Interview with Edgar W. Schneider

Edgar, I’ve known you for almost 15 years and known of you for more than that. The scope of your interests and research—and output!—never ceases to amaze me. You are a world leading scholar in sociolinguistics, corpus linguistics, and variety studies. Many probably see your work on Early African American Vernacular English (AAVE) as the starting point of your career in all these domains, but your fascination with diversity and usage seems to have stretched wider than that. I am thinking of your work in dialectology. What was it that brought you into dialect geography and what took you from there to the focus on Early African American Vernacular English?

Thank you, Viveka!

Well, I’d love to say it was all strategically planned, but it wasn’t at all—where I grew up, in a small town in rural Austria, globalization was a world away, and when I was ten and started learning English at school the only reason I could imagine for doing that was in order to understand the lyrics of English pop songs. As a boy growing up in such a monolingual and monocultural environment I was fascinated by far-away lands and foreign cultures, and digested books about such topics.

At the University, then, I fell in love with linguistics, something I hadn’t heard of before. My first course was on generative grammar, in fact, the fashion topic then, and I was intrigued by the strange combination of intellectual concept-shuffling and empirical attestations. And then it was a series of lucky coincidences, and I guess me taking opportunities when they came up, and very soon I was pulled deeply into this fascinating field. I was a very good student and wrote a final thesis which my professor liked, so I got a job offered at a German university (Bamberg, a place I also hadn’t heard of before) with a chance to work on a PhD, which was thrilling, so I embarked on it with enthusiasm. And my love for foreign languages and different cultures received a lot of food.

My dissertation supervisor and then boss was a dialect geographer, so I learned a lot about linguistic atlases and their methodology, and as a research assistant I worked out data for a project in that domain. And that fascination was deepened by experiencing real-life language in Britain and the US, the South in particular, and by meeting eminent scholars and strong characters like Raven McDavid or Larry Davis, and by building a friendship with Bill Kretzschmar and later getting an opportunity to collaborate with him for a longer period of time, working on LAMSAS in Georgia (Kretzschmar & Schneider 1996).

Similarly with earlier African American English, my dissertation topic. I had been interested in African American music and culture, like many young people my age, but from a distance, really. But the attraction was there, and my dissertation supervisor directed my attention to the corpus of ex-slave narratives which George Rawick had edited and published (1972), and which promised to be a fruitful source on earlier Black dialect. Jeutonne Brewer, who I met much later, had written a UNC dissertation on the copula in these narratives (1974), but a comprehensive analysis of the characteristics of the speech represented there had not been carried out, so I embarked on this project enthusiastically. It was linguistically rich and also personally moving material, life stories written down largely as narrated, in interesting linguistic forms, a great project for me to carry through. The big diachronic issue in those days was how African or how British earlier Black dialect was, and these recordings promised and provided a window into the past, allowing an investigation of this question. And I was lucky because I had a friend in the US Army in Bamberg, so in hanging out with him and his buddies I did get some exposure to the dialect as spoken by African Americans as well.

And then things just got rolling, somehow. I published an article on some core results on verb inflection in American Speech in 1983 (Schneider 1983a), which attracted some attention precisely because its findings were controversial, contradicting the then orthodox view (the general assumption was that earlier AAVE, a putative creole, was largely uninflected, but I showed that it was actually unexpectedly rich in inflectional suffixes). I presented that material at my first conference, the Sociolinguistics Symposium in Sheffield in 1982, with Bill Labov in the audience giving extended comments. On the basis of this Salikoko “Sali” Mufwene invited me to his conference on Africanisms in Afro-American Language in Athens, GA (cf. Schneider 1993), so I had a chance to meet the leading creolists of the time, including Bill Stewart, Ian Hancock, John Rickford, Sally Thomason, and many more. And Ron Butters encouraged me to translate the dissertation and bring it out in the US, and its 1989 publication with Alabama Press (Schneider 1989) gave me a lot of mileage. It brought me in touch with many American sociolinguists, like Walt Wolfram, who reviewed it in Language, and Shana Poplack and Sali Tagliamonte recognized it as having substantially influenced their thinking (e.g., Poplack & Tagliamonte 1989:51-52). It kind of suited the spirit of the time of going back to the roots and searching for authentic historical records, and with others it triggered the diachronic turn of AAVE studies.

Well, there are several strands here that I would like to pursue. Before I move on to your work in the diachronic domain: your early encounters with dialect geography and atlases brought you into the area of data and methodology, which in itself led to a large output on the use of computational methods as well as corpus building and statistical methods in dialectology and sociolinguistics, culminating in your book with William Kretzschmar ( Kretzschmar & Schneider 1996 ). Now, some 25 years on, as computational methods, corpus building, and especially statistics have been more integrated in the general linguistic discourse, how would you compare the climate for such discussions and quantitative methods then and now?

That has changed dramatically, along with the availability of personal computers and their power to manipulate large datasets, which then was turned into software for statistics or corpus analysis. In the 1980s this was barely beginning—I bought my first Commodore 64 computer, ridiculous by today’s standards, in 1987, and email became available early in the 1990s, but only on a single computer in the department’s basement in Berlin. In 1988/89 I was an Adjunct Associate Professor at the University of Georgia, and Bill Kretzschmar and I were considering the suitability of statistical techniques in the computerization of LAMSAS, the Linguistic Atlas of the Middle and South Atlantic States, a huge set of data. We used to meet on a weekly basis and each time discuss a single chapter from a statistics textbook (Ott 1988, still on my shelf) and ask what it meant for atlas data: what is the nature of the informant sample?; how can responses or communities be understood as trials in a statistical sense?; where do we get a binomial distribution, a normal distribution? Consequently, which tests can be applied, and how can the atlas data be tuned to lend themselves to semi-automatized statistical analysis? The notes from these meetings were then expanded to yield the manuscript of Atlas by the numbers (Kretzschmar & Schneider 1996), an interesting process.

Bill continued to move along this path, with exploring things like GIS software and large-scale data building and analysis, which ultimately led to his Linguistics of Speech Cambridge University Press book in 2009 (Kretzschmar 2009). For me, one of the immediate consequences was to fuel my interest in chaos theory—I gave a paper on its suitability for the understanding of language variability at a Methods conference (Schneider 1997a). And the topic attracted a lot of interest when I gave a lecture tour in Australia and New Zealand in 2001. But then I got “distracted”—by editing (English World-Wide and the Handbook of Varieties of English) and working on World Englishes and the “Dynamic Model.” Interestingly enough, in the long run and in the recent past independently of each other Bill and I have reached the same goal, extrapolated from quantitative data observation and thinking about variability—the idea that language should best be understood as a “Complex Dynamic System” in the sense of the science-derived systems theory. Last year I gave a plenary on this topic at the Changing English conference in Helsinki (Schneider 2015a; and I plan to pursue this line of thinking), and Bill, well ahead, published his book Language and Complex Systems (Kretzschmar 2015).

Statistical techniques have become central to some branches of linguistics (you know that better than I do, Viveka, because you teach this), and the level of sophistication that some corpus linguistics bring in, or that you find at the annual New Ways of Analyzing Variation (NWAV) conferences is admirable. This is important and helpful for the assessment of one’s data—though I do think it shouldn’t be an end in itself. Occasionally you find publications where it seems the ease and attraction of number crunching have become overwhelming. I have to admit when I read that the most frequent bigram of American English is and the I ask myself if this is helpful in understanding how language “works”—which I do think is the ultimate goal of linguistics. So despite all my admiration for the stats whizzes in the younger generation of colleagues I still think that qualitative documentation, quoting and analyzing individual structural examples, is important, and even distributions below a conventional level of statistical significance can be meaningful and indicative—of initial stages in a process of change, for instance. I firmly believe that ultimately any linguistic analysis must be issue-driven: you have to have a structural hypothesis which can then be translated into a quantitative testing procedure, not the other way round.

I guess the leitmotif behind all of this, and a central cause of my fascination with language, is the relationship between unity and diversity, the sameness of speakers’ shared utterance forms and linguistic knowledge (needed in order to communicate successfully) and the difference between speakers’ or dialects’ performance habits, features not shared and hence potentially endangering successful communication (which leaves room for indexicality, the socially and contextually symbolic function of linguistic forms). The coexistence of the idea of one “English” with the reality of so many different “Englishes” is a most remarkable fact to me—seemingly trivial but opening up so many facets and questions.

I agree, and I certainly share your fascination with the relationship between unity and diversity in and across languages. But taking a step back: after the publication of your dissertation, and parallel to your work on quantitative methods, the diachronic angle stayed with you, didn’t it? Not only the diachronic angle of AAVE studies, which has been pursued and deepened by your students, but also your work in historical linguistics and with the Helsinki corpora. It seems to me that the combination of your thinking on quantitative methods and data crunching and your thinking on the evolution and development of language are in a sense interwoven. What are your thoughts on that? And might this also have contributed to your fascination with chaos theory as a model for variability and change?

Yes, you’re definitely right. There are several topics and branches in our field which I have always seen as intrinsically related and which have continued to fascinate me and probably to fuel each other: the diachrony of English, including its varieties; evolutionary thinking and the principles behind the ongoing and never-ending patterns of variation and change; and the fact that much of this proceeds incrementally, via some quantitative waxing and waning of formal choices but also allowing for the occasional catastrophic changes and qualitative leaps. Somehow that’s how language works and evolves.

As a student I was trained in Old and Middle English—not as thoroughly as earlier generations of scholars, but more substantially than later ones, since in recent years at least in Germany the history of English has tended to be fairly marginalized as a field of study, which is a pity, I think. Agreed, you may not need to know each and every formal variant of some strong verb form in some Old English dialect, but developing a sense for the regularity but also the specific conditionality of long-term processes is something I have always found illuminating and instructive. The same applies to other related phenomena—the tension between the persistence of long-term developmental trends and the danger of sudden collapse, or the continuation of developmental trajectories across varieties, irrespective of whether they are regarded as a “standard” or “just” some kind of dialect. To illustrate this by means of an example—if we look at the past tense formation of English verbs with or without vowel alternation you see a beautiful interplay of regularity and irregularity, of systematicity and idiosyncrasies across millennia. “Strong verbs” with ablaut were once, in Indo-European, a perfectly regular, phonetically predictable set of positionally conditioned variants, but in the course of about two thousand years or more and under the impact of all kinds of locally conditioning interfering factors the system got gradually and increasingly disrupted, to yield a largely unpredictable set of “irregular verbs” today. In dialects, however, you find all kinds of regularizations and attempts at local systematicity being either retained or re-instituted, as in bring—brought / brung / bringed, help—helped / holp, sneak—sneaked / snuck. Or take demonstratives—a sub-system where the change from order to chaos and back happened much more quickly: a highly complex but fully systematic set of very many formal variants by gender, case, number, and syntactic position in Old English fell to pieces rather quickly afterwards, but another such system re-emerged, with just a few cornerstone forms retained, as a rather simple but clear two-dimensional set of choices (this/these, that/those) in Early Modern English (and some speakers and some dialects can then complexify things by adding further options such as this here / them there, or yon and yonder). And so on. Yes, I do think that chaos theory and complex systems theory, developed in quite different scientific contexts, perhaps alien to linguists, describe principles, processes, and relationships which we see at work quite effectively in the evolution of language varieties as well.

Generations of historical linguists and variationists or sociolinguists have investigated and documented such processes, of course—with varying theoretical perspectives and interests, often focused on specific data sets and contexts or also, at other times, with an eye on more fundamental principles (though, if I see this correctly, hardly ever with the chaos or complex systems approaches in mind). My own take on this has materialized in various surface forms indeed, but somehow, I think, they have all been connected by a similar common thread. My dissertation work on earlier Black English looked at possible influences on and changes within this fascinating dialect, which shows both similarities with other varieties of English and also distinctive properties of its own. For a while I did use historical corpora to investigate specific long-term changes, in Early Modern English in particular, and in later varieties of English, analyzing, for example, certain lexico-semantic choices associated with syntactic complementation patterns (the verb consider, with all of its variant uses, has been one of my favorites), or the loss of endings in wh-pronouns, such as who(m). And together with colleagues, PhD students, and junior colleagues we have built a fairly unique set of nonstandard text corpora here at Regensburg, which allow investigations across ethnic and social varieties and time periods—the Southern Plantation Overseers Corpus (SPOC) of letters, inspired by Michael Montgomery; the early Blues Lyrics Corpus (BLUR), analyzed mainly by Ulrich Miethaner, and, also with Michael Montgomery and mainly built by Lucia Siebers, the Corpus of Older African-American Letters (COAAL). And somehow I see my later work on World Englishes as a continuation of this line of evolutionary thinking as well. All of these look at the interplay between synchronic relationships and diachronic developments, starting points in time and trajectories along which they move ahead, of both sub-systems of linguistic forms and choices on the one hand and social relationships within and across speech communities on the other. All of these processes are similar in some respects, and different in others, and it always depends on whether we decide to focus on the similarities or the differences, the details and data on the one hand or the “grand picture,” the large-scale patterns and evolutionary principles, on the other.

You did in fact delve into chaos theory at the turn of this century. Can you maybe expound a little on what made you join the various threads we’ve been talking about and view them through the chaos looking glass?

The chaos line of thinking was actually something like a side line for a while, a project intention which excited me but which was then backgrounded because of too many other, equally challenging, options and opportunities. The paper on chaos theory as a possible model for dialect variability and change mentioned above was printed in the conference proceedings edited by Alan Thomas (Schneider 1997a), but the book was not produced by a commercial publisher, so it didn’t reach a wide audience. I had read some pertinent publications, like Gleick’s book Chaos (1987), and in my paper suggested that many principles recognized in that theory—like the alternation between areas or periods of relative stability and areas or periods of high variability, the introduction of large-scale systemic differences by minimal initial distinctions (as predicted by the “butterfly effect”), formal splits via “bifurcations,” etc.—operate in speech performance and linguistic evolution in a rather similar fashion. In the following years I tried to apply these principles to examples from the history of English (I hinted at some of that in my response to your previous question). Since then, however, most of this has remained on the back burner, because of other obligations and mainly because new topics and orientations attracted and demanded my full attention—related ones via the core notion of variability but quite different ones in actual detail.

That was mainly my move into World Englishes, triggered by the fact that Manfred Görlach had offered me the role of editor of the journal English World-Wide and the book series Varieties of English around the World, which I took over in 1997 and held for more than fifteen years. I had done a little work in that area before (cf. Köppl et al. 1983; Schneider 1983b; Schneider 1997b), though only marginally, but I was immediately thrilled by the vibrant developments of these new varieties of English, mainly in Asia and Africa, and I was happy to fully embark on this topic. In many respects it opened a new world and innovative, most attractive opportunities to me. With connections established by Thiru Kandiah from Sri Lanka, I attended my first conferences of the International Association of World Englishes (IAWE) in Singapore in 1997, and later, 1999, in Japan (the IAWE conference in Tsukuba and the AILA Congress in Tokyo) and Hong Kong (visiting Kingsley Bolton and his ICE Hong Kong project), and I met Braj and Yamuna Kachru, the late Larry Smith, and many other colleagues who since then have become dear colleagues or friends. I have become a regular at those conferences (one of which, in 2007, I then organized myself, together with my junior colleagues in Regensburg; cf. Hoffmann and Siebers 2009). Through contacts and personal exposure the IAWE meetings became immensely important learning environments for me personally (and I simply have to confess I love to travel and to experience new cultures and countries, and try to sense what is happening there linguistically). So rather than continuing thinking on chaos and the history of English, I was fully occupied by working myself into World Englishes, journal editing (hugely interesting, wonderful in terms of establishing global contacts, but also immensely time-consuming), editing the 2004 two-volume Handbook of Varieties of English mainly with Bernd Kortmann (cf. Kortmann & Schneider 2004; Kortmann et al. 2004; Schneider et al. 2004), and developing the “Dynamic Model” over the years.

So, in keeping with the chaos theory, you in a sense took a new turn due to a series of events and ended up on the path which, arguably, is what you are best known for, the Dynamic Model. Yet I wonder if the germs for the Dynamic Model might not lie in the combination of your work on quantitative methods and on the evolution and development of language, now with the added ingredients of the chaos theory, all projected on the backdrop of World Englishes. What are your thoughts on that? And could you maybe trace for us how you first came to formulate the earlier version of the Dynamic Model ( Schneider 2003a )?

The relationship between these strands of my research interests is an indirect one at best, I suppose, linked via an underlying similar set of beliefs on how linguistic systems evolve. I suppose Sali Mufwene, with his 2001 book The Ecology of Language Evolution and his “feature pool” idea, has influenced me substantially in this, and ideas culled from chaos theory have also helped shape my thinking on what has happened to New Englishes. Yes, I do believe that chaos and complex systems principles have been instrumental in the evolution of World Englishes as well. This includes the core idea of transmission of language forms across generations with the option of incremental changes being introduced in these cases of replication, and also the potential for new patterns and complexity to auto-emerge. But basically these notions would have to be pinned down on the level of individual human utterances, in a usage-based framework. Essentially the core of usage-based evolutionary thinking as I see it is that every utterance is shaped by earlier utterances and at the same time constitutes the germ for subsequent ones and contributes to the overall shape of the system. And the relationship between the set of all relevant utterances, speech forms produced at a given setting defined by space and time, is sometimes systematic and distinctive, sometimes variable, sometimes chaotic, sometimes emergent, and so on, influenced by and interacting with many parameters, essentially cognitive and social ones. But in the end to apply chaotic modeling, ideally mathematically, effectively we’d need something like a complete record of all relevant utterances in a given area and over a longer period of time—which is no more than a thought experiment, I’m afraid. The Dynamic Model as such is much broader and more abstract in orientation, highlighting sociopolitical and sociolinguistic aspects of the emergence of New Englishes rather than structural analysis. Of course, complexity principles could be seen as operating on these social levels as well.

As you know, in essence the Dynamic Model claims that despite all differences between world regions and historical settings there are fundamental similarities, an underlying uniform sociolinguistic process, which unify basically all World Englishes which have emerged in colonial history. It suggests a unilateral causal relationship, such that historical and political background facts were decisive for the identity constructions of the parties involved in a colonial process, which draw a line between who belongs to “us” and who is “them”; these identity settings determined the precise sociolinguistic conditions of linguistic contact and use as well as language attitudes, which in turn were the main determinants of linguistic developments, with certain forms and structures turning out to be more successful than others, ultimately to become adopted in a given social community. The two main parties in a colonial process, and consequently the two main perspectives or “strands” from which this evolutionary pattern was experienced, were the indigenous population of a territory which fell under outside colonial rule and the settlers who migrated to new lands as colonizers or representatives of the “motherland” or “metropolis.” And this process of both social and linguistic evolution, ultimately the birth of a new language variety, proceeds through five subsequent stages, which I called “foundation,” “exonormative stabilization,” “nativization,” “endonormative stabilization,” and “differentiation,” respectively. The basic trajectory of the process is such that the social and linguistic distance between colonizers and colonized is large in the beginning but then continuously decreases, with ties to the homeland (and, consequently, metropolis-directed identities of colonizers and their descendants) getting weaker, especially after independence and nation building, and new shared identities across ethnic boundaries becoming stronger.

The model as such, and details associated with it and with its application cases, kept growing and getting more complex in the course of time, also in my mind and in my computer. Again, it was not strategically planned; in fact, it rather developed and expanded over a number of years, and I see this as an interaction of intuitive hunches and insights, a coincidence of occasions and opportunities, and of course then also simply a lot of work. For me, it grew and kept expanding in three successive stages.

The intuitive insights into similarities between many World Englishes and their evolutionary processes started growing throughout the late 1990s, as a consequence of my increasing familiarity with all these varieties through journal and book editing, research work, and personal experience. The first version of the model was still rather simple. In 2001 Pam Peters invited me to give a plenary to the Australian Style Council in Sydney, and as this is not an audience of linguists, rather than coming up with some sort of structural analysis, which I used to do at conferences, I decided to present a systematized but not yet very complex version of this idea of cross-variety evolutionary similarities (later published as Schneider 2003b). The idea of cyclic thinking and of evolutionary stages in new varieties of English was not mine—influential earlier, though considerably simpler versions, with different stages suggested, included Moag’s proposal for Fiji English in the 1992 Kachru volume and similar suggestions by Josef Schmied in his book English in Africa (1991), and of course it was also indirectly influenced by theories in pidgin and creole linguistics on the pidgin-to-creole lifecycles established in pidgin and creole linguistics. The Sydney presentation of 2001 discussed the conceptual background of language variation studies under contact conditions, and then suggested the evolutionary idea for New Englishes for the first time. It introduced the five phases and some of their essential characteristics, and hinted at specific applications to about a handful of countries around the Pacific Rim, but only in about a paragraph in each case. But the presentation and the idea met with a whole lot of interest and support—on the side of the audience and also with Simon Elmes, a British journalist who was preparing a popular series of books on World Englishes himself (Elmes 2001), happened to be attending this meeting, and interviewed and encouraged me afterwards. And this feedback turned out to be valuable for me and motivated me to carry on.

So in the beginning I really saw this as a “light,” non-technical (i.e., not structurally analytic) topic, and it was only then that I recognized its potential and decided to test my ideas more systematically by reading up substantially on relevant aspects, comparing details from many different countries, and elaborating on both the components of the theoretical framework and the documentary details of the application cases. The result was a rather lengthy article text, which, since I thought it was of rather general and fundamental interest, I offered to Language, with many more details and application cases and a much more substantial theoretical framework. And I confess I was extremely happy to get it accepted, despite its unusual length and practically without revisions. Clearly the 2003 publication of the model in Language helped to make it much more widely known and, with the journal being what it is, the most recognized linguistics journal of all with highly competitive acceptance rates, it helped to get the model seen, recognized, diffused, and accepted in the long run. Consequently, I kept the momentum going and decided to work it out in full, and the result is the best-known, fully elaborated version of the model in my 2007 Cambridge University Press book Postcolonial English. Again, Salikoko Mufwene was extremely supportive and influential in all of this. The book has a lot of background discussion, extensive theoretical reflections on aspects of the model, a lengthy chapter with an application of the model to the growth of American English, seventeen rather detailed applications to individual countries from almost all continents, and discussions of structural and social issues. And I’ve been lucky—it has turned out to be highly successful in many ways.

Well, it certainly was no mean feat with the Language article—and hardly any revisions! I can well imagine that such recognition made you happy. And the model was very well received indeed. I think it’s fair to say that you not only represented a new watershed in the field of varieties, but ushered in a wave of new research. It has been and is being applied widely to and tested on a number of different Englishes. When you see it used in practice like that, do you still “recognize” your model or do you maybe feel that it is in a sense being reinterpreted and tweaked as it is applied?

Yes indeed, there has been a long series of references, applications, discussions, and also suggestions for modification, often with respect to specific application cases. And that’s great, I think—if an idea is valid and useful the best thing that can happen is that it takes up a life of its own and moves beyond its original confines. And it is most welcome to see suggestions for improvements and modifications, both in principle and with respect to individual case studies. After all, as I have frequently emphasized, a model is an abstraction—it is meant to help us understand an underlying, complex reality, it is supposed to highlight interesting or important aspects of a process or similarities across varying instantiations, but it is not meant to be slavishly applied with respect to every tiny detail in all cases, and it is not conceived of as carved in stone.

Actually, I have found it very interesting to see all the various innovative applications and the ways of adapting the model and reacting to it. One of the earliest constructive reactions was a 2007 article by Joybrato “Danny” Mukherjee (2007) in this very journal, in which he expanded and revised my case study of Indian English, suggesting it has reached a “steady state equilibrium” stage in phase 4 (endonormative stabilization). One of the most creative and surprising ones was by a colleague from South Africa who I do not know in person, Brenda Spencer, who (amongst other things) identified the FIFA soccer World Cup as an identity-turning “Event X” (as posited in my framework) in South Africa (Spencer 2011). There have been a few sophisticated technical attempts at testing predictions derived from the model with statistical machinery, e.g., with respect to the emergence of complex and distinctive verb patterns in later developmental phases (e.g., Mukherjee & Gries 2009), and I was pleased to see most of them found the predictions to be borne out by the data at hand. Weston (2011) suggested a few adjustments when applying the model to Gibraltar; Bautista (2010), Borlongan (2011), and Martin (2014) applied it to the Philippines, Huber (2014) to Ghana, Tan, Kuang, and Low (2010) to Malaysia, Imm (2016) to Singapore, Bennui and Hashim (2014) to Thailand, Biewer (2015) and Britain and Matsumoto (2015) to the South Pacific, Marsden (2013) to regional diversification of New Zealand English as a sign of phase 5 (differentiation), Collins (2012) to a range of varieties on a comparative basis, and so on (and you see I’ve been keeping track of the application cases that I have become aware of!). A class project at Northeastern Illinois University directed by Richard Hallet discussed the applicability of the model to the case of Malta in great detail, and the result, authored by a group of twelve students, was then published in 2009 in the journal English Today (Thusat et al. 2009). Bekker (2009), van Rooy and Terblanche (2010), and du Plessis (2015) broke aspects of the model down to specific ethnic groups in South Africa. Stephen Evans investigated the growth of Hong Kong English on the basis of historical demographic and also corpus-based data in a recent series of articles (e.g., Evans 2014), and my impression is that while he repeatedly wishes to sound critical in fact he regularly states that my statements describing the history of English in Hong Kong in the model’s framework are largely confirmed except for some dividing dates between subsequent phases, which he adjusts. And no, I wouldn’t fight for every single detail in all the application cases of my 2007 book, so that’s fine, and it is an improvement, but of course it is good to see the basic developmental trajectory broadly and repeatedly confirmed.

Actually, I think the relationship between details (relevant but secondary in my view) and the “grand picture” is important and interesting insofar as somehow it reflects changing realities and boundaries getting blurred. And in this respect the discipline of World Englishes struggles to follow and mirror rapidly changing social realities. Some three decades ago, varieties, or the status of English in nation states, could still be categorized rather neatly—into Kachru’s (1992) “Three Circles” (“Inner,” “Outer,” and “Expanding,” respectively) or the ENL (English as a Native Language)—ESL (English as a Second Language)—EFL (English as a Foreign Language) types, for instance. But that is not the case any longer, with differences getting increasingly less clear-cut, and rapid developments in many nations and contexts changing linguistic settings quite drastically. A junior colleague of mine here at Regensburg, Sarah Buschfeld, showed in her work on Cyprus English (2013) that individual varieties can move back and forth, from ESL back to EFL, for instance, and she argues (rightly, I think) that in fact it makes much more sense to see this as a continuum rather than a binary distinction (and right now, in another exciting research project, she’s tracking down another fluid boundary, the change of English in Singapore from second-language to native-language status).

It has been one of the distinctive properties of the Dynamic Model that it refrained from positing such boundaries but in fact integrated all kinds of differences under the overarching umbrella of identifying many of them as manifestations of different states along a developmental trajectory. That’s both a strength and a possible source of contention, I realize that. Raj Mestrie and Rakesh Bhatt in their 2008 textbook World Englishes criticize the model for trying to treat ENL and ESL settings jointly (cf. Mesthrie & Bhatt 2008:31-36). I see and recognize their point—whether a country was essentially settled by a majority of British people, such as New Zealand, or linguistically influenced by just a rather limited number of representatives of the English Crown, as in Nigeria, for instance, makes a difference, of course. But my point is that it is all the more remarkable that still substantial similarities in the language contact conditions and effects and in developmental directions can be identified, and so with others (e.g., Melchers & Shaw 2011 in the second edition of their textbook World Englishes) I do believe that as long as we maintain a reasonable degree of reluctance in pointing out and interpreting parallelisms this is actually one of its exciting features and strengths. One interesting sideline of this is whether the model really has a “predictive” value, the implication being that a nation at phase n will be moving on to phase n+1, which somehow seems to be implied. And I think that’s an empirical question for future observation—I think so, but only time will tell.

More generally, a growing recent trend appears to have been the desire to integrate Expanding Circle countries within such general explanatory frameworks, i.e., to identify similarities between ESL nations and those where English is essentially still a foreign language but growing strongly. Ike’s (2012) Melbourne dissertation, for instance, applied the model to Japan—which clearly goes beyond its original confines and its intended reference to “postcolonial” settings. The strongest and most fruitful application along these lines is currently the dissertation and now new book by Edwards (2016), in which she finds many features predicted by the model also present in the Netherlands, clearly an Expanding Circle country, and she rightly suggests that in contexts such as this one the force of globalization seems to have taken over the former role of colonialism. In a 2016 article in World Englishes my friends and colleagues from Regensburg Buschfeld and Kautzsch (forthcoming) propose an expansion, or adjustment, or supplement, of the Dynamic Model, which they call the “Extra- and Intra-Territorial Forces” model, arguing along rather similar lines, proposing a more encompassing framework, and applying these ideas to interesting new case studies such as Namibia (where, strangely enough, English was proclaimed the country’s official language at the time of independence in 1990, despite the absence of any British colonial past or a substantial presence of English in the country at that time; and the substantial effects of that decision are becoming visible now).

So yes, I recognize the model in all of these instances, and I very much appreciate and enjoy seeing it growing and being “tweaked,” as you call this—like a child maturing, developing features, properties, and abilities of its own. (And, honestly, I’ve found this much more interesting than to rehash it on my own, as I have been requested to do repeatedly for some of the highly popular handbooks—popular at least with publishers, I suspect—that have been produced by almost all major presses in the field over the last decade or so.) It’s great to see it alive and kicking, and it clearly is, at least in the branch of linguistics known as “World Englishes” research.

Yes, I imagine that it must be gratifying to have the model validated by such extensive application, and to see it take on a life of its own. It seems to me that the field of Varieties of English has in the last few years taken a more diachronic turn, in that earlier stages of individual World Englishes are being looked at more closely. I am thinking of, for example, Magnus Huber’s work using early recordings to trace the phonological history of English in Ghana (e.g., Huber forthcoming ), as well as Thorsten Brato’s, Sebastian Hoffmann et al.’s, and Biewer et al.’s work using early newspaper articles to build historical corpora of Ghanaian English (e.g., Brato in preparation ), Singapore English (e.g., Hoffmann, Sand & Tan 2012 ), and Hong Kong English (e.g., Biewer et al. 2014 ) respectively, and several other similar projects. Could that be a kind of spin-off effect of your model? And what are your thoughts in general about this diachronic turn?

Absolutely, this “diachronic turn,” the search for archival evidence of earlier stages of language use and in a few cases the construction of systematically compiled historical corpora, is one of the major trends in World Englishes research today. And I think you’ve mentioned the most ambitious and important ongoing projects, perhaps in addition to Ariane Borlongan’s compilation of an early Philippines English corpus to match the American Brown Corpus (e.g., Collins, Borlongan & Yao 2014). Calling this recent trend a spin-off effect of the Dynamic Model would be way too much, I suppose. I guess the model did have a triggering effect in going diachronic in a much more explicit fashion than all earlier approaches to global varieties of English; it invites empirical testing; and some of the authors and researchers you mentioned refer to it explicitly. But beyond that and to some extent independently of it the desire to unearth historical evidence, to investigate how a system has come to be what it is now, just comes natural after an initial stage of recognizing something, in our case language varieties and linguistic settings, as distinct. Social systems, including language, emerge and grow and change in time (a sideline of Complex Dynamic Systems thinking, if you so wish), and so in order to understand and reliably investigate the conditions and processes of evolution we need first-hand, original evidence. And that is what these projects intend to compile and provide, a most welcome step which promises much needed and more detailed insights into the growth of New Englishes. Actually, the upcoming conference of the International Society for the Linguistics of English (ISLE) in Poznan, Poland, this coming September [2016] will host a workshop on diachronic corpora of World Englishes, and I look forward to attending it and to participating in the summary discussion, which I’ve been invited to chair.

I think the fact that such a diachronic turn comes natural in a maturing social or sociolinguistic sub-discipline has been evident in related fields as well. Think of pidgin and creole studies—this aspect, like so many others, is authoritatively covered in your recent textbook as well (Velupillai 2015). Soon after the discovery of structural similarities between creoles across different lexical bases many colleagues have worked hard to identify and analyze historical documents and thereby to test the competing hypotheses on creole genesis empirically. Work on the Surinamese creoles by Jacques Arends (e.g., 1989, 1993), and also many others, has motivated the gradualist model of creole genesis. Sarah J. Roberts’s work (e.g., 2000, 2005) on diachronic data from Hawai’i has turned out to be most important for a realistic assessment of the Derek Bickerton’s Language Bioprogram Hypothesis (e.g., 1981, 1984). And we have seen a range of similar collections and investigations of old texts—on Negerhollands, recorded rather extensively in early texts by missionaries (see, e.g., Stein 1996 and van Sluijs 2013 with further references); on Jamaican Creole in the collection by Lalla and D’Costa (1989); on Guyanese in a section of John Rickford’s 1987 book; and so on. We have also experienced that diachronic turn in the study of African American English, where my own dissertation investigating ex-slave narratives was the first comprehensive diachronic text analysis in its time (Schneider 1981, 1989), soon followed by others, such as Poplack’s and Tagliamonte’s (1989, 2001) work on speech islands like the Samaná peninsula and on letter corpora, Kautzsch’s (2002) study including the Hyatt “Hoodoo” recordings, and others. Sonja Lanehart’s recent Handbook of African American Language (Lanehart 2015) has an article of mine in which I’ve attempted to summarize this aspect, the growth of sources on the diachrony of the dialect (Schneider 2015b). And more broadly, this ties in with what we have been experiencing in dialect geography as well, where there has also been a strong desire to build systematic and large-scale documentary resources of earlier vernacular speech. This started with individual activities, such as Eliason’s (1956) Tarheel Talk compilation of early North Carolina dialect, and has led to larger compilations and projects such as Michael Montgomery’s and my SPOC (Southern Plantation Overseers Corpus) and the other diachronic dialect corpora we’ve compiled at Regensburg University, mentioned above, or Michael’s ongoing activity of building CACWL, a Corpus of American Civil War Letters, in collaboration with Mike Ellis. This collection is now being made available online, with the support of some historians, in an “eHistory” project (www.ehistory.org), so clearly activities along these lines will carry on and are embedded in wider frameworks.

And of course all of this ties in with historical linguistics as such and historical sociolinguistics in particular, so somehow we’ve come full circle. The basic interest in finding direct evidence to better understand historical processes, including linguistic evolution, is the same across the disciplines referred to here. And many methodological issues and problems are largely identical as well—the World Englishes diachronic projects have to face them in much the same way as, say, researchers on Early Modern English have always done. As we all know, the difficulty results from what Labov (1972) called the “bad data problem”—most of the time when it comes to earlier vernacular speech we simply do not have the kind and the amount of original data that we would want and need in order to understand natural change in these periods and contexts. The directors of the projects on the diachrony of World Englishes, some of whom I’ve talked to about this, are confronted with the very same problems, give or take some local specifics, as historical linguists and historical sociolinguists have always been. Textual evidence that happened to have been preserved and that has come down to us in writing is usually not what we’d hope for, and is not directly comparable to what can be systematically elicited in a present-day context. Most of the time in terms of quantity it may not be enough, and in terms of quality it may not be the right kind of data. There is the issue of validity—documents which have been preserved in archives tend to represent formal language and contexts, not vernacular speech. And the issue of representativeness –the writers and the text types represented in such documents are usually not typical of the “average speaker,” in whose performance we’d be interested when it comes to natural, informal language use. We have learned a lot, and developed improved methodological skills—there are solid concepts and approaches to tentatively fill the gaps (like the “uniformitarian principle,” assuming that we can project present-day observations and relations back in time in certain ways), assess the sources we have, open new resources (cf. Schneider 2013a)—but gaps they remain, somehow. These are challenges. But challenges are there to be overcome as well as we can, they make life and scholarship interesting.

Yes, we have indeed come full circle. And this ties in with the thinking on Complex Dynamic Systems that you’ve mentioned above, which seems to me a logical extension to the Dynamic Model and the direction you are heading at the moment. Could you maybe tell us a bit more about your ideas on Complex Dynamic Systems as a framework for language evolution in general and World Englishes in particular?

Yes indeed, adopting the Complex Dynamic Systems approach most likely constitutes some sort of a continuation of my earlier lines of thinking, my earlier activities across sub-fields, all of which have to do with language variability and evolution, I suppose. The theory offers an overarching framework which somehow promises to bring it all together and which places linguistic usage also in the wider framework of other systems which have grown naturally. It is related to, but is more comprehensive than, chaos theory, which we talked about earlier, and some of the principles and examples I mentioned before, on the interaction of chaotic and ordered sub-systems at different times and places, can be integrated here as well, but as I see it, Complex Systems Theory is more powerful, more fundamentally explanatory. I’ve been fascinated by it for a while. I would not claim to be a real expert, but I have read a few books on the subject—for a start, I’d recommend relatively concise introductions like Johnson (2009) and Holland (2014) but also the monumental and most impressive survey volume by Mobus and Kalton (2015). And there is a small number of applications to, or thoughts on its applicability to, languages, notably work by Diane Larsen-Freeman from an applied perspective and by Nick Ellis, with a strong psycholinguistic twist (e.g., Larsen-Freeman & Cameron 2008; Ellis & Larsen-Freeman 2009), and of course now there is Kretzschmar’s (2015) book which I mentioned earlier, inspired by distributions observed in dialectological data. My own take on this, so far, is a keynote conference presentation, also referred to above (Schneider 2015a).

The theory of Complex Dynamic Systems (CDS), also known as “Systems Science,” “Complexity Science,” or theory of “Complex Adaptive Systems,” has been established and growing vigorously over the last few decades in the natural and social sciences and in mathematics, and it has seen a wide range of applications in biology, medicine, social organization, in explanations of natural phenomena, technology, business processes, and so on. At this point it is probably best seen as a meta-science, one which is not yet fully developed and does not have a “canonical” form, but it insightfully captures many basic properties of many domains in life and of living or social organisms. Its approach is holistic and cross-disciplinary, so in a sense it is also directed against reductionist, categorical or deterministic thinking as is found, I think, also in the early leading theories in linguistics such as Saussurean structuralism or Chomskyan generativism. When it comes to views of language, CDS theory seems much better in line with recent usage-based approaches—Bybee (2010) refers to some similarities between language and “Complex Adaptive Systems,” for instance.

CDS are characterized by a set of typical interacting and partly interrelated properties, some of which are shared with and perhaps emerge from chaotic systems, such as alternations between order and chaos, nonlinearity, or fractal patterns (which are independent of scale). They are systemic (with several types of constituent components interacting with each other), complex (with large numbers of objects or agents building hierarchical organizational levels), organized in networks of relations, and perpetually in flux, dynamically changing all the time, via reproductions and replications of their constituent entities. (For instance, I found it remarkable to see that the well-known S-curve pattern of the diffusion of linguistic innovations in sociolinguistics has an exact match, a logistic function with an exponential rise and then deceleration, in population biology; cf. Mobus & Kalton 2015:215). Most importantly, such systems are emergentist and auto-organizing, so that the constant interaction of less complicated entities in an evolutionary, self-organizing capacity builds higher-order, more complex clusters and sub-systems. Clearly we find manifestations of such processes in languages as well, for example in phenomena such as chunking and compositionality, in hierarchies of units and unit clusters and increasingly complex functions associated with them, or in processes of adjustments and adaptations of language forms, both internally, mutually to each other, and externally, to conditions of social usage. In my 2015 lecture I gave a range of examples drawn from the history of English and from a few World Englishes in which I see these basic principles and evolutionary processes manifesting themselves. And I am convinced there are many more—and that is something I hope to work out and show in some future work.

One of the beauties of the CDS approach is that it unites it all—the history of English, the history of varieties, the evolution of World Englishes, you name it; these fundamental principles operate all across the board. And much like in the sciences, where we find anything from rather loose and metaphorical applications of such thinking in social systems, the business world, etc., to highly formalized and strict applications in mathematics and biology, when we look at language the framework will also cover both sociolinguistic, language-external developmental processes and language-internal processes of structural emergence. The Dynamic Model, with its built-in extralinguistic, historical, and sociolinguistic components, largely operates on the basis of a broader, socially oriented understanding of CDS. On the other hand, structural processes in the narrow sense illustrate the more mechanical, subconscious processes of auto-organization (in case you want examples, fairly well documented in the literature on the historical evolution of English(es), consider the manifold changes and principled, usually well-motivated developments and reorganizations of the systems of modal verbs or of relativizers in English, or the changing functions of the verb do in the course of time). The real challenge would be to trace such developments of language forms across time and space in greater detail and in line with specific quantitative principles and predictions under precisely determined contextual conditions—but that would require huge amounts of appropriately stratified performance data across time (in addition to a much higher level of mathematical sophistication than I possess). A long-term goal for the field, perhaps, to be worked out by savvy corpus linguists. Until that point, and for the time being, applications will tend to remain somewhat metaphorical by necessity, pointing out the operation of certain principles in rather general terms. But still, as I stated earlier, I find the approach convincing and fascinating. And somehow I think it is important for our discipline to understand the holistic perspective on all of this, including all varieties of English. In that sense World Englishes like, say, Singlish or Kenyan English or Jamacian Patwa are all legitimate descendants of Old English somehow (or of other speech habits further back in time and space, of course), just with (as Sali Mufwene would possibly phrase this) different parameters filtered into a “contact equation,” contributing to the continuously ongoing process of change and adaptation and emergence of English(es) in varying environments and under varying conditions.

These are thoughts that thrill me, and that I’d like to pursue in the future, and perhaps pin down in greater detail, in a less abstract fashion. Alas—too little time can be devoted to that for the time being, with all kinds of duties to be fulfilled and other activities and interests also eating some of my time. But who knows—perhaps there is still another book in me, and if so it will be on this topic.

Well, that would certainly be a highly fascinating and most welcome book! Reflecting on the future, what do you picture for the field of varieties and World Englishes? Would you say that this kind of holistic thinking of languages as complex dynamic systems is the general direction that you would wish for?

As a final note, I know you not only as a colleague but also as a generous and supportive mentor. I’d like to ask you what you, as a mentor, would give as a nutshell piece of advice to the new generation of researchers in the field of varieties and World Englishes?

Let me focus on World Englishes here—the notion of “varieties” encompasses other branches like sociolinguistics, the study of language variation and change, or dialect geography, which have different concerns and orientations and are older, in some respects more established. And in principle—yes, as you’re implying, I believe holistic thinking, recognizing the complex, dynamic, and constantly interacting nature of linguistic evolution is something the field would and will greatly benefit from.

I suppose the topic as such will continue to grow in importance, both within linguistics and from a practical perspective. The unbroken force of globalization, increased travel, and continuously growing international ties, building upon English as the default communicative tool, will keep boosting the transnational attraction associated with English, and will consequently strengthen both the establishment of new local varieties and the process of striving for some degree of competence on the side of millions of individuals. True, there are competitors—strong regional languages like Malay or Spanish, or Chinese, which is increasingly acquired as a foreign language and being pushed; but for the time being I do not see any factor likely to break the predominance of English on a global scale. And we as societies and scholars need to understand what is happening here linguistically and culturally—how and why these new varieties emerge, which properties they have, what people think about this, which symbolic meanings are implied by certain usage forms, what all of this means for intercultural communication, intelligibility, language teaching, and so on. So I trust it will remain a strongly growing field for the foreseeable future. As Crystal (2004) predicted, the globalization of English and the establishment of World Englishes is truly a major component of a “language revolution” which will transform what “English” means, for good.

The field of World Englishes, with journals, conferences, textbooks, handbooks, and research monographs, has been growing exponentially for some thirty-five years, and I think it’s maturing right now, in the sense that paradigms and branches which were established independently of each other are now increasingly coming to interact, to exchange ideas and results, to respect and take notice of each other. Across Asia, and in the United States (to the extent that World Englishes are established there as a field, with a main center at the University of Illinois, where Braj Kachru taught, and building on graduates of that school) the Kachruvian school, building upon his “Three Circles” model but also allowing for sociocultural and literary perspectives, has been strong. Mainly in Europe a more strictly linguistically minded tradition has grown, with work on historical and sociolinguistic usage conditions of varieties of English (by Manfred Görlach, for instance) and, more recently, with many excellent and increasingly sophisticated corpus-based studies (often employing the many sub-corpora of the International Corpus of English [ICE] project or Mark Davies’ huge Global Web-Based English corpus; http://corpus.byu.edu/glowbe/). New Zealand has produced important sociolinguistic work. In Australia, with the linguistics scene strongly characterized by Michael Halliday’s influence, and in many Asian and African countries local concerns are strongly applied in nature, focusing on effective practices and strategies meant to improve the teaching of English. And the last few years have seen the growth of “ELF,” the study of “English as a Lingua Franca,” as a distinct (but related) sub-field. While these have been separate traditions so far, I see practitioners increasingly interacting, attending a wider range of thematic conferences, exchanging methods, integrating various lines of thinking, and I think this is a healthy development, likely to continue.

Within World Englishes in the narrow sense, we have mentioned some of the research trends which are strong right now: diachronic corpus-based studies (just like corpus compilation and corpus analysis in general), and the integration of foreign-language and second-language contexts, plus many individual investigations of specific locations, structures, and contexts based on corpora or fieldwork data. Other trends and topics need to and are beginning to be investigated—the increasing numbers of native speakers of New Englishes, for example; the many mixed and hybrid varieties growing in so many countries (cf. Schneider forthcoming); the “grassroots” acquisition of English by so many speakers globally, for instrumental motivations (cf. Schneider 2016); the many new usage contexts like cyberspace, “transcultural flows” (Pennycook 2007), ESP (“English for Special Purposes”; Schneider 2013b), and so on. A treasure trove of new research opportunities! And I hope we’ll see more creative work transcending established boundaries between research traditions. Just as an example—as we all know, there is an intense debate going on about Singapore’s colloquial variety, “Singlish”; there are many publications on its properties, nature, and status, and there is superb contact-based typological work (most recently Bao 2015). But I know of not a single solid documentation of how regularly, by who and how, and in which contexts, it is actually used—I am convinced a Labovian-style sociolinguistic investigation of Singlish, based on conversational fieldwork and quantitative variationist techniques of analysis, would make a great dissertation topic. And finally, I suspect the future of World Englishes also has a strong sociopolitical component associated with it—a critical perspective, if you so wish, even if I do not consider myself a “critical” linguist in the sense defined by Melchers & Shaw (2011). Indigenized varieties are calling for sociopolitical recognition (also in their respective education systems), and I do think in the long run there is a need for them to be accepted; upholding exonormative standards is mostly just unrealistic, and socially divisive. And that’s the other side of the coin—accepting local varieties of English as appropriate would also contribute to removing class barriers, would serve the needs and concerns of less affluent strata of many societies. Access to English should not constitute a camouflaged social barrier, as it still often does.

Now, as to the mentoring question. Indeed, one of the best parts of the job as a university professor, certainly in my country, is the constant contact with young scholars, the exchange of ideas with them, the opportunity to mentor them and support their careers; and you get so much in return. I am very proud of and happy about the many former students and junior colleagues and friends who have successfully established themselves as academic teachers, researchers, and professors elsewhere. The same holds true now—the bunch of younger, bright and good-spirited researcher colleagues and friends around me in my department are a constant source of joy, a great gift to have. (And I’m also very grateful for another gift they all gave me a while ago, a wonderful festschrift—Buschfeld et al. 2014.)

I’m not sure if I have words of advice which might be valuable on a general scale—personalities, research branches, opportunities, and conditions are so different and highly variable. But here are some things on a meta-level which I find important. Watch out for exciting research topics—exciting to you, that is, that’s all that counts and will drive you. Share with others, go to conferences, which is extremely motivating, and talk to people (including “real” people, not just academics and fellow linguists!). Work to acquire cutting-edge methodological skills—but don’t overestimate this ability and their role; don’t forget that research must be driven by issues, the desire to understand something, not by the availability of some tool. And finally—enjoy what you are doing! I am happy to say I have always done that—and that’s an immense privilege.

I think that is a very inspiring note to end on. Thank you, Edgar, for this interview, for sharing your thoughts and reflections with us, and for your many years of dedicated scholarship. Thank you especially for your openness and generosity towards your friends, colleagues, and students.

Thank you very much, too, Viveka—also for steering me through this series of questions so that it all becomes quite meaningful to me, tying several strings of my research activities together.