Logic,evidence and consensus: Towards a more constructive debate on gaming disorder

In their most recent ANZJP debate paper on the topic of gaming disorder (GD) as a mental disorder which, for context, was preceded by two other debate papers (Dullur and Starcevic, 2018; King et al., 2018), Schimmenti and Starcevic (2019) argue that our paper, which outlines arguments in support of GD as a diagnostic category in the Diagnostic and Statistical Manual of Mental Disorders (DSM) and International Classification of Diseases (ICD) classification systems, was ‘biased’ and based on a series of ‘logical fallacies’. While we share the authors’ concerns regarding the need for conceptual clarity and strong evidence in this growing field, we believe that their paper’s attempt to frame this GD debate as a contest between logical and fallacious thinking has some weaknesses. In our view, many of their examples of ‘faulty logic’ are derived from misinterpreting our arguments or taking a less than charitable interpretation and represent a failure to account for the caveats of a brief debate format. Furthermore, their views concerning logical fallacy could also be applied to Dullur and Starcevic’s (2018) arguments in opposition to GD much to the same effect.

Limitations of the logical fallacy approach

The principal argument advanced by Schimmenti and Starcevic to rebut our views in support of GD is that our arguments are founded on errors in applying logic. With reference to the practice of psychotherapy, they explain that such errors may arise ‘without the knowledge of the arguer, [or] they may also be used to manipulate or deceive the audience or opponent’. Through this argumentative lens, it is our impression that the authors are attempting to critique our paper by taking a less than charitable interpretation of our statements and, presumably, the intentions of our arguments. While we do not dispute the value of examining logical reasoning in scientific communication, we believe their ‘logical fallacy’ approach presents some shortcomings when applied to the task of critiquing a short debate paper. As per the journal’s conventions and guidelines for ‘Debate’ papers, these are articles which are generally intended to provide a provocative account that reinforces one side of a debate (i.e. not to provide a complete overview of the topic). Thus, the reader should endeavour to receive these arguments in the most rational way possible (i.e. considering the best or strongest possible interpretation). This may extend to making some reasonable concessions out of consideration to the obvious constraints of this format of debate, such as the need for arguments to be brief and succinct.

Expectations and representation

Rather than considering our paper in this spirit of debate, Schimmenti and Starcevic have evaluated our paper as though it was a research report or review containing the totality of our thinking, and awareness of other viewpoints, on the topic of GD. From this position, they criticise our paper for demonstrating the cherry-picking fallacy or ‘presenting data that confirm a particular position … [and that] King et al. ignore other perspectives’. This conclusion demonstrates expectations that are beyond the purpose and scope of a short debate paper.

The authors also demonstrate the straw man fallacy by distorting a simple proposition and then making this the target of criticism. One of our most straightforward and concise arguments (i.e. ‘IGD does not pathologise or stigmatise normal gaming … [because] the ICD-11 and DSM-5 do not state that gaming is inherently harmful’, p. 615) leads the authors to raise new propositions that were not directly related to our original point (i.e. ‘Can this diagnosis then stigmatise only abnormal gaming? The presumed ability of the GD diagnosis to stigmatise or not stigmatise reflects a reification fallacy because GD is treated as a concrete entity not as a construct or abstraction’). The stigmatisation of ‘abnormal gaming’ was not the focus of our argument. We would not have suggested that GD stigmatises abnormal gaming, just as we would not suggest that the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5) stigmatises depression. Our statement does not specify whether GD is a construct or a real entity; however, to clarify, our view is that GD is a valid conceptualisation of extreme forms of problematic gaming associated with harms, and that this conceptualisation may be further refined over time, informed by accumulated data.

Another example of misrepresentation of our views in the authors’ search for logical fallacies arises from their claim that we have demonstrated the begging the question fallacy by assuming that ‘support for GD’ is already ‘proven’. Here, we are simply stating that the GD classification has received strong support, referring to the research and clinical evidence that has accumulated and recently informed the International Classification of Diseases (11th Revision; ICD-11) GD category. As an example, we cite Dullur and Hay’s (2017) study that reported stronger support than opposition for GD and other non-gaming online addictions among Australian and New Zealand psychiatrists. At the same time, we temper our statements about support for GD by noting that there are still some unresolved issues and inconsistencies; specifically, the potential lack of fit of some components of the addiction model to gaming (e.g. ‘tolerance’). Our statements are consistent with our other published review work (King and Delfabbro, 2018), where our conclusions about support for GD have proceeded from a careful examination of the literature.

Our opponents’ views demonstrate the same logical fallacies

Following Schimmenti and Starcevic’s (2019) example, which we do not endorse universally, we could draw attention to the first debate paper by Dullur and Starcevic (2018) and argue that it contains arguments that contain many of the same logical fallacies. For example, the authors present an argument based on the appeal to consequences fallacy in their statements that the GD diagnosis would lead to the use of ‘potentially harmful treatments, such as certain medications and coercive methods’ (p. 110) and ‘a premature closure in our thinking about problematic online gaming’ (p. 111). The latter example could also be considered an example of the fallacy of false choice. Similarly, one could argue that the authors’ claims that there are ‘significant flaws with almost every diagnostic criterion’ and that GD is ‘more akin to a disorder of impulse control’, where the authors cite only two narrative-style reviews, are examples of the cherry picking fallacy.

Along similar lines, Schimmenti and Starcevic’s argument that ‘we fail to cite any research’ is another unhelpful criticism to make for a short paper that has a maximum limit of only five references. Following their example, which would not be constructive, we could argue that they have failed to provide an appropriate citation to the research they mention.

Towards more constructive debates on GD

Here, we have only briefly communicated some of our points of disagreement with Schimmenti and Starcevic’s (2019) paper. While we are in agreement that logical arguments are vitally important in all scientific endeavours, we believe that framing our arguments as a series of logical fallacies is not particularly productive. Unfortunately, this approach may provide a framework or point of reference for future communications within and outside of the GD field, as the authors recommend. This would likely generate even more divisive rhetoric in the field, which may be capitalised on by opportunistic entities with a vested interest in discrediting the GD diagnosis or field as a whole. As a final note and important point of clarification, we are not suggesting that researchers should avoid identifying errors and/or limitations that arise in others’ (and/or their own) arguments. Rather, we are arguing that debates are more beneficial to all parties when participants endeavour to be constructive, interpret opponents’ statements in the most rational ways, apply the same standards to their own arguments and consider the medium through which arguments are delivered. Following such guidelines would be more conducive to stimulating new ideas and valuable discussion of the available evidence in the GD field, if not for consensus then for progress.