Building ‘Applied Linguistic Historiography’: Rationale, Scope, and Methods

In this article I argue for the establishment of ‘Applied Linguistic Historiography’ (ALH), that is, a new domain of enquiry within applied linguistics involving a rigorous, scholarly, and self-reflexive approach to historical research. Considering issues of rationale, scope, and methods in turn, I provide reasons why ALH is needed and argue that, while it can borrow from Linguistic Historiography, it should also distinguish itself, for example, by paying more attention to histories of practice as well as ideas, with corresponding methodological emphases and challenges. Making specific reference to the histories of applied linguistics and of language learning and teaching, I identify ways in which theories, theory–practice links, and practices themselves can be investi-gated historically in a more rigorous and ultimately useful manner. Overall, I show that innovation to establish this new domain of enquiry in applied linguistics involves reflection on research methods but also on more fundamental concerns.


INTRODUCTION
I shall attempt here to write into existence a field which is not yet recognized as such-that of 'historical research within applied linguistics' or, as I shall term it, 'Applied Linguistic Historiography' (ALH). In doing so, I provide an example of innovation in nascent form, showing that innovation to establish a new domain of enquiry inevitably involves reflection on the research methods to be used, but also that the nature of research methodology depends partly on initial clarification of the rationale for research in that area and of its scope. The first step in establishing a new domain of enquiry like this one might need to be 'talking it up', in other words, providing a rationale for its coming-into-being (Section 2). Then, the scope of the new field needs delimiting (Section 3)-there is a need for clarification of what is being talked 'about'-and in the case of ALH, this is not at all an easy task (Section 3). Only then can methods-innovative or otherwise-be properly considered (Section 4).
In arguing for the establishment of ALH, I shall derive support from previous, pioneering work in the area of 'Linguistic Historiography' (that is, principled historical research in and into the field of linguistics, or of 'linguistic ideas'). However, I also argue that, while the scope and methods of the two fields will to some extent be comparable, they need to be distinguished, for example, in the attention ALH needs to pay to practices as well as ideas, with concomitant methodological emphases and challenges.
My argument is advanced here mainly with reference to the histories of applied linguistics itself and of (second) language learning and teaching. However, the argument is also relevant to other areas of practice with which applied linguistics (henceforth, 'AL') is concerned-speech therapy, machine translation, language policy, language assessment, and so on. While these areas have witnessed some historical research, substantial bodies of work have not so far emerged except in relation to first-language education (largely within the field of History of Education) and lexicography and language standardization (both strongly related to History of Linguistics).

The current state of play
Within AL, historical research has been, and still is, a pursuit without a pedigree, without an obvious mandate, and without commonly recognized or followed methodological procedures. This contrasts with the relatively high status history has gained in adjacent fields, for example in education and linguistics, and with the methodological reflections which have been pursued there.
For the History of Education, a number of journals and parent learned societies have been established, for example, History of Education (British History of Education Society), History of Education Quarterly (United States History of Education Society), and Paedagogica Historica (International Standing Conference in the History of Education, based in Continental Europe). The field of History of Linguistics (HoL) began to develop as far back as the late 1960s and had become a recognized area of scholarship by the late 1970s. Its milestones have included the launch of Historiographia Linguistica (1974), the First International Conference on the History of the Language Sciences (1978), the foundation of the Henry Sweet Society for the History of Linguistic Ideas (1984), and the launch of its journal, Language and History, in 2009. Historical research is very far from being comparably established within AL, although there have been sporadic attempts to highlight its importance. Among early, relatively abortive efforts can be counted Stern's (1983: 44) call for history to be placed alongside linguistics, sociology, psychology, and educational theory as a foundation for the theorization of language teaching. His arguments for more research have not, though, been followed up in any kind of consistent fashion. Colloquia on the History of Applied Linguistics (henceforth, 'HoAL') were arranged at the annual meeting of the Henry Sweet Society in Edinburgh in 2000 (see Smith 2000) and at the 2001 AAAL (American Association for Applied Linguistics) convention in St Louis. Both Angelis (2001) and McNamara (2001) wrote up their contributions to the latter colloquium, but neither event was built on in a consolidated way.
More recently, however, there have been some promising signs of revival: Andrew Linn, Professor in the History of Linguistics at the University of Sheffield, organized a symposium on HoAL at the annual BAAL (British Association for Applied Linguistics) meeting in 2012, having co-edited a themed issue of Histoire-Epistémologie-Langage, also on HoAL (Linn et al. 2011). Subsequently, a themed issue of Language and History on 'Building the History of Language Learning and Teaching (HoLLT)' was co-edited by McLelland and Smith (2014) and an international conference on HoLLT was organized, also in 2014, at the University of Nottingham (McLelland and Smith forthcoming). Within the International Association of Applied Linguistics (AILA), a Research Network on the History of Language Learning and Teaching ('HoLLT.net', for short) was established at the beginning of 2015, while two further very recent indications of a rise in interest in history within AL also deserve mention: first, Carter and McCarthy's (2015) paper on spoken grammar in the present journal, which puts the case-via reference to sources going back to Tudor times-that 'applied linguistics is too heavily absorbed in the present' (p. 14); and, secondly, de Bot's (2015) interviewbased monograph on recent developments in AL. My argument here will, I hope, help to consolidate these achievements, moving HoAL and HoLLT further into the applied linguistic mainstream.

A preliminary argument
As can be seen from the way I have been referring to HoLLT as well as HoAL above, in this article I shall advance the view that establishing the importance of historical research within AL is not the same thing as establishing (only) the importance of researching HoAL per se: the histories of areas of practice with which applied linguists are concerned, including but of course not restricted only to HoLLT, are equally involved. This view has various ramifications, as we shall see. First, it provides a preliminary justification for my emphasis throughout this article on 'Applied Linguistic Historiography' rather than on the 'History of Applied Linguistics'. I intend the former to have a broader reference than the latter, in particular where the scope or 'range' of the field of enquiry is concerned, as will be further explained in Section 3.
A second, more strategic reason for coining and using the term 'Applied Linguistic Historiography' is that my argument can thereby at once draw sustenance from previous, foundational arguments in favour of establishing what was-and still generally is-termed 'Linguistic Historiography' for the field of HoL and can distinguish itself from these arguments, asserting the difference R. SMITH 73 and autonomy of the new domain of enquiry (to the same extent that AL is itself distinct from Linguistics). In what follows, then, I discuss, in turn, the rationale, the scope, and some possible methodological procedures for ALH by borrowing arguments made 40 years ago for Linguistic Historiography, but complementing this, throughout, with indications of additional, specifically applied linguistic dimensions.

A 'paucity of studies'
E.F.K. (Konrad) Koerner, formerly Professor of General Linguistics at the University of Ottawa, has been the best-known advocate of the establishment of what he termed 'Linguistic Historiography'. His first line of argument (e.g. in Koerner 1978) concerned at once the lack of quantity of existing studies and the idea that 'a scientific field reaches its maturity only by becoming aware of its history and by taking a serious interest in having it documented' (as repeated in Koerner 1995: 119). Both sides of this argument can be borrowed where HoAL is concerned. It is now almost 70 years since the phrase 'applied linguistics' was first officially used, in 1948, in the sub-title of the new, Michigan-based journal Language Learning, and 60 since its first UK manifestation (in 1957, when the School of Applied Linguistics was founded in Edinburgh). There has still, though, been relatively little research published into HoAL. The same holds true, although to a lesser extent, where the history of language teaching is concerned. More than 30 years ago, Stern (1983: 76) lamented a 'paucity of studies' and-while there are pockets of substantial work in Continental Europe, especially in Germany, and in relation to the history of teaching French (see Besse 2014)-original research, particularly in Anglophone countries and into the history of English language teaching, has continued to be rather sparse overall [see McLelland and Smith (2014) for relevant overviews].
Recent stirrings of greater interest in both HoAL and HoLLT are perhaps just nowadays indicating a certain 'coming of age' in the fields of AL and language didactics, respectively. As when Koerner felt compelled to make his argument for Linguistic Historiography 40 years ago, ALH now needs to be better promoted, and more widely undertaken, recognized, and reflected upon. An increase in original research on a sounder footing really would indicate that AL has begun to 'reach its maturity'.

Some weaknesses in existing research
Linguistic Historiography, for Koerner (e.g. 1995: 3), refers not just to a field of research (HoL) but to a scholarly, rigorous approach to research in that field: a 'principled manner of writing the history of the study of language' (my emphasis), which 'naturally includes the discussion of questions of methodology and epistemology', as distinct from 'History of linguistics', that is, the 'actual recording of [. . .] linguistic research through the ages' (ibid.). Indeed, the major thrust of Koerner's argument 40 years ago for the establishment of 'Linguistic Historiography' was not just that there was not enough research but that historical work already occurring was of insufficient quality, often being methodologically uninformed, over-partisan, anachronistic, and overliteral.
A similar argument can be made to justify the establishment of ALH here. I will briefly indicate some ways in which existing work can be critiqued for weaknesses in method or stance, while in Section 3, following an initial consideration of reasons for historical research, I consider further weaknesses, relating to anachronism, lack of contextualization, and over-literalism.
First, it can be demonstrated that an over-reliance on previous secondary accounts or, worse, referring simply to hearsay or handed-down mythologyis endemic in many existing historical narratives, and that original, primary evidence-based research has been relatively rare. As critiqued recently by Hunter and Smith (2012) and Howatt and Smith (2014), this is exemplified in the way typical 'potted' accounts of language teaching methods tend to feed off one another, continually reproducing a kind of decontextualized 'mythology' which bears little relationship to attested, concrete, contextualized realities. Such accounts tend to stereotype and demonize the past in a 'progressivist' manner, serving to assert the supposed superiority of current conceptions (see also Stern 1983: 77).
Even relatively scholarly, though short, accounts of recent developments in AL (e.g. Davies 1999; Grabe 2010; Kaplan 2010) tend to be heavily reliant for the period prior to the 1950s on existing secondary accounts (especially Howatt 1984), and not particularly assiduous in their citing of primary sources for the more recent period. Two exceptions are the well-focused studies by Mitchell (1997) and, more recently, de Bot (2015), a book-length treatment based on a series of interviews with leading protagonists in the very recent history of AL.
The influential monographs by Phillipson (1992) and Pennycook (1994Pennycook ( , 1998) deserve a more detailed appreciation and critique than is possible here but, in brief, they can be seen, on the one hand, to have given a boost to the idea that history has value, by referring to primary sources and opening up the field to new historical perspectives. However, in writing history, there are ever-present dangers of selecting facts to 'make an argument', and neither of these authors--in their evident desire to bring about change in current conceptions-succeeds in avoiding what Koerner (1995: 5) terms 'propagandistic' history, in other words, writing which selects historical facts to support a thesis. Their arguments that AL and ELT have been political, not neutral, activities have been persuasive and, to many, welcome; however, they both cite rather selectively for the purposes of argument, offering insufficient context for sources. Their assertions about the colonial and neo-colonial roots of current activity therefore need to be balanced with more 'historiographical' treatments.

Potential uses of history
So far I have followed Koerner in providing some arguments from a relatively 'academic' perspective for a quantitative increase and a qualitative improvement in historical research within AL. However, such arguments may be insufficient on their own for ALH to become better established. Perhaps, for this to occur, its usefulness within this applied field requires particular emphasis. That is, historians within AL-more, perhaps, than historians of linguisticsmight need to be relatively brash in highlighting the current, practical relevance of their work.
In the field of language teaching, for example, it can be argued that historical evidence is needed as a basis on which to build appropriate reform efforts (cf. Smith with Imura 2004), or, more generally, that developing 'historical sense' is an important aspect of language teacher education (Smith 2013). Historical research into applied linguistic antecedents can place present-day conceptions of AL activity in perspective and reveal their historically constituted limitations (Smith 2009(Smith , 2011). An argument which might need to be further elaborated, then, is that HoLLT and HoAL can both be engaged in for their potential practical relevance, not just for their intrinsic academic value.
There is potentially a contradiction here, of course. If history is to establish itself as part of AL, in other words to prove itself as useful in relation to current applied linguistic concerns, then it could be argued that 'abuses' of the past will continue to be committed in the service of current positions. This could involve propagandistic history or milder forms of 'presentism'-viewing the past through the prism of the present-such as over-literalism or anachronism (to be discussed in Section 3). One response to this might be that better reference to primary sources will enable assertions to be more easily falsified and/ or countered, thus guarding against extreme forms of presentism. Also, it is advisable to be self-reflexive-as conscious and explicit as possible, in other words, not only about one's evidence base but also about the extent to which a particular standpoint on current issues is influencing the focus, procedures, and presentation of one's historical research.
Discussions of epistemology within the general field of history are relevant in this connection, of course [for useful overviews see Berger et al. (2010), Green and Troup (1999), and-more explicitly attempting to bridge the gap between theory and historical practice-Gunn and Faire (2012) and Jordanova (2000)]. While it is clear from these accounts that history can never be completely impartial, this is far from the same thing as saying that 'anything goes'. Indeed, there is a strong 'craft' tradition among professionally trained historians- Thomas (2010), for example, offers an honest and entertaining atheoretical account of '[immersing] myself in the past until I know it well enough for my judgment of what is or is not representative to seem acceptable without undue epistemological debate' (p. 37). As has been the case for Linguistic Historiography, the priority for ALH, I would suggest, is to find a sure footing within this existing source-driven, 'practical' tradition. In other words, while keeping an eye open for possible 'lessons of history', historians of HoAL and HoLLT should at least aim in the first instance at 'objective history', defined by Evans (2000: 272) as 'history that is researched and written within the limits placed on the historical imagination by the facts of history and the sources which reveal them, and bound by the historian's desire to produce a true, fair, and adequate account of the subject under consideration'.
3. THE SCOPE OF APPLIED LINGUISTIC HISTORIOGRAPHY Following Koerner, and as explained above, I am deliberately using the term 'Applied Linguistic Historiography' with an ambiguous meaning, both to designate the practice of historical research into applied linguistic areas and the adoption of a rigorous, scholarly, and self-reflexive approach to such research which makes it worthy of being seen as a form of AL in its own right. In Section 4, I explore further what this 'scholarly approach' might involve in relation to methods. Here, though, I first attempt to demonstrate rigour in relation to necessary definitions of scope, exploring what might be meant by 'research into applied linguistic areas' in the above definition.
There are two main issues to be considered here. The first, corresponding with one of Koerner's major further criticisms of pre-existing work in HoL, concerns the need to avoid both anachronism and over-literalism. The other issue-not considered by Koerner since he was not operating in an applied arena-concerns the need for the remit of ALH to extend beyond ideas, theories, and research to links between these and language-related practices, and the desirability for such practices also to be researched in their own right, not just as an adjunct to the history of ideas.

Avoiding anachronism and over-literalism
As Stern (1983: 95) points out, some accounts 'impose modern conceptualizations on historical developments and [thereby] oversimplify the underlying theories'. Anachronism of this kind, as critiqued also by Koerner (e.g. 1978), is often tied to the desire to find antecedents or counter-examples to bolster a new or current conception, accompanying presentism, progressivism, or propagandism, in other words. A relatively scholarly approach to history, in contrast, involves attempting to see past phenomena in the context of their own times, viewing them, as far as possible, on their own terms. How this might be achieved methodologically, via immersion in both primary and secondary sources (cf. Thomas, cited above), will be considered further in Section 4.
The form of anachronism I wish to mainly discuss here, relating to the overall scope of ALH, is over-literalism-that is, being concerned only with past phenomena which bear the exact name of a current focus of concern. This extends (but is by no means limited) to how we conceptualize the overall field to be researched. Koerner proposed use of the label 'history of linguistic ideas' R. SMITH 77 rather than 'history of linguistics' to ensure coverage of phenomena from previous ages up until the term 'linguistics' started to be widely used, in the 20th century. Similarly, there are many phenomena from the past which could be viewed as relevant to HoAL (the Tudor Vulgaria considered by Carter and McCarthy 2015 being a case in point) but which were naturally not-in their own time-labelled as forms of 'applied linguistics'. If we took a completely literal view, such phenomena would be anachronistically excluded from the remit of HoAL, to deleterious effect.
This has, in fact, been quite common in recent brief treatments of the history of AL. Thus, Catford (1998) notes uses of the phrase 'applied linguistics' prior to 1948 by Baudouin de Courtenay (in Russian) and an associate of C.K. Ogden (in English) and in this over-literal manner identifies Courtenay and Ogden as significant precursors. However, many writers before 1948, including more influential figures than those mentioned by Catford, engaged in reflections and activities which we might term 'applied linguistic' in nature today.
In order to move beyond over-literalism and thereby expand the scope of HoAL, we do, though, need a current conception of what 'counts' as AL which will take us back to when the term was not used. This is where the task becomes harder than in the case of HoL, due to the fact that the meaning of 'applied linguistics' has been so disputed.
Thus, if the original (1948, Michigan / Language Learning) 'linguistics applied' conception is adopted, late-19th-century Reform Movement figures like Henry Sweet who were concerned to apply the techniques and findings of phonetics to language teaching, must be seen as important applied linguists avant la lettre (cf. Howatt and Smith 2002;Linn 2008). Indeed, the actual influence of the pan-European Reform Movement on the involvement in wartime language teaching by linguists like Leonard Bloomfield, and thence on the early development of AL in the USA, would repay further study.
There is also a need, though, to move beyond viewing the past through the particular post-war prism of 'linguistics' being the primary source of, for example, language teaching theory, and it is at this point, I would suggest, that ALH begins to differ and liberate itself from purely Linguistic Historiography. In crude terms, as Davies (1999: 12) has indicated, a distinction needs to be made 'between two fundamental views [. . .], that of ''linguistics applied'' and that of ''applied linguistics.'' The first starts with theory, the second with practice'. Thus, a less linguistics-driven, more interdisciplinary, and problem-oriented conception can be identified, whereby the applied linguist is seen as a mediator between practice and a variety of possible source disciplines. Adopting this second, interdisciplinary view of sources for theory enables us to identify precursors who might be otherwise ignored if we have a 'linguistics applied' conception, for example, Claude Marcel, who mined mid-19th-century educational science in a substantial two-volume work on the theory of language teaching (Smith 2009), or, in the 17th century, Jan Amos Komensky (Comenius) (Caravolas 1984;Smith and McLelland 2014).
Of course, the applied linguist is no longer seen these days simply as a consumer of theories or descriptions but as a producer of relevant theories on the basis of systematic investigation. In relation to language teaching, the growth of Second Language Acquisition (SLA) as a field of applied linguistic research perhaps best epitomizes this third conception. Viewing AL as an autonomous discipline rather than one tied to background disciplines which may or may not have been identifiable in the past, in fact, allows us to view many theorists as applied linguists avant la lettre. Indeed, Thomas (2004) has made a study of the history of SLA theories which goes much further back than the Reform Movement, while an interesting, practice-oriented alternative to a 'linguistics applied' conception of theory-practice links was provided in the early 20th century by Harold E. Palmer-another figure whose work has been almost completely neglected within over-literal previous accounts of the development of AL (though see Smith 2011; Carter and McCarthy 2015).

Concerns with context and history of practice
Needing to make my own conception explicit here, I view AL as essentially practice-oriented, autonomous from linguistics, interdisciplinary in its inspiration, and a field of activity (involving active attempts to bridge theory and practice), not just an intellectual arena. From this standpoint, I wish to argue now that, where ALH is concerned, there is a need to complement Koerner's emphasis on the history of linguistic ideas with a focus on relatively practical, applied aspects. As we shall see, this will bring with it a need to develop research approaches for accessing histories of practice in ways which go beyond purely Linguistic Historiography. It is, then, to the historiographical implications of this overall conception of AL that I now turn.
First, it should be noted that most existing accounts of applied linguistic and language teaching history emphasize the development of theories in the abstract rather than paying much attention to practice (including contexts for the production of theories), and the false impression can thereby be conveyed that history has everywhere been the same. Despite the undoubted difficulties of accessing data in this area, attempts should at least be made to situate ideas in particular contexts of practice (asking 'what gave rise to them?') and to ascertain their impact on practice, for example, on policies, on learning materials, or (most difficult, admittedly, to establish) on teaching and learning activities in particular settings. Smith (1998), Howatt and Smith (2002: Vol. 2), and Linn (2011) all offer examples of this sort of history.
Practice can also be looked at historically in its own right, of course, not just as part of HoAL; indeed several free-standing learned societies exist for HoLLT research in countries including Portugal (APHELLE), Italy (CIRSIL), Japan (HISET), and Spain (SEHEL), and, internationally, for the history of French language teaching (SIHFLES). Despite coming together recently within HoLLT.net, a new Research Network of AILA, the members of these associations are likely to have their own conceptions of historical research which do R. SMITH 79 not necessarily make reference to AL at all, and which may be influenced more by perspectives from the humanities than from social science. As Berthet (2011) indicates, there has, indeed, been a decline in the importance of AL in France, alongside establishment of the field of 'didactique des langues étrangères', which may make historians of French language teaching, in particular, question the value of the notion of Applied Linguistic Historiography argued for in the present article. Nevertheless, in contexts where AL does remain the major academic domain within which language-related practices are theorized, and where AL constitutes the main arena for explicit discussion of research methodology, I suggest that it is appropriate to consider ALH as having a coverage which extends to past language-related practices (e.g. HoLLT), even in cases where such practices are or were not obviously linked to AL or to equivalents of 'applied linguistic theorizing' (avant la lettre).

METHODS OF APPLIED LINGUISTIC HISTORIOGRAPHY
As we have seen, one symptom of-and perhaps continuing reason for-the general lack of interest in HoLLT and HoAL within AL is that few arguments have been previously presented in their favour, but to this might be added the observation that there has not been much discussion of appropriate historical methods either [apart from a few general comments in Stern (1983), the only directly relevant source I am aware of is Ruisz et al. (forthcoming)]. Careful consideration of methodological procedures-whether in themselves innovative or not-needs to be at the heart of any attempt to establish a new field of study. Thus, as Koerner and others [e.g. Schmitter (2003) and Hü llen (2005)] have done for Linguistic Historiography, there is a need to propose possible methods for ALH if only to show research students and others a 'way in' to the field, and to provide a means for evaluating work that is done.

Fundamental general precepts
I have emphasized that a major failing of many existing studies which adopt a historical perspective on language teaching or AL is that they tend to be overreliant on, and insufficiently critical of, existing secondary sources and that they involve no, little, or only very selective reference to primary sources. We can begin this consideration of methods, then, with some basic principles of historical research which relate to these existing insufficiencies and which can be considered domain-independent-that is, which are borrowed from the general field of history. Building on Stern's (1983: 87-88) rather mild advice in this area, the following can, therefore, be proposed as an initial set of 'precepts' for ALH: Be critical of existing secondary accounts. Refer to primary sources (do 'original' research). Attempt to make only evidence-based assertions. Explicitly state sources of evidence.
There is a need to compare and contrast existing secondary accounts, aiming both to correct inaccuracies in and considerably 'fill out' these accounts and assessments, but historical research should also, almost as a sine qua non, make reference to primary sources. What kinds of primary source evidence are to be consulted, then, and how are primary sources to be sought out, evaluated, and utilized? This is where the above considerations of the specific scope of ALH come firmly into play.

Primary sources
Relevant sources will depend on the nature of the ideas or practices being investigated, and on the rationale for and intended scope of the research. Continuing to focus on HoLLT for illustrative purposes, I shall provide some examples here of types of possible primary source (only some of which were mentioned by Stern 1983: 87-88), indicating how they might correspond to different kinds of focus and then exemplifying how some of them might be referred to, via some further 'precepts' for ALH: History of ideas: Theoretical and polemical writings-treatises; prefaces of, and advertisements for textbooks; professional journal articles (late 19th century onwards); lecture notes; unpublished writings. Contextualization of ideas: Written memoirs or reports, published or unpublished; oral accounts; university course descriptions; letters; photographs; newspaper advertisements, articles, cuttings. Policy and curriculum: Government papers; policy documents; curricula; syllabus documents. Learning materials: Manuals/textbooks for learning languages; pedagogic grammars and dictionaries; audio-visual materials (20th century). Teaching and learning practices: Reports of public commissions; published or unpublished memoirs; eyewitness accounts and observation reports; teachers' reports on their own practice; project evaluation reports; minutes/ reports of exam boards; minutes of school staff and Local Education Authority meetings; videos of classroom events (late 20th century onwards); lesson plans; exercise books; marginalia in textbooks. The above lists, it should be emphasized, are illustrative only and by no means exhaustive; also, there will be considerable overlap across the somewhat arbitrarily established headings representing different types of focus. I hope, nevertheless, that the lists provide an indication of how ALH can be opened up to new kinds of evidence. In particular, they show some types of unpublished source which have hardly ever been referred to in our field [even Howatt's authoritative A History of English Language Teaching (1984) made little reference to archival or unpublished sources].

Accessing practice
The above typology of sources may be useful, specifically, in the two areas which, I have suggested, could characterize ALH but which are not usually given coverage-that is, contextualization of ideas/research/theories in relation to contemporaneous practice, and assessments of the impact of ideas on practice. Admittedly, it is far easier to find statements of theory than representations of practice, but the search can pay dividends-for example, much more is now known about the reasons why H.E. Palmer developed his 'Science of Linguistic Pedagogy'-a clear early-20th-century precursor of AL-due to discoveries of newspaper advertisements for schools he ran in Belgium and university calendars detailing the classes he gave in London (Smith 1999), while the continuing impact of his ideas on at least some teachers in Japan can be established with reference to videos of demonstration lessons given at recent conferences of the Institute he founded in Tokyo in 1923.
As I have argued, there is currently a surfeit of attempted histories of ideas as opposed to practices in the domains of both HoAL and HoLLT. From the perspective of AL being a problem-or practice-oriented discipline, we should at least attempt to situate ideas in contexts of practice. In addition to the general historiographical precepts suggested earlier, then, I suggest the following three, which are targeted relatively specifically at the field of ALH: Situate ideas in contexts of practice. Attempt to ascertain the impact of ideas on practice.
[Also] chart developments in practice on their own terms.
Regarding the last of these, quite a common pursuit within the learned societies for HoLLT referred to earlier has, in fact, been to analyse learning materials, their authors, and their contexts of production, often on their own terms and with little reference to 'applied linguistic' theory. A challenge from ALH to this kind of pursuit might be for better assessments to be carried out regarding the wider significance of the materials under consideration, for example, by relating them to applied linguistic ideas, on the one hand, and to contexts of actual use-to ascertain impact-on the other.

Final general precepts
Finally, we return to some further important principles from the wider field of historical research, but ordered according to what I believe are priorities for ALH, and which are geared, then, to its current state of (under-)development: Scope available sources/seek out further relevant sources. Immerse yourself in primary sources, and secondary sources beyond AL. Triangulate by comparing and contrasting different sources.
Rather than foregrounding techniques of principled selection and source criticism in the manner of Stern (1983: 77) and Ruisz et al. (forthcoming), I place emphasis here on the preliminary basic need to access, indeed immerse oneself in (a variety of) primary sources-but also secondary sources from other fields relating to the history of the context and period under investigation. In the first instance it is desirable to 'scope' what sources might be available, taking previous secondary sources as a starting point, perhaps, but noting inconsistencies between them or gaps in their accounts. When textbook analysis is to be engaged in, it is important to establish what the entire corpus of materials available might consist of (e.g. Klippel 1994), while in order to understand the work of a particular author it is equally important to establish a comprehensive bibliography (e.g. Smith 1999Smith , 2009. Details need to be checked, wherever possible, against copies of the writings themselves, and sources indicated when this is not possible. Beyond published sources, however, consideration can also be given to identifying and arranging visits to key archives, as appropriate to research questions, and establishing-for relatively recent, person-centred history-whether family members can be contacted, since they can be a particularly rich source of primary source data (letters, photographs, etc.). All of these are likely to be iterative processes, however, since one difficulty is that it is not always possible to know in advance what will be useful, nor what is available. The need for immersion emphasized here, and for continually asking oneself questions which require further data, contradicts, to some extent, existing strictures about the need for 'systematic selection', or at least-I would argue-immersion is an important prior step that should be undertaken before narrowing down of focus and selection occur.
Here, then, are two final precepts from the field of history 'proper': Select in a principled way from available sources. Be aware of the limitations of primary sources.
As Stern (1983: 77) notes, in existing work we need but rarely find a 'discussion of the reasons for the selection of the events, books or names'. Providing a reason for one's focus seems particularly important when a corpus (e.g. of past learning materials) is being analysed. Also, it should be recognized, primary sources need to be evaluated and not taken at face value, and there is a whole tradition of historical training in 'source criticism', particularly in German-speaking countries, which is of relevance but can only be touched on here. Ruisz et al. (forthcoming) provide further useful information about how insights from this tradition might specifically apply in the fields of HoLLT and HoAL. I agree with these authors that it is important for any historian to develop a critical awareness of the biases of accounts, the fallibility of human sources, and the genuineness or otherwise of archival documents. However, just to encourage any use of primary-in particular, unpublished-sources at all is probably the main task facing ALH at present, and, in the absence of specific training, experience of triangulation via immersion in data will often contribute a lot to developing the critical awareness that is needed. There are also, of course, practical introductions available in the general field of historical research methods (e.g. Howell and Prevenier 2001;Brundage 2013), which can provide further useful advice for novice applied linguistic historians. The main thrust of my argument has been that the research approach of ALH will necessarily be something new within AL, since HoAL, HoLLT, and-it would appear-other forms of history of language-related practice are currently only in a relatively embryonic state. This is not to say, of course, that the methods of ALH will themselves be particularly 'new', being largely imported from the domain of history, and comparable, in many respects, to those employed in the contiguous, better-established field of Linguistic Historiography. I have, though, placed particular emphasis on some aspects of source and method which distinguish ALH from Linguistic Historiography, specifically in the former's focus on practice as well as theory.
There is even a possibility that ALH could develop some truly 'applied linguistic' research methods of its own, which could then be useful within the wider field of history. This has not been the focus of the present article, but in recognition of the fact that ALH could be taken to mean 'historical research using applied linguistic techniques', I present below a speculative list of research methods ALH could develop and potentially export to the broader field of history: techniques of (critical) discourse analysis applied diachronically (e.g. to track change in representations of national identity, race, religion, gender, sexuality, class, and so on in textbooks or policy documents); uses of corpus linguistics-for example, keyword analysis (see below); application of techniques from forensic linguistics (to 'source criticism'); introduction of an applied linguistic awareness of the nature of spoken discourse within oral history methodology (beyond thematic analysis as in de Bot 2015).
As just one example of original research innovation, I shall focus on the work of Duncan Hunter, whose PhD thesis (Hunter 2009) reported on a keyword analysis of past issues of ELT Journal (see also Hunter and Smith 2012). Hunter followed Stern's recommendations for methodological rigour and built up a corpus of texts according to a systematic procedure of selection which was carefully described and recorded. By means of a computer-based keyword analysis procedure using WordSmith Tools (Scott 2004), the potential for human bias in selection of 'key concepts' for further investigation was mitigated. In this manner, Hunter attempted to avoid the over-emphasis on theory and the distorting effects of bias he had identified in existing accounts of the development of communicative language teaching. Overall, by adopting a rigorous, clearly 'applied linguistic' corpus-based approach, Hunter both shed new light on the changing discourse of ELT and developed procedures which could be applied to the study of the history of any discipline. ALH is a good example of the opening of a new sub-field bringing with it the need to find appropriate ways of exploring hitherto neglected areas, even if the methods themselves are not particularly innovative outside AL. At the same time, approaches can emerge or develop in new ways because of technological advances, adding to existing methods in ways which have not been attempted before. As we have seen, corpus-based keyword analysis has already been utilized-indeed, corpus linguistics and discourse analysis might be particularly promising sources for contributions that ALH could make to other historical domains. Rather than becoming fixated on any one methodological approach, however, we should remember that history needs to be imaginatively reconstructed 'in the round', ideally from multiple kinds of source, and from multiple perspectives.

CONCLUSION
In arguing here for the establishment of 'Applied Linguistic Historiography', I have suggested that more historical research is required, and that it needs to be methodologically better-informed, involving avoidance of over-literalism, presentism, and propagandism via immersion in and careful reference to primary as well as secondary sources. I have also argued that there are particular needs, which distinguish ALH from Linguistic Historiography, for looking not only at applied linguistic theories but also at related practices. Thus, sources will include representations of practice, where possible, and not just material relating to the history of ideas. Indeed, to the extent that a focus can be maintained on practice as well as on theories, and to the extent that practical benefits of historical research can be emphasized without prejudice to scholarship, HoAL and HoLLT can justify their independence from HoL and assert their worth within contemporary AL. A final possibility for ALH has simply been touched upon: the prospect of applied linguistic tools serving historical research more generally. In 19th-century Germany, after all, the 'science' of history (and specifically the notion of source criticism) was established on the basis of methods then in use by philologists (Evans 2000: 8). Evans (ibid.: 9), for one, has explicitly invited applications of linguistic analysis to contemporary mainstream history, and this could happen if a more diachronic perspective is established within our own field.
In relation to the wider concerns of this themed issue, I have shown that innovation in research methods does not simply involve new tools or techniques. When a new field is in the course of establishment, issues of rationale and scope need to be considered alongside methods. I have also shown that methods already established in other areas can be imported; thus, appropriation and adaptation should be seen as forms of innovation in their own right. In the case under consideration here, methods have, of course, been appropriated particularly from the wider field of history, and a major challenge is how applied linguists can acquire the skills possessed by historians-how general historical research methods can be taught and learned, in other words. At the same time, I have shown how an existing approach (Linguistic Historiography) needs to be adapted to different, more practice-focused ends, R. SMITH 85 and have ended by proposing some ways in which-once the innovation which is ALH itself (as domain and overall approach) has become sufficiently established-it might proceed to develop its own, more specifically applied linguistic historical research methods into the future.