Avoiding translation

Discussion in 'General Discussion' started by tastyonions, May 27, 2014.

  1. emk

    emk Member

    Joined:
    Jun 3, 2014
    Messages:
    23
    Native Language:
    English
    Advanced Languages:
    French
    Basic Languages:
    Egyptian_ancient
    There are definitely ways to use in-context flash cards with unrelated languages. They serve two purposes: tampering with the frequency of natural input to emphasize certain features, and building memory of collocations and partial phrases.

    I think that fossilization does exist, but many people worry too much about it. I suspect most fossilization is caused by a combination of factors:
    1. Needing to speak a lot very early (A2 or less).
    2. Not getting enough native input.
    3. Getting enough native input, but never paying attention to the details (a weak form of the "noticing hyphothesis").
    4. Having low "integrative motivation", that is, very little desire to identify as a member of a group of native speakers.
    5. Not receiving formal education in the TL. (I recently saw one study the claimed that the grammatical correctness of non-native speakers was highly correlated with years of formal education received in the language—and not so much with age, at least once years of education were controlled for properly. But I can't find this study right now.)
    6. After screwing up some combination of (1–5), I suspect that serious, long-term fossilization still requires building deeply ingrained bad habits. This wouldn't even need to be a purely linguistic process: if you do anything wrong for enough years, it will generally be hard to unlearn it.
    Interestingly, I've actually seen native toddlers with weird, fossilized errors that remained utterly resistent to correction for as much as six months before finally improving. This doesn't happen with every child of that age, but it can happen. And if it persists, most US schools will eventually assign a speech therapist. And as I've noted before, I've seen both good speech therapists and good foreign language tutors working, and the similarities are rather striking. To give a fun example, if native French children make an unusually large number of gender errors, there are actually games that speech therapists can use to tackle that.

    So: fossilization. Yeah, I think it's a real thing, but I think a lot of people worry about it far too much. If you get lots of exposure to native media (or native speakers), if you try to notice grammatical details (even if you can't explain the actual rules), and if you have high integrative motivation, then I think fossilization is a minor danger. Just in case, it's probably worth getting either written or spoken corrections occasionally, just to be aware of any consistent errors.

    Because of my weird history, I actually had a small, solid core of French even at A2. It consisted almost exactly of those things which a parent would say to a toddler. But within those very narrow limits, my comprehension was almost fully automatic, to the point that 10 seconds after someone spoke to me, I often couldn't remember what language they had used. But outside of those narrow limits, my listening comprehension was pretty typically A2: I got maybe 40% of RFI Français Facile on a good day, and I could barely understand anybody other than my wife.

    So it was specifically the non-household French that I had learned from my flashcards that was giving me massive trouble. Every time I hit one of those chunks, my brain immediately switched from transparent comprehension to word-for-word translation. I needed to replace the slow, painful L2->L1 "dictionary" in my head with automatic understanding of those L2 words.

    Like I said, I can easily believe that some people never encounter this problem. But I've learned that at least in my case, my TL needs to stand on its own, and even when I use translations, my ultimate goal is always to understand actual, raw TL in context without translation. I certainly would never say anything like, "Well, sure translation seems to be working for you, but you're really damaging your TL in secret, and it will be hard to recover." Hah, no. My advice on this issue is, "If translation is obviously an obstacle for you, and it's not helping, then maybe it's time to try something else."
    Big_Dog likes this.
  2. Big_Dog

    Big_Dog Administrator Staff Member

    Joined:
    Jan 11, 2014
    Messages:
    1,039
    Native Language:
    English
    Advanced Languages:
    Spanish
    Intermediate Languages:
    French, Japanese, Mandarin, Russian, Swahili, Thai
    Basic Languages:
    Korean
    Interesting post. I liked your description of how you got into the word by word translation habit, and understand now why you have a bit of a phobia of L2 to L1 translation. If someone is prone to doing this during listening, then taking away the ability by never allowing L2 to L1 translation in any situation might be a reasonable fix. Fortunately, I don't think many learners fit into this category.
    Actually, flashcards are out of context. You can say words within a sentence have some context. You can make flashcards from something you have studied in context. But there are no in-context flashcards.

    And learning languages in different families has made me more aware of this. This is because, at least for the difficult languages I've learned, nothing works as well, and it's more apparent that flashcards are merely tools for building up the language in my brain. So I see that a sentence deck in Anki is just a bunch of isolated sentences, and is still quite different from reading a book, for example. A sentence deck is not without it's benefits, of course.
    Last edited: Aug 18, 2014
  3. Iversen

    Iversen Member VIP member

    Joined:
    May 1, 2014
    Messages:
    91
    I agree with emk that the risk of fossilized errors is overestimated - the real danger is becoming complacent as you get better, and if you can avoid that it is easier to correct a bad habit from the vantage point of being a a more or less fluent speaker than it is to avoid errors when you still are a poor struggling beginner.

    As for word by word translations I find them useful, even when dealing with idiomatical expressions, but their role is to allow you to grasp the thinking patterns of the target language through something you know better - and in the beginning that would be your own language. The same applies to the memorization of single words. As you progress you can use references within the target language itself and cut down on the translations.

    And finally about context: In my opinion the real benefit of 'large' contexts is the training in keeping your momentum while reading rather than the limited use of it as a memory hook (as when you remember that you saw a certain word for the first time in a certain book). The contexts which are relevant for learning words and idiomatic expressions are much smaller - rarely more than four-five words. If you want to remember the endings of for instance the instrumental plural of masculine adjectives AND nouns in a Slavic language then it is certainly relevant to memorize a few combinations of an adjective plus a noun with their respective endings even though those particular combinations don't have any particular idiomatic value. And for those who have problems with verbs and their typical prepositions in English it can't be a bad idea to turn a couple of typical cases a couple of times in your head to 'fixate' a certain pattern. To some extent you can also include short word combinations on wordlists (and on ANKI cards), but the purpose of those tools is first and foremost to teach you the single words that enter into the expressions, and cluttering them with details will NOT make that easier. The details come by themselves when you meet the words again in real life .. or context, as real life is called in this context.

    My position on the learning of expressions may not be shared by all, but it has served me well, and it goes as follow: you shouldn't learn expressions without also knowing all the words (and smaller items in them) in them. Without this rule you would actually be trying to learn your target language as if it consisted of words that were 20-30 phonemes long, and you wouldn't have a shred of an idea about the real meaning of those expressions.
    Last edited: Aug 18, 2014
    Big_Dog and Cainntear like this.
  4. Peregrinus

    Peregrinus Active Member

    Joined:
    May 27, 2014
    Messages:
    613
    Native Language:
    English
    Intermediate Languages:
    German
    Basic Languages:
    Spanish
    I agree with this 100% and find it odd that someone would advocate otherwise. It's OK to not explain a conjugated or declined form early on in a course as long as the base word is taught, but I have run across modern courses introducing idioms when even the main word it is based on has not been taught. While it could well be that the most frequent use of such a word is in that idiom, it still helps to know the meanings of all the constituent words. Even just introducing such a meaning parenthetically would be fine.

    Besides the the problem Iversen mentions of turning a language into extremely long groups of phonemes, there has to be the danger that without explicitly teaching the meaning of a word in a phrase or idiom, that the learner will intuit its meaning wrongly, and transfer that wrong meaning to another context.
  5. Cainntear

    Cainntear Active Member VIP member

    Joined:
    Apr 29, 2014
    Messages:
    343
    Native Language:
    English
    Advanced Languages:
    Catalan, French, Italian, Scottish_Gaelic, Spanish
    Intermediate Languages:
    Corsican
    Basic Languages:
    Dutch, German, Irish, Polish, Russian, Welsh, Sicilian
    Backing up Iversen's "long words", there's an example that stuck in my mind years ago. In Scottish Gaelic, "Where are you from" is actually "who/which are you from". In many beginners' courses, it's the only "who" question. I was at a conversation circle, and somebody had just been to a residential course. The native speaker facilitating the conversation asked who the teacher was. The learner answered with where she was from. By parroting the same phrase over and over, she'd learned not to pay attention to the details, and not to analyse the sentence (I am of course referring to subconscious analysis here).

    The simplest example of overgeneralising from idiom would be my first-year French at high school, which started with "my name is..." "I am ... years old" etc. The teacher kept telling us that j'ai meant I have, not I am, but a lot of the class ended up saying j'ai instead of je suis and others said j'ai instead of je. I remember the teacher howling at them in frustration, when she could have avoided the problem by not teaching the age idiom (I have X years) until we had properly learnt je, je suis and j'ai in clear unambiguous contexts.
  6. Peregrinus

    Peregrinus Active Member

    Joined:
    May 27, 2014
    Messages:
    613
    Native Language:
    English
    Intermediate Languages:
    German
    Basic Languages:
    Spanish
    I don't know that I would characterize this phenomenon as overgeneralizing from idiom, but rather overgeneralizing from the internal logic of the grammar of the L2, where its literal meaning is also its literal meaning in both languages. In German one says "mit ...[x]... Jahren" for "at age [x]", i.e. "with" so many years. There are frequent posts on HTLAL over time where learners have great difficulty with not just idiomatic non-literal meanings, but also the literal meaning that is used which differs from the internal logic of their native language. It is the process not only of translating, but also not grasping the literal grammatical logic of the L2 that is the problem, instead of just accepting it, even without understanding it.
  7. Cainntear

    Cainntear Active Member VIP member

    Joined:
    Apr 29, 2014
    Messages:
    343
    Native Language:
    English
    Advanced Languages:
    Catalan, French, Italian, Scottish_Gaelic, Spanish
    Intermediate Languages:
    Corsican
    Basic Languages:
    Dutch, German, Irish, Polish, Russian, Welsh, Sicilian
    I think our disagreement might stem from the meaning of "idiomatic". A lot of people use it to talk about proverbs, fixed phrases etc, but at uni I was taught it as "anything which doesn't follow directly from the grammar", which in practice boils down to "anything that can't be translated literally".

    But that's a minor academic quibble. Whichever way you classify it, if you're comfortable with "I am" and "I have" in general use before you encounter the age expression/internal logic/idiom, you're not going to overgeneralise, and I think we agree on that.

    Going back to the question of "logic", though, the only language I've really felt carries through a consistent age metaphor is Spanish, because you "have" years, and you describe yourself as "being with" years, and the age is consistently treated as an "object" that you carry with you.

    In French and Italian you "have" (item in possession) but then you "are at" (time). In English the metaphor is a personal attribute (I am X) and a time (at X). German goes from personal attribute (ich bin) to physical possession (mit). That ain't no logic!
  8. Iversen

    Iversen Member VIP member

    Joined:
    May 1, 2014
    Messages:
    91
    Let's take an expression which for me is quintessentially idiomatic: "boils down to". The logic is clear: all the 'noise' in a discussion is somehow disappearing, which just leaves the core message or information.

    In the expression this process is illustrated by comparing it to longtime boiling, which typically results in all the meat falling off the bones. And the expression is idiomatically charged because a language learner couldn't have guessed that the native speakers had adopted precisely this parallel. The expression can definitely be translated literally into Danish (".. hvilket i praksis koger ned til..." - but we just don't express ourselves in the same way), and it is obviously grammatical.

    If I was a language learner who met this expression for the first time I would definitely like to know what "boil" and "down" meant, and then I would understand the comparison involved. I wouldn't be satisfied with a loose rendering as for instance "can be reduced to" or "basically means that".

    Of course there are other expressions which with time have become virtually opaque even to native speakers, but they must have been transparent once to those who invented them.
    Last edited: Aug 19, 2014
  9. Big_Dog

    Big_Dog Administrator Staff Member

    Joined:
    Jan 11, 2014
    Messages:
    1,039
    Native Language:
    English
    Advanced Languages:
    Spanish
    Intermediate Languages:
    French, Japanese, Mandarin, Russian, Swahili, Thai
    Basic Languages:
    Korean
    I would ask my African students "Who are you?" to which they would almost always respond "Fine sir, and how about you?"
    Cainntear likes this.
  10. Iversen

    Iversen Member VIP member

    Joined:
    May 1, 2014
    Messages:
    91
    The example with "dont" vs. "of which" is actually also interesting from a totally different perspektive: I would argue that "of which" is the best general translation in English, and even though an advanced Anglophone learner in most cases would reformulate the sentence structure in their L2->L1 translation to avoid it, it still gives the beginner a decent understand of the role and meaning of "dont". So I would not hesitate to put it in a wordlist or on an ANKI card (if I used them). And what is the alternative? Force newbees to try to extract the role of "dont" from X example sentences and and find the common denominator? But that has already been done and the result was "of which", so why not just hand people the solution on a silver platter? Besides any L2-L1 translation is a crutch, and if at some point you have acquired a decent overview over the constructions where "dont" is used you will drop that crutch when you feel you don't need it anymore. No reason to keep an elegant solution secret - it will just prolong the learning time.

    In spite of this it could be relevant to remember that "of what" can be rendered as one ord ("dont") in French. And as octroyenne and emk have pointed out, forcing yourself to find good translations for things you often say in your own language is certainly a relevant exercise - and not only for the C1 segment among the learners. Actually the old discussion about connectors and 'chunks' is the same thing under another name, and it is probably even more relevant for lower skill levels. When you have found good expressions in the target language for things you know you need in your native tongue you can train yourself in using them, and once again: when the detour around a translation isn't longer needed, you will quite automatically drop it and jump directly to the relevant target language expression.

    Basically the cure against too much use of translation is fluency training - you can't be fluent if you first formulate a sentence and then translate it, so if you force yourself to bulldoze ahead you will soon dicover that there isn't time to bother with explicit translations.
    Last edited: Aug 22, 2014
  11. Cainntear

    Cainntear Active Member VIP member

    Joined:
    Apr 29, 2014
    Messages:
    343
    Native Language:
    English
    Advanced Languages:
    Catalan, French, Italian, Scottish_Gaelic, Spanish
    Intermediate Languages:
    Corsican
    Basic Languages:
    Dutch, German, Irish, Polish, Russian, Welsh, Sicilian
    You're conflating presentation with practise here. By all means present the most readily comprehensible explanation/translation/description, but memorising the description is different from learning the meaning.

    So the alternative is to present the explanation, then practise applying it in context. (Did you forget I'm in the "early production" camp?)

    But the point is that "of what/which" isn't something most English speakers would say. Seeing "of which" doesn't provoke recall of any meaningful concept in the way "pineapple", for example, does.

    So while you can use SRS for anything you like, and in the absence of decent materials it may be the best option, you still have to recognise that it is doing two different things. With the concrete and the directly translatable, you're learning directly. With the abstract and untranslatable, you're basically memorising the rulebook in order to learn later. Both are valid strategies, it's just good to recognise the difference.

    Again, though, if it's something you actually use, then you're tying the foreign version to a mental concept that is clearly evoked by the L1. There will initially be some translation, but that's often a question of confidence.

    ...except that people who translate (i.e. almost every inexperienced learner) will simply try to translate faster and faster, and obviously this won't work.

    You're looking at the problem the wrong way round. The problem isn't translation per se, but overly conscious processing, something which is a risk in all fields of learning -- it just so happens that we overthink in our L1, which makes translation appear to be the problem. Overthinking is not the problem though -- it's a symptom. The root cause is that the material is overcomplicated.

    A (near) effortless learning path will avoid overthinking becoming a habit.
  12. Iversen

    Iversen Member VIP member

    Joined:
    May 1, 2014
    Messages:
    91
    I totally agree about the pernicious influence of overthinking, and I have several times warned against it - for instance here, where overthinking would be the inevitable result of inviting learners to discover the many roles of "dont" from a series of examples.

    I don't see the big pedagogical difference between using the word "pinapple" and using words combinations like "of which" in wordlists or anki - except of course that you can draw a pinapple and put the drawing on an Anki card instead of a translation if that's what you want to. It is not easy to make a drawing of "of which". But I'm not drawing now, I just want to know what "pinapple" and "of which" can be rendered as in French, and I'll only use these associations from L1 to L2 until I have reached the point where the French words "dont" and "ananas" present themselves in my mind automatically. Well, actually there is a difference, but not one that is relevant for my memorisation of these words. "Ananas" is a simple substantive, and it behaves like 99% or more of other French substantives. "Dont", on the other hand, is a 'grammar word', and there are special rules for its use in a sentence. You can roughly equate its meaning with "of which" in English, but its syntax is coupled to the uses of a totally different word, "de", and you can't expect this pattern to have an exact counterpart in English. That's why you need to spend time reading and listening, but you don't need to know everything about a word to memorize it.
    Last edited: Aug 22, 2014
  13. Cainntear

    Cainntear Active Member VIP member

    Joined:
    Apr 29, 2014
    Messages:
    343
    Native Language:
    English
    Advanced Languages:
    Catalan, French, Italian, Scottish_Gaelic, Spanish
    Intermediate Languages:
    Corsican
    Basic Languages:
    Dutch, German, Irish, Polish, Russian, Welsh, Sicilian
    I can actually see two possible outcomes, so I wouldn't say "inevitable"... but as the other possibility is that you just end up learning it wrong, I think we both agree that it's not a good strategy....

    The idea of whether it can be drawn is only half the story, and the real distinction isn't between substantives and other words, or between lexical words and function words. I can't draw "them", but it labels a concept that is immediately recognisable to me. Or consider yes/oui/ja/... These are words which are purely functional, but the concept immediately recognisable to a huge part of the world's population. Meanwhile, if you're learning Hawai'ian, you might be taught the word Koki'o. It's a substantive, and you can draw a picture of it. But it's a rare flower endemic to Hawai'i and I for one had no concept of it at all until I googled "rare flower", and even though I now know a fair bit about the recent history of the flower and its semi-extinction, I don't really have a concept of the flower as such.

    This is more a matter of a concept existing at a level above the single word. For example. English has distinctions equivalent to ser and estar, and saber and conocer, even though conventional instruction ignores this.

    Ser: ¿Como és? - What is he/she/it like?
    Estar: ¿Como está? - How is he?

    Saber: ¡Lo sé! - I know!
    Conocer: Lo conozco. - I know him/it.

    It's these sort of phrasal concepts that are behind the argument for learning in context. The problem is that when we try to apply that to list/SRS, we get stuck in one form of the concept.

    Let's get even more abstract and look at differences in frame-of-reference. If you have a movement, you can have manner and direction. Verbs can typically encode either (and sometimes both - "climbed" on its own tends to imply a manner of hands-and-feet and a direction of up, but either of these can be overridden by something else in the context of phrase). So if we have a choice of encoding in in the verb and one in the "satellites" (adverbials), which goes where?

    English prefers to put manner in the verb and direction in the satellite - eg "he ran out [of the room]" - whereas Spanish prefers the inverse - "salió corriendo".

    This concept of "head framed" vs "reference framed" languages can't be put on a card independently, as you need a concept of manner, a concept of direction and a concept of person, and this is where I make the distinction between "learning in context" and "learning in a context".
  14. Iversen

    Iversen Member VIP member

    Joined:
    May 1, 2014
    Messages:
    91
    My basic premise is that my chances of understanding a word in a given context and understanding what it has to do there are much better if you already know a word - even though this only consists in knowing the basic form, a core meaning or two and maybe its morphological class. And the wordlists etc. don't stand alone - there will be grammar studies too, which should teach me how to analyse just about any context I may see that word in. I'm actually more intent on making the memorizing situation simpler than in complicating it.
  15. iaing

    iaing New Member

    Joined:
    Jun 22, 2014
    Messages:
    1
    What do people think of Julien Gaudfroy's comments from HTAL?

    "Avoiding any translation is the only way to really feel a language and think it naturally. After one year of Chinese I basically refused to use anything else than a 100% Chinese dictionary. It's just a matter of time, at the beginning you'll be tired of not knowing the meaning of things, but that will force you to develop hearing abilities and a great feeling for the language. And you'll think naturally in the language because you have no choice."

    "You develop the ability of understanding only when you don't understand and have to guess all the time. "

    http://how-to-learn-any-language.com/forum/forum_posts.asp?TID=7171&PN=0&TPN=6

    *first time poster, avoiding Big Dog's lurker-purge, this is the thread I thought most interesting to me to post on, but not competent enough to have a strong opinion of my own
    Big_Dog likes this.
  16. Cainntear

    Cainntear Active Member VIP member

    Joined:
    Apr 29, 2014
    Messages:
    343
    Native Language:
    English
    Advanced Languages:
    Catalan, French, Italian, Scottish_Gaelic, Spanish
    Intermediate Languages:
    Corsican
    Basic Languages:
    Dutch, German, Irish, Polish, Russian, Welsh, Sicilian
    This sort of discussion tends to encourage circular logic and internal inconsistency,
    This is definitely circular logic. "Knowing" a language means being able to create sentences with little conscious thought, and no translation. Is it productive to say "the only way to know a language is to know a language"? Not really.

    The question is whether translation can help you get to that stage -- I believe it can: "fake it till you make it", and all that. This isn't to say it isn't risky -- in my first foreign language (high school French) I came to realise that I had learnt (or "acquired", if you prefer Krashen's terminology) most of the French I'd been taught, and that I'd been slowing myself down by translating. I stopped translating and learned to trust this sort of "second instinct" I felt about the language, and I was OK. I noticed it for myself, and others don't, so never stop doing it. That's the danger.

    But the reason it's needed is because of the complexity of the material. If you are given too much to learn in one go (and even the conjugations of a French verb are at least 6 things, not 1) then you need to file it all away and have it at your conscious command so that you can teach yourself it.

    If you were a Spanish speaker, and you wanted to know what "oak" means, which definition would you prefer:
    a) roble
    b) A large tree which bears acorns and typically has lobed deciduous leaves. Oaks are dominant in many north temperate forests and are an important source of durable timber used in building, furniture, and (formerly) ships.

    Because now you may be wondering about "acorn" "to lobe" (which isn't even a real verb) and "deciduous".

    Using a bilingual dictionary isn't "translating" per se, it's just the most efficient way of unambiguously determining meaning.
    That's an unsupported assertion, and while it is true that one of the big barriers for learners to get over is the fear and panic associated with not understanding, the more you know, the easier it is to intuit unknown; therefore, the quicker and more efficiently you learn stuff, the quicker you will learn to deal with stuff you haven't learnt yet.
    Big_Dog likes this.
  17. Big_Dog

    Big_Dog Administrator Staff Member

    Joined:
    Jan 11, 2014
    Messages:
    1,039
    Native Language:
    English
    Advanced Languages:
    Spanish
    Intermediate Languages:
    French, Japanese, Mandarin, Russian, Swahili, Thai
    Basic Languages:
    Korean
    Welcome to the forum; really nice post. Actually, it might feel like a "lurker-purge", but it's designed to be a "dormant spammer purge". If you saw the patterns in the non-posting membership that I saw, you'd know that there's a whole lot of auto-generated members here. From now on everybody who wants to join has to answer a simple question I send them in the form of an email. But existing spammers still need to be purged.

    On topic, I agree with Cainntear regarding bilingual vs monolingual dictionaries, and the dangers of translating while using a language. I don't have much to add, except that I learn several languages, and it would suck to have to limit myself to never being able to translate something. I use my native language, English, as a tool when I learn other languages. This tool isn't always available, and it's not always the best choice, but I'm not about to throw it out of the toolbox.
  18. Peregrinus

    Peregrinus Active Member

    Joined:
    May 27, 2014
    Messages:
    613
    Native Language:
    English
    Intermediate Languages:
    German
    Basic Languages:
    Spanish
    I also agree with Cainntear and Big Dog that it is best to learn the definitions of L2 words through one's L1, or perhaps a very well-known other L2. There are dangers with this especially with online dictionaries, but you not only get a clear understanding, but also an understanding that allows you to more clearly see the differences between words in two languages, as in one or the other being more limited or broad.

    If you have reached at least an intermediate stage where for a given L2 source vocabulary is not holding you back (i.e. appropriately graded comprehensible input), and still translate in your head too much, then try reading a little faster or selecting audio/video sources with faster speech.
  19. Iversen

    Iversen Member VIP member

    Joined:
    May 1, 2014
    Messages:
    91
    One's native language or some other wellknown language can actually be used to ease the understanding of constructions which are very different from anything in it, but you need to be willing to treat it rather cruelly if need be - or in other words: to use hyperliteral translations of characteristic phrases in the target language. My favorite example right now is the Serbian (and Croatian) compound past tense, which consists of a present tense form of the verb "biti" (to be) plus a past active participium which has no corresponding form in English. But it is the same form as in Russian, where it is used on its own as the one and only past tense form, so when I see a sentence like "poručila je", I know that the speaker was a female because of the -a, and I got accostumed to the construction by paraphrasing it in my head as *"(she) said 'is'" instead of the uninformative "(she) said" or the correct, but misleading translation "she has said". It sounds alien even to myself, but actually that's how the Serbians and Croats think, and having such a crutch to lean on in Danish or English helps me to internalise the construction.
    Last edited: Sep 8, 2014

Share This Page