Monday, September 12, 2022

The Overlapping Dispositional Profiles of Different Types of Belief

Last spring at a workshop in Princeton, Neil Van Leeuwen presented some of his work on the differences between "beliefs" and "credences".  Beliefs, in Van Leeuwen's sense, are cognitive states that play the kinds of causal and epistemic roles that Anglophone philosophers normally associate with belief.  A belief like "there's beer in the fridge" arises in response to evidence (e.g., looking in the fridge and seeing beer) and is liable to disappear given counterevidence (e.g., your housemate tells you they finished the beer).  It governs practical action in a straightforward way (you'll go to the fridge if you want a beer, and you'll fail if your beer belief happens to be false).  And having or lacking the belief is not particularly distinctive of group membership (no one will kick you out of the beer-lovers club for realizing your fridge is beerless).

Credences work differently, on Van Leeuwen's view.  (Don't confuse "credence" in Van Leeuwen's sense with the more common use of "credence" in philosophy to mean something like degree of confidence.)  Typically, credences have religious or political or group-affiliative content.  They're the kind of thing that come readily to mind for most people when you ask them about their "beliefs": You believe that Jesus rose from the dead, that Black Lives Matter, and that Thousand Oaks High School is the best.  Unlike ordinary beliefs, your group identity isn't independent of whether you affirm these propositions: You're not a Christian if you don't hold that Jesus rose from the dead; you're not a good political liberal if you don't affirm that Black Lives Matter; you don't have proper school spirit if you don't agree that (in some hard to specify sense) your high school is "the best".

Also unlike ordinary beliefs, on Van Leeuwen's view, credences are not straightforwardly connected to evidence: You don't have these credences primarily because they are well supported by the evidence, nor are you likely to revise them in the face of counterevidence.  It's not always clear what counterevidence would even look like.  Nor is the connection to action as straightforward as for ordinary beliefs.  If you're wrong about Jesus, or Black Lives, or your high school, you can go about your ordinary life just fine.  No ordinary action plan depends on the truth of these propositions.

I think Van Leeuwen somewhat overdraws the distinction.  Credences can have some responsiveness to evidence, and their truth or falsity can matter to our actions.  And ordinary beliefs can also get tangled up in one's group identity (e.g., scientific beliefs about climate change or the age of the Earth).  Still, Van Leeuwen is onto something.  Different types of belief probably can have somewhat different functional roles.

In the question period after Van Leeuwen's presentation, Thomas Kelly posed an interesting challenge: If credence and belief really are different types of attitude, why does it seem like there's rational tension between them?  Normally, when attitude types differ, there's no pressure to align them: It's perfectly rational to believe that P and desire or imagine that not-P.  You can believe that it's raining and desire or imagine that it is not raining.  But with belief and credence as defined by Van Leeuwen, that doesn't seem to be so.  There's something at least odd, and arguably just straightforwardly irrational, in saying "I have religious credence that Jesus rose from the dead, but I don't believe that Jesus rose from the dead."  What explains this fact, if credence and belief really are distinct attitudes?

I confess I don't recall Van Leeuwen's reply.  But the discussion did trigger some thoughts of my own, grounded in my dispositionalist metaphysics of belief.

According to dispositionalism about belief, to believe some proposition, such that there is beer in the fridge, is nothing more or less than to have a certain dispositional profile.  It is to be disposed, ceteris paribus (all else equal or normal or right) to say, if asked, "yes, there's beer in the fridge" and to go to the fridge if one wants a beer.  It is be disposed, also, to think to oneself in silent inner speech, if the occasion arises, "there's beer in the fridge", and to feel surprise should one go to the fridge and find no beer.  It is to be disposed to draw related conclusions, such as that there is beer in the house and that there is something in the fridge.  And so on.  It is, in general, to have the behavioral, experiential, and cognitive dispositional profile that is characteristic of someone who believes the proposition in question.

To believe some proposition P, according to dispositionalism, is not to have some interior object, the belief or representation that P, stored discretely in some location in one's mind.  Dispositionalism is not strictly inconsistent with the existence of discrete interior representations, since in principle the dispositional architecture could be underwritten by such discrete interior representations.  But there no need to posit such representations, and not positing them helps you escape various thorny puzzles.

Two interesting features of dispositional profiles are:

(1.) They can overlap.

(2.) There can be more central and less central dispositions.

Intuitively, it is easiest to see this with personality traits.  Ordinary people -- or rather I should say ordinary people of the rather extraordinary sort who will read this far into an article or blog post on the metaphysics of belief -- don't appear to find dispositionalism intuitive for beliefs.  But they do for personality traits.  So it's generally a useful exercise in thinking about the structure of dispositionalism to start with personality cases and then analogize.

Think about the traits of being bold, courageous, and risk-tolerant.  These aren't exactly the same thing.  Someone who wagers big on a poker hand would be somewhat more aptly described as bold or risk-tolerant than courageous.  Someone who quietly risks their career to help out a junior colleague who is being mistreated would be somewhat more aptly described as courageous than bold or risk-tolerant.  But it's not exactly like the career-risker is not also risk-tolerant or even bold, and there's a kind of courage in the poker player.  To be bold, courageous, or risk-tolerant is to be disposed to act and react certain ways in certain situations, to have a certain general posture toward the world; and these postures have considerable overlap -- for example, none of the three will easily be daunted by the prospect of small losses.  More central to boldness is swift, decisive action -- but this is also somewhat characteristic of the courageous and risk-tolerant.  More central to courage is tolerating risk when morality demands risky action -- but this will also typically be true of the bold and risk-tolerant.  One can be risk tolerant without being especially bold or courageous, but flat out timidity and cowardice seem to be inconsistent with high risk tolerance.

The thought is not that there are three ontologically completely distinct personality traits that tend to correlate with each other.  Rather it's that the personality traits are not completely ontologically distinct.  Each comprises a similar, overlapping suite of dispositions such that in virtue of completely fulfilling one you also partly fulfill the others.  Compare also: being extraverted, sociable, and assertive, or being grumpy, irascible, and irritable.

Back to belief.  What Van Leeuwen calls beliefs and what he calls credences are both constituted, if we accept dispositionalism, by clusters of dispositions.  These clusters are somewhat different in emphasis -- like boldness and courage are different in emphasis -- but they overlap.  Central to ordinary belief is the cognitive disposition to structure mundane plans around the truth of the proposition (such as planning a trip to the fridge in a way that relies on the truth of the belief that there is beer in the fridge).  This is less central to credence in Van Leeuwen's sense.  Central to credence in Van Leeuwen's sense is affirmation in social contexts.  But affirmation in social contexts is also characteristic of ordinary belief, if not quite as central.

Imagine two overlapping networks of dispositions.  Anyone who possesses all of the dispositions in one network automatically possesses some of the dispositions in the other network.  Of course you needn't have all the dispositions to have the belief in question (compare: the extravert needn't be extraverted in every respect all the time to count as an extravert).  Typically, having enough of the dispositions in one set will also mean having enough of the dispositions in the other.

The figure below might serve as a representation.  The red dots are the dispositions constitutive of one attitude (belief, desire, personality trait, etc.), the blue dots are the dispositions constitutive of another attitude, and the size of the dots signifies their centrality to the dispositional structure constitutive of the attitude.  (The dots make dispositions look more discrete than they are, but let's not worry about that for this illustration.)

As the illustration suggests, you might be able to draw a figure around most of the largest dots for Attitude A while excluding many of the dots from Attitude B.  That would represent having most of the dispositions constitutive of Attitude B while lacking many of those constitutive of Attitude B.  The more overlap the attitudes have, the more careful the carving will have to be to generate that result.  

On this model it's not just that there's rational pressure not to have an ordinary belief that P alongside a credence that not-P.  It's actually ontologically impossible to be a full-on typical believer that P without also to a substantial extent having a matching credence and vice versa.  There might be some cases where "credence" captures things better than "belief" or the other way around (just like "courageous" might be a better fit than "bold"); but you won't find any cases where the person has 100% of the dispositions constitutive of the credence and also 100% of the dispositions constitutive of the opposite belief.  That would be like being a perfectly stereotypical example of a courageous person who is also a perfectly stereotypical example of a risk-avoidant person.

That's not to say that there's nothing to Van Leeuwen's distinction.  Ordinary belief and religious or political credence do differ.  But it's not that there are two discretely separate attitude representations stored in the mind that can freely agree or conflict with each other.  Rather, belief and credence are closely related, overlapping, but not identical patterns in our dispositional structure.

Similar remarks apply to other related attitudes, including:

  • knowing intellectually how to steer your car and having procedural knowledge of how to steer your car;
  • caring about justice, valuing justice, wanting justice, and thinking justice is good;
  • worrying there might be a war, fearing that there will be a war, and hoping that there won't be a war.



A Phenomenal, Dispositional Account of Belief (Nous, 36 [2002] 249-275).

Desiring, Valuing, and Believing Good: Almost the Same Thing (Aug 30, 2012).

Do You Have Infinitely Many Beliefs about the Number of Planets? (Oct 17, 2012).

A Dispositional Approach to Attitudes: Thinking Outside of the Belief Box (in N. Nottelmann, ed., New Essays on Belief; Palgrave, 2013).

It's Not Just One Thing, to Believe There's a Gas Station on the Corner (Feb 28, 2018).

Love Is Love, and Slogans Need a Context of Examples (Mar 13, 2021).


David Duffy said...

" need to posit such representations": The Law of Requisite Variety entails that there has to be such a representation in any system that is attempting to control the (interior or) outside world. "if a system S adapts to a class of external signals U, in the sense of regulation against disturbances or tracking signals in U, then S must necessarily contain a subsystem which is capable of generating all the signals in U. It is not assumed that regulation is robust, nor is there a prior requirement for the system to be partitioned into separate plant and controller components."

In the case of personality, and its relationship to belief, a dispositional model is perfectly reasonable. At one end of normal variation, we see paranoid belief systems that have much similarity to Van Leeuwen's "credences", and understand they are an exaggeration of "motivated reasoning". The authoritarian personality model for political conservatism is just a bit closer to the median of Reinforcement Sensitivity or Defensive Direction.

Paul D. Van Pelt said...

Credence = political or group affiliative content. Which, it follows, could include religious belief, as with the illustration of Christ's rising from death. Christians might typically group such beliefs under the heading of faith. The Professor's idea is interesting enough, but is it really all that novel? Seems to me that faith-based belief or believing was always already foundational to religious experiences, even before James ( that's William) wrote about them. Wherefore it also seems to follow as unnecessary to re-label religious belief as credence. Or, as you suggested, roughly, such re-labelling may amount to something like an over distinction. Cogent philosophy or hair-splitting? I leave that to your discretion. I'm just lookin' out my back door right now---literally.

Arnold said...

Propositions (Stanford Encyclopedia of Philosophy)....good reads thanks...

Did Socrates' propose attitudes for allowing dispositional self knowledge... epistemical empirical evidential are...
...sensational emotional mental mindful overlapping dispositional states...