Boroditsky - Lost in Translation

THE IIALI STREET JOURNAISATURDAY. JULY 24.2010 O 2010, Dow Jones & Company, Inc. All Rights Reserved

Lost in TrarclationNew cognitiue research suggests thatlnngwqe proloundly infltrcnces thn wry peopln see the world,; a

dilferent sense of blnme in lapanese and Spanish

Now, a flurry of new cognitive scienceresearch is showing that in fact, lan-guage does profoundly influence how wesee the world.

The question of whether languagesshape the way we think goes back centu-ries; Charlemagne proclaimed that "tohave a second language is to have a sec-ond soul." But the idea went out of favorwith scientists when Noam Chomsky'stheories of language gained popularityin the 1960s and'70s. Dr. Chomsky pro-posed that there is a universal grammarfor all human languages-essentially,that languages don't really differ fromone another in significant ways. Andbecause languages didn't differ from oneanother, the theory went, it made nosense to ask whether linguistic differ-ences led to differences in thinking.

The search for linguistic universalsyielded interesting data on languages,but after decades of work, not a singleproposed universal has withstood scru-tiny. Instead, as linguists probed deeperinto the world's languages (7,000 or so,only a fraction of them analyzed). innu-merable unpredictable differencesemerged.

Of course, just because people talkdifferently doesn't necessarily meanthey think differently. In the past dec-ade, coglitive scientists have begun tomeasure not just how people talk, butalso how they think, asking whether ourunderstanding of even such fundamen-tal domains of experience as space, timeand causality could be constructed bylanguage.

For example, in Pormpuraaw, aremote Aboriginal community in Aus-tralia, the indigenous languages don'tuse terms like "left" and "right." Instead,everything is talked about in terms ofabsolute cardinal directions (north,south, east, west), which means you saythings like, "There's an ant on your

southwest leg." To say hello inPormpuraaw, one asks, 'Where are yougoing?", and an appropriate responsemight be, "A long way to the south-southwest. How about you?" If you don'tknow which way is which, you literallycan't get past hello.

About a third of the world's lan-guages (spoken in all kinds of physicalenvironments) rely on absolute direc-tions for space. As a result of this con-stant linguistic training, speakers ofsuch languages are remarkably good atstaying oriented and keeping track ofwhere they are, even in unfamiliar land-scapes. They perform navigational featsscientists once thought were beyondhuman capabilities. This is a big differ-ence, a fundamentally different way ofconceptualizing space, trained by lan-oltga

Differences in how people thinkabout space don't end there. People relyon their spatial knowledge to build manyother more complex or abstract repre-sentations including time, number,musical pitch, kinship relations, moral-ity and emotions. So if Pormpuraawansthink differently about space, do theyalso think differently about other things,like time?

To find out, my colleague Alice Gabyand I traveled to Australia and gavePormpuraawans sets of pictures thatshowed temporal progressions (forexample, pictures of a man at differentages, or a crocodile growing, or abanana being eaten). Their job was toarrange the shuffled photos on theground to show the correct temporalorder. We tested each person in two sep-arate sitttngs, each time facing in a dif-ferent cardinal direction. When asked todo this, English speakers arrange timefrom left to right. Hebrew speakers do itfrom right to left (because Hebrew iswritten from right to left).

BY LERA BORODITSKY

Do the languages we speak shapethe way we think? Do they merelyexpress thoughts, or do the structures inlanguages (without our knowledge orconsent) shape the very thoughts wewish to express?

Take "Humpty Dumpty sat on a..."Even this snippet of a nursery rhymereveals how much languages can differfrom one another. In English, we have tomark the verb for tense; in this case, wesay "sat" rather than "sit." In Indonesianyou need not (in fact, you can't) changethe verb to mark tense.

In Russian, you would have to marktense and also gender, changing theverb if Mrs. Dumpty did the sitting. Youwould also have to decide if the sittingevent was completed or not. If our ovoidhero sat on the wall for the entire timehe was meant to, it would be a differentform of the verb than if, say, he had agreat fall.

In Turkish, you would have toinclude in the verb how you acquiredthis information. For example, if yousaw the chubby fellow on the wall withyour own eyes, you'd use one form of theverb, but if you had simply read orheard about it, you'd use a differentform.

Do English, Indonesian, Russianand Turkish speakers end up attendingto, understanding, and rememberingtheir experiences differently simplybecause they speak different lan-guages?

These questions touch on all themajor controversies in the study ofmind, with important implications forpolitics, law and religion. Yet very littleempirical work had been done on thesequestions until recently. The idea thatlanguage might shape thought was for along time considered untestable at bestand more often simply crazy and wrong.

THE PUBLISHER'S SALE OF THIS REPRINT DOES NOT CONSTITUTE OR IMPLY ANY ENDORSEMENT OR SPONSORSHIP OF ANY PRODUCT, SERVICE, COMPANY OR ORGANIZATION

cusroM REpRt/vrs 609) 520-4x1 PO, Box 3oo PRtNcEToN, NJ 08543-0300. Do Nor EDIT oR aLTER REPRINTS, REPRoDucrtoNs Nor PERMTTTEDDOWIoNES==

Pormpuraawans, we found,arranged time from east to west. Thatis, seated facing south, time went left toright. When facing north, right to left.When facing east, toward the body, andso on. Of course, we never told any ofour participants which direction theyfaced. The Pormpuraawans not onlyknew that already, but they also sponta-neously used this spatial orientation toconstruct their representations of time.And many other ways to organize timeexist in the world's languages. In Man-darin, the future can be below and thepast above. In Aymara, spoken in SouthAmerica, the future is behind and thepast in front.

In addition to space and time, lan-guages also shape how we understandcausality. For example, English likes todescribe events in terms of agents doingthings. English speakers tend to saythings like "John broke the vase" evenfor accidents. Speakers of Spanish orJapanese would be more likely to say"the vase broke itself." Such differencesbetween languages have profound con-sequences for how their speakers under-stand events, construct notions of cau-sality and agency, what they rememberas eyewitnesses and how much theyblame and punish others.

In studies conducted by CaitlinFausey at Stanford, speakers of English,Spanish and Japanese watched videos oftwo people popping balloons, breakingeggs and spilling drinks either intention-ally or accidentally. Later everyone gota surprise memory test: For each event,can you remember who did it? She dis-covered a striking cross-linguistic differ-ence in eyewitness memory. Spanishand Japanese speakers did not remem-ber the agents of accidental events aswell as did English speakers. Mind you,they remembered the agents of inten-tional events (for which their languagewould mention the agent) just fine. Butfor accidental events, when one wouldn'tnormally mention the agent in Spanishor Japanese, they didn't encode orremember the agent as well.

In another study, English speakerswatched the video of Janet Jackson'sinfamous "wardrobe malfunction" (awonderful nonagentive coinage intro-duced into the English language by Jus-tin Timberlake), accompanied by one oftwo written reports. The reports wereidentical except in the last sentencewhere one used the agentive phrase"ripped the costume" while the othersaid "the costume ripped." Even thougheveryone watched the same video andwitnessed the ripping with their own

eyes, language mattered. Not only didpeople who read "ripped the costume"blame Justin Timberlake more, theyalso levied a whopping 53% more infines.

Beyond space, time and causality,patterns in language have been shown toshape many other domains of thought.Russian speakers, who make an extradistinction between light and dark bluesin their language, are better able tovisually discriminate shades of blue.The Piraha, a tribe in the Amazon inBrazil, whose language eschews numberwords in favor of terms like few andmany, are not able to keep track of exactquantities. And Shakespeare, it turnsout, was wrong about roses: Roses bymany other names (as told to blind-folded subjects) do not smell as sweet.

Patterns in language offer a windowon a culture's dispositions and priorities.For example, English sentence struc-tures focus on agents, and in our crimi-nal-justice system, justice has beendone when we've found the transgressorand punished him or her accordingly(rather than finding the victims and res-tituting appropriately, an alternativeapproach to justice). So does the lan-guage shape cultural values, or does theinfluence go the other way, or both?

Languages, of course, are humancreations, tools we invent and hone tosuit our needs. Simply showing thatspeakers of different languages thinkdifferently doesn't tell us whether it'slanguage that shapes thought or theother way around. To demonstrate thecausal role of language, what's neededare studies that directly manipulate lan-guage and look for effects in cognition.

One of the key advances in recentyears has been the demonstration ofprecisely this causal link. It turns outthat if you change how people talk, thatchanges how they think. If people learnanother language, they inadvertentlyalso learn a new way of looking at theworld. When bilingual people switchfrom one language to another, they startthinking differently, too. And if you takeaway people's ability to use language inwhat should be a simple nonlinguistictask, their performance can change dra-matically, sometimes making them lookno smarter than rats or infants. (Forexample, in recent studies, MIT stu-dents were shown dots on a screen andasked to say how many there were. Ifthey were allowed to count normally,they did great. If they simultaneouslydid a nonlinguistic task-like bangingout rhythms-they still did great. But ifthey did a verbal task when shown the

dots-like repeating the words spoken ina news report-their counting fell apart.In other words, they needed their lan-guage skills to count.)

All this new research shows us thatthe languages we speak not only reflect0r express our thoughts, but also shapethe very thoughts we wish to express.The structures that exist in our lan-guages profoundly shape how we con-struct reality, and help make us assmart and sophisticated as we are.

Language is a uniquely human gift.\Mhen we study language, we are uncov-ering in part what makes us human, get-ting a peek at the very nature of humannature. As we uncover how languagesand their speakers differ from oneanother, we discover that humannatures too can differ dramatically,depending on the languages we speak.The next steps are to understand themechanisms through which languageshelp us construct the incredibly complexknowledge systems we have. Under-standing how knowledge is built willallow us to create ideas that go beyondthe currently thinkable. This researchcuts right to the fundamental questionswe all ask about ourselves. How do wecome to be the way we are? Why do wethink the way we do? An important partof the answer, it turns out, is in the lan-guages we speak.