Fake News vs. Echo Chambers

ABSTRACT

I argue that there is a prima facie tension between solutions to the problem of fake news and solutions to the problem presented by various cognitive biases that dispose us to dismiss evidence against our prior beliefs (what might seem to be the driving force behind echo chambers). We can guard against fake news by strengthening belief. But we can exit echo chambers by becoming more sensitive to counterevidence, which seems to require weakening our beliefs. I resolve the tension by arguing against an injunction to weaken belief in the required ways. In particular, there is no injunction to be open-minded in all circumstances toward various counterarguments, even those whose premises seem compelling and in which we can’t expose a flaw. On the contrary, there are circumstances in which you should, in the relevant sense, be closed-minded toward such counterarguments.

KEYWORDS:

• Echo chambers
• fake news
• belief
• open-mindedness
• attitude-certainty

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. Nguyen’s echo chambers come with epistemic discrediting of non-members. But if that epistemic discrediting is accurate, then Lackey’s arguments about the legitimacy of truth-conducive echo chambers (in her sense) should apply to echo chambers in Nguyen’s sense, as well.

2. Nathan Ballantyne (2019) proposes that you become less sure of yourself when confronted with significant amounts of evidence that you haven’t yet explored (he calls this the ‘problem of unpossessed evidence’). But Ballantyne notes that his injunction can be defeated if, for example, you have ‘compelling evidence’ in favor of your position. For example, Ballantyne singles out ‘human-caused climate change’ as a case in which ‘the available evidence seems to demand a particular doxastic response’: namely, belief (127). Likewise, Bishop and Trout (2005, 146–47) recommend that when you are reasoning about a highly significant problem in which short term reliability is very important, you explicitly consider why your belief might be false. Suppose you are a Republican senator considering whether to vote to impeach President Trump for inciting an insurrection. You might believe that the members of the riot are Antifa infiltrators, that voting for impeachment will damage your political career, that there was massive fraud on election day, etc. All of these beliefs seem relevant to the decision at hand, and because the decision is momentous and short term reliability is important, Bishop and Trout’s injunction seems to rightly recommend considering the opposite.But it’s also relevant to your decision whether the Capitol building exists, whether voting for impeachment will inspire a Martian invasion of the earth, whether you have a material body (and thus are capable of raising your hand to vote), and a number of other beliefs about yourself, the mechanics of U.S. government, and the state of the universe. These, I presume, you need not subject to Bishop and Trout’s strategy. Why not? The natural answer is that your evidence for them ‘demand[s] a particular doxastic response’: namely, belief.

3. There may already be a tendency to approach alternative sources closed-mindedly. Current work in social epistemology and social psychology – in particular, in discussions of how people respond to the views of others over social media – emphasizes the defenses we set up around our beliefs to protect them from the influences of others. Belief bias and myside bias – both plausibly categorized as kinds of confirmation bias (see (McKenzie 2004, 207–209) and (Stanovich, West and Topak 2013, 259) – unwarrantedly dispose us to lend greater credibility and attend more closely to arguments and evidence that support our pre-existing beliefs. We have a tendency to ignore or discredit arguments and evidence against our pre-existing beliefs. But regardless of whether we have this tendency generally, we will certainly have this tendency if we approach other sources closed-mindedly.

4. See (Nyhan and Reifler 2010) for an influential early argument for the backfire effect. See (Sippitt 2019) for a critique. Again, whether there is a general tendency to be subject to a backfire effect, the issue here is whether approaching sources closed-mindedly disposes you to the backfire effect. I assume that if there is a general tendency to experience a backfire effect, there is also a tendency for the closed-minded to experience one.

5. See (Fantl 2018, Ch. 1) for a defense of this conception of open-mindedness.

6. See also (Lammenranta 2011) and (Wieland 2013).

7. See (Fantl 2018, Ch. 6) for further details on this argument.

8. See (Jastrow 1902, 49).

9. This is related to what philosophers sometimes call ‘stability’ or, following (Skyrms 1977), ‘resiliency.’

10. More precisely: 1) You should strongly believe that p. (assumption for conditional proof) 2) If you should strongly believe that p, then you should strongly believe that counterarguments to p are misleading. (see argument in the body of the article) Therefore, 3) You should strongly believe that counterarguments to p are misleading. (from 1 and 2) 4) If you should strongly believe a counterargument is misleading, you should do whatever that fact is a decisive reason for doing. (assumption, but see (Fantl 2018, Ch. 6) and (Fantl and Matthew 2009, Ch. 3) for details) Therefore, 5) You should do whatever the fact that a counterargument to p is misleading is a decisive reason for doing. (from 3 and 4) 6) In standard situations, that a counterargument is misleading is a decisive reason to be unwilling to reduce your confidence in response to the counterargument (even if you can’t figure out what’s wrong with it and all the steps seem compelling). (from the argument in the body of the article)Therefore, 7) In standard situations, you should be unwilling to reduce your confidence in response to counterarguments to p (even if you can’t figure out what’s wrong with it and all the steps seem compelling) 8) To be open-minded toward a counterargument is, in part, to be willing to reduce your confidence in response to the counterargument (even if you can’t figure out what’s wrong with it and all the steps seem compelling). (from the definition of open-mindedness) Therefore, 9) In standard situations, you should not be open-minded toward counterarguments to p. (from 7 and 8) Therefore, 10) In standard situations, if you should strongly believe that p, you should not be open-minded toward counterarguments to p. (from 1, 9, conditional proof). Therefore, 11) If you should be open-minded toward counterarguments to p, you should not strongly believe that p. (contrapositive of 10) And 11 entails C – that if you should always be open-minded toward counterarguments, then you shouldn’t ever have strongly held beliefs.

11. See (Fantl 2018, Ch. 2) for a more complete development of this argument.

12. See (Kortepeter 2020).

13. See https://www.facebook.com/maxshimbaministries/posts/power-against-evil-spiritual-marriagesthis-tormenting-spirits-are-responsible-fo/1622923351259433/.

14. See (Lockhart 2018).

15. For example, thirty one percent of respondents believe that the Coronavirus ‘was created and spread on purpose.’ (Uscinski and Enders 2020), and forty nine percent believe that antifa is a terrorist organization (see https://www.rasmussenreports.com/public_content/politics/current_events/social_issues/49_say_antifa_is_a_terrorist_organization).

16. See, e.g. (Swindlehurst 2020), (Littlejohn 2013), and (Sutton 2007).

17. See, e.g. (Greco 2015) and (Clarke 2013).

18. See, e.g. (Fassio 2020), (Hughes 2017), and (Fantl and Matthew 2009).

19. See, e.g. (Gao forthcoming).

20. See, e.g. (Lackey 2008) and (Coady 1992).

21. See, e.g. (Kawall 2013) and (Stroud 2006).

22. See, e.g. (Jones 2002) and (Goldman 2001).

23. See, e.g. (Hinton 2019) and (Coady 2006).

24. See, e.g. (Elliott forthcoming) and (Elliott and Resnik 2014). Thanks to Elias Yuan for alerting me to this literature.

25. See, e.g. (Intemann and Inmaculada 2014) and (Longino 1990).

26. See, e.g. (Rolin 2009).

27. See, e.g. (Harman 1986) and (Chisholm 1980).

28. See, e.g. (Huemer 2007).

29. Many thanks to David Liebesman, Matthew McGrath, attendees of the 2019 Canadian Society for Epistemology annual meeting, and three anonymous referees for this journal, who provided many thoughtful and helpful suggestions.