University of California San Diego What is Belief Philosophy Essay

Content: Please choose one of the following topics:

Topic 1Do all of the following, in the format of a polished, persuasive essay. What is a belief? Give a clear, detailed definition. Make sure that you accurately describe each element involved. Clearly state the role beliefs are supposed to play in epistemology. Finally, describe realism about belief, and contrast it with eliminativism about belief. Which of these two positions do you think is more persuasive? Defend your choice with reasons. Topic 2Again, do all of the following, in the format of a polished, persuasive essay. What is truth? Based on our readings, give a clear, detailed definition (noting for example, possible differences between the concept and the property of truth). Accurately describe each element involved. Clearly state the role truth is supposed to play in epistemology. Finally, do you find Lewis’s arguments against correspondence theory of truth persuasive? Defend your answer with reasons.   634
the journal of philosophy
I
ALIEF AND BELIEF*
n March 2007, 4000 feet above the floor of the Grand Canyon, a
horseshoe-shaped cantilevered glass walkway was opened to the
public. Extending 70 feet from the Canyon’s rim, the Grand Canyon
Skywalk soon drew hundreds of visitors each day, among them New York
Times reporter Edward Rothstein, who filed the following dispatch:
A visitor to these stark and imposing lands of the Hualapai Indians on
the western rim of the Grand Canyon knows what sensation is being
promised at the journey’s climax. After driving for a half-hour over
bone-jolting dirt roads … you take a shuttle bus from the parking lot … .
You deposit all cameras at a security desk, slip on yellow surgical booties
and stride out onto a horseshoe-shaped walkway with transparent sides
and walls that extends 70 feet into space, seemingly unsupported.
Below the floor’s five layers of glass (protected from scratches by the
booties) can be seen the cracked, sharp-edged rock face of the canyon’s
rim and a drop of thousands of feet to the chasm below. The promise is
the dizzying thrill of vertigo.
And indeed, last week some visitors to this steel-supported walkway anchored in rock felt precisely that. One woman, her left hand desperately
grasping the 60-inch-high glass sides and the other clutching the arm of
a patient security guard, didn’t dare move toward the transparent center
of the walkway. The words imprinted on the $20 souvenir photographs
taken of many venturesome souls herald completion of a daredevil stunt:
“I did it!!!”1
Though some readers may find this story politically or aesthetically disturbing, none—I take it—find it perplexing.2 While the sarcasm of
* I am grateful to the Yale University faculty lunch group for comments on a very
early draft of this paper, and to audiences at Princeton University (March 2007), the
Central American Philosophical Association meeting in Chicago (April 2007), and the
Mind & Language Pretense Conference at University College, London ( June 2007) for
excellent questions, comments, objections, and suggestions regarding the talk which
served as its immediate predecessor. For more recent discussion and comments, I thank
John Bargh, Paul Bloom, Richard Brooks, Carolyn Caine, David Chalmers, Greg Currie,
Paul Davies, Andy Egan, Roald Nashi, Elliot Paul, Eric Schwitzgebel, Ted Sider, Jason
Stanley, Zoltán Gendler Szabó, and Jonathan Weinberg.
I discuss additional aspects of the notion of alief in a companion article, “Alief in
Action (and Reaction),” Mind & Language, xxiii, 5 (November 2008): 552–85.
1
Rothstein, “Skywalk Review: Great Space, Glass Floor-Through, Canyon Views,” The
New York Times (May 19, 2007).
2
Indeed, the story is a slight variation on the early modern “problem of the precipice,” discussed—among others—by Hume (Treatise 1.3.13.10, 148), Pascal (Pensées,
0022-362X/08/0510/634–63
ã 2008 The Journal of Philosophy, Inc.
alief and belief
635
“venturesome souls” is surely well placed, and the price of the “‘I did
it!!!’” photo is surely excessive, the basic phenomenon—that stepping
onto a high transparent safe surface can induce feelings of vertigo—is
both familiar and unmysterious.3
How should we describe the cognitive state of those who manage to
stride to the Skywalk’s center? Surely they believe that the walkway will
hold: no one would willingly step onto a mile-high platform if they
had even a scintilla of doubt concerning its stability. But alongside
that belief there is something else going on. Although the venturesome souls wholeheartedly believe that the walkway is completely safe,
they also alieve something very different. The alief has roughly the following content: “Really high up, long long way down. Not a safe place
to be! Get off!!”4
•
In a series of ingenious studies spanning several decades, psychologist Paul Rozin has demonstrated a widespread tendency for
well-educated Western adults to exhibit behaviors consonant with a
commitment to the existence of “laws of sympathetic magic:”5 that
“there can be a permanent transfer of properties from one object
… to another by brief contact” (contagion) and that “the action taken
on an object affects similar objects” (similarity).6
So, for example, subjects are reluctant to drink from a glass of juice
in which a completely sterilized dead cockroach has been stirred,
hesitant to wear a laundered shirt that has been previously worn by
section 44) and Montaigne (Essays, Donald Frame, trans. (Stanford: University Press,
1957), p. 250). See Saul Traiger, “Reason Unhinged: Passion and the Precipice from
Montaigne to Hume,” in Joyce Jenkins, Jennifer Whiting, and Chris Williams, eds., Persons and Passions: Essays in Honor of Annette Baier (Notre Dame: University Press, 2005),
pp. 100–15. I discuss precipice cases in more detail in Gendler (op. cit.).
3
The physiological explanation, of course, is that there is a mismatch in input between
the visual, vestibular and somatosensory systems. For discussion, see Thomas Brandt
and R.B. Daroff, “The Multisensory Physiological and Pathological Vertigo Syndromes,”
Annals of Neurology, vii, 3 (1980): 195–203; and Thomas Brandt, Vertigo: Its Multisensory
Syndromes (New York: Springer, 1999/2003, second edition).
4
Throughout my discussion, I am using the term ‘content’ in a somewhat idiosyncratic way, for want of a better term to describe the general notion that I wish to capture. As I am using the term, content need not be propositional, and may include—as
the example above makes clear—affective states and behavioral dispositions.
5
Cf. J.G. Frazer, The New Golden Bough: A Study in Magic and Religion (abridged) (New
York: Macmillan, 1959; edited by T.H. Gaster, 1922; original work published 1890); Marcel
Mauss, A General Theory of Magic, Robert Brain, trans. (New York: Norton, 1972; original
work published 1902) (as cited in Paul Rozin, Linda Millman, and Carol Nemeroff,
“Operation of the Laws of Systematic Magic in Disgust and Other Domains,” Journal of
Personality and Social Psychology, l, 4 (1986): 703–12).
6
Rozin, Millman, and Nemeroff, op. cit.
636
the journal of philosophy
someone they dislike, and loath to eat soup from a brand-new bedpan. They are disinclined to put their mouths on a piece of newly
purchased vomit-shaped rubber (though perfectly willing to do so with
sink stopper of similar size and material), averse to eating fudge that has
been formed into the shape of dog feces, and far less accurate in throwing darts at pictures of faces of people they like than at neutral faces.7
How should we describe the cognitive state of those who hesitate to
eat the feces-shaped fudge or wear their adversary’s shirt? Surely they
believe that the fudge has not changed its chemical composition, and
that the shirt does not bear cooties8—just as they believe that that the
newly purchased bedpan is sterile and that the fake vomit is actually
made of rubber: asked directly, subjects show no hesitation in endorsing such claims. But alongside these beliefs there is something else
going on. Although they believe that the items in question are harmless, they also alieve something very different. The alief has roughly the
following content: “Filthy object! Contaminated! Stay away!”
•
Last month, when I was traveling to the APA Program Committee
meeting, I accidentally left my wallet at home. I noticed its absence
when I arrived at the check-in desk at the Hartford Airport, and fully
expected to be turned away from my flight. Much to my surprise, the
desk agent simply wrote the words “No ID” on my boarding pass, and
told me to allow for a few extra minutes at security.9 The various scans
showed nothing amiss, so I boarded my plane, flew to Baltimore, and
made my way to the meeting site.
7
The descriptions of the cases make it clear that the experimenters go out of their
way to avoid the possibility of any sort of confusion. In the fudge study, for example,
“subjects were offered a piece of high-quality chocolate fudge, in a square shape, on a
paper plate [and then] ate the piece . … [Next] two additional pieces of the same fudge
were presented, each on its own paper plate.” Subjects were made explicitly aware that
the two pieces come from the same initial source, and that the only difference between
them is that “one piece was shaped in the form of a disc or muffin, the other in the
shape of a surprisingly realistic piece of dog feces.” Despite recognizing that they contained identical ingredients, subjects showed a striking reluctance to consume the fecesshaped piece. See Rozin, Millman, and Nemeroff, op. cit., p. 705.
8
For definition, see: http://en.wikipedia.org/wiki/Cooties. Apparently, a roughly
equivalent British term is ‘lurgi’.
9
Legally, one is not required to carry identification in order to fly. Rather, the Transportation Safety Administration (TSA) requires that airline passengers either “present
identification to airline personnel before boarding or be subjected to a search that is
more exacting than the routine search that passengers who present identification encounter.” Cf. Gilmore v. Gonzales, 04-15736 D.C No. CV-02-03444-SI Opinion. (Full text at
http://www.ca9.uscourts.gov/ca9/newopinions.nsf/A6AE4C85241C517C88257101007
B72EB/$file/0415736.pdf ?openelement.) As a quick internet search for “flying without
identification” will reveal, however, there is a gap between the law and the practice: there
were, no doubt, additional features of my particular circumstance that led me to be offered
this option.
alief and belief
637
Though the TSA may not require identification, restaurants and
hotels do require payment, so when I got to Baltimore, I arranged to
borrow money from a friend who was also attending the meeting. As
he handed me the bills, I said: “Thanks so much for helping me out
like this. It is really important for me to have this much cash since I
don’t have my wallet.” Rooting through my bag as I talked, I continued:
“It’s a lot of cash to be carrying loose, though, so let me just stash it in
my wallet ….”
How should we describe my mental state as my fingers searched for
my wallet to house the explicitly wallet-compensatory money? Surely I
believed that I had left my wallet in New Haven; after all, the reason I
was borrowing so much money was because I knew I had no credit
cards or cash with me. But alongside that belief there was something
else going on. Although I believed that my wallet was several hunded
miles away as I rooted through my bag, I simultaneously alieved something very different. The alief had roughly the following content:
“Bunch of money. Needs to go into a safe place. Activate wallet-retrieval
motor routine now.”
•
Charles is watching a horror movie about a terrible green slime. He
cringes in his seat as the slime oozes slowly but relentlessly over the earth
destroying everything in its path. Soon a greasy head emerges from the
undulating mass, and two beady eyes roll around, finally fixing on the
camera. The slime, picking up speed, oozes on a new course straight towards
the viewers. Charles emits a shriek and clutches desperately at his chair.10
How should we describe Charles’s cognitive state? Surely he does
not believe that that he is in physical peril; as Kendall Walton writes
“Charles knows perfectly well that the slime is not real and that he is
in no danger” (ibid., p. 6). But alongside that belief there is something
else going on. Although Charles believes that he is sitting safely in a chair
in a theater in front of a movie screen, he also alieves something very
different. The alief has roughly the following content: “Dangerous
two-eyed creature heading towards me! H-e-l-p …! Activate fight or
flight adrenaline now!”
i. introducing alief
I.1. Belief-Behavior Mismatch and Belief-Discordant Alief. In each of the
cases presented above, it seems clear what the subject believes11: that
10
Kendall Walton, “Fearing Fictions,” this journal, lxxv, 8 ( January 1978): 5–27,
see p. 5.
11
Although belief is clearly one of the central notions in epistemology, the question
of what belief is has been (with important exceptions) underexplored in this context.
638
the journal of philosophy
the walkway is safe, that the substance is edible or potable, that the
wallet is in New Haven, that the theater is in no danger of being invaded by slime, and so on. Ask the subject directly and she will show
no hesitation in endorsing such claims as true. Ask her to bet, and this
is where she will place her money. Ask her to think about what her
other beliefs imply and this is what she will conclude. Look at her overarching behavior and this is what it will point to. At the same time, the
belief fails to be accompanied by certain belief-appropriate behaviors
and attitudes: something is awry.
When else do we find this sort of belief-behavior mismatch? One
sort of case is that of deliberate deception. If I believe that I have a
winning hand, but I am trying to mislead you into thinking that I do
not, I will behave in ways discordant with my belief. But clearly, this is
not a good model for the cases just considered: Charles is not trying to
fool the movie-maker; Rozin’s subjects are not trying to mislead the
experimenters. In contrast to the cases of deliberate deception, the
belief-behavior mismatch in our cases is not the result of something
other-directed and deliberately controlled.
Perhaps, then, it is akin to a case of self-deception? A self-deceived
subject believes, say, that her child has committed some terrible crime,
but somehow brings herself to represent the situation—both to herself
and to others—as if she believed precisely the opposite, resulting in
(Of course, there have been extensive discussions of this question in the context
of philosophy of mind (for an overview, see section 1 of Eric Schwitzgebel, “Belief,”
The Stanford Encyclopedia of Philosophy (Fall 2006 Edition), Edward N. Zalta, ed., URL 5
http://plato.stanford.edu/archives/fall2006/entries/belief/).) But (with some important exceptions) this literature has remained largely insulated from the literature in epistemology). One might think a simple characterization would suffice—something like:
“To believe a proposition is to hold it to be true” (Simon Blackburn, The Oxford Dictionary
of Philosophy (New York: Oxford, 1996), p. 40). But, for reasons that David Velleman
brings out nicely (Velleman, “On the Aim of Belief, ” in The Possibility of Practical Reason
(New York: Oxford, 2000), pp. 244–82), this will not quite do (at least, not without a
careful spelling out of what “hold to be true” amounts to, which just pushes the question one step back). Moreover, the issue is complicated by there being at least two
apparently different fundamental notions of belief: what H.H. Price calls the “occurrence” or “traditional” view—that to believe a proposition is to be in a mental state
with a particular sort of introspectively available feature, such as “vivacity” or “liveliness”
or “solidity” (a view he attributes to, among others, Descartes, Hume, Spinoza, Cardinal
Newman and Cook Wilson)—and what he calls the “dispositional” or “modern” view—
that to believe a proposition is to be disposed to act in certain ways (a view he attributes
to, among others, Alexander Bain, R.B. Braithwaite, and Gilbert Ryle). See Price, Belief
(London: Allen and Unwin, 1969). I will have more to say about this matter below. In the
meantime—as the astute reader will have suspected by now—I invoke this legacy as
much to exculpate as to inform: though I will offer more details in subsequent sections,
for the time being, I will leave the notion of belief undefined. (For further discussion,
see section 2 of Gendler, “Alief in Action (and Reaction)” (op. cit.).)
alief and belief
639
the requisite belief-behavior mismatch. 12 This is an improvement
on the previous model; it corrects the problem of other-directedness,
and—to some extent—the problem of deliberate control. But it still
misrepresents the structure of the situation: it is not that the reluctant
walker on the Hulapai Skywalk believes that the surface is safe, but has
somehow deceived herself into thinking that it is risky; it is not that
Rozin’s subject believes that the bedpan is sterile, but somehow deceives herself into thinking that there’s some reason not to drink from
it. The mismatch runs two directions: unlike in cases of self-deception,
the subjects in our cases show no reluctance to endorse explicitly the
belief with which their behavior fails to accord. And unlike in cases of
self-deception, their behavioral responses do not result from some deliberate or quasi-deliberate process of misrepresentation.
Perhaps, then, the subjects’ hesitation to act on their beliefs is the
result of some sort of doubt or uncertainty? In setting out for the day,
I might dither a bit before leaving my umbrella at home: “it’s not
going to rain,” I might aver—though I am not completely certain that
I am right. Though the action-pattern is strikingly similar to some of the
cases above, the model is still inadequate. Stepping onto the Skyway,
eating the stool-shaped fudge, or staying seated in the theater is not
like willing oneself to play Russian roulette: it is not a case of discounting a low-probability outcome and hoping for the best. Charles does
not leave the theater thinking: “Phew! It’s lucky the slime stayed on
the screen this time!” Rozin’s subject does not breathe a sigh of relief
that the dart hitting the photograph did not actually harm her friend. I
was not rooting around on the off-chance that maybe my wallet really
was in my bag after all.13
Perhaps, then, the belief is temporarily forgotten? When I reach for
my wallet, perhaps it is that I just do not remember that it is not with
me. When I hesitate before the fudge, perhaps I have just lost track of
the fact that it is not dog feces. When I step timidly on the walkway,
perhaps I have just forgotten that it is solid. Perhaps. But I do not
think this could be the full story. Rozin’s subjects hesitate to eat the
soup even if they are vividly and occurrently entertaining the thought
12
I discuss these issues in more detail in Gendler, “Self Deception as Pretense,” Philosophical Perspectives: Mind (2008).
13
Nor are these cases of what Schwitzgebel (“In Between Believing,” The Philosophical
Quarterly, li, 202 (2001): 76–82) calls “in-between beliefs”—attitudes “that are not quite
accurately describable as believing that P, nor quite accurately describable as failing to
believe that P” (op. cit., p. 76)—cases such as “gradual forgetting, failure to think things
through completely, and variability with context and mood” (op. cit., p. 78). They are
closer to some of the cases that Price calls “half-beliefs” (op. cit., pp. 302–14); I discuss
Price’s examples in more detail below.
640
the journal of philosophy
“this is a completely sterile bedpan,” fully, consciously and with explicit attention to its meaning and implications. I was rooting around
in my bag for my wallet at the exact moment that I was vividly and
occurrently entertaining the thought “I left my wallet in New Haven,”
fully, consciously, with explicit attention to its meaning and implications. And certainly the Hulapai Canyon steppers have not forgotten
that the platform is safe, else they would do something a good deal
more dramatic than hesitate before taking the next step.
But if it is not a case where the subject is deceiving others, or selfdeceived, or uncertain, or forgetful, then why is stepping onto the
Skywalk different from stepping onto the back porch? The reason,
of course, is that each activates a different set of affective, cognitive,
and behavioral association-patterns. When the subject steps onto the
wooden porch, input to her visual system affirms her explicit conscious belief that the surface is solid and secure; this sets into motion
a train of associations and activates a number of motor routines. But
since these motor routines coincide with those activated by her explicit intention to walk across a surface that she believes to be solid,
there is no belief-behavior mismatch. When she steps onto the glass
platform, by contrast, input to her visual system suggests that she is
striding off the edge of a cliff. This visual input activates a set of affective
response patterns (feelings of anxiety) and motor routines (muscle
contractions associated with hesitation and retreat), and the visualvestibular mismatch produces feelings of dizziness and discomfort,
leading to additional activation of motor routines associated with hesitation and withdrawal.14 These motor routines compete with those activated by her explicit intention to walk across a surface that she
believes to be solid; the result is the belief-behavior mismatch adverted
to above.
Nor do we need anything so dramatic to make the point. The same
phenomenon occurs when I set my watch five minutes fast. The effectiveness of the strategy does not depend on my forgetting that the watch
is inaccurate, or on my doubting that it is really 9:40 rather than 9:45,
or my deceiving myself or others into thinking that it is five minutes
later than it is. Rather, as with the glass-bottomed Skywalk, when I look
at my watch, input to my visual system suggests that I am in a world
where the time is t15. This visual input activates a set of affective response patterns (feelings of urgency) and motor routines (tensing of
the muscles, an overcoming of certain sorts of inertia), leading to the
14
For detailed discussion, see Brandt, Vertigo (op. cit.), chapter 29 (“Visual Vertigo:
Visual Control of Motion and Balance”), pp. 409–40.
alief and belief
641
activation of behavior patterns that would not be triggered by my explicit, conscious, vivid, occurrent belief that it is actually only 9:40.15
The activation of these response patterns constitutes the rendering
occurrent of what I hereby dub a belief-discordant alief. The alief has
representational-affective-behavioral content that includes, in the case
of the Skywalk, the visual appearance as of a cliff, the feeling of fear
and the motor routine of retreat.16 Similar appeal to belief-discordant
alief can be made in each of the other cases. The visual appearance of
the feces-shaped fudge renders occurrent a belief-discordant alief
with the content: “dog-feces, disgusting, refuse-to-eat”—an alief that
runs counter to the subject’s explicit belief that the object before
her is composed of a substance that she considers delicious and appealing. The visual-motor input associated with throwing a dart at a
representation of a loved one renders occurrent a belief-discordant
alief with the content: “harmful action directed at beloved, dangerous
and ill-advised, don’t-throw”—an alief that runs counter to the subject’s
explicit belief that damaging a representation has no effects on the entity represented. The visual-motor input associated with handling cash
rendered occurrent my belief-discordant alief with the content: “Bunch
of money. Needs to go into a safe place. Activate wallet-retrieval motor
routine now”—an alief that ran counter to my explicit belief that my
wallet was in Connecticut while I was in Maryland. And so on.
I.2. A Provisional Characterization of Alief. In the remainder of the article, I argue for the importance of recognizing the existence of
alief—so-called because alief is associative, automatic, and arational.
As a class, aliefs are states that we share with nonhuman animals; they
are developmentally and conceptually antecedent to other cognitive
attitudes that the creature may go on to develop. And they are typically also affect-laden and action generating.17 I will argue that any
15
Examples of such cases are manifold. I think, for example, that many of the cases
of motivation by imagination discussed in David Velleman’s “On the Aim of Belief ” are
actually cases of motivation by alief. Likewise, I think that many of the cases of heuristicbased reasoning discussed by Daniel Kahneman and Amos Tversky are cases of decision
on the basis of alief. Cf. Kahneman, P. Slovic, Tversky, eds., Judgment under Uncertainty:
Heuristics and Biases (New York: Cambridge, 1982); Kahneman and Tversky, eds., Choices,
Values and Frames (New York: Cambridge, 2000); cf. also Veronika Denes-Raj and
Seymour Epstein, “Conflict between Intuitive and Rational Processing: When People
Behave against Their Better Judgment,” Journal of Personality and Social Psychology, lxi,
5 (1994): 819–29; and other work in the “dual processing” tradition. For additional
discussion, see Gendler, “Alief in Action (and Reaction)” (op. cit.).
16
Of course, stepping onto the wooden deck also renders occurrent an alief—
indeed many aliefs—but since those aliefs accord with the subject’s explicit beliefs,
we do not need to make appeal to them in order to explain her subsequent behavior.
17
An alternative term might be prelief, but this expression is already spoken for (cf.
J. Perner, S. Baker, and D. Hutton, “Prelief: The Conceptual Origins of Belief and
642
the journal of philosophy
theory that helps itself to notions like belief, desire, and pretense
needs to include a notion like alief in order to make proper sense
of a wide range of otherwise perplexing phenomena. Without such
a notion, I will contend, either such phenomena remain overlooked
or misdescribed, or they seem to mandate such a radical reconceptualization of the relation between cognition and behavior that traditional notions like belief seem quaint and inadequate. In short, I
will argue that if you want to take seriously how human minds really
work, and you want to save belief, then you need to make conceptual
room for the notion of alief.
Because alief is a novel notion, introduced to make sense of a cluster
of otherwise baffling cases, most of the paper will proceed by examination of specific examples. The heart of the paper lies in that discussion, and in the claim that consideration of such cases brings to light
issues of philosophical importance. At the same time, I will tentatively
offer a more abstract characterization of the concept that I am introducing, so that the general claim that I making can be properly assessed.
The account that follows is explicitly provisional. I have little doubt
that I have gotten some of the details wrong—and perhaps a good
deal more than the details. But it seems to me better to make an honest mistake by attempting to be precise than to avoid error by refusing
to be explicit. With that in mind, I offer the following tentative characterization of a paradigmatic alief:
A paradigmatic alief is a mental state with associatively linked content
that is representational, affective and behavioral, and that is activated—
consciously or nonconsciously—by features of the subject’s internal or
ambient environment. Aliefs may be either occurrent or dispositional.
Nearly every clause in this characterization merits a quick remark
or highlighting:
(1) Alief is a mental state…
Since I incline towards physicalism, this means that I think alief is also
a physical state. But it is a special sort of physical state—one that occurs in the brain of a conscious subject. And it occurs in her brain as
the result of her (or her genetic ancestors) having undergone certain
Pretence,” in Charlie Lewis and Peter Mitchell, eds., Children’s Early Understanding of
Mind (Hove, UK: Erlbaum, 1994), pp. 261–86). And in any case, it lacks the resonance
of the chosen term. One might also want to leave room for a notion related to desire in
something like the way that alief is related to belief. Had ‘prelief’ been available, one
might choose presire; since it is not, a suitable expression is cesire. (I remain utterly
agnostic about what sort of attitude cesire might be.)
alief and belief
643
sorts of experiences—experiences that result in the creation of clusters of associations with representational-affective-behavioral content.
(2) Alief is a mental state …
Alief is a state and not, say, an attitude. It is (I think) roughly what
Aristotle would call a hexis.
(3) … with associatively linked content …
That is, a cluster of contents that tend to be co-activated. The contrast
here is with discrete contents that fail to be linked through such
an association.
(4) … that is representational, affective, and behavioral …
In paradigmatic cases, an activated alief has three sorts of components:
(a) the representation of some object or concept or situation or circumstance, perhaps propositionally, perhaps nonpropositionally, perhaps
conceptually, perhaps nonconceptually; (b) the experience of some affective or emotional state;18 (c) the readying of some motor routine.19
(5) Paradigmatic alief is a mental state with content that is representational, affective, and behavioral …
Notwithstanding the characterization offered in (4), I do not want to
rule out the possibility of there being aliefs that involve the mental ac18
Our affective processing mechanisms seem to be fairly insensitive to the question
of whether the scenario under consideration is real, imagined, supposed or denied. (To
the extent that there is a difference in the intensity of our responses, this can be largely
traced to a difference in the intensity of the stimulus.) (Cf., for example, the literature
surveyed in Anthony R. Damasio, Descartes’ Error: Emotion, Reason and the Human Brain
(New York: Grosset, 1995), and The Feeling of What Happens: Body and Emotion in the
Making of Consciousness (New York: Harcourt Brace, 1999).) For discussion of this in
the context of fictional emotions, see Gendler and Karson Kovakovich, “Genuine Rational Fictional Emotions,” in Matthew Kiernan, ed., Contemporary Debates in Aesthetics
and the Philosophy of Art (Malden, MA: Blackwell, 2005), pp. 241–53; Paul Harris, The
Work of the Imagination (Malden, MA: Blackwell, 2000); T. Schroeder and C. Matheson,
“Imagination and Emotion” in Shaun Nichols, ed., The Architecture of the Imagination
(New York: Oxford, 2006), pp. 19–39.
19
This gives rise to a potential worry: that alief is not a fundamental mental state, but
instead an amalgam of several more primitive mental states: those of entertaining content R, experiencing affect A, and activating behavioral repertoire B. I reply: the fact
that our current vocabulary requires us to describe alief-content using three separate
terms does not show that the state is an amalgam of three others. Indeed, one might
even argue that it is out of these more primitive association patterns (“Mama, warmth
and comfort, purse lips to drink”) that the less fundamental differentiated attitudes like
belief, desire, and imagination are constructed. These are cognitive attitudes that rely on
the notion of representation (and misrepresentation), a distinction between seeming and
being, one that is largely absent from the more primitive state of alief. I discuss this issue
further in Gendler, “Alief in Action (and Reaction)” (op. cit.). (Thanks to Andy Egan for
raising this concern.)
644
the journal of philosophy
tivation of a different sort of associative cluster. Perhaps there are cases
where the activation occurs at a sufficiently low level to render the notion of representation inapplicable. Perhaps there are states that lack
an obvious affective ingredient, or that do not include the clear activation of a motor routine, but that nonetheless sufficiently resemble our
paradigm cases that we want to count them as aliefs. Perhaps there are
cases where the most noticeable associations are not easily subsumed
under the three categories offered—cases that primarily involve the
heightening or dampening of certain sorts of attention, or the heightening or dampening of certain perceptual sensitivities.
(6) Alief is a mental state with … behavioral … content.
That is: alief itself does not involve the execution of these motor routines; it merely involves their activation (alief is a mental state). At the
same time, this activation renders it more likely that the routine will
actually be performed.20
(7) Alief … content … [may be] activated … consciously or nonconsciously.
That is: a subject may (occurrently) alieve something with or without
being aware of being (put into) in such a state.
(8) Alief … content … [may be] activated … via features of the subject’s internal
or ambient environment.
That is: the activation of an alief may be the result either of (conscious or
nonconscious) (quasi-)perception, or of (conscious or nonconscious)
nonperceptual thought.21
(9) Aliefs may be either occurrent or dispositional.22
20
William James calls the principle that “the mere act of thinking about a behavior
increase[s] the tendency to engage in that behavior” the principle of ideomotor action. He
writes: “We may then lay it down for certain that every [mental] representation of a
movement awakens in some degree the actual movement which is its object; and awakens it in a maximum degree whenever it is not kept from so doing by an antagonistic
representation present simultaneously to the mind” ( James, The Principles of Psychology
(1890), available on-line at http://psychclassics.yorku.ca/James/Principles/). Or again:
“Merely thinking about a behavior makes it more likely to occur, even if it is unintended
… the mere act of thinking about a response, even when the thought involved is meant
to help prevent the response, has the automatic effect of increasing the likelihood of
that response” ( John Bargh, Mark Chen, and Lara Burrows, “The Automaticity of
Social Behavior,” Journal of Personality and Social Psychology, lxxi, 2 (August 1996):
230–44, see p. 232, discussing work by Daniel Wegner).
21
As Bargh, Chen, and Burrows write: “Recent research has shown that attitudes and
other affective reactions can be triggered automatically by the mere presence of relevant objects and events … without conscious attention or awareness …. [They] then
exert their influence on thought and behavior” (ibid., p. 230, citations omitted).
22
For discussion of this distinction in the case of belief, see Price, op. cit.; David M.
Armstrong, Belief, Truth and Knowledge (New York: Cambridge, 1973); William G. Lycan,
alief and belief
645
A subject has an occurrent alief with representational-affective-behavioral
content R-A-B when a cluster of dispositions to entertain simultaneously
R-ish thoughts, experience A, and engage in B are activated—consciously
or unconsciously—by some feature of the subject’s internal or ambient
environment. A subject has a dispositional alief with representationalaffective-behavioral content R-A-B when there is some (potential) internal or external stimulus such that, were she to encounter it, would
cause her to occurrently alieve R-A-B.23
(10) Tentative characterization …
Despite all that I have said in this section, I continue to waver on
whether it would be better to think of the term as two-place (S alieves
R) rather than four-place (S alieves R-A-B) relation. Had I opted for the
former, I might have introduced the expression as follows:
S (occurrently) alieves R when S’s R-related associations are activated
and thereby rendered cognitively, affectively and behaviorally salient.
In most of the discussion that follows, I will make use of the expression
in its four-place version, occasionally noting cases where the two-place
version seems more appropriate.
I.3. Examples and Usage. How does the terminology just introduced
help us with our opening examples? Consider, for example, Rozin’s
subject who shows reluctance to put a piece of vomit-shaped rubber
in her mouth. When the visual experience as of vomit awakens in the
subject the entertainment of vomit-related trains of thought, the affective experience of disgust, and the activation of motor routines associated with behaviors like retreat and avoidance, Rozin’s subjects
come to alieve occurrently the representational-affective-behavioral content: “Vomit! Disgusting! Stay away!”24 And anyone whose inclinations
to feel disgust and avoidance would be activated by encountering a
vomit-like visual stimulus (a class which for evolutionary reasons is
“Tacit Belief,” in R.J. Bogdan, ed., Belief: Form, Content, and Function (New York: Oxford,
1986), pp. 61–82; John R. Searle, The Rediscovery of the Mind (Cambridge: MIT, 1992);
and Robert Audi, “Dispositional Beliefs and Dispositions to Believe,” Noûs, xxviii
(1994): 419–34. (References thanks to Schwitzgebel, op. cit.)
23
Obviously, there need to be some restrictions on what this causal relation looks
like: the connection must be nondeviant, and the encounter must not in itself bring
the dispositional alief into existence.
24
In fact it is likely that you right now—prompted by the associations set into play
through imagining such a case—occurrently alieve something with similar (though decidedly milder) content.
646
the journal of philosophy
likely to include nearly everyone) dispositionally alieves what Rozin’s
subjects occurrently alieve.
Of course, occurrently alieving “Vomit! Disgusting! Stay away!” is
fully compatible with occurrently believing that there is no vomit in
one’s vicinity. An occurrent alief whose content is P may well be accompanied by an occurrent belief whose content includes not-P. Indeed, it
is precisely when they are belief-discordant that aliefs tend to be evident to us. It is because Rozin’s hesitating subjects occurrently believe
something like: “the object in front of me is made of sterilized rubber
and poses no risk to my health” that we need to explain their reluctance
in terms of their alief. (Actually, I think that alief plays a major role in
explaining behavior even when it is belief-concordant, an issue to
which I return briefly in the closing section. But since the most convincing cases are those involving belief-discordant alief, I will focus
primarily on those in making my initial argument.)
One final remark concerning usage. Given that I have opted for
the four-place characterization, I need to say that Rozin’s subjects occurrently alieve something like “Vomit! Disgusting! Stay away!” while
believing that there is no vomit in their vicinity. Had I opted for the
two-place characterization, I might have said instead: Rozin’s subjects
believe that that the object before them is a piece of rubber, but they
alieve that it is a mound of vomit. This usage seems particularly tempting in cases where the associational clusters are awakened by the presence of a particular object or situation, and where the associations
awakened tend to be similar across individuals. Indeed, there is a natural tendency to loosen usage yet further, saying, for example, that
visitors to the Skywalk believe that the glass surface is safe, but alieve
that it is dangerous; that Rozin’s dart-throwers believe that damaging
the picture will not harm their loved one, but alieve that it will; that
Rozin’s shirt-avoiders believe that their enemy’s laundered chemise is
utterly harmless, but alieve that wearing it is ill advised; that Charles
believes that he is at no risk from the slime, but alieves that it is about
to attack him. I consider it a live possibility that careful reflection on
natural patterns of usage will reveal that I have made the wrong decision in opting for the four-place characterization. But for the time
being, I will explore the advantages of employing the term in the way
that I have characterized it thus far.
This ends the official introduction of the notion of alief. In the
remainder of the paper, I do three things. In section ii, I offer some
brief additional general remarks about the relation between the state
of alief and propositional attitudes such as belief, desire, and pretense. In section iii, I offer a series of examples—drawn from recent
empirical work in psychology—that played a central role in convincing
alief and belief
647
me that appeal to the notion of alief is crucial if we wish to hold on to
a notion like belief that relates to action in anything like the way philosophers have traditionally assumed. In section iv, I close with a few
speculative remarks about ways that appeal to the notion of alief may
be help us to make sense of two apparently unrelated phenomena: the
tendency of examples to affect us in ways that abstract descriptions do
not; and the role of habit in Aristotelian ethics.
ii. alief and other attitudes
II.1. Alief, Belief and Imagination. Why can’t alief be assimilated to one of
the more familiar cognitive attitudes—belief, for example, or imagining?
There are a number of reasons that I think that it cannot, which I will
present in the remainder of this section.25
Alief differs from both imagining and believing along certain crucial dimensions. If I believe that P, I believe that it is true that P, and
my belief is nondefective only if, as a matter of fact, it is true that P. If
I suppose or imagine or pretend that P, I suppose or imagine or pretend that it is true that P, but the actual truth or falsity of P is explicitly
irrelevant to my successfully supposing or imagining or pretending it
to be. Both classes of states, then, involve what Velleman helpfully calls
accepting a proposition: to believe or imagine or suppose or pretend
that P is to regard P as true (in some way26). But though they coincide in
this dimension, they differ in another: whereas belief is reality-sensitive,
supposition and imagination and pretense are explicitly reality-insensitive.
It is this latter disparity that is typically taken to underlie one important
difference between belief on the one hand, and supposition, imagination, and pretense on the other: whereas (modulo certain complications) we can imagine pretty much any content, we can (without
acrobatics) believe only what we take to be true.
How does alief fare along these dimensions? Strictly speaking, it lies
in another plane altogether. Believing and supposing and imagining
and pretending are all (at least on certain uses of the expressions in
question) propositional attitudes, whereas alieving (as I am provisionally using the expression) is not. But we can, by employing the “loose”
usage adverted to above, make reasonable sense of the notion of
25
For additional discussion, see Gendler, “Alief in Action (and Reaction)” (op. cit.).
He writes: “Regarding-as-true [is] … involved in … believing … [in] supposing or
assuming, and in propositional imagining as well … . To imagine that p is to regard p
as describing how things are …. Imagining is therefore a way of regarding a proposition
as true—or, to introduce a term, a way of accepting a proposition” (Velleman, “On the
Aim of Belief,” op. cit., p. 250). Note that Velleman’s use of the term ‘acceptance’ is somewhat different than that of L. Jonathan Cohen (An Essay on Belief and Acceptance (New York:
Oxford, 1992)) and Michael Bratman (“Practical Reasoning and Acceptance in a Context,” as reprinted in Bratman, Faces of Intention (New York: Cambridge, 1999), pp. 15–34).
26
648
the journal of philosophy
alieving that P, and we can ask—keeping in mind that our usage is
loose—whether alieving that P involves accepting that P. We will need
to be a bit more careful when we ask whether alief is reality-sensitive or
reality-insensitive, and whether we are in a position to alieve at will. But
again, we will be able to draw certain fairly sharp contrasts between
alief and other attitudes.
Let us begin with the question of acceptance. Does alieving that P
involve accepting that P ? (That is, does being an alief state with the
content R-A-B involve regarding it as true in some way that R is part
of one’s real or imagined environment?27) Interestingly, the answer to
this question turns out to be: no, and the way in which it turns out to
be no reveals something important about the nature of alief. Unlike
belief or pretense or imagination or supposition, alief does not involve
acceptance. Though the point can be made on conceptual grounds
alone, it is helpful to begin with a specific example.
In a 1986 study by Rozin, subjects saw “sugar poured into two bottles,
and then applied labels of sugar and sodium cyanide, each to one of the
bottles, making their own choice.” Despite having applied the labels
themselves, subjects “showed a reluctance to consume sugar from
the cyanide labeled bottle.”28 So far, the case is a familiar one: while
Rozin’s subjects believed that both bottles contained sugar, consideration of the second rendered occurrent an alief state with the content
“cyanide, dangerous, avoid” associated with the second bottle—and
this belief-discordant alief played a role in governing their behavior.29
Up to this point, there is no reason to posit a case of alief without acceptance: in alieving “cyanide, dangerous, avoid” the subject is regarding as true (perhaps in imagination) that the bottle contains cyanide.
The interesting case comes from a follow-up study four years later.
In that study:
Subjects faced two empty brown 500 ml bottles. In the presence of the
subject, the experimenter opened a container of “Domino” cane sugar,
27
I am here skating over the difficult question of whether there is a uniform rule for
stating what one (loosely) alieves when one (strictly) alieves R-A-B.
28
Rozin, Maureen Markwith, and Bonnie Ross, “The Sympathetic Magical Law of
Similarity, Nominal Realism, and Neglect of Negatives in Response to Negative Labels,”
Psychological Science, i, 6 (November, 1990): 383–84, see p. 383, reporting results from
Rozin and Carol J. Nemeroff, “The Laws of Sympathetic Magic: A Psychological Analysis
of Similarity and Contagion,” in J. Stigler, G. Herdt, and R.A. Schweder, eds., Cultural
Psychology: Essays on Comparative Human Development (New York: Cambridge, 1990),
pp. 205–32.
29
As Rozin reports, subjects “knew this response was foolish, but felt the reluctance
anyway. This suggests a ‘low-level’ gut feeling, that can influence behavior in spite of
countering cognitions”—“The Sympathetic Magical Law of Similarity, Nominal Realism,
and Neglect of Negatives in Response to Negative Labels,” p. 383.
alief and belief
649
and poured some into each bottle, so that about ¼ of each bottle was
filled. The experimenter informed subjects that she was pouring sugar
into each bottle. The experimenter then presented the subject with two
typed labels. One had not sodium cyanide, not poison written on it, with a
red skull and cross bones preceded by the word not. The other label had
sucrose, table sugar typed on it. The subject was invited to put one label
on each bottle, in any way he or she chose. The experimenter then set
out two different colored plastic cups, one in front of each bottle, and
poured unsweetened red (tropical punch) “Kool-Aid” from a glass pitcher
into both, until they were about half full. Now, using separate, new plastic
spoons for each bottle, the experimenter put a half spoonful of powder
from one sugar bottle into the glass standing in front of that bottle, and
repeated this with the other glass for the other sugar bottle (ibid.).
Subjects then faced the choice of drinking from the cup containing
the sugar that had been labeled “sucrose, table sugar” or from the cup
containing the sugar that had been labeled “not sodium cyanide, not
poison.” Though the effect was somewhat less pronounced than in the
original study, subjects showed considerable reluctance to drink from
the latter.
Here again, while Rozin’s subjects believed that both bottles contained sugar, consideration of the second bottle rendered occurrent
an alief state with the content “cyanide, dangerous, avoid.” But in this
case, the label read precisely the opposite: it “had not sodium cyanide,
not poison written on it, with a red skull and cross bones preceded by
the word not.” So, although these subjects were in an alief state with the
content “cyanide, dangerous, avoid,” the content they were prompted
to imagine was exactly the opposite. They did not—as the acceptance
condition requires—regard it as true in some way that cyanide is to be
found in the vicinity; instead, it was the negated presence of the word
“cyanide” that rendered occurrent their cyanide-associated aliefs.
Can we explain this with the resources of only belief and imagining?
Clearly, belief cannot do the work: it is implausible to suggest that the
subject believed that the bottle she had labeled “not sodium cyanide,
not poison” contained cyanide. But what about imagining? Can’t we
say that the source of the subject’s hesitation is that she first imagines
that the bottle does contain poison, and that she then somehow negates this, and that this enables her (perhaps in some special Sartrean
fashion) to imagine the absence of poison?30
30
As in the following joke. Jean-Paul Sartre was sitting in a cafe when a waitress approached him: “Can I get you something to drink, Monsieur Sartre?” Sartre replied,
“Yes, I’d like a cup of coffee with sugar, but no cream.” Nodding agreement, the waitress walked off to fill the order, returning a few minutes later. “I’m sorry, Monsieur
Sartre,” she said, “we are all out of cream—would you like your coffee with no milk
instead?” (Taken with slight variation from http://www.workjoke.com/projoke70.htm.)
650
the journal of philosophy
Perhaps this is indeed what happens. But how is this supposed to
explain the subject’s hesitancy to drink the liquid? Is the reason for
her hesitancy supposed to be that she had been imagining that the bottle
contained cyanide, though now she is not—and that what she imagined in the past (though fails to imagine now) somehow explains her
action at present? Or that her current imagining that the bottle does
not contain cyanide somehow contains within it (in not-fully-aufgehoben
form) the antithetical imagining that the bottle does contain cyanide?
And that somehow this negated semi-imagined content—content that
she has, throughout the entire process, been fully consciously aware of
explicitly disbelieving—sneaks into the control center for her motor
routines and causes her to hesitate in front of the Kool-Aid?
Really? Is this really what you think imagining is like? Or have you
just described a case of belief-discordant (and imagination-discordant)
alief: a case where the subject believes that the bottle does not contain
cyanide, imagines that the bottle does not contain cyanide, yet has an
occurrent alief with the content: cyanide, dangerous, avoid? Is it not a
lot more natural to describe this as a case of alief-motivated behavior
than as a case of motivation by (past or negated) imagination? And if it
is alief that is doing the explanatory work here, is it not plausible that
alief is doing the explanatory work in the cases above as well?
For those unconvinced by examples or lines of rhetorical questioning, there is a more general argument for why alief can occur without
acceptance. At its core, alief involves the activation of an associative
chain—and this is something that can happen regardless of the attitude that one bears to the content activating the associations. (Indeed,
since alief may be activated nonconsciously, one may bear towards that
content no attitude at all.) This means that alief contexts are what we
might call hyperopaque: they do not permit salva veritate substitution
even of expressions that the subject explicitly recognizes to be coreferential.31 Even if I believe that the phrases “not poison” and “safe to consume” pick out coextensive classes of substances, even if I focus on that
belief and hold it vividly before my mind, even if the synonymy of these
two terms is crucial to my views about some other matter, still the aliefs
activated by the two expressions may be wildly dissimilar.32 Imagination,
by contrast, is not hyperopaque in this way. If I explicitly recognize that
P and Q are synonymous, and I imagine P while focusing explicitly on
31
Note that they are not hyperopaque in a stronger sense: they do permit salva
veritate substitution of expressions with respect to which the subject holds corresponding patterns of alief. (Thanks to Dave Chalmers for pointing out this stronger reading.)
32
Likewise (in a slight variation on a Kantian theme), my triskaidekaphobia may be
elicited by ‘13’, but not by ‘716’. This feature of alief will turn out to be important in
the discussion in section iv below.
alief and belief
651
the co-referentiality of P and Q, then in imagining P I imagine Q. Alief
just is not imagination.
The same features that explain alief’s hyperopacity and the possibility of alief without acceptance explain why we are not in a position to
alieve at will. If I believe that P, and subsequently learn that not-P, I will
revise my belief. If I imagine that P, and subsequently learn that not-P,
I will make no such revision. But what if I (loosely speaking) alieve
that P, and subsequently learn that not-P ? What happens then? At first
glance, alief seems to behave like imagination and its kin: after all, the
cases above are all cases where the subject truly and consciously believes P while actively alieving not-P. But this does not quite capture
the full story. If I believe that P and imagine that not-P, I am violating
no norms. But if I believe that P and alieve that not-P, something is
amiss. Learning that not-P may well not cause me to cease alieving that
P—but if it does not, then (though other considerations may override
this) I am violating certain norms of cognitive-behavioral coherence.
No such criticism is possible in the analogous case of imagining.
To the extent that action is supposed to be responsive to reality, the
well-functioning aliever is one whose aliefs and beliefs largely coincide
(or one whose ability to suppress contrary impulse is strong33). But
alief just is not reality-sensitive in the way belief is. Its content does
not track (one’s considered impression of) the world. At the same time,
it is not reality-insensitive in the way that imagination is. For while we
can (for the most part) imagine at will, we do not seem to have the
same sort of freedom in alief.34 We may be relatively unconstrained
in which of our dispositional aliefs we render occurrent—at least in
the case of those aliefs that can be rendered occurrent through contemplation alone—but we are far from unconstrained in which dispositional aliefs we have in the first place. Our dispositional aliefs
depend on the associational patterns that have been laid down in
our minds as the result of our experiences and those of our genetic
ancestors. We are not in a position to generate such patterns of association merely at will.
So it looks like, just as it is (something close to) conceptually impossible to believe at will, it is practically impossible to alieve at will. Of
33
As William James writes: “To make our nervous system our ally instead of our
enemy … we must make automatic and habitual, as early as possible, as many useful
actions as we can” (op. cit.).
34
It is the reality-sensitivity of belief that is typically taken to explain the impossibility of believing at will. Cf. Bernard Williams, “Deciding to Believe,” reprinted with
new pagination in Williams, Problems of the Self (New York: Cambridge, 1970/1973),
pp. 136–51, for a classic articulation of this view. (Thanks to Ted Sider for suggesting
that I consider this issue in the context of alief.)
652
the journal of philosophy
course, in both cases we might use all sorts of tricks to bring ourselves
to be in a certain sort of mental state—“roundabout routes” involving
processes that we ourselves deliberately initiate.35 But if we use such
tricks to cultivate beliefs, we need to cover our tracks;36 if we use them
to cultivate aliefs, we can do so under conditions of full disclosure.
This concludes the brief survey contrasting alief with attitudes like
belief and imagining. We now turn to the second issue of this section,
the relation between these attitudes, and the bringing about of behavior. I will suggest that alief’s special structure—its being a mental
state with affective, representational, and behavioral content that is
activated by features of the environment—means that it poses problems for behavioral accounts of belief that are especially severe.
II.2. Alief and Behavior. According to what Velleman has dubbed the
“purely motivational view of belief,” “all that’s necessary for an attitude to qualify as a belief is that it disposes the subject to behave in
certain ways that would promote the satisfaction of his desires if its
content were true. An attitude’s tendency to cause behavioral output
is thus conceived as sufficient to make it a belief.”37 Or, again: to believe that P is to be disposed to act in ways that would tend to satisfy
one’s desires, whatever they are, in a world in which P (together with
one’s other beliefs) were true.38
There are at least three sorts of marginal cases where this sort of
analysis seems to go awry—two that pose problems for necessity, the
third for sufficiency. The first sort are cases where (arguably) a subject
35
Cf. Williams (op. cit.); for instructions, see Pascal, for example, “The Wager,” from
Pensées, reprinted in Gendler, Susanna Siegel, and Steven M. Cahn, eds., The Elements of
Philosophy: Readings from Past and Present (New York: Oxford, 2007).
36
For traditional discussions in addition to Williams (op. cit.), see Barbara Winters,
“Believing at Will,” this journal, lxxvi, 5 (May 1979): 243–56; Jonathan Bennett,
“Why Is Belief Involuntary?” Analysis, l, 2 (March 1990): 87–107; cf. J.T. Cook, “Deciding to Believe without Self-Deception,” this journal, lxxxiv, 8 (August 1987): 441–46.
There has been a recent resurgence of interest in these issues: see, for example, Philip
Pettit and Michael Smith, “Freedom in Belief and Desire,” this journal, xciii, 9 (September 1996): 429–49; Paul Noordhof, “Believe What You Want,” Proceedings of the
Aristotelian Society, ci (2001): 247–65; Matthias Steup, “Doxastic Freedom,” Synthese, clxi,
3 (2008): 375–92; and essays cited therein.
37
Velleman (op. cit., p. 255). Velleman rejects this view, for reasons related to the ones
discussed here, but notes that the view has been widely endorsed, by philosophers as
diverse as R.B. Braithwaite, “The Nature of Believing,” Proceedings of the Aristotelian
Society, xxxiii (1932–33), pp. 129–46; David Armstrong, Belief, Truth, and Knowledge
(New York: Cambridge, 1973); W.V. Quine and J.S. Ullian, The Web of Belief (New York:
Random House, 1978, second edition); Robert Stalnaker, Inquiry (Cambridge: MIT,
1984); Lynn Rudder Baker, Explaining Attitudes: A Practical Approach to the Mind (New
York: Cambridge, 1995); and Daniel Dennett, “Intentional Systems,” this journal,
lxviii, 4 (February 25, 1971): 87–106, and “Do Animals Have Beliefs?” in Herbert
Roitblat, ed., Comparative Approaches to Cognitive Sciences (Cambridge: MIT, 1995), pp. 111–18.
38
Stalnaker, op. cit., p. 15; cf. Dennett, op. cit.
alief and belief
653
believes that P, but where this belief does not bring with it a disposition
to act in P-concordant ways because of some feature of the subject.
(Think, for example, of an immutable omniscient purely contemplative God, a permanent paralytic, a subject built to act with utter randomness, a character under an unbreakable spell that causes him to
act contrary to his first-order intentions, a hopeless akratic, or an agent
who aims always to deceive.) The second are cases where (arguably) a
subject believes that P, but where this belief does not bring with it a
disposition to act in P-concordant ways because the belief itself has
no behavioral implications. (Think, for example, of a subject who believes in causally-inert invisible goblins, or of a subject who believes
that she inhabits a space that is distorted but Euclidian (rather than
undistorted but non-Euclidian).39) The third are cases where, although
the subject is disposed to act in the requisite ways, she nonetheless fails
to believe that P because she lacks beliefs (either locally or globally).
(Think, for example, of a super-stoic who acts and has desires but always withholds assent, or of a hyper-Van Fraassenite who extends his
constructivist commitments to the realm of the observable.)
Five-finger exercises that they are, these marginal cases do not show
that there is anything deeply wrong about the motivational view. All
that is needed to avoid them are a few tweaks to the notion of disposition and a reiteration of the irrelevance of mental states. The big guns
come loaded with a different sort of ammunition: not with the suggestion that the view is wrong in certain far-fetched contrived cases, but
with the assertion that it is problematic through and through because of
a wide range of attitudes—among them acceptance (Michael Bratman),
imagination (Gregory Currie, Velleman) and pretense (Tyler Doggett
and Andy Egan, Velleman)—may motivate P-concordant behavior.40
Even here, I think there is room for the defender of a neo-behaviorist
account. Restrict yourself to nondeviant subjects, and retreat, say, to
betting behavior or high-stakes situations. Once again, you can save the
letter of the view that belief and behavior go hand in hand.
To some extent, this strategy works for alief as well. (If it did not, it
would be hard to maintain that the paradigmatic cases above are ones
39
If you are worried about verbal reports counting as behavior, add the requisite caveat that they never speak about this particular belief.
40
See Bratman, “Practical Reasoning and Acceptance in a Context,” op. cit.; Currie,
“XI—Imagination as Motivation,” Proceedings of the Aristotelian Society, cii, 1 (2002):
201–16; Velleman, “The Aim of Belief,” op. cit.; Egan and Doggett, “Wanting Things
You Don’t Want,” Philosophers’ Imprint, forthcoming. For my own take on these issues,
see “On the Relation between Pretense and Belief,” in Matthew Kieran and Dominic
McIver Lopes, eds., Imagination, Philosophy, and the Arts (New York: Routledge, 2003),
pp. 125–41, “Imaginative Contagion,” Metaphilosophy, xxxvii, 2 (April 2006): 183–203,
and “Self-Deception as Pretense,” op. cit.
654
the journal of philosophy
in which the subject believes that P but alieves that not-P.) H.H. Price,
whose underappreciated discussion of related examples deserves
more detailed attention than I have space for here, employs such a
strategy. Defending his account of a case of what he calls “half-belief,”
Price writes:
It might be suggested that the man who avoids walking under ladders
does just believe (however unreasonably) that walking under ladders
has bad consequences … . After all, these people act as if they believed,
and they often go to considerable trouble in consequence. They step off
the pavement into a muddy street or even into a street full of traffic, to
avoid the ladder …. Moreover they show the emotional symptoms of
belief, for example, discomfort or unrest if there is … no way of avoiding
the ladder …. Of course, these people will not admit that they … believe
these propositions; not even to themselves, and still less in public …. But
one may hold beliefs … without admitting to oneself that one holds them
(op. cit., p. 310).
Price rejects this account—a proposal, he suggests, to “dispense with
the concept of half-belief altogether”—because, while
… no doubt there are some who do wholly believe that their chances of
suffering misfortunes are increased if they walk under a ladder …. I do
not think that this is the usual situation, … the ordinary person who
avoids walking under ladders does not seriously believe that walking under ladders does any harm, or at any rate he does not believe it with
complete seriousness. We notice that if it is very important for him to
get to his destination quickly (for example, if he will miss a train if he
does not hurry) he does not seem to mind the ladder at all. He sees it—
there it is, in front of his nose—but he goes straight under it without
hesitation. He himself, if he thinks about his experience afterwards, will
be able to notice that he felt no qualms at all about doing the thing
which he ordinarily avoids so carefully (op. cit., pp. 310–11).
“A half-belief,” he concludes, is “something which is ‘thrown-off’
when circumstances alter…. [I]n some contexts to which the proposition is relevant one is in a belief-like state about it, but in other contexts to which it is equally relevant one disbelieves it or disregards it.”
This is so even though “in both sorts of contexts, the evidence for the
proposition … remains the same, and the probability of the proposition
is as great, or as little, as it was before” (op. cit., p. 312).
I agree with Price that the ladder case might well proceed as he
describes. But I am not so clear that his analysis will work for the cases
presented on the opening pages. Suppose it is very important for
me to get to my train, but that the station lies across a chasm fifty feet
wide and 1000 feet deep, bridged by a transparent glass walkway. Even
if I “will miss a train if [I do] not hurry,” I do not think it is true that I
alief and belief
655
would “not seem to mind the [apparent chasm] at all,” crossing it
“without hesitation” even though the visual stimulus is “right under
my nose.” I very much doubt that in “think[ing] about [my] experience
afterwards,” I would “be able to notice that [I] felt no qualms at all
about doing the thing which [I] ordinarily avoid so carefully.” (Indeed,
in my own case, I am not sure I could make it across the bridge at all
without closing my eyes—which would be, of course, to suspend the
occurrent alief by suspending the feature that activates it.)
Suppose we raise the stakes. My child is on the other side of the
chasm, and I need desperately to reach him to prevent some dreadful
occurrence. Here I suspect I could make it across the bridge—eyes
open—to perform the rescue: after all, I believe that he is in danger,
and I believe that the bridge is safe. But even here, the hesitation
would not fully dissipate. And not because I doubt in any way that
the surface is sturdy: I see others walking across it and am about to
do so myself. I am 100% certain that I will make it safely—as certain
as I would be if the chasm were only 5 feet deep, as certain as I would
be if the bridge were made of opaque material. Still, I hesitate; still, I
shudder. My behavior reflects something other than my belief. It is my
alief in action.
The reason Price’s explanation fails for our paradigm cases is that
the mechanisms they exploit are not under our direct control. We
are not in a position to “throw them off … when circumstances alter.”
This is not because we are in doubt about what we believe. There is no
question in my mind that the fudge has not been transformed into
dog feces; there are few things of which I am more certain than that
hurling darts at a photo of my baby will do no harm to the baby itself.
Still—even in high-stakes situations—there is a hesitation to my beliefconcordant actions.41
The problem with the belief-behavior picture is that at its heart lies
a faulty picture of what makes us act.42 I do not doubt that the account
could be made extensionally adequate: limit the cases that count as
“behavior” in the relevant sense, fuss with the notion of disposition,
make the fate of the world depend on the subject’s actions. Belief and
behavior can be made to match up, so long as one is free to make
41
Of course, I may become accustomed to performing the alief-averse action, and
my hesitation may dissipate. But this is a way of changing alief (by creating new patterns
of representational-affective-behavioral association patterns)—not a way of “throwing
it off.”
42
A nice recent defense of such an account can be found in Eric Funkhouser and
Shannon Spaulding’s “Imagination and Other Scripts,” where they defend what they
call the “Belief-Desire Thesis: For every intentional action, there is a belief-desire pair
that both causes and rationalizes that intentional action” (manuscript, p. 2).
656
the journal of philosophy
relevant alterations from both directions. But deep down, the account
misses something very important about human behavior. This is something to which both Aristotle and Hume were especially well attuned
(I will return to this in the final section), and which contemporary
psychology has begun to explore in detail. It is to cases from the latter
domain that I turn in the next section.
iii. automaticity
Recent work on “automaticity” has produced a remarkable series of
widely publicized results suggesting that alief plays a larger role in behavior than many had thought. Indeed, one of the main projects in
social psychology over the last two decades has been to document systematically the ways that behavior-inducing mental representations
may be activated by awakening the associative patterns that have come
to be linked with some object, stereotype, protocol, or mental image.43
A few examples will suffice for giving a sense of their flavor. But it is
important for the reader to realize that this is a massive research program and that while it may be possible to come up with alternative
explanations for one or another of the examples I discuss, the basic
phenomenon I am describing here has been established beyond any
reasonable doubt in hundreds of published studies.44
Much of the work in this area has been pioneered by John Bargh
and his colleagues, who, in a typical task present subjects with some sort
of association-inducing stimulus. This is often a “scrambled sentence”
task—a standard technique in psychology used to “prime” particular
concepts.45 In one such study, subjects faced one of three conditions:
either the collections of words from which they were asked to form
43
I discuss these and related cases in greater detail in “Imaginative Contagion”
(op. cit.); some of the material in this section draws on the discussion in that essay. In
the earlier paper, I suggested that these cases were examples of a phenomenon that I
called “imaginative contagion.” I now think that the phenomenon that I identified there
is a special case of an alief-like phenomenon. Readers interested in additional examples
of these sorts of cases may find them in that essay, and in the works cited therein.
44
I am gliding over many important distinctions about exactly which sorts of primes
tend to generate which sorts of responses: whether they tend to elicit assimilation or
contrast, whether they involve goals or nongoals, and so forth. In a full-fledged account
of alief, it will be important to address these subtleties in proper detail.
45
In such a task, subjects are presented with a list containing a number of five-word
sets, and asked to come up with a sentence for each set that contains at least four of the
designated words. So, for example, one such set might contain the words “snow, roof,
cat, cheerful, red” and the subject might write: “The cat stood in the snow atop the red
roof.” For original presentation of the scrambled sentence task, see Thomas K. Srull
and Robert S. Wyer, “The Role of Category Accessibility in the Interpretation of Information about Persons,” Journal of Personality and Social Psychology, xxxvii, 10 (1979):
1660–72, and “Category Accessibility and Social Perception,” Journal of Personality and
Social Psychology, xxxviii, 6 (1980): 841–56. For discussion of priming, see (concept)
alief and belief
657
sentences contained only neutral terms, or they also contained a number of terms associated either with politeness (for example, respect,
honor, considerate, patiently, courteous) or rudeness (for example,
aggressively, bother, disturb, intrude, brazen). Subjects were instructed
that, after completing the task, they should come out into the hallway
and find the experimenter, who would then give them the next task to
complete. When they emerged, they found the experimenter engaged
in a conversation with another “subject” (actually a confederate), a
conversation that continued either until the first subject interrupted
the conversation, or until 10 minutes had passed.
The action-patterns of the three groups differed markedly. Of those
who had been primed with the rudeness concept, most interrupted in
the allotted time; those in the neutral condition interrupted in less
than half of the cases; whereas those in the polite condition interrupted in almost none of the cases.46
One might maintain that the various groups differ in their beliefs, or
that they differ in their desires, or that the subject’s interruption of
the experimenter is not an action of the sort that belief-desire explanations are designed to cover. I have no doubt that such a story could
be told. One might say, for instance, that all three groups share the
same desire—to interrupt the experimenter only if doing so would be
socially acceptable—but that they differ in their belief about whether
it is. (Note that this would involve attributing to the subjects an odd
sort of belief—one that is formed as the result of mechanisms that are
not themselves sensitive to any subject-independent truth attitudes.47)
L.H. Storms, “Apparent Backward Association: A Situational Effect,” Journal of Experimental Psychology, lv, 4 (1958): 390–95; (term) S.J. Segal and C.N. Cofer, “The Effect
of Recency and Recall on Word Association,” The American Psychologist, xv (1960):
451; (discussion) J.H. Neely, “Semantic Priming and Retrieval from Lexical Memory,”
Journal of Experimental Psychology: General, cvi (1997): 226–54; (review) J.P. Toth and
E.M. Reingold, “Beyond Perception: Conceptual Contributions to Unconscious
Influences of Memory,” in G. Underwood, ed., Implicit Cognition (New York: Oxford,
1996), pp. 41–84.
46
Bargh, Chen, and Burrows, “The Automaticity of Social Behavior,” op. cit., p. 236.
47
For either there is no fact of the matter whether interruption in such circumstances is socially acceptable (in which case there is no truth for the mechanisms to
be sensitive to), or there is a fact of the matter, which is either independent of or dependent on the subject’s attitudes in the situation. If it is independent of those, then
the belief-forming mechanism is clearly truth-insensitive, for the three groups using
the same mechanism respond in three different ways to the same scenario. (See next
note.) And if it is dependent on those attitudes—say: interrupting is socially unacceptable iff the interrupter takes it to be socially unacceptable—then the belief-desire
explanation to which we are appealing becomes close to vacuous. (This is not to deny
that there are all sort of interesting instances of self-fulfilling beliefs and assessment-
658
the journal of philosophy
Alternatively, one might try to explain the phenomenon in terms of
imagination or pretense. Perhaps engaging in the scrambled sentence
task causes the subjects to fantasize that the experimenter is rude, or
polite—or that they themselves are rude or polite—and, carried away
by this fantasy, perhaps they begin to act as if it were true. Perhaps.48
But why would engaging in the scrambled sentence task cause the subjects to engage in this sort of fantasy (unless, of course, the explanation runs through something like the notion of alief)? And even if we
have an answer to that question, why would engaging in such a fantasy
make them act as if it were true (again, unless the explanation runs
through something like alief )?
Rather, what Bargh and his colleagues have done, I want to argue, is
to induce in their different sets of subjects different sorts of occurrent
alief. As the result of the pre- or quasi-conscious activation of the cluster of affective tendencies and behavioral repertoires associated with
the notion of rudeness, subjects in the third condition find themselves
more likely to act in ways that they would act in the presence of rudeness; as the result of the pre- or quasi-conscious activation of the cluster of affective tendencies and behavioral repertoires associated with
the notion of politeness, subjects in the second condition find themselves more likely to act in ways that they would act in the presence
of politeness.49
Nor is this an isolated anomaly. Example after example reveals the
subtle role of alief in guiding behavior. In another widely publicized
dependent attitudes. But subliminal primes altering perceptions of rudeness are hardly
instances of the cogito.)
48
Actually, there is experimental evidence suggesting that the behavior is not the result of any sort of conscious process. “To assess whether the priming manipulation had
resulted in different perception of the experimenter’s politeness, Bargh, Chen, and
Burrows examined the ratings participants made on” a scale where they were explicitly
asked to rate the experimenter’s degree of politeness. They found “no reliable difference in the ratings made in the three priming conditions”—all three groups ranked
him as neither especially polite nor especially impolite. They continue: the “fact that
the behavioral measure showed quite strong effects of the priming manipulation,
whereas the effect on the judgment measurement was nonexistent, argues against
the … interpretation … that the priming manipulation affected consciously made judgments about the experimenter, which then determined behavioral responses to him.
The results instead point to a direct effect on behavior that is not mediated by conscious perceptual or judgment processes” (Bargh, Chen, and Burrows, op. cit., p. 235).
49
See Bargh, Chen, and Burrows (op. cit.). Of course, subjects in the first (neutral) condition also have various aliefs rendered occurrent, but none that systematically affects the
likelihood of their interrupting the experimenter; they are like the visually induced aliefs
associated with stepping onto the back porch (as opposed to the Skywalk); they are present, but we do not need to appeal to them to explain otherwise discordant behavior.
alief and belief
659
Bargh experiment, subjects performed a scrambled sentence task in
which one group confronted sentences containing terms associated
with the elderly (for example, wrinkle, bingo, and retired), whereas
the second group’s unscrambling task involved only neutral terms.
After completing the experiment, subjects were surreptitiously timed
as they walked down the hall to the elevator. Those primed with the
elderly stereotype took significantly longer to walk to the elevator
than those who had not been so primed.50
It seems implausible (to say the least) that Bargh’s elderly-primed
subjects believed that they had suddenly turned into a bunch of geezers
who needed to dawdle lest they overtax themselves. It is slightly less
absurd to suggest that Bargh’s elderly-primed subjects imagined themselves as old—or imagined someone else who is old—and, having so
imagined, began to act in some ways as if the imagined content should
govern their own actual behavior. But even this is a rather far-fetched
explanation.51 (Among other things, in well-designed scrambled sentence tasks, subjects remain unconscious of the fact that a particular
notion is being primed.52) Rather, I want to suggest, Bargh’s elderlyprimed subjects occurently alieved below the level of conscious awareness something like: “Old. Tired. Be careful walking to that elevator
…”—and the activation of this behavioral repertoire made them more
likely to act in accord with it.
Additional research within this paradigm has reinforced and
expanded the lessons of these early experiments. So, for example,
showing suitably primed subjects a picture of a library leads them
to speak in quieter tones; showing them an image of an elegant dining room—or exposing them to the smell of soap—leads them to eat
more neatly.53 Subliminal visual priming with an image of an Africanfeatured face leads subjects to respond more aggressively to certain sorts of provocation.54 Priming subjects with thoughts of their
50
For discussion of how these results can be reconciled with neuropsychological evidence suggesting that simple motor actions are impervious to high-level mental processes such as stereotype activation, see Jane F. Banfield, Louise F. Pendry, Avril J.
Mewse, and Martin G. Edwards, “The Effects of an Elderly Stereotype Prime on Reaching and Grasping Actions,” Social Cognition, xxi, 4 (August 2003): 299–319.
51
Though one that I tacitly appealed to in my discussion of this case in “Imaginative
Contagion” (op. cit.).
52
In this particular case, “inspection of the responses” to a similar priming task
“revealed that only 1 of the 19 participants showed any awareness of the relationship
between the stimulus words and the elderly stereotype” (Bargh, Chen, and Burrows,
op. cit., p. 237).
53
Henk Aarts and Ap Dijksterhuis, “The Silence of the Library,” Journal of Personality
and Social Psychology, lxxxiv, 1 ( January 2003): 18–28.
54
Bargh, Chen, and Burrows, op. cit.
660
the journal of philosophy
(achievement-oriented) mother leads them to persist longer at wordfind tasks; priming them with thoughts of a friend makes them more
likely to help a stranger.55
Indeed, alief may be activated in even more striking ways. Recently,
psychologist Lawrence Williams “hypothesized that a simple experience
of physical, spatial distance would trigger feelings of psychological
distance and that those feelings, in turn, allow people to enjoy aversive
media.” Subjects were first asked to plot a pair of points on a cartesian
plane: the points were either quite close to one another (occupying
less than 1/4 of the plane) or quite far apart.
All the participants then read an embarrassing passage from a novel—in
which a woman opens a magazine to find that her ex-boyfriend has written an article about her, called “Loving a Larger Woman”—and rated
how much they enjoyed the story. Just as Williams had expected, the
participants who drew the dots far apart liked the passage more.
In his next study, after the volunteers drew the dots, they read a book
excerpt in which a man beats his brother with a rock after a car crash.
When the readers rated their emotional experience, Williams found,
people who were told to draw the dots close together reported feeling
more negative emotions.56
In all of these cases, it is perhaps possible to explain what is going on in
familiar terminology. Perhaps Bargh’s interruption subjects imagine
that there is rudeness afoot in their dominion, and adjust their behavior accordingly. (Really? Even though the priming takes place at the
unconscious level?) Perhaps his elevator subjects imagine that they
are old and gray and full of sleep, and consequently slow their pace.
Perhaps Williams’s subjects imagine that they are far away from the
stories they hear, and therefore feel their emotional tug less strongly.
Perhaps. Or perhaps what is happening in each of these cases is the
activation of a low-level cluster of associations—representational, affective, behavioral—an activation that renders the subject more likely
to exhibit behavior of a certain sort. To a reasonable approximation, it
looks like all depictive representations—even those that we explicitly
disavow as false—feed into our behavioral repertoires, and that it is
55
G. Fitzsimmons and J.A. Bargh, “Thinking of You,” Journal of Personality and Social
Psychology, lxxxiv (2003): 148–63.
56
Polly Shulman, “Priming the Mind,” Science: Science Careers (March 2007). In addition to showing greater enjoyment of embarrassing media and less emotional distress
from violent media, distant-dot drawers offered lower estimations of calories in unhealthy food, and weaker reports of emotional attachments to family members. (See
Lawrence Williams and Bargh, “Keeping One’s Distance: The Effect of Spatial Distance
Cues on Affect and Evaluation,” Psychological Science, xix, 3 (2007): 302–08.)
alief and belief
661
only through a process of conscious or habit-governed inhibition that
representations whose accuracy we endorse come to play a distinctive
role in governing our actions.
If so, there is something deeply wrong about the traditional picture
of the relation between belief and behavior that we discussed in section ii. But of course, this is not the only way philosophers have
thought about these matters. In the final section, I briefly examine
one competing philosophical strand.
iv. alief, persuasion, and habit
Despite certain protestations to the contrary, philosophers have been
exquisitely sensitive to the ways in which contemplation of an imaginary particular may have cognitive and motivational effects that differ
from those evoked by an abstract description of an otherwise similar
state of affairs.57 (Think of Plato’s cave, the ring of Gyges, twin earth,
the Chinese room, teletransportation, Thomson’s violinist, the veil of
ignorance, Mr. Truetemp, the fat man on the bridge, and any of the
myriad other examples). A particularly vivid presentation of this claim
can be found in Hume’s Treatise on Human Nature, where Hume writes:
There is a noted passage in the history of Greece, which may serve for
our present purpose. Themistocles told the Athenians, that he had
form’d a design, which wou’d be highly useful to the public, but which
`twas impossible for him to communicate to them without ruining the
execution, since its success depended entirely on the secrecy with which
it shou’d be conducted. The Athenians, instead of granting him full
power to act as he thought fitting, order’d him to communicate his
design to Aristides, in whose prudence they had an entire confidence,
and whose opinion they were resolv’d blindly to submit to. The design of
Themistocles was secretly to set fire to the fleet of all the Grecian commonwealths, which was assembled in a neighbouring port, and which
being once destroy’d wou’d give the Athenians the empire of the sea
without any rival. Aristides return’d to the assembly, and told them, that
nothing cou’d be more advantageous than the design of Themistocles
but at the same time that nothing cou’d be more unjust: Upon which
the people unanimously rejected the project.58
Hume goes on to note that his contemporary Charles Rollin found it
astounding that the Athenians would reject—merely on grounds of
57
I discuss this issue in more detail in “Philosophical Thought Experiments, Intuitions
and Cognitive Equilibrium,” Midwest Studies in Philosophy, xxxi, 1 (2007): 68–89. The discussion in the next three paragraphs draws on the discussion from the opening pages of
that paper.
58
David Hume, A Treatise on Human Nature, L.A. Selby-Bigge, ed. (New York: Oxford,
1978), II.iii.6.4.
662
the journal of philosophy
injustice—a strategy so “advantageous” that it would give them “the
empire of the sea without any rival.” But Hume himself is not surprised:
For my part I see nothing so extraordinary in this proceeding of the
Athenians. … [T]ho’ in the present case the advantage was immediate
to the Athenians, yet as it was known only under the general notion of
advantage, without being conceiv’d by any particular idea, it must have
had a less considerable influence on their imaginations, and have been
a less violent temptation, than if they had been acquainted with all its
circumstances: Otherwise `tis difficult to conceive, that a whole people,
unjust and violent as men commonly are, shou’d so unanimously have
adher’d to justice, and rejected any considerable advantage (ibid., II.iii.6.4).
Hume’s story brings out the way in which engagement of the cognitive
mechanisms associated with vivid imagining may lead a subject to reverse a prior commitment, selecting as preferable the option previously
rejected, and shunning the option previously embraced.59
For the reader who has gotten this far, it should be apparent what
lesson I want to draw from this case. Ever sensitive to the role of habit
and association—“If any thing can intitle the author to so glorious a
name as that of an inventor, ‘tis the use he makes of the principle of
the association of ideas”60 —Hume is here pointing out that judgment
about a particular case may be driven as much by alief as by belief.
Like his K Street counterpart, Hume recognizes the citizen who believes that wealth should be redistributed across generations alieves
that the death tax is unfair; like his Madison Avenue foil, Hume recognizes that a customer who believes that a $9.99 scarf costs nearly ten
dollars alieves that it costs only nine. When the citizen votes against
the amendment does this show that he really opposes redistribution?
Or does it show that action is often governed by alief?
If so, then Aristotle is right: In order to live well, we must work to
bring our habits in accord with our reflective beliefs:61
Men become builders by building and lyre-players by playing the lyre; so
too we become just by doing just acts, temperate by doing temperate
acts, brave by doing brave acts … states of character arise out of like
59
In the paper on thought experiments, I go on to explore how this phenomenon might help explain both the effectiveness and the limitations of philosophical
thought experiments.
60
Hume, op. cit., “Abstract,” pp. 661–62.
61
For exploration of this connection in a related context, see the final paragraph
of J. Thomas Cook, “Deciding to Believe without Self-Deception” this journal,
lxxxiv, 8 (August 1987): 441–46; cf. also Myles Burnyeat, “Aristotle on Learning to
be Good,” in Amélie O. Rorty, ed., Essays on Aristotle’s Ethics (Berkeley: California UP,
1980), pp. 66–92, as well as Bill Pollard, “Explaining Actions with Habits,” American Philosophical Quarterly, xl, 1 (2006): 56–69.
alief and belief
663
activities .… It makes no small difference, then, whether we form habits
of one kind or of another from our very youth; it makes a very great
difference, or rather all the difference.62
My conclusion should not be a surprising one. I think that alief
governs all sorts of belief-discordant behavior—the cases with which
I began the paper, and the ones that I have presented along the way.
But if alief drives behavior in belief-discordant cases, it is likely that it
drives behavior in belief-concordant cases as well. Belief plays an important role in the ultimate regulation of behavior. But it plays a far
smaller role in moment-by-moment management than philosophical
tradition has tended to stress.
tamar szabó gendler
Yale University
62
Aristotle, Nicomachean Ethics, J.L. Ackrill, J.O. Urmson, and David Ross, eds. (New
York: Oxford, 1998), pp. 1103–04. Somewhat simplistically, one might say that Aristotelian ethics is an ethics of alief, whereas Kantian ethics is an ethics of belief. I hope to
explore this issue in more detail in further work.
Received: 18 February 2020
Accepted: 29 March 2020
DOI: 10.1111/phc3.12668
ARTICLE
The relationship between belief and credence
Elizabeth G. Jackson1,2
1
School of Philosophy, Australian National
University, HC Coombs Building 9, Canberra,
Abstract
Australia
Sometimes epistemologists theorize about belief, a tripartite
2
attitude on which one can believe, withhold belief, or disbe-
Department of Philosophy, Ryerson
University, Toronto, Canada
lieve a proposition. In other cases, epistemologists theorize
Correspondence
about credence, a fine-grained attitude that represents one’s
Elizabeth G. Jackson, School of Philosophy,
subjective probability or confidence level toward a proposi-
Australian National University, HC Coombs
Building 9, Canberra 0200, ACT, Australia.
Email: lizjackson111@gmail.com
Funding information
Australian Research Council, Grant/Award
Number: D170101394
tion. How do these two attitudes relate to each other? This
article explores the relationship between belief and credence in two categories: descriptive and normative. It then
explains the broader significance of the belief-credence
connection and concludes with general lessons from the
debate thus far. Video Abstract link: https://www.youtube.
com/watch?v=3eOSlPVYxI8&feature=youtu.be
KEYWORDS
belief, belief-first, credence, credence-first, degree of belief,
dualism, epistemic rationality, Lockean thesis, lottery paradox,
preface paradox, reduction, statistical evidence
1
|
I N T RO DU C T I O N : BE L I EF A N D C R E D E N C E
Belief is a familiar attitude. To believe something is to regard it as true or take it to be the case (Schwitgebel, 2019). I believe
1 + 1 = 2 and that it will be sunny tomorrow. According to the tripartite model, there are three doxastic attitudes one can
take toward a proposition p: believe p, disbelieve p, and withhold belief, being effectively undecided on whether p.
But consider: I believe both 1 + 1 = 2 and that it will be sunny tomorrow, but my attitude toward these propositions is not exactly the same—the former is more probable. To capture this, epistemologists appeal to another propositional attitude, called credence (also sometimes called partial belief or degree of belief, but see Moon, 2017).
Credences are more fine-grained than beliefs and are often given a value on the [0,1] interval, where 1 represents
maximal confidence p is true, and 0 represents maximal confidence p is false. For example, I have a ~0.9999 credence 1 + 1 = 2, but only a ~0.9 credence it will be sunny tomorrow. Unlike belief, there are (in principle) an infinite
number of credences one can take toward a proposition. While the concept of credence grew out of work on subjective probability (see Ramsey, 1926, p. 166ff, Jeffrey, 1965, de Finetti, 1974, and especially Eriksson & Hájek, 2007),
Philosophy Compass. 2020;15:e12668.
https://doi.org/10.1111/phc3.12668
wileyonlinelibrary.com/journal/phc3
© 2020 John Wiley & Sons Ltd
1 of 13
2 of 13
JACKSON
many contemporary epistemologists posit a close connection between credence and the familiar notion of confidence (Moon, 2019, pp. 276–277; Schupbach, 2018, p. 191).
This article is about the relationship between belief and credence. Section 2 covers the metaphysics of belief
and credence, focusing on potential descriptive connections between the attitudes. Section 3 turns to the normative:
what is the relationship between rational belief and rational credence? Section 4 concludes with upshots and outstanding issues for further research.
2
2.1
B E L I E F A N D C R E DE N C E : D E S C R I P T I V E QU E S T I O N S
|
|
The ontological question
A natural starting point is the ontological question: do we have both beliefs and credences? Eliminativists maintain that
belief or credence (or both) do not exist. Belief-eliminativism the view that we do not have beliefs. Several philosophers—
for example, Jeffrey (1970), Maher (1993, pp. 152–155), Stich (1996), and Pettigrew (2016)—express sympathy for
belief-eliminativism. They claim that beliefs are left over from folk psychology—despite appearances, humans do not have
beliefs. While belief-eliminativism is simple and explains fine-grained features of our mental lives by appealing to credences, it requires an extensive error theory about commonsense psychology and everyday discourse.
Another eliminativist view is credal-eliminativism (Holton, 2008, 2014; Horgan, 2017). Holton (2014, p. 14) writes,
“I argue that we cannot form credences at all. The Bayesian approach is not an idealization of something we actually
do. Instead, it is quite foreign to us. Just as our core native deliberative state is that of the simple intention, so our core
native epistemic state is that of simple, all-out belief.” Holton nonetheless maintains that we have another attitude he
calls “partial beliefs” that comes in degrees and stands in contrast to full beliefs. Depending on how broad one’s notion
of credence is, then, virtually no one defends credence-eliminativism. After all, we are more confident in some of our
beliefs than in others, and it is not obvious how to capture this with a belief-only ontology. Further, if we understand
credence as closely connected to degree of confidence, credal-eliminativism also requires revision of folk psychology
(Eriksson & Hájek, 2007, p. 209). On a third view, neither belief nor credence exists, but, except for those skeptical of
all intentional mental states (Churchland, 1981; Rosenberg, 1999, 2018), most will find this implausible.
2.2
|
The reduction question
Realists about belief and credence affirm that both attitudes exist, but are divided on a second question: the reduction question. Is one attitude more metaphysically fundamental? There are three answers to the reduction question:
credence is more fundamental, belief is more fundamental, and neither is more fundamental.
First, consider the credence-first view, on which credence is the fundamental attitude, and belief reduces to credence. There are two versions of this view. On the first, belief is maximal credence or credence 1. While they differ in
the details, Levi (1991), Roorda (1995), Tang (2009), Wedgwood (2012), Clarke (2013), Greco (2015), and Dodd (2016)
all defend versions of this view. One reason to favor this view is that belief and credence 1 have similar functional
profiles—they both have content that we tend to treat as true in our reasoning. Also, as Greco (2015, p. 179) emphasizes, the view that belief is credence 1 is clean and simple. On the other hand, it seems like many common beliefs
(e.g., it will be sunny tomorrow) are held with less than maximal certainty. It also does not seem like we should take bets
at extraordinary odds on all our beliefs, although, according to decision theory, we should take such bets on propositions
in which we have credence 1. Further, the belief-as-maximal-credence view cannot capture the plausible datum that we
are more confident in some of our beliefs than in others. Defenders of this view reply to these worries by, among other
things, denying that credence 1 is certainty and weakening the connection between betting and credence.
On a second credence-first view, belief reduces to credence above some threshold; this is often called the
threshold view.1 Simple views posit a fixed threshold, often between 0.5–1; the challenge is to set the threshold at a
3 of 13
JACKSON
place that does not look ad-hoc.2 On other views, the threshold for belief is context or stakes-dependent
(Bach, 2008; Ganson, 2008; Pace, 2011; Weatherson, 2005). This contextualist view looks less ad hoc, but implies
that whether one believes something is context- or stakes-sensitive. Another potential credence-first view says that
belief that p does not reduce to one’s credence in p, but to facts about one’s cre…