二语习得中的年龄效应和技能习得理论——专访马里兰大学教授Robert DeKeyser

马 拯

(上海海事大学,上海,201306)

Robert DeKeyser博士,美国马里兰大学语言、文学和文化学院教授,在二语习得研究领域久负盛名。本专访尝试以简单易懂的语言,以一名普通的二语学习者的身份与DeKeyser进行交流,由此希望借助最专业的教授的视角,帮助读者更清楚地理解二语习得研究的一些抽象而复杂的概念和理论。DeKeyser本人来自于比利时,作为美国大学的教授,也算是一名二语学习者。他热情地接受了采访,并就二语习得的一些关键问题分享了他的观点和看法。

外语学习看似非常简单,它不像学习数理化那样高门槛。评价外语学习的效果看起来也很容易,轻易就能找到本族人作为参照。这也是为什么随处可见市场化的语言培训广告和机构,却很少能看到类似的数理化培训机构。然而,事实上外语学习又不那么容易。在二语习得领域,有一个广为所知的事实,那就是几乎没有一个成年的外语学习者外语水平已经或者能够达到像本族人一样(native-like),成年的外语学习者甚至被认为是“残疾”(disabled)的语言学习者,这就是二语习得中的年龄效应(Age Effects)。

DeKeyser教授认为,孩童的隐性学习能力(implicit learning capability)随着年龄而逐渐下降,因此与年龄相适应的显性学习(explicit learning)和系统性的练习非常重要。年龄效应研究的启示不在于说我们越早教小孩学习语言则越好,而在于它强调教学活动应该与年龄相适应。对小孩来说,不应该教给他们过多的显性信息,使他们负担过重,而应该给他们提供高质量的语言输入;相反,对于青少年和成年学习者,则应该给他们提供显性的信息,再加上大量交际性、系统性的练习,从而促进陈述性知识(declarative knowledge)的程序化和自动化。他认为,任何形式的语言教学和学习的成功,不管是在外语课堂里、在国外学习的环境里,还是在移民语境下的学习,都取决于与学习语境和学习者年龄及学能相适应的显性信息、语言输入和练习的正确组合。

因此,DeKeyser教授认为,“native-like”可以作为语言学习的理想(ideal),但以此作为语言学习的目标却有些不切实际,这反而容易让学习者感到沮丧、失望。同样,外语教师与其过多强调语音的完美,不如更多地关注语言表达的“准确与流利”。因为既然语音是语言习得最难的方面,不如更切实际地关注语法、词汇等层面的提高,从而达到流利、准确。DeKeyser以自己为例:虽然他口语表达“流利、也相当(reasonably)准确”,写作上达到跟本族人很难区分的程度,但他仍具有很强的、很容易就能跟本族人区分开来的口音,可他仍是一位成功的语言学习者。在美国课堂,经常可以看到很多教授,操浓重口音的英语,教授美国大学生语言、数学、物理等课程,他们可能来自印度、日本等国,但从没有一个学生抱怨他们的语音。DeKeyser教授认为,就是本族人自己也有各自的口音,作为语言教师,“以自己都无法达到的标准去让学生坚持或者去评价他们,其实是愚蠢的”。美国大学尚且如是,英语作为外语的中国大学就更应如此了。从这一点来看,我们在试图开展以内容为依托的教学时,应该弱化教师的语音考察,而注重内容表达的流利和准确。因年龄效应研究的开展目前存在许多技术上的难题,直到今天,学界对年龄效应产生的原因仍然众说纷纭。

二语习得领域另一个著名的理论就是技能习得理论(Skill Acquisition Theory),这也是DeKeyser教授研究的一个重点。他认为,任何学习,不管是学习骑车、学习基本的计算、学习开车,还是学习游泳等等,从一开始,我们都会对正在学的这些事情想得很多,而且会有专门的人,如教练或者教师等,在一旁指导。再后来,通过训练、实践,我们不仅能更快地做这些事情,而且犯的错误也更少,也更不费劲。最终做这些事情时,也不会再想得很多。正如一个熟练的司机,他能够一边开车一边与人对话,但仍然不会闯红灯,或者跟别的车相撞等。“自动化的力量是神奇的”,DeKeyser教授认为“相同的事情,也一样可以应用于语言学习”。技能习得理论一开始强调显性知识(explicit knowledge)的学习,把显性知识视作为通向隐性知识(implicit knowledge)的工具手段。同时,DeKeyser认为,在强调“熟能生巧”时,我们也应该注意练习(practice)的层次,需要对练习进行精致化(fine-tuned),具体说来就是要让练习的层次与学习者个体层次相一致,而非仅仅强调练习的量,这样才能达到练习的最佳效果。

DeKeyser教授也对语言教学的“技术主义”、大规模的“电脑化”现象发表了自己的看法。他认为,语言最根本的是用来交流意义(meaning),而这正是电脑很难实现的地方。放开语言学习的其他方面不说,单是用语言表征意义就是一件很难做到的事情。把句子或者文本放在上下文语境来理解,这是电脑很难做到的事情,或者至少无法做到像人那样。从语言学习来看,电脑在基础层面上有它的作用,比如用于基础的操练、用于展示、重复等等,但是根本上,语言是一项很复杂的技能,是为人与人之间进行交互的,也只有通过与人的交互,才能最终掌握这项技能。“如果说有什么电脑不擅长的,语言教学就是其中之一”。从这个角度看,语言学习和教学的“大规模的技术主义和消费主义”倾向(杨枫、吴诗玉2013),以及由此导致的课堂教学的“娱乐化”倾向(秦秀白2013)值得深思。

用Ortega(2013)的话来说,今年是二语习得研究四十周年(尽管很多人不赞同这个观点)。本访谈也请DeKeyser就这过去的四十年进行了总结。就此,他首先回顾了前辈学人如Dulay,Burt,Krashen和Schumann等对二语习得这门学科创立所做的开创性工作,然后主要谈到了在二语习得领域已经得到较为满意解决的一些进展和一些仍然需要学界进行更多研究才能获得答案的问题。这些观点将帮助人们更深入地认识二语习得这一重要研究领域。

DeKeyser教授曾担任二语研究的著名期刊LanguageLearning的编辑,并即将成为Bilingualism的副主编。因此,本访谈也特别提及了中国二语研究者在国际期刊发文困难的问题。他认为,这并不是由于歧视造成的,实际上中国的学者已经和正在国际期刊上发表越来越多的研究成果,而且中国学者群体大,从而必然获得国际关注。但是DeKeyser教授又指出,要做好研究,我们仍然需要获得专业的训练,这包括基本研究方法的训练和写作技巧等能力的训练,如:如何引用、如何组段成文,等等,“这是一门专门的艺术”。从他的审稿经验看,有些研究者对研究论文的写作过于生疏,缺乏经验和训练。但他相信,假以时日,将会有更多中国学者的研究获得国际二语研究界的认可。

DeKeyser还对《当代外语研究》的发展提出建议。他认为,由于期刊界所形成的一种惯性和潜在规则,好期刊总是能获得更多的资源,而小期刊却又总是被忽视。要打破这一怪圈,一方面需要时间的积累,另一方面,可以尝试在学术界先建立一种联系,邀请国际作者先从短文供稿开始,从而最终形成更多的融合。

DeKeyer教授治学勤奋、严谨。尽管他在学界早已成名,但仍常常只行走于家庭与学校办公室的两点一线。一方面为研究仍然不知疲倦地付出,另一方面正如他自己说的,“学生优先”,这一理念也给了他不竭的动力。在马里兰大学陪先生访学的日子,偶尔路过时,或在白天见他在办公室与造访学子侃侃而谈,或已天黑,他亮灯的办公室让人感受到的是学府里特有的宁静。此时也不禁有些感慨起来。

Ma: Robert, in some sense, you are also an ESL learner?

DeKeyser: Yes, of course, yes. I started learning when I was in middle school in Belgium.

Ma: Do you think you are successful as an ESL learner?

DeKeyser: Hm, well yes, by the standard of most learners. You could say well I am fluent and reasonably accurate, and from my writing, most people couldn’t tell it is not by a native speaker. But of course there are people who are more successful. I do have an accent that is easy to recognize; some people have a better accent. The aptitude for having the perfect accent is a bit different from the aptitude for learning vocabulary and grammar. So yes, successful, above average, but certainly not perfect.

Ma: Yes, you mentioned “accent” this sort of things, can I use a term SLA researchers often use—“native-like”—to describe your proficiency?

DeKeyser: You know we make a distinction between near-native and native-like. In that sense, no, I am not native-like, because native-like in a strict sense means that one cannot be distinguished from a native. That is not the case; a native speaker can tell that I am not a native speaker.

Ma: Then do you think “native-like” should be the goal of a language learner?

DeKeyser: Hm, I think it’s good to say that is an ideal, but given that basically no language learner ever gets there, I am not sure that we should tell our students, our colleagues, that that’s where they should get, because it is very, very difficult to do; and one of the reasons, I think, why people sometimes get so disappointed in language learning is that they have an unrealistic goal. They think within just a few years they will sound like a native speaker,and of course when they find out that is not the case at all, they tend to be disappointed.

Ma: Yeah, ok, I really hope that we, teachers and educators in China, could hear what you are saying, ‘cus I think we sometimes just put too much emphasis on factors like accent; people get a lot of practice in this, trying their best, spending a lot of time, imitating the accent of a native speakers, yes too much time. And some teachers and educators judge their students by these factors, and of course, they themselves were also judged in this way when they began to learn a language. But here, the situation is quite different: I noticed that some professors from other countries speak with accents, yet people never mind this.

DeKeyser: Well, yes. That’s an interesting question you are getting at, because underlying what you’re asking, I think, is the question of how important accent is compared to grammar, compared to vocabulary, and that is a hard question because how you answer the question always depends. Obviously, if you have a very very bad accent, that is a problem; if you have very bad vocabulary, that is a problem; if you have very bad grammar, that is a problem. But if all of them are reasonable, and the question is which is most important to improve, then it still depends. For instance, on whom you are going to talk with. Like here in this academic environment, everybody is used to accents. None of undergraduate students have ever complained about my accent. They always have other teachers who have stronger accents, teaching them subjects like chemistry and physics, who are from Indian or Japan, wherever. So that is not an issue. But if you are dropped in a really rural area, there you would stand out more because of a foreign accent. People may care less about your grammar there, whereas in academia, having an accent is normal, but if your grammar is bad, that would sound bad because that is often associated with a lack of education. So it really depends on whom you are talking to. As a tourist for instance, there I would say accent is more important because if you have a strong accent people might have a hard time understanding you. If you make grammar mistakes, it is unlikely that it will interfere with comprehension. So it really depends on whom you are talking to for how long.

More generally speaking, what you can say about accent, and grammar, and vocabulary, it is that often we are obsessed too much with correctness. It is not just pronunciation, but maybe even more in grammar, because it takes a very long time before people can be perfectly fluent and perfectly accurate. So that is not really a realistic goal. For most people, being able to communicate fairly fluently, at a more or less normal speed, is a very important goal; otherwise they feel embarrassed that they speak slowly, and the native speakers get impatient, so it is important to have a certain degree of fluency, even if that means you might make grammatical mistakes.

Ma: Ok, so we could put more attention to fluency as opposed to so much on accent...

DeKeyser: Yeah, it is hard to generalize, but in many cases I think that is the problem when people come out of class with very little fluency and also with little confidence in speaking.

Plus talking about pronunciation again, if it is one thing that is almost impossible to teach well for a non-native speaker, it is, of course, pronunciation. You know non-native speakers typically have accents themselves. So they cannot be the perfect model. So how much sense does it make to hammer on correct accent, if you cannot do it perfectly yourself. It is very difficult to have perfect accuracy, pronunciation or grammar. It is silly for teachers to hold students to norms that they themselves cannot meet.

Ma: Ok. Maybe also related, let’s move on to another thing, let’s talk about age effects, for which you are well known. In the beginning, you seem very radical, like, you once said that an adult language learner is a disabled language learner?

DeKeyser: I didn’t use that term, certainly not in writing, but I know what you are saying. To some extent, it is true in the sense that I don’t believe that an adult can learn a language completely as a child does because, as you know, a child does not sit there, think about grammar; a child learns things implicitly, without consciousness of what he/she is learning and still eventually ends up like a native speaker of his or her age. So in that sense, I really do believe that an adult cannot do that anymore. And I don’t think I have changed my mind on that point...

Ma: (Laugh)...

DeKeyser: Not at all.

Ma: I have read your recent article published inLanguageLearningin March...

DeKeyser: Ok, I see why you are asking that, because that paper is a very different paper from the other ones. It is not an empirical study, but just recommendations for future research. That was our goal. We had a little conference in Michigan a couple of years ago. We were all asked to talk about one of our research specialties, and to say where we thought the field is going. So I tried not to say anything in that article about what I think about age effects, only to say things about methodology for research on age effects. That’s why I am deliberately taking a very neutral stand, so that the article is equally interesting, I would hope, for people who believe in age effects, and for those who don’t believe in age effects, for people who think they are maturational and for people who don’t think they are maturational. So that’s why that article is very neutral. It is not because I have changed my mind.

Ma: Ok, so it is still a “very controversial yet poorly understood topic”...

DeKeyser: Yes, it is very controversial and I would say poorly understood, yes, because if it were well understood, it wouldn’t be so controversial. The only thing I think we all agree on is that typically the adult learner does not get as far as the much younger learner. But then the big controversy is why that is. Is it because all sort of things that we have found change with age: that people become less interested in sounding like a native speaker, that they have less good input, all these kinds of things? Or is all this because of the cognitive maturation, the psychological development as such that it is not possible to learn exactly the same way anymore? Well, I don’t happen to think that it is just a matter of confound... but it is not easy to show that; that is why I try in that article to tell people “Look, whether you believe in age effects or not, much of the research is a little bit naïve”.

I don’t think that there is any study so far that is really perfect. And one point I am hammering on in that article you are quoting, is that so often we take convenience samples. That means in this case people who are living in a community that speaks predominantly their native language, and who speak the local language as a second language. That means you can fairly easily access a large number of people which you need for statistical purposes, but of course, these people are not the ideal second language learners in the sense that they spend quite a bit of their time using their native language, and that might have an effect. So, ideally, we would do research with people who are isolated, like parachuted somewhere in a middle of very, very small town with nobody else speaking their language, and at the same time, we should have many many many people so that we could do appropriate statistical analyses. The combination of these two is, of course, almost impossible to reach.

Ma: Yeah, still concerning the age effect. You know, in China, very interestingly, many parents, especially those in big cities like Shanghai, send their kids to learn English at rather young age, at kindergarten, at primary school or through private tutors. Some parents come to me, saying that “their kids should not lose at the starting line” and asking me for advice, “What should I do to help my kid learn English; should I have a private tutor for my kid?” People do not seem to mind the huge amounts of money and time put into it. What’s your advice for them?

DeKeyser: All right, it is important for people to understand that age is only one of the factors, though it is an important factor. But if everything else is the same as before, for instance that people learn a foreign language four hours a week in the classroom, if you simply start doing that early, the effect is not going to be very big. You can see in some European countries or other countries, many countries where people have for many years now started at young age, and really what they have achieved after several years of school teaching is not that much. Of course starting early is going to be better than starting late, because you would have many more hours in total at the end. But at the beginning these hours are not used very efficiently, because there is a certain incompatibility between the way a child learns at that age and the way you have to teach in the classroom. So I would say, the important question is not so much at what age people start, but more how many hours a week people have and what is done during these hours. So if a young child gets 20 hours a week, talking to a native speaker, about topics they are interested in, that should have a big effect. But if that same child at the same age, spends 4 hours a week in a classroom going through the same old drills as years ago, then that would hardly have any effect. So age, in combination with other factors, is very important.

Ma: Another big worry is that: Will it have detrimental effects on the learning of their native tongue?

DeKeyser: No, there is no evidence of that. If you look, for instance, at the Canadian immersion programs, which started decades ago, they have always been very well documented, you would see that immersion has never led to a deficit in their native language, that it never has led to an academic deficit. The only difference is that they do much better in their second language. But of course I am talking about immersion programs from a very young age, where teaching is largely in a second language, many hours a week, teachers are native speakers, are fully bilingual, have been trained to be bilingual teachers and so on. That is the ideal way of doing things probably, and so they received no detrimental impact, not on the native language learning, not on academic learning, but still we see that the learning of a second language is not perfect. This is much better than traditional foreign language learning, but they still are not like native speakers when they are 18 years old.

DeKeyser: Yeah, some of the literature in that domain is very technical, but the basic idea I think is not really difficult to grasp. In any kind of learning we do, whether it is learning how to ride a bicycle, or learning how to carry out computations, or learning how to swim or learning how to drive a car, at the beginning we think very much about what we do, and often we learn to do these things, through verbal information, that is we have a coach that tells us how to throw the ball, we have a teacher to tell us how to work with a computer or we have teachers tell us what computations to do in what order, and so on. Then through practice, we not only can do these things much faster, we can also do them with fewer errors, and with less effort, eventually without thinking about them. An experienced driver can drive a car while having a conversation about any topic but still not go through any red light, not bump into any other car, and so on and so on.

So the power of automatization is really amazing. When you see how easily and well you can do something later on that, at the beginning, you could not do very well at all with great efforts and with a lot of mistakes. So that is something we know from many areas of learning. I don’t see any reason for saying “No, it does not apply to language learning”. It doesn’t apply to language learning maybe the way a little child learns, Ok, fine. But when dealing with an adult who is learning a language at school, that is a very different situation. There, I think, skill acquisition theory has something to say.

Ma: It seems that we should not always believe that “practice makes perfect?”

DeKeyser: Hm, well, it depends on what kind of practice, because often people think that practice simply means lots, lots, lots of practice. But practice has to be appropriate for the level you are at. If you are a fairly advanced person and you do mechanical drills, that is a waste of time; if you are still a beginning learner, and you are told, “Go, practice, talk to that native speaker over there”, that practice is not very useful either. So you have to have practice and start from where you are and then let you become a more fluent user of the knowledge you have, and at the same time maybe acquire a little bit more knowledge. So practice has to be fine-tuned and that is not easy to do, because even in a given classroom, different students might be at different levels. So more generally speaking, when you are talking about any kind of learners, of course the kind of practice they need is very different from one to another.

Ma: I see. So how is skill acquisition theory related to implicit and explicit learning?

DeKeyser: Skill acquisition theory very much insists on explicit knowledge of the beginning, and sees explicit knowledge as a tool to eventually get to something like more implicit knowledge.

Ma: You say “like”...

该方案的优点：①系统独立，降低外界干扰；②缩短自动燃烧控制系统的设计和制造工期；③自动燃烧控制系统组态编程（复杂的燃烧计算及控制逻辑） 和调试（相关程序根据现场情况需不断进行修正）由焚烧炉独立完成，成套集成度高，可有效缩短现场调整作业时间。ACC就地控制柜和DCS可以实现数据通信，常规设备运行状态监控也可以由DCS系统完成；④作业培训后，上位机操作及现场控制柜维护比较简单（ACC柜不需经常维护，核心PLC免维护）。

DeKeyser: “I say like” is because some psychologists make a distinction between highly automatized explicit knowledge and the real implicit knowledge. The difference between the two in practice can be impossible to tell, or it’s only through a very careful set of experimental techniques that maybe you can tell the difference. So from the psycholinguistic point of view, these two are very important; from the practical point of view, the only thing important is that you can use that explicit knowledge at the beginning, engage in a kind of practice that then allows you eventually become fluent and accurate enough so that you may sound like you were an implicit learner. That is what really matters: Are you functionally equivalent to an implicit learner? But that does not mean you learn it implicitly.

Ma: All right. Probably that is also the significance of classroom teaching; the learner develops kinds of automaticity.

DeKeyser: Yes. Because in the classroom we typically don’t spend a sufficient amount of time on this. The reason for that I think is classrooms are, of course, part of an institution, even a bureaucracy, and it is very important for teachers and for students to be able to show to their parents and to their principal, to the ministry and so on that “look, we have taught this much and they have learned this much.” And if you teach a number of rules, the teachers have done their job and if the students have paid attention you can show their learning. Whereas, if you deal with fluency, with skill, it can take a long period of time before you have measurable differences, and people in school contexts won’t be able to show that every single week the students perform noticeably better than before. And that’s why the teachers are strongly inclined to constantly teaching new things instead of making sure what was taught can be integrated and can be used in their work.

Ma: I see. Ok, now let’s talk about second language acquisition in general. In the words of Ortega, “SLA has completed 40 years of existence and moves forward into the 21stcentury”. I think if Ortega is true, these should be uneventful 40 years for SLA.”

DeKeyser: Hm, are you talking about her article that is in the sameLanguageLearningthat shares with my article?

Ma: Right.

DeKeyser: So some things in it I agree with, other things I don’t agree with.

She pointed out many things that I disagree with. I would say one of the things she insists on there is that we should realize that second language learners are bilingual to some extent and you cannot expect the same things from bilingual as you can expect from monolinguals in many areas of language. I agree with all that in principle, but I think that is really overblown here, and comparing learners to what a native speaker can do is truly a useful thing. That does not mean we think that bilinguals are two monolinguals in the same mind. No, it just means that if you are going to have a standard for what people can do, well, that standard, in my opinion, it’s ok if it comes from a native speaker, as long as you don’t exaggerate that standard, for instance, as long as you don’t demand that a non- native has a perfect accent and so on. I also disagree with what she says about research methodology; I think she emphasized the sociocultural a little bit too much, compared to the cognitive paradigm anyway.

Ma: Ok. If we say 40 years, and if we could just look back on what problems, you think, in SLA, have been resolved in a satisfactory way, then what important problems are still awaiting there, to be resolved?

DeKeyser: 40 years, that would mean first started in the early 1970s, as now we are in 2013. And indeed, in the early 1970s SLA as a discipline was born, so to speak, because people like Dulay, Burt, Krashen, Schumann, Hakuta,and many others, started documenting second language acquisition outside of the classroom, and they showed that things are learned in a certain order by people at different ages, in different native languages, and that often what they learned was in an order that was very different from what people were teaching in the classroom at the time. So that is what made the field, because people started realizing that “Aha, there are certain things that we can generalize about SLA”, and they are not quite the same as what we can say about first language acquisition. So that’s how the field was born, since then we have learned certain things. We have learned to understand better what aspects of language learning are harder than others, and why, not a hundred percent, but we understand that better. We have learned how important it is to make a distinction between implicit and explicit knowledge and between the automatized, procedural and declarative. We have learned that teaching has to be as communicative as possible. We have also learned that it is important to pay as much attention to form as possible because just being 100% communicative is not ideal either. So we have learned we have to focus on word form and meaning. And some people stress the meaning aspect, some people would stress the form aspect, and I think that kind of distinction is a bit silly because anything in language is about conventional linking of form and meaning. So you cannot leave either of these out of the picture. So I think we understand all these things better now. Sometimes one gets the impression from outside, that we go back and forth that grammar is important, then grammar is not important and so on. I don’t think that is really the case. In fact, we understand better how grammar is learned, how it could be learned, how important AND not important it is. Another problem is what are the implications of what we know, because even we understand language acquisition process completely that still does not mean that problems are all resolved, like even we understand a certain kind of disease, that does not mean we have a cure, ok. Even if we have a cure, it does not mean we can always use it efficiently, with everybody everywhere. So it is a bit like that too: that language learning is a very involved process, it is not something that we could do in a couple of weeks, it is something that some people are better at than others. If we have unrealistic expectations, we are just going to make ourselves unhappy. So I think that is one thing SLA has contributed to: having more realistic expectations showing that yes, you can learn to communicate fairly efficiently, as an adult, but you are likely to always be different from a native speaker, especially in the area of accent. Your accent is certainly going to be different from a native speaker.

Ma: Yes, SLA in some sense is abstract, yet you have made it easy to follow. Robert, do you mind that we talk about things in China?

DeKeyser: Go ahead.

Ma: In China, these days, English teaching in many universities is largely computerized. With many computerized self-learning centers constructed and computerized English courses provided, educators and, also university authorities, seem quite confident in the efficiency of computers in language education. But you don’t seem so confident?

DeKeyser: No. Because given what I said before, given that language is a skill that is about communicating meaning, that is what computers have trouble with, because even leaving aside for a second language learning aspect, just representing meaning in language is very difficult to do. Getting computers to do machine translation, or to interpret meaning of texts, is difficult because computers are not as good as humans, at taking into account the whole context in understanding sentences in context and so on, so that is one thing. I do believe that computers can be very useful for the basic levels, for doing basic drills, for presentation, for repeating something, for a certain kind of individualization, i.e. in terms of speed. All that is easy to do. Then what is a little bit harder is individualizing in terms of how students learn, having a student model. What a student really knows declaratively and procedurally, what is automatized and how can we give different exercises to this student compared to that one, that sort of more sophisticated use, what people might call computerized tutoring and teaching. But ultimately language is really a very complex skill for human interaction that you can only fully learn from interaction with other humans. So I have nothing against the use of computers but they should not be seen as a replacing human interaction. That wouldn’t work. If there is anything that a computer is bad at teaching, it is probably a language.

Ma: Yeah. I think this could serve as good practical guidelines for us to follow. Here I want to mention college English teaching (CET) in China. You know it is really a huge thing, large in scale, with many learners and also teachers. Yet, CET is constantly under public criticism, mainly for its low efficiency. Much time and money is spent on it, yet our students are not proficient, especially in speaking and listening. And now changes are taking place there. Some people hold that traditional CET, GAP in nature, should be replaced by ESP. What college students need, they believe, is English for specific purposes, not the present system, English for general purposes. If this is correct, it would have far-reaching influences on English education in China.

DeKeyser: A couple of things about that. So first of all, I agree that the outcomes of many years of language teaching in high school or college are very disappointing. But I think it is easy for the average person to see how imperfect it is because they can say “Look, this is a native speaker, this is how this person communicates. That one obviously cannot do that.” When it comes to a subject matter like physics, it is much harder for people to tell where they would like to be and where the students get. So that is one thing. Another thing is that definitely things can be improved, given what I said about instilling more skill, not insist so much about ever more words, ever more grammatical detail, but make sure instead that people can fluently use what they have learned for speaking and listening and so on. And then, of course, finally I would also agree that different people are using the language for different purposes, that it is good to adapt teaching to that. If somebody is going to be mainly a reader of English texts, that is one thing; if somebody is going to attend business meetings, that is something else. Yet everybody who is learning English will, I assume, at some point, learn it or use it for basic interaction with other English speakers. So it is not that all teaching is for special purposes. There is a common core of words and structures that all teachers cannot avoid. But it is certainly true, I think, that at some point, you need to start to think about what people are using the language for, or likely to use it for, and there you can individualize more. That is probably easier to do in China than in other countries with such a large number of people learning English.

Ma: Another topic concerns second language researchers in China. You once served as editor ofLanguageLearning, and soon will be associate editor ofBilingualism. Why, do you think, is it so hard for Chinese researchers to get published in international journals? What could they do about it?

DeKeyser: One thing is that they are beginning to publish more and more. If you look at how many Chinese names you see in journals likeLanguageLearning, orBilingualism, just to mention a few, you see that there are quite a few these days, where there were almost none twenty years ago. These are not just people living abroad: there are people from Mainland China or Taiwan who publish in these journals. But what is certainly true is that, first of all, the standards in these journals go up all the time, because the field has grown so much, grown more than the number of journals and the number of pages. So it is more competitive than it used to be. And typically the journals will have a rejection rate of between 75 to 90 %, so it is only the top 20% that get in. So that means every aspect has to be pretty good, from the research question, to the research design, to the analysis of the data, to the writing. It all has to be very good. And I think sometimes when people from China have been studying for a PhD in this country just for a few years, they have learned some of these aspects very well, but there is something else sometimes they still have to learn. Don’t forget that typically in graduate school you don’t have a class that teaches you how to get published. Everybody has to know how to publish, but few people have a course about that, and few people work closely with an adviser who says “Well, in order to get published do this and do that”, so I think there is something lacking in PhD education there. Looking at my own case, I went to graduate school at Stanford, which is certainly a very good university, and still I failed the first time I tried to publish something in a good journal, because I wasn’t fully prepared for that even though I thought I had a good idea of what that was like. So I think people need a bit more training specifically on that point and I think any kind of research methods course, or the end of the course at least, should try to deal with that: What is the difference between a dissertation and a journal article, not just shorter, you have to focus differently, and there is specific writing style affecting everything from organizing the text into paragraphs, to how to quote people. All that is really a specialized art, and all this reminds me that when I grade undergraduate papers from students who are native speakers of English, I see how clumsy they can be at quoting something. It is not because these students are Chinese, or Arab, no. It is because they haven’t learned this very particular style of writing.

So I don’t think there is a bias in the Western publishing world, against people from China or Asia. I think all publishers have realized that China is an enormously important market; that is certainly the case for applied linguistics, certainly the case for a journal likeLanguageLearning, because they have detailed statistics about how many copies are sold in China, who has a subscription, and what articles they read and so on. So I think it is a matter of time, before more and more people who maybe got their PhD in China and did not study in North America or Europe, before more and more of these people start publishing in these journals. It is a bit like, if you look at what was the case 50 years ago and see that in American academia there were hardly any women professors, and now in many cases there are more women than men. But you can’t change it overnight. It takes a couple of decades for changes like that.

But having been an editor, I must say that very often you just glance at a paper, and you see the authors are extremely inexperienced and really do not know what it takes to write an article, both in terms of content and in terms of writing. So I think in PhD courses, especially for foreign students, we should pay more attention to that, make sure they are on top of everything they learned, the theories, the statistics and so on, that they really learn what is specific to writing articles for journals in the field.

Ma: Ok, I see. You know, compared to the journals where you served as an editor, this journal,ContemporaryForeignLanguageStudies(CFLS), is trivial. Its dream, yet, is big. Besides publishing papers in Chinese, it plans to have international issue. Difficulties, however, are always there. International contributions are hard to get...

DeKeyser: That’s probably because of two things. First of all, academics tend to be overworked, they are overscheduled, or much behind schedule, as in how am I going to write this for last month, and so on. So they are not exactly looking for more work. And secondly, what is happening more and more, is that there is an elaborate system for measuring how much people have published, where and what the impact factor is for that journal. In other words, how important, how prestigious is this. And of course that sort of thing leads to a vicious circle. More people want to publish in the leading journals, because they are the leading journals, and that makes them the leading journals even more clearly. So it is difficult to break out of that circle, but I think it is a good idea to invite people to at least provide short contributions to your journal, to create these links, and here again, over time there can be more integration. It won’t happen overnight.

Ma: I am very much impressed by your devotion to your research. I’ve learned that, besides a very good teacher, you are very much devoted to your research. When I just arrived here, people told me that professor Robert is the most “accessible” professor, because people could always find you in your office, working hard at your research. As a matter of fact, you are already established in the field of SLA research, what, then, drives you...

DeKeyser: Yeah, I come to my office basically every day, in part because I am a habit person, in part because I live nearby. So it’s not a big effort to come in. But I come to the office a little less often now, because I have a dog now, and I don’t want it to be alone at home all day long. But still I come to the office a lot, and I always think students are our first priority. It’s hard to juggle these things sometimes, whether you should work on a draft of a manuscript that you promised to a colleague or on a draft of a dissertation that you promised to read for a student, but I try to give the students priority I think because most people would

still see our first duty as professors as teaching and mentoring...and also it’s a little bit easier to do once you have tenure or are a full professor, and you don’t have to worry as much as before about exactly how much you get published. There is still a lot of pressure to publish, but you are not going to lose your job because of one article more or less.

Ma: Very impressive, because you are already well established in SLA and still you work so hard.

DeKeyser: Oh well, that is, I think, fairly typical of American academia, not for everybody everywhere, but still most people tend to work pretty hard once they have tenure. It becomes a habit (laugh). Also, there are still pressures on us to bring in research money and to sponsor students and so on, so there are many reasons why we need to keep working hard even if we have a permanent job.

Ma: The last question. What are you doing currently?

DeKeyser: Currently I am trying to get some new grants to do research on age effects, but I am also, as always, writing various chapters that people want me to write for various books. That’s part of the problem. Of course that also takes time and it’s not really empirical research, just reading and writing, so a good chunk of my summer always goes into that. Right now I am going to start writing a chapter for a new edition of this book on theories of SLA, and they require a section on implicit and explicit learning, so as that is one of my areas of interest I see that as a sign of progress, and I am eager to write it.

Ma: Many thanks, Robert.

Ortega, L. 2013. SLA for the 21st century: Disciplinary progress, transdisciplinary relevance, and the bi/multilingual turn [J].LanguageLearning63 (Suppl. 1): 1-24.

秦秀白.2012.警惕课堂教学娱乐化[J].当代外语研究(7):1-2.

杨枫、吴诗玉.2013.大学英语教学通识化转向的“逻各斯”[J].外语电化教学(1):9-14.