Graduate schools and a job

From my positive experiences as an EFL teacher in Morocco, I decided to get an M.A. in teaching English as a second language (TESL) and went to Ohio University. As do many students in U. S.-based M.A. programs in TESL, I took quite a few linguistics courses and got turned on by the combination of syntax and sociolinguistics. So, having never been west of the Mississippi, I decided that I had to pursue a linguistics Ph.D. at the University of Southern California (USC).

My experiences at USC strongly shaped my future path as an academic. Aside from being in the midst of a vibrant, exploring, and challenging academic environment, I was strongly influenced by three professors and by fellow students. For three years, I took every course offered by Bernard Comrie and became very interested in syntactic typology as a research field. The most important lesson I held onto from those courses, even now after all these years, is the power of descriptive linguistics. There was always much more to learn simply by understanding form and variation in the actual linguistic data, without over-committing to any specific linguistic theory. My courses with Comrie probably led me away from standard Chomskean linguistics fairly early on, even if that wasn't ever his overt intention.

From my first semester in 1978, I met and got to know both Bob Kaplan and Dave Eskey. In different ways, they both became my mentors. Bob brought me to see the real value of written discourse analysis, even if it was not a well-established area in 1979, and, of course, he introduced me to ways of looking at written discourse variation across cultures. I also learned the value of allowing an idea to evolve, and for Bob, the notion of Contrastive Rhetoric was continually evolving over time, rather than being the static concept that critics often complained about. Throughout this time, I also taught writing fairly regularly at the American Language Institute (at USC), leading to a continuing interest in second language (L2) writing development, primarily in English for academic purposes (EAP) environments. I also met Dave Eskey in 1978. At the time he was Director of the American Language Institute. From the beginning, and over six years while I was at USC, he encouraged my interests in reading research and instruction. He introduced me to the early research of Keith Stanovich (Reference Stanovich1980) and Charles Perfetti (Lesgold & Perfetti Reference Lesgold and Perfetti1981), radically altering my view of how reading comprehension worked.

By 1982, I had gone to China for a year as an early Senior Fulbright Lecturer in Shanghai (1980–1981) (an early wanderlust continued), and after returning, had decided that I was much more interested in L2 reading and writing development than in typological linguistics. A key turning point was going to the 1982 Teaching English to Speakers of Other Languages (TESOL) Conference (it was not a ‘Convention’ yet) and having a great time in Honolulu. This very stimulating experience was quite different from three Linguistic Society of America (LSA) (and early American Association for Applied Linguistics) meetings I had been to between 1979 and 1982. By direct comparison with TESOL at the time, I found the LSA conferences stultifying. Audience members seemed to need to attack speakers to prove their self-worth. Frankly speaking, the ‘rough and tumble’ bit of blood-sport that many engaged in seemed immature. I'm sure many might disagree with my perspective, but these were the accumulating experiences shaping my thinking.

In contrast, the 1982 TESOL Conference in Hawaii, at that time, included many applied linguistics researchers, and many leading second language acquisition (SLA) people were participating actively. There were friendly discussions, and audiences were supportive even while pointing out weaknesses and limitations in research talks. As a graduate student at the time, I found the leaders of the field very approachable, and I was impressed by the easy interplay among theory, research, and application (ah, impressionable youth). So by the end of 1982, I went to Bob Kaplan and told him I was interested in changing topics and switched my focus to written discourse analysis and L2 writing development. (And in how many departments today can a Ph.D. student switch among a formal linguistics focus, a sociolinguistics focus, and an applied linguistics focus, even in the fourth year, and stay in the same department? Times were stimulating at USC in the late 1970s and early 1980s.)

My last two years at USC were both more focused and also more broadening. My dissertation, on linguistic variation in expository texts, combined analyses of text structure with the new corpus linguistic opportunities being generated by Doug Biber, one of my more interesting and accomplished peer graduate students at that time. Doug ran his programs on my corpus and helped me identify key patterns of variation across text types. I may have been Doug's first corpus linguistics ‘student’ (see Grabe Reference Grabe, Connor and Kaplan1987). At the same time, from 1982 to 1984, I explored and expanded two related lines of scholarly interest: (1) the development of L2 reading comprehension skills (Grabe Reference Grabe, Dubin, Eskey and Grabe1986, Reference Grabe, Carrell, Devine and Eskey1988, Reference Grabe1991) and (2) the ‘application’ of my literacy interests through content-based EAP instruction. I quickly developed an applied interest in the idea that coherent content learning went with comprehensible input to improve academic literacy skills, and build student engagement and motivation for learning (Stoller & Grabe Reference Stoller, Grabe, Snow and Brinton1997).

One final thread from USC has grown increasingly larger in my thinking over the three decades since leaving USC: that is the importance of practice, the type of practice engaged in, and particularly the importance of extensive reading (especially thanks to Dave Eskey). I can't stop myself from ending most of my talks and workshops on reading with the simple statement ‘We only learn to read by reading, and by reading a lot.’ I believe that now more than ever. There are no miracles or short-cuts in learning to read (even if the process of reading comprehension itself is somewhat miraculous).

From all of these experiences until heading off to Northern Arizona University (to actually work at a full-time job for the first time at the age of 34), I learned two very important lessons from my student experiences. First, we don't always get to choose the opportunities we want, but when we do have opportunities, we can't be afraid to walk through the door, so to speak. Second, we can worry about failing, but we shouldn't be afraid of failing. What was the old Nike slogan? – ‘Just do it’. In 1984 I headed to the mountains of Northern Arizona, and Northern Arizona University (NAU) in ‘Big Sky’ country. As it turned out, we (as an applied linguistics group) built a core faculty of great scholars and dedicated teachers, started a Ph.D. program in applied linguistics in 1989, and mentored many exceptional graduates who are productive contributors in applied linguistics and TESOL.

Research touchstones and understanding constructs

After arriving at NAU, I continued to pursue my interests in L2 reading, L2 writing, and content-based instruction pretty consistently, along with occasional forays into vocabulary (the lifeblood of reading and writing development), as well as regular explorations over the years of what it means to be an applied linguist. It has been in the explorations of L2 reading that I have been able to expand my own intellectual inquiry in so many different directions. Because educational psychologists and cognitive psychologists have worked so intensely in trying to understand the nature of reading comprehension since the 1970s, there is a major research reservoir with strong quantitative approaches available. This massive resource is not equally available in any of the other three major language skills; it has been a natural foundation for me to draw on and develop my own perspectives on both L1 and L2 reading abilities and reading development.

In the course of working through the large research data resources from L1 reading, and drawing on the work of L2 colleagues over the years, I have learned the importance of cognitive component skills for reading comprehension, the combination of cognition and context in learning to read, the similarities and differences between L1 and L2 reading development, the distinct role of L2 reading within applied linguistics and its distance from SLA, the role of extensive reading and fluency as parts of reading development, the fundamental link between reading and vocabulary development, and ways to reconcile my interests in language skill learning (reading) with language knowledge learning (SLA).

L2 reading

Exploring reading has been a means by which I have actually extended my interests across multiple disciplines beyond linguistics, including cognitive psychology, educational psychology, curricular innovation and teacher education, and language assessment (of and for learning). From psychology (both cognitive and educational) I was able to explore and connect theories of learning, individual differences in learning (from learning difficulties to developing expertise), motivation and engagement, component skills, comprehension as construct, memory, practice (repetition and variation), fluency and automaticity, implicit and explicit learning, and ways of assessing learning outcomes. One of the things that I have come to realize through all my years of engaging with L2 reading is that learning to read fluently, and especially understanding and using academic information, is miraculous, yet most people think of learning to read fluently and critically as an expectation.

So here is the bottom line: the ability to read fluently is, in fact, quite miraculous. Our brains were not designed to be reading brains by evolution. But we have learned, from one generation to the next, to take graphic forms on a page and mentally interpret them into our own language, no matter what language. We do this by recruiting neural pathways to comprehend these graphic symbols automatically and connect that visual input to the lexical storage that was originally intended to be aurally processed. As we become fluent, we do this faster than we speak or listen, and we involve a larger receptive vocabulary than we use for speaking.

I need to explain what happens in two seconds of reading to explain why this is miraculous. First, as humans, we are extraordinary pattern recognizers. The early humans who were not great pattern recognizers early on probably did not recognize the striped patterning of the tigers in the woods, and all of them were eaten, leaving only good pattern recognizers around. And as we become good readers, we recognize words at about 230 milliseconds (about a quarter of a second) per word that we focus on visually (keeping the story simple). We can actually recognize a word in about 100 milliseconds (one tenth of a second), but we do a lot more during the time we ‘stare’ at the word forms. As good readers, we focus on about 4–5 words per second, every second that we are reading text. (Do the math; 4 times 60 seconds is about 240 words per minute.) We actually look at and focus on almost all of the content words of a text and more than 50% of the smaller function words. So we really look at, and process, most of the words on a page when we read (unlike skimming, where we really do sample parts of the text).

Second, we do a lot of other things as fluent readers while we read a text. We draw on grammatical information to help understand and disambiguate text meaning (using word ordering, time marking, pronoun tracking, etc.). We disambiguate meanings of words and phrases when there are possible multiple meanings, we connect sentences together to make a network of more important ideas (building main ideas). We monitor our comprehension of the text and use various reading strategies to maintain comprehension, or enhance our comprehension. We use our knowledge of prior texts and texts types that we have encountered before. We decide if we believe the information, if the information is worthwhile, if we can use the information. We decide if we like the ideas, or if we like the author, or wonder who the author is. We decide if we are interested in the ideas or if we should stop reading.

Third, we do all of these remarkable ‘other things’ more or less at the same time that we are recognizing and accessing the meanings of 4–5 words every second, and this is multiplied within every minute. Many of these ‘other things’ are happening in the same two seconds while we are also focusing on and recognizing eight words. (Think of how long two seconds lasts.) When we contemplate all the things that we do as we read fluently, it is hard to fathom how efficient and fast our brain is as a meaning processor. We should wonder why we don't get a headache after reading for two or three minutes – but here is the miracle. We can do this for hours at a time if we choose to, and we often do this for enjoyment!

We also need to recognize that reading (and writing) is not a cognitive skill that we simply inherit genetically as humans. The fact that many humans around the world are not literate shows that processing written language for comprehension is not a genetically driven process but still a culturally transmitted process. If someone (teacher, parent, sibling, or other) does not help a learner to read, that person will not become a reader. So the real issue is not ‘why some people do not learn how to read,’ but rather, ‘how do so many humans learn to read so fluently’? On top of that, how can so many humans do this fluent reading trick in a second or third language! Now this is getting to be miraculous (see Grabe Reference Grabe2009).

L2 writing

My interests in L2 writing arose at about the same time. A bit was brewing during my years at Ohio University, but a real desire to understand ‘how it works’ grew while I was at USC. Of course, over the years, even up until now, I have only gradually come to recognize that how we understand fluent reading is going to be more definitive than how we produce writing appropriate to the level of task demand, or better. This imbalance may seem counter-intuitive, because we can actually see written output, but the difference lies in the ultimate performance levels, the impact of context and task, and the inability to identify precisely what counts as a good example of writing ability versus an example of less than exemplary writing ability. The fundamental difference here between reading and writing may be in the performance consistency of building implicit, automated skills (reading) versus building explicit visual output through a process of cognitive reflection and revision. Nonetheless, trying to better understand the nature of writing abilities, and the interaction of the cognitive and contextual, is a worthy undertaking. And it might help produce better student writers, but how is still an open question.

Bob Kaplan pretty obviously has been a catalyst over decades in thinking through many of these questions about writing, and about L2 writing development in particular. Much of my thinking about writing development and writing instruction came together fairly intensely between 1992 and 1996 with the writing of Theory and practice of writing (Grabe & Kaplan Reference Grabe and Kaplan1996). As it turned out, saying almost everything that we wanted to say about writing created a book that was too long and perhaps too intense for teacher training and graduate school courses, but it was certainly also an intensely shaping experience for me. Writing is always a process of discovery, and writing over 700 manuscript pages was a lot of discovering.

Over the years, I have been a proponent of some version of formal preferences in writing that is attributable to cultural influences, and often to national experiences (contrastive rhetoric or intercultural rhetoric). There is no reason to apologize for seeing L2 student writing through this perspective. Researchers know, or fairly quickly learn, that no single or small set of factors provides ‘the explanation’ for writing differences, or distinctly observable patterns of variation in writing. However, it is also reasonable to identify interesting patterns when they arise in an appropriately large and controlled set of data at varying levels of task proficiency. After all, finding patterns is typically what researchers do in one way or another. However, finding a plausible interpretation for observable patterns is not always a case of essentialism or reductionism. In the last 15 years, in particular, I do not think that serious researchers come to conclusions where, for example, all Chinese write a certain way because they have had Chinese experiences and live in a Chinese culture. But there is no reason not to explore carefully how prior educational experiences, cultural preferences, and other national factors might generate patterns of variation that are less common or not as pronounced in some other group of learners from a different L1 background or a different country.

Content-based instruction

However, other aspects of L2 writing development have been much more prominent in shaping my own thinking about writing. The two most prominent influences are the interactions of reading and writing abilities as parts of academic-oriented L2 development and the role of engagement in reading and writing to develop academic writing abilities. In both cases, the role of content-based instruction (CBI) and the need to teach for student motivation are critical issues. As a teacher of L2 students in academic settings, I have been interested in both integrated reading-writing instruction and CBI since the early 1980s (going back to USC days). In the 1990s I started exploring an application of CBI, along with my colleague (and spouse), Fredricka Stoller, which we called the Six-Ts Approach (Stoller & Grabe Reference Stoller, Grabe, Snow and Brinton1997). Once again drawing on Steve Krashen for inspiration, if comprehensible input was an obvious winning alternative (to incomprehensible input), we thought that a guiding principle of coherent content input was preferable to non-coherent content input in an L2 curriculum, even though most L2 courses and many textbooks were/are relatively incoherent in terms of having engaging content for students that build to a higher level of mastery. The goal was to put in a curricular framework many of the insights we had developed through teaching and developing CBI curricula. (We have also found out over the past 30 years that building really effective CBI curricula is a difficult and challenge effort for both teachers and programs.)

A key attraction of CBI, and our own effort to build engagement through more coherent content, was that it could, in principle, address a number of academic skills development issues at the same time. A stimulating and engaging content-oriented academic language curriculum could provide motivation to read more extensively, develop effective reading and writing strategies while working with more challenging texts, exchange real information in course activities rather than artificially constructed information-gap activities, develop integrated reading-writing skills, and give students pride in building some level of mastery over a content area. After all, if a student can say, ‘I learned some things about nanotechnology and advanced hybrid materials,’ that is probably more motivating than saying, ‘I learned how to make complex relative clauses and practiced making them.’ Of course, the other goal of providing students with integrated academic reading-writing tasks – addressing academic skills development – follows from this type of coherent academic curriculum as well if the curriculum is well designed.

As a last set of comments about L2 writing development, I can say that I am now much more interested in exploring how, instructionally, to bring reading and writing together for more advanced academic skills development and how to integrate such instruction into standard L2 reading and writing courses without asking teachers to abandon common textbooks (Grabe & Zhang Reference Grabe and Zhang2013; Zhang Reference Zhang2013; Grabe & Zhang Reference Grabe and Zhang2016). This orientation is a worthy challenge for applied linguists more generally, and not just a small group of reading and writing specialists. I've also come to appreciate, in the past decade, well-designed and -executed qualitative research on L2 writing development, especially when the study is longitudinal (e.g. Cumming Reference Cumming2006; Leki Reference Leki2007; Tardy Reference Tardy2009). Just about any well-designed longitudinal study of L2 writing development is critically important. A final observation about writing more generally is that academic writing abilities are definitely developed incrementally over (long periods of) time. I know that I am a better writer now than I was ten years ago, and certainly much better than I was 30 years ago.

Language assessment

A third major shaper of my thinking has been an expanded interest in L2 language assessment. My developing work in the areas of both L2 reading and L2 writing led to a long connection with the TOEFL program at Educational Testing Service (ETS). I now firmly believe that one cannot be an applied linguist without a strong grounding, at least conceptually, in language assessment and evaluation: the what, how, and why of measuring student (or participant) performances (and the responsibilities that come along with language ability measurement – both for language performance and for research purposes). Perhaps one of the bigger drawbacks in SLA research, up until recently, has been the relative disregard given to the role of rigorous assessment practices and methodology in designing SLA research studies. Certainly in the past 15 years, I have come to appreciate that how data are collected is at least as important as how the data are interpreted. (Perhaps I am a latecomer to this appreciation.)

My first real experiences in thinking about language assessment began in 1988, when I was asked to serve on ETS's TOEFL Committee of Examiners. I have more or less had a connection to TOEFL/EFL ever since. The connection with TOEFL/ETS has been an extraordinary educational experience for me all along these past 27 years. Early on, beginning first in 1990, I was part of the initial discussions on building a different type of TOEFL exam, exploring ideas for TOEFL 2000, which eventually became TOEFL iBT. These experiences gave me a strong appreciation of the importance of well-designed, valid, and responsible measurements of language abilities. It also drove home for me the centrality of appropriate and valid measurement practices for all applied linguistics research.

As the blueprint for TOEFL iBT transitioned into production, I began a second cycle on the TOEFL Committee of Examiners (2002–2007). From 2005 to 2010, I also became a member of the Visiting Panel on Research, a senior committee that reviews all research areas of ETS. This committee led me to consider areas of assessment beyond L2 language situations and introduced me to the work of Dylan Wiliam, then a senior researcher at ETS (e.g. Wiliam Reference Wiliam2007/Reference Wiliam2008, Reference Wiliam2011; Black & Wiliam Reference Black and Wiliam1998, Reference Black, Wiliam and Gardner2012). I quickly came to realize that assessment should be as much, or more, about supporting learning than measuring learning. I still find this shift in conceptualizing assessment, as assessment for learning, very attractive and liberating when it comes to thinking about the relationship between language assessment and language learning. I have to say that part of the ‘education’ I've received in all my years interacting with TOEFL/ETS is being in rooms with very smart, very professional researchers, test developers, and policy experts, all of whom are seeking to do language assessment fairly, responsibly, and innovatively.

L2 reading and SLA

Since about 1994, I came to realize a fairly obvious fact within applied linguistics. There is a very large gap between SLA researchers and researchers/scholars interested in literacy skills development (whether reading, or writing, or both in some combination). It is obvious when one looks at standard major SLA textbooks and the ever growing set of handbooks over the past 15 or so years. It seems, to put it simply, that learning language structures and some conception of developing overall language proficiency are what interest SLA researchers while learning how language skills develop (as cognitive skills in and of themselves, e.g. Anderson Reference Anderson2000) is not of interest. I'm not entirely sure how this has come about generally in applied linguistics except to believe that researchers, like most other people, are products of their educational environment (or their linguistics training in the SLA case). However, this makes me an odd one out, coming from a linguistics background rather than an educational or cognitive psychology background. I think that this separation of reading (and also writing) from SLA is largely why I do not consider myself an SLA person.

There are, of course, SLA researchers who have bridged this gap. This was most forcefully brought to my attention when I first heard Nick Ellis give a talk in 1996 at the University of Arizona, Second Language Research Forum (SLRF) conference if I recall. Not only did Nick fully understand the centrality of language skills processing as a key to SLA but he also was the first person to introduce me to emergentism and the beginnings of usage grammar as a way to understand first language (L1) language development (as opposed to the Chomskyan black box theory of language acquisition). And, of course for me, I found a truly persuasive theory of language acquisition that drew on positive data and empirical quantitative evidence rather than a flawed ‘logical’ argument. Thereafter, I have read just about everything Nick has written up until recently (but that's OK, I'm convinced). Nick's articles on frequency effects in language learning (Ellis Reference Ellis2002) in Studies in Second Language Acquisition, and his three articles on psychological experimentation evidence about learning theories (one with Diane Larsen-Freeman) appearing in Applied Linguistics in 2006 and 2007 were cornerstones for me in seeing a newer bridge between SLA and literacy development (Ellis Reference Ellis2006a, Reference Ellis2006b; Ellis & Larsen-Freeman Reference Ellis and Larsen-Freeman2007).

His work also led me to Michael Tomasello and especially his Reference Tomasello2003 book Constructing a language, a book that I've read more than once. Tomasello's arguments and evidence strengthened my own view on the importance of learning from environmental input over time, the need for carefully designed empirical evidence, a renewed sense of the power of human associational learning skills, the centrality of the human capacity for pattern recognition as a fundamental cognitive concept in all learning, and the importance of implicit learning and automaticity development (see also Holger Diessel Reference Diessel2004, for L1 development). I now see compelling ways to connect SLA research and the development of literacy-related language skills, although in the past ten years, I've come to think this re-connection is better left to someone else to build. That's because another applied linguistics concept has engaged my interest and imagination.

I am now strongly committed to understanding better how to bridge research ideas and results with student learning, the language classroom, and teaching practice (see Grabe Reference Grabe2009; Grabe & Stoller Reference Grabe and Stoller2011). This commitment is not simply with identifying implications for teaching or learning, but with how ideas about learning can become the core of teaching practices (for better or worse), which practices become commonly accepted best practices, how new ideas can become new common best practices, how common teaching practices can actually be assessed for their effectiveness, and how teaching and learning can change for the better as a result of research on teaching practices and textbook curriculum design formulae. For me at least, this is now how I see my most useful contribution as an applied linguist, taking the concept of a mediator between language (linguistics) knowledge and real-world language issues to a more logical conclusion in the areas of L2 reading and L2 writing.

This most recent line of inquiry is, for me, a central question for applied linguistics as a mediating interdisciplinary field involving, on the one hand, what we know about how languages are learned, and what seems to be practiced in language learning classrooms. Simply put, this is the question of how applied linguistics moves from research to widespread instructional practices, and potentially back to research/theory (Grabe & Jiang Reference Grabe, Jiang and Cohen2009, Reference Grabe, Jiang, Evans, Anderson and Eggington2015). Most often applied linguists carry out research and find seemingly useful implications and interpretations for practitioners (primarily teachers). But what if these implications are sterile (limited to the research content), insufficiently robust, unfeasible in real-world learning and teaching settings, or just not of interest to practitioners? And what happens when the implications are engineered to be accessible for real-world applications, but teachers don't buy in, or publishers won't build textbooks from these ideas?

If we are really interested in mediating research and language knowledge to address language issues, we should want to take the research implications through to their being applied in many contexts and in typical classrooms with all of the real-world constraints on teaching and learning (Jiang Reference Jiang2012; Zhang Reference Zhang2013). Then we should find out if the ideas actually lead to any improvements – that is, have we actually addressed the real-world issue in a way that means learners and teachers do things differently, and the results amount to a better outcome? At such a point, researchers and teachers could decide together if there is improvement, and if so, how, and where, and when, and why. Of course, such an end point should circle back to new questions and directions for applied linguistics research. Having said this, I also realize that this logical progression is almost impossible to carry out. We already know that teaching is often criticized for not responding to the more recent research findings. At the same time, efforts to translate research concepts to easily useable classroom approaches without specialized training or resources often does not work (e.g. Klinger et al. Reference Klinger, Vaughn, Arguelles, Hughes and Leftwich2004; Hitchcock et al. Reference Hitchcock, Dimino, Kurki, Wilkins and Gersten2010).

Even if we could devise concepts that can be adaptable in textbook-sized frameworks and that teachers could use easily in many different contexts (like using reading comprehension questions after students read a text, or learning and then practicing 20 or so important words after reading a text, or doing daily ten-minute free-writing exercises, or writing the narrative-biography essay), there are additional challenges. Assuming that a set of ideas that should improve students’ language learning were packaged in such a way that an excellent L2 language textbook series was created, who would want to assess its effectiveness (aside from teacher and administrator impressions of effectiveness)? Why would a publisher encourage teachers and program administrators to find out, somewhat more objectively, if the new textbooks are any better, or even as good as, earlier textbooks in terms of student learning outcomes? Why would textbook writers want to participate in such research activities and potentially lose sales and contracts? And how can such assessments be carried out in ways that mitigate the many types of irrelevant sources of variance in many real classroom settings? So the challenges in moving from ‘implications for teaching from research’ to demonstrations that ‘new’ teaching approaches actually are better for learners is a daunting challenge. But it is one that is certainly worth contemplating, and perhaps acting on.

Task-based learning, content-based learning, and strategies-based learning

Of course, some approaches to learning and teaching simply don't fit well within the framework of published textbook resources. So they present yet different challenges for going beyond implication from research statements. It is at this juncture that larger approaches to language instruction play key roles, and here I am thinking about task-based learning, CBI, and strategies-based instruction (Pressley et al. Reference Pressley, El-Dinary, Gaskins, Schuder, Bergman, Almasi and Brown1992; Cohen Reference Cohen1998; Pressley Reference Pressley2006). These are ‘larger approaches’ because they really do not fit well within textbook structures. Task-based learning is a very important way to conceive of the relationship between research and learning improvements. For example, certain reading-writing academic tasks, such as writing synthesis papers or synthesis literature reviews, are not so much components of writing skills generally as much as they are the actual writing tasks that need to be mastered to demonstrate abilities for learning academic content and producing new content. In this way, the task itself is a goal that also happens to represent a type of academic writing ability. Having said this, I am skeptical that task-based language teaching, overall, can be a programmatic approach to language proficiency development across long spans of curricular time.

The choice between starting from the task or starting from the content to engage students with is one that has intrigued me for some time, and one that I had a few interesting discussions about with Joan Carson in the 1990s. Certainly, successful CBI needs sequences of effective tasks to drive learning; conversely, task-driven instruction needs meaningful, stimulating, and engaging content if task sequences are to maintain motivation and lead to long-term language development. I am not sure which comes first, although I know that finding truly stimulating content for a particular group of learners is very difficult to do, so that is why it comes first for me. In either case, I do not think that task-based learning or content-based learning can be tightly programmable. Rather, each will always take creativity, local context accountability, and teacher ‘artwork’ to make it successful.

The third, less common, variation I have noted is strategies-based instruction. This is an intriguing idea that has been around in one version or another since the 1980s. To take reading as an example, the goal of becoming a successful reader in academic settings requires a learner to become a strategic reader, and all that this idea entails. Reading strategies typically work in combination to help students understand and critically interpret challenging academic texts. Some approaches to developing students as strategic critical readers places the development of comprehension strategies at the center of the curriculum. For example, in L1 contexts, the description of the secondary-level Benchmark School in the U. S. A. centers on the uses of comprehension strategies across all hours of the day and across all classes in the day (see Pressley et al. Reference Pressley, Gaskins, Solic and Collins2006).

Andrew Cohen's 1998 book, Strategies in learning and using a second language, moves toward this concept for L2 contexts. While some experiments are reported along this line of curricular development with L2 students, strategies-based instruction has never, to my knowledge, been extensively implemented as a practical curriculum design in L2 settings. However, it is a well-developed component of content-based research in both L1 and L2 settings (but not as prominent with task-based instruction explanations). It could fairly easily be an overarching design concept that then draws upon excellent content and a series of core academic tasks to enable the development of strategic readers with academic texts. The potential for this strategy's ‘third way’ is that both content-based learning and task-based learning might be a means toward achieving more strategic language learners as the end goal (see Pressley Reference Pressley2006; Grabe Reference Grabe2009).

Where is applied linguistics after all this time?

Finally, I have participated in an issue that has stayed with me for most of my 30 years as a faculty member and applied linguist: What is applied linguistics? How am I defined by this field and how do I contribute? In considering these questions, my thinking has been strongly shaped over the years by Bob Kaplan, Dick Tucker, and Henry Widdowson, not that they all agree with each other all of the time. (I've also been influenced by that young whippersnapper, Norbert Schmitt, but I'm not sure I want to admit it.) First, I need to say that applied linguistics is decidedly not ‘linguistics applied’. Most applied linguists start with a strong grounding in language and linguistics knowledge, and that knowledge is necessary. But this foundation is a far cry from the idea that all we need is linguistic knowledge to address real-world language problems in effective ways. That is why applied linguistics must be interdisciplinary. Addressing real-world problems requires supporting knowledge and resources from multiple associated fields, and applied linguists almost always have reasonable knowledge resources in one or more related fields. Our ‘problems’ arise from issues in (incomplete) language learning, or language disadvantages, or language assessment, or language policy, or language variation and effective communication, or language loss. Addressing these problems will require more than linguistics applied.

Applied linguistics also emphasizes the need for a descriptive foundation to language explanations rather than singular theory-internal explanations and descriptions. If we, as applied linguists, are going to intersect with the real world, we need to do so with explanations that are readily understandable and useable. We also need to account for both the cognitive and the contextual patterns of variations and differences for realistic results that groups of people can use.

Not to be too blunt (or too controversial), there are some things that really don't matter much for applied linguistics. Below is a sampling from my perspective:

1. 1. Theory-internal arguments about language and learning

2. 2. ‘Accuracy, fluency, and complexity’ as the singular guiding construct for L2 proficiency development

3. 3. Loosely interpreted implications from research for teaching and learning

4. 4. Learning styles as sources of differential language learning (especially in academic settings)

5. 5. The concept of multiliteracies, which includes ideas such as spoken-language literacy, emotional literacy, media literacy (the problem here is with the term ‘literacy’)

6. 6. Research on learning specific grammatical structures as an end goal.

However, from my perspective, some other topics and themes really are important. A number of these have been touched upon in earlier sections. These key concepts include more general cognitive theories of learning, the centrality of vocabulary for using language, the role of practice and what counts as effective practice, the need for more longitudinal research, and valid assessments of language growth and change, as well as ethics and fairness in measurement. I would suggest that a few other concepts also need major prominence in applied linguistics research:

1. 1. The complementary, but equally important, roles of implicit learning and explicit learning in understanding language development and language use

2. 2. The importance of motivation and engagement as a major predictor in reading development, and perhaps in different ways, in language learning more generally

3. 3. The need for replications in L2 research if the field is going to derive legitimate generalizations from research questions. Too often in L2 settings, a few studies dominate a research topic and lead people to believe that the topic is a settled matter. However, little should be viewed as settled unless there are dozens of replications of a key finding.

In the past year, I have had a few exchanges with Kees de Bot, who is carrying out a research project that I'll call a ‘living’ history of applied linguistics. That is, he has been interviewing and communicating with at least 50 applied linguists to develop a participatory description of the field of applied linguistics. The project is endlessly stimulating and I've enjoyed a number of exchanges with him over his work. At the heart of his project is how applied linguistics matters, and for whom, and what counts as productive applied linguistics: in brief, what is the impact of applied linguistics and does applied linguistics make a difference? The answer appears to be that applied linguistics does make a difference. And in all its messiness – probably being no different from other messy fields such as psychology, biology, chemistry, and education – it impacts people's lives in many ways. With this last idea, I believe I have found a fitting way to conclude my personal perspective on 40 years as a teacher and applied linguist.

Oh, and in closing, remember: ‘We only become readers by reading, and by reading a lot.’