TSElosophers meeting 21.2.2017
Otto Rosendahl, Kari Lukka, Jonathan Van Mumford, Milla Wirén

On the Virtues and Vices of Combining Theories: The Case of Institutional and Actor-Network Theories in Accounting Research (working paper)
Sven Modell, Eija Vinnari and Kari Lukka

Quick summary:

The article discusses the paradigmatic tensions that (may) underlie combining diverse method theories through looking at accounting research that draws from both Institutional Theory and Actor-Network-Theory. Our discussions centered on the following themes:

  • The importance of consistency, not only in theorizing but in all problem solving efforts utilizing insights from existing knowledge.
  • The very special kind of role of theory in ANT and the implications that follow from following logically through with its suggestions in scholarly pursuits.
  • ANT is useful in pointing out the issues of the very typical and widely accepted ‘normal science’ tendency (which is prominent also in IT, among other theoretical constructions) to view phenomena through a set of assumptions. However taking Latour’s quips too far may result in a paralyzing level of reflectivity in actual theory-building/refining research efforts.

The importance of episteme, the theories in research seems to be an ongoing theme in our discussions, and warrants further looking into. Can there be a relevant practical contribution about a phenomenon if the theoretical foundations are shaky? This is an especially important question in disciplines with a strong practitioner focus.

Longer outline of discussion:

Modell, Vinnari and Lukka (2017) article “On the Virtues and Vices of Combining Theories: The Case of Institutional and Actor-Network Theories in Accounting Research” problematizes combining method theories with very different ontological and epistemological backgrounds. Method theories can be used in studies as lenses – the study does not aim to contribute to these theories, but only employ them as their resources (Lukka & Vinnari, 2014). However, these theories are used to focus one’s study to something.

IT and ANT are arguably tectonic plates in the geography of science but arguably they are located far from each other regarding their ontological and epistemological underpinnings. For example, ANT rejects structures and embedded agency that IT emphasizes. Instead, ANT concentrates on actors and distributed agency. Whereas IT emphasizes “previous theory and its extensions”, “theory has no role in ANT” (Modell, Vinnari & Lukka 2017). Nevertheless, Modell et al. (2017) found many accounting articles that either entirely ignored or unreflexively assimilated these differences. In general, the motivation to mix elements of ANT and IT is an attempt to get the best of both worlds: flexibly following both structures and actors in research. There have been some claims of contribution in mixing IT and ANT, but they may have been built on epistemologically and ontologically shaky ground.

Elaborating the geography analogy, most scientists live on the IT tectonic plate and, if they have heard about ANT, they see it distant and foreign. Although Latour (2005) himself proudly proclaims that ANT is inhabitable (“You cannot apply it to anything”), there are also many business researchers who successfully dwell on it (e.g. Quattrone & Hopper 2001; Mennicken 2008; Christner 2016). Because of its ‘smaller size’ as compared to IT, ANT has a larger risk in losing the characteristics that make it special.

Therefore, retaining the provocativeness of ANT can be good strategy to spread its ideas. ANT ideas are already spreading in a weaker form in many social sciences, e.g. sociomateriality (Orlikowski & Scott 2008), market devices (e.g. Muniesa, Millo & Callon 2007) and market shaping (Harrison & Kjellberg 2016). Why not celebrate ANT for something that overdoes its arguments so that it stays an interesting read, especially to nonconformists? Needless to say, Latour is a master provocateur. In our group we had different opinions on how strong scientific influence we would hope for ANT, but all of us would certainly hesitate to commit to using pure ANT methodology in our own research.



The article was very strict in its view that theory has no role in ANT. I believe that although this is a useful generalization, it is not so straightforward. If ANT-driven researcher were to use theory, he or she would only have to admit that all scientific articles that have an effect to the research are actors. These articles would no doubt dilute the role of other actors. Moreover, in Latour’s (2005) strict view, this dilution would often equate to laziness or admitting that the research wasn’t very interesting to begin with. Laziness in the case of adding theory to research is used to reduce the number of primary actors that is being studied; not very interesting in case the study needs pretty theoretical frames to help it capture attention. One could also try to argue that ANT itself is not completely devoid of theory. ANT would defend itself by depicting itself as a negativistic methodology: ANT merely states what the researcher should avoid and gives guidance to the researcher how to trace associations between other actors (Latour 2005).

Although ANT only claims to make research more difficult, not easier in any way, it suits me. I have been very carefully trying to understand the essence of science from the start of my doctoral studies. ANT states that there is no essence, but there can be some things which are more essential than others. I feel ANT is helpful in understanding relativity without making you a relativist. Referring to our group’s previous topic, I believe ANT is a solid anti-performative foundation which eventually helps me to reach critical performativity (Alvesson, Spicer & Kärreman 2009) in a way that I am most comfortable with.



I don’t like ANT. This article helped me in understanding why that is the case: the underlying naïve realism as its ontological basis and the subsequent lack of in-depth vision joined with the blatant a-theoretical orientation make me want to shout like the proverbial fire brigade captain that the fire was extinguished the wrong way. On the surface one may ask what does it matter as long as the fire was put out, but if the method of pouring water on the visible structures allow the sparks for an even worse fire smoulder beneath the surface, the worry is merited.

For me the fire is the human tendency to enter any situation with a set of existing contact lenses through which everything is perceived and made sense of. I agree full-heartedly that one of the most valuable skills of a scholar is to be able to acknowledge (at least a set of) ones own interpretive frames, and to try to overcome them to see also the gorilla on the basketball field (you know the famous cognition experiment, right?). To me the truly dangerous sparks left kindling the worse fire result from the attempt to flatten everything into pixels – from the loss of the 3D vision that enables us to tell the foreground from the background.

Essentially to me it’s not the brushstrokes that we should zoom on in the painting we see, but the stories behind the obvious figures captured on the canvas. Rendering everything (yet another irritant, as to think anyone able to capture “everything” reminds me of an ikarosian attempt at a godlike omniscience) into the atoms forming any level substance is to me a waste of effort that could be spent better in trying to understand the emerging image and the meanings attached to it. In my view ANT takes a 180° wrong turn in suggesting which direction to focus. It’s like counting the dots and analyzing their colours in a pointillist painting instead of taking a step back to see what’s really happening in the artwork – not only what’s painted, people, flowers, lake, but the story, sentiment, deeper meaning of positioning the people, flowers, lake in their exact positions.

Yes, I understand why ANT has been welcomed – it points out both individual and systemic level flaws in “traditional” normal science approaches – but following it to its logical conclusion takes us too deep into not seeing the forest from the trees, and as such ending up tearing down knowledge about the forest for the sake of being able to provide the “astonishing” revelation of “wow, there are trees!”. One could argue that both are needed, yes, but for me, individually, life consists of such a flux of events, thoughts, things, people, that from my scientific endeavors I’m looking more for the theories that help me make sense of what I see and experience, not an a-theoretical listing and itemizing of “everything” that flows me by.