Two good interview questions: Mobilizing the ‘good farmer’ and the ‘good day’ concepts to enable more-than-representational research

In this paper, I assess the utility of two targeted qualitative interview questions: descriptions of a ‘good farmer’ and a ‘good day’, for eliciting rich textual data. Studies where farmers have been asked to describe or define a good farmer have been utilized across a growing range of international contexts in order to identify farmers’ cultural scripts, symbolic capital and social norms. The constitution of a ‘good day’ is new to rural studies but has been employed academically to assess perceptions of well-being and job satisfaction. I employ document analysis to analyse the multiple uses of the ‘good farmer’ question in the rural studies literature and introduce a contrasting empirical application the ‘good day’ question in a rural case study in the United Kingdom. Findings demonstrate that both interview questions can generate rich textual descriptions of embodied performances. ‘Good farmer’ definitions may also include a moral judgement, whereas the ‘good day’ question specifically yields descriptions of affect. Farmers are reluctant to identify ‘bad farmers’, but asking about a ‘bad day’ can open up discussion of the vulnerabilities of farming life. Both questions are thus suited to more-than-representational gaining utility from their congruence between common parlance and academic conceptualization. peers as ‘bad farmers’, although they were happy to describe their neighbours’ bad farming practices. This distinction may reflect language or cultural differences: in his Belgian study, De Krom (2017) described farmers who indicated that they believed their peers were identifying them as ‘bad farmers’ for participating in agri-environmental measures, and one withdrew from the programme for that reason. 8 An assessment based on my own experience in asking both questions, and review of transcripts from this and my other published work.


INTRODUCTION
More-than-representational theory has gained considerable traction in the rural studies literature, in line with a broader trend across the social sciences (Vannini, 2015;Phillips, 2014). Increasing numbers of rural researchers are seeking to go beyond the values, symbols and textual meanings emphasized in social constructivist approaches, to explore how daily life is experienced, practiced and performed; and challenge conceptual divisions between mind and body, people and place (Maclaren, 2019). Non-representational theory is an umbrella term utilized to refer to work that seeks to go beyond what can be consciously expressed, to make sense of the more-than-human, more-than-textual, multi-sensual world (Lorimer, 2005, p. 83). In rural studies, this body of work is more commonly termed 'more-than-representational theory', to demonstrate that the research includes but goes beyond representational approaches (e.g. Carolan, 2008;Halfacree and Rivera, 2011;Phillips, 2014). There are some differences in the definition of non-representational and more-than-representational approaches, but these are not directly relevant to this paper. I utilize 'more-than-representational' to describe my own approach and that of other academics who have adopted it, but 'non-representational' when it is the term utilized in the associated publication. In essence, more-than-representational research attempts to yield deeper, richer understandings by attending to the messiness and materialities of human and more-than-human interactions. In this paper, I critique two interview questions that can assist rural researchers to undertake empirical research grounded in more-than-representational thinking.
More-than-representational theory is built on the principle of relational materialism (from Actor-Network Theory, Vannini, 2015). The particular value of more-than-representational work is the attention it draws to unconscious, more-than human relations, and to the importance of lived, everyday experiences. Within this journal, Michael Carolan (2008) has drawn attention to the way that we 'think with our bodies', taking aim at social constructivism to argue for the importance of materiality -how our actions, bodies and worlds are constructed not only with our minds but with our bodies. What we do -in his example climbing trees, chopping them down or observing them -influences how we know them. Halfacree and Rivera (2011), also in this journal, demonstrated the utility of more-than-representational thinking for understanding post-migration rural experiences. Their article illustrated how the expectations of rural life held prior to migration (i.e. representations) were renegotiated through everyday acts of living. They posited that representational approaches are helpful but stop too soon. They argued that more-than-representational approaches offer 'fuller stories', including insights into the affective dimensions of rurality.
Although more-than-representational theory comprises a broad church of competing and sometimes conflicting approaches, it is widely agreed that theorical advancement has far outpaced methodological development. Maclaren (2019, p. 7) contended that "How you undertake, or 'do', a study with non-representational theories is probably the most fundamental question emerging from much of the work calling for non-representational theories". More-than-representational approaches represent considerable methodological challenges. As Phillips (2014) pointed out, by definition, interviews and other standard data collection methods are inherently 'representational', rendering the 'non-representational' represented. The emphasis in more-thanrepresentational theory on pre-conscious actions, performances, becomings and sensations, as well as the inclusion of more-than-human actants also present substantial challenges. Vannini (2015) drew on Dewsbury (2009) to conclude that non-representational methods will inevitably fail to achieve their ideal outcomes, but argued that success is found in 'failing better'. That is, we should, nevertheless, attempt to improve social research methods by applying more-thanrepresentational theory, because of the potential advancements in understanding we can gain.
The need for more targeted development and critique of research methods is also recognized in the broader rural studies literature (Woods, 2012). The few methodological papers published in rural studies to date have emphasized data collection approaches, such as focus groups (Pini, 2002), Q method (Duenckmann, 2010), 'deep mapping' (Bailey and Biggs, 2012), mapping and photo elicitation (Sutherland, 2019) and mixed methods (Akimowicz et al., 2018), with some papers also addressing sampling (e.g. Sutherland, 2020a;Burton and Wilson, 1999). Although visual and sensory methods have been introduced, the majority of qualitative studies utilize interviews to elicit or co-construct verbal responses to a wide range of topics. Mark Riley has specifically examined interview-based methods, demonstrating the discourses emergent from joint interviews, which might otherwise be hidden in a one-to-one interview (Riley, 2014) and the utility of locating interviews on the farm, where other farm workers, activities and objects can be included (Riley, 2010). Riley's work situated interviews as sites of performance -of both the interviewer and the interviewee(s) -with different locations and participants collaborating to produce the resultant data. In this paper, I similarly take a relational approach, understanding interview data as co-constructed. I focus on the theoretical positioning of the researcher: how theoretical concepts can be explored comfortably with study participants, yielding rich data, which can be utilized for both representational and more-than-representational analysis.
I focus the paper on the employment of two specific interview questions. Overarching research questions are typically identified in academic journal articles, as a standard of good practice and orientation for the research findings presented. However, the specific wording of the questions asked of study participants is primarily evident in quantitative studies, where questions are often listed in appendices, or included in tables of standardized responses. These questions are carefully considered, pilot tested, and sometimes developed into standardized scales for repeated use. In qualitative research -where semi-structured interviews are typically conducted -pre-planned questions act as a 'guide' to the interview. These questions may be explicitly formulated but are often generalized into a list of interview guide topics in the description of methods. There is an argument that it is poor practice to develop a formal questionnaire, as it restricts the flexibility of the researcher and potential for new concepts to emerge (Mason, 2002), encouraging standardization and 'quasi quantification' of qualitative data (King et al., 2019). It is also easier to simply start asking questions. However, there is a risk that interviewers will ask poorly formed questions if required to spontaneously formulate questions in situ during interviews. Mason (2002) argued for substantive advanced preparation of possible topics and subquestions to enable the interviewer to make rapid but well-informed decisions during the interview process.
The two questions presented in this paper are intended for use in this advanced planning for qualitative interviews. The 'good farmer' question asks farmer interviewees (and sometimes other industry professionals such as veterinarians) to define a good farmer, or to describe someone they think of as a 'good farmer' (in general, or in their locale 1 ). Studies utilizing the good farmer concept have been undertaken in empirical research in the United Kingdom (e.g. Burton, 2004;Sutherland, 2013;Riley, 2016a,b;Shortall et al., 2018;Thomas et al., 2019), Belgium (de Krom, 2017, Germany (Burton et al., 2008), Finland (Huttunen and Peltomaa, 2016), France (Saugeres, 2002), New Zealand (Haggerty et al., 2009;Enticott et al., 2021), Switzerland (Burton et al., 2021), Sweden (Saunders et al., 2016;Fischer et al., 2019) and the United States (Stock, 2007), demonstrating its popularity and utility across a range of contexts.
The Cambridge English Dictionary identifies 15 potential meanings of the term 'good' when utilized as an adjective, demonstrating that its usage can evoke an array of meanings. 2 In our recent book, Burton et al. (2021, p. 17-18) described the ways in which the term 'good farmer' has been used. When farmers utilize the term, it is primarily employed as a common-sense category, utilized to refer to farmers whose farming practices demonstrate a level of competence (i.e. 'good' is a positive assessment of skills). Burton et al. (2021) posited that the term 'good farmer' is an emic category, demonstrating the intrinsic cultural distinctions, which are meaningful to members of a social group. Farmers' appraisals thus have social significance, which has been conceptualized from multiple academic perspectives, most commonly 'symbolic' or 'cultural capital', drawing on Bourdieusian concepts (e.g. Burton et al., 2008;Sutherland and Darnhofer, 2012;Riley, 2016a,b). However, the moral implication of the term 'good' has led to alternative conceptualization of the 'good farmer', which emphasize the morality of farming practices (e.g. Stock, 2007).
The good farmer literature has also gained sufficient acceptance in academia that it has become a common place referent among scientists for social norms in agriculture (with or without reference to the underlying theoretical literature, e.g. Swinkels et al., 2015;Sutherland, 2011). Particularly important for this paper, Burton et al. (2021) recognized that the 'good farmer' term has also been developed as a methodological tool for eliciting responses relating to social norms, in a way which is comfortable for farmers. The grounding of the term in farmers' own usage makes it a familiar concept, whereas direct questions about 'social norms' are more challenging and inconsistent with the value of independence and sense of individualism associated with farming across the global West (see Emery, 2015). 3 However, to date, the specific methods of employing the good farmer question have not been systematically addressed.
The 'good day' question (asking study respondents to describe a good day, typically in relation to a day spent in particular location such as a workplace, for example: "can you remember a good day and can you describe it" Jackson, 2005) has been employed in the United Kingdom (Jackson, 2005), Chile (Aravena, 2017), Germany (Knabe et al. 2010), Singapore (Cleary et al., 2013), the Netherlands (Ouweneel et al., 2012) and the United States (Maier, 2011). In these studies, 'good' has usually been interpreted by respondents to mean 'satisfactory' or 'enjoyable'. Similar to the good farming concept, the usage of the 'good day' in colloquial speech has led to a range of applications, with varying degrees of conceptualization. A 'good day' may simply appear in a paper title as a reference to a statement made by a study participant, rather than being formally conceptualized (e.g. Vargo and Petróczi, 2016). When it is actively conceptualized, analyses have tended to focus on wellbeing, particularly noting the connection between psychological health and feelings of competence and autonomy, evident in study subjects' descriptions and assessments of a 'good day' (e.g. Sheldon et al., 1996). In its application to nursing studies, respondents' descriptions have referred to professional experiences, the ability to do something well and to achieve a high standard (e.g. Jackson, 2005). As such, both the good farmer and good day questions can lead to descriptions of skilled role performance.
In introducing the 'good day' question to rural studies and reviewing the published literature on the 'good farmer', I have two aims. The first is to facilitate the implementation of empirical research, which utilizes more-than-representational thinking. Linked to that aim, methodological critique of the two questions is intended to enable qualitative interviewers to co-produce thick descriptions, yielding rich datasets, which can be utilized in both representational and more-thanrepresentational research. My overall aims are to assist researchers who are new to qualitative research with farmers, and to offer insights to more experienced researchers on how to evoke rich textual descriptions and engage in more-than-representational work.

WHAT IS A GOOD INTERVIEW QUESTION?
A good interview question is one element of a much larger process of qualitative research design.
The key principles of qualitative research design are well rehearsed in method textbooks (e.g. Silverman 2016; Creswell and Creswell, 2018), so I will not reiterate them here, except to note the general principles that qualitative studies are iterative and reflexive, oriented towards increasing the understanding of processes and mechanisms, rather than proving or disproving particular hypotheses. Qualitative research relies heavily on social theory at all stages of the research process, but new concepts and findings are also expected to emerge organically through data collection and analysis. Miles et al. (2013, p. 37;Miles and Huberman, 1994, p. 34) identified four criteria for evaluating qualitative research methods: the potential to generate rich data on the type of phenomena studied, contribution to the generalizability of the findings, high ethical standards and feasibility. The potential to generate rich data is the primary evaluation criteria utilized in this paper; in this section, I evaluate the two interview questions on all four criteria, setting the foundation for more in-depth analysis of rich data production in the Findings Section.

Rich data
Rich data can be defined as "deeply contextual, nuanced and authentic accounts of participants' outer and inner worlds" Schultze and Avital (2011, p. 1). It is the hallmark of strong qualitative research. The successes of the 'good day' and 'good farmer' questions in terms of generating rich data are perhaps already evident in their associated volumes of qualitative research; the content of specific quotations will be analysed in the Findings Section. 4 The important question here is the extent to which the data produced through employing these questions is consistent with morethan-representational theory.
Although some proponents have argued for the development of specifically nonrepresentational methods (e.g. Thrift, 2008), to date there has been no preferred form of data collection for more-than-representational research. Proponents can and do use interviews, focus groups, participant observation, archival research and a range of other methods utilized by researchers affiliated with other paradigms (Vannini, 2015). Maclaren (2019) pointed out that in rural studies, more-than-representational analyses by Carolan (2008) and Phillips (2014) were based on data collected from a representational perspective. Where more-than-representational research differs is in its temporal orientation (Vannini, 2015). Thrift (2008) in seminal work, described non-representational methods as aiming to capture the on-flow of everyday life and focusing on practice, action and performance. Whereas traditional qualitative approaches rely on descriptions of past events or experiences, more-than-representational work seeks to emphasize the present and future. This emphasis on flows of practices which are ongoing and yet to be reflected upon, rather than collecting post hoc representations, privileges tacit over scientific knowledge, and performance and reaction over conscious reflection (Dirksmeier and Helbrecht, 2008). The paradox of more-than-representational methods is thus that data -as it is understood in scientific terms -are arguably less important than in other social research. At the same time, more-than-representational approaches aim to produce deeper, more holistic understandings (i.e. richer data). The utility of the two questions for more-than-representational research will thus be evident in the richness of the data they yield.
The distinctions between representational and more-than-representational methods are also evident in the subjects of the research, which I utilize in this paper to assess the relevance of data generated from the interview questions. Vannini (2015) identified five primary subjects of nonrepresentational investigation: • Events: the happenings (e.g. accidents, adventures, mishaps, crises and occasions) which reveal and alter expectations, raising the possibility of alternative perceptions and outcomes. • Relations: the entanglement of actors (human and non-human) which comprise the lifeworlds where meaning is negotiated. • Doings: the practices and performances (physical and mental actions) which produce particular effects. • Backgrounds: the situated spatial contexts in which practices and events unfold.
• Affect: "a set of flows moving through bodies of humans and other beings" (Thrift, 2008, p. 236) -automatic, relational and unreasoned responses to places and events, including but not limited to emotions.
The inclusion of affect is perhaps the most distinctive subject of more-than-representational research, focusing attention on bodily responses, going beyond emotion to encompass atmospheres and shared sensations. As a set, the five subjects attend to the experiences and practices of everyday life which are the focus of more-than-representational approaches, and are also inherent in the notions of how someone comes to be recognized as a 'good farmer' and to experience a 'good day'.

Generalizability
Generalizability in relation to qualitative research relates to the applicability of the associated concepts to other sites or subject areas. The generalizability of qualitative research is thus analytical or 'theoretical' -reflecting the utility of the analysis to apply beyond the study setting, for example by advancing theoretical concepts or yielding new concepts (Yin, 2014, p. 41). Good qualitative research must, therefore, be grounded in or produce convincing theory.
The 'good farmer' concept is rooted in social constructivism, most commonly Bourdieusian concepts of symbolic capital and habitus. Although Burton's initial (2004) conceptualization of the good farmer was grounded in symbolic interactionism (particularly drawing on Mead, Stryker, and Goffman), his later work developed Bourdieusian concepts of habitus and symbolic capital exchange (Bourdieu, 1984(Bourdieu, , 1986. This approach was followed by Sutherland (Sutherland and Darnhofer, 2012;Sutherland, 2013Sutherland, , 2019Shortall et al., 2018), who focused on Bourdieusian concepts of cultural capital and taste (Bourdieu, 1984). Mark Riley's (2014;2016a, b; work similarly utilized Bourdieusian concepts of capital exchange. The most recent published work (at the time of writing) similarly utilizes Bourdieusian concepts (e.g. Cusworth, 2020; Burns, 2021).
Boudieusian approaches to the 'good farmer' are also often integrated with other concepts, such as practice theory (Huttunen and Peltomaa, 2016), social identity theory (Naylor et al., 2018), materiality (Thomas et al., 2019) and assemblage (Sutherland and Calo, 2020). The emphasis within this work on performances -particularly visual appraisal but also how farmers' own performances influence their assessment of others -suggests that there is potential for more-thanrepresentational analysis of the associated datasets.
The 'good day' question has a less coherent body of literature. Where the concept has been most developed is in relation to a good day for medical staff -particularly nurses (Jackson et al., 2005;Maier, 2011;Cleary et al., 2013). Jackson et al.'s (2005) widely cited paper on a good day for new nurses utilized a Heideggerian phenomenological methodology. Similar to the good farming literature, Jackson et al.'s analysis demonstrated the importance of skilled role performance. It also drew attention to nurses' affective experiences (e.g. the "wonderful feeling at the end of a good day", Jackson et al., 2005, p. 85), demonstrating the utility of the 'good day' question for eliciting descriptions of feelings and sensations.
The growing number of studies utilizing the 'good farmer' concept demonstrate its utility across a range of topics and questions. The initial conceptualization was grounded in farmers' apparent resistance to afforestation (Burton, 2004) and environmental actions (Silvasti, 2003;Burton et al., 2008;Sutherland and Darnhofer, 2012;Sutherland, 2013). Recent work has demonstrated its utility for understanding livestock disease management (Naylor et al., 2018;Shortall et al., 2018;Enticott et al., 2021), retirement practices (Riley, 2016b) and fishing (Gustavsson et al., 2017). In contrast, the 'good day' has been utilized in medical and well-being studies (e.g. Howell et al., 2017), as well as studies of drug use (Vargo and Petróczi, 2016), unemployment (Knabe et al., 2010 5 ) and the influence of horoscopes (Clobert et al., 2016). Both concepts thus appear to generalize well across contexts and topical applications, although the 'good farmer' question is clearly restricted to agricultural populations.

Ethics
Research ethics have had limited exploration in relation to either the good farmer or the good day studies. This reflects (in part) the apparently innocuous nature of the associated questions. Most study participants to date have not been 'vulnerable populations': they have been predominantly working aged adults, and the questions are unlikely to lead to the discussion of sensitive issues (e.g. politics and sexuality). However, there is growing recognition of mental health issues within farming populations (Torske al., 2016;Jones-Bitton et al. 2019). Attempts to elicit descriptions of experiential responses, particularly affect, may open up deeply personal issues and raise the researcher's responsibility to break confidentiality to report mental health or animal welfare concerns. The 'good' element of the good farming concept also implies that some farmers achieve this standard moreso than others, thus suggesting that the term has a social boundary-making role, which may inadvertently be reinforced or challenged by researchers in the phrasing of their questions.

Feasibility
The particular value of both the 'good farmer' and the 'good day' interview questions is their familiarity to study participants. Farmers and others can and often do use the term 'good farmer' unsolicited within interviews -it emerges organically as part of their discourse. This is indeed how the meaning and potential power of the term as a conceptual and methodological tool came to be recognized (Burton, 2004). Similarly, it is common to mention a particularly good or bad day in casual conversation. As will be demonstrated, some of the research on these concepts is based on these emergent descriptions of the 'good farmer' and the 'good day'. However, the focus in this paper is on the intentional employment of these terms by researchers for the purposes of (co)producing empirical data.

METHODS
The findings presented in this paper are based on a combination of document analysis (of the method and finding sections of published good farmer research), and new empirical analysis of a dataset where the 'good day' question was developed. 'Good farmer' papers were identified for review through Science Direct and Google Scholar searches of papers with 'good farmer' in the title or abstract, with the addition of two other 'good farmer' papers written by its leading proponent, Rob Burton, where the term is developed within the body of the text (Burton et al., 2008;Burton, 2012). Some 30 articles were identified, comprising all of the 'good farmer' publications by leading authors (Burton, Sutherland and Riley) and all 6 of those published in leading rural studies journals at the time of writing. There is no similar body of rural studies literature available for the 'good day' question. The 'good farmer' is presented as a well-established term, which can be reworked for use in more-than-representational research. The 'good day' question is contrasted as a novel question, which is specifically designed to elicit affective responses.
The empirical data presented in this paper are from a case study in Scotland, not far in geographical terms from where Burton (2004) developed his initial conceptualization. The purpose of the empirical research was to evaluate agrarian change processes, particularly in relation to gentrification. The study developed a novel methodology, which involved seeking to engage with all of the holders of land over two hectares in size in the parish (township), and thus involved farmers at a range of scales, but predominantly smallholders (for further detail, see Sutherland, 2020a).
The good day question was developed during the course of the research, as a means to elicit rich text on the practices and experiences of smallholding. Subsequent literature review revealed that the good day question had not been utilized previously in a rural context. A dozen study participants responded to the question of what constituted a 'good day' on their holding; of these three also described 'bad days' and three more described 'not so good days', yielding the empirical data for this paper. Interviews were recorded, transcribed and analysed thematically utilizing NVIVO qualitative data analysis software. Pseudonyms have been utilized in the text to ensure anonymity. The empirical data presented in this paper are intended to be illustrative, rather than exhaustive, demonstrating the types of responses that arise from these interview questions, and how these data can be interpreted, rather than a systematic analysis of the dataset.

A note on feasibility and generalizability
A variation of the good farmer question was also employed in the empirical case study. However, it quickly became apparent that non-farmers in the study -that is, the majority of study participants -were uncomfortable with the question because they had no basis for evaluating the role performance of local farmers. They instead relied on a different meaning of the term 'good', discussing their neighbouring farmers in relation to their level of helpfulness or the visual appearance of their holding (particularly the orderliness with which it was kept). This demonstrates how different populations interpret the question and underlines the significance of 'learning with our bodies' argued by Carolan (2008). Non-farmers did not have the practical experience of farming to inform their appraisals. It is important that interviewers utilizing the 'good farmer' question are intentional in assessing the positioning of their interviewees, as this influences the interviewees' understanding of and responses to the interview question.
I also attempted to employ the good day question in a separate study addressing backyard poultry keeping, and discovered that it was less useful, owing to the time scale implied. For most participants, poultry keeping was a small part of their day: participants were confused by the good day question because they could not see its purpose, and resultant responses had little to do with poultry keeping (i.e. the research topic). I make this observation to demonstrate the importance of matching the question to the study population. The good farmer question is primarily suited to farmers and other participants in the agricultural sector who are in a position to evaluate the skilled role performance of farmers; the 'good day' question has broader reach, suited to farmers but also a range of rural and indeed urban inhabitants, when used in reference to how they experience a whole (work) day.

FINDINGS
In this section, I present an analysis of how the 'good farmer' interview question has been utilized in the rural studies literature, in order to demonstrate its utility for eliciting rich text, and to assess how these texts can be reconsidered from a more-than-representational perspective. I then present new empirical data produced through employment of the 'good day' question, similarly considering it's utility for more-than-representational research.

Asking the 'good farmer' question
To assess the utility of the 'good farmer' question, it is first necessary to ascertain how the question has been employed. Some published papers are straightforward about the usage of the term. For example, Burton's (2004) paper clearly stated that the 'good farmer' term emerged at an earlier stage in the research, and second interviews were conducted specifically to interrogate its meaning. Stock (2007) and Shortall et al. (2018) explicitly stated that the 'good farmer' concept emerged from their interviews. Papers that utilized references to being a good farmer which emerged from the interviews also include Gray (1998), Saugeres (2002) and Sutherland and Burton (2011). However, the term 'good farmer' did not appear in the direct quotes presented in McGuire et al. (2013), Kessler et al. (2016), Thomas et al. (2019) or Fischer et al. (2019). In Huttunen and Peltomaa (2016), Riley et al. (2018), Wheeler et al. (2018) and Enticott et al. (2021), it appeared only once. There are thus three primary ways of utilizing the 'good farmer' concept: • emergent from farmers' (and others') statements • included in the question guide for the interview • academic interpretation of farmers' statements These approaches are not mutually exclusive -most social science papers employing the 'good farmer' term utilize it for academic interpretation, but there is often ambiguity regarding whether the concept emerged from the data, or descriptions were actively solicited. In conducting the analysis, I found only 10 datasets (comprising 12 papers) where it was clear in the description of methods that the good farmer question was specifically utilized in data collection. These articles rarely reported the precise wording of the question utilized during the interview (see Table 1), in keeping with the flexibility of semi-structured interviewing practices. The 12 papers which explicitly utilized the 'good farmer' interview question are analysed here, in order to ascertain the responses which are elicited through intentional use of the good farmer question.
Differences in apparent application may also reflect language differences. To date, the 'good farmer' concept has primarily been employed among native English speakers. Applications in Belgium, Finland, France and Sweden suggest that it has slightly different meanings in these languages. In Finland, France and Belgium, 'good farmer' appears to imply 'real farmer' ('paysan', in French). For example, Silvasti (2003) described Finnish farmers' perceptions that 'real farming' is 'real work': physical and visible. In her case study, hard work and morality were equated. A good farmer was understood as a 'real farmer'; a 'real farmer' is hard-working; production is a moral activity. De Krom's (2017) Belgian study participants similarly used the term 'real farmer'. Although none of these papers drew attention to translational issues, they demonstrate the importance of situated understanding of the specific terms utilized in interview questions. Gkartzios and Remoundou (2018) drew attention to the language politics of contemporary rural research, arguing for the importance of embracing original language and associated meetings, in order to counter the dominance of Anglophone research. I raise it here for attention in the international applications of the interview questions.

Re-interpreting the good farmer from a more-than-representational perspective
In considering the potential of the 'good farmer' question for more-than-representational research, I return to Vannini's (2015) five foci: events, relations, doings, backgrounds and affect. As described earlier, the good farmer question elicits descriptions of farming practices (i.e. skilled role performances), or the 'doings' of competent farm work. Farming is a profession, with evident professional competencies. Proponents of the good farmer question then delve into the symbols and associated meanings of those practices (i.e. the visible representations of embodied farming practices). Burton's (2004) article emphasized the importance of yield as a symbol of skilled role performance. Subsequent studies have expanded his initial list to include: keeping the farm tidy, productivity, timeliness, 'doing the job right', hard work, livestock condition and quality, soil and wildlife maintenance, making good use of resources, being progressive, securing farm succession and remaining on the farm into older age Darnhofer, 2012, Riley, 2016b). From these descriptions, the underlying mechanisms -symbolic capital (Burton et al., 2008), habitus (Riley, 2016a,b) and capital production (Sutherland, 2013) have been conceptualized and applied to advance understanding of the associated research questions. Consistent with morethan-representational thinking, these mechanisms are largely unconscious. The implications are thus that the good farmer interview question is suited to more-than-representational data collection, but also that representational research using the good farmer question has already been successful in illuminating these semi-conscious mechanisms.

TA B L E 1 Analysis of articles which intentionally utilized the 'good farmer' (GF) question in qualitative interviews
Author/Article

Presentation of GF question in findings
Silvasti (2003) Social scripts (Money) Not stated "When asked to describe a good farmer" p. 145 Burton (2004) Symbolic interactionism (Mead, Stryker, Goffman) "The interviews were based around a number of broad themes, in particular, behavioural symbols of commitment to agriculture, status symbols and the process of status acquisition, entry criteria into the farming social group (i.e. point of admission as a 'good farmer')" pp. [200][201] No specific reference to question structure or direct quotes of questions. Burton et al. (2008), also Burton (2012) Cultural and symbolic capital (Bourdieu) "Study participants were presented with a series of cards, which they were asked to order in terms of the extent to which an activity was able to display their skills to other farmers and the accessibility of the display to other farmers" (2008, p. Social Identity theory (Taifel and Turner), self-categorization theory (Ellemers) "interviews lasted approximately one hour and asked participants to recall in detail their routine disease management practices" p. 5.
"during a face-to-face interview, when asked to describe what he meant by the term 'good farmer'" p. 7 Birge and Herzon (2019) Social scripts (Simon and Gagnon), habitus and cultural capital (Bourdieu) "interviews focused on presentation and discussion of the approach and practicalities of the hypothetical bonus scheme, exploration of the intersection of social and cultural capital elements with uptake of the bonus scheme, and notions of good farming and the place of NMG within those conceptualisations." P. 4 "We asked farmers directly what 'good farmer' means to them and whether NMG fits into good farming." P. 6 Cusworth (2020) Habitus (Bourdieu) Question guide questions not stated.
"Consider the version of good farming given by I02, who ran a large mixed farming operation. . . I021. . . reported that good farming is: . . . " (p. 167) Sutherland and Calo (2020) Symbolic capital, habitus (Bourdieu) and assemblage (Delanda) "The importance of 'good crofting' identity emerged early in the study and was included in the question guide. . . The interview guide specifically asked study participants to describe people they considered to be 'good crofters' and the type of croft they would like to emulate" p. 535 "When asked to identify 'good crofters' in the area, and the type of crofter they might seek to emulate" p. 538 Burton's (2004) analysis also addressed the practices through which the 'good farmer' assessment was made and communicated. He elaborated on Seabrook and Higgens' (1988) observation of 'roadside' or 'hedgerow' farming, drawing attention to the physical performances and visual appraisal of farmers: It's nice to be able to look over the hedge of somebody who's doing worse than you because you feel really self-satisfied . . . I bet we get more pleasure at looking at a bad field of someone's than we get from looking at a good field. (farmer 37) Wherever it is, if I see a field of sheep I always have to slow down and look at them . . . If their lambs look as good as mine or -you know. If they look better than mine we drive on and if they look worse than mine I feel quite pleased. (farmer 44) Burton 2004, p. 205 Similar observations about roadside farming have been taken forward in Burton's subsequent work (particularly Paragahawewa, 2011, andBurton, 2012). Utilizing a more-thanrepresentational lens to consider these quotations draws attention to the emotions described by the farmer -self-satisfaction and pleasure -and how these performances and affective responses are co-produced in relation to crops and livestock (i.e. the agency of non-human actors in the formation of 'good farmer' identity and experiences).
Affective responses are also elicited through questions or discussions of bad farming. Saunders (2016, p. 396) included a quotation from a farmer who "tried organic farming, but could not cope with thoughts of weeds. I had nightmares. Good farmers have no weeds". There is a clear stigma associated with poor performance (Burton et al., 2008) and a negative psychological impact of failing to reach this standard (Haugen and Brandth, 2017). However, few researchers to date appear to have specifically asked their study participants about 'bad farmers'. The adjective 'bad' has been utilized in the published literature, but this has typically been in reference to farming practices, rather than to a specific farmer. Farmer 37 above commented on a 'bad field' rather than a bad farmer. Sutherland and Burton (2011) demonstrated that farmers were less likely to share equipment with farmers who did not meet good farming standards, but their respondents did not use the term 'bad farmers'. Enticott et al.'s (2021) New Zealand study respondents appeared more enthusiastic in offering descriptions of the poor practices of their neighbours, but they reported that farmers used terms like 'dodgy', 'rough' or 'dumbarse', although the term 'bad farmer' did appear once in the quotes presented. Native English-speaking farming respondents appear to steer clear of labelling their peers as 'bad'. 7 Researchers instead may imply it from farmer statements. For example, in his American study, Stock (2007, p. 88) stated that "the organic farmers I interviewed. . . distinguish between good farmers and those who are not -in other words, bad farmers". Achieving the status of a 'good farmer' is an identity predicated upon but not reduceable to the performance of good farming practices. The failure to produce symbols of good farming does not necessarily equate to identification as a 'bad farmer' by other farmers. These nuances are important for understanding the mechanisms of performance and farming identity formation, and the sensitivities of farmers about the label 'bad farmer'.
Research utilizing the good farmer concept also demonstrates relationships. Joking and banter about farming mistakes reinforces and promotes good farming practices (Burton, 2004;Sutherland and Calo, 2020), but adds to the challenge of assessing how seriously these criticisms are intended and taken. These interactions occur at industry events and day-to-day encounters.
Consistent with Vannini (2015), events are important for altering perceptions. Sutherland and Darnhofer (2012) and Sutherland (2013) in particular drew attention to the reflexivity of 'good farmers' who converted to organic farming, finding that farmers undertook processes of active reflection in response to critical incidents (e.g. disease outbreak and low commodity prices) prior to and during their conversion to organic farming. These conscious reflection processes undertaken in the past tend to be de-emphasized in more-than-representational theory, in favour of current and ongoing sensations and experiences. This suggests the importance of the 'more-than' as opposed to 'instead of' approaches implied by the term 'non-representational': that is, to seek to build on and go beyond but not exclude representational analysis.

The 'good day'
In contrast to the 'good farmer' question, which typically elicits the description of another person, the 'good day' question focuses attention on the experiences of the study participant over a specified period of time. Perhaps owing to the duration of a day -which necessarily involves a series of events -the responses tended to be longer than those to the good farmer question. 8 The following quote from Betty (female farmer) demonstrates the length of some responses, also encompassing many of the recurring themes in the interviews. Responses to shorter framings of the question, for example: "Tell me what a good day is for you around here?" yielded shorter responses, but progressed with minimal encouragement, demonstrating the feasibility of the interview question. Study participants typically understood the question easily, although one clarified whether their description should be "fantasy or truth", opening up the avenue for descriptions of the good day they would like to have in future, or in an idealized world.
Lee-Ann: So tell me what a good day is here: if you come to the end of the day and you're thinking "oh today was a really good day" what would that likely have involved?
Betty (part-time farmer): Okay well the first thing I would say a good here is a day without wind because at the top of the hill here it's almost always windy. And it's really noticeable from where we used to live which was at the bottom of the hill the difference is marked, even within half a mile. So I think you know a good day for me would be really weather dependent, a good day for, for me would be a day where work didn't take so much precedence that I could indulge some of the things that I think of as not being work. So if I had time to ride, or time to spend time in the garden, that would be a good day for me. But there are lots of little good things as well, the light here can be absolutely amazing so especially very first thing in the morning, or in the evening you get amazing light so that's a good thing. Then there are other things that make good days, you see something really fabulous. I don't know, you see the osprey or you see, days when we see something unusual like a kingfisher, or an otter, or something, that's a good day. A good day for me is a day when I ride round the hill here and back and come back in one piece -that's a very good day! A good day is a day when we actually have time to enjoy, a good day is a day when we have time to enjoy the place and not just work on the place so, you know sometimes on a Friday night we'll get something to eat and a couple of bottles of cider or something and go down to the river and have our tea there. Or we'll take coffee on a Sunday morning and go down to the river and take uh, blimmey reading a newspaper is a good day here! But you know to actually have time to stand back and appreciate the surroundings instead of only ever looking ahead and seeing a 'to do' list that's impossible to achieve. I think it's that isn't it?
Betty's response provides entry points for interviewers probing about all five of Vannini's (2015) more-than-representational subjects: she described practices (riding, reading a newspaper) and events (Friday night meals by the river, seeing osprey). Relations are apparent in the interactions with her horse and wildlife. Backgrounds are evident in the wind, weather and the farming context, where the experiences she described are located and shaped. Affective responses were clearly apparent in her description of early morning light, sense of indulgence, slowing down and appreciating her surroundings. She also alluded to the risks associated with her activities, coming 'back in one piece' from a horseback ride, suggesting a sense of satisfaction and relief. Other study participants described the sound of birdsong, the different intensities of light in morning and evening, and the 'softness' of the air at particular times of the year. Food featured in some descriptions (e.g. a nice meal and wine). These pleasures were described as 'simple', in keeping with broader rural idyllic imagery about wholesomeness of the 'good life' in the country (Sutherland, 2020b;Halfacree and Rivera, 2011).
As with Betty, almost all respondents in this case study included some mention of the weather in response to the 'good day' question. Owing to the outdoor location of most farming and smallholding activities, the weather has a direct impact on what is feasible, and the enjoyment of those activities. A 'good day' is co-constructed in relation to the atmospheric conditions, linked to the primary characteristic of a 'good day' identified -the freedom to choose how to spend it. 'Freedom of choice' was a theme for land holders at the range of scales, an affective sense which is both located in space -dependent on local weather conditions -but separate from it. This freedom to choose also demonstrated respondents' preferred activities, particularly those that are "not work". A good day was an opportunity to indulge and enjoy where they were living. These activities included horse riding, fishing, walking in the woods (alone or with pets or family members) and simply sitting in the sunshine. Responses thus demonstrated the materiality of the 'good day', in relation to the affordances of rural living.
As pointed out by Elva (female smallholder), the 'good day question' is highly personal: Elva: But to get back to what you were asking about a day we enjoy, for me that is actually quite a personal question because it's, it's what turns people on really. I'm, I'm a goal driven person to some extent . . . For me. . . what I'm enjoying now that I've retired is seeing all the things . . . starting to come to fruition. . . . Now we've lost so many shrubs there because it's a frost pocket and it's also a wind tunnel so for me satisfaction is being able to move a shrub in there that I know will survive on our land, . . . it is very much the feeling of working along with nature in the fact that I never really think about it as this is ours, we own it! It's more like we're living with it and living in it and if I've planted a shrub that I know that the bees and the butterflies and things like that will like, and it takes, and. . . it's that sort of thing. Seeing Ben [her husband] out virtually everyday in his orange overalls and his tackety [old, worn] boots and his silly hat out working away because I know that Ben's happier working outside.
The good day question immediately draws out the personal feelings, interests and values that are enacted through daily activities. It does so in a way that naturally elicits a descriptive response, in Elva's case the competence she feels from having worked on her holding for a number of years -long enough to learn what works, and what attracts wildlife onto her property. Consistent with Carolan (2008), what Elva has learned with her body through working on her holding became evident. She recognized what species would and would not thrive and expressed the agency of the natural environment. The good day question also elicited statements about the relationships Elva values -the experience of a good day linked to and interrelated with her husband's emotional state and practices. The agency of animals within a 'good day' was particularly evident among the livestock farmers in the study, where "if you go around in the morning and a cow has new calved and there's a nice healthy calf up and sucking like, that gives you a boost" (Ross, male farmer). The good day question thus reveals the everyday experiences that are embodied in the affective landscapes of the farm.
Affect and role performance were particularly evident in descriptions of bad days. These were days that were particularly emotive, often involving animal interactions. One respondent described a "particularly hellish" weekend when three cattle died for different reasons. Another described a routine bedtime visit to the stables, where she discovered "blood all over the snow", from an injury to her horse. Bad weather -particularly when it impacts on production -was also experienced in the body: Oh actually the very worst thing of all is when we've had hay or silage just ready for baling and rain has come on it, that is torture isn't it? . . . And you're waiting for something to dry and you get rain it's just hugely depressing, that undoubtedly is the worst thing of all, and there's nothing you can do about it. (David, male farmer) These descriptions painted clear pictures of the experiences of daily life on the holdings. Descriptions of bad days also led to comments about potentially leaving the holding -"packing it in" (Dennis, male farmer) or moving somewhere that all the to-dos were not "staring her in the face" (Jeanne, female smallholder). This was particularly evident for the one respondent in the study who was disabled -her bad day was limited by her pain levels, the primarily enabling/disabling condition for her.
The bad day and not so good day questions yielded noticeably shorter responses than their more positive counterparts, comprised of statements describing specific instances or conditions which may have lasted minutes or hours but ruined the remainder of the identified day. The bad day question was thus particularly useful for bringing into discussion the strongest negative experiences of rural life. It is temporally distinct from the 'good day', which was comprised of a series of positive events. Like the distinctions between a 'good farmer' and a 'bad farmer', a 'bad day' is not necessarily the opposite of a 'good day', and tends to be employed as a description made by academics of what their respondents are experiencing rather than a targeted interview question (e.g. Fifield et al., 2004;Cummings and Connelly, 2018).

DISCUSSION
In this paper, I have investigated the data produced utilizing two specific interview questions: what it is to be a good farmer, and what constitutes a good day on a smallholding or farm. Both questions directly or indirectly yield descriptions of events, relations, doings, backgrounds and affects; opening up the opportunity for interpretation from a more-than-representational perspective. The 'good day' question is particularly useful for eliciting descriptions of affect. Study participants freely described how it 'feels' to live on a farm or small-holding, and how daily experiences and performances situate their understanding of rural spaces. This is consistent with Halfacree and Rivera's (2011) argument that staying in place is, nevertheless, a process, one where everyday lives are entangled with a range of actors and agencies, leading to anticipated and unanticipated experiences where residents are immersed in their farms, rather than simply being located there.
Both the 'good farmer' and the 'good day' questions are particularly helpful for eliciting rich text on topics which are otherwise challenging to approach. Farmers in particularly pride themselves on their independence and are unlikely to identify themselves as influenced by others (Emery, 2015). Asking about definitions of a 'good farmer' or descriptions of a good farmer in their locale elicits data on symbols, cultural scripts and social norms in a way which is consistent with how farmers naturally approach these issues. Farmers themselves discuss who is a good farmer, promoting and socially enforcing these norms. Asking someone if they have had a good day is a standard social question; inviting study participants to describe a good day is a small step further which brings out rich descriptions of the experiences of the remembered day. Although study participants recognized that this question is inherently personal, they were, nevertheless, comfortable answering it in detail. The 'good' questions are thus highly feasible, drawing their utility from the linkage they form between academic concepts and colloquial speech.
In contrast, the 'bad farmer' question has not been substantively explored as an interview question, and instead has been inferred through academic analysis. This present analysis suggests that failure to produce good farming symbols, and performing 'bad farming practices' do not necessarily equate to identity as a 'bad farmer'. Farmers also appear uncomfortable with this term. These subtle distinctions, as well as differences in language (e.g. the meanings of 'good' and 'bad' outside of the English language) are areas for further research.
The 'good day' and 'bad day' questions also have considerable potential for gaining entry into issues where farming populations may be particularly vulnerable, such as retirement, gender identity and mental health. As described in the introduction, the good day question is already in use in studies of mental health and wellbeing (e.g. Sheldon et al., 1996;Howell et al., 2017). The bad day question appears to be more palatable than asking farmers about bad farmers. It shows particular promise in providing insights into strongly affective experiences of farming life. Haugen and Brandth (2017) argued that failure to achieve the expectations of being a 'good farmer' can be damaging to mental health. Farming as a profession is high stress (Jones-Bitton et al., 2019), with higher than normal rates of depression (Sanne et al., 2004;Torske al., 2016) and suicide (Fraser et al., 2005;Simkin et al., 2003). Farmers are also characteristically stoic and unlikely to seek help (Roy et al., 2014). Gaining insight into farmers' private worlds with simple questioning about good and bad days may be an important entry point for studies of wellbeing, or indeed for practitioners working with farming populations.
The evidence for utility of the two 'good' interview questions has some caveats. The two questions are not universally applicable to rural populations. Non-farmers in the empirical study were uncomfortable with responding to the good farmer question, clearly feeling incompetent to identify good farmers, beyond what made for a good neighbour. When they attempted to describe a good farmer, their notions were quite different from those expressed by practicing farmers. Nonfarmers were outside of the social collective of commercial farmers, who reinforced and upheld good farming standards. Non-farmers lacked visual appraisal skills, supporting Carolan's (2008) contention about the ways in which performance and knowledge are intertwined. The good day question was more accessible, similarly identifying the importance of skills, but this attention was focused on their own holding, not comparison to others. It is thus important to recognize that both the 'good farmer' and the 'good day' questions are situated and suited to particular study cohorts and overarching research questions.
The academic popularity of the good farmer concept also raises the risk of it becoming 'academic shorthand'. The good farmer term does not appear to have been utilized in the interviews (either by participants, or by the interviewer) for a number of 'good farmer' studies. Instead, it appears to have been employed post hoc in order to apply the associated conceptual lens to the data. This is not necessarily indicative of poor practice -new concepts and ideas are an expected outcome of qualitative research, and this was indeed how I came to utilize the good farmer concept for the first time (Sutherland and Burton, 2011). I mention it here to argue for the value of employing social theory in the initial research design. Agee (2009) andBrinkman (2013) have argued that qualitative researchers typically spend disproportionate amounts of time collecting data, paying insufficient attention to planning their research and analysing their data. My promotion of the 'good farmer' and 'good day' questions is not intended to offer a shortcut past detailed interview planning, but to enable planning which increases the richness of the text generated.
The feasibility of qualitative interviewing (in comparison to more time-consuming techniques such as photo elicitation) suggests that it is likely to remain a standard rural sociological practice. In writing this paper, my intention is to contribute to developing a corpus of established interview questions, which will help rural sociologists advance existing approaches and critically appraise new research. Newcomers to the field need not learn the hard way how to enable farmers to talk about cultural scripts, symbolic capital and social norms -or worse, because farmers often do not, come to the conclusion these norms do not exist, or are not influential.

CONCLUSION
More-than-representational approaches have gained popularity across the social sciences, leading to assertions of an 'affective turn' (Carolan, 2015;Hayes-Conroy and Hayes Conroy, 2010;Clough and Halley, 2007). This is a critical area in which rural scholars can add particular value. Rural studies have a strong empirical orientation and track record of attending to what it is to live and work in highly affective environments and engage with a range of more-than-human actants. This expertise offers important opportunities for the field testing of more-than-representational approaches, which is urgently needed. My aim in writing this paper is to assist academics working with farmers and other landholders to pursue methodological excellence and engage with morethan-representational theory in empirical research.
I have demonstrated the utility of the 'good farmer' and 'good day' interview questions for both representational and more-than-representational data collection, along with some caveats, but I wish to insert some further notes of caution. The paper plays at the edges of morethan-representational thinking. There are multiple methods beyond qualitative interviewing to elicit more-than-representational data. I am also not suggesting that more-than-representational research is simply a matter of reinterpreting existing datasets and publications -there is a growing volume of complex theory through which to animate these data, which can illuminate it in radically new ways. Nor is more-than-representational research necessarily preferable to established approaches -clearly, there is a corpus of rural sociological research, which has and continues to take the field forward without adopting this perspective. In particular, the 'representational' research conducted to date addressing the good farmer has enabled the analysis of unconsciously embedded ideals and symbols achieved through farming practices and experiences: it has already achieved much of what more-than-representational research aims to accomplish. My argument is that more-than-representational thinking offers the potential to take our qualitative research deeper into farmers' lived experiences, to critically develop our research methods and to improve our understanding and professional engagement in the events, relations, doings, backgrounds and affects of rural life.  Table 1, precise wordings of this question vary. 2 As one of the reviewers noted, this flexibility in definition is why it is a good term to use in qualitative research, but quite problematic for quantitative research, where it makes it difficult to interpret responses. 3 Other uses of the 'good farmer' term identified by Burton et al. (2021) include references by agricultural scientists, policy makers and indeed some industry professionals, such as veterinarians (e.g. see Shortall et al., 2018) to those farmers whose practices follow expert scientific knowledge or industry standards. Burton et al. also describe historic uses of the term 'good farmer' to assert the validity of one's own experimental practices, and as a rhetorical means of promoting a particular agricultural approach as the 'right' way to improve agricultural production. 4 As of March 2021, google scholar identified some 786 articles have cited Burton's (2004) paper, placing it among the most highly cited papers in this journal. 5 Knabe et al. (2010) utilized the Day Reconstruction Method, also reviewed by Diener and Tay (2014) -a standardized psychometric method. As a quantitative approach, it was not specifically reviewed for this paper. 6 My aim was to be comprehensive, but it is possible that a few articles have been missed. 7 This conclusion is also based on personal experience interviewing UK farmers who clearly did not want to label their peers as 'bad farmers', although they were happy to describe their neighbours' bad farming practices. This distinction may reflect language or cultural differences: in his Belgian study, De Krom (2017) described farmers who indicated that they believed their peers were identifying them as 'bad farmers' for participating in agrienvironmental measures, and one withdrew from the programme for that reason. 8 An assessment based on my own experience in asking both questions, and review of transcripts from this and my other published work.

A C K N O W L E D G E M E N T S
This research has been funded by the Scottish Government's Rural Affairs, Food and the Environment Strategic Research Program (2016-2021. Open access publishing has been funded by the United Kingdom's Biotechnical and Biological Sciences Research Council through their agreement with the James Hutton Institute. I wish to thank Orla Shortall of the Hutton Institute, two anonymous reviewers and the editor for helpful feedback on earlier versions of this manuscript.

D ATA AVA I L A B I L I T Y S TAT E M E N T
The 'good farmer' papers reviewed are included in online Appendix A. The author elects not to share the qualitative dataset utilized in this paper, to protect confidentiality.

C O N F L I C T S O F I N T E R E S T
There are no conflicts of interest associated with this paper.