Hacker News new | past | comments | ask | show | jobs | submit login
Lost in Translation (wsj.com)
109 points by grellas on July 24, 2010 | hide | past | favorite | 47 comments



The strong form of the Sapir-Whorf hypothesis has fallen out of favor, but research and results like this seem to make it clear that language has an effect on thought and thinking. Linguistic relativity is a dirty word (well, dirty term) but there's clearly something going on.

Perhaps it's time to move beyond the original statement(s) by Sapir and Whorf, beyond the limited interpretation of Linguistic relativity, and to start talking about the bi-directional effects of thought and language.

Those of us who routinely program in significantly different languages know that some languages better suit some tasks. Pure imperative as fundamentally different from pure OO or pure functional. "Thinking in" one langauge colors the way one views/creates solutions and algorithms.

Time to view Chomsky as one, rather limited, point of view, at odds with some of the more recent research.

Some reading:

http://en.wikipedia.org/wiki/Linguistic_relativity

http://www.google.co.uk/#q=sapir+whorf+hypothesis

http://www.enformy.com/dma-chm0.htm


Those of us who routinely program in significantly different languages know that some languages better suit some tasks. Pure imperative as fundamentally different from pure OO or pure functional. "Thinking in" one langauge colors the way one views/creates solutions and algorithms.

I was wondering how programming languages shape the way programmers think about the world.

For example, does it affect the ways we write?


For sure. Functional programming helps to separate a problem in small elements and to think of all kinds of creative ways to combine those elements. I basically came from imperative to functional straight away, so I cannot tell a lot about OO programming, but I definitely see the consequences of OO thinking in a lot of code. Clojure was a great way for me to simplify programming in my mind. Although it's a very capable programming language, the concepts are so simple that it puts your mind into creative thinking :).


So you think about separating problems in small elements in non-programming contexts?


Absolutely. Often I can think of not directly related elements that help to solve something.

Slightly off-topic: I made a shortlink doeslanguageinfluence.tk to make it easier to pass this article on.


First real language I learned was Scheme in college. There were a couple weeks where I had to stop myself from putting parens in place of periods)


I think perhaps the most pervasive element to the phenomenon of linguistic relativity is the existent literary material of a person's respective language.

Has this been considered by linguists?


I speak 9 languages by now, and it's definitely true. I cannot talk well about feelings in other languages than Portuguese, explaining technical stuff and counting I do best in Esperanto (it has a very intuitive regular number system and the agglutinative grammar makes it very convenient to explain the finer details of complicated stuff), when I'm tired I answer casual stuff in my native Dutch, philosophy is a lot easier with toki pona and to be short and clear, English is very useful as it's words generally are very short.

Pro-tip: if you want a quick way to get more control over your language usage, and a way to escape your national language habits, try to learn Esperanto. It definitely helped me.

And as I know this stuff out of experience for quite some time, you can probably imagine why I don't like Chomsky that much (and for similar reasons, Richard Dawkins... Populistic science sucks :(.)


Just out of curiosity, how did you manage to learn 9 languages (and how long did it take you, if you actually "studied" them explicitly somehow)? I'm trying to learn French and it still sucks, after almost 3 months in Paris (but not much social contact, admittedly, as I came here without knowing any people). Any tips from an actual doer? Anything that is possible to do by oneself (i.e. not the standard "get a French girlfriend")?


I am Dutch so I learned English, French and German at school, but actually those are my worst languages. I really started to learn languages with Esperanto. The first merit of Esperanto was to show me that I actually am capable of learning languages (because at school, languages were my worst topics, as far as that I could conditionally go to the next class with a task for French...) Then I started to discover that really the only thing that matters with learning languages is to have fun and to speak, and Portuguese, which I learned last, I speak actually better than practically all other languages, I studied least grammar for it and I use it with my fiancee daily. http://fi3m.com describes mostly how you can actively start learning languages quickly, although details can differ with my path.

I am 21 now, I learned Spanish and Portuguese about a year ago, and most languages earlier.


I speak 2 languages natively, but have recently studied a third (Spanish) and speak it at a beginner, but passable level. I really recommend the Pimsleur method for learning languages. It's a set of audio lessons, where the basic principle makes a lot of sense: learning the same way you learned your native language. Simply by hearing things said, over and over, and taking part in mock conversations. There is exactly zero time spent on teaching you rules, it's all taught by actual talking.

Link to wiki for more info: http://en.wikipedia.org/wiki/Pimsleur_method


On the one hand, there's been research showing that people who move to a foreign country and explicitly learn the grammer learn the language faster than those who only learn naturally.

On the other hand, I think both methods can get at the same aim: massive memorization of both new material and the natural use of patterns.

When we learn a language "from birth" (of course it is more of a 6-year full-time immersion program, which isn't all that amazing in itself than something that happens the day you're born) do we learn the patterns without realizing, or do we literally have to memorize each new adjective ending for nouns in different cases and genders (eg, in German)?

I assume our brains pick up on the patterns, hence why kids make cute mistakes like "I'm the goodest". BUT, as an adult, why not just learn the pattern explicitly AND listen to and participate in lots of conversations AND memorize a lot?

They ALL help. Do as much as you can as often as you can and your learn will spiral out of control awesome ;-)

EDIT: I'm also not sure whether only-memorization to learn patterns works or not. When people say they're "conversational" in 3 months, how much can they achieve in 3 more months? It's a question of whether memorization or whatever process they're using is exponentially scaling. I think that most grammer patterns can be made clear with 3-50 examples, so in the end that doesn't seem that difficult. But I've only had experience memorizing grammer first, with the help of hundreds of examples at the same time, and then producing hundreds of examples myself to become fluid.


Try watching familiar movies that are dubbed into French _and_ have French subtitles. "Children's" movies may be best for this. One caveat is often the audio and subtitles will vary (e.g. tu vs vous); I'm not sure why they do this.

Anyway, watch the movie, and pause it whenever something comes up that you don't understand. You can also try watching English movies with French subtitles but it doesn't develop your French ear. Also Google Translate is good practice. Write something in one language and translate back and forth--it does a pretty decent job. Writing in French will force you to use the proper grammer and vocabulary to make the translation come out OK.

I've in my mid 30's and have learned French fluently and advanced Spanish in the past 6 years. Aside from being immersed in the language, finding a girlfriend definitely helps... :)

Finally Meetup.com and Couchsurfing.org can be good resources to meet locals (of both sexes...).

Good luck!


> One caveat is often the audio and subtitles will vary (e.g. tu vs vous); I'm not sure why they do this.

I'm not sure of the reason behind this, but there are also often minor differences in the English audio and subtitles of films, even when the film was originally recorded in English.


Same-language captions are often taken from the script, which often varies from the final cut. If you have a copy of the dialogue already in-hand, why go to the trouble of doing a transcription from scratch?


Check out the audio cd's by michel thomas. I was trying to learn french a few years ago and this was the best learning aid I found. It's entirely conversatinal - listen and speak along; you never have to write stuff down or try to memorise things.

The best thing, imo, about this system is that you can forget about it for a month and then, when you try it again, you get back into it in no time. It definitely seems to have the edge over other systems I've tried in getting things into your permanent and automatic memory.


what were your end results for the amount of time you invested? are you conversational? fluent? can you read news articles or fairy tales (many languages are quite different when casually spoken versus formally written....usually there are more tenses for verb past-tense conjugations)

(what does "conversational" even mean? that you've memorized common stranger-conversations, or that you know how to conjugate all present tense verb forms?)


  > I'm trying to learn French and it still sucks, after 
  > almost 3 months in Paris (but not much social contact, 
  > admittedly, as I came here without knowing any people).
Well get to know some people! Drop by my office in Suresnes :)


If you don't mind me asking, at what ages did you learn each language (e.g. < 3 years, Dutch)


- Dutch from birth

- English from around 10 years, needing about 7 years or more to get to speak it comfortably

- French and German from around 12 years, never liked it really and couldn't speak comfortably yet, only the last 3 years speaking it with more confidence

- Esperanto since about 15 years, got fluent in less than a year and conversational level in a month

- Ido and toki pona in free time online that follows, mostly in mailing list conversations, probably don't really count as they are conlangs

- Spanish and Portuguese since my 20th, both basic conversational level in about a month, my spanish is still conversational but my portuguese is quite reasonable, to the point where non-brazilians accuse me of a brazilian-like accent.


Thanks for a Toki Pona reference. Very interesting.


What was interesting to me was the use of agents in the English language. In my English classes, using passive voice would get red marks on your paper. So even if there were a natural tendency to think in agent-less terms, it would be drilled out of you in class.

Running my own business, I now find that passive voice is the most diplomatic way to discuss user error. The customer didn't do the wrong thing, e.g., "you made a typo in the shipping address" becomes "the shipping address had a typo". The wrong thing just happened, with no agent bearing direct blame. :)


I'm afraid I must mention that while that example does indeed demonstrate how different phrasing can avoid placing blame, it's not a use of the passive voice. The verb 'to have' is just as active as the verb 'to make'. A passive construction would be, 'a typo was made in the shipping address.'


> Running my own business, I now find that passive voice is the most diplomatic way to discuss user error. The customer didn't do the wrong thing, e.g., "you made a typo in the shipping address" becomes "the shipping address had a typo". The wrong thing just happened, with no agent bearing direct blame. :)

Great observation. My rule - when I screwed up or something even peripherally related to me got screwed up, I took full accountability. "That's 100% my error, that's not acceptable, and steps A, B, and C are being taken to fix it. Do you think that will fix things?" When a screw up was on the client side, then it was, "Yeah, things get messed up sometimes, things get broken, it happens..." followed by switching to active voice to work it out.


You just reminded me of my own experience of this. All the way into university and writing papers, my natural tendency was to write about subjects in a very passive voice, and it was only slightly reduced because it disagreed with my instincts.

Now I live in a culture where the spoken language is dramatically passive and usually actor-less "Q: walk to the store? "A: Sure, walk together!"

and I am often told that the way I think seems very much like the locals think themselves. But I was already thinking this way in my own language, I was just considered abnormal...


Great article. It reminds me of when I was growing up. My parents would often speak to me in Korean. My (english speaking) friends would sometimes ask me "what did she say?" Strangely, though I understood clearly in my mind what my mother said, I could not translate what she said into English. I had to describe what she said instead. This friction never popped up in my head until I was explicitly asked to translate.

As Charlemagne is quoted in the article to say, "to have a second language is to have a second soul." In a weird way, that's true.


"I speak Spanish to God, Italian to women, French to men, and German to my horse." —Charles V


* My parents would often speak to me in Korean. My (english speaking) friends would sometimes ask me "what did she say?" Strangely, though I understood clearly in my mind what my mother said, I could not translate what she said into English.*

I'm English-Spanish bilingual and I have the same problem. I can chat along in Spanish and then try to translate what was said into English but it doesn't work. I have to rethink it and say it all over again in English.

Spanish and English are a lot closer than English and Korean. Since it happens in both pairs, it's probably a universal problem. Translating is a different skill from bilingualism.


Speaking now as a linguist, I am somewhat bothered by articles of this type—not because they're inaccurate, because they're entirely accurate, but because they walk a narrow line between fascination with and fetishization of language. Research does indicate that language and cognition are linked in a lot of ways, that's true; but the problem with overly aggressive Sapir-Whorf pushing is that it becomes all too easy to simplify the "language affects thought" maxim until it is no longer accurate.

For instance, the statement, "Languages affect how people conceptualize time," is true, according to research, while the statement, "[ethnic group] is the only group of people to feel [some overly specific emotion]," is never true, or at least hasn't been true all the many times I've seen it brought up. (Usually, such claims end up evoking the No True Scotsman fallacy when someone else suggests that maybe members of other ethnic groups are capable of feeling the emotion.) "Cognitive tasks like counting are affected by a person's language," is true, while Orwell's "You can make people more think less by making their language have a staccato rhythm," is not. There's a fine line to walk between recognizing that language and thought are interrelated, and thinking of language as the magical thought-producing-machine that completely determines all mental (and sometimes physiological) functioning.

Finally, a lot of people read "language" and therefore forget the importance of things like dialect and culture. There's a common tendency to become obsessed with 'untranslatability' as a marker of language and forget that a lot of untranslatable utterances are untranslatable not because of an inherent feature of their host language, but because language and culture are also deeply intertwined. If in Russian I made a joke about "preved" or discussed samizdat, and you didn't understand it in translation, it wouldn't be because of any inherent differences in the mental processes of Russian-speakers versus English-speakers, but instead because of cultural differences, which are an incredibly important part of communication.

This is my two cents, I suppose—the Sapir-Whorf hypothesis is a seductive idea and it's produced a lot of compelling research, but it's very easy to go too far with Sapir-Whorf and turn language into something much more powerful and more magical than it really is.


Interesting:

    In recent studies, MIT students were shown dots on a
    screen and asked to say how many there were. If they were
    allowed to count normally, they did great. [...]
    if they did a verbal task when shown the dots—like
    repeating the words spoken in a news report—their
    counting fell apart. In other words, they needed their
    language skills to count.
I find that external tapping destroys my ability to perform or enjoy listening to music. Something about the interaction of a rival rhythm shuts down the part of my brain that handles music. But most other singers I know don't have any issue with this, and I've wondered for a while what's wrong with me.

An important aspect the article doesn't get into is idiomatic languages. English has a rich vocabulary. Some languages with weak vocabulary compensate for this by heavy linking into a story culture. It would be like if we had no word for beautiful, and instead expressed this by saying "your face launches a thousand ships". I'm told Chinese dialects are like this.


If, when my wife tries to wake me up in the morning, I grunt in reply I can usually fall back asleep. But if I respond using actual words, I'm generally awake for the day. My hunch is that this phenomenon is somehow related to the study.


Very true. I have worked in 15 languages and each one colours the world differently. It does become a bit of a chicken-and-egg (or nature vs nurture) question, though.

I definitely agree on Chomsky. Common structures and dynamics don't necessitate a quasi-Platonic unity.


> In one study, Spanish and Japanese speakers couldn't remember the agents of accidental events as adeptly as English speakers could. Why? In Spanish and Japanese, the agent of causality is dropped: "The vase broke itself," rather than "John broke the vase."

If the author translated "se le rompío el vaso" then it is not quite "the vase broke itself". The "le" is very important: it indicates who did the breaking, albeit indirectly and indicating an accident. I would translate it as "the vase broke on John" (it broke while he was using it.)

If the author translated "se rompío el vaso", still, "the vase broke itself" is too literal, just as "se venda" is not translated as "it sells itself" (it means "for sale.") "The vase broke" sounds better.


Oftentimes, but not always. If John picked it up and it just broke (due to being fragile, I suppose) he could just as reasonably say "se me rompió" (note where the accent goes, by the way) and we might translate that is "it broke on me".

(Also: it's "se vende".)


"Take "Humpty Dumpty sat on a..." Even this snippet of a nursery rhyme reveals how much languages can differ from one another. In English, we have to mark the verb for tense; in this case, we say "sat" rather than "sit." In Indonesian you need not (in fact, you can't) change the verb to mark tense."

It seems that in many languages verb tense plops the "when" into one of a few or dozen general buckets. the "when" is then further specified using more words, such as "yesterday" or "next month" or "a long time ago".

I expect Indonesian to simply rely entirely on such "further specification". Or is the sense of time fundamentally different than, say, English? Does anyone know more?


yes and no

In Dutch (technically not too far from English) we only use past and present tense, and although we can mark the future with a special verb like in English, we often don't do that. It is perfectly possible to mark the time with a time marker then, but often this is not necessary to be understood. Relative time can be enough to be clear for the other. (Like "I am short of bread. So I go to the market and buy some. Then I eat it." is grammatically incorrect English, but easily understood, and a correct formulation in Dutch.)

Esperanto has past, present and future and has the possiblity to note the completeness of an action, but it's not obligatory as in English.

toki pona has no inflection and doesn't indicate time at all. If you want to indicate time, you end up using a complicated grammatical construction of causation. Mostly you just make a general statement or indicate time as something coming before or after something else.


I've been wondering for some time why some languages 'mark' (eg, verb tenses, declining nouns, adjectives) more/less than others.

Over time some mark more and others mark less, so what is the trend? How does this stuff arise??


Well, Spanish marks subjects of verbs with inflection and requires detailed accounting of objects with pronouns and clitics. In return Spanish gets the opportunity to build a subtle set of emphasis points in sentences.

The final position in a Spanish sentence may belong to the subject, an object of any kind, the verb, or a subordinate clause equally easily. Which thing gets that honored position becomes the topic of emphasis for the sentence. Then whichever one comes first (also flexible) gets the second strongest mark. Things that come along in the middle can get much smaller emphasis depending on order. Adjectives and relative pronoun phrases can change emphasis based on order, too.

In Latin, there is almost no natural order needed. Nouns specify their function in the sentence with declension and verbs attach to subjects with their conjugations. Latin poetry can hold a soup of words in almost any order.

In English and Mandarin order matters. But there is little chance to mark the function of words morphologically. Instead the words stay the same and you rearrange them. That means you can't reorder sentences like you could in Spanish or Latin.


I agree that language influences way of thinking. In Russian language words (verbs, nouns, etc.) often can freely swap their position in sentence without changing meaning of the sentence (just slightly changing emphasis). Questions usually are made using intonation only. English has very rigid structure when verbs, nouns and other parts must be put in a certain order. This results in clear, sharp and unambiguous way of expressing thoughts typical for english-speakers, that always amazes me (native russian).


FTA: "Languages, of course, are human creations, tools we invent and hone to suit our needs. Simply showing that speakers of different languages think differently doesn't tell us whether it's language that shapes thought or the other way around."

Umm.... I disagree. English was around well before I was, and I haven't known anything but it, so to say that I am effecting the language more than it is effecting me seem inherently incorrect.


This was interesting except for the pointless Chomsky bashing. The fact that languages are different does not mean that they do not share a universal grammar. Languages can be very different and yet still share some underlying rules.

This article was written on a very different conceptual level than the universal grammar theories so it is really pointless to even mention those theories.


Bashing?

Dr. Chomsky proposed that there is a universal grammar for all human languages - essentially, that languages don't really differ from one another in significant ways. And because languages didn't differ from one another, the theory went, it made no sense to ask whether linguistic differences led to differences in thinking.

I don't see that as bashing. The author is just explaining the chain of reasoning.


It's actually an oversimplified (biased?) view of Chomsky's theories. Chomsky stated that language acquisition relies upon innate abilities and that all languages share some basic "structures" (the theory dates back from times when structuralism was all the rage). The theory has been proven false or inaccurate in many details, but the driving concept is definitely true.


Heinlein covered this subject in his book 'Stranger in a strange land'


This is actually true. Some languages use passive pronouns to tell a story while English mostly uses pronouns in the active format. That alone can transpose a meaning to a whole different level. While English does have a wide array of lingo to explain ideas and circumstances; languages such as that of France's can intricately indicate the gender modality of a sentence by the linking of pronouns and verbs.

Interesting, indeed.


This is utter gibberish.


Kind of odd I got minus points; because some of us speak more than 3 languages. Oh well, lesson learned =)




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: