Ethical Judgments About Social Entrepreneurship in Sub-Saharan Africa: The Influence of Spatio-Cultural Meanings

Within this paper, we adopt a qualitative process approach to explore how ethical judgments are influenced by spatio-cultural meanings applied to social entrepreneurship in the context of Mozambique. We analyse how such ethical judgments emerged using data gathered over a 4 year period in Maputo. Our findings illustrate three modes used to inform ethical judgments: embracing, rejecting and integrating. These describe how ethical judgments transpire as participants evaluate social entrepreneurship drawing upon related global normative meanings and those embedded within the local context. This analysis offers a critical contribution regarding how ethical judgments regarding social entrepreneurship evolve from negotiation and interaction within a context of multiple spatio-cultural meanings.


Introduction
As Sparks and Pan (2010) note, it is difficult to define what constitutes an ethical judgment; this presents a challenge to those attempting to analyse and investigate this issue. Thus, in an effort to clarify argument and so, advance research, Sparks and Pan suggest a broad approach that defines ethical judgments as the 'evaluation between one or more possibilities with respect to a specific set of evidence or goals' (2010, p. 407). As such, judgment is activated when ethicality-that is suitability and desirability-is recognised, a process informing preferences, beliefs or opinions shaping an 'individual's personal evaluation ' (2010, p. 409). Whilst useful in combining the notion of complex decision making with reference to the personal evaluation of possibilities, this definition does not fully acknowledge that contextual influences also shape such processes. Thus, the notion of context remains somewhat tacit within this model although there are two areas identified which point towards its importance in shaping ethical judgments. First, how the 'ethical content' of a decision scenario is filtered through social interaction and comparisons of various decision options. Second, how 'deontological evaluations' are used as individuals draw upon laws, rules, codes and norms to reduce the cognitive burden involved in ethical judgments. Both of these areas draw upon notions of context and how it informs the processes that produce ethical judgments. To contribute to this emerging debate, we explore how ethical judgments regarding social entrepreneurship (SE) are influenced by spatiocultural meanings within the context of Maputo, the capital city of Mozambique.
We develop a conceptual framework to study the influence of spatio-cultural meanings on ethical judgments by first; drawing upon the onto-epistemology of process theory (Tsoukas and Chia 2002;Langley and Tsoukas 2016) to study how ethical decisions are constructed within the moral texture of practice (Tsoukas 2018). Process theory shifts the locus of theorising about ethics away from individual cognition towards the evaluative role of language as it 1 3 is used to make evaluative distinctions during social interactions (Tsoukas 2009). One particular use of language is values articulation work (Tsoukas 2018), where meanings are negotiated to define the ethics of situated practices, such as those typically associated with SE (Dey and Steyaert 2016).
As values articulation work is context specific, our framework builds on the move to integrate context into theory development (Johns 2006(Johns , 2017. We focus on the confluence of spatial and cultural embeddedness (Baker and Welter 2018) that combine to generate context specific spatiocultural meanings. Our research focuses upon Mozambique, a sub-Saharan African country with specific contextual dynamics reflective of this region of Africa (George et al. 2016). These include colonialism, acute poverty, informality and ethnic identity (Rivera- Santos et al. 2015). Accordingly, our study is heterochthonous (Zoogah et al. 2015), as we study how ethical judgments about SE are influenced by spatio-cultural dynamics present within a sub-Saharan African context. Accordingly, our objective is to contextualise the process of ethical judgment making by analysing: How are ethical judgments regarding SE in the context of sub-Saharan African influenced by spatio-cultural meanings ? We apply a qualitative process approach to study how ethical judgments emerge as multiple spatio-cultural meanings are managed to inform ethical judgements about SE in sub-Saharan Africa. As such, we seek to advance theory by challenging the normative assumption that SE is homogeneous and universal (Baker and Welter 2018;McMullen and Bergman 2017). Consequently, we contribute to theory building by acknowledging the effects of context as articulated through spatio-cultural meanings that configure modes of values articulation work (Tsoukas 2018) as ethical judgments are applied to SE. We develop three modes by analysing material gathered from fieldwork undertaken in Maputo over a 4 year period during the implementation of a social enterprise project (2012)(2013)(2014)(2015)(2016). 1 This longitudinal approach enabled immersion in the local context during a period of ongoing interaction with multiple actors involved in SE in the Maputo region. A variety of methods, including participant observation, seventy-five semi-structured interviews, informal conversations and visual recording (video, photography) were used to gather accounts of local engagement with the SE project including ethical judgements regarding its meaning and value. The empirical material was analysed using a processual qualitative approach (Hengst et al. 2019;Jarzabkowski et al. 2017) to illustrate how spatio-cultural meanings informed by a normative global sense of what constitutes SE, and also locally embedded beliefs, influenced ethical judgments regarding SE. Our findings outline three modes (embracing, rejecting and integrating) to illustrate how spatio-cultural meanings within Maputo influenced the process of making ethical judgments about SE. Thus, our study analyses the influence of a sub-Saharan African context upon the process of constructing ethical judgments. As such, we offer a more granular understanding of how spatio-cultural meanings are connected to ethical judgments through different modes.
To address these issues, the paper is structured as follows. First, we explain our conceptual framework drawing upon process theory and calls to integrate context into theory development. Second, we describe the methodology and methods used during fieldwork to study the process of producing ethical judgments about SE in Maputo. Third, we describe the data analysis used to create three modes of values articulation work presented in our findings section. We conclude by discussing the implications for advancing knowledge about business ethics by integrating spatio-cultural meanings present in sub-Saharan Africa.

A Process Approach to Ethical Judgment Making
SE is defined as entrepreneurial activity with a social purpose (Austin et al. 2006) with assumptions that such universal beliefs and values translate across borders (Karanda and Toledano 2012). Encoded into SE are meanings that reflect ethical practices framed within neo-liberalism (Dey and Steyaert 2016); although these provide valuable insights to the genealogical influences affecting SE in Western contexts, they offer little insight into how prevailing definitions of SE are transposed to other contexts (Scholtens and Dam 2007). This is relevant to business ethics research if SE is defined as a set of practices (Dey and Steyaert 2016) drawing upon "socialised moral norms that reflect the social systems in which morals are embedded" (Anderson and Smith 2007, p. 4). The extant literature suggests that institutional contexts influence how SE is defined (Dionisio 2019;Kerlin 2010Kerlin , 2012, but focuses less on the process within which SE is exported and filtered into other contexts. This creates a gap concerning the extent to which assumptions about the ethics of SE are relevant beyond Western settings and more specifically, in sub-Saharan Africa (Zoogah et al. 2015). To analyse how ethical judgments about SE are produced in the sub-Saharan context, we develop a conceptual framework that draws upon process theory and calls to contextualise theory.
As noted, Sparks and Pan (2010) suggest that integrating the effects of social context may advance knowledge regarding ethical judgments. This necessitates recognition of how ethical content is recognised within social interaction and how cultural meanings are utilised to ameliorate cognitive burden during decision making (Sparks and Pan 2010).
We address this by drawing upon the onto-epistemology of process theory, which describes the world as emerging through ongoing interactions (Tsoukas and Chia 2002). This is relevant for conceptualising ethical judgments not as the property of an individual's biological cognitive system, but as mutually constituted through dynamic interactions connecting people with environments (Bansal et al. 2018).
Using process theory we conceptualise ethical judgments as emerging and unfolding as behaviour is evaluated according to an idea of what is 'good' and 'acceptable' (Sayer 2011). This definition focuses on the social processes within which the meaning of ethical judgments is negotiated as values articulation work is undertaken to evaluate the moral texture of situated practices (Tsoukas 2018). As such, this stance emphasises the importance of language, not as a way of transmitting ethical judgments, but as a space in which cultural meanings present in the situation at hand are used to make evaluative distinctions (Tsoukas 2009). From a process view, ethical judgments should not be conceptualised as static entities, but as dynamic and emerging as interaction occurs within specific contexts with negotiation upon what is considered good and acceptable (Tsoukas 2018).
We apply a process approach to conceptualise the process of how individuals engage in making ethical judgments. The ethical content of evaluation, or what is recognised as requiring judgment, is constructed through the language used to articulate the moral texture of situated practices as values articulation work occurs (Tsoukas 2018). This situates ethical judgment making within specific contexts and suggests that judgments about SE practices are not pre-ordained (Dey and Steyaert 2016). Instead, we expect ethical judgments about what is considered good and acceptable SE practices to emerge and so, be influenced by spatio-cultural meanings bounded by context. Such spatio-cultural meanings are expedient in reducing the cognitive burden involved in making ethical judgments.

Spatio-Cultural Meanings and Ethical Judgments About SE
The processual conceptualisation of ethical judgments emphasises the role of language in values articulation work (Tsoukas 2018). This is useful for conceptualising how the process of making ethical judgments is influenced by cultural meanings bounded by specific settings which in turn, responds to calls to incorporate context into theory (Johns 2006). Such contextualisation integrates a range of issues including cultural, institutional, social and spatial factors, to uncover voices that may remain unheard and practices that may be left unseen as they do not align with dominant views on entrepreneurship (Baker and Welter 2018). As the status of context is being debated and invites the risk of theory becoming unbounded (Johns 2017), it is necessary to state which aspect of context is being studied (Felin et al. 2015) and why investing in contextualising is a productive pathway.
Responding to such arguments, first, we focus upon the specific influence of spatial and cultural embeddedness drawing upon this to contextualise the process of ethical judgment making. Spatial dynamics are foundational to entrepreneurship theory (Marshall 2009) in terms of analysing regional influences upon entrepreneurial activity with a notable focus on sectoral clustering and technology ventures (Keeble and Wilkinson 2017). Steyaert and Katz (2004), Welter et al. (2017) note this has narrowed debate so, have encouraged more diverse analyses upon the impact of spatial context upon entrepreneurial activity illustrated by for example, how indigenous community relations influence entrepreneurial activity (Peredo and McLean 2006). We build on this spatial turn by exploring how contextual influences prevalent in sub-Saharan African countries influence SE (George et al. 2016, Rivera-Santos et al. 2015. Our study also integrates the influence of cultural meanings which we define as schemata, values and beliefs that create predictability (DiMaggio 1997), but are unevenly distributed and thus, capable of both enabling and constraining human activities (Patterson 2014). Spatial context influences culture by affecting which meanings are accessed and how they are filtered to meeting the pragmatic goals of individuals and groups (Patterson 2014). For example, the process of making ethical judgments about SE requires access to cultural meanings in order to engage in values articulation work and evaluate what SE is, and should be.
We combine spatial and cultural embeddedness in this paper to develop a spatio-cultural perspective of the process of making ethical judgments. We assume ethical judgments emerge as individuals draw upon both local and supra-local meanings (Baker and Welter 2018). The latter include cultural meanings developed by dominant paradigm builders, such as Non-Governmental Organisations (NGOs), elite universities and Governments (Nicholls 2010). These typically define SE as the creation of blended value by simultaneously addressing social and environmental problems, whilst also generating profit (Zahra and Wright 2016). This is premised on beliefs that SE is driven by individuals (Mair and Martí 2006;Zahra et al. 2009) who pursue a social mission (Mair and Martí 2006;Battilana and Dorado 2010) and are opportunity-driven and innovative (Zahra et al. 2008;Bull 2008;Kickul et al. 2012;Zahra et al. 2009). As such the ethics of SE are related to practices undertaken by heroic (Schwartz 2012) and transformative individuals (Alvord et al. 2004). At this supra-local level, definitions of SE are embedded in Western cultural meanings which are exported (Dart 2004;Dacin et al. 2011) via elites including transnational development organisations (Battilana and Dorado 2010;Rivera-Santos et al. 2012), philanthropic foundations, (for example, Bill and Melinda Gates Foundation, Ashoka, Ford Foundation) and university business schools (Schwartz 2012). We define this level as part of the global circuit of reflexive knowledge about capitalism (Thrift 2005). SE occupies a specific node within this broader circuit, one that branches into sustainable capitalism (Dey and Steyaert 2016), development (Seelos and Mair 2005) and blending of prosocial goals (Miller et al. 2012) alongside competitive behaviours (Calvo and Morales 2015). This paradox (Jay 2013) produces ethical tensions and conflicting cultural meanings (Smith et al. 2013).
The extant literature also suggests the cultural meanings shaping SE are locally embedded, for example in institutional differences (Kerlin 2012), community and indigenous cultural meanings (Peredo et al. 2004;Peredo and Chrisman 2006). Such research shows how cultural meanings about SE, for example a focus upon individual problem solving, do not necessarily translate to other spatial contexts (Kerlin 2010; Dionisio 2019). This is relevant given that arguments pertaining to contextual influences in sub-Saharan Africa such as acute poverty, informal institutional norms, resource constraints, the legacy of a colonisation, corruption and ethnic group identities are more prominent and so, likely to influence the validity of dominant normative theory (George et al. 2016). We also know these influences affect SE (Rivera- Santos et al. 2015;Santos 2012) and so, raise the question of how locally embedded cultural meanings influence SE (Karanda and Toledano 2012;Chell et al. 2016). Accordingly, our study is heterochthonous (Zoogah et al. 2015), as we analyse the degree to which locally embedded cultural meanings challenge normative constructions of SE derived from dominant cultural interpretations.
We argue that the spatio-cultural meanings informing evaluations about SE in sub-Saharan Africa will reflect challenges such as institutional voids-prominent in the region (Amaeshi et al. 2016). Such voids are relevant to SE as they are caused by state deficiencies, resulting in higher levels of informality, an absence of well-functioning markets and limited access to credit (George et al. 2016;Littlewood and Holt 2015;Rivera-Santos et al. 2015). As these reduce the effectiveness of formal institutions, such as the rule of law, regulations, tax systems and policy, they increase reliance upon informal institutions, such as social capital, cultural norms, and societal attitudes and values (Mair et al. 2012;Welter 2011;George et al. 2016;Holt and Littlewood 2017). As previous studies illustrate, this can motivate communities to exert direct control over such localised challenges through SE initiatives (Karanda and Toledano 2012).
Other influences include the prevalence of informal practices, such as Rotating Savings and Credit Associations (De Vletter 2001). ROSCAs, known as Xitique in Mozambique, enable communities to become both agent and beneficiary. This contrasts with SE in which the agent is typically separated by transactions from the beneficiary (Desa 2012;Mair and Martí 2009;Battilana and Dorado 2010;Rivera-Santos et al. 2012). In Xitique, all members donate a predetermined amount of money to a common 'pot'. The proceeds are allocated to one individual at each meeting who uses the money at their own discretion. Such localised practices are often seen as a precursor to modern microfinance group lending models (Armendáriz and Morduch 2010), but are largely unexplored in SE research, This is problematic as they reflect locally embedded cultural meanings associated with communitarian philosophies (West 2014;Woermann and Engelbrecht 2017) and ethical practices rooted in trust and reciprocity (Nhambi and Grest 2007). The African philosophy of Ubuntu, which underpins ethical practices (Lutz 2009), was considered significant as it is linked to values of mutual-assistance within families and communities to preserve togetherness and indebtedness to kin networks (Lux et al. 2016).
Summarising, we combine insights from process theory and contextualisation to study the influence of spatio-cultural meanings within sub-Saharan Africa on ethical judgments regarding SE. Our conceptual framework assumes spatiocultural meanings will be used within social interactions to make the ethical content of decision making visible (Sparks and Pan 2010) and reduce cognitive burden. As our conceptualisation of ethical judgments focuses upon the language used in values articulation work (Tsoukas 2018), we assume spatio-cultural meanings about SE occur simultaneously and shape actors practical knowledge which may be linked with global knowledge of SE or more locally embedded. We assume both global and local meanings and indeed, overlaps between the two, may influence ethical judgments; we now explore this premise through empirical investigation.

Methodology
A process approach (Langley 1999;Bansal et al. 2018) is utilised to address our research objective. This is relevant for responding to calls to advance business ethics research by moving beyond variance approaches that identify and analyse antecedent effects, for example race and gender, upon ethical judgments (Sparks and Pan 2010). In addition, whilst there are multiple approaches to process theory (Langley et al. 2013), an underpinning rationale for this stance is the opportunity to acknowledge the influence of context (Langley 1999). We draw on a qualitative process approach to study spatio-cultural meanings applied to SE and to develop understanding about the drivers of the process underlying ethical judgment making. Reflecting contemporary qualitative process studies, we analyse how conflicting spatiocultural meanings are negotiated and in some circumstances, integrated to address tensions related to such decision making (Hahn et al. 2015;Hengst et al. 2019).
Empirical material illustrating the process of ethical judgements pertaining to SE was gathered through immersion within the local context (Langley 1999). Fieldwork was undertaken to observe the flows, changes and relationality within which individual ethical judgments and spatio-cultural meanings became mutually constitutive (Bansal et al. 2018). This required observing social interactions in which values articulation work (Tsoukas 2018) was undertaken as ethical judgments were applied to the contradictory goals and ethical dilemmas that shape SE (Smith et al. 2013). Our aim was to 'do context' (Baker and Welter 2018) by entering a sub-Saharan African setting and harnessing a mix of intimacy and distance (Langley et al. 2013) as participants filtered and configured (Patterson 2014) spatio-cultural meanings. Accordingly, we critically evaluate the influence of spatio-cultural meanings present in a sub-Saharan context that reflect the dual influence of practices within global assumptions of what constitutes SE and those locally embedded and rooted in sub-Saharan Africa contexts.

The Research Setting
As an Eastern sub-Saharan urban setting Maputo combines a historical legacy (Bromley et al. 2004) with the growing presence of multi-cultural influences as diversity increases. Mozambique had a colonial past, gaining independence in 1975, followed by a brutal civil war (1976)(1977)(1978)(1979)(1980)(1981)(1982)(1983)(1984)(1985)(1986)(1987)(1988)(1989)(1990)(1991)(1992). The country experienced political and economic transformations as socialism, liberalisation, rapid privatisation and industrial restructuring created the contemporary economy and dominant party system (Dibben 2010). Mozambique is one of the poorest nations in the world and in sub-Saharan Africa (see World Bank data in Table 1). In addition to economic inequality, the country has limited state resources, weak governance, market failures, aid-dependency, corruption (Hanlon 2004), and an informal economy estimated to occupy 95% of the labour force (LO/FTF Council 2014). The country has been a target for multiple international development programmes with Maputo hosting one of the highest concentrations of donor agencies and NGOs (DFID 2012). The setting was appropriate to study the contextual dynamics expected to be relevant for developing theory in Africa (Rivera- Santos et al. 2015;George et al. 2016).

Data Gathering
To integrate context researchers use methods enabling the exploration of context. This includes ethnography as an approach to become immersed in everyday social interactions (Hammersley and Atkinson 2007). We drew upon a multi-sited ethnographic study within Maputo from 2011 as one of the research team worked in the area to initiate a local SE project. Four field visits provided access to observe the process of making ethical judgments about SE. The participants ranged from international aid workers to local Maputo residents whilst the field researcher adopted a range of roles including social entrepreneur, member of an expat family, judge at a social entrepreneur boot camp, guest lecturer at a local university and volunteer. Between 2012 and 2016, three shorter trips and one extended stay of 3 months were undertaken. The short trips were conducted to implement a social entrepreneurship project, but proved useful for developing contacts, conducting interviews and participating in various SE related events. The longer trip involved 3 months in the field to gather interviews, undertake observations and test initial analysis.
Seventy-five semi-structured open-ended interviews were audio recorded and transcribed. These lasted thirty to 90 min; most were conducted in Portuguese with some in English and Spanish. The field researcher is fluent in all these languages. The interviews were translated into English for consistency and analysis by the research team. Interviewees are summarised in Appendix 1 and included members of Mozambique governmental agencies, the UK Department for International Development (DFID), World Bank and United Nations representatives, local and international NGO employees, members of civil society organisations (CSO), MNC executives, SMEs, microfinance institutions, religious organisations, local academic researchers, journalists, SE promoters, and social entrepreneurs. Informal conversations were also held with other residents, beneficiaries, and members of Xitique groups. Participant observation included field visits, voluntary work, and participation in local official events, workshops, and Xitique ceremonies.
Material gathered included notes in a research diary; audio recordings of interviews and conversations (approximately 66 h), events and field visits (approximately 7 h), audio reflections (approximately 5 h); photographs (approximately 900); video recordings (3 h). These are summarised in Appendix 2. We also gathered documents including local case study posters composed by university students and analysed 42 secondary reports (e.g., policy documents, promotional materials). As multiple field trips were undertaken, it was possible to review initial findings as the data analysis, which is described below, was initiated.

Data Analysis
Building on the qualitative process approach, the data analysis involved multiple stages of refinement (Hengst et al. 2019). These were conducted to analyse the underlying threads (Bansal et al. 2018) that connected individual ethical judgments to spatio-cultural meanings about SE in Maputo. The three stages proceeded as follows. First, the data was structured to identify the range of practices people defined as SE. This included practices associated with the global circuit, such as social innovation, sustainable development and blended value creation. The initial analysis also showed how local practices, such as Xitique, were also included in people's evaluations. First-order codes were generated as the fieldwork was ongoing. This enabled live feedback as the researcher tested initial codes in the field. This iterative process continued throughout the research and was useful for generating an extensive  with what was published. This was useful for stage 3 to analyse the influence of global spatio-cultural meanings range of practices, projects and terms used in Maputo to define SE. Stage two initially involved categorising the data by demographics. Our aim was to connect the ethical content, or what people recognised as acceptable and good (Sayer 2011) about SE, into specific groups based upon participant backgrounds. Codes were developed to categorise the projects and practices identified in stage one with the aim of creating aggregate dimensions (Gioia et al. 2013). It was expected these would explain how ethical judgments were influenced by, for instance, religious orientation, international experience, gender, education and occupation. Whilst this stage increased familiarity with the data set, it departed from the process approach. Rather, it reflected a variance approach whereby we identified antecedents and explored their consequence on the outcomes of ethical judgments.
Subsequently, stage two was amended to focus on the physical context where an evaluation was observed. This enabled us to analyse how tensions between spatio-cultural meanings influenced ethical judgments about SE, rather than focussing upon who was making a judgment. This analysis enabled us to identify patterns across the data that showed how spatio-cultural meanings influenced the process of recognising ethical content (Sparks and Pan 2010). For example, we analysed instances in which participant language use linked practices of SE to poverty, informality, resource constraints, ethnicity and colonial history. We then connected these spatio-cultural meanings to the range of practices developed in stage one. This was useful for illustrating how spatio-cultural meanings were drawn upon to reduce cognitive burdens related to making ethical judgments (Sparks and Pan 2010). Stage two showed how context (spatiocultural meanings) and individual ethical judgments were mutually constituted as language used in social interactions enabled evaluative distinctions (Tsoukas 2009). These examples showed how values articulation work (Tsoukas 2018) undertaken to evaluate the moral texture of situated practices such as those associated with SE, involved using multiple and potentially conflicting spatial-cultural meanings.
In stage three, we refined the relationships between spatio-cultural meanings and individual ethical judgments by exploring whether the distinctions participants made in values articulation work drew upon meanings associated with the global circuit, or locally embedded meanings. This was initially a binary categorisation used to analyse how spatial and cultural embeddedness influenced the emergence of ethical judgments. For example, we analysed instances where judgment suggested exposure to practices legitimated by paradigm building actors (Nicholls 2010) and juxtaposed these to evaluations tightly coupled with meanings that suggesting local embeddedness. This stage was useful for analysing different states (Langley et al. 2013) of ethical judgment, or how the use of global or local spatio-cultural meanings influenced what people recognised as ethical content and how they used meanings to ameliorate the cognitive burden of making ethical judgments about SE. From this, we developed two modes of values articulation work which were labelled embracing and rejecting, depending upon how they related to global or local meanings. During subsequent analysis and later periods of fieldwork we added a third mode, integrating, to categorise ethical judgments in which global and local spatio-cultural meanings were co-enacted.
The three modes of values articulation work were developed to show how people made ethical judgments about practices considered as SE. The modes are based on analysis of the language used during evaluation and the degree to which individuals drew on spatio-cultural meanings from the global circuit of SE, locally embedded forms of SE, or indeed both. The three modes are intended to inform understanding about how context influenced individual ethical judgments about SE. Thus, rather than simply assuming meanings from a global circuit are transposed on to a sub-Saharan African context, using a process approach we analysed how spatio-cultural meanings used in ethical judgment making evolve (Bansal et al. 2018). The findings section illustrates how the three modes of values articulation work were mutually constituted in dynamic interactions between individuals making ethical judgments and spatio-cultural meanings co-present in Maputo.

Embracing
Ethical judgments reflecting the embracing mode typically demonstrated a strong relationship with locally embedded meanings and practices. They included references to Xitique and Ubuntu that participants linked to SE; these were used to depict SE as desirable and good as it provided a response to immediate needs and challenges for basic survival. For example, some evaluated Xitique as a form of local SE as it enabled marginalised communities to mobilise their resources and survive. Maputo residents used embracing to evaluate SE as practices they associated with coping with resource scarcity, surviving poverty and informality.
By acknowledging local challenges and opportunities to respond to such, embracing evaluated SE in a similar manner to that associated with community-based and indigenous entrepreneurship (Peredo and Chrisman 2006). For example, as the examples below illustrate, embracing involved evaluating SE by drawing upon locally embedded spatio-cultural meanings to promote community cohesion, resilience (necessity) and informality (not structured). Embracing suggested a mode of values articulation work (Tsoukas 2018) in which SE practices were evaluated using locally embedded meanings closely associated with the substantive sub-Saharan contextual dynamics (Mair and Martí 2009;Welter 2011;George et al. 2016).
In this country, social entrepreneurship, if we want to put a label on it, is born spontaneously and not in a structured way. Within communities, people are always doing little things to earn money they need. (Local development consultant) There are certain values and norms that facilitate certain practices, especially if we consider that there are small communities, large families, scarce resources, and an absent state… (International NGO project coordinator) Social entrepreneurship is almost everything. We never called it that here… the social is what motivates entrepreneurship. People start these things out of necessity! Social entrepreneurship in a poor country is about people solving their own problems.

(Social entrepreneur)
Social entrepreneurship here is a matter of survival and need, not of opportunity or growth. (Maputo based co-founder of an entrepreneurship incubator) These excerpts illustrate that what is recognised as ethical content is shaped by language related to locally embedded spatio-cultural meanings. It was not only local Maputo residents who used embracing in their ethical judgments of SE. As the following quotes suggest, global resource-rich actors also considered local practices as a form of SE as they provided an informal way to alleviate poverty. People need to create their own sources of income. I'm always asking myself where social entrepreneurship begins and ends and what is the commercial or individual aspect. There's a limit to the term here as in this reality people are interested and willing, but they have to make a living too. (International SE promoter) Embracing was used in ethical judgments to frame SE using locally embedded meanings that reflected the sub-Saharan African context. Embracing was open to a range of practices, such as Xitique, that were used to reduce the effects of poverty upon survival. Reflecting research from other deprived settings, community action is often the only way to meet basic needs whilst also creating mutual benefits (Peredo and Chrisman 2006;Karanda and Toledano 2012). As such, embracing was also used to promote and celebrate locally embedded practices as a form of SE.
Social entrepreneurship is implicit in what people do here, it's implicit because it promotes collaboration, common thinking. People find a way to solve each one's problems together. (Local Maputo serial entrepreneur and founding partner of 'Women Savings Bank') There is a social conscience here! People want to contribute to the community and the family and the issue of trust is important. (Public servant) Social entrepreneurship is from the grassroots! It challenges the microfinance sector because there aren't interest rates! (Mozambican liaison officer at an international foundation) Xitique is definitely an initiative of social entrepreneurship and you have to explain that to people. (Local Mozambican general manager of community relations for a multinational company) We have a sort of African socialism, which is one of permanent mutual-assistance in the family, in the community, and in society (Government Official) These comments illustrate how participants used embracing to express what they saw as good and acceptable (Sayer 2011), and so, the ethical purpose of SE. Embracing encapsulates the desirability of practices that preserved bonds and sustained collective responses to the tensions faced in the local context. As such, it demonstrated how spatio-cultural meanings associated with practices and beliefs present in the local context, such as Xitique and Ubuntu, were used in ethical judgments about SE. There was no mention of individual agency, social transformation and scalability, language typically associated with the global circuit of SE.

Rejecting
The rejecting mode was used to define ethical judgments about SE that involved distance from local spatio-cultural meanings. For example, participants recognised the ethical content of SE as being aligned with definitions flowing in the global circuit of SE. Typically this included ethical judgments about what SE practices were expected to achieve. Early interactions with IPEME (the Mozambican state institute for the promotion of entrepreneurial activity), for example, showed how SE was defined using terms such as charity and so, was associated with NGOs. As such, the rejecting mode was used to dismiss SE as a form of charity and thus, not real business and enterprise. Local practices were also rejected as being subsistence based. The rejection was made to affirm how SE should seek to achieve the same goals as those legitimised in the global circuit of SE. Consequently, it was evaluated as good and acceptable when it harnessed market-based sustainable business models to address poverty (Zahra and Wright 2016). As the quotes illustrate, individual actors, as opposed to collective and community action, were seen as the main driver of SE.
Xitique is not a business that's making money and having a social impact; it's an informal way of encouraging and facilitating savings. (North-American independent consultant in cross-sector collaborations) Xitique is not sustainable…but social entrepreneurs like Muhammad Yunus and possibly the inventor of Xitique…they are in fact the ones causing social transformation! (Maputo resident and social entrepreneur) Our mission is also to identify and distinguish existing high impact social entrepreneurship initiatives in the country. Our (bootcamp) winners are selected based on models which have to be: innovative, high impact, disruptive, scalable, replicable, and sustainable. (Bootcamp organiser) Xitique is not entrepreneurship, it is making do for the present day and lacks planning. (Social Entrepreneur) In these instances, participants drew upon concepts related to spatio-cultural meanings from the global circuit of SE to evaluate and de-value local practices of SE. Accordingly, rejecting drew upon global definitions of SE to criticise the value and efficacy of local practices. Rejecting therefore, involved articulating values such as social mission, transformation, scalability, planning and disruption. These aspirational goals rejected local practices for coping with survival, informality and sustainability. Instead, rejecting promoted the desirability of SE by accessing cultural meanings associated with resource-rich paradigm building actors. Accordingly, it acted to generate distance from informal practices based upon community responses to local issues.
The language used in rejecting suggested incompleteness to local practices. Such practices were described as lacking qualities, such as individual ownership, entrepreneurial ambition and opportunity-driven behaviours. This distancing from local spatio-cultural meanings was critical of dependency on informality which was seen as limiting SE from achieving the impact and growth good SE should create. The following excerpts illustrate how spatio-cultural meanings from the global circuit were used to distance meanings and practices rooted in local African context. Such practices were rejected for being non-market-based responses to poverty by marginalised community members and so, misaligned with what participants expressing the rejecting mode considered good ethical practices.
There's a place for social entrepreneurship because there's a place for business […] As opposed to NGOs, a social venture shouldn't transfer its operations to the local community. A social venture is a business that takes over and grows the country into a more developed nation. I don't think it's realistic that the community would assume control of all of these businesses! (North-American social entrepreneur) There's a clear separation between for-profits and nonprofits. The perspective people wrongly adopt here is that social entrepreneurship is for assisting a community and not an opportunity. (Government official) SE for me is an activity, an innovation that provides something to society and carries a social outcome. But what exists here in terms of SE is unknown, not recognised. (Public servant)

Integrating
The integrating mode represents instances when participants made ethical judgments referencing local practices and those associated with the global circuit of SE. As has been found in other contexts where conflicting practices are used, such as in CSR (Hahn et al. 2015(Hahn et al. , 2018 and sustainable strategy implementation (Hengst et al. 2019), integration is used to co-enact conflicting meanings without necessarily resolving differences. Integration in our study referred to ethical judgments that indicated familiarity with spatio-cultural meanings bridging locally rooted practices and those in the global circuit. For example, local practices, such as Xitique, were framed as both a collective form of survival and a means of encouraging entrepreneurial activities and innovation. If the members of Xitique are operating autonomously amongst themselves… we probably could call it an intracommunal social enterprise! (Advisor at DFID) SE is a new concept, but it has been implemented here already. For a long time it was perceived as charity. There's no legal form for it yet. People used to think that doing good was philanthropy and donating money. (Development Agency) The quotes are illustrative of conversations in which participants drew upon local and global spatio-cultural meanings 1 3 in their ethical judgments. Integration required awareness of local and global meanings without the aim of seeking to resolve differences. Instead, integrating was used as a source of flexibility rather than indecision. This created a space for linguistic invention, for example as one participant used the term 'intracommunal social enterprise', in an attempt to integrate their awareness of local practices with that of global practices of SE. Integrating suggested how change is possible as it created new ways of enacting spatio-cultural meanings, bridging spatial and cultural embeddedness in multiple contexts. This created ambivalence and increased cognitive burden, but also suggested how ethical judgements about what is, and is not, considered good SE practices may emerge and change (Bansal et al. 2018). The following quotes illustrate how integrating was used to enact change by filtering and configuring (Patterson 2014) cultural meanings from global and local sources to promote SE as space of possibility and openness in Maputo.
In my view, social entrepreneurship is more frequent where there is more know-how and people with different natures, origins… different cultures converging.
There are internal entrepreneurial practices that can be adapted through Western input. We utilise western models as references, but in reality there are difficulties in adopting them. There is the social framework here which is different and no legal structure, which is a deficiency. So we need a new model to substantiate social entrepreneurship here. (CSO1) Social entrepreneurship is a slightly artificial title right? Nobody starts a business thinking 'I'm going to be a social entrepreneur'. Individuals looking for livelihoods are not saying if what they do is going to be social or not. It just happens! The definition of SE taken in the western context is narrower but when taken in this country then almost everything could technically be classed as social entrepreneurship. (DFID's private sector development advisor) There are opportunities here for coupling social entrepreneurship initiatives with CSR programmes. This is gaining momentum and I intend to make the most out of it! (Social Entrepreneur Expat)

Discussion
Our findings are summarised in the process map (Langley 1999) represented in Fig. 1. This illustrates how context influences the process of making ethical judgments about SE. The three modes of values articulation work (embracing, rejecting and integrating) illustrate different ways in which, by drawing on spatio-cultural meanings, ethical judgments about SE were made during fieldwork. From this analysis, we suggest that the ethical content, used to recognise what is good or acceptable (Sayer 2011) and deontological evaluations used to reduce cognitive burden (Sparks and Pan 2010) are both influenced by the degree to which language used in interactions is connected to local and/or globalised spatiocultural meanings.
Embracing demonstrated how ethical judgments were shaped by direct connection to the contextual dynamics in sub-Saharan Africa such as poverty, informality, ethnic identity. Embracing required awareness and even promotion of local practices such as Xitique and African philosophies, such as Ubuntu. This mode of values articulation work (Tsoukas 2018) was premised on the desirability of SE when framed as practices that preserve collective responses to poverty and community cohesion. Within embracing we found no reference to meanings about SE based upon those within the global circuit, which promote social entrepreneurs as heroic individuals capable of generating both economic value and social innovation (Zahra and Wright 2016). The embracing mode indicated how spatio-cultural meanings rooted in local sub-Saharan African contextual dynamics influenced ethical judgments. Embracing showed little resonance with the ethical practices of SE constructed to create freedom from neoliberal governmentality (Dey and Steyaert 2016).
Rejecting was more closely aligned with the ethical content of practices associated with the global circuit of SE knowledge. Rejecting was observed as participants used language expressing the desirability of transactional exchanges, individual agency and scalable interventions expected to create social transformation. This mode shows how spatiocultural meanings from the global circuit of SE were drawn into ethical evaluations in Maputo. Rejecting suggests how access to unevenly distributed cultural meanings (Patterson 2014) can be used to create distance from, and seek to displace, local practices considered as somehow incomplete and less ethically acceptable. This evaluation was made upon the basis that globalised SE practices are expected to achieve transformational goals, for example through disruption and scalability, than those of local community responses.
Third, the integrating mode indicates that it is possible to co-enact multiple spatio-cultural meanings simultaneously without the need for resolution. Building upon research into how tensions in CSR and sustainable development are managed (Hahn et al. 2018;Hengst et al. 2019), integrating shows how spatio-cultural meanings from multiple spatial contexts can be accessed, filtered and applied during ethical judgments. This demonstrates how, as SE is not simply exported and transposed through a global circuit (although this is possible as rejecting suggests), but how ambiguity creates interstices in which ethical judgments emerge. The integrating mode suggests that ethical judgments become fluid and open to new possibilities when multiple spatio-cultural meanings come into contact. This is indicative of how cultural meanings do not determine ethical judgment making, but are filtered and configured (Patterson 2014) within spatial contexts. Interestingly, integrating was not used to resolve differences between spatio-cultural meanings, rather, it provided a space in which global and locally rooted sub-Saharan African contextual dynamics could be co-enacted to create new possibilities for ethical judgment making.

Limitations and Future Research
Whilst a multi-site ethnography is appropriate for studying how context influences entrepreneurship, it has limitations. The fieldwork required time to build interactional expertise (Langley et al. 2013) required to engage with participants as they made ethical judgments about SE. Thus, it was only practical to focus upon Maputo. This physical boundary means our study has a moderate generalisability, common in qualitative process theory (Langley 1999). Accordingly, we cannot make claims about how spatio-cultural meanings may influence ethical judgments about SE in other African locations. Our theorising is limited if judged against the aim of developing pan-African theory (George et al. 2016), for instance by making cross-country comparisons. We acknowledge the need to develop theory that can generalise across African countries and indeed between global regions (Dionisio 2019). Such research may be better served by future studies employing variance-based theory to identify antecedents and study their influence on outcomes across regions and between countries. However, such research designs may fail to articulate how context influences the process of ethical judgment making, which is identified as a limit business ethics theory (Sparks and Pan 2010).
We identify multiple pathways whereby future research may advance knowledge about how contextual influences in sub-Saharan Africa influence the process of ethical judgments about SE. Researchers may explore different levels of analysis, for instance by using cognitive theories to explore the micro-foundations (Felin et al. 2015) of how ethical judgments are contextualised. Meta-theory, such as institutional logics (Thornton et al. 2012), may also be useful for expanding knowledge about how contextual meanings change over time through institutionalisation. Micro and macro levels may also be combined, for example to study how attention is altered as institutional logics interact and alter what is recognised as ethical SE content. Thus, multilevel and in depth process based studies may continue to develop knowledge about how sub-Saharan African contextual dynamics influence the emergence of ethical judgments as the region undergoes periods of stabilisation and change.

Conclusion
By integrating spatial and cultural embeddedness, our study analyses how the spatio-cultural meanings within a sub-Saharan context influences the production of ethical judgments. We draw upon a qualitative process approach to analyse how multiple modes of values articulation work (Tsoukas 2018) emerge and influence ethical judgment making. Accordingly, our analysis demonstrates how theory development can be extended by integrating context (Johns 2006(Johns , 2017, particularly as it reveals the effects of normative assumptions (George et al. 2016;Zoogah et al. 2015).
We contribute to business ethics research by analysing how ethical judgments are influenced by the language used in social interactions to make evaluative distinctions (Tsoukas 2009(Tsoukas , 2018. By exploring a sub-Saharan context our study reveals how spatio-cultural meanings about SE are filtered and configured (Patterson 2014) through an interrelationship between local and supra-local levels (global) of interpretation (Baker and Welter 2018). By adopting a qualitative process approach, we develop a fine-grained understanding of how context and spatio-cultural meanings influences the process of ethical judgment. This advances knowledge by showing how ethical judgments about SE are not easily exported and transmitted to sub-Saharan Africa. Instead, by studying the process of making ethical judgments about SE, we show how local and global spatio-cultural meanings create heterogeneous modes of values articulation work (Tsoukas 2018). As our study demonstrates, spatiocultural meanings were used to promote and displace local practices; import globalised practices and to co-enact both global and local cultural meanings to create linguistic space for new practices to be framed as ethical.
We advance debate in this paper by contextualising the process of ethical judgment making in sub-Saharan Africa; in so doing, we analyse how context captured within spatio-cultural meanings influences ethical judgment making. This develops understanding by showing how spatio-cultural meanings are used in values articulation work (Tsoukas 2018), to influence what is recognised as ethical content and to reduce cognitive burden. Through our analysis, we demonstrate how ethical judgments emerge as a dynamic social accomplishment as language is used to draw upon spatiocultural meanings. Our model shows how different modes of values articulation work unfold as individuals make ethical judgment about practices they consider as good, ethical and desirable. This finding has implications for developing future theories pertaining to SE by challenging assumptions that there exists a universal SE phenomenon (McMullen and Bergman 2017); rather, recognising the influence of context and related heterogeneity regarding how ethical judgements are generated has considerable scope to advancing normative SE theorising. In effect, researchers should remain critical and reflexive to explore the process of making ethical judgments about SE as a contextually embedded phenomena (Grimes et al. 2013).
Our final point considers the practical implications for implementing SE in sub-Saharan Africa. The normative view of SE assumes initiatives will assist in poverty alleviation, although many projects, as we found through direct experience, fail during implementation. Our study could assist those implementing SE projects in sub-Saharan Africa by developing sensitivity to the heterogeneous modes through which context influences ethical judgments. Increased awareness of different modes of values articulation could help to manage tensions between spatio-cultural meanings more effectively. Acknowledging there are differing modes, such as those we identify, to interpreting SE should be considered when designing strategies for implementing SE. These may, for example, help to mitigate against the rejecting mode, harness the embracing mode and leverage the integrating mode. The aim is not to try and resolve differences between spatio-cultural meanings, but to work with opposing and co-enacted modes to increase the fit between SE projects and local realities. Consequently, we contribute to the design and management of SE projects such that they become more effective at navigating contextual dynamics influencing ethical judgment making processes within sub-Saharan African contexts.

Compliance with Ethical Standards
Conflict of interest All authors declared that they have no conflict of interest.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.