Language: The DangersCATEGORY / Language AUTHOR / Tio DATE / January 3, 2017
[showhide type="post" more_text="" less_text="Hide Plain Text"]
So far in our series on language we have seen how little sense the languages that we speak around the world make, how many inconsistencies there are, and overall what a mess this has become.
This time we will showcase why talking about language and trying to improve ‘it’, is not only a good idea, but a necessary one. Before I prove it to you let me just say that even if we were able to build a worldwide saner society where people have access to whatever they need and want without any trade involved, and this society is ‘educated’ and so forth, language would pose a ton of issues to everyone out there if it is left in the present state, and because of that people would die, scientific research would be massively slowed down, and overall many humans would live a miserable life. The danger lies in the fact that languages, whichever ones, are poorly constructed, but that’s not the main issue, the main issue is with the fact that there are so many languages out there.
Let’s look at the offline and online impact.
In May this year (2016) I started to feel pretty sick. Indigestion, fever, chills, abdominal pains, and so on. I live in Spain and people here rarely speak English from my personal and so many other people's experiences, though I was lucky that my GP (general practitioner) was speaking a broken English so I was able to go to the hospital there. My dialog with the doctor was so simple that I could not explain much about my symptoms, but since I went to the hospital many times in a month period (mainly to make sure that he understands what my symptoms were), one of these times my doctor wasn’t available and I had to go and see another one who knew no English at all, and my Spanish is almost non-existent. So we barely communicated, and based on what he understood he sent me to do a few tests that my GP didn’t quite see as necessary as he explained to me later on, and this ‘spanish-only-speaking’ doctor also put me on a pill for 8 months that my GP recommended not to take anymore.
I went to the hospital many times over the course of 3 months, and at the end not much was ‘discovered’ about my health, and my problems disappeared, but the entire situation was very very stressful for me, and here’s why it could have been dangerous: if you experience indigestion for a few weeks time, it could be a stomach bug or cancer, that’s how extreme it could be. The difference in the diagnostic could lay in the details you give to the doctor. If your doctor only knows that you have indigestion and not any other ‘alarming’ symptom, then he/she will ask you to wait more and then do some tests if necessary, but for example if you also experience weight loss, or you see blood in your stool, then that’s alarming and must be checked out immediately since it could be a symptom of a much more serious problem, even cancer. Since the main thing doctors rely upon is what you tell them in order to send you to do more tests (or try to diagnose you), or ask you further questions, then whatever you tell them must be properly explained by you and properly understood by them.
I was also losing weight rapidly besides the indigestion that I had, and I explained that to my GP (the one who understood a bit of English), and he didn’t see a cause of concern after doing some simple physical checkups (though I wonder if he properly understood all of the symptoms I explained to him), but when I saw the other doctor who didn’t speak English I mistakenly told him in a broken Spanish that I have these issues for the past 5 months instead of 5 weeks, and I noticed that mistake only after the visit ended. So maybe that’s why this doctor sent me to do those tests and put me on a pill for 8 months, because having indigestion and losing weight for 5 months is, for sure, a cause of concern. Such little details can make a huge difference when it comes to a doctor consultation - serious health issues can be ignored and treatment delayed, and on the other side unnecessary tests and treatments may be recommended.
When I went for an abdominal ultrasound test while investigating my health problems. The specialist, who didn’t speak English, told me in a Spanish-Catalan that she saw something “red” on my gallbladder but she didn’t know what it was, and said I should check with my GP…. That’s what I understood. That scared me as hell because ‘finding something red’ on the gallbladder doesn’t sound too ‘happy’ of a news. I went to my GP and he explained very briefly in a broken English that these are some polyps and I should not worry about it, but we should do another ultrasound to make sure what they are. And he sent me back home. So I knew that I had some polyps on my gallbladder and that’s all, because my doctor couldn’t explain more to me. I also knew that gallbladder polyps can mean cancer, and gallbladder cancer is one of the deadliest out there (imagine the stress). So I went back the next day to my GP and I managed to pull some more words from his mouth that made me less concerned about the situation. But the stress this caused me was immense.
I went through a few other tests and they were all ‘painful’ in terms of communication. I could barely understand what the doctors said because none of them spoke English, and I could not explain my symptoms well at all because my Spanish is ‘minimal’ at best and as I explained in the previous parts there is no Spanish out there but many varieties of, so even if I were to speak some Spanish (Castilian) it would not have helped much in an area where people mainly speak Catalan. So I went through 3 months of huge stress and so many visits to the hospital mainly because of communicational issues. Luckily for me my health improved a lot on its own and I didn’t have to go to the hospital again. But in this period of time I also visited the Emergency Room two times and both times I needed my sister to translate for me (from Spanish to English and vice-versa), and it is bloody hard to ‘function’ in this state of ‘hot potato’ game with sentences and words, moving from you to the interpreter, then to the doctor, back to interpreter, then to you, and so forth, especially in an emergency situation.
On top of all that I was not able to take advantage of vaccines here in Spain because I had to bring a paper from Romania (where I am from) with the vaccines I had done in Romania so they know which ones I still need, but they were unable to decipher it as it was written in Romanian and said they can’t do much about that.
Am I the unluckiest one out there though? Not at all. Actually I am lucky since nothing severe happened to me so far, but let’s have a look at these examples from the US tribe that are not in the ‘lucky’ spectrum:
In the US a 9 year old Vietnamese girl was brought to the hospital by her family. She had “an infection with a rapid onset and great severity, low blood volume, and a heart attack”, but because no one from her family spoke English and the doctor and nurses didn’t speak Vietnamese, they struggled to communicate and the doctor wrongly assumed that the girl had gastroenteritis, an inflammation of the digestive tract, and he prescribed the drug Reglan. Reglan’s side effects are not something to be ignored and the doctor said that he informed the parents about that and told them to bring the girl immediately to the hospital if they notice side effects. However, he explained that in English...and the drug’s description was in the same English language together with all of the documents that the parents came in contact with and signed. So because the parents didn’t understand much at all and the girl was misdiagnosed and had experienced side effects of the drug, she died shortly after that. The doctor and the hospital were found guilty and the ‘language barrier’ was the reason why this girl passed away.
Another girl originally from Taiwan was hit with a tennis racquet in the head and after 2 weeks she went to the emergency room because she wasn’t feeling well. Three days of fever and intense headaches followed whilst a doctor tried to communicate with her but was unable to due to language barriers once again. She died from a brain abscess that, if detected earlier (something the doctor could have deduced if he could properly talk to the girl), could have saved her life.
In another case a 59-year-old non-english-speaking male “suffered from an overdose of a chemotherapy drug that resulted in a toxic reaction causing heart and systemic damage” because the doctor relied on his son to be the interpreter and, of course, this resulted in a misunderstanding of the treatment.
A Spanish speaking patient “presented at the hospital with complaints of dizziness, nausea and vomiting. Her past medical history included kidney infection and chronic abdominal pain. But before Mrs. Jimenez could be treated, she went into cardiac arrest. As a result of untreated fluid in the brain, she suffered irreversible brain damage, became comatose and lapsed into a vegetative state.” The problem? Miscommunication, again. The doctors were struggling to talk to her daughter through the telephone to use her as an interpreter, but her daughter requested for the mother to be seen by another neurologist because “the treating neurologist allegedly did not answer the family’s questions”. So they were unable to have a proper discussion and this woman suffered enormously because of that.
In Florida a teenager collapsed but managed to say to his girlfriend “Me siento intoxicado.” - When the paramedics came, the girlfriend, who spoke limited English, repeated intoxicado, which the paramedics, who spoke minimal Spanish, interpreted as “intoxicated.” They brought the teenager to the emergency room, where he was treated for drug abuse. But after the boy spent 48 hours in a coma, the hospital staff ordered a CT scan, which revealed that the teenager’s head had flooded with blood. It turns out that feeling intoxicado can also mean “sick to the stomach,” which is a symptom of a brain aneurysm. This communication breakdown led to a $71-million-dollar malpractice lawsuit. Another 7-year old kid whose parents only spoke Spanish was misdiagnosed and that resulted in “damage to the child’s organs”.
A 78-year old Russian woman was recovering in a hospital from a stroke and the doctors could not evaluate her medical condition because they were unable to talk to her, but they noticed that she was “holding her leg at the time”. Because they could not communicate with the woman, they decided to amputate her leg after observing that her leg “was cold below the knee [and it had] been since the middle of the night”. “With the use of an interpreter, the patient could likely have described the severity of her pain and the lack of sensation that might have led to a timely intervention.”
And the story goes on and on and on. You can read more such examples here.
There are also issues in which the doctors/nurses do not know what interpreters to ask for because they do not know the distinction between dialects. For example “Chinese” people are often regarded as speaking “Chinese” but as we’ve showed in this series they speak at least 7 different main dialects that are so different from each other, so the doctor/nurse may be completely confused as to what kind of interpreter they need.
Over 63 million Americans speak a language other than English at home, and over 25 million self-identify as having limited English proficiency (understanding). (source) More than 90 percent of hospitals in the US say that they see Limited English Proficiency (LEP) patients frequently, and 97 percent of these hospitals see them daily. (source) - Please memorize those numbers because, for one, 88 million people (one third of the US population) who do not speak proper English (or at all) is not a small number, that’s the population of Argentina and Spain combined. So it is almost like all of the Spanish and Argentinian people being in danger when it comes to healthcare. Now if 90% of hospitals see such people daily, then that’s almost like “all the time”. If I don’t see 90% of the time I can be labeled as blind, that’s how big of a number 90% is.
A study done in 2006 shows that very few hospitals in the US have bilingual health care providers or other such solutions, so the situation is not great at all for the millions who cannot speak English in the US. But even where they can contact an interpreter they may choose not to do so because of monetary related causes (or other causes) as this important doctor says: “Doctors often don’t call interpreters when they need to. Given the time constraints that providers are under, if it takes one extra iota of time to use an interpreter, they will try and get by with their own rudimentary language skills.” And these ‘rudimentary’ skills are often truly ‘rudimentary’ as one important attorney said: “Not many people who had high school or college language training or studied abroad would be able to translate specialized medical terminology like describing cancer treatment options. So there is definitely an overconfidence many providers have about their language skills.” - And the same goes for automated translators as we argued in the previous part about how they are unable to make much sense of these situations and cannot be relied upon, especially in this context of healthcare where small mistakes can result in dramatic consequences.
Studies showcase that those who do not speak ‘proper’ English in the US are more likely to not be able to apply for an insurance or understand the benefits of it, are more likely to not receive timely eye, dental, and physical examinations; have a poorer understanding of the care they have received; are less likely to follow recommendations for treatment and follow-up visits; are more likely to be admitted to the hospital, to have longer hospital stays, and to receive insufficient anesthesia when admitted to the hospital. A study found that ‘lations’, people who speak only Spanish, suffer more when dealing with diabetes because of language barriers in the same US tribe. (source) Also the asthmatic children are more likely to be intubated. More than that, these patients are also at risk of receiving unnecessary diagnostic testing and may be at greater risk of suffering medical errors compared with those who speak English well. Other studies found that those who cannot speak the language properly of the tribe they live in are not engaging in ‘screening’ tests to detect or prevent diseases like cancer. Overall the studies found that these people who do not speak the local language are not satisfied with the healthcare and same goes for their doctors who struggle to communicate with these patients. You can find all of these studies here.
On top of all that think about ‘mental health issues’ that are primarily or entirely diagnosed and treated via ‘communication’ (therapies, sessions, etc.). Basically how well can an English speaking psychologist ‘treat’ a non-English speaking patient with mental health problems?
In Canada a 2001 report by the “Department of Community Health Sciences, University of Manitoba” concluded that:
- There is compelling evidence that language barriers have an adverse effect on initial access to health services. These barriers are not limited to encounters with a physician and hospital care. Patients face significant barriers to health promotion/prevention programs: there is also evidence that they face significant barriers to first contact with a variety of providers.
- In many cases, language, rather than cultural beliefs and practices of patients, may be the most significant barrier to initial contact with health services.
- Language barriers may result in failure to protect patient confidentiality, or to obtain informed consent.
Keep in mind that words like “compelling” or “significant” are rarely used by scientists, so when you come across them in such studies they might weigh a few kilograms.
And the situation didn’t improve since 2001 in Canada. More recent studies showed the same trend and also similarities to the situation in the US tribe. A 2012 interview with leading doctors from Canada could be best summarized by two of their quotes: “There’s a saying that without language, medicine is veterinary science. It’s absolutely crucial.” and “We encounter it [miscommunication] all the time — so does Ottawa, so does Calgary, Vancouver and Montreal”.
In Saudi Arabia for example studies show that around 42% of the patients are not happy with the communication between them and the nurses/doctors because of language barriers. The same studies looked at work-related violence against nurses, and 36% of nurses perceived that inability to understand the language was the main cause of violence against them.
You will find similar situations in Australia, UK, all throught Europe, and any other tribe. The simple fact that people travel from tribe to tribe, something that’s going to increase in our ‘globalized’ world, means more problems for these people when they will have to face a healthcare system that does not speak their language.
And even when you see how many issues language barriers create in healthcare, it is even worse than that because such problems are usually overcome with the help of family members or friends, or are not reported (Look at my case, did I report those situations to anyone!? No! Now how many more people experience those issues like I did or worse, and do not report them anywhere!? Plenty for sure.). In essence if you go to the hospital but you don’t know the tribe’s language and no doctor there knows yours, then more than likely family members or friends who are more familiar with both languages will try to help with the translation (so it’s an ‘alright’ situation for you - hopefully -, only because you are lucky to have family members who can help out, not because the system helped you in any ways), and in most cases where people face issues because of language barriers this is not reported anywhere, so overall the situation is even worse than these studies showcase. On top of this, even if you know the language of the tribe, this is not enough because medical terms are a different creature altogether - meaning, no matter how well you understand and speak English (or any language), if you don’t understand what a treatment is all about, what chemotherapy is, what antibiotics are, what vaccines represent, what exercise means in a medical term or a healthy diet, or overall what the doctor explains to you because you never came across those terms/words and ‘ideas’ (notions) before, then you are in the same pond of people who are having issues with language barriers when it comes to healthcare. Therefore, based on these last points, the situation is way worse than any report reports.
There is no point in continuing with examples on this as it is beyond proven that language barriers in healthcare is a cause of death, poor health care, disatisfaction, and overall waste of time, resources, and energy for both the patients and doctors.
This entire situation affects many many millions of people, so it is not something to be ignored. This is not like a ‘fluke’. This is something very serious. You can find way more details in this study that looks at many cases around the world, from Australia to Netherlands, Canada to US, and so on. I am sure there are similar reports from Spain, Japan, China, Russia, and many other tribes out there, but they might only be written in their native language making them inaccessible to me, and this is the next point we should tackle: language barriers break or slow down science.
In the 1930s German scientists found a strong link between smoking and cancer but because their research was published only in the German language, it didn't get much attention at all. 30 years later the same links between smoking and cancer were found by British and US scientists and because their research was published in English, it had a bigger impact worldwide in the scientific community. This is a classic example of how much of a difference language makes in conducting science. We may wonder if the original study was published in English and had a big impact in the scientific community since 1930, how many lives would have been saved resulting from that. Basically a delay in 30 years in exposure for such a study is a huge amount of time.
English is the main language used by scientists to publish their research and although this represents a struggle for many scientists who do not speak English or don't properly speak English, they have to write in English nevertheless so that they can get recognition and be able to share their work on a global scale, something that is vital in science for it to progress. (source) But even if this seems like a good thing, as it forces people to share one common language, the reality is different.
Brazilian scientists publish around 50,000 scientific papers per year (2007 statistic), but only about 18,000 of these are published worldwide in renowned scientific journals, with 17,500 of them published in English (translated), and only 500 in Portuguese (their official language). So, only about one third of Brazilian scientific papers are published worldwide, and the vast majority of these are published in English. (source) So, overall some 33,000 Brazilian scientific papers per year are not accessible to people speaking other languages than Portuguese, nor published worldwide. Multiply that to several years of Brazilian research and you will quickly realize that there are millions of Brazilian scientific papers that are not accessible to those who do not speak Portuguese. Then also multiply that for other tribes out there where the situation is similar.
And the reverse is also true: scientists from Brazil (for example) who are not familiar with English, cannot take advantage of research papers that are only written in English (or other languages that they don’t understand). And this applies for any tribe and any language and any researcher. Top 50 scientific worldwide journals where scientific papers are published, are written in English (from the US or UK mainly). A 2012 study by the largest scientific journal out there came to realize that more than 80% of their scientific papers are written entirely in English. So regardless if there are so many scientific papers written in any other language than English, most of the worldwide accessible and promoted scientific papers are entirely written in English. Basically if you are from Brazil and you are a scientist, but you do not understand English, you have access to some 50,000 scientific papers from your own tribe per year, written in your language, but you won’t have access to a gazillion more from other languages, and considering that the ‘cream’ of scientific papers is written in English (because they are published in these worldwide top scientific journals), you will miss most, and the most important papers out there. How does this situation sound to you!?
So, on one hand we have a situation in which tribes produce scientific content but only a fraction of it is translated into English. On the other hand we have a world dominated by English when it comes to worldwide exposure of scientific papers. So if a scientist wants to take advantage of what humans have discovered and tested so far on this planet so that he/she can discover and explore more (do research), then he/she must either rely on her/his tribe for research papers if that's the only language they understand, or learn English if that’s not their main language. But the basic idea is that if you don’t know English and you are a scientist, you’re fucked (to put it in simple terms), because you miss the ‘cream’ of the scientific world that is written in English and exposure to that world (your scientific papers written in any other language than English will be quite unknown). And if you are a scientist who speaks only English, then you’re also fucked because you are missing a ton of scientific papers that are not written in English. And if you are a scientist who speaks 10 languages including English, you’re also fucked because there are millions or billions of scientific papers written in the other 100/200+ languages out there.
Did I mention that scientists also lose a ton of time trying to learn other languages to try and navigate this situation!? They could spend that time researching and doing experiments, rather than learning a new language that they will never master. And did I also mention that the VAST majority of these scientific papers are behind locked doors that can only be unlocked with ‘money’? Yes, most scientific research is only accessible if you pay for it, that makes them even more inaccessible for research and the public.
So, to summarize with a ‘nice’ imaginary, yet based on reality story: Bono is a scientist from Brazil. He studies the marine life and the impact climate change has on it. Bono is 32. He learned English in school and college but as it happens more often than not, he didn’t get much from it. School is too inefficient to teach you anything, more so when it comes to a new language. His culture, the Brazilian tribe, is also dominated by the Portuguese language, their language. Movies, books, songs, you name it, most of them are in Portuguese. Because of all of that, Bono was unable to learn English. So Bono starts to investigate the climate change impact on marine life and he needs to know what other research was done in this domain. He finds some Portuguese papers after a battle with technology and search engines, or asking colleges about them. It’s hard to find these papers with so many journals out there publishing them. But he finds some! Aaaaawesome! But his excitement is quickly squashed when he sees that he must pay to access them. “Fuck it!” he says. “How many research papers and/or scientific journals should I buy in order to access them, until I find what I need!?” he rhetorically asks, because indeed when you do science you may need access to many scientific papers, and that could become costly. Bono may or may not pay from his own money to access them, but even if he doesn’t, it is hard to make a company or state pay for it unless he has a ‘good’ reason for why he needs access to those papers (like working on something that can bring a profit for a company). “Fuck it!” he says.
Ok, so he gets access to the papers he needs and he is happy about them, but then a colleague of his tells him about this popular research paper that is only written in English and was published recently about the impact of climate change on marine life. Bono is so eager to read the paper, but it’s only in English… “Fuck it!” he says. He tries to find someone to translate it but can’t really find anyone. After struggles, as he needs someone with expertise to translate such a paper, he finds someone. That someone costs a lot of money. Bono pays, and “Fuck it!” he says. A second “Fuck it!” follows as he realizes that he needs more translations for the sources mentioned in this paper for further research….all of that costs him money to access those resources and also to translate them.
Now Bono has access to several Portuguese scientific papers and a ‘good’ English one that is translated into Portuguese. He does some research, he is happy with his findings and conclusions, he writes his scientific paper, and now he is ready to publish it in Portuguese. After being forced to pay for publication in several Brazilian journals (“No fuck it!” he says, “I was expecting that”), he realizes that his work is largely ignored because it is written in Portuguese. “Fuck it!” he says.
Bono will have to translate his work into English and publish it in a renowned global journal for his work to gain some attention. Bono is one of the millions of scientists out there, from any tribe you can point on the map.
And guess what: many scientific papers are rejected because of “grammar errors”. Basically, because, as we’ve showcased, language has become a religion, scientific papers are now rejected because non-native English speakers, who struggle to translate scientific papers or write them directly into English, may make some typos. (source)
But the issue with language barriers in science is not limited to researchers/scientists, although that could be where the most damage is done. This is also about you and me, and the A.I.. ‘Non-scientist’ people are also interested in science and need access to such scientific papers, and A.I. is the latest kind of tech that is truly based on data, so not having access to billions of scientific papers that an A.I. can analyze and draw powerful conclusions from, is a very ‘bad bad’ thing. Especially because the vast majority of scientific data in such papers consists of ‘normal language’. You have to understand that this is a HUGE issue, since A.I. based decisions in science will be the future (if it’s not already) because with so many scientific papers in the world it is nearly impossible for scientists to analyze them even if they were all to be written in a single language. So they need software to do that for them and the only way to take advantage of science papers written in ‘natural’ human language is to develop software that is able to make sense of this ‘natural language’. Quite ‘logical’. So what about the situation where we have billions of scientific papers out there inaccessible for such A.I. software? Who knows what research is out there that could help us find a cure for some cancers, or extend human life, or whatever, and are not discovered because they are written in some ‘unpopular’ languages for which there is no A.I. development.
Projecting that such an A.I. that can analyze scientific papers can be built for every language out there is quite unreal and unnecessary. Even small A.I. samples like google translator are way better for certain known languages like English and Spanish than the other ones, because the focus of those developing them is on such popular languages. Also, since ‘grammar’ differs so much from language to language, it’s not a matter of building a piece of software that can read a certain language and draw powerful conclusions off of it, and then take that and apply the same technique for other languages. Meaning that if you want to interpret scientific papers in multiple languages using A.I. it is not going to yield good results because you have to basically program such pieces of software for each language individually. And even if you achieve that, and let’s say that there is an A.I. for each language out there, and this A.I. reads scientific papers and draws powerful conclusions from them, then we are still faced with the same issues since language barriers will mean that now we have 200+ such A.I.’s, doing ‘research’ for 200+ languages, yet unable to communicate with each other because translating from one language to another is such an impossibility and always will be, even for complex A.I.’s.
The point is that in order for science to work it needs globally accessible data. One scientific study done in Spanish can relate to one done in Romanian, and then with 50 done in German, and also 7 in Portuguese. Science is universal, so using multiple languages to define scientific research is purely a mess that doesn’t help anyone, and A.I. won’t help us in this situation.
It could be hard to grasp the severity of this problem because it may be hard to imagine what could be achieved if all scientific papers shared the same language. But think about how many lives would have been saved if that research paper written in German in 1930 would have had the impact that the English one had 30 years later. What research paper is out there showing new connections between climate change and the environment, or between some foods that we eat daily and human health, or some ideas that could revolutionize our understanding of the world, save lives, help us progress much faster….
I want to force this a little bit more because I suspect it is hard to understand that this situation kills people and damages the environment and people’s lives. Let’s say that the paper on cancer and smoking was released in 1930 in English and had a substantial impact in the world in the sense that people would have been made aware of the connection between smoking and cancer. This world moves pretty slowly due to the tangled trade craziness, but still people would have heard of that study and some measures would have been taken/implemented together with the spark of more studies on the matter. If Genie, a 27 year old smoker, would have heard of these studies in 1930, she could have made the decision to quit smoking in her 20s. But since the studies were visible only 30 years later, now Genie is 57 years old and the chances of developing all kinds of cancers are very high for her because she is a long time smoker now. She may die in a few years as a consequence of smoking. How many Genies could have been out there? A ton! What about other studies that suffered or are still suffering from the same lag? What about similar situations that are happening right now and we are not aware of because that research is buried beneath the weight of a particular language!? Maybe you, or I are Genie today, and millions of others.
Scientific papers are not the only important bunch of data out there, digital data of any kind can be of a huge importance in any domain. Analyzing millions of tweets with automated software or people's searches can help us deduce flu outbreaks or how to help people trapped in natural disaster situations and so forth; being able to properly ‘read’, in automated ways (A.I.), billions of articles instead of relying on keywords and tags could revolutionize the search engines and allow for far more accurate search results; from wikipedia articles to youtube videos we can find a ton of useful information in regards to what people want, what they need, or find new ideas that can lead to discoveries, better managing a society, and so forth. Around 80% of the internet data is this kind of unstructured data (blog posts, comments, articles, videos, and so forth) - basically data based on normal language. Combine that with the following statistics:
- Google translates some 70 million words
- Siri answers 100,000 questions
- 350,000 messages are sent on twitter
- 400 hours of new youtube videos are uploaded
- 2,600 new articles are produced
- 5 new books are written
All of these are happening every minute (take that in for a second....or a minute) and these statistics may only represent English content, and of course that’s a small percentage of the total online content produced each minute (think about comments, videos uploaded on other websites, websites that are not indexed, and so forth).
Now you have to realize that the only way to extract meaning from this data is to automate the process. You need smart software to crawl through all this material and extract meaning. But the materials’ multitude both in form and quantity is adding another layer of complexity: not only that there is so much data produced, but it is now converted from one form to another and also translated from one language to another. Thereby, both processes either lose bits of meaning or distort bits of meaning - and this is primarily the fault of how language is structured today (from English to Chinese and so forth). This means that if you want to extract meaning from videos, your software must understand spoken language first, something so hard to do when “My ant died” is so similar to “My aunt died” in pronunciation. Add hundreds or thousands of such scenarios to the list, and you can see how difficult it is to extract meaning from audio files (videos, recordings). Also add to the list the fact that if you want to extract meaning out of text directly, the multitude of inconsistencies that we hugely discussed so far are going to be existent in any language, and you can realize how hard this is to make sense of such unstructured data. Basically the state of language in today’s world is completely ruining the potential we have with the use of A.I..
Secondly, if you want to extract meaning from stuff that is already translated (automated or not) from another language, then what you extract from there would be even less accurate.
Meaning: Imagine we want to translate Spanish articles into English and then extract data from those translated articles because they represent a huge chunk of the internet’s unstructured data and we want to tap into it. The process of translation will result in small errors (with so many kinds of Spanish and inconsistencies in both English and Spanish it’s not hard to imagine that). Small errors multiplied with millions of articles can result in a huge mess. Now you are faced with trying to make sense of these articles with the use of an automated software. But how can you design such a software to avoid those errors and not be confused by them? Who knows what messed up data you will get now by analyzing these improperly translated articles. And when it comes to video/audio files it is even more difficult trying to (for one) transcribe the Spanish videos in Spanish (from audio to text), then translate that transcription (the text) into English. It is crazy! And this is already happening today since A.I. software is already used to crawl through the internet and find articles, yet of course the software does not know if it is reading a translated article or not. So whatever they extract from the web with the use of A.I. can be truly messed up because of the way languages are today.
Therefore, there is a ton of data produced each minute, and each day, and the only way to make sense of it is with automation (software), but since languages are poorly constructed (inconsistencies) and there are so many languages out there that either create inability to access some data because of the language it is written in or it forces the software to access poorly translated data, then this is a majestic mess that we will have to face more and more. The only reason these so called A.I. systems sparked lately with so much advancement in any domain is because of DATA. No fancy computational power, no fancy code, but MAINLY data. Therefore DATA is CRUCIAL, yet today’s data is a mess because of language. Keep in mind that these powerful A.I. systems could very well find patterns in this data to find better cancer treatments, even cures, or who knows what mind blowing discoveries they could make. So if this situation with language affecting big data does not seem so dangerous to you, it may be because neither you or I can properly understand what we are missing if there would be a better designed language and a main (or single) one for all of us.
Now let’s go back to us, the other ones, the ‘non scientists’. In order for the world to become saner, people must be exposed to saner information, and scientific information is the definition of that. We need scientific articles, documentaries, videos, books, podcasts, and so forth to reach more and more people. I watch a ton of English documentaries because there are almost none in my native language, Romanian. Or at least not at that scientific quality (nothing close). I am lucky that I know English (and I know how to select good documentaries 🙂 ), but what about my fellow tribal Romanians who don’t? Rarely you can find Romanian subtitles for any of the documentaries I’ve seen so far, and I’ve seen a lot. VideoNeat is my website where I list the best of what I’ve seen so far (that is some 20/30% of the total documentaries that I’ve seen so far), and perhaps only 5% of the total that I’ve seen (not only those on VideoNeat) have a Romanian subtitle somewhere on the internet. I suspect this is the case for all languages out there in regards to these documentaries. So if I would not know English I could not see more than 5% of the documentaries that I’ve watched so far. And that applies for any content out there for sure: videos, lectures, articles, podcasts. Remove all of that English content from my brain right now and I’ll start to roar and hunt pigeons 🙂
In the previous part of this series I showcased that the world is moving more and more towards adopting solely their tribal language because of many reasons and one of them is because it is easier to translate any materials into their local language. That could have been great in this situation as those scientific materials would be translated for many tribes out there so that they can enjoy them too, but the interesting fact is that tribes (and people) mainly translate materials that are popular, not scientific materials. To make it simple: you will find Romanian subtitles for almost any English movie out there, but almost no subtitle for most of the good quality English documentaries. I cannot find any statistics on that and I’m unsure if there are any out there, so this is from my experience of running a documentary website for the past 3-4 years and trying to find subtitles for them in any other language, but unable to find much (if any at all). So, again, something popular, even a documentary (something like Planet Earth 2) will have subtitles in multiple languages, but more unknown ones won’t, and from my experience the unknown (unpopular) documentaries are the most scientific ones out there. And, to make sure your head explodes, think again about the fact that translated stuff is not going to represent the original one very well - If I see the same English documentaries in Romanian I will get a slightly different picture of what they are all about.
To summarize what I am trying to explain: good quality English documentaries are not translated in other languages. And this applies to so many materials out there. There are many great educational YouTube channels, amazing books, and podcasts, that are only available in English. And the opposite is also true. I am always wondering what awesome documentaries, articles, podcasts and so on I am missing that are only in Japanese, or a Chinese dialect, or Italian, and so forth….because as in the case with scientific papers, anything that’s not in English is even more unlikely to be translated in English. An awesome Japanese documentary, or a youtube channel, and so forth, is more unlikely to be translated into English (or another language) than the opposite (that popular ‘thing’ to be in English and then to be translated into Japanese or whatever other language).
Derivando is a Spanish youtube channel that teaches maths in a fun way; Date un Voltio does the same job in Spanish, but this time for physics; and Que Significa explains words/terms to Spanish speaking people.
What about Hebrew science experiments on Youtube as this channel does? What about this other Hebrew channel that ‘debunks’ religion and promotes a skeptical mindset? Or Universities from Israel that put up lectures on Youtube only available in Hebrew, like this one.
There are amazing Youtube educational channels in French (1, 2, 3), Arabic, Italian (1, 2, 3), Russian (1, 2, 3), Romanian (1, 2, 3), and for any other language out there. ‘Chinese’ video content for example could be mainly hosted on Youku.com a sort of Youtube of theirs (see examples 1, 2, 3). Same applies for books, podcasts, or articles that are in these other languages than English and therefore unavailable for the rest who do not speak those languages.
And science education is not solely about the information, it is about the medium through which it is transmitted. Carl Sagan inspired many to pursue science and transformed many inert minds into sharp and scientific ones through his way of explaining things, something that cannot be translated in any other language:
George Carlin put funniness into promoting critical thinking and skepticism and I doubt any subtitle or dubbed voice can in any way match that delivery of his message:
Working with TVP for 3-4 years and seeing many of their materials being translated (including many such videos), I realized two things: one was that it is impossible to translate them or properly translate them (too much stuff, too much work, too much to interpret from one language to another); second - they lose Jacque’s charisma completely - it is like seeing a music video without sound and only reading the on screen subtitle. Such minor things can be super important, it can be the difference between a plain text article (like a wikipedia one), and a great documentary (both about the same subject) - in the sense that not many may read the plain text article or get inspired by it, yet the documentary will be watched by way more people and have a much bigger impact for them. Remove Carl Sagan’s voice and writing style and the way he talked, and read the original Cosmos in Russian or another language...do you think it will have the same impact?
And, for the last time, the opposite is also true: how many more Carl Sagan, George Carlin, Jacque Fresco, and so many other non-English speaking people are out there that we, the ones who only/mainly speak English, are missing?
To summarize: even if we were to build a society of abundance for all, where there are no countries, no laws, no ownerships, and such, so we are a global species and everyone has access to everything they need and want, if we are still going to rely on so many languages that are all so poorly constructed then we will see many dead people, many unhappy people, many scientific papers ignored and solutions to problems delayed, education being handicapped, and so forth. All because of language barriers that no A.I. can solve. Imagining that in the future we can talk whatever language we want and complex software will instantly translate it so that we can communicate with anyone (including making sense of scientific knowledge and so forth), is a pure fantasy that can never be achieved, unless we are talking about simple communication and not something like healthcare where any bit of information in communication can be crucial, or normal day to day conversations that are so complex. Today we see so many scientists being forced to work together due to common interests in pursuing science, yet they struggle to communicate with each other because they come from different tribes, and despite the fact that science can still be done using maths and physics alone, the inability to have a conversation between scientists is a major flaw for science.
If we take laws into consideration and the inability for foreigners to make sense of them, or recognize scams, creating a bank account, looking for a job, reading food labels and so forth then the situation quickly escalates and becomes even more severe - I don’t even have more words to emphasize the severity of this situation. But also add in emergency situations like earthquakes, a flu outbreak, natural disasters and so forth that those not speaking the local language may not be able to save or protect themselves from. And with the impact of climate change and conflicts nowadays where people are forced to migrate from one tribe to another, this situation is unaccceptable https://www.youtube.com/watch?v=dz_ZIm63nuU
To make sure you understand the point of this article, imagine you ‘land’ in China. You only speak English and people there only speak the Chinese dialects. Now start! Play your life!
Want to read a book? This is how a bookstore looks like http://cdn.theatlantic.com/static/mt/assets/peter_osnos/RTXD7QH.jpg . Want to buy some food? Most restaurants are solely in Chinese (whatever dialect) and those who have a translated menu in English are doing such a bad job at translating food menus that you may have to choose between a “God with Vanilla” or a “fuck the duck until exploded” for your dinner (source). Same happens when translations are ‘in action’ for all kinds of ‘things’ in China, from street signs to warning signs or messages of any kind http://www.boredpanda.com/funny-chinese-translation-fails/
So, what book are you going to read? And what food are you going to eat? Are you allergic to certain foods or don’t like some ingredients? How can you avoid those? Do you know the laws, and/or if a policeman stops you, will you be able to communicate with him/her? Will that get you into trouble?
If you want to navigate, the street signs will be in the same ‘alien’ language together with your phone that, if you have an internet connection, will probably display apps in Chinese automatically based on your location. So even if you open an English app for directions, or whatever, it may display it in Chinese automatically.
Want to rent a place to stay, a car, buy stuff, ask for directions, or for help of any kind, and you will realize that you can’t do any of these. You feel lost and that’s only a few hours of experience. What if you have health issues? Where do you get help from? How do you communicate your problems?
Navigating your life online or offline in such a place is going to force you to know the local language. If you don’t then you’re fucked!
Now think about the fact that so many people are in this situation all over the world where they feel lost. They can’t ask for help or be helped, they can’t get a job, a health insurance, friends, and so forth. Even the tiny details like not being able to read the details from a medication package is going to cause you problems. You are vulnerable in such a situation.
Doctors say that medicine without communication is veterinary medicine because you can’t listen to a cat to see what ‘her’ symptoms are, but extending this to the worldwide society, a world without a global language is like a jungle with millions of different species that are forced to live together - expect misunderstandings, angry and dead people, slowness of progress, and so forth.
Therefore, if people might think that “language” is not something worth addressing if we want to create a saner society because it is not such a big problem (there are other more important issues), perhaps if those people would be aware of the many problems language creates and how many people suffer and die because of language barriers, then they could see how important it is to address this subject.