CommTalk Podcast September 2006

Intoduction

This is a transcript of the September edition of the CommTalk Podcast. Audio files for the current show and archive episodes can be accessed from the CommTalk Podcast index page ( http://www.thecomwellgroup.com/commtalk.htm), which also contains details on how to subscribe to the CommTalk Podcast for FREE. Send us your feedback to mailto:info@thecomwellgroup.com

The ComWell Group: Welcome to CommTalk. A podcast for exploring the topic of business communications, and technical communications… anything to do with communicating with your internal and external customers.

This episode is all about Controlled English. In this three part series, we will be interviewing John Smart of Smart Communications about Controlled English, Simplified English, Plain English and the differences therein. For those of you who have never heard of Controlled English, it is basically a simplified version of English, in essence a subset, used for creating technical documentation intended for an international audience.

John M. Smart is President of Smart Communications, Inc. A computer scientist, technologist and visionary for Controlled English, Mr. Smart has degrees in computer science and electronic publishing from the Rochester Institute of Technology, NY. He has spent 19 years in software development, systems integration, electronic publishing and information technology. An author and frequent speaker at technical documentation conferences around the world, John has dedicated his career to Controlled English.
For more information on John, you can visit his website at www.smartny.com.

So lets get started.

TCG: Hello John. Let’s begin by telling our audience a little bit about yourself and what brought you to Controlled English.

John Smart: That’s an interesting question. The accent you hear is Australian, or Australian. I’ve been here over 30 years. I can actually work in British English, American English, New Zealand English, Australian English and South African English. Today I’m going talk to you about Controlled English.

TCG: Great.

John Smart: My love is the English language and technical writing. I have a degree in printing and publishing from the Rochester Institute of Technology and spent some intern time with Xerox and Kodak. So really come into the printed word, and that’s really brought me to the whole thing. Many moons ago, oh, I was working on globalizing a lot of documentation and I could see that the problem was not the translations, which were quite good, except the problem was in the fact that the input was really terrible. And bad English is just going to get you bad French. So I decided to do something about it and at that particular point, ran into the Caterpillar Company, then known as the Caterpillar Tractor, and worked on Caterpillar Technical English for a while and we were the licensees and we went off into the heavy construction area and started producing technical English that was simplified to roughly a 1200 word vocabulary, which I will describe a little bit later on, how you get to a 1200 vocabulary, when the English language is well over 900,000 words. Now we gotta find a way to get down there and I’ll explain that to you a little later. That’s how I got into it. We’re based in New York City, a city of about 174 different languages.

TCG: Wow, that’s a lot of languages for anyone trying to communicate to New Yorkers. How can you begin to manage your communications...

John Smart: I claim that they could all work in Controlled English and we are doing that. We have recently simplified the New York State Department of Motor Vehicles website. Put that all in – I’ll speak about this a little later- called Plain English, which is really for consumer. Then also speak a little in the tape here, a little further on, about the Simplified Technical English in the Aerospace and Aircraft Industry. For the listeners who might be flying in and they’re listening to this on an iPod, if you just reach in the front pocket of your aircraft, there, and if it says an Airbus, you are flying on two and a half million pages of Simplified Technical English.

TCG: Wow, that’s a lot of documentation!

John Smart: And you hope it’s correct. But it is correct, it’s been checked by our tools. But that’s generally how I got to the business.

TCG: So you’ve been at this for quite some time. But, for our listeners who are new to this topic, can you tell us a little bit about exactly what Controlled English is and describe some of the differences between, as you mentioned, Controlled English, Simplified English and Plain English?

John Smart: Okay. I’d like the audience to listen to get the three definitions that I have. I’ve got a Controlled English. Now, Controlled English is a vocabulary that promotes readability and usability in English technical documentation. Note my emphasis on technical documentation. This is not going to work for poetry, letters home to mom or the girlfriend; (Laughter) totally useless. But we’re talking today about technical English and publications managers who are facing the challenge of ‘how do I communicate with just the Economic Union, which requires 20 languages.’ And even more, Microsoft is localizing in 76 languages. So we’re going to use a Controlled language where we’ve picked roughly 1200 words.

The second definition is what we call, and what the world calls, Simplified Technical English. It has a standard, and it’s called A-S-D-S-T-E 100 Simplified Technical English, and that is the language that Airbus, Boeing, Embria Air, Bombay Air, and about 700 airlines all comply. That has 985 words as a small core vocabulary. You’re probably wondering ‘what’s he talking about, a core vocabulary’? If you could all imagine, out there, an apple; and you’ve got a core of an apple. You can take that core out and you can have different colored apples. The idea is that the core of the apple always remains the same and that’s your 800 to 900, roughly 1200 words in there. And then you’re wrapping nomenclature, or the part names, of any product or service around that and they are linking. Technical nouns, of course, make the vocabulary swell to probably 8500 words.

The third one, is what we call Plain English. And there is a movement afoot to put most of e-government into Plain English because of the tremendous diversity of this nation. We’re talking about, in these cities like Los Angeles, of 172 different languages being spoken; everything from Lao and Mong, to Balinese, to Mauri to these languages. All we have to do on there is we’re trying to get a common language that we can use administratively. It’s interesting that the United States of America does not have an official languages act. There is no official language of the United States. Where Canada, of course, many of you know, is English and French.

TCG: Right. But there is an effort in the U.S. government to move toward a simpler, plain language. I remember in 98, President Clinton issued an executive memo requiring agencies to write in plain language. Our listeners can go to www.plainlanguage.gov for more information.

John Smart: So we’re trying to form a Plain English. And the reason why people would do Plain English is because it’s becoming more difficult to communicate and we’ve become far more reliant on the web. So the World Wide Web, or the internet, we’re forcing people to go to the internet. And if you want a driver’s license in the State of New York, you will go to the web, and, I say, we’ve recently spent a lot of time, many years, simplifying that, so people could actually understand how to get a driver’s license in Plain English. We try to remove all the regulatory information and the legalese and the mumble jumble that the lawyers put, because nobody’s ever gonna read it anyway. So we want to get it very simple.

So we’ve got these three: We’ve got our Controlled English, which is probably what most listeners are interested in, which is for any product or service. We’ve got our Simplified Technical English because of the Aerospace Industry, and then we’ve got our Plain English, which is interesting to the people doing e-gov, forms development, insurance companies, who recently simplified an entire truck lease for Ford. And the reason why they wanted that simplified is because they’re leasing to the Hispanic audience and they wanted to get a very simple English and a very simple Spanish. So now there is an Hispanic specific auto lease, simplified Spanish that matches the simplified English. So you can begin to see the trend here and the key to this is, you’ve got to get control of the original English language and it can be in one of those three.

TCG: So, we’re trying to present information in a way that makes it as easy as possible for people to understand.

John Smart: And generally the idea is, a simple explanation is preferred over the complex. I’m only using the words of our Patron Saint, the philosopher Occam, known as Occam’s razor, that’s O-C-C-A-M R-A-Z-O-R. And he was born around about 1285, and he’s been running around the world for a long time. You know, the simple explanation’s preferred over the complex. In other words, if you’re going to write it very clear and concise, there’re going to be big returns on that investment. The investment’s modest, but on a global world, and most of our customer’s are doing business in 208 countries.

I’m seeing this more and more often, you know, there somebody’d be in Turkistan, the middle of Afghanistan; saw one the other day to Pashto. Pashto is one of the oldest languages of the world, but one of the languages that the United States Army had to focus on. It never did before, and they don’t have more than 50,000 words in Pashto, so obviously, with this reduced vocabulary, we could get to that level, to communicate with the people there.

TCG: So John, you spoke a little bit about the benefits of writing in a clearer Plain English, and the benefits for for not only your English speaking customers, but also your ESLs, or English as a Second Language readers. Let’s talk now about the benefits moving into translations and localization. Could you touch on some of those benefits?

John Smart: Yes, I think and I know your company’s involved in localization, and for you people out there listening, what localization is, it’s basically a translation, but more advanced services, and moving from one language to another and making sure all the screens and the prompts, and all these are in the languages. Now, where the Controlled Language really helps, and this does pay an enormous dividend, is that we remove a lot of the gobble-de-gook. We stopped serious errors that Apple had a few years ago when they had in their technical manual “type: enter”. I’ll repeat that again, think about it as I speak. “Type: enter.” Well, they did. E-N-T-E-R. And after 50,000 calls a month of “I am typing or entering into the computer the word ‘enter’ and nothing is happening on my Mac”. People began to see the problem wasn’t with the Mac, it was the fact that the technical documentation was incorrect. So, therefore, we do not use the word ‘type’ T-Y-P-E because in modern society, you really have to go to a museum to see a typewriter. There are no more typewriters.

TCG: Right. (Laughs.)

John Smart: A very similar word is ‘anti-clockwise’, ‘counter-clockwise’, I can’t remember which of these. All these clockwises are vanishing. In the digital society, you’re either going to turn it to the left or turn it to the right. Now, some of the older technical writers are going to be sort of annoyed, and I can sort of feel them ready to turn off their iPods at this point, but hang in guys. This is going to be interesting. The younger people coming along are entering a digital age. We're blogging, a brand new word; we got a new word, this is recorded in 2006, called mashups. I asked Google what a mashup was and they said it’s basically taking a whole bunch of Google functions, putting them through some Ajax code, and then stirring twice and you get these services like Google Maps out of it. So, we’re seeing this digital age form new words where the clockwise is vanishing, and going rather rapidly.

We’re also seeing a very interesting trend. There are only three countries in the world that are non-metric. And by, you listeners outside the United States, you may not understand, but, the United States does not use metric system. We’re a little old fashioned. And we are joined in the great company of Myanmar, which I call Burma and the Fijian Islands. So here we have the Fiji Islands, Myanmar and the United States of America are all on the non-metric system. Anybody under 40 today in the world, basically is trained in the metric system. Well, in the United States, everybody is trained not to understand the metric system, and there are tremendous problems that are being caused.

TCG: Right, it caused NASA to lose a $125 million Mars orbiter because Lockheed Martin was using English units while NASA was using the more conventional metric system.

John Smart: Recently, the wheels on the Disney Magic Mountain Ride in Tokyo fell off, and Disney did an investigation. Hold tight. No passengers were hurt; it was only a test run. And they wondered, “why would the wheels fall off”. The reason is that the Japanese had used United States English measure to refabricate a series of wheels in the metric system and there was this slight, ever so slight difference, which was enough for the wheels to fail. This is a big problem. And when we go in there and talk about temperature and measurements and so on, we’re beginning to see dual dimensioning also going along with the Controlled English.

So the Controlled English we’re trying to bring in both the metric system and an international vocabulary. We’re using words like ‘help’ or ‘assistance’ rather than other terms that might be used in there for getting medical assistance. We would prefer to say, let then, “call a doctor”, for instance, in the manual. If you see “Call a doctor”, well, nobody’s going to do that. And the other thing that comes into play in here with the metric system and simplification is this great thing that in America, we’re getting more famous for is, we sue everybody. We sue them for product liability. And that’s the failure of your product.

So we often deal with our clients in, going out and looking at the technical documents. Because a good plaintiff’s lawyer, that’s the mean lawyers – they drive around in very fancy cars. They do, well, because they’ve gotta afford the cars – So, what they do is, they normally get hold of your technical documentation. And your audience out there who are publication managers, once you print that part number on the back of your manual, you have now created a part of that equipment. Whether it be a digging excavator, or a financial software handbook, you’ve created a part. And that part can be deemed by a court to be faulty, and if it’s faulty, you’re liable for it. So, a lot of times, we’re looking at, “can we use a ‘might’, an ‘aught’, a ‘should’ or a may?” No. We’ve removed all of those from our Controlled English vocabulary. And they were already removed from Simplified Technical English. Because, I wouldn’t want to fly on an Airbus where is says “You might want to adjust the ailerons every now and then.” (Laughter) We don’t like that type of language, so we want to get very precise communication. It is functional but not elegant.

TCG: So why don’t more people do this? It seems to solve a myriad of problems from liability to localization.

John Smart: Well, I think the questions we’ve got a lot of times is people will come to us and say “Wow. You want to take away my creativity.” Well, our reply to that is “Yes, we’re gonna take away some of the writer’s creativity. If they would like to write the great American novel, they should do it on their own time, on their own payroll.” We’re trying to write technical documentation, clear, concisely and accurately that’s going to be global. We’re going to enable it for this massive globalization as we jump forward into this millennium where we’re becoming more and more global; everything comes from somewhere else. So that’s some of the reasons why, and some of the benefits and that does affect directly into the translations because we’ve got the key, to this little discussion here, is that you fix that English, you fix 29 languages and that’s really the key for people who are doing documentation, localization.

Your problem is not with your translators, your problem is with your office, and I would like to put the point of pain on the office and have them do a better job, and it’s really an educational job. But that author -. That author is in Dayton, Ohio; that author is writing for somewhere around the world, probably somewhere, where I’m familiar, Tasmania. And of course in Tasmania, its 24 hours of a time clock away, so a lot of these procedures are computer driven and of course, there’s no customer support that you can call from Tasmania to Dayton, Ohio, and anybody will be awake. So we’re beginning to see call centers; we’re seeing this in India where we’re going more from the voice over to the printed word. We’re seeing pdfs out on webs, so we’re beginning to make our authoring works truly global. Ten years ago, this was not possible; we would fax somebody something. But today you can go on most of the major websites and download the documentation and you can be in a different time zone. So all these things are beginning to influence why a Controlled English approach and why spend the time and do it. That gives you some of a few ideas.

TCG: Great. So, as a technical publications manager, I can ensure my writers are using good technical writing practices to create simple, easy to read documents that are ready for translation or use in the international market. Sounds great! How do I get started? Are my writer’s going to need to be trained to write in this style? Is there software? What’s my monetary output, as well as training to get started?

John Smart: Okay. Let me do a quick walk through the best I can put words into an audio picture. The first step, which is, we use a new word, it’s called ‘text mining’ and it’s very similar to mining for gold. What we want to do, is we want to pass a lot of text through what we call a text miner and we have a particular tool called Smart Text Miner. We get a Frame maker file, a Word file, SGML, XML and as you can imagine, we pour this into this into the computer and we want to see what words are going to come out, so then we can analyze our words. So what we’ll do is we’ll do the text mining and then we’ll go back to the client and say “Look, you use this particular word and we don’t think that’s carbium sulfur, we think it’s this, but you tell us.” and “You use this weasel word phrase and we just don’t like that phrase, can we remove it?” So, there’s a little negotiation between the client and then we suggest the client form a terminology committee to get agreement.

I think the world record for the ‘o-ring’ a little humble thing called the ‘o-ring’ found on most engines? We found 29 ways to write ‘o-ring’. Obviously, that’s 29 ways to confuse the end user. We got o dash ring, little quotes, and we got no standardization. So we’re trying to standardize on that.

So once we’ve got the text mining, to get the Technical English vocabulary, we then apply that core. Remember that apple core I talked about? We’ll drop that core in. That’s not going to change. It contains most of the common words like ‘push’ and ‘pull’ and ‘do’ and ‘go’. All the common words that, if the person claims to English literacy in a low literacy situation, maybe at the sixth grade, they will have that 1200 core, or they could be taught that 1200 core. Say they were working in a factory somewhere in the United States and they have low literacy. They could be taught that core language. Then, from the text mining, we’re able to wrap around the nomenclature. And if you listen to this iPod in the UK, that’s known as the nomenclature. Somebody said the other day “How did the nomenclature get into the nomenclature of English?” And actually, I’ve heard of a way of describing what technical nouns are.

In the old days of the Roman Senate, there was a person that stormed the US Senate, too, who stood at the door and announced, “this is Caesar Augustus from Syracuse”, and “this is Senator Caesar Somebody from…” -. That was announcing. ‘Nomen’ means I put a name on you. As you come through the door, I put a name on you. We want to do the same thing with the parts, field names, on software, and we want to form that standardization. So, now the first step, the hardest step, is we’ve got agreement on the technical nouns. We’ve got agreement from legal. We want to remove any words that – we call them weasel words, ‘cause the weasel is a North American animal, so it sneaks around and once they get into your manuals, they can create absolute havoc. So, we want to remove the weasels, and we want to remove a thing called gobble-de-gook, which writers are mostly familiar with. We get rid of all of that and we're going to save money later on because when we go to translate this, you know, if the translation place is full, you’ll be charged for that gobble-de-gook, so there are some savings in that, but more importantly, we want to get that clarity.

Now, having gotten that dictionary and everybody’s agreed; we’ve got legal approval on it, engineering approval on it, we’re ready to go. You could check all the documentation you write by hand, same as you can navigate an aircraft without any flight instruments. You can do it, but it’s quite difficult. So, what we developed many, many years ago, we actually developed it for the Nortel Networks which is actually making this iPod recording possible over a giant switch. Nortel’s problem at that particular point, is they had to check millions of pages. What you may not know is just the documentation to operate the ring tone on your cell phones or your mobiles, that’s a quarter of a million pages of documentation. And Nortel sold all this to the Brazilians and of course, it would take a long time to translate a quarter of a million pages in the time they wanted. So we adopted a Controlled English or Nortel Standard English program so the Brazilians at Embartel in Brazil could immediately understand the new cell phone technology while the translations were coming along years later.

So once we have developed this vocabulary, we needed a way to check it and we created a thing called the Maxit Checker. This is a tool that uses another science called Artificial Intelligence. I can hear a few people guffering, – they got a guy being interviewed here, his name is Smart and he’s into Artificial Intelligence and what’s his biggest product? Fuzzy logic. (Laughter.) Well, if you combine all those together, you do get Artificial Intelligence and you do get fuzzy logic, but what the computer can do, and I know many of you are probably really thinking, “Wow, this is insane. How can a computer analyze text?” Well, we’re getting some very fast computers with 286 gigahertz Pentiums, a lot of memory, yes, you can do it.

So we formed a model of English grammar and what it does, it just reads the text against a model. It’s a mathematical model of English and when it finds an exception, if you write “This is good.” Well, you don’t know what ‘this’ is. Nobody knows what I’m thinking about ‘this’ is. So the computer’s going to put a little mark on there and it marks the text and says “Please go off, writer, and find what is the subject of this sentence?” And if you say “This recording is good.” now I have informed the reader. So the computer has about nine and a half thousand rules; very similar to somebody who may at one time struggled with the Chicago Manual of Style, that giant book of grammar?

TCG: Oh, yeah. I think it’s up to a thousand pages now? A little light reading...

John Smart: I don’t recommend anybody trying to wrestle with it every day. We’ve essentially said the Chicago Manual of Style, modified for technical writing, not expository writing; inside a computer, it’s going to check it in seconds. It takes a second a page to analyze it. It remembers everything and then it’s making suggestions. So, on that case of the O-ring, you can say “Please dear writer, this is the way we want to spell ‘O-ring’. A big ‘O’, a dash, and a small R-I-N-G.” and it can detect right down to that particular level of capitalization and spelling. It can also find gobble-de-gook. If we said something like “In this case.” That’s confusing because as soon as it gets translated in Chinese, the Chinese want to know where you put the suit case. (Laughter.) I’ve seen better than that in China. I’ve actually seen a Coca-Cola mistranslated as “Wax Tadpole”. (Laughter.) I hear a few laughter, but that’s good. You’re probably wondering how can the Chinese get a wax tadpole?

TCG: How can anybody get one?

John Smart: Well, I don’t know how they got a wax tadpole, but the idea is that the Maxit Checking tool would never allow that to happen and we’re able to avoid a lot of embarrassment with the checking tool. It’s very fast. It is providing the guidance. The writer, of course, has the final say on the computer and the computer is not going to put any writer out of business, but it’s very helpful. It actually helps people who are learning to write English as a second language. We have a lot of computer engineers who have immigrated to the United States, I’m speaking of the US market, who are sort of struggling with English and they don’t know a participle from a gerund and an article, but the Maxit computer will tell them, and they soon learn that when they see an I-N-G word, which is a gerund, and that appears on the computer in green, that doesn’t mean good to go. That means I've got a problem and I’ve got to turn that Gerund form, that I-N-G, into an infinitive form of the verb, or put a subject in to show who does what to whom.

After a little while of doing this, they soon learn the computer expertise. It’s as simple as that. We do suggest that some training is needed because we’re all good writers, I’ve been at it for over 18 years and I know you people have been at it for a while. But we all tend to forget our principles of grammar, so we’ve got some courses that reinforce the grammar of the technical authoring capacity, and we can do it by e-learning in about 20 hours, or you can go what we call “commando style”, we do go out and actually teach technical authors with the Maxit Checker on their computers, in a classroom, lock them up and say “You’re not leaving until we’re finished.” (Laughter.)

But it works very, very well. That’s the commando style. We’ve taught probably 10,000 authors, technical writers around the world. We have only lost one writer, and I like to do exit interviews of writers who don’t like this, there are a few, and I asked that writer, I said “Why will you not continue?” and he said “It’s against the will of the Lord” and I said “Well, I’m not prepared to dispute the will of the Lord, but that’s a pretty good excuse.” So I figure one out of 10,000 isn’t a bad thing.

So, most writer’s initial reaction to it, Controlled English, is absolute horror. The first four hours of our seminar, they sit there stone faced, and all of a sudden the ice sort of breaks and then they begin to see it’s quite simple. You have less choices to make. You can pick the nomenclature that’s all been pre-selected for you in the computer. You’re sentences are short. We don’t want long sentences. I saw one the other day, world record, won’t mention the company; 176 words in that sentence.

TCG: Wow.

John Smart: Yeah, wow. I mean and the author was proud. We explained to the author that this document was destined to become part of an oil drilling rig in Saudi Arabia in Hassan. By the time he got to 17 words, had basically given up and then would decide to take the hammer to the rig in the appropriate place that he thought would be the place. Obviously 176 words in a sentence is not good. I saw sometime ago where a major automotive company made, I think, a major mistake. They had used very small, nine point type and they had achieved a miracle of getting 1200 words on a page. So we sat down with the purchasing agent and he said “Well, I don’t know if I like this whole Controlled English thing because you know, we’re gonna have to print bigger books because you want more white space, and we’d been working diligently getting the white space down.” Well obviously, a good technical author and good technical writing does rely on a good use of white space. If you’ve got 1200 words on a page and it’s jammed together like a legal agreement, it’ll be read like a legal agreement. Nobody’s gonna read it.

So obviously, we’re beginning to change both the language and we’re changing the design of the document. So we want to design for usability; we want to design for readability. If there’s something we can’t explain in Controlled English, let’s put a picture out there. Pictures are becoming cheap with digital cameras; we could put a picture, so often a picture will be inserted in there. Let’s make a bulleted list. We don’t like large blocks of text. In fact, in the Simplified Technical English Standard that Airbus relies on, a paragraph can only contain six sentences and the Maxit Checking pool is sitting, monitoring that, and no writer’s going to spend the time doing that. And they want 21 words in a sentence. And if you’re writing for the United States Navy, you’ve gotta go down to 17 words. I think that’s quite a good rule because the Navy has quite a few missiles that could go wrong at the wrong time if they were not understood.

But, in light of that, we do a lot of work and I would encourage anybody doing a lot of work in the area of mission critical documentation. A lot of it’s there for mission critical, is really look at your sentence length and adopt this type of technology, because if it’s gonna go wrong, I think somebody said it was Murphy’s Law, it’ll go wrong. And if it’s wrong in the books, it’s gonna be wrong all the way through. I think in the area of successes, we had a wonderful success with Nortel Networks, the people that do bring you your internet connection. When we arrived on the scene, they’re professional technical writers and they’ve been doing it a long time, since 1919. That’s a pretty long time to be writing documentation, and most of it’s pretty bad. And we sort of looked at it and we said, “Wow. How do we make this?” And the Nortel audience before that was the seven Bell operating companies, and of course they only had to sell to the Bell operating companies. They really didn’t care and the Bell operating companies could call them any time, but soon as we went global, we had to change that and that was mission critical.

In the first year, we were able to remove 42 percent of the procedural errors out of that technical documentation. That’s a lot of errors. And that translated into millions of dollars worth of savings in time of customer support. They weren’t getting these calls from panicked operators in Croatia, saying “I’ve got the switch; I’m pulling the circuit card, and I’m casting J-9 and nothing’s happening. And it says in the book five volts.” When somebody had made an error, it was minus five volts or eight volts. So we began to see that the technical documentation was far more important than anybody at Nortel, the original management, thought. Before, I think, they produced it by the pound or the kilogram to impress the boss. And a lot of it was written by an engineer to impress his boss. Well, we changed that paradigm, and we said, now, you’re the writer, you’ve got to understand the customer and be the voice of the customer. You’ve got to go down to the voice of the customer, which is basically a six sigma principle, which people like GE, Motorola and Toyota have all gone into six sigma. So we’re able to actually implement six sigma quality control over the top of Controlled English, so that was my third step.

Really my fourth step, I’ve got my dictionaries, I’ve got my electronic checking, which is really my electronic navigation. I’ve got my training; I’m bringing new people on line. Train a few people in the company then they’ll train the others. And then the fourth one is “Let’s get the quality metrics going because if you’re an ISO 9000 shop or you want to have six sigma, this is the way to go and we’re actually able to quantify documentation and say this is a five sigma or four sigma. Anywhere over a four sigma is good. I won’t belabor today, the points of six sigma, but you can get a very good introduction by doing a Google search on the word six sigma and also, you can to the General Electric website where they have a very nice tutorial. But essentially, six sigma means that you want 3.4 failures in a million.

Well, writing is a little different from making widgets. So, we want to remove the strange English and the complicated English and the awkward English from there, which counts as a quality metric. And when you do look at Controlled English, it’s really as a joy to look at it. And one of the most interesting things about it, as we work around the world, we’ve had people writing Controlled English in South Africa, New Zealand, Australia, and the United Kingdom. I can hold those texts from those countries and you cannot tell where it was written. People ask, is it British? Is it American? No, it’s an homogenized form of English. We may use a few American spellings, for instance color without the U, the British will be rather annoyed, but to avoid Her Majesty’s rough, we have put back the U in color but you can see that that is not going to destroy the meaning of everything. We could use the U in color in the United States and people would understand what we meant. So those are the words, there’s only probably 40 or 50 words between the two English languages that we use in Controlled English that would have different variant spelling that might be confusing but most of them are not. Most of them are very, very minor on it. But essentially we are shooting for a global audience and we want our documents to be as easy to read in Beijing as they are in Dayton, Ohio.