The moment one starts thinking of language as discourse, the entire landscape changes, usually forever.
- McCarthy & Carter (2014) Language as Discourse: Perspectives for Language
Introduction
Developed countries have seen increased imbalances between the supply of and demand for health workers. Demographic factors such as population aging and the rise of life expectancy have created increasing job opportunities in the sector. According to The European Centre for the Development of Vocational Training, in the EU alone, employment in the sector grew by 14.3% between 2006 and 2018 and has been projected to grow by a further 5% over the period 2018 to 2030, an increase of more than 300,000 new jobs (CEDEFOP, 2019). In order to provide for the vacancies and replace healthcare workers who will leave the occupation for one reason or another by 2030, almost 4 million job openings will need to be filled (CEDEFOP, 2019).
Abundant employment opportunities have intensified the international mobility of doctors and nurses and instigated a continuous battle to attract health professionals by the healthcare providers (Mara, 2020). The COVID-19 pandemic reminded everyone with unprecedented urgency about the essentiality of the role that health workers play in modern society and made the examination of the factors that influence their international mobility particularly topical. This includes recruitment procedures, which the literature argues are crucial for productive employment (Andresen, 2015; Reiche et al., 2019). As a result of the complexity of the operational environment (Mayo et al., 2021) and the specific tasks and responsibilities that the provision of healthcare services requires, when hiring, healthcare organisations have to apply particularly strict requirements to test the preparedness of the applicants to fulfil their duties. Under these circumstances, the choice of selection criteria acquires particular importance and has profound implications for both the hiring organisation and the career of the applicant.
In this article, we investigate the challenges that surface when language proficiency is tested as a part of the process of hiring overseas healthcare workers. Our analytical focus is on the challenges that the available options used to demonstrate the required competence pose to the recruiting organisation. We consider the case of the UK National Health Service (NHS), possibly the largest publicly funded national healthcare system in the world. Over many years, the NHS has been experiencing staff shortages. Consequently, its growth has been accompanied by an increase in the number of medical professionals coming from abroad: in 2018, 144,074 of the NHS employees (12.7% of all staff whose nationality was known) were classified as non-British nationals (Baker, 2018). In the foreseeable future, the NHS will continue facing a shortage of medical staff in all categories (Iacobucci, 2017), accentuating the need for research on the bottlenecks that impede the absorption of skilled workers from abroad. In this respect, this article offers an original contribution to the debate by addressing a facet that is rarely in the spotlight: the effects and implications of testing foreign recruits’ language proficiency. Although interest in the language aspects of global professional mobility has seen a surge, the extant literature tends to have a narrow scope. Typically, the linguistic aspects are investigated in the context of multilingual environments (Vulchanov, 2020), most often in multinational firms, and concentrate on communication barriers caused by linguistic diversity (Ahmad & Barner-Rasmussen, 2019; Harzing & Pudelko, 2013) and the enablers of communication and knowledge flows across language barriers (Ciuk et al., 2018; Kuznetsov & Kuznetsova, 2014; 2016).
The focus of this article is different. It examines a situation in which an organisation seeks to maintain a certain standard of just one working language used by the multinational and multicultural staff for whom this is not a native language. To complicate things, the workers’ professional duties require them to partake in multiple discourses, which, using the classification proposed by Linell (1998), may be labelled as intraprofessional (discourse within specific professions), interprofessional (discourse between representatives of different professions at workplaces, in meetings, public debate etc.) and professional-lay discourse. This situation is typical for healthcare professionals and raises questions whether standard language proficiency testing is a reliable screening tool in organisations in which expatriate workers are expected to perform in an array of social, cultural and physical contexts.
Conceptually, we base the scrutiny of the effects of language testing on the premise of the contextual complexity of organisational practices (Ashkanasy et al., 2011; Schein, 2010). Further, following Heracleous and Barrett (2001), Taylor and Robichaud (2004), Thomas et al. (2011), and more recently Gunnarsson et al. (2014) and Kong (2014), we problematise the notion of ‘profession’ as a discursive construct. Accordingly, we view language as an inherent internalised professional tool imbued with values, assumptions and traditions, which supports ongoing meaning-making in a specific context. Following from this, when organisations employ language-based recruitment filters, these should purposely account for the fact that new entrants ‘do not diffuse into cultural void but, rather, into a perplexing cultural universe that delineates the roles and responsibilities of its respective actors and the boundaries of appropriate behaviour’ (Ansari et al., 2010: 78).
A fit with professional and organisational cultures is achieved through socialisation, of which language practices are an important component (Irimiea, 2017; Ochs, 1991). In this article, we focus on socialisation into a professional context through discourse. Undoubtedly, for expatriates to achieve a discursive fit is of particular importance in professions such as the medical profession, in which ‘cognitive shortcuts’ (Tversky& Khanemann, 1974) may have fatal consequences. This article argues that the established international recruitment process in the NHS is predisposed to ignore – with detrimental consequences for its staffing needs and, potentially, the well-being of the expatriate employees – the importance of the discursive fit by overlooking the differences between language and discourse in their language testing praxes.
This research centres on expatriate nurses within the NHS. For this professional group, the dynamics of change in recruitment and retainment have been particularly volatile ( The Times, 2017). For many of them, English is not a ‘first language’ 1 . This creates a problem for the NHS because poor command of English by the staff may be risky to the patients. As a result, a test of English language competence has become a condition for entry into the nursing profession for non-British nationals. There is, however, a substantial controversy regarding the nature of the test and the kind of linguistic ability to be tested and at what level (Müller, 2016; Sedgwick and Garner, 2017). As a contribution to this debate, we argue that the existing testing system suffers from important shortcomings because of its disconnection from embedded communication practices. We maintain that in certain circumstances a contextualised communicative language ability (a discursive ability) may be more important than formal language proficiency. Although this article refers to the NHS, healthcare is not the only professional field in which language testing is a condition of employment. This gives this analysis a wider relevance within the debate on business organisations’ reliance on a globally scattered workforce.
The concern of this study is the aptness and compatibility of the available screening instruments in their aspects related to language proficiency. Our study mostly relies on conceptual arguments supported by evidence obtained during a controlled experiment conducted in a guarded environment. Although necessarily limited in scale, it provides some valuable pointers because of its randomised nature (Creswell, 2013). In terms of scholarship, this study of the NHS case delivers what Langley et al. (2013:4) call actionable knowledge: ‘knowledge about how to produce the changes that the evidence suggests are desirable’. Our evidence reveals inconsistencies in the three language-related recruitment paths employed by the NHS, indicating that the organisation has not found a coherent solution to the challenge of maintaining a required standard of communication skills of staff for whom English is not their ‘first language’. The use of a standard off-the-shelf test, we argue, appears to be part of the problem.
This investigation results in a research proposition relevant to organisations with an international workforce: reliance on language testing that prioritises formal characteristics of language proficiency over appropriate contextual awareness may result in choices that hinder the professional functionality of an organisation. We conclude that although testing language proficiency helps to establish the relevance of the incoming workforce, organisations should be aware of the limitations of such tests with respect to the specific demands of their operational context. Because professional communications are anchored in contexts, which off-the-shelf language tests are unlikely to reconstruct, organisations should seek to complement them, when necessary, with further instruments of selection more attuned to their own specific needs.
The linguistic foundations of professions
Professions may be defined as paid occupations which involve prolonged education, training and a formal qualification. Linguistically, a profession is demarcated by a particular variety of language developed and applied in order to share specialist knowledge (Gunnarsson et al., 2014). In the literature, it is described as the ‘language for special purposes’ (LSP). It draws on specialised terminology and has distinctive stylistic features which provide linguistic foundations for a specific knowledge domain (Alcaraz, 2000; Engberg, 2006). LSPs enable purposeful and effective communication and make it possible to capture meanings that everyday language is not equipped to do. An LSP, however, is more than an agglomeration of professional terms. It has a prominent social function: apart from signalling membership of a selected professional group, the use of an LSP enhances the users’ image and professional credibility, and gives weight to what is being said.
It is significant that a distinction should be made between a professional language and professional discourse. The latter is broader and qualitatively different in comparison to LSP; it is never restricted only to the use of terminology. Rather, it is a form of shared practice that exists as experience, norms, concepts, forms of representation, group ideologies and language that has a sociocultural origin and is communicatively sustained (Fairclough, 2007; Grant et al., 2009). Discourse creates a contextualised and situated understanding of connected statements, concepts, terms and expressions pertinent to a specific knowledge domain (Watson, 1994). Discourse is more than just an application of words; it is also about how and when the words are used, and how they are related to a social and professional situation in which they are applied. This requires an appreciation of culturally specific ways of speaking, writing and organising thoughts (Lune, 2013; Paltridge, 2006). In other words, professional discourse involves situated and socially constructive use of LSP conditional on contexts-of-use, which, it is important to note, frequently are not limited to communication within professional groups only but include insider-outsider interactions and meetings (Watson, 1994). Partaking in discourse implies an ability to choose and apply a linguistic register that corresponds to the social situation in which communication is taking place (Sedgwick & Garner, 2017).
The linguistic foundations of a profession have, therefore, several interrelated language-based aspects (Tietze, 2008): first, a technical aspect, as specialist lexical units constituting an LSP have to be precise, exact and unambiguous; second, the application of an LSP as practice in specific contexts of use as a discourse; and third, the LSP must be regarded as a subsystem of the national language, the language for general purposes (LGP), so to say, characterised by aspects such as distinctive grammar, syntax and punctuation. It follows that the LSP and the LGP necessarily overlap and a balance between the two depends on communication situations.
To put this in the context of the NHS, an expatriate nurse whose native language is not English would need a level of competence in English that will make possible both social and occupational interactions within the very particular and demanding professional environment of a medical establishment. She is expected to understand, reproduce and apply the specialist vocabulary associated with her job as a nurse, that is, to apply the English language as an LSP. In dealings with patients, she will need to be able to respond with understanding, compassion and expertise, thus simultaneously drawing on the specialist vocabulary of the occupation and, importantly, activating it as appropriate within particular social settings (e.g., matching the language to the patient’s background). This may require an interaction referencing certain established rituals and cultural traditions. These communications, ceremonies, responses and interactions constitute discourse, that is, the active use of language. A nurse also needs to be able to apply discretion and circumspection to execute her responsibilities well. In other words, an expatriate nurse has to be able to draw on and activate several linguistic elements: English as the LGP, the LSP, and discourse as language in action.
Within the boundaries of a nation-state, professional lexical systems are rooted in the national language and cultural conventions. A link between professional and native languages is important in the light of the findings of psychological and cognitive science research, which show that the latter profoundly influences how people perceive the world and affects cognition and behaviour (Liang et al., 2018). For example, Chen (2013) provides evidence that languages in which it is mandatory for speakers to grammatically mark future events (e.g., in English by using either ‘will’ or forms of ‘be going to’) foster future-oriented economic behaviour. The implication is that LSPs practised in different countries, even when they are terminologically close, may embody a different worldview reflecting that of the native languages on which they are based.
That LSP is linked to national culture, history and mentality. It has great importance for expatriate professionals who are non-native speakers and were brought up and educated in a linguistic environment that has different evocative and representation properties and, most likely, dissimilar cultural underpinnings pertinent to knowledge acquisition and information processing. Despite the standardising effect of the internationalisation of knowledge and the advance of English as a lingua franca in many professions, these two factors combined do not fully mediate the persistent variances in meaning that reflect not just differences in vocabulary, but also cultures and social conditions, as well as any other factor that constitutes the environment in which language is practiced (Angouri, 2018; Mautner, 2016). If an expert in his/her field were only ever to communicate with fellow experts in highly specific and limited contexts, discourse-based issues would be of lesser importance. However, in the vast majority of cases an LSP is practised in social interactions with broader settings involving both members of the profession and outsiders. The argument then is that LSP intertwines with discourse, and that being an effective professional depends simultaneously on having high competence in both LGP and the discursive aspects of his or her work. These often relate to circumstances of symbolic signification specific to a situated culture that relies strongly on ‘the unspoken assumptions of social life’ (Davies, 2001:135), putting the often tacit social aspect of communication centre stage. What follows is that in relation to the quality of communication, that is, the ability to make the exchange of information complete and exact, and the information accessible, the contributions of language knowledge and background knowledge are very difficult to distinguish in practice (Douglas, 2013). The question is whether a language test may be trusted to evaluate both language proficiency and appropriate discursive abilities.
A recruitment conundrum: language proficiency vs language skills
With communicative abilities being central in many professions, employers may want to be reassured that non-native speakers among their staff meet specific language benchmarks. In certain sectors, these requirements are standardised and formalised across the field, for example, the medical profession in the UK, USA, Canada and Australia and the legal profession in England and Wales. Some companies set their own criteria. Either way, testing language skills faces a huge conceptual issue: should it be a test of a professional language or of general language competence? Some fundamental arguments have been presented to justify either approach. On the one hand, the maxim that the use of language belongs ‘to practices not to individuals’ (Davies, 2001:135) lends support to a focus on testing language for special purposes (LSP). On the other hand, professional discourse is never restricted to the use of terminology. From this perspective, a more inclusive and generalised approach to testing that combines language ability and background knowledge makes a good case for itself.
The pros and cons of these two approaches to language testing have been discussed at some length in the language assessment literature (Davies, 2001; Douglas, 2000). The discussions have produced strong theoretical arguments in favour of applying a generalised language test rather than targeted tests. The most fundamental problem with LSP tests is that the logic which guides them drives towards an over-restriction of language requirements (Knoch & Macqueen, 2020). This narrowness follows from the need to link the test material to specific situations and demands of the profession which the examined person aspires to join. Outside professions that apply formulaic and, therefore, restricted language, for example air traffic controllers, specialised language testing may prove inadequate for at least two reasons. First, terminological competence represents only a small fraction of the required communicational skills. Second, setting LSP exams is problematic because it is very difficult to draw a line between a language proficiency test and a test of professional skills on the one hand and between languages used in different professional domains, for example, ‘Chemical English’ or ‘Biological English’, on the other. Davies (2001:143) concludes that: ‘LSP testing cannot be about testing for subject-specific knowledge. It must be about testing for the ability/abilities to manipulate language functions appropriately in a wide variety of ways. This might mean no distinction between a general proficiency test and an LSP test’.
The literature, though, highlights one significant weakness of a generic test: it breaks the link between language and its context (Davies, 2001; Kuznetsov & Kuznetsova, 2016), which is a substantial limitation considering that ‘language is not only content; it is also context and a way to recontextualise content’ (Boje et al., 2004:571). Because it seeks to be neutral towards any specific form of professional activity, a generic test (e.g., the International English Language Testing System (IELTS) ends up examining grammatical accuracy and competence in a decontextualised and, therefore, essentially sterile ‘dictionary’ language not grounded in a realistic communicative setting (Spolsky, 1985). Deprived of a context, ‘for-test’ language loses its discursive constituents in the form of ideas and expressions utilised by members of professional groups to make sense of their situations. The following section illustrates the challenges of finding the balance between formal language proficiency and context awareness pertinent to effective international recruitment.
Methods and empirical context
We studied non-UK trained expatriate nurses seeking employment with the NHS. We considered the period since January 2016 when a stipulation was introduced that all such nurses should demonstrate language competency as a condition for employment and reviewed the NHS procedures in terms of their consistency with the stated goal of maintaining core professional standards in the organisation. The study’s background information came from open sources, including a variety of documents and public communiques produced by the NHS and the Nursing and Midwifery Council. Collectively, these documents constituted an essential database that provided us with a critical insight into the formal requirements of the NHS concerning how linguistic competence of the nurses was to be assessed.
Our analysis was informed by an approach favouring inductive reasoning, when ideas arise from the data rather than from assumptions reflecting available theoretical constructs (Padgett, 2017). Because there were no existing theoretical frameworks from which we could derive a research proposition, we started with the collection of qualitative data. As we reviewed the evidence, ideas and concepts began to emerge. Our analysis was particularly influenced by constructivist principles that position the researcher as the author of a reconstruction of experience and meaning (Mills et al., 2006). We began our analysis by structuring the data into open codes. Eventually, this led us to identifying two key themes, one being the actual language-related tasks faced by healthcare professionals and the other the simulated tasks that form the centre-piece of language testing (Table 1). Finally, we set an experiment intended to test our interpretation of the data. Ours was a version of a lab experiment where subjects (the NHS nurses in our case) perform tasks under controlled conditions (Hamel & Birkinshaw, n.d.). Experiments of this kind are widely used in the social sciences (e.g., Demaj, 2017; Hipes et al.,2015). Their purpose is to test theoretical relationships rather than to discover empirically grounded generalisations (Lucas, 2003; Webster & Sell, 2014). This approach removes the requirement of sample representativeness as a condition of external validity.
Language-related tasks: actual versus simulated
Sources: Hearnden, M. (2008) ‘Coping with differences in culture and communication in health care’, Nursing Standard (through 2013) 23(11):49–57; O’Hagan, S., E. Manias, J. Pill, R. Woodward-Kron, T. McNamara, G. Webb and G. McColl (2013) ‘What counts as effective communication in nursing? Evidence from nurse educators and clinicians’ feedback on nurse interactions with simulated patients’, Journal of Advanced Nursing 70(6):1344–1355; Petronio, S. and J. Sargent (2011) ‘Disclosure predicaments arising during the course of patient care: nurses’ privacy management’, Health Communication 26(3):255–266; Rayo, M. F., A. F. Mount-Campbell, J. M. O’Brien, S. E. White, A. Butz, K. Evans and E. S. Patterson (2014) ‘Interactive questioning in critical care during handovers: a transcript analysis of communication behaviours by physicians, nurses and nurse practitioners’, BMJ Quality and Safety 23(6):483–489; Sedgwick, C., M. Garner and I. Vincente-Macia (2016) ‘Investigating the Language Needs of International Nurses: Insiders’ Perspectives’, IELTS Research Reports Online Series No.2/2016.
The NHS selection practices
To be eligible to practise within the UK, nurses must be registered with the Nursing and Midwifery Council (NMC), the statutory body regulating the profession in the country. The core professional guidance for nurses and midwives in the UK is The Code: Professional standards of practice and behaviour for nurses and midwives, developed and enforced by the NMC. The Code (§7.5) states that nurses and midwives must ‘be able to communicate clearly and effectively in English’; ‘Guidance on Registration Language Requirements’, published by the NMC, clarifies that the necessary knowledge of English is a knowledge ‘which is necessary for the safe and effective practice of nursing or midwifery in the United Kingdom’ (NMC, 2020:1).
To implement this requirement, in January 2016, the NMC introduced a stipulation that all new applicant nurses who were not trained in the UK should demonstrate language competency as a condition for registration. Several options for doing this were provided, one of which was achieving a minimum score of 7.0 in the four elements (listening, reading, writing and speaking) of the IELTS Academic 2 .
The decision to use the IELTS Academic, originally developed as an English language test for university entry in the UK, was taken after a public consultation that revealed a conflict of opinions among the stakeholders regarding the test’s suitability. Those in opposition to IELTS commonly expressed views that the test did not necessarily give an indication of competence in a clinical context and that the minimum score of 7.0 in each element of the test was too high (NMC, 2015). In the end, the NMC opened to applicants three options to demonstrate competence in English. They need to present evidence of: (a) a recent overall score of 7.0 in the IELTS Academic test (or grade B in the Occupational English Test after 2017); (b) a recent pre-registration nursing or midwifery programme that has been taught and examined in English in a country where English (in the terminology of the NMC) is ‘the first and native language’; or (c) a registration and two years of registered practice with a nursing or midwifery regulator in a country where English is ‘the first and native language’ and a language assessment was required for registration.
What is central from the point of view of this investigation is that the three options are neither equal nor consistent in terms of the implicit criteria applied and what this inconsistency suggests about a generic language test as a professional filter. The IELTS-based option (a), as was demonstrated, reflects the ability of the person to perform well in a test environment and to exhibit a certain level of competence in the use and understanding of what is essentially a pro forma decontextualised language. Because the assessors have to be guided by clear and replicable rules to maintain the rigour and compatibility of results, the assessment criteria are necessarily limited and focus on easily demonstrable qualities of language proficiency. For example, in the IELTS Speaking module, the score depends on fluency and coherence (e.g., talking without unnatural pauses or hesitation), lexical resource (e.g., how broad the range of used words and sentence structures are) and pronunciation (e.g., how easy it is to understand the candidate) (British Council, 2018). In turn, options (b) and (c) are based on the supposition that the applicant has had sufficient practice in using the language during either study or as a practising nurse to warrant that her English is equal to the IELTS band 7 as a minimum.
The cogency of this conjecture is questionable at least on two counts. First, as will be demonstrated later, existing and working in an English language environment is no guarantee of acquiring language skills equitable with the IELTS band 7. Second, the successful performance of an individual as a student or practitioner in a foreign language environment only marginally depends on language proficiency. Other qualities described in the literature as affecting professional success (see, e.g, Bingle & Davidson, 2014; Boudreau et al., 2001), for example, are social and emotional intelligence, perceptiveness, adaptability (openness to experience), determination and persistence, professional acumen, and other traits are equally, if not more, important. None of these may be revealed or evaluated through a language test. Conversely, when present, they do not necessarily translate into a prescribed level of language competence. The inevitable inference is that the three paths to the NMC registration available to nurses and midwives from outside the UK are vastly unequal in terms of the demands and criteria they represent.
Inconsistencies in the three paths vividly demonstrate, in the opinion of the authors, the language versus discourse dichotomy. As noted, the NMC demands that nurses/midwives communicate in English ‘clearly and effectively’. The IELTS option assumes that passing the test is sufficient proof of this ability. The soundness of this assumption is questionable. The demands on the communicative ability of nurses are extremely diverse and expansive, as is attested by extensive literature (see Table 1). Notably, they include an ability to express medical concepts in professional and everyday English, adapt the linguistic pallete to the needs of both colleagues and patients; use linguistic strategies to elicit specific information from patients and colleagues, reassure anxious patients, and other duties (Sedgwick et al., 2016). Meeting these demands is what compounds effective communication in English in the nursing environment. None of them is appraised by the IELTS. By contrast, by accepting the evidence of training/practicing in an English language environment as a path to registration, the NMC shifts priorities from passing a language proficiency test as a token of a professional aptitude to the ability to communicate with authority in the workplace. These two priorities are not fully compatible, because professional communication is discourse-based and, as such, is much richer and more diverse in what it represents than what a language test may reveal.
The experiment
To verify the analytical conclusions developed in the study, an experiment was set by the authors. Its purpose was to confirm our supposition of a gap between the discursive fit and formal language proficiency that undermines the validity of the existing practice of language testing as a selection tool. Thirteen randomly selected volunteers from among NHS nurses were invited to take the IELTS Academic test. The number of participants was unavoidably limited because of the cost (the exam fee of circa £160 plus expenses of up to £30 per participant) and the demands the experiment imposed on the free time of the volunteers. At first glance, the size of the sample may appear a hindrance. In reality, the size in this case is irrelevant because the purpose was not to prove a point statistically or establish frequencies. Our intention was not to seek generalisation but instead to look, based on a descriptive inference, for discernible traces of incongruity between the two qualitative categories, language proficiency and discursive ability, within the context of the NHS. From this perspective, what matters is the presence of the evidence of a possibility of occurrence. Accordingly, more observations do not lead to a more precise description of the event; from this perspective, the randomness of the sample is more important than its size (King et al., 1994).
Professional success (employment with the NHS at an advanced professional grade) was assumed as a proxy for a discursive ability while the test results were assumed as a proxy for language competence. The purpose was to establish whether a discrepancy between the two may be detected even in a small arbitrary sample. All the participants were NMC-registered nurses working in UK hospitals. The first cohort consisted of seven native English speakers (four from England, two from Scotland and one from Wales) born, raised and educated in the UK and with university degrees. The second cohort included six EEA-trained nurses. Each of these had over three years of experience working in UK hospitals and was in the NHS pay band six (out of a possible nine); three of the second cohort had post-graduate qualifications from UK universities. The participants were informed about the structure and requirements of the test and provided with materials and advice to help with preparations for the exam. All participants signed a letter of commitment in which they promised to perform at the exam to the best of their ability. The expenses (the test fee, travel) were met using the private funds of one of the authors. The test results are shown in Table 2 and Table 3.
Two out of seven native speakers (participants 5 and 6) failed; yet, the native speakers excelled in the speaking part of the test (the average score was 8.71). The writing section was a weak spot for both native and non-native speakers 3 . All foreign participants failed, but they showed a strong performance in the listening and speaking components of the test.
Discussion and conclusions
Our evidence based on the UK practice highlights the limitations of a language proficiency test as an instrument used to gauge the professional abilities of the testees. Even native speakers may on occasion fail a test in English, for example, to achieve the score of 7.0 in Academic IELTS as required by the NMC. This demonstrates a distance that exists between test conditions and real-life demands. This is indirectly recognised in the existing recruitment system that makes nurses of foreign origin trained in the UK eligible to join the register without passing the IELTS test, although the UK universities’ English entry requirements for international students applying to study nursing can be lower (Cardiff, 2019). Similarly, nurses who become midwives through an 18-month NMC-accredited course acquire an automatic NMC registration without being required to provide a level 7 IELTS in each part of the exam. It may be pointed out further that it is apparent from the description of the bands that language command at a level 7 allows for some misunderstandings on the part of the testees. Even level 8 (out of 9) accepts misunderstandings in ‘unfamiliar situations.’ It is apparent, therefore, that the pass grades for IELTS chosen by the NMC are both arbitrary and restricted as a predictor of the readiness of a nurse to safely operate in the hospital environment. Even more important in our opinion are the inconsistencies in the language-related recruitment criteria employed by the NHS. The three paths to the NMC registration available to expatriate nurses and midwives vary greatly in terms of the focus of the screening process and the criteria applied. The organisation has not found a coherent solution to the challenge of maintaining a required standard of communication skills of staff for whom English is not their ‘first language’. The language skills of the recruits employed through different paths cannot be the same, casting doubt on the validity of this approach. A two-stage process that combines a language test to establish basic language skills and an introductory period, during which discursive experiences are built under supervision would combine the benefits of all three paths and preserve equality of selection.
A recognition that the pool of talent and skills has undergone globalisation has changed the organisational approach to recruitment. Awareness of differences across national contexts as an important factor of successful recruitment strategies has acquired a prominence that was not appreciated previously (Gallardo-Gallardo et al., 2020; Silvanto & Ryan, 2014). Importantly, alongside other manifestations, these differences reveal themselves through diverse discourse traditions within the same profession, which a generic test of language skill on its own is unable to account for because it does not relate to the ideological and psychological elements of professions. As a result, organisations hiring across borders can experience professional functionality loss because of the rift between language as a formalised system of words and grammatical rules, and discourse as ‘language in action’, which is characteristic of generic language testing. Therefore, the employer concerned with maintaining cognitive coherence and a shared ideology between recruits who are not native speakers and the existing staff may find that a generic test cannot be trusted to choose the right individuals. For example, in her study of foreign nurses in Australia, O’Neill (2011) reports that even those respondents who successfully passed prescribed English language exams felt frustrated and isolated because of the impact that difficulties with language had on their sense of belonging and ability to participate once in the hospital setting. Rosa González et al. (2021) relate similar findings for Spanish nurses in Germany. It is on this basis that the power that certain organisations and professional bodies accord to the IELTS and similar is challenged (Pilcher & Richards, 2017).
Another apparent downside of relying on a decontextualised test includes rejecting valid candidates based on inadequate assumptions and sending a wrong message about the priorities of the hiring organisation. At the same time, those expatriates who managed to get the job have to cope with a potential disruption of their identity construction as the new discursive environment challenges their confidence as competent professionals. And yet, there are, of course, strong practical considerations favouring a standardised one-serves-all approach to testing that are hard to ignore. By using ‘generalised tests’, organisations benefit from the economy of scale and save money on preparation, validation and application of tests. Test of English as a Foreign Language (TOEFL) and the International English Language Testing System (IELTS) are the two most well-known standardised English language tests. Millions take them every year and the results are accepted by academic institutions, professional organisations and immigration authorities across the world. The scale of these operations generates enough income to make these exams affordable and accessible without compromising their validity, rigour and integrity.
Our conceptualisation of the language-based selection process implies that there is no single winning default strategy that can be prescribed to all organisations. The choice of an approach needs to reflect practical and contextual considerations. The case of the NHS highlights the difficulty of finding a selection tool that would provide a model for international recruiting that accounts for both professional competence and linguistic skills. The results of this analysis reflect that when choosing between an LGP and an LSP test, organisations cannot ignore the difference between language and discourse.
The example of the NHS demonstrates that the ability of employees to adapt discursively to variable social and professional contexts may be more significant than the richness of their vocabulary and their command of grammatical rules. In such situations, testing that emphasises formulaic language may be misleading and, eventually, counterproductive. It follows that recruiting organisations must be flexible and imaginative when approaching the task of language testing rather than buying into the off-the-shelf products. And yet, in practical terms, moving away from the IELTS and similar international testing systems is not likely to happen any time soon. The numerous advantages they offer (accessibility, affordability, well established international infrastructure, to name some) are, however, superficial. Probing deeper into the contextual foundations of the discursive fit reveals the untapped potential of the benefits of international mobility and encourages innovative resolutions. One possible solution is to adopt a multi-tier system in which an LGP exam is an initial filter, followed by a period of socialisation into the organisation during which a dedicated effort is to be made to raise the awareness of the new recruits of the nuances of the socio-cultural environment in which they operate professionally.
Organisational socialisation is an important human resources practice (Saks & Gruman, 2014). Its design has implications for the scale of funds required to recruit needed skills and, importantly, the productive allocation of these funds, for example, towards pre-entry filtering or into post-entry development. Another consideration that emerges from the analysis is the choice of the exam components. For instance, in professions for which the social, cultural and emotional aspects of communication are particularly important, it is sensible to prioritise speaking and listening skills during testing by giving the results in these categories greater weighting. In the experiment described earlier, to give one example, the writing test was responsible for most failures but was not directly relevant to what nurses have to do (Sedgwick et al., 2016).
The NHS is an ideal case of an organisation that can reap benefits from a more situated approach to testing language and discursive abilities. Although not all organisations may face the challenges that the NHS has to deal with, because of the specific nature of the services it provides, there is a primary theme that emerges from this analysis: organisations should be concerned with the social aspects of language proficiency. They need to recognise the conflict that exists between ensuring the practicality of an assessment and its appropriateness and make an effort to resolve this conflict in accordance with the organisational priorities. Our analysis leads us to propose to organisations that face recruitment challenges similar to the NHS to consider investing in developing their own testing models and protocols, which will prioritise contextualisation over a simplified notion of language proficiency.
The experience of the NHS demonstrates the importance of recognising the social aspects of linguistic acculturation. Passing a formal test is only a beginning of the process of professional socialisation. There is a potential threat that both the employee and the employers may interpret a successful test as sufficient evidence of language proficiency. As we have demonstrated in this article, discursive ability and practice often matter much more. There is a real risk that foreign workers may find themselves not sufficiently equipped as far as terminology and jargon are concerned to participate in professional discussions with colleagues on equal terms. This may produce a feeling of exclusion, a discourse-related segregation when non-native speakers may feel that certain information and knowledge is not accessible to them. This is potentially a very undesirable situation as it is fraught with multiple serious consequences for both employees, who may have to deal with anxiety caused by self-doubt and organisations as they may face the issues of low staff morale and staff turnover. This is an important and yet underresearched topic that deserves far more attention from the academic community than it has received so far.
© Andrei Kuznetsov, Olga Kuznetsova and Jaime Fernández de Simón de la Cruz, 2022