Introduction
As a first approximation, argumentation can be described as a communicative activity consisting of giving and asking for reasons to support claims (Bermejo Luque, 2011). Good arguments come in many kinds: they may be deductively, inductively, abductively, or defeasibly correct, but the key requirement is that the premises support the acceptability of the conclusion (van Eemeren, Garssen, Krabbe, Snoeck Henkemans, Verheij, & Wagemans, 2014), in particular because an argument ideally offers epistemic justification for the conclusion (Goldman, 2004). Given the connection with epistemic justification, argumentation is typically expected to deliver truths and lead to consensus.
However, there is much discussion on what exactly argumentation is, which goals those who engage in argumentative practices seek to achieve, and the extent to which these goals can indeed be thus achieved. As a matter of fact, there are many instances in which argumentation does not seem to achieve its presumed goals: instead of consensus, it leads to polarisation; instead of the circulation of reliable information, it leads to the propagation of falsehoods; instead of fostering sound decision-making, it leads to suboptimal choices.
One important strand in discussions on the nature and purposes of argumentation is the analysis of the different metaphors used to conceptualize argumentation. A seminal source is Lakoff and Johnsonâs influential book Metaphors We Live By (1980), where the ARGUMENT-AS-WAR metaphor is discussed at the very beginning. They famously argue that metaphorical conceptualizations of practices and phenomena shape the ways we think and act, and this is true in particular of argumentation conceived as a kind of battle, a metaphor that reinforces argumentative aggression. Cohen (1995) further developed and criticized the conceptualization of argumentation as war, seemingly suggesting that the metaphor should go if we are to improve our argumentative practices. Kidd (this volume) in turn argues that there is nothing wrong with the martial metaphor per se, but rather with how it tends to be interpreted in relation to argumentation.
The ARGUMENT-AS-WAR metaphor is not the only conception of argumentation extensively discussed in the literature (other conceptions are not always presented explicitly as metaphors). Some other prominent conceptualizations of argumentation include the âmarketplace of ideasâ as a way to maximize truth (Goldman, 1999; chap. 7), and Socratic argumentation as âmidwifery/therapyâ aiming at epistemic cure (Sedley, 2004; Moss, 2007). These three conceptions of argumentation â as war, as economic exchange, as therapy â differ quite substantially from each other, so much so that it might seem that they cannot all simultaneously be correct.
Alternatively, it is conceivable that argumentation in fact corresponds to multifaceted and heterogeneous phenomena, in which case perhaps each of these conceptualizations may be partially right (and also partially wrong). If so, then it makes sense to reassess them so as to establish where each of them gets things right (and wrong). This is precisely the goal of the present contribution. This analysis will allow us to address two of the central questions mentioned above: which goals those who engage in argumentative practices seek to achieve, and the extent to which these goals can indeed be achieved by means of such practices.
In this chapter, I discuss these three familiar conceptualizations of argumentation in turn.1 I start with the Millian conception of argumentation as the free exchange of ideas in order to obtain (or come closer to) the truth, and argue that, while it successfully captures the idea of epistemic gains that argumentation may bring about, it fails as a realistic account of argumentative practices. I then discuss the argument-as-war conception; relying on earlier critiques of this model, I argue that it has problematic moral as well as epistemic implications, in particular in that it implies that argumentation must be a zero-sum game with winners and losers (rather than a mutually beneficial activity). However, I argue that the adversarial component of the model does capture a significant feature of argumentation, provided that it is not restricted to mere aggression. Thirdly, I turn to argumentation as therapy in a Socratic vein; I argue that there is much to be commended in this conceptualization in that it highlights a different kind of (epistemic, personal) improvement that argumentation may bring about. But it attributes an asymmetrical relation to the participants of an argumentative situation (patient-therapist), which in turn does not do justice to many instances of argumentation among equals/peers.
In the final part of the chapter I introduce a different conception of argumentation as epistemic exchange, one that borrows elements from each of the previously discussed conceptions while seeking to avoid their pitfalls. On this conception, argumentation consists in exchanges of epistemic resources that can be mutually beneficial, thus rejecting the zero-sum conception of argumentation as war, as well as the inherent asymmetry of argumentation as therapy. The conceptualization of argumentation as epistemic exchange can be viewed as belonging to the Millian tradition (broadly speaking) in that it emphasizes the potential epistemic gains brought about by argumentation in social settings, but it avoids the overly idealized account of argumentation as straightforwardly truth-conducive. I conclude with some observations on the conditions under which argumentation is likely to lead to fruitful epistemic exchange rather than to polarisation.
1. Argumentation as straightforwardly2 truth conducive
That there must be a strong connection between practices of argumentation and the pursuit of truth/avoidance of error is a pervasive view. The thought is that, by examining critically relations of support between premises and conclusions, one will arrive at better supported, suitably justified beliefs, and this in turn increases the likelihood of these beliefs being true, as the unjustified ones have been weeded out. While this is a millennia-old idea, it has recently received an exceptionally clear articulation in Goldmanâs account of the social epistemology of argumentation as a social quest for true belief and error avoidance.
In other words, when the participants in argumentative situations comply with the norms of good argumentation (which Goldman understands as including the requirements that participants speak truthfully, and that they are justified in believing that the premises support the conclusion), the process should naturally lead to an increase in the accuracy of their beliefs: they will end up with more true beliefs and fewer false ones, in particular thanks to the âmutually correctiveâ component. But why is it that argumentation has the power to be truth conducive, especially in interpersonal settings? What are the exact processes/mechanisms involved?
An influential account of these processes can be found in John Stuart Millâs On Liberty (1999). His position is thus aptly summarized:
Dissenters thus force us to stay epistemically alert instead of becoming too comfortable with existing, entrenched beliefs. But for this process to be successful, dissenters of all stripes must be permitted to voice their opinions and criticism freely, and indeed Millâs defence of free speech not only on moral but particularly on epistemic grounds is one of his most celebrated positions. He emphasizes the role played by the free expression of ideas in facilitating the growth of knowledge in a community: the more dissenting views and arguments in favour of each of them are exchanged in a community, the more likely it is that the âgoodâ ones will prevail (Halliday & McCabe, 2019).
Millâs position is often interpreted as a defence of a completely unrestricted âmarketplace of ideasâ, where a competitive market for speech will maximize truth just as competitive markets for other goods maximize prosperity; if there are no external interferences with this market, knowledge will be maximized (Goldman, 1999; chap. 7). In truth, however, Mill was much more cautious than many interpreters take him to be, and defended certain forms of state regulation in the epistemic lives of citizens; in fact, it has been argued that the âmarketplace of ideasâ metaphor does not accurately reflect Millâs position at all (Gordon, 1997).
We must thus distinguish between two ideas here: the attribution of truth-conduciveness to argumentation, and the more contentious claim that it is the competition between different views in an open epistemic âmarketâ that will maximize truth and accuracy, by analogy to the free exchange of goods in a market economy. (Notice that the second one is more straightforwardly a metaphorical argument, whereas the first is rather a non-metaphorical claim about the truth-conduciveness of argumentation.) There are (at least) two obvious objections to the claim that the âmarketplace of ideasâ will indeed maximize truth: (i) the purported structural similarity between the economic realm and the epistemic realm does not hold; (ii) the purported maximization of wealth in a free market economy does not in fact occur, and so there is no reason to think that maximization of truth will occur in a free marketplace of ideas.3 The proponent of the view that argumentation is truth conducive can accept both objections, and still maintain that, when properly conducted â which may include forms of external interference and regulation â argumentation does indeed promote the pursuit of truth and avoidance of error.
Indeed, Goldman has argued on conceptual grounds for the truth-conduciveness of argumentation, at least when conducted by the norms of good argumentation. Furthermore, there is formal (Betz, 2013) as well as empirical (Mercier, 2018) evidence for this claim. However, there is equally compelling evidence (strengthened by the mere observation of the current dismal state of public discourse) against argumentationâs potential to maximize truth. A case in point would be the massive increase in vaccination hesitation in the last decade, despite vast amounts of reliable scientific evidence attesting to its safety and efficacy, which however is not having much uptake among so-called âanti-vaxxersâ (Pierik, 2018).
One explanation for the existence of conflicting evidence on the truth-conduciveness of argumentation is this: argumentation can indeed be truth conducive, but only under special (perhaps rare) circumstances. A closer look at the positive available evidence lends support to this hypothesis: Betzâs (2013) simulations, for example, depend on a number of very strong assumptions, such as that all agents in an argumentative community converge in their assessment of the correctness and cogency of specific arguments. Goldman also explicitly states that his account aims at capturing the norms for good argumentation, which may however deviate substantially from actual practices of argumentation. We might in fact describe Goldmanâs account as an ideal theory of argumentation (in the sense of the Rawlsian ideal versus non-ideal theory distinction) possibly with limited applicability to real-life scenarios. The same may be said of Millâs account, even on the more nuanced interpretation that does not attribute to him a crude âmarketplace of ideasâ conception.
The conclusion for now is that there is some truth to the idea that argumentation can be truth conducive, but in real life matters are much more complex than a number of the available âideal theoriesâ seem to suggest.4 To investigate these complexities, a more realistic social epistemological approach to argumentation is required, one which retains the epistemic focus of the accounts discussed so far, but has the resources to deal with ...
