Ovi -
we cover every issue
Apopseis magazine  
Ovi Bookshop - Free Ebook
Ovi Greece
Ovi Language
Michael R. Czinkota: As I See It...
WordsPlease - Inspiring the young to learn
Tony Zuvela - Cartoons, Illustrations
Stop human trafficking
BBC News :   - 
iBite :   - 
Solutions to unsolved linguistics puzzles Solutions to unsolved linguistics puzzles
by Joseph Gatt
2019-03-01 09:05:30
Print - Comment - Send to a Friend - More from this Author
DeliciousRedditFacebookDigg! StumbleUpon

Having spent most of my life learning foreign languages in different settings and different countries, I experienced firsthand how languages evolve, how argots, dialects or languages evolve and the different relationships people have to a language.

First thing I noticed, which linguists never really point out, is how little we tend to use language. Language needs a social setting to evolve, be it schools, churches, workplaces, families, couples, associations, groups or any social setting. Though one might assume that such social settings make constant use of language, that is converse constantly, one might notice that in a lot of social settings language is not used as profusely as one might think.

lingua01_400Take schools for example, an ideal breeding ground for argots to form. While in some schools school children and teachers hang out with each other a great deal, in many schools, individualism is the norm and students don't interact as much as one would imagine. To form a new dialect, you would need a few hours a day of close, tight-knit interaction among school students and a network of interactions that would help the argot evolve. Unfortunately, in many cases conversations are not frequent and students don't spend as much time conversing with each other as they should.

The trend is the larger the group setting, the less one is inclined to use argots. Since a lot of people spend time conversing in group settings on the Internet, a lot of time such group settings don't see the emergence of new words, expressions, idioms or grammar structures simply because some members of the group might not understand them. Argots form in small groups, and gets diffused within the small group's network. There is a secrecy element, as the small group will want to prevent, for example, teachers from understanding what is being said.

There are three types of argots that can exist: when the intruder just comes in, when the intruder is present in the room and when the intruder is absent. When the intruder just comes in, argots will revolve around calling for attention that an intruder just came in. Arabs will use expressions like “how is Mussa” or “where is Mussa” to change the topic on whatever sensitive topic was being discussed. When the intruder is present in the room, it's either existing words that change their meaning (using “he” instead of “she” for example) or inventing new expressions so the intruder won't understand. When the intruder is not present, people will invent codes to simplify, shorten, ridicule or show admiration for a certain concept.

I admit I am a unique kind of linguist. Until now, linguistics focused on narrative forms of conversation that is focused on narrated text. I, however, focus on narrative as well as emotional text, as most linguists analyze words and sentences as if they were devoid of emotions. To me, emotions play an important role in linguistics, to the point of influencing grammar, syntax, semantics and the rest.

Last time I discussed at length the solutions to some of linguistics’ problems, I will hear discuss the remainder of the solutions to linguistics' problems. I admit it took me almost a year to resume this work in progress, and I kept procrastinating when it came to this specific work. The reason is I did not have time to catch up with linguistic literature and there were some more pressing problems to fix. When I see Asians battling unemployment, North Koreans, Iranians, Palestinians or unsolved economic problems, I tend to put the priority on those problems. I finally found time to share with you some insights on how linguistics problems can be fixed. Hurray!

As usual the solutions are not set in stone. These are not really solutions to problems; they are more like hints to solutions to problems listed by Wikipedia. The mindset is always that of if I were a professor teaching linguistics and that some student had asked me a question on what the solution to the problem were, here's what I would tell them. 

What role does linguistic intuition play, how is it formed and how does it function?

Life is a series of instincts. A lot of the actions we perform on a daily basis are responses to those instincts. We eat instinctively, work instinctively, rest instinctively, sleep instinctively. We don't put too much thought into those actions, and we don't think too hard before instinctively performing those actions.

Same goes for language. Language is something we use instinctively before we start thinking about the way we use language. We use language through imitation, or through a synthesis of the different forms of language we hear around us, or through describing what we witnessed originally.

Children are not as proficient language users as we might think. They use the kind of language they need, the kind that their social settings allow them to use. One of the things adults tend to do when they spend time with children, that is if they like spending time with children, is identifying what language patterns the child is comfortable with before they proceed to using the simplified form of language that the child understands.

Here's the best way to explain language intuition. I remember this colleague of mine who uttered what sounded like a sequence of unintelligible facts. He was saying something about something about something, then I left, realizing that the guy was just trying to look smart, he wanted me to think that he was smart, but he wasn't actually trying to teach me anything. I got that when he said something about writing 36 chapters of a book in a few hours.

So what role does linguistic intuition play? The truth is if we don't have a vast knowledge of the context of what is being said, we won't understand what is being said. To understand something, we need to know a few things about the person who is talking, the objects that are being discussed, the history of the concepts being discussed. This is why Google comes in so handy, because a lot of times things get discussed on television that we don't really understand and need to research to be able to conceptualize.

Unfortunately, Google does not have many of the answers we are looking for in our friends and acquaintances. Sure social media provides some clues, but otherwise, we either have to guess what the context is regarding what the person is saying, or we have to probe and ask questions. More often than not, we forget what was being said. In sum, most communication is misunderstood, and you need to go on several dates before you understand exactly what is being said.

Now of course most of us communicate in familiar contexts, so we understand most of what is being said. If we communicate with our colleagues, friends or family we will tend to understand most of what is being said. When we meet new friends, we spend a great deal of time discussing what their background is before we can understand what is being said. I'm sure you've all had that awkward moment where you walk into a shop and order food, and decide to make small talk before you realize what you are saying is of little interest or not being understood by the person you are talking to.

How is linguistic intuition formed? At the basic level when children learn one verb, one verb in one tense or one form, they figure out pretty quickly what a verb is and what a tense is, even when verbs, tenses or words don't have clearly marked forms. I remember when I was probably 3 or 4 years old some adult came to me and asked me “are you upset?” I had no idea what “upset” meant, but I instinctively knew that it referred to a state of mind or an emotion, because the person was staring at my face when asking me that. The body language referred to my facial expression.

When we communicate, we communicate with words but also with eyes and hand gestures. We tend not to pay attention to the eyes when we communicate, but one of the reasons sunglasses are banned in many places is that we lose a great deal meaning when we can't see the other person's eyes. We point at things with our eyes, we look at things with our eyes, how wide we open our eyes express a great deal of content.

Same goes with hand gestures and body gestures. We also communicate with our front, with our nose, with our lips and mouth, with our hands and with our body. Our body complements some of the words we use, while children learn the meaning of words specifically because they are looking at people point at objects, either with their hands, fingers, body or eyes.

How does linguistic intuition function? Imagine I just used the sentence “I moff brof.” Now you know word order, and you know that English is a subject verb object language, so you know “moff” is certainly a verb and brof is almost certainly a word, probably a noun. But you would have to guess the meaning of “moff” to figure out whether “brof” is a noun of an adjective. Now you would need someone to show you an image or use body language to describe what moff and brof are before you get an accurate idea of what the meaning of those words are.

What are the best ways to ensure communication without misunderstandings?

There are several factors that play out in conversation. Most conversation is instinctive, and most people don't examine their instincts. There can be misunderstandings because someone was trying to be polite. There can be misunderstandings because someone was too self-absorbed. There can be misunderstandings because someone resents the other person, or because a person feels superior to another person. There can be misunderstandings simply because someone omitted to explain something, or deliberately distorted something.

Again most linguists assume that language is a narrative affair. If there was a misunderstanding, it's because the narrative was not clear enough. Linguists tend to omit that language is also an emotional affair. The emotional filter can block the speaker from speaking clearly, or the listener from listening attentively.

So the best way to ensure communication without misunderstandings is to have a clean emotional filter, one that does not get in the way of communication. The other factor in misunderstandings is the knowledge of the topic being discussed, and the willingness to give clear and accurate narrative to what is being discussed.

In a lot of armies, some think that merely change some elements of the language can ensure clear communication. But if you say 1400 instead of 2 PM, that won't change much of the clarity or understanding. Saying 1400 instead of 2 PM merely puts the emphasis on the time by changing the wording in time so people pay more attention. But in the army as anywhere else, the emotional factor plays a role in misunderstandings.

The other factor for misunderstandings is that in some cases language can be ambiguous or the nature of the wording can be misunderstood. Yesterday I asked the guy who delivered my beer “where were you in November” as he did not pick up the phone during the month of November. But he understood “where will you be in November” and told me that November is still several months away.

If you say “he keeps annoying me” without specifying who, either deliberately or out of omission, it won't be clear who that is who is annoying you. Another factor is merely knowledge of the context. If you say “my car broke down” there can be a million meanings depending on the context. If you're telling that to your boss you're implying you will have problems with transportation and might be late for work. If you tell that to your wife you might be implying you need to spend money fixing it. If you tell that to your mechanic, the mechanic will tell you to try to be more specific.

What are the best ways to compare linguistic performance and linguistic competence?

You can only measure performance and competence if you test them. Unfortunately you can't just open someone's brain and look for performance, or measure competence. Tests are a social measure before all else, that is tests tend to determine your place within society. When you take a test there are two factors: your personal desired place in society and how you prepared for it, and the test maker's idea of who should acquire a certain social status. If the two factors combine well, you will score high in tests.

Tests also involve going beyond your intuitions and using reasoning, which requires effort from the brain. So when you measure performance, several factors will be involved including health, willingness to succeed and overall preparedness for the test.

How can we really measure someone's performance or competence? If you're like me, you use different types of speech depending on your audience. You can dumb yourself down a great deal, or you can be overly pretentious and discuss things you had little previous knowledge of. We all do that at certain points. That's when you have intellectuals who binge on cocaine, and blue collar workers reading Darwin or Tolstoy's War and Peace.

Performance and competence are also elastic. They can grow as you grow older, or they can recede. There are certain things we know but never use, certain things we don't know yet readily use. In sum, performance and competence questions are really more about testing than about real-life applications.

How does time and age influence linguistic competence?

We forget certain things and we use other things constantly. We evolve within our social circles, take on different roles, use certain things more than others. When I was an English as a Foreign Language teacher I kept discussing past simple and past continuous, two years after quitting my last EFL gig I'm no longer sure what past simple or past continuous are.

Instinctively, or ideally, one never really forgets anything and your linguistic competence keeps growing. Others would say that if you forget certain things your linguistic competence shrinks. But the truth is it doesn't grow and it doesn't shrink, it evolves.

Linguistic competence is the product of the social situations you happen to encounter. If you move to Korea, your linguistic competence will evolve a certain way, because you will be discussing concepts or ideas that are unique to Korea. If you work for Apple or Halliburton, your linguistic competence will be greatly influenced by your workplace. Same goes for if you are a freelance, or if you date Jane as opposed to Sarah or Norah, or James or Mohamed.

So it's not time and age that just influence linguistic competence, but also the social settings. If you're unemployed or retired that will affect your linguistic competence, along with whether you work at a workshop, using machinery, or chatting all day long about deadlines or new markets. Some use language more than others, some for more purposes than others. And yes, most people don't spend time reading books or trying to be smart.

What are the relations between standard language and dialects?

A standard language is a collection of mutually intelligible dialects. That is Canadian French is also French, French from France is French with all its regional dialects, Belgian and Swiss French are French, and the French varieties spoken in African countries are French, along with those spoken in the French West Indies and the French Pacific and Indian Ocean territories.

Same goes for English, with the British, Canadian, American, West Indies, Guyana, Africa, South Asia, and the Indian Ocean and Pacific Ocean territories. Take any language and the same principle applies. A standard language is the collection of all the mutually intelligible dialects of a language.

Now you have offices where there are arbitrary uses of standard language, which usually focuses on the details of language use rather than on the bigger picture. That is things like you shouldn't start a sentence with “and” or that you should spell “center” or “centre” or whatever. Those are small details that organizations like to bother their employees with and focus on form rather than on content. The truth is no one speaks a true, standard variation of any given language, everyone has their mannerisms and quirks in language use.

What causes language to change in some dialects and not others?

Again language is something used at the social level. You need a group of intimate people with a large network group of other intimate people to inspire language change. In some countries, or in some regions, there are few groups of intimate friends and networks are weak, for cultural reasons or for political reasons. In other regions groups of intimate friends are more dynamic and have bigger networks, which mean that language change is more frequent.

As I said before language changes either because there's a present intruder to any given group, because an intruder is absent to any given group, or because an intruder just showed up or just left any given group. Word changes and language change start off as a code, and gradually spreads through the population depending on the strength of the networks, or just dies out.

Is perfect computable word-sense disambiguation attainable by using software? Is accurate computational word-sense induction feasible?

In strict terms you can't attain perfect, computable word-sense disambiguation because different words are used differently by individuals, depending on the regions, and depending on the context. However, in general terms, software can provide a large pallet of the different words, their senses and their meanings.

Now the question here is with the word “computable.” That is in computer dictionaries, you usually have to input words and their definitions for them to show up. Computing would imply that with Artificial Intelligence, computers would go around the internet searching for words and computing their meanings.

Unfortunately with language, you usually need a human being to explain what the different words mean. That is if I start using the word “trafts” to mean “beer” you would need me to explain that by trafts I mean beer for you to get the idea. A computer however could compute by looking at the words that I associate with “trafts” to try to guess the meaning of the word, and the computer's guess will be an educated guess. That is if I associate “trafts” with 5%, drinking, drunk, relaxing, and “a can of trafts” the computer will eventually guess it's beer. But those words could also be associated with “makkeoli” a type of Korean rice ale which also tends to be 5% and be sold in cans, yet has nothing to do with beer.

In sum computers can guess the meanings of new words through word associations, but there can always be misunderstandings and wrong estimates of what the word actually means. It can throw excellent guesses, but mistakes are always possible.

What makes a good dictionary?

If Wikipedia had a dictionary in addition to its encyclopedia, that would make a pretty decent dictionary. The real question behind what makes a good dictionary, is what is the line between a dictionary and an encyclopedia?

A dictionary provides basic definitions and limits itself to defining certain words. Encyclopedias provide broader meaning and cultural context behind the definitions of certain words and concepts.

I believe it's not the dictionary that counts but the user's intentions that count. Is the user trying to identify the meaning of a word of which he does not know the meaning. Or is the user trying to gain broader information on the use of the word. Or is the user trying to write a paper on the details of certain customary uses of the concept?

A dictionary could never account for all the words, their meanings and their uses. During Brett Kavanaugh's confirmation hearing, Kavanaugh used a word in his school year book which was the verb to “boof.” Now no dictionary had the word “boof” in it, except for “urban” dictionaries which are slang dictionaries which disagreed on what the verb to “boof” actually meant. According to Kavanaugh “boof” is merely another word for “fart” which is hard to believe provided the exact sentence was “have you boofed yet?” Urban dictionaries provide colorful definitions such as inserting alcohol up your sphincter.

Again it's not dictionaries that are at fault. If I write in public, first I will try to use words that most people are familiar with so they don't need to use a dictionary. Second, if I use dictionary words, I try to make sure that the words are actually in the dictionary and that I'm using them while conforming to the definition presented in the dictionary. Third, if I use a word that's not in the dictionary or most people tend not to be familiar with, I provide context and a definition for the word.

If you live in France, Japan or South Korea, dictionaries can be frustrating. Not because of the dictionaries themselves, but because of the way writers tend to write. That is a lot of writers don't take the public's knowledge or assumed knowledge into account, and use words that are often not in the dictionary or don't match the dictionary's definition. Furthermore French, Japanese and Korean dictionaries try to save time and space by providing short definitions, leading to circular use of the dictionary, needing to find definitions for words in definitions.

This is because France, Japan and Korea are elitist societies where writing is done to prove that one is part of the elite, rather than writing to inform the public.

To what extent are dictionaries reliable in their universality?

Again it's not dictionaries that have to be reliable it's writers who should adapt their writing by using words that are matched by the dictionary.

Dictionaries have played a role in shaping and influencing the general meaning of words used in everyday life. Then of course there are people who have never seen a dictionary, yet who make very accurate use of words in everyday life. While people tend to have a general sense of what different words used mean, in a lot of cultures where there are no dictionaries there tends to be greater variety in regional and local differences in the words used.

So indeed dictionaries and education shape the way we tend to use different words. Education has done a great deal to reduce regional variations in word use and has standardized language to a certain extent. Let me give three examples to illustrate my point.

In languages like Korean, Japanese, French or English, the media and standardized education has reduced a great deal the variations in the words used among people. In some cases, dialects and regional languages have died or are only used by older people because of the force of education and the media.

In Arab countries, there was too much of an effort to standardize the language. In Japan the Tokyo dialect was used, making it easier for people to identify with speakers. In South Korea the Seoul dialect was used, in France the Paris dialect was used, while in English countries a more vague form of dialect is used. However, in Arab countries, the form used in the Quran was adopted as the official language, a form that has not been spoken in 13 centuries. So there are vast variations in dialects from city to city and from region to region.

In Kabyle, where there is no official dialect, which is not taught in schools and only has two television channels and no widely circulated newspapers, regional variations of the language are very vast, and in the cases of Tizi Ouzou and Bejaia, barely intelligible.

In sum, it's easier to compile a dictionary in English-speaking, French-speaking countries or in Korea and Japan than it is for Arabic or Kabyle.

What are good practices to avoid circular definitions in dictionaries?

The more information a dictionary contains, the better. If a dictionary contains pictures and examples, users are more likely to understand the definition. But it is a question, again, of the user.

I read a lot of books about a little bit of everything so it's easy for me to understand most dictionary definitions. But for someone with little or no interest in, say, psychology, astrophysics or politics, some definitions can be hard to comprehend, and in the end only with in-depth study of the topic will the person be able to comprehend dictionary definitions.

Is total terminology standardization attainable in science and education?

The problems with science and education are that practitioners tend to disagree a great deal and each definition and concept has its counter-definitions and concepts. Think of it this way, I know I'm exaggerating, but some people believe the world is flat, others believe the world is round and revolves around the sun.

Since new discoveries are made in science and education all the time, dictionaries will have to adjust to new discoveries. Furthermore an academic trend can erupt or disappear, a trendy concept can become outdated. And given the millions of scholars around the world, you cannot expect them to all agree on a definition.

So the best way to standardize terminology in science and education is simply to include the different schools of thought and to include some notes on how they agree and how they disagree. Some concepts of course form a large consensus, but when it comes to the details, there are often disagreements.

To what extent are term bases reliable?

Different countries with different administrative systems use completely different terminology and the terminology has different meanings depending on the context. People call Bibi Netanyahu the “Prime Minister of Israel” but a more accurate definition of “rosh hamemshala” would be “President of the government.”

Let's look at more precise examples. In Canada and the United States, the leader of a State or a Province is a governor and each state or province has different regulations as to what the governor's prerogatives are. In most countries a university leader is called a rector, but many universities prefer the term “president” of the university while others use the term “chancellor.” Each university defines the prerogatives of the university leader.

Examinations are called “qualifiers” in some countries and “partials” in others and each institution has its own precise definition of what examinations should be. In construction, some use the term “survey” while others prefer the term “study” and each institution will in some cases force the use of the word “survey” or “study.”

Translation in the end is for the reader, and if the reader is an institution, you better find the right terminology. Unfortunately term bases don't always take institutions into account, leading to bureaucratic hassles and delays because the terminology has to be changed. In some cases contracts can be cancelled, because they ordered a “survey” and you did a “study.” It's the exact same thing, but somehow they think it's different.

How do you successfully reduce lexicographic errors?

The more experience a translator has, the better. To reduce lexicographic errors, you need knowledge of both cultures, and more importantly, of both institutions. The French government has its preferred wording, while the Belgian government will use different terminology and the United Nations will use different terminology. African French-speaking countries are trickier, because they all have completely different terminologies.

Furthermore a good translator is one who researches terminology, not just on lexigloos or trados, but who actually examines documents from Belgium, France, the UK or different states in the US to make sure he is giving the most accurate translation he or she can possibly give.

Furthermore, translation is done for the reader. And the reader does not always know that the translator didn't know. As I described above, if you go to certain states and say you did a “study” of the ground, your project will be rejected because regulations mandates a “survey” of the ground. It's exactly the same thing, but administrators don't always know that.

Is there an objective gauge to the quality of translation? What are the best strategies for quality translations?

Again, translation is done for the reader and the institution reading the translation. That's assuming they actually read the document. That's a pretty big assumption, because a lot, if not most, translated documents go unread.

A good translation is one which is able to not only carry out the wording and the meaning, but also the cultural context, and more importantly the administrative context.

When a document is being translated, it is usually written in its original form by a person who is familiar with a set of cultural and administrative assumptions. So it is the translator's job to be able to read into those, and to ask themselves the following question: who is more important, the writer or the reader? If the writer is more important, as in literature, emphasis has to be put on the writer's culture. If the reader is more important, as in construction contracts, culture and administrative terminology has to fit that of the reader in the translated version. There are a lot of gray zones, where both the reader and the writer are important, as in presidential speeches intended to another president. That is for the translator to sort out. But there is certainly not objective gauge for the quality of a translation. 

How to best deal with translation loss when translating from a translation?

In a lot of academic and institutional settings, they use a system of translation and back-translation to try to guarantee accuracy. That is a document gets translated from French to English, the English version gets translated back to French to check for accuracy. Then the final version of the English translation will get translated to Spanish then back to English and so on.

But this system does not guarantee that cultural and administrative terminology will be accurate. Again the reader and the writer are the most important elements of a translation. That is, is the reader more important, or is the writer more important? In either case, you will need translators who are familiar both with the reader and the writer's culture, if possible one who worked and lived in both countries.

Can machine translations ever achieve a high degree of reliability?

Here's how machine translation works. There's a huge database of translated documents. The algorithm then looks at the sentence and its translated version. If the sentence already exists, it will give you the exact translation of the sentence that was used in the previous document. If the exact sentence does not exist in the database, machine translation will look at portions that have already been translated and piece them up like a puzzle, which results in funny, in some case hilarious translations.

There is no other way to get machine translation to work. Eventually the database will get bigger and bigger and the translations will be more and more accurate, but if you use slang, or original sentences, you will get very funny translations. If you put this article up for translation, you will get very funny, and unintelligible results.

The reason machine translation doesn't work is because grammar and semantics work completely differently in different languages. Machines can't choose from a set of terminology, nor can they translate the grammar accurately. How do you translate the Spanish subjunctive into English for example.

What are effective ways to achieve localization and internationalization?

Localization is a subject I have discussed at length when it comes to human translation. Unfortunately, when it comes to machine translation, administrative and cultural terminology changes all the time. You have to constantly keep up with governments, new laws, new institutions and new cultures to be able to keep up with localization. As for internationalization, even, suppose there were a world government, different regions and institutions of that world government would still use different terminologies.

Print - Comment - Send to a Friend - More from this Author

Get it off your chest
 (comments policy)

© Copyright CHAMELEON PROJECT Tmi 2005-2008  -  Sitemap  -  Add to favourites  -  Link to Ovi
Privacy Policy  -  Contact  -  RSS Feeds  -  Search  -  Submissions  -  Subscribe  -  About Ovi