Free language! Why we should promote de-Englishization, transdisciplinary linguistic tolerance and greater stylistic freedom, and why we should do so in international business particularly

With the “death” of Latin and the retreat of German and French, English became the lingua franca across close to all scientific communities in the globalized academia of the 20th and 21st century, International Business (IB) not being different. Before that, science flourished in many parts of the world in a wide plethora of different languages, such as Arab, Chinese and Persian. Hence, although we relate our argument particularly to the domain of management and organization studies generally and IB, particularly, what we argue here applies much more broadly. Of course, sharing a common language comes with many advantages. In principle, this stimulates international knowledge sharing, reduces global mobility barriers and promotes worldwide interaction. Hence, one might hope that working with a shared language across the globe will boost scientific progress. However, in reality, we have developed a persistent set of language practices that may well produce the opposite outcomes. Our core argument is that there are two phenomena that strongly interact with Englishization in producing a negative effect on scientific progress: transdisciplinary language policing and stylistic language impoverishment. This is even more counterproductive and damaging in a field so intrinsically global as IB. Clearly, this persistent status quo frustrates the effective response to the call for diversity of perspectives, paradigms and cognitive approaches in IB, as so convincingly and forcefully argued for by Tung (2023) in the field’s leading journal.

Much of scholarly writing is a very boring read, indeed (this tiny piece of quasi-rhyme is an accidental exception). We are not the first to notice this – absolutely not (see e.g. Dane, 2011; Helin, 2015; Meier and Wegener, 2017; Van Maanen, 1995). Many before us across business disciplines broadly made the very same point, arguing that we should adopt a much more playful writing style. Other well-known examples are Weick (1979, 2007) and Boisot and McKelvey (2010). We are expected, if not forced, to remove our own voice from our writing: “We cultivate and teach a writing style of nonstyle that values limited metaphor, simplicity and a formal, if not mathematical precision. Much of our writing is washed by a thick spray of claimed objectivity since artful delights and forms are seen by many, if not most writers (and readers) in the field to interfere with the presentation of what is actually there in a given social world” (Van Maanen, 1995, p. 134). We aim to contribute to this conversation by fundamentally discussing the implications of the way we use language. Moreover, we argue that this is particularly problematic for IB, being an intrinsically global discipline, to which crossing boundaries, not just of countries and languages but also of cultural practices, including modes and genres of language use, is at the core of the field.

This essay is a personal reflection on what we believe goes wrong with our current language practices, based on our own and extensive experience with writing for different audiences within our broad domain of management and organization studies, including IB, as well as across the social sciences and beyond. This is immediately clear from our referencing strategy. As said, we are not the first to argue in favor of creative and playful language freedom in the context of academic work. Nor are we the first to emphasize the critical role of language in IB, which is home to a rich tradition of debates regarding language sensitivities. Clearly, if we would have the ambition to be taken seriously in these academic circles, we could also cite, discuss and reference not only our esteemed colleagues in our own communities who “problematized” language, but preferably also domain-transcending authors such as Deleuze or Kant [in English translation, of course, leading to the classic comedy of “see Kant (2018)”]. However, we will not do so systematically (but still quite extensively) as prescribed in the scholarly cookbook, also because, as we will argue, such referencing is part and parcel of our impoverished language practices, discriminating against the majority of the world’s population.

Below, we develop our argument in eight steps. First, we reflect on the poverty of our language and the main associated downsides. Second, we discuss the role of global Englishization of the academic discourse, including the impact on power asymmetries. Third and fourth, we argue that, on top of Englishization, we have developed a pair of language practices that further degrade the effectiveness of our writing: transdisciplinary language policing and stylistic language impoverishment. We discuss this pair in turn. For a field that is so intrinsically global as IB, this is particularly problematic. Hence, fifth, we continue by arguing that IB should lead by example by opening the gates to the use of multiple languages, transdisciplinary linguistic tolerance and greater and more effective stylistic freedom, as is exemplified by poetry. Sixth, we take a step aside by emphasizing that language freedom comes with language responsibilities. Seventh, by way of intermezzo, we suggest translation in time and over time as a writing quality test. Finally, we conclude by providing a tentative roadmap in terms of practical guidance, proposing a few actionable changes that we, as the IB community, could implement.

In advance, we would like to emphasize that much of what we have to say is not brand new – certainly not. All three cornerstones of our argument – de-Englishization, transdisciplinary linguistic tolerance and greater stylistic freedom – are associated with a rich palette of debates across a wide variety of social sciences. Specifically, in IB, we can benefit from a decades-long discourse on the multifaceted impact of language sensitivities, as triggered by the classic contribution of Marschan-Piekkari et al. (1999). What we add to this large and rich set of literature, though, is twofold. First, we attempt to bring together insights from different strands of literature, from within but also from outside IB. Second, we hope that our reflective and normative intervention will trigger awareness and change, however modest, such that IB as a field will benefit even more from the many advantages of language freedom and plurality. Indeed, as we will argue, IB is not only particularly sensitive to such language-related issues but is also well-positioned to lead by example.

As often is the case in our scientific communities, bad news never comes alone. It is not just that every subcommunity uses its particular scholarly dialect, understood in the broadest sense of language use that denotes particular communities of speakers, an issue to which we return below, but also that the dialects developed in our chosen field are decidedly not equivalent to other mature languages: they are poor cousins. One has to write in ways that minimize surprise, enjoyment, or beauty. We as writers must disappear behind a wall of boring, formalizing and disengaging language, void of any personality. We must use a series of short and grammatically simple sentences to help us through the argument. We should not use the passive mode. The personal “I”, “me” and “mine” are out of order. Do not make jokes. Use the approved structure, which has to be announced properly (and, at the end, a discussion and conclusion should round off the paper). Use your empirical material to test your hypotheses or answer your questions, not just as an incomplete illustration of a new idea. And so on and so forth.

Our prose tends to be impenetrable for non-academic stakeholders, and even for scholars operating at a distance from our niche. And because we are expected to come up with hypotheses in the theory, at least in the context of the dominant hypothetico-deductive tradition (but see Van Witteloostuijn, 2020), we pretend to have developed all our spelled-out predictions in advance, before embarking on our actually inevitably iterative empirical analysis, even when these hypotheses emerged unexpectedly during the empirical analysis journey. Worryingly, we are frequently asked by reviewers to engage in HARKing (= Hypothesizing After Results are Known) – a deadly sin indeed from a methodological perspective. In fact, the reader is treated as a child needing permanent guides, markers of intentions and affiliations and lots of reassurance that nothing needs to be changed in one’s worldview or next week’s classes. To do all this, we oftentimes present whatever we believe we have to present dishonestly, leading to questionable research practices, and we do so without actively and deeply engaging in Open Science practices (Meyer et al., 2017; Beugelsdijk et al., 2020).

Additionally, there is much that we are (not) supposed to write down. Regarding what we must do, an example is symbolic referencing, which is all over the place in what we publish. If we refer to institutional theory, any reviewer will blame us for not citing DiMaggio and Powell (1983) if we fail to do so; in the case of sense-making, referring to Weick (1979) is a must; we cannot discuss issues of international competitive advantages without citing Dunning (1977); and so on and so forth. An example of a “do not dare to do this” is (excessive) self-referencing. To kick against one of these sore legs, one of us engaged in rather exuberant self-referencing whilst writing this reflection. Another example is that up-front attacking established earlier work is banned to the no-go zone. Would we wholeheartedly disagree (which only one of us does, but only part of the time) with Barney’s (1991) famous four sustainable competitive advantage quod non’s, we must first subtly praise his contribution before equally subtly criticizing what we believe might perhaps be wrong. Certainly, we should not say that an esteemed colleague writing in the same journal wrote it up in a way that made it “not only not right, but not even wrong” [a saying, in the original German “nicht nur nicht richtig, es ist nicht einmal falsch”, attributed to Wolfgang Pauli]. As a result, we are forced to write very boringly and dishonestly, using excessive referencing and flattering language in order to avoid rejection.

One reason for this sad state of affairs is the deep and highly persistent Englishization of academia, generally, and management and organization studies, specifically. A powerful body of literature in management and organization studies has produced a stream of arguments regarding Englishization’s very damaging downsides. Here, we cannot help but discuss the tip of the iceberg. A central argument involves issues of Anglophone colonization of global academia and linguistic imperialism (Boussebaa and Brown, 2017; Grey, 2010; Tietze and Dick, 2013), producing forceful power asymmetries along the way (Piekkari and Zander, 2005; Tenzer et al., 2024). Not only do English native speakers benefit from an intrinsic language advantage, but also the power is in the hands of those who master the lingua franca well. This implies “significant language-based inequalities and … demanding quasi-colonial forms of identity work by those being Englishized” (Boussebaa and Tienari, 2021, p. 59). All this is amplified and solidified by a web of US-based institutions such as the field’s leading associations (with the Academy of Management as the dominant norm-setter), (close to all) top journals, HRM practices and accreditation bodies (particularly the AACSB). The result is conformity and homogeneity in tandem with hegemonic US-centric assumptions (Tipton, 2008), of which non-stylish formalistic writing is a prime manifestation.

This speaks to another, and related, critical downside of Englishization: this implies a drastic reduction of the plurality of voices – whether or not the (in)famous Sapir–Whorf hypothesis holds for language in general, it seems very much the case that the characteristics of academic English determine to a substantial degree what can be written about in that language. The hegemony of English and the US standard leads to “conformity of thought and an impoverished intellectual landscape… through the colonizing effects of the English language, specifically its influence in shaping and constructing shared understandings” (Tietze and Dick, 2013, p. 124). Of course, there are exceptions, of which the critical management tradition is a prominent example. But these exceptions are marginalized as heterodox niches, not really influencing the mainstream. Within this mainstream, we all take the same set of very influential US-centric theories as our benchmark, which come with explicit and implicit assumptions rooted in the Anglophone and US culture and which have to be expressed with what the toolkit of the English language has on offer. We do not have an issue with these theories per se (at least, not with all of them), but we do so with their pressing and worrying dominance that suppresses the plurality of voices that we need to really understand the complex social reality of the whole global place we call Earth.

In IB, ever since the seminal contribution of Marschan-Piekkari et al. (1999), we are well aware of the multifaceted and critical role of language, both in how multilingual businesses operate, as well as in how we should carefully reflect on the impact of language plurality in the context of multilingual research. Indeed, by working honestly and transparently with multiple languages, which is standard in IB, we can give voice to multiple perspectives, breaking through the power dominance of English and native English speakers (see e.g. Couper and Piekkari, 2025). In so doing, we can escape from the many downsides of “the unreflexive use of English in academic practices” (Steyaert and Janssens, 2013: 131) by recognizing that language diversity is not a problem, but an opportunity (see e.g. Tietze, 2021).

The observation that English is our lingua franca does at least suggest that we share the same language across our scholarly communities – an impoverished language, regrettably, but still one we share so that global communication is running efficiently and effectively. However, we do not. Our scholarly profession is highly fragmented, with a plethora of (sub)disciplinary niches defending their territory. As we have published across many niches within management and organization studies, from entrepreneurship to strategy and from IB to organizational behavior and beyond (e.g. economics, formal logic, political science, public administration, psychology and sociology), we have substantial experience with language policing. What we have found out the hard way is that each niche uses its own idiosyncratic dialect of English, full of very specific jargon and ditto grammar rules, with the latter being defined broadly, including guidelines as to how the argument is supposed to be structured, what is seen as theory, how much empirical evidence should go into the mix and in which way, what analytic methods we are allowed or supposed to apply, whom, how and how many to reference, et cetera. To protect our (many) guardrails, in each (sub)discipline, we are very efficient in organizing and training an impressive language police force in the form of an army of oftentimes thousands and thousands of editors and reviewers.

Take the example of industrial organization and organizational ecology, by way of nonrandom illustration. One of us traveled between both domains over many decades, including a series of fruitless efforts to build bridges between both in the hope, naïvely and in vain, that these brave attempts would be appreciated by both sides of the divide. Industrial organization (IO) is a branch of economics. To publish there, profit or utility functions must be maximized, with economic agents preferably settling in an equilibrium, with hard mathematical proof (see e.g. Jansen et al., 2007). Organizational ecology (OE) is rooted in sociology. There, the opposition against any form of agentic rationality with an economics flavor is close to dogmatic, emphasizing evolutionary processes instead, with algebra being a no-go (see e.g. Boone et al., 2002). Hence, unsurprisingly, an attempt to develop an integrated IO-OE theory ended up in Google Scholar’s digital archive without any serious follow-up (van Witteloostuijn and Boone, 2006), with a very meager 129 cites (including self-references and that for a piece published by the Academy of Management Review; accessed on January 28, 2026) accumulated over an extensive 20-year period. Today, in the mid-2020s, both IO and OE continue to live their hopefully happy lives in separate silos.

This is one example among many. Basically, scientific communities operate like any other social tribe full of exemplars of homo sapiens, actively developing, promoting and policing an impressive series of identity markers, as we know very well since at least the publication of Kuhn’s (1962) classic, with many of these markers being rooted in subtleties of language use. For an outsider, entry into a tribe like this is anything but an easy endeavor, as mistakes in the accent used immediately signal the outsider’s identity as being of someone not belonging to “us”. Being able to identify outsiders easily and bar their effective entrance into the community can seem to have lots of advantages, especially to two types of actors: those whose position in a wider community is related to their high status in the focal subcommunity and who fear that weakening the boundaries of the subcommunity will dilute that status; and those who rely on the resources of the subcommunity and fear outsiders may capture part of this. These two possibilities positively reinforce each other when the high-status individuals in a subcommunity mobilize worries about sharing resources with outsiders of the lower-status kind. Here, the parallels with modern politics are hard to miss. There is an endless variety of cultural practices that can function as identity signals, from how to dress to how to eat, but language always plays a part – and often a dominant one.

Status hierarchies abound in science. Within the social sciences, for instance, we frequently experience that the field of business is seen by, say, economists and psychologists as inferior. The higher the collective status of a subcommunity within the larger community, the more likely it will be that identity signals will be used to police boundaries. The obvious reason is that the higher-status ones have more to lose and hence more to protect. This can create another self-reinforcing process that further strengthens the negative effects; if higher-status subcommunities do police more stringently, policing stringently will become a signal of higher status (think also of journals proudly mentioning their rejection rates) and subcommunities eager for status (and/or overestimating their present status) will demand ever more precise use of the language “as we do it around here”. Note, importantly, that we are not suggesting that this language policing strategy is consciously developed and maintained by anyone in particular – certainly not. In the daily practice of living in any tribe, much of what the tribe’s members are supposed to do, and hence keep imposing upon others, is developed and maintained subconsiously and through social learning, having evolved over long time spells in many subtle ways (see e.g. Chalub et al., 2006).

Identity markers are difficult and costly to obtain. Learning new languages or even new pronunciations as an adult is very hard work. The career and tenure system in our field helps quite a lot to make the effects of language-based identity signals more pernicious. With the widespread diffusion of North American career and tenure track policies, job entry and career progress in academia are highly dependent upon publication success in peer-reviewed English-language journals, the latter being assessed on the basis of impact factor metrics. This has a wide range of negative consequences (cf. van Witteloostuijn, 2016), including the sine quo non to publish in these top outlets that display all these identity markers. Hence, scholars are trained, both explicitly in PhD courses and implicitly by trial-and-error learning-on-the-job, to internalize the rules of the language police of their chosen subcommunities. If a scholar has obtained tenure, she has been forced to invest heavily in her language skills; and if she does not want to consider these investments sunk costs, she better keep putting them to use, as well as making sure as best as possible that others who do not invest as she did are being kept down. All this is happening all the time, and mostly subconsciously. Hence, to escape from this powerful web of subtle forces, conscious and persistent action is needed.

Moreover, if we succeed in keeping outsider-scholars at bay, we should not be surprised that non-academic stakeholders completely stay away from even browsing through what we have to write in all our prestigious journals. That is, interdisciplinary policing is very likely to translate into a transdisciplinary barrier. Funny enough, we act as if those non-academic stakeholders cannot wait to learn about our groundbreaking insights. After all, what else can be the point of this other language norm of summarizing the implications for practice at the back end of our manuscripts? However, as we write in dialects that make even interdisciplinary boundary-crossing an uphill battle, we cannot help but be sure that the transdisciplinary barriers are difficult to penetrate indeed. Although our prestigious journals enforce language practices that pretend otherwise, we are very well aware of this, as is clear from the many outlets specialized in translating our scholarly abracadabra into prose that we hope readers from the world of practice not only read, but even understand and implement in their daily work ^[1]. Of course, another sign of our self-knowledge about how problematic these translations can be is the ongoing and endless conversation regarding the science-practice bridge or rigor–relevance gap (Tsui and McKiernan, 2022).

To escape from our multifaceted language straitjacket, we have to embrace stylistic freedom. Particularly in the realm of qualitative work, many have argued in favor of such freedom. And indeed, in a niche such as critical management studies (oftentimes inspired by post-modernism), playful language is stimulated. Other examples are Tsoukas and Dooley (2011) and Cornelissen (2025), arguing in favor of an ecological, and a configurational and process style of writing (and theorizing, by implication). Doing so, they seamlessly fit within the tradition of qualitative scholars such as Van Maanen and Weick, arguing that we should work with playful language to capture our complex social world. For instance, “[h]ow organizational worlds are represented in print is not thought to be much of an issue. Writing is seen as a secondary or mop-up activity in our professional pursuits. This is, I think, a mistake and overlooks what might be learned if we were to take the textuality of our organizational theories and facts more seriously: What might we learn if we were to explore the terra incognita of our literary practices?” (Van Maanen, 1995, p. 134). Basically, these and other pleas for a change in the way we write are all motivated by the need to have greater variation and freedom in respect to our language use to be able to capture the complexities of real (organizational) life much better than we can if we remain in our customary lingual straitjacket (would we or at least our writing otherwise become a danger to others?).

Without doubt, as far as we are concerned, we should recapture the freedom, richness and power of our language – or rather, our languages, plural. And we should do so throughout, and not only in our community’s qualitative and constructivist niches. Language is what makes us the only animal that can hope to increase its expected (literary) longevity by submitting papers to top journals. Making language poorer does not just diminish our ability to communicate effectively; even more importantly, this drastically reduces our chances to understand social reality. After all, much reality is captured and constructed by language. Maybe mathematics does underlie the reality of physics [see Eugene Wigner’s (1960) “The unreasonable effectiveness of mathematics in the natural sciences”], but social reality is largely structured by our language, so natural language should be the instrument of choice to make sense of this very reality. The Sapir–Whorf hypothesis is everywhere, in whatever form or shape. In IB, this language effect has another layer of complexity, not only because different languages are used across the world but also within internationally operating enterprises, generating a plethora of foreign language effects (see e.g. Akkermans et al., 2010; Haans and van Witteloostuijn, 2024; Tenzer et al., 2017).

If one wants to apply mathematics, we are supposed to use an explicitly circumscribed formal language in which each concept has a specified and very precise meaning. This brings huge advantages, because doing so allows us to come up with proofs that also use such a language and thereby provide a wonderful mode of control. Formal languages can be very useful in describing social phenomena, especially when a formal framework can suggest foundational general relations (Conradie et al., 2021), as well as when the formal framework fits with the empirical methodologies – from statistics or, increasingly, from computer science (Bosma and van Witteloostuijn, 2024). But if one does not do mathematics (or formal logic), as is largely the case in IB, then we have to take care that natural language does what natural language does best, as a toolbox to speak or write (and listen and read) about inherently complex and essentially fuzzy phenomena using inherently complex and essentially fuzzy concepts in inherently complex and essentially fuzzy structures. Actually, we cannot do without natural language in mathematical pieces either, if we aspire to communicate the intuition behind the battery of formulas, proofs and symbols to a broader audience of non-mathematically trained scholars.

Releasing the full potential and power of our writing requires promoting stylistic freedom and hence stylistic variety. To increase the plurality of voices in academia, we have to encourage the use of multiple languages. And whilst doing so, we can and should escape from the straitjacket of our formalistic, voiceless and non-stylish style of academic writing. As said, with this lament, we join a choir with many co-singers. A few telling examples should suffice here. Dane (2011) argues in favor of writing as sounds and tunes, Lehtonen and Putkonen (2023) promote the use of graphic novels in the context of multimodality, and Kiriakos and Tienari (2018) suggest understanding writing as love, all three in the spirit of a plea for writing differently in academia. The key is that we should embrace stylistic freedom, from fiction and journalism to dialogue and poetry (cf. Czarniawska, 2009). What this implies exactly depends critically upon the nature and topic of the scholarly work at hand.

For instance, Meier and Wegener (2017) develop their argument on the basis of, and with reference to, a reflexive exchange of arguments and ideas in the form of a chain of personal letters, as reproduced across their published text. Interestingly, in line with earlier proposals of prominent management and organization scholars such as Van Maanen (1995), March (2006) and Weick (2007), a flourishing stream of scholarly literature applies and promotes what is referred to as poetic or social inquiry (see e.g. Cunliffe, 2002; Cloud and Faulkner, 2019; Galvin and Pendergast, 2016; Hopkinson, 2015; Prendergast, 2015). As Geers (2024, p. 3) explains, “[i]n poetic inquiry, as in practice-based research poetry, poetry functions both as a process (methodology), in a way data is collected and created during the writing process, and product (result), in the final poem that is produced” (emphases in original). Whatever the style of writing, the key is the freedom to adopt a personal writing voice, however formal or literary, that fits with the aim of the study at hand, and that communicates the obtained insights with regard to complex reality effectively ^[2].

As said, the issue of language should be even more important in IB, particularly, compared to business generally, as a topical lens for research because of the evident reason that operating globally cannot but involve the use of multiple languages (cf. Brannen and Mughan, 2017). Above, we briefly referred to the example of the foreign language effect. Another example is the popularity of the language or linguistic distance construct (see e.g. Beugelsdijk et al., 2018; Cuypers et al., 2015). The importance of the multiplicity of languages is so intrinsically relevant from an IB perspective that we assume we can rest our case in this respect. Indeed, in IB, a rich discourse evolved regarding the many subtleties associated with the need to contextualize translation (see e.g. Piekkari et al., 2020; Westney et al., 2022; Wilmot and Tietze, 2023). Additionally, we would agree with earlier observations that the importance of at least two other highly influential and complementary reasons should not be underestimated, one relating to the social construction of reality through language and the other to the very international composition of the scholarly IB community.

Language, and hence social reality, is rooted in and does produce meaning and structure. We do not need to consult the work of the likes of Chomsky, Deleuze, Foucault or Wittgenstein to be aware of this. In this context, again, the classic Sapir‐Whorf hypothesis ^[3] is a very well-known benchmark (Kay and Kempton, 1984). A very simple stylized fact is telling here: probably all languages have words without any real equivalent in English. For instance, Dutch has the word zwijger (“the silent”). The 16^th century “father of the Netherlands”, the French-speaking German nobleman leading Holland’s revolt against the Spanish ruler of the Low Countries, even having received the nickname Willem de Zwijger. In English, there is no “silent” equivalent of the noun speaker (spreker, in Dutch). What does this tell us about how Dutch versus English-speaking people view the world and act accordingly? Differences in grammar matter, too. How does lack of a past tense impact the associated culture’s take on the importance of future orientation? Chen (2013), for instance, reports evidence as to the impact of this linguistic feature upon a population’s average saving behavior.

In IB, these language effects must be taken extremely seriously, as countries and their businesses and inhabitants will behave differently in different contexts depending upon the culture and perspective rooted in their language. This implies that we oftentimes should study IB issues through the frame of local languages, using these local languages both as our analytical material and as our writing workhorse (see e.g. Couper and Piekkari, 2025, for a recent reflection). For instance, a process case study of post-acquisition outcomes of a German enterprise’s takeover of a Dutch firm cannot really identify the essence of this process if both sides of the acquisition are not studied in their mother tongue, both as the language used in the respective part of the case study, as well as by studying differences in the use, meaning and impact of both languages (see e.g. Monin et al., 2013). In terms of research practices, this implies that we have to be very careful in our use of language, systematically reporting and registering what we do regarding all languages used. Practices such as back translation and survey language equivalence testing are just two rather superficial examples out of many (Zander et al., 2011). Careful translation requires contextualization, often implying that literal translations are off mark (Chidlow et al., 2014). Within multilingual research teams, this means that much reflexivity is needed to develop good working practices that facilitate mutual understanding (Thomas et al., 2009).

Another issue involves the supposedly global nature of the IB research agenda. Notwithstanding the ambition to be really global, IB as a field is not. Much of the research agenda is still highly US-centric, as is the network of scholarship (see e.g. Haans and van Witteloostuijn, 2019). It is not that the IB community is not investing in this – quite the contrary. Both IB associations and IB journals are very well aware of this issue and do much to make the field, in terms of both scholars and topics, truly globally inclusive. However, with English as the overly dominant lingua franca, the community hits a ceiling here. Native English speakers can benefit from a fundamental “competitive advantage” and many US-centric practices and values are dictated in detail by our leading journals, of which many originate from and are firmly rooted in the USA, as to how we are supposed to conduct research and how we are expected to write in our publications.

Here, the impact of modern generative AI tools may be argued to give a boost to creating linguistic equity across the globe, as the quality of the translation algorithms is (or soon will be) so high that authors can write their original drafts in their mother tongue to subsequently feed this better-written text into a translation tool, with an equally better-written English text as the result. However, while this may reduce the competitive advantage of native English speakers, the AI tools will be trained on the existing literature. Moreover, competent prompts (an activity that still advantages the native English speakers, especially those best-integrated in their niche) will make the generative AI algorithm pay more attention to the writing in the target niche, thereby strengthening the many (tacit) strictures as to how to write proper journal paper English in a particular niche.

Hence, freeing language use will have to involve much more than just passively accepting support from current AI translation tools. Luckily, modern-day journals have more than enough space in the cloud to store and give access to articles in the native tongue of the authors as well, which will enrich future training data substantially. The bottom line is that the use of multiple languages should be common practice in IB across the board. Much original research has to be conducted in the language of the study’s setting to really identify underlying mechanisms, although the use of multiple or other languages within a single study should always be carefully accounted for. By taking language even more seriously than we already do, the IB field will become truly international in terms of fields of study, of the spectrum of topics, of the contextualization of findings and of the internationality of the scholarly community. And importantly in the context of this essay, our use of language will become much richer, and hence less boring and more creative.

But of course, with language freedom, and the associated language power, comes language responsibility. After all, a well-functioning scientific community requires effective communication. And we have to be very aware of the fact that natural language imprecision is unavoidable. Imprecision is an inherent characteristic of any natural language. So, again, our argument must be that language is key. Specifically, with more language freedom, we are even more obliged to safeguard clarity and precision, just like poetry combines freedom and power. In the same sense, we speak of power in statistical analysis – the extent to which the results can be meaningful. And in dreams begin responsibilities, as William Butler Yeats wrote (and after him Delmore Schwartz). So, what should be done, as Vladimir Lenin put down in writing, plagiarizing Nikolai Chernyshevsky?

We are not proposing to replace all modes of academic writing in our scientific communities by poetry or any other literary form, although we do argue that poetry and other literary genres should be accepted as completely valid modes of communicating scientific insights (see e.g. van Witteloostuijn, 2026). What we propose is that journals and conferences, beginning with the most highly ranked, should declare and practice a much greater tolerance for differences in modes and styles of writing, as well as in respect to the broader aspects of what constitutes good writing, like how and when to refer to earlier work; that while the advantages of using English as a “shared” language should be conserved, there must be greater hospitality ^[4] to papers in other languages in our (top) journals, as well as of translations (submitting a great article originally published in Chinese, now translated into English, should not lead to an instant desk rejection); and tenure decisions should also take work in other languages into full account. We briefly return to these issues below.

And through the lens of IB, there is an important point: We should take the impact of language extremely seriously in everything we do. Language differences impact the very issues and processes we as IB scholars are interested in. It is not that we are not aware of this – we are. And it is not that we do not study the impact of language – we do. However, we argue here that we can and should do much more here and much more fundamentally so. We have to turn the inevitable fact that we have to work in and with so many languages across the globe into our advantage. Indeed, language plurality is much more an opportunity than a problem. We must do this in terms of how we conduct research in our multiple-language field settings, of how we develop better and better multiple-language research practices, of how we apply AI-based translation tools in our writing, reporting and publishing and so on and so forth. Here, intrinsically, we can expect IB to lead by example within the broader community of management and organization disciplines. Luckily, much excellent groundwork has already been done in the language-sensitive IB tradition, as referenced quite extensively throughout this essay.

So, our plea is to maximize the clarity and precision of our language, irrespective of the stylistic writing mode. Here, as we will explain below, the translation method provides a very effective instrument of control. Yes, we should be free to use the language and the style that we regard as providing the best fit with the slice of complex reality that we try to capture. But no, we should not do so in a way so idiosyncratic and sloppy that close to no one beyond ourselves can really understand what we want to say to the rest of the scientific community. This is even more important if we seek to stimulate cross-disciplinary work, of whatever type – inter, multi or trans. Our definitions should be as sharp as possible; it should be as easy as possible to determine whether our arguments are right or wrong. Our writing should be as simple as the reality we write about allows for, which means that, yes indeed, we sometimes will write in ways that are more difficult to read, but only if we must, not as a signal that we belong to a particular subcommunity or to artificially try to boost our own status. If we use our language creatively and freely but well, we can start to capture a much wider spectrum of our complex reality, particularly if this comes with the demolition of inter-niche mobility barriers across our many scientific tribes. Free language and the rest will follow.

As implied by the above, all this does not and cannot imply that anything goes. Our language must have a few qualities to pass the tests of scholarly appropriateness and effectiveness. We would like to propose one particular way of testing regarding how well language is used, a particular way that rather well coheres with our overall argument about the negative reinforcement embrace of Englishization, excessive transdisciplinary language policing and impoverishment of language. It is the test of contextualized translation, in the widest possible sense of the word – i.e., over time in the same language and in time in another language. This particular test was also suggested by the experience of one of us in poetry. (And yes, one of us is actually much better known as a poet than as an academic, having received prestigious awards and prizes – hence, happily and proudly boasting all the quality signals equivalent to a H-score or i-index among researchers.)

In doing so, we seamlessly jump on an interesting IB bandwagon that reflects on issues relating to translation. The key message of this literature is that the usual mechanical view on translation – as, for example, reflected in the standard practice of back translation – is off mark. We do not need to add yet another reference to the Sapir-Whorf hypotheses to understand that translation involves much more than looking for equivalent wording. Contextualization is critical. After all, translating texts into any other language inevitably implies that “[o]riginal meanings are inadvertently lost (and new meanings created […])” (Boussebaa and Tienari, 2021, p. 63). As Couper and Piekkari (2025) argue convincingly, IB should be prudent in avoiding such mechanical translation practices because such a lack of contextualization and reflexivity is inextricably bound up with (a) loss of rich nuances embedded in multilingual data and (b) many important and insightful voices being silenced ^[5]. Hence, if the IB community does not adopt language-sensitive (translation) practices, knowledge production will be seriously hampered. This is not only true for research practices but also for organizational practices that move across language boundaries (see e.g. Piekkari et al., 2020).

Moreover, languages change all the time. This has the maybe underappreciated effect of making any text written before today, to some extent, a text in another language that we consciously or unconsciously translate into our modern day’s own time-bound dialect. The richer the language, the more our speaking and writing can work with multiple layers of meaning and multidimensional relations between concepts and hence the more the passage of time can disturb or destroy these structures and relations. Poetry puts this richness to maximal use, which can make poems extremely fragile with respect to the passage of time; at the same time, the continuing relevance of a poem of the past becomes a test of quality. There are words in Homer that we do not know the meaning of, and there are lots of jokes and maybe nods that we are missing. Homer wrote about a world that had disappeared long before his own time, but much of what he wrote continues to help us understand our modern-day world. Contemporary poets, across the globe, still reflect on his work and respond to what they believe he has to say in his poetry. So, equivalently, the test of the passage of time and continued use by other and later researchers (do not throw away your time-evolving H-scores) should still be an important signal of quality, also of our use of language.

The second criterion involves another survival test: How well does the original text’s quality stand the test of translation in another language? Here, we are lucky that we ourselves are not native speakers of English. This has the great side-effect that we, as multilinguals, are probably more sensitive to language differences than most of those mastering only their mother tongue. Without any doubt, we both are much better writers in Dutch than in English. Interestingly, seriously investing time in translating the original text into another language is instrumental to identifying where the original wording needs to be improved to (a) increase the quality of the original text and (b) ensure that the meaning in the original text translates well to the other language. Returning to the example of poetry. The most important poetry is the poetry that people keep trying to translate in other languages. It is a mistake to think that poetry is what is lost in translation; strong poems are the ones that survive translations, even though one can never capture exactly the same structures and exactly the same relations of meaning as in the original (but, as we argued above, that is also the case when one reads a poem, as any other text, of the past in the original language).

Moreover, translation is almost as wonderful an instrument of testing and controlling the quality of the original text as proofs are in mathematics (or formal logic, which is essentially a branch of mathematics). If you try to translate a poem you have written or if you are discussing attempts to do so with a translator, you have to be willing to continually answer questions about what your words really mean and what you really want to say by making particular language choices. This can make translation a working tool to write better, also in your native language. By improving the original text in your mother tongue on the basis of these experiences, this original writing becomes clearer – a better read that is more likely to survive the passage of time. Hence, both aspects are strongly related. Translation is a very effective method to control and test the quality of any text. Luckily, with the arrival of excellent deep AI translation algorithms, we can now more easily conduct translation and back translation checks, even if these are just first – mechanical – steps in the process.

The above implies a great advantage for IB scholars. Not only are many of us bilingual (at least), but oftentimes do we also collaborate in the context of multi-language teams. Hence, by translating and writing our key output in multiple languages, we can substantially improve the quality of our writing. With the arrival of high-quality AI-based translation software, all this is much easier to do compared to the old non-AI days. Here, however, two critical comments are worth emphasizing. First, as we know from insightful work in cross-cultural management (see e.g. Chidlow et al., 2014), literal translations are very likely to be off mark. Any translation must be contextualized, accounting for differences across languages and the associated cultures. Hence, although AI tools will be extremely helpful, a human check and manual finetuning are inevitable. Second, current AI tools are trained on unknown training data – that is, unknown to us, as users (we may hope the developers do know what they are doing). Any biases in the training data will reemerge in the AI tools’ output (Tenzer et al., 2014). So, key is to enrich the training data maximally, by adding as much cross-cultural, cross-language and cross-style material as can be reasonably managed by the AI tool developers and their algorithms. In this context, we as the IB community can be instrumental in this enriching process by feeding our AI-based translation software with our pluriform material.

A hopefully insightful analysis of the multifaceted language problem – or rather, opportunity – is one thing, but coming up with practical solutions is quite another matter. After all, we are far from the first to put the finger on the sore spot (this is a literal AI translation of de vinger op de zere plek leggen), but apparently, we still need a lament as ours in 2026. Part of the action requires institutional change and collective effort. After all, a system of dominance is very hard, if not impossible, to change by one or a few individuals. Regarding institutional change, the dominant Anglophone career and tenure track system, in combination with the battery of Anglophone HRM practices in the spirit of Anglophone New Public Management, with an emphasis on Anglophone career competition and performance measurement, has to be radically reformed. This has been discussed elsewhere by many, also through the lens of the downsides of Englishization (see e.g. Boussebaa and Brown, 2017). Additionally, in tandem, a grassroots movement such as Responsible Research in Business and Management (Tsui, 2021) can further speed up the process of systemic change.

We believe the time is ripe to seriously invest in implementing change, for at least two reasons. The first reason is cultural or institutional. In many countries around the world, opposition against the US-inspired Englishization and HRM system in academia is on the rise. For one, many countries seek to reduce the number of foreign students in their English-language programs, with many outside the Anglophone world doing so by promoting studies in their native languages. Moreover, and related, in many countries, the highly competitive and research-biased tenure track system is increasingly unpopular, being replaced by a plural career path policy set that is less competitive, more inclusive and more open to career plurality. On top of this, Trumpism’s recent, very aggressive and forceful attack on universities will undermine US dominance in academia. All this implies that in many countries, the Zeitgeist is supportive of the kind of changes that we propose in the current essay.

The second reason is technological. Above, we already reflected upon the use of AI-based algorithms in the context of translation exercises aimed at increasing the quality of our writing. Many non-native speakers of English are much better at the type of creative and playful writing that we seek to promote in this essay in their mother tongue. Hence, we hope that using such AI-based translation software by first starting to write creatively and playfully in our mother tongues will impact our English text favorably ^[6]. Paradoxically, this may imply that the current and persistent power asymmetry between native speakers of English versus their academic colleagues mastering non-English mother tongues will slowly be eroded. After all, over time, with more and more material being published that is deeply influenced by the pluriform cultural and linguistic background of the authors, both the AI tools and the academic literature will be more and more penetrated by insights from other cultures, originating and written in other languages. Actually, slowly, academic English may change as a result as well.

Given the above, the IB community provides the ideal seedbed for change. After all, the IB field is intrinsically interested in and aware of issues related to cross-cultural and multi-lingual contextualization. To close the circle of our argumentation, we therefore would like to provide three (sets of) examples of actionable solutions – one at the level of an individual, two at that of an institution and three at that of a journal. At the personal level, we hope that many readers of this essay are inspired to start engaging with stylistic variety in an attempt to better capture the contextual richness under study, including the translation exercise suggested above. The more of us who start doing this, the more we will feed the AI translation algorithms with contextually rich multi-lingual material and the higher the likelihood that the writing norms in our community will start to change in the direction argued for in this essay. Of course, for those who never tried to write like this, progressing along a learning curve is inevitable. To speed up this learning process, attending a creative writing course may well be instrumental.

At the institutional level, not only must HRM practices move away from the US-centric career and tenure track system emphasizing and promoting competition, as discussed above, but also should universities and their business schools adapt their study and training programs accordingly. For instance, our advice is to make creative writing (including poetic inquiry) a mandatory course in all BSc, MSc and PhD programs. More broadly, the type of transformative change that we suggest here can only succeed through a collective institutional effort, which will inevitably take time to evolve. Training novices only will not suffice to crack the stronghold of senior power. For example, as a collective, we must train reviewers (and editors) to evaluate unorthodox writing styles honestly, without stubbornly falling back on our usual orthodox set of boring and formalistic standards. Learning to be tolerant and to accept freedom might look good on paper, but it is likely to turn into an uphill battle on the ground.

At the journal level, we suggest changing language policies. Without any pretension to be exhaustive, we would like to propose three such policy changes. The first suggestion is to introduce stylistic experimentation corners in our journals (see the Academy of Management Discoveries for examples). These corners are dedicated to thought-provoking and insightful contributions that experiment with the stylistic mode of writing. The second suggestion involves multiple-language publications. With so much online capacity, our journals can publish, say, at least one article per issue with a reference to the online version in the cloud written in the original non-English mother tongue. In the English version, a short section can be added that explains the nature of the differences between the two versions due to translation difficulties. The third suggestion regards investment in AI-based translation tools. Our journals can provide access to and advice regarding AI-based translation software that is most adequate in the translation into English of specific (families of) languages. With measures like this, we may hope that our IB community will slowly develop much better language practices over the next decade or so, specifically if we as scholars and our universities invest in launching flanking measures along the way. Then finally, to reiterate Van Maanen’s words from 1995 (!), “perhaps, […] we can loosen up some of the writer’s cramps that seem so prevalent in our field” (p. 139). Free language and the rest will follow! (Table 1)