until eventually I simply clear it.
I just watched Steve’s YouTube video about this thread and it made me think about my experiences with Anki. I used Anki as my main way of learning vocabulary starting around the beginning of December and ending mid January when I realised that it could not keep up with LingQ. Anki keeps good statistics. I used it a lot, and was doing about 45 minutes per day, entirely while travelling to and from work. In that time, I had got around 1200 words to what could be called the known level. This corresponds to about 25 words per day (and this is not counting different word forms separately). All of the words I added myself, and I only added words that were unknown to me.
As a side note, the reason I stopped liking Anki was not that I found it to be a slow way of learning flashcards, but that I found it to be an ineffective way to learn words. Anki was great at helping me to memorise words such that I could recall them within Anki, but the moment I saw the word outside of Anki, most of the time I was unable to understand it. I remember sitting on a tram in Vienna studying on Anki, reviewing a word and being able to recall the meaning of it instantly, and then 10 seconds later looking up and reading the word on the wall of the tram and having no idea what it meant.
When I used Anki, I usually just did a clumsy (but plausibly effective) import of sentences wherein I had seen a given word. I didn’t bother to prepare a definition for the other side. If I grasped the word upon seeing it again, that was good. In a way that was a lot like a portion of the process here. But the LingQ process is much more complete.
@ Steve
I have been thinking a bit more about how you got the 400 words per day number. You wrote the following.
"In the case of Romanian, I acquired over 24,000 words in 60 days according to my statistics at LingQ. This includes the 21,000 “known words” on my stats, and some portion of the over 10,000 LingQs that I created. (I assume that I know half of my “LingQs.) This means I learned, from scratch, over 400 words a day.”
Looking at your stats, I see you have 21,048 known words, but only 641 of these were words that you originally LingQed, and then set to known. This means that of the 21,048 known words, 20,407 were set to known when you first came across them. Given how you use the system, you would have set these to known because you knew their meanings the first time you saw them. In what way does this count as learning the words? Surely when you set a word to known the first time you see it, all you are doing is telling the system that you already knew this word. For example, if I opened an English lesson with 1000 words in it, chances are I would know all of the words in there. If I then set them all to known, could I claim that I just learned 1000 words?
I would say that based on your statistics, it is only fair to say that you learned 641 words. In 60 days, that is then not much more than 10 words per day. There is also an unknown fraction of the 10,674 LingQs that you have yet to set as known but have learned in the process, but I think it is difficult to know what this fraction is (in my experience, it is a small fraction). Probably 10 words per day is a bit small but sounds to me more reasonable than 400 per day.
@ creimann
I tried doing all this stuff with learning words in sentences through Anki but I could never be bothered to keep it up. Probably it would have been better if I had done it that way, but then it would have slowed me down a lot. I was able to get through so many words in Anki because I did not spend time on each individual word. It is the same in LingQ. I never read the phrases that are saved with the LingQs, even when I do flashcards.
@ColinJohnstone
I think when Steve comes across words and they are marked known without him having to LingQ them is what he would call the incidental learning.
You have to consider that LingQ does not consider declensions. So eat and eaten can be different. So if someone came starting English from zero and had to LingQ eat, later on they may not have to LingQ eaten because they understand based on context.
@ Colin,
You cannot discount the words that I learned incidentally. When I started Romanian I knew no words of Romanian. Soon I realized that there were many cognates with Italian and French. Soon I got to know which ones have the same meaning as their cognates and which don’t. Some I could guess from context, and as brichardson2x says, many were different forms of words that I had learned.
I see no reason to discount incidentally learned words. Most words are learned incidentally. The more you listen and read, the more incidental words you learn. That is the advantage of this form of learning as opposed to heavy Anki learning.
The analogy with English does not work. You already speak English. You cannot claim to learn a language you already know. I didn’t know Romanian.
@ Steve
What exactly do you mean by learning a word incidentally?
You said earlier in this post on this thread
“My goal in learning words is to learn to recognize them when I hear them and see them in context.”
From this, and from other statements of yours, I assume you meant that you consider a word ‘known’ if you are able to recognise one of its uses when you come across it in a context. Please correct me if I am wrong by what you mean when you say you know a word.
Presumably in order to ‘learn’ a word, you need to first not know it, and then after learning it, know it. By your definition of knowing a word, that means that you could not at first recognise the word in context and then after learning the word, you could recognise it in context. But if you were able to understand a lot of the vocabulary already from other Romance languages or even English before you saw them in Romanian, in what way can you claim to have learned those words, given your definition of what it means to know a word? You may not have known Romanian, but you knew most of the vocabulary, by your own definition of what it means to know a word.
The analogy with English works in the way I meant it. I don’t think I can claim to have learned words that I already know, no matter what language it is. For example, if I come across ‘to import’ in an English text, and set it to known on LingQ, clearly I have not just learned that word. Similarly, if I come across ‘importieren’ in a German text, and know what it means because I know it from English, I can’t claim to have learned the word just because I have seen it for the first time; or at least not by your definition of what it means to know a word (as I understand it).
Wow. Rereading my last post, I am struck by how much it sounds like a poor quality philosophy essay that tries to find a clear answer to an fuzzy ambiguous question using poorly defined concepts. Maybe it is silly to try to discuss this at all. Maybe we should just give up and accept that the question cannot really be answered in any unambiguous way. Astrophysics is simpler than this.
Colin, we are talking about learning languages. You cannot learn a language you already know. You can learn Romanian you cannot learn English.
The first time we hear a new language we don’t understand much. If the language has cognates we may think we hear a recognizable word but we aren’t sure. The same is true when we read, although in reading we may recognize more words that we think may be cognates. In learning all languages, things that appear opaque, fuzzy or difficult at first, slowly become clearer and more understandable. In my 10th or 20th or 30th lesson I may skip over a word and not LingQ it, because I think I know it. I might very well have saved or LingQed this word if I had come across it earlier. This is true for all languages. It is true for me in Romanian, Russian, Czech and Korean. The higher the percentage of cognates, the faster we can learn new vocabulary especially incidentally.
Our ability to notice, our ability to infer meaning, our ability to relate new word to words we have already learned, this all increases over time. That is why it is so important, in my view, to focus on massive input rather than studying individual lists of words. You will learn more words, and especially more incidental words, if you devote yourself to massive listening and reading.
Not at all, Colin. I believe this issue of the words we learn incidentally is very important. It is a major reason why the massive input approach to vocabulary acquisition is so powerful.
Some people are happy with getting the gist and using whatever knowledge they have (and still keep improving little by little). Others are obsessed with getting everything perfect from day 1.
I’ve met both kinds of people in all aspects of life, sports, music, languages… My experience is that people from group 2 never seem to get anywhere, despite the “solid base”.
For example, I have been studying a biology lesson I have imported. It had many new words, but all of them were cognates for me: “monera”, “protoctista”, “fungi”, “carbon atoms”, “lipids”, “nucleic acids”, “colonies”, etc. So I didn’t lingQ any of them. I learnt them immediately. I know I can recognize them in the future in spoken and written language. They are now part of my “known words”. Summing the cognates, the words with an obvious meaning for the context, and the different forms of some words (set of words with the same meaning) , I suppose that the number of known words may be dramatically increased in these cases.
@ Steve
“Colin, we are talking about learning languages. You cannot learn a language you already know. You can learn Romanian you cannot learn English.”
We are talking about learning words, and my point is that we can’t learn words we already know.
I will write more tomorrow when I am at a real computer. I hate typing on the iPad.
If you’ve never seen the word, but you understand what it means did you already know it? How could you already know a word without having seen it? I believe your brain, with a combination of context and prior knowledge seemingly ‘instantaneously’ learned a word.
I’ll use an example in a language extremely far from English.
In Japanese I knew 俳優 meant actor. I had never seen the word 女優 but I knew that the word meant actress, I mark it as known. How would you explain this phenomena(Knowing this will give me a better perspective on where you’re coming from)?
My point is that many of these cognates are not obvious at first. Besides, let’s not be fixated on Romanian. Not all incidentally learned words are simply cognates.There are far fewer obvious cognates with English in Czech or Russian not to mention Korean. You may have noted that I suggested 100 words a day as a reasonable number for a learner to achieve at LingQ if they follow my approach, counting words the way we count them here… Romanian is an aberration.
However, if we do not count cognates, then we would have to discount any English person’s achievements in learning French, which I would be loth to do. See the example below.
I have taken a text from Le Monde. I have put in the cognates from English below each line of French text. Am I to understand that you would consider these cognates not to be “newly known words” nor “learned words” for someone studying French. I think that many people, myself included, would LingQ some of them, if coming across them for the first time.
Sauver l’Italie ou Berlusconi ? Au lendemain d’une journée marquée par la décision de la Cour de cassation de statuer dès le 30 juillet sur la condamnation en
save Italy journey marked decision court cassation statute July condemnation
appel de Silvio Berlusconi à quatre ans de prison et cinq ans d’interdiction d’exercice de charge publique pour fraude fiscale, les parlementaires du Peuple de la
appeal quartet annual prison annual interdiction exercise charge public fraud fiscal parliament people
liberté (PDL droite) ont décidé de boycotter, mercredi 10 juillet, les travaux de la Chambre et du Sénat malgré la détérioration de la situation économique. La
liberty adroit decided boycot July travail chamber senate deterioration situation economic
veille également, l’agence de notation Standard & Poor’s (SP) avait abaissé d’un nouveau cran (BBB) la note de l’Italie à la suite de prévisions plus sombres
surveillance equal agency notation base new note Italy suite prevision plus somber
que prévu (moins 1,9 % du PIB en 2013, selon SP, - 1,8 % selon le Fonds monétaire international).
preview minus funds monetary international
I don’t think it actually worth worrying about what it really means to know a word. There is no correct answer. I was just going by how I thought Steve defined it.
I was only fixated on Romanian because I thought Steve’s claim to have learned 400 words per day was a little excessive, but that is certainly not an important point here.
The 2X man’s example is a good one. While the process of recognizing the meaning of new words may be more obvious in languages that use characters as in his example, the process of acquiring new words by default is the same in other languages. We learn them from context. We learn them by associating them with words we already know. We learn them because they are new forms of words we already know. These incidentally learned words are our largest source of new words or learned words in every language that I have learned.
Colin, what are your views on the French cognates for an English speaker? Would they count as new words? As learned words?