Data collection is now underway for a new corpus of ‘Dutch English’ within the broad scope of World Englishes. This news is often met with suspicion from ELT practitioners, SLA researchers and the average person on the street, Dutch and English L1s alike. How could a Dunglish-style interlanguage arising from ‘imperfect learning’ be cast as legitimate regional variation? Yet this has been a fruitful field for many decades across Asia and Africa, and researchers in Europe are starting to follow suit (see e.g. Erling, Reference Erling2004; Erling & Bartlett, Reference Erling and Bartlett2006 for the case of Germany). With English being used for intra-national purposes on the continent all the more frequently, especially in higher education, it is not hard to find examples of regionally flavoured English being more appropriate than any native ‘norm’.

A case in point is the word beamer. Usually used in colloquial standard English to refer to a BMW, it is also used by the Dutch and others to mean ‘projector’. In this case it is a literal translation from the Dutch word beamer – and the transfer is hardly surprising, as it looks and sounds like an English word, follows the usual rules of English morphology, and makes logical sense. As such, it appears to have become the accepted term, both in the Netherlands and elsewhere in Europe. This can give rise to confusion when used with interlocutors unfamiliar with the ‘new’ usage, as demonstrated in the following passages from a forum thread on the Canadian-based website Digital Home.Footnote ¹

Here we have a new forum user (Zerogate) using the unconventional sense of beamer, which clearly baffles the forum moderator (who, with his/her almost 40,000 posts, is apparently the chief authority on the forum). The moderator is based in Toronto, as shown in the profile details. Zerogate's location is not shown, but we are given a clue in the line ‘in german language’ above.

Some time later, a new player joins the conversation, also located in Canada.

The moderator then rejoins the thread, apparently having now understood Zerogate's initial question. S/he posts an answer, without making mention of either beamer or projector. In the subsequent response, however, Zerogate now refers to ‘projector’ – having apparently ‘corrected’ his/her ‘error’ for the benefit of the interlocutors.

Several days later, another person joins the dialogue, again from Canada. This response seems to show a measure of accommodation by (at least initially) using both the standard and new forms:

The final post in the thread comes from another Canadian-based interlocutor, and seems to give off somewhat mixed messages. In the first line below, s/he shares an in-joke with the other NSs, implicitly addressing only them by explicitly referring to Zerogate in the third person (‘he’). Given that the post is, of course, also visible to Zerogate (where s/he is in effect being treated as the ‘other’), this comes off as somewhat exclusionary. Yet the second line acts to soften the blow, suggesting an awareness that learning and using other languages could be a worthy endeavour rather than a ‘deviant’ condition.

The issue of intelligibility is at stake here, and many mainstream SLA researchers will readily assert that in the exchange above, the NNS's use of a nonstandard form (i.e. Zerogate's ‘imperfect’ lexical proficiency) led to a breakdown in communication. Yet this line of argument is brought into question when we extrapolate to the situation in the Netherlands.

Until recently I worked as an English NS editor within a Dutch university, and found my daily practice plagued with what Erling (Reference Erling2002: 8) aptly termed a ‘moral quandary’: how to reconcile on the shop floor the conflicting personas of sociolinguist versus ELT practitioner (or in my case, editor). To ‘correct’ recurring and perfectly intelligible (but nonstandard) features, or to retain them as legitimate varietal features that simply add local flavour? Or, in the case at hand:

1. (a) to ‘correct’ the word beamer to the more usual English term, projector, based on the quasi-ethical reasoning that editors are employed to do precisely such things; that is, to impart their NS ‘authority’ on ‘deficient’ NNS writing (‘just doing my job’)?

2. (b) to allow the word to stand, taking into account the (predominantly) Dutch readership and the fact that beamer will pose no intelligibility threat to Dutch readers, whereas projector might?

More to the point, if empirical research showed projector to be a difficult (i.e. lesser-known) word for the Dutch, but beamer to be easily understood by everyone else in the environment concerned, should editors then still persist in replacing it? After all, why can't the Dutch have their beamer, when Canadians can have their tuques and loonies and Australians their dunnies and tucker?

The corpus

To turn to the ‘Dutch English’ corpus currently being compiled at the University of Cambridge, another quandary exists in how to position the corpus within the prevailing WEs-meets-corpus-linguistics paradigm. For reasons of comparability, the corpus replicates the design of the written sections of the International Corpus of English (ICE), thus 200 texts of approximately 2000 words each. As in ICE, these texts are divided among the categories academic writing, non-academic writing, instructional writing, creative writing, press writing and correspondence. A ‘Dutch person’ is defined as anyone who has lived in the Netherlands since before the age of 10, and who has not spent more than 10 years or over half their lifetime abroad (see e.g. Holmes, Reference Holmes and Greenbaum1996, for ICE-NZ). As in ICE, too, the contributors ought to be ‘educated’ speakers of the relevant variety, in order to find more reliable evidence of a settled linguistic system rather than an interlanguage. For this reason, a HAVO diplomaFootnote ² has been agreed on as the minimum education criterion, as this will exclude those who work in a trade, for example, but include anyone who could potentially produce a text deemed suitable for one of the corpus categories. Naturally, the final key criterion is that the texts have not been professionally translated or edited by an NS editor. All contributors are required to fill out a questionnaire and a consent form, so that the corpus can be searched along demographic parameters and ultimately made publicly available.

Users or learners?

The difficulty arises in that, when basing the design on that of ICE, inevitable implications arise. ICE is expressly only for ENL and ESL varieties of English (national corpora within the ICE framework have so far been released for Great Britain, Canada, India, Singapore, and so on). On the project website,Footnote ³ researchers interested in examining the English of countries where English is not the native or an official language are directed to join the ICLE (International Corpus of Learner English) project. This is ICE's sister corpus, led by Sylviane Granger at the Université Catholique de Louvain. To date, ICLE corpora have been released for China, Brazil, Bulgaria, France and numerous other countries, including the Netherlands (developed by researchers at Radboud University Nijmegen).

As its name suggests, however, ICLE is explicitly profiled as a ‘learner’ corpus, and as such only includes undergraduate essays. In contrast, the present Dutch English corpus is being composed on the hypothesis that it may show evidence not merely of learner language but of potentially stable characteristics. It is not meant to be an overstretch of the original initiative, but merely a new and exploratory way of approaching a WEs corpus. As indicated above, it is being composed predominantly of academic and professional texts from well-educated contributors with long exposure to English in a country where it is not formally a second language, but arguably is functionally, and which is repeatedly referred to in the WEs literature as being in transition from the expanding (EFL) to the outer (ESL) circle. Thus, it would seem to be better profiled as a ‘user’ rather than a ‘learner’ corpus, and as such does not sit well within the ICLE paradigm.

Various factors point to this notion of Dutch academics and professionals as potentially being users rather than learners. For example, Björkman (Reference Björkman2008: 36) points out that one is only a ‘learner’ when English is the object of study. Identity, too, is a key factor in shaping a new variety (Brutt-Griffler, Reference Brutt-Griffler2004). Ridder (Reference Ridder1995) reports that the Flemish historian Sophie De Schaapdrijver once described the Dutch use of English with people who attempt to communicate with them in their own language as a form of ‘repressive tolerance’ (‘My English is always going to be better than your Dutch’). This may point to a growing sense of Dutch identity in English: their collective competence in English (and other languages, incidentally) is seen as a source of national pride.

Another key factor in varietal development is intra-national use of the language (see e.g. Mollin, Reference Mollin2006), and here the change is well underway in the Netherlands: master's programmes are increasingly offered only in English, and even in Dutch-language undergraduate programmes, lectures and reading materials are frequently offered in English. Competence in English is a prerequisite for almost any career, from secretary to scientist. English is also pervasive in popular culture: Dutch talk show hosts, for instance, regularly interview foreign guests in English and such programmes are rarely (if ever) subtitled. Indeed, without reasonable competence in English (ranging from more to less active depending on career path, for example), one would have difficulty functioning as a full-fledged and unhindered citizen in Dutch society. Thus, while English may not have the official status that it often holds in other ESL (and ICE) countries, such as Kenya and the Philippines, with nearly 90% of Dutch people claiming to be able to converse in English (Eurobarometer, 2006), its de facto position in the Netherlands is unchallenged.

Agency

Another good indicator of this status as ‘user’ versus ‘learner’ is a certain agency in manipulating English purposefully to one's own environment and ends. At the southern Dutch university where I used to work, Dutch colleagues writing in English sometimes intentionally opted for a ‘nonstandard’ form if they felt it better suited the situation. For example, the 24-hour clock was preferred over the use of am and pm; the staff felt that this would be more readily recognisable by their students (mostly Dutch and German L1s), and this consideration was deemed to outweigh the fact that in standard English the 24-hour clock remains mainly the domain of soldiers and pilots (Burrough-Boenisch, Reference Burrough-Boenisch2004: 58). Another example was the deliberate transfer of the double-barrelled honorific Prof. Dr. X. In standard English, only the highest title is used. But in Dutch, all qualifications are made visible (e.g. Prof. dr. ir. X), and thus many Dutch writers will purposefully opt for the ‘nonstandard’ formulation rather than strip someone of an earned qualification.

Another experience well known to NS translators and editors in the Netherlands is that of being ‘corrected’ not infrequently by Dutch clients. In the late 1990s, Joy Burrough-Boenisch published a pair of articles in the newsletter of the Society of Native-English-Speaking Editors in the Netherlands, entitled ‘L2s correcting L1s’ and ‘More on L2s correcting L1s’. In these articles she described her own experiences and those of other members of the society with being corrected by Dutch clients. In 2010, I conducted a similar, very informal survey of members of the same society, to get a sense of the situation 10 years on. One wrote:

Another had been asked to translate the phrase Let maar niet op de troep, which is said by the owner of a house when receiving a guest:

The two dozen or so responses I received ranged from the mildly amused to the downright infuriated. A number pointed out that this phenomenon is partly the result of the generally perceived English-language competence of the Dutch population as a whole, which gives rise to the confidence to challenge the purported authority of the NS. Some also saw it as a form of cultural transfer of apparently typically ‘Dutch’ characteristics such as assertiveness and directness. In any event, Dutch agency in English should certainly be examined in more thorough and systematic studies.

Conclusion: a starting point

The positioning of the Dutch English corpus as a full-fledged national component of the International Corpus of English would contribute to the discussion on the nature of nativised varieties, the distinction between error and feature and so on. Data collection is around half-way through, but well before the analysis phase starts many issues still need to be resolved. The aim is to look for empirical evidence of certain grammatical and lexical properties that

• – have been impressionistically reported as characteristic of DuE

• – are reported as common features of all New Englishes

• – emerge during the POS-tagging process.

More precise hypotheses are currently being formulated regarding the grammatical and lexical properties that we may expect to find. There are technical issues to address, too, which will include the choice of word classes and annotation software, and how to deal with ‘deviant’ forms in the tagging process. Not to mention the many issues that will arise in the analysis stage: Can the findings be explained by the phenomena of L1 transfer or universal/angloversal processes (e.g. of regularisation)? With respect to any novel characteristics, is there evidence of stability and systematicity in the vein of Mollin (Reference Mollin2006) or Kirkpatrick (Reference Kirkpatrick2007), or are we looking at merely a random collection of learner errors? In the event that certain systematic characteristics can be provisionally catalogued, it is hoped that this will provide a basis for the development of intelligibility and attitude studies, to be conducted among international stakeholders, but most especially among the Dutch themselves. After all, as Mollin (Reference Mollin2006: 159) put it, ‘unless a variety is accepted, it is not a variety’.