The symbolic costs of advice: how social situations explain the occurrence of unplanned knowledge sharing interactions

Abstract

Purpose

Informal knowledge sharing interactions (IKSI) are of particular value for innovation projects. This is especially true for unplanned IKSI, because they are even more likely to provide non-redundant knowledge and new perspectives than planned IKSI. Seminal studies have shown that the formation of unplanned IKSI can be explained on the basis of spatial structures. Strictly speaking, however, these studies only explain unplanned encounters. Whether unplanned IKSI result from these unplanned encounters, though, cannot be satisfactorily explained on the basis of spatial configurations alone. The purpose of this paper is to tackle this explanatory gap by unraveling the fundamental social processes by application of the symbolic interaction theory.

Design/methodology/approach

For this purpose, the formation of 132 IKSI on innovation projects from three research and development departments of large companies was recorded in detail using a combination of diaries and interviews. The data were analyzed using qualitative content analysis.

Findings

The analysis reveals that IKSI cause symbolic costs (image damages), and that these costs vary between types of social situations. Because actors anticipate situation-specific costs, their propensity to initiate IKSI can be explained in terms of the situations in which they encounter one another. Furthermore, the analysis reveals six particularly relevant characteristics of situations and further elaborates the basic argument by analyzing their functioning.

Originality/value

The paper complements previous explanations of unplanned IKSI by opening up the social processes underlying their formation.

Citation

Roth, P. (2023), "The symbolic costs of advice: how social situations explain the occurrence of unplanned knowledge sharing interactions", European Journal of Innovation Management, Vol. 26 No. 7, pp. 1-22. https://doi.org/10.1108/EJIM-08-2021-0381

Introduction

Informal knowledge sharing interactions (IKSI) allow actors to receive advice from colleagues who are formally not in charge of dealing with the task at hand. Thus, IKSI provide access to a wider knowledge base. They are, therefore, a key success factor for accomplishing knowledge-intensive tasks such as innovation projects (Allen, 1977; Irving et al., 2020; Reagans and McEvily, 2003). Recent research has revealed that it is useful to distinguish between planned and unplanned IKSI, and that each type accounts for about half of all IKSI (Appel-Meulenbroek et al., 2017; Irving et al., 2020; Roth, 2022). Planned IKSI are the result of partner selection decisions. Research shows that actors tend to select partners with whom they maintain strong ties and which are similar to themselves (Rivera et al., 2010; Lazega, 2020). Further, due to the preference for personal IKSI and on the basis of space-theoretical analyses, it has been shown that partners are preferred who are easy to reach (Small and Adler, 2019; Catalini, 2018; Kabo et al., 2014). By contrast, unplanned IKSI arise more occasionally and bring together heterogeneous and weakly tied actors. As a result, such IKSI tend to provide non-redundant knowledge and new perspectives (Roth, 2022; Toker and Gray, 2008; Irving et al., 2020). This gives them particular value, and it is, therefore, of great interest to understand how they come about and how to foster them in practice.

Previous research has built appropriate theories to explain planned IKSI. However, these theories are insufficient to explain unplanned IKSI. Because relationships and characteristics of the actors play a subordinate role in the formation of unplanned IKSI, the underlying social processes can only be inadequately explained by the theories explaining planned IKSI. The situation is different for space-theoretical approaches. Because unplanned IKSI presuppose unplanned encounters, spatial relations are even more important for unplanned IKSI than for planned ones. However, the corresponding research shows that the sophisticated analysis of spatial structures alone is not sufficient because it does not explain why only some encounters become IKSI and most others do not (Rashid et al., 2006; Appel-Meulenbroek et al., 2017; Khazanchi et al., 2018). The explanatory approach for unplanned IKSI is missing a key puzzle piece.

This paper argues that the puzzle has not yet been solved because the occurrence of unplanned IKSI is also structured by social processes that have been neglected in the absence of an adequate theorization. This paper aims to outline such a theory and to provide empirical evidence for its usefulness. As a fundamental social theory for the development of such a middle range theory, I will rely on the symbolic interaction theory (Blumer, 1969; Stryker and Burke, 2000; Carter and Fuller, 2016). The symbolic interaction theory is particularly appropriate here because it takes a micro-sociological perspective and focuses on explaining interaction processes. With regard to the question of why only some encounters result in unplanned IKSI, it is first assumed that IKSI can cause image damages and thus symbolic costs. Depending on the anticipated symbolic costs, actors are more or less likely to initiate IKSI. Second, symbolic costs are assumed to vary across types of situations because situations operate as frames for interpretation and action. Thus, the situation in which actors encounter one another can be used to explain why unplanned IKSI occur in some cases and not in others. Beyond the elaboration of this theoretical argument, the value of the concept in this field of research is empirically demonstrated by drawing on rich data on the formation of 132 IKSI concerning innovation projects recorded by 41 developers from three research and development (R&D) departments of large companies. The exemplary application of the concept reveals six characteristics of situations that hinder or favor unplanned IKSI. Here they serve to illustrate the value of the developed concept. Beyond that, however, these characteristics help to further elaborate our understanding of IKSI and opportunities. With its conceptual and empirical work, the paper, thus, contributes to our understanding of the social processes underlying the occurrence of unplanned IKSI and thus complements the space-theoretical approach.

State of research: explaining informal knowledge sharing interactions

Explanations of planned IKSI largely focus on social selection processes (Rivera et al., 2010; Borgatti and Cross, 2003). Based on network theoretical concepts, partner preferences are explained by the existence of strong, reciprocal or indirect ties between actors. Furthermore, actors prefer IKSI partners with whom they share personal values and characteristics. Following these strong preferences, they tend to ask persons from their close social environment for advice. Planned IKSI are, therefore, likely to connect individuals with similar knowledge and perspectives. While planned IKSI can be explained to a large extent in terms of enduring social structures (social networks and personal characteristics), these structures are less important when it comes to unplanned IKSI. Research shows that such often occur even between heterogeneous and weakly tied actors (Roth, 2022; Irving et al., 2020; Toker and Gray, 2008). The social selection processes that explain planned IKSI to a large extent can, therefore, only insufficiently explain unplanned IKSI.

Additionally, IKSI are also explained on the basis of spatial structures. Thereby, two dimensions of space have been considered in particular: proximity and openness. Proximity promotes IKSI by creating opportunities and reducing effort (Allen, 1977; Catalini, 2018; Kabo et al., 2014). In early studies, spatial proximity was conceptualized statically and operationalized as the distance between workstations. More current studies, however, increasingly conceptualize proximity in dynamic terms. They consider how individuals move through space and which proximity–distance relationships result from this (Small and Adler, 2019; Catalini, 2018). More specifically, spatial layouts and the arrangement of places relevant to everyday life (such as offices, laboratories, entrances, meeting rooms, coffee kitchens, etc.) are analyzed. This allows to create and compare individual movement profiles. The greater the overlap of the movement profiles, the less additional effort is required to visit colleagues, and the more likely it is that they will encounter each other. The dynamic analysis of space, thus, helps to explain both planned and unplanned IKSI. In the case of planned IKSI, distance is a cost factor insofar face-to-face interactions are preferred. However, since distances within organizations are usually small and can be bridged by media in the case of planned IKSI, the impact of spatial proximity is much smaller than that of social proximity. The opposite is true in the case of unplanned IKSI. Since they presuppose encounters without an appointment for IKSI, the space-based explanation of unplanned encounters is central to their understanding. However, it also turns out that explaining encounters is not sufficient, since only a small proportion of those encounters result in unplanned IKSI, and, so far, it has not been possible to explain why (Rashid et al., 2006; Appel-Meulenbroek et al., 2017; Irving et al., 2020). Openness is the other dimension of space that has received greater attention in research on interactions in organizations. The starting point for this research is the assumption that open space designs reduce physical contact barriers and thus promote the informal sharing of knowledge (Allen and Henn, 2007; Boutellier et al., 2008). In the meantime, however, numerous studies have shown that open spaces also have negative effects on the informal exchange of knowledge (Khazanchi et al., 2018; Bernstein, 2012; Fayard and Weeks, 2007). Informal interactions in public spaces are avoided because actors value privacy and do not want to disturb colleagues at work. The increased number of encounters through openness, thus, does not result in more informal communication, due to the associated lack of privacy. Privacy describes the possibility to control the limits of an interaction (Altman, 1975). Places where uninvolved third parties could overhear or enter the setting lack privacy. To conduct planned IKSI, actors choose private places. Unplanned IKSI, however, require encounters. By definition, these are extremely unlikely to occur in private places. Thus, taking privacy into account can only explain to a very limited extent why some encounters result in unplanned IKSI. Further research is needed here, which explains why some encounters result in unplanned IKSI (Figure 1) (Appel-Meulenbroek et al., 2017; Irving et al., 2020; Khazanchi et al., 2018).

The studies on openness and proximity have been criticized for considering space as absolute and thus ignoring the close intertwining of space and social processes (Cnossen and Bencherki, 2019; Löw, 2008; Carlile et al., 2013). Some seminal studies have already succeeded in overcoming this shortcoming. Fayard and Weeks (2007) show that comparable copy rooms in three different research organizations are differently integrated into the everyday social practice of employees and thus have different effects. Roth and Scheidgen (2022) show that open spaces and canteens in three different R&D departments are associated with department-specific practices that enable or prevent IKSI. Both studies show that the concrete impact of space depends on the structure of social practices and can vary between different social contexts. These differences show how important social processes are for the effect of material arrangements on the formation of IKSI. In addition to the differences, however, it seems plausible that IKSI are formed according to more generalizable social rules, and that taking these into account helps to explain under which conditions encounters result in unplanned IKSI.

Conceptual framework: the symbolic costs of advice and their situational variance

Drawing on basic ideas from the symbolic interaction theory (Blumer, 1969; Stryker, 2017), in this section, I will develop the argument that symbolic costs are crucial to the willingness to initiate IKSI, and that these costs vary among types of situations. Thus, situation-specific symbolic costs can explain why some encounters result in unplanned IKSI and others do not. I will first explain how symbolic costs influence the formation of IKSI and then how these costs are moderated by situations.

Symbolic costs

Key to the explanation of action by the symbolic interaction theory is the assumption that actors, even if they pursue other, more immediate goals, always also strive to cherish their own image (Goffman, 1966; Stryker and Burke, 2000). Accordingly, they strive to create positive impressions and avoid negative ones. Symbolic costs I define as the negative impressions concerning an actor's image.

The interest in one's own image derives from two assumptions that Blumer (1969) describes as fundamental to the symbolic interaction theory. First, he notes that actors do not refer to an objective environment, but ascribe meanings to other people, things and events. These meanings include not only abstract ideas but also practical impulses for action. Accordingly, material and social elements usually do not simply stand for themselves and act mechanically, but become symbolically effective. Second, he assumes that these meanings are produced, reproduced and mediated through processes of interaction. Accordingly, meanings are not simply given, but are interactively made by actors. This refers both to the long-term development of abstract concepts in the course of numerous interactions and to the enactment of such concepts in an interaction.

The meanings attributed in this way structure action fundamentally, above all because they remain largely unquestioned in their everyday application, and actors, therefore, reliably align their actions according to them. At the same time, however, actors reflect that others also act this way, and that they themselves are constructed interactively. Although these are “only” constructions, they are existential insofar as they cause how others behave toward themselves. Because the construction of their self is also an interactive process in which they themselves participate, they strive to bring about advantageous attributions and avoid disadvantageous ones (Alexander and Wiley, 1990; Stryker, 2017). Actors are able to do this because they learn to see themselves through the eyes of others over the course of numerous interactions (Mead, 1934; Cooley, 1902). Who they appear to be to others is something actors can only influence to a limited extent because, for example, they are tied to physical dispositions, competencies or courses of interaction (Goffman, 1961; Alexander and Wiley, 1990). Within these bounds, however, actors shape their actions toward how others will perceive them and develop corresponding routines (Alexander and Wiley, 1990; Stryker, 2017). Expectations about how others will perceive themselves are, therefore, an important motive of action, and their consideration helps to explain it.

Individual participation in various social relations should be analyzed not only in terms of the material interests that operation of the institutions serves, but in terms of the symbolic meaningfulness of that participation (Friedland and Alford, 1991, p. 250).

Symbolic costs describe the negative consequences for the perception of themselves by others. They are called costs because actors anticipate them and put them in relation to other goals. Recurrently, actors are confronted with conflicting goals when pursuing immediate objectives while trying to maintain their image. Hungrily serving yourself at a buffet that has not yet opened may satisfy your hunger but could harm your image. Symbolic costs occur whenever actors face undesirable attributions through their actions. Insofar as actors pursue various goals beyond image maintenance, it is possible that symbolic costs are anticipated and willingly accepted. In other cases, such goals are put aside because the anticipated symbolic costs appear too high.

Several empirical studies show that symbolic costs are of particular importance for the occurrence of IKSI because asking for advice in general “is risky in the sense that the individual seeking knowledge necessarily reveals his ignorance, which could undermine his standing in the organization” (Tortoriello et al., 2012, p. 1,026). This is especially true when a specific expertise is significant in an organization and employees identify with it. Thus, IKSI carry the risk of symbolic costs, and there is, therefore, a tendency to keep the number of IKSI low. In addition, Peter M. Blau showed already in 1955 that symbolic costs influence the choice of partners. In his investigation, he shows that IKSI partners are selected on the basis of competence and trust. With respect to competence, he concludes that colleagues are preferred as advisors who are not highly competent, but who have a similar level of competence as the person seeking advice. This seems irrational at first because more competent colleagues could give better advice. However, because they are expected to reveal gaps in their knowledge, higher symbolic costs are anticipated and, in consideration of these, less competent colleagues are preferred (Blau, 1955).

Further, the anticipated symbolic costs depend not only on whether knowledge gaps are disclosed, but also on how sympathetically and discreetly they are treated. Colleagues who tend to show a lack of empathy and attack others based on their knowledge gaps are eschewed as advisors, regardless of their competencies (Casciaro and Lobo, 2008). Instead, colleagues with whom relationships of trust exist are preferred. Such relationships are again likely to exist between similarly competent actors who have mutually disclosed their knowledge gaps to each other in previous IKSI and have thus developed more trust in each other IKSI by IKSI (Borgatti and Cross, 2003; Rivera et al., 2010; Lazega, 2020).

In addition to the risk of revealing knowledge gaps, Casciaro et al. (2014) show that revealing a morally less legitimate motivation also causes symbolic costs. If the impression might arise that one colleague is taking advantage of another, IKSI are more likely to be omitted to avoid symbolic costs. Strong ties, in which actors repeatedly reciprocate, prevent the impression that one is taking advantage of the other. To protect their image, actors therefore prefer partners with whom they maintain close ties.

This research shows that anticipated symbolic costs are key in predicting willingness to initiate IKSI, and that they vary based on relationships and personal characteristics of potential advisors. I argue that the symbolic costs associated with IKSI can also vary between social situations, and that systematic consideration of social situations can explain why some encounters result in unplanned IKSI and others do not.

Social situations

Although social situations have not been considered so far, their importance is already indicated in Blau's seminal investigation. He describes consultations in disguise as a way to reduce symbolic costs. Consultations in disguise are characterized by the fact that advice is not openly asked for. Instead, an anecdotal report of a project is given first. Following Blau's observations, this often leads to listeners giving unrequested advice on the project described (Blau, 1955, p. 133f.). Symbolic costs are lower because the reporting actor does not have to openly express his need for help to get advice. According to Blau's brief analysis, the prerequisites for such an approach are interesting cases that legitimize a report. The data on which Blau bases his findings were obtained by observations of collective work breaks in an agency. They are, therefore, limited to a very specific type of situation. A systematic comparison between types of situations has not been undertaken. I argue that it is due to the focus on this specific setting that the importance of interesting cases has been overestimated and the relevance of the situation has been neglected. In what follows, I will introduce the concept of social situation, which is also a key category of the symbolic interaction theory, and argue that situations can moderate symbolic costs.

For the first part of my argument, I have emphasized, following the symbolic interaction theory, that meaning is interactively produced and shared. For the second part of the argument, it is now important that this is also true in reverse: interactive action is oriented toward symbols.

Actors generally have an interest in how others will interpret their actions because how they will behave toward them in the short and long term depends on their interpretations. They are able to shape their actions to trigger certain interpretations because they share knowledge about what certain symbols and combinations of symbols mean. At the same time, the ability to express themselves is limited by the fact that interactions take place in situations that also provide meaning (Goffman, 1974; Blumer, 1969). Accordingly, what an actor's actions mean to others also depends on the situation in which they occur. Characteristics of situations include, for example, other actors, things or abstract concepts as well as their temporal and spatial arrangement. Because actors share knowledge about what specific combinations of these characteristics mean and, moreover, indicate to each other how they interpret the situation, they usually come to a shared definition of the situation (Schulz-Schaeffer, 2008). Situations thus act as a shared framing of interpretation and action (Blumer, 1969; Kroneberg, 2014).

Situations enable and constrain interactive action at the same time (Kroneberg, 2014; Alexander and Wiley, 1990). On the one hand, specific actions with certain meanings only become possible through symbolic or material references to characteristics of situations. On the other hand, the symbolic and material framing provided by a situation limits the possibilities of expression. David could not have hit Goliath with his slingshot in a situation without a stone and if the scene had not been embedded in a war, and if Goliath had not challenged David to fight, David's murder would not have been interpreted as a heroic deed. The meaning of the situation, which becomes clear here in retrospect, also influences action in the acute situation. The basis for this are mutual expectations. Given the situation, interaction partners can estimate how others will interpret and react to their behavior (Alexander and Wiley, 1990; Goffman, 1959). Following the interest in interpretations and reactions, actors therefore shape their actions in a situation-specific way and develop corresponding practical knowledge. The application of this shared practical knowledge reliably leads to the reproduction of situation-specific patterns of action, here termed interaction orders.

With the concept of the situation, matter becomes tangible through its practical meaning. Things and their spatial orderings are considered in their entanglement in social processes (Cnossen and Bencherki, 2019; Löw, 2008; Carlile et al., 2013). Following Goffman's (1959) theater analogy, things are understood here as stage props and spaces as stage sets. As elements of situations, they enable and stimulate actions (with specific meanings).

Symbolic costs are anticipated by actors seeking advice, and IKSI are undertaken or omitted as a function of them. I argue that symbolic costs also vary depending on the situation in which actors encounter each other because specific types of situations facilitate favorable initiation practices and/or suggest favorable interpretations of such. Accordingly, the symbolic costs associated with IKSI are lower in such situations. Because actors anticipate this, the willingness to initiate IKSI is greater in these situations. Situations can, therefore, be used to explain why some encounters result in unplanned IKSI and most others do not (Figure 2).

Setting and method

Why do some encounters result in unplanned IKSI and others do not? I argued that this question can be answered by considering how situations moderate symbolic costs. Subsequently, the empirical value of this theoretical argument will be examined. To this end, I conducted an exploratory study of the occurrence of IKSI. In doing so, it will be examined whether the expected interrelationships can be identified through the lens developed in this paper. By achieving this primary goal, the abstract argument will be further specified based on the empirical data so that it can be described more precisely how situations affect symbolic costs and how these affect the formation of IKSI. In the following section, I will outline the empirical procedure.

Selection of participants

To be able to investigate the emergence of IKSI, I decided to recruit participants for whom IKSI play an important role, and in whose everyday life, IKSI can emerge in a larger number and variety. For this purpose, it seemed significant to me, on the one hand, that the activities of the persons induce a high and diverse need for IKSI. Therefore, people were selected who are responsible for innovation projects in the central development department of innovative companies in knowledge-intensive industries. Although these individuals are highly qualified, their projects repeatedly take them to the limits of their knowledge, which means that they can benefit from the advice of other experts (Allen, 1977). In addition to the above-mentioned need for advice, the number and range of observable IKSI depend on the opportunities to conduct such in everyday life. Due to confidentiality agreements, developers are incentivized to primarily find IKSI partners within their own organizations. Therefore, companies with large numbers of employees in R&D and related fields were selected. In selecting the department, I also ensured that it was located in a central site so that various opportunities for unplanned IKSI would arise in everyday life. Within these requirements, a list of 31 companies was compiled whose research centers were within acceptable proximity to my location. The final factor in selecting among these was the willingness of those responsible in the organizations. To keep the effort associated with data collection within manageable limits, and to ensure greater comparability between cases, all project managers from the R&D department were included in each case. The sample consists of 41 developers, distributed among three corporate R&D departments of large companies located in Germany and Switzerland and which are primarily active in the fields of material science and electrical engineering. Table 1 gives an overview.

The participating companies in total employ between 8,000 and 10,000 people in R&D and invest between €2bn and €5bn per year in this area. Sociodemographically, the participants are relatively homogeneous. The vast majority are male, have a doctorate and are middle-aged.

Data collection

In the seminal studies on unplanned IKSI, especially observation methods were applied (Fayard and Weeks, 2007; Backhouse and Drew, 1992; Irving et al., 2020). Following on from the conceptual considerations developed in the previous section, alternative data collection methods appear necessary and promising. Four main requirements guide the design of the procedure:

1. It is necessary to record encounters occurring in as many different places as possible to analyze the greatest possible variance of situations. The insightful research of Fayard and Weeks (2007), for example, is limited to copy rooms so that only very specific types of situations can be analyzed.

2. Given the importance of privacy, it seems appropriate to avoid observer participation in the situation.

3. Since situations and their meaning are defined by actors, it is necessary to reconstruct the situations from the actors' perspectives. Recourse to observation data alone limits the possibilities of analysis substantially.

4. The reconstruction of the emergence and situation of IKSI from the perspective of the actors, however, carries the risk of memory gaps or memory errors. The methodological procedure should, therefore, be designed to counteract such.

To meet these requirements, I combined event-based diaries, focused interviews and ethnography (Roth, 2015). To be able to capture a wide range of IKSI independently of time and space and without influencing them as observers, they were recorded by the participants themselves using digital diaries prepared for this purpose (Nezlek, 2012; Kunz, 2018). Participants were instructed to document interactions whenever they discussed topics related to current development projects with people formally not involved in them. The documentation sheet included semi-standardized questions about how the IKSI came about, the situation and the outcome (Appendix 1). After recording, the completed documentation sheet was immediately sent to me. To explore the occurrence and the situation of the IKSI in more detail and from the perspective of the actors, interviews were conducted with the participants. The information from the diaries was used to prepare the interview guideline (Appendix 2). Contrary to the interviewees' tendency to talk mainly about the ideas addressed in the IKSI, my questions prompted them to give a detailed and in-depth account of each step from the occurrence of the encounter to the IKSI and to shed light on the specific situation and its significance in this regard. Usually, the interviews were conducted on the same day and in person. In addition to the short duration between IKSI and interview, recall biases were counteracted by the process of filling out and the presence of the diaries in the interviews (Roth, 2015; Alaszewski, 2006).

Participants were also asked in the interviews to contextualize the described events and situations in their everyday experience. Through this contrast, the interview partners were encouraged to bring out further characteristics of the situation and to describe their significance for the occurrence of the IKSI. To be able to classify and contrast the described cases even better and more independently, I occupied an office in the respective department during the research period, participated in everyday activities and made observations at places that seemed relevant. In doing so, I took field notes that primarily described spatial conditions, patterns of everyday practice and the resulting occurrence of encounters and interactions. In addition to my general considerations, indications from the interviews were pivotal for the selection of these places. The ethnographic data allow an appropriate contextualization of the highly focused information about IKSI. In particular, they make it possible to understand situations in their everyday embedding and to compare them with situations in which encounters did not result in IKSI.

Data collection was conducted in each department for one month. Altogether, 132 IKSI and about 15 h of interview were thus recorded. Considering the corona crisis and the related changes in working conditions, it seems important to mention that the data were collected in 2017.

Data analysis

To analyze the data with regard to the situational variance of symbolic costs, a three-step procedure was applied. This enables the different data sources to be integrated and cases to be systematically compared to identify patterns that indicate causal relationships.

In the first step, I structured and merged the data from the various sources. First, interviews were divided into individual IKSI cases. Based on the information from the interviews, further information from the diaries and field notes was assigned to these. By integrating the data in this way on a case-by-case basis, all the information collected that pertained to the case was immediately available when the cases were analyzed (Yin, 2012; Eisenhardt, 1989).

The second step aims to systematically describe the cases and thus create comparability. Following the established procedure of qualitative content analysis, each individual case was analyzed in detail (Miles et al., 2014; Eisenhardt, 1989). The focus was on the questions of which characteristics of situations were relevant for the interactive formation of IKSI and how. To answer this question, I developed codes to describe each case abstractly and with sufficient precision. Individual codes refer to elements of situations such as types of activities, things or people (Appendix 3). Whenever cases could not be described satisfactorily with the existing codes, the code system was supplemented (e.g. new type of activity) or made more precise (e.g. distinction of things) (Miles et al., 2014). To apply the elaborated code system to all cases, the entire material was coded again after the first run.

In the third step, the cases were systematically compared with each other on the basis of the codes to identify generalizable interrelationships (Eisenhardt, 1989). For this purpose, abstract descriptions of how characteristics of situations influence symbolic costs and thus the occurrence of unplanned IKSI were developed on the basis of individual cases. The basic vocabulary of these descriptions were the assigned codes. These abstract rules were then tested using the data material. For this purpose, the codes were used to find comparative cases in which similar characteristics of situations or similar mechanisms were identified. In comparing these cases, the abstract descriptions were challenged. If they were confirmed by a comparative case, they were retained. If comparative cases contradicted the rule, a reformulation of the rule was sought that would allow the contradictory cases to be explained as well. If this was not successful, rules were dropped. As a result, statements about the influence of characteristics of situations on symbolic costs and the occurrence of IKSI were developed on the basis of the data. These statements are valid for all cases considered here, and at the same time are so precise that they contain empirically substantial information.

Findings

Asking colleagues for advice generally bears the risk of generating symbolic costs. First of all, this is because every developer is formally responsible for a project and should be able to manage it independently and acquire the necessary knowledge because of his or her high professional qualifications. Asking others for advice runs the risk of giving the impression that a developer is not able to cope with his/her task and is in urgent need of support from colleagues. Further, informal advice usually requires the disclosure of the developer's problems and knowledge gaps. IKSI are therefore delicate in general.

As expected, the data show nevertheless that open requests for advice do exist. However, a comparative analysis of these cases shows that in all cases, conditions were given that reduce symbolic costs. This is the case when:

1. Developers maintain a close and trustful exchange relationship with the contact person (A1802; A2002; B1035; C0735f.);

2. Developers ask colleagues who are recognized as clearly superior because of their substantially greater experience (B0406; B0904; C0304);

3. Decisions relating to the work of the addressed colleague are discussed (A1605; C0312; C0206);

4. Developers ask questions on issues for which they are clearly not experts but the contact person is (A0348; A1812; C0615ff.; E0204);

5. Very specific questions are asked that show the developers high expertise (A0332; A0501; C0211; C0304; C0504; C0722; D0202); and

6. The contact persons could not negatively influence the image of a requesting developer because they do not belong to the developer's professional community or maintain contact with other members of the community (A1726; A2002; C0620; C0632).

Accordingly, open requests have specific prerequisites. They only come about when (1) very specific questions arise that can meet the mentioned requirements and (2) a developer seeking advice can reliably determine that the conditions mentioned are met based on his/her knowledge about the expertise of a potential advisor. Considering the symbolic costs, the willingness to openly ask for advice is limited to a small circle of questions and persons.

However, the analyses further show that the symbolic costs are significantly lower under certain situational conditions, and that the above-mentioned limitations can thus be overcome. A crucial factor here is that IKSI (apparently) result from the situation. If IKSI result from the situation, the symbolic costs that arise for the recipient of the advice are low because they have not openly asked for advice. An urgent need for advice is not expressed, and the expertise of the developer is not challenged. This also applies when current problems and open questions are expressed. Since the IKSI have resulted from the situation, it is not attributed to an urgent need. Rather, it then seems obvious that the advice receiving actor could have solved his/her current problem even without advice, if the encounter had not happened by chance and the advice was only taken because the opportunity occurred.

Seeking opportunities

The impact of this mechanism can be recognized on the one hand in the characteristics of situations in which encounters result in IKSI, which I will discuss in detail below. On the other hand, the importance of symbolic costs is demonstrated by actors' strong interest in IKSI that (apparently) arise from the situation. It is expressed above all by the fact that they seek for adequate situations and sometimes deliberately try to create such. An example for this is the initiation of IKSI that took place during an internal fair that one of the participating developers visited. There he noticed a colleague who works as a developer in another division. They know each other because they worked together years ago on a cross-divisional development project.

I did not go there OK, I now see X and want to talk with him about the topic. But I go there, encounter X and then ahh, ok, now I could talk with him about the topic. Of course, I did not contact him immediately. I did it skillfully: First of all, the poster Oh interesting, then you request some explanations by the employee of X and then you get to the point. He comes up and from the conversation one slowly comes to the topic. The poster was primarily a good measure to initiate the interaction! (E12I1:08.13)

Other examples of such skillful initiation strategies are a fictitious conversation with a receptionist, with the aim of casually entering into an interaction with her supervisor (A0215), scheduling one's own lunch break to ensure that a certain colleague can be encountered in the canteen at that time (B0704) or keeping an eye out for contact persons in the coffee kitchens by developers who can see or hear who is there from their office (A0721). On the one hand, these examples demonstrate how important symbolic costs are to the actors. On the other hand, they also show how much they depend on characteristics of situations that promote the occurrence of unplanned interactions and IKSI as well. Regardless of how much an actor intervenes dramaturgically or is carried by the situation, the same characteristics of situations are crucial because they trigger the unplanned occurrence as well as they facilitate their deliberate initiation. Hence, more productive than the question to what extent actors planned IKSI intentionally seems to be whether they appear unplanned. In line with the interactionist assumptions, it therefore seems worthwhile to identify the situational characteristics that enable this appearance.

The importance of such characteristics is further highlighted by the fact that all situations that have led to IKSI in our dataset can be defined by individual characteristics or combinations of these, and that, in contrast, numerous encounters in situations without these characteristics have not led to IKSI. In the following, I will describe these characteristics.

Fleetingness

To enable IKSI to emerge from the situation, an interaction must be initiated without expressing the intention to receive advice. If colleagues are connected by strong ties, open interactions are common, and their initiation is not linked to special conditions. In all other cases, adequate situations are required. The data clearly show that most everyday encounters are inappropriate because they are fleeting. Despite the large number of fleeting encounters I observed, none of the IKSI analyzed resulted from such an encounter.

Considering the symbolic costs of initiating IKSI, this finding seems highly plausible. The initiation of an interaction from a fleeting encounter would require the potential partner who might typically be on his/her way to an appointment and therefore in a hurry, to stop. This interruption requires a justification. The need for IKSI must be articulated accordingly. Moreover, the interruption of the other's movement gives the impression of urgency. Accordingly, initiating IKSI from fleeting encounters causes high symbolic costs and is, therefore, rather omitted.

Availability

By contrast, situations in which actors are close to each other and appear available offer favorable conditions. Especially, when actors are at least slightly acquainted with each other, such situations are defined as interaction situations. It is even perceived as awkward if it is not managed to initiate and maintain an interaction. There is, therefore, a mutual interest in an adequate topic of interaction. Among colleagues, it is obvious to discuss the current R&D projects, because, even without closer knowledge of the person, it can be assumed that they are able to comment on it, and at the same time, the topic is not intrusive or banal. Talking about a current development project and taking the advice is, therefore, perceived as the solution to the shared action problem (overcoming awkward silence) and not as an expression of urgent need for help.

For IKSI to arise from unplanned encounters, therefore, situations are needed in which actors are continuously in one place and are not very busy. That both characteristics are present is by no means a matter of course in the development department context, since the developers are very busy or at least interested in giving this impression. This is stated in many interviews (e.g. A1518; A1412; B0119). Therefore, legitimization is needed to enable potential IKSI partners to be barely busy and in closeness, proximity to each other. In the cases analyzed, legitimacy is given by actors being held up rather than staying voluntarily. This is mainly due to machines and equipment whose use requires little attention from the users, but nevertheless encourages them to stay. Examples of this are coffee machines, copiers and traffic lights (e.g. B0115; C0508f. B0303; A0614; B0506; B0308; B1020; D0102; A0719). For illustration purposes, I will briefly describe one case (A1507):

In one of the companies, it is necessary to cross a larger street to reach the gate from the car park or bus stop. The crossing is controlled by a traffic light. On the way from the car park, a developer encounters a casually known colleague from another area. They just greet each other casually at first. “It would have stayed that way if we had not stopped next to each other at a red light.” There the colleague starts an interaction by asking about the developer's current project. On the shared way across the street and through the turnstile, the interaction gains a certain depth. Although their paths separate behind the gate, they stayed together for about five minutes to conclude the interaction.

It is only by being held up by the traffic light that availability is created, which provokes interaction and leads to the unplanned IKSI.

Things inviting interactions

Another characteristic of situations that fosters the initiation of open interactions are events or things in the shared environment that are worth talking about. Through the shared attention directed at such an artifact, interests are communicated nonverbally. Thereby, interactions and topics are suggested. In particular, when artifacts are thematically related to a developer's work, they facilitate the initiation of IKSI. Because the IKSI arise from the situations, the symbolic costs are correspondingly low. A particularly significant example of this in the cases analyzed are experimental setups (B0402; B0404; C0204; C0347; E0202). They are very much present in the everyday life of all development departments and directly invite discussion about the current project. The special importance of such objects is evident from the fact that they play a major role in the few IKSI with strangers.

The conversations arose in the experimental field. This is a public area. Anyone who works in the company can walk through it. Colleagues are not only there because they have something to do there, but also simply because they are curious and want to see what is happening there. But those who come by are often people who conduct tests and then have a look at what the others are testing. The colleague with whom I exchanged ideas there works in purchasing. He wanted to do something completely different in the test field, but then he saw me, looked at the assembly and asked questions (B0402ff.).

As the example shows, the advantage of having artifacts representing the project is not only that they make it obvious to talk about the project, but that it puts the developers in the very comfortable position of being approached. To assume that the developers are in urgent need of help seems absurd in this constellation, and the risk of causing symbolic costs is correspondingly low.

Restrictive interaction orders

So far, I have mainly focused on the initiation of interactions that can become IKSI. However, situations still influence the symbolic costs of IKSI when actors are already interacting. In this respect, the interaction orders associated with the situation are particularly important. While the interaction orders are relatively open in unplanned encounters, many other everyday situations in organizations have restrictive interaction orders that prevent informal interactions.

This is especially true for formal meetings. In such meetings, interest in asking (specific colleagues) for advice arises, especially often because colleagues with complementary expertise come together and show their expertise in professional discussions. At the same time, the definition of the situation also justifies the assumption that the actors involved are primarily interested in achieving the (usually explicit) objectives of the meeting. The initiation of IKSI, therefore, appears to be a violation of the situational interaction order. The break generates symbolic costs, as it is interpreted as an expression of urgent need for help. The tensions that arise here are particularly evident in the fact that a comparatively large number of IKSI are realized in the dissolution phase of meetings, when the restrictive interaction order is no longer valid, but the colleagues and the desire to ask them for advice are still there:

Well, that happens quite often when it's a big meeting. There's a very strict agenda. You think to yourself, there are actually three colleagues with whom I actually wanted to talk about something, which does not happen. Then it is advantageous if you have a little time afterwards (A2010).

The effect of restrictive interaction orders does not necessarily require a formalization of the procedures. This becomes apparent in the case of lunch breaks. Here, a violation of the informal interaction order would cause the same symbolic costs as in the case of formal meetings:

During the lunch break people have other topics they want to talk about and not about one's project. Then more entertaining topics are appropriate (C0628).

However, if the number of people involved in the situation allows a deviation from the interaction order to be coordinated casually, deviations without generating symbolic costs are possible:

The conversation digressed and since we both had an interest in following the conversation, it was fine. We were two of us at that moment. […] If the others had come earlier, that would not have been possible (A0616).

Time horizon

Furthermore, the analysis shows that the time horizon of a situation is important for whether a conversation turns into IKSI. If actors assume that the shared situation will only be of short duration, IKSI seem inappropriate because it would take more time due to the complexity:

Sometimes there is a jam at the machine. Then it happens that there are conversations. But then it's usually not about the projects. We talk about something trivial. Because there is not enough time for you to immerse yourself in a longer conversation and get into the subject matter (C0344).

Considering the situation, the IKSI could be perceived as a hold-up of the partner in dialogue. This would express urgency and give rise to symbolic costs. In situations with a longer time horizon, the willingness to initiate IKSI is, therefore, higher.

Publicity

Finally, in line with previous findings on the importance of privacy, the data show that the publicity of situations can prevent IKSI. Using the concept of symbolic costs, it is possible to describe more precisely how privacy encourages IKSI. As expected, the ethnographic observations show that many encounters occur in public situations, and that communication occurs regularly in these situations. However, both the observations and the data on the IKSI show that IKSI are proportionately rare in these situations. The crucial factor here is that actors in public situations are more likely to expect harm to their image. However, the effect of privacy is different from the other characteristics of situations described so far. Privacy does not work by encouraging IKSI to (apparently) emerge from the situation. Rather, such situations encourage the expectation of image damage through sharp criticism or indiscretion.

When problems are addressed in private situations, this indicates the confidentiality of the information. Furthermore, by taking a particular person into confidence, a commitment to treat the matter confidentially and constructively is imposed. This is not the case in public situations. Especially for bystanders, who were not directly addressed but can still participate, the mentioned commitments do not arise. Because competent criticism of the shortcomings of others is a widespread way of showing one's own superiority, there is an incentive to take advantage of corresponding opportunities – especially if there is an audience. Asked about their unwillingness to conduct IKSI in public situations, the developers refer to the increased risk of attacks by bystanders:

When you address your own problems in public, you put yourself in a position in which you are vulnerable, or in which you can be attacked. And unfortunately, there are people who use the situation to do just that (A1723).

In addition, there is the expectation that information about shortcomings and errors that are disclosed in public situations could be handled indiscreetly by bystanders. Symbolic costs can thereby arise from the mere dissemination of information. Moreover, individual pieces of information can be taken out of context and communicated in such a way that an even more unfavorable picture results:

Somehow all projects here are in a certain competitive situation. There's a struggle for resources, for staff, for money, for budget, and if there are problems, nobody wants them to be a topic of discussion here or to be played on as a Chinese whispers. If we have bad results in a particular aspect, it is then suddenly said that the whole project is a mess (A0728).

Because of the wide reach of this communication practice, the image damage for a developer is particularly high. Thus, in light of the symbolic costs, it is possible to specify why less public situations are preferred for IKSI.

Conclusion

In current organizational research, informal processes and structures are understood as an integral part of organizations (Ahrne and Brunsson, 2011). Unplanned IKSI are essential for the distribution of knowledge in organizations. Although they often seem random to the actors, systematic analysis reveals that they can be explained by tracing the genesis of opportunities. This paper contributes to sharpening the understanding of what defines good opportunities for IKSI. So far, it has been assumed that good opportunities are characterized by a low logistical effort for initiating IKSI (Small and Adler, 2019). Based on this assumption, it seemed sufficient to account for encounters to explain IKSI, and it seemed promising to stimulate them by reducing physical barriers and distances (Boutellier et al., 2008). Although the assumption is certainly true, it turned out to be insufficient. Encounters explain IKSI only to a very limited extent, and stimulating encounters in practice has all too often not led to an increase in knowledge transfer (Khazanchi et al., 2018; Rashid et al., 2006). So, what characterizes good opportunities beyond copresence?

To answer this question, the paper draws on the findings from research on knowledge transfer. This research has shown that the bad habit of some individuals to strive for a positive self-presentation can hinder the transfer of knowledge (Tortoriello et al., 2012). By contrast, this paper follows the symbolic interaction theory in its assumption that self-presentation is an integral part of everyday life and hence a key driver of action in organizations. It then conceptualizes negative impressions concerning the image as symbolic costs and argues that (1) the initiation of IKSI bears a particular risk of causing such costs, (2) these costs vary across types of situations and (3) actors anticipate the situation-specific costs and tend to initiate IKSI to a greater or lesser extent. Unplanned IKSI can, therefore, be explained in terms of the social situations in which actors encounter each other (Figure 3).

By developing this argument theoretically and empirically, the paper makes two important contributions.

The main contribution is to offer an approach to explaining the social processes underlying the occurrence of unplanned IKSI. To this end, previous research has focused solely on spatial structures. In addition to explaining encounters, these have been found to be relevant because they can facilitate privacy and actors prefer IKSI in private situations. This preference has been attributed to the reluctance to disturb and to be disturbed while conducting IKSI (Bernstein, 2012; Khazanchi et al., 2018). This paper shows that the privacy of the situation further affects the benevolence of IKSI partners in the situation as well as the number of people discrediting information will reach. Hence, the privacy of a situation can reduce symbolic costs. Accordingly, the previous findings on privacy can be consistently integrated and deepened.

However, the explanatory approach outlined here significantly goes beyond this. The empirical analysis has shown that developers often do not ask their colleagues for advice because interrupting them while they are working carries the risk of expressing an urgent need for help and thus causing symbolic costs. By contrast, symbolic costs are low when IKSI (apparently) arise out of the situation, because the IKSI is then more likely not to be attributed to the need of the advised developer. Characteristics of situations allow IKSI to be initiated out of the situation because they enable actions with specific meaning by providing props through which certain actions become possible and by acting as shared interpretive frames that influence the meaning attributed to actions. Actors anticipate the situation-specific symbolic costs and their willingness to initiate IKSI varies accordingly. Based on situational characteristics, this approach explains why some encounters turn into unplanned IKSI.

A particular strength of the approach is that it does not eclectically juxtapose social processes to spatial structures, but rather systematically integrates spatial and material characteristics with the concept of the social situation. Although all the characteristics identified are linked to social interactions, some of them clearly refer to material characteristics of situations. Therefore, they also allow to make assumptions about how IKSI are encouraged by the spatial design of work environments. Three aspects seem particularly significant here. First, a space conducive to IKSI should legitimize that employees spend time there without pursuing an occupying task. This is facilitated by holding them up with less challenging activities (e.g. by operating a copier). Second, a medium level of privacy seems adequate. On the one hand, a room should be open enough to allow diverse actors to meet there unplanned. On the other hand, the room should be closed enough that actors can reliably determine who can hear and see them. Coffee machines or photocopiers, which legitimately hold up actors, only fully unfold their IKSI-promoting effect if they are placed in spaces providing a medium level of privacy and not, for example, in the middle of an open-plan office. Third, interactions or even IKSI can be stimulated by artifacts that give such occasion and subject. In our cases, the things were primarily experimental setups. However, it can be assumed that regularly updated posters with reference to current developments placed in coffee kitchens would have a comparable effect. Hence, the approach developed here not only complements previous explanations of unplanned IKSI based on space by revealing the underlying social processes, but also coherently integrates materiality.

This paper has succeeded in deductively validating and inductively elaborating the theoretical assumptions about the impact of social situations on the occurrence of unplanned IKSI. To reveal the contours of the underlying mechanisms, six characteristics of situations and their role in this context have been outlined. As a consequence of the exploratory approach, however, the integration of these does not yield a complete and elaborated model of the interrelations between social situations and unplanned IKSI. On the basis of the insights about the general link, however, it seems promising to develop such a model through further empirical research, building on the approach provided here. In addition to this consolidation of the work begun here, it also seems worthwhile to take a closer look than has been done here at how differences between actors affect their scope for action in situations.

The second contribution concerns the virtualization of work. Scholars in this field have repeatedly emphasized the capability of advanced media to bridge social and spatial distances between potential IKSI partners. This was based on the assumption that IKSI particularly presuppose awareness about possible advisors and their easy accessibility. Both can be well provided by appropriate media. However, numerous empirical studies have reported that the implemented media failed to match the high expectations (Leonardi and Vaast, 2017). Also, in the study presented here, such media did not play a role, despite being available in all cases. This paper suggests that the impact of media is limited because it marginally helps to reduce symbolic costs. This first leads to the assumption that the virtualization of work (as driven by the corona crisis) will result in a strong downsizing and localization of advice networks. Second, it raises the question of which media features can yet contribute to the reduction of symbolic costs. Leonardi and Meyer (2015) already pointed out that enterprise social media applications can provide information about potential advisors that can serve as conversational fodder for initiating (apparently) unplanned IKSI. Beyond that, Knorr Cetina (2009) and Subramaniam et al. (2013) proposed to analyze media-mediated interaction settings as social situations as well. Following on from this, it seems very promising to investigate which specific characteristics of virtual social situations promote or hinder the formation of IKSI. Corresponding findings promise a deeper understanding of the impact of such media and enable the stimulation of IKSI by their design.

Figures

Figure 1

Research gap in explanation of unplanned IKSI

Figure 2

Symbolic costs explain why encounters result in IKSI

Figure 3

Summary of empirical findings

Table 1

Overview of participants

R&D department A	R&D department B	R&D department C
Material science	Electrical engineering	Electrical engineering
Germany	Germany	Switzerland
17 participants	10 participants	14 participants

Table A1

Interview guide¹

General Narration request		Please describe to me how the interaction came about
Follow-up questions about the occurrence of the consultation	Knowledge about interaction partners and its origin	Please describe to me how it became apparent to you that the person you were talking to could help you
		What did you know about the contact person?
		– When you recognized the knowledge gap?
		– Before you made the appointment?
		– Before you met by chance?
		How did you know this?
		Did you specifically search for this information about the interlocutor?
		Where/how did you search?
		Have there been several people shortlisted based on your knowledge? Why did you approach this person and not others?
		Have you been aware before the encounter that the person was a suitable contact?
		How did this become apparent in the situation?
	Occurrence of unplanned encounters	Please describe to me how it happened that you met
		Where did you encounter each other?
		Describe the place. What is generally done there?
		Why were you at the place?
		Why was your contact there?
		Was it reasonable for you to expect to meet the person there?
		Who was addressing whom?
		What was the person approached busy with at that moment?
		How did you come to talk about your project?
		If you specifically sought out the contact person, did you know they were in that place? How did you know that? Would you have tried even if you did not know?
	Characteristics of the situation with relevance for the occurrence of interactions	Please describe to me how you got into the conversation and onto the specific topic
		Was it a good opportunity to approach the person on the subject?
		Why?
		Have you been doubtful whether the situation is the right occasion to have such a conversation?
		What prompted you to do so?
		Were there other people present? What were they engaged in? If not, would that have made a difference?
		Were the people present in any way relevant to how the conversation came about?
		Would it be conceivable to you that someone might criticize your behavior in the situation?
	Expectations and outcome of the interaction	What were your expectations about the interaction, and what did it actually result in?
		In what way did you benefit from the interaction?
		Did you expect that?
		With what specific expectations did you go into the interaction?
		Can you say what triggered these expectations for you?
		If you were referred to another person, how were they able to help you?
	Questions about the context	Are there places, facilities or events where such interactions would seem to be particularly well suited or should explicitly take place?
		In your experience, is the “offer” being used? Why not?
		Is there someone you would have liked to talk to but did not? Why did it not happen?
		Were there any planned interactions that did not actually take place?
		How many years have you been working in the function?
		How many years have you been working on the issue?
		Where are you located?
		How long have you been with the company?

Note(s): ¹ The selection of the questions is made in addition to the course of the interview on the basis of the information given in the documentation sheet

Appendix 1

The appendix file is available online for this article.

Appendix 2

Appendix 3

Coding scheme

1. Contact persons are recognized as such

   • Hints from colleagues

   • Utterances

   • Presence

2. Media

3. Purposeful approaching

   • Knowledge about availability

   • Knowledge about competence

   • Knowledge about willingness

   • Expected effort

   • Symbolic costs

4. Unplanned encounters

   • Everyday movement patterns

   • Intentionality

5. Conversation arises unplanned/occasional

   • Enduring copresence

   • Action script/norm

   • Social control

   • Common ground

   • Willingness to talk

     • ‒From the situation

     • ‒Interactive signals

6. Task-related interaction arises spontaneously (from the conversation)

   • Preliminary small talk

   • Boundary object

     • ‒As a stimulus

     • ‒As means

   • Action script

     • ‒Promote

     • ‒Hinder

   • Interest

     • ‒Expressed in conversation

     • ‒Is evident from the situation

   • Symbolic costs

   • Interest of third parties

   • Discretion

     • ‒Social control

     • ‒Information control

   • Time