Two puzzles for shared-reason accounts of persuasion

ABSTRACT

Dowding has developed an account of deliberative persuasion leveraging ‘reliability conditions’ distinguishing persuasion from manipulation. The central idea is that proper reasons offered for deliberation must be shared reasons: reasons the speaker holds as grounding the proposition, and believes should motivate the receiver’s assent too. I present two puzzles for this account. The first queries the level at which reasons should be shared to count as persuasive. The second concerns the breadth of reasons for a given proposition – should a speaker have many, only some he sees as shared, may he choose just any of these to present?

KEYWORDS:

• Deliberation
• persuasion
• manipulation
• justification
• reasons

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. See, for instance: Daniels (1996); Estlund (1998); Friedman (2006); Vallier (2011b); Boettcher (2012); Hertzberg (2015).

2. A disregard or lack of care around the application of standards of truth and accuracy that one would otherwise hold as normative (Cassam 2018).

3. Dowding himself suggests this complicating possibility but doesn’t pursue the role of `reasons about reasons’ in his discussion (2016, p. 14).

4. I will for the most part leave the second reliability condition aside, since it does less work than the first, and alone would not rule out paternalistic manipulation of all kinds.

5. Dowding alludes to but does not pursue what he calls meta-reasons (Dowding2016, 14), a promising solution to this puzzle that I elaborate below.

6. See, among others, Ware (1981); Riker (1986); Goodin (1980); Noggle (1996); Baron (2003); Barnhill (2014)

7. See especially Baron (2003); Barnhill (2014); Gorin (2014); Whitfield (2020).

8. See (Korsgaard 1996, chs 2 & 10) and (Scanlon 1998, chs 1, 5, & appendix) for especially prominent examples.

9. Note that this does not offer the possibility of salvage for Devil’s advocacy, structured debate, and other forms of communication aimed at securing assent (whatever the reasons proffered) to a proposition one does not oneself assent to.

10. For present purposes put aside beliefs held without clear connection to an articulable reason. Of course it’s true that almost all people endorse some propositions without clear reason or for reasons we would fail to articulate if pressed. Such beliefs have less relevance to this level of thinking about justification, where it is precisely grounds rather than the mere fact of assent which matter.

11. .One could of course resolve this puzzle outright by requiring i to share the full set E = {e₁, ... e_n} (n ≥ 2), however large nmight be. While that would resolve the worry about strategic choice among shared reasons, it would present avery significant burden on communication in cases where nis large, itself not asmall set of cases. This discussion presumes this is not apath shared reason accounts can take out of the puzzle. $E=\{e_1,\dots e_n\}$

Additional information

Notes on contributors

Gregory Whitfield

Gregory Whitfield is a postdoctoral fellow with the Research Group on Constitutional Studies at McGill University, writing on liberal and democratic theory and philosophical/ethical issues in social science research. His recent works on political manipulation and research ethics have appeared in the European Journal of Political Theory and Political Research Quarterly.