Ask HN: Do you think primarily in English? | Hacker News Reader

Ask HN: Do you think primarily in English?

56 pointspalish85 comments

Rather, do you think primarily in your native language, or more in terms of abstract imagery?

I've noticed that rather than forming images in my head to represent ideas, I will instead often think in English. For example, if I'm working through a logic problem, I will repeat the various elements of the problem in my head in English repeatedly until I solve it.

I'm unfortunately unilingual. English is the only human language I have ever learned. I've noticed that when I speak phrases in Spanish, I'm actually speaking English in my head. It's like a lookup table. By that I mean, let's say I want to communicate "How are you?". Rather than attempting to reason through this task in Spanish (by trying to communicate the individual elements "how", 'in what way', "are", 'your state of being', and "you" 'the person I am communicating with'), I try to remember the phrase in Spanish that corresponds to the ordered set of words "How are you?". So it is rote memorization -- I'm not thinking in Spanish, I'm translating from English to Spanish, speaking it, then hearing a response and translating that back into English. No wonder I was never able to learn another language. There's always this complicated layer of indirection.

Anyway, back to the topic: do you find that you think more in terms of words or of images?

Here's one reason I can't imagine thinking through a problem without reciting it to myself in English several times. Let's say for example I need to describe the concept 'empathy', which is to say, understanding another person's situation by virtue of having been in a similar one yourself. I'd imagine it would be very difficult to conjure up an image which succinctly describes empathy. It seems like the verbal encoding "empathy" is probably the most succinct way to describe it. So why try to think in terms of images when our native languages are so very precise?

And yet, "thinking in English" can be a burden. I can feel how it limits me to thinking through problems in certain predefined ways, just as a given editor forces you to edit text in certain predefined ways. I would imagine that most extremely creative people (such as DaVinci, Tesla, etc) probably don't have this limitation imposed on their neocortex.

More importantly, my reading speed is significantly degraded. I tend to recite each sentence to myself as I read it. I've been unable to find much information about how to break this limitation, so any references would be very appreciated.

Thoughts?

View on HN

I think in 6502 op-codes. That's probably why many people think I'm 'emotionally distant'.

People who are fluent in a foreign language actually learn to think in that language as well. They switch back and forth between modes of thought. This is why they can rattle off phrases in the language rather than speak it in the very broken manner that the less fortunate (myself included) would.

I'm Dutch and I often find myself thinking in English. If I work in English all day I context switch and English becomes my "default" language. When I wake up the next day my first thoughts will then be in English. Until I consciously switch back I'll continue to function in en_us mode.

It can be really annoying at times. I constantly mix up proverbs and language constructs when I'm context switching, I often know a word in only one of the two languages and can't think of a good translation, and so forth. And more often than not there is no 1-to-1 correspondence between English and Dutch words, so it's easy to get frustrated knowing exactly what you want to say, but being unable to say it because the words don't exist in your native language.

I don't think it matters much whether you think only in English, in several languages, or in pictures/emotions. A game I sometimes played as a kid was to take a concept and repeatedly think of more succinct definitions for it with a friend. And the cool thing is that despite getting the original definition wrong, in the end you always end up with a definition both parties agree on. People rarely think about the exact definition of any every-day word, and yet people are capable of intuitively picking the "right" definition from a list.

To completely derail this thread I'll challenge you to define the word "chair". It's not "something you sit on", because you can side on a couch and a couch is not a chair. Nor is it "something you sit on with legs", because then it could be either a stool or a bench.

The point is, when you think of the word "chair", you're not thinking in English. You're thinking about the concept chair, and that's one you can't define without some effort. And defining it in words is completely unnecessary because you know what a chair is. It's no different for the word "empathy".

ps: Your definition empathy doesn't pass the sociopath test. A sociopath notices somebody else is hurting, and knows what pain feels like, which is for your definition sufficient. But textbook sociopaths lack empathy. So empathy must be more than mere recognition of an emotion. Yes, this is nitpicking, but that's kind of the point.

I'm Dutch as well and with work (software dev) and most of the stuff I read (HN, books) mostly in English, the language of my thoughts tends to follow the language of my current activity.

Whenever I'm speaking Dutch and an English word better conveys what I'm trying to say, I just use the English word. Most people barely notice. For extra fun use German and French words too!

Nobody thinks in English or any language. I know this, because I actually thought about the problem you are describing when I was a child of about 8 - I was faced with the problem because some guy said he thinks in words. I realised that I didn't think in words, but in concepts.

Later, when I was much older, I realised that parts of my thinking seemed to be words - but just seemed. With some effort, it's possible to remove the words from the thinking and just think in concepts - the words are automatically placed on your thoughts, but you are not thinking in them.

Let me explain again : You have a thought, and then the words matching that thought are created. Then you reprocess those words as input. You can also skip the reprocessing step, but for that you have to avoid a fatal mistake - the replacement for words is not images. Rather, you need to simply not try to inteprete the concepts you are working with in your head.

For example, if I say 'empathy', you know what it means without a dictionary definition of empathy. If I write out a sentence - the man felt empathy with his begging son, you can imagine the situation without needing words or pictures, and only when the entire concept is finished do you convert it into words. I.e, in the sentence above, you are not breaking empathy down, you are breaking the entire sentence down.

There is a simple way to discover how to do this - think of some random object. Then think of something related to that object, then something else related to it. For example Book, Page, Letters, Reader. And so on. Just go through a list of interconnected things, but as you do so, try actively to avoid putting words on the concepts. After a while, also try not putting pictures either. You should see your thinking speed up significantly as you improve.

Nobody thinks in English or any language.

That's a pretty bold statement that many eminent linguists and psychologists would disagree with.

From my limited knowledge of linguistics what I've gathered is that current thoughts are that people think in symbols, which are often words, and that our abilities to think in many types of abstract thought are directly connected to the language faculty of the brain. Modern linguistics is driving towards there being a fundamental underlying universal grammar in which human languages can be seen as specializations of.

The unification of cognitive psychology and linguistics seems to be one of the goals at the moment, and there is the notion that abstract thought and language aren't fundamentally different things.

Alright, what about a child who sees cars everyday, but has never been told what a car is. He can still manipulate the car concept as easily as if he knew the word for it.

Or some people have very small vocabularies, simply because they have been exposed to a small number of words - does that affect their cognitive skills?

Abstract though and language may not be fundamental different things, I don't have enough information to say. But what I am saying in my post is that first you think in a concept, then you translate it into a word. You don't think in the word first. So you can think without needing to translate it into words. The relationship between the words and the concepts is of course strong, because we use words to vocalize the concepts we have.

Think about it - what is a word actually? It's just a series of sounds we create in the air. So the sound of a word is quite irrelevant to the concept behind it - the construction of the word is not relevant. Imagine we did not have mouths but instead we communicated by flashing a series of pictures in the air. The part of us that creates these pictures is just a sensory IO appliance like our eyes or mouth. Would that change us in any fundamental way? No - we would construct a similar grammar and communication method with pictures as we would with words.

What this implies is that the words themselves are unimportant, they are just representations of concepts we are dealing with. This representation can take many forms - but because we are used to vocalising our thoughts, when we do indeed think, we automatically prepare the things we are dealing with for vocalisation, leading to the illusion that words are actually involved.

What happens this is that this vocalisation becomes a sort of feedback loop for your thoughts - you think of concepts, convert them into words, then reabsorb the words to be reprocessed as if someone told you them. That is, they come back in as if you were listening to them from other people. You can observe this sometimes with people when they mutter to themselves. When we rehear our own thoughts, we can sometimes process them better (for reasons I don't know). But this step is optional, one can also do it without reprocessing, though this will fail with very abstract things that actually need concepts alien to us.

You might find this interesting:

http://en.wikipedia.org/wiki/Pirahã_language#Pirah.C3.A3_and...

A lot of what you're saying is intuitive, but modern research suggests that it's wrong.

That could be, but if everybody always stuck with what the current scientific dogma is, we would still think the earth were flat.

Likewise, if people ignored scientific evidence in favor of their intuition we would still think the earth was flat.

It turns out we're all better off if we first understand the "current scientific dogma" and then challenge it with scientific evidence.

Well, but the way that people proved that the world was round was with evidence, not anecdotes. I'm willing to admit that current theories could be wrong, but not based on a thought experiment.

All snarkiness aside, I see you're in Berlin, so you're welcome to drop in for dinner on Wednesday. I'm slightly less bull-headed in person. :-)

http://news.ycombinator.com/item?id=388042

Alright, what about a child who sees cars everyday, but has never been told what a car is.

If a child can see cars but there's no one there to name them for the child, the child is being neglected... and perhaps even in physical danger.

And haven't you observed children name things ad hoc by the sounds they make?

So I don't buy the idea that a child "can still manipulate the car concept as easily as if he knew the word for it".

Also, yes, people with "very small vocabularies... because they have been exposed to a small number of words" do have their cognitive skills affected, usually for the worse.

Haven't ever happen to you that you have the concept in your mind but you are not able to remember or even know the word for it? And nevertheless you are able to think about it!

How can you not attach a picture to a word or concept? And if you do not attach a picture to something - what do you see?

Tennis.

So, did you see a racket, people playing tennis or something else as you read the word? And if not, what did happen as you skimmed from paragraph to paragraph?

edit:

Hmm. I just realized that when I read a list of words [ball, car, frisbee] I don't visualize anything, but when I think of any of the words individually I do visualize. Even when just writing down a simple list. That's peculiar.

edit 2:

It can only mean one thing - that when reading you don't think at all about the nouns. Obvious in retrospect. This also explains why you immediately forget a list when just reading over it, but recall perfectly when you visualize each item for an instant.

edit 3:

> Just go through a list of interconnected things

For most people this should be really difficult when listening to music with lyrics, and really easy in a quiet environment. "proving" that you're still using the linguistic part of your brain, even when you're painting a mental picture.

When a word or concept 'ends', you will translate it. Before it ends, you can can deal with it without converting it into a picture or a word.

For example, the sentence: "there is a lion coming". How do you deal with the sentence. Do you imagine a lion coming? If so, where is the lion coming to? To me or to you? What's the context?

There are too many variables, so it's difficult to convert that sentence directly into a picture - but you still understand what it means. It's just a fragment, but it represents some type of mental concept, and this mental concept could be related to danger. As soon as I can complete the fragment:

"There is a lion coming up behind your desk", you can immediately convert it away from a a mental concept into pictures or words.

That's remarkable. This is exactly the sort of information I hoped I would be fortunate enough to receive. I want to say "write more", but I'm not sure there is anything left to say. But you should definitely write a blog post about this, if you're interested.

I find it interesting that you discovered this at such a young age. Maybe that's one reason it's so natural for you not to think in a language.

By the way, roughly how fast do you read? (Extremely fast, average, very slow, etc.)

Native and non-native (even if fluent) speakers recruit different areas of the brain when speaking or comprehending a language. This is now well known from fMRI studies. http://www.sciencedirect.com/science?_ob=ArticleURL&_udi... : From the abstract (Listening to comprehensive but non-native language seems to demand more networked co-processing.) Another one: http://www.sciencedirect.com/science?_ob=ArticleURL&_udi...

Non-native speakers recruit more brain regions, suggesting that more "concepts/whatever" are being invoked in the formulation or parsing of a sentence. With this in mind, it would seem that even people who assume they are thinking in a certain language, are only accessing a post-conceptual process (language formulation). Conceptual relations between objects or other concepts may still have been put in place in a ways influenced by the language in which you picked them up.

For the same reasons as above, verbal encoding of empathy is probably not the succinct representation the brain uses; you just happen to have conscious access to the verbal encoding only. I would guess though, that something like the Implicit Association Test would uncover more. http://en.wikipedia.org/wiki/Implicit_Association_Test

As for your Spanish, you will stop using the English-to-Spanish pre-processing crutch once you get more fluent (with practice and repetition).

I was raised in Central America by US parents. At 4, I knew more Spanish than English. We moved back to the US when I was 14. Since then, I've lost quite a bit of my Spanish, but I'm still fluent.

When I was younger, I would think in Spanish and occasionally dream in Spanish. I don't do that any more. I've been in the US for so long that the transition to thinking in Spanish requires effort.

But, yes, there are concepts and shades of meaning in one language that are hard to communicate in the other.

I think that being monocultural is much more constraining than being monolingual, however. When you grow up between separate cultures, you're not only acutely aware of multiple ways of talking. You're also acutely aware that there are multiple ways of living, of relating to people and family, much different sets of priorities, completely different ways of problem solving and of being. It's hard to appreciate unless you've made a big cultural transition like emigrating or growing up multicultural.

If you want to get past it, move to another country for a couple of years.

Your Spanish may be a bit rusty. But I think you will be able to become fluent very fast if you had a chance to practice more. edit:grammar

I too think is English. It's the only language I'm fluent in so that much is natural. One interesting thing I've noticed is that when I'm conceptualizing something, I will have an idea in my head, but out of some bizarre habit I will think the words through.

My point is that I already know what I'm going to say in my mind, but there is no reason to actually hear my brain think them. Basically I go through 3 steps in my thought process:

1. Think or conceptualize,

2. recite what I've thought to myself (in my mind, in English of course),

3. process what I've just heard myself think and act on it.

My question is this: why is step 2 necessary? It would be faster just going from conceptualization to processing, wouldn't it? Maybe it's a habit from reading at a young age when I sound the words out in my mind as I read.

Or perhaps putting it in words makes you feel confident enough about the conceptualization to proceed with action. I often find myself doing far too much "planning" and "checking for possible errors/pitfalls" before taking action on something I have conceptualized (probably in part due to the perfectionist in me).

In case you are unaware, the question of how much of our thinking is language-constrained has been extensively studied in linguistics and cognitive science. http://en.wikipedia.org/wiki/Sapir-Whorf_hypothesis is a good starting point. Anecdotal evidence is great, but you should also look into the studies.

My personal perspective: I am natively trilingual, as an Indian person. (someone else mentioned this as well.) My two Indian languages have fallen to near-complete disuse, to the point that I'm only borderline fluent in them. Nevertheless, growing up with two or more languages forces a child to seek common denominators, and thus think in terms of imagery.

This is definitely the case for me. As soon as I focused on the word "empathy" for a few seconds, I actually had two different images come to mind. These were both exemplars, as in, specific instances of people in my life showing empathy, one from recent memory and one from long ago. That's the trick with imagery -- it can often be crude, and only cover a special case of the concept, but it still does the job much better than words. You mention precision. Precision can in fact be a disadvantage, given the inherent ambiguity of human thought.

This is learnable. It gets harder as you age, but never impossible. To improve thought imagery, try this. Buy a ginormous whiteboard and cover one wall of your room/office with it. Draw everything. You might initially have to struggle with coming up with any sort of image. Your perfectionism, and your poor perception of your drawing skills, if you suffer from those, might hinder you. Try and lose your inhibitions. When you're at your computer, keep looking at the whiteboard once in a while. You want to get to the point where if there's a concept you're working on over a few days, every time you think of that concept, you should immediately be able to see the corresponding picture in your mind's eye. Also keep reams of paper around. Again, draw everything. I'll even start drawing on napkins if a thought comes into my head that I need to draw and I can't find paper. Hopefully, you will eventually be able to 'draw' effortlessly in your head.

Here's a solution for your "lookup table" problem. This is a technique that is used in teaching accents, but I have also used it in learning new languages. Pick a character. It could be a friend, or a character from a TV show, just anyone who speaks Spanish (preferably exclusively). When you're trying to speak Spanish, be that person. Imagine you're them, to the extent possible. It will make it a lot easier to push the English out of your head.

Finally, subvocalization (i.e, reciting sentences to yourself). Also (un)-learnable. Start gliding your eyes over the text rapidly. At first, you won't understand anything. Try the same sentence again. Force yourself not to say the words. If it's too hard, go fast enough that you can't possibly say the words. Keep doing that until you derive at least some meaning from the sentence. Start from there, and hopefully you'll get better slowly.

These are all things I've used to various degrees. Thinking in English is not a big problem for me, but more imagery helps, no matter what. YMMV, but hope this all helps!

Not a linguist, but I took some classes ;)

I think the Sapir-Whorf hypothesis is a pretty imperfect fit for the question. Even the weaker form doesn't really gel well. The question seems more do certain modes of thought make you tend to think of certain languages, and not so much "do certain languages impact behavior".

Basically, its generally formulated in the opposite manner, and like many things, cause and effect can not usually be reversed and still have the reasoning hold ;)

Also, I grew up effectively trilingual as well. I simply associate the different languages with different areas of thought. Math/science is English, food/home-y things is Chinese (and er, well, Taiwanese+Japanese also food/home), and I really don't think in terms of imagery. So I think you're definitely in anecdote territory. (BTW: I actually count in Chinese, because its a tad more uniform and the words are shorter, but operators I use English.)

I wonder if language learning methods will vary from person to person, because most of your advice totally disagrees with my experiences learning Spanish later in life. I learnt through rote memorization, but followed up with a conversation partner, so, to continue the lookup table analogy, I simply used the lookup table until it became so commonly used that it stayed in cache and not memory. Then over time, segments were pulled out and their structure transformed (length of cache value increased, keys switched to a general emotion), until more and more of it was "fluid".

(The other possibility, which I'm very inclined to believe, is that describing your own learning process is a crap shoot. You find analogies you're pleased with, and selection bias takes over.)

I take a medium-strong form of S-W. It makes sense in the connectionist model (which I also take a medium-strong interpretation).

The strongest form says that concepts in one language are completely alien to another, such that if you don't know the concepts in one language, you cannot perceive them. This is obviously false.

However, as related concepts are evoked by spread-activation, it is completely plausible that one language allows more efficient retrieval of certain concepts, such that very strong biases result.

The famous example of the many Inuit words for snow comes to mind. It is true that a non-Inuit speaker is able to discern between different kinds of snow, but the language guides the mind in terms of where to look for details. Without this guide, it is easy to get lost and confused.

Here's the familiar example for hackers: code in C/Basic/Pascal for a while, and switch to Python/Ruby (or vice versa). After you get used to the target language, and look back at the code right after the switch, you'll notice that a lot of programming constructs are out-of-place and "counterintuitive." This is, so to speak, Sapir-Whorf "at work."

This is also why when LISPers say that you should learn lisp, because it will make you a better programmer, they are correct. :)

The comparison with programming languages is extremely misleading, because, unlike natural languages, programming languages do differ in the expressive power. An algorithm coded in Haskell or Lisp may well be 5x less code than an equivalent thing in C or Java. On the other hand, when a text is translated from one natural language to another the result is typically of the same length and typically there's even an almost one-to-one mapping between individual sentences.

To anyone interested in the relations between the language and thinking, I recommend an excellent book "The Stuff of Thought" by Steven Pinker. One chapter is devoted to the Sapir-Whorf hypothesis, more specifically, to debunking it.

Two things: again, I don't espouse the strongest form of Sapir Whorf. The interpretation of differences in speed of access to concepts, based on spread-activation, however, is solid. When I learned Sapir Whorf we were given the original interpretation and a brief of how it has been reinterpreted more recently -- but it still goes by the name of Sapir Whorf. I am not making a case for the possibility of expression. I am make a case for the salience of expressions in different languages. It should be easy to show this: take two languages and find two words of identical meaning, and measure the response speed to comprehension of these words. All other things equal, the higher-frequency word will be accessed faster. Hence, if one culture says Apple more than Orange, Apple is more salient in the former culture. You will then expect more efficient access to Apple-related concepts.

Second, what makes you think that natural languages do not differ in expressive power? Certain things are more easily said in one language than another, just as some constructs may be easier to represent in one language or another, although they may be able to create the same results.

I also don't know where you get the "typically the same length" from. Stephen Pinker wrote about the relative length of English in one of his books. I don't remember which he was comparing to, but English is longer. Pinker explained this along the lines of English being more "fault tolerant" due to more "redundancy" (my words).

This is probably more a function of the fact that languages change over time, so if a common task is hard to talk about, new forms will appear that make it easier. Programming languages see very little change (and even rarer are ambiguity or "non backwards compatible" changes introduced.)

Stuff like this is why I became interested in PL theory in the first place actually.

Wasn't Inuit snow words story just an urban legend?

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

My fault for not confirming the veracity of that story in particular, but I was using it as an illustration of accessibility to concepts in different language, independent of its truth. I just supposed that this story is widely known, but I could have (perhaps should have) easily used another example. There are several words in Chinese that describe taste without direct English translations that I know of (it is harder to sustain these ideas nowadays, given the quickness of word-borrowing like "umami"). One of them is "fish stink," which really isn't fish stink. A strong S-W will say the concept is inaccessible to the non-Chinese speaker, and, of course, that isn't true.

In fact you don't even need to cross language boundaries to see the same effect: layman versus technical terminology reveals analogous differences in "perception." Me not knowing the name of industry colors does not mean I cannot perceive these colors (let's assume a single-sex population), but I will gladly group together large groups of colors into single words.

I suggest you read the wikipedia article. Theres a pretty good summary of its current status. "well, dub it has some effect, but it doesn't make certain thought impossible"

A better example is that some languages have more words for color, and thus people are able to differentiate between slightly different more shades. Though I'd argue thats more a function of training than the mere existence of more word.

Strangely, for me, in imagining empathy, all that comes to mind for me, is an abstract understanding of it. I can "feel" its meanings, almost in the sense that the meaning flows through my fingertips, but I don't experience imagery or episodes or sounds.

Similarly, when I need to use an abstract word, I translate these feelings into words. Often times, possibly as a result, I come up with strange words that fit the context but in an unconventional manner.

Nice post!

While this probably sounds a bit strange, babbling in freefrom gibberish and stretching your mouth / jaw muscles can also help to relax sub-vocalization. It seems to relax an underlying vocal fidgetiness the same way a good stretch/workout does for the rest of the body.

In other words, additionally approaching sub-vocalization as a physical phenomenon might be helpful.

Steven Pinker pretty convincingly demolished S-W (for me) in "The Stuff of Thought". He argues that there is an abstract linguistic framework upon which the concepts in a language are built, things like content-locatives, container-locatives, causatives, and much more. Recommended book.

Are you familiar with the literature dealing with subvocalization and reading? Any recommendations?

Sure, ask this while I'm sitting here with a Linguistics text studying for a final. This post is going to be a bit long.

Empathy is a hard word to bring up an image of because it's meaning is largely connotative rather than denotative. On the other hand, there are many words ('Book', 'Clarinet', 'Web Browser', and all the other things I can see from where I'm sitting) for which it's simple to form an image. The words are largely symbols which denote an entity.

Now to drift from the topic:

I think in either my native language (English), or in images. Flashes of insight tend to come as images. Generally the hardest problems I work on are design situations rather than algorithmic, so I think images are the best way for fully formed answers to pop into my brain.

I'm also unilingual. Canadian English is the only natural language where I regard myself as being fluent. I speak some French (Quebecois), but by and large it's parroting full phrases when I speak. My reading comprehension is sufficient that I can interpret arbitrary French when the vocabulary is in my limited lexicon, however my grammar is quite weak when it comes to forming new phrases. My formal education in French was poor, as I was taught essentially arbitrary vocabulary by non-fluent teachers. All through high school I read French in much the same way as you speak Spanish, by treating another language as something that I was just lacking the mental dictionary for.

I lived in southern Quebec for a couple of years while attending University. Though it was an English town and school, a fantastic amount of the signage was in French.

Some time ago a friend and I were driving through Montreal. More accurately he was driving and freaking out about the traffic, I was navigating and translating the road signs for him. Having been exposed to French road signs for a couple years I'm competent in that area. I realized that I wasn't translating into English in my head when I started getting frustrated at how _slow_ it was to translate the signs to him. I would read them, have to find the equivalent concept in English, and then say it. In the end I just started directing him where to turn instead of reading him individual signs. The layer of indirection doesn't last forever. One day I'd like to be able to speak the language like I can read the road signs.

My view is that people think in (1) words (2) pictures and (3) concepts. How much of each depends upon the individual.

As for myself, I think mostly in pictures and concepts, very rarely in words.

I am bilingual, but since I rarely use words for thinking, languages don't matter as much.

I completely agree with this -- I am bilingual too, but the words seem to spring out of the flow of thought, not to "be" the flow. And when I'm learning a new language (or enough rudiments not to be impolite when visiting, say, Lisbon) I never have this "translation table" issue the OP mentions -- at least, I don't think I do. There is no transitional linguistic step between the thought-flow and the words, new or not. (Of course, that doesn't mean that language pickup is necessarily easier -- just different.)

I'm actually learning Russian right now, while living in St. Petersburg, so I think about this a lot. The program I'm in is intended to take us to a high level of fluency-- Superior, or Level 3-4 on the various scales. As a result, there is a lot of talk about thinking in the target language. I, sadly, still feel like I very rarely do this. I can handle very simple contexts rapidly, but I don't know if I'm actually thinking directly in Russian, or simply using the lookup table you describe at an extremely rapid rate. When I read texts or hear other people speaking in Russian, I almost always still reform the sentences into English in my head.

Some of the other students seem to not have this problem. By their own admission, they're able to operate in many, if not most, contexts completely in Russian-- though, of course, there's no way to know what this means. They mention dreaming in Russian, for instance. This hasn't happened to me yet. There are times when I can intuit what someone wants without actually translating word for word, but usually that's just because I didn't catch all of their words to begin with.

It might be background-- I'm the only engineer, the others are all Russian majors. Maybe they have less need of precise rules and translations. It's very hard for me to adopt a new Russian phrase without having found an English equivalent for it. It might be time spent learning the language. The other students have been at it 2, 3 times longer than I have. It might just be that I'm older: 30, as opposed to 22, 23. When I was in the Peace Corps, the volunteers in their 50s and 60s had huge problems learning the local language. So it's probably true that, at least for most people, the older you get, the more ingrained your modes of thought. I guess that means that the earlier you can expand this, the better.

I think in the language that I am using or have experience on the subject - my native language is Bulgarian, but sometimes I think in an another language when I face a concept I experienced in another language. Those cannot be only natural languages but also formal ones - its too confusing and slow to translate mathematical, psychics, programming or any notation into a natural language.

I sometimes only know English terminology on a subject and I am forced to think and even use English words.. I consciously try to avoid that. Usually people with only ostensible knowledge of English AND the subject pollute their native language with unnecessary foreign words. Unfortunately those people are very common here and annoy me daily, but on the other hand its fairly easy to separate the signal from the noise by looking at how often such fancy words are used.

I really recommend this essay in which the unnecessary use of foreign words is only part of a bigger problem: George Orwell, "Politics and the English Language," 1946

English is my primary language, but I also picked up a fair bit of German and Danish as a child from my grandparents.

I mostly think in English when programming, but when I get into either specific relationships between elements of the program or high level conceptualization I tend to slip into German. Names for objects often come out in Danish if I don't stop and translate them into English.

In college I had some unfortunate partners on a programming project who were subjected to code written late at night and very quickly with comments in both English and German and with many Danish identifiers. Fortunately they were up to the task, and as they were Italian and Norwegian with English as a second language, they gave as good as they got with the multi-lingual coding.

Our professor requested a full English translation to accompany our source when we turned it in.

In my experience I think in English 90% of the time. When working problems requiring a lot of spatial insight, I pretty easily switch to something more visual and visceral. This also happens when I'm drawing and sometimes when I'm playing music.

I'm also learning Chinese. It is anything but an easy language to learn, I'll say, and I don't feel like I'll be thinking fluently in it anytime soon. That being said, if I concentrate and force myself to think in Chinese I can get by with a huge hit to rate of thought. It's something I practice from time to time despite the frustration.

It depends. English is my third lamguage. French is the second. Wolof the first. I think in all three depending on the situation. When I first moved to the States I would speak in French in my head and translate it in English from brain to mouth. As the years went by, I started to automatically think in English. When I speak with my mom I think and process all the info in Wolof. If the conversation becomes more political or academic with her, then I start to process the information in French because that is the administrative language where I come from.

context: Most educated Indians are trilingual (the provincial language - there are more than 20 of these, with a few hundred distinct dialects- , Hindi (the national language), and English. I can read and write French and German but I don't really think in these.

I find that the "voice in my head" speaks different languages at different points in time. While programming I use mostly English. I find that sometimes when stuck on some programming problem, trying to explain/talk about it in another language often "unsticks" it. No idea why, really.

Hindi is not THE national language. It is "A" national language.

Hindi is just the language used for the transacting business of the Central government, for which English is a constitutionally and legally acceptable alternative.

Coming back to the topic, I think in English most of the time, just because a majority of my waking hours are spent thinking about tech and programming, for which English is a natural fit.

Note: I'm trilingual (actually more like 3 + 0.5 + 0.1 + 0.1 + ..., the fractions being other south indian languages).

I associate certain activities with certain languages. Food -> Chinese. Programming -> English.

At least for me, the language that people usually speak to me in when I do that task or learn that task is usually what sets the language I think in. I have somewhat similar experiences with music (certain music makes me want to code), and even handedness (I yo-yo left-handed because thats how I learnt it)

An amusing side effect of that is that though my native language is English, after 7 years of living in Germany, if I've been drinking I have trouble forcing myself to switch into English. I catch myself even speaking German to friends who speak English better than German. Since I'm so used to "social activities are in German" I can't shake it.

That's interesting -- I never thought of it in terms of context, but I had similar issues when I returned to the states after 7 years in Germany -- I would go into shops and order things in German, because my "talk to shop people in German" bit didn't turn off right away. Some confused/amusing looks when you say "Ich moechte gern ein Hamburger, bitte" in a restaurant; and it takes a few months to switch back.

I think is is a perfect example to explain how our brain stores memories and information (images, ideas, concepts, new things, etc). I'm not bilingual (I only speak English) but from reading everyone's comments on this page a clear pattern emerged:

To your brain, language is just another piece of information, normally used for expressing ideas, concepts, or new things you have learned. If you learn about a new thing while you're living in a country speaking the native language, your brain stores it together with the associated language, which you will later use to express that thing, even if you're expressing it after you've somewhat forgotten the language it was learned in.

This very clearly explains why most programmers will "think" in English when they're programming -- the majority of documentation and code is written in English and when they learn programming concepts, the concepts are stored in the brain along with the English language used to express it.

Our brain is constantly looking for relationships. I notice whenever I play a game on the computer, I instantly feel hungry -- even if I just ate. My brain associates "eating" with the activity "gaming" and tells my body it's hungry. In the exact same way, the brain stores the associated language with the new concept or thing when it is being learned. Taming and controlling this functionality of the human brain is very powerful and I think it's the basis of all habits and addiction. (I say "I think" because I have absolutely no formal education on this topic.)

This entire subject fascinates me as I have always wondered why learning a new language seemed so difficult to me. I often catch myself reading aloud in my head and quickly realize how much it's slowing me down. When I switch to "absorbing" the words instead of repeating them in my head, my reading speed instantly increases two-fold.

I'm going to write a post on my blog compiling and summarizing all the ideas and discussion on this page, so feel free to look for it on raamdev.com if you're interested.

I find that I use English for everything except 2 major classes of thoughts:

1. Art. Specifically, if I'm trying to compare some work of art (music/movie/art), I have a virtual "media player/audiovisual search engine" in my mind that quickly lets me locate other works that are similar, related, or otherwise relevant. This similarity is often on the basis of the emotional content of the work, but I find that I don't need to put the concept into words to locate it. To use your example, if I'm looking at something that provokes empathy, I'll think of other works that provoke the same emotion, but without ever thinking of the word 'empathy' (or any synonyms/etc.).

2. Science. If I'm trying to understand some scientific concept or if I'm working on something new (I'm a researcher), I find that I make almost no progress unless I can form a mental picture of the problem. Only after creating this picture can I do anything. Then most frequently I'll imagine myself "flying" through this mental model, tweaking it as I take into account various other factors, sometimes destroying parts and rebuilding, etc. Usually, I will come back to this same model over and over until finally I understand the concept or come up with some new realization.

I speak english fairly well, I think... and have an IELTS certificate, so in theory I could even teach it. This is strange to say here, since most of you either speaks english as your primary language or also had several years of exposure to it.

Anyway, I can think in English. I told this to a co-worker and he mocked me, "hahaha you're so full of yourself", but he asked a friend, an english teacher, and she told him that she also does it.

My guess is that this s something that comes with usage. I read a lot of stuff written in english. Heck, sometimes I know the english term but not the portuguese one, something specially true for things related to my profession. Sometimes I am not sure how to spell an word in english, or even if it exists, and I google for it, and the first result is from a dictionary... meaning the word isn't even used that much.

Here's an anecdote: I have a friend who is a musician and works mixing movies. Once I told him that I can't think, or make sounds in my head, using multiple tracks. By that I mean, when I think about a drum sound, and try to imagine a guitar riff with it, the drum always "stop". The notes don't intercalate. And well... he told me that he can, which is way cool.

Just practice more Spanish.

I am a Spanish native speaker who later learned English, French and Italian for work. In my case, I don't have the feeling of translating sentences/ideas/concepts to English in my mind. They just come naturally, or don't come at all. When they don't come naturally, it's normally because of a lack of vocabulary, so I switch to Spanish and lookup that word/expression in a dictionary.

Whenever there is a hard problem, I try to visualize it and form a mental movie out of it. For example, if it is some sort of algorithm or optimization problem, I like to ride through the function/algorithm landscape. Visualization really helps in clearing thought.

But in everyday thought process, I do think in Hindi or English.

I don't think people are capable of introspecting with enough accuracy to tell if they think in English. When you try to describe how you think, you might say you think in English, but perhaps you just don't have words accurate enough to explain how thought really works.

Do you speak more than one language? The people I know who do, are usually very clear that they find themselves thinking in one or another language.

Yes, and at least for me it takes some effort to switch from one to the other. But to be fare with the previous post I do not know if it is the same process when you are speaking and when you are thinking about a problem.

When thinking about numbers and music, it's primarily colors and shapes. Every tone and digit has a different color-shape. "Blue notes" really are blue, flats are brown, harmony is a slurry-grey pattern that's hard to describe. :)

I think primarily in English or Spanish words, then pictures when I am ramping up into a problem. I'll start talking then stop, and my hands reach for a pad & pen of their own accord.

Once I'm "in", my thinking is not really words and not really pictures. It's... I don't know. There is a saying: when you want to find a lost horse, you have to think like the horse. Ideally I identify with whatever it is and I "think like the horse". Afterwards I discover/rationalize explanations for it.

Thats not being unilingual, that's just not being fluent. The whole definition of fluent, is that you can speak both languages w/o translating it in your head.

As far as the whole needing to describe empathy, thats just a lack of vocabulary...you don't know the word for empathy, so you try to describe it with other words.

As far as thinking in one language, that's not that uncommon, I'm fluent in 3 languages, and know 2 so-so, but the voice in my head is set to the language I use most in my day to day life. Sure I can switch to another language on a whim, but it takes an extra effort.

I'm fluent in two languages and know a minimal conversational-level with three more (and enough that I "think" in those languages, as opposed to a "lookup table" of sorts). What happens is I usually think in English (most used, maybe first (still disputed by parents, heh), most knowledgeable in), but once in a while I'll think in whatever language has a word I know that seems to match whatever it is best. There's also those loan words in every language that for whatever reason might seem to fit. I unfortunately can't think of great examples right now, but one would be schadenfreude (although I don't know German at all, quite possibly my favorite word ever, although rarely actually used). It isn't necessarily because I don't know the appropriate vocabulary in the language, just that whatever word seemed to have the right definitions and connotations for me to want to use it. Perhaps those word-equivalents in a different language were extraordinarily obscure, and I don't know of it.

In addition to the above, which is typically my "average" thinking, if I'm talking to someone in a different language, I don't really "think" in English until I reach the point I made above, if I come across an English word that made more sense. I don't know if it takes extra effort, but the language used to think does seem to change.

Once in a while I'll think of images instead of words. One example are those..cough..shock sites out there.

I'd noticed this sometimes too, but I think it's something that you can change if you want to. When I was learning german, after a couple years of high school classes I went to a month long full immersion program. When I finished that, I was thinking entirely in german. This stopped over time as I went back to English, but I can still make myself do it sometimes. So with practice/time, you could probably teach yourself to think in something besides English. Though English seems like it might be the most efficient mode for a native speaker.

010011100110111101110000011001010010110000100000 010010010010000001110100011010000110100101101110 011010110010000001101001011011100010000001100010 0110100101101110011000010111001001111001

If you have time check out Korzybski's general semantics. NLP also deals with this. One of the classics is Benjamin Worf's 'Language, thought and reality. Generally speaking, it seems we describe the world to ourselves with internal dialogue and then react to the words of our description rather than responding directly to reality.

I think in Italian occasionally, but of course English is my native language so that's what gets used most. When programming, it's English only, and speaking Italian doesn't come as naturally. I started trying to think in Italian as a way to practice the language, a long time ago, and found that it was pretty effective.

Does anyone know if there is anything open-source that is comparative to Rosetta Stone? I would like to learn Spanish and also brush up on my Italian a bit. I mention Rosetta Stone because I hear that it's the best language learning tool out there but I'm open to other learning systems as well.

While it is possible to "think in pictures", language is required to form complex thoughts.

A long time ago, people who were born deaf were called dumb. This is because, being deaf, they were never able to learn a language, and therefore could not form complex thoughts.

I'm a visual learner, and I think in terms of images, even when programming. I am also decent at drawing/painting.

The concept of "empathy" makes me think of mirror neurons and makes me imagine nervous systems of human bodies.

I am bi-lingual: English and Russian (having lived in US since 1996). Often times, however, I can't quite tell whether it is Russian or English I am thinking in.

Born in China, with my native language being Mandarin. Moved to USA at 4 years old. I think in English, though–I rarely use Mandarin day-to-day.

I'm similar, except I moved to US at 6. When I'm around English speakers, I think in English. When I'm around Mandarin speakers, I think in Mandarin. The funny thing is when I'm around a localized Chinese dialect, say Taiwanese speakers, the voice in my head adopts that accent too. It's probably because I don't usually think in Chinese.

Born in China, still in China, I think in neither Chinese nor English, people don't think in human languages, human languages are only for communications.

I think in which ever language I have been using the most recently. This varies between English, Welsh, French, Japanese, etc.

I speak 7 languages. Whenever I am in company with this specific linguistic group I start thinking in their culture.

Language = Culture (for e.g Japanese)

Native english speaker, but I've known Spanish long enough that sometimes I think in Spanish.

Hazoo fzeem! Wallenmacher bibble zark zark wachoon gra faZAMble!

Mek holgen Schwatris! Zoonkar bes limforg.

Einstein famously thought in pictures, but very few humans do.

Temple Grandin also famously speaks of thinking in pictures; could this + Einstein be related to Simon Baron-Cohen's thesis that autism can be defined as an "oversystematizing" brain?

From what I understand, Einstein taught himself to. Is it the same with Grandin?

But in some cases, you have to think in pictures. Like trying to solve a Graph problem, it is even hard for me to describe a specific Graph problem in proper language...

As a child, I used to think in pictures too, but the school system beat it out of me.

Would you expand on this?

The most famous example is his discovery of relativity coming from envisioning himself riding along a beam of light.

But most people when hungry think "I'm hungry" (if they have a conscious thought on the matter at all, which they may not) rather than picturing a hamburger. Or so my psychology professor said.

I would say most people don't actually "think" they're hungry rather than perceive they're hungry. The hunger response is not a higher-level cognitive function; it is regulated by the lower brain. When it reaches the level at which you can reason about (and picture a hamburger), it is more than likely that you already are aware of your hunger.

English

The language/culture influences quite a bit.

I know this quite well, because I grew up with a mother language quite different from my current language (both are not English), so learning other languages (preferably really different from your mother tongue) is always a mind opener.

I made this experience: while slowly growing into my 2nd language/culture, I extended my horizon, without loosing what I learned so far. So, it's not that I'm now inside my 2nd language, and lost my 1st experience/capabilities, but: my horizon is widened, I'm more logic and more flexible in both languages, and in my whole thinking process.

BTW, I never explicitly think in the words of my human language, words may accompany the thinking process, but thinking is so complex and fast, that you would not be able to follow directly; it mostly 'happens'...

Reply on news.ycombinator.com