Perspectives and ambitions of interdisciplinary connectivity researchers

Different scientists have different aspirations; the challenge arises when they assume a shared understanding of their research perspective, which often results in confusion and unintentional miscommunication (Bracken and Oughton, 2006). This is especially so in an interdisciplinary environment such as the connectivity community where cross-fertilisation carries a large potential for scientists to improve their research practises using knowledge from beyond their own discipline. However, very different motivations exist to do this, and it is often not clear what a scientist intends to achieve by applying the knowledge of connectivity methods such as indices, modelling approaches or field designs from neighbouring disciplines.

Öberg (2011) identified four different perspectives that are common in environments where people deal with the interactions of human and natural systems while working across disciplinary boundaries: the pragmatic, conceptual, epistemological and ontological perspectives. While we acknowledge that other terminologies and classifications are possible, in the following we review Öberg's (2011) four perspectives in regard to their interdisciplinary applicability to connectivity research.

1. Pragmatic perspective: to solve a practical academic problem. Hydrologists saw the similarity between a rainfall–runoff equation for catchments and the waiting of customers in a serving queue (see for example Harel and Mouche, 2014). Subsequently a hydrologist used a model from queuing theory, which was developed in operational research for telecommunications using a simple probabilistic approach or map equations describing queue length and waiting time. Harel and Mouche (2013) applied a queuing model to study connectivity features of rainfall–runoff processes along hillslopes using corresponding terms of the waiting time of queues for the separation of water flow. Although these researchers were perhaps initially motivated by curiosity in exploring the parallels of the two applications, this example illustrates how disciplines may borrow methods, theories and models from other disciplines to enhance their toolbox in proceeding with a certain research objective (Öberg, 2011). The pragmatic approach is probably the most common one in current connectivity research and strives towards the cross-fertilisation of methods from different environmental disciplines, as established by the EU COST Action Connecteur (Connecteur, 2018). However, one has to be aware that this approach holds the danger of severe misjudgement when using methodologies without understanding the underlying theories, assumptions, boundary conditions and resulting consequences. An example might be engineering the application of erosion models reproducing sheet erosion (such as USLE) to assess reservoir infilling by sediments in regions where most of sediment originates from gully erosion.

2. Conceptual perspective: to contribute to a new or emerging field. Connectivity research can be viewed as forming a new, emerging science field, which goes beyond the traditional disciplinary boundaries of single environmental disciplines such as hydrology, ecology or geomorphology and even stretches to current efforts in life science research and beyond. The establishment of new theories with novel concepts for any connected systems is at the heart of this perspective, which include, for example, the study of brain network organisation and function connectivity in neuroscience (Stam et al., 2016); social networks for opinion formation in social science (Grabowski, 2009); interacting, adaptive or self-organisational sensor or power networks in electrical engineering (Manjunath and Mohan, 2007); or connectivity index tools for big data analysis. The setting up of overarching theories requires a deep understanding of the core of existing connectivity methods and concepts in a range of science disciplines (e.g. Callagero and Ursino, 2018). The conceptual perspective, therefore, has the great potential to identify much more innovative applications of knowledge than just borrowing single methods as described above, but this will only be possible if deep communication and the exchange of information between disciplines are ensured.

   Scientists adopting the conceptual perspective are likely to belong to a specific speech community associated with their discipline. In this context, Bracken and Oughton (2006) called for a critical, reflexive awareness of how scientists use language in their interdisciplinary work as a crucial step towards establishing a shared language. For example, they showed that differing usage and understanding of common terms such as “dynamic” was rooted in differences between disciplinary use of term and their everyday meaning. The background of the research group, research approach, geographic setting of the study, language and national scholarly background (Bracken et al., 2013; Smetanová and Dąbrowska, 2009) can further influence understanding of common terms and development of connectivity concepts in interdisciplinary and international groups.

3. Epistemological perspective: to analyse the way in which disciplinary structures cause problems. With an epistemological approach, the focus of study is knowledge generation itself, e.g. through analysing the implications of studying, understanding and describing a problem from particular disciplinary viewpoints (Öberg, 2011). The comparison of particular disciplinary viewpoints (from biology, neuroscience, geomorphology, social network science and ecology) on the definition of the fundamental unit of connectivity, structural and functional connectivity; emergent behaviour of complex systems; and measuring connectivity using epistemological approach was provided recently by Turnbull et al. (2018). Connectivity research further opens an interesting arena for interdisciplinary scholars to study the practices of interdisciplinary environmental projects and analyse how and to what extent the involved disciplines connect their knowledge with each other and with society. Essentially, this article attempts to use an epistemological perspective on connectivity research to understand how multiple mental models of connectivity scientists differ and the measures that might be necessary for a shared understanding to be gained.

4. Ontological perspective: to analyse the consequences of societal perceptions of an environmental issue. The way environmental issues are described guides our understanding and perception of the environment (Öberg, 2011), thereby reinforcing how environmental issues “are” (in an ontological sense) through particular management responses. “Connectivity” is a term that is currently widely used in the hydrological and ecological sciences, but scientists actually have very limited knowledge on the perceived relevance of connectivity (or lack of thereof) for water and land managers and policymakers outside academia. An example of connectivity perception outside academia was given by the Unites States Supreme Court (547 US 715 – 2006) case Rapanos vs. United States. The legal notion of a “significant nexus” was introduced by US Supreme Court Justice Anthony Kennedy and was further criticised while acknowledging that tangible evidence of water, sediment, chemical and biological connectivity needs to be obtained before specific wetlands, lakes, riparian areas and other water bodies are protected by the federal government.

   We claim that if the concepts of connectivity methods – both theoretical aspects (e.g. Bracken et al., 2015; Cossart et al., 2018; Keesstra et al., 2018) and practical aspects regarding monitoring the design, model and index implementation adapted to planning applications (e.g. Clauzel et al., 2013; Foltête et al., 2015; Tannier et al., 2016; Ahlmer et al., 2018) – have not yet fully entered the mindset of water and land managers, they cannot understand how to monitor, model and subsequently manage environmental problems. But how relevant is connectivity to water and land managers? This question can only be answered by studying the perceived relevance of connectivity issues by stakeholders across the environmental sector. Perceived relevance of connectivity seems to be rooted in the experience and everyday challenges of water and land managers and contributes to heterogeneities in the potential to manage connectivity and apply methods adapted for management purposes (Smetanová et al., 2018, questioned 85 stakeholders in 19 EU countries). The studies of perceived relevance of connectivity may radically alter (in an ontological sense) the nature of connectivity as a research problem (Freeman et al., 2007; Nadeau and Rains, 2007; Leibowitz et al., 2008; Golden et al., 2017; Ali et al., 2018).

   The four very different viewpoints applied to connectivity research signify the diversity of thinking in the connectivity community and call for a structured way for scientists from different viewpoints to communicate with each other. It suggests, for example, that scientists with a pragmatic and an ontological perspective might develop serious communication and understanding problems if they start working together on connectivity issues. The next section will present the results of our mental model elicitation as a way forward.

The first step to enhance mutual understanding in a group gathered around a specific research concept such as connectivity is to be aware of the different individual mental models that exist in that group. Mental models are closely linked to different research philosophies, concepts and methods, as they represent how people understand the world around them; they are the internal, cognitive representations of the external system. Or in other words, mental models are specific mental representations of information about reality (Pahl-Wostl and Hare, 2004). Our mental models are shaped by our previous experience and, in turn, shape our behaviour and approaches for reasoning, solving problems and carrying out tasks (Lynam and Brown, 2012). Mental models allow human beings to survive and act in a complex world (Pahl-Wostl and Hare, 2004), though for the most part they are incomplete representations of reality and are often inconsistent among people – which is arguably one of the key reasons for understanding and communication problems in interdisciplinary research groups.

As we cannot directly access other people's thinking, a process of elicitation is used to encourage a person to externalise her or his mental model (van der Bossche et al., 2011; Jones et al., 2014). Mental models can be elicited to explore the similarities and differences in understanding of a specific concept, e.g. regarding connectivity, in order to improve understanding and communication among scientists from different disciplines. The majority of elicitation techniques are based on the assumption that an individual's mental model can be represented as a network of concepts and relations (Jones et al., 2011). Methods for eliciting mental models comprise oral methods, such as textual analysis and inferences from interview data or questionnaires (see e.g. Carley, 1997) and visual methods using diagrammatic interview techniques that let a person externalise their mental model through the graphical representation of concepts and interactions, e.g. as a mind map (e.g. Kearney and Kaplan, 1997; see also Mohammed et al., 2000, for an excellent review on elicitation methods).

Shared mental models refer to the overlapping mental representations by members of a group or, in other words, the meta-knowledge that goes beyond the various research and personal perspectives of individual team members (van der Bossche et al., 2011; Godeman, 2011). Our study demonstrates similarities and differences in mental models of connectivity researchers, which was apparent even in a small group. Carley (1997) suggested three major areas of contention in shared knowledge production, (i) the uniformity of sharing – whether knowledge must be uniformly shared by group members; (ii) degree of sharing – how widely the knowledge must be shared; and (iii) awareness of sharing – whether the individual group members must be aware that the group's mental model is shared. According to group discussions during the workshop, we consider the last area the most important for a truly interdisciplinary research field such as connectivity science.

How then can we achieve a shared understanding or a shared mental model in interdisciplinary connectivity research? According to van der Bossche et al. (2011), it appears insufficient to attempt knowledge convergence solely based on conversation or by simply paying attention and acknowledging a contribution as we usually do in keynote lectures and workshop presentations during scientific meetings. Instead, van der Bossche et al. (2011) call for active interactions; three of such efforts documented in this study will be discussed here in turn.

First, co-construction of specific or general connectivity terminology is required, even if parts of the group might consider it a waste of time. On this basis, co-construction of knowledge can be understood as the group members' attitude towards knowledge which allows them to query it. Challenging each other's views, definitions and divergences with respect to a specific aspect of their joint work might become essential – especially given that no coherent definition for connectivity itself has been agreed upon (see list of references with possible definitions in the introduction section). For example, in our group discussion we began to co-construct knowledge regarding the concept of a basic unit of connectivity, a concept that some scientists had a very clear opinion on (e.g. discharge of water in m³ day⁻¹), whereas some were not aware that there was a basic unit, and others rejected the idea of a basic unit of connectivity altogether; in their research, the focus lay on the linkages of multiple human–environment aspects where a basic unit concept would only constrict their perspectives (see Turnbull et al., 2018, for a review of basic units of connectivity).

Second, constructive conflicts may help to improve group communication, e.g. by unravelling different points of view (De Dreu and Weingart, 2003; Krueger et al., 2016) that affect how interdisciplinary group approaches open questions in connectivity science. Although the colour contours of Fig. 3, representing Öberg's (2011) four research perspectives (pragmatic – yellow; conceptual – orange; epistemological – red; ontological – violet), were only reproduced by the individual profiles of the scientists' mental models to some extent (lines in Fig. 3), it was possible to identify certain groupings of profiles around one of the four perspectives. For a constructive conflict, scientists need to be aware of the mere existence of other research perspectives; based on our group discussion, we claim that this awareness normally does not exist among connectivity researchers. Non-existing awareness about other research perspectives might be an inherent trait of the natural sciences, since their education does not emphasise different research positions as done by the interpretative social sciences. The process of constructive conflict will expose, among other things, what the intentions of scientists are to use techniques from neighbouring disciplines. When one research tradition opposes the methods of another, a window of opportunity for reflection and improvement of one's own research tradition opens. Though when cross-fertilisation in connectivity research is attempted without a clarification of existing (parallel, convergent or divergent) research perspectives, any further discourse might quickly become both patronising and frustrating.

Figure 4Graphical representation of potential mental model overlaps between pairs of stylised types of connectivity scientists. Stylised types of connectivity scientists (Fig. 3) after Öberg (2011) are represented here by pragmatic (yellow, A) and conceptual (orange, B). Remaining mental models (from group of 13 scientists) were classified as mixed types according to perspective: (C) between epistemological (red) and ontological (violet), (D) pragmatic and conceptual, and (E) conceptual and epistemological. Grey bars show overlap, and shaded grey bars show partial overlap. There can be considerable overlap between the pragmatic (A) and the conceptual (B) perspectives. There is no potential overlap between the pragmatic and the ontological perspectives. Several overlaps occurred between mixed types (D and E), but those differed from A–B overlaps. We elicited some overlap between the epistemological and the ontological perspectives, but the number of individuals consulted here was too small to draw any conclusions.

Download

Third, the process of building a shared mental model (methods in Sect. 4.1, results in Sect. 4.2) can be supported by a detailed interpretation of overlaps of individual profiles, and lacks thereof, on the basis of Fig. 3. In our case study, 13 profiles (black lines in Fig. 3) could be grouped into five profile types (A–E) in Fig. 4. The five profile types A–E are further represented by colour shading of the vertical bars (corresponding to colour bands in Fig. 3). Colour shading of the bars (A–E) contains information as to whether a profile type exhibits attributes which were associated with only one of the four stylised research perspectives (pragmatic – yellow; conceptual – orange; epistemological – red; ontological – violet; A–B in Fig. 4) or with mixed perspectives (C–E in Fig. 4). Grey and shaded grey vertical bars represent the overlap between attributes of profile types. For the yellow (pragmatic, A) and red–violet profiles (epistemological–ontological, D) paired in Fig. 4 literally no overlaps exist in their mental models of connectivity research. For the other two paired profiles (A and B and D and E), several overlaps exist, though for very different attributes. We suggest that the graphical profile chart as depicted in Fig. 4 can be used as a tool to identify gaps and overlaps of mental models for all participants of an interdisciplinary research group as a way of speeding up the building of the group's meta-knowledge (van der Bossche et al., 2011) and the awareness of the group members' eventual sharing of a mental model (Carley, 1997).

With this article, we did not aim to maximise the group performance as business and military managers or team scientists using similar methods. We also did not aim to further develop the theory of connectivity, mental models or models of perception in environmental science (e.g. Öberg, 2011). Rather, we intended to encourage natural scientists active in connectivity research to become more familiar with literature on interdisciplinarity and to become aware of the existence of collaboration techniques, such as shared mental model building. Previous studies demonstrated that different or even diverging perspectives do not negatively influence the knowledge creation processes when interactions between the actors are repeated, positively perceived and sufficiently adjusted to encourage relationship building (e.g. Dewulf et al., 2007). The approaches and results of our study have been presented to connectivity scientists in EU COST Action ES1306 and closely discussed with the leaders of the action's working groups in order to facilitate effective communication within the working groups and the network. The principles of mental model analysis were in a different form applied within the collaborative work of EU COST Action ES1306 and led to interdisciplinary studies within (e.g. Connecteur WG3 Think-Tank Team, 2018; Heckmann et al., 2018) and without (e.g. Turnbull et al., 2018) research of connectivity or with actors outside academia (Smetanová et al., 2018).

Data collected in the questionnaires contain private data. They cannot be published under GDPR policy.

The authors declare that they have no conflict of interest.