When entrepreneurship becomes a tool for political hegemony

or used interchangeably, and this perpetuates the assumption that all phenomena labelled as ‘entrepreneurship’ essentially include the same activity (Smit & Pretorius 2022). The conceptual umbrella has been stretched to Orientation: From a critical entrepreneurship perspective, this article examines the potential hegemonic nature of the mainstream entrepreneurship discourse. Research objectives: (1) To determine what mainstream entrepreneurship assumptions and resulting discourses are being reproduced in the South African media discourse. (2) To determine if and how this current discourse is naturalising knowledge claims about ‘entrepreneurship’. (3) To highlight the hegemonic possibilities of producing and reproducing this discourse in a South African context. (4) To provide recommendations that could mitigate the potential hegemonic entrepreneurship discourse in South Africa. Motivation for the study: Everything labelled ‘entrepreneurship’ is not contributing to job creation and economic development equally. From a critical standpoint it is becoming evident that the assumptions underlying mainstream entrepreneurship discourse – especially when entrepreneurship is used as a development tool – have to be challenged. Research design, approach and method: Random purposive sampling is employed by searching English print media articles containing relevant key words that were published in South African English newspapers between 01 January 2018 and 31 December 2018. The sampling frame was obtained from Sabinet through the University of Pretoria Library SA Media database. A critical discourse analysis is conducted on the final sample of 63 articles. Main findings: The South African media discourse replicates the mainstream entrepreneurship knowledge claims that any and all type of entrepreneurial activity is essentially the same and that anything labelled ‘entrepreneurship’ will lead to economic development and job creation. These knowledge claims have become normalised in the South African discourse. The dominance of this mainstream discourse on entrepreneurship is opening the way for political hegemony in a South African context. Practical/managerial implications: The South African scholarly community has to take up its responsibility as actors of social change and challenge the reigning public discourse in the field of entrepreneurship that is resulting in political hegemony. Contribution/value-add: This article shows that the failure to distinguish between different types of entrepreneurship is providing a fertile ground for political hegemony.


Introduction
Research and theory development related to the drivers of entrepreneurship abound and various theoretical currents have emerged (Da Costa & Silva Saraiva 2012).From the earliest theories proposed by Cantillon in the mid-1700s, say in the 1800s and Schumpeter in 1934, the term 'entrepreneurship' has continued to change through different schools of thought and expand over time.It has come to include many different contents, domains, origins and destinations (Baker & Welter 2020;Bögenhold 2020;Filion 1998;Poole 2018).It 'has become a broad label under which a hodgepodge of research is housed' (Shane & Venkataraman 2000).Today, the theoretical framing of the entrepreneurship theory includes a number of diverse phenomena under one, polysemous conceptual umbrella (Smit & Pretorius 2022).The boundaries of this conceptual umbrella have been extended to now include 'more or less all of humanity' (Poole 2018).There is little consensus on what entrepreneurship is (Bernold 2020).Terms included under the conceptual entrepreneurship umbrella are often conflated or used interchangeably, and this perpetuates the assumption that all phenomena labelled as 'entrepreneurship' essentially include the same activity (Smit & Pretorius 2022).The conceptual umbrella has been stretched to such an extent that the word 'entrepreneur' has become an empty signifier -it now means 'both everything and nothing' (Da Costa & Silva Saraiva 2012).
However, this confusion surrounding the definition and boundaries of the entrepreneurship umbrella is largely ignored (BöPenhold 2020;Smit & Pretorius 2021;Talmage & Assert 2020:318).This terminology confusion and conflation inherent in entrepreneurship theory is not only promoting a weak paradigm for research but it is also influencing economic development policy (Pretorius et al. 2021;Smit & Pretorius 2021).Based on the perpetuated assumption that anything and everything included under this conceptual umbrella has equal potential to contribute to economic development and job creation, government decision-makers appropriate entrepreneurship as a development apparatus (Pretorius et al. 2021;Smit & Pretorius 2020).This view is supported by non-critical, mainstream entrepreneurship publications worldwide that hail the entrepreneur as a superhero for economic development and job creation (Ahl & Marlow 2021;Audretsch & Moog 2020;Luiz 2010;Pretorius et al. 2021;Williams & Nadin 2012).Furthermore, it is also supported by the South African literature that describes entrepreneurs as 'playing an important role in most businesses.They contribute significantly to employment, job creation and wealth creation' and that 'economic development can be directly attributed to the level of entrepreneurial activity in a country and entrepreneurial businesses ensure growth in the economy' (Nieuwenhuizen 2019).The prevailing argument is that everything included under the conceptual entrepreneurship umbrella is the saviour of economic development and job creation.

Critical entrepreneurship studies
When a concept, such as 'the entrepreneur', has been clouded with such an aura of mystification and superhero status, it creates a fertile ground for critical study (Fournier & Grey 2000).However, entrepreneurship has hardly been investigated from a critical perspective (Da Costa & Silva Saraiva 2012), especially from a South African perspective where critical scholarship remains largely underexplored (Goldman 2021).Authors in critical entrepreneurship studies argue that the current theoretical conceptualisation of entrepreneurship is in fact standing in the way of several mainstream assumptions being questioned (Ahl & Marlow 2021;Calás, Smircich & Bourne 2009;Spicer 2006).Furthermore, focusing on entrepreneurship as a desirable activity and 'the entrepreneur' as a superhero obscures important questions from being asked (Tedmanson et al. 2012).One such question is how this ideology of entrepreneurship may be contributing to relations of power and hegemony in a society.
Classic entrepreneurship research views 'the entrepreneur' as an agent of the capitalist economic system (Ogbor 2000).These mainstream perspectives on entrepreneurship aim to reproduce a capitalist ideology (Calás et al. 2009;Goldman 2021).Capitalism, and more specifically neo-liberalism, guarantees riches without limits for everyone (Fairclough 2013).It constructs a new, agentic citizen who can capitalise on market opportunities to become a successful entrepreneur (Ahl & Marlow 2021;Da Costa & Silva Saraiva 2012).What is more, entrepreneurship is hailed as a noble cause, 'the attitude of a people who seek the social and economic development of their country' (Da Costa & Silva Saraiva 2012).The word 'entrepreneur' thus evokes an image of a superhero, the cornerstone of economic growth, the saviour of the job-creation crisis (Smit & Pretorius 2020;Williams & Nadin 2012).Globally, however, current neo-liberalist policies are leading to greater economic and social inequality, widening the gap between the rich and poor and placing unsustainable pressure on the environment (Da Costa & Silva Saraiva 2012;Fairclough 2013).Societies are becoming increasingly disillusioned by neo-liberalism and the current system needs to be repaired or replaced (Fairclough 2013).However, critical studies that militate against the false promise of neo-liberalism -even and especially in a South African context -are still the exception and entrepreneurship is increasingly sold as the answer to inequality and poverty (Ahl & Marlow 2021;Goldman 2021;Smit & Pretorius 2021;Tedmanson et al. 2012).
As discussed earlier 'entrepreneur' is by no means a homogenous concept.From a critical standpoint, it is becoming evident that 'entrepreneurship' cannot be used as an umbrella term, assuming that any and all types of entrepreneurs will equally contribute to economic development and job creation (Hartmann, Krabbe & Spicer 2019;Kuada 2015;Nightingale & Coad 2014;Smit & Pretorius 2022).However, the majority of mainstream contemporary entrepreneurship research continuously seems to conceptualise it as a homogenous concept and thus protect a number of mainstream assumptions in the field of entrepreneurship from being questioned (Ahl & Marlow 2021;Calás et al. 2009;Pretorius et al. 2021;Smit & Pretorius 2022).
Seminal critical management authors argue that the mainstream approach in entrepreneurship research imposes a priori and taken-for-granted definitions and meta-theoretical assumptions onto an ambiguous social reality (inter alia) (Achtenhagen & Welter 2007;Alvesson & Deetz 2000;Alvesson & Willmott 2003;Calás et al. 2009).The bulk of the empirical studies performed to determine the link between entrepreneurship and economic development have been conducted in Global North economies in North America and Europe, and the findings of these studies have very limited use for answering questions about economic development in the Global South (Naudé 2011;Smit & Pretorius 2020).In the theoretical assumptions underlying the mainstream entrepreneurship theory, the vast contextual differences between Global North and Global South settings are largely overlooked (Baker & Welter 2020;Verver, Roessingh & Passenier 2019).
Entrepreneurship is hailed as a one stop shop for job creation and economic growth, yet it is continuously failing to create significant empowerment or upward mobility for the marginalised communities it claims to emancipate (Honig 2017;Naudé 2011;Nightingale & Coad 2014;Shane 2009).Existing policies in relation to entrepreneurship development are failing to create jobs and alleviate poverty (Edoho 2016;Hartmann et al. 2019;Pretorius et al. 2021).
Voices critical of mainstream entrepreneurship theory agree that the conventional discourse eulogises 'the entrepreneur' as the saviour of economies and the champion of job creation worldwide (Achtenhagen & Welter 2007;Ahl & Marlow 2021;Alvesson & Deetz 2000;Alvesson & Willmott 2003;Calás et al. 2009;Smit & Pretorius 2021).What is not yet known is how this mainstream entrepreneurship ideology may be contributing to political hegemony in a South African context.This article joins the fledgling critical entrepreneurship voices by examining the potential hegemonic nature of the mainstream entrepreneurship discourse, specifically when it is used as a development tool in a South African context.
To examine this research problem, this article will proceed with a short discussion on hegemonic discourse and how it can be used as a tool for political manipulation.

Hegemonic discourse
Language is not merely a collection of a few words that are strung together unintentionally.It is also not always neutral and objective.By choosing a specific word over another or omitting some details in a description while including others, a specific version of the world is intentionally created (Rapley 2011).Language can thus serve 'as the mediator for constructing reality' (Achtenhagen & Welter 2007).Consider this example by Smit and Pretorius (2022): In 1990, Nelson Mandela is freed after 27 years in prison.One newspaper headline reads, 'Freedom fighter finally released after 27 years'.Another headline reads 'Terrorist released after only 27 years'.The two headlines not only present two vastly different realities, but they also create two vastly different realities.(pp. 721) The French philosopher Foucault introduced the term discourse as a discursive structure, a system of thought or knowledge claims, which assume an existence independent of a particular speaker (Foucault 1972(Foucault , 1980;;Stoddart 2007).Discourse thus not only refers to the words we use but also how these words are used in the context of embedded social practices (Gee 2004).Through the systematisation of ideas and opinions, certain ways of talking, thinking or actingcertain social practices -are institutionalised, regulated and normalised (Jäger & Maier 2009).Normalised ways of talking and acting then create the impression that there are 'appropriate' (and 'inappropriate') ways of talking and acting (Gee 2004).These normalised ways of talking can be used to craft realities -because the way we speak or write constructs that very reality (as illustrated by the given example).Discourse can thus be used intentionally to craft what is considered appropriate (and thus acceptable) versions of a normalised social reality while excluding others (Greckhamer & Cilesiz 2014).
Discourses are mainly produced, circulated and embedded through three mechanisms: (1) institutions and organisations, (2) individuals and (3) inter-individual interaction and, finally, (4) the media (Achtenhagen & Welter 2007).Societies take up these transmitted discourses and incorporate them into their subjectivities -all critical consciousness towards the merits of the discourse is dissolved.The social network becomes convinced that the accepted status quo has always been the norm, it is how it should be and it cannot be nor should it be changed.Because the status quo is viewed as appropriate, the subordinate class remain uncritical and politically passive (Stoddart 2007).The need for coercive actions is mitigated and power is created through hegemony.To ensure that it maintains hegemonic power, dominant discourses prevent oppositional discourses from taking root (Brigg 2002).In a South African context where entrepreneurship is continuously hailed as the silver bullet for job creation and economic development, it has to be recognised that the (intentional) production and circulation of a specific discourse can be used as a mechanism of social power (Foucault 1980).

A critical approach to entrepreneurship discourse
In South Africa, as in the rest of the world, much has been written on how individuals can increase their economic power by becoming 'entrepreneurs'.As Ahl and Marlow (2021) writes, 'the foundational neoliberal market logic 'releases' the individual to exploit their potential through an entrepreneurial way of being'.However, less than 25% of start-up ventures in South Africa survive past the first 3 years of existence (Haltiwanger, Jarmin & Miranda 2013).Furthermore, 80% of so-called entrepreneurs in the informal sector work only for their own account and are merely selfemploying (Small Enterprise Development Agency [SEDA] 2016).South Africa's total early-stage entrepreneurial activity (TEA) remains below average for the African region and its business discontinuance rate is also higher than the established business ownership rate for the same period (Bowmaker-Falconer & Meyer 2022), implying that 'there are more businesses being closed, sold or otherwise discontinued than there are businesses being continued' (Bowmaker-Falconer & Herrington [2019] 2020).At 29.1%, South Africa's unemployment rate is the highest amongst all the member states of the Brazil, Russia, India, China and South Africa (BRICS) bloc, and one of the highest in the world (The World Bank 2018).If one would include the discouraged work effect, the unemployment rate in South Africa is actually closer to 38.5% (Bowmaker-Falconer & Herrington [2019] 2020).In spite of the visible failure of start-up ventures to survive past the first three years or move beyond mere self-employment, entrepreneurial promotion activities such as training, incubation and funding are continuously being endorsed and promoted by political role players (Honig 2017), especially in marginalised communities with little or no other options for employment, assuming that high levels of 'what is labelled', entrepreneurial activity will necessarily lead to economic growth and job creation (Nightingale & Coad 2014;Smit & Pretorius 2020).Honig (2017) extends on this phenomenon by introducing the term compensatory entrepreneurship, which refers to: [T]he political endorsement of entrepreneurial promotion activities, including training, incubation, and media dissemination, for the primary objective of maintaining political and/or economic control of one population over another.
Compensatory entrepreneurship is used as a 'method of promoting symbolic justification for inequity of economic and political power' without creating alleviation or upward mobility for the beneficiaries it is aimed at.The mainstream entrepreneurship discourse drives and normalises the assumption that the beneficiaries of entrepreneurship development initiatives will become successful entrepreneurs who contribute to job creation and economic development.Through the implementation of these initiatives, especially in marginalised communities where unemployment is rife, the responsibility of creating jobs is thus shifted from political actors to the supposed entrepreneur (Honig 2018).
The entrepreneurship ideology is thus used and manipulated to reach a political goal (Honig 2017) -political actors reckon that they have implemented entrepreneurship development initiatives and the beneficiaries' failure to capitalise on these initiatives are because of their own shortcomings.In Foucauldian terms, the entrepreneurship discourse is used and manipulated to maintain political hegemony.The normalised discourse convinces the marginalised and unemployed that their inability to become successful entrepreneurs following their participation on an entrepreneurship development initiative is a result of their own shortcomings.Because the political elite can now claim that they did attempt to emancipate the unemployed and create jobs through 'so-called' entrepreneurship interventions, they become exempted from any further responsibility towards these marginalised masses.When the beneficiaries take up this transmitted discourse and accept it as appropriate, the status quo -poverty, inequality and unemploymentbecomes acceptable.The subordinate classes then have no reason to protest against the ruling elite, because they are convinced that the social and economic inequality they experience is because of their own inability to realise the entrepreneurship dream.This article proposes that compensatory entrepreneurship is an example of a hegemonic entrepreneurship discourse.
To confirm this, one would need to determine if the entrepreneurship discourse in South Africa indeed has the characteristics of a hegemonic discourse.These would include answering questions such as: Has critical consciousness towards assumptions in the discourse been dissolved?In other words, has the discourse been normalised?Do different players in the social network seem satisfied that the current social system is organised in the way it should be?Is this dominant discourse preventing oppositional discourses from taking root?When one is interested in the way discourses produce power relations of specific groups over others, a critical analysis of the discourse has to be performed (Gee 2004;Van Dijk 2009).Such a critical discourse analysis (CDA) plays an important role in social and organisational research (Achtenhagen & Welter 2007;Fairclough 2013;Greckhamer & Cilesiz 2014).Critical discourse analysis is also becoming a valuable tool in critical management studies (Ahl & Marlow 2021;Fairclough 2013).In more recent years, the potential of CDA to generate new knowledge in entrepreneurship research has increasingly been recognised (Achtenhagen & Welter 2007).A CDA can contribute to the field of entrepreneurship research by exploring how socially constructed entrepreneurship-related phenomena can have societal implications (Achtenhagen & Welter 2007;Ahl & Marlow 2021).To date, not a lot of research has been carried out to highlight the potential of the entrepreneurship discourse to contribute to unequal power relations, particularly in a South African context where entrepreneurship is widely appropriated as a development tool.The relationship between discourse, ideology and entrepreneurship also remains largely underexplored (Ahl & Marlow 2021;Da Costa & Silva Saraiva 2012;Ogbor 2000).
In South Africa, we have, on the one hand, the statistics of entrepreneurs neither surviving nor contributing significantly to job creation -what Ahl and Marlow (2021) call the false promise of entrepreneurship (Bowmaker-Falconer & Herrington [2019] 2020; Haltiwanger et al. 2013).On the other hand, we have the normalised view that individuals have the responsibility to take advantage of market opportunities and become successful entrepreneursthe vehicle for economic development and job creation (Ahl & Marlow 2021;Smit & Pretorius 2021).This obvious contradiction begs critical analysis of the mainstream discourse on the causal relationship between everything labelled 'entrepreneurship' and job creation in South Africa.This study attempts to add a voice to the nascent field of critical entrepreneurship research by critically examining the mainstream discourse on the causal relationship between entrepreneurship and job creation in order to answer the following question: Is the conventional entrepreneurship ideology producing a discourse in South Africa that is promoting political hegemony?

Agenda and research objectives
Based on previous studies critical to mainstream entrepreneurship research, this article takes the position that the mainstream entrepreneurship discourse is based on a number of assumptions.From this position, it proposes that the mainstream entrepreneurship discourse may be contributing to political hegemony in South Africa.This article has the following specific objectives: • To determine what mainstream entrepreneurship assumptions and resulting discourses are being reproduced in the South African media discourse.
• To determine if and how this current discourse is naturalising knowledge claims about 'entrepreneurship'.• To highlight the hegemonic possibilities of producing and reproducing this discourse in a South African context.
As a CDA should not merely focus on what is 'wrong' within a society or institution, but should also strive to be normative by proposing possible ways of mitigating the identified assumptions and resulting social wrongs identified during the discourse analysis (Fairclough 2013), the final objective of this article will be the following: • To provide recommendations that could mitigate the potential hegemonic entrepreneurship discourse in South Africa.

Research methods and design
This article was orientated by the methodological perspective of CDA.

Critical discourse analysis
No method is without challenges and CDA is no different.Like many qualitative methods, CDA does not follow rigid rules or a formula to be followed to the letter (Fairclough 2013;Greckhamer & Cilesiz 2014;Jäger & Maier 2009).A number of approaches have been described, but none is considered as a superior best practice approach.Furthermore, a comprehensive discourse analysis would ideally look at the history, the present and the future of a specific discourse (Wodak & Meyer 2001).It would identify and analyse all the different discourse planes such as the sciences, politics, media, education, everyday life, business and the like and determine how these different planes influence each other, identify the different ideological positions of the subjects in the discourse and exhaust all the possible discursive fragments that constitute a specific discourse.However, this would be an enormous project that could realistically only be completed by conducting a number of smaller projects.Every small contribution to the overall understanding of a particular discourse creates reliable knowledge that can influence the future direction of a specific discourse strand (Jäger & Maier 2009).More specifically, studying a discourse that produces hegemonic views will contribute to our understanding of the entrepreneurship theory (Ahl & Marlow 2021;Da Costa & Silva Saraiva 2012).This study can therefore be considered as a small contribution to a larger possible critical analysis of the discourse on entrepreneurship presented as the vehicle for economic emancipation and job creation in the Global South.
To overcome these challenges, this study clearly demarcated the sampling of data from the onset.Firstly, this study did not focus on multiple discourse planes, but rather analysed only one discourse plane.Newspaper and magazine articles are considered the most ubiquitous and accessible sources of texts (Rapley 2011) and the media plays an influential part in reinforcing stereotypes about entrepreneurship (Anderson & Warren 2011).A CDA of entrepreneurship in print media can add to the understanding of this phenomenon (Achtenhagen & Welter 2007).For this reason, the analysis in this study was limited to the media discourse plane, specifically news media articles printed in South African English newspapers.
Another challenge is that the range of documents in which to conduct a discourse analysis is potentially never-ending (Rapley 2011).It was therefore decided to limit the analysis to a specific time period: 01 January 2018 to 31 December 2018.Fairclough (2013) finds that it takes a surprisingly small amount of qualitative data to reach a point of complete analysis.The analysis was regarded as exhaustive when arguments began to repeat themselves and no longer derived new insights.Although the timeframe falls in 2018, it is argued that findings and conclusions in this article are still relevant.The business discontinuance rate of businesses in South Africa is (still) lower than the rest of Africa, yet 'entrepreneurship' is (still) evoked as 'an essential driver of societal health, wealth creation, and a formidable engine of economic growth' (Bowmaker-Falconer & Meyer 2022).
Because the range of texts in which to conduct a CDA is potentially never-ending, it is important to pre-define the specific discourse for critical analysis from the onset (Ahl 2007;Gee 2004).The word 'entrepreneur' evokes different images, but the phenomenon under investigation in this study is the discourse surrounding Entrepreneurship as a development apparatus (EDA) (Smit & Pretorius 2020).To ensure that the analysis focused on this specific discourse (entrepreneurship when it is appropriated as a development tool), the key term 'entrepreneur' was substituted with 'small business', one of the synonyms used for entrepreneurship when referring to the phenomenon in a development context.As the analysis regarding the terminology confusion inherent in the entrepreneurship discourse will illustrate here, other synonyms will have rendered a similar (if not identical) sample of articles.With the term 'job creation' being at the centre of the South African question of economic development, this was the second key term used to find news articles specific to the discourse being analysed.
To ensure a methodical analysis of the data, this study was guided by, but not limited to, the sociological discourse analysis approach as described by Ruiz (2009).This is specifically relevant when ideological discourses are used to construct social inequalities and power relations.In this study, the CDA guidelines developed by Ahl in 2007are used as a departure point for the analysis (Ahl 2007): • Identify the knowledge claims and assumptions that are perpetuated in the text.• Identify the potential influence of the institutional context as well as writing and publishing practices on the questions that are being asked during the analysis.• Is the investigated text celebrating or criticising the takenfor-granted knowledge claims?• Which (if any) writing and publishing practices are shaping and delimiting the discourse?• Identify what is counted as 'proper' knowledge and 'proper' methods for entrepreneurship research.
• What voices are left out of the discourse?Or who is allowed to speak on the topic?• Are there (explicit or implicit) ontological and epistemological premises that lead to the conclusions?

Sampling
Random purposive sampling was employed by searching English print media articles that were published in South African English newspapers between 01 January 2018 and 31 December 2018 and contained the key words 'small business' and 'job creation'.The sampling frame was obtained from Sabinet through the University of Pretoria Library SA Media database.Sabinet, previously known as SA ePublications, is a database that contains records from South African newspapers and periodicals from 1978 onwards.Articles that contained the key words 'small business' and 'job creation' were then searched for this period.
The search with these key words rendered 122 articles, but these included articles that were duplicated in different newspapers.The duplications were eliminated, putting the sample at 95 articles.Thereafter, articles that were not relevant to the discourse on the relationship between entrepreneurship and job creation were excluded.These included, for example, articles on the fashions worn by Members of Parliament at the State of the Nation Address or a report of a student winning an entrepreneurship competition.The final exclusions left a sample of 63 articles on which the analysis was conducted.One identified discourse strand posed the need for further investigation and this was performed through critical case sampling of data not included in the original sample.
The texts were imported into Atlas.tias portable document format (PDF) documents in their original format and numbered from 1 to 63.An initial round of open coding was undertaken with articles 1 to 15.This initial round of open coding was not performed to establish emerging themes, but rather as a method of identifying major codes and code categories that could inform the hegemonic entrepreneurship discourse.From this first round of open coding, an initial thematic template was developed.However, this thematic template was not meant to be exhaustive and the coding and analysing of the data remained an iterative process.
Because CDA is a highly reflexive process, Ruiz's ( 2009) and Ahl's (2007) guidelines were followed to ensure rigour and minimise researchers' bias.These guidelines merely served as a departure point (as discussed in the methodology).Furthermore, because the discourse analysis was carried out through a critical lens it moved beyond not only being merely descriptive but also strove to be normative by proposing possible ways of mitigating the identified social wrongs (Fairclough 2013).Ruiz (2009) focuses on a field within discourse analysis called sociological discourse analysis, but the basic direction that his levels of analysis provide guided the analysis of entrepreneurship discourse (similar to Anderson and Warren [2011]).
Therefore, the analysis was conducted on three levels.Firstly, on a textual level, the aim was to objectively find patterns that may emerge from the data, for instance, by looking at the frequency of specific words or phrases.Discourses are recognised when certain patterns emerge from the text at certain frequencies (Achtenhagen & Welter 2007).In this first phase, meaning was not yet allocated to these texts yet, but it formed a reduced version of the discourse that guided the further analysis of the discourse.During this phase, categories for classifying the information were determined.These categories are indicated in the tables following in the discussion of the findings.
Although texts serve as the primary source from which data are extracted (Gee 2004), a critical approach to discourse analysis moves beyond a mere textual analysis.It takes into account the social, political and economic context (and existing power structures) from which these texts were produced (Wodak 2011).Secondly, the contextual analysis, thus meant taking into account this social and political context that gave rise to the specific discourse being studied and its influence on the discourse that was produced.In this second contextual level, the discourse is added as the subject (Anderson & Warren 2011).
Thirdly, at the interpretive level, the discourse is analysed as information, ideology and social product.In this study, the aim was to interpret the discourse to 'identify possible ideological developments of relationships between discursive formations' (Da Costa & Silva Saraiva 2012) and to establish how the production of a specific entrepreneurial discourse leads to hegemony in a South African context.
The process of analysis was not a linear process from level one to level three, but rather a circular, iterative flow between the three levels.The analysis thus moved between, identifying, contextualising and interpreting discursive items in the discourse, as well as relationships between different items.From these conclusions it could be determined if these discursive items and the relationship between them had any ideological implications (similar to Da Costa & Silva Saraiva 2012).It incorporated both quantitative 'mostly in the textual analysis', and qualitative analyses, with emphasis on the latter in the contextual and interpretive levels of analysis.

Ethical considerations
Ethical clearance to conduct this study was obtained from the University of Pretoria Faculty of Economic and Management Sciences Research Ethics Committee.(No. EMS176/20).

Terminology confusion surrounding entrepreneurship
From the first textual analysis, it became clear that a number of alternating terms are being used when referring to 'entrepreneurship', often interchangeably.A distinction was made between primary concepts (terms used more than five times) and secondary concepts (terms used fewer than five times).These included no less than 28 variations, shown in Table 1.In one article, as many as eight alternating terms were used to refer to the same concept.
The use of these terms as synonyms could have an inferred suggestion that all kinds of entrepreneurial activity, or anything labelled 'entrepreneurship' are in essence the same.One example explains how the information and communications technology (ICT) Small, medium and micro-enterprise (SMME) Development Strategy would 'actively support and develop entrepreneurs and small businesses in ICT' (Article 51, See Smit 2021).Throughout the article, a different array of terms is used to refer to the ventures that would be supported through skills development, funding and incubation.These include 'small business', 'enterprises in rural areas and townships', 'SMME', 'entrepreneurs', 'new businesses', 'start-ups', 'cooperatives' and 'youth and women entrepreneurs'.No definition of any of the terms is given and no distinction between the different terms are made.
In one instance, the article explicitly states that: [O]ur drive boost small businesses across all sectors of the economy is inspired by our National Development Plan which highlights that small businesses can generate up to 80% of our growth and create 90% of the jobs we need by 2030.
When this statement is made in the context of an article where seven other terms are used as synonyms for small business, it can be read that any one of these types of ventures can create 90% of the jobs needed by 2030.When these terms are used interchangeably, it is inferred that, for instance 'start-ups in rural areas and townships will create 90% of the jobs we need by 2030'.
Another example states: 'As a chamber focused on SMEs [Small and medium-sized enterprise] and entrepreneurs, we are pleased that President Ramaphosa has acknowledged the role that small businesses and start-ups play in our economy' (Article 34, See Smit 2021).This can be read to imply that small businesses and start-ups play similar roles in the South African economy.The same article also refers to 'SMMEs', 'cooperatives' and 'township and rural development enterprises' without making any distinction between the terms.
This finding is in line with Griffiths et al. (2012) stating: 'Policy makers, media, and practitioners increasingly recognise the value of entrepreneurship to economic growth but fail to distinguish among different types of entrepreneurship'.When the concepts included under the entrepreneurship umbrella are used interchangeably or conflated without consistency, it reinforces the knowledge claim that anything and everything included under the broader umbrella term 'entrepreneurship' will deliver the same results in terms of job creation and economic development (Poole 2018).Basing the mainstream discourse of entrepreneurship on commonly accepted definitions (Anderson & Warren 2011) that present any type of 'entrepreneurship' as essentially the same activity is protecting the theoretical assumption from being questioned that anything labelled 'entrepreneurship' will naturally translate into economic emancipation and job creation (Calás et al. 2009).This assumption hides the fact that only a small proportion of high-performing ventures account for the majority of innovation, job generation and wealth creation (Nightingale & Coad 2014), and as stated by Shane ( 2009): [O]nly a select few entrepreneurs will create the businesses that will take people out of poverty, encourage innovation, create jobs, reduce unemployment, make markets more competitive, and enhance economic growth.(pp.146) Viewing all types of entrepreneurship as essentially the same activity, and assuming that all will contribute equally to economic development, is a knowledge claim in the mainstream entrepreneurship discourse.Based on the aforementioned, it is therefore proposed that: P1: The South African media discourse replicates the mainstream entrepreneurship knowledge claim that all type of entrepreneurial activity is essentially the same.

Causality between concepts
The data were further analysed to determine whether the current media discourse in South Africa promotes Start-ups 9 Co-operative 6

Secondary synonyms ‡
New business and/or enterprise 4 Black, women and youth-owned enterprises 3 Early-stage business and/or enterprise 3

Black entrepreneur 1
Black-owned company 1 Business 1 Early-stage small enterprise 1 Emerging black business 1 Enterprises 1

Rural development enterprise 1
Rural entrepreneur 1

Small employer 1
Small enterprise 1 Young black entrepreneur 1 †, occurring more than five times; ‡, occurring less than five times.SME, Small and medium-sized enterprise; SMME, Small, medium and micro-enterprise.
http://www.actacommercii.co.zaOpen Access entrepreneurship (and all its synonyms) as a panacea for the lack of job creation and economic growth in the country.
To determine if the current discourse draws a link between small business and job creation or economic development, the following concepts were identified: 'small business' (and all synonyms), 'entrepreneurial development', 'job creation' and 'economic development'.Through the coding process, it became clear that the claimed relationship between these concepts was in some instances stated explicitly, but in other instances rather inferred or implicitly stated.Table 2 shows the number of statements drawing a causal relationship between the different codes -either explicitly or implicitly.
Of the 63 articles analysed, 50 claimed some causal relationship between small business or entrepreneurial development on the one hand, and job creation or economic development on the other.These positive relationship claims were stated 90 times in these 50 articles, either explicitly or implicitly.Some examples are given in the next section.

Implicit statements
Right now small-business owners are proceeding mostly on the basis of hope and promises.If government is to succeed in accelerating growth and job creation it can't afford to fail them this time.(Article 36, See Smit 2021) It does not explicitly state that small businesses accelerate growth and create jobs, but it is inferred.
Another article reports on the launch of a Centre for Entrepreneurship Rapid Youth Incubator.In this article, a politician states that the purpose of the centre was to move graduates of the centre from 'being job seekers to job creators' (Article 3, See Smit 2021).It is not explicitly stated that young people who are trained as 'entrepreneurs' will be job creators, but it can be inferred.

Explicit statements
We know that small business is the engine room for job creation.

Naturalised ideological position
An interesting finding showed that, although politicians are by far the highest number of persons claiming a link between small business and growth, other role players have also voiced the same opinion.Table 3 shows the number and origin of different voices that have done so.
Another key finding is that the bulk of these claims are made without supporting it through statistics or data.Of the 63 articles, only six show any type of statistics to support the claim that small business contributes to job creation, yet these statistics appear largely irrelevant to the claim being made.
One example, for instance, refers to findings of a study carried out in the USA as support for the case made in a South African context:  Only five of the 50 articles that draw a relationship between small business or entrepreneurial development on the one hand, and job creation or economic development on the other, substantiate the claims with local statistics.Two of these were the reporting on the completion of two separate entrepreneurial development programmes, not general statements about small business and job creation.Two others do not indicate from where their statistics were obtained, and only one article bases its claim on numbers made available from national treasury.In the large majority of articles, however, the claims that small business will necessarily create jobs and thus lead to economic growth are based on assumptions at best.Greckhamer and Cilesiz (2014) state that discourse can be used to naturalise ideological positions and win their acceptance as being common sense.When an entire community adheres to this position, the ideology is maintained and protected (Ogbor 2000).The given findings show that it is no longer only politicians that are claiming causality between entrepreneurship (including entrepreneurship development) and job creation, but that the ideological position has been 'naturalised' and is being accepted as a fact by a wider South African community.Furthermore, the validity of the claims is taken for granted and unreflectively reproduced, and in this manner legitimised (Ogbor 2000).When certain ideological positions are continuously reproduced, it starts to create a hegemonic sense of reality (Da Costa & Silva Saraiva 2012).We therefore propose that: P3: The knowledge claims that entrepreneurship leads to economic development and job creation has become normalised in the South African discourse.

Signs of political hegemony
Four articles explicitly stated that small business was not creating any jobs.All these four articles (12, 13, 32 and 53) refer to the same discourse strand: a phase 1 report on the SME sector in South Africa conducted by the Small Business Institute (SBI).
The SBI is a not-for-profit organisation that has been in operation for 75 years.According to its website: SBI stands for a free-market approach to grow SMEs, create jobs, and provide ethical leadership.We envision a country where small-and medium-sized businesses are supported by thoughtful, evidence-based policy that limits obstacles to their success and sustainability and the private sector works to eliminate structural obstacles to competition, access to markets and effective entry into supply chains.
In 2018, the SBI announced that it would be raising funding to complete South Africa's 'first baseline study on the nature of the small business segment of the economy'.The preliminary findings of the first phase were made available in 2018.These findings include, inter alia, that the number of SMEs in the country are significantly lower than estimated and contribute very little to job creation in South Africa.The report raises concern about the lack of a common definition when talking about SMEs across laws, regulations and strategic policy documents.Furthermore, it asks for the closing down of the Department of Small Business Development (DSBD), saying that it is ineffective and fiscally wasteful.
Because this report led to the only four articles opposing the notion that small business creates jobs, further articles following from this report were identified through critical case sampling.The three media articles that were identified showed a public argument between the SBI and DSBD following the publication of the mentioned phase 1 report.A hegemonic discourse establishes itself as the dominant discourse by not allowing for oppositional positions to be voiced (Brigg 2002;Stoddart 2007).This is the only example of the explicit silencing of an oppositional discourse, but the finding is in line with a study performed in Brazil, which found the discourses on entrepreneurship to be hegemonic in nature (Da Costa & Silva Saraiva 2012).It also supports the compensatory entrepreneurship phenomenon that Honig observed in Brazil and South Africa (Honig 2017).
When oppositional voices are patronised or silenced, especially as explicitly as found in this study, one has to acknowledge the hegemonic nature of a discourse.
In light of the aforementioned, it is proposed that: P4: The dominance of the mainstream discourse on entrepreneurship is opening the way for political hegemony in a South African context.

Conclusion
The authors' words, and the relationship of these words to other social practices, can be used to craft certain realities while excluding others.When the same discourse is produced and reproduced often enough and by enough social actors, society becomes subjective in thinking that this perpetuated discourse is the appropriate way of thinking, talking and acting about a certain theme.Critical consciousness towards this dominant discourse is dissolved and the masses believe that the status quo is the way that things have always been and therefore cannot (or should not) be changed.To ensure that it maintains hegemonic power, the dominant discourse does not allow for oppositional discourses to be voiced.
By studying a discourse that produces hegemonic views, this study contributes to the critical entrepreneurship theory.From a position critical to the assumptions in the mainstream entrepreneurship discourse, this article aimed to answer the question: Is this conventional entrepreneurship ideology producing a discourse in South Africa that is used to maintain political hegemony?In order to do so, a CDA was conducted on the claimed causal relationship between entrepreneurship (as embedded in the term 'small business') and job creation, specifically in the South African media discourse.The conclusions of this analysis produced the following propositions for further research: P1: The South African media discourse reproduces the mainstream entrepreneurship knowledge claim that all types of entrepreneurial activity are essentially the same.
P2: The South African media discourse reproduces the mainstream entrepreneurship knowledge claim that entrepreneurship will lead to economic development and job creation.
P3: The knowledge claim that entrepreneurship leads to economic development and job creation has become normalised in the South African discourse.
P4: The dominance of the mainstream discourse on entrepreneurship is opening the way for political hegemony in a South African context.
This study confirms that the media plays an influential part in reinforcing stereotypes about entrepreneurship.It reproduces the knowledge claim that anyone can be a successful entrepreneur, especially when given a chance on an entrepreneurship development initiative.As this discourse has become embedded in societies subjectivities, the masses are led to believe that their inability to capitalise on these opportunities are as a result of their own lack of abilities.
Because they believe that they themselves are to blame, the political actors promising economic emancipation and job creation are absolved of further responsibility in this regard.
The masses remain voluntarily subordinate and the social inequality becomes accepted.
Emancipation is a crucial focus of critical scholarship (Goldman 2021).Therefore, from a normative standpoint, this study proposes the following mitigating actions and directions for future research.This article has shown that the failure to distinguish between different types of entrepreneurship is providing fertile ground for political hegemony to develop.In a South African context, it could be argued that not all types of economic activity that are currently included in the umbrella term 'entrepreneurship', should in fact be included.By not clearly defining the boundaries in the field of entrepreneurship, a weak paradigm for entrepreneurship theory is promoted and the way for political hegemony is paved.However, if discourses can exercise social power, they can also challenge the exercise of social power.Discourses can challenge the assumptions inherent in the entrepreneurship discourse that are driving marginalisation and inequality.The South African scholarly community has to take up its responsibility as actors of social change and challenge the reigning public discourse in the field of entrepreneurship and strive for the emancipation of people from the hegemonic effects resulting from the assumptions underlying mainstream entrepreneurship theory.

TABLE 1 :
Interchangeable terminology used in the analysed discourse.

TABLE 3 :
Persons voicing the causal link between entrepreneurship and job creation and/or economic development.