The language of science q

I am honored by President Joachim Hasse and the European Association for Cardio-thoracic Surgery (EACTS) to have been invited to speak to you today. It is a special pleasure because to celebrate the millennium the Society of Thoracic Surgeons (STS) and EACTS are honoring one another, and the seeds for this coming together began on both sides of the Atlantic during the year I had the privilege of presiding at the STS. Why did I hope the language of science might be of interest to you? The EACTS, with English as its of®cial language, is one of the great international melting-pots of thoracic surgery. For many of you English is a second language, and I supposed that this might present you with more challenge in presenting your work than it would be to do so in your own languages. Why is this topic of interest to me? I was stimulated to think about differences between the research environment in the US and elsewhere when David Nahrworld sounded an alarm in 1995 with the observation that the number of publications from US authors in peer-reviewed surgical journals had been decreasing [1]. He said `The most striking ®nding was the dramatic, signi®cant fall in the number of1⁄4research pages from the US and the increase in1⁄4pages from countries other than the US1⁄4' Most of the American Surgical Association audience believed this re ected a decrease in time and funds allotted to research in the US. It took a French colleague, during the discussion period, to reassure us Americans and to make us chuckle. Professor Henri Bismuth of Villejuif, France responded by saying that this `1⁄4may re ect the increased ability of non-American surgeons to write English1⁄4It is like the French complaining of more Americans in French restaurants. It does not mean that French taste is losing.' John Waldhausen, editor emeritus of our sister journal, added perspective by saying, `1⁄4we should be thankful that these manuscripts are1⁄4in English rather than in Mandarin Chinese, the language spoken by more people than any other in the world.' Shortly after the Nahrwold±Bismuth±Waldhausen exchange that called attention to language, my interest was further stimulated when I lectured in Vienna. My offer to speak in Viennese German, my ®rst language, was politely declined in favor of English ± their of®cial language. I then realized how much more work it would have been for me to speak in my mother tongue than in my daily language. During the discussion period, which was in German, my audience understood better, and expressed itself more clearly and succinctly, in its own language than in English. This was true despite their sophistication and good knowledge of English. Subsequent visiting professorships in the Netherlands, France and Japan enhanced my awareness of the language burden faced by English second language scientists. My personal editorial experience for the Annals of Thoracic Surgery (ATS), the European Journal of Cardiothoracic Surgery, and a number of other journals has made me sensitive to today's topic. In addition, friends in Europe and Asia, who have impeccable skill in their own languages, have repeatedly asked me to review or edit manuscripts from the language viewpoint. Ladies and gentlemen, with native tongues other than English, my hat is off to you. You have accepted and coped admirably with an English language burden, and my hope is that you will see my remarks as constructive. I began my audacious venture into language studies by consulting the late Professor Victoria Fromkin, former chairperson of our UCLA Department of Linguistics. She reminded me that mathematics is the most durable language of science and she stressed a commonalty among written European Journal of Cardio-thoracic Surgery 18 (2000) 642±648


Introduction
I am honored by President Joachim Hasse and the European Association for Cardio-thoracic Surgery (EACTS) to have been invited to speak to you today. It is a special pleasure because to celebrate the millennium the Society of Thoracic Surgeons (STS) and EACTS are honoring one another, and the seeds for this coming together began on both sides of the Atlantic during the year I had the privilege of presiding at the STS.
Why did I hope the language of science might be of interest to you? The EACTS, with English as its of®cial language, is one of the great international melting-pots of thoracic surgery. For many of you English is a second language, and I supposed that this might present you with more challenge in presenting your work than it would be to do so in your own languages.
Why is this topic of interest to me? I was stimulated to think about differences between the research environment in the US and elsewhere when David Nahrworld sounded an alarm in 1995 with the observation that the number of publications from US authors in peer-reviewed surgical journals had been decreasing [1]. He said`The most striking ®nding was the dramatic, signi®cant fall in the number of¼research pages from the US and the increase in¼pages from countries other than the US¼' Most of the American Surgical Association audience believed this re¯ected a decrease in time and funds allotted to research in the US. It took a French colleague, during the discussion period, to reassure us Americans and to make us chuckle. Professor Henri Bismuth of Villejuif, France responded by saying that this ¼may re¯ect the increased ability of non-American surgeons to write English¼It is like the French complaining of more Americans in French restaurants. It does not mean that French taste is losing.' John Waldhausen, editor emeritus of our sister journal, added perspective by saying,`¼we should be thankful that these manuscripts are¼in English rather than in Mandarin Chinese, the language spoken by more people than any other in the world. ' Shortly after the Nahrwold±Bismuth±Waldhausen exchange that called attention to language, my interest was further stimulated when I lectured in Vienna. My offer to speak in Viennese German, my ®rst language, was politely declined in favor of English ± their of®cial language. I then realized how much more work it would have been for me to speak in my mother tongue than in my daily language. During the discussion period, which was in German, my audience understood better, and expressed itself more clearly and succinctly, in its own language than in English. This was true despite their sophistication and good knowledge of English. Subsequent visiting professorships in the Netherlands, France and Japan enhanced my awareness of the language burden faced by English second language scientists.
My personal editorial experience for the Annals of Thoracic Surgery (ATS), the European Journal of Cardiothoracic Surgery, and a number of other journals has made me sensitive to today's topic. In addition, friends in Europe and Asia, who have impeccable skill in their own languages, have repeatedly asked me to review or edit manuscripts from the language viewpoint. Ladies and gentlemen, with native tongues other than English, my hat is off to you. You have accepted and coped admirably with an English language burden, and my hope is that you will see my remarks as constructive.
I began my audacious venture into language studies by consulting the late Professor Victoria Fromkin, former chairperson of our UCLA Department of Linguistics. She reminded me that mathematics is the most durable language of science and she stressed a commonalty among written and spoken languages. Professor Fromkin also said that language is a political weapon, modi®ed by social and economic differences. Christine Holten of our Department of Applied Linguistics, encouraged and guided my start and she introduced me to my co-author, Kathy Howard. Kathy is completing her doctoral thesis and was recently awarded a Fulbright Scholarship for linguistic studies in Thailand. She has been teaching English to international UCLA graduate students who scored low on a routine pro®ciency test, and she brings the perspective of a language professional to today's topic.

Historical background
The evolution of the language of science from Egyptian, Greek, Arabic, Latin, French and German to English has certainly been a function of the amount of information that was being transmitted in a particular language. The changes that occurred were entirely compatible with Professor Fromkin's contention that the choice of language is related to politics and economics. The oldest medical writing is the Kahun Papyrus which deals with veterinary medicine and women's diseases [2]. The earliest known written communications about surgery are the Edwin Smith Papyri, which came from Egypt in about the 17th century BC. These relics can now be found at the Royal College of Surgeons of England. Hippocrates (460±370 BC) illustrates the important place of Greek among the languages of science, albeit that the enduring oath attributed to him was probably not his writing, and some scholars have even questioned his existence. Maimonides (1135±1204), a great Hebrew physician, philosopher and ethicist, wrote in Arabic and Hebrew. Perhaps if Maimonides were alive today, the tensions in the Middle East could be resolved. As the strength of the Romans increased, the language of science became Latin. Latin remained a prerequisite for medical education in Europe until about 20 years ago, and to this day most scien-ti®c terms and names are still based on Greek or Latin. In the 17th and 18th centuries French was perhaps the most important language of science, sharing its prominence with German. In the latter part of the 19th century and the early part of the 20th century, the Bismarck era, French remained the language of diplomacy, but German became the dominant language of science. Around the turn from the 19th to the 20th century Bismarck is said to have been asked what had been the most signi®cant development in his lifetime. He replied`the fact that North America speaks English' [3]. This was an amazingly farsighted comment! German deserves special attention because it is the immediate predecessor of English as the language of science. Professor Ulrich Ammon of Duisburg has pointed out that from 1908 to 1937 there were 8 Nobel Laureates in Physiology and Medicine from Russia, Sweden, Great Britain, Denmark, the Netherlands and Hungary who had significant portions of their scienti®c education in Germany [4].
Indeed, when my father earned his Doctor of Medicine degree from the University of Vienna in 1924, German was clearly the language of science and US academic surgeons regularly took periods of their education in the great surgical clinics of Europe partially in quest of a working knowledge of German. For example, my chairman, Lester R. Dragstedt of the University of Chicago, was a physiologist who had essentially all of his surgical education in Europe before he became one of the world's great gastrointestinal surgeons. After World War II, when political and economic strength centered in the English speaking nations, English became the language of science. The continuing growth of the use of English in medicine is shown in Table 1. We in the US have been enriched by the in¯ux of international scientists into our laboratories and clinics.

English as the language of science
How have publications and scientists adapted to English as the language of science? With no effort to be complete, we identi®ed 11 journals relevant to our specialty that changed from their native languages to English between the years 1963 and 1998. It was not surprising that Hans Borst, in 1978, was the ®rst one in our specialty to shepherd the change from its native tongue to English. The journals that changed their languages to English included nine in Germany and two in Japan. Three of these 11 journals are in thoracic and cardiovascular surgery. We found it interesting that the current editor of the European Journal of Plastic Surgery, which changed from German to English in 1986, is I.T. Jackson, a Fellow of the Royal College of Surgeons, who works in South®eld, Michigan, and who Table 1 Languages used in medical publications cited in medline [5]  appears to be native English speaking. Editing and publishing in science has indeed become international! The 11 editors of the journals that had changed their languages to English were queried by mail as to why the language change had been undertaken and whether the change had achieved the desired goals. We received ®ve responses from the survey ( Table 2). These included three from the current or past editors of publications in thoracic and cardiovascular surgery, and two from the editors of the interdisciplinary journals that focus upon cancer research and animal science. Three of the ®ve editors said that the change had accomplished the goals of greater readership and attracting better, more international manuscripts (Table 3). One editor of a journal that was founded in 1953, which changed its name in 1998, expressed his concern that the name change had not yet fully accomplished the goal of attracting more`world class' manuscripts to his journal. While he may have been unrealistic in his hope to achieve his journal's goals within two years, this editors' answer suggests the possibility that a change in language may not be enough to alter manuscript submission patterns. None of the journals suffered any resignations from their editorial boards as a result of the language change. However, the Journal of Experimental Animal Science elected to appoint an entirely new editorial board as part of the change, and the Japanese Journal of Thoracic and Cardiovascular Surgery added 13 new members to its board. Up to four additions to the support staffs were made as a result of the change from native tongues to English.
What has been the experience of the ATS, which I have had the privilege of serving for 22 years, with international authors? Thomas B. Ferguson, the Editor Emeritus, and Carol L. Blasberg, the Administrative Editor Emeritus provided me with the information I needed to address that question, and I am grateful to them for this. From 1995± 1999 we have been delighted to see ATS become more international as illustrated by an increase of manuscripts from international authors from 52% of 1778 manuscripts to 59% of 2185 papers. During the same ®ve-year period the number of manuscripts received from non-native speaking (NNS) authors increased 40% as compared to a modest 5% increase in manuscripts from authors in the US and Canada.
During the year 1999, our most proli®c international contributors were Japan, Germany, Italy, England and France; the next ®ve were Sweden, Turkey, Switzerland, India, and Australia. Thus, 80% of our top ten international contributors were from non-native English speaking (NNS) nations. The acceptance rates of original articles submitted to ATS in 1999 are shown in Table 4. The mean acceptance rate was about 49% among the top ten international contributors as compared to a mean acceptance rate of 78 (71± 93)% for native speaking (NS) authors. Among NS nations, Australia, submitted 14 original articles of which 13 were accepted, and thus the Australians had the highest acceptance rate of 93%.

Language burdens of NNS authors
We wanted to focus upon the language burdens of NNS authors. Toward this goal, we reviewed the manuscripts and all correspondence for all 50 original articles about lung cancer from volumes 65±68 of ATS. There were 27 papers from NNS authors from Japan, France, Italy, the Netherlands, Austria, Taiwan, Germany, Switzerland and Korea. The 23 NS papers were from the United States, Canada, Australia and the United Kingdom. Case reports, review articles and letters to the editor were not included. The signi®cance of differences between NS and NNS manuscripts was evaluated using Student's t-test. This statistical method was imperfect because the number of reviewers for the various manuscripts differed from one to four, but we chose to use an imperfect method in preference to using none at all.

Acceptance rates
The reviewer's recommendations for the 50 lung cancer articles are shown in Table 5. The rate at which the manuscripts were given an accept, revise or reject recommendation by individual reviewers was calculated. In the initial reviews, NNS authors' manuscripts were given an accept evaluation by 24% of the reviewers, on average, as compared to 41% for NS authors (P , 0:05). However, all manuscripts, including those that had been accepted, were required to be revised once before they were published. There was no signi®cant difference between NS and NNS manuscripts in the rejection rates.

Ratings
The originally submitted manuscripts had been rated by reviewers as a routine part of the ATS editorial process on a scale of 1±4. This scale, wherein 4 is the highest rating, had been applied in the categories of scienti®c accuracy, originality, interest to readers. The mean ratings of NS manuscripts varied from 2.76 to 2.81 for NS manuscripts and from 2.57 to 2.83 for NNS manuscripts. Thus, there was no difference in quality between NS and NNS manuscripts as judged by the reviewers by these criteria.

Reviewers' comments
We categorized these comments according to language, organization (genre) or writing quality, using criteria for each of these categories as described below. Table 6 shows that the number of NNS manuscripts that received at least one comment was apparently far greater than the number of comments for NS manuscripts. However, only the numbers pertaining to language and writing quality achieved statistical signi®cance.

Language
Any comment that mentioned a problem with the language, including any speci®c or general comments about grammar, word choice, and inappropriate register or style, as well as suggestions for the reformulation of speci®c sentences or passages. In this most frequent category, eight of 23 NS manuscripts or 35% prompted comments as compared to 20 of 27 or 74% of NNS manuscripts (P 0:003). The total number of errors in the category was 11 for NS manuscripts and 20 for NNS papers. The number of comments per manuscript varied from 0±3 for NS manuscripts and 0±23 for NNS papers. The following are examples of comments in this category: One reviewer said`Some of the English structure and spelling needs revision. 'Another reviewer for a different manuscript said`I would suggest you send it to your`grammar police'.' More speci®c comments, with which we did not always agree, included`The term metastatic lung cancer is ill advised' or`The sentence`¼does not make any sense.'

Organization (genre)
Any comments about problems with the appropriate structuring of the manuscript, including comments about organization, length, or appropriateness of any section of the manuscript. The frequency with which reviewers commented about organization was apparently greater, but without statistical signi®cance, for NNS manuscripts as compared to NS. The following are examples of comments in this category: One reviewer said`The discussion is not precise, but rambles'. For another article, a reviewers said The discussion is long and is often more a thesis and testi-mony¼rather than a discussion of the author's data.' An example of a more speci®c comment was`The selection criteria should be in the Materials and methods and not in the Results.'

Writing quality
Any positive or negative comment about the quality of the writing that did not ®t into either the category of language or organization. The frequency with which reviewers commented about writing quality in NNS manuscripts was greater than that for NS manuscripts (P 0:03). The following are examples of comments in this category: One reviewer said`This section needs to be re-written so that the reader can understand what the authors are trying to a The differences in recommendations for acceptance and revision were signi®cant (P , 0:05). There was no signi®cant difference in the rate of recommendations for rejection. b Native speaking authors. c Non-native speaking authors. explain.' and another said`The paper is very poorly written and is very dif®cult to understand.'

Error analysis
A randomly selected group of 20 manuscripts, including ten each from NS and NNS authors, was chosen for error analysis by a language professional (KMH). The reviewers comments regarding linguistic issues were compared to the ®ndings of the language professional. When error analysis of the ®rst ®ve of the NS manuscripts revealed no`errors', we decided not to pursue analysis of the remaining ®ve NS manuscripts and to focus our efforts upon the ten NNS papers. A total of 263 errors, distributed as shown in Table 7 were found. The three most common types of errors were as follows: incorrect use of an article, improper word choice, sentence structure that was grammatically incorrect. An example of a typical missing article was`Therefore metastases which relapse in the lung in a surgically treatable manner should be considered as indication for reoperation. ' Improper word choice was illustrated by`Our study votes for surgical treatment of patients with recurrent resectable pulmonary metastases.' An example of an ungrammatical sentence structure was`These studies¼have promoted us use MVP chemotherapy in this study.'

Correlation of reviewer comments and error analysis
The peer reviewers made 51 comments about language as compared to 263 errors that were identi®ed by the language professional. The difference in the focus of the peers and the language professional was shown by the fact the type and number of errors identi®ed by the peers and the language professional did not uniformly coincide. Incorrect usage of articles was found by the language expert to constitute the largest number of errors in the manuscripts. Peer reviewers focused more often on word choice, and the need for correction of grammatical errors in a general way. An example where one reviewer noted improper word choice was when he noted a sentence that read`The rate of silent metastasis is considered to be actually smaller...' and suggested that this be rewritten to read,`The rate of silent metastasis may actually be lower¼' 4.9. Comparison of changes made by language professionals versus professional peer Several unclear passages were chosen from NNS manuscripts that were from ATS, and from two NNS manuscripts not from ATS. Many passages noted to be unclear or suboptimum by the peer reviewers went unnoticed by language experts. For example, an article that had been edited by a German language professional included a sentence that read,`A 31-year-old female presented with relapsing, severe hemoptysis after suffering from coughing and asthma-like attacks since 1.5 years.' This was revised by one of us (JRB) to read`A 31-year-old woman presented with recurrent severe hemoptysis after having suffered from coughing and asthma-like attacks since she was 18 months old.' The peer reviewer was better equipped to advise the authors about word choice, especially regarding precision in using medical terminology. In the same paper, the original version said`¼intra-operatively, no mention was made to any injury to the ipsilateral bronchial system.', and the peer reviewers said`¼ the operative report made no mention of any encounters with the bronchial system.' Here, the translator's passive voice formulation`intra-operatively no mention was made¼' was changed to the active voicè the operative report made no mention¼' This change also required a change in word choice from`intra-operatively' to`the operative report'. The word choice changes in the above examples would not likely have been made by non-expert editors or translators because they require a technical knowledge of the ®eld of medicine.
We found that changes made more easily by the peer reviewers often clari®ed passages in which it was unclear to what a noun or pronoun referred. In English we use reduced versions of the full noun phrases-inde®nite noun phrases, pronouns, and demonstratives to refer to earlier mentioned objects or persons. However, these conventions differ from language to language and they are dif®cult to master. Sometimes these reduced nouns or pronouns are ambiguous, i.e. one can no longer identify to what they refer. For this reason peer reviewers are better equipped than language professionals to identify the missing information and to correct the text. Examples we found are as follows: The NNS author and language professional said Tracheobronchial injuries are usually caused by blunt, penetrating, inhalation or iatrogenic traumas and most of them occur within 2.5 cm from the carina; lobar or segmental bronchi are seldom affected.' The peer NS editor said Tracheobronchial injuries are usually caused by blunt, penetrating, inhalation or iatrogenic trauma. Most injuries from blunt trauma occur within 2.5 cm distal to the carina; lobar or segmental bronchi are seldom affected.' In the language professional's ®rst version, the noun phrasè most of them' could refer either to all tracheobronchial injuries, or to those caused by blunt trauma. The peer reviewer was able to clarify this passage by replacing the ambiguous pronoun`them' with a full noun phrase modi®ed Table 7 Error analysis of ten manuscripts by non-native speaking authors a Total number of errors 263 Missing or incorrect article 52 Improper word choice 49 Ungrammatical sentence structure 28 a The above were the most common types of errors; 134 other errors each comprised less than 10% of the total number found. by a prepositional phrase`injuries from blunt trauma'. It is possible that a language expert could notice this ambiguity of reference, but, he or she would need clari®cation from the author or from a medical expert before reformulating the passage. The point we wish to make is that authors and language experts should work together.
Our observations are in keeping with those of an independent language professional who said that`¼feedback given by non-expert writing teachers¼addresses¼problems of form and/or presentation, whereas¼the feedback of experts¼focuses on ideas and content' [6]. This reinforces our contention that language professionals and peer reviewers have separate but complementary roles in the publication process.

Discussion
What has our study shown? Scienti®c publishing and the use of English have become more international than ever before. NNS authors in ATS have adapted well to the use of English, but they have had to work harder than NS authors. Language professionals can identify patterns of dif®culties for NNS authors. Professional peers are better than language professionals in helping NNS authors to convey their messages well. In general, neither peers nor language professionals alone suf®ce to make the best of NNS manuscripts, and so input from both should be obtained.
In the ideal world we would have an eternal international language of science that everyone would master. However, dialects, jargons and subtleties interfere with this throughout the world and also among English speaking people. Winston Churchill addressed this issue when he said that the English and the Americans are two people separated by a common language. The thesaurus alludes to subtle differences in language by giving the words dialect, speech, idiom, vernacular, jargon as synonyms. My experience when we visited the Jacquemart Andre Museum in Paris while on our way to participate in the May meeting of the European Association of Science Editors (EASE) also illustrated the issue of language subtleties. The place mats for the museum restaurant acknowledge the importance of English by forbidding cigarette smoking in two languages. It says,`Votre amour de l'art est certainement plus grand que votre de Âsir de fumer. Merci de vous abstenir.' In another corner it says,`Please accept our thanks for not smoking. As an art lover you will appreciate the need to refrain.' The French version started with the love of art and the English version started by asking one not to smoke. In Chicago we might simply have had à No smoking' sign, and in Frankfurt it might say`Rauchen verboten'. My point is that some of the beauty of language lies in its subtleties. In any language, the ability to express subtleties and to appreciate them, is a mark of mastery. However, it is unrealistic to expect all NNS scientists to master the subtleties of English.
How well can one expect a non-native English speaker to do with English? We have con®rmed that some NNS authors do superbly with English, and for others it is very hard work. This variable ability to learn second languages, as well as the difference between learning to write and to speak, is illustrated by a remarkable Chinese author who came to Brandeis University in 1985 as an adult, having previously been an soldier in Mao's army. He is now Professor of English at Emory University. In 1996 he won the Flannery Award for Short Fiction. In 1997 he won the PEN/Hemingway Award, and last year he was the National Book Award Winner for a book entitled`Waiting' which I recommend to all of you. Had time permitted, I would have read you some passages from his book to illustrate the beauty of his writing. However, he apparently cannot speak English very well because in the New York Times it said,`The e Âmigre Â novelist, a former soldier under Mao, still has trouble speaking English. So how can he write like Henry James?' [7].
In thoracic surgery some of our NNS authors write well, but they do not speak well. That's quite alright, and we need to be tolerant of that, particularly since most of us don't speak their languages at all. I suggest that we be so tolerant as to establish a mechanism whereby we encourage thoracic surgeons who write English well but cannot speak it as well can have their papers presented by a colleague who does speak well. It's better to have a delegate present one's work clearly than to impair the message through language that the audience cannot understand.
Before closing, I would like to tell you some of what I have learned in the past year about how some of our NNS colleagues have coped with the English language burden. During last Fall's meeting of the Japanese Association for Thoracic Surgery, where I was given the courtesy of delivering my address in English, I learned that there is an Institute Medical Communications Center in Tokyo which was established`to increase the contributions to the international research literature from Tokyo Medical University and....to develop their medical English ability and to enhance their ability to communicate internationally.' The founder and still leader is the American, Professor J. Patrick Barron. Barron wrote me that he was at ®rst reluctant to undertake projects in thoracic surgery, but decided to accept when Professor Hayata assured him that he would teach Barron enough thoracic surgery to do his job. At the National Cancer Hospital in Tokyo each of the regular weekly conferences is once each month held completely in English. As a result of this, the participants in a conference to which my hosts took me without forewarning were able to change to English essentially as I walked in the door. In Europe there is a remarkable group of language professionals and editors called the European Association of Science Editors (EASE) which publishes its bulletin entitled European Science Editing. When I participated in the EASE meeting in Tours last May, Professor Jacques Lansac welcomed the group on behalf of the University of Tours with a wonderful description of the history and the mystery and beauty of the region. He spoke well in English, but his message would have been even better in his classic, beautiful French. A wonderful wine tasting session that followed was conducted by a charming young French lady, assisted by an American translator who lives and works in Paris. The point is that Europeans, Asians and NNS authors in general have met us more than half way in adapting to English as an international language, and we should be grateful for those efforts. As we returned from Tours to Paris, Philippe Dartevelle, of the Marie Lannelongue Hospital and the University of Paris Sud was kind enough to allow me some insight into his own formula for success. Dartevelle was preparing to give a series of lectures in Toronto and he invited me to participate as he and a language professional, Mr. Thomas Lowry, were reviewing his slides. I learned that Dartevelle had ®rst retained Lowry in 1994 to help prepare his address as the Honored Guest of the STS. Lowry's program for Dartevelle included making arrangements for Dartevelle to spend a week living with a family in England. How many Americans have made an investment of this kind in order to improve their skill in a language other than English?
It is a privilege for native English speakers that the language of science is currently English. With privilege comes responsibility. To meet this responsibility, I suggest that we take steps to enhance editors' and reviewers' awareness of the added burden English imposes upon non-native speakers. I also suggest that each publication consider developing a mentoring service wherein native speaker thoracic surgeons are made available to non-native speakers when they ask for assistance. My experience with such interchange indicates that the intellectual and personal bene-®t from such international activity far outweighs the expenditure of effort. In addition, our organizations, such as STS and EACTS should offer workshops, at least in conjunction with our annual meetings.
In closing, I return to Victoria Fromkin who introduced me to linguistics when President Hasse accepted my proposal to speak about the language of science. Fromkin's book, which is one of the basic texts in linguistics, says,`A basic property of human language is its creative aspect ± a speaker's ability to combine the basic linguistic units to form an in®nite set of`well formed' grammatical sentences, most of which are novel, never before¼heard.' [8] That's it, ladies and gentlemen. That's what we need in international thoracic surgery in order to communicate and to be sure that our patients continue to get the best and that our speciality continues to thrive despite political and economic considerations.
Mr. President, colleagues and friends, I thank you for your attention.