New ways of investigating coaching: linguistic research on executive, business and workplace coaching – a systematic scoping review

ABSTRACT In this review, we give a systematic overview of linguistic coaching process research. This research represents a relatively recent inductive, descriptive-phenomenological approach allowing to linguistically document and analyse the in-situ co-construction of the coaching conversation by coach and client. Applying qualitative, linguistic methods and working with transcribed, video/audio-recorded authentic coaching sessions as data, this approach allows insights into what coach and client actually do together on a moment-to-moment basis. Using a scoping review methodology, the aim of this paper was to systematically collect, present, synthesise and critically assess for the very first time what we know about linguistic process research on business, workplace, and executive coaching to date; to map its most important concepts and findings; to carve out advantages and limitations as well as research gaps. Following a rigorous search strategy, 46 contributions were reviewed. Findings are divided into quantifiable results regarding, for instance, coaching format, methods, and data used etc., and non-quantifiable results. The latter give an overview of the most important research foci and findings on the macro, meso and micro levels of coaching interactions to provide readers with a comprehensive picture of the depth and richness as well as the complex opportunities of linguistic coaching research.


Introduction
As recently argued by Grant and O'Connor (2019) or Graf andDionne (2021a, 2021b), coaching has left its embryonic, belief-based origin and is gaining an evidence-based status.Particularly the past two decades have witnessed a substantial growth in research and various international meta-analyses on coaching outcomes have been published (see, for instance, Wang et al., 2022).The overall positive effect on clients and their wish for learning, development, and change has therein been empirically established.
Yet, various challenges regarding the professional, theoretical, and empirical foundations of coaching continue to exist: With respect to coaching as distinct helping format, so far little standardised conceptualizations and only few established definitions of coaching are available.Serious definitions promote a reading of coaching as a collaborative and egalitarian relationship between a coach, who is not necessarily a domain-specific specialist, and a client, which involves a systematic process that focuses on collaborative goal setting to construct solutions and employ goal attainment process with the aim of fostering the on-going self-directed learning and personal growth of the client.(Grant & Stober, 2006, p. 3) As regards its more specific organisational form, Kilburg (2000, p. 142) talks of a helping relationship formed between a client who has managerial authority and responsibility in an organization and a consultant who uses a wide variety of behavioural techniques and methods to help the client achieve a mutually identified set of goals to improve his or her professional performance and personal satisfaction and, consequently, to improve the effectiveness of the client's organization within a formally defined coaching agreement.
Besides these conceptual difficulties, significant research gaps persist, especially regarding the coaching process, i.e., 'the interactions between coach and coachee in vivo' (Fillery-Travis & Cox, 2018, p. 530).Such insights would be particularly rewarding for practitioners, who could monitor and critically reflect on their own professional doing (Stein, 2009).Gaining insights into the coaching process itself and the underlying interactive and communicative mechanisms between coach and clienta core goal in coaching process research (Wegener, 2019;Graf & Dionne, 2021b)will also allow to understand what exactly enables and facilitates the (positive) outcome in coaching (Gessnitzer et al., 2018;Ianiro & Kauffeld, 2015) attested to in various meta-analyses.In this sense, De Haan et al. (2010, p. 110) argued that … to understand the impact and contribution of executive coaching … it is not enough to just understand general effectiveness or outcome.One also has to inquire into and create an understanding of the underlying coaching processes themselves, from the perspectives of both clients and coaches.
So, for over a decade, researchers like De Haan et al. (2010), Ianiro and Kauffeld (2015), Fillery-Travis and Cox (2018), Gessnitzer et al. (2018) or De Haan (2019) have called for analyses into coaching activities, (un-)successful interventions, communicative practices etc. from both coach and client, to gain more detailed insights into the coaching process itself.Yet, unlike the prolific coaching outcome research, coaching process research is still relatively scarce, due to, among other things, the complex nature of the process itself (Fillery-Travis & Cox, 2018;Schermuly, 2019).
Linguistics, the core discipline to study language and communication, however, is particularly suited to generate both in-depth insights into this so-called 'black box' (Ianiro & Kauffeld, 2015, p. 42), i.e., into the 'underlying interactive mechanisms of coaching' (Graf & Dionne, 2021b, p. 6), and to tackle and unravel the complexity of the coaching process on the interactional micro-level.In analogy with the 'change process research paradigm' established by Elliot (2010) for psychotherapy, linguistic coaching research thus operates within the 'microanalytic sequential process design' (Wegener 2019).It approaches coaching as talk-in-interaction and more specifically, as a professional helping interaction (see, e.g., Graf & Spranz-Fogasy, 2018a).Such interactions depend on a purposeful interpersonal relationship between help provider and help seeker and are primarily communicative in nature (Miller & Considine, 2009;Scarvaglieri et al., 2022).By means of a helping conversation, professionals provide (emotional and informational) support to address the physical, social, psychological, intellectual, and, in the case of coaching, professional, needs of their clients.Like other professional helping formats, coaching transpires in and through a series of helping conversations with language as primary medium and means for the coachclient interaction (Graf et al., 2019;Graf & Spranz-Fogasy, 2018a).These conversations are key to results, i.e., they guarantee the coaching intervention's effectiveness (De Haan & Page, 2013).The linguistic-constructivist perspective conceptualises coaching as emerging in-situ, which means that both coach and client create coaching as a communicative event via their mutually depending (non-)verbal contributions to the coaching conversation that eventually allow for learning and development in clients (Graf & Dionne, 2021b).By focusing on this local interaction within a qualitative approach, valuable insights into how coach and client 'do coaching', and of the change processes that clients undergo regarding emotions, experiences, and actions, i.e., of the local effectiveness of coaching, emerge (Graf et al., 2019;Graf & Dionne, 2021b).This means that '(t)his form of qualitative research stays as close as possible to the events in the sessions themselves' (De Haan, 2019, p. 228).It does not rely on pre-formulated hypotheses or categories, since these are inductively developed and iteratively adapted based on indepth data (re-)examinations.This allows for the analytical flexibility needed to capture the complexity as well as the internal organisation of (unique) interactions.Similarly, findings represent interactional or discursive patterns that have (repetitively) emerged from the data, but not distilled themes or abstracted and simplified models.Such research is based on naturally occurring data sets, i.e., audio and/or video recordings of authentic coaching processes, which allow immediate access to the actual communicative activities.Following strict guidelines, recordings are transcribed (and anonymized), including pauses, overlaps, back channelling, repairs, false starts, stress etc. that offer valuable insights into relational aspects (Sidnell, 2010).Compared to other helping formats such as psychotherapy (see, e.g., Peräkylä et al., 2008), linguistic research has only recently become interested in coaching; the communicative activities of coaches and clients within sessions and processes or the effectiveness of discursive interventions represent relatively new research foci in linguistics.To this day, no systematic review of the existing linguistic research on coaching has been carried out.
The purpose of this review is to close this gap and to systematically collect, present, synthesise, and critically assess what we know about qualitative, linguistic coaching process research to date.The focus is on business, workplace, or executive coaching, the most frequently researched type of coaching so far.This complements existing knowledge on qualitative psychological research as found in the systematic review by De Haan (2019).Our aim is to offer a first understanding of the nature, the quality, the complexity, and the extent of this research alongside its gaps and limitations (Siddaway et al., 2019).To carry out this exploratory endeavour, we use a scoping review methodology (Sherrer & Rezania, 2020;Willemsen et al., 2023) as one possible way to systematically synthesise existing literature, map an entire area of research for the first time, and locate its key themes and concepts (Siddaway et al., 2019).Our review is guided by the following research question: What does the existing academic literature tell us about the goals, methods, findings, limitations, research gaps etc. of linguistic research on (business, workplace, executive) coaching?The scope of the question is intentionally broad to provide as comprehensive and complete an overview as possible.On the basis of such a review, it will become evident how and what qualitative linguistic coaching process research can contribute alongside quantitative and qualitative psychological outcome research on coaching; concurrently, the review will also carve out and critically discuss its (methodological and theoretical) limitations.Research gaps and desiderata will be identified.All in all, this systemic, scoping review will add a so far missing rigour, transparency, and systematization (Siddaway et al., 2019) by providing a novel, comprehensive overview of qualitative linguistic coaching research to date.

Methodology
The scoping review methodology chosen for this article is particularly suited for broad topics and (new) areas of study that have not yet been (fully) reviewed as is the case for linguistic research on business, workplace, and executive coaching (Sherrer & Rezania, 2020;Willemsen et al., 2023).In line with systematic reviews (Siddaway et al., 2019), it follows a rigorous methodology to identify, select, and critically examine relevant research contributions.Its aim is to help especially non-linguists to better understand a still relatively unknown but promising research approach, its concepts, methodological procedures, research foci, analytical possibilities, major findings etc.This seems particularly relevant given that coaching research is still dominated by (quantitatively operating) psychological (outcome) research and in light of urgent calls for more research on the coaching process as well as both participants involved.
In order to systematically locate relevant studies, a three-stage search strategy was applied.First, two large databases were searched.Due to the lack of topic-specific, linguistic databases 1 , 'Scopus', a comprehensive database covering disciplines such as life sciences, social sciences, health and physical sciences, and the psychological databases 'PSYNDEX/ PsycINFO', to include both German and English research, were chosen.Based on initial searches, the broad search keywords 'coaching' and 'linguist*' 2 were used in 'all fields' for Scopus and in 'keywords' for PSYNDEX/PsycINFO to identify a maximum number of relevant studies.The use of additional keywords narrowing the scope of the search, such as 'executive', 'business' or 'workplace', was discarded as it limited the number of relevant results.This way, a total of 6,647 documents (5,132 in Scopus and 1,515 in PSYNDEX/PsycINFO) were identified for screening.Duplicates were immediately removed.After completing the screening of abstracts, titles, and keywords according to the inclusion criteria (see below), 59 items were read more thoroughly during the eligibility stage by both authors.This reduced the number of included items to 17. Second, reference lists of the 17 relevant studies from the data base search as well as the reference lists of eligible work identified via those references were consulted.This resulted in 26 additional contributions.In a third step, a Google Scholar search using the same keywords as well as an individual search of relevant journals 3 was conducted, leading to 3 more results.The overall process of item identification, screening, eligibility, exclusion, and eventual inclusion is outlined in Figure 1.The number of items used for the analysis eventually amounted to 46.
The criteria consistently applied for results to be included in the scoping review are as follows: Regarding languages and year of publication, works published in English or German (the native language of both researchers) between 2000 and 2023 were included.The starting point was set in the year 2000 according to Grant and O'Connor's (2019) primer on evidence-based coaching, which sees the beginning of serious coaching research in the year 2000.The review included all qualitative and interdisciplinary study designs (with a clear contribution of linguistics) as well as theoretical articles with a qualitative, linguistic approach.For empirical research, authentic coaching interaction data had to be used since this is a prerequisite for any kind of process-oriented linguistic inquiry (Sidnell, 2010).In contrast to other more encompassing reviews, the types of contributions included were limited to academic articles, books, chapters, and dissertations, excluding practicebased literature, unpublished works, grey literature, conference abstracts, etc.This decision was made due to the multitude of practice and experience-based literature that exists on language in coaching, which lacks empirical and scholarly backing.As mentioned above, the type of coaching in focus was business, executive or workplace dyadic coaching.This excluded peer and group coaching formats, as well as other types of coaching such as health, sports, language/dialectal or family coaching as well as coaching in pedagogical settings, where the term is often used synonymously for training, teaching, or even instruction.This contrasts with our definition of coaching as help for self-help at eye level.Linguistic contributions on other helping formats such as mentoring, supervision, mediation, consulting etc. were also excluded since they display other interaction-type specificities.Most importantly, to differentiate the results from other language-related but not linguistic research, a process-based, data-driven, inductive focus on the coach-client interaction using linguistic methods or theoretical concepts was a necessary requirement for inclusion.
The citations (author(s), year, title, source, keywords, abstracts) were exported from the data bases and complemented with the additional results from the secondary search (see references with *).The aims, research question(s), data, method(s), results, and concluding sections were used to identify, compare, contrast, and assess concepts, research phenomena and key findings as well as contradictions and research gaps (Siddaway et al., 2019;Timulak, 2009).

Results
The results section first illustrates quantifiable results related to the authors, the types and forms of coaching investigated, or the data and research methodologies using diagrams and tables.In the second part, non-quantifiable results such as the focus or phenomenon under scrutiny as well as essential theoretical points and major findings extracted from the 46 reviewed contributions are presented in a more narrative way to pay tribute to the complex and rich nature of qualitative research results.The discussion section, then, assumes a conceptual perspective by synthesising the main concerns and themes of the contributions and by critically reflecting on the benefits and limitations of linguistic coaching process research in general.

Quantifiable results
Figures 2 and 3 illustrate the number of authors per included contribution (N = 46 in total) and the number of different contributors to those contributions (N = 26).Results show that most contributions were written by a single author (N = 27), while only one contribution was co-authored by four authors and none by more than four (see Figure 2).Figure 3 illustrates that 26 different authors are responsible for the body of research included in the scoping review, with most of them (N = 18) authoring a single paper.One author was involved in more than five publications (i.e., 31 in total) as single or coauthor.
Figure 4 gives the number of publications on linguistic coaching research per year and decade included in the review.The years 2011 and 2016 with 7 and 5 contributions  respectively constitute peak years, while no contributions were published in 2009 and (until the beginning of March) 2023.
Figure 5 visualises findings regarding the (specific) type of coaching within the overall area of business, executive, or workplace coaching referred to by the authors: most contributions focus on executive coaching (N = 21), followed by business (N = 7) and management (N = 5) coaching.Two contributions relied on different corpora and thus 2 formats (career and workplace coaching).Some authors did not provide a clear definition of the type of coaching (N = 4) though the overall placement within workplace coaching could be inferred via quotes, data, or sources used.Additionally, Figure 6 showcases the kind of interaction formatwhether face-to-face (N = 44) or distant (e.g., via telephone or computer-mediated communication) (N = 2)chosen for investigation.Since the default case is face-to-face interaction, texts not clearly stating another focus were included here.The results offer a clear orientation towards real-life, face-to-face encounters.
Finally, methodological aspects of the included items were quantitatively assessed (see Figures 7-9).From the 46 included contributions, most were empirical (N = 22), 14 were theoretical and 10 were mixed in the sense that they used analyses of authentic examples to illustrate concepts, phenomena, or arguments (Figure 7).
From the empirical and mixed (N = 32) contributions 4 , most (N = 12) apply an integrative discourse analysis, i.e., a mix of discourse analytic methods and research-relevant linguistic concepts.This is closely followed by a Conversation Analytic (CA) or a CA-based methodology (N = 11). 5Other methods include (Close) Discourse Analysis (N = 4) and (Critical) Discursive Psychology (N = 4).In one contribution, authors also carried out a multimodal analysis (see Figure 9).The final point of comparison concerned the amount or scope of data used as a basis, i.e., corpus, for the linguistic analysis.Figures 10 and 11 document this continuum within empirical (N = 22) and mixed articles (N = 10).While all researchers use authentic, dyadic language data (from face-to-face, chat and telephone coaching sessions), the continuum spans from in-depth analyses into a single session or process to using larger corpora as a basis for linguistic analysis or exemplary illustration.

Non-quantifiable Results
In order to (a) pay tribute to the in-depth analyses and considerations of linguistic, qualitative approaches and (b) to adequately collect and map its key themes and findings for the very first time, non-quantifiable results are presented in a more narrative fashion (see  also Willemsen et al., 2023) in Tables 1-3.This seems necessary due to the nature of the research, which in its focus on the variability and complexity of individual phenomena is difficult to break down into numbers, a few keywords or themes.This way, we seek to avoid the risk of being 'too relativistic (not having a clear method) but with insightful findings or too realistic (being very methodical) but with superficial results' (Timulak, 2009, p. 593).Such a course of action also adheres to the research field's disciplinary vision.
Table 1 synthesises the central arguments made in the theoretical works (see *Deplazes et al., 2018;*Graf, 2007a*Graf, , 2007b*Graf, , 2008*Graf, , 2011b*Graf, , 2015a*Graf, , 2016*Graf, , 2018   Tables 2 and 3 focus on empirical and mixed articles (N = 32).These explore a plethora of phenomena and foci ranging from the macro structure of the conversation or process, to the construction of roles or (gender, sexual, professional) identities at the linguistic meso-level, to important micro-phenomena such as questions.22 contributions address the macro (N = 11) and meso (N = 11) levels of coaching, while 7 focus on the linguistic micro-level.Three carry out analyses at several analytical levels.Table 2 presents a thematically arranged overview of the foci and phenomena, ordered according to the types of contributions and levels of analysis.
Finally, Table 3 succinctly summarises key findings and messages of the above-mentioned research so as to give a comprehensive overview of the existing knowledge on business, executive and workplace coaching gathered via linguistic analyses.It complements Table 2 in that it can be consulted to gain further insights into the various research foci and to gain a quick overview thereof.The findings are again thematically arranged and organised according to the level of analysis; similar research was condensed and subsumed under one heading.

Discussion: critical reflections on existing linguistic coaching research
This scoping review of linguistic research on business, executive or workplace coaching published since 2000 has analysed 46 contributions.It complements De Haan's (2019) systematic review of qualitative yet psychological coaching literature and thereby constitutes the first of its kind offering a hitherto unprecedented overview of linguistic research.Having given detailed insights into the complexity and diversity of the existing body of linguistic research in the previous chapter, we now zoom out to synthesise its most Table 1.Key arguments from theoretical contributions.
1. Coaching as a co-constructed, person-centred, process-oriented, and systemic interaction type emerges, inter alia, in and through coaches' and clients' talk.2. Communication is thus at the heart of coaching and client change is facilitated through this talk-in-interaction, which speaks for (more) linguistic research into the actual practices of 'doing coaching'.3.Such analyses, regarding communicative goals or discursive practices, do more than simply describe the interaction.
They also consider the interactional context and the coaching-specific function of linguistic actions (such as questions), i.e., their interaction-type specificity, and how these are negotiated between coach and client.4.This allows carving out differences and commonalities between coaching and other counselling formats on the linguistic micro-, meso-and macro-level.5. On the macro level, for instance, a linguistic, process-and activity-based view has led to the conceptualisation of coaching as transpiring along recurring Basic Activities rather than in temporally ordered, stepwise, and static phases.6. Linguistic coaching process research encompasses various methods and approaches with different advantages and limitations: 6.1.A discourse-analytic approach enables an analysis at three interconnected discursive levels of coaching: (1) coaching as a co-created conversation; (2) coaching as an organisational intervention that is constructed and legitimised by different stakeholders; (3) coaching as a socio-historical macro-phenomenon embedded and influenced by dominant (professional, economic, political, social) discourses and power structures.6.2.A conversation analytic approach details the turn-by-turn, sequential organisation of the contributions by coach and client and the (un-)successfulness of interventions.For instance, the contribution of questioning sequences (i.e., question-answer-response sequences) to local, moment-by-moment change processes become empirically observable.6.3.An interdisciplinary approach combining, for instance, normative-theoretical psychology and descriptivephenomenological linguistics augments existing models and theories as exemplified in the Turn-Sequence-Phase-Process (TSPP) model that serves as a blueprint to analyse 'what', 'how', 'how often' and 'when' of different phenomena in coaching.7. Linguistic, inter-and transdisciplinary insights may enrich coaching practice and support its standardisation and professionalisation.
important themes before critically reflecting on its benefits and disadvantages.Following the discussion (4.1),we elaborate on research gaps and desiderata (4.2.) and end with some limitations of this review and concluding remarks (4.3).Focus Findings

Macro-level analyses
The Basic Activity model of coaching / the discourse of (management and executive) coaching ( *Graf, 2011a*Graf, , 2017a*Graf, , 2017b*Graf, , 2019;;*Graf et al., 2010) .The heuristic meta-model of the 'Basic Activities of Coaching' has mapped the 4 major constituents of coaching (Defining the Situation, Building the Relationship, Co-Constructing Change, Evaluating the Coaching) by analysing what 'doing coaching' actually means in the interaction.The model conceptualises coaching not in phases, but in repetitive and interdependent activities that occur in loops and whose realisation accounts for the discourse of coaching.The activities are realised via language-based practices that allow tackling various communicative tasks (e.g., 'Securing Transfer') and their core components. .The four Basic Activities in their specific interrelation and formation on the structural, thematic, and interactive level allow for the professional format coaching and differentiate it from, e.g., psychotherapy. .A typical coaching morphology emerges regarding the Basic Activity 'Defining the Situation'.The step-bystep negotiations and recurring positionings of coach and client within the core task 'Methodological and Procedural Framing' form an endemic part of the coaching process. .In a meta-discourse coach and client need to negotiate their understanding of coaching as well as their social and discursive identities/roles therein.Coaches use considerable time to provide orientation and establish a common ground regarding their encounter and its procedure because they cannot make use of pre-existing knowledge or pre-defined role constellations.
The Basic Activity 'Co-constructing Change' (*Graf, 2015b; *Graf & Jautz, 2019) .Analysis illustrates how the core of change communication, i.e., successfully working on the client's concern, is interactively and thematically linked to other Basic Activities (inter-dimension) and how coach and client tackle its core tasks (diagnosing, intervening, and securing transfer) (intra-dimension).It reveals the clientoriented and processual character of 'Co-constructing Change'. .From a linguistic perspective, the effectiveness of the interaction, i.e., client change, can be evaluated by looking, for instance, at the (un-)successful processing of the communicative task 'intervening' within the Basic Activity 'Co-constructing Change'.By analysing the introduction of themes and their interactive processing in one first session, the coach is shown to be very directive and dominant, which leads to misunderstandings, irritation, dispreferred responses, client withdrawal and other communicative troubles that constantly need repairing.Overall, interactants display a lacking consensus regarding the concern/goal.
The Basic Activities 'Building the Relationship' and 'Defining the Situation' and their role in working alliance and client design ( *Graf & Jautz, 2022) .Focusing on the Basic Activities 'Building the Relationship' and 'Defining the Situation', a contrastive analysis of two processes investigates how coaches initiate the coaching, how clients voice expectations regarding the procedure, and how participants negotiate these expectations.Analysis reveals that the specific working alliance requires great flexibility from the coach.Coaches must discursively manage knowledge discrepancies and contestations as well as different role expectations.They therefore customise their informing practices and attune to clients' experiences, emotional and relational needs, and expectations.The working alliance emerges as an adapting discursive achievement.
Negotiating and conceptualising 'help' in executive coaching (*Graf, 2021) .Due to tensions arising from vague understandings of coaching and its localisation between consulting and therapy, coach and client actively and explicitly negotiate and topicalize their understanding of 'help' in coaching interactions, as well as their roles, responsibilities, and (professional) expertise.The coach thereby uses 3 main strategies to orient the client and construct coaching as 'help for self-help': information delivery sequences, dispreferred reactions to coaching non-conforming client projections, and individual client design.
The (re-)production of gender ideologies and discourses in coaching (* Graf & Fleischhacker, 2020) .Gendered discourses, assumptions and representations are omnipresent/-relevant in coaching, where hegemonic discourses of gender differences are often interwoven with ideas of organisation and leadership.However, the gendered issues are not explicitly addressed in the analysed coaching conversation, but mainly implicitly (re-)produced by coach and client alike.Individualising such structural problems puts additional pressure on the female client and is counterproductive for empowerment and professional development.
Feminine discourse in coaching vs. psychotherapy ( *Pawelczyk & Graf, 2011) .In coaching and psychotherapy, similar communicative strategies that have been linked to the private sphere of female friendship and symbolic feminine discourse are applied to enable client change.They are used to foster rapport and build a relationship that allows for intimacy, disclosure, and self-reflection as steppingstones towards change.Strategies include mirroring, (re-)focusing on emotions, expressing empathy and the creation of spaces that allow for less dominant discourses and self-presentations. .Focusing on the role of emotions, i.e., whether or not they are topicalized and how they are processed via discursive practices, eight conversational patterns used by coaches to work on emotions are found.They either reinforce/emphasize or control/de-emphasize emotions within the interaction.Patterns used depend on the degree of displayed affect by clients and the coach's communicative goal. .Narratives of illness, distress or emotions are constitutive of coaching as psycho-managerial discourse.They are frequently employed (a) when clients present their concern and stage their identities, (b) when they are implicitly or explicitly elicited by coaches to process the concern, or (c) to transfer insights from private narratives to professional contexts.These emotional experiences are, however, always used strategically for business-related goals. .While coaching and psychotherapy employ similar strategies to engage in emotional work, the functionalization of 'feelings talk' for the organisation or the client's professional goals is constitutive of coaching.Emotional learnings are transferred from the private to the professional realm to enhance the client's performance, or to achieve the overall goal.

Meso-level analyses
Topicalizing client change in coaching ( *Graf, 2011c) .Change is the communicative goal in coaching.It entails changed ways of speaking about concerns as well as the explicit topicalization of (ongoing) change processes.For this, coach and client use different strategies to co-construct talking about change, e.g., narratives about changed behaviour or positive evaluations thereof. (Continued)

COACHING: AN INTERNATIONAL JOURNAL OF THEORY, RESEARCH AND PRACTICE
Table 3. Continued.

Focus Findings
While the clients remain the experts of their development, the topicalization is initiated and supported by the coach via questions, extensions, alignment, metapragmatic comments or positive ratifications.
(Re-)structuring client knowledge and addressing client (dis-)alignment ( *Sator & Graf, 2014) .This analysis focusses on knowledge management during solution development, a core task in coaching.It reveals client resistance during goal elicitation and the development of concrete steps towards said goal.The client's resisting strategies include agency, disengaging with the (thematic) agenda, or changing the modality, while the coach either endorses or challenges such resistance or disalignment.Non-directive, affiliative attempts to overcome resistance (as suggested by coaching practice), however, remain ineffective.
Interaction management in chat coaching ( *Winkler, 2017) .Interaction management, another core element of successful coaching, operates at different levels of the interaction.It relates to organising the interaction via turn-taking or (local/global) structuring, steering the thematic unfolding, e.g., using reformulations, and managing the action, e.g., via questions.Conversation Analysis is used here to identify trainable moments and potential challenges in chat coaching.
Symmetry / asymmetry and the 'dialogue at eye level' ( *Jautz, 2017) .Identity constructions via self-and other-positionings are also important for relationship building.The situated, relational identities 'coach' and 'client' thereby necessitate various corresponding activity identities such as 'guidance giver' and 'guidance taker'.Within these, yet again, different discursive identities, e.g., 'explainer' or 'questioner', can be found.A proposed model of situated, activity, and discursive identities helps to describe these complex, mutual positioning processes. .The construction of competence in coaching is based on yet another relational pair, that of 'expert' and 'nonexpert/layperson' with shifting areas of expertise.Four competence domains for coaches, i.e., professional expertise, knowledge of the procedure, empathy, and experience, are identified; these are continuously constructed via different activity identities of both coach and client.

Micro-level analyses
Questions requesting examples in coaching vs. psychotherapy (*Spranz-Fogasy et al., 2019) .Though examples are often used in coaching, questions that elicit them are infrequent since clients often volunteer them.Clients thereby show little resistance (compared to psychotherapy) to work with examples and dive deeper into the subjective realms of experiences and feelings, demonstrating a greater willingness to tackle and change (problematic) behavioural patterns.
Questions for solution generation in coaching vs. psychotherapy (* Kabatnik & Graf, 2021) .Solution-focused questions (SFQ) are four times more frequent in coaching than in psychotherapy.While formal, thematic, and functional characteristics display considerable overlap, in coaching, SFQ are asked from the very beginning, they can also be used to co-construct next steps and are often very concrete.Their primary function is to support clients to independently generate own solutions.Clients are thus attributed a higher degree of agency and decision-making competence.The frequency of SFQ in coaching makes further differentiation into sub-types necessary.
Semi-responsive answers and coaches' reactions in third position ( *Winkler, 2022) .Semi-responsiveness plays an important role in revealing clients' ambivalences, conflicts etc. in coaching interactions.Semi-responsive answers are identified and analysed using a conversation analytic, empirically developed manual based on the categories: 'topic', 'intention', and 'linguistic markers' such as modality.
Coaches' reactions to semi-responsive answers include insisting on the initial question, repeating the question via repair or rephrasing, or abandoning the question altogether.This suggests yet another type of responsiveness in third positions.Focus Findings hand, spark moments of realisation and provide immediate possibilities for new interpretations.The changes on the linguistic surface level may then be indicators of (immanent) client change.
Lexical and gestural particularities of career coaching (* Konat & Juszczyk, 2015) .The lexical analysis of (Polish) career coaching sessions revealed an overuse of verbs describing mental states, desires and feelings, words indicating a speaker's uncertainty, and abstract referents (such as work, aim) compared to other corpora.Four co-speech gestures accompanying key words are also analysed focusing on function as well as temporal alignments between hand movements (e.g., grasping an imaginary object) and words.
Linguistic features of telephone coaching (*Lynden & Avery, 2016) . Based on three themes identified as relevant by participating coaches, research finds that 'building rapport and attending to the coachee' ( 1) only via audible cues sometimes resulted in difficulties.Changes in the tone of voice, however, can serve coaches as feedback prompts.Due to the lack of visual cues, 'attendance to awkward silences' ( 2) is also essential in telephone coaching.Coaches thereby differentiated between (not) awkward types of silences using breathing, tone or pitch as indications, and clients often avoided longer pauses altogether.'Offering challenges' (3) was done differently in telephone coaching interactions, compared to other formats.Instead of using the first name of clients as preface or signal to a challenge, it was placed at the end of the turn by coaches, as a softened call for attention.
Functional listening in coaching (* Albrecht & Perrin, 2016) .Functional listening is another communicative practice of coaches.It means supporting client change via verbal withdrawal and a lack of intervention, which allows the coachee to expand on own reflections.To find the right place for this practice, it is necessary to pay attention to incongruences regarding (non-)verbal signals, the meaning level of feelings and thoughts, as well as the action level of the interaction.

Multi-level analyses
The multifaceted relational work in coaching (*Behn-Taran, 2014) . Relational work in coaching is co-constructed, dynamic, and multi-faceted occurring simultaneously at different levels of the coaching interaction.At the micro-level, interventions such as questions, feedback practices, or advice are important tools to steer the conversation, manage the flow of information or activate coachees.At the meso-level, issues of identity construction, self-presentation and face work are at play.On a macro-level, relational patterns emerge as quite unique with dyads displaying considerable variety in their communicative systems.
The interplay of managerial and psychotherapeutic discourses ( *Schulz, 2013) others within these, as well as the relational and processual development over time.On the macro-level, coaching is found to apply both managerial and therapeutic discourses and methods.On the meso-level, 5 different discourses of problem construction are found that either locate the problem outside or inside of the managers The reoccurring discursive devices used by the coach on the micro-level of the conversation form 3 categories: interpreting backwards, forwards interpreting and setting interpreting.
The discursive unfolding of emotion management in coaching ( *Schulz & Steyaert, 2014) 4.1.Synthesis of theoretical and empirical findings The theoretical linguistic contributions reviewed in this article predominantly address three concerns: (a) defining and carving out coaching as a distinct helping professional interaction, (b) arguing for the relevance of linguistic research, and (c) distinguishing the linguistic approach from other research fields and (methodological) approaches.From a linguistic perspective, coaching is defined as a co-constructed communicative event, i.e., a helping professional encounter, that emerges in-situ in and through the consecutive contributions by coach and client.Via linguistic, bottom-up analyses of these interactions recurring communicative practices and language strategies to tackle important tasks in coaching (e.g., building the working alliance or establishing the coaching realm) can be identified.While some practices are typical for professional encounters in general, others are coaching-specific and offer insights into its interaction-type specificity.This allows to distinguish coaching from other professional formats, such as psychotherapy or counselling.Since coaching is communicative in nature and predominantly based on talk-in-interaction, linguists argue for the necessity and the empirical as well as the practical value of further in-depth research into the (language-based) mechanisms that make up coaching.Due to its flexible, data-driven, and exploratory nature, linguistic analysis can, for instance, carve out phenomena that coach and client orient to, but that have not been addressed or made relevant in professional theory or in deductive quantitative research.Using linguistic insights as a basis for evaluation, quality standards or training programmes, such findings may contribute to (further) promote an evidencebased coaching.In contrast to other research approaches, linguistic analyses are based on authentic language data.Such data reveal the actual unfolding of the interaction, unlike pre-and post-intervention interviews or questionnaires.Thus, by looking into the activities of coach and client, a true process-orientation and reconstruction becomes possible that allows for a meticulous description of the structural, interactional, and thematic components of coaching.On the macro-level, this has already evinced a different conceptualisation of the overall structure of coaching processes in terms of Basic Activities instead of static phases.On the linguistic micro-level, the turn-by-turn set-up of the coaching interaction allows to assess the local effectiveness of interventions such as questions and helps to understand how client change emerges and becomes empirically observable via transformative sequences along and across sessions on a moment-bymoment basis.
The empirical linguistic research puts theoretical reflections and arguments into practice thereby operating at different levels of focalisation (i.e., macro, meso, micro levels).Research predominantly focusses on the interaction-type specificity of coaching, the coach-client relationship as well as client change, the communicative goal of coaching, alongside the tasks, communicative practices and strategies contributing to it.Various empirical articles carve out the endemic features of what makes up the (overall) interactional structure and the discourse of coaching.At the macro and mesolevels, coaching draws on managerial and psychotherapeutic meta-discourses.It also becomes a 'playing field' for discursive power struggles and ideologies regarding, e.g., gender/sexuality, leadership, or emotionality.Coaching, however, is also constructed as a distinct phenomenon at the micro-level of the coaching conversation by addressing specific features and interventions such as functional listening, solution-oriented questions, requesting examples, attending to silences etc.Sometimes labelled as the success factor in psychological coaching outcome research, the working alliance and the underlying coach-client relationship also receive special attention in linguistic analyses.Conceptualised as a co-constructed, momentarily shifting, and flexible discursive achievement by both coach and client, (dis)alignment, (non-)hierarchy, (a)symmetry, (non-)directiveness, client design, as well as (professional) identity and role constructions are the heart of such empirical contributions, carried out again at different levels of granularity.The co-construction of change as well as change processes for clients are empirically addressed in different ways, too.Various contributions look into some of coaches' major communicative tasks and the discursive practices involved that contribute to goal achievement, client change, and development.Apart from explicitly topicalizing client change, these include managing the interaction, (re-)structuring knowledge, topicalizing concerns, or addressing and managing emotions.On the micro-level, the impact of certain interventions, e.g., questions, on the local effectiveness of the sequentially organised coaching conversation are also investigated.
Linguistic research thus generates relevant and novel findings regarding the coaching process, i.e., the 'black box' of coaching.The analyses cannot only document, analyse, and describe the concrete linguistic realisation of established concepts such as 'active listening' or 'dialogue at eye level', but may also reveal interactive phenomena not yet mentioned in coaching practice literature or research (e.g., the methodological, procedural, and temporal framing of coaching).Concerning established success factors (e.g., the working alliance), empirical insights into 'how' and 'where' these are communicatively and linguistically co-constructed by coach and client on the interactive micro-level can be provided.Regarding coaching practice, linguistic findings display great potential for coaches and training institutions to (a) identify trainable moments, best practice, or potential interactional challenges, (b) to carve out evaluation or quality criteria, and (c) to increase professionals' language awareness and conversational skills via sensitising for the spectrum of communicative options found in naturally occurring coaching interactions instead of normative or prescriptive instructions.
However, authentic data generation and linguistic transcription as well as in-depth, iterative, qualitative (re-)analyses are very time-consuming and generating findings takes time.For that very reason, only small data sets are analytically manageable and case studies are common.No generally valid statements can be made, nor can linguistic research focus on frequencies and significant results in ways that psychological research can.Similarly, due to their focus on the local interaction, qualitative linguistic analysis cannot assess the global effectiveness of coaching measures.Given that conversations are (to some extent) unique social events, replicability does not play a role in linguistic research either.While this is not considered problematic from a qualitative perspective, whose exploratory empirical interest lies in reconstructing the social event in a detailed and bottom-up manner, the standards of quantitative research, i.e., objectivity, validity, and reliability, are not met.For those focussing on the 'bigger picture' of the coaching process, investigating (very specific) isolated interactional phenomena such as questions requesting examples may at times seem insignificant.However, such phenomena may still play a decisive role when carving out change-inducing or interaction-type specific coaching practices.
In sum, considering the existing body of linguistic coaching process research, both the theoretical and empirical contributions speak to its status as promising, though still developing, research area.It presents itself as fragmented and little systematic, with few references to and building on previous research and findings.The still limited number of researchers and studies, as well as the diversity of linguistic findings prevent the drawing of robust conclusions (for theory building and modelling) at this point.Most importantly, as is natural for a relatively new field, many authors are concerned with arguing for the relevance and benefits of their research and with finding an adequate place within a larger research framework.While early contributions embedded linguistic coaching research within the more linguistic area of professional communication and professional helping interactions, more recent publications locate themselves in the coaching process research paradigm (see Figure 12), developed in analogy with Elliot's (2010) paradigm for psychotherapy (see Wegener, 2019, and Fleischhacker and Graf, under review, for a further differentiation and embedding of linguistic process research).

Limitations of the review
While the aim of this review was to give a comprehensive and full overview of existing linguistic research on coaching, there are omissions and limitations.The necessary reliance on non-linguistic databases may have led to omissions.Publications written in languages other than English and German were not considered for review.The focus on business, executive, and workplace coaching excluded research on other types of coaching, e.g., health coaching.In an attempt to not gloss over the uniqueness of the individual contributions (see Willemsen et al., 2023, p. 17), the summaries and presentations of the qualitative findings in section 3.2 are to some extent subjective and do not allow for exact replicability.Moreover, since this constitutes the first scoping review on linguistic coaching research, further conceptualisation, theory building and modelling was not the aim at this point.

Recommendations for future research and conclusion
The following research gaps and desiderata emerged from the scoping review: Since most projects were dedicated to the macro and meso levels of coachinga necessary requirement for new research fields -, more research is now needed on the multitude of interactional micro-phenomena, as well as on the sequential organisation and processing of coaching interventions apart from questions, such as formulations, interpretations, or extensions.Some unfinished project include detailing all Basic Activities as carved out by *Graf (2019), since the Basic Activity 'Evaluating the Coaching' has not yet been empirically investigated, or addressing other identity factors apart from gender, such as ethnicity, age, etc. Especially in post-pandemic times with an increase in digital counselling, (more) research on digital and computer-mediated coaching is needed.Both digital and face-to-face coaching research could be complemented with a focus on multimodality, i.e., on non-verbal communication, gestures, facial expressions etc.Additionally, more comparative research with helping professional formats other than psychotherapy is needed to carve out the communalities and differences between these interactions and to further detail the interaction-type specificity of coaching.Apart from intra-disciplinary research, linguistic research could benefit from interdisciplinary projects, e.g., in the context of established success factors with psychology, or from transdisciplinary projects to work on the implementation of findings and the generation of practice-relevant but linguistically-sound training guidelines.
To conclude, as this scoping review documents, linguistic coaching process research has ventured into the 'black box' of coaching (Ianiro & Kauffeld, 2015) despite its empirical and methodological challenges and time-consuming analyses.It is particularly well suited to explore and describe the communicative activities and tasks that take place during the interaction.By looking into the co-constructed coaching conversation as it unfolds in real time, linguistic research can provide valuable insights that cannot be generated with quantitatively oriented, deductive outcome research based on questionnaires or interviews.Though the foundation is laid, (linguistic) process research is still scarce and represents 'one of the most pressing research gaps' (Fillery-Travis & Cox, 2018, p. 529) in coaching research.The contributions reviewed above are pieces of a mosaic yet to be completed and systematically put together.Despite the limitations of qualitative approaches, linguistic analyses and findings should become points of departure for or points of intersection with other empirical approaches to round out our understanding of coaching.

Figure 1 .
Figure 1.Search strategy process for the scoping review.

Figure 2 .
Figure 2. Number of authors per contribution.

Figure 3 .
Figure 3. Number of different contributors / contributions per author.

Figure 4 .
Figure 4. Number of publications per year and decade.

Figure 7 .
Figure 7. Type of research contribution.

Figure 11 .
Figure 11.Amount of data used for mixed contributions.

Figure 10 .
Figure10.Amount of data used in empirical contributions.

.
The co-construction of change, conceptualised as change in the way clients address concerns, is supported via various linguistic, change-inducing interventions.Questions, for instance, are shown to support a gradual, stepwise intensification of an issue, eventually leading to a change in perspective.Keywords, on the other (Continued) COACHING: AN INTERNATIONAL JOURNAL OF THEORY, RESEARCH AND PRACTICE

Figure 12 .
Figure 12.Embedding linguistic research within a larger framework of coaching research.

Notes 1 .
Since systematic literature reviews are not a common practice in linguistics, data bases focussing on linguistic research do not exist or are a) too small or b) are provided by specific publishers, which again limits the output of the search.2. The asterisk was used to capture both German ('Linguistik', 'linguistisch' etc.) and English contributions.3.These German andEnglish journals included: Coaching | Theorie & Praxis, Organisationsberatung, Supervision, Coaching, International Coaching Psychology Review, International Journal of Evidence-Based Coaching and Mentoring, Coaching: An International Journal of Theory, Research and Practice, The Coaching Psychologist, Philosophy of Coaching.4. The theoretical contributions were excluded for Figure 8, Figure 9, and Table 2.

Table 2 .
Overview of foci and phenomena.

Table 3 .
Summary of key findings of mixed and empirical research.